Contents
Preface
vii
Preface to the Thid Edition Inoduction to the Chinese Edition
1
Analycal Index
5 9
Inoduction Pats1-20
14
Postscipt on Relativism
268
Index
273
Preface
In 1970 lmre Lakatos, one of the best friends I ever had, coered me at a party. 'Pau, he said, 'you have such sange ideas Why dont you write write them them down? I sha sha write write a repy, repy, we pubish the whoe thing thing and promise you - we sha have ots of fun I iked iked the sugeson suges on and I promise started working. The manuscript of my part of the bk was ished in 1972 and I sent it to London There it disappeared under rather mysterious circumstances lmre Lakatos, who oved dramac gestures, noed Inteo and, indeed, Inteo found my manu script and retued it to me I reread it and made some a changes In Februa 1974, ony a few weeks after after I had nished my revsion, I was informed of lmres death. I pubished my part of our common enterise without hs response A year ater I pubished a second voume, Sce in Free Soe, containing addiona materia and repies to cricism This histo expains the form of the book. It is not a systemac ease; it is a etter to a friend and addresses his idiosyncrasies. For exampe Imre Lakatos was a raonaist hence raonism pays a arge roe roe in the bk. He aso admired Popper and therefore Popper curs much more frequenty than his 'objecve importance woud warrant lmre Lakatos, somewhat jokingy, caed me an anarchist and I had no objecon to putng on the anarchists mask Finay, lmre Lakatos oved to embarrass serious opponents with jokes and irony and so I too, casionay wrote in a rather ironica vein. An exampe is the end of Chapter 1: 'anything goes is not a 'principe I hod - I do not think think that that 'principes 'principes can be used used and fruiuy fruiuy discussed outside the concrete research situaon they are supposed to aect - but the terried excamaon of a raonaist who takes a coser ook at histo Reading the many thorough, serious, oninded and thoroughy misguided cricisms I received after pubicaon of the rst Engish edion I often recaed my exchanges with lmre; how we w e woud both have aughed had we been abe to read these eusions together The new edion merges parts ofinst Metowith excets from Sce in Free So I have omitted materia no onger of interest,
V
PREACE
added a chapter on the a of Gaieo and a chapter chapter on o n the noon of reaity reaity that seems s eems to be required required by the the fact that knowedge knowedg e is is part of a compex historica press, eiminated mistakes, shortened the argument wherever possibe and freed it from some of its earier idiosyncrasies. Again I want to make two points: rst, that science can stand on its own feet and ds not need any hep from raonaists, secuar humanists, arxists and simiar reigious movements; and, secondy, that nonscienc cutures, procedures and assumpons can aso stand on their own feet and shoud be aowed to do so, if ths is the wish of their representaves. Science must be protected fr om fr om ideoogies; and siees, especiay democrac sociees, must be protected from science Ths does not mean that sciensts cannot prot from a phiosophica educaon and that humanity has not and never wi prot from the sciences However, the prots shoud not be imposed; they shoud be examined and freey accepted by the pares of the exchange In a democracy scienc instuons, research programmes, and suggesons must therefore be subeted to pubic cono, there must be a separaon of state and science ust as there is a separaon between state and reigious instuons, and science shoud be taught as one view among many and not as the one and ony road to uth and reaity There is nothing in the natur of science that excudes such instuona arrangements or shows that they are iabe to ead to disaster None of the ideas that underie my argument s new y interetaon of scienc knowedge, knowedge, for exampe, was a viaity for fo r physicists ike ach, Bomann, Einstein and Bohr. But the ideas of these great thikers were distorted beyond recognion by the rodnts of neoposivism and the compeng compeng rodents of the church church of 'crica raonaism Lakatos was, aer Kuhn, one of the few tikers who noced the discrepancy and ed to eimnate it by means of a compex and ve interesng interesn g theo of raonaity I dont thik he has succeeded in ths But the attempt was woth the eort; it has e to interesng resuts in the histo of science and to new insights to the imits of reason I therefore dedicate aso this second, are ady much more oney version of our common work to hs memo Earier materia reang to the probems in ths book is now coected in my Phloshil Pape} Fare to Reon2 contains hstorica materia, especiay from the eary hsto of raonaism in the West and appicaons to the prbems of today Berkeey, Sepember 1 987 2 vs, vs, Camb Cambd dge ge,, 198 1 9811 . 2 London 987
Preface to the Third Edion
any things have happened since I rst pubished Against ainst Method A for short) There have been dramac poica sia and ecoogic ecoogica a changes changes Freedom Freedom has increas increased ed - but itit has brought hunger, insecurity, naonaisc tensions, wars and saightfoard murder. Word eaders have met to dea with the deterioraon of our o ur resources; as is their habit, they have made speeches and sied agreements The agreements are far from sasfacto; some of them are a sham However, However , at east verbay, the environment environm ent has become a wordwide conce Physicians, deveopmenta agents, priests working with the poor and disadvantaged have reaized that these peope peope know more about their condion than a beief in the universa exceence of science or ornized reigion had assumed and they have changed their acons and their ideas accordingy (iberaon theoo; prma environmenta care, etc) any inteectuas have adapted what they have eaed at universies and specia schoos to make their knowedge more ecient and more humane On a more academic eve historians (of science, of cuture) have started approaching the past in its own terms Aready in 1933, in his inaugura ecture at the Coge de France, Lucien Febvre had ridicued writers who, 'sitng at their desks, behind mountains of paper , having cosed and covered their windows, made profound judgements about the ife of andhoders, peasants and farmhands In In a narrow ed historians of science ed to reconsuct the distant and the more immediate past without distorng it by mode beiefs about truth (fact) and raonaity Phiosophers then concuded that the various forms of raonaism that had oered their serices had ot ony produced chimaeras but woud have damaged the sciences ad they been adopted as guides Here Kuhns masteiece payed a decisi v v e r ole. 1 It led aged Unfortunately fortunately it also encour aged to new ideas. ideas. Un led to new ots o trash. Kuhns main terms ('paradigm, 'revouon, 'norma 1
e c cre re oc oc Rol Rolio io,, Chcago, 1962
P R EA C E TO T H E T H I R D E D I T I O N
science, 'prescience, 'anomay, 'puzzesoving, etc tued up in various forms of pseudoscience whie his genera approach confused many writers nding that science had been freed from the fetters of a dogmac ogic and epistemoo they ed to e it down ain ths me with sioogica ropes. That end aste we into the eary sevenes By conast there are now historians and sioogists who concenate on parcuars and aow generaies ony o the extent that they are supported by siohstorica conecons 'Nature, says Bruno Latour, referring to 'science in the makng is 'the consequence of [a] settement of 'conoversies 2 Or, as I wrote in the rst edion of 'Creaon of a thng, and creaon pus fu understandig of a coe a of the thig, are v o pas ofone and the same ndble procs and cannot be separted without bringg the press to a stop 3 Exampes of the new approach are Andrew Pickering Constng Quar, Peter Gaison, How Eem End, arn Rudwick, e Great oan Conte, Arthur Fine, he Sha Game and others. 4 There are studies of the various adions (reigious, styisc, aonage, etc that inuenced scinsts and shaped their research; 5 they show the need for a far more compex account of scienc knowedge than that which had emerged from posivism and simar phosophes. On a more genera eve we have the oder work of cha Poanyi and then Puam, van Fraassen Cartwrght, arceo Pera 6 and, yes, lmre Lakatos, who was sucienty opsc to beieve that histo hersef - a ady he tk ve seriousy - oered simpe rues of theo evauaon In socioo the attenon to detai has ed to a situaon where the probem is no onger why and how 'science changes but how it keeps together Phosophers, phiosophers of bioo especiay, suspected for some me that there is not one enty 'science with ceary dened principes but that science contains a great variety of (high eve theoreca, phenomenoogica, expermenta approaches and that even a parcuar science such as physics is but a scattered coecon of subjects (eascity, hydrodynamics, rheoo, thermo dynamics, etc., etc each one containing cona tendencies (exampe Prandt vs Hemhotz, Kevin, Lamb, Rayeigh; Truesde 2 S in in, Mlton eynes, 1 987, pp. 4 and 98f London, 1975, p. 26, repeated on p. 1 7 of the present edon - original emphass 4. Al Chicago Unversi Press. 5 An example s Maro Bagioli, Gi Coi, forthcoming. 6 Sce a Rhetoc, forthcomng.
P R AC E TO T H T H I R D E D I I O N
xi
Prandtl; Birkhoff vs 'physcal commonsense Knsman llusang all ends - n hydrod�amics). For some authors this is not only a fact t s also desrable7 Here ain conbuted, in a small way, n Chapters 4 and 1 1 of A8 in secon 6 of my conbuon to Lakatos and Musgraves Ctsm and the Growth of Knowledge (cricism of the unformity of paradgms n Kuhn)9 and already n 19 62, in my conbuon to the Delaware Studr the Phloshy of
S ce.10 Unity fuer disappears when we pay attenon not only to breaks on the theorecal level, but to experiment and, especially, to mode laborato science. As Ian Hacking has shown in his pathbreakng essay Rrtng and /nten � 11 and as emerges from Pickerings Sce a Prace and Culture, 2 terms such as 'experment and 'obseaon cover complex presses containing many sands. 'Facts come from negoaons between dierent pares and the nal product - the publshed report - is inuenced by physcal events, datapressors, compromses, exhauson, lack of money, naonal pride and so on. Some microstudies of laborato science resemble the 'New Jouaism of jimmy Breslin, Guy Talese, Tom Wolfe and others researchers no longer sit back and read the papers in a cerain eld they are not content with sient vists to laboratories either- they walk right in, enge sciensts in conveaon and make thigs happen (Kuhn and his collaborators started the predure in their interiews for the histo of quantum mechancs). At any rate we are a long way from the old (Platonic) dea of science a a system of statements growing with experiment and obseaon and kept in order by lasng raonal standards. A s sll partly proposon oriented however, I also had my sane moments. My discussion of ncommensurabilty, or exaple, does not 'reduce the dierence to one of theo a Pickering writes.1 3 It includes art forms, percepons (a large par of hapter 16 s about the ansion from Greek geomec art and py to the classcal perod), stages of chid development and asserts 'that the views of sciensts and especally their views on basic matters are often as dierent from each other as are the ideologies of derent 7 J Dupr, The Disun of Scence', Mind, 92, 198. 8 Present edion Taken over unamended from rst edon 9 I. Lakaros and A Musgrave (eds), Ctim and the Gwth of Knowledge, Cambrid ge, 1965 10 How to be a d Empiricst', Deware Stud, Vol 2, 196 Cambrid ge, 198 12 A Pickering (ed.), Sce Price and Culture, Chicago, 1992. 1 . bd, p 0
xii
PR E A C E TO T H E T H I R D E D I T I O N
cultures'} 4 n this connecon I examined the praccal aspects of logic, the way, that is, in which ideas are related to each other in ongoing research rather than in the nished products (if there ever are such products) My discussion of the many events that constute what is being obseed 1 5 and especially my discussion of Galieo's telescopic discoveries 1 6 agree with the requirements of the new laborato siolo except that Galileo's laborato was rather smal by comparison This case shows, incidentally, that like the older phiosophies of science the new microsiolo is not a universal account but a descripon of prominent aspects of a special period It does not matter A universal descripon of science at any rate can at most oer a list of events 1 7 It was dierent in anquty It is clear that the new situaon requires a new phiosophy and, above all, new terms Yet some of the foremost researchers in the area are sll askng themselves whether a parcular piece of research produces a discove', or an invenon', or to what extent a (tempora) result is objecve' The problem arose in quantum mechanics; it is also a problem for classical science Shall we connue usig ouoded terms to describe novel insight or would it not be better to start using a new language? And wouldn't poets and joualists be of great help in nding such a language? Secondly, the new situaon ain raises the queso of science' vs demracy For me this was the most important queson My main reason for wring the book', I say in the Inoducon to the Chnese Edion, 18 was humanitarian, not intelectual I wanted to support people, not to "advance knowledge' Now if science is no longer a unit if dierent parts of t preed in radically dierent ways and if conecons between these ways are ed to parcular research episodes, then scienc projects have to be taken indivdually This is what govement agencies started doing some me ago In the lat sixes the idea of a comprehensive science policy was gradually abandoned It was realized that science was not one but many enterises and that there could be no single policy for the support of all of them 1 9 Goveent agencies no longer nance sience, they nance parcular projects But then the word scienc' can no longer exclude unscienc projects we have to look at matters in 1 4 , rs edion, p. 274 1 5 ibid, pp 1 49 Reprnted n the present edion 1 6 Chapters 8 to 0 of the present edion 1 7 Cf my conibuon to the 1 992 Erasmus Symsum, Has the Scienc View of the Wold a Specal Satus Compared With Other Views?, fothcoming 1 8 Conaned n the present edon 19 J Ben-Daid, Stc Gwth, Bereley, 1991 , p 525
PR AC O IR D D I I O N
xii
etal Are the new philosophers and sologsts prepared to consider ths consequence of their research? here have been many other changes Medcal researchers and techologsts have not only invented useful nsuments (such a tose emplong the prncples of bre opcs which in many contexts replace the more dangerous methods of X-ray diaosc) but have become more open towards new (or older) deas Only twenty years ago the dea that the mind aects physcal well-being, though supported by experence, was rather unpopular today t s anseam Malpracce suts have made physicians more careful, someme too careful for the good of their paent, but they have also forced them to consult alteave opinions (In Switzerland a belligerent plurality of vews s almost part of culture and I used t when arrangng public confrontaons between hardheaded scensts and alteave' thnkers 2 However, here as elsewhere, simple philosophes, whether of a dogac or a more lberal knd, have their limts There are no geral solutons An ncreased liberalism n the denion of fact' can have grave repercussons, 2 whle the dea tha uth is concealed and even peerted b the processes that are meant to establsh it makes excellent sense 2 I therefore agan wa the reader that I dont have the ntenon of replacng ld and dogmac' prncples by new and more lbertaran ones' For example, I am neither a populist for whom an appeal to the people' s the basis of all knowledge, nor a relavst for whom there are no uths as such' but only uths for ths or that group and/or indvdual All I say s that non-experts often know more than experts and should therre be consulted and that prophets of uth (ncluding those who use arguments) more often than not are carried along by a vson that clashes wth the ve events the vson s supposed to be explorng. There ests ample evidence for both parts of ths assron A case I already menoned s development professionals dealng wth he ecologcal, social and medcal parts of developmental ad have by now realzed that the imposon of raonal' or scenc' predures, though casonally benecial (removal of some par astes and nfecous diseases), can lead to serous material and
20 Cf he seres edted by Chrsan Thomas and myself and publshed by he
Verlag derFachvereine, Zurch, 1 983-87. 2 1 . Cf. Petr W. Huber, Gal' Rge, Ne w York, 1 9 9 1 .
22 or a conal account, cf To Wolfes he ore ofhe Vani, Ne Yor, 1987
x
PR AC TO TO TH TH IRD D IT IO N
sprtual problems They dd not abandon what they had leaed in ther univeres, however they combned ths knowledge wth lal beliefs and customs and thereby established established a much needed lnk with the problems of lfe that surround us everywhere n the Ft Second, and Thrd Worlds The present edon contans major changes changes (Chapter (Chapt er 19 and part of Chapter 16 have been rewrtten, the old Chapter 20 has been omtted), addions (a paragraph here, a paragraph there), stylisc changes ( I hope they are mprovements) and correcons as well as addons n the references As far as I am conceed the man deas of the essay (e the deas expressed n talcs n the Inoducon to the Chinese Edion) are rather val and appear val when expressed expressed n sutable terms terms I prefer more mor e paradocal formulaons, however, for fo r nothng dulls the mnd as as thoroughly as heang famiar words and slogans It is one of o f the merts of deconsucon to have undermned phlosophical commonplaces and thus to have made some people thk Unfortunately it aected only a small crcle of nsders and t aected thm n ways that are not always clear, not even to them th em That's why I prefer prefer Nesoy, who w ho was a great, popular and funy deconsucteur, whle Derrda, for all hs goo ntenons, can't even tell a sto Rome, July 1992
Introduction to the Chinese Edion This book proposes a thess and draws consequences rom it The thesis is: the ts ts proce procedur dur and and rults rults that that constitute constitute the the sces sces he he no coon sture; there are no elements that cur in eve scienc ineson but are missing elsewhere Concrete develop ments (such as the ovethrow of steady state cosmologies and the discove of the sucture of DNA) have disnct features and we can often explain why and how these features led to success But not eve discove can be accounted for in the same manner, and predures that paid o in the past may create hav when imposed on the future Successful Succes sful research does not obey general standards; stan dards; it relies now on one ck, now on another; the moves that advance it and the standards that dene what counts as an advance are not always known to the movers Far-reaching changes of outlook, such as the so-called Copeican Revoluon' or the Dainian Revoluon', aect dierent areas f research in dierent ways and receive dierent impulses from them A theo of science that devises standards and suctural elements or a scienc acvies and authorizes them by reference to Reason' or Raonality' may impress outsiders but it is much too crude an insument for the people on the spot, that is, for scensts facing some concrete research problem In this book I y to support the thesis by historical exaples Such support does not establish it it makes it plausible and the way in which it s reached indicates how future statements about the nature of science' may be undermined given any rule, or any general statement about the sciences, there always always est developments which are praised by those who support the rule but which show that the rule does more damage dama ge than gd. One consequence of the thesis is that stc succs nnot be laned n a sip siple le way way We cannot say the sucture sucture of the atomic atomic nucleus nucleus was found because people did did A, B, C ' wher� A, A , B and C re prcedures which can be understood independent of ther use can do s to give a historical historic al account of the nuclear physcs Al we can
1
2
AGAINST METHOD
detals, ncluding sal circumstances, accidents ad personal dosyncrasies Another consequence is that the the suc su ccs cs of sce sce'' nnot nnot be used
an am amtt r treat treatin ing g yet unsoed pbls pbls in a stan sta nrrd ded ed way way
That could be done only if there are predures predures that can be detached detac hed from parcular research situaons and whose presence guantees success The thesis says that there are no such predures Referring to the success of science' in in order order to jus, jus, say, quaning human behaviour is therefore an argument without substance Quanca on works in some cases, fails in others; for example, it ran into dicules i one of the apparently most quantave of all sciences, celesal mechanics (special region stability of the planeta system) and was repaced by qualitave (topological) consideraons non-stc'proced cedur ur can cannnot be bepushed i It aso follows that non-stc'pro at. To say the procedure you used used is non-scienc, threfore we cannot cannot ust your results and and cannot cannot give you money for research' assumes tha science' is successful and that it is successl because it uses uniform procedures The rst part of the asseron (science is always succesful') is not ue, if by science' we mean thngs done by scien sciensts sts there are lots lots of failu failure ress also also The second second part part that successes are due to uniform procedure proceduress is not ue because there are no such predures Sciensts are like architects who build buidings of dierent sizes and dierent shapes and who can be judged only aer the event, ie only after they have nished their sucture sucture It may stand up, itit may fall down nobody knows But if scienc achievements can be judged only after the event and if there is no absact way of ensuring success beforehand, then there ests no special way of weighing scienc promses either sciensts are no better o than anybody else in these matters, they only know more details This means that that thepublic n pai paipate pate in the the disssion wihout distuing isting to succs (there are no such roads) In cases where the sciensts' work aects the public it even should parcipate: rst, because it is a conceed party (many scienc decisions aect public life); secondly, because such parcipaon is the best best scienc educaon the the public can get a full democrazaon of science (which ncludes the protecon of minories such as scensts) s not n conict with science It is in conct with a philosophy, often called Raonalism', that uses a frozen image of science to terrorize people unfamiiar with its pracce A consequence to which I allude in Chapter 19 and which is closely connected conn ected with its basic thesis is that there can can be man many dert n ofsce People starng from dierent sial backgrounds will
I N TR O D U C T I O N O T H E C H I N E S E E D I I I O N
3
approach the world in dierent ways and lea dierent things about it People surived millennia before Weste science arose to do this they had to know their surroundings up to and including elements of asonomy Several thousand Cuahuila Indians never exhausted the natural resources of a desert region in South Califoia, in which today only a handful of white families manage to subsist They lived in a land of plenty, for in this aparently completely barren territo, they were familiar with no less than sixty kinds of edible plants and twenty-eigh others of narcoc, smulant or medical properes' 1 The knowledge that presees the lifestyles of nomads was acquired and is preseed in a non-scienc way (science' (sci ence' now being mode natural science). Chinese technolo for a long me lacked any Weste-scienc undeinning and yet it was far ahead of contempora Weste technolo It is ue that Weste science now reis supreme all over the globe however, the reason was not insight in its inherent raonality but power play (the colonizing naons imposed their ways of living) and the need for weapons: Weste science so far has created the most ecient ecient insuments of death The remark that without Weste science scienc e many many Third World naons would be staring is correct but one should add that the oubles were created, not alleviated by earlier forms of develop ment. It is also ue that Weste medicine helped eradicate parasites and some infecous diseases but ths does not show that Weste science scienc e is the only adion adion that has ha s good thngs to oer and that other forms of inqui are without any merit whatsoever Ft worlds world sce ce is is one sce among many many by claiming to be more it ceases to be an insument of research and tus into a (polical) pressure roup More on these matters can be found in my book Fare to
Reon 2
My main mov in wring the book was humanitarian, not intellectual. I wanted to t o support people, not to to advance knowledge' People all over the world have developed ways of suriving in partly danerus partly areeable surroundings surroundings The stories they told and the acvies they enged in enriched their lives, protected prote cted them and ave them meaning. The proress of knowledge and civilizaon' as the process of pushng Weste ways and values into all coers of the lobe is being called called desoyed these these wonderful products of of human ing enui enuity and compassion without a single glance in their direcon Prgress of knowledge' in many places meant killing of minds day old adions are being revived and people ain to adapt C L-Sass, Mind, London, 19, pp 4 2 London 987
4
AG AI NS T T TOD
their lives lives to t o the ideas of their their ancestors ancestors I have tried t show by an analysis of the apparently hardest parts of science, the natural sciences, that science, properly understood, has no argument against such a procedure. There are many sciensts who act accordingly Physicians, anthropologists and environmentalists are starng to adapt their procedures to the values of the people they are supposed supposed to advise. I am not against against a science so understood Such a science is one of the most wonderful invenons of the human mind. But I am against against ideoogies that use the name of science for cultural murder.
Analycal Index Beng a Skech ofthe Man Amt
nucn 9 S s an stl anarchc tee theoretl anarchsm s more humantaan and more ke to couragepros than ts w-and-orr altt. 1 I Th sho both an amnaton ofhstol o ad an abstra anas ofthe relaton bwe ia and aon The on pnple that not nhbt ps s g gs 20 For ample we m ue hypoth that contrad we-coed theo anor we-tablhed emtal rults We m ance s pceedng countendue 2 The coist condtion whch man that n hypoth aee wth cted eoes unreonable beue t pre the or theo and not the better theo Hypoth contradng we-coed theo e u nce that nnot be obtaned in any other w Proleraon of theo s balr sce whle un mpa ts tl power Uni ao nge theee elmt of the nddual 33 There is no a hower ant and aur that s not pable of mpng our knowledge The whole hsto ofthought s absoed nto sce and ued r mpng e sngle theo Nor poltcal inteerce reed It may be need to ercome the chanism ofsce that rsts altat to the statu quo 39 o theo er ae wth a thefas n man yet t s not alws the th eo that is o bme Fas are sttuted or olo ad a ch 5
6
AGAI NST METHOD
betwefas and theo may beproofofpess It s ao at st n our attpt tond the pnpl mpt n mar obseatonal notons
As an ample ofsuch an apt I amne the tower argument whch the Astotelans used to rte the moton ofthe eah The amt o natural interetaons so ose conneed wth obseatons that t nee a speal to reale ther stce and to tene their contt Galleo nt the natural nteretatons whch are nconsstt wth Cs and rl th othe 65 The n natural nteretatons consttute a n and high abstra obseaton lanage Th are ntroduced and concealed so that one to notce the change that h tak p e method ofanamns Th contan the a ofthe relavity of all moon and the law f circular inera. 77 In ton to natural nteretatons Galeo ao chang sensaons that se to nger Cs He admts that there are such ssatons he pras Csr hng dsregard th he a to he red th with the help of the telescope Hower he oe no theorecal reons why the telce should be eed to e a te pure ofthe s 86 Nor the ntal experience wth the telcepr such reons The t telic obseatons of the s are ndstn enate contrado and n con wth what eone n see wth hs unad And the on theo that could he helped to sarate telc usonsfm vedcal phoma w rted smple tts 0 103 On the other hand there are some telcc phoma whch are pn Cn Galleo ntduc the phoma nnt nce r Cs whle the stuaton s rather that one ruted v Cnsm h a cean smla wth phoma eng fm another rted v the a that telc phoma are thl mag ofthe s
II 106 Such atonal' metho ofsuo are need beuse of the un elmt ' Ma Ln ofdert pas ofsce Cn sm and other stal nedts ofmo sce sued on because reon w equt eled n therpt
A N A L Y TI C A L I N D E X
7
123 Gaeo method wor in otherel wel For ample it n be used to eiminate the isting amts against matealism and to put an d to the philosophical minbo probl the oponding scienc pbl rain untouched hower It not low that it should be uneal alied 125 The Church at the time ofalileo not on kt oser to reon ned th nd in pa now it also consired the ethil and soal nsequc ofGalileo 's vis Its indimt ofGaeo w rational and on ounism and a lk ofpepee n mand a rision 135 Gaileo 's inqui ed on a sma pa ofthe socaled Cin Rolution Aing the raining elts mak it sti more dlt to reconle the elmt with miliarpnpl oftheo aluation 17 I The rults obtained so r suest abolishing the dtinion bwe a ntt of disce and a contt of justcation nos and fas obsational ts and theoretil tes None of the distinions plays a role in stc praice Apts to rce th would he distrous consequces Per til rationalism i r the same reons ppn I
159
16 Final the nd ofcompason that unrli most metholog is possible on in some rather simple It brea wn wh we t to compare nonstc vis with sce and wh we consir the most and most geral and therre most mytholocal pas ofsce itse ppn
209
21 ither sce nor rationali are uneal meures ofcece Th are ilr traditions unaware oftheir histocal ounin 230 et it possible to aluate stanr ofrationali and to impre th The pnpl of impt are neither abe tradition nor bond change and it is impossible to nail th
8
AGAINST METHOD
238 Sce is neither a sngle tradition nor the best tradton there is ctr pele who he become stomed to its prce its bts and its dsantag In a mo it should be saratedfm the state just church are now saratedfm the state 252 The pont ofvi unring ths book is not the rult ofa wel-pnned tran ofthought but of ats prompted ntal counte Anger at the wanton ton ofltural hts om which we al could he leaed at the conceted surance wth whch some nteleua nteere with the l ofple and contpt r the tre phr th use to belsh ther mse w and st is the mote rce behnd my work
Inoducon Sce an stl anarchc tee theoretl anarchm s more humantaan and more lke to couragepros than w-and-orr altat
Ordnung st heuutage mestens dort, wo nchs st Es st ene Manechenung BRCHT
The followig essay is written in the convicon that anarchm, whle perhaps not the most atacve poltcal philosophy, is certainly excellent medicine for tolo, and for the phloshy ofsce The reason is not dicult to nd. Histo generally, and the histo of revoluon in parcular, is always richer in content, more varied, more many-sided, more lively and subtle than even' the best historian and the best methodologist can imagine. • Histo is full of accidents and conunctures and curious uxtaposions of events' 2 and it demonsates to us the complety of human change and the unpredictable character of the ulmate consequences of any given act or decision of men'. 3 Are we really to believe that the naive and simple-mnded rules which methodologists take a their guide are capable of accounng for such a maze of interacons'?4 And is it not clear that successful 'Hstory as a hole, and the hstory of revoluons in parcular, s aays rcher n content, more vared, more mulfo, more lively and ngenious tan is imagned even the bes pares, the most conscious vanards of the most aanced classes (VI Lenn, 'LeWing Comunsm An Infane sorder, Seed W, Vol. , London, 1967, p. ). Lenn is addressng pares and revoluonary vangurds rther than scensts and methodologists; the lesson, hoever, s the me. Cf. fote 5. 2 Herbe Buereld, e ig Interetion ofto, Ne York, 1 965, p. ibd. , p 2 4 ibd, p 25 , cf Hegel, Phishie Gichte, W, Vol. 9, ed. Edard Gans, Berln, 187, p. 9 'But hat erence and history teach us is this, that naons nd
9
10
A G A I N S T T H O D
paton in a process of this kind is possible only for a ruthless
opporunist who is not ed to any parcular philosophy and who adopts whatever procedure seems to t the occasion? This is deed the conclusion that has been drawn by intelligent and thoughtful obseers. Two ve important praccal conclusions follow from this [character of the historical press],' writes Lenin, 5 connuing the passage from which I have just quoted. First, that in order to full its task, the revoluona class [i.e. the class of those who want to change either a part of society such as science, or siety as a whole] must be able to master a forms or aspects of sial acvity without excepon [it must be able to understand, and to apply, not only one parcular methodolo, bu any methodolo, and any variaon thereof i can imagine] . . . ; second [it] must be ready to pass from one to another in the quickest and most unexpected manner.' The exteal condions', writes Einstein, 6 which are set for the scienst] by the facts of experence do not permit him o let himself be too much restricted, in the consucon of his conceptual world, by the adherence to an epistemological system. He, therefore, mus appear to the systemac epistemologist as a type of unscrupulous opporunist. . . .' A compex medium containing surising and unforeseen developmens demands complex predures and dees analysis on the basis of rules whch govements have never leaed anythng from hstory, or acted according to rules that might have derved from it. Every perod has such pecular crcumstances, is in such an ndidua state that decsons ill have to be made, and decsons can ony be made, n t and out of it 'Very clever 'shred and very clever 'NB rtes Lenn n his margnal notes to this passage (Coected Wor, Vol 8, London, 1 96 1 , p. 07.) 5 ibd. We see here very clearly ho a fe substuons can tu a litica lesson nto a lesson for methl Ths s not at all sursng. Methodolo and lics are both means fo mong from one hstorcal stage to another. We also see how an nddual, such as Lenn, ho s not nmdated by tradonal boundares and hose thought s not ed to the ideolo of a pacular profession, can gve useful adce to everyone, phlosophers of scence ncluded In the 19th century the idea of an elasc and hstorcaly infoed methodolo as a maer of course Thus Est Mach rote n hs boo Erkntnis und um, Neudruc, Wssenschafche Buchgesell schaft, Dastadt, 1 980, p. 2 'It s often sad that research cannot be taught That s qute correct, n a cean sense. The schemata oflogc and of indue logc are oflte use for he ntellectual stuaons are never exacy the me. But the examples of great scensts are very suggesve They are not suggestive because e can absact rules from them and subject future research to ther jursdcon; they are suggestive because they mae the mnd nmble and capable of nvenng enrey ne research tradons or a more detailed account of Machs phlosophy see my essay Farewe to Reon, London, 1987, Chapter 7, as ell as Vol 2, Chapter 5 and 6 of my Philoshil Pap, Cambrdge, 1981 6 Albe Ensten, Albe Einstein Philhr Stt, ed PA Schlpp, Ne Yor, 195 , pp68
NTRODUCTON
11
have been set up in advance and without regard to the ever-changing condions of histo Now it is, of coue, ssible to simp the medium in whch a ienst works by simpliing its main actors. he histo of science, aer all, does not just consist of facts and conclusions drawn from facts. It also contains ideas, interetaons of facts, problems created by conicng interetaons, mistakes, and so on. On closer analysis we even nd that science knows no bare facts' at all but that the facts' that enter our nowledge are already viewed in a certain way and are, therefore, essenally ideaonal. This being the case, the histo of science will be as complex, chaoc, full of mistakes, and entertaining as the ideas it contains, and these ideas in tu will be as complex, chaoc, full of mistakes, and entertaining as are the minds of those who invented them. Conversely, a little brainwashing will go a long way in mang the histo of science duller, simpler, more uniform, more objecve' and more easily accessible to eaent by strict and unchangeable rules. Scienc educaon as we know it today has precisely this aim. It simplies science' by simpliing its parcipants rst, a domain of research is dened The domain is separated from the rest of histo (physics, for example, is separated from metaphysics and from theolo) and given a logic' of its own. A thorough aining in such a logic then condions those worng in the domain t makes ther ons more uniform and it freezes large parts of the hsol pcs as well. Stable facts' arise and persevere despite the vicissitudes of histo. essenal part of the aining that makes such facts appear consists in the attempt to inhibit intuions that might lead to a blurring of boundaries A person's religion, for example, or his metaphysics or his sense of humour (his natural sense of humour and not the inbred and always rather nasty nd of jularity one nds in specialized professions) must not have the slightest connecon with hs scienc acvity. His imaginaon is resained, and even his language ceases to be his own. This is again reected in the nature of scienc facts' which are experienced as being independent of opinion, belief, and cultural background. It is thus possble to create a adion that is held together by sict rules, and that is also successful to some extent. But is it rable to support such a adion to the exclusion of everything else? Should we ansfer to it the sole rights for dealing in knowledge so that any resul tha has been obtained by other methods is a once ruled out of cou And did sciensts ever remain within the boundaries of the radins they dened in this narrow way These are the quesons I nend ask in the presen essay. And o these quesons my answer will be a r m and resounding NO.
12
AGAINST METHO
here are two reasons why such an answer seems to be appropriate he rst reason is tha the world whch we want to explore is a largely unknown enty We must, therefore, keep our opons open and we must no resict ouelves n advance Epistemological prescripons may lk splendid when compared with other epistemoogica prescripons, or with genera prncipes but who can guarantee that they are the best way to discover, not ust a few isolated 'facts, but aso some deeplying secrets of nature? he second reason is that a scienc educaon as described above (and as pracsed in our schls) cannot be reconcied with a humntarian attude It is in conict 'with the culvaon of individualty which alone produces, or can produce, welldeveloped human beings 7 it 'mas by compression, like a hnese ladys ft, eve part of human nature whch stands out prominently, and tends to make a person markedly dierent in outine8 from the ideas of raonaity that happen to be fashonable in science, or in the phlosophy of science he attempt to increase liberty, to lead a full ad rewarding life, and the corresponding attempt to discover the secrets of nature and of man, entais, therefore, the reecon of univera standards and of al rigid adions (Naturally, it aso entais the reecon of a large part of contempora science) It s sursng to see how rarely the stulng eect of the aws of Reason or of scienc pracce is examined by professional anarchists Professional anarchsts oppose any knd of resicon and they demand that the individual be permitted to develop freely, unhampered by laws, dues or oblions And yet they swalow without protest all the severe standards whch sciensts and logicians impose upon research and upon any kind of knowledgecreang and knowledgechanging acvity Occasionally, the laws of scienc method, or what are thought to be the laws of scienc method by a parcular writer, are even integrated into anarchism itself 'Anarchism s a world concept based upon a mechanica explanaon of all phenomena, writes Kropotkin9 'Its method of inveson is 7. John Stua Mill, 'On Lberty, in The Phlos oohn Sua M, ed Marshall Cohen, Ne York, 1961, p 258. 8. ibd., p. 265 9 eter Aexeich Krotkn, 'Mode Science and narchism, Kon Runa Pah, ed RW. Badin, Ne York, 1970, pp. 152. 't is one of Ibsens great dsncons that nothng as vad for hm but scence. B Sha, B to Mheh, Ne York, 1921, p xcv Commenng on these and smlar phenomena Sndbeg rites ntbaa 'A generaon that had the courage to get rd of God, to crush the state and church, and to overthro se and morai, sll boed before Scence And in Science, here freedom ought to reign, the order of the day as beleve in the authories or o ith your head.
NRODUCTON
13
that of the exact natural sciences . the method of inducon and deducon' 0It is not so clear,' writes a mode radical' professor at olumbia, 1 that scienc research demands an absolute freedom of speech and debate Rather the evidence suggests that ertain nds of unfr eedom place no obstacle in the way of science ' There are certainly some people to whom this is not so clear' Let us, therefore, start with our outline of anarchsc methodolo and a corresponding anarchsc science There is no need to fear that the dimnished conce for law and order in science and siety that characterizes an anarchism of this nd wil lead to chaos The human neous system is too well organzed for that 11 There may, of course, come a me when it wll be necessa to give reason a tempora advantage and when it wll be wise to defend its rules to the exclusion of everything else I do not think that we are living in such a me today 1 2
10 R. Wol, The Pe ofLibalm, Boston, 1 968, p 1 5 For a cricism of Wol see fote 52 of my essay 'Aganst Method, in Minno Sti in the Philoshy ofSce, Vol 4, Minnealis, 1970 1 1 . Even n undetermined and ambious stuaons, unformi of acon sn achieved and adhered to tenaciously. See Muzafer Sherf, e Pho ofSl No rs, New York, 1 964.
12 This as my opnon n 1 970 hen I rote the rst verson of this essay. Tmes have changed Considerng some tendencies in US education ('lcay coect, academic menus, etc.), n phlosophy posodeism) and in the orld at large I thnk that reason should no be gven greater eght not because t s and alays as fundamental but because t seems to be needed, n cicumstances that cur rather frequenty today (but may dsappear tomoo), to create a more humane approach
1 Ths s sho th an amnaton of htol so and an abstra anass ofthe relaton betwe a and aon The on pnple that not nhbt pros anything goes.
The idea of a method that contains , unchanging, and absolutely binding prnciples for conducng the business of science meets considerable diculty when confronted with the results of hstorcal research We nd, then, that there is not a single rule, however plausible, and however rmly grounded in epistemolo, that is not volated at some me or other. It becomes evident that such violaons are not accidental events, they are not results of insucient knowledge or of inattenon which might have been avoided On the cona, we see that they are necessa for pro gress Indeed, one of the most sikng features of recent discussions in the histo and philosophy of science s the realizaon that events and developments, such the invenon of atomism in anquity, the Copeican Revoluon, the rse of mode atomism nec theo; dispersion theo; stereochemisy; quantum theo), the gradual emergence of the wave theo of light, occurred only because some thnkers either d not to be bound by certain obvious' methodological rules, or because they unwng bke them This liberal pracce, I repeat, is not just a fa of the hsto of science. It is both reasonable and absolute necsa for the growth of knowledge More specically, one can show the folowing gien any rule, however fundamental' or raonal', there are always cir cumstances when it is advisable not only to iore the rule, but to adopt its opposite. For example, there are crcumstances when it is advisable to inoduce, elaborate, and defend hoc hypotheses, or hypotheses whch conadict well-established and generly accepted experiental results, or hypotheses whose content is smaller than the 14
ONE
15
content of the esng and empirically adequate alteave, or self inconsistent hypotheses, and so on 1 There are even circumstances and they cur rather frequently when amt loses its foard-lkg aspect and becomes a hindrance to progress. Nobody would claim that the teaching of sma chldr is exclusively a matter of argument (though argument may enter into it, and should enter into it to a larger extent than is customa), and almost eveone now agrees that what looks like a result of reason the maste of a language, the estence of a richly arculated perceptual world, logical abiity is due partly to indocinaon and partly to a press of wth that preeds with the force of natural law And where arguments do seem to have an eect, this is more often due to their pscal retton than to their antc ntt
Having admitted this much, we must also concede the possibiity of non-argumentave growth in the adult as well as in (the theorecal parts o) nstitutons such as science, religion, prostuon, and so on. One of he few hiners to undersand his feature of he development of nowledge was Niels Bohr:. . . he would never to outine any nished picture, but would patienty go oug e phses of e development of a poblem, sng from some apparent pardox, and gdually leading to its elucidaon. n fac, he never regarded achieed results in any oher light han as sang ints for further exploron. In speculating about he prospects of some line of invesgation, he woud dismiss he usua consideraon of smplici, elegnce or even consisten wih he remar hat such quaies can only propery judged [my itcs] he event. . . . L. Rosenfeld in e Bohr L and Wk se h F and Cola, S. Rosenta (ed.), ew or, 1967, p. 1 1 7. Now science is never a completed press, herefore it is awaybefore he event. Hence simplici, elegnce or consisten are n necesry condions of (scienc) prcce. Considerons such as hese are usuy cricied by he chidish re hat a conadicon enais eveng. But condicons do not enail aning une people use hem in cein ways. nd people will use hem as enaiing eveing ony if hey accept some rher simple-minded rules of derivaon. Sciensts proing heories wih logca faults and obaining interesng results wih heir hep (for example: he results of early fos of he caculus; of a geome where lines consst of ints, planes of lines and volumes of planes; he predicons of he older quntum heory and of eary fos of he quantum heory of rdiaon nd o on) eidenty preed accordng to dierent rules The cricism herefore fals bac on its auhors unless it can be shown hat a logcaly deconaminated science has beer resuts. Such a demonstraon s impossible. Logcaly pefect versions (if such versions est) usually arrve only long after he mpefect versions have enriched cience by heir conibuons. For example, wave mechanics was not a logcal reconsucon of precedng heoies; it was an aempt to presee heir achievements and to solve he physical problems hat had arisen from heir use. Boh he achievements and he problems were produced in a way very dierent from he ways of hose who want to subject everything to he ranny oflogc.
16
AGANST ETOD
We certanly cannot take t for granted that what s possble for a small chld to acqure new modes of behavour on the slghtest provaon to slde nto them wthout any noceable eort s beyond the reach of hs elders One should rather expect that catasophc changes n the physcal envronment wars, the breakdown of encompassng systems o moralty, polcal revolu ons, w ansform adult reacon pattes as well, ncludng mportant pattes of argumentaon Such a ansformaon may agan be an enrely natural press and the only funcon of a raonal argument may le n the fact that t ncreases the mental tenson that preceded and used the behavoural outburst Now, f there are events, not necessarly arguments, whch use us to adopt new standards, ncludng new and more complex forms of argumentaon, s t then not up to the defenders of the status q to provde, not just counter-arguments, but also cona us (Vrtue wthout terror s neecve,' says Robesperre) And f the old fos of argumentaon tu out to be t weak a cause, must not these defenders ether gve up or resort to songer and more rraonal' means? (It s ve dcult, and perhaps enrely mpossble, to combat the eects of branwashng by argument.) Even the most purtancal raonalst wll then be forced to stop reasonng and to use pragan and coeron, not because some of hs reons have ceased to be vald, but because the pcholocal conditions whch make them eecve, and capable of nuencng others, have dsappeared And what s the use of an argument that leaves pople unmoved? Of course, the problem never arses qute n ths form The teachng of standards and ther defence never consss mere ly n putng them before the mnd of the student and makng them as ear as possble. The standards are supposed to have mal usal as well. Ths makes t ve dcult ndeed to dsnush between the lolrce and the ateal of an argument. Just as a well-aned pet wll obey hs master no matter how great the confuson n whch he nds hmself, and no matter how urgent the need to adopt new patte of behavour, so n the ve same way a wel-aned raonalst wll obey the mental mage of his master, he wll conform to the standards of argumentaon he has leaed, he ll adhere to these standards no matter how great the confuson n whch he nds hmself, and he wll be qute ncapable of realzn that what he rerds as the voce of reason' s but a usal aer- of the anng he had receved. He wll be qute unable to dscover that the appeal to reasn to whch he succumbs so readly s nothng but a potical anore
ONE
17
hat interests, forces, propanda and brainwashing techniques play a much greater role than is commonly believed in the growth of our nowledge and in the growth of science, can also be seen from an analysis of the relation betwe ia and aion. It is often taken for granted that a clear and disnct understanding o new ideas precedes, and should precede, their formulaon and heir instu onal expression Fit, we have an idea, or a problem, th we act, ie either speak or build, or desoy Yet hs is cerainly not the way in which small children develop hey use words, they combine them, they play with them, unl they grasp a meaning that has so far been beyond their reach And the inial playful acvity is an essenal prerequisite of the nal act of understanding here is no reason why this mechanism should cease to funcon in the adut We must expect, for example, that the ia of liberty could be made clear ony by means of the ve same acons, which were supposed to eate liberty reaon of a thing, and creaon plus full understanding of a e ia of the hng, are v o pas ofone and the same indisible pcs and cannot be separated wthout bringing the press to a stop he process itself is not guided by a welldened programme, and cannot be guided by such a programme, for it contains the condions for the realizaon of all possible programmes It is guided rather by a vague urge, by a passion (Kerkegaard) he passion gives rise to specic behaviour which in tu creates the cir cumstances and the ideas necessa for analysing and explaining the press, for makng it 'raonal he development of the opeican point of view from Gaileo to the 20th centu is a perfect example of the situaon I want to describe We start wth a song belief that runs counter to contempora reason and contempora experience he belief spreads and nds support in other beliefs which are equally unreasonable, if not more so (aw of inera the telescope) Research now gets deected in new direcons, new kinds of insuments are built , 'evidence is related to theories in new ways unl there arises an ideolo tha is rich enough to provide independent arguments for any parcular part of it and mobie enough to nd such arguments whenever they seem to be required We can say toda that Gaieo was on the right ack, for his persistent pursuit of what once seemed t be a silly cosmolo has by now created the material needed to defend it against all those who wll accept a view only if it is told in a cert ain way and who will ust it only if it contains cerain magical rae, called 'obseaonal reports And this is not an excepon t the normal case: theories become clear and 'reasonable only aer ncherent parts of them have been used for a long me Such
18
A G A N S T H O D
unreasonable, nonsensical, unmethodical foreplay thus tus out to be an unavoidable precondion of clarity and of empirical success. Now, whn we attempt to describe and to understand develop ments of ths knd in a general way, we are, of course, obliged to appeal to the esng forms of speech which do not take them into account and which must be distorted, misused, beaten into new pattes in order to t unforeseen situaons (withou a constant misuse of language there cannot be any discove, any progress) Moreover, since the adional categories are the gospe of eveday thinking (including ordina scienc thinking) and of eveday pracce, such an attempt at understanding] in eect presents rules and forms of false thinking and acon false, that s, from the standpoint of (scienc) common sense' 2 Ths is how da/el thnng arises as a form of thought that dissolves into nothig the detailed determinaons of the understanding', 3 formal logic included. (Incidentally, it should be pointed out that my frequent use of such words as progress', advance', improvement', etc, does not mean that I claim to possess specia knowledge about what is good and what is bad in the sciences and that I want to impose hs knowledge upon my readers. Eveone n read the tes n hs own way and in accordance with the adion to which he belongs Thus for an empiricist, progress' will mean ansion to a theo that provides direct empirical tests for mos of its basic assumpons Some people believe the quantum theo to be a theo of hs knd. For other, progress' may mea unicon and harmony, perhaps even at the expense of empirical adequacy This is how Einstein viewed the general theo of relavity And
ths s that anarchs helps to he ps n any one of the ss one car to choose. Even a law-and-order science will
succeed only if anarchisc moves are casionally allowed to take place) It is clear, then, that the idea of a xed method, or of a xed theo of raonality, rests on too naive a view of man and his sial surroundings. To those who look at the rich material provided by histo, and who are not intent on impoverishing it in order to please their lower insncts, their craving for intellectual security in the form of clarity, precision, objecvity', uth', it wil become clear that there is only one principle that can be defended under a 2. Herbe Marcuse, Reon and Rutn, London, 1 941 , p. 1 3 0. 3 Hegel, Wsscha Lk, Vol. 1, Hamburg, 1965, p. 6.
ONE
19
circumstances and in all stages of human development. It is the principle: anything goes. This absact principle must now be examined and explained in concrete detail
For aple we ay use hotheses that contrad weed theo anor we-establshed etal rults. We ay ance sce pceedng countendue
Examinng he prncple n concrete detal means acng he consequences of counterrules' whch oppose famlar rules of he scenc enerse. To see how hs works, let us consder he rule hat t s experence', or he facts', or expermental results' whch measure he success of our heores, hat agreement between a heo and he data' favours he heo (or leaves he stuaon unchanged) whle dsagreement endangers t, and perhaps even forces us to elminate t Ths rule s an mporant part of all heores of conrmaon and corroboraon. It s he essence of emprcsm. The counterrule' correspondng to t advses us to noduce and elaborate hypoheses whch are nconsstent wh well-establshd theores anor well-establshed facts It advses us to procee untendue.
The counternducve procedure gves rse to he followng quesons Is counternducon more reasonable han nducon? Are here crcumstances favourng ts use? What are he arguments for t? What are he arguments anst t? Is perhaps nducon always preferable to counternducon? And so on These quesons wll be answered n two steps. I shall rst examine he counterrule hat urges us to develop hypoheses nconsstent wih accepted and hghly conrmed theo Later on I shall examne he counterrule hat urges us to develop hypoheses nconsstent wh well-establshed. The results may be summarzed as follows. In he rst case t emerges hat he evdence hat mght refute a heo can often be unearhed only wh he help of an ncompable alteave the advce (whch goes ack to Newton and whch s sll ve popular today) to use alteave ony when refutaons have aready dscredted h orhodox heo puts he car before he 20
O
21
horse so, some of he most important formal properes of a heo re fond by conast, and not by analysis A scienst who wshes to mamize he empirical content of he views he holds and who wants to understand them as clearly as he possibly can must therefore inoduce oher views hat is, he must adopt a pluraltc metholo He mus cmpare ideas wh oher ideas raher han wh 'eperience and he must to imprve raher than discard he views hat have faled in he compeon Preeding in his way he will retain he heories of man and cosmos hat are found in Genesis, or in he Pimander, he wll elaborate hem and use hem to measure he success of evoluon and oher 'mode views He may hen discover hat he heo of evoluon is not as good as is generally ssumed and hat it must be supplemented, or enrely replaced, by an improved version of Genesis Knowledge so conceived is not a series of self consistent heores hat converges towards an ideal view it is not a gradual approach to he h It is raher an ever increasig ocean of utual ncopatble altt, each single heo, each faitale, each myth hat is part of he collecon forcing he ohers into greater arculaon and all of hem conbung, via his process of compeon, to he development of our consciousness Nohing is ever settled, no view can ever be omitted from a comprehensive account Plutarch or Diogenes aerus, and not Dirac or von Neumann, are he models for presenng a knowledge o ths knd in which he histo of a science becomes an inseparable part of he science itsel- it is essenal for its furher elt as well as for giving cont to he heories it contains at any parcular moment prt and lamen, professional and dilettan, hfreaks and iar he all are invited to parcipate in he contest and to make eir conbon to he enrichment of our culture he task of he cient, however, is no longer 'to search for he truh, or 'to praise d, or 'to sstemaze obseaons, or 'to improve predicons h r bt ide effect of an acvity to which his attenon i no mainl directed and which is to ake the weaker ce the strone a he sophits said, and there to sustan the moton of the hoe
he cond 'counterrule which favours hypoheses inconsistent ith obseatons, fs and etal rults, needs no special ef ence, for there is not a single intersng heo hat agrees wh all e own facts in its domain he qeson is, herefore, not wheer conterinducve heories should be aditted into science e eson is, raher, wheher he stng discrepancies between eo and fact should be increased, or dimnished, or what else should be done with them.
22
A A I N S T E T O D
To answer this queson it suces to remember that obseaonal reports, expermental results, factual' statements, either contain theorecal assumpons or se them by the manner in which they are used (For this point cf the discussion of natural nerpretaons in Chapters 6.) Thus our habit of saying the table is brown' when we view it under normal circumstances, with our senses in gd order, but the table seems to be brown' when either the lighng condions are poor or when we feel unsure in our capacity of obseaon expresses the belief that there are familiar cir cumstances when our senses are capable of seeing the world as it really is' and other, equally familiar circumstances, when they are deceived I expresses the belief that some of our senso impressions are veridical whie others are not We also take it for granted that the material medium between the object and us exerts no distorng inuence, and that the physical enty that establshes the contact lght carries a ue picture Al these are absact, and hgy doubtl, assumpons which shape our view of the world without beig accessible to a direct cricism Usualy, we are not even aware of them and we recognize ther eects only when we encounter an enrely dierent cosmolo prejudices are found by conast, not by analysis The material which the stist has at hs disposal, hs most sublime theories and hs most sophiscated technques icluded, is suctured in exactly the same way It ain contans prciples whch are not known and whch, if known, would be eemely hard to test (As a result, a theo may clash wth the evidence not because it is not correct, but because the evidence is conamnated) Now how can we possibly examine somethg we are using all the me? ow can we analyse the tes in whch we habitualy express our most simple and saightfoard obseaons and reveal their presupposions? ow can we discover the knd of world we presuppose when proceeding as we do? The answer is clear we cannot discover it from the insi. We need an tal standard of crcism, we need a set of alteave assumpons or, as these assumpons w be qute general, constung, as it were, an enre alteave world, we need a dream wor in or to discer theatur ofthe real world we think we inhabit
(and which may actually be just another dream-wold) The rst step in our cricism of familiar concepts and procedures, the rst step in our cricism of facts', must therefore be an attempt to break the circle We must invent a new conceptual system that suspends, or clashes with, the most carefully established obseaonal results, confounds the most plausible theorecal principles, and inoduces
TO
23
percepons that cannot form part of the esng perceptual world This step is again counterinducve Counterinducon is, therefore, always reasonable and it has always a chance of success In the following seven chapters, this conclusion wll be developed in greater detail and it will be elucidated with the help of historical eamples One might therefore get the impression that I recommend a new methodolo which replaces inducon by counterinducon and uses a mulplicity of theories, metaphysical views, fai-tales instead of the customa pair theo/obseaon. 2 This impression would certainly be mistaken My intenon is not to replace one set of general rules by another such set my intenon is, rather, to convince the reader that a eholo the ost oous on he ther lits The best way to show this is to demonsate the limits and even the irraonaity of some rules which she, or he, is likely to rerd as basic In the case of inducon (including inducon by falsicaon) this means demonsang how well the counterinducve predure can be supported by argument Aways remember that the demonsaons and the rhetorics used do not express any deep convicons' of mine They merely show how easy it is to lead people by the nose in a raonal way An anarchist is like an undercover agent who plays the game of Reason in order to undercut the authority of Reason (Truth, Honesty,Jusce, and so on) 3
. Clashes' or sspends is meant to be more genera than conadcts' I shal y hat a set ofdeas or acons lashes' wth a concepal system fit s ether nconsstent ih t, or makes the system appear absrd For deals cf Chapter 1 6 below. 2. This s how Professo Ean McMllin intereted some earler papers of ie SeeA Taonomy of the Relaons between Hstory and Phlosophy of Science', nota Stui, Vol 5, Mnneals, 1971 3 Dada', says Hans Richter n D A an Anti-A, not ony had no programme, t was aganst al programmes.' Ths ds not eclde the sifl defence ofprogra mmes o show the chimerical character of any defence, however raona' (In e sa e way an actor or a laywright cold prodce the ote manifesaons of deep le ' n order to debnk the dea ofdeep love' tself Eample Pirandello.)
3 The consist ndition which an that n hypotheses aee with cted heores is unreonable because it pre the olr theo and not the better theo Hypoth ntradiing wel-coed theo e us ince that cannot be obtained in any other way Pleraton of th is balr sce while uni ipai its til power Uni also nge thefree elt ofthe ndidual
In ths chapter I shall present more detaled arguments for he counterrle' hat urges us to noduce hypoheses whch are innsistt wh well-establshed theoes The arguments wll be ndrect. They wll start wh a crcsm of he demad hat new hypoheses must be consstent wh such heores. Ths demand wl be called he consist ndition Pa fae, he case of consstency condon can be dealt wh n a few words. t s well known (and has also been shown n detal by Duhem) hat Newton's mechancs s nconsstent wh Galeo's law of free fall and wh Kepler's laws; hat stascal hermodynamcs s nconsstent wh he second law of he phenomenologcal hat wave opcs s nconsstent wh geomecal opcs; and so Note hat what s beng asserted here s local nconsstency; t may well be hat the derences of predcon are t small to be detected by experment. Note also hat what s beng asserted s not he innsist of, say, Newton's th and Galeo's law, but raher he nconsstency o soe nsequces of Newton's heo he doman of valdty of Galeo's law, and Galleo's law In he last case, he stuaon s especally clear. Galeo's law asserts hat he acceleraon of free fall s a constant, whereas applcaon of Newton's heo to he surface of he earh gves an acceleraon hat s not constant but The conssten condition goes back to Arstotle at least. It plays an mnt pa in Newtons phlosophy (thogh Newton hmself constantly volated it) It is taken for granted by many 20thcentury scentists and phlosophers of scence. 2 Perre Dhem, Th Ai and tur o Physical Tho, New ork, 1962, pp 180
24
R
25
rees (alhough mpercepbly) wh he dstance from he centre
f he earh To speak more absactly consder a heo T hat successfully descrbes the stuaon nsde doman D T agrees wth a nite number of obseaons (et her class be F) and t agrees wh hese obseaons nsde a margn M of error Any alteave hat contradcts T outsde F and nsde M s supported by exactly he same obseaons and s herefore acceptable f T was acceptable (I shall assume hat F are he only obseaons made). The consstency condon s much less tolerant It elmnates a heo or a hypohess not because t dsagrees wh he facts; t elmnates t because t dsagrees wh anoher heo, wh a heo, moreover, whose conrmng nstances t shares It hereby makes he as yet untested part of hat heo a measure of valdty. The only derence between such a measure and a more recent heo s age and famlarty Had he younger heo been here rst, hen he consstency condon would have worked n ts favour The t adequate heo has he rght of prorty over equally adequate aftercomers' 3 In hs respect he eet of the consstency condon s raher smlar to he eect of he more adonal mehods of anscendental deducon, analyss of essences, phenomenologcal analyss, lngusc analyss It conbutes to he preseaon of he old and famlar not because of any nherent advantage n t but because t s old and famar Ths s not he only nstance where on closer nspecon a raher sursng mlarty emerges between mode emprcsm and some of he school phlosophes t attacks Now t seems to me hat hese bref consderaons, alhough leadng to an nteresng tical crcsm of he consstency condon, and to some rst shreds of support for counternducon, o not et go to he heart of the matter. They show hat a alteave to he acceped point of view whch hares ts conrmng nstances cnnot be eliminted by factual reasonng They do not show hat such n lteave s ctable and even less do hey show hat t should be se It s bad enough, a defender of he consstency condon mght pont out, hat he accepted vew does not possess full emprcal uppo Addng new heores of an equal unsatisfo charr wll not mprve he tuaon; nor s here much sense n tryng to rle e ccepted heores by some of her possble alteaves Such replcement wll be no easy matter A new formalsm may have to be lee and familar roblems may have to be calculated n a new 3 C Tresdell, A Program Toward Redscoverng the Raona echancs of the Ae ofReaon, Archor the to o c, Vol. 1, p. 14.
26
GAINST ME
way. Textbooks must be rewritten, university curricula readjusted, experimental results reintereted. And what will be he result of all he eort? oher heo which from an empirical standpoint has no advantage watsoever over and above the theo it replaces. The only real improvement, so he defender of he consistency condion will connue, derives from he ton ofn fs Such new facts will eiher support he current heories, or hey wil force us to modi hem by indicang precisely where hey go wrong. In both cases hey will precipitate real progress and not merely arbia change. The proper predure must herefore consist in he confrontaon of he accepted pont of view wih as many relevant facts as possible. The exclusion of alteaves is hen simply a measure ofexpediency: heir invenon not only does not help, it even hinders progress by absorbing me and manpower hat could be devoted to better hings. The consistency condion elimnates such fruitless discussion and it forces he scenst to concentrate on he facts which, after all, are he only acceptable judges of a heo. This is how he pracsing scienst will defend is concenaon on a single heo to he exclusion of empirically possible alteaves It is worhwhile repeang he reasonable core of ths argument. Theories should not be changed unless here are pressing reasons for doing so. he only pressing reason for changing a theo is disagreement wih facts. Discussion of incompable facts will herefore lead to progress. Discussion of incompable hypoheses w not. Hence, it is sound predure to increase he number of relevant facts. It is not sound predure to increase he nuber of factually adequate, but incompable, alteaves. One mght wish to add hat formal improvements such as increased elegance, simplicity, generalty, and coherence should not be excluded. Bu once hese improvements have been carried out, he collecon of facts for he puose of tests seems indeed to be he only thig left to the scienst. And so it s - provided facts st nd re ble nnt of wheher or not one cons ltt to the theo to be tted This assumpon, on which he validity of he foregoing argument depends in a most decisive manner, I shall call he assumpon o he relave autonomy of facts, or he utonomy pnple It is not asserted by ths principle hat he discove and descripon of facts is independent of heorizing. But it is asserted hat he facts which belong to he empirical content of some heo are available wheher or not one considers aleaves to ths heo. I am not aware hat ths ve important assumpon has ever been explicitly formulated as a separate postulate of he empirical mehod. However, it is clearly implied in almost all invesons which deal wih quesons of
THREE
27
conrmaon and test. All hese invesgaons use a model in which a sngle heo is compared wih a class of facts (or obseaon statements) which are assumed to be 'given somehow. submt hat this is much too simple a picture of the actual situaon. Facts and heories are much more inmately connected han is admitted by he utonomy principle. Not only is he descripon of eve single fact dependent on some heo (which may, of course, be ve dierent from he heo to be tested), but here also est facts which cannot be unearhed except wih he help of alteaves to he heo to be tested, and which become unavailable as soon as such alteaves are excluded This suggests hat he mehodological unit o which we must refer when discussing quesons of test and empirical content is constuted by a whole set ofp erlng ful qute but utul nconsstt theo In he present chapter ony he barest outlines will be given of such a test model. However, before doing his, I want to discuss an example which shows ve clearly he funcon of alteaves in he discove of crical facts. It is now known hat he Brownian parcle is a peetual moon machine of he second kind and hat its estence refutes he phenomenological second law. Brownian moon herefore belongs to he doman of relevant facts for he law. Now could this relaon between Brownian moon and he law have been discovered in a dre manner i.e. could it have been discovered by an examnaon of he obseaonal consequences of he phenomenological heo hat did not make use of an alteave heo of heat? This queson is readily divided into two: (1) Could he relnce of he Brownian parcle have been discovered in his manner? (2) Could it have been demonsated hat it actually rut he second law? The answer to he rst queson is hat we do not know. It is impossible to say what would have happened ifhe kinec heo had not been inoduced into he debate. It is my guess, however, hat in hat case he Brownian parcle would have been rerded as an oddity - in much he same way as some of he late Professor Ehrenhaft s astounding eects were rerded as an oddity, and hat it would not have been given he decisive posion it assumed in contempora heo. The answer to he second queson is simply No. Consider what he discove of an inconsistency between he phenomenon of Brownian moon and he second law would have req uire d. It would have required: (a) measurement of he exact oton of he parcle in order to ascerain he change n its kinec enery pus he ener spent on overcoming he resisance of he uid; and ( precise measurements of temperature and heat transfer he urrounding medium in order to estabish hat any loss
28
AGAINS OD
occurring there was indeed compensated by the increase in the ener of the moving parcle and the work done against the luid. Such measurements are beyond experimental possibilies;4 neither the heat ansfer nor the path ofthe parcle can be measured with the desired precision. Hence a 'direct refutaon of the second law that considers only the phenomenological theo and the 'facts of the Brownian moon is impossible. It is impossible because of the sucture of the world in which we live and because of he laws that are valid in this world. And as is well known, the actual refutaon was brought about in a ve dierent manner. It was brought about via the kinec theo and Einsteins ulizaon of it in his calculaon of the stascal properes of Brownian moon In the course of this predure, the phenomenological theo ) was incoorated into the wider context of stascal physics (T) in such a manner that the consst condton a volted, and it was only th that crucial experiments were staged (invesons of Svedberg and Perrin) 5 4 For details cf R Zs Physik Vol 81 1933 pp 43 5 For these invesgaons (hose philosophical backound derives from Bolmann) cf A Enstein Itigations on th Tho oth Bian Motion ed R Frth New Yo 1956 hch contains all the relevant pape by Entein and an ex hausve bbloraphy by R Frth For the experimental ork of J Perrn see Di tom Lepzg 920 For the relaon beteen the phenomenologica theory and the
kinec theory of von Smoluchosk see Expermentell nachesbare der blchen Thermodnamk dersprechende Molekulahnomene Physikalch Zs Vol 8 1912 p 1069 as ell as the brief note by KR Popper Irreversblty or Eno snce 1905 Bth Joual r th Philosh oSc Vol 8 1957 p 1 5 1 hich summarizes the essenal arments Despite Einsteins ephmaing discoveres and von Smoluchoskis splendd prsentation of their consequences (uvr Ma Smoluchowski Craco 1927 Vol 2 p 226 316 462 and 530 the present situaon n thermodynamcs is eremely unclear escally in vie of the connued presence of some very doubtful deas of reducon To be more specc the attempt is frequently made to determne the eno balance of a complex stattil press by refeence to the (refuted) phoolocal la after hch uctuaons are inseed in an hoc fashion For this f my noe On the Possibilt of a Peetuu obile of the Second Min MattrandMthod Mnneals 1966 p 49 and my paper In Defence of Classical Physics Studi in th Histo and Philhy o S No 2 1 970 It ought to be menoned ncidentally that in 1903 hen Einsten staed hs ork in thermodynamcs there exsted empirical evidence suggesng that Bonian moon could not be a molecular phenomenon See FM Eer Noz zu Browns Molekularbeeng Ps No 2 19 p 843 Exner claimed that the moon as of orders of magntude beneath the value to be expected on the equipaon principle Enstein (Itigations in th Th oth Bian Mt pp 63 esp p 67) gave the folloing theorecal explanaon of the discrepancy: since an obseer orang th dente means of obseaon n a dente manner n never perceive the actual th ansversed n an arbarly small me a ceain mean velit ll
TRE
2
t seems to me m e h hat at his his example example is i s typica typica of o f he relaon between between fairly general heories, or points of view, and he 'facts Boh he rlvance and he refung character of decisive facts can be established only with the help of other theories which, though factually adequate, 6 are not in agreement wih he view to be tested. his being bei ng he case, ca se, he he invenon invenon and arculaon of alteaves may have to precede prec ede he producon produc on of refung refung facts. facts. Empiricism Empir icism,, at least leas t in some of its more sophiscated versions, demands hat he empircal content of whatever whatever knowledge knowledge we possess be increased as much as possible. Hce the nvton oflttes to the v t the ctre ofdssson consttutes n stl p ofthe pl method
Conversely he fact hat he consistency condion elimnates alteaves now shows it to be in disagreement not only wih scienc pracce but wih empiricism as well. By excluding valuable tests it decreases he empirical content of he heories hat are permitted to remain (and hese, as I have indicated above, wll usually be he heories whch were here rst); and it especially decreas dec reases es he number of hose facts hat could show heir limtaons. Ths is how empiricists (such as Newton, or some proponents of what has been called he orhodox interetaon of quantum mechanics) who defend he consistency condion, being unaware of the complex nature of scenc knowledge (and, for that matter, of any form of knowledge) are voiding heir favourite heories of empirical content and hus hus tung tung hem into what hey hey most despis des pise, e, metaphysical docines. 7 John Stuart Mill has given a fscinang account of he gradual ansformaon of revoluona ideas into obstacles to hought. When a new view isis proposed it faces faces a hosle audience and exce excelen lentt reasons are nee n eede dedd to gain gain for for it an even moderately fair fair hearng. hearng . The reasons are produced, but hey are often disregarded or laughed out of court, and unhappiness is he fate of he bold inventors. But new generaons, being interested in new hings, become curious; hey consider consider he reasons, reasons, pursue hem fur furhe herr and groups groups of researchers researcher s iniate detailed studies. The studies may lead to surisig successes successe s (hey also raise lots of dicules). Now nothg succeeds like always appear o hm as an insantaneous velty But t is clear that the velty sceaned thus corresponds to no objecve property of the moon under v vson Cf also Mary Jo Nye Molr Rali London London 1972 19 72 pp 98 98 6 The condion condion of factua factuall adequacy adequacy wll be removed n Chapter Chapter 5 5 The most most dramac dramac conrmaon conrmaon of o f the orthodox view whch made mad e its emprical nure obvious came came by way of Bell Bel lss theorem But Bell was on the sde of Enstein not of hr whom he regarded as n obscuranst Cf Jeremy Bestein Quantu rol Princeton Princeton 1 99 1 pp 3 (for (for Bells Bell s background) background) and p 84 (fo (forr obscuranst) obscuranst)
30
AGANST MOD
success, even if it is success surrounded by dicules The heo becomes acceptable as a topic for discussion it is presented at meengs and large conferences. The diehards of he status quo feel an oblion to study one paper pap er or another, another, to make mak e a few few grumbling comments, comment s, and perhaps perh aps to to join in its exploraon. There Th ere comes hen a moment when he heo is no longer an esoteric esoteric discussion topic for for advanced seminars and conferences, but enters he public domain There are inoducto texts, popularizao popularizaons ns examinaon examinaon quesons ques ons start dealing wih problems to be solved in its terms. Scensts from distan distantt elds and philosophers philosophers,, ying ying to show o, o, drop a hint here her e and here, and his h is ofte oftenn quite uninformed uninformed desire des ire to be on he right side is taken as a furher si of he importance of he heo. Unforunately, his increase in importance is not accompanied by better understanding he ve opposite is he case. Problemac aspects which were originally inoduced wih he help of carefully consucted arguments now become basic principles doubtful points tu into slons debates wih opponents become standardized and also quite unrealisc, for he opponents, having to express hemselves n terms which presuppose what hey contest, seem to raise quibbles, or to msuse words. Ateaves are sll employed but they no longer contain realisc counterproposals hey only see as a background for he splendour of he new heo. Thus we do have success succ ess - but it isis he succes su ccesss of a manoeuvre manoeuvre carried out in a void, overcom overcomng ng dicules dicu les hat were set up in advance for for easy soluon. soluo n. empirical heo such as quantum mechanics or a pseudo empirical pracce such as mode scienc medicne wih its materialisc background can of course point to numerous achieve ments but ny view and ny pracce hat has been bee n around for for some so me me has achievements The queson is whose achievements are better or more important and ths queson cannot be answered for here are no realisc rea lisc alteaves altea ves to provide provide a point of of comparison compa rison A wonderf wonde rful ul invenon has tued into int o a foss fossilil.. There est numerous historical eamples of he press I have just described and various auhors have commented on it. The most important recent auhor is Professor Thomas Kuhn. In his Sture ofStc Rolutons, 8 he disnguishes between science and pre science and, an d, wihin wihin science, between revoluons revoluons and normal nor mal science. scienc e. Prescience according to him, i pluralisc hroughout and herefore in danger of concenang on opinions raher han on hings hings (Bacon (Ba con made ma de a simlar simla r point) point) The two two components of mature 8 Ch 1962
THREE
31
science perfectly agree with the two stages menoned above except that Kuhn doubts that science or, for that matter, any acvity that claims o produce facual knowledge can do without a normal component. Fossis, he seems to say, are needed to give substance to that cur in the the revol revoluona uona compon component ent - but he adds the debates that hat the latter canot advance without alteaves. Two earlier authors are Mill and Niels Bohr. Mill gives a clear and compelling o f the ansion from from the earl e arlyy stag sta g of o f a new ne w view to its descripon of orhodo. Debates and reasoning, he writes, are features belonging to periods of transion, when old noons and feelings have been unsettled and no new dtrines have yet succeeded to their ascendancy At such mes people peopl e of any mental acvity, acvity, having havin g given given up their od beliefs, and not feeling quite sure that those they sll retain can stand unmodied listen eagerly to new opinions But ths state of things is necessarily transito: some parcular body of dtrine in me rallies the majority round it, organes sial instuons and modes of acon confoably to itself, educaon wthout the mtl impresses ths new creed upon the new generaon wthout pcs tht he led to t and by derees it acquires the ve same power of compression, so long exercised by the the creeds of whch it had taken place.
account of the alteaves replaced, of the press of replacement, of the arguments used in its course, of the sength of the old views and the weaknesses of the new, not a 'systemac account but a hstol count of eh stge ofknowledge, can aleviate aleviate these drawbacks and increase the raonaity of ones thorecal commients. Bohrs presentaon of new discoveries has precisely this patte; i contains preliina suaries sueyng the past, move movess on to the 'prese ' present nt state of knowledg knowledge e and ends up by making general suggesons sugge sons for the fu ture. ture . Mils views and Bohrs predure are not only an expression of their libera atude; atude ; they also als o relect relec t their convic convicon on that a plurim of ideas and forms of life is an essenal part of any raonal inquiry conceing the nature of thigs. Or, to speak more generaly
Unnm of non my beng bengr r d chu chu rch r the thefghted fghted or eedy ms of some so me nt nt or mo mo myth yth or r th the wek nd n d 9 Autobography quoted from Esstl Wo ofJohn Stua Mill ed M Leer New Yor Yor 1 965 p 1 1 9 my emphasis 10 For a more detaled account cf my Philoshil Pa Vol 1 Chapter 16 secon 6
32
AA NS TO
iino iinow w ofso ofsome me rn rntt e ofi ini nion on is necess necess r objee nowle nowledde And method method tht cour cou rg e is o o the the on on method method ext ent to which tht is comptible with humnitn outlook o the extent
the conitency condion delimit variety, it contain a theological element which lie, of coure, in the worhip of fact' o characteric of nearly all empiricim. 11
1 1 It is nterestng to see see that the the platudes platude s that drected the Protestant Protestantss to the ble are often almost idencal wth the platudes which drect empicists and other fundamentalist to thir foundaon vz experence Thus in his Num anum Bacon demands that all preconceived noons (aphorsm 36) opinons (aphorisms 42) even r (aphorsms 59 121) be adjured and renounced wth and olemn resoluon and the understandng must be completely freed and cleared of them so that that the access to the the kngdom kngdo m of man which is i s founded founded on e sciences may resemble that to the kngdom of heaven where no admission s conceded except to chldren (aphorsm 68) In both cases dsputaon (which is the consideraon of alteaves) is cricized n both cases we are invted to dense wth it and n both cases we are promised an mmedate percepon here of God and here of Nature For he he heorecal background of his smilary cf my essay Classical Cl assical Empircsm n Mtho olo lol lHta Htag gooNwton wton Oxfo RE RE Buts (ed) ThMth Oxford rd and Toronto 1 970 970 For the song connecons between Puritansm and mode science see RT Jones n and Ms Califoa 1 965 965 Chapters 57 A horoug horoughh exminaon of he factors hat inuenced he rs of mode empircism n England is found n RK Sc Tc Tch nol and So So in Stth Ctu Ctu Enland Enland New York Meon Sc York 1 970 (bk version of he he 1 938 aricle)
4 hower nt nt nd nd bs bsurd tht tht s not n ot c cpble pble of of There s no mp mpng ou our nowl nowledge The whole hsto hsto of thought thought s bsoed bsoed nto nto sce nd used r mprng mprng e sng snglle theo. theo. Nor s poltcl poltcl reed.. It my my be need to er ercome come the the ch chn nsm sm ofsce sce nteerce reed tht rests lttes to the sttus quo
This nshes the discussion of part one of counterinducon dealing with the invenon and elaboraon of hypotheses inconsistent with a point of o f view that is highl conrmed and generaly g eneraly accep a ccepted. ted. The result was that a thorough examinaon of such a point of view may involve incompable alteaves so that the {ewtonian) advice to to postpone alteaves alteaves unl afte afterr the rst diculty diculty has arisen means putng the cart before the horse. A scienst who is inter ested in mamal empirical content, and who wants to understand as many aspects of his theo as possible, will adopt a plural isc methodolo, he will compare theories with other theories rather than with experience, data, or facts', and he will y to imprve rather than discard the views that appear to lose in the compeon. 1 For the alteaves, which he needs o keep the contest going, going , may may be taken from from the past as well. As a matter of fact, they may be taken from wherever one is able to nd them from ancient myths and mode prejudices; from the ucubraons of experts and from the fantasies of cranks. The whole histo of a subject is ulized in the attempt to improve its most recent and most advanced' stage. The separaon between the hiso of a science sc ience its philosophy and and the science itself dissolves into thin It is therefore imant that the alteaves be set against each other and not e isolated or emasculated by some form of demythologizaon. Unlike Tillch Bulann and their followers we should regard the worldvews of the Bible the altat smolo smolo which can be ilmesh epic the Ilad the Edda as fully edged altat ue ue t mdi and even to to replace the scienc scien c cosmoloies cos moloies of a iven iven period p eriod
33
3
AGA N S METHOD
air and so does the separaon between science and non-scence 2 This poson, which is a natural consequence of the arguments presented above, is frequently attacked not by counterarguments, which would be easy to answer, but by rhetorical quesons. f any metaphysics goes,' writes Dr Hesse in her review of an earlier essay of mine, 3 then the queson arises why we do not go bk and exploit 2 An account and a uly humanitaran defence of ths son can be found inS Mlls L Poppers philosophy which some people would lie to lay on us as the one and only humantarian ronalism n existence today s but a le reecon ofMill It s specialized formalsc and elist and devod of the conce for ndvidual happness that s such a charcterisc feature of Mill We can understand ts peculiares when we consider (a) the background of logical sivism which plas an imant role n the Logc ofStc Dse (b) the unrelenng purtanism of ts author (and of most of his followers) and when we remember the inuence of Hariet Taylor on Mlls life and on his phlosophy There is no Hariet Taylor n Poppers life The foregoing arments should aso have made t clear that I regard proliferaon not just as an exteal catyst of progress as Lakatos suggests n his essays (History of Science and Its Raonal Reconsucons Boston Stud Vol. 8 p 98 Popper on Demarcaon and Inducon MS 1 970 p 21 ) but as an essenal ofit Ever since Explanaon Reducon and Emprcism (Mnnota Stud Vol 3 Minneals 1962) and especialy in How to Be a Good Empircst (DewareStud Vol 2 1 963) I have ared that alteaves ncrease the empirica content of the views that happen to stnd in the cene of attenon and are therefore ncsa s f the falsng press (Laatos History fn. 27 descrbng his own sion) In Reply to Cricsm (Bostn Stud Vol 2 1965) I nted out that the princple of proliferaon not only recommends invenon of new alteaes it also prevents the eliminaon of olr theores which have been refuted The reason is that such theores conibute to the content oftherictorous rvals (p 224) This agrees wth Laatos obseaon of 971 that alteaves are not merely catalysts whch can later be removed n the ronal reconsucon (Hstory fn 27) ct that Lakatos attrbutes the psychologisc view to me and my views to himself Considering the agument in the text it s clear that the ncreasing separaon ofthe hstory the philosophy ofscience and ofscience itselfs a dsadvantage and should be terminated in the nterest of al these tree discplnes Otheise we shall get tons of mnute precse but utterly barren results 3. Mary Hesse Rato No 9 1967 p. 93 cf B.F Sinner Bond Frm and D New York 1971 p 5 No mode physicist would tu to Arstote for help. This s nether ue nor would t be an advantage if it were ue. Aristotelan ideas nuenced research long after they had allegedly been removed by eary mode asonomy and physics any history of 17th or 1 8thcentury science wlshow that (example John Helbronns maellous El n th th and th Ctu Berkeley and Los Angeles 1979) They resufaced in bolo in the termodynamcs of open systems and even in mathemacs Arstotes theory oflomoon (whch has the consequence that a movng object has no precise length and that an object having a precse laon must be at rest) was more advanced than the Galean view and showed that des whch in our me emerged from emprical researh can be obtained by a careful anayss of the problems of the connuum (detals on ths int n Chapter 8 of my Farew to Ron London 1987). Here as elsewhere the propagandsts of a naive scensm give themselves the air of presenng arments when al they do s spread unexamned and llconceived rumours.
FOUR
35
the objecve cricism of mode science avalable in Arstotelianism, or indeed n Vood?' and she insinuates that a cricism ofthis kind would be altogether laughable. Her insinuaon, unfortunately, assumes a great deal of ignorance in her readers Progrss was often achieved by a cricism from the past', of precisely the kind that is no dismissed by her After Aristotle and Ptolemy, the idea that the eath moves that sange, ancient, and enrely ridiculous',4 ythagorean view was thrown on the rubbish heap of histo, only to be revived by Copecus and to be forged by him into a weapon for the defeat of its defeaters. The Hermec wrings played an important art in this revival, which is sll not suciently understood,5 and they were studied with care by the geat Newton himself 6 Such developments are not surising No idea is ever examined in all its ramicaons and no view is ever given all the chances it desees Theories are abandoned and superseded by more fashionable accounts long before they have had an opportunity to show their virtues Besides, ancient docines and primive' myths appear sange and nonsensical only because the informaon they conta is either not known, or is dstorted by phiologists or anthropologists unfamiliar with the smplest physical, medical or asnomca knowledge7 Vood, Dr Hesse's pce rstnce, is a 4 Ptolemy, Syn quoted aer the anslaion of Manius, D Cudus Ptole Hanuch rAstnome Vol. , Leipzg, 963, p 8 5 or a sive evauaon of the role of the hermec ings during the Renaissance cf Yates, Gorno Bno and the Hetc Trtn London, 963, and the ltertue given there or a cricism of her iton cf. the acles by Mary Hesse and Edward Rosen in Vol 5 of the Mnnota Stu fr the Phl ofSce ed. Roger Stuwer, Minnesota, 970 RS Wesan andJ.E. McGure, Hetm and the Stc Rutn Los Angeles, Clark Memoria Lbrry, 977, as well as Brian Vickers,JouofModHitor 5 1 , 1 979.
6. CfJM Keynes, ewton the Man', in Essays and Setches in Biogphy, ew York, 956, and, in much greater detail, McGuire and Raansi, ewton and the "Ppes of Pan Not and Rer ofthe Ral Soe Vol. 2 , o 2, 966, pp. 08 or more detailed accounts cf rank Manuel, e Relon ofIs Newton Oxford, 974, RS Westfall's monument biography, N at Rt Cambridge, 980, wth literatu re, as well as Chapters x and xi of R opin, e rd For n tethtu Thought Leiden and ew York, 992. 7 . or the scienc content of some myths cf. C de Sanllana, e n of tc Thought ew York, 96, especialy the Proloe. We can see then', ites de Sanllana, how so many myths, fantsc ad aba n semblance, of which the Grek tale of te Argonauts is a late ospring, may provide a termnolo of image mofs, a kind of code which is beginning to be broken. It was meant to low those who kne (a) to determine unequivaly the ion of given plnets in respect to the eah, to the rmament, and to one another (b) to present what knowledge there was of he fabric of the world in the form of tes about how the world began' There are two reasons hy ths code was not discovered earler One is the rm convicon of
36
AGAIS MEOD
case in point Nobody nows it, evebody uses it as a paradigm of backwardness and confusion And yet Voodoo has a rm though sll not sufciently understood material basis, and a study of its manifestaons can be used to enrich, and perhaps even to revise, our nowledge of physiolo 8 A even more interesng example is the revival of adional medicine in Communist China We start with a familiar develop ment 9 a great couny with great adions is subjected to Weste domnaon and is exploited in the customa way A new generaon recoizes or thinks it recognizes the material and intellectual superiority of the West and aces it back to science Science is imported, taught, and pushes aside all adional elements Scienc chauvinism umphs What is compable with science should live, what is not compable with science, should die' Science' in ths context means not just a specic method, but all the results the method has so far produced. Things incompable with the results must be eliminated Old style doctors, for example, must either be historians of scence that science did not sta before Greece and that scientic results can only be obtaned wth the scienc method as it is prcsed today (and as it was foreshadowed by Greek scientists). The other reason is the asonomical, geologica, etc., ionce f most Assyriologss, Aeptologss, Old estment holars, and so on: the apparent primitism of many myths s just the reection of the primive asonomica, biologica, etc, etc., knowedge of their collectors and trnslators Since the discoveries of Hawins, Marshack, Seidenberg, van der Waerden (Gmet and Algebra n Ant Clatons New York, 1983) and others we hae to admit the existence of an inteaonal palaeolthic asonomy that gave rse to schls, obseatories, scientic tradons and most interesting theories. hese theories, which were expressed n siological, not in mathemacal, terms, have le their aces in sagas, myths legends and may be reconsucted in a twofold way, by going f�ad into the present from the material remains of Stone Age asonomy such as marked stones, stone obseatories, etc., and by going k into the past from the lterry remains which we nd in sagas, legends, myths An example of the st method is A. Marshack, e R ofClton New York, 1972 An example of he second s de Sanllana-von Dechend, Hamlets Mll Boston, 1969 8 Cf Chapter 9 of LiStrauss, Stural Antholo New Yor, 1967 or he physiologic basis of Vood cf CR Rchter, 'he Phenomenon of Unexplained Sudden Deah, in Gan (ed), The Physol B ofPcht s we as WH Cannon, B Chang n Pan Hung Fear and Rage New Yor, 1915, and ' "Vood Deah, n Amen Anthologst ns, xliv, 1942 The deailed biologica and meteorological obseaons made by socalled 'primives are reed in LviSauss, TheSageMnd London, 1966 9 R.C. Crozier, Trtonal Medne n Mo Chna Cambridge, Mass, 1968. he auhor gives a very nteresng and fair account of developments wih numerous quotaons from newspapers, bks, pamphlets, but is often inhibited by his respect for 20thcentury science. 10 Chou Shao, 1933, as quoted n Croizier op. cit., p 109 Cf. also D.WY Kwok, Stsm n Chne Thought ew Haven 1 965
U
37
reoved from medical pracce, or they must be re-educated Herbal edicine, acupuncture, mobuson and the underlyng philosophy are a thing of the past, no longer to be taken seriously This was the aude up to about 1954 when the condemnaon f bourgeois eleents in the Minis of Health started a campai for the revival of adional medicine No doubt the campai was polically inspired It contained at least two elements, (1) the idencaon of Weste science with bourgeois science and (2) the refusal of the party to exempt science from polical superision 1 1 and to grant experts special privileges. But it provided the counterforce that was needed to overcome the scienc chauvinism of the me and to make a plurality (acually a duality) of views possible. (This is an iportant point It often happens that parts of science become hardened and intolerant so that proliferaon ust be enforced from the outside, and by polical means Of course, success cannot be guaranteed see the Lysenko aair But this does not remove the need for non scienc conols on science) Now this polically enforced dualism has led to most interesng and puzzling discoveries both in China and in the West and to the realizaon that there are eects and means of diaosis which ode medicine cannot repeat and for which it has no explanaon It revealed sizeable lacunae in Weste edicine Nor can one expect that the customa scienc approach will eventually nd an answer In the case of herbal medicine the approach consists of two steps 12 irst, the herbal concocon is analysed into its chemical constuents hen the specc eects of each constuent are deterined and the total eect on a parcular organ explained on their basis This neglets the possibility that the herb, taken in its enrety, changes the state of the whole organism and that it is this new state of the whole oganis ather than a specic pa of the herbal concon, a magic ullet', as t were, that cures the diseased organ Here as elsewhere owledge is obtained from a mulplicity of views rather than from e etermined applicaon of a preferred ideolo And we realize at polifeaon may have to be enforced by non-scienc agencies wose ower is sucient to overcome the most powerul scienc isons xamples ae the Curch, the State, a polical party, uli isontent, or money the best single enty to get a mode siest away from what his scien conscience' tells hi to pursue is l r (or, more recently, the Swiss franc) r the tensions between 'red and 'ee cf. . Schurmann, Ilo and a ition ommunt hina Berkeley, 966. 2 Cf. M.B Krieg, Gre Medne New York, 1964.
38
A A I S M E H OD
Pluralism of theories and metaphysical views is not only important for methodolo, it is also an essenal part of a humanitarian outlook Progressive educators have always tried to develop the ndividualty of ther pupls and to bring to fuion the parcular, and somemes quite unique, talents and beliefs of a child Such an educaon, however, has ve often seemed to be a fule exercie in day dreamng For is it not necessa to prepare the young for life a it ul is Does this not mean that they must lea one pilr set of is to the exclusion of everything else? And, if a ace of their imaginaon is sll to remain, will it not nd its prper applicaon in the arts or in a thn domain of dreams that has but little to do with the world we live in? Will this predure not nally lead to a split between a hated realty and welcome fantasies, science and the arts, careful descripon and unresained self-expression? The argument for prolferaon shows that this need not happen It s possble to retin what one might call the freedom of arsc creaon nd o use it to the fu, not just as a road of escape but as a necessa means for dscovering and perhaps even changing the features of the world we live n Ths coincidence of the part (ndividual man) wth the whole (the world we live in), of the purely subjecve and arbia with the objecve and lawful, is one of the most important arguments in favour of a pluralisc methodolo For detas the reader s advised to consult Mill's maicent essay On Libe 3
13. Cf. my account of ths essay n Vol , Chapter 8 and Vol 2, Chapter 4 of my Phloshl P Cf also Appendix of the present essay.
5 o theo ees with the facts in its min yet it is not lways the th tht is to blme Fs re constituted olr iolo nd h bewefs and theoes my beproofofpess It is lso t st in our ttept tond thepnpl implit in milir obsetionl notions.
Considerin now the invenon, elaboraon and the use of theories ic are inconsistent, not ust with other theories, but even with emts, fs, obsetions, we may start by poinn out that no single theo e with l the nown fs in its min And the ouble is not created by rumours, or by the resut of sloppy rcedure t is created by experiments and measurements of the hihest precision and reliability It will be convenient, at this place, to disnuish two dierent nds of disareement between theo and fact numerical disaree ment, and qualitave faiures The rst case is quite familiar a theo makes a certan numerical redicon and the value that is actually obtained dirs from the redion made by more than the marin of error Precision istruments are usually involved here Numerical disareements aound in science They ive rise to an ocean of anomalies' that suounds eve sinle theo 1 Tus the Copeican view at the me of Galileo was inconsistent with fts so plain nd obvious that Galileo had to call it surely se Tere is no limit to my astonishment,' he writes in a or the ean and various ways of dealing wth it cf. my 'Reply to Cricism oo ol. 2 1965 pp. 224. 2 allo alilei The Ass quoted in S. Drake and C. D. OMalley (eds) The ot o the ome of London 1960 p. 185 . The 'surely fase refers to the condemnation by Church authorities. But as wll be explaned in the course of the book and epecally n Chapter 1 3 the condemnaon was based in pa on the 'philosophcal burdi of the idea of a moving earth i.e on its empirical failures and its theorecal •ndequay. See also the next quotaon and fote. 'As to the system of tolemy te alileo on this point (184) 'neither Tycho nor other asonomers nor even
39
0
A A I N S T M E T HO D
later work, 3 when I reect that starchus and Copeicus were able to make reason so conquer sense that, in deance of the latter, the former became misess of their belief' Newton's theo of gravitaon was beset, from the ve begining, by dicules serious enough to provide material for refutaon 4 Even quite recently and the non-relavisc domain it could be said that there est numerous discrepancies between obseaon and theo' 5 Bohr's atomic model was inoduced, and retaed, in the face of precise and unshakeable cona evidence. 6 The special theo of relavity was retaed despite Kaufmann's unambiguous results of 1906, and despite D.C Miller's experiment7 The general theo of relavity, Copeicus could cleary refute it, inasmuch as a most imnt agument taen from the movement of Ma and Venus aays stood n their ay The 'most imnt agument and Gaeos resoluon are discussed in Chapters 9 and 10 3 Gaeo Gaei, D Cnng the T Ch W Systs Bereley, 1953 p 328 4 According to Non the 'mutua acons ofcomets and planets un one another give rise to 'some nconsiderble irrelities hich wlbe apt to increase, the system nts a eformaon, t New or, 1 952, p. 42 Wat Non me that gvitaon dtubs the plnets in a ay that is ey to blo the plnetary system ap. Babylonin data as used by tolemy show that the planetary system has remaned stble for a long me Neton concluded tat it ws beng peroicy 'reformed by divine nteentions God acts as a stabiing force in the planetary system (nd in the orld as a hole, ch s constanty losng moon through preses such as inelasc coions) One of the 'rrearies considered by Neton, the great inequai of Jupiter and Satu (np ansl. Moe ed. Cajor, Bereley, 1934, p. 397) was shon by Laplace to a periodic dstubance th a la period Then oncar found that the series developments customary in the caculatons oen diveged aer they d shon some convegence he Bruhns discovered that no quntitave methods other thn series expnsons could resoe the n- problem This as the end ofthe purey quantave period n celesa mechanics (detils nJ. Moser, Anna ofMath Stud Vol. 77, 1973, rinceton). See aso M. Ryabov An Elta Su ofCe Mean Ne or, 196 1, for a suey and quantitae results ofvarious methods of caculation. The quatave approach is brey described on pp. 1 26f Thus it t mo than to hundred yea before one of the many dicues of ths rther succes theory was n resolved 5. Broer-lemence, Meth ofCeltl Mechan Ne or, 1961 Aso RH ice, 'Remars on the Obseaona Basis of Generl Relativi, in H. Chiu nd W.F. Homan eds), Gratn and Ret New or, 1 964, pp. 1 -16. For a more detaed dscusson of some of the dicules of clssica celestia mechanics, cf. J. Cha, La e rettet Mchanque lte Vol. 1 , Chaptes 4 and 5 , aris, 1928. 6. Cf Max Jammer, e Cntual Delmt ofQuantum Mechan Ne or 1966, section 22. For an anayss cf secon 3c/2 of Laatos 'Fascaon and the Methodolo of Scentic Research Progmmes in LaatoMusgve (eds), Ctcm and the Gwth ofKnowledge Cambridge 1970. 7. W. Kaufmann 'ber de Konstuon des Eleons, Ann Ps No. 19 16, p. 487. Kaufmann stated his concluson quite unambiously and in italics e
I
4
tog surisingly successful in a series of occasoally rather dramac tests, 8 ad a rog me in areas of celesal mecanics dierent from the advance of the perielion of Mercu 9 In the es the aguments and obseaons of Dicke and oters seemed to endanger even this predicon The problem is sll unresolved 1 r lts ofthe meur are not atble wth thenmtal sumpton ofLor and Enstn Lorens reacon ' . . t seems very lkely that we shall have to relinquish this dea altogether (Th ofEns second edion, p. 213). Ehrenfest
Kaufmann demonstrates t Lorens deformable elecon s ruled out by the measurements ('Zur Stabilittsfrge bei den Bucherer-Langevn Elekonen Ps Zs Vol. 7 1 p. 302). Poincars reluctance to accept the 'new mechanics of Loren can be explained at least n pa by the otcome of Kafmanns experment. Cf. Sce and Meth New ork 19 Bk I Chapter 2, secon v, where Kafmanns experment is discussed, the conclusion being that the 'prnciple of relavi . . cannot have the fndamental imance one was nclined to scribe to t. Cf. also St Goldbeg, 'Poncars Silence and Einsteins Relav, Bth Joul rthe Hto ofSce Vol. 5 1970, pp. 73 and the literture given there. Einstein alone regarded the results as 'mprobable because their basc assumpons, om which he mass of the moving elecon is deduced are not sgested theoretical systems hich encompass wider complexes of phenomena ahuch Rktund lekttt Vol 4, 17, p 349) Millers work was studied Loren for many years, bt he could not nd the oble It was only n 1955 twenve years aer Mller had nished his experments, that a sasfactory account of Millers result was found Cf RS Shankan, 'Convesations t Einstein Am Jou . Ps Vo 31 1963, pp 7-57 especially p 5 1 , as well as fotes 19 and 34 cf also te inconclusive dscssion at the 'Conference on the Mchelson-Morley Experment, Astsl ol ol 68 1928 pp 341 Kafmanns experiment as analysed by Max Planck and found to be not decisive what had stopped Ehrenfest Poincar and Loren did not stop lanck Wy My con ecture s that Plancks rm belief in an obecve reali and hs sumpon that nsteins theoy was abot sch a reali made him a lile more crical Details in Capter 6 ofElie Zahar Enstn Roluton La Salle Ill 1989 Sc as the test ofthe eects of rav pon liht that was carred ot in 199 by ddnon and Crommelin and evalated by Eddnon or a colofl descripon of e even and its impact, cf. CM Will W Enstn Rght? New York 1986 pp 75. 9 Ca op cit p 230. 0. Repeang consideraons b Newcomb (reed for example in Cha op . pp. 204) Dicke inted ot that an oblateness of the sn wold add classical e o Mercrys moon and redce te excess (compared wth Newtons theory) advance o its perihelion. Measrements by Dicke and Goldenberg then fond a eence of 52 km between the equatoral and lar dameter of the sn and a oeponding redcon of three seconds of arc for Mercry - a sizeable deviaon m e elavisc vale. This led to a considerable conover conceng the aca of the Dicke-Goldenberg experment and to an increase in the nmber of nn Enteinan theores of grvtaon. Technical details in C.M Wll Th and t n Grtatonal Ps Cambrdge 198 1 pp. 1 76 a plar sey cldng later developments in W Enstn Rht Chapter 5. Note how a new theory Enein s theory of grvitation) which is theorecally plasible and well conrmed an e endangered by exploing its 'refted redecessor and caring ot approprate exements. C. also R.H. Dicke o. ct.
42
A A I N S T ME T H O D
O the other hand there est numerous new tests, both nsde the planeta system and outsde of t 1 1 that provde conrmaons of a precson unheard of only twenty years ago and unmagned by Ensten n most of these cases we are dealing with quantave problems which can be resolved by dscoverng a better set of numb but whch do not force us to make qualtave adusents 12 1 1 . Tests outsde he planeary system (cosmolo, black holes, plsars) are needed to examne alteatives hat agree wh Einstenan relavi inside he solar system There now ests a considerble number of sch alteaves and special steps have been aken to classi hem and to elcidate heir similarties and derences. Cf. he inodcon to C.M Will, op. ct. 12. The staon jst descrbed shows how silly t wold be to approach scence from a naive-falsicaonist persective. Yet his is precsely what some philosophers have been ting to do Ths Herbe eigl (Mota Stud Vol. 5, 1971, p. 7) and Karl Popper (e Kowlee p 78) have tred to tu Einstein into a naive falsicaonst. eigl tes: ' I f Enstein reled on "bea, "harmony, "mmet, "elegance n consucting . . his general heory of relav, it mst nevertheless be remembered at he also said (n a lecture n Prae in 1920 - I was present hen as a very yong student) "If he obseations of he red shift in he spectra of massve sars dont come ot qanavely in accordance wh he princples of gner rela, hen my heory wlbe dst and ashes. Popper tes 'Einstein . . . said hat ifhe red shift eect . was not obseed in he case of whte dwar, his heory of gener relat would b refuted Popper ges no sorce for his story, and he most likely has it om eig Bt egs story and Poppers repetion conct wh he nmeros casons where Ensten emphases he 'reason of he maer ('de Ven der Sache) er and above 'vercaon ofile eects and hs not only in casa remarks drng a lecure, bt in tng Cf he quoation in fote 7 above, whch deals wh dculties of he specia heory of rela and precedes he alk at which eig was present. Cf. aso he leers to M Besso and K Seelig as quoted in G Holton, 'Inuences on Einstens Eary Work, an No 3, 19, p. 242, and K Seelig, A Einst Zurch, 9 p. 27 . In 952 Bo ote to Einstein (BoEitn L, New ork 97 p. deaing wh Freundlihs analysis of he bending o flight near he sun and he red shi) 'It realy ls as f your formula is not quite orrect It lks even worse in he ase of he red shi [he rial ase referred to by Feig and Popper] his is muh smaller han he heoreal value towards he ene ofhe suns disk and muh lar at he edges . . Could his be a hint ofnonlinear? Einstein (eerof 2 May 952 op. it. p. 92) replied. 'Freundlih . . ds not move me in he slightest. Even if he deeon oflight he perhelia moveent or line shift were unknon he grviation equaons would sll be onvining beause hey avoid he nea system (he phantom whih aets evehing but is not tself aeted) It real strange ta
human bn are noal afto the stngt aum while th are alws inclined to tmate meung r' (my ils) How s hs onit (beeen Feigs tesony and Einstens writings) to be explained? It annot be explaned by a change n Einsteins attde. is disrespetfl attude towards obseaon and experent was here from he very beginning as we have seen. It might be explained eiher by a isake on Fegs pa or else as anoher nsane of Einsteins 'oprtnism - f. text to fote 6 of he Intduion On he last page (p. 9) of his b die Spiee und allne Relattthe,
FIVE
43
The second case, the case of qualitave faiures, is ess famiar, but of much greater interest In ths case a theo is inconsistent not with a recondite fact, that can be uneathed with the help of complex equipment and is known to experts only, but wth circumstances whch are easiy noced and whch are famliar to eveone The rst and, to my mind, the most important example of an inconsistency of this knd is Parmenides' theo of the unchanging and homogeneous One Ths theo ilusates a desre that has propelled the Weste sciences from their incepon up to the present me the desire to nd a unity behnd the many events that surround us Today the unity sought is a theo rich enough to produce all the accepted facts and laws; at the me of Parmenides the unity sought was a substance. Thales had proposed water, 3 Heracltus re, Anamander a substance which he caled the apein and whch ould produce all four elements without being idencal with a single one of them Parmenides ve what seems to be an obvious and rather ivial answer the substance that underles eveg that is is Being But ths ivial answer had surising consequences For example, we can assert that (rst prnciple) Being is and that (second principle) Not Being not Now consider change and assume it to be fundamenta Then change can only go from Beng to Not Beng Bt according to the second prnciple Not Being is not, whch mea th at there is no ndamenta change Next consider derence and assume it to be fundamenta Then the dierence can only e between Beig and Not Being But (second prnciple) Not Being not and therefore there est no dierences in Being it a singe, unchanging, connuous blk. Parmenides knew of coe that people, helf included, perceive and accept change and dierence; but as hs argument had shown that the perceived presses could not be fundamenta he had to rerd them as merely apparent, or decepve Ths is indeed what he sad thus ancipang al those sciensts who conasted the real' world of science with the eveday world of quaes and emoons, declared the latter to be mere appearance' and ied to base ther arguents on ojecve' experiments and mathemacs exclusively He also ancipated a Brunsk 922 Ensten rtes ' If the red shi of the speca lnes aused by the gravtaona tena dd not est then the gener theory of relav ould be ntenabl e. Ds ths onict th Enstens avaer attude toards obseaon as desbed above? It ds not. The pasge speaks of the red sh not of oations ofit 3 The follong aount higy speulae Dets n Vols and 2 ofW.KC. Gthe A Histo of Greek hilos, Cambrdge 962 and 965 as we as in
Chapters 1, 2 and 3 ofmy Farewel to Reon.
44
AGAN ST MTHOD
popular interetaon of the theo of relavity which sees all events and ansions as already prearranged in a fourdimensional connuum, the only change bein the (decepve) jouey of consciousness along its world line Be that as it may, he was the rst to propose a conseaon law (Bng is), to draw a bounda line between reality and appearance (and thus to create what later thnkers caled a theo of knowledge') and to give a more sasfacto foundaon for connuity than did 9th and 2centu mathemacians who had to invoke intuion' Using Paendes arguments Aristotle consucted a theo of space and moon that ancipated some ve deep-lying properes of quantum mechanics and evaded the dicules of the more customa (and less sophiscated) interetaon of a connuum as consisng of indivisible elements 5 Parmenides' theo clashes with most mode methodological pinciples but this is no reason to disrerd it A more specic eample of a theo with qualitave defects is Newton's theo of colours According to this theo, light cosists of rays of dierent refagibility which can be separated, reunted, refracted, but whch are never changed in their inteal constuon, and which have a ve small lateral etension in space Consideing that the surface of mirrors is much rougher than the lateral etension of the rays, the ray theo is found to be inconsistent with the estence of mrror images (as is admtted by Newton hmsel): if light consists of rays, then a mrror should behave like a rough surface, ie. it should lk to us like a wall Newton retained his theo, eliminang the diculty with the help of an hoc hyothesis: the reecon of a ray is eected, not by a single point of the reecng body, but by some power of the body which is evenly diused all over surface'} 6 1 4 A vvd desrption of the Parendean avour of the theory of relavi s given by . Weyl, Philoshy ofMathati and Natural nce, Prneton, 1949, p 1 1 6 Einstein hiself rote For us ho are onvined physists the disnon beeen past, pesent and future has no other meanng than that of an llusion, though a tenaious one' Coponce ec Michele Bso, Pars, 1979, p 312 Cf alsop 292 In a ord: the events of a human life are illusions, though tenaious ones' 15 For Arsote f the essay quoted in Chapter 4, fooote 3 Moe aempts to get onnui out of olleons ofindvsible eleents are reed in A Gruenbau, A Consistent Conepon of the Extended Lnear Connuu as an Aggegate of Unextended Eleents', Philos ofS, o 19, 1952, pp 283 Cf also W Samon (ed), Zo 's Parx, e York, 1 970 1 6 Sir Isaa Neton, Oti, Bk 2, pa 3, proson 8, e York, 192, p 266 For a disussion of ths aspet of Neton's ethod f y essay, Classa Epiris', Philoshil Pe, Vol 2, Chapter 2
IV
45
In ewton's case the qualitave discrepancy between theo and ac was removed by an hoc hypothesis In other cases not even this imsy manoeuvre is used one reains the theo and t toet t shortcomings An eample is the atude towards Kepler's rule accordingto which an object vewed through a lens is perceived at the point at which the rays travelling from the lens towards the eye ntersect 7
n
-
-"
The rule implies that an object situated at the focus will be seen innitely far away. But on he cona,' writes Barrow, Newton's eacher and predecessor at Cambridge, commenng on ths predicon, 8 we are assured by eperience that [a point situated close to the fus appears variusly distant, according to the dierent situaons of the eye . . . And it does almost never seem futher othan it would be if it were beheld wth the naked eye; but, on the cona, it does somemes appear much nearer . . . Al which does seem repuant to our prnciples' But for me,' Barrow connues, neither ths nor any other diculty shall have so great an inuence on me, a to make me renounce that which I know to be manifestly agreeabe to reason.' een bet nnty el be u
17 ohannes Kepler, Ad Vitellion Paraipoma Johann K, Gammelte e, Vol 2, Mnh, 1939, p 72 For a detaled discusson of Kepler's ule and is nene see Vasco Ronhi, Oti e Sce ofVion, e York, 1957, Chapters 3 Cf also Chapters 91 1 belo 1 eion X Cant in Scholio pulic hait n qui Otim nonon guinae Ration itgantur ontur, London, 1669, p 125 The
assae s sed by Berkeley in hs aak on he radonal, obetvst' opts n Essay oa a Theo oVision, Works, Vol , ed Frazer, London, 1901, pp 137
46
AGAIST MTOD
Barrow mtions the qualitave dicules, adding that he will not abandon the theo. This is not the usual procedure. The usual predure is to forget the difcules, never to talk about them, and to proceed as if the theo were without fault. This atude is ve common today. Thus the classical elecodynamics of Maell and Lore ntz implies tha the moon of a free parcle is selfaccelerated Considering the self-ener of the elecon one obtains divergent expressions for point-charges while charges of nite extension can be made to agree with relavity onl� by adding untestable sesses and pressures inside the elecon 1 The problem reappears in the quantum theo, though it is here parally covered up by renormalizaon' This predure consists in crossing out the results of certain calculaons and replacing them by a descripon of what is actually obseed Thus one admits, implicitly, that the theo is in ouble whie formulang it in a manner suggesng that a new principle has been discovered 20 Small wonder when philosophic ally unsophiscated authors get the impression that all evidence points with merciless deniteness in the . direcon . . [that] all the processes involving . . unknown interacons conform to the fundamenta quantum law' 21 A siking example of qualitave failure is the status of classical mechanics and elecodynmics after Boltzmann's equiparon theorem According to this theorem ener is equally disibuted over all degrees of freedom of a (mechanical or elecodynamic) system. Both atoms (which had to be elasc to rebound from the walls of a container and from each other) and the electromaec eld had innitely many degrees of freedom which meant that solids and the 19 Cf W Heier The Quantu Theo oRdaton, Oxford, 1954, p 3 1 20 Renoalization has n the eante beoe the basis of qantu eld
theory and has led to preditions of srsing ara (re th literature in A Pais, Inward ound, Oxford, 196) This shos that a point ofvie hih, lked at fro afar, appears to b hopelessl rong ay ontan exellent ngredents and that ts exellene ay rean nrevealed to oe gided by strit ethodologal rules Alays reeber that y exaples do not rtize siene they ritize those ho ant to sbe it to their siplended rles b shoing the disasters sh rles old reate Eah of the exaples of foootes 317 an be sed as a bass for ase studies of the kind to be arred ot in Chapters 12 (Galileo and he Copeian Revoltion) This shos that e ase ofGalleo s not an exeption haraterzng the begnning of the soalled sienti revoltion' (G Radniky, Theorienplralss Theorenoniss', in Deer Meisenhei (ed), D Metho und eo pluralmus n Wisssch, 1971, p 164) bt is pcal of sienti hange at all ties 2 1 Rosenfeld in saton and Inteetaton, ondon, 157, p
FIVE
47
elecomaec elds should have acted as nsaable snks ofener Yet [a]s so often n the hsto ofscence, the conct between simple and generaly known 22facts and current theoreca deas was recogned nly slowly' Another example of mode physcs s qute nsucve, for t ght have led to an enrely derent dvelopment of our knowledge conceng the mcrosm Ehrenfest proved a theorem accordng to whch the classcal elecon theo of orentz taken n conjuncon wth the equparon theorem excludes nduced maesm 23 The reasonng is exceedngly simple accordng to the equparon theorem, the probabity of a gven moon s proporonal to exp ( R, where s the ener of the moon Now the rate ofwork of an elecon movng n a constant maec eld B s, accordng to Lorentz, =Q(+ VxB) where Q s the charge of the movng parcle, V its velty and the elecc eld Ths magntude reduces to Q whch means that the ener and, therefore, the probabty remans unaected by a maec eld (Gven the proper context, ths result songly supports the deas and expermental ndngs of the late Felix Ehrenhaft) Occasonally t s impossble to suey al the nteresng consequences, and thus to dscover the absurd results of a theo This may be due to a deciency n the esng mathemacal methods t may also be due to the orance of those who defnd the theo Under such crcumstances, the most comon predure to use an older theo up to a cetain pont (whch s oen qute arba) ad to add the new theo for calculang renements. Seen from a methodologcal pont ofvew the predure a veritable nghare. et us explain t usng the relavisc calcuaon of the path of Mercu as an example. The perielon of Mercu moves along at a rate of about 5" per centu. Of ths value, 502" are geomec, havng to do with the movement of the reference system, whe 53" are dynacal, due to perturbaons n the solar system Of thes perrbaons but th 22 K. Gotried V.F Wesskopf Cnc ofPaice Psi, Vol 1, ord nd ew York 1984, p. 6 2 The dil was reaized by Bohr n hs dtora thess f Nels hr oeed Wor Vol. I Amsterda 1972, pp. 158, 381 e inted ot that the
veli hanges due to the hanges in the eea eld would equaie aer the ed W established so that no agneti eets ould ise Cf aso eilbron and T.S uhn 'The Genesis of the Bohr Ato Htol Studi in the PhysilSc, No 1 969, p. 22 1 The aguent in the te s taen fro e Fnman Leur, Vol 2, Califoa and London 1965 , Chapter 346 For a soewhat leare aount f R Beker eoe Elettt, Leipzig 1949, p. 132
48
A A S M O
famous 43" are accounted for by classical mechanics This is how the situaon is usually explained. The explanaon shows that the premise from which we derive the 4 3" is not the general theo of relavity plus sutable inial condions. The premse contains classical physics in ition to whatever relavisc assumpons are being made. Futhermore, the relavisc calculaon, the so-called Schwaschild souon', does not deal with the planeta stem as it ests in the real world (i.e our own asymmec gaaxy); it deals with the enrely conal case of a central symmeical universe containing a singularity in the mddle and nothig else. What are the reasons for employing such an odd conjuncon of premises? The reason, according to the customa reply, is that we are dealing with appromaons The formulae of classical physics do not appear because relavity is incomplete. Nor is the cenally symmec case used because relavity does not oer anything better. Both schemata ow frm the general theo under the special circumstances realized in our planeta system prid we omt maitudes too small to be considered. Hence, we are using the theo of reavity throughout, and we are using it in an adequate manner. Now in te present case, making the required appromaons would mean calculang the full -body problem relaviscally (including long-term resonances between derent planeta orbits), omitng magnitudes smaller than the precision of obseaon reached, and showing that the theo thu curtailed coincides with classical celesal mechanics as corrected by Schwaschild. This predure has not been used by anyone simply because the relavisc body proble has as yt withstood soluon. When the argument started, there were not even appromate soluons for important problems such as, for example, the problem of tability (one of the rst great stumbling blocks for Newton's theo). The classical part of the explanans, therefore, did not cur just for convenience, it a absoute necsa And the appromaons made were not a result of relavisc caculaons, they were inoduced in order to make relavity t th case One may prpery call them hoc aroimations. 2
24 Today the soalled paraetrzed postewtonan foralis satises ost of
the desiderata oined in the tet (detals in C.M. Wll ). My nt is that this was a later aheveent whose absene did not prevent scentists fro arging and aing we, abot the new ideas Theories are not only sed as preises for derivations; they are evn ore freqently sed as a general bakgrond for novel
FV
4
d hoc apprmaons abound in mode mathemacal physics They play a ve important part in the quantum theo of elds and they are an essenal ingredient of the correspondence principle. At the moment we are not conced with the reasons for ths fact, we are only conceed with its consequences: hoc appromaons conceal, and even elminate, qualitave dicules. hey create a false impression of the excellence of our science. It follows that a philosopher who wants to study the adequacy of science as a picture of the word, or who wants to build up a realisc scienc methodolo, must look at mode science with special care. In most cases mode science is more opaque, and more decepve, than its 1 6th- and 7th-centu ancestors have ever been. As a nal example of qualitave dicules I menon agan the helienic theo at the me of Galileo. I shall soon have occasion to show that this theo was inadequate both quatavely and quantavely, and that it was also philosophcally absurd. o sum up this brief and ve incomplete list wherever we lk, whenever we have a little paence and select our evidence in an unprejudiced manner, we nd that theories fai adequately to reprduce certain quantitate rults, and that they are qualitate incompett to a surising degree. Science gives us theories of great beauty and sophscaon Mode science has developed mathe macal suctures whch exceed anything that has ested so far in coherence generality and empircal success But in order to achieve his mracle all the esng oubles had to be pushed into the relation between theo and fact, 25 and had to be concealed, by hoc hypotheses, hoc appromaons and other predures gesses whose oral relaton to the bas asspons s dlt to certain 'I st onfess rites Descaes in hs coue on Meth (Lbrry ofLibera A 1965, p 52), 'that the wer of nature s so aple and so vast and these prnples [the theoreal prniples he had developed for hs ehanal niverse] siple and so general that I aost never noe any parlar eet sh that I do not see rght away that t an [be ade to onfor to these prniples] n any derent ways and y greatest dty is saly to dscer n whih of these ways the eect is derved ode theoretal physsts nd theselves in eay the sae situaon 25 Von eanns work n qantu ehans s an espealy nsve eple of ths predre In order to arrve at a sasfatory pof of the epanson theore in ilbe Spae von eann replaed the qasiintuive noons of r (and Bohr) by ore ople noons of his own The theoretial relaons beeen the new notions are aessible to a ore rgoros eaent than the theoreal relaons eween the noons that preeded the ('ore rgoros fro the pont ofew ofvon eann and hs followers) It is dierent wth ther relaon to eperental pred res o easrng nsents an be speed for the great aorty of osees gner Amen Joual ofPhys, Vol 3 1 1 963, p 14), and where seon s ssble t beoes neessary to odi wellknown and nrefted
50
A G A S E T O
his being the case, what shall we make of the methodological demand that a theo must be judged by experience and must be rejected if it conadicts accepted basic statements? What attude shall we adopt towards the various theories of conmaon and corroboraon, which all rest on the assumpon that theories can be made to agree wth the known facts, and which use the amount of agreement reached as a principle of evaluaon? his demand, these theories, are now all seen to be quite useless hey are as useless as a medicine that heals a paent only if he is bacteria-free In pracce they are never obeyed by anyone Methodologists may oint to the importance of falsicaons but they blithely use falsied theories, they may semonize how important it is to consider all the relevant evidence, and never menon those big and drasc facts which show that the theories they admre and accept may be as badly o as the older theories which they reject I price they slavishy repeat the most recent pronouncements of the top dogs in physics, though in doing so they must violate some ve basic rules of their ade Is it ossibe to proceed in a more reasonable manner? Let us see! 2 According to Hume, theories cannot be ed fm facts he demand to admit only those theories which follow from facts leaves us without any theo Hence, science a we now t can est only if we drop the demand and revise our methodolo According to our present results, hardly any theo is consstt wth the facts he demand to admit only those theories which are consistent wth the available and accepted facts agan leaves us without any theo (I repeat wthout any theo, for there is not a single theo that is not in some trouble or other) Hence a science as we know it can est only if we drop this demand also and ain revise our methodolo, now dmtting countenduon n ton to dmittng unsuoed hypoth he right method must not contain
laws n an arbary way or else to admit hat some qute ordnary problems of quantum mehanis suh as he saerng problem do not have a soluon M Ck Jou ofMathatil Phyi Vol 36 957) Thus he heory beomes a veriable monster of rigour and preision while its relation to eperiene is more obscure han ever It is interesting to see hat similar developments ur in primive thought The most strking features of upe sand dining writes SF ader in Ne Relin 954 p 63 is he ontrast between its pretentious heoretial framework and its primve and slipshod appliation in pratie It does not need a siene to produe eumannan nghtmares 26 The estene of qualitative diulties o pockets of resistane (St Augsne ontra Juanum V v 5 Mie Vol ) was used by he Church fahers to defuse obeons whih he sciene of their time had raised against pa of he Christian faih suh as he dtrine of he orporeal resurreon
FIVE
51
any rules hat make us choose between heories the b of falstion Raher, its rules must enable us to choose between heories which we have already tested and which arefaed o pred further Not only are facts and theories in constnt disharmony, hey are never as neatly separated as eveone makes hem out to be Mehodological rules speak of 'heories, 'obsea ons and 'experimental results as if hese were well-dened obects whose properes are easy to evaluate and which are understood in he same way by all sciensts However, he material which a scienst aual has at his disposal, his laws, his experimental results, his mahemacal techniques, his epistemological preudices, his atude towards he absurd con sequences of he heories which he accepts, is indetermiate in many ways, ambiguous, and ner ful sarated om the htol bk und It is contamiated by pinciples which he ds not know and which, if known, would be exemely hard to test Quesonable iews on cognion, such as he view hat our senses, used in normal circumstances, give reliable informaon about he world, may invade he obseaon language itself, constung he obseaonal term as well as he disncon between veridical and iuso appeaace As a result, obseaon languages may become ed to older layers of speculaon which aect, in his roundabout fashion, even he most progressive mehodolo Example he absolute space-me fae of classical physics which was codied and consecrated by Kant) he senso impression, however simple, contains a component hat expresses he physiological reacon of he perceiving oism and has no obecve correlate his 'subecve component oen merges ih he rest, and forms an unsuctured whole which must be subdivided from he outside wih he help of counteinducve procedures ( example is he appearance of a xed star to he aked eye, which conains he eects of irradiaon diracon, diusio, rescted by he lateral ihibio of adacent elements of he rena and is furher modied in he brain) Finaly, there are he alia premises which are needed for he derivaon of testable coclusios, and which casionaly form enre aia sc Cosider the case of he Copeican hypohesis, whose inveno, defece, ad paral vindicaon runs counter to almost eve ehodological rule one might care to thik of today he auia �iences here contained laws describing he properes and he uence of he terresal aosphere meteorolo); opcal laws dealig wih he sucture of he eye and of telescopes, and wih he behaviour of light and dynical laws descrbing moon in moving sstes Most imporantly, however, he aulia sciences contaed
52
A G A I S T M E T H OD
a heo of coion hat postulated a certain simple relaon between percepons and physical objects Not all aulia dis ciplines were avalable in explicit form Many of hem merged wih the obseaon language, and led to the situaon desribed at the beginning of he preceding paragraph Consideraon of all hese circumstances, of obseaon tes, senso core, auia sciences, background speculaon, suggest hat a heo may be inconsistent wh he evidence, not because it is incorrect, but beuse the ince is contaminated he heo is hreatened because he evdence eiher contans unanalysed sensa ons which only parly correspond to exteal presses, or because it is presented in terms of anquated views, or because i is evaluated wh he help of bacward auia subjects he Copeican heo was in ouble for a hese reasons It is hs hstoco-physiolol charer ofthe ince, he fact hat it does not merely describe some objecve state of aairs but ao rs suee mythical and long-o vis conceing hs state of aairs, hat forces us to ake a fresh look at methodolo It shows hat it would be exemely imprudent to let he evdence judge our heories directly and whout any furher ado A saighoard and unqualied judgement of heories by 'facts is bound to eliminate ideas smp beuse th not t into the framork of some o cosmolo kng experimental results and obseaons for grante and putng he burden of proof on he heo means takng he obseaonal ideolo for granted whout having ever examned it {Note hat the experimental results are supposed to have been obtained wh he greatest possible care Hence 'taking obseaons, etc, for granted means 'taking hem for granted aer he most careful examnaon of heir reliabiity for even he most care examnaon of an obseaon statement does not interfere wh he concepts in which it is expressed, or wh he sucture of he senso image) Now how can we possibly examine somethng we use all he m e and presuppose in eve statement? How can we crcize he term in which we habitually express our obseaons? Let us see he rst step in our cricism of comonly-used concepts is to create a measure of cricism, somehing wh whch hese concep can be compared Of course, we shall later want to know a little more about he measurng-sck itself; for example, we shall want to kno wheher it is better han, or perhaps not as good as, he mate examned But in order for ths examinaon to start here must be measurng-sck in he rst place herefore, he rs step in o cricism of customa concepts and customa reacons is to ste
IV
53
outsde th crcle ad either to ivet a ew cocepal system, for xample a ew theo, that clashes wth the most carefully established obseaoal results ad cofouds the most plausible theorecal ricples, or to import such a system fom outside scece, from rligio, from mholo, from the ideas of com tets or the rambligs of madme. This step is, agai, couteriducve Couteriduco is thus both a scece could ot est without it ad a legimate ad much needed me i the e of scece.
27 It is nteresng to see tha Phlolaos who disregarded the evidene of the senses and set he earth in motion was an unmathematial onfusionist It was the onfusonst who found the ourage laking in many great obseers and atheatally wellinformed scientists to disregard the immediate edene of the se�ses in order to reman in agreement with priniples he rmy believed von F ndprobler Gchichter antik Wsscha Berlin-ew York 97 p 5 It i therefore not surising that the ne step on this path was due to a man w ose wrngs as far as we know them show him as a talented sylist and pularizer asionaly neresting ideas of hs own rather than as a profound hinker or exat ibid p 84 Confusionists and superia intelletuals me whle the thinkers cd into the darker rego s of the status quo or to express it in a
W
way, they remain stuck in the mud.
6 As an ample ofsuch an apt I amne the tower arment whch the Astotelans used to rte the moton ofthe eah e amt o atural interetaons so ose conneed wth obseatons that t nee a speal to reale ther istce and to tene their contt. Galeo nt the natural nteretatons whch are nconsstt wth Cs and rc th oth.
It seems to me tht [Glleo] suers gretly from onnul dgressons nd tht he does not stop to epln ll tht s relevnt t eh point; whh shows tht he hs not eamned them in order nd tht he hs merely sought resons for prulr ees wthout hvng onsdered rst uses ; nd thus tht he hs ult wthout foundon
I m (ndeed) unwllng to ompress phlosophl dnes nto the most nrrow kind of spe nd to dopt tht s onse nd greless mnner tht mnner re of ny doment whh pure geomens ll their own not utterng sngle word tht hs not een gven to them y st neessity I do not regard t s fult to tlk out mny dverse thngs even in those eses whih hve only single top . for I eleve tht wht gves grndeur nolty nd eellene to our deeds nd invenons does not le n wht s neess - though the sene oft would e gret mistake ut wht is no
LL But where ommon sense eleves tht ronlzing sophists hve the ntenon of shkng the ve fundment of the ommonwel then t would seem to e not only resone ut pessle nd even
54
SIX
55
prseworthy to d the gd use wth shm resons rther thn levng the dvntage to the opponent
K
As a concrete illusaon and as a basis for furher discussion, I shall ow briey describe he manner in which Galileo defused an importat argument against he idea of he moon of he earh I say defused', and not refuted', because we are dealing wih a chaging conceptual system as well as wih certain attempts at cocealment. According to he argumet which covinced Tycho, and which is used agaist he moon of he earh in Galileo's own rattato l ra, obseaon shows hat heavy bodies . falling down from on high, go by a saight and vercal line to he surface of he earh Ths is cosidered an irrefutable argument for he earh being moonless. For, ifit made he diual rotaon, a tower from whose top a rk was let fall, being carried by he whirling of he earh, would avel many hudreds of yards to he east in he me he rock would consume in its fall, ad he rk ought to sike he earh hat distance away from he base of he tower 2 The three quoations are Desaes, leer to Mersenne of Otober 638, r , p 380 Galeo, leer to Leold of Tosana of 64, usuay quoted under the Sul anr Lunare Oe Favoro, VII I, p 49 For a deiled discussion ofGalileo's style and its onnetion th his natural philosophy f L Olschki, G und sine Zt chichte neusprachch wissschalich Litatur Vol III, Halle, 927, reprinted Vaduz, 965 The leer to Leold is quoted and dsussed on pp 455 Desaes' leer is disussed by Salmon as an example of the issue beteen rationalism and empiriism in The Foundations of Sienti Inferene', Mind and osmos ed Coodny, Pittsburgh, 966, p 36 It should rather be regarded as an example of the issue beteen dogma methodologies and oprtunis methodolo gies, earing in mind that empirism an be as strit and unyielding as the most rigrus ypes of rationalism.
The ant quotaon is from the tique ofPure Reon 777, 8 (the quoation as brught to y aenon by Professor Sanley Rosens ork on Plato's Symposium) Kant ontinues Hoever, I ould think that there is nothing tha goes less ell together th the intention of asserng a gd ause than subtefuge, oneit, and deeption one ould take only this muh for grnted, then the bale of speulave reason ould have been onluded long ago or ould sn ome o an end Thus th e purity of a ause often stands in the inverse proron to its uth ' One should al noe tha ant explains the rse of cilation on the basis of disingenuous moves hh have the funtion to raise mankind above its rude past, ibid, 776, 4f Smilar deas cur in his account ofworld history . 2. Dialoge op. cit., p. 1 26.
56
AGAINST METHOD
In considering the argument, Galileo at once admits the correcess of the senso content of the obseaon made, that heavy bodies falling from a height, go peendicularly to the surface of te earth' 3 Considering an author (Charmon) who sets out to convert Copeicus by repeatedly menoning this fact, he says: I wish that this author would not put himself to such ouble ng to have us understand from our senses that ths moon of falling bodies is simple saight moon and no other knd, nor get ang and complain because such a clear, obvious, and manifest th ng should be caled into queson For in ths way he hints at believing that to those who say such moon is not saight at al, but rather circular, it seems they see the stone move visibly in an arc, since he cals upon their senses rather than their reason to clari the eect. This is not the case, Simplicio; for just as I . . . have never seen nor ever expect o see, the rock fall any way but peendicularly, just so do I believe that it appears to the eyes of eveone else. It s, therefore, better to put aside the appearance, on which we lagree, and to use the power of reason either to conrm its reality or o revea its fallacy.'4 The correcess of the obseaon is not in queson. What is in queson is its reality' or falacy'. What is meant by t expression? The queson is answered by an example that curs n Galileos next paragraph, from which . one may lea how esily anyone may be deceived by simple appearance, or let us say by the impressions of one's senses. This event is the appearance to those who avel along a seet by night ofbeing followed by the moon, with steps equal to theirs, when they see it go gliding along the eaves of the roofs. There it looks to them just as would a cat really runnng aong the les and putng them behind it; an appearance which, if reason did not inteene, would only too obviously deceive the senses.' In this example, we are asked to start with a senso impression and to consider a statement that is forcefully suggested by it (he suggeson is so song that it has led to enre systems ofbelief and to rituals, as becomes clear frm a closer study of the lunar aspects of witchcraft and of other cosmological hypotheses.) Now reason inteenes'; the statement suggested by the impression s examned, and one considers other statements in its place. The nature of the impression is not changed a bit by this acvity. (This is onl appromately e but we can omi for our present purpose the complicaons arising from an interacon of mpression and 3 ibid, p 125 4 ibid, p 256
SIX
57
proposion. But it enters new obseaon statements and plays new, better or worse, parts in our knowledge What are he reasons and he ehods which regulate such an exchange? To start with, we must become clear about the nature of the total henomenon appearance plus statement There are not two acts one, nocing a phenomenon; he oher, expressing it with he help of he approprate statement but on one, . saying in a certain obseaonal situaon, he moon is following me', or, he stone is falling saight down' We may, of course, abstractly subdivide his process into parts, and we may also y to create a situaon where stateent and phenomenon seem to be psychologically apart and waing to be related. (This is raher dicult to achieve and is perhaps enrely impossible.) But under normal circumstances such a division does not occur; describing a familiar situaon is, for he speaker, an event in which statement and phenomenon are rmly lued togeher. This unity is he result of a press of leaing hat starts in one's childhood. From our ve early days we lea to react to situaons wih he appropriate responses, linguisc or oheise The teaching procedures boh shape he appearance', or phenomenon', and establish a rm conneion wih words, so hat nally he phenomena seem to speak for hemselves wihout outside help or exaneous knowledge. They are what he associated statements assert hem to be The language hey speak' is, of course, inuenced by he beliefs of earlier generaons which have been held for so long hat hey no longer appear as separate principles, but enter he term of eveday discourse, and, after he prescribed aining, seem to emerge from he hings hemselves At his point we may want to compare, in our imaginaon and quite absactly, he results of he teaching of dierent languages incoorang dierent ideologies. We may even want consciously to chane some of hese ideologies and adapt hem to more mode' points ofview It is ve dicult to say how ths will alter our situaon, unls we make he furher assumpon hat he quality and sucture f ensaons (percepons), or at least he quality and sucture of ose sensaons which enter he body of science, is independent of eir linguisc expression. I ve doubtful about even he aprmate validity of his assumpon, which can be refuted by ple examples, and I am sure hat we are deprivng ourselves of ew and surising discoveries as long as we remain wihin he limits de ned by it Yet, I shall for he moment, remain wihin these limits . aking he addional simpling assumpon, we can now dsuish between sensaons and hose mental operaons which
58
AGAINST METHOD
follow so closely upon the senses', 5 and which are so rmly connected with their reacons that a separaon is dicult to achieve. Considering the origin and the eect of such operaons, I shall call them natural nteretations. In the hsto of thought, natural nteretaons have been regarded either as a po pruositions of science or else as prudic which must be removed before any serious examnaon can ben. The rst view is that of Kant, and, in a ve dierent manner and on the basis of ve dierent talents, that of some contempora lnguisc phiosophers The second view is due to Bacon (who had predecessors, however, such as the Greek scepcs) Galileo is one of those rare thinkers who wants neither forever to retain natural interetaons nor altogether to eliminate them. Wholesale udgements of this knd are quite alien to his way of thinkng. He insists upon a tical dssson to decide which natura interetaons can be kept and which must be replaced. This is not always clear from hs wrings. Quite the cona. The methods of reminscence, to whch he appeals so freely, are desied to create the impression that nothing has changed and that we connue expressing our obseaons in old and famliar ways. Yet his attude is relavely easy to ascertain natural interetaons are necsa The senses alone, without the help of reason, cannot gve us a ue account of nature What is needed for arriving at such a ue account are the . . senses, companied reonn. 6 Moreover, in the arguments dealing with the moon of the eah, it is this reasoning, it is the connoaon of the obseaon tes and not the message of the senses or the appearance that causes ouble. It is, therefore, better to put aside the appearance on which we all agree, and to use the power of reason either to conrm its reality or to reveal its fallacy.' Conrming the reality or revealing the fallacy of appearances means however, examning the validity of those natural interetao which are so inmately connected with the appearances that we no longer regard them as separate assumpons. I now tu to the rst natural interetaon implicit in the argument from fallng stones According to the Copeican view as presupposed n the tower argument the moon of a falling stone should be mixed saight and-circular'8 By the moon of the stone' is meant not its moon relave to some visible mark in the visual eld of the obseer, or its 5 6 7 8
Franis Baon, Nu anu Inoduton ialoe op t, p 255, ils ibid, p 256 bd, p 248
SIX
59
obse ed moon, but rather its moon in the solar system or in absolute) space, ie its real moton. he famliar facts appealed to in the argument present a dierent knd of moon, a simple vercal oon his refutes the Copecan hypothesis only if the concept of moon that occurs in the obseaon statement is the same as the concept of moon that curs in the Copecan predicon he obseaon statement 'the stone is falling saight down must, therefore, refer to a movement in absolute) space It must refer to a real moon Now, the force of an 'argument from obseaon derves from the fact that the obseaon statements involved are rmly connected with appearances here is no use appealing to obseaon if one does not know how to descrbe what one sees, or if one can oer ones descripon wth hesitaon only, as if one had just leaed the language in which it is formulated Producing an obseaon statement, then, consists of two ve dierent psychological events 1) a clear and unambiguous ssaton and (2) a clear and unambiguous conneon between this sensaon and parts of a language his is the way in which the sensaon is made to speak Do the sensaons in the above argument speak the language of real moon? hey speak the language of real moon n the context of 17th centu eveday thought At least, this is what Galieo tels us He tells us that the eveday thinkng of the me assumes the 'operave character of ll moon, or, to use well- known phlosophcal terms, it assumes a nae realm wth rpe to moton: except for casional and unavoidable illusions, apparent moon is idencal with real absolute) moon Of course, ths disncon is not explicitly drawn One does not rst disnguish the apparent moon from the real moon and then connect the two by a correspondence rule One rather descrbes, perceives, acts towards moon as if it were already the real thing Nor ds one preed in this manner under all circumstances It is admtted that objects may move whch are not seen to move and it is also admtted that certain moons are iluso cf the example of the moon menoned earler in ths chapter) Apparent moon and real moon are not always idened However, ere are paradatc c in which it is psychologically ve dicult, f not planly impossible, to admt decepon It is from these paradigmac cases, and not from the excepons, that naive realism th respect to moon derves its sength hese are also the aons in whch we rst lea our knemac vabula From our ve chidhood we lea to react to them with concepts which have ae realism buit rght into them, and which inexicably connect
60
AGANST OD
movemen and he appearance of movemen The moon of the sone n the ower argumen, or the alleged moon of the eath, s such a paradgmac case How could one possbly be unaware of the swf moon of a large bulk of atter such as the ea s supposed to be! How could one possbly be unaware of the fac tha the fallng sone aces a vastly extended rajeco through space! Fr the pon of vew of 17th-centu though and language, the argument s, therefore, mpeccable and qute forceful However, noce how teo (operave character' of all moon essenal correcess of sense reports) whch are not formulated explctly, enter the debate n the guse of obseable events We realze agan that such events are Trojan horses whch must be watched most carefully. How s one supposed to proceed n such a scky stuaon? The argument from fallng stones seems to refute the Copecan vew. Ths may be due to an nherent dsadvantage ofCopecansm; but t may also be due to the presence ofnatural nteretaons whch are n need of mprovement. The rst task, then, s to der and to solate these unexamned obscles to prgress. It was Bacon's belef that natural nteretaons could be dscovered by a method of analyss that peels the o, one aer another, unl the senso core of eve obseaon s lad bare Ths method has serous drawbacks Frst, natural ntereons of the knd consdered by Bacon are not just ad to a prevously esng eld of sensaons They are nsumental n constituting the eld, a Bacon says hmself Elmnate all natural ntereons, and you also elmnate the ablty to thnk and to perceve Second, dsregardng ths fundamenal funcon of natural nteretaons, t should be clear that a person who faces a perceptual eld wthout a sgle natural nteretaon at hs dsposal would be comple disoted, he could not even sta the busness of scence The fact that we srt, even after some Baconan analyss, therefore shows that the analyss has stopped prematurely It has stopped at precsely those natural ntereons of whch we are not aware and wthout whch we cannot proceed. It follows that the nenon o start fro scratch, after a complete removal of all natural neretaons, s self defeang. Futhermore, t s not possble even pa to unravel the cluster o natural ntereons. At rst sght the task would seem to be smple enough One takes obseaon stements one after the other, and analyses ther content. However, concepts that are hdden · obseaon statements are not lkely to reveal themselves n the mo absact pars of language If they do, t wll sll be dcult to na them down concepts, just lke percepts, are ambguous an
SX
61
dependent on background. Moreover, the content of a concept is deterined also by the way in which it is related to percepon. Yet, how can this way be discovered without circularity? Percepons must be idened, and the idenng mechanism will contain some of the ve same elements whch gove the use of the concept to be invesgated We never peneate t concept completely, for we always use part of it in the attempt to nd its constuents There is only one way to get out of this circle, and it consists in using an tl meure ofmpason, including new ways of relang concepts and percepts Removed from the domain of natural dscourse and from all those principles, habits, and atudes whch constute its form oflife, such an exteal measure wil look sange indeed This, owever, is not an argument against its use On the cona, such an impression of sangeness reveals that natural interetaons are at work, and is a rst step towards their discove. Let us explain his situaon with the help of the tower example The example is intended to show that the Copeican view is not in accordance with the facts' Seen from the point of vew of these facts', the idea of the moon of the eath is outlandish, absurd, and obviously fase, to menon only some of the expressions which were frequently used at the me, and which are sll heard whenever professional squares confront a new and counter-factual theo This makes us suspect that the Copeican view is an exteal measuring rod of precisely the kind described above Let us therefore tu the argument around and use it as a teing ice that helps us to discover the natural interetaons which exclude the moon of the eath Tuing the argumen around, we t s the moon of the eath and th inquire what changes will remove the conadicon Such an inqui may take considerable me, and there is a good sense in which it is not nished even today he conadicon may stay with us for decades or even centuries ll, it must be upheld unl we have nished our examinaon or else the examinaon, the attemp to discover the antediuvian com onen of our knowledge, cannot even start This, we have seen, is ne of the reasons one can give for retining, and, perhaps, even for �ting, theories which are inconsistent with the facts Ideological nedients of our knowledge and, more especially, of our obseaons are discovered with the help of theories which are refuted by them h are discered countendue Let me repeat what has been asserted so far Theories are tested, and ossibly refuted, by facts Facts conin ideological components, older views which have vanished from sight or were perhaps never orulated in an explicit manner Such components are highly
62
AGA NS HOD
suspcous Frst, because ofther age and obscure orgn w e do not know why and how they were noduced secondly, because ther ve nature protects them, and always has protected them, from crcal examnaon In the event of a conadcon beween a new and nteresng theo and a collecon of rmly establshed facts, the best predure, therefore, s not to abandon the theo but to use t to dscover the dden prncples responsble for the conadcon. Counternducon s an essenal part of suc a process of dscove. (Excellent hstorcal example the arguments aganst moon and atomcty of Parmendes and Zeno Dogenes of Snope, the Cync, tk the smple course that ould be ken by many contempora scensts and all contempora phlosophers he refuted the arguments by rsng and walkng up and down The opposte course, recommended here, as led to muc more nteresng results, as s wessed by the hsto of the case One should not be t hard on Dogenes, however, for t s also reported that he beat up a pupl who was content wth hs refutaon, exclaming that he had gven reasons whch the pup sould not accept wthout addonal reasons of hs own Havng dsered a parcular natural ntereon, how can we amne t and tt t? Obvously, we cannot preed n the usual way, e. derive predcons and compare them wth results of obsea on' These results are no longer avalable The dea that the senses, employed under normal crcumstances, produce correct reports of real events, for example reports of the real moon of physcal bodes, has been removed from all obseaonal statements (Remember that ths noon was found to be an essenal part of the an Copecan argument) But wthout t our senso reacons cease to be relevant for tests Ths concluson was generalzed by some older raonalsts, who decded to buld ther scence on reason only and ascrbed to obseaon a qute nscant aula funcon. Galleo does not adopt ths predure If one natural nteretaon causes ouble for an atacve vew, and f ts elmnaton removes the vew from the doman of obseaon, then the only accepble predure s o use oth nteretaons and to see what happens The ntereaon whch Galeo uses restores the senses to ther poson as nsuments of exploraon, t on wth rpe to te real ofrelate moton Moon among thngs whch share t n common' s non-operave', that s, t remans nsensble, mpercepble, and without any eect 9 Hegel, Voung die Gchichte Philoshie, , ed Michelet, Berlin, 18, p 289
SIX
63
watever' 10 Galileo's rst step, in his joint examinaon of he Copeican docne and of a familiar but hidden natural intereta on, consists herefore in rlacng the latter a dert nteretaton. In other words, he ntduc a n oseaton lanage. This is, of course, an enrely legimate move In general, he obseaon language which enters an argument has been in use for a long me and is quite familiar Considering he sucture of comon idioms on he one hand, and of he Aristotelian philosophy on he oher, neiher his use nor his familiarity can be rerded as a test of he underlying principles These principles, hese natural intereta ons, occur in eve descripon Exaordina cases which might create dicules are defused wih he help of adjustor words', 11 suc as like' or analogous', which divert hem so hat he basic ontolo remains unchallenged A test is, however, urgently needed. It is especialy needed in hose cases whee he principles seem to hreaten a new heo. It is hen quite reasonable to inoduce aleave obseaon languages and to compare hem boh wih he original idiom and wih he heo under examinaon. Preeding in 0 ue op ct, p 1 7 1 . Galleo's inemac relasm s not consstent In the passage quoted he poses the ew () that shaed moon hs no whao Moon,' he says, n so far as it s and acts as moon, to that extent ests relavely to thngs that lack it; and among things which all share equaly in any moon, t ds not act and s as f it did not est' (. 116) atever motion comes to be abuted to the earth must necesrly remain impercepble so long as we lk only at terresal objects' (. 114); moon that s common to many mong thing s s idle and nconsequena to the relation of those movables among themseles ( 116). On the other hand, (2) he aso suggests that nothing m in a straight line nature The moon of a celesa objects in a circle; shps, coaches, horses brds, a move in a circle around the earth; the moons ofthe pa ofanmals are al crcular; n sum we are forced to assume that only gria oum and lia suum move apparenty in a saight lne; but even that is not cein as long as it has not been proven that the earth s at rest' (. 19) Now, if (2) is adopted, then the lse pa of systems mong in a saight line wll tend to describe circular paths, thus contradcng () It is th nconsstency whch has prompted me to split Galleo's arment nto two steps, one dealng wth the rela of moon (only relative moon notice, the other deaing wth neral laws (and only neral moon le the retion betwe the pas ofa st unaed assuming, of course, that neghbourng neral moons are appromately parallel) For the two steps ofthe arment, see the next chapter One must also reae hat accepng relavi of moon for neral paths means giing up the impetus th, which proides an (nner) cause for moons and therefore assumes an absolute space n which ths cause becomes manfest Ths Galleo seems to have done by now, for his arment for he exstence ofboundless' or peetual' moons which he ouines on 1 47 of the ialogue appeas to moons which are neua, ie neiher natural nor orced, and whch may herefore (?) be assumed to go on for ever J Ausn, Sse and Ssibilia, New York, 1964, p 74 Adustor words play an porant role in he Aristotelian phlosophy
64
A G A I N S M E TH O
hs way, we must make sure that the comparson s r Tat s, we must no crcze an dom tha s supposed to funcon as an obseaon language because t s not yet well known and s, herefore, less strongly connected wth our enso reacons and less lausble than s another, more common' dom Supercal crcsms of ths knd, whc ave been elevated nto an enre losopy, abound n dscussons of the mndbody problem. Phosopers wo want to ntroduce and to test new vews thus nd themselves faced not wth amts, whc they could most lkely answer, but wth an mpeneable stone wall of well-entrenched reons. Ths s not at all derent from the attude of people orant of fore languages, wo feel that a certan colour s much better descrbed by red' than by rosso'. As opposed to such attempts at converson by appeal to famlarty (I now what pans are, and I also know, from nospecon, that they have nothg whatever to do with materal processes'), we must emphasze that a comparave judgement of obseaon languages, eg materalsc obseaon languages, phenomenalsc obseaon languages, objecve-dealsc obsaon languages, theologcal obseaon languages, etc, can srt only wh a of th are spok equal
ut
Let us now connue wth our analysis of alleos reasonng
7 Te n naural nteretatons consttute a n and hgh astr oseaton lanage Th are ntduced and concealed so that one to notce the change that h tak ple method ofanamns Th contan te a of the relavity of all moon and the law of circular inera
alileo replaces one natural interetaon by a ve dierent and as yet 160) at least partly unnatural interetaon How does he preed? How ds he manage to inoduce absurd and counter inducve asserons such as the asseron that the eath moves, and yet get them a just and attenve hearing? One ancipates that arguments will not suce an interesng and highly important imtaon of raonalism and Galileo's utterances are indeed arguments in appearance only For Galileo uses pragan He uses pcholocal tc in addion to whatever intellectual reasons he has to oer Tese icks are ve successful they lead im to victo But they obscure the new attude towards experience that is in the makng, and postpone for centuries the possibility of a reasonable philosophy. Tey obscure the fact that the experience on whic Galleo wants to base the Copeican view is nothing but the result of is own ferle imaginaon, that it as been ted They obscure this fact by insinuang that the new results whic emerge are known and conceded by all, and need only be called to our attenon to appear as the most obvious expression of the uth alileo reminds' us that there are situaons in which the non operave character of shared moon is just as evident and as rmly believed as the idea of the operave caracter of all moon is in other circumstances (This latter idea is, therefore, not the only natural interetaion of moon.) Te situaons are events in a boat, in a soothly moving cariage, and in other systems that contain an obseer and permit im to car out some simple operaons Sae: There has just urred to me a ertan fantasy whh assed through my magnaon one day whle I was salng to Aleppo,
65
66
AGA NST ETHOD where I was going as onsul for our oun I f the point of a pen had een on the ship during my whole voyage from Venie to Alexandretta and had had the property of leaving vsile marks of its whole p, what ae what mark what line would it have left? implio It would have left a lne extending from Venie to there not perfetly saight or rather, not lying n the perfet ar of a rle ut more or less lutuang aordng as the vessel would now and again have rked ut this ending in some plaes a yard or o to the right or left, up or down, length of many hundreds of mles, would have made little alteraon n the whole extent of the line These would sarely e sensile, and, without an error of any momen, it ould e alled pat of a perfet ar e So that if the lutuaon of the waves were taken away and the moon of the vessel were alm and anquil, the e and preise moon of that pen would have een an ar of a perfet irle Now if had had that same pen onnually in my hand, and had moved it only a little somemes his way or that, what alteraons should I have rought into the main extent of his line? implio Less than that whih would e given to a saight line a thousand yards long whih deviated from asolute saiess here and there y a lea's eye e Then if an arst had egun drawing with that pen on a sheet of paper when he left the port and had onnued dong so all the way to Alexandretta, he would have een ale to derive from the pens moon a whole narrave of many gures, ompletely aed and skethed in thousands of dreons, with landsapes, uildings, animals, and other things Yet the atual real essena movement marked y the pen point would have een only a line long, indeed, ut ve simple ut as to the arsts own aons, these would have een onduted exatly the same as if the ship had een standing sll The reason that of the pens long moon no trae would remain exept the mars drawn upon the paper s that the gross moon from Venie to Alexandretta was ommon to the paper, the pen, and everything else in the ship ut the small moons ak and forth, to right and left ommuniated y the arsts nges to the pen ut not to the paper and elonng to the former alone, ould therey leave a ae on the aper whih remained staona to those moons 1
lo, op ct, pp 171
SEVEN
67
Or iti Imagine yourself in a boat with your eyes ed on a point of te sail yard Do you think that because the boat is moving along briskly, you will have to move your eyes order to keep your ision always on that point of the sail and follow its moon? implio I am sure that I should not need to make any change at all not just as o my vision, but if had aimed a musket I should never have to move it a hairsbreadth to keep it aimed, no matter how the boat moved iti And this comes about because the moon which the ship confers upon the sail yard, it confers also upon you and upon your eyes, so that you need not move them a bit in order to gaze at the top of the sail yard, which consequently appears moonless to you (And the rays of vision go from the eye to the sail yard just as if a cord were ed beeen the o ends of the boat Now a hundred cords are ed at dierent ed points, each of which keeps its place whether the ship moves or remains sll
It s clear that these stuaons lead to a non-operave concept of moon even wth common sense. On the oher hand, common sense, and I mean 17th-centu Italan-arsan common sense, also contans the dea of the erate character of all moon. Ths latter dea arses when a lmted object that does not contan too many parts moves n vas and stable surroundngs for example, wen a camel ots through the desert, or wen a stone descends from a tower ow aleo urges us to remember' the condons n whch we asse the non-operave character of sared moon n ths case also, and to subsume the second case under the rst Tus, the rst of the two paradgms of non-operave moon enoned above s followed by the asseron that It s kewse ue that the eath beng moved, the moon of the stone n descendng s actually a long setc of many hundred yards, or even many thousand and had t been able to mark ts course n moonless ar or uon some other surface, t would ave left a ve long slanng lne. ut that art of all ths moon whch s common to the rock, the 2 bd, pp 249 That phenomena of se moon depend on rete moon has bee aered by Euclid n his Oti, Theon red par 49 old scholion of par 50 ue he example of a boat leang he harbour Heiberg, vi, 283 The example is repeated by opecus n Bk I, hapter ii, of e Rl t as a commonplace i edaeval opcs f Witelo, Ppea, v par 1 38 (Basel, 1 572, p 1 80)
68
A G A N S T M E TH O D
tower, and ourselves remains insensible and as if it did not est. There remans obseable only that part in whch neither the tower nor we are parcipants in a word, that with which the stone, in falling, measures the tower.' 3 And the second paradigm precedes the exhortaon to ansfer argument to the whirling of the eath and to the rk placed on top of the tower, whose moon you cannot disce because, in common with the rock, you possess from the eath that moon whch is requred for followng the tower; you do not need to move your eyes. Next, ifyou add to the rock a downward moon whch is peculiar to it and not shared by you, and whch is med with this circular moon the circular poron of the moon whch is common to the stone and the eye connues to be impercepble. The saght moon alone is sensible, for to folow that you must move your eyes downwards.' 4 Ths is song persuasion indeed. Yieldng to this persuasion, we now qute automatl start confoundng the condions of the two cases and become relavists. This is the essence of Galieo's icke As a resut, the clash between Copecus and the condions aecng ourselves and those in the ar above us' 5 dissolves into th air, ad we naly realize that al terresial events from whch it is ordinay held that the eath stads sl and the sun and the ed stars are moving would necessary appear just the same to us if the eath moved and the other std sll' 6 3 bd pp 7 2 4 bd p 250 5 Ptolemy, Synt, , 1, p 7 6 logue, op ct, p 416 c the ialogu Cncing T N S, nsl Henry re and Afonso de Saio, Ne York 1958, p 164: The same expermen hch at rst gance seemed to ho one thng, hen more carefully examned ssur us of the con' Professor McMulln, n a crique of this ay of seeng thin, ants more local and bioraphcal juscaon' for my asseon hat Galleo no ony agued, but also cheated [A Taxonomy of he Relaon beteen History d Phlosophy ', Minnota Studi, Vol 5, Minneapolis, 1971, p 39, nd e obects to he ay in hch let Galileo inoduce dynamica relasm Accordn hm hat Gaeo ares is hat snce his opponent alre nterets obsea made in such a context [movements on boats] n a "relasc ay, ho can e consistenty do ohese in he case of obseaons made on he earth's suface' (bid, p ) This is ndeed ho Galileo ares But he ares aganst an opnen ho, according to him, feels a reat repugnance toards recognzng hs non operave quali of moon among he hng hich share t n common' (iaue, op cit, p 1 7 1 ), ho is connced hat a boat, apa from hang relae moons, h aolute pit and motins well (cf Arstote, Physcs, 208b8), and ho at any rae has developed e a of usng dierent noons on dierent casions whout runn into a conadon No i f th s he son to be attacked, hen shong ha
SN
6
et us now look at the situaon from a ore abstract point ofview start with o conceptual sub-systems of ordina' thougt (see th following table) One of them rerds moon as an absolute rocss whih always has eects, eects on our senses included The dscripin of this concetual system gven here may be somewhat idalized but the arguments of Copeicus' opponents, which are quoted b alile himself and, accordng to him, are ve lausible', show that there was a wdespread tendency to think in its terms, and that ths tendency was a serious obstacle to the discussion ofalteave ideas. Occasionally, one nds even more primive ways of thinkng, where concets suc as up' and down' are used absolutely. Examples are: the asseron that the eath is t ea1 to clmb up over the sun and then fall eadlong back down ain', or the asseron that after a sort me the mountans, snkng downward wth the rotaon of the terresial globe, would get nto sch a poson that whereas a little earler one would have had to clmb steeply to their peaks, a few hours later one would have to stp and descend in order to get there' 9 alleo, in his margnal notes, calls these utterly childsh reasons which] sufced to keep imbecles believing in the ty of the eath' and he k t unnecessa to bother abou suh men as those, whose name s lon, or to take noce of ther fooleries Yet t s cear that the absolute idea of moon was well-enenched', and that the attempt to replace it was bound to enconter song resistance 2
opponent has a relatve dea of moton, or frequenty uses the relative idea n his everyday aars, s not at all pof of nconsistency n hs own "paradigm ' (McMulln, op ct, p ) It just reveals one of that paradgm thout touchng the other The arment tus nto the desred prf ony f the absolute noon s ether uppressed or sprted aay, or else dened th the relasc noon and thi hat Galleo actualy does, though surrepously, as I have ed to sho 7 ialogue, op ct, p 328 8 bd, p 327 9 bd , p 330 0 bd, p 327 II bd, p 327, talcs added 2 The dea that there s an absolute drecon n the unvese has a very nteen htory It rests on the stucture of the grataonal eld on the suface of the earh, or of that of the earth hch the obseer knos, and generalzes the epeence made there The generalzaon s only rarely regarded as a separate othe, t rather enters the grammar' of common sense and gves the terms up' and don' an absolute meanng (Ths s a natura nteetaon', n precsey the ense hat was eplaned n the tet above.) Lactanus, a Church father of the fourth cenry, appeals to ths meanng when he asks (i titutin, I , De Fasa Sapena) s one realy gong to be so confused as to assume the estence of humans ose feet are above ther heads? Where ees and ft grow not upwards, but
70
AGAINST METHOD
The second conceptual system s but around the relavty of moon, and s also well-enenched n ts own doman of applcaon Galeo ams at replacng the rst system by the second n a cases, terresal as well as celesal Nave realsm wth respect to moon s to be complete elmnated Now, we have seen that ths nave realsm s on casons an essenal part of our obseaonal vabula On these occasons aradgm I), the obseaon language contans the dea of the ecacy of a moon Or, to express t n the materal mode of speech, our experence n these stuaons s an experence of objects Pri / Moon of ompat ojets in stae surroundings of great spaa etenson deer oseed y the hunter
Pri Moon ofojets n oats oahes and other moving systems
Nturl interettion moon operae
Nturl interettion Ony reae moon s operae
Fang stone p
Earth at rest
Moon ofearth predis
Oique moon ofstone
Fang stone p
No rete moon eeen starngpont and earth
Moon of ea rth predis
No reae moon eeen starngpoint and stone
downwards' he me use of lanage s presupsed by that mass of untutod men' who rse the queson why the anpodeans are not faling o te ea rth (n, Natural Hto 11, pp 1616, cf also Ptolemy, Syn 1, 7) The aempts ofe, Anamenes and Xenophanes to nd sup for the earth which prevents it fro faling down' (Arstote, e Coelo 294a2) shows that amost eary phlosophers, wth the sole ecepon of Anamander, shared n ths way of thinng (or the Atomsts, who assume that the atoms orignally fall down,' cf Jammer, Cnc Spe Cambrdge, Mass, 1953, p I I .) Even Galleo, who thorougly ridcules the dea of the falng anpodes (logue op cit., p. 33 1 ), casionaly es of the upper haf of te mn', meanng that pa of the mn whch s invsble to us' nd let us not forget that some lnsc phosophers of today who are t stupd to recognze ther own lmaons' (Galleo, op. ct., p. 327) want to revve the absolute meanng of updown' at least lol Thus the wer over the minds of contemporaries ofa prmve conceptual frme, assumng an ansoopc world, wch Gaeo had aso to ght, must not be underesmated For an eamnaon of soe aspects of Brsh common sense at the me of Galeo, including asonom common sense, see EMW. Tillyard, The Elethan Wor Piure London, 19 The agreemen between pular opnon and the cenally symmec unverse frequenty asseed by Arstote, eg n e Coelo p 308a23f
SVEN
71
whch move absolutely Takng ths nto consderaon, t s apparent that Galleo's proposal amounts to a paral revson of our obseaon language or of our experence experence whch partly contrads the dea of the moon of the eath s tued nto an experence that cons t, at least as far as terresal thngs' are conce ed Ths s what aual has But Galeo wants to persuade us that no change has taken place, that the second conceptual system s already unversally kno, even though t s not unversally used Salva, hs representave n the Dalogue, hs opponent Smplco and Sagredo the ntellgent layman all connect Galeo's method of argumentaon wth Plato's theo of anamns a clever taccal move, typcally Gallean one s nclned to say Yet we must not allow ourselves to be deceved about the revoluona developmen that s actually takng place The ressance aganst the assumpon that shared moon s non operave was equated wth the resstance whch forgotten deas exhbt towards the attempt to make them known Let us accept ths nteretaton of the resstance! But let us not forget ts tce We must then admt that t rescts the use of the relavsc deas, conning them to pa of our eveday experence Outs ths part, e n ntelaneta space, they are forgotten and terefore not acve But outsde hs part there s not complete chaos Other concepts are used, among them whose ve same absolusc concepts whch derve from the rst paradgm We not only use them, we must also admit that they are enrely adequate No dcules arse as long as one remans wthn the lmits of the rst paradgm Experence', e the totalty of all facts from all domans, canot force us to car out the change whch Galleo wants to noduce The move or a change must come from a derent source It comes, rst, from the desre to see the whole [correspond] to ts pars wth wonderful smplcty', 4 as Copecus had already 3 ialogue op ct, pp. 32 and 4 6 bd, p 341 Galileo quotes here from Copeicus' address to Pope Paul n e outioni cf. also Chapter 10 and the Naatio Pma (quoted from E. Rosen, ee in Treat New York, 1 959, p. 1 65) For these phenomena appear to e linked most noby together, as by a golden chain; and each of the planets, by its oion, and order, and every nequai of its moon, bears wess that the earth ove and that we who dwell un the globe of the earth, instead of accepng its chane of poson, believe that the planets wander in all sos of moons of their o' Note that empircal easons are absent from the arment and have to be, for Copeicus himself admits (Commolu op ct, p 57) that the Ptolemaic theory s consstent wth the numercal data'.
72
AGAINST METHOD
expressed hmsel It comes from the typcally metaphyscal urge' for unty of understandng and conceptual presentaon And the move for a change s connected, secondly, wth the ntenon to make rm from the moon of the eath, whch Galleo accepts and s not prepared to gve up The dea of the moon of the eath s closer to the rst paradgm than to the second, or at least t was at the me of Galeo Ths gave sength to the Arstotelan argumens, and made them plausble To elmnate the plausblty, t was necessa to subsume the rst paradgm under the second, and to extend the relave noons to all phenomena The dea of anamns funcons here as a psychologcal crutch, as a lever whch smooths the process of subsumpon by concealng ts estence As a result we are now ready to apply the relave noons not only to boats, coaches, brds, but to the sold and well-establshed eath' as a whole And we have the mpresson that ths readness was n us all the me, although t took some eort to make t conscous Ths mpresson s most certanly erroneous t s the result of Galeo's propandsc machnaons We would do better to descrbe the stuaon n a derent way, as a change of our conceptual system Or, because we are dealng wth concepts whch belong to natural nteretaons, and whch are therefore connected wth sensaons n a ve drect way, we should descrbe t as a change ofece that alows us to accomodae the Copecan docne It s ths change whch underles the anson from the Arstotelan pont of vew to the epstemolo of mode scence For experence now ceases to be the unchangeable fundament whch t s both n comon sense and n the Arstotelan phlosophy The attempt to support Copecus makes experence ud' n the ve same manner n whch t makes the heavens ud, so that each star roves around n t by tsel 15 An emprcst who starts from experence, and buds on t wthout ever lookng back, now loses the ve ground on whch he stands Nether the eath, the sold, wel establshed eath', nor the facts on whch he usualy reles can be usted any longer It s clear that a phosophy that uses such a ud and changng experence needs new methodologcal prncples whch do not nsst on an asymmec judgement of theores by experence Clscalphys ntuvely adopts such prncples at least ts great and ndependent thnkers, such as Newton, Faraday, Boltzmann proceed n ths way But ts oal ne sll clngs to the dea of a stable and unchangng bass The clash between t 15. ioe op ct., p 120.
SEEN
73
doctrne and the actual pocedue s concealed by a tendenous esentaon of the results of esearch that hdes the evoluona gn and suggests that they arose fom a stable and unchangng souce These methods of concealment start wth Galeo's attempt noduce new deas unde the cove of anamness, and they culmate n Newton 1 6 They must be exposed f we want to arve at a bette account of the pogressve elements n scence My dscusson of the an-Copecan argument s not yet comlete So far, I have tred to dscover what assumpon wll make a stone that m alon a mng tower appear to fall saght down', nstead of beng seen to move n an arc The assumpon, whch I shall call the ret pnple, that our senses noce only relave moon and are nsensve to a moon whch objects have n common was seen to do the c What remans to be explaned s why the stone stas wth the tower and s not left behnd In order to save the opecan vew, one must explan not only why a moon that esees the relaon among vsble objects rans unnotced, but also, why a common moon of varous objects does not aect ther elaon That s, one must explan why such a moon s not a causal agent Tug the queson around n the manner expaned n the text to fooote 10, page 63 of the last chapter, t s now apparent that the an-Copecan argument descrbed thee rests on two natual nteretaons , the tolocal sumpton that absolute moon s always notd, and the dynaml pnple that objects (such as the fallng stone) whch are not nterfered wth assume ther natura moon For stotelans the natural moon of an object not nterered wth s rt, e constancy of quales and of poson 1 7 Ths corresponds to our own experence where thngs ave to be pushed around to move The dscove of seeds, bactera, ruses woud have been mpossble wthout a rm belef n the qualtave part of the law and t conrmed t n a most mpressve way Usng ths law scensts nferred that a stone dropped from a towe stuated on a movng eath would be left behnd Thus the elavty rncple must be combned wth a new law ofnera n such a fason that the moon of the eath can sll be asserted One sees at once that the followng law, the pnple ofrlar nea as I shal call t, ovdes the requred soluon an object that moves wth a gven anula velty on a frconless sphere around the cene of the earh wll connue movng wth the same angular velty for ever 6
lassic Empcsm', op cit This s he account of moon n the smolol account we have cula oon above and up-and-down moons on earth
74
A G A N S T M E T HO D
Combnng the appearance of the fallng stone wth the relavty prncple, the prncple of crcular nera and some smple assumpons conceg the composon ofveles, 1 8 we obtan an argument wch no longer endangers Copecus' vew, but can be used to gve t paral support The relavty prncple was defended n two ways The rst was by showng how t helps Copecus ths defence s hoc but not obeconable, because necessa for revealng natural ntereta ons The second was by ponng to ts funcon n comon sense, and by surrepously generalzng that funcon (see above) No ndependen argument was gven for ts valdty Galleo' s support for the prncple of crcular nera s of exactly the same knd He noduces the prncple, an not by reference to experment or to ndependen obseaon, but by reference to what eveone s already supposed to know imp So you have not made a hundred tests or even one And yet you so freely delare t to e ertain iti: Wthout experimnt I am sure that the eet wll happen as I tell you eause it must happen that way and I mght add that yu yourself also know that it annot happen otheise no matter how you may pretend not to know it But I am so handy at pkng peoples rains that I shall make you onfess ths in spte ofyoursel 19
Step by step, Smplco s forced to admit that a body that moves, without frcon, on a sphere concenc with the cene of the eath 18 hese assumpons were not at al a maer of course, but conicted wth some very basic ideas ofrstotelian physics The princple ofcirular nea s related to the mpetus theory but not denca wth it The impetus theory retans the dea that it needs a force to bring about change, but t puts the force nsde the changng objet Once pushed, an object connues moving in the me way in which a heated objet stays warm boh contain he cause ofher new state. Galileo modies his idea in two ways. First, he circular moon s supsed to on forever whle an object kept movng by mpetus wll gradualy slow down, just as a heated object, its anaoe, gradually becomes colder The arment for hs modicaon is gven in e tet below t is purely rhetorica. Secondly, he eteal circular moons must preed whout a cause if relave moons are not operave, hen inoducng a moon wh he me cene and he me anlar veli as a crcular moon upheld by ipetu cannot elminate forces we are on he way from impeus to momentum (cf. A. aie, ie Voruf Ga/i/ im Jahrhun Rome, 1949) Al hese changes are overlooked by ose who assume hat he transion was he simple result of a new and beer dynamics and hat he dynamics was already available, but had not yet been applied in a determned way. 19 logue op ct, p 145.
SEVEN
75
ll car out a boundless', a peetual' moon We know, of course, especally after the analyss we have just completed of the non-operave character of shared moon, that what Smplco accepts s based nether on experment nor on corroboraed theo It s a darng new suggeson nvolvng a emendous leap of the magnaon 20 A lttle more analyss then shows that ths suggeson s connected wth experments, such as the experents' of the Dsco2 by hoc hypotheses (The amount of frcon to be elmnated follows not from ndependent nvesgaons such nvesons comence only much later, n the 1th centu but from the result to be acheved, the crcular law of nera ) Vewng natural phenomena n ths way leads to a re-evaluaon of all experence, as we have seen. We can now add that t leads to the nvenon of a n nd ofece that s not only more sophscated ut ao r more spelate than the experence of Arstotle or of 20. For a Copeicn the only leap involed was the idencaon of the eh as a celesa object. According to Arstote celesa objects move in ciles and a body that moves n a crcle has neither heavness nor lghess for t cannot change ts disance from the cene neither n a natur nor in a forced way' e C, 269b34f. 2 1 ncidenty, many of the experences' or experments' used the arments about e moon of e eart are enrely ous. Thus Gileo, in h Trat l Sfa (Oe Vo , pp. 21 I), which follows the opnon of istote and of Ptolemy' ( 223), uses ts arment aganst a roon of the earth: . object which one lets fal from hgh places to the ground such as a stone from the top ofa tower would not fall towards the f of that tower for durng the me whch the stone cong recineary towards the ground, spends in the air, the earth, escapng it, and moving towards the east would receve it in a p far remed from the ft of the tower in the same
mann in which a stne that dedm the mt ofa ri ming ship ntfall towar i ft but more towar the ste' ( 224) The ticzed reference to the behavour of stones on ships s agan used n the gue ( 1 26), when the tolemaic arments are dscussed, but t s no longer accepted as correct t seems to be an approprate me,' ys Savia (bd., p. 18 0), to ake noce of a cen generosi on the of the Copeicans towards their advesares when, wth pehas t much lbe rl, they concede as tue and correct a number of experments whch their opponents have never made. Such for example s that of the body falling from the mast ofa shp whle t is in moon . . ' Earlier, p. 1 54, t s mplied rather than obseed, that he stone wll fl to the ft of the mast, even if the shp should be n moon whle a oible experment s dscussed on p 186 Bruno (La C le C Oe liane , ed Giovann Genle, Bar, 1907, p 83) akes t for granted that the stone w arive at he foot of the mast. t should be noted that the problem dd not readiy lend tself to a epermenal soluon. Eperments were made, but ther result were far from cocluive Cf. A. Armage, The Devaon of Fang Bodes', nna ofS 5, 9 7, pp 342, and A. Koyr, Metsi and Meurt Cambdge, 1 968, pp. 9 The tower arment can be found n Arstote, e oelo 296b22, and Ptolemy, nti , Copecus dscusses t n the me chapter of e R but es to defue it n the next chapter ts role in the Middle Ages is descrbed n M. Clage, The ce ofMechani in theMileg Madison, 1959, Chapter 10
76
AGANST MTHOD
common sense Seang aadocally, but not ncorrectly, one may say that Galleo ts an ece that h metaphscal nedts It s by means of suc an eperence tha te ranson fom a geostac cosmoloP o the ont of view of Coecus and Kele s acheved 2 22 Aan Chalmers, n an interesng and well-ared paper (The Galileo That Feyerbend Missed An mproved Case Against Method' n A Schuster and R.R. Yeo (eds), e Politi and Rhetoc ofStc Method Dordrecht, 1986, pp I), disnishes between Galileo's conbuons to a new science, on the one hand, and the queson of he sia condions n which that science is developed and prcsed, on the other', admits that propaganda' (though much less than suggest) may have been pa of his aempt to change the laer, but emphasizes that it ds not aect the former The main source for Galeo's conibuon to scence itsel, says Chalmers, is his Two New Sc' Ths s the work should have studied to explore Gaieo's predure But he Two New Sc do not deal with the topic was disussing, viz the anson to Copeicus He Galileo used predures rther derent from those of his later work Lynn Thodke, who shares Chalmers' evaluaon of the le wished that Gaileo had wren a systemac textbk on that subect Hto of M and emtal Sce Vol 6, New York, 1 94 1 , pp 7 and 62 Galieo might have done beer to wrte a systemac textbk than his provave daloes') Now for such a textbk to have subsance it would have to be as generl as its Arstotelian iv and t would have to show how and why Aistotelian concept needed to be replaced at the most elemenary level Aristotelian concepts, thoug absact, were closey related o common sense Hence t was necessary to replace some common noons by othes ( am now speakng about what Chalmers calls perceptual relavi' p 7) Two quesons arse how big were the changes? and was propaganda (rhetoric, were iona moves') needed to car them out? My answer to the laer queson is that dscourse aempng to bring about maor conceptual changes is a normal pa of scence, common sense, and culturl exchange (for the laer cf Chapter 16 and Chapter 1 7, tem vi, open exchange'), and that t ders from the discourse carred out within a more or less stable frmework Personally, am qute prepared to mae t pa of ronal But there est philosophical schls that ope it or call it ncoherent (cf Chapter 10 of Farewel to Ren which discusses some of Hilary Puam's view) Using the tinol ofthe sch speak of Galleo's ckery', etc And add that science conans ingredents that casionaly need such ckery' to become accepable The dierence between the Sc and the aloe terefore, is not between scence and siolo but between techncal changes in a narrow eld and basc changes, reaiscally intereted My answer to the rst queson is that pereptual relatvi, though acknowledged by many scholars (and by Arstote hmsel, was not a common ssesson ( Galileo ints out that even some of his fellow scensts stumbled at this nt) and thus had to be argued for Tis is not at al susng, as my discusson of qualiave dicules n Chapter 5 show Besides, is it realy tue that a aveller on a boat sees the harbour as receding as f t were removed by some sange force? conclude that Galeo's ckery' was necesry for a proper undersanding of the new cosmolo, that t s ickery' only for philosophes that set narrow condons on conceptual change and that t should be extended to areas sll resicted by such condions (in Chapter 12 are that the mind-y problem s one such area)
8 In ton to natural nteretatons Galleo also chang sensaons that se to nger Cs He admts that there are such ssatons he pra Csr hng dsregard th he ams to he red th wth the help of the telescope Hower he oe no theorecal reons why the telce should e eed to e a te pure ofthe s
I repeat and summarze argument s proposed that refutes Copecus by obseaon The argument s nverted n order to dscover the natural nteretaons whch are responsble for the conadcon The oensve nteretaons are replaced by others, propanda and appeal to dstant, and hghly theorecal, parts of common sense are used to defuse old habts and to enthrone new ones. The new natural nteretaons, whch are also formulated explctly, a auary hypotheses, are establshed partly by the support they gve to Copecus and partly by plausblty consdera ons and ad hoc hypotheses enrely new experence' arses n ths way There s a yet no ndependent evdence bt ths s no drawback; t takes me to assemble facts that favour a new cosmolo. For what s needed s a new dynamics that explans both celesal and terresial moons, a theory of sold objects, aero dynamcs, and all these scences are sll hdden n the future But ter tk now we-ned, for Galeo's assumpons, hs ad hoc hypotheses ncluded, are sucently clear and smple to prescrbe the drecon of future research Let t be noted, ncdentally, that Galleo's predure drascally reduces the content of dynamics Arstotelan dynamics was a general I Galleo's circular la is not the right dnamics t ts nether the epicycles hich sll cu in Copeicus, nor Kepler's eipses n fact, it is reted by both Sl, Galileo regards it an essena ingredient of the Copeican int of ie and ies to �emove bodes, such comets, hose moon quite obviousy is not circuar, om ntelanetary space n his Gaieo tked about comets [and intereted them as illusions, smilar to rainbow] in order to protect the Copeican system from ossible falsicaons' P Redondi, Gall Hetc, Princeton, 1 987, pp 1 45, 3 1
77
78
AGANST METHOD
theo of change, comprising lomoon, qualitave change, generaon and corpon, and it could also be applied to mental presses. Galileo's dynamics and its successors deal wth locomotion only, and here ain just with the lomoon of matter Other kinds of moon are pushed aside with the promisso note (due to Demritos) that locomoon will eventually be capable of explaining all moon. Thus a comprehensive empirical theo is replaced by a narrow theo plus a metaphysics of moon, 2 just a an empirical' experience is replaced by an experience that contains speculave elements. Countenduion, however, is now seen to play an important role both v-vis theories and vis-vis facts. It clearly aids the advancement of science. This concludes the consideraons begun in Chapter 6 I now tu to another part of Galileo's propanda campai, dealing not with natural interetaons but with the sso core of our obseaonal statements Replying o an interlocutor who expressed his astonishment at the 2 The socalled scienc revoluon led to astoundng discoveres and considerbly etended our knoledge of physics, physiolo, and asonomy This as achieved by pushng aside and regarding as irrelevant, and o nntt, those facts hich had suped the older philosophy Thus the evdence for tchcr, demonic ssession, he estence of he del, etc, s disregrded eth wh he supersons' i once conrmed The result as that toards the close of the Middle Ages science as fored aay from human psycholo, so that even the great endeavour of Erasmus and his frend Vives, as the best represenaves of humanism, did not suce to brng about a reapproachment, and psychopatholo had to ail centuries behnd the developmenal end of general medicine and surgery a maer of fact the divorcement of medical science from psychopatholo as so denite that the laer as alays toally relegated to the domain of theolo nd ecclesiasc and civil la - to elds hch naturally became further and further removed from medcine ' G lbrg, MD, The Medl Man and the Witch, Balmore, 935, pp 3 and 70 Asonomy advanced, but the knoledge of the human mind slipped back into an earlier and more prmive sage Another example is asolo n he early sages of he human mind,' writes A Comte (Cou Phshie Posite, Vol , pp 27380, ed Lit, Paris, 836), ese connecng links beteen asonomy and biolo ere studied from a very dierent int ofvew, but at /et hey ere studed and not le out of sight, as is he common tendency in our own me, under he resicng inuence of a nascent and incomplete sivsm Beneah he chmerical belief of he old philosophy n he physiological nuence of he stars, here lay a song, hough confused recognion of he uh hat he facts of life ere in some ay dependent on he solar system ike al primive inspirons of man's intelligence hs feelng needed reccation by sive sence, but not desucon hough unhappily in science, as in litics, it is often hard to reorgne hout some bref perod of overthro' A hird area is mahemacs Aristotle had developed a highly sophiscated heory of he connuum hat overcame he dicules rised by eno and ancipated quantum heorecal ideas on moon (see fote 5 and tet of Chapter 5) Most physicists retued to he idea of a connuum consisng of indivsible elements - if hey considere uch recondite maers, hat is
79
EGHT
small number of Copecans, Salvia, who act[s] the part of Copeicus', 3 gives the following explanaon You wonder that there are so few followers of the Pythagorean opinion [that the eath moves] whl I am astonished that there have been any p to this day who have embraced and followed it Nor can I ever sufciently admre the outstanding acumen of those who have taken hold of this opnon and accepted it as ue they have, through sheer force of intellect, done such violence to their own senses as to prefer what reason told them over that whch sensble experience plainly showed them to be the cona. For the arguments anst the whirling [the rotaon] of the eath we have already examined [the dynamcal arguments discussed above] are ve plausble, as we have seen; and the fact that the Ptolemacs and the Aristotelians and all their disciples took them to be conclusive s indeed a song argument of their eecveness. But the experiences which overtly conadict the annual movement [the movement of the eath around the sun] are ndeed so much greater in their apparent force that, I repeat, there s no lmit to my astonshment when I reect that Aristarchus and Copecus were able to make reason so conquer sense that in deance of the latter, the former became misess of their belief. A little later Galieo notes that they the Copeicans] were condent of what their reason told them And he concludes his bref account of the origins of Copecansm by sayng that with reason as his gude he [Copeicus] resolutely connued to arm what sensble experience seemed to conadict'. I canot get over my amazement', Galileo repeats, that he was constantly willing to persist n sayng that Venus might go around the sun and might be more than sx mes as far from us at one me as at another, and sll look always equal, when t should have appeared forty mes larger' The experiences whch overtly conadict the annual movement', an which are much greater in their apparent force' than even the namical arguments above, consist n the fact that Mas, when it s close to us . . . would have to look sixty mes as large as when t is most istant. Yet no such dfference is to be seen. Rather, when it is in oosion to the sun and close to us t shows itself only four or ve
3 oe, op cit, pp 3 and 256 4 ibid, p 328 At other tmes Galileo speaks much more belligerently and o acally, and apparently thout any aareness of the dicules menoned here Cf his prepartory notes for the leer to Grand Duchess Chrisna, Oe, V, pp 3 67 5 ibi p 335 6 ibid, p 339
8
G A N S T M HO
mes as large as when, at conjuncon, it becomes hden behnd the rays of the sun.' 7 Another and greater dculty s made for us by Venus which, if it crculates arund the sun, as Copecus says, would now be beyond it and now on this sde of it, recedng from and approaching towards us by as much as the dameter of the circle it describes. Then, when it is beneath the sun and ve close to us, its disc ought to appear to us a little less than forty mes as large as when it s beyond the sun and near conjuncon. Yet the dierence is almost impercepble.' In an earier essay, The Asser, Galileo expressed himself sll more bluntly Replying to an adversa who had raised the issue of Copecansm he remarks that neither Tycho nor other tronome nor Cis could ear rute Ptoly] inasmuch as a most important argument taken from the movement of Mars and Venus stood always in their way'. (Ths argument' is menoned ain n the Dialoe, and has just been quoted.) He concludes that the two systems' (the Copeican and the Ptolemaic) are surely false'. 8 We see that Galleo's view of the origin of Copeicansm diers markedly from the more familar historical accounts. He nether points to n fas whch oer inducve suo to the idea of the moving ea, nor does he menon any obseaons that would rute the geenc pont of vew but be accounted for by Copecansm. On the cona, he emphasizes that not only Ptolemy, but Copecus as wel, is refuted by the facts, 9 and he prases Aristarchus and Copecus for not havng given up in the face of 7 ibid, p 334 8 eAss, quoted from The Conte the Come of op cit, p 85 9 This refers to the perod before the end ofthe 6th century; cf Derek de S Prce, ConaCopeicus: A Crcal ReEsmaon of the Mathemaca Planetary Theory oftolemy, Copecus and Kepler, in M Clage (ed), CtPbs n theHtoof S, Madison, 959, pp 972 8 Prce deals ony th the ntc and the tl diculties ofthe ne vew (A consideration ofthe dynamical dicules ould rther senen his case) He ints out that under the best condions a geostac or heliosc syste using eccenc circles (ortheirequivaents) th cenal epicycles can account for al anlar moons of the planets to an accuracy beer than 6 excepng only the specia heory needed to account for Mercury and excepng also the planet Mars hich shos devaons up to 30 from a theory [This is] ceinly beer than the accuracy of 0 hich Copeicu himselfstated as a sasfactory goa for his own theory' hich as dicult to test, especialy in ve of the fact that refracon (most o on the horion) as not taen into account at the me of Copeicus, and that the obseaona basis of the predicons as less than sasfatory Carl Schumacher (Untuhung de ptolche e unt Pn, Mnster, 9 7) has found that the predicons conceing Mercury and Venus made by Ptolemy dier at most by n amount of 30 from those of Copeicus The devaons found beteen mode predicons and those of tolemy (and Copeicus),
GHT
81
such emendous dicules He praises them for havng proceeded
count"ndue
Ts, however, is not yet the whole sto. For while it might be conceded that Copeicus actd smply on fath, t may also be said that Galileo found hmself in an enrely dierent posion. Galieo, after all, invented a new dynamics. And he nvented the telescope. The new dynamics, one might want to point ot, removes the inconsistency beeen the moon of the eath and the condions aecng ourselves and those in the air above us'. And the telescope removes the even more glaring' clash beeen the changes in the apparent brighess ofMars and Venus as predicted on the basis of he Copeican scheme and as seen with the naked eye. This, ncidentally, is also Galleo's own view. He admits that were it not for the estence of a superior and better sense than natural and common sense to join forces with reason' he would have been much more recalcant towards the Copeican system' . The superior and better sense' is, of course, the telce and one is inclined to remark that the apparently counterinducve procedure was as a matter of fact inducon (or conjecture plus refutaon plus new conjecture), ut one ed etter ·ce, containng not only better natural interetaons but also a better senso core than was available to Galieos Arstotelian predecessors. 2 This matter must now be examined in some detail. The telescope is a superior and better sense' that gives new and more reliable evidence for udging astronomical matters. How is this hypothess examined, and what arguments are presented n its favour? In the Sis Nuncius, 3 the publicaon which contains hs rst hich in the case of Mercury may be as large as 7°, are due mainly to rong constants an inial condions, including an incorrect value of the constant of precesson or e vesality o the Ptolemaic scheme cf N R anson, Is, No 5 , 960, pp 58 0 Ptolemy Syn, i, 7 I I ale op cit, p 328 2 or this vie cf Ludovico Geymonat, Galil Gall, ans Sllman Drake, Ne York, 965 (rst talian edition 957), p 84 or the story ofGaileo's nvenon and ue of the telescope cf RS Westfall, Science and Paonage', Is, Vol 76 985, I According to Westfall, Galileo sa the telescope more as an insument of paonage than as an nsument of asonomy' (p 26) and had to be pushed into some aonoical applicaions by his pupil (and staunch Copecan) Casteli Galileo's telscopes ere beer than others n circulaon at the me and ere much in demand u h rst satised the demands ofpotenal pans Kepler, ho complained about q uality of teescopes (cf ne chapter, fote 2 nd te) and ho ould have lo to possess a beer insument, had to ait 3 e Sil Msg ofGall Galilel, ansl E St Carlos, London, 880, reisu by Dasons of Pall Mall, 960, p 0
82
AGA NST METHOD
telescopic obseaons, and which was also the rst important conibuon to his fame, Galieo writes that he succeeded (in building the telescope) through a deep study o the theo o refracon' This suggests that he had theoretil reo fr preferg the results of telescopic obseaons to obseaons with the naked eye. But the parcular reason he gives hs insight into the theo of refracon is not coe and is not sut either. The reason is not correct, for there est serious doubts as to Galieo's knowledge of those parts of contempora physical opcs whch were relevant for the understanding of telescopic phenomena In a letter to Giuliano de Medici of 1 October 1610, more than half a year after publcaon of the Sire Nunus, he asks for a copy of Kepler's ti of 160 5 poinng out that he had not yet been able to obtain it in Italy. Jean Tarde, who in 161 asked Galleo about the consucon of telescopes of pre-assied maicaon, reports in his dia that Galieo rerded the matter as a dicult one that he had found Kepler's ti of 16 1 1 so obscure that perhaps its own author had not understd it' In a letter to Lice, written o years before hs death, Galieo remarks that as far as he was conceed the nature of light was sll in darkness. 8 Even if we consider such utterances wth the care that is needed in the case of a whimsical author like Galileo, we must yet admit that his knowledge of opcs was inferior by far to that of Kepler.9 This is also the 4 Gaileo, Oe, Vol X, p 44 5 A d Vteon Paralpoma qubus Astnom Pa Ot Trtur, Frnu, 604, to be quoed from]ohann Kl ammelte Wk, Vol , Munich, 939, ed Fr Hammer This pacular ork ill be referred to as the opcs of 604' t as the ony useful opcs that existed at the me The reason for Galileo's curiosity as most likey the many references to ths ork in Kepler's reply to the Sr Nunus For the history of this repy as ell as a anslaon cf Krs Catn th all Srl Msg, ansl E Rosen, Ne York, 965 The many references to eaier ork contained in the Coatn ere intereted by some of Gaieo's enemies as a sign that his msk had been to from his face' (G ugger to Kepler, 28 May 60, Galileo, Oe, Vol X, p 36) and that he (Kepler) had ell plucked hm', Maesin to Kepler, 7 Aust (Galileo, Oe, Vol X, p 428) alileo must have received Kepler's Catn before 7 May (Oe, X, p 349) and he acknoledges receipt of the printed Coaton in a leer to Kepler of 9 Aust (Oe, X, p 42 ) 6 tce Augsburg, 6 , We, Vol V, Munich, 94 This ork as wien after Galileo's dscoveres Kepler's reference to them in the peface has been anslated by E St Carlos, op cit, pp 37, 79 The problem referred to by Tarde s eated n Kepler's tce 7 Geymonat, op cit, p 37 8 Leer to Lice of23 une 640 Oe, V, p 208 9 Kepler, the most knoledgeable and most lovable of Gaieo's contem aies, gives a clear account of the reasons hy, despite his superior knoledge o
83
EGHT
onclusion of Professor E. Hoppe, who sums up the situaon as follows aeos asseron that havng heard of the Dtch teescope he reconscted the apparaus by mathemaca cacaon mst of corse be nderstd wth a gran of sat for hs wrngs we do not nd any cacuaons and the repot, by etter, whch he gves of hs rst eot says that no better enses had been avaabe s days ater we nd hm on the way to Vence wth a better pece to hand t as a o the Doge Leonard Dona Ths does not k ke cacaon t rather ks ke a and eror The cacaon may we have been of a derent knd, and here t scceeded, for on 25 Agust 1609 hs saa was ncreased by a factor of three
g
Trial and error this means that in the case of the telscope it was ece and not mathemacs that led Galileo to a serene faith in the reliability of his device'. 2 This second hypothesis on the origin of the telescope is ao supported by Galieo's tesmony, in which he wrtes that he had tested the telescope a hundred thousand mes on a hundred thousand stars and other objects' 22 Such tests produced great and surising successes The contempora lterature letters, books, gossip columns teses to the eaordina impression which the telescope made as a means of improving ttal vision
opcal maers, he refrained from aempng to consuct the device' You, hoever,' he addresses Gaileo, desee my praise Putng aside al misgivings you tued directly to visua expermentaon' (Coatn, op cit, p 8) t remains to add that Gaileo, due to his lac of knoledge in opcs, had no misgivings' to overcome Galileo totaly ignorant of the science of opcs, and it is not t bold to assume at is as a most happy accident bo for him and for humaniy at large', Ronchi, Stc Change, ed Crombie, London, 963, p 550 20 ie Gchihte Otik, Leipzig, 926, p 32 Hoppe's udgement conceing e venon of the telescope is shared by Wolf, inner and others Huyghens ints out at superhuman inteigence ould have been needed to invent te telescope on e basis of the available physics and geome Aer al, ys he, e sll do not uderstd the orings of the telescope (Diopca', Hugii Ous Postuma, Ludg Bat, 903, 63, paraphrased after AG Kser, Gchichte Mathatik, Vol V, Gtngen, 8, p ) 2 Geymonat, op cit, p 39 22 Leer to Caroso, 24 May 66, Oe, X, p 357 leer to P Din, 2 May Oe, X p Nor can it be doubted that , over a perod of to years no, have tested my insument (or rather dozens of my insuments) on hundreds and housads ofobects near and far, la and smal, brght and dar; hence do not see ho it can enter the mind of anyone that have simplemindedy remained deceived in y obseaons' he hundreds and thousands of experments remind one of ooe, and are most liey equay spurous Cf fote 9 of Chapte 9
84
A G A N S T T HO D
Julius Caesar Lagalla, Professor ofPhilosophy in Rome, describes a meeng of 1 6 Aprl 161 1, at which Galileo demonsated his device We were on top oftheaniculum, near the city gate named after the Holy Ghost, where once is said to have stood the vla of the poet Maral, now the propety of the Most Reverend Malvasia By means of this insument, we saw the palace of the most illus ious Duke Atemps on the Tuscan Hills so disnctly that we readily counted its each and eve window, even the smallest and the distance is sixteen Italian mies. From the same place we read the letters on the lle, which Sixtus erected in the Lateran for the benedicons, so clearly, that we dsnguished even the periods caed between the letters, at a distance of at least two mles.' 23 Other reports conrm this and simlar events. Gaileo hmself points to the number and importance of the benets whch the insument may be expected to confer, when used by land or sea'. 4 23 Legala, De phom in oe lun ni telii usa a D Gai Galilei nunc itm sust psi dputatio (Venice, 62), p 8 quoted from E Rosen, e Naming ofthe Tee, Ne Yor, 947, pl 54 The relar re of the uchy of Urbno on events and gossip in Rome conain the follong noce of the event Geo Gei e maemcian, arved here from Forenc bfore ter Formery a Professor at Padua, he is at present reaned by the Grnd Due of Tuscny at a salary of scudi He has obseed the moon of te srs th the i, hch he invented or rther improved Aganst the opnion of l ancent phosophers, he declares that there are four more sars or planets, hch are tetes of Jupiter and hch he cls the Medcean ies, as ell as to compnons o Satu He has here discussed this opinion of his ith Father Claius, the Jesut Thursday evenng, at Monsignor Maavasia's esate outsde the St Panus gate a hgh and open place, a bnquet as gen for him by Frederc Cesi the marqu o Moncei and nephe of Cardinal Ces, ho as accompanied by his iman, Pau Monadesco In the gatherng there ere Gaieo a Feming named Terren Persio, of Cardina Cesi's renue, [] Gala, Professor at the Unersty here e Gree, ho s ardna Gonga's mathemacan P, Professor at Sena, and as mny as eght others Some of them ent out expressy to peform s obseaon and even though they sayed unl one o'cl in the moing, they s did not reach n agreement n the ves' (quoted from Rosen, op cit, p 3 ) 24 Sal Msg, op cit, p ii According to Bereus (De V toe, Hae, 655, p 4), Prince Mor immedatey reaed the m vaue o the telescope and ordered that its nvenon - hch Bereus abues to ch Jansen - be ept a secret Thus the telescope seems to have commenced as a secret wean and was tued to asonomcal use ony later There are many ancpao o the telescope to be found n the terture, but they mosty belong to the domain o naturl magic and are used accordngy An example s Agippa von eeshem, who in his b on cult phosophy (wien 1 509, Bk II, chapter 23), ites et ego no ex s mrnda concere, et specula n quibus quis videre potert quaecunque voluer a longissima dsana' So may the toy of one age come to be the precous easure o another', Henry Morley, e ofoelius Aa Nettheim, Vol II, p 166
EIGHT
85
Te tetral success of the telescope was, therefore, assured. It applicaon to the sta, owever, was an enrely dierent matter.
9 or the ntal experience wth the telcepr such reons e t telcc oseatons of the s are ndstn ntenate contrdo and n co wth what eone n see wth hs unad es And the on theo that could he helped to sarate telescc lusons om vedcal phoma w rted smple tests
To start with, there is the problem of teescopic vision This problem dierent for celesal and terrestria objects and it was also thought to e dierent in the two cases. It was thought to be dierent because of the contempora idea that celesal objects and terresial objects are formed from dif ferent materials and obey dierent laws Ths idea entails that the result of an interacon of light (whch connects both domans and has special properes) with terresial objects cannot, without futher discussion, be extended to the s To this physical idea one added, enrely in accordance with the Aristotelian theo of knowledge (and also with present views about the matter) the idea that the senses are quanted wth the close appearance of terresal objects and are, therefore, able to perceive them disnctly even if the telescopic image should be vastly distorted, or disgured by coloured frnges. The srs are not known from close by. Hence we cannot in their case use our memo for separang the conibuons of the telescope and those which come from the object 1 is hardly ever realied by those who are (wth Kser, op cit, p 133) that one does not see how a telescope can be good and useful on te earth and ye t deceive in the sky' Kser's comment is directed against Horky See below, tet to fotes 16 of the present chapter 2 hat the senses are acquainted wth our everyday surroundings, but are lable to gve misleadin repos about objects outside this domain, is proved at once by the aearanc ofthe moon On the earth large but distant obects in familiar surround ings, such as mountains, are seen as being large, and far away he appearance of the mn, however, gives us an enrely false idea of its dstance and its sie
86
IE
87
itself. 3 Moreover, all the famliar cues (such as bacground, overlap, nowledge of nearby size, etc.), which constute and ad our vision on the surface of the eath, are absent when we are dealing wth the sy, so that new and surprising phenomena are bound to cur. 4 Only new theo of vision, containing both hypotheses conceing the behaviour o light with the telescope and hypotheses conceng the reacon of the eye under exceponal circumstances, could have brdged the gulfbetween the heavens and the eath that was, and sll i, such an obvious fact of physics and of asonomcal obseaon 5 e sall soon have occasion to comment on the theories that were available at the me and we shall see tht they were unt for the tas and were refuted by plain and obvious facts For the moment, I want to stay wth the obseaons themselves and I want to comment on the conadicons and dicules which arise when one ies to tae the celesal results of the telescope at their face value, as indicang stable, objecve properes of the things seen Some of these dicules already announce themselves in a report of the contempora A whch ends wth the remar that even thou they (the parcipants in the gatherng described) went out expressly to perform ths obseaon (of "four more stars or planets, which are satelltes of Jupiter as well as of two companions of Satu 7 ), and even though they stayed unl one n the mong, they sll did not reach an agreement in their views' 3 It is not t dcult to separate the leers of a famiiar aphabet from a background of unfamiliar lines, even if they should happen to have been wren wth an almost illegible hand o such separaon is ssible wth leers whch belong to an unfamiliar aphabet he pas of such leers do not hang together o form disnct paes which sand out from the background ofgenera (opca) noise (in the manner described by K Koa, Pcho/ ull, 9, 922, pp 55 , pay reprinted in MD Veon ed), emts n Vsual Pcton, ondon, 966 cf aso the arcle by Gochaldt in te same volume) 4 or the imnce of cues such as diaphragms, crossed wires, background, etc, in the lalizaion and shape of the telescope image and the sange situaons arising when no cues are present cf Chapter of Ronchi, Ot, op cit, especialy pp 5 , 74, 89, 9 1, etc Cf aso R Gregory, , ew York, l 966, sm ad p 99 on the autoinec phenomenon) P Klpaick (ed), Eloratons n ransonal Pchol, New York, 961, contains ample material on what happens in he absence of familiar cues It is for ths reason that the deep study of the theory of refracon' which Gaileo re te ded to have carried out (te to fote 3 ofChapter 8) would have been quite ut for esablishing the usefulness of the telescope cf aso fote 6 of th preset chapter 6 Deai ls in Chapter 8, fote 23 This is how the ring of Satu was seen at the me Cf also R Grego, h I ntelgt Eye, p. 1 1 9.
88
AGA NS HOD
Another meeng that became notorious all over Europe maes the situaon even clearer About a year earlier, on and April 610, Galieo had aen his telescope to the house of his opponent, Magii, n Bolona t demonstrate it to twenty four professors o all facules. Hor, Kepler's overlyexcited upil, wrote on this occasion 8 I never slept on the th or th April, da� or night, but I tested the insument of Galileo's in a thousand ways, both on thngs here below and on those above. Below t wor wonul; in the heavens t deceives one, as some xed stars [Spica Virgins, for example, is menoned, as well as a terresal lame] are seen double I have as wiesses most excellent men and noble doctors . . and all have admitted the insument to deceve . This sienced Galieo and on the 6th he sadly left quite early in the moing . . not even thanng Magini for his splendid meal . Magini wrote to Kepler on 6 May: He has achieved nothing, for more than twenty leaed men were present yet nobody has seen the new planets disnctly (nemo perfecte vdit); he will hardly be able to eep them.' A few months later (in a letter sied by Runi) he repeats: Only some with sha vision were convinced to some extent.' 2 After these and other neve reports had reached Kepler from all sides, lie a paper avalanche, he ased Galileo for wiesses 3 I do not want to hide it from you that quite a few Italians have sent letters to Prague asseng that they could not see those stars [the mns oupiter] with your own telescope I as mysef how it can be that so many deny the phenomenon, includig those who use a telescope. Now, if I consider what casonaly happens to me then I do not at al rerd it as impossible that a single person may see what thousands are unable to see . . . 4 Yet I reget that e 8 aieo, Oe, Vo X, p 342 (my acs, referng to th derence commene un above, between celesa and terresa obseatons) 9 he hundreds' and thounds' of obseaons, as, ec, whch we nd here agan are hardy more than a rhetorcal oursh (corresndng to our have told ou thousand mes) hey cannot be used to nfer a lfe of ncessnt obseaon 1 0 Here agn we have a case where eteal clues are mssng Cf Ronc Oti, op ct, as regads the appearance of ames, sml lghts, etc 1 1 Leerof26 May, Oe, 12 bd, p 96 13 Leer of 9 Aust 1 610, quoted from CasparDc, Johann K in S , Vol 1 , Munch, 1930, p 349 14 Kepler, who suered from Poyopa (nstead of a snge small obect at a gre dsance, two or three are seen by those who suer from ths defect Hence, nstead o a sngle mn ten or more present themselves to me', ation, op ct, foe 94 cf aso the remander of the fote for further quoatons), and who was fam wth Plaer's anatomcal nvesgatons (cf SL Polya, e Retina, Chcago, 1942 pp 134 for deals and lterature), was well aware of the need for a physio
tim oftnomil oeations
NINE
89
conrmaon by others should take so long in tuing up . Therefore, I beseech you, Galileo, give me wiesses as sn as possible . . ' Galieo, in his reply of 1 9 August, refers to hmsel to the Duke ofoscana, and Giuliano de Medici as wel as many others in Pisa, Florence, Boloa, Vence and Padua, who, however, reman silent and hesitate Most of them are enrely unable to disngush Jupiter, or Mars, or even the Mn as a planet ' 5 - not a ve reassurng sate of ar, to say the least. Today we understand a little better why the direct appeal to telescopic vsion was bound to lead to disappoinent, especialy in the nal stages he man reason, one already foreseen by Aristotle, was that the senses applied under abnoal condions are liable to give an abnormal response Some of the older historans had an inling of the situaon, but they speak nate, they y to explan the asce of sasfacto obseaonal reports, the p of what is seen in the telescope. 16 hey are unaware of the possibilty that the obseers mght have been dturbed by stng poste us also he extent of such ilusions was not realized unl qute recently, mainly as the result of the work of Ronch and his school. 1 7 Here sizeable varaons are reported in the placement of the telescopic image and correspondingly in the obseed maon Some obseers put the mage rght insde the telescope mang it change 15 Casparyc op cit p 352 16 hus E Wohl, Gali und sein Kamfr die Keine Lehre, Vol , Hamburg, 19, p 288, rites o doubt the unpleasant resuts ere due to the
lack of aining in telescopic obseaon, and the resited eld of vision of the Gailean telescope as e as to the absence of any ssibili for chnging the dtnce of the gasses in order to mae them t the pecuies of the eyes of the eed me ' A simlar udgement, though more drmacay expressed, is found in ur Koester's Slea, p 369 7 Cf Ronchi, Oti, op ct Htoire Lumie, Pas, 1956; St l annhle, Vacan C, 1964 t Fnmti Ati e tt, Rome, 1964 cf. aso E ntore's summa in Ar dhi s ecember 1966 pp . 333 I oud lie to acknoedge at this plae at rofer Ronchi's invesgaons have greay inuenced my thining on scienc method For a bef historica account of Gaieo's ork cf Ronchi's cle in AC Crombie (ed), t hange, ondon, 1963, pp. 54261 Ho lie this eld exlored omes clear from S olans's bk Ot ins, ondon, 1964 ons a physicist ho n his mcroscopc research (on crysts nd me) as distrated by one usion aer another He rites his tued our interest to the sis of other saons, th the ulmate unexpeted discoery that opca usions can, and do, play a very real pa in aeng many daily scienc obeaons his ed me to e the lkout and as a result I met more uso thn I had brned for' he usons of direct vision', hose role in scienc researh i sloly being redered ere e non to medaea rters on opcs, ho eated them n specia chapters of the ir texbs Moreover, they eated le-images as pcho phenomena, as
90
AGAIS MEHOD
ts lateral poson with the lateral posion of the eye, exactly a would be the case wth an after image, or a relex insde the telescope - an excellent proof that one must be dealng wth an illusion 8 Others place the mage in a manner that leads to no mancaon at all, although a linear macaon of over thirty may have been promsed 9 Even a doubling of mages can be explained as the result of a lack of proper fusng 2 Addng the many mperfecons of the contempora telescopes to these psycholocal dicules,2 one can well understand the scarcty of sasfacto repots and one s rather astonshed at the speed with whch the reality of the ne w phenomena was accepted, and, as was the custom, publcly acnowledged 22 hs development becomes even more puling results of a msapprehension, for an mage s merely the appearnce of n object outside ts place' as we read n John Pecham (cf David Lndberg, he "Perspeca Communs ofJohn Pecham', Arch ntiona dhtoire s, 1965, p 5 1 , as we as the lst paragph of Proposion 19 of Pecham's Ppea ommun, whch is to be found in John Peam and the Sce of Oti, D indberg (ed), Madison, 1970, p 171) 1 8 Ronch, Oti, op ct, p 189 hs may explan the frequeny uered desre to lk ns the telescope o such problems arise n the case of tt objects whose mages are relarly placed in the plane of the obje' (bd, p 1 82) 1 9 o he agnicaon of Geo's telesop cf he S Msg, op t, p 1 1 , cf aso A Sonnefeld, De Opschen Daten der Hmmelsferohre von Gaeo Gaile', Runchau, Vol 7, 1962, pp 207 he old rule that te size, siton nd gement according to whch a thng s seen depends on the size of the ane through whch it s seen' (R Grosseteste, De , quoted from Crombe, R Gste, ord, 1953, p 120), whch gs back to Eucld, at as s remember my dspnent when, havng built a reector wth an leged lne mancaton of about , I found that the mn was ony about ve mes ed and stuated quite close to the ular (1 937) 20 he imge remns sha and unchnged over a consderble intea - the lack of fusg may show tself n a doubling, however 21 he rs uble telescope which Kepler receved om Elector E of K (who n tu had receed it from Gaieo), and on whch he based hs Na oa a se qarJs se, rn, 1 61 1, showed the sas as sqar d ntensey red (G We, , p 61) Et von Kln himself ws unable to nything wth he telescope and he asked Clavius to send hm a beer nsuen rio l Pont Uni Grna, 530, f 182r) rncesco onana, who from 1643 onwards obseed the phases of Venus, notes an unenness o the boundary (nd infers mounains), cf R Wolf, Gchite rAstnomie Munich, 1 8 p. 398 or the dosncrasies of contemrry telescopes and descripve literture c Est Zinner, Dche und Nierndche Astnomische nstmte 11 1 Jahu, Munich, 1956, pp. 21621. Refer also to the author caaloe n the second pa of he bk. 22 ather Clavus eer of 17 December 61 0, Oe, X, p 485), the asonoer of the weful Jesut Collegium Romnum, prises Galeo as the rst to hae obseed the mns ofJupiter and he recognzes their rea. Magin, Grenber and others sn followed suit. It is clear that, in dong so, they dd not preed
91
IE
when we consider that many reports of even the best obseers were either plainlyfae, and capable ofbeing shown a such at the me, or else sentradio. hus alileo reports unevennes, vast protuberances, deep chass, and sinuosies 3 at the inner bounda of the lighted part of the moon while the outer bounda appear[s] not uneven, rugged, and irregular, but perfectly round and circular, as shaly dened a ifmarked out with a pair of compasses, and without the ndentaons of any protuberances and cavies 4 he moon, then, seemed to be according to the methods prescribed by their own philosophy, or else they were very lax in the nvesgaon of the maer. Professor McMullin cit., fote 32) maes much of this quick acceptance of Galileo's telescopic obseaons he relar perods obseed for the satellites and for the phases of Venus songy indicated that they were not aefacts ofphysiolo or opcs. here was surely no need for "aulia sciences . .' here was no need for auliary sciences,' wites McMullin, while using himself the unexamined auliary hypothesis that asonomica events are disn ished from physiologca events by their relari and their intersubjecv But this hypothesis is fe, as is shown by the mn illusion, the phenomenon of fata morgana, the rainbow, haloes, by the many microscopic illusions which are so vvdy described by olans, by the phenomena of witchcra which suive n our texks ofpsycholo and psychia though under a dierent name, and by numerous other phenomena he hythesis was aso befe by Pecham, Witelo, and other mediaeva schola who had studied the la and inteubecve iusions' ceated by lenses, mirrors, and other opca convances In anqui te fasehood of McMullin's hythesis was mm. Galileo explicity discusses and repudiates it in his bk on comets. hus a new theory of vsion was needed, not just to the Gailean obseaons, but aso to provide am for their asonomica reai Of course, Clavus may not have been aware ofthis need his is hardly sursing Aer some of his sophiscated 20thcentury successors, such as rofessor McMun, not aware of it either In addion we must int out that the relar periods' of the mns ofJupiter were not as well known as McMullin insinuates For his whole life aileo ed to detemine these periods in order to nd beer ways of detemining longtude at sea He did not succeed. ater on the me problem retued in a dierent fo when the aempt to detemine the veli of light wit more than one mn led to concng results his was found by Cassini shory aer Rmer's discovery - cf IB. Cohen, Rmer and the rst deteminaon of the vel of light (1 676)', s, Vo 3 1 (1 94), pp. 347. For the attude ofCavus and the sciensts of the Collegum Romanum cf the very interesng b G in Chin by asquae M. d'Elia, SJ., Cambridge, Mass, 19. he early obseaons ofthe asonomers ofthe Coll egu m are contained in their own Nuncius Sidereus', , II , pp 291-8 23. The i Msg, op. cit, p. 8 24 op cit., p 24. - cf. the drawing on page 97 which is taen from alileo's publication Kepler in his of 16 wites (on the basis of obseaons with the unaided eye) t seemed as though something was missng in the crculari of the ouost perphery' (Wke, Vol. II, p. 219). He retus to this asseon in his (op. cit., pp. 28), crciing alileo's telescopic results by what he hmselfhad seen with the unaided eye You as why the mn's outeost crcle ds not .also appear irrelar. I do not know how carefully you have thought about ths subect or whetheryour query, as is more lely, is based on popular impression. For in
kno to
t
Otc
{oeaton
92
A G A I N S T M ET H O D
full of moutais at the iside but perfectly smooth at the periphery, ad hs despite the fact that the peri hery changed as the result of the slight librao of the luar body The moo ad some of the plaets, suh as for eample Jupiter, were elarged while the apparet diameter of the ed stars decreased: the former were brought earer whereas the latter were pushed away. The s,' writes alieo, ed as well as errac, whe see with the telescope, by o· meas appear to be icreased in magtude n the se proporo as other objects, ad the Moo itself, ga increase ofse but i the cse of the stars such icrease appears much less, so that you may cosider that a telescope, whch (for the sake of ilusao is powerful eough to mag other objects a hudred mes, w scarcely reder the stars maged four or ve mes.'
:
my bk [he Ot of 1] I sate hat here surey some mpefeon in t outeost cirle during mn Study he maer, and once ain te how t ls to you ' Here he results of naked eye eaon are quoted ainst Gaeo telescopc re - nd h pefecy g reaon, as e shl ee beo The read ho remembers Kepler's yopia (cf fote 14 to his chapter) may onder ho h could ust his enes to such an extent The reply i conained n he follo quoaon (We, II, pp 194) Wen eclipses of he mn begn, I, ho suer fro his defe, become aare of he eclipse before al he oher obeers Long before eclpse s, I ven deect e direon from whch e shadow is approachg, he ohers, ho have very acute son are sl n doubt The aoremeno avine of he mn [cf he prevous quoaon] stop for me hen he m approches he shado, and he songest pa of he sun's rays is cut o ' Gi has o explanaons for he conradctory appearance ofhe mn The one no lunar aosphere (Msg, op cit, pp 26). The oher explanaon (ibid, pp 25 hich nvolves he angena appearance of eries of mounains lyng behind e oher, is not rely very plausible as he dsibuon of mounains near he vsible sie of he lunar gobe ds not sho he arrangement hat ould be needed (hs s no even beer esblshed by he publcaon of he Ruan mn photograph o 7 Oober 1959; cf Zdenek Kopa, An ntuin t the Stu f the Mn, Hond, 1966, p 242). 25. The lbrons ere noced by Galeo C G Rghin, Ne Light on Ga Lunar Obseaons', n ML RghinBonei nd R Shea (eds), Ren an Mystcm in the Stc Rutn, Ne York, 1975, pp 59. Thus it n sloppiness of obseaons but he phenomena hemselves hat msided Geo In o leers to he joua S (2 May and 1 0 ober 1980) TH accued me of ving a msleadng account ofGaleo's obseaona ski- I ed a r obseer hen his lunar obseaons ere n fa raher impressve accuon s refuted by he text to fotes 29 and 30 and by fote o present chapter Wtaker obviously hought my quoaons from Wolf (te fote 28) reeed my on opnon He aso nts out hat he coppelate o Gaileo's obseaons are much beer, from a mode int of vie, dcuts hch accompaned he Nunus. Ths is ue but ds not invaidate descrpon of e debate hich as based on he published account 26. Msg, op cit, p 38; cf aso he more deaied account in , op pp 336. The telescope, as t ere, removes he heavens from ' tes A Ch
93
NINE
The sages features of the early histo of the elescope emerge,
however, whe we ake a closer look at alileo's pu res ofthe moo I eeds oly a brief look at alieo's drawgs, ad a photograph of similar phases, to covice the reader that oe of the features recorded ca be safely ideed with ay kow markgs of the luar ladscape' 27 Lookig at such evidece it is ve easy o thik that alileo was ot a great asoomical obseer or else that the excitemet of so may telescopic discoveries made b him at that e had temporarily blurred his sll or crical sese' 8 Now his assero may well be ue (though I rather doubt it view of the quite exaordia obseaoal sll which alileo exhibts o other casios) But it is poor cotet ad,
n hs edon of Klme, De Kbewng Gte (Leipzg, 927, p 90), commenng on he decrease of he apparent diameter of sars wh he sole excepon ofhe sun and he mn Later on, he derent magnicaon of planets (or comets) and xed sars was used as a means of dsnishng hem. From experence, I know', wrtes Herschel n he paper reporng his rst oseaon of Urnus (Phl rans., , 78 , pp. 493 he planet s here idened as a cmet), hat he dameters of he xed sas are not proronaly magnied wh hgher wers, as he planets are; herefore, I now put on he wers of 460 and 932, and found he dameter of he comet increased in proron to he power, as it ought to be ' It is noteworthy hat e rule did not invaably apply to e telesopes in use at ileo's me. hus, commenng on a comet ofovember 1618, Horao rassi ('On he hree Comets of 1618', in he Cnte ft he Cme f, op cit., p. 17) ints out hat when he comet was obseed hrough a telescope it suered scarcely any eargemen', and he infers, pefecty in accordance wh Herschel's experience', hat it wlhave to be d hat it more remote from us han he mn. . . ' In his Astnml B (ibid., p 80) he repeats at, according to he common experence of illusious asonomers' om many pa of urope' he comet obseed wh a very extended telescope received scarcely any increment . . alileo (ibid, p. 177) accepts his as a fact, crczing only he conclusions whch rss wants to drw from it. hese phenomena refute aieo's asseron (syer, op cit., p. 204) hat he telescope works always in he same way'. hey also undeine his heory of idiaon (cf fote 56 to is chapter).
al
27.
Kopal, op. cit., p.
207
28. R Wolf (Gchchte Astnme, p. 396) remarks on he r quai of
aileo's drawngs of he mn ( . . seine Abbildung des Mondes ann man . aum · eine Kae nennen'), whie Znner ( Gchchte St, Berlin, 1931 , p. 4 73) cls alileos obseaons of he mn and Venus typca for he obseaons of a beginner' Hs pcture of he mn, according to Znner, has no simiri wh he �n (ibid p 472). Zinner also menons he much beer qua of he almost smultaneous obseaons made by he Jesuits (ibd., p 473), and he nay asks wheer alileo's obseaons of he mn and Venus were not he result of a ferle rain raher han of a careful eye (sollte dabe der Wunsch der Vater der oachtung gewesen sen?') - a pernent queson, especialy n view of he P enmena rey described n fote 34 to his chapter. 29 he discovery and idencaon ofhe moons ofupiter were no mean acheve ments especially as a useful sable suppo for he telescope had not yet been developed
9
AA NS MHOD
I submit, ot ve iteresg. No ew suggesos emerge for addioal research, ad the possiblity of a tt is rather remote There are, however, other hypotheses whch do lead to ew suggesos ad whch show us how comple the stao was at the me of alileo. Let us cosder the followig two. Hypothess I. alileo recorded fathfully what he saw ad i ths way le us evidece of the shortcomgs of the rst tele scopes as well as of the peculiares of cotempora telescopc visio Itereted i this way alleo's drawgs are repo of eactly the same kd as are the repo emer g from the eperiets of Satto, Ehrisma, ad Kohler - ecept that the characteriscs of the physcal apparatus ad the u famliarity of the objects see must be take ito accout too We must also remember the may colcg views whch wer held about the surface of the m, eve at aleo's me, ad whch may have iueced what obseers saw.34 What would be eeded order to shed more light o the matter is a em pirical colleco of all the early telescopc results, preferab parallel colums, icludg whatever pctorial represetaos
p
30. he reaon, among other thngs, is the great varaon of telescopc vison fro one obseer to the next, cf. Ronchi, Ot, op cit., Chapter IV 3 1 For a suey and some inuctory lterature cf. regory, op. cit., Chapter 1 1
For a more detaed discusson and literature cf. K.W Smith and W.M Sth Ptn and Mtn, hiladelphia, 1962, reprnted n pa in MD Veon, op ct he reader should aso consult Ames' arcle Anseionic lasses' ratns ransina Pch, whch deas wth the change of n vison caused b o slighy abno optca condons. A comprehensive account s given by I. R Nature fPcual Aain, New Yor, 1966. 32. Many o the old insuments, and excellent descrpons of them, are s avalable. Cf. Znner, Dche und Nierndchetnme nstmte. 33. For nteresng nfoaon the reader should consult the relevant pasges o Kepler's Catin as we as of hs Smnium (the laer is now avalable n a n translaon by E Rosen, who has added a considerable amount of bacground materia Kls Smnum, ed. Rosen, Madson, 1967) he standard wor for belefs of the e s sll lutarch's Fe n the Mn (it wll be quoted o Cheiss' ansaon ofMoral XI, London, 1967). 34. One describes the mn aer objects one thinks one can perceive on i surface' (Kser, op. ct, Vol , p 167, commenng on Fontanas obseao re of 1646) Maesn even saw ran on the mn' (Kepler, Catn, op c pp. 29f, presenng Maesn's own obseaona re); cf. also da Vnci, note quoted from.. Richter, he Nte fLnar Vn, Vol. II, New Yor, 19 p. 1 67: Ifyou eep the detals ofthe sts of the mn under obseaon you wloen nd great varaon in them, and this I myself have proved by drawng tem And th caused by the clouds that rse from the waters in the mn . ' For the nstabi o te images of unknown objects and their dependence on belief (or nowledge') Ronchi, Oti, op cit, Chapter IV
NINE
95
have surived. 3 5 Subacng insumental peculiaries, suc a ollecon adds fascinang material to a yet-tobe-written isto of ercepon (and of science) 36 Tis is e content of Hyoesis I Hpotess II is more spei tan Hypotesis I, and evelops it in a ceain direcon I ave been considering it, wi varying degrees of nsiasm, for e last two or ree years and my interest in it as en revved by a letter from Professor Stepen Toulmin, to wom I am grateful for is clear and simple presenaon of e vew t seems to me, owever, at e ypoesis is confronted by many dicules and must, peraps, be given up Hoess II, just like ypoesis I, approaces telescopic repors from e pont of view of e eo of percepon; but it adds at e pracce of telescopic obseaon and acquaintance wi e ew telescopic reports canged not only wat was seen roug e telescope, ut ao what w seen wth the naked e It is obviously of impoance or our evaluaon of e contempora atude towards Galileo's repors Tat e appearance of e stars, and of e moon, may at some me ave been muc more indenite an it is today was originally suggested to me by e estence of various eories about e moon wic are inompable w wat eveone can plainly see wi is own eyes Aamanders eo of paral stoppage (wc aimed to explain e pases of e moon), Xenopanes' belief in e estence of dierent suns and dierent moons for dierent zones of e eat, Heraclitus' assumpon at eclipses and pases are caused by e tuin of e basins, wic for im represented e sun and e moon all ese views run counter to e estence of a stable and plaily vsible surface, a face' suc as we know' e moon to possess Te same is ue of e eo of Berossos wic occurs as late a Lucreus 38 and, even later, in Alazen 3 Chapter 15 of Kopal, op cit, contains an interesng colleon of exacy this kind Wider scope has W Schulz, De Anschauung v Mn und sin Gtalt n Myts u Kunst lk, Berlin, 912 36 Oe must, of course, also invesgate the dependence of what is seen on the
urrent methos of pitorial representaon Outside asonomy ths was done by E mrch, and Iusn, London, 1960, and L Choulant, Ht aBblgrap Anatcal Iustratin, New York, 1945 (translated, wth addions by Singer and thers) wh deals wth aatomy Asoomy has the advatage that side of the uzze vi the stars, is fairly simple i stucture (much simpler tha the uterus, for emle) d relavely well kown c also Chapter 16 below 3 7 r these theories ad further literature cf JLD Dreye Hist f Atn al t Kl, New Yrk, 193 r ersos, cf oums ace Is N 3 , 967, p 65 Lucreus te n t aur fThin, asl Leard, New Yrk, 1957 p 26) Agai, she
96
AGAI NS MEHOD
Now such dsrerd for pheomea whch for us are quite obvious may be due either to a certa derece towards the esg evdece, whch was, however, as clear ad as detaled as it is today, o ee to dece n the ce e It ot eay to chse betwee these alteaves. Havg bee ueced by Wttgestei, Haso, ad others, I was for some me iclied towards the secod versio, but t ow seems to me that it s ruled out both by physioo (psycholo ad by storcal iformao We eed o remember how Copecus disrerded the dicules arsig fro the variaos i the brighess of Mars ad Veus, whch were wel kow at the me d as rerds the face of the m, we se that Aristotle refers to it quite clear whe obseg that the s do ot For rog ivolves rotao: but the "face, as it s caled, of the m s always see' 4 We may ifer, the, that th casoal dsrerd for the stablty of the face was due ot to a lac of clear impressios, but to some widely held vews about th ureliabiity of the seses The iferece s supported by Plut's discusso of the matter whch plaly deals ot with what is see (except as evdece for or agast ceta viws but with ce tons of pheomea otheise sumed to e we o 4 To begi with,' he says, t is absurd to cal the gure see i the m aeco ofvso . a codo whch we call bedaemet (gare yoe who asserts hs does ot obsee that t pheomeo should rather have cued relao to the su, sice the su l upo us kee ad volet , ad moreover does ot explai why dl ad weak eyes disce o disco of shape i the m but her o for them has a eve ad ll lght whereas those ofkee ad robt may revolve un herself le to a bl's sphere - f pehace to be - oe haf of he dyed o'er wth gowng lght ad by the revoluto of that sphere she may beget f us her varyg shapes u she tus that ery pa of her to the sght nd eyes ofme ' 39 f text to fotes of my Reply to Crtsm', op ct, p 246 I ntqu the dereces the magntudes of Veus nd M e regded as beg obous to our eyes', Smpus, De C, II, 12, Heberg, p 5 olemahus here consders the dcutes of Eudoxos' theory of hom sphees, tht Venus and Mars appear in the mdst of the rede mee may es brghter, that Venus] on mnless nghts causes es to shadows' (objeon of utolycus) nd he may wel be appeag to the b of a deepton ofthe senses (wch was equenty dscusd by ancent schs) e who must have been famar wth l these facts, ds not menon them n De Cl or n the Mi, though he gives an account of Eudoxos' stem the mprovements of olemarchus nd ippus Cf fote 7 of Cpter 8 41 De C 225 42 op ct, p 37, cf aso S Samburs, e i W fthe Gre, o, 1962, pp 244
NIN
FIURE 1 The shape of a lunar mounain and a wled plan, from aileo S Nunu, Venice, 1 610 (cf p 1 1 1)
97
98
AAINST ETHOD
vision make out more precisely and disnctly the patte of fac features and more clearly perceive the variaons' The unevenness also enrely refutes the hypothesis,' Plutarch connues, 4 for the shadow that one sees is not connuous and consed, but is not badly depictured by the words of Agesian "She gleams with re encircled, but within Bluer than lapis show a maiden's eye danty brow, a visage manifest In uth, the dark patches submerge beneath the bright ones which they encompass and they are thoroughly entwined with each other so as to make the deieaon of the gure resemble a pang' Later on the stabiity of the face used as an argument aganst theories whch rerd the mn as bein made of re, o air, for air is tenuous and without conguaon, an so it naturaly slips and ds not stay in place' The aearan o the mn, then, seemed to be a wellknown and dn phenomenon What was n queson was the relance of the phenomenon for asonomical theo We can safely assume that the same was ue at the me o Galieo 4 But th we must admt that Ga/leo s osatons could e checd the nad e and uld n ths w e osed a luso Thus the circular monster below the cene of the disk of the moon is well above the threshold of naked eye obseaon i 43 ibid, cf however, fote 1 7 to ths chapter, liny's remark (t Nat, II, 43,
) that the mn is now stted and then suddeny shining clear', as we as da Vnc'
re, referred to in fote 34 to this chapter ibid, p 50 45 A song arment n fur of this contenon is Kepler's descrpon of the mn his Ot of 1604 he comments on the broken charcter of the boun between lght and shadow (Wke, II, p 218) and descrbes the dark pa of the mn durng an eclipse as lking lke to esh or broken wood (bid, p 2 1 9) He retu to these passages n the Catn (op ct, p 27), where he tells aleo that th very acute obseaons ofyours do not lack the sup of even my own tesmony For [in my] Ot you have the halfmn diided by a wavy lne From this fact I dedue peaks and depressons n the y of the mn [ater on] I descrbe te mn dun an eclipse as lkng lke to esh or broken wood, with bright seas eneang into the region of the shadow' Remember aso that Kepler crices alleo's telescopc re on the bss ofhis own naed eye obseaons cf fote 24 of this chapter 46 There s one other int whch I must on no account forget, whch I h noced and rther wondered at t It s this he mddle of the n, as t see, cupied b a cen cai larger than all the rest, and n shape pefecy round I h lked at this depression near both the rst and the thrd uaers, and I represented t as well as I can in the second illusaon aready given It produce th me appearance as to eects of light and shade as a act like Bohemia would produc on the Earth, f t were shut n on al sides by very loy mountains anged on th crcumference of a pefet circle; for the act n the mn i waled n with peas o such enormous height that the furthest sde adacent to the dark on of the mn
99
NINE
dameter is larger than 31/ minutes of arc), whe a single glance convinces us that the face of the mn is not anywhere dgured by a blemsh of ths kind It would be interesn to see what contempora obseers had to say on the matter 4 or, if they were arsts, what they had to draw on the matter I summarze what has emerged so far Galileo was only slghtly acquainted wth contempora opcal th. 48 His telescope ve surising results on the eath, and these results were duly praised. Trouble was to be expected in the sky, as we know now Trouble promptly arose the telescope produced spurious and conadicto phenomena and some of its results could be refuted by a simple lk wth the unaided eye. Only a new theo of telescopc vision could brng order into the chaos (which may have been sll larger, due to the dierent phenomena seen at the me even wth the naked eye) and could separate appearance from reaty Such a theo was developed by Kepler, rst in 1 6 and then ain in 161 1 4 Accordng to Kepler, the place of the image of a puncform object s found by rst acing the path of the rays emergng from the object aordng to the laws of (reecon an) refracon unl they reach the eye, and by then usng the prnciple (sll taught today) that the image wll be seen in the point determined by the bacward
.
seen bathed in sunlight before the boundary between lght and shade reaches haf way across the circular space ' sg, op cit, pp 21) This descipon, I think, dentely refutes Kopal's conjecture of obseaona l It is interesng to note the dierence between the woodcuts in the Nunu ( 131, Fire ad ieo's original drawng The wcut corresnds uite closely to the descron whie the orginal drawng wth its impressionisc features ( Kaum eine Kae,' ys Wol) is vage enough to escape the accusaon of gross obseaona error 7 I canno help wondering about the meaning of that large circular cai n hat I usually call the left coer of the mouth,' wites Kepler (Catn, op cit, p 28), nd then preeds to mae conjectures to its origin (conscious eo b ntelligent beings included) 8. Contemporry academic opcs ent beyond simple geomeical consucons (hich alileo may have known) and included an account ofwhat se when ling at a mirror, or through a lens, or a combinaon oflenses. Excepng rradiaon aileo nohere considers the propees of telescopic vn Arstotelians wing aer Galileo's telescopic obseaons did Cf. Redondi, op cit, pp 69 I have here disregded the work of della o (De rne) and of arolycus who both ancipated Kepler in ceain respects (and are duly menoned b him) Maurolycus maes the imnt step [Photm Lumne, ansl Henry Cre, Ne York, 1, p 5 (on mrrors) nd p 74 (on lenses) of consderng only he cusp of th e causc but a connecon wth what is seen on dr ision is snot esblished or the dicules which were removed by Kepler's smple and ingenious hothesis cf Ronch, tore Lume, hapter III
100
AGANST METHOD
intersecon ofthe rays of ision from both eyes' or, in the case of monocular ision, from the two sides of the pupil This rule, which preeds from the assumpon that the image is the work of the act of vision', is partly empirical and partly eomecal t bases the posion of the image on a meical iangle' 2 or a telemeic iangle' as Ronchi calls it,4 that is consucted out of the rays whch nally arrive at the eye and is used by the eye and the mnd to place the image at the proper distance Whatever the opcal system, whatever the total path of he rays from the object to the obseer, the ule says that the mnd of the obseer ulizes its lt pa n and bases its isual judgement, the percepon, on it The rule considerably simplied the science of opcs However, t needs only a second to show that it is false take a maing glass, determine its fus, and look at an object close to it. The telemec iangle now reaches beyond the object to innity A slight change o distance brgs the Keplerian image from innity to close by and back to innty No such phenomenon is ever obseed We see the image, slightly enlarged, in a distance that is most of the me idencal with the actual distance between the object and the lens The isual dstance of the image remains constant, however much e may va the distance between lens and object and even when the image becomes distorted and, nally, diuse
50 W, II, p 72. The Ot of 1 604 has been pary nslated nto Germn b F lehn, K Gndg gch Otk, Leipzg, 1922 The relent pasges cur n secon 2 of Chapter 3, pp. 38 5 1 ibid, p. 67 52. Cum imago st vsus opus', bid, p. In visone tenet sensus comm ulom suom disanam ex assuefacone, anlos vero ad ilam disanam noat ex sensu contoonis ulom', ibd, p. 66. 53 Tranlum disanae mensorum', ibd, p. 67 54 Ot, op t, p ne should aso consult the second chapter of this k for the history of pre-Kepleran opcs. 55 bid, pp 182, 202 Ths phenomenon was known to everyone who had use magniing gass ony once, Kepler included ich show that dsregad of f phenomena ds not enal that the phenomena were seen derenty (cf te t fote to this chapter). Iaac Barrow's account of the dculty of Kepler's was menoned above (text to fote 18 to Chapter 5) crding to Bkeley ( cit, p. 141) this phenomenon enrey subve the opnion of those who wlh us judge of disnces by lines and anges. ' Bereley replaces this opinion b his theo according to which the mnd judges dsances from the clar or confusion o the prmary impressions. Kepler's idea of the telemeic iange was adopted at onc by amost al thnkers in the eld It was given a ndament son by Desca according to whom Dsanam . . . dscimus, per mutuam quandam copraone ulom' (Dt, quoted from at D Spema Phosh, Amsterd, 1657, p 87). But,' says Barrow, neitherthisnor any other diculty shall . . make renounce that which I now to be manfesty agreeable to reason.' I is this attu
NIN
101
his, then, was the actual situaon in 1610 when Galileo published his telescopic ndings How did Galieo react to it? The answer has already been given he raised the telescope to the state f a superior and better sense' 5 6 What were his reasons for doing hich ws responsible for the slow adnce of a scenc theo of e gs nd of visua opcs n gene. The reon for this peculiar phenomenon,' wites Mor von Rohr (D g tch tmt, Brn, 934 p. ), is to be sought in the close connecon beteen the eye gass nd the eye nd it is im ible to ge an accepable theoy of eye gss thout undersanding what hapns in the pess of vison itself. . . ' The telemeic iange omits preey this pess, or rather gves a simplsc and fase account of it. The sate of at the beginnng of the 2 century is well descrbed in A Gund's Appendices to a I' of Helol's reate n Physi Ot, ansl Southall, New Yok, 962 pp 26. We read here how a retu to the pcho-physoloca press of sion enabled physits to arrve at a more reoble account een of the physi of ca img: The reon hy the laws of at ca imagery have been, so to speak, summoned to life b the requirements of physioloca op is due py to the fa that b mens of igonomeical cacuaon, tedous to be sure, but esy to peform, t has been ssible for the op engineer to get closer to the reaes of his problem Thus, thans to the laours of such men as be nd hs schl, techni opcs has aned ts present splendd deelopment; hereas, th the ienc mean avable, a comprehensive gp of the inicate relaons n the c of the image n the eye been actuay imible' 56. Nichols Cicus, what a plesue t would have ben fo you to e is pa ofyour stem conrmed by so clear an experment!' wites Gleo, impng that the new telescopic phenomena are addiona sup for Copeicus (D, op cit, p 339) The derence n the aprnce ofplanets nd ed s (cf fote 26 to this chapter) he expns by the hythesis that the very nsument of seeng [the eye] noduces a hndnce of its on' (bd, p 335), nd that the telescope remoes s hindrance, tn, pemtng the eye to see the s nd the pnets as ey realy are (Mro Giuducc, a foower of Gaieo, abed idiaon to refraon b mosture on the suface of the eye, De n the C of, op ct, p 47 ) Ths explanaon, plausible as it may seem (eiay in view of Geo's aempt to show how idaon can be removed by mens other than he telesc) s not as saightfoard as one mght wh. Guand (op. ct., p. 426) ys hat ong to he propees of he wve sufae of he bundle of rays refred he eye . . . it is a mahemaca imb for any ssecon to ut he usc suface in a smh cue n he form of a crcle concenic h he pupi'. er auhors t to inhomogeneies in he vrous humours, and above n he csine len' (onch, Ot, op. ct, p. 04) Kepler giv account (Catin, op ct, pp. 33) Pont sources of ght nsmt hei cones to he cse len There recon takes place, nd behind he lens he cones an cona to a int. Bu his t ds not reach as far s he rena Therefore, he light is diped once more, nd reds er a small area of re, whereas t should mpinge on a int. Hence he telope, b inoducng anoher refracon, makes hs nt coincde h he rena . . . . ' P?lyak, in his classi wok e ina, abutes idaon party to defects of he dopca media nd to he pefe acommon' but chie' o he pecar suctur constuon of he rena il (p. 76) addg hat it may be a ncon of e bran also (p. 429). one of hese hyheses covers a he facs knon about •rradaon. Guand, Ronchi, and Poyak (if we omt reference to he bn
102
AAINT METHOD
so? This queson brngs me back to the problems raised by the evidence (against opeius) that was reported and discussed in hapter
which can be made to explain anything we want) cannot explain the dippearance of idiaon n the telescope. Kepler, Gullsand and Ronch also fail to give an account of the fact, emphaszed by Ronchi, that la obects show no diaon at their edges (Anyone undertang to account for the phenomenon of irradiaon must admit that when he ls at an elecic bulb from afar so that t seems lie a int, he sees it surrounded by an immense crown of rays whereas from nearby he sees nothing at al around it,' , op. cit, p 05) We now now that large objects are made denite by the lateral inhibitory interacon of rena elements (which s further increased by brain funcon) cf Rai, an, p 46 but the varaon ofthe phenomenon wth the dameter of the object and under the condions of telescopic vision remains unexplored Galieo's hypothesis receved sup mainly from its agreement wth the opeican point of view and was, therefore, largely .
10 n the ther hand there are sme telcc phma whch are plan pecan Gale ntrduc the phma a nnt nce r es whle the stuatn s rather that ne ruted cansm has a cean smla t phma e·ng frm anther ruted the a that telcc phma are thful ma fthe s
According to he opeican heo, Mars and Venus approach and reede from he earh by a factor of 6 or respecvely (These are appromate numbers) Their change of brighess should be 0 and 60, respecvely (hese are Galileo's values) Yet Mars hanges ve litle and he variaon in he brighess of Venus is almost impercepble' These experiences overly conadict he annual movement [of he earh] ' The telescope, on he oher hand, prduces new and sange hma, some of hem exposable as illuso by obseaon wih he naked eye, some contradicto, some having even the appearance of being illuso while the only the hat ould have brought order into his chaos, Kepler's heo of \sion is refuted by evidence of he plainest knd possibe But and i his come to what think is a cenal feature of Galileo's prcedure there are telescc hma, namely he telescopic ariaon of the brighess of he planets, whch aee mre cse wth s than the rults fnaede seatn Seen hrough he le sope Mars does indeed change as it should according to he opeian view ompared wih he tota performance of he eleope hs change is sll quite puzzling t is ust as puzzling as is e opeican heo when compared wih he pretelescopic ide nce But he change is in harmony wih he predicons of opeius t s ths hann aher han any deep understanding of
I
Th acta variaons of Mas and enus ar four maiudes and on ade especvely 2 loe, op cit., p. 328
0
104
A GA I N S T M E T H O D
cosmolo and of opcs whch r Gale pres s and the er fthe telesce in terresial we celesal matters And it s this harmony on whic he builds an enrely new iew of he universe Galileo,' wrtes Ludovico Geymonat, referrng to this aspect of the stuaon, was not the rst to tu the telescope upon the heavens, but he was the rst to grasp the enormous interest of the thngs thus seen And he understood at once that these thigs tted in perfectly with the opeican theo whereas they conadicted the old asonomy Galeo had believed for years n the uth of opeicanism, but he had never been able to demonsate it despte hs exceedingly opmsc statements to friends and colleagues [he had not even been able to remove the refung instances, as we have seen, and as he says hmsel Should direct proof[shoud even mere aet with the eidence] be at last sought here The more this conicon took root in hs mnd, the clearer to hm became the importance of the new insument n Galeo's own mind faith in the reliablty of the telescope and recoion of its importance were not tw sarate as, rather, they were tw pes fthe same prcs Can the absence of independent eidence be expressed more clearly? The Nunus, wrtes Franz Hammer in the most concse account I have read of the matter, 4 contains two unknowns, the one being solved wth the help of the other Ths is enrely correc, except that the unknows'�ere not so much unknown as known to b false, as Galieo on casions says hmself It is this rather pecular situaon, ths haony between two interesng but refuted ideas, whch Galieo exploits in order to prevent the elmnaon of either Exactly the same procedure is used to presee hs new dynacs We have seen that ths science, t, was endangered by obseable events To enate the danger Gaileo inoduces fricon and othr disturbances with the help of hc hypotheses, eang them as tendencies ned by the obious discrepancy between fact and theo rather than as physica events ned by a theo of frcon for whch new and ndependent eidence might some day becom
3 op. ct., pp 38 (my tcs) 4 J W, op. ct., Vol. , p. 7 Kepler (, op. ct., p. 4) speaks of mutuay self-supporng evidence' Remember, however,
that what s mutuly self-suprng' are two refuted hytheses and two hytheses which have in the doman ofbasc statements In a leer to Hearth of 26 March 598 Kepler speaks of the many reasons' he wants to adduce for the moon of the earth, adding that each of these reasons, taken for itsef, would nd only scant belief (asparyc,J , Vol Munich, 930 p . 68).
TEN
105
available (such a theo arose ny much later, in the 1 8th centu) et the agreement between the new dynamcs and the idea of the moon of the eath, whch Galileo increases with the help of hs method of anamns, makes both seem more reasonabl The reader wll realze that a more detaied study of hstorical phenomena such as these creates considerable dicules for the view that the ansion from th preCopcan cosmolo to that of the 7th centu consisted in the replacemnt of rfutd thoris by more gneral conjecturs whch explaned the refug instancs, made new predicons, and were corroborated by obsaons carried out to test these new prdicons And he w prhaps se the merits of a dierent view which asserts that, whe th pre opecan asonomy w n ule (was confronted by a series of refung instnces and implausibiies), the Copecan tho w n eater tule (was confrnted by even more drasc refug instances and implausibiies) but that being in haony wth stl fuher nquate the it aned sength, and was retand, the refutaons bing made inecve by hc hypotheses and clevr techniques of persuasion Ths would seem to be a much more adequate dscripon of the developments at the me of Gallo than i oered by almost all alteave accounts I shall now intupt the hstorcal narrave to show that the descripon s not only faual quate, but that it is also pe renale, and that any attempt to enforce some of the more famar methodologies of the 20th centu would have had disasous consequences
11 Such atnal meth f supp are need ecause f the un elmt Ma n fdertpas fsce cansm and ther stal nedts m sce sued n ecause ren wfrequt led n her pt
A prevalent tendency in philosophical discussions is to approach problems of knowledge su spee ettats, as it were Statements are compared with each other without regard to their histo and without considering that they might belong to dierent historical sata For example, one asks given background knowedge, inial condions, basic principles, accepted obseaons what conclu sions can we draw about a newly suggested hypothesis The answers va considerably Some say that it is possible to determine degrees of conr maon and that the hypothesis can be evaluated with their help Others reject any logic of conrmaon and judge hypotheses by their content, and by the fasicaons that have actually curred But almost eveone takes it or granted that precise obseaons clear principles and wellconrmed theories are alrea se; that they can and must be used here and nw to either eliminate the suggeste hypothesis, or to make it cceptable, or perhaps even to prove it Such a procedure makes sense only if we can assume that the elements of our knowledge the theories, the obseaons, the principles of our arguments are tmels tt which share the same degree of perfecon, are all equally accessible, and are related to each other in a way that is independent of the events that produce them This is, of course an exemely common assumpon It is taken for granted by most logicians; it underlies the familiar disncon between a ontext of discove and a context of juscaon; and it is often expressed by saying that science deals with proposons and not with statements or sentences However, the procedure overlooks that science is a complex and heterogeneous hstcalprcs which contains vague and incoherent ancipaons of futur ideologies side by sde with highly sophiscated theorecal 106
ELEEN
107
stems and ancient and peied forms of thought Some of its elements are available in the form of neatly written statements while others are submerged and become known only by conast, by comparison with new and unusual vews (Ths is the way in which the inverted tower argument helped Galileo to discover the natura interetaons hose to opeicus And this is also the way in whch Einstein discovered certain deeplyng assumpons of classical mechanics, such as the assumpon of the estence of innitely fast sials For general consideraons, cf the last paragraph of hapter 5) Many of the conicts and conadicons whch cur in science are due to this heterogeneity of the material, to this unevenness' of the historical development, as a arst would say, and they have no immediate theorecal signicance 1 They have much in common with the problems which arise when a power According to Ma secondary pa of the sa press sch as demand arsc prodcon or ega reaons may get ahead of materal prodcon and drg it along: cf. e Pe ofPhloshy bt especialy the Intun to the Ctue of Poll En, Chicago 98 p. 309: The neqal reaon between the deeopment of matera prodcon and a for insance. n genel te concepon of progress is not to be taken in the sense of the sal absacon. n the cse of a, etc., it is not so mnt and dclt to ndend hs disproon s in hat of pcticl sial relations, e.g. the relation between edcation in the U.S and Eope The really diclt int, howeer, that is to be discssed here is that of the neqal development of relaons of prodcon as legl relaons' Trotsk descrbes the same sitaon The gist ofthe aer lies n ths, that the dierent aspects ofthe hitorcal pross economcs, lcs, the sate, the groh of the working class do not develop simlaneosly long pallel lines' (The Schl of Revoltionary Sate', speech delvered at the general par membershp meeng of the Moscow Ognzon of Jly 1921 , pblished e Ft Fe Yea ofthe Communt Intal Vol , ew Yor, 1953, p 5) See also Lenin, L- Wng Communm an Innt Dorr (op cit, p. 59), conceng the fat that mlple cases of n event may be ot ophase nd have an eet only when they cr together n a derent form, the tesis ofneen deelopment' deals wth the fat that capitlsm has reached dierent sages in dierent conies, nd even in dierent pa of the me con This second of neven deeopment may lead to inverse relao between the accompang ideoloes, so that ecien in prodcon and rdical ltical ideas develop n nverse proons n civlzed Erope, wth its igly deeloped mache inds, its rch, mlform cltre and ts constitons, a nt of history has been reached when the commndng booisie, fearng the groh and increasng sengh of the proleariat, comes ot in sp ofevehingbacward, morbnd, and medieal Bt all yong sia grow a mgh demrc movement, spreading and gaining in senth' (Lenn, Bacward Erope and Advanced sia', Coled Wo Vol 19, op cit, pp 99) For ths very nteresng station, which desees to be exploited for the philosophy of science, cf A C Meyer, Lnm Cambridge, 1 957, Chapter 1 2 and L Athsser, For Ma London and New Yor, 1 970, Chapters 3 and 6 The phlosophcal bacgrond s splendidly explained in Mao Tse-tng's essay onrdn (Seleed Ren Peking, 1 970, p 70, especially seon V)
108
AGAINS
METHOD
staon is needed right next to a Gothic cathedral. Occasionally, such features are taken into account; for example, when it is asserted that physical laws (statements) and biological laws (statements) belong to dierent conceptual domains and cannot be directly compared But in most cases, and especially in the case obseaon vs theo, our methodologes proect the various elements of science and the dierent historical sata they cupy on to one and the same plane, and preed at once to render comparave udgements Ths is lke arrangng a ght between an infant and a grown man, and announcing iumphantly, what is obvious anyway, that the man is going to (the hsto of science is full of inane cricisms of this knd and so is the histo ofpsychoanalysis and of Marsm) In our examnaon ofnew hypotheses we must obvously take the historical situaon into account Le us see how ths is going to aect our udgement! The geenic hypothesis and Aristotle's theo of knowledge and percepon are well adapted to each other Percepon supports the theo of lomoon that entails the unmoved ea and it is in tu a special case of a comprehensive view of moon that includes lomoon, ncrease and decrease, qualitave alteraon, generaon and coupon This comprehensive iew denes moon as the ansion of a form from an agent to a paent which termnates when the paent possesses exactly the same form that characterzed the agent at the begnning of the interacon Percepon, accordingly, is a press in which the form of the obect perceived enters the percipient as precisely the same form that characterzed the obect so that the percipient, in a sense, assumes the properes of the obect. A theo of percepon of this kind (whch one migh rerd as a sophiscated version of naive realism) does not permt any maor discrepancy between obseaons and the things obseed That there should be things in the world which are inaccessible to man not only now, and for the me being, but in prnciple, and because of natural endowment, and which would therefore never be seen by was qute inconceivable for later anquity as well as for the Middle Ages.' Nor ds the theo encourage the use of 2 F Blumenberg, al/ Galle S Nunu Nhchtn n St Vol. , Frnfu, 1 965, p 13 Aristotle himself more open-minded The eidence (conceing ceesal phenomena) is fuished but scantily by sentions, whereas respecting perishable plants and animals we have abundant information, living as we do in ther midst DePa An , 6b26. In what follo, a higly delzed account is given oflater Astoteliansm Uness otheise stated, the word ristole' refers to dealization For the dicules in forming a coherent pcture of Arstotle hecf. rng, A tote Hedelberg, 1966 For some dierences between ristotle and h is medaeal folowers cf Wolfgang Wieland, DeA toele Psk Gtngen, 1970.
VN
10
instruments for they interfere with the presses in the medium hese presses car a true picture only as long as they are left undisturbed Disturbances create forms which are no longer idencal wih the shape of the obects perceived they create usons Such illusions can be readily demonsated by examining the images produced by cued mirrors, 3 or by crude lenses (and remember that the lenses used by Galileo were far from the level of perfecon achieved today) they are distorted, the lensimages have oloured fringes, they may appear at a place dierent from the place of the obect and so on Asonomy, physics, psycholo, epistemol o all these disciplines collaborate with the Aristotelian philosophy to create a system that is coherent, raonal and in agreement wth the results of obseaon as can be seen from an examinaon of Aristotelian philosophy in the form in which it was developed by some mediaeval philosophers. Such an analysis shows the inherent power of the Aristotelian system The role of obseaon in Aristotle is quite interesng Aristotle is an empiricis His inuncons against an overlytheorecal approach are as militant as those of the scienc' empiricists of the 17th and 1th centures But while the latter take both the uth and the ontent of empiricism for granted, Aristotle explains the nature of experience and why it is important. Experience is what a normal obseer (an obseer whose senses are in good order and who is not drunk or sleepy, etc) perceives under normal circumstances (broad daylight no nterference with the medium) and describes in an idiom that ts the facts and can be understood by all. Experience is poant r knowledge because, given normal circumstances, the percepons of the obseer contain idencally the same forms that reside in the obect Nor are these explanaons hoc. They are a direct consequence of Aristotle's general theo of moon, taken in conuncon wth the physiological idea that sensaons obey the same physical laws as does the rest of the universe. And they are conrmed by the evidence that conrms either of these two views (the estence of distorted ensimages being part of the evidence). We understand today a little better why a theo of moon and percepon which is now regarded as false could be so successful (evoluona elanaon of the adaptaon of organisms movement in media). 3 Already a plain mirror gves rise to an interesting illusion. To noce it, rst loo at yoursef n a plain mirror. You wll see your face at its normal' size. en let some steam condense on the suface of the mirror and draw the out lne of your face in the steam The ouine wll loo about half the sze of your face
110
A G A I N S T HO D
he fact remans that no decsve emprcal argument could b e rased aganst t (though t was no free from dcules) hs harmony between human percepon and the Arstotelan cosmolo s regarded as lluso by the supporters of te moon of the eath In the vew of the Copecans there es largescale presses wch nvolve vast cosmc masses and yet lee no tre n our experence he estent obseaons therefore count no longer as tests of the new basc laws that are beng proposed hey are not drectly attached to these laws, and they may be enrely dscon nected Toy aer the success of mode scence led to the belef that the relaon between man and the unverse s not as smple as s assumed by nave realsm, we can say that ths was a correct guess, that the obseer s ndeed separated from the laws of the world by the specal physcal condons of hs obseaon platform, the movng eath (graitaonal eects; law of nera Corols forces nuence of the aosphere upon opcal obseaons; aberraon stellar parallax; and so on ), by the dosyncrases of hs basc nstrument of obseaon, the human eye (rradaon; aftermages mutual nhbon of adjacent renal elements; and so on ) as well as by older vews whch have nvaded the obseaon language and made t spea the language of nave realsm (natural nteretaons) Obseaons may contan a contrbuon from the thng obseed, but ths contrbuon merges wth other eects (some of whch we have just menoned), and t may be completely oblterated by them Just consder the mage of a xed star as vewed through a telescope hs mage s dsplaced by th e eects of refracon, aberraon and, possbly, of gravtaon It contans the spectrum of the star not a t s now, but as t was some me ag (n the case of exagalacc supeovae he dfference may be mllons of years), and dstorted by Doppler eect, nteenng galacc matter, etc oreover, the extenson an the nteal structure of the mage s enrely determned by the telescope and the eyes of the obseer: t s the telescope that decdes how large the dffracon dss are gong to be, and t s the human eye that decdes how much of the sucture of these dss s gong to be seen It needs consderable sll and uch theo to solate the contrbuon of the orgnal cause, the star, and to use t for a test, but ths means that nonArstotelan cosmologes can be tested only after we have sarated obseaons and las wth the help of aula scences descrbng the complex presses that cur between the eye and the object, and the even more complex presses between the coea and the bran We must sud what we perceve to nd a core that mrrors the smulus and nothng else In the case of Copecus we need a new eteolo (n the good ol
ELEVEN
FGURE 2 . Mn, ae seven days (rst quaer).
111
112
AGAINS HO
sense of the word, as dealing with things below the moon), a new science ofphysolocl t that deals with the subjecve (mind) and the objecve (ight, medium, lenses, sucture of the eye) aspects of vsion as well as a new dynam stang the manner in which the moon of the eath might inuence the physical presses at its surface. Obseaons become relevant only the presses described by these new subjects have been inserted between the world and the eye. The language in which we express our obseaons may have to be revised as well so that the new cosmolo receives a fair chance and is not endangered by an unnoced collaboraon of sensaons and older ieas. In sum: what s needr a tt ofCs s an tre n world-v contanng a n v of man and ofhs pct ofknowng 4 4 Bacon relzed that scientic change involves a reformaon not ony of a few ideas, but of n enre world-ew and, perhaps, of the very nature of humans. For the senses are weak and errng', he tes in Num anum Aphorsm 50 For man's sense is falsel asseed to be the standard of thngs; on the conary, all the percepons, boh of the senses and of the mind bear reference to man and not to the universe, and te humn mind resembles those uneven mirrors whch impa their own propees to dierent objects from which rys are emied and disto and disre them (Aphorsm 41) Bacon repeatedly comments on the dullness, ncompten and erro of te senses' (50) and prmits tem only to udge . . . te experment' whle t s the experment that funcons as a judge of natue and the thing itself ) Thus when Bacon speaks of the unprejudiced senses' he ds not mean sense-data, or mmediate impressions, but reacons of a sense ogn that h be rebult in order to mrror nature in the right way Research demands that the tre human ng rebult This idea of a physical and mental reform o humani has relgious features A demolishing brnch' ( 1 1 5), an expatory press', a puricaon of the mind' (69 ) must precede the accumulation of nowledge. Our only hope of salvaon is to begin the whole labour of the mind agan' (Preface) ut only aer hang cleansed, lished, and levelled its surface' (1 1 5) Preconceived noons (36), opnio (42), even the most common words (59, 121 ) must be abured and renounced wth n and solemn resoluon so that the access to the ingdom of man, which is founded on the sciences, may resemble that to the ingdom of heaven, where no admission is conceded except to children' (68) A reform of man is necessary for a correct science ut it is not sucient Science, according to Bacon, not only orders events, it is also supsed to give physical reasons. Thus Ptolemy and Copeicus give us the number, situaon, moon, and perods of the stars, as a beauful outside of the heavens, whilst the esh and he enails are wanting; that is, a well fabrcated system, or the physical reasons and foundations for a ust theory, that should not only solve phenomena, as almost any ingenious theory may do, but show the substance, motions and inuences of the heavenly ies as th really are' Aant ofLeang Chapter 4 quoted from Wiley New Yok, 19, p 85 Cf. also the Num anum op. cit., p. 3 71 For let no one hope to determine the queson whether the earth or heaven revolve n the diual moon, unless he have rst comprehended the nature of sntaneous moon' the new needs a new physics n order to give substance to his asonomy Gaileo dd no succeed in proding such a physcs.
ELEVEN
113
It is obious that such a new worldiew will tae a long me appearing, and that we may never succeed to formulate it in its enrety It is exemely unlely that the idea of the moon of the eath will at nce be followed by the arrval, in full forml splendour, ofall the sciences that are now said to constute the body ofclassical physics Or, to be a little more realisc, such a sequence of events is not only exemely unlely, t s possle n pnple, given the nature of humans and the complees of the world they inhabit oday Copecus, tomorrow Helmholt this is but a Utopian dream Yet it is only aer these sciences have arrived that a test can be said to mae sense his need to wat, and to ore large masses of crical obseaons and measurements, is hardly ever discussed in our methodologies Disrerding the possibility that a new physics or a new asonomy might have to be judged by a new theo ofnowledge and might require enrely new tests, empirically inclined sciensts at once confront it with the status quo and announce iumphantly that it is not in agreement with facts and received prnciples hey are of course rght and even ivially so, but not in the sense intended by them For a an early stage of development the conadicon only indicates that the old and the new are dert and out ofphe. It does not show which view is the eer one A judgement of ths nd presupposes that the competors confront each other on equal terms How shall we preed in order to brng about such a fair comparison? he rst step is clear we must retan the new cosmolo un it has been supplemented by the necessa auia sciences e must retan it n the face of plain and unambiguous reng facts e may, of course, y to explan our acon by saying that the crcal seaons are either not relevant or that they are iluso, but we cannot support such an explanaon by a single objecve reason Whatever explanaon we give is nothing but a veal gure, a gentle invitaon to pacipate in the development of the new philosophy Nor can we reasonably remove the received theo of percepon which says that the obseaons are relevant, gives reasons for ths asseron, and is conrmed by independent evidence hus the new iew is arbiarily separated from data that supported its predecessor Scienceloving philosophers, including those who call themsees crcal', are quic to crcze thnkers who do not share their pet ideas. Bacon oen crcized for not at once flling for Copecus. He was crcized for this unspeakable crme by philosophers whose own ronalsm' would neer have alowed Copeis to live. n example is K.R. Popper, e O S a I En Vol. 2, p 1 6.
114
AGAINST METHOD
and is made more metaphysical a new period in the histo of science coences wth a ard t that retus us to an earlier stage where theores were more vague and had smaller emprcal content his bacward movement is not just an accident it has a dente ncon it is essenal ifwe want to overtae the status quo, for it gves us the me and the freedom that are needed for developing the main view in detal, and for nding the necess aula sciences 5 hs bacward movement is indeed essenal but how can we persuade people to follow our lead? How can we lure them away from a welldened, sophiscated and empirically successful syste and mae them ansfer their allegiance to an unnished and absur hypothesis? o a hypothesis, moreer, that is conadicted by one obseaon aer another if we only tae the ouble to compare it with what is planly shown to be the case by our senses? How can we convince them that the success ofthe status quo is only apparent and bound to be shown as such in years or more, when there not a single argument on our side (and remember that the illusaons I used two paragraphs earlier derive their force from the successes o classical physics and were not available to the Copeicans 6 It clear that allegiance to the new ideas w have to be brought about b means other than arguments It wll have to be brought about b atonal eans such as propaganda, emoon, hoc hypotheses, and appeal to prejudices of all nds We need these irraonal means in order to uphold what is nothng but a blind faith unl we have foud the aula sciences, the facts, the arguments that tu the faith into sound nowledge It is in ths context that the rise of a new secular class with a new outl and considerable contempt for the science of the schls, i methods, its results, even for its language, becomes so importt he barbaric Lan spoen by the scholars, the intellectual squalor o academic science, its otherworldliness whch is sn itereted as uselessness, its connecon with the Church all these elemen ts are 5 example of a bacward movement of ths kind is Galleo's retu to the kinemacs of te ommtaolus and hs dsregard for the machinery of epcles developed n te De Rol For an admrble ratnal account of ths step cf I Lakatos and E Zar, y Dd Copeicus' Research Progrmme Supeede Ptolemy's?', n mre Latos Phshl P Vol I, Cambrdge 198 6 They were available to the scepcs, especally to Aenesdemus, who nts ou folowng Phlo, that no object appears as it is but s moded by beng combned th ar, lgh humd, heat, etc cf D Lus IX, 84 Howeer, seems that scepl ew had ony le nuence on the development of mode sonoy, d undersandably : one ds not st a movement by beng reonable
ELEVEN
115
now led together wth the Aristotelian cosmolo and the contempt cont empt one feels ee ls for them is ansferre ansferredd to eve single Aristotelian argument. This guitbyassiaon ds not make the arguments less ratonal or less conclusive, ut t reduc ther uce on the mnds of those who are wiling wiling to follo follow w Cope Co pecus cus.. For Copecus Cope cus now stands stand s fo for progre progress ss in other areas are as as well, he is i s a symbol symbo l fo for the ideals of a new class c lass that tha t lks back to the classic cla ssical al mes of Plato and Cicero and foard to a free and pluralisc siety. The assiaon of asonomcal ideas and historical and class tendences does not produce produce new arguments arguments either. But it engenders a rm commient to the the helienc helienc vie view w an andd is all that is needed neede d at this stage, as we have seen We have also seen how masterfully Galileo exploits the situaon and how he amplies it by icks, jokes, and non-sequtu of hs own. We are hee dealng with a situaon that must be analysed and understood f we want to adopt a more reasonable attude towards the issue between reason' and irraonality' Reason grants that the ideas which we inoduce in order to expand and to mprove our 7 For these sal sal pressures pressures cf cf Olschks magn magn cent Ghte nsprh /ch ssch/ch Lttur For the role of Puriansm cf RF Jones op cit Chapte V and 8 In a remarkable b G Hetc, Prnceton 1987 (rs publshed n Itaian n 1 982) 982) Peo Redond has described described the grou groups ps both both nsde te Church Church (and ncudng the Pope hmse) and outside of it who led favourbly un new scenc developments the vews on percepon connu maer and moon that had been expaned by Galeo Gal eo n hs ss� among them Being n drect drect conct wth o f the Eucharist the most mnt sacrment these iews the tradional account of were consderby more dangerous than Copeicanism and could tolerted only as ong as the gups and the Pope hmsef had the upper hand in the complex lica deveopments deveopments of the me (hirty (hir ty Years War French and Spanish cs cs the French Frenc h aance wth the Pope) he lcal reversal of the Popes fortunes the accuons of eniency towards herecs that were rased aganst him on lca grds cast a shadow on hs attude towards scienc maers as wel (here t he seemed to sup heresy) and made protecve measures necessary Redond ies to show (a) that the physics of the me was connected wth theologcal dines such as the ine of the Eucharist and that a history of science that neglects the connecon ecomes incomprehensible and ) that the atttude towards scienc problems cau caused y the connecon and thus the atttude towards innovaon changed wth the polical climate he second pa of ) may well ue ut there is only wea evdence to sup the rest what Galileo ys about atomism in the ss� is much t brief and inde i nden nite ite to conict wth anssubsanation anssubsanation (it is an aside almost not an elaborate elaborate satement) and wth the excepon excepon of a rther problemac du dumen mentt no such such of G, Vatican conict was pereived (Cf RS Westfall Ess�s the T of bseatory Publcaons 1989 pp 84) Wat is valuable in Redonds account s that he wdens the domain of ssible inuences and thus undermines the anachnstic) belef that then as now scientic raonali was resied to the nte nteal problem situaon of a scentic dscplne
116
A G A I N S M E H OD OD
nowledge may ase in a ve disorderly way and that the on of a parcular point of view may depend on class prejudice, passion personal idiosyncrasies, quesons of style, and even on error, pure and simple But it also demands that in judng such ideas we follow certain welldened rules our aluaton of ideas must not be invaded by irraonal elements Now, what our historical examples seem to show is this there are situaons when our most liberal j udgements and our ou r most liberal rules would have have eliminated a point point ofview whch we regard today as essenal esse nal for science, sci ence, and would not hae hae permtted permtted it to prevail prevail - and such situaons situaons cur quite frequently he ideas suved and they now are said to be in agreement with reason hey suved because prejudice, passion, conceit, errors, sheer pigheadedness, in short because all the elements that characterize the context of discove, osed osed the dictates dictates of reason and an d euse the aton atonal al el el ts we petted petted to he their way o express it dierently Cns and other rat raton onal al'' vs vs ist toy toy on euse re reon w eled at soe soe te te in their p (he opposite is also ue witchcraft and other irraonal irrao nal vews have ceed to be inuenal only because reason was overruled at some me n their past) 9 Now, Now, assuming that Copecanism is a Good hing, hing, we must also admit that ts suval is a Good hng And, considerng the condions of its suval, we must futher admit that it was a Go hing that reason was overruled in the th, th and even the th centurie cent uries s oreover, the cosmologists of the the th an andd th centuries centu ries did not have the nowledge we have today, they did not now that Copecani Cope canism sm was capable ca pable of giving giving rise rise to a scienc scien c system system that is acceptable from the the point of view ofs of sci cien en cc method meth od hey did d id not now whch of the many views that ested at their me would lead to ture reason when defended in an irraonal way Being wthout such guidance guid ance they had to mae mae a guess and in maing ths guess they could only follow their inclinaons, as we have seen Hence it advisable to let le t ones inclnaons go against reason n any any rs rstan tan, , for it maes lif l ifee less consained and science may prot from from it 9 hese consderaons rete rete J Dorlng who n Bth Joualr oualr the Phlos ofS, Vol 2 1972 189f presents my onalsm as a presupon of m research not a result result He connues connues one woul wouldd have have thought thought at phlosopher of cience would be most nterested in pckng out and analysing n de those scenc arments arm ents whch did seem to be raonay reconsucble reconsucble One woul woul have thought that the phlosopher of scence would be most nterested n pckng ou and analysing n deta det a those moves whch are necesry for the at of sence Such moves I have ed to show oen resst raonal raonal reconsucon
LVEN
117
t is cear that this argument, that advises us not to let reason overrule our inclinaons and casionally to suspend reason altogether, does not depend on the historical material which I have presented If my account of Galileo is historically corrct, then the argument stands as formulated. If it tus out to be a faitale, then this faitale tels us that a conict beeen reason and the precondions precondions of progress is possle, it indicates how it might arise, and it forces us to conclude that our chances to progress y be obsucted by our desire to be raonal. And note that progress progress is here he re dened as a raonalisc over of science would dene it, i.e as entailing that Copecus is better than Aristotle and Einstein better than Newton Of course, there is no need to accept ths denion, which is ceainly quite narrow I use it only to show that an idea of reason accepted ac cepted by the the majorit maj orityy of raonalists raonal ists may prevent progress as dened by the the ve same same majority I now resume the discussion of some so me details deta ils of the th e ansion from from Aristotle Aristotle to Cope C opeicus icus cosmolo , I have said, is a step The rst step on the way to a new cosmolo, c apparently relevant evidence is pushed aside, new data are brought in by hoc connecons, conneco ns, the the empiric empirical al content content of science is drascally reduced. Now the cosmolo that happens to be at the cene of attenon and whose adopon causes us to car out the changes changes just jus t descr des crbed bed diers diers from from other views views in one respect res pect only on ly it has features which at the me in queson seem attracve to some people. But there is hardly any idea that is totally without merit and that might not also become the starng point of concenated eort. No invenon is ever made in isolaon, and no idea is, therefore, completely without (abstract or empirical) support Now if paral support and paral plausi plausibili bility ty suce to start a new end - and I have suggested that they they do - if stang a new end means taking a step back from from the evidenc e vidence, e, if any idea can become plausible and can receive paral support, then the step bac is in fact a sep foard, and away from the tyranny of ghtlyknit, hghly corroborated, and racelessly presented theorecal systems. Another dierent error', rites Bacon on precisely this point, 10 is the . perempto reducon of o f knowledge into arts and m methods ethods,, fm whch me the sciences are seldom improved; for as young men rarely grow in stature after their shape and limbs are fully formed, so knowledge, hilst it lies in aphorisms and obseaons, remains in a growing edon) New Yor Yor 1 9 p 2 1 Cf Cf also also he he Aanct ofLeang ( 1 5 edon) 0 Aanct
Num anum, Aphor Aphorsms sms 79 86 as well well as W WN N Wans Wa ns splendd le le obb obb Syst Syst of, f, London 1965 p 169
118
A G A I N S M E HO HO D
state; but wen once fasioned into metods, toug it may be futer polised, illusated and tted for use, is no longer increased in bulk and substance' Te similrty wit te arts wic as often been asserted arises at exactly tis tis point poin t Once it as been realized realize d tat a close empirical empiri cal t is no virtue and tat it must be reled in mes of cange, ten style, elence of expression, simplici simplicity ty of o f presentaon, tension of plot plot and narrave, and seducveness of content become important features of our knowledge Tey give life to wat is said and elp us to overcome te resistance of te obseaonal material 1 1 Tey eate and maintain interest in a teo tat as been partly removed from te obseaonal plane and would be inferior to its rivals wen judged by te customa standards It is in tis context tat muc of alileo's work sould be seen Tis work as oen been likened to paan 1 2 and propaganda it certainly is But propanda of ts kind is not a marginal aair aair tat tat surrounds alleg a llegedl edlyy more more substanal substan al means of defence, and tat sould peraps be avoided by te professionaly onest scienst' In te circumstances we are considering now, paan paan s of the the sc scee It is of te essence because interest must be created at a me wen te usual metodological prescripons ave no point of attack; and becaue tis interest must be maintaied, peraps for centuries, unl new reasons arive It is also clear tat suc reasons, ie te appropriate auia sciences, need not at once tu up in full formal splendo Tey may at rst be quite inarculate, and may even conict wit te esng evidence Agreement, or paral agreement, wit te cosmolo is al tat is needed in te beginning Te agreement sows tat tey are at least relant and tat tey may some day produce full-edged posive evidence Tus te idea tat e telescope sows so ws te world world as itit really is leads to many many dicule dic uless But te support support it lends to, and receives rec eives from, from, Copecus Cope cus s a nt tat we migt be moving m oving in te rig rigtt direcon We ave ere an exemely interesng relaon beeen a general view view and te te parcular ypot ypotese esess wic constute constute its evidence evid ence It i often assumed tat general views do not mean muc unless te relevant evidence can be fully specied Caap, for example, asserts tat tere is no independent interetaon for [te language in terms of wc a certain teo or world-view is formulated] Te system [te aoms of te teo and te rules of derivaon] is itself I I . at resttes to scenc phenomenon ts lfe is a (The D of Nn, Vol I p 277) 1 2 Cf A Koyr Koyr Et Et Ga/ Ga/n, n, Vol III Paris 199 pp 5
ELEVE
119
an unintereted postulate system [Its] terms obtain obtain ony a n indirect and incomplete interetaon by the fact that some of them are connected by correspondence rules with obseaona terms 3 here is no n o independent nterp nterpret retaon, aon, says says Caap and a nd yet an idea such as the dea of the moon of the earth, whch was inconsistent with with the conempora evdence, whch was upheld uph eld by declaring declar ing hs evdence to be irrelevant and which was therefore cut from the most important facts of contempora asonomy, managed to become a nucleus, a cstallizaon point for the aggregaon of other inadequate views which gradually increased in arculaon and naly fused into into a new cosmolo including including new inds of o f evid eviden ence ce here is no better account of hs press p ress than th an the descripon desc ripon whch ohn Stuart Stuart ill has left left us of the vicissitudes vicissitudes of hs educaon educa on Referrng Referrng to the explanaons whch his father gave h on logica matters he writes he explanaons did not mae the matter at at all clear to me at the me but they were not therefore useless they remaned as a nucleus for my obseaons and reecons to cstallize upon the import of hs general remars being intereted to me, by the parcular instances whch came under my noce aar 4 In exactly the same manner the Copeican view, though devoid of coive content co ntent from the point of view of a sict empiricism empirici sm or else el se refuted, was needed in the consucon of the supplementa sciences even before it became testable wth their help and even before it, in tu, provded them with supporng evidence of the most most forcel orcel in ind d here is a further element in this tapesy of moves, inuences, beliefs beliefs whch is rather interesng inter esng and and whch received rec eived atenon only recently recently - the the role of paonage oday most most researchers gan a reputaon, a sala and a pension by being assiated with a anor a research laborato hs involves cerain university anor condions such as an abiity to wor in teams, a wlliness to subordinate ones one s idea i deass to to those of a team leader, l eader, a harmony between betw een ones ways of doing science and those of the rest of the prof p rofession ession,, a cer cerai ainn style style a way way of of presenng the the evidence - and so on on Not eveone ts condions such as these able people remain unemployed because b ecause they fal fal to sas some of them Conversely the reputaon of a university or a research laborato rises with the
he Methodoogca Character of heoreca Concepts', Mnno Mn no Stud Stud the Phlos ofSce, Vo I Mnneas, p 47 1 4 Autobo, quoted from Esstl Wo ofJohn Stua M, ed Leer,
New York, l 96 S , p . 2 1 .
10
AGAIS MEHOD
reputaon of its members. In Galileo's me paonage played a similar role. There were certain ways of gaining a paon and of keeping him The paon in tu rose in esmaon only if he succeeded o attract and to keep individuals of outstnding achievement. According to Westfall, the Church permitted the publicaon of Galileo's Daloe in the full knowledge of the conoversia matters contained in it [n]ot least because a Pope [Urban VIII] who gloried in his reputaon as a Maecenas, was unwilling to place it n jeopardy by saying no to the light of his mes', and Galieo fell because he violated his side of the rules of paonage. Considering all these elements, the Rise of the Copecan WorldView' becomes a complicated matter indeed. Accepted methodological rules are put aside because of sial requirements (paons need to be persuaded by means more eecve than argument), insuments are used to redene eperience nstead of being tested by it, local results are eapolated into space despite reasons to the cona, analogies abound - and yet all this tus out in reospec, to have been the correct way of circumvenng the rescons implied by the human condion. This is the materia that should be used to get better insight into te comple press of knowledge acquision and improvement. To sum up the content of the last ve chapters: When the Pythagorean dea' ofthe moon of the eath was revived by Copeicus it met with dicules which eceeded the dicules encountered by contempora Ptolemaic asonomy. Sctly speak ing, one had to regard it as refuted. Galieo, who was convinced of the uth of the Copecan view and who did not share the quite common, though by no means universal beliefin a stable eperience, lked for new kinds of fact which might support Copeicus and sl be acceptable to all. Such facts he obtained in two dierent ways First, by the invenon of his telce, which changed the sso core of eveday eperience and replaced it by puzzling and uneplained phenomena; and by his pnple ofrelat and hs na, which changed its conctual coponts Neither the telescopic phenomen a 15 op cit, p 7 1 6 Further detais on these maers in Chapter 8, fote 12 ofthe present esy , Westfa, op cit, and M Biagio, Ga// Cou M Fincharo, Ga// and the ofReonng, Dordrecht, 1980, has commented on Gaeo's use of rhetorc, whe M Pera and W R Shea (eds), Pung Sce The ofStc Rhetoc, 1991, d especilly Marceo Pera, Sce and Rhoc, forthcoming, comment on scenc rhetoric n gener
LV
11
nor the new deas of moon were acceptable to common sense (or to the Arstotelans) Besdes, the assated theores could be easly shown to b false. et these false theores, these unacceptable phenomena were transformed by Galleo and converted nto strong support of Copecus The whole rch reseor of the eveday eerence and of the ntuon of hs readers s ulzed n the argument, but the facts whch they are nvted to recall are arranged n a new way, apprmaons are made, known eects are omtted, derent conceptual lnes are drawn, so that a n n ofece arses, anufure almost out of thn ar. Ths new experence s then sole by nsnuang that the reader has been famlar wth t all the me It s solded and soon accepted as gospel uth, despte the fact that ts conceptual components are vastly more speculave than are the conceptual components of common sense Followng posvsc usage we may therefore say that Galleo's scence rests on an ustrate etaphys The dstoron permts Galleo to advance, but t prevents almost eveone else frm makng hs eort the bass of a crcal phlosophy (for a long me emphass was put ether on hs mathemacs, or on hs alleged experments, or on hs frequent appeal to the uth', and hs propagandsc moves were altogether neglected) I suggest that what Galleo dd was to let refuted theores support each other, that he bult n ths way a new worldvew whch was only loosely (f at all!) connected wth the precedng cosmolo (eveday experence ncluded), that he establshed fake connecons th the perceptual elements of ths cosmolo whch are only now beng replaced by genune theores (physologcal opcs, theo of connua), and that whenever possble he replaced old facts by a new e of experence whch he smply nte for the puose of supporng Copecus Remember, ncdentally, that Galleo's procedure drascally reduces the content of dynamcs: Arstotelan ynamcs was a general theo of change comprsng lomoon, qualtave change, generaon and corrupon. Galleo's dynamcs an ts successors deal wth locomoon only, other knds of moon eng pushed asde wth the promsso note (due to Democrtos) that locomoon wll eventually be capable of comprehendng a oon. Thus, a comprehensve emprcal theo of moon s replaced by a much narrower theo plus a metaphyscs of moon, ust as an emprcal' experence s replaced by an experence that contans speculave elements Ths I suggest was the actual procedure followed by Galleo Preedng n ths way he exhbted a sle, a sense of humour, an elasc and elegance, and an awareness of the valuable weaknesses of human thnkng, whch has never been equalled n the hsto of scence Here s an almost nexhausble
1
A G A I S M E H OD
source of material for methodological speculaon and much more importantly, for the recove of those features of knowledge which not only inform, but which also delight us. 1 7
17 A few years ago Marn Gardner, the pitbu of sciensm, pubished an arce wth the e n-Scence, the Strange Case of Pau Feyerabend' Ctl Inqu Winter 1988 he vaant ghter seems to have overked tese and other pasges I am not against scence I praise ts foremost praconers and (next chpter suggest that ther predures be adopted by phiosophers Wat I object to is now mnded phlosophca inteference and a narrow-mnded extenson of the att scenc fashons to a areas of human endeavour n sho what I object to raonasc nteretaon and defence of science
Galeo ethod wor n otherel a we For aple t can e used to elnate the stn ats aanst ateals and to put an d to the philosophical nody prol he condn scienc pls ran untouched hower It notow that t should e ne al appled
Galileo made progress by changing familiar connecons between words and words (he inoduced new concepts), words and impressions (he inoduced new natural interetaons), b using new and unfamiliar principles such as his law of inera and his principle of universal relavity, and b altering the senso core of his obseaon statements. His move was the wish to accommodate the Copeican point of view Copeicanism clashes with some obvious facts, it is nconsistent with plausible, and apparently well established, principles, and it does not t n with the grammar' of a commonl spoken idiom It does not t in with the form of life' that contains these facts princples and grammacal ules But neither the rules, nor the princples, nor even the facts are sacrosanct. The fault ma lie with them and not with the idea that the ea moves. We ma therefore change them, create new facts and new grammacal rules, and see what happens once these rules are avaiable and have ecome famliar Such an attempt ma take considerable me, and in a sense the Galilean venture is not nished even toda But we can alrea see that the changes were wise ones to make and that it would ave een foolish to sck with the Aristotelian form of lfe to the exclusion of everythng else. ith the mnbod problem, the situaon is exactl the same We ave again obseaons, concepts, general principles and gram acal rules which, taken together, constute a form of life' that aarentl supports some views, such as dualism, and excludes oers, such as materialism (I sa apparentl' for the situaon is c less cear here than it was in the astronomical case) And we a aain proceed in the Galiean manner, look for new natural
1
14
AGAI S MEHOD
interetaons, new facts, new grammacal rules, new principles wic can accommodate materialism and ten compare te total systems materialism and te new facts, rules, natural intereta ons, and pinciples on te one side; dualism and te old forms of life' on te oter Tus tere is no need to y, like Smart, to sow tat materiaism is compable wit te ideolo of common sense Nor is te suggested procedure as desperate' (Armsong) as it must appear to tose wo are unfamiliar wit conceptual cange Te predure was commonplace in anquity and it occurs werever imaginave researcers ske out in new direcons (Einstein and Bor are recent examples) 1 So far te argument was purel intellectual I ed to sow tat neiter logic nor experience an limit speculaon and tat outstanding researcers often ansgressed widel accepted limits But conceps ave not onl a logical content; te also ave assiaons, tey give rise to emoons, tey are connected wit images Tese assiaons, emoons and images are essenal fo te way in wic we relate to our fellow uman beings Removg tem or canging tem in a fundamental way may peraps make ou concepts more objecve', but it often violates important sial consaints It was for tis reason tat Aristotle refused to abandon an intuive view of uman beings simply because a more pysiologic approac sowed successes in a limited domain For im a person was a sial enty and dened by is or er funcon i te city no matter wat atomists or pysicians involved in teo migt say Similarly te Roman Curc, being interested in souls and not o in asonomical icks, forbade alieo to present is badly founded guesses as uts and punised im wen e violated te proibion Te al of alieo raises important quesons about te roe products of specialists, suc as absact knowledge, are supposed to play in siety It is for tis reason tat I sall now give a brief accout of ts event
For a more detaed dscusson the reader s referred to my Vo , Chapters 9 and 0
3 e Church at the te ofGaeo not on t coser to reon ned th and n now t ao cored the ethcal and soal conseuc ofGalleo s Its ndt ofGalleo w ratonal and n uns and a lc ofpepee can and a rson
Ther were man trials in the 1 7th centu The proceedings started either with accusaons made private pares, with an ocial act a public ocer, or with an inqui based on somemes rather vague supicions. Depending on the locaon, the disbuon of jurisdic on and the balance of power at a parcular me, crimes might be examined b secular courts such as the courts of kings or of free cies, b Church courts, such as the spiritual courts attached to eve episcopate, or b the special courts of the Inquision. After the iddle of the 12th centu the episcopal courts were greatl aided b th stud of Roman law. Laers became so inuenal that, even if wholl unained in canon law and theolo, the had a much better chance of high preferment than a theologian The inquisitorial process removed safeguards provided b Roman law and led to some wellpublicized excesses What has not been publicized to the same extent s that the excesses of roal or secular courts ofen matched os of the Inqusion It was a harsh and cruel age. B 600 the I or ths compant (made b Roger Bacon) cf H Ch Lea, A Hto ofthe nision ofhe Mle Ag, Vo I, p 09 Chapters x expan the detais of the qustora procedure, the ways n which they dered from other predures and the resos for the derence Cf aso GG Couton, Inition and Li, Boston, 9, Chapters XIXV 2 Chares Henry Lea, the great iber historan, wrtes On the whoe we may concud that the secret prsons of the Inquson were ess intoerabe paces of abode than th pscopa and pubc gaos The genera icy respecng hem was more uan and enghtened than that of other jursdcons, whether in Span or elsewhere, although neglgent supeson alowed of abuses and there were ample resources of rgor n resee, when the obsnacy of the mpentent ws to be broken do.' Hito ofthe Inquition in Spain, Vol. 2, New York, 1906, p. 34. Prsoners accused before secular cous csonally commed crmes under the jursdcon of
125
126
AGA NS MEH OD
Inquision ad lost muc of its power and aggressiveness Tis wa ue especialy in Italy and more parcularly in Venice 3 Te cours of te Inquision also examined and punised crimes conceing e producon and te use of knowledge T can be explained by teir orign: tey were supposed to exterminate , ie complexes consisng of acons, assumpons and talk making people inclined towards certain beliefs Te surised reader wo asks wat knowledge as to do wit te law sould remember t many legal, sial and nancial obstacles knowledgeclaims face today alileo wanted s ideas to replace te esng cosmolo, but e was forbidden to work towards tat aim Today te muc more modest wis of creaonists to ave teir views taugt in scls side by side wit oter and compen views runs into laws setng up a separaon of Curc and State Increasing amounts of teorec and engneering informaon are kept secret for milita reasons and are tereby cut o from inteaonal excange 5 Commerial interests ave te same resicve tendency Tus te dscove o superconducvity in ceamcs at (relavely) ig temperae, wc was te result of inteaonal colaboraon, sn led to protecve measures by te American goveent 6 Financ aagemens can make or break a researc progrme and an en profesion Tere are ay ways to sience people ap fro fobidding tem to speak and of tem are being ud today press of knowledge producon and knowledge disbuon w
the Church so hat hey mght be handed over to he nquson: Henry Kamen, Sane I"un, Munch, 1 980, p. 1 7 3 n 1356 e secular ocas of Vence fobade he nqustor ofreso to h prsoners, seed hs nformants and tortured hem on he cha of pferng te porty of he accused Lea, "utn n the Mg, Vol. , p 273 4. A comprehee re of one of he as hat resuted from he con been publshed n S, Vol. 2 15, 1982, pp. 934. Many oher s foowed. 5 t seems hat he need for secrecy n nuclear maers was rst sed b te sensts hemselves Cf he re and he duments n Spencer R. Wea n Geude Wess-Szlard (eds), L Szr H Vn ofthe F, Cambrdge, M 1978, esp. Chapters 2. Cf aso he matera on he Oppenhemer c. he ven of he telescope was foed to secrecy as he mlary mnce of he connce sn reaed. Cf. Chapter 8, fote 24. Research teams come very secreve when approachng what hy hnk s a g scovery Ae what sat stae are patents, consulances n ndus, mon n pes, he honour of a Nobel ze. For a speca case cf RM S, London, 1988 he manpulaon of nowledge b he co used, wh many exmples, by Peter W Huber, Gal Rge, New Yo 1991 6 S, Vol 237, 1987, pp. 476 nd 593f An mn step t exclueness coted g pa of he resech to he mary.
HREEN
127
never the free, obecve', and purely intellectual exchange raonalists make it out to be The ial of Galieo was one of many ials It had no special features except perhaps that Galieo was eated rather mildly, despite his lies and attempts at decepon 7 But a small clique of intellectuals aided by scandalhung writers succeeded in blowing it up to enormous dimensions so that what was basically an altercaon between an expert and an instuon defending a wider vew of thgs now looks almost like a battle between heaven and hell Ths is chldish and also ve unfair towards the many other vicms of 1 7th centu jusce It is especially unfair towards Giordano Bruno, who was bued but whom sciencaly mnded intellectuals prefer to forget It is not a conce for humanty but rather party interests whch play a major role in the Galleo hagiography Let us therefore take a closer lk at the matter The socaled ial of Galleo consisted of two separate preedings, or ials The rst curred in 1616 The Copecan doctrine was examned and cricized Galieo received an order, but he was not punshed The second ial tk place in 1 6333 Here the Copecan dtrine was no longer the point at issue Rather, what was considered was the queson of whether Galileo had obeyed the order given h n the rst ial, or whether he had deceived the inquisitors into believing that the order had never been issued e preedings of both ials were published by Antono Favaro in Vol 19 of the Naonal Edion of Galiean material The suggeson, rather popular in the 19th centu, that the proceedings contained 7. An example is Galileo's reply to the nquires of 12 Aprl 1633 Maurce A. Fincharo, e Gali ir, Berkeley and Los Angeles, 1989, p. 262, the rst two lnes he reacon of an admrer s charactersic hs absurd pretence . . .'
Geymonat, op. cit., p. 1 49.
8. t cannot be dened that pressure grous, persona grevances, envy, the fact that Gaeo, beng t infatuated wth his own genus' was unsuerble' estl, op. ct., 52, 38) and the rules of paonage played an mnt role s they, or simlar crcumstances, do at every al. However, the tensons between varous groups of the Church on the one sde nd the demands for scenc autonomy on the other were real enough aer al, ther mode successors (should the scences be ven the run of our educaona instuons and of sie as a whole or should they be ated like any oher specal nterest group?) are sll wth us. Here the Church dd the rght thng the scences do not have the last word n humane maers, knowledge ncluded. he man duments penng to the ial were assembled and anslated wth comments and an noducon by Fincharo, op ct. shall use hs anslaons. ccounts of the ias and ther problems are found in G. de Sanllana, The Cme of Galil, Chcago, 1954, Geymonat, op. ct., Redond, op. ct., and, most recenty, in estfal.
128
AGA S MEHOD
falsied documents and that the second ial was therefore a farce, seems no longer acceptable The rst ial was preceded by denunciaons and rumours, in whch gree and envy played a part, as in many other ials The Inquision started to examine the matter Experts (qualcator) were ordered to give an opinion about two statements whch contained a more or less correct account of the Copeican docne Their decision conceed two points what would today be caled the stc contt of the docne and its ethil soa implications On the rst point the experts declared the docine to be flih and absurd in philosophy' or, to use mode terms, they declared it to be unscienc This judgement was made without reference to the faith, or to Church dne, but was based exclusvely on the scienc situaon of the me It was shared by many outstanding sciensts (Tycho Brahe having been one of them) and it a coe 2 when based on the facts, the theories and the standard of 9. One of the authors ofthe suggeson was the Galleo scholar Emil Wohw Hs reasons, rather mpressve at the me are given in hs D nqutonsps G G, Berln, 870. Accordng to Wohlwll two duments of the preedngs, date 25 Feb 1616 and 26 Feb 1616 (Fnchao, op ct., pp. 147 ae mutu conadctory. The rst adises Galleo to eat Copecus as a mathemaca mode should he ree the advce, then he s forbidden to menon Copecus n any form whatsver. n the second dument Gaileo s adsed as above and mmedatey forbdden (.e. wthout waing for hs reacon) to menon Copecus Wohl thought the second dument to be a forgery. hs seems now efuted. f e Sanllana, Chapter 13. Sllman Drke (appendx to Geymonat) deised a very ning hyhess to explan the dscrepancy I 0. Some crcs used idiosyncrses n the formulaon as prf o a lack of com prehension on te pa of the expe. But there was no need for the inqusitors to stick closely to the lanage of the authors they examned. heir account of Copeicnim was clear enough wthout such textual puritanism. I I Fnchiaro, op. cit., p 16 1 2. Note that in rendering my judgements rely on standards subscrbed to b many mode sciensts and philosophers of scence. Retued to the eary 7 century, these champons of raonali would have judged Galeo as te Astote judged him then. Michelson, for example, would have been aghast at Gaileo's ae to get nowledge out of an insument as lile understood s the telescope n Rutheford, who was never t happy about the theory of rela, would v produced one of his charctersc rude remars. Savador Lura, an outstd mcrobologist who favours theores decdable by clear-cut experment step[] would have relegated the debate to outelds of scence' lie siolo' and wo have stayed away from it Slot Mhne Bk Tt Tube, New York, 985, pp 1 9). For what Gaileo suggested was no less than to regard as ue a teory whch only naogies n ts favour and which suered from numerous dcues. And made ths suggeson in public while even today it i a deady sn for a sien address the public before haing consulted his peers (example in A Pickeri
HREEN
12
the me Compared with those facts, theories and standards the idea of the moon of the eath was as absurd as were Velkovs's ideas en compared with the facts, theories and standards of the fes A mode scienst really has no choice n this matter He cannot clng to his own ve sict standards and at the same me paise Galleo for defending Copeicus He must either agree with the rst part of the udgement of the Churc experts, or admt that standards, facts and laws never decide a case and that an unfounded, opaque and ncoerent dine can be presented as a fundamenta uth. Only few admrers of Galieo have an inking of ths rather complex stuaon Te stuaon becomes even more complex when we consider that the Copeicans changed not only views, but also standards for udgng views Aristotelians, in this respect not at all unlike mode epidemologsts, molecular biologists and empirical' siologists who nsist eiher on the examnaon oflarge stascal samples or on clearcut expermental steps' in Luria's sense, demaded song emprical support whle the Galeans were content with far reachng, unsupported and paraly refuted theories 13 I do not cricze them for that on the cona, I favour Nels Bohr's ths is not crazy enough' I merely want to reveal the conadicon in the acons of those who praise Galieo and condemn the Church, but become as sict as the Church was at Galleo's me when tug to the work of their contemporaries On the second point, the sial (ethical) implicaons, the experts declared the Copecan docine to be formally herecal' Ths means it conadicted Holy Scripture as intereted by the Church, and it did so in full awareness of the situaon, not inadvertently (that would be material' heresy) Consants on Conoversy the Case of the Magnec Monole', Soal St of , Vol 1 1 , 1 98 1, pp 63) All this s reazed neither by pogresse' (.e sciencally inclned) prnces of the Church nor by scensts, so the discussion of the a of Galileo curs in a dream world wth only lle relaon to the real world we and Galleo nhabt Further arments on that int are found in Chapter 9 of Farewell to Reon and Chapter 19 below 13 As ndicated n Chapter 8, fote I , Galileo's law of inera was n conict h the Copeican as well as the Kepleran eaent of planetary moon Galeo hoped for future accommodaons. hat was a sensible thing to do but not in greement wth some standards of hs me and of today oday a similar clash between heorecans and emprcists curs in the eld of epdemiolo here are theorecal reson to expet that -rays and other forms of parculate radaon constute a cncer-rsk down to the smallest dose Many epdemologists demand emprca prf, hoever, though t s clear that events, when currng below a cen threshold of frequency, cannot be deteted in that way
130
AG AIN S MEHOD
The second point rests on a seres of assumpons, among them the assumpon that Scripture is an important bounda condion of human estence and, therefore, of research The assumpon was shared by all great sciensts, Copecus, Kepler and Newton among them According to Newton knowledge lows from two sources - the word of God - the Bible - and the works of God - Nature and he postulated divine inteenons in the planeta system, as we have seen 4 The Roman Church in addion claimed to possess the exclusive rights of exploring, intereng and applying Holy Scripture Lay people, according to the teachng of the Church, had neither the knowledge nor the authority to tamper with Scripture and they were forbidden to do so Ths comment, whose rigidity was a result of the new Tridenne Spirit, should not surise anyone familiar with the habits of powerful instuons The attude of the American Medcal Assocaon towards lay praconers is as rigid as the attude of the Church was towards lay intereters - and it has the blessing of he law Experts, or ioramuses having acquired the formal insiia of experse, always ied and often succeeded in securng for themselves exclusive rights in special domains Any cricsm of the rigidity of the Roman Church applies also to i mode scenc and scenceconnected successors Tuing now from the form and the adminisave bacng of the objecon to its content we noce that it deals with a subject that ining increasing importance in our own mes - the quality o human estence Heresy, dened in a wide sense, meant a deviaon from acons, attudes and ideas that guarantee a wellrounded and sanced lie Such a deviaon might be, and occasionally was, encouraged by scienc research Hence, it became necessa to examine the herecal implicaons of scienc developments Two ideas are contained in this attude First, it is assumed that the quality oflife can be dened ndependently of science, that it may clash with demands which sciensts regard as natural ingredients o their acvity, and that science must be changed according Secondly, it is assumed that Holy Scripture as intereted by th Holy Roman Church adumbrates a correct account of a well rounded and sanced life 14 Chapter 5, fote 4. See aso he literature in fote 6 to Chapter Accordng to Galeo (leer to Grand Duchess Chrisna) he idea of he two sour gs ack to ertullan Maron (E. Evans, ed.), , 18 1 5. For he exact wording see DenzingerSchnmeer, Enhon 36h edion, Freburg, 1976, pp. 365f
HREE
131
The secod assumpo ca be rejected without deyig that the Bible is vastly richer i lessos for humaity tha aythig that might ever come out of the scieces Sciec results ad the sciec ethos (if thee is such a thig are simply t a foudao for a life woth livig May sciests agree with ths judgemet They agree that the quality of life ca be deed idepedetly of sciece - which is the rst part of the rst assumpo At the me of Galieo there ested a istuo - the Roma Church - watcig over this quality i its ow parcular way We must coclude that the secod poit - Copeicus beig formaly herecal' - was coected with ideas that are urgetly eeded today The Church as o the right ack But was it perhaps mistake i rejecg sciec opiios icosistet with its idea of a Go Life? I Chapter 3 I argued that kowledge eeds a plurality of ideas, that wellestablished theories 16 hus Konrad Lorenz in hs interesng if somewhat superc bk De ht Tonn Zl Mshhet, Munch, 1984 (rst publshed n 973), p 70
wites he erroneous belef hat ony what can be ronay sped or even what can be proved in a scenc way constutes he sold knowledge of manind has dious consequences t prompts he sciencly enlightened" younger generon to dscard the mense esures of knowledge and wsom that are contined n the adons of every ancient culture and in he teachings of he reat world relgions Whoever hinks hat l his s whout signicance naturaly succumbs to anoher, equally peicious mstae, living in he concon hat science is able, as a maer of course, to create from nohing, and n a rona way, n enre culure wh l ts inredents' n a simiar veinj eedham, iniator and paauhor of a reat hstory of Chnese science and technolo, speas of scienc opum', meanng by it a blindness to he suerng of ohers' Tme the Rhng R, otngham, 1986. Raonalism', rtes Peter Medawar to a Young Stt, ew o, 1979, p I 01 ), fas sho of answering he many smple and chdlie queso people e to ask quesons about orgins and purses such as are oen contemptuously dismissed as nonquesons, or pseudoquesons, ahough people undersand hem clearly enough and long to have an answer hese are ntellecal pains hat ronaists - lke ad physcians confronted by alments hey cannot dagnose or cue - are apt to dsmss as mainaon"' he clearest and most percepve satement s found nJacques Monod, Chan and Nes, ew ork, 1972, p 170 (te in brckets from p 169): Cod and autere,' rites Monod, prong no explanaon but msng an ascec renunciaon of all oher spirtua fare, [he dea hat obecve knowledge is he ony auhenc source of uh ] was not of a ind to alay ane but aggvated it nstead By a singe soke t claied to sweep away he adions of hundreds of hounds of years, which had becoe one wh humn nature itself t rote an end to he ancient anmist convenant between man and nature, leaving nohing in place of hat precous bond but n anous ques in a frozen unerse of solitude Wih nohing to recommend t but a cein puran arrogance, how could such n idea wn accepance? t did not t s has not t has however conded recognion but hat it dd ony because of its prigious wer of peformance '
132
AGANS MEHOD
are never song enough to termnate the estence of alteave approaches, and that a defence of such alteaves, being almost the only way of dscoverng the errors of highly respected and comprehensive points of view, is required even by a narrow phlosophy such as empiricism Now if t should tu out that t is also required on ethcal grounds, then we have two reasons instead of one rather than a conlict with science' Besides, the Church, and by this I mean its most outstanding spokesmen, was much more modest than that It did not say what conadicts he Bible as intereted by us must go, no matter how song the scienc reasons in its favour A uth supported by scienc reasonng was not pushed aside It was used to revise th e interetaon of Bible passages apparently inconsisent with it There are many Bible passages which seem to suggest a lat eath Yet Church dtrine accepted the spherical eath a matter of course On the other hand the Church was not ready to change just because somebody had produced some vague guesses It wantd pof scienc prf in scienc matters. Here it acted no dierently from mode scienc instuons unverses, sch and even research instutes in various counies usually wait a long me before they incoorate new ideas into ther curricua rofessor Stanley Goldberg has descrbed the situaon n the c of the special theo of relavity) But there was as yet no convincing prf of the Copecan dtrine Hence Galleo was advised to teach Copecus a hypoths; he was forbidden to teach it a tth Ths disncon has surived unl today But while the Church was prepared to admt that some theories mght be ue and even that Copecus' mght be ue, given sufcient evidence 17 there are 17 n a wdely discussed leer which Cardna Robeo Bellarmino, master of conoversial quesons at the Collegio Romano, rote on 12 Apri 1615 to Pao Antonio Foscani, a Carmelite monk from Naples who had inqured about the re of the Copeican system, we nd the followng passage (Fncharo, op ct, p ) ifthere were a ue demonsaon that the sun is at the center ofthe world nd th earth in the thrd heaven, and that the sun ds not circle the earth but he earth cr the sun, then one would have to preed wth great care in explaining the Scriptu that appear conary, and � rath that not untand th than that w ontrated e But wll not beleve that there is such a demonsaon, un it shown me Nor is it the same to demonsate that by supsing the sun to be at th center and the earth in heaven one can save the appearances, and to demonsate tht in uth the sun is at the center and the earth in heaven; for beleve the demonsaon may be available, ut have very great doubts about the second, and case of doubt we must not abandon the Holy Scripture as ntereed by the o Fathers' n his Conraton n the Cn Onn, Finchiaro, op cit, pp. 7
HREEN
133
now many sciensts, especially in high ener physcs, who view a theores as insuments of predicon and reject uthalk as being metaphysical and speculave Their reason is that the devices they use are so obviously desined for calculang purposes and tha heorecal approaches so clearly depend on consideraons of elegance and easy applicability tha the generalizaon seems o make ood sense. Besides, the formal properes ofappromaons' ofen dier from those of the basic principles, many theories are rs seps owards a new poin ofview which a some future me may yield them as appromaons and a direc inference from theo o reality is therefore rather naive } Al hs was known o 16th and 1 7th centu sciensts. Only a few asonomers though of deferents and epicycles as real roads in the s mos rerded them as roads on paper which migh aid calculaon bu which had no couneart in reality. The Copeican poin of view was widely inereed in the same way - as an ineresng, novel and rather ecien model. he Church reqused, both for scienc and for ethcal reasons, tha alileo accep hs ineretaon. Considering the dicules the model faced when rerded as a descripon of reality, we mus admi tha l]ogic was on the side of . . Bellarmine and no on the side of alileo,' as the hisorian of science and physical chemis Pierre Duhem wroe in an ineresng essay. 1 9 To sum up: the judgemen of the Church experts was sciencaly correc and had the righ sial inenon, . o proec people from the machinaons of specialists. I waned o proec people from being corruped by a narrow ideolo tha migh work in resiced domains bu was incapable of sustaining a harmonious life. A revision of the judgemen migh win the Church some friends among sciensts bu would severely impair its funcon as a preseer of importan human and superhuman values. 2 esp pp 85f, Galeo addresses precisely these ints He agrees that ifhe Copeican asonomers are not more han nine percent right, hey may be dismssed' but adds ha f al hat is produced by phiosophers and asonomers on he opsite sde is shown to be mosy fase and wholy nconsequena, hen he oher sde should not be dsparaged, nor deemed paradoca, so o hin hat it could never be clearly pred' research should be permed even if demonsaons are not yet avaable Ths does not conct wh Bellarmino's suggesons it did conit and to a cein een sll ds conict wh he attude of many mode research instuon 18 More on his int in ancy Cawright, How the La oPs Le, ford, 1983 19 To S the Pha, Chicago, 1 963, p 78 20 After some apparent wllingness to consider he maer (cf he address of Pope John Paul II on he centena of Enten's birth, publshed an Epoe in Paul Cardinal Poupard (ed), Ga Gal Towar a Rolutn o Yea oDebate,
134
AGAIST METHOD
Pisburg, 198 Cardina oseph Rainger, who holds a sion similar to that once held by Bellarmine, formulated the problem in a way that would mae a reision ofthe judgement anachronisc and intess Cf. his t in Paa of Mah 19, p reed in I Sabato, 31 March 1990, pp 80 wiesses the Cardinal quoed Est Blh (being merely a maer of convenience the scienc choice beeen geenism and helienism cannot overrule the prccal and religious cenici of the earth), CF von Weicker (Gaileo leads direcy to the atom bomb) and yself (the chapter heading of the present chapter) I commented on the speech in o inteiews, /Sabato, 1 2 May 1 9, pp 54 and La Rubbl, 14 July 19, p 20
4 aeo s nques ed on a sma pa ofthe so-caled Ccan Roluton Ang the ranng elts mak t st more dlt to renle the elmt wth mlar pnpl oftheo aluation.
alileo was no the only sciens involved in the reform of physics, astonomy and cosmolo. Neither did he deal with the whole area of astonomy. For example, he never studied the moon of the planets in as much detail as did Copeicus and Kepler and he probably never read the more echnical parts of Copeicus' great work. Tha was no unusual. Then as now knowledge was subdivided ino specialies an expert in one eld rarely was also an expert in another and disan eld. And then as now sciensts with widely diverg ing philosophies could and did commen on new suggesons and developments. Tycho Brahe was an outstanding astonomer obseaons conbued o the downfall of generally aceped iews. He noced the importance of Copeicus' cosmolo ye he reained the unmoved ea on physical as well as on theological grounds. Cpeicus was a faithful Chrisan and a gd Arisoelian he ied o resore cented circular moon o the prominence i once had, postulaed a moving eah, rearranged the planea orbits and ve absolue values for their diameers. The astonomers surrounding Melanchthon and his educaonal reform acceped and praised the rs part of his achievemen bu (with a single ecepon - Rhecus) either disregarded, or criized, or rein ereed (Osiander!) the second And they ofen ied o tansfer ope icus' mathemacal models o the Polemaic sysem. aestlin, Kepler's eacher, rerded comets as solid bodies and ied calculang the orbi of one of them. His (incorrec) resul made h ccep the Copeican arrangemen of the planeta orbits (i sll
I Detals and literature n RS Wesan, The Melanchthon Crle, Rhecus, and the Wenberg Interetaon of the Coeican Theory', Is, Vol 66, 1975, 16 13
136
A G I S M E T HO D
inluenced Kepler). Maestlin respeced Arisotle bu rerded mathemacal correcess and harmony as sis of physical ruth. alileo's approach had is own idiosyncrasies, i was more complex, more conjetural, partly adaped o the greaer role theological consideraons played in Ialy, parly deermined by the laws of rheoric or patonage. Many dieren personalies, professions and groups guided by dieren beliefs and subjeced o dieren constaints conibued o the prcess tha is now being described, somewha summarily, as the Copeican Revoluon'. As I said a the beginning this process was no a simple thing bu consised of developments in a variety of subecs, among them the following cosmolo; physics; asronomy; the calculaon of astonomcal ables; opcs; episemolo; and theolo. I draw these disncons no in order o be precise' bu because they relec actually esng subdivisions of research. Physics, for example, was a general theo of moon tha described change withou reference o the circumsances under which i curred. I comprised lomoon, the growth of plants and animals a well a the tansion of knowledge from a wise eacher o an ioran pupil. Arisotle's Phys and the many mediaeval commentaries on i give us an idea of the problems eaed and the soluons proposed. Cosmolo described the sucture of the universe and the speial moons that are found in i. A basic law of phys in the sense jus explained was tha a moon withou moor comes o a sandsll - the natural' situaon of a body is res (this includes lack of qualiave change). The natural' moons of cosmolo were those tha occurred withou noceable inerference; eamples are the upward moon of re and the downward moon of sones. Arisotle's On the He and the many mediaeval commenaries on i give us an dea of the problems and the view discussed in this domain The books I jus menoned were for advanced studies only. Inroduco exts omied problems and aleave suggesons and concentaed on the bare bones of the ideas then held One of the mos popular intoduco exts of cosmolo, Sacrobosco's sp, conained a skech of the world, and described the main Qheres withou giving the deails of their moons - the res is silence. 2 Sll, 2 Cf ynn Thodyke (ed), The phe o aoso a Its Commtat , Chcago, 1949 The elements and ther motons are briey mentoned in the rs chapter together wth a smple argument n favour of the unmoved earth the ea rth situated in the centre (ths is shown earler by optica argments, including the fa a the constellatons have the same se, no maer where the daily roaton puts the) and quicqud a medio movetur versus crumferenam scendit Terra a edio movetur, ergo ascendit, quod pro impossbile relinqutur' (p Equant, deferen
FORTE
137
i as used as a basis for rather advanced crcal commens down o alileo's own me Pysics and cosmolo claimed o make ue staemens. Theolo ich also clamed o make ue saemens was regarded as a bouda condon for research in these elds though the sength of is requiremen and ofits instuonal backingvared n me and with e locaon. I was never a necessa bounda condion for asonomy whch deal with the moons of the stars, bu withou caing uh for its models Asronomers ere ineresed in models a migh correspond o the actual arrangemen of the planets, bu ey were no resiced o them Handbooks of asronomy such as Polemy's handbook and the varous popularaons based on i conained deailed asonomical models preceded by skechy cosmologcal inoducons. As far as these nroducons were conceed, there esed only one cosmolo- Arsotle's Some ofthe adbooks also contained ables. Tables were a fuher sep away from reality' They no only used hypotheses', ie models tha gh no relec the sucture of reality, they also used appromaons. Bu an asonomer's appromaons did no always correspond o the ecellence of his models Advanced' (from our stadpoin) models migh be combined with crude appromaons and thus gve worse ables than er older counearts 3 The separaon beteen physcs and cosmolo on he one side an asonomy on the other was no only a praccal fac i also had a philosophical backing. According o Arsotle4 mathemacs oes o deal with real things bu conains absracons. There ess therefore an essenal dierence beteen physical subjects such as psics, cosmolo, biolo and psycholo and mathemacal bjecs such as opcs and asronomy. In e encyclopaedias of the arly Mddle Ages the separaon was a maer of course Opical etbooks only rarely deal with asonomical maers 5 Asrnomy used basic opcal laws such as the law of linear prpagaon, bu the more complicaed parts of opcal theo were and epcycle are menoned n he fourth chapter togeher wh he miraculou nature of he solar eclipse accompanying Chrst's deah 3 The example of PtolemyCopeicus s eated by Sanley E Babb Jr in Is, ol 68, September 1 977, especially p 432 4 M, k i, Chapter 2 Ps, k ii, Chapter 2 For an account and efence of Arsotle's heory of mahemacs c Chapter 8 of my Farewe to Ren As an example I menon John Pecham's opcs (quoted from David Lindberg (ed), John Pham and the S o Ot, Madison, 1970) asonomical maers cur here on pages 53 (mn illusion nd he northward displacement of he sun and he xed sars, explained by vaurs near he horzon), 2 (scinllaon of he
138
A G A I S T M E TH O D
no well own he same is ue of episemolo alileos argumens (and the arguments of Copeicus on which they are based) brough episemolo bac ino science (the same happened many years laer, in connecon with the quantum theo) Now is i o be expeced tha a collecon of relavely ndependen subects, research staegies, arguments and opnons such as the one us menoned will develop in a uniform way? Can we realy assume tha all the physicists, cosmologists, theologians and philosophers who reaced o the Copecan dine were guided by the same moves and reasons and tha these reasons were no only acceped by them, bu were also rerded as being binding for any scien enering the scene? he ideas of an individual sciens such as Einsein may show a certain coherence 6 and this coherence may be releced in his standards and his theorizing Coherence is o be expeced in otalitarian surroundings tha guide research either by laws, by peer pressure or by nancial macnaons Bu the astonomers a the me of Copeicus and aer did no live in suh surroundings; they lived a a me of dissension, wars and genera upheaval, a a me when one city (Venice, for example, and the ies under its ursdicon) would be safe for a progressive sciens while another (such as Rome, or Florence) oered considerable dangers and when the ideas of a single individual ofen faced groups of sciensts not in agreemen with his monomania o show this, le us l a two astonomers who parcipaed in the developmen Copecus himself and aestlin, Keplers eacher Copeus waned o reform astonomy He explained misgivings and the ways in which he ied o overome them e wroe 7 sars explained by unevennesses of their rotang sufaces which reect the sun) 218 (imsibili to determine the sie of the sars from their appearances), 233 (sars appear to be smaer than they actually are), 225 (they are displaced tows the north at the hoion, and the more so, the greater their disance from te merdian) 6 The case of Einstein shows that een this modest assumpon gs much t ar Einstein recommended a lse oprtunism as the best research sate (c te quoaon in the te t to fote 6 of the Inoducon) and he waed hat a good joe (such as the consderaons leading to the specal theory of relai) shoud no repeated t often Philipp Frank, Etn H L and Tm, ondon, 19, p 26 7 Commolu, ed E Rosen, ree Cn Treat, 3rd edion, N York, 1971, anslaon pay changed in accordance wth F Kra, Cope Reoversus ', Coloqua Cna III and , Preedings of the oint Sym of the IA and the IHPS, Torun, 1973, p 1 1 9 In what foows I shall Kra, Copeicus Reoversus II', J ct
139
FORTEE
Te planeta teores of Ptolemas and most oter asonomers seemed to present no small dulty For tese teores were not adequate unless ertan equants were also oneved t ten appeared tat a plaet moved wt unfo velty neter alog ts own deferent nor relave to an atual ene Havng eome aware of tese defets I often onsdered weter tere ould peraps e found a more reasonale arrangement of rles from w eve apparent nequalty ould e derved and n w eveg would move unformy aout ts proper ene as te le of aomplsed moon requres
v
The crique of Copecus coces he followig model ha was used for calculag he logiudes of Mars, Jupier ad Sau. The plae P moves o a small circle, he epicycle, whose cete is locaed o a larger circle, he defere. The cee of he epicycle preeds wih cosa agular veloiy wih respec o E, he equa poi. The plae is obseed from he eah T. E ad T are o opposie sides of he cee of he defere, havig he same disace from i. Copeicus does o queso he empirical adequacy of he odel O he coa, he admis ha he plaea heories of he olemaeas ad ohers are cosise wih he umerical daa'. 8 or does he believe ha hese daa are i eed of correco. Isead of ioducig ew obseaos he emphasizes ha 8 Ren, op cit p
140
AGAIST METHOD
we must follow in teir [te ancient Greeks'] ftsteps and hold fast to ter obseaons bequeated to us lke an nertance. And if anyone on te cona tnks tat te ancients are untrustworty n tis regard, suely te gates of ts art are closed to hm9
Neither new obseaons nor the inability of Polemy o ake care of wha was known o him are the reason for Copeicus discomfort. The diculty he perceives lies elsewhere. In his accoun Copeicus disnguishes between absolue moons and apparen moons. The second inequality of planeta moon, i.e the fac tha a plane may run ahead in its path and then reverse its direcon, is apparen' - i mus be reduced o other moons. According o Copeicus these other moons are moons on centred crcles with a constan angular velocity aroud the centre. Polemy vioaes the condion he uses equants. Equants expla apparen moons no by ue moons bu again by apparen moons where the plane moves with uniform velity neither along its own deferen, nor relave to an actua centre . . .' For Copeicus (and for many other astronomers) real moon is a circular moon around a centre with consan angular velity. Copecus removes excentre and equan and replaces them by epicycles. In the Polemaic sceme each plane has now three epicycles the old epicycle and two fuher epicycles for replacing the ecenic and the equan. In order o avoid this accumulaon of epicycles (which occasion ally pushed the planets far ou ino space) Copeicus lks for a dieren explanaon of the second inequality. He is helped by the fac tha the second inequality agrees with the posion of the sun. I can therefore be inereed as an apparen moon creaed by a real (and, of course, circular) moon on part of the eah. The argumen as reconsuced so far (afer Kra) contais o 9. Letganst W, n Rosen, op. ct., p. 99 0 Erasmus Renhold rote on the e page of hs pesona copy o
Rutn xma stnomm Motus et qua/ t r a qua/ et r mptus Quoted from Wesan, The Melancho
Cle', op ct, p 176 . Ths s ue of the Commo/us In hs man wor he agan uses exce deferents Only the equnt repaced by an epcyce. Ths beron fom the e (Erasmus Renhod), aso n lunar theory, greay mpressed some of Cope admrers who pad no aenon to hs ne cosmolo and the moon of the Wesn, The Melanchthon Crcle, op ct., pp. 17, 177. 12. The mean sun n Copecus. Kepler eects a reducon to the ue sun d thereby senens the Copecan aangement.
FORTE
141
elemens; a purely formal elemen and a reality asseron. Formally i is requesed tha any periodic moon be reduced o cenred circular moons. The reques is conneced with the assumpon tha inequalies re apparen while circular moons alone ar real. Le us call this the t rea sumpton. Bu Copeicus also discovered tha his procedure allowed him o incorporae eve planea path ino a sysem, conaining the large circle', the circle of the eah, as an absolue measure. Al these phenomena', Copecus wries in his main work, are conneced with one another in a mos noble way, as if by a golden chain, and each plane with its posion and order is a wiess tha the eah moves while we, who live on the erresrial globe, failing o recoize its moon, ascribe all sorts of moons o the planets.' I is this inner connecedness of all parts of the planea sysem tha convinced Copeicus of the reality of the moon of the eah. I call this the sed rea sumpton. The rs reality assumpon was part of the Plaonic tadion; Arisotle gave i a physical basis The second reality assumpon conliced with Arisoelian physics and cosmolo. Arisotle had already cricized an earlier (Pythagorean) version ofi: mathemacal harmonies, which are abstacons, relec uth only if they agree with wellconrmed physical principles. This is a reasonable reques; i was used in our own centu o rejec Schroedingers ineretaon ofwave mechanics. I is reasonable especially for those thinkers who rerd mathemacs as an aulia science tha may describe bu canno constue physical presses. I is unreasonable for a Plaonis or a Pythagorean. The resulng clash between two ineretaons of the nature of mathemacal saements played an importan role in the Copeican Revoluon'. Copeicus stengthened the second reality assumpon by referring o tadions such as the Hermec tadion and the idea of the exceponal role of the sun 4 and by showing how i could be reconciled with the phenomena. He made two assumpons. Firs, tha the moon of a body is appropriae o its shape. The eah is spherical, hence its moon mus be circular. Secondly, objects such as sone say with the ody (the eah) from which they were 3. De R/, Preface to Pope Paul Kra assumes that Copecus dscovered ths armony n the course of hs aempts to remove the equant and ony later tued t no a fundamental argment in favour of a real moon of the earth. 4 The prase and n the mddle stands the sun' was not new In the older asonomy the sun was indeed n the mddle of the planets, wth Mars, Jupter and Sa standng above t, and Venus, Mercury and the mn below It aso ruled' the lanets n the sense that ts moon was mrrored n the moons ofal planets (the mn exeted). Cf. eg. Macrobus, omnum pon
14
AGAIST METHOD
separaed ence te falling sone stays close o te ower According o Arisotle te natural moon of obects, ie te upward moon of re and te downward moon of sones, was deermined by te sucture of space (central syey Accordig o Coper nicus i is deermined by te disbuon of matterCopecus saves penomena' suc as te free fall of eavy bodies bu provides neiter independn arguments nor sic laws ta could lead o a detailed comparison His predure is hoc. Tis does no mea ta i is bad i only means ta i canno be reconciled wit te leading metodologies of oday My second example is Micael Maestlin, Kepler's eacer Maes was an expert asonomer and is udgement was generaly respected. He only reluctantly abandoned' te Ptolemaic dibu on of te speres but e was forced to do so by ccumstance beyond is conol. 1 far as we can see, te ccumstances were, rst, te nova of 1572 Maesti obseed it, measued its pa and put it beyond te spere of te mn into te spere of te ed stars. e rst part (beyond te mn folowed for Maesti ro te missing paral, te second part (ed stars from te abence of any proper moon. According to Copeicus, woe ideas Maeti used at ts oint, a planet moves more slowly te greater its dtance from te sun. Obseg te canges of colour and brige Maesti (ad Tyco wo saw te new star on te way to alcemical aborato inferred tat te region above te mn cannot be witout cange, as Aristotle ad assumed. However, i would be ra to conclude tat Maestlin (and Tyco rerded te nova as a blow ainst te peripatec pilosopy' . 1 6 Many Curc people, Teodore Beza among tem, rerded te penomenon as a retu of te star of Betleem, i.e. as a supeatural event. 17 Tyco tougt tis comparison too modest; ere, e said, is te greate miracle since te beginning of te world, comparable at least to Josua's stopping of te sun. 1 8 Tis means tat as far as Tyco w conceed mracles refuted te idea of te autono of te laws of 1 5 In hat foUow I am using the disseon by RA.JarreU, eLadS W fthe Tng Astnm Mhl Mtln, Toronto, 1972, eU R Wesn, Mihae Maesin's Adopon of the Copeian Theory', C Cna , Ossiineum, 1 975, p. 53 1 6 JarreU, op ct., p 08. 17. Cf. the iterture in P H. Kher, S and Relgn n Elhan En New or 969, pp 1 74f, fotes 1 2 nd 1 3 Cf. ao ol. , Chapter of Thoie, t fMa and l S, New o, 1 941 18. , p 548
ORTEEN
143
nature (which as an Aristotelian idea), they did not refute spec laws Maestin, on the other hand, being perhaps more scepcal bout miracles, may indeed have regarded the case as a blow against' Aristotle The next queson is how serious a blow it was for him The idea of permanent heaven was part of cosmolo and contained the special hypothesis of a fth element The falsehood of ths hypothesis impaid neither the remaining laws of moon nor the tower argument Both Clavius and Tycho accepted a changing heaven but sll used the tower argument to exclude the moon ofthe eath If Maestlin's doubts reached futher then this was due either to an idiosyncrac interetaon of the Aristotelian dtines, or to personal inclinaons towards a nonAristotelian worldview It seems that we must assume the latter The next decisive event on Maestlin's jouey towards Copecus was the comet of 1 577 Ain Maestlin, prompted by numerous obseaons', puts the comet into the superlunar region 20 The idea that this region is free from change has now denitely been dropped Maestlin also ed to determine the ajecto of the comet He found it to be moving in the path of Venus as described in Bk 6, hapter 12 of Rolutionius Somewhat hesitangly he now accepts the Copecan orderng of the spheres But, so he adds, e was forced to do so by exeme necessity' 22 This exeme necessity' arises only when geomecal considera ons are given the force of cosmological arguments Many years later alieo cauoned against this way of reasoning rainbows, he said, cannot be caught by angulaon Maestlin had no such doubts He accepted the adional disncon between physics and asonomy and idened asonomy with mathemacs opecus wrote his enre book not as a physicist, but as an asonomer' is his comment on the margn of his copy of Rolutionius 23 He then intereted the results of mathemacal arguments by using the second reality ssumpon This means that he did not ercome an Aristotelian sstance ainst such an interetaon, he acted as if such a sistance did not ist This argument', he wrote in hs marginal otes, 24 is wholly in accord with reason Such is the arrangement of 9 or Claus cf his commentary on Sacrobosco's sphere, 1593 edion, pp 0. C also Westfall, op. it, p 0. Jarrell op. cit, p 1 2 2 Ibid , p. 1 7
. Iid , p 20 3 Wesan, op ct, p 59 4 Iid
144
AGAINST METHOD
ths enre, mmense mhina that t permts surer demonsaons, ndeed, the enre unverse revolves n such a way that nothng can be ansposed wthout confuson of ts parts] and, hence, by means of these surer demonstraons] all the phenomena of mon can be demonsated most exactly, for nothng untng curs n the course of ther orbts' Kepler too became a Copecan because of ths harmony and because of the comet, the nteresng fact beng that Maestln' s calculaons of the path of the comet contan serous mstakes t dd not move n the orbt of Venus Now let us compare these events and the stuaons n whch they cur with some once popular phlosophes of scence We noce at once that none of these phlosophes consders all the dscplnes that conbuted to the debate Asonomy s n the cene A raonal reconsucon of the developments n ths area s thought to be a raonal reconsucon of the Copecan Revoluon tself The role of physcs (the tower argument), the fact that theolo occasonaly formed a song bounda condon (cf Tycho's reacon to hs nova and to the dea of the moon of the eath) and the role of derent mathemacal phlosophes shows that ths cannot possbly be ue Ths fatal ncompleteness s the rst and most fundamental objecon aganst all reconsucons that have been oered They sll depend on the (posvsc) pjudce that obseaons alone decde a case and that they can judge a theo all by themselves, wthout any help (or hndrance) from alteaves, metaphyscal alteaves ncluded Moreover, they even fal n the narrow doman they have chosen or reconsucon, asonomy To show ths, let us consder the following accounts 1 Nae pism the Mddle Ages read the Bble and never looked at the s Then people suddenly looked upwards and found that the world was dent from the opnon of the schools Ths account has dsappead from asonomy - but ts analogue surves n other areas (for example, n some parts of the hsto o medcne) The man argument anst t s that Arstotle was an arch emprcst and that Ptolemy used carefully collected data 2 2 Shisticated pism new obseaons forced asonome to mod an already emprcal docne 25. Carefuy' has been contested b R.R Newton, e Cme fCudu Pt Bamore, 1977 Newton show that many of Ptolemy's data' we manufud to his model. or his opcs ths has been nown for a long me.
ORTEEN
145
This certainly is not ue for Copeicus and his followers in the 16th centu As we have seen, Copecus thought the Ptolemaic stem to be pil aquate - he cricized it for theoretical ons Ad his obseaons' are essenally those of Ptolemy, as he says himself Mode comparisons of Copecan and Ptolemaic predicons with the facts', ie with 19th and 20thcentu calculaons, show, futhermore that empirical predicons were not iproved and actually become worse when the compeng systems are rescted to the same number of parameters 2 The only new obseaons made were those ofTycho Brahe - but they already led beyond Copeicus to Kepler Galieo's obsea ons belong to cosmolo, not to asonomy They lend plausibilty to some of Copecus' analo A compeing proof of the moon of the eath did not emerge, however, for the Galilean obseaons could also be accommodated by the Tychonian system 3 Facationism new obseaons refuted important assump ons of the old asonomy and led to the invenon of a new one This is not correct for Copeicus and the doman of asonomy (see bove, coments on 2) The refutaon' of the immutabiity of the heavens was neither compellig nor decisive for the prblem of the moon of the eath Besides, the idea of the moon of the eath was in big ouble or, if you will, reted' It could surve only if it was eated with kindness But if it could be eated with kindness, then so could the older system We see here ve clearly how misguided it is to y reducing the process Copecan Revoluon' to a single principle, such as the principle of falsicaon Falsicaons played a role just as new bseaons played a role But both were imbedded i a complex patte of events which contained tendencies, attudes, and cnsiderons of an enrely dierent nature
otionalism: the old asonomy became more and more
complicated - so it was in the end replaced by a simpler theo It is is ssumpon that led to the mocking remark of the epicyclical
2 Stanley E. Ba, Accuracy of Planetary Theores, Parcularly for Mars', Is,
Sep 1 977, pp 426 Cf aso the earlier arcle of Derek de Sola Prce, Cona Copeicus', n M Clage (ed.), Ctl Pb fthe t fSce, Madison, 1959, pp. 1 97; NR. Hanson, Is, No. 51 , 1 90, pp 1 50 eU Owen Gingerich, Cri v ethec in the Copeican Revoluon', n Beer (ed), Vt n Astn, ol. 17, 1974 Gingerich compares the l of Stoeer, Stadu, Maesn, Magin and Origanu and nds of them beset b errors of roughy the same magntude (ough not of te ame dsbuon aong the eclipc).
146
A G A I N S T M E T HO D
degeneraon' The theo overlooks the fact that the Copeican scheme has about as many circles as the Ptolemaic one 27
The theo ofs asonomy was in a crisis The crisis led to a
revoluon which brought about the iumph of the Copecan system The answer here is the same as under 2 pl here was no crisis and no crisis was resolved A crisis did cur in cosmolo, but only the idea of the moon of the eath received a seriou hearing The many complaints about the inexacess of asonomical pdicons that pceded Copeicus (Regiomontanus, for example cricized the lack of precise inial condions and accurate tables, nt basic theo, and such a cricism wold have been quite unjust, as the later exanaon of these theories shows 28
27 The reader should consult the very nstrucve dagrms n de San edon of Gaileo' s , Chcago, 1 964 28. Cf. fote 26 above.
5 e rults otaned so r sut aolhng the dtnon we a contt of de and a contt of justcaton no and fas oseatonal ts and theoretl tes None ofthe dtnons pys a le n stc pra Ap to r th wou he dstus consequc Per's tl' ratonalm fa r the same reo
Let us now use the material of the preceding secons to throw light on the following features of contempora empiricism (1) the disncons between a context of discove and a context of uscaon - no and facts, obseaonal terms and theorecal terms (2) Popper's crical' raonalism (3) the problem of incommensurability The last problem will lead us back to the problem of aonality and order vs anarchism, which is the topic of this essay One of the obecons which may be raised ainst my attempt to draw methodological conclusions from historical eamples is that it confounds two contet which are essenally disct a context of discove, and a context of uscaon Dsce may be iraonal nd need not follow any recoized method Justton, on the other hand, or - to use the Holy Word of a dierent school - tsm, starts only aer the discoveries have been made, and it proceeds an orderly way It is one thig,' writes Herbert Feigl, to reace the historical orgins, the psychological genesis and development, the sociopolicaleconomic condions for the acceptance reecon f scienc theories and it is quite nother thing to provide a logical reconsucon of the conceptual sucture and of the tesg of scienc theories' These are indeed two dierent thn, especially s they are done by two dierent dpln (histo of science, philosophy of science), which are quite ealous of their indepen dence But the queson is not what discons ferle mind can e Orodox Vew of eores, n RadnerWinour (eds), s f
and Meh fPs Pch, Minnea, 970, p 4
17
148
AGA INST METHOD
dream up when confronted with a complex press, or how some omogeneous material may be subdivided; the queson is to what extent the discon drawn reects a real dience and whether science can advance wthout a song nteracon between the separated domains. (A river may be subdivided by naonal boundaries but this does not make it a disconuous enty) Now there is, of course, a ve noceable dierence between the rule of tesng as 'reconsucted by phosophers of science and the predures which sciensts use in actual research T dierence apparent to the most supercial examnaon. On the other han, a most supercial examiaon also shows that a determied applica on of the methods of cricism and proof which are said to belong o the context of juscaon would wpe out science as we know i and would never have permitted it to arise.2 Conversely, the fa that science ests proves that these methods were frequently over ruled. They were overruled by predures which belong to the context of discove Thus the attempt 'to reace the torc origns, the psychological genesis and development the so policaleconomc condions for the acceptance or rejecon of scienc theories, far from being irrelevant for the standards ofte actualy leads to a cricism of these standards - prid the domans, hstorcal research and discussion of test predure, are not kept apat by at. n another paper Feigl repeats his arguments and add o ther points. He 'astonished that . . . cholars uch as NR Hanon, Thoas K, Mchael Polanyi, Paul Feyeraben Sigund Kh et a/, conider the dcon as invaid or at le mseading3 And he points out that neither the psycholo o invenon nor any siarty, however great, between the sciences an the arts can show that it does not est n t he ce r Even the most sursing stores about the ner in whch scien aive at their theories cannot exclude the possibility that th preed in a enrely dierent way once they have foun them th possiili realed Inveng theores and contempla them in a reed and 'arsc fashion, sciensts often e moves that are forbidden by methodological es For eample, th interet the evidence so that it ts ther fanciful ides, ete dicules by hoc predures, push them aside, or simply rese to take them seriously e acves which according to Feigl belon 2 Cf. the exmpes in Chpter 5 3. mprcsm t By', MS, 1 972, p. 2
ITEEN
149
the context of discove theefoe, not just dert fom what phlosophes say about juscaon, th are in with it Scienc pcce does not conta o contexts moving s s, it is a compliated mtre of pedues and we ae faced by the queson if hs ue should be le as it , o if it should be eplaced by a moe 'odely aangement. T p one of the agument. Now we have seen that science as we know i today could not est without a fequent oveuling of the context of juscaon. Ths is pat o of the agument. e conclusion clea P one shows that we do not ave a dieence, but a e. P o shows that eplacing the ue by an ode that con ove on one side and juscaon on the othe would have ed sience : we ae deang with a uom pacce al of whose ingedients ae equ impotant fo the gowth of science. T poses ofthe discon. A simla agument appes to the itual dcon een methodological prpt and toical po Methodolo, it is said, deals with what sh be done and cot be cied by efeence to wt But we mt of cou ke sue that ou pescipons have a nt ofak in the toical teial, and we must also make sue that the detemned applicaon leads to desiable esults. We ake su by consideing toical, sio logical, physcal, psychologcal, etc) t and ws whch tel us what is possible and what not possible unde the given ccumstances and thus sepaate feasible pescipons fom those whch ae gong to lead nto dead ends. Again, pogess can be de o the discon beeen the oght and the is egaded as a tempoa device athe than as a fundamental bounda ln e. A discon whch once may have had a poin but whch has now denitely lost it is the dscon beeen osatiol te and theoretil te. It now geneally adtted that t dcon is not as sha as it was thought to be only a few decades ago. It is also admitted, in complete ageement with Neuaths oignal views, that th theoies and obseaons can be abandoned: theoies y be emoved because of concg obseaons, obseaons may be emoved fo theoecal easons. Finaly, we have oveed that leaing ds not go fom obseaon to theo but aways involves both elements. Eeience aises together with theoecal umpons not beoe them, and an expeience without theo jt as ncompehensible as is (alegey) a theo without eience: elimnate pat of the theoecal knowledge of a sensng subject and you have a pson who is compete oiented and ncapable of Crg out the simplest acon. Eate the knowledge and senso wod (hs 'obseaon languae) w s integ,
150
A G A I N S T M E TH O D
colours and other simple sensaons will disappear unl he is in a stage even more primve than a small child. A sml child, on the other hand, does not possess a stable perceptual world whch he uses for makig see of the theories put before h Quite the cona he passes through various peeptual stages whch are only lsely connected with each other (earlier stages dsaear when new stages take over - see Chapter 16) and whch embody all the theorecal knledge avaiable at the me Moreover, the whole press starts only because the child reacts correctly towards sials, ntere th coe, because he possesses meas of interetaon een before he has experienced hs rst clear sensaon Al these discoveries c out for a new termnolo that no longer separates what is so inmately connected in the development both of the individual and of science at large Yet th� disncon beeen obseaon and theo is sl upheld But what is its point? Nobody w deny that the sentences of science can be classied into long sentences ad short sentences, or that its statements can be classied into those whch are intuively obvious and others whch are not Nobody w deny that such discons n e Bu nobody w put great weight on them, or w even menon them, r th not now p a e le n the usns ofs. (Ts was not awa so Intuve plausibty, for example, was once thought to be a m important de to the uth; it disappeared from methodolo the ve moment intuin was replaced by experience, ad by fo consideraos) Does expeence play such a role? It ds not, as we have seen Yet the inference that the dcon beeen theo an obseaon has now ceased to be relevant is either no drawn or explici rejeed4 Let us ake a step foard and let us abandon ths las ace of doam in ience! Incoensbty, whch I shal diss net close connected wth the queson of the raonaty of ience Indeed one of the most genera objecons not merely to the use of incoen surable theories but even to the idea that there are such theories to b found in the to of science the fear that they would severe rect the ecacy of adional, nondialeccal amt. Le therefore, lk a ttle more cloely at the crical stanr whc, according to some, constute the content of a 'raonal arent More especialy, let us lk at the standards of the Popperian schl, 4 ur fs o gv rus [wi dsnis mpiral samns ors] and us unwngly row mpirsm ovoard KR Popr e L f St D, Nw York and London 959, p 97.
FIFN
151
wich are sll being aken seriously in the more backward regions of nowledge. This will prepare us for the nal step in our discussion of the issue beteen law-and-order methodologies and anarchism in cience. Some readers of my arguments in the above text have pointed out that Popper's crical raonalism is suciently liberal to accom modate the developments I have described. Now crical raonasm is either a meaningl idea or it is a colecon of slogans that can be adapted to ay situaon In the rst case it must be possible to produce rules, standards, rescons which pert us to separate crical behaviour (thinkg, singng, wrig of plays) from other types of behaviour so that we can discer irraonal acons and coe them wth the help of concrete suggesons. It is not dicult to produce the standards of raonality defended by the Popperian schl. These standards are standards of tism raonal discussion consists in the attempt to ccze, and not in the attempt to prove or to make probable. Eve step that protects a view from cricism, that makes it safe or well-founded, is a step away from raonality. Eve step that makes it more vulnerable is welcome In addion, it is recommended to abandon ideas which have been found want ing and it is forbidden to retain them in the face of song and successful cricism unless one can present suitable counter arguments. Develop your ideas so that they can be criczed; attack them relentlessly; do not y to protect them, but exhibit their weak spots; eliminate them as sn as such wea spots have become manifest these are some of the rules put foth by our crical raonalists. These rules become more denite and more detaied when we to the phiosophy of science and, especially, to the phiosophy of the naral sciences. Within the natural sciences, cricism is connected with experi ent and obseaons The content of a theo consists in the s oal of those basic statements which conadict it; it is the class of its potenal falsiers. Increased content means increased lnerabilty, ence theories oflarge content are to be preferred to theories of small content. Increase of cotent is welcome, decrease of content is to be voided. A theo that conadicts an accepted basic staement must given up. Ad hoc hypotheses are forbidden and so on. A science wever, that accepts the rules of a crical empiricism of this nd ill develop n the following manner. . We start with a , such as the problem of the planets at the e of Plato. This problem (which I shal dicuss in a somewhat
152
AGAI NS MHOD
idealized form) is not merely the result of osi, it is a thortil rlt. t is due to the fact that cetan aon have been disappointed: on the one hand it seems to be clear that the stars must be divine, hence one epects them to behave in n orderly nd lawul mer. On the other hand, one cannot nd any easiy dieble regularity. The planets, to intents and puoses, move in a quite chaoc fashion. How can ths fact be reconcied with the epectaon and with the pinciples that underlie the epectaon? Does it show that the epectaon is mistaken? Or have we failed in our analysi of the facts? This is the problem t i impot to see that the elements of the problem are not simply The 'fact of ieuarity, for eample, i not accesible without futher ado t cannot be icovered by just anyone who healthy eyes and a gd d. t is only thrugh a ce epectan that it becomes an obect of ou attenon Or, to be more accurate, t fact of ieuarity because there i an epectaon of reuarity and because there are ideas wch dene what it me be 'reuar. er , the term 'ieuarity makes sense only we have a rule. In ou case the rule that denes reuarity asserts c moon with constant anuar velity. The xed stars agree with t e and so ds the sun, we ace its path relave to the xed s e planets do not obey the e, neither decy, with ree to e e, nor indiecy, with respect to the xed s. In the problem we are eg now the e i formulae epli and it can be discussed. T i no aways the Recog a colour as red i de possible by deep- pattes conceg the sucture of our surroun and rec nion ds not cur when these patte cease to et) To s up t p of the Popperian de: research s th a problem The problem i the result f a coct beeen epectaon and an obseaon wch i constute by e epectaon It i clear that t de ers from the de f inducm where obecve facts enter a passe d a leave e aces there. It was prepared by Kt, Mach, Poincar, er, by M O L). Hag foulated a problem, one ies to o it So problem means inveng a theo tat i relevant fable (to degree larger than any alteave), but not yet falsed In the c menoned above (planets at the me ofPlato) the problem i to cicular moons ofconstant anuar velity for the puse of the plane phenomena. A rst soluon was provided by Eudx and then by eracleides ofPontos. Net comes the tim of the theo that has been put foth in e
153
IF
additionl predictis
succes of th ld the (t f the tth contnt f th new th)
failre the old th a of th flsity contnt of the w the)
attemp to solve the problem. uccessful cricism removes the theo once andr a and creates a new problem, . to eplan (a) why the theo was successl so far; (b) why it failed. Tng to solve th problem we need a new theo that reproduces the successful consequences of the older theo, denies its mistakes and makes addio predicons not made before. These are some of the al ditions whch a suitale successor of a ruted theo must sas. Adopng the condions, one proceeds conecure ad refutaon fm less genera theories to more genera theories and epands the content of human knowledge. More and more facts are discered (or consucted with the help of epectaons) and are then eplained theories. There is no arantee tat siensts will solve eve problem and replace eve theor that has been refuted with a successor sasing the formal condions The invenon of theories depends on our talents and other foruitous circumstances such as a sasfacto se life But as long as these talents hold out, the enclosed scheme is a correct ccount of the growth ofa knowledge tat sases the rues of crical roalis. o at ths point, one my raise o quesons. 1 . s it irale to live in accordance with the rules of a crica ronai sm? 2. s it possile to have both a science as we know it and these rules? As far as conceed, the rst queson is far more impot an the second. True, science and related insuons play an iportant p in our culure, and they cupy the cene of interest for any phiosophers (most phlosophers are opporunists). Thus
154
AG AI S MHOD
the ideas o the Popperan hool were obtained by geneazing soluons for methodological and epistemological problems. Crical raonalism arose from the attempt to understand the Einsteinian revoluon, nd it was then etended to polics nd even to the conduct of ones private life uch a predure may sas a schl philosh, who looks at life thugh the spectacles of his wn technical pblems and recozes haed, love, happiness, only to the etent to which they ur in these problems But i we consider human interests and, above all, the queson of human freedom (freedom from hunger, despair, from the tyrany of conspated systems of thought and not the academic 'freedom of the wll), then we are preeding in the worst possible fashion For is it not possible tat science as we know it today, or a 'search for the uth in the style of adional philosophy, w create a monster? s it not possible that an objecve approach that frowns upon personal connecons beeen the enes eamed wll ha people, tu them into mserable, unfendly, self righteous mechasms without cha and humour? 's it not possible, asks Kierkegaard, 'that my acvity as an obecve [or a cco-raonal obseer of ature w weaken my sength as a human being? suspect the answer to many of these quesons aave and believe that a reorm o the iences tha es them more anarchic and more subjecve (in Kierkegaards sense) ugen needed. But these are not the problems want to dscuss now. n the present essay shal resct myself to the second queson and ask: is it possible to have both a science as we know it and the es f a crical raonai as just described? nd to this queson the er sees to be a rm and resounding NO. To s with we have seen, thou rather briey, that the a develpment of instuons, ideas, pracces, and o on oen stafm a p but rather from some eaneous acvity, suc playng, which, as a side eect leads to developments which later o can be intereted as soluons to unrealized problem. 6 e uc developments to be ecluded? nd, if we do eclude them, w not considerably reduce the number of our aapve reacons and the quality of our leag press?
J
5 , d Hbrg, V, P , s A, o 1 82. Ml is o sow ow sn mod an b undrsood pa ofa ory of man and us gvs a siv awr quson rsd by rgaard f fooo 2 o Cpr 4 6. Cf bf ommns on rlaon bn da and aon in Cpr For dals f fos 3 1 ofAga Mod', , Vol 4, 97
FIF
155
Secondly, we ave seen, in Chapters 1, that a st pnpl of fion, or a 'aive falsicaonism as Lakatos cals it, 7 would wipe out science as we know it and would never have permitted it to start. The demand for ind contt is not sased either. Theories which eect the ovethrow of a comprehensive and well-enenched point ofview, and take over aer its demise, are inially resicted to a fair narrow domain of facts, to a series of paradigmac phenomena which lend them support, and they are only slowly extended to other areas. This can be seen from historical examples (fooote 1 of Chapter ), and it is also plausible on genea grounds: ng to develop a new theo, we must rst take a st k from he evidence and reconsider the problem of obseaon (this was discussed in Chapter 11) Later on, of course, the theo is extended to other domains; but the mode of extension is only rarely determined by the elements that constute the content of its predecessors The slowly emerging conceptual apparatus of the theo s stas ning i o p rl, and earlier problems, facts, and obseaos are either forgotten or pushed aside as irrelevant. This is an enre natural developmen, and quite unobjeconable For why should an ideolo be consaned by older problems which, at any rate, make sense only in the abandoned context and which lk siy and unnatural now? Why should it even consi the 'facts that ve se to problems of kind or played a role in their soluons? Why should it not rather peed in its own way, devsing its own k and assembling its own doman of 'facts? A comprehensive theo, after all, is supposed to conta also an ontolo that determines what ests and thus delmits the doman of possible facts and possible quesons. The develop ment of science agrees with these consideraons New views sn sike out in new direcons and frown upon the older p rl (what is the base upon whch the e rests? what is the specic weight of phlogiston? what is the absolute velity of the e?) and the older fas (most of the facts described in the M M Chapter , fooote the facts of Vdoo Chapter , fote - the prperes of phlogiston or those of the ether) which so much exercised the ds of earlier thikers. And where they pay attenon to preceding theories, they y to accommodate thei factual core in the manner already described, with the help of hoc hypotheses, hoc appromaons, redenion of terms, or by simply 7 Falsaon and Modolo of Sn Rsar Progrmms', in LakaosMusgv (ds), Ctm and the Gwth fKnwledge, Cbrdg, 197, 93 (av fasaonsm' s r aso alld dogma')
156
A G A I N S M H O D
sing, without any more detailed study of the matter, that the core
'folows from the new basic pinciples. 8 They are 'grafted on to older programmes with which they [are] blatantly inconsistent. 9 The result of all these predures is an interesng tolo iusion the imagind content of the earlier theories (whch is the intersecon of the remembered consequences of these heories with the newly recoged doman of problems and facts) shn and y decrease to such an extent that it becomes smaler than the imand content of the new ideologies (which are the actual conequences of these ideologies plus al those 'facts, laws, pinciples which are ed to them by hoc hypotheses, hoc appromaons or by he sayso of some inuenal physicit or philosopher of science and whic properly belong to the predecessor) Compaing the old and the new it thus aa that the relaon of empiical contents is ke this n
or, perhaps, lke this
whie in actual fact it is much more lke this
8 'nsns o i br an . Nwons o nno 1 . �e
xpland vng a Nwons o ad sussly xpland op . p. 24 9 Lks ussng Coiu and or bid. p. 143
FIFTEN
157
domain D represenng the problems and facts of the old theo which are sll remembered and which have been distorted so as to t into the new famework. It is ths iusion which is responsible for the 10 peristent surival of the demand for increased content Finally, we have by now seen quite disctly the need for hoc hypoth hoc hypotheses and hoc appromaons create a tentave area ofcontact between 'facts and those parts of a new view which seem capable of explaiing them, at some me n the ture and aer addion of much ther material They speci possible explananda and explanaa, and thus determne the decon of ture reseach ey y have to be retaied forever if the new framework p unhed (ths happened in the case of the quantum theory, which needs the clasical concepts to tu it into a complete theo) Or they are incoorated into the new theo as theorems, leading to a redeion of the basic tes of the preeg idelo t happened in the cases of Galeo and of the theo of relavity) e ded that the uth-content of the earlier theo as ed wh the earl theo red spr be included in the uth-content of the successor violated in either case To sum up: wherever we lk, whater epes we consider, we see that the prnciples of crical raonm (take ficaons seriously; increase content avoid hoc hypoteses; be honest whatever th me; and so on) and, a o, the prnpes of logical empim (be prece; base your teories on measuements avoid vague and untestable ideas and so on), though praced in speal areas, give an inadequate account of the past development ofscience as a whole and are able to hnder it in te ture ey give an inadequate account of ience because ence much more
uion i or o ar' n ar on dmn om Lrn o org o Z, udd Ln 0
panaon of on of M 95) Bu 95 noy d a udd ng a aun o aon o o dgr ofaroon rad b nd o�, nd o oLrn on aomi l (n o o m) wr no und for ir, bu wr gadu rd b qunum o: Lrn udd' no b on, u b a la o n nd muuay mmb , n n ron o domn o o Cpcu from m o R no g bu 0 bau o nd a rob nd on on mac, ur nd Swondr a bo Z nd ar undr ron a nn ondon i d C m o no ar on n', Bh Jf hh fS, M 9 a w a RN N S Ra a a Sag omn o nu o', t Sum No 34, 988, 5
15
AGAINST METHOD
'sloppy and 'rraonal than ts methodologcal mage And they are lable to hnder t because the attempt to make scence more 'raonal and more precse s bound to wpe t out, as we have seen e derence between scence and methoolo whch s such obvous fact of hsto, therefore, ndcates a weakness of the latter, and perhaps of the 'laws of reason as well or what appears 'sloppness, 'chaos or 'oppounsm when copared wth such law has a most mportant ncon n the development of those ve theores whch we today rerd as essenal parts of our knowledge o nature e tons the eo are precondtos ofpro. ey permt knowledge to surve n the complex and dcult world whch we nabt, they permt u to reman free and happy agen Wthout 'chaos, no knowledge Wthout a frequent dsmssal o reason, no progress Ideas whch today form the ve bass o scence est oy because there were such thngs as prejudce, concet passon because these thngs osed reon; and because they we peed to he thr w. We have to conclude, then, that wth scence reason cannot and should not be allowed to be comprehen sve and tha t must oen e overruled, or elmnated, n favour o other agences There s not a sngle rule that remans vald under l crcumstances and not a sngle agency to whch appeal can always made 1 1
n Lkaos' ngnous modolo ds no sap nd
os sms lbral baus forbds ll and sms ron baus s forbds somng Bu ony ng forbids s o dgnr rsr rogrmm', a rsar rogrmm laking n nol rdon lud h adaon, rogr H ds no forbid us B m a i sndard p a mn o omm n rims rdd nr li abou m es in m hhl , Vol 2, C
Append
Havng lstened to one of my anarchsc sermons, Professor Wer exclamed: 'But surely, you do not read all the manuscrpts which people send you, but you throw most of them nto the wastepaper basket I most certanly do 'Anythng goes does not mean that I shall read eve sngle paper that has been wrtten God forbd! t means that I make my selecon n a highly ndvdual and dosyncrac way, partly because I cant be bothered to read what doesnt nterest me and my nterests change from week to week and day to day partly because I am convnced that humanty and even Scence will prot from eveone dong his own thig: a physcst mght prefer a sloppy and partly ncomprehensble paper ful of msakes to a cstal-ear exposon because t s a natural extenson f his own, sl rather dsorganzed, research and he mght achieve success as well as clarty long before hs rval who has vowed never to read a sngle woolly lne (one of the assets of the Copehagen Schl was ts abty to avod premature precson) On other cons he might look fr the most perfect proof of a prciple he s about to use n order not o be sdeacked n the debate ofwhat he consders to be man results There are of course so-called 'thikers who subdvde ther mal n exactly the same way, come ran, coe sunshine, and who also mitate each others princples of choce but we hall hary admire them for ther unformty, and we sll certanly not thik ther behavour 'raonal Scence needs people ho are adaptable and nvenve, not rgd mitators of establshed behavoural pattes In the case of nstuons and organiaons such as he Naonal Scence oundaon the stuaon s exactly the same. e physoomy of an rnzaon and ts ecency depends on ts ebers and t mproves wth ther mental and emoonal aglty ven Prter and Gamble realied that a bunch ofyes-men s nferor compeve potenal to a group of people wth unusual opnons and busness has found ways of ncoorang the most amazng nonconformsts nto ther machne Specal problems arse wth
159
160
A A N S METHOD
foundaons that dsbute money and want to do ths n a just and reasonable way Jusce seems to demand that the allaon of funds be carred out on the bass of standards whch do not change from one applcant to the next and whch reect the ntellectul stuaon n the elds to be supported The demand can be sased n an hoc manner wthout appeal to unal 'standards of raonalty any free assaon of people must respect the lusons of ts members and must gve them nstuonal support The lluson of ratona becomes especally song when a scenc nstuon opposes polcal demands In ths case one class of standards s set agat another such class and ths s qute legmate each organzaon, each pa, each relgous group has a rght to defend ts parcular form of lfe and all the standards t contans But st go much fuher. Lke the defenders of The ne True Relgon before them they nsnuate that ther standards are essenal for arrvng at the Truth, or for getng Results and they deny such authorty to the demands of the polcan They oppose lpolcal nterference, and they fall over each other ng to remnd the lstener, or the reader, of the dsasous outcome of the Lysenko aar Now we have seen that the belef n a unque set of standards that has always led to success and wl always lead to success s nothng but a chmera The theoretcal authorty of scence s much smaller than t s supposed to be Its soal authorty, on the other hand, has by now become so oveowerng thatpoltl nteerce necsa to rtore a alanced elmt And to udge the s of such nterference one must study more than one unanalysed case ne must remember those cases where scence, left to tself, commtted grevous blunders and one must not for et the nstances when polcal nterference dd mpe the stuaon Such a balanced presentaon of the evdence may even convnce us that the me s overdue for addng the separaon of state and scence to the by now qute custo separaon of state and church Scence s only one of the many nsuments people nvented to cope wth ther surroundngs It s not the only one, t s not nfallble and t has become too powerful, t pushy, and too dangerous to be left on ts own Next, a word about the pracal am raonalsts want to realze wth the help of ter methodolo Raonalsts are conceed about ntellectual polluon I share t conce Illterate and ncompetent bks d the market emp verbage ful ofsange and esoterc terms clas to express profoun
9
1 An xaml w disussd n t txt to fots 12 of Car 4
161
APP ENDI X
insights, 'eerts without brains, character, and without even a modicum of intellectual, stylisc, emoonal temperaent tell us about our 'condion and the means for improving it, and they do not only preach o us who might be able to look through them, they are let loose on our chldren and permitted to drag them down into their on intellectual squalor. 2 'Teachers using grades and the fear of failure mould the brains of the young unl they have lost eve ounce of imagnaon they mght once have possessed. Ths is a disasous situaon, and one not easiy mended. But I do not see how a raonalisc methodolo can help far as I am conceed the rst and the mos pressing problem is to get educaon out of the hands of the 'professional educators The consaints of grades, compeon, regular examnaon must be removed and we must ao sarate the pcs of leangfm the praraton r a par tra. I grant that business, regions, special professions such as science or prostu on, have a right to demand that their parcipants anor praconers conform to standards they rerd as important, and that they should be able to ascertain their competence. I also admt that this implies the need for special types of educaon that prepare a man or a woman for the corresponding 'examaons The stanards taught need not be 'raonal or 'reasonable in any sense, though they wll be usuay presented as such; it suces that they are ted by the groups one wants to join, be it now Science, or Big Business, or The One Tue Religion. Aer all, in a demracy 'reason has just as muchright to be heard and to be expressed as 'unreason espealy in view of the fact that one mans 'reason the other s inaty. But one thg must be avoided at all costs: the special standards hch dene special subjects and special professions must not be alowed to permeate geral educaon and they must not be made the denng prope of a 'well-educated person. Gener educaon should prepare ciens to choose etwe the standards, or to nd ther way in a society that contas groups cotted to various standards, ut t must unr condton d thr m so that th co to the stanr ofoneparup. The standards w be cosed, they ill be dssed, chldren will be encouraged to get prociency in the ore mportant subjects, ut on a one ge p n a game, that is, without serious coent and without robbing the mind of its abilty to play other mes as wel Having been prepared in t way a young person may decide to devote the rest of hs life to a parcular profession and he may start takng it seriously fothwith T 2 En t law now sms to su ts ndns, Hubr's G Nw York, 1991
is sown n Ptr
162
AGAIST METHOD
'comment should be the result o f a conscous decson, on the bass of a farly complete knowled of alteaves, and not a gone conuson. Al ths means, of course, that we must stop the scensts from takng over educaon and from teachng as 'fact and as the one ue method whatever the myth of the day happens to be Agreement wth scence, decson to work n accordance wth the canons of scence should be the result of examnaon and choce, and not ofa parcular way of brngng up chldren It seems to me that such a change n educaon and, as a sult, n perspecve wll remove a great deal of the ntellectual polluo raonalsts deplore The change of perspecve makes t clear that there are many ways of orderng the world that surrounds us, that the hated consants of one set of standards may be broken by freely accepg standards of a derent knd, and that there s no need to reect al order and to allow oneself to be reduced to a whnng sea of conscousness A sety that s based on a set ofwell-dened and rescve rules, so that beng human becomes synonymous wth obeyng these rules, rc the dsster nto a noman sland ofno at a and thus s hm ofhs reon and hs human. s the parado of mode rraonalsm that ts proponents sently den raonalsm wth order and arculate speech and thus see themselves forced to promote serng and absurdty many forms of 'myscsm and 'estenalsm are mpossble wthout a rm but unrealzed comment to some prncples of the despsed deolo ust remember the 'theo that poey s nothng but emoons colourfully expressed) Remove the prncples, admt the possbl of many dent forms of lfe, and such phenomena wll dsappear lke a bad dream My daoss and my suggesons concde wth those of Lakatos up to a pont. Lakatos has dened overly-rgd raonalty prncples as the source of some versons of rraonalsm and he has urged us to adopt new and more lberal standards I have dened overly-rg raonalty prncples as well as a general respect for 'reason as the source of some forms of myscsm and rraonalsm, and I also ur the adopon of more lberal sandards But whle Lakatos great 'respect for grea scence 3 makes hm look for the standards wt the connes of mode scence 'of the last two centures 4 I recommend to put scence n ts place as an ntesng but by n means exclusve form of knowledge that has many advantages but 3 'Hsto 113 4 ibd
APPED IX
163
also many drawbacks: Athough science taken as a whole is a nuisance, one can sll lea from it 5 Aso I dont believe that charlatans can be banned just by ghteng up rules Charlatans have ested at l mes and n the mos ghtlykt professions Some of the examples whch Lakatos menons 6 seem to indicate that the problem is created by too much conol and not by too little. 7 Ths is especially ue of the new revoluonaries and their reform of the universies. Their fault that they are uritans and not that they are libernes. 8 Besides, who would expect that cowards w improve the intelectual clmate more readiy than w libernes? (Einstein saw t problem and he therefore advised people not to connect their research with ther profession: research has to be free from the pressures whch professions are likely to impose. We must also remember that those rare cases where liberal methodologies encourage empty verbiage and loose thnkng (loose from one point of view, though perhaps not from another) may be inevitable in the sense that the guity liberalism is o a pcondion of a free and humane life. Finally, let me repeat that for me the chauvinism of science is a much greater problem than the problem of intellecual poluon It may even be one of its major causes. Sciensts are not content with runng their own playpens in accordance wth what they rerd as the rules of scienc method, they want to universalze these rules, they want them to become part of society at large and they use eve means at their dposal argument, propanda, pressure taccs, inmidaon lobbying to achieve the as The Cese ommunsts recoized the dangers iherent in ths chauvsm and they proceeded to remove it In the process they restored important parts of the intellectual and emoonal heritage of the Cese people and they also improved the pracce of medicine 10 • It would be of advantage if other govements followed suit.
Gottfried Benn, eer to Ge Miha Simon of Otober 1949, quoted from ottfried Benn Lyk und Ps B und Dkumte, Wesbaden, 1 962, p. 235 6 Fasaon', p 1 76, fote Cf. aso h remars on fase onsousness' History', pp 94 1 08. 8. For an oder exame, f. the BEitein Lett, ew York, 197 1, p 9. ibid., pp. 05. 1 0. Cf. text to fotes 912 of Chapter 4
16 Final th nd ofcompason that unrli most mtholog possi on in som rather simpl It ra wh w t to compar non-stc vis with sc or wh w consir th most an most geral and therr most mytholol pas ofsc its
I have much sympathy wth the vew, formulated clearly and elegant by Whorf (and ancpated by Bacon), that languages and the reacon pattes they nvolve are not merely nsuments for ing events (facts, states of aar), but that they are also shap of events (facts, states of aars), 1 that ther 'grammar contas a cosmolo, a comprehensve vew of the world, of sety, of the stuaon of which nuences thought, behavour, percepon 3 Accordng to Whorf the cosmolo of a language s expressed partly by the ove use of words, but t also rests on classcaons 'which ha[ve] no ove rk but which operate [] through an nvisble "cenal exchan of lnkage bonds n such a way as to determine other words whi rk the clss 4 us '[t]he gender nouns such as boy l father wife, uncle, woman, lady, ncludng thousands of gven names e George, red, Ma, Charle, Isabel, Isadore, Jane, Jon, Ace Aoysus, Esther, Lester, bear no dsngushng ark of gender e the Lan - u or -a withi each motor process, but nevetheless ea of these thousands of word has an nvarable lkage connecng t with absolute precson ether to the word "he or to the word "she which, however, does not come nto the overt behavio Aording to of 'the bakground nis system (n oher word h gramm) of eah anage is not merey a reproduing system for voing deas b raher is sef a shaper of deas he proamme and de for he ndvduas me av for hs anaysis of mpessions for hs ynhess of hi ment stk in de Lanage ght and Reali, Cambridge Mas 1956, p. 121 See aso Append 2 an exampe f. ors anayss of Hop Mephysis n bd. pp. 57 3. 'Users of markedy derent gmm are nted by her gmmars towa derent types of obseaons . ibid. p. 221 4 bd. p. 69
16
SIXTEE
165
picture unl and unless special situaons of discourse require it' C classicaons (wic, because of teir subterranean nature, are s ensed rater tan compreended aareness of [em] as an intuive quality'6 wic are quite apt to be more raonal tan overt ones' 7 and wic may be ve subtle' and not connected wit any grand dicotomy'8) create patteed resistances to widely divergent points of view' 9 If tese resistances oppose not just te ut of te resisted alteaves but te presumpon tat an alteave as been presented, ten we ave an instance of incommensurabilty I also belive tat scienc teories, suc as ritotle's teo of moon, te teo of relavity, te quanum teo, classical and mode cosmolo are sucient genera, suciently deep' and ave dveloped in suciently complex ways to be consdered along te same lines as natural languages Te discussions tat prepare te ansion to a new age in pysics, or in asonomy, are ard ever rescted to te ove features of te otodox point of view Tey often reveal idden ideas, replace tem by ideas of a dierent nd, and cange overt as wel as cove classicaons alleo's analysi of te tower arent led to a clearer formulaon of te rstotean teo ofspace and it also revealed te dierence between impetus (an 5 bd p 68 6 bd p 70 Even '[a] phoneme may assume dente en dues as pa of
ts rap. In Engsh the phoneme ['tho] (the voed sound of urs nlly ony n the pto [ove assaon not onneted wth ny d ihotomy p 70] of demonsve pares (the this there than et) Hene here s a hi prsure aganst aeng he voed sound of n new or magna words ig ay o uzze et not havng demonsve meaning Enounterng suh a new word (e.g. o) on a page we wil "nsnvey gve t he voeess sound of n "hink ut t s not "nsnt. Just our od frend ns rp again (p. 76 my iis) 7 ibd p. 80. e pasge onnues ' . . . some rher foa and not very meanngfu ins group marked b some ove feature may han to onde very roughy wh some onatenaon of phenomena n suh a way as to sust a raonazaon of hs presm In he ourse ofphone hnge he dsnsing mark endng o what not s ost and he ass passes m a fo to a en one Its reane is now what dsnishes t as a as nd ts dea is what unes it me and use go on it beomes nreasiny ornd ound a ronae it ats �emanay suibe words and o foer members hat now re en napproprate Log s now what hos it togeher Cf Mils aount of hs eduaona deveopment as esrbed text to fote 4 of Chapter . 8. Whof op t p 70 Suh sube assao are ed tos by Whof. A pto is 'a submed sube nd eusive meng oesng to no actua word yet shown by is nalys to be nonay mt he gramar 9 bd p 247.
166
AGAIST METHOD
absolute maitude that inheres in the object) and momentum (which depends on the chosen reference system). Einstein's analysis ofsimul taneity uneathed some features ofthe Newtonian cosmolo which, though unown, had inuenced larguments about space and e, while Niels Bohr found in addion that the physical world could not be rerded as being enrely separated from the obseer and thus ve content to the idea ofndependence that was part of classical phsics Attending to cases such as these we realize that scienc arguen may indeed be subected to patteed resistances' and we expect that incommensurability will also occur among theories. (As incommensurability depends on cove classicaons and involves major conceptual changes it is hardly ever possible to give an explicit denion of it. Nor wil the customa reconsucons' succeed in binging it to the fore The phenomenon must be so, the reader must be led up to it by being confronted with a great variety of instances and he must then judge for himself. This w be the method adopted in the present chapter.) Interesng cases of incommensurabilty cur already in the domain ofpeton Given appropriate smuli, but dierent syste of classicaon (dierent mental sets'), our perceptual apparatus may produce perceptual objects which cannot be easiy compared10 A dect judgement impossible. We may compare the two obects in our mo, but not while attending to the sam pu. The t drawing below goes one step ther It gives rise to perceptu objects which do not ust nat other perceptual objects - th reg the basic categories - but prevent the formaon o an object whaver (note that the ylnder in the middle fades into nothiess as we approach the inside of the twopronged smul Not ven memo can now give us a l view of the alteaves
0. A mster of inospeon, Kenneh Ck, has reent desbed to vv how even he was defeated when he aempted to "s an iuson. Lkg great Veuez, he wnted to obsee what went on when he bsh sokes and f pigment on he nvas ansfoed hemseves nto a vson ofansred re e stepped bak But as he might, stepping bakward nd foard, he oud n er d boh vso at e same me ', E Gombrh, and Illi, Prneton, 56 p 6
SN
167
Eve picture with only a modicum of perspecve exhibits this penomenon: we may decide to pay attenon to the piece of paper on whic the lines are drawn - but then there is no threedimensional patte; on the other hand we may decide to investe the properes of ths patte, but then the surface of the paper disappears, or is integrated into wat can only be called an illusion. Te is no way of catching' the ansion from the one to the oter. In all these cases the perceived image depends on mental sets' that can be changed at will, without the help of drugs, ypnosis, recondioning. But mental sets may become frozen by illness, as a result of ones upbringing in a certain culture, or because of pysiological determinants not in our conol. {ot eve change of language is accompanied by perceptual canges.) Our attude towards other races, or towards people of a dieent cultural background often depends on fzen' sets of the second kind aving leaed to read' faces in a standard way we mae standard udgements and are led asay. A interesng example of physiologically determined sets leading to incommensurability is provided by the lmt of human rcton. As has been sugsted by Piaget and his school, 2 a cild's percepon proceeds throug various stages before it reaches is relavely stable adult form In one stage, obects seem to behave ve muc ike afterimages and are eated as suc. Te chid 11 Cf R Gregory he Inteit Eye, London 1 970 Chapter 2 Cf aso the non between ekon and pntsma n Pato Sht, 235b8 's "aar ng or seemng without reay "beng a these expressons have aways been and re deepy nvoved n ex Pato s about the dstorons n stues of ooa ze hh were nodud to make them ear with the proper prorons annot make use of an iuson and wath t says Gombrih n suh ases op it 1 2 J Paget he Ctin Real in the Chi, ew York 954 pp. .
168
AGAIST METHOD
folows the obect with his eyes un it disappears he does not make the slightest attempt to recover it, even if ths should require but a mnimal physical (or intellectual) eort, an eort, moreover, that already within the chids reach The is not even a tendency to search and this is quite appropriate, 'conceptualy speakg For it would indeed be nonsensical to 'look for an aerimage ts 'concept does not provide for such an operaon e arrival of the concept, and of the perceptual image, ofmateral objects, changes the situaon quite dramacaly There curs a drasc orentaon of behavioura pattes and, so one may conjecture, of thought Aerimages, or thigs somewhat ie them, sl est but they are now dicult to nd and must be disovered by special methods (the earlier visual world therefore literaly d aa). 13 Such methods preed from a new conceptual schem (afterimages occur in humans, they are not parts of the physical world) and cannot lead back to the exact phenomena of the previo stage (These phenomena should therefore be caled by a dierent name, such as 'pseudoaerimages a ve interesng perceptual analogue to he nsion from, say, Newtonian mechacs to spe relavity: relavity, too, ds not give us Newtonia facts, but relavisc aalogues of Newtonia facts) Neither aerimages nr pseudoaerimages have a special posion in the new world Fr eple, they are not eated as n on whch the new noon f a materal object supposed to st Nor can they be used to t noon: aerimages a togher wth t, they depnd on i an are absent from the minds of those who do not yet recoge te obecs and pseudoaerimages daar as sn as such recognion takes place e perceptual eld never con er ges together with pseudoaerimages t to be atted tat eve stage ossesses a kd of obseaonal 'bas to whch sp attenon is paid and from which a multude of suggesons are received However, ths bas (a) chang fm stage to stage, and t is pa of the conceptual apparatus of a given stage, no its one only source of interetaon as sme empiricts would ie to e us believe Considerng developments such as these, we may suspect that e famly of concepts cening upon 'material obect and the fa f concepts cening upon 'pseudoafterimage are incmmensurabe 3. is seems to be a gene feature of the aquison of new peept wo
e oder represeno for the mos t p have to suppressed rther refoed,' writes Son eh-mang essay Vson wthout nversion of te Rena Image' e Pl R , 897 p. 47.
SIXE
16
in pcisely the sense hat is at issue ere; hese families cannot be sed simulteously and neiher logical nor perceptual conneons can established between hem Now is it reasonable to expect that conceptual and perceptual canges of his ind cur in childhd only? Should we welcome he fact, if it is a fact, hat an adult is stuc wih a stable perceptual world and an accompanying stable conceptual system, which he can modi in many ways but whose general outlines have forever become immobilized? Or is it not more realisc to assume hat fundamental cans, enailing incommensurability, are sll possible and hat hey should be encouraged lest we rmain forever excluded from what might a higher stage of nowledge and consciousness? Besides, he queson ofhe mobiity of he adult stage is at any rate an empirical queson hat must be attaced by rarch, ad cannot be setled by mehodological}4 Te attempt to brea hrough he undaries of a given concepual system is an essenal part of suc search (it also should be an essenal part of any interesng life). Such an attempt involves muc more han a prolonged crical discussion' 5 as some relics of he enlightenment would have us lieve. One must be able to produ and to p new perceptual and conceptual relaons, including relaons whch are not mmediately apparent (covert relaons- see above) and cannot be achieved by a crical discussion alone (cf also above, Capters 1 and 2). The orodox accounts neglect he covert relaons hat conibute to heir eaning, disrerd percepual canges and eat he rest in a rigidly sndardized way so hat any debate f unusual ideas is at once stopped by a series of roune responses. But now his whole array of responses is in doubt Eve concept hat curs in i is suspect, especially fundamental' concepts suc as obseaon', test', and, of cors he concept heo' itself. And as regards he word uh', we can at his stage only say hat it certainly has people i a zzy, but as not achieved much else. Te best way to preed in such ciumstances is to use examples whic are outside he range of he rune responses. It is for his reason hat I have decided to examine eans of representaon dient frm languages or heories and to dve lop my termiolo in connecon wih hem More especially, I all examine styles in painng and drawing. t will emere hat here 4 A Lakatos aempts to do Fason', p 179, fooote nomen urbe theories are neither innsistent with eah other, nor omparabe for ontent But we an me them, by a dionary, nonsistent and their ontent omparabe.' I · Poper in Ctim d the Gwth Knwlee, p 5
170
AGAI ST METHOD
are no 'neual objects which can be represented in ay style, and which measure its closeness to 'reality. The applicaon to languages is obvous. e 'archaic s le as dened by Emanuel Lwy n his work on ancient Greek art 6 has the folowng characteriscs. (1) The sucture and the oveent of the gures and of thi parts are lited to a few typical schees; (2) the indivdual forms are styized, they tend to have a certan regularity and are 'executed with . . precise absacon; 7 (3) the representaon of a form depends on the contour which ay retai the value of an independent line or form the boundaries of a sihouette. 'The sihouettes coud be given a nuber of potures: they coud stand, arch, row, drive, ght di, ament. . . . But aays ther essenal sucture ust be clear; 8 4 cour appea in one shade only, and gradaons of light and shad are missing; S as a rule the gures show their parts (and the la episodes thei eleents in their most ple pe even if t eans awwarness in coposion, and 'a ce disred of spaal relaonships The pa are given their known value evn when s conicts with ther seen relaonship to the whole; 9 thu (6) with a few wel-determined excepons the gures which form copoion are aanged in such a way that erps are od an obects situatd behnd each other are presented as bing side side; (7) the imt of an acon (ouns, clouds, es, tc) is eithr copetely direrded or it i omitted to a le xtnt. acon form self-coned unts of typical scenes attes, , etc.). 20 Thee styc eeents which are found, in various odic, in th drags of children, in the 'frontal art of the E, early Geek as wel as among socaled Pves, ar elan
6. u in t Ge Kut Rom c a g t cvrng phnomna Ean Gee ve the drwngs of chdn nd of untutod oers n G h re appy to the g s to 7 B do to the a 5 B wch eats the humn re in greater det nd vo t epes. f. aso F M a Gh G Kut Vol 950 w B nd shmole Greek Sture an Painting ambrd 9 hapters . 7. Webster c to Hom ew or 96 p. 292. Webster r this u ofsimple nd clear pae reek geomec the forener ofla delopments a umatey the nvenon of prspcve) mthema plophy. 8 Webster cit p. 205 . 9. bd. p. 207. 20. Be d shmole op. ct. p. 3.
SIXTEEN
171
y oewy on the basis of psychological mechanisms: 'Side by side with the images which reality presents to the physical eye there ests an enrely enrel y derent derent world of images images which live or, better bet ter come to life life in our mind only and a nd whch, although althoug h suggested by realy, realy, are totally totally ansf ansformed ormed Eve Eve primiv primivee act of drawing drawing . . e ess to reproduce reproduce these images and them alone wth the insncve regularity of a psy psychical funcon. 2 1 The archaic style changes as a result of 'numerous planned obseaons of naure which modi the pure mental images, 22 niate the development towards realism and thus start the histo his to of art. Natural, physiological reasons are given for the archaic style and for its change. Now it is not clear why why it should be more 'natural to copy memo memo images than images of percepon which are better dened and more permanent.23 We also nd that realism oen pece more schemac form of presenon. This is ue of the Old Stone Age, 24 of Epan 25 of Atc Geomeic Art. 26 In these cases the 'archaic style is the result of a coo (which ay of course be aided, or hindered, by unconscious tendencies and physiological laws rather than a natural reacon to inteal deposits
2 Loew Loewy y op t t p 4 22 ibd ibd p 6 23 The fats of perspeve are noed ut they do not enter the pitora presentaon tis is seen m lite despons f H Shfer At ed. Kut Wiesaden 963 pp 88 where the problem s rther ded. 24 f Paoo ios Plth A ew ok 9 nd Andr Le Treu ofPreht Preht A A ew or 96 Gourhan Treu 967 7 both wth exeent luaons These results were not known to Lwy as 'Mea ulpa dun sepue for example example appeaed appeaed ony n 2 25 f the hange n the preentaon of n the ou of the on from predynas mes to the Frst nas The er on r hes hes Museum Nr 224) s wid threatening uite derent in expresson nd xuon om the majes anma of the Seond nd Thrd nases. The laer eems to e more a representaon of the t lon thn of any ndvidua lon. f. also the derene beeen the faon on the vitory tablet of Kng er asde) asde) and on the buria stone of Kng Wadj jet) of the Fit nas 'Evewhere one advaned to pure li the foms were sngened and made Shfer op t pp 2 espe espeay ay p 5 where rter detas are smple smple Shfer gven 2 'At geome a should not be led primve athough t t has not the kind of photograp h reaism whh lterry sholars seem to demand n panng t is a highly sophsated a with ts own onvenons onvenons whh e e ts own puses As with he shap hapes and the oamentaon a revoluon separtes it from late Myenaean pnng n ths revoluon res were redued to ther mnmum souees nd out of hese mnmum slhouees the new a but up. Webster op it p 20
172
S METOD
of exteal exteal smuli. smuli . 27 Instead of looking for for the psychologcal caus of a style' we should therefore rather to dscover its elts, analyse therf ther fu non, non , compare them wth other phenomena of the same culture (tera style, sentence consucon, graar, deolo) and thus arrve at an outlne of the underlyng world-v ncludng nclu dng an account of the the way n wh whch ch ths ths world-vew nluenc nl uences es percepon, thought, argument, and of the lmts t mposes on the roamng about of the magnaon. We shall see that such an analyss of outlnes provdes a better understandng of the press of conceptual change than ether a naturalsc account whch recognzes only one realty' and orders artworks artworks by by ther closeness closen ess to t, or te slogans such a a crcal dscusson and a comparson of . . . varous varous fr framework ameworkss s alwa always ys possble'. possble'. 28 Of course, course, some kind of comparson s alws possble (for example, one physcal theo y sound more melodous when read aloud to the accompanment of a gutar gutar than another physcal theo). But Bu t lay downspec down spec rules for th press of comparson, such as the rules of logc a appled to the relaon of content classes, or some smple rules of perspecve and you wll nd nd excepon exc epons, s, undue rescons, resco ns, and you wil be forced forced to talk your way out of ouble at eve tu It s much more nter n teres esng ng and nsucve to examne what knds of thngs can be sad (represented (repre sented)) and what kinds of thngs thngs cannot be said (represente (re presented d the compason h to take ple wthn a cean speed and hstol we-trched framork. For such an examnaon we must go beyond generales and study frameworks n detal I start wth an account of some examples examp les of the archac archac styl style. e. The human gure shows the followng characterscs: the men are ve tall and the unk of a angle taperng to the th e wast, the head of a knob wth a mere excrescence excresce nce for a face face towards towards the end e nd of the style style the head s lt up the head knob s drawn drawn n n outlne, outln e, and a dot ses the eye'. 29 Al, or almost all, parts are shown n proe and they are sung together lik the lmbs of a puppet or a rag dol. They are not ntegrated' to form an organc whole. Ths addve' feature of the archac archa c style becomes be comes ve clear cle ar fro from the eaent ea ent of the eye. The eye ds not parcpate n the acons of the body, t does not gude the body or establsh contact wth the surroundng stuaon, stuaon, t does not look'. loo k'. It s s added on to the the prole head hea d like pa 27 Ths thesis s further suped by the obseaon that so-called Prmi oen t ther ac to the objects they want to to draw; Schf Sch fer er op. ct. p. p . 02 ae
onze etc. p 56 28 Popper n Ctim etc. Beazy y and Ashmole op. cit p. 3. 29. Beaz
SXTEE
173
of a notaon as if the arst wanted to say: 'and beside all these other hings such as legs, arms, feet, a man has also eyes, they are in the head, one on each side. Smilarly, special states of the body (aive, dead, sck are not no t indcated by a special arrangem arrangement ent of its parts, but by putng the same standard body into various standard positions. Thus the body of the dead man in a funeral carriage is arculated in exactly the same way as that of a standing man, but it is rotated through 90 degrees and inserted in the space between the bottom of the shroud and the top of the bier.30 Being shaped ike he body of a live man it is in tion tion put into the death posion Another instance is the picture of a id half swalowed by a lion. 3 1 The lion lks ferious, the id looks peacel, and the act of swalowing is simply tked on to the presentaon of what a lion and what a kid . (e have what is called a paraic ic aate: the elements of such an aggregate are al given equal importance, the only relaon between them is seque se quena nal,l, there is no hierarchy, hierarchy, no part part is presented prese nted as being subordinate to and determined by others The pcture re ferious lion, peacel id, swallowing of kid by lion The need n eed to show eve essenal essena l part of a situaon ofe ofen n leads lea ds to a separaon of parts which are actualy in contact The picture becomes a lst Thus a charioteer standing in a carriage shown as standing above the lr (whch is presented in its fulest view) and unencumbered by the rais so that his feet, the lr, the rails can al be clearly seen No ouble arses if we rerd the pang as a vl taloe of the pa of an event rather than as an illuso rendeing of the evnt itself (no ouble ases when we say: his touched the oor which is renr, and he was surrounded by a railng . )3 2 But such an interetaon must be leaed, it cannot be simply read o the picture The amount of leag needed y be consideable. Some Epan drawings and panngs can be decoded only with the help of either the represented obect itself or with the help of three dimensional copies of it (statua in the case of humans, als, etc. Using suc informaon we lea that the chi in Figure A represents th obect of Figure C and and not the obect of Figure Figure B and that it must be read 'chi with backest and four legs, legs connected 30 Webster op. ct. p 2 The pater feels the need to y hat he has o as o legs and a many many chest chest 3 R Hampl Die Ghnse H und d d Bit sen sen et et Tbingen 952. 32 l geomec pctures of charots show at least one of these dstorons Webs Webste terr op op cit. cit . p 2 Late Mycenaean ery on the th e other nd has the legs l egs of the cupant cupantss conceaed by the side. s ide.
174
G I S T M T T O O
by support' where t s understood that the front legs are connected wth the back legs and not wth each other. 33 The nteretaon of groups groups s complcated and some som e cases are not yet understood. 34
B
(Beng able to read' a certan style also ncludes kowledge o what features are elant. Not ve feature of an archac lst has representaonal representao nal value ust as not eve feature feature of o f a wrtten sentence sent ence plays a role n rculang ts content. Ths was overlooked by the Greeks who started nqurng nto the reasons for the ded postures' of Epan statues (already Plato commented on ths. Such a queson mght have suck an Epan arst as t would strke us f someone nqured about the age or the ood of the kng \ on the chessboard' 35 ) So far a bref account of some peculares of the the archac' style style.. A style can be descrbed an analysed n various ways The descrpons gven so far pa attenon to nal atur the archac style provdes vsle lsts whose parts are arranged n rughly the same way n whch they cur n nature' except when such an arrangeme arrangemen n s lable o hde mpoant mpoant elements eleme nts Al parts are on on the Schfer op cit p 23 33 Schfer 34 bd pp 223 35 Gomrch op cit p 34 lteture
SIXTEE
175
same level, we are supposed to read' the lsts lsts rather than see' se e' them o f the stuaon stuaon 36 The lsts are not organzed n as lluso accounts of any way way except sequenally, seq uenally, that that s, the shape of an eleme e lement nt does not depend on te presence of other elements (addng a lon and the act of swallowng does not make the kd lk unhappy; addng the process of dng does not make a man look weak) Archac pctures are paratac hypotacc systems systems The elements elemen ts of the the paratac aegat, t, not hypotacc aggrete may be physcal parts such as heads, arms, wheels, they stat es of of aar aar such as a s the fact that a body body s dead de ad,, they may be may be states acons, such as the acon of swallowng Instead of descrbng the formal features of a style, we may e lement entss ontolocalatures of a world that conssts of he elem descrbe the ontolocalatures represented represen ted n the t he style style,, arranged n the approprate way way and we may also descrbe the mpresson such a world makes upon the vewer Ths s the predure of the art crc who loves to dwell on the pecular behavour of the characters whch the arst puts on hs canvas and on the nteal lfe' the behavour seems to ndcate Thus GMS Hanfmann37 wrtes on the archac gure: No matter how anmated anmated and agle archac heroes may be, they do not appear to move by ther own wll Ther gestures are explanato formulae mpose mposedd upon upon the actors actor s from from wthout wthou t n order to expla what sort of acon s gong on And the crucal obstacle to the convncng portrayal of nner lfe was the curously detached character of the archac archac eye. t shows that a person s alve, but t cannot adjust adju st tself to the demands of a specc stuaon. Even when the archac arst succeeds n denong a humorous or gc mood, these factors of extelzed gesture and detached glance recall the exaggerated anmaon anmaon of a puppet play' play ' An ontologcal descrpon frequently adds just verbage to the 36 We come closer to the factual factual content off of front ront teg] drwngs of obects f we st by ng o ther para contents n the fo of natve declartv declartvee sentenc se ntences es The front ront mode of representaton gves us a "vsua "vsua concept represented chf chfer er op ct p p 8 f aso [Sehb of the thng (the stuaton) represented Webster op ct p 202 about the natve nd eplnatory charter of ycenaean and geomec a But cf HA GrnewegenFrnko At and Mt London 95 pp. 33 3 3 f the scenes scenes from from daly lf l fe on the wals of Ean tomb tombss should shoul d be "rea " read d haesng ents ploughng sowng and reapng; reapng; care of cale cale entls fording fording of seams seams and mlking mlk ing . . the seuence seuence of scenes s purey con conceptua ceptuall not nave nor s the wrtng whch whch curs wth the scenes d c n caracter. The sgns remars names songs nd nd eplanaons whch illumnate il lumnate the aco do not ln events events or or eplan eplan ther development development;; they are pical sayn sayngs gs beongin in to pca stuaaon pcall stu ons. '
37 arraon in Gree A Amen Joual ofArhlo Vol. 6 Janu Januar aryy 957 p 74
176
AGAI ST METHOD
formal analyss t s nothng but an exercse n sensvty' and cuteness. However, we must not dsregard the possblty that a parcular style a prese count ofthe r a it is perced the aist and his contpora and that eve formal feature corresponds to (hdden or explct) assumpons nherent n the underlyng cosmolo. (n the case of the archac' style we must not dsrerd the possblty that humans then actualy lt themselves to be what we today would cal puppets guded by outsde forces and that they saw and treated others accordngly.) Such a realstic interetion of styles would be n lne wth Whors thess that n addon to beng nsuments for ing events (whch may have other features, not covered by any descrpon) languages are also shape of events (o that there s a lngusc lmt to what can be sad n a gven language, and hs lm concdes wth the lts of the thg tsel but t would go beyond t by ncludng non-lngusc means of representaon. 3 The rec nteretaon s ve plausble But t mut not be aken for granted. 3 It must not be aken for granted, for there are techncal falure, specal puroses (carcature) whch may change a style wthout changng the cosmolo We must also remember that humans have roughly the same neurophysologcal equpment, so that percepon cannot be bent n any drecon one chooses. And n some cases we can ndeed show that devaons from a fathful renderng of nature' cur n the presence of a detaled knowledge of the obect and de by sde wth more realsc' presentaons the workshop of th sculptor Thuoss n Tel al-Amaa (the ancent Achet-Aton) contans sks dectly taken from lve models wth all the detals of the formaon of the head (ndentaons) and of the face ntact, as wel as heads developed from such masks Some of these heads presee the detals, othrs elmnate them and replace them by smple form An exeme example s the completely smooth head of an Ep man. It proves that at least some arsts remaned concou ndependent of nature'. 41 Durng the re of Amnophs (B 1 361 37) the mode of representaon was changed twce the change, towards a more realsc style, curred merely four ye 38 Cf. fote and tet of the present chapter 39 For a sketch of the problems that arse in the case of psi th cf. y Repy to Crcism', Btn St in the hs oSce, Vol 2 1 965 seco 8 nd especily the list of problems on p 234 Hnson, Popper nd others take t f
grnted that realism is correct It may dierent wth drug-induced states, especiay when they are e pa of a stemc course of educaon 4 1 Schfer, op. cit, p. 63
SIXTEE
177
after his ascension to the throne which shows that the technica abiity for reaism ested, was ready for use, but was intenonay eft udeveoped An ercefm sle or lanage to smolo and mo
ofpecton therre nee speal amt t nnot e ma a a maer of ue (A simiar remark appies to any inference from popuar
theores in science, such as the theo of reavity, or the idea of the moon of the eath, to cosmoo and modes of percepon.) The argument (which can never be concusive) consists in poinng to charactersc features in distant eds. If the idiosyncrasies of a parcuar stye of paing are found aso in statua, in the grmar of contempora anguages (and here especiay in covert cassica ons which cannot be easiy twisted around), if it can be shown that these anguages are spoken by arsts and by the common fok aike, if there are phosophca principes formuated in the anguages which decare the idiosyncrasies to be features of the word and not just arfacts and which y to account for their origin, if man and nature have these features not ony in painngs, but aso i poey, in popuar sayigs, in common aw, if the idea that the features are parts of norma percepon is not conadicted by anything we know from physioo, or from the psychoo of percepon if ater tinkers attack the idosyncrasies as 'errors resung om an iorance of the ue way, then we may assume that we are not just deag with technca faiures and parcuar puoses, ut wth a hert of l, and we may expect that peope invoved in ths way of ife see the word in the same way in which we now see their pictures It seems that these condions are sased in archaic Greece: the forma sucture and the ideoo of the Greek c as reconsucted both frm the text and om ater references to it repeat a the pecuiaries fthe ater geomec and the eary archaic stye42 To start with, about nine-tenths of the Homeric epics consist of nul which are prefabricated phrases extending in ength from a singe word or two to severa compete ines and which are repeated at appropriate paces43 One-fth of the pms consist of ines whoy repeated from one pace to another; in 28, Homeric ines there are about 25, repeated phrases Repeons cur aready in Mycenaean court poey and they can be aced to the py of 42. Webster, op. cit., pp. 294 43 In the 20 century the role of foulae was described and tested by Milman Pa, L Epthe trtnel ch Home, Paris, 1928 aard Stud n Csl hl, Vols. 41, 1930, 43, 1932. For a brief account cf. Dl Page Hto and the Il, Berkeley, 1966, Chapter as we as GS Kirk, Hom and the Epc, ambridge, 1965, Pa I
178
AGA INST MET HOD
easte courts 'Ttles of gods, kngs, and mn must be givn correctly, and n a courtly world the princple of correct expresson y be extended futher. Royal correspondence s hghly formal, and hs formalty s extended beyond the messenger scenes of py o the formulae used for noducng speeches. Smlarly, operaons are reported n he tes of the operaon order, whether the operao order tself s given or not, and hs techque s extended to othe deripons, whch have no such operaon orders behnd them These compulsons derive ulmately from the cour of the kg and t s reanable to suppose that the court n tu enjoyed uch formalty n poey4 The condons of (Sumerian, Babyloni, Hurran, Hethc, Phoenican, Mycenaean) courts also explan th currence of standardzed elements of ntt (typcal ene the kg and the nobles n war and peace; future descripon of beaul thgs) whch, movng from cty to cty, and evn across naonal boundaries are repeated, and adapted to lal crcumstances The slowly arisng combnaon of constant and varable elemen that s the result of numerous adaptaons of hs knd was ulized b the lterate poets of the 'Dark Age of Greece who developed language and forms of expresson that best see the requrements o oral mpositon The requrement f mo demanded that there b readymade descripons of events that can be used by a pot wh composes hs mind, and wthout the ad of wrng requrement of mere demanded that the basc descripve phrases b t for us n the various parts of the lne the poet s about to complet 'Unke the poet who wrts out hs lnes . [the oral pt] ca thk wthout hur about hs next word, nor change what h made nor, before gong on, read over what he has just wrtten . . H must have for hs use word groups all made to t hs verse Eomy demanded that given a stuaon and a ce mec consant (begnnng, mddle or end of a lne) there be only one w of connug the narraon and hs demand s sased to surisng extent 'Al the chief characters of the iad and th Odyss, f ther names can be tted nto the last half of the verse alon wth an epthet, have a nounepthet formula n the noiav , beginnng wth a smple consonant, whch lls the verse betwen the hac caesure of the hrd foot and the verse end for stance Jo� o� OE� In a lst of thrtysevn characters who have formulae of hs type, whch ncludes thos Webster op. ct pp. 75f 45 M. P, HaardStud Cl Phil, 41 1 930 p 77
SIXTEEN
179
having any mportance in the pms, there are ony three names which have a second formua which coud repace the st 'If you ake in the ve grammac cases the singuar of a the noun-epithet formuae used for Achies, you wi nd that you have forty-ve different formuae of which none has, in the same case, the same meca vaue47 Being prvided for in this manner, the Homeric poet 'has no interest in originai� ofexpression, or in varety. He uses r adapts inherited formuae4 He does not have a 'choice, do[es] not even think in terms ofchoice for a given part of the ine, whatever decension case was needed, and whatever the subject matter mght be, the formuar vabua suppied at once a combinaon of words ready-de . 4 Using the formuae the Homeric poet gives an account of pical sc in which objects are casiona described by 'adding the parts on in a stn ofwor in apposion 0 Ideas we woud oday regard as being ogicay subordinate to others are stated in separate, grammacay co-ordinate proposions. Exampe 9.556: eeagros 'ay by his wedded wife, fair Ceopaa, daughter of fair anked aessa, daughter of Euenos, and of Ides, who was the songest of men on eath at that me and he aganst ord Phoebus poo took up hs bow for the sake of the far-anked maid: her then in their has did her father and ady mother ca by he name of Aon because . . and so on, for ten more ines and two or three more major themes before a jor stop. This rataic feature of Homeric poet which paraes the absence of eaborate systems of sbordinate causes in eary Greek 5 aso makes i cear why
ibid, pp 86f 47 bid., p 89
48 Page, op ct., p 230 49 bd., p. 242 50 Webster op ct, pp. 99f; my italcs. 51 . Cf Raphael Khner, A hrliche Grammatik Gechch Sprhe, 2 Tel, rprinted Dastadt, 1966 In the 2 century such a paratacc or simultanisc', ay of presenton was used by the early epressonsts, for eample by Jacob von Hoddi in his pm Welt Dem Brger iegt vom spien Kopf der Hut, In allen Lften hallt es e Geschrei Dachdecker sten ab und gehn enei, Und an den Ksten liest man stei die Flut. Der Stu st da, die lden Meere hupfen n Land, u dicke Dmme u erdrcken Die mesten Menschen haben enen Schnupfen ie Eisenbahnen faen von den Brucken
180
AGA INST METHOD
Aphrodite is called sweetly laughing' when in fact she complains tearfully (Il, 5.375), or why Achilles is called swift fooed' when he is sitng talking to Priam (Il, 2.559). Just as in late geomec potte (in the archaic' style of Lwy) a dead body is a live bo brought into the posion ofdeath (cf above, text to fote 30) or an eaten kid a live and peaul d brought into the appropriate relaon to the mouth of a ferocious lion, in the ve same way Aphrodite complaining is simply Aphrodite and that is the laughng goddess nseed into the situaon of complaining in which she parcipats only exteally, without changng her nature. The te treatmt of events becomes ve clear in the case of (human) moon. In Ili, 22.298, Achiles drags Hector along in th dust and dust arose around him that was dragged, and his dark h owed lse on either side, and in the dust lay his once fair head' that is, the pcs of draing contas the state of lying as an independent art which together with other such parts constutes the moon. 5 Speaking more absactly, we mght say that for th pt me is composed of moments' .5 3 Many of the simles assum that the parts of a complex enty have a life of their own and can separated wth ease Geomecal man is a visible list of parts and posions; Homeric man is put together from limbs, surfacs connecons which are isolated by comparng them with inaniat obects of precisely dened shape the unk of Hippolhos ro through the battle eld like a log after Amemnon has cut o
;
Von Hoddis claims Homer as a precursor, explaining hat smulanet was used Homer not in order to make an event more ransparent ut in order to create a feen of immeasurabe spaciousness. Wen Homer descrbes a bale and compares he nose of he weans wh he beat of a wcuer, he merely wants to show hat whe here s bale here s also he quieess of ws, nteupted only by he work of he wcuer Catasophe cannot be hought whout simultaneously hinking of soe uerly unmant event. The Great is mxed up wh he Smal, he Imant he Tral (For he re cf J.R. Becher n Ersinm , ed. P Raabe, Olten an Freiburg, 1965, pp ; his sho acle also conains a descrpon of emendous impression von Hodds' eight-liner made when t rst came out n 191 One cannot infer hat he me impression was created in he lstener of he Hoe singers who did not ssess a complex and romanczing medum hat deteorated nto teaful senmentalt as a background for comparson. 52. Cf Gebhard Ku, Datellun mschli Bewung in l Hedelberg, 1966, p. 53. This is e heory ascrbed to Zeno by Arstote, Psi, 239b, 3 1 The he comes forth most clear in he arment of he arrow The arrow at ight is at r For, if everyhng is at rest when it cupes a space equa to tself, and what s in t at any gven moment always cupies a space equal to itself, it cannot move' (aer Physi, 239b) We cannot y hat he heory was held by Zeno himself, but we conecture hat it played a role n Zeno's me
181
SIXTEEN
ams and s head (I, 116 o�, round stone of lindrical shape), the body of Hector spins like a top (, 1 12), the head of Gorythion drops to one side like a garn being heavy with fuit and the showers of spring' (I, 8302) 4 and so on lso, the formulae of the epic, especially the noun-epithet combina ns, are frequently used not according to content but according to meical convenience Zeus changes from counsellor to storm montain god to pateal god not in connecon with what he is doing, but at the dictates of mee He is not nhelegerata Z when he is gathering clouds, but when he is lling the mecal unit, 55 ust as the geomecal arst may distort spaal u u - u u ', relaons inoduce contact where none ests and break it where it curs in order to tel the visual sto in hs own pacular way Thus the poet repeats the formal features used by the geomec and the early archaic arsts Neither seems to be aware of a underlying substance' that keeps the obects together and shapes their parts so that they reect the higher unity' to which they belong Nor is such a higher unity' found in the concepts of the language For example there is no expression that could be used to descrbe the human body as a single enty 56 Soma is the cose, is accusave of specicaon, it means in sucture', or as rerds shae, reference to ms curs where we today spea of the body (, limbs as moved by the joints; , limbs in their bodily sength; o , his whole body embled; Qo� b V QV, his body was lled with sength) l we get is a puppet put together from more or less arculated parts The puppet does not have a soul in our sense The body' an aggrete of limbs, unk, moon, the soul' is an aggrete of mental' events whch are not necessarily private and whch may belong to a dierent individual altogether Never ds Homer in his descripon of ideas or emoons, go beyond a purely spaal, or quantave denion; never does he attempt to sound their special,
54 Ku, J cit. 55 R Latore, eIl ofHm, Chcago, 1 9 5 1 pp 39f. 5 6 For the following cf. B Snell, e D ofthe Mnd, Hr Torchbs, 1960 Chapter . Snell's iews have been crczed bt eem to sue the crticsm. Cf. the re n F. K e Unthung zu H und H mn, He 6 Gnn, 193, pp. 25 In h Gam Sch, Gtnn, 1966 p. 18 Sne ao argues that in Homer we never nd a pena decison, a consious choice made by an acng human beng A humn beg who s
faced wth various ssbies never thnks "now t depends on me, t depends on hat I decde to do.'
182
AGAI NST METHOD
non-physical nature' 57 Acons are iniated not by an autonomous I', but by futher acons, events, occurrences, including dvine interference. And this is precisely how mental events are peced 58 Dreams, unusual psychological feats such as sudden remembering, sudden acts of recoion, sudden increase of vital ener, durig battle, during a senuous escape, sudden ts of anger are not only laned by reference to gods and demons, they are also lt as such Amemnon's dream listened to his [Zeus'] words and descended' (I, 216)- the dream descends, not a gure in it and it stood then beside his [Agamemnon's] head in the likeness of Nestor' (I, 220) One does not he a dream (a dream is not a subjecve' event), one se it (it is an objecve' event) and one also sees how it approaches and moves away 59 Sudden anger, ts of sength are described andlt to be divine acts 60 Zeus buids up and Zeus diinishes sength in the way he pleases, since power is beyond all others' (I, 2021) is not just an objecve descripon (that may be extended to include the behaviour of animals) it aso expresses theelng that the change has entered om the outside, that one has been lled with song courage' I, 13.60) Today such events are either forgotten or rerded as purely accidental 61 But for Homer or for early thought in general, there no such thng as accident 2 Eve event is accounted for T makes the events clearer, sengthens their objecve features, moulds them into the shape of known gods and demons and thus tus them into powerful evidence for the divine apparatus that is 57 Sne, Gammelte Sch, p 18 58 Cf. ds, The Gree and the Iatinal, Boston, 1957 Chapter . 59 Wth some eo ths experence can be repeated even today Step 1 e do, close your eyes, and aend to your hypnagogc halucnaons. Step 2: pemt the
hallucnaons to preed on ther own and accordng to ther own tendences. wll then chan from events in front ofthe eyes into events that grdualy surround the ewer but wtout yet making hm an acve partcpnt of an acon in a ee dimensional dreamspace. Step 3: swtch over from ving the haucnatory event to ingpa of a complex of rea events which act on the ewer and can be aed un him Step 3 can be reversed either by the act of an amost nonestent w or b outsde noise The threedmensional scenery becomes twodimensona, together into an area in front ofthe eyes, and moves away It would be interesng to s how such al elements change from culture to culture 60 Today we say that somebody s overcome' by emoons and he may feel i anger as an alen thing that invades him aganst his wi The daemonic ontolo of the Greeks conans obecve temnolo for descrbng ths feature of our emoon a the stail it 61 Pchoanayss and related ideologes now again conbute o makng sc
events pa of a wder context and thereby lend them subsanalty. 62 Dodds, op ct., p. 6
SIXTEE
183
used for expanng them The gods are present. To recoge ths a a gven fact for the Greeks s the rst condon for comprehendng ther relgon and ther culture. Our knowledge of ther presence rests upon an (nner or outer) experence of ethr the Gods themselves or of an acon of the Gods. '6 3 To sum up the archac world s much less compact than the world that surrounds us, and t s also experenced as beng less compact Archac man lacks physcal' unty, hs body' conssts of a multude ofparts, lmbs, surfaces, connecons and he lacks menta' unty, hs mnd' s composed of a varety of events, some of them not even mental' n our sense, whch ether nhabt the body-puppet a addonal constuents or are brought nto t from the outsde. Events are not shaped by the ndvidual, they are complex arrangements of parts into whch the body-puppet s nseed at the appropriate place. 64 Ths s the world-vew that emerges from an analyss of the al features of archac' art and Homerc poey, taken n conuncon wth an analyss of the con which the Homerc poet used for descrbng what he perceves. Its man features are 63 Wilamowi-Mllendo, D Gue He, I , 1955, p 17 Our concepons of the wold subdiide otheise unifo tei d ceate dierences in perceived brghess where obecve brghess has no gradient e same press is resnsible for the orderng of the rather chaoc impressions of our inner life, leadng to an inner) percepon of diine inteerence, and it may even noduce daemons, gods, sprtes into the domain of outer percepons At any rte there is a sucent number of daemonc experences not to reect this conecture out of hand 64 Ths means that success is not the result of an eo on he pa of the indiidual but the founate tng together of circumsances This shows iself even in words like QT which seem to designate iti In Homer such words emphasize not much the eect of the agent as the fact that the resul comes about in the rght way, hat the press that brngs it about ds not encounter t many disturbances i s into the other presses that surround it (n the Atc diaect EUT sll means I am doing well') Similarly XEV emphasizes not much a personal achievement as the fact that things go well, hat they t into their surroundings The same is ue of the acquision of knowled Odysseus has seen a lot and experenced much, moreover, he is the OJX
� £oao6m:
184
AGAINS T METHOD
ed by the indivduals using the concepts. e indid le ined in the same nd ofr that is consted ther a.
Futher evdence for the conjecture can be obtaied om an eaon of 'meta-attudes such as gene regius attudes and 'theories of (attudes to) knowledge For the lack of compacess just describd reappears in the ed of ideoo There is a tolerance in relgious tters wch atr generaons found moay and theorecaly unacceptable and whch even today is reded as a mfestaon of frivolous and simpe ids 65 chaic is a regious eclecc, he ds not object t forei gods and myths, he adds them to the esg futue of the word without any attempt at synthesis, or a removal of conad ons There are no priests, there no doa, there are no categorical statements abut the gds, huans, the wod toleace ca s be found with the Ionian phosophers of natu who develop thei ideas side by side with myth without g o elmnate the latter) There no regious 'morty in ou sense, n are the gods absact emboents of eteal pincipes 67 T th became later, duing the archac age and as a result they 'lost [thei] huty Hence Olypianism in its moralized form tended o become a religion of fear, a tendency whch reected n th regous vabu ere is no word for "o-feg in th 68 hs s how e was dehuazed by what some people ar pleased to cl 'moa progress or 'sienc progress Simlar remarks apply to the 'theo of knowledge tat s imp in ths early world vew. The Muses in , 2284, have knowle because they are ose to thigs they do not have to rely on ours and because they know all the many thigs that are of iterest to th witer, one aer the other 'Quanty, not intensity Homes standard of judgement and of knowledge,69 as becomes clar from such words as OAU
Eape: F Schacheayer, , S 9 C WaowMendo, op t MP Nsn, , ord, 949 p 52. odds, op t, p 35 Sne, , p 8 Hercs, er ogenes aerus, I,
SIXTN
185
laned n ts own parcular way and wthout the use of unversal prncples, perssts n the coastal descrpons of the th and th (and later) centures (whch smply umerate the trbes, bal habts, and coastal formaons that are successvely met durng the uey), and ven a thnker such as Thales s sased wth makng many neresng obseaons and provdng many explanaons wthout tng to e them together n a system 7 (The rst thnker to consuct a system' was Anamander, who followed Hesod) nowledge so conceved s not obtaned by yng to grasp an essence behndthe reports of the senses, but by ( 1 putng the obseer n the rght poson relave to the obect (press, aggrete), by nserng hm nto the approprate place n the complex patte that constutes the world, and (2) by addng up the elements whch are noted under these crcumstances It s the result of a complex suey carred out from sutable vantage ponts One may doubt a vague report, or a fth-hand account, but t s not possble to doubt what one can clearly see wth one's own eyes The oje depcted or descrbed s the proper arrangemets of the elements whch may nclude foreshort enngs and other perspectod phenomena 72 The fact that an oar looks broken n water lacks here the scepcal force t assumes n another deolo 73 ust as Achlles stng does not make us doubt that he s swft-footed as a matter of fact, we would start doubng hs swfess f t tued out that he s n prncple ncapable of stng n the ve same way the bent oar does not make us doubt that t s perfectly saght n ar as a matter of fact, we would start doubg s straghess f t dd not lok bent n water 74 The bent oar s not The idea that Thales used a principle expressing an underlyng unt of naral phenomena and tha he idened this principle with water is rst found in Aristoe, Metapsi, 983b 12 and 26 A closer look at ths and other pasges and consultaon of Herodotus suggests that he sll longs to the group of those thinkers who dal with numerous exraordnary phenomena, wthout ting them together n a syste Cf the iid presentaon in F Krat, Gchichte Natusscha, I, Freiburg, 1 971 Chapter 3 72 Perspectoid phenomena are sometmes ated as f they were special properes of the objects depicted For example, a container of the Old Kingdom (Ancent Ept has an ndentaton on top, indicatng perspecve, but the indentaon s presented as a feature of the objet itself, Schfer, op cit, p 266 Some Greek arsts to nd situaons where perspectve ds not need to be considered Thus the peculiarit of th socalled redre stle that arises n about 530 BC ds not so uch consst in the fact that foreshoenngs are drawn, but in the new and hghly vare was to circumvent them', E Pfuhl, Malei und Zeichnung Gech, Vol I, Munich, 1 92 3, p 378 .
3 Cf the discussion in Chapter I of AJ Aer's Fountions o Empical nowledge The example was familiar in anquit 4 This s also the way in which JL Ausn takes care of the case Cf Sse and
186
AG AIN ST METHO
an pe that denies what another pe says about the nure of the oar, it is a parcular pa (situaon) of the real oar that is not only cple with its saighess, but that demands it: the objects of knowledge re as addive as the visible lists of the rchic arst and the situaons described by the archaic pt. Nor is there any unform concepon of knowledge 75 A great variety of words is used for expressing what we today red as dierent fom of knowledge, or as dierent ways of acqug knowledge O<(a 76 means experse in a ce professon (cener, singer, general, physician, charioteer, wrestler) inclug the arts (where it praies the arst not as an outstanding creator but as a master of hs craft); EtbEVaL, literaly 'having seen, rers to knowledge ined from pcon V, especia the Il, though often anslated as 'teg or 'undestang, songer, it contans the idea of following and obeyg, on abso somethig and acts in accordance with it (haing may play an important role) And so on Many of these eressios en recepve attude on the part of the knower, he, as itwere1 acts out th bhavour of the thigs around h he follows them, 7 he acs as bets an enty that is inserted at the place he cupies To repeat and to conclude: the modes of rpresntaon usd duing the early archaic period in Greece are not ust reecons o incompetence or of special arsc interests, they giv a fai account of what are felt, seen, thought to be fundamntl feas of the word of archaic man This world is an open world. Its elemn are not formed or held together by an 'underlyng substance, are not appeances from which t substance may be errd wi diculty ey casionaly coalesce to form assembags relaon of a single element to the assemblage to which it belongs like the relaon of a part to an aggrete of parts and not like relaon of a part to an oveoweing whole. The parcuar aggr caled 'man visited, and casionaly nhabited y 'mental een Senba, Ne ork, 962 It is clear that problems such the prob o estence o theoretca enes' cnnot ae under these ces eth these problems are ed b the ne approach that superseded the adde do of rchc nd pre-archc mes 75 B Sne, Die r B Ws n Phsh, Bern, 924 sho account is gen Sne, D Jr, pp 4 Cf a von F, Phihie und sprhlh Dot, P oe, LeipzigPariLondon, 1938 76 On urrence Homer, Il, 5, 42, conceg te o caenter an expe caenter' ansltes Latmore) 77 Cf Sne,, p 50
SIXTEEN
187
Such events may resde n hm, they may also enter from the outsde Lke eve other obect man s an exchange staon of nuences rather than a unque source of acon, an I' (Descartes cogto' has no pont of ttack n ths world, and hs argument cannot even start) There s a geat smlarty between ths vew and Mach's cosmolo except that the elements of the archac world are recognable physcal and mental shapes and events whle Mach's elements are more absact, they are as yet unknown a of research, not ts oe In sum, the representaonal unts of the archac world vew admt of a realsc nteretaon, they express a coherent ontolo, and ors obseaons apply. At ths pont I nterrupt my argument n order to make some comments whch connect the precedng obseaons wth the problems of scenc method 1 It may be obected that foreshortengs and other ndcaons of perspecve are such obvous features of our perceptual world that they cannot have been absent from the perceptual world of the Ancents The archac manner of presentaon s therefore ncom plete, and ts realsc nteretaon ncorrect Reply: Foreshortenngs are not an obvous feature of our percep tual world unless specal attenon s drawn to them (n an age of photography and lm ths s rather frequently the case) Unless we are professonal photographers, lm-makers, panters we perceve thn, not pes Movng swftly among complex obects we noce much less change than a percepon of aspects would permt Aspects, foreshortegs, f they enter our conscousness at all, are usually suppressed ust as after-mages are suppressed when the approprate stage of perceptual development s completed 78 and they are noced n specal stuaos only 79 In ancent Greece such specal stuaons arose n the theae, for the rst-row vewers of the mpressve producons of Aeschylus and Atharchos, and there s ndeed a school that ascrbes to the theae a decsve nuence on the development of perspecve 80 Besdes, why should the perceptual world of the ancent Greeks concde wth ours? It needs more arument than reference to a non-estent form of percepon to consoldate the obecon.
C fooote 12 and text of the present chapter 9 C fooote 13 0 Cf Pa II of Hedwig Kenner, D eat und Realmu in Gechch ut, Vienna, 1954, especiay pp 121 f
188
AGANS TOD
2. The procedure used for establishing the peculiaries of the
archaic cosmolo has much in common with the method of an anthropologist who examines the world-vew of an associaon of trbes The dierences are due to the scarcty of the evidence and to the parcular crcumstances of its origin (written sources; works of art; no personal contact). Let us take a closer look at ths predure! A anthropologist ying to discover the cosmolo of his chosen ibe and the way in which it is mrrored in language, i the arts, in daily life, st leas the language and the basic social habits; h inquires how they are related to other acvies, including such ma fe unimportant acvies as milkng cows and cooking meals; 8 h ies to iden key ideas. 82 His attenon to minuae is not th result of a msguided urge for completeness but of the realzaon that what looks insiicant to one way of thinking (and perceiving) a play a most mportant role in another (The dierences beeen th paper-and-pencil operaons of a Lorentzian and those of an Einsteinian are oen minue, if disceible a all; yet they reect a major clash of ideologies.) Having found the key ideas the anthropologist ies to untand them Ths he does in the same way in which he originaly ined an understandng of his own language, including the language of th special profession that provides him wth an income He ntl the ideas so that their connecons are rmly engraved n his memo and his reacons, and can be produced at will. The nave siety has to be in the anthropologist himself and not merely in his notebks f he is understand it' 83 Ths process must t eeom ta nteerce. For example, the researcher must not to get a bette hold on the ideas of the tribe by likening them to ideas he alread knows, or nds more comprehensible or more precise On n account must he attempt a logical reconsucon'. Such a predue would e him to the known, or to what is preferred by cetain groups, and would forever prevent him frm grping the unknown wold view he is examning Having completed his study, the anthropologist carries wit himself both the nave society and his own background and he ma now start comparng the o The comparison decides whether th nave way of thinkng can be rproduced in European te (provided there is a unique set of European terms'), or whether it a logic' of its own not found in any Weste language . In the cou 81 EvansPrtchard, SAnthol, New York, 1 965, p 80 82 ibid, p 80 83 ibid, p 82
SIXTEEN
189
of the comparison the anthropoogist may rephrase certain nave ideas in Engish This ds not mean that Engish a spok nt ofthe compason aready contans nave ideas It means that anguages can be t in many diecons and that understnding does not depend on any parcuar set of rues. 3 The eamaon of key ieas passes through vaious stages, none of which eads to a compete caricaon Here the researcher must exercise rm cono over his urge for instant carity and ogica perecon He must never y to make a concept cearer than is suggested by the materia (except as a tempora aid for futher research. It is this materia and not his ogica ituion that determines the content of the concepts To ake an empe The Nuer, a Nioc ibe which has been examed by EvansPritchard, have some interesg spaotempora concepts 4 The researcher who is not oo famiar with Nuer thought w nd the concepts 'uncear and insucienty precise To improve matters he might expicag them, using mode ogica noons That might create cear concepts, but they woud no onger be Nuer concepts If, on the other hand, he wants to get concepts which are both cear and Nuer, then he mus keep his key noons vague and incompete untl the ght ation com along, ie un ed study tus up the missing eements whch, taken by themseves, are ust as uncear as the eements he has aready found Each item of infoaon is a buiding bk of understanding, which means that it has to be caried by the discove of futher bks from the anguage and ideoo of the ibe raher than by premature denions Statements such as ' the Nuer cannot speak of me as though it was something actua, which passes, can be waited for, can be saved, and so foth I do not think that they ever experience the same feeing of ghg against me, or of having to coordinate acvies with an absact passage of me, because the points of reerence are mainy the acvies themseves, which are generay of a eisurey character 5 are either buiding bks - in this case their own content is incompete and not fuy understd or ey are preimina attempts to ancipate the arrangement of the tot aity of a bks. They must then be tested, and eucidated by the disco ve of futher bks rather than by ogica caricaons (a chid eas the meaning of a word not by ogica caricaon but by 84 Evans Prtchard e N Oxford 9 Pa III cf aso the bref acount n /Anthl pp 02 85 he N p 03
190
I N S T M E T H OD
realizing how it goes together with things and other words) Lack of clarity of any parcular anthropological statement reects the scarcity of the material rather than the vagueness o the logical intuions of the anthropologist, or of his ibe. 4 Exactly the same remarks apply to any attempt to explore important mode noons such as the noon ofincommensurabiity Within the sciences incommensurability is closely connected with meaning. A study of incommensurability in the sciences wil therefore produce statements that contain meaning-erms but these terms will be only incompletely understood, ust as the trm me' is incompletely understd in the quotaon of the precedin g paragraph. Thus the remark that such statements should be mad only producon of a clear theo of meaning86 is as sensible a the remark that statements about Nuer me, whch are the material that le to an understanding ofNuer me, should be written dow ony after such an understanding has been acheved 5 Logicians are liable to object They point out that an exama on of meanngs and of the relaon beeen terms is the task of loc, not of anthropolo Now by logic' one may mean at least o dierent thngs Logic' may mean the study of, or results ofthe study of, the suctures inherent in a certain type of discourse. And it y mean a parcular logical system, or set of systems A study of the rst kind belongs to anthropolo. For in ordr to see, for exple, whether AB v A = A is part of the logic of quantum theo' we shall have to study quantum theo And a quantum theo is not a divine emanaon but a human product, we shal have to study it in the form in which human products usually are avaiable, that is, we shall have to study historical records textb, original papers, records of meengs and private conversaons, letters, and he like (In the case of quantum theo our posion improved by the fact that the ibe of quantum theorecians has nt yet died out Thus we can supplement historical study w anthropologcal eld work such as the work of Kuhn and collaborators 87 ) 86 chinsten Minnta Studi in the Phil fSe Vol 4 Minneal 970 p 224 sas that Feerabend owe[s] us a theor of meaning and Hempe prepared to accept incommensurabilit onl the notion of meaning involved n has been made clear op cit p 56 87 Re in TS Kuhn JL Heilbron P Forman and L len Sur fr H fQuanum Pi merican Phiosophica Siet Philadelphia 967 e matera assembled under the progrmme described in this re can be consule various universes the Universit of Cafoia in Berkele among them
SIXTEE
191
It is to be admitted that these records do not, by themselves, produce a unique soluon to our problems But who has ever assumed that they do? Historical records do not produce a unique sol uon for historical problems either, and yet nobody suggests that they be neglected There is no doubt that the records are necsa for a logical study in the sense examined now The queson is how they should be used. We want to discover the sucture of the eld of discourse, of which the records give an incomplete account We want to lea about it without changing it in any way In our example we are not interested i whether a peed quantum mechanics of the ture employs AB v AB = A or whether an ition of our o, whether a little bit of 'reconsucon which changes the theo so that it confos to some preconceived prnciples of mode logic and readiy provides the answer employs that pinciple We want to know whether quantum theo a aual praised physists employs the pinciple For it is the work of the physicists and not the work of the reconsuconists we want to examine And this work y wel be l of conadicons and lacunae Its 'logic (in the sense i which I am now using the term) may well be 'ilogical when judged from the point of view of a parcular system of formal logic Putng the queson in this way we reaize that it may not admit of any answer There may not est a single theo, one 'quantum theo, that is used in the same way by all physicists The dierence between Bohr, Dirac, Feynman and von Neumann suggests that this is more than a distant possibility To test the possibility, ie to either eliminate it or to give it shape, we must examine concrete cases Such an examinaon of concrete cases may then lead to the result that quantum theorecians dier from each other as widely as do Catholics and the various types of Protestants: they may use the same texts (though even that is doubul ust compare Diac with von Neumann), but they sure are doing dierent thigs with them The need for anthropological case studies in a eld that inialy seemed to be dominated by a single myth, always the same, always used in the same manner, indicates that our common knowledge of science may be severely defecve It may be enrely mistaken (some mistakes have been nted at in the preceding chaptes) In these circumstances, the only safe way is to confess iorance, to abandon reconsucons and to start studying science from scratch We must approach science lke an anthropologist approaches the menta contorons of the medicine-men ofa newly discovered assiaon of ibes And we must be prepared for the discove that these contorons are wdly ilogical (when judged from the point ofview of
192
G IST METHOD
a parcular system of formal logc) and he to e wldly llogcal n order to funcon as they do 6 Only a few phlosophers of scence nteret 'logc n ths sense, howeer Only few phlosophers are prepared to oncede that the basc suctures that underle some newly dscoered dom might der radcally from the basc suctures of the more famlar systems of formal logc and absolutely nobody s ready to admt that his mght be ue of scence as well Most of the me the 'logc (n the sense dscussed so far) of a parcular language, or of a theo, mmedately dened wth the features of a parcular logcal system wthout consderng the eed for an nqu conceng the adequacy of such an dencaon Professor Gedymn, for example, 88 means by 'logc a faourte system of hs whch s farly comprehense, but by no means all-embracng (For example, t does not contan, nor could t be used to formulate, Hegels deas And there hae been mathemacans who hae doubted that t can be used for expressng nformal mathemacs A logcal study of scence as Gedymn and hs fellow logcans understand t s a study of sets of formulae of system, of ther sucture, the properes of ther ulmate con stuents (nenson, extenson, etc), of ther consequences and of possble models If ths study does not repeat the features an anthropologst has found n, say, scence then ths ether shows that scence has some faults, or that the anthropologst does not know any logc It does not make the slghtest derence to the logcan n second sense that hs formulae not look lke scenc statements, that they are not ed lke scenc statements and that scence coud not possbly grow n the smple ways hs ban s capable of understandng (and therefore regards as the only permssble ways) He ether does not noce the dscrepancy or he regards t as beng due to mperfecons that cannot enter a sasfacto account Not once does t occur to hm that the 'mpefecons mght hae pose funion, and that scenc progress mght be mpossbe once they are remoed For hm scence is aomacs plus mode theo plus correspondence rules plus obseaon language Such a predure assumes (wthout nocng that there s an assumpon noled) that an anthropologcal study whch famlar izes us wth the oert and the hdden classcaons of scence been completed, and that t has decded n faour of the ao (etc etc) approach No such study has eer been carried out And the bts and peces of eld work aalable today, manly as the result of 88 Bth Jual f the Philo f S ust 970 February 97 pp 39
pp 257 d
SIXTEE
193
the ork of Hanson, Kuhn, Lakatos and the numerous historians who remained untouched by posivisc prejudices, show that the logician s approach removes not just some inessenal embroideries ofscience, but those ve features whch make scienc progress and thereby science possible 7 he discussions of meaning I have aluded to are another ilusaon of the deciencies of the logicians approach For Giedymin, this term and its derivaves, such a the term 'incommen surabiity, are 'unclear and insuciently precise. I agree Giedymn wants to mae the tes clearer, he wants to understand them better. A agreement He ies to obtain the clarity he fees is lackig by explicaon in terms of a parcular kid of formal logic and of the double language model, resicg the discussion to 'intension and 'extension a explained in the chosen logic It is here that the disagreement starts For the queson s not how 'meag and 'incommensurabilty cur withi a parcular logical ystem he queson is what role they play in actual, ie non-reconsucted) science Claricaon must come from a more detaied study of ths role, and lacunae must be lled with the results of such sudy And a the lling takes me the key terms will be 'unclear and isuciently precise for years and perhaps decades See also items 3 and 4 above.) 8. Logicians and phosophers of science do not see the situaon in this way Being both unwilling and unable to car out an informal discussion, they demand that the main terms of the dscussion be 'claried. Ad to 'clari the terms of a discussion does not mean to study the tonal and as yet unknown properes of the doman in queson whch one needs to make them fully understd, it means to ll them with istng noons from the enrely dierent doman of logc and common sense, preferably obseaonal ideas, un they sound common themselves, and to take care that the press of lling obeys the accepted laws of logic he discussion is permitted to preed only a its inial steps have been modied in hs manner So the couse of an inveson is deected into the narrow channels of things already understd and the possibity of fun daental conceptual discove or of fundamenta conceptual change) is considerably reduced Fundamenta concepual change, on the other hand, presupposes new world-views and new languages capable of expressing them. Now buiding a new world-view and a correspondig new language, is a press that takes me, in science as well as in meta-science he terms of the new langage become clear only when the press is faly advanced, so that each single word is the cene of numrous ines connecg it wth other words,
194
GI ST METHOD
sentences, bts of reasonng, gestures whch sound absurd at rst but whch become perfectly reasonable once the connecons are made Arguments, theores, terms, ponts of view and debates can therefore be claried n at least two derent ways (a) n the manner already descrbed, whch leads back to the famlar deas and eats the new as a specal case o thngs already understood, and (b) by ncooraon nto a language of the future, whch means that one must lea to a with unlaned tes and to use stcesr which no cear l ofusage are a yet ailale Just as a chld who starts usng words wthout yet understandng them, who adds more and more uncomprehended lngusc fragments to hs playful acvity, dscovers the sense-gvg princple only aer he has been acve n ths way for a long me the acvity beng a necessa presupposon of the nal blossomng foth of sense n the ve same way the nventor of a new world-view (ad the phlosopher of scence who es to understand hs predure) must be able to talk nonsense unl the amount of nonsense created by hm and hs frends s bg enough to gve sense to all ts parts There s agan no better account of ths process than the descripon whch ohn Stuart lhas left us of the vicsstudes ofhs educaon Referring to the explanaons whch hs father ve hm on logical matters, he wrote 'The explanaons dd not make the matter at clear to me at the me but they were not therefore useless tey remaned as a nucleus for my obseaons and reecons to cstallse upon the mport of hs general remarks beng ntereted to me, by the parcular nstances whch came under my noce aar 89 Buldng a new language (for understandg the world, or knowledge) s a press of exactly the same knd t that the nial 'nucle are not given, but must be nvented We see here how essenal t s to lea talkng n riddles, and how dsasous an eect the drive for nstant clarty must have on our understandng addon, such a drve beays a rather narrow and barbarc mental 'To use words and phrases n an easy gong way wthout scrunzng them too curiously s not, n general, a mark of l breedng on the cona, there s somethng low bred n beng too precse All these remarks are rather vial and can be llusted
'
89 There is much more randomness in this press than a rtionalist would e permit or suct or even noce Cf von Kleist ber die lmhliche Vefen der Gedanken beim Red en Metke Dch Ltatutk ed Hans Me Stuga 962, pp 747 Hegel had an ining of the situation Cf K Lwth an Redel (eds) Hel Studauge I Frnkfu 968 p 54 For Mill cf Chapter fooote 3 . Plato Theatet 84 Cf also I Attel Heidelberg 966 p 79, cricing ristotes demand for instant precision
SIXTEE
195
by obvous examples Classcal logic arrved on the scene only hen there was sucent argumentave materal (n mathemacs, rhetorc, polcs) to see as a starng pont and a a tesng ground Arthmec developed without any clear understandng of the concept of number such understandng arose only when there ested a sucent amount of arthmecal facts' to give it substance In the same way a proper theo of meanng (and of ncommen surablty) can arise only after a sucent number of facts' has been assembled to make such a theo more than an exercse n concept pushng Ths s the reason for the examples n the present secon 9 There s sll another dogma to be consdered before retung to the man narraon It s the dogma that all subjects, however assembled, qute automacally obey the laws of logic, or ought to obey the laws of logic If ths s so, then anthropological eld work ould seem to be superuous What s ue n logic s ue n psycholo n scenc method, and n the histo of scence,' wrtes Popper91 Ths dogmac asseron s nether clear nor s t (n one ofts man nteretaons) ue To start with, assume that the expressons psycholo', histo of scence', anthropolo' refer to certain domans of facts and regularies (of nature, of percepon, of the human mnd, of sety) Then the asseron s not ear a there s not a sngle subject LOGIC that underles all these domans There s Hegel, there s Brouwer, there are the many logical systems consdered by mode consucvsts They oer not just derent nteretaons of one and the same bulk of logical facts', but derent facts' altogether And the asseron s not te a there est legmate scienc statements which volate smple logical rules For example, there are statements whch play an mportant role in establshed scenc dscplnes and which are obseaonally adequate only f they are self-condcto xate a movng patte tat has jus come to a standsll, and you wll see t move n the opposte drecon, but wthout changing ts poson The only phenomenologically adequate descrpon s t mes, n space, but t does not change place' and ths descripon s self-conadcto92 9 ee Knwledge Oxford 972 p 6 ncipated eg b Comte Cue 52° e on and of course ristote 92 It has been objected (er GEL Owen) that we are deaing wih appearances n?t th actual events and that the correct descripon is it appears to move But the dcut remains For ifwe inoduce the appear we must put it at the beginning ofthe entence which will read it appears that it moves and ds not chang place nd a�earances beong to the domain ofphenomenologica pscholo we have made our pont z that this domain contans self-inconsistent elements
19
IST METH OD
ere are examples from geoey: 93 thus the enclosed (whch need not appear in the same way to eve person) is seen as an isoseles iangle whose base not hved by the peendicular. there are exmples with a = b b = c a � c as the n phenomenologicaly adequate descripon. 94 Moreover, there n a single science or other form offe that use progressive as wl as in agreeent with logical demands Eve science con theories whch are inconsent both with facts and with oth theories and whch reveal condicons when analysed in de Ony a domac belief in the principles of an alegey ur disciple 'Logic w mke us disregrd hs situan d th objecon that logical principles and prinples of say arim dier from empirical principles by not being accessible to the meth of conjectue and retaons (or for that matter, any oth 'empirical method) has been defused by ore recent research in eld95 Secony let us assue that the expressions 'psycholo 'ano polo 'hso of science 'physics do not refer to facts and l but to certan metho of assebling facts including ce w of connecg obseaon with theo and hypothes That is us coider the ai 'science and its various subdivsions we ay lay do ial man of knowledge and knowle acquision and we ay y to consuct a (sial) mchne obeys these demds. Almost all epteologists and phosophers science preed in ths way Occasionaly they succeed in nding 93 E Rubin Visua Fires ppeny ncompable wh eome P VI 950 pp 365 Cf aso he drwngs on pages 67 94 E Trnekjaer-Ramusen Perspeod isces A P 955 p 297 95 Mny b he work ofe t Proofs and Refuaon BthJu the Ph S 9663
SIXTEEN
197
achine that ight work in certain ideal condions but they never inquire or even nd it woth inquiring whether the condions are sased n this real world of ours Such an inqui on the other hand will have to explore the way in which sciensts deal ith their surroundings it will have to exaine the actual shape of their product viz 'knowledge and the way in which this product changes as a result of decisions and acons in coplex sial and aterial condions In a word such an inqui wil have to be anthropological There is no way of predicng what an anthropologica inqui will bring to ligh. In the preceding chapters which are rough sketches of an anthropoogical study of parcular episodes it has eerged that science is full of lacunae and condicons that iorance pgheadedness reliance on preudice lying far fro ipeding the foard arch of knowledge ay actually aid it and that the tradional vrtes of precision consistency 'honesty respect for facts au knowledge under given circustances, if pracsed with deterinaon ay bring it to a standsl It has also eerged that logcal principles not only play a uch saller role in the (arguenave and non-arguentave) oves that advance science but that the attept to enforce the would seriously ipede science (One cannot say that von Neuann has advanced the quantu theo But he certainy ade the discussion of its basis ore long winded and cubersoe 96) Now a scienst engaged in a certain piece of research has not yet copleted al the steps that lead to denite results His future is sll open Will he follow the barren and illiterate logician who preaches to hi about the virtes ofclarity consistency experiental support (or experiental falsicaon) ghess of arguent 'honesty and so on or will he iitate his predecessors in his own eld who advanced b breaking ost of the rules logicians want to la on hi? Will he rely on absact inuncons or on the results of a study of concrete episodes I hink the answer is clear and with it the relevance of anthropological eld work not ust for the anthropologists but also for e ebers of the siees he exaines I now connue y arraon and proceed to describing the ansion fro the paratacc iverse of the archaic Greeks to the substance-appearance universe of eir followers 10 96 Besdes the mprecisions which he removes from the formasm now reappear
the relaon between theory and fat Here the corresndence princple s reigns supreme Cf fooote 25 of Chapter 5
19
S MEH OD
The archaic cosmoo (whch from now on I sha ca cosmoo A� contains thigs, events, their parts it does not contain appearances 9 Compete kowedge ofan obect is compete enumeraon of its parts and pecuiaries Humans cannot have compete knowedge er are too many thigs, t many events, too many situaons (, 2. ), and they can be cose to ony a few of them (l, 2) But athough humans cannot have compete knowedge, they can have a szeabe amount of it The wider ther experience, the greater the number of adventures, of things seen, heard, read, the greater thei knowedge 98 The new cosmoo (cosmoo B) that arises in the th t th centuries BC dishes between much-knowig, OAJ6(, and ue kowedge, and it ws against us 'cusm bo of manifod experience, £eo� OAUEQOV Such a discn and such a wag make sense ony in a word whe sucture diers from the sucture ofA. In one version which paed a ae roe in the deveopment of Weste cion and whih underies such probems a the probem of the estence oftheore enes and he probem of aienaon the new events fom what e might ca a T Wor, whe the events of eveday ie are aearan that are but its dim and miseag reecon.0 e True Word simpe and coherent, and it can be described in a uiform way. So can eve act by which its eements comprehended: a few absact noons repace the numerous cone that were used in cosmoo A for describing how humans migt e 'inserted into their surroundings and for expressing the eq numerous types ofinformaon thus gained. From now on there one important type of infoaon, and that is owledge. 97 Sne, Ausdke, p 28 referring to omer), speaks of a knowed preed om appeanc nd draw their mutude together a ut whch sited as their ue essence' ma appy to the resracs, it ds not a omer n the case of omer the word is comprehended as the sum ofngs, space, nd not as reon ang inteey' bd., p 67, dussg mpedo the es foowng the uotaon for a rther eaboraon of the theme). 98. Sne, e alt Ge und Wr, p 48 99. f eractus, fr. ienz) 1. edes, 7, 3 ere for the rst me sense and reason are cone WK. Guthre, A Ht fGreek Phs, Vo , ambrdge, 1965, p. 25. I 0 I is dsncon is chctersc of cen mythoogica vew as we. o thus diers both om the preceeding mythoogies and from he succe phisopies is nt of vew is of great orn n he 2 cenur J .L. has deeoped simar deas nd he has crized he deveopment from ato to he present essenm f. he st chapter of Se and Sbl h of Fare e cont detais
STN
199
The conceptual totalitarianism hat arses as a result of he slow arrival of world B has interesng consequences, not all of hem de sirabl e. Situaons which made sense when ed to a parcular type ofconion now become isolated, unreasonable, apparently inconsis tent wih oher situaons: we have a 'chaos of appearances The 'chaos is a direct conseuence of he simplicaon of language hat accompanies he belief in a True World 1 2 Moreover, all he manifold abilies of he obseers are now directed towards his True World, hey are adapted to a un aim, shaped for one pailar puose, hey become more similar to each oher which means hat huans become impoverished togeer wih heir language They become impoverished at precisely he moment hey discover an auonomous 'I and preed to what some have been pleased to call a 'more advanced noon of God (allegedly found in Xenophanes), which is a noon of God lacking he rich variety of typcally human features. 1 03 'Mental events which before were eated n analo � wih events of he body and which were eced cording 1 become more 'subjecve, hey becoe modicaons, acons, revelaons ofa spontaneous soul he disncon between appearance (rst impression, mere opinion) and reality (ue nowledge) spreads evewhere. Even he k of he arst now consists in arrangng his shapes in such a manner hat he underlyng essence can be grped wih ease In painng t leads to he development of what one can only call systemac mehods for deceiving he eye he archaic arst eats he surface on which he paints as a writer might eat a piece of papus; it a real surface, it is supposed to be se as a real surface (hough attenon is not always directed to it) and he ars he draws on it are comparable to he lines of a blueprint or he letters of a word They are sybols hat i he reader ofhe stureofthe oe, ofits parts, of he way in which he parts are related to each oher. The smple drawig overleaf, for exaple, may represent hree pahs meeng at a point. The arst using perspecve on he oher hand, egads he surface and he mars he puts on it as stimuli hat ggr he usion of an arrangement of hree-dimensional objects. The illusion curs because he human mind is capable of producing illuso experiences when properly smulated. The drawing is now seen 102 Snell, Aud, pp 8 von Fr, Phhe und rhl Aud Dt P undA, LeipgaLondon, 1938, p. I I I 03 '. . n bcoing the emiment of cosic use los hu. ce Olympansm n its moed form tended to become a ion of fear . . ', ds, Gree, p 35. For enophnes cf. hapter 2 of Fare t e 1 04. Snell, D, p. 69.
�
200
A G A N S T M T O D
eher as he coer ofa cube hat etends towards he vewer, or as he coer ofa cube hat ponts away from hm (and s seen from beow), or ese as a pane oang above he surface of he paper caryng a o dmensona drawg of three paths meeng.
Combng ths new way of seeng wh he new concept of knowedge hat has just been descrbed, we obta new enes, physca objects as hey are understd by most contempora phosophers. To epan, et me agan take he case of he oar. In he archac vew 'he oar s a compe conssng ofparts some of whch are objects, some stuaons, some events. It s possbe to say 'he saght oar s broken (not 'appears to be broken just as t possbe to say 'swft-footed Aches s wakg sowy, for he eements are not set anst each oher ey are pa o a paratac aete Just as a aveer epores lparts ofa sange couny and descrbes hem n a 'peregess hat enumerates ts pecuares, o by one, n he same way he student of smpe objects such as o, boats, horses, peope nserts hmsef nto he 'major oar-stuaons, apprehends them n he approprate way, and reports hem n a st of properes, events, reaons And just as a detaled peregess ehaus what can be sad about a couny, n he sae way a detaled t ehausts what can be sad about an object 1 05 'Broken n water beongs to he oar as does 'saght to he hand; t s 'euay rea. cosmoo B, however, 'brken n water s a 'sembance t contri what s suggested by he 'sembance of saghess an d hus shows he basc unustworhness of a sembaces. 1 h concept of an object has changed from he concept of an aggrete of eu-mportant percepbe parts to he concept of an mpercepb I 05. The idea that knowledge conssts n l reaches back far into the Sume
past. f. von Soden, Ltung und Gr SumchBlnch Wc, n edn, Darmsad, 1965. e dierence between Bablonian and Greek mathemat and asonom ies precse n this. e one deveops methods for the presenao n o what we toda ca phenomena' and whch were interesng and reent events n s, wie the other ies to deveop asonom, whie eavng the heave aon ato, ., 530bf L, 818a) 1. enophanes, fr. 34
SIXTEEN
201
essence underying mutude of decepve phenomena (We may ess that the appearance of an object has changed in a simar way, that objects now ook ess 'at than befre.) Consdering these changes and pecuiaries, it s pausbe to assume that the comparison of A and B ntereted thepapants (rather than as 'reconstructed by ogicay we-ained but otherwise iterate outsiders) wi raise various probems In the remander of this chapter ony some aspects of some of these problems wi be discussed Thus I sha barey menon the psychoogica changes that accompany the anson from A to Band whch are not just a matter of conjecture, but can be estabshed by independent research. Here s ric material for the detaied study of the roe of frameworks (menta sets, anguages, modes ofrepresentaon) and the imits ofraonaism. To start with, cosmos A and cosmos B are built from dierent elnts
The eements ofA are reavey independent parts of objects which enter into etea reaons They parcipate in aggregates without changing their intrnsc properes. The 'nature of a parcuar aggregate is determned by its parts and by the way in which the parts are reated to each other. Enumerate thepas in theprer orr andyou he the oje. This appies to physica aggregates, to humans (mnds and bodies), to animas, but it aso appies to sa aggretes such e hnour of a warrior. Te eements of B fa into two casses essences (objects) and appearances (of objects what foows s ue ony of some rather seamned versions of B). Objects (events, etc.) may an combine. They may form harmonious totaies where each part gves meanng to the whoe and receves meaning from it (an eeme case is Parmendes where isoated parts are not ony unrecoizabe, but atoether unthinkabe). Aspects propery combined do not produce ojes, but psychoogica condions for the apprehension ofphantoms wh are but other aspects, and parcuary miseading ones at that (ey ook so convincing). No numeration ofpes is ntil with the oje (probem of inducon) The ansion from A to B thus inoduces new enes and new aons between enes (this is seen ve ceary in painng and statua) It aso chnges the concept and the sef-eperience of ans. An archaic individua is an assembage ofimbs, connecons, tnk, neck head 1 07 (s)he is a puppet set in moon by outside forces 1 07 To be precise, omer does not even have an words for the arms and the legs he speaks of hnds, lower as, upper arms, feet, caves, and thghs Nor is there a comprehensve term for the unk.' Sne, , hapter , fooote 7.
202
AGANST METOD
such as enemies, social circumstances, feelings (whch are described and perceived as objecve agencies see above): 1 08 'Man an open target ofa eat many foes which impinge on h and peneate ve core 09 He is an echange staon of materal and spirtual, but always objecve, causes. And this not just a 'theorecal idea, it a social fact. Man is not only in way, he piu in this way, and he e hmself to be constuted in this manner. He ds not possess a cenal agency of acon, a spontaneous 'I that produces o ideas, feeings, intenons, and diers from behaviour, sial situaons, 'mental events of type A. Such an I is neither menoned nor is it noced. t is nowhere to be found within A. But it plays a ve decisive role withi B Indeed, it is not mplausible to assume that some outstandin ecuaries ofB such as aspects, semblances, abiuty of feeling 1 enter the stage as a result of a sele nee ofse
yJ
consousns
Now one might be inclined to eplain the ansion as follows: ahaic man has a limited cosmolo; he discovered some thigs, he missed others. His universe lacs important objects, his laguage lac important concepts, his peepon lacs mportant suctures. Add the missing elements to cosmos A, the missing terms to laguage A, the missing suctures to the perceptual world ofA, and you obtan cosm B, language B, percepon B. Some me ago I called the theo underlying such a eplanao the 'hole theo or the 'Swiss cheese theo oflanguage (and other means of representaon). According to the hole theo ve cosmolo (eve langua, eve mode of percepon) has szeab lacunae whch can be lled, leng ething ee unchanged. The h theo is beset by numerous dicules In the present case there th dicuty that cosmos B does not conta a single element ofcosmos A Neither comonsense terms, nor phosophcal theories; nth paing and statua, nor arsc concepons; neither region, theological speculaon contan a sngle element of A once th ansion to B has been completed Ths s a histof 1 1 1 Is 08 moons do not sprng sntaneousl from man, but e bestowed on h b the gods,' ibd., p. 52 See also the account earlier n the present chapter 1. bd., p 20. 0. f. Sappho's bier-sweet ros', bd, p . e fact s not easy to establsh. Mn presentaons of A, ncudng soe ver deted and sophscated ones, are infeted b B-concepts. exampe uoted n fooote 97 to the present chapter. ere elsewhere on e nthrologica method can lead to nowledge that s more thn a reecon of thning. A smar stuaon in the course ofndivdua development s derbed e text to fn. 1 2 of the present chapter.
STN
03
fact an accident, or has A some suctua properes that prevent the co-estence of A-situaons and B-situaons? Let us see! I have already menoned an exple that mght give us an g of a reason as o why B does not have room for Afacts: the drawg below may be the intersecon of three paths as presented in accordance with the principles of A-pictures (whch are visual ts). Perspecve having been inoduced (either as an objecve method or as a mental set), it can no longer be seen in ths maer. Instead oes on paper we have the illusion of depth and a three-ensioa panorama, though of a rather simple d. ere is no way of incoorang the A-picture into the B-picture except as part of ths lusion But an ilusion of a vsual lst is not a visual lst.
The situaon becomes moe ansparent when we tu to concepts. I have sad above that the 'nature of an object (=aete) in A is detered by the elements of the aete and the relaon between the elements One should add that ths deteraon 'closed in the sense that the lements and th relaons ttute the objct; whn they are given, then the object given as wel. For example, the 'elements descrbd by Odysseus in hs spech in , 225 constitute honour, grace, respect. A-concpts are thus ve sar to noons such as 'checate: given a ce arrangemen of pieces on the board, there is no way of 'discoverng that the me can s be conued. Such a 'discove would not a p, it would not add to our knowledge of possible chess posions, it would put an end to the game . And so would the 'discove of 'real meangs behnd other oves and other constelaons. Precisely the sae remarks apply to the 'dcove of an individual I that is dierent from faces, behaviour, objecve 'mental states of the type tha cur in A, to the 'dscove of a substnce behnd 'appearances (formerly elements of A) or to the 'disce that honour may be lacg despite the presence of its outer anifestaos A statement such as Heractus 'you could not nd e limts of the soul though you are aveling eve way, so deep is its logos (Diels B 45) does not just to cosmos A, it unts
204
AGANST METHOD
the prnciples which are needed in the consucon of A-ype 'mental states while Heraclitus rejecon of 1OAa6C' and Parmendes reecon of an eo� onQOV undercuts ules that gove the consucon of single f of A A enre worldvew, an enre universe of thought, speech, percepon is dissolved It is interesng to see how this press of dissolvng manifests itself in parcular cases. In his long speech in , 9 .308, Achlles wants to say that honour may be absent even though all its outer manifestaons are present. The terms of the language he uses are so inmately ed to denite sial situaons that he 'has no language to express hs disilusionment Yet he expresses it, and in a remarkable way. He does it by msusing the language he disposes of. He quesons that cannot be answered and makes demands that cannot be met.1 12 He acts in a most 'irraonal way The sae irraonality is found in the wrings of all other ear authors. Compared wth A the Presracs speak sagely indeed So do the lyrcal poets who explore the new possibiies of selo they have 'discovered Freed from the fetters of a wellconsucted and unambiguous mode of expression and thinig, the elements of A lose their faliar funcon and start loang around ailessly the chaos of sensaons arses Freed from rm and unambiguous sial situaons feelings become leeng, ambivalent, conadicto: ' love, and I love not; I rave, and I do not rave, wrtes Anakreon} 3 Freed from the rules of late geomec painng th arsts produc sange mixtures of perspecve and blueprint. 1 14 Separated from well-determned psychological sets and freed of their reac import, concepts may now be used 'hypothecally wthout any odium of lyng and the arts may begin exploring possible worlds in an imaginave way1 15 This is the same 'step back which was earlier 1 1 2. A. ar, e Lanage of Achilles', Trans ( P Am Phil Ass, 7, 1 956, p. 6. f. te discussion of the case in Farewe t e, hapte 10. 1 1 3 . , Anth/ga Ly, fr. 79. 1 1 4. hl, op. cit., cf. also J. ite, Ppee in Ant Drawng and Pann London, 1965. 1 1 5. luarch repo the following stor in h L fSn: en the compan o espis began o exhibit aged, and its novel was aacng the pulace but h not yet got as far as public compeons, Solon, being fond of listening and e and being rthe gven in is old age to leisure and amusemen, and indeed to d pes nd music, went to see Thespis a in his own play, as the prcce ancient mes oon approached him aer the pefo ce and ased i he not asmed to te so many es to so many people en Thespis sid the nothng dreadl in represenng such works nd acons in n, Solon su e ground violenty with hi wing sc "If we applaud these things in n, he si
SITEEN
05
be a necessa presupposon of change and, posbly, seen to progress only t now does not ust dscard obseaons, t dscards somemportant standards of raonalty as wel Seen from A (and also fro the pont of vew of some later deologes) ll thse e, poets, ats, are ravng macs Remember the ccumstances whch are responsble for ths stuaon We have a pont ofvew (theo, framework, cosmos, mode of representaon) whose elements (concepts, facts, pctues) are but up n accordance wth ce prncples of consucon The prncples volve somethg le a closure there are thgs that cannot be ad, or dcovered, wthout volang the prncples (whch does not mean conadcng them) Say the thgs, make the dscove, and the prncples are suspended Now take those consucve prncples that underle eve element of the cosmos (of the theo), eve fact (eve concept) Let us call such pncples unal pnpl of the theo n queson Suspeng vel prncples means suspendng all facts and all concepts Fnally, let us call a dscove, or a statement, or an attude nommura wth the cosmos (the theo, the framework) f t suspends some of ts unversal prncples Heracltu B 45 s ncomensrble wth the psychologcal part of A t suspends the rules that are needed for constung ndvduals and puts an end to all A facts aut ndvduals phenomena correspondng to such facts may of course persst for a consderable me as not all conceptual changes lead to changes n percepon and as there est conceptual changes that never leave a ace n the appearances however, such phenomena can no longer be ed n the customa way and caot therefore count as obseaons of the customa 'obecve facts) Note the tentave and vague nature of ths explanaon of 'ncommensurable and the absence of logcal termolo e reason for the vagueness has already been explaned (tems and 4 above) The absence of logc s due to the fact that we deal wth phenomena outsde ofts doman My puose s to nd termolo "we sha sn nd oursees honourng them in eest' The story sees historcay imsibe yet elucidates a wdespread atude (for ttude cf Chapter 8 of ohn Forsde Gree b , New or, 196) Solon sf seems to have been somewhat es impressed by aditona fos of thought nd he may have been one of the t dmc ators (othe ca vare) G E, e n and Ear fTrage, Cambrdge, 1965, pp The opste atude, hich reveas te secure and aead somewhat conceited cien ofB i epresed by Sonides who nwered te ueson why the Thesns were not deceied b him saying Becaue tey are t stupid' Pluh, Dea , D 116 Chapte I I , tet to fote 5
206
AGA NST METOD
for describing certan compe historica-anthropoogical phenomena which are ony mperfecty understood rather than dening properes ofogica systems that are speced in deta Tes, such as 'unversal prncpes and 'suspend, are supposed t summarze anthropoogica nformaon much in the same way n whch Evans Pritchards account of Nuer me tet to fooote 85) summarizes the anthropoogica informaon at hs disposa c aso the brief dscussion in item 3 above) The vagueness of the epanaon reects the ncompeteness and compety of the materia and nvites arcuaon by futher research The epanaon has o have some content otherwse it woud be useess But t must not have too much content, or ese we have to revise t eve second ine Note, aso, that by a 'prncpe I do not smpy mean a statt such as 'concepts appy when a nte number of condons is sased, or 'nowedge s enumeraon of dscrete eements whch form paratacc aggretes but th ammatcal hat corresponding to the statement. The two statements just uoted descrbe the habit of reding an object as given when the ist of its parts has been fuy presented Ths habit suspended though not conadcted) by the coeure that even the most compete st does not ehaust an object it s ao suspended but an not conadcted) by any unceasing search for new aspects and new properes. It s therefore not feasibe to dene 'ncommensurabty by reference to state ments 7) I the habt s suspended, then A -objects are suspended with it one cannot eamine A-objects by a method of conjectures and refutaons that nows no end How s the 'rraonaty of the ansion period overcome? It is ercome n the usua way c tem 8 above), e by the determned producon ofnonsense un the materia produced is rch enough to permt the rebes to revea, and eveone ese to recogne, new universa principes Such reveang need not consst in wring the princpes down n the form of cear and precse statements) Madness tus into sanity provided t s sufcenty rich and sucenty reguar to funcon as the bass of a new word-vew And when that happens, then we have a new probem how can the od vew be compared with the new vew? From what has been said t is obvous that we cannot compare the contts of A and B. A-facts and B-facts cannot be put sde by side, not even n memo presenng B-facts means suspending prncipes 1 1 7 This taes care of a crcism in fote 63 of Shapere's arcle in Mnd a Cms, Pisbugh, 1966 The classicaons achieved by the principles are cove n the sense ofof cf above fote 4 and text down to fote 9
SITEEN
07
assumed in the consucon of A-facts. All we can do is draw B pictures of A-facts in B, or inoduce B-statements ofAfacts into B. We cannot use A-statements of A-facts in B. Nor is it possible to transt language A into language B. Ths does not mean that we cannot dsss the two views but the discussion w lead to szeable changes of both views (and of the languages in whch they are expressed). ow it seems to me that the relaon beeen, say, classical mechancs (ntereted realiscaly) and quanum mechancs (inter preted in accordance with the views of Niels Bohr), or between Newtonian mechancs (intereted realcaly) and the genera theo of relavity (o intereted reacaly) is in many respects similar to the relaon between cosmolo A and cosmolo B. Thus eve fact of Newton's mechcs presumes that shapes, ses, peros are changed only by physical interacons and ths presumpon is suspended by the theo of relavity Sar the quantum theo constutes facts in accordance with the uncety relaons whch are suspended by the classical approach At this point it important to interpret the situaon in a sensible manner of else scienc (cultural) change becomes an inexplicable macle The idea that comprehensive ways of g, acng, perceiving such as cosmolo A (and, n a much more naow domain, classical physics) and cosmolo B (relavity or quantum mechanics) are closed framewor with ed rules creates an unbridgeable gulf between situaons whch, though derent in surising ways, are yet connected by arents, alusions, borrowgs, analogies, general pnciples of the nd explaned in the text above Logicians who conne the term argument' to cha of reasoning involving stable and prece concepts and who recosuct theories and world-views using equaly precise and unambiguous terms are forced to cal such connecons raonal' whe their opponents can report the discove' that science, that aleged songhold ofreason, often volates reason in a decisive way. Both are talkng abou chimaeras, not about science and culture as they realy are. Thngs change when we use scienc pracce or cultural reaty and not logic as our informants, in other words, when we enge in sociologcal research, not in reconsucon We then discover that scienc concepts (and concepts, shapes, percepts, styles in general) are ambiguous in the sense that decisive events can aect their appearance, their perceived implicaons and, with them, the logc' they obey. Achles (see the text to fooote 112 above) misuses' the language he has at his dposal by asserng a dierence between real' honour and its sial manfestaons Asserng
208
AGA NS T MET HOD
derences s not n conlct wth vew A; for eample, there s a great derence between the knowledge, the power, the acons of the gods on the one sde and the knowledge, the power and the acons of humans on the other. Assumng that honour s n the hands of gods who don't gve a damn about the aspraons of humans devalues the socal manfestaons of honour, makes them seconda. The assumpon ts well nto the general outlnes of vew A but Achlles s the rst to make t. Why? Because hs anger, suerng makes hm see connecons whch, because of a wdespred opmsm, are not part of the general vews about honour and do not conbute to ts denon' He seems to volate basc sal rules but vewed wth the anety caused by Amemnon's acons such rules gve way to a derent dea that s rerded as beng mplct n the esng materal but as not havng surfaced so far Generalzng, we can say that concepts have potenales over and above the usages that seem to ene them; t s ths feature that makes them capable of connecng enrely derent conceptual systems More about ths n my (I promse!) last bk, The Conqut ofAunnce
Append 2 Whorf speaks of Ideas', not of events' or of facts', and it is not always clear whether he would approve of my extension of his views On the one hand he says that me, velity, and matter are not essenal to the consucon of a consistent picture of the universe, 1 and he asserts that we cut up nature, ornize it into concepts, and ascribe sicances as we do, largel} because we are paral to an agreement to ornize it in this way', 2 which would seem to imply that widely dierent languages posit not just dierent ideas fo the orderng of the same facts, but that they posit also dierent facts e linguisc relavity principle' seems to point in the same decon It says, in informal terms, that users of markedly dierent grammars are pointed by their grammars towards dierent types ofobseaons and dieren evaluaons of exteally similar acts ofobseaon, and hence are not equivalent obseers, but must arrive at somewhat dierent views of the world'. 3 But the more formal statements'4 of the principle already contains a dierent element, for here we are told that all obseers are not led by te same pyscal nce to the same picture of the universe, unless their lnguisc backgrounds are simlar, or can in some way be calibrated', 5 which can either mean that obsees using widely dierent languages wll post dtfas under the same physical circumstances in the same physical world, or it can mean that they will aange smlarfas n dt ways The second interetaon nds some support in the examples given, where dierent isolates of meaning in English and Shawnee are said to be used in reporng te same ece6 and where we read that languages classi tems of experience dierently' 7 experience is 1 o, op cit, p 216 2 ibid, p 2 13 ibid , p 221 4 ibid, p 2 2 1 5 ibid, p 2 1 4, italics 6 ibid, p 208 7 ibid, p 209
209
210
AGA N ST METHO
rerded as a uniform reseoir offacts which are sed derently by dierent languages. It nds fuher support in Whors descripon of the ansion from the hor-v account of bameic phenomena to the mode theo: If once these sentences [Why does water rise in a pump? Because Nature abhors a vacuum.] seemed sasing to logic, but today seem diosyncrasies of a parcular argon, the change dd not come about because science has discovered new facts. Science has adopted new lingusc formulaons of the old facts, and now that we have become at home in the new dialect, certain aits of the old one are no longer binding on us'.8 However, I rerd these more conseave satements as seconda when compared with the great inluence ascribed to grammacal categories and especially to the more hdden rapport systems' of a language 9 Whorf and those who follow him rerd language as the main and perhaps as the only shaper of events'. That· is much too narrow a point of view Anals have no language in the sense of Whorf, yet they do not live in a shapeless world. Planets, at least as conceived today, are not even alve, but they aect their surroundings and react to them in a lawful manner In humans rituals, music, the arts, adapve behaviour that occurs without the inteosion of wors mae important conbuons to the way n whch the world aea and, to those living accordingly, s In the sciences we have not o statements (the old idea that science is a system of statements has b now been thoroughly discredited), but obseaons, experimental equipment, an intuive relaon between obseers and the equipment that has to be leaed n a praccal way and cannot wrtten down, the wor of expementasts which has much in common with the wor of arsts what they want are not mere results, but results that emerge in a simple, compellng and aesthecally pleasing way and so on A concenaon on langua alone, or on texts', can easily lead into absurdity, as s sho by Ausn and by the pracce of deconsucon on the oe hand philosophers produce texts, lie poets; on the other hand they tae it for granted that their texts reveal a realty beyond the thoughts, impressions, memores, gures of speech, etc, etc. fro whch they arose (Scienc realists to a certain extent share in predicament) Fnally, soe commens on what I thin about incommensurabil and how I arrived at the idea 8 bd , p 222 9 ibd, pp 68
APPE NDI 2
11
I thk that incommensurabiity tu up when we shaen our concepts in he manner demanded by the logical posivsts and their osprng and that it unin their ideas on explanaon, reducon and progress. Incoensurability disaea when we se concepts as sciensts use them, in an open, ambiguous and ofen counter ntuive manner Incomensurabiity is a problem for phlosophers not for sciensts, though the latter may become pcho consed by unusual thgs. I arrived at the phenomenon whie sudyng the arly literature on basic statements and by considerng the possibiity of percepons radically dierent from our own In my thesis 10 I eed the meang of obseaonal statements I considered the idea that such statements describe what s given' and ied to iden this given' Phomolocal this did not seem to be possible; we noce objects, their properes, their relaons, not the given' It is of course ue that we can give quick reports on the pperes of eveday objects but ths does not change them into non-objects but only shows that we have a special relaon to them. Phenomenologically what is given consists of the sae hgs which can also est unobseed it is not a new knd of object. Special arrangements such as the reducon screen inoduce new con dions, they do not reveal ingredients in objects we already know Result the given cannot be olated by obseaon The second possibility was to isolate it by logical means what given can be asceed with cean, hence I obta the given contained in the table before me by removing from the statement there is a table' all the consequences that make ture correcons possible Th shows that the given is the result of an unreasonable decision unestable statements cannot see as a basis for science Followg this argument I inoduced the assumpon that the meaning of obseaon statements depends on the nature of the objects described and, as this nature depends on the most advanced theores, on the content of these theories Or as I formulated it in my rst English paper on the topic the interetaon of an obseaon language is determined by the theories which we use to explain what we obsee, and it changes as sn as these theories change 11 In a word obseaon statements are not just theo-/ the views of Toulmin, Hanson and apparently also Kuhn) but theoretil and the disncon between obseaon statements protol 10 Venna, 195 1 wren aer two years of etensive discussion in the Kra Crcle and supesed by Professor Vtor Kraft of the Unersi of Vienna II Aept at a Reastc Interetaon of Eperence', P At S 198, reprnted n Phshl P, Vol I e passage (in iacs) i on p 3 1
212
AGAINST METHOD
stateents' n the termnolo of the Vienna Circle) and theorecal statements is a pragmac disncon, not a semanc disncon; there are no special obseaonal meanngs' Thus in the same year as Hanson (Hanson's o[Dc appeared in 18) and four years before Kuhn I formulated a thesis a weaker fom of which became ve popular later on Moreover, my thesis not only was songer than the thesis of theo-ladenness, it also came from a dierent source For while Toulmn and Hanson were inspired by Wittgenstei's Philoshil tigations I started from and retued to ideas that had been developed in the Vienna Circle and I said so. 12 Quine, whose phlosophy shows close connecons to the phlosophy of the Vienna Circle, 13 also used a criteron of obseabity that is rather sar to mne 14 Now when Feigl heard of these ideas he pointed out that intereg obseaons in terms of the theores they are obseaons of makes nonsense of crucial experments for how can an experment decide between two theores when its interetaon already depends on these theores and when the theores themselves have no comon elements, such as a common obseaon language? In the paper just menoned and in Explanaon, Reducon and Empircism' published in 162, I took up the chalenge I rst increased it by consucng cases where portant terms of one theo cannot in any way be dened in another which, moreover, es to do its job. My example which I found in Anneliese Maer's i Voruf Glilis im 14 Jahrhun was the relaon of the te impetus' ad momentum' I also developed a theo of test to answer the challenge In 162 I called theores such as those cong mpetus' and momentum' incomensurable theores sad that ony a special class of theores, so-called non-instanal theores could be (but need not be) incommensurable and added tat successive incomensurable theores are related to each other b replacement, not by subsumpon The year 162 is also that of Kun's grea book but Kun used a dierent approach to apply the same term to a similar (not an idencal) situaon Hs approach was historcal, whie mne was absact In 160 I started the studies descrbed in chapters 8, and 16 They revealed tht percepon and expermentan obey laws of thei 12 Phhl Pap, Vol I , pp 49, 125 13 Deails n Dr Koppelberg, De Auung Anatch Phlse, Franfu, 1987 14 Phshil Pap, Vol I, pp 17f
APP EN DI 2
13
own which cannot be reduced to theorecal assumpons and are therefore beyond the grasp of theo-bound epistemologes I also joined Kuhn in demanding a hstorical as opposed to an epistemlogical grounding of science but I sl dier from him by opposing the polical autonomy of science Apart from that our views (i.e my published views and Kns as yet unpublished recent phiosophy) by now seem to be aost idencal, 1 5 except that I have little pathy for Ks attempt to e up hsto with phiosophical or nusc, but at any rate with theorecal ropes a connecon with theo just brings us back to what I at least want to escape from the rigid, though chimaerical (deconsucon!) boundaries of a conceptual system'
Cf my eaism and the istoc of nowedge', eJu fPh, Vo , 1989, pp 353, e fote 26 nd the pt to the psent ey
7 Neith sce nor rationali are uneal meur ofcece Th are pailar traditions unaware oftheir htocal undin
So far I have ied to show that reason, at east in the form in which it is defended by ogicians, phiosophers and some sciensts, ds not t science and coud not have conibuted to its growh This is a gd argument against those who admire science and are aso saves of reason They must now make a choice They can keep science they can keep reason; they cannot keep both But science is not sacrosanct The mere fact that it ests, is admired, has resuts is not sucient for making it a measure of eceence Mode science arose from goba objecons ast earier views and raonaism itsef, the idea that there are genera rues and stndards for conducng our aairs, aairs of knowedge incuded, arose from goba objecons to common sense (eampe Xenophanes ast Homer) Are we to refrain from enging in those acvies that ve rise to science and raonaism in the rst pace Are we to rest content with their resuts Are we to assume that eveg tht happened after Newton (or aer ibert) is perfecon? Or sha we admit that mode science may have basic fauts and may be in need of goba change And, havig made the admission, how sha we preed How sha we aie fauts and car out changes Dont we need a measure that is independent of science and conicts with it in order to prepare the change we want to bring about And wi not the rejecon of rues and stndards that conict wth science forever prevent us from nding such a measue? On the other hand have not some of the case studies shown that a bunt appicaon of 'raona predures woud not have given us a better science, or a better word but nothing at a And how are we to judge the resuts themseves Obviousy there is no simpe way of guiding a pracce by rues or ofcricizing standards of raonaity by a pracce The probems I have sketched are od ones and much more 24
SEVENTEN
215
neral than the problem of the relaon between science and raonality They occur whenever a rich, well-arculated and familiar pracice a pracce of composing, of painng pictures, of stage ducon, of selecng people for public oce, of keeping order and punishing criminals, a pracce of worship, of organizing society is onfronted by a pracce of a dierent ind that can intract with it Te intraons and their results depend on historical condions and a from one case to the next A powerful be invadig a coun may impose its laws and change the indigenous adions by force only to be changed itself by the remnants of the subdued culture A rler may decide, for reasons of convenience, to use a popular and stabilizing religion as the basic ideolo of his empire and may thereby conbute to the transformaon both of his empire and of the religion chosen A individual, repelled by the theae of his me and in search of something better, may study forei pays, ancient and mode theories of drama and, using the actors of a friendly ompany to put his ideas into pracce, change the theae of a whole naon A roup of painters, desirous of adding the reputaon of being sciensts to their already enormous reputaon as silled craftsmen, may inoduce scienc inredients such as geome into painng and thereby create a new style and new problems for pinters, sculptors, ahitects A astronomer, crcal of the difference between classical prnciples of asonomy and the esng racce and desirous to rstore asonomy to its former splendour, ma nd a way to achieve his aim and so iniate the removal of the classical principles themselves In all these cases we hae a pracce, or adion, we have certain inluences upon it, emerging from another pracce or adion and we obsee a change The change may lead to a slight modicaon of orginal pracce, it may eliminate it, it may result in a adion t barely resembles either of the interacng elements Interacons such as those just descrbed are accompanied by hanging degrees of awarns on part of the parcipants Copeicus new ve well what he wanted and so did Constanne th reat I m now speaing bout the inial impulse, not about the tnsfaon that followe) The inusion of geome into inng is less easily ccounted for in terms of awareness We have no idea why iotto ed to achieve a compromise between the surfc of the painng and the corporeality of e things painted specially as pictures were not yet regarded as studies of a material reality We can surmise that Brunelleschi arrived at his consucon b a natura extension of the architects' method of represenng ree- dimensional objects and that his contacts with contempora
216
AGAINST METHOD
scienss were not without consequence It is sll more dicult to understand the gradually rising claims of arsans to make conibuons to the same kind of knowledge whose principles were explained at universies in ve dierent terms Here we have not a crical stu of alteave adions as we have in Copecus, or in Constanne but an prson of the uselessness of academic science when compared wth the fascinang consequences of the joueys of Columbus, agellan and their successors There arose hen the idea of an America of Knowledge', of an enrely new and as yet unforeseen connent of knowledge that could be discovered, just as the real America had been discovered by a combinaon of skill and absact study arsts have been fond of confounding insucient informaon conceing the awareness that accompanies such presses with irrelevance and they have ascribed only a seconda role to indivdual consciousness In ths they were right but not the way they thought For new , though often necessa, were not sucient for explaining the chang that now curred and that depended also on the often unknown and unrealzed) rmstanc under whch the ideas were applied Revoluons have ansformed not only the pracces their iniators wanted to change but the ve principles by means of whch, intenonally or unintenonaly, th carried out te change Now considering any interacon of adions we may ask two kinds of quesons whch I shall call obseer qutons and papant qutons respecvely Obseer qutons are conceed with the detais of an interacon They want to give a hstorical account of the interacon and, perhaps, formulate laws, or rules of thumb, that apply to all interacons Hegel's ad posion, neon, synthesis negaon of the neon) is such a rule Papan qutons deal with the attude the members of a pracce or a adion are supposed to take towards the possible) inusion of another The obseer asks what happens and what going to happen The parcipant asks what shal I do? Shal I support the interacon Shall I oppose it? Or shal I siply foret about it In the case of the Copecan Revoluon, for example, the obseer asks what impact did Copecus have on Wittenber asonomers at about 60 How did they react to his work Did they change some of their beliefs and if so, why Did ther change of opinion have an eect on other asonomers, or were they an isolated group, not taken seriousy by the rest of the profession The quesons of a parcipant are ths is a sange book indeed
SVENTEE
217
should I take it seriously? Should I study t detai or only supercially or should I simply connue as before? The main theses seem absurd at rst sight but, maybe, there is something in them? How shall I nd out? d so on It is clear that obseer quesons must take the quesons of the parcipants into account and parcipants will also sten most carefuly (if hey are inclined that way, that is) to what obseers have to say on the matter but the intenon is dierent i both cases Obseers want to know what is going on, parcipants what to do An obseer describes a life he does not lead (except accidentay), a parcipant wants to aage his own life and asks himself what attude to tke towards the things that may inluence it Parcipants can be unis and act in a saighoard and praccal way In the late 6th centu many pinces becae Protestants because ths futhered their interests and some of the subjects becae Protestants in order to be left in peace Wen Brish colonial ocials replaced the laws and habits of forei bes and cultures by their own 'civiized laws the latter were oen accepted because they were the laws of the ng, or because one had no way to oppose them, and not because of any inisic excelence The source of their power and 'validity was clear understd, both by the ocials and by the more astute of their unfortunate subjects In the sciences and especially in pure mathemacs one often pursues a parcular line of research not because it is rerded as inisicaly perfect, but because one wants to see where it leads I shal call the philosophy underlying such an attude of a parcipant a praatic philoshy A pragmac phlosophy can lourish only if the adions to be juged and the developments to be inluenced are seen as tempora makeshis and not as lasng constuents of thoughts and acon A parcipant with a pragmac phiosophy views pracces and adions much as a aveller views forei counes Each couny has features he kes and things he abhors In deciding to settle down a aveller will have to compare climate, landscape, language, temperament of the ihabitants, possibilies of change, privacy, looks of male and female populaon, theae, opportunies for advancement, quaity of vices and so on He w also remember that his inial demands and expectaons may not be ve sensible and so permit the press of choice to aect and change his 'nature as wel which, aer all is just another (and minor) pracce or adion enteing the press So a pragmast must be both a pacipant and an obseer even in those exeme cases where he dedes to live in accordance with his momenta whims enrely
218
A G A I N S T M E T H OD
Few individuals and groups are pragmasts in the sense just described and one can see why it is ve dicult to see ones own most cherished ideas in perspecve, as parts of a changng and, perhaps, absurd adion Moreover ths inabiity not only ts it is also couraged as an attude proper to those enged in the study and the improvement ofman, siety, knowledge Hardly any religion has ever presented itself just as something woth tryng The claim is much songer the religion is the uth, everything else is error and those who kow it, understand it but sll reject it are rotten to the core (or hopeless idiots) Two elements are contained in such a claim First, one disnguishes between adions, pracces and other results of individual anor collecve human acvity on the one side and a dierent doman that may act on the adions without being one Secondly, one explains the sucture of this special domain in detai Thus the word of God is powerful and must be obeyed not because the adion that carries it has much force, but because it is outside all adions and provides a way of improving them The word of God can s a adion, its meanng can be handed on from one generaon to the next, but it is itself outside all adions The rst element the belief that some demands are 'objecve and adion-independent plays an important role i ratonals which is a seculaed form of the belief in the power of the word of God And ths is how the opposion reasopracce obtans i polemical sng For the two agencies are not seen as two pracces which, while perhaps of unequal value, are yet both imperfect and changng human products but as one such product on the one side and lasg measures of excellence on the other Early Greek raonalsm already contans this version of the conict Let us exame what circumstances, assumpons, predures what features of the historical press are responsible for it To s with the adions that oppose each other Homeri comon sese and the various forms of raonalism that arise in the 6th to 4th centuries have dert ntal stur 1 On the one hand we have complex ideas that cannot be easiy explained, they 'work but one ds not know how, they are 'adequate, but one ds not know why, they apply in special circumstances only, are rich in content but poor in simiaries and, therefore, i deducve connecons On the other side there are relavely clear and simpe concepts which, having just been inoduced, reveal a good deal of I For detals see Chapter 16
SEVENTEEN
219
their sucture and whch can be lnked in many ways. They are pr in content, but rch in deducve connecons The dierence becomes especially sg in the case of mathemacs In geome, for example, we s with ules of thumb applying to physical objects nd their shaes under a great varety of circumstances Later on it can be ped why a given rule applies to a gven case - but the prfs make use of new enes that are nowhere found in nature n anquity the relaon between the new enes and the faiar world of common sense ve rse to various theores One of them whch one mght call Plat assumes that the new enes are real whle the enes of common sense are but ther mperfect copies. Another theo, due to the Shists, regards natural objects as real and the objects of mathemacs (the objects of 'reason) as smpleminded and unrealisc mages of them. These two theores were also applied to the dierence between the new and fair absact idea of knowledge propated by Plato (but found already before) and the comonsense knowledge of the me (Plato wise uses a distorted mage of the latter to give substance to the former) Ain it was either sad that there ested only one ue knowledge and that human opnon was but a pale shadow ofit or human opnon was rerded as the only substanal knowledge in estence and the absact knoledge of the phosophers as a useless dream ( can see horses, Plato, sad Ansthenes, but nowhere see your ideal horse). t would be interesg to follow ths ancient conct through hsto down to the present. ne would then lea tha the conct tus up in many places and has many shapes Two exmples must suce to ilusate the great varety of its mafestaons When Goched wanted to reform the German theae he lked for plays woth imtang That is, he lked for adions more orderly, more dignied, more respectable than what he ound on the stage of hs me He was attracted by the French theae and here manly by Coeile. Being convinced that such a complex edice of poe (as agedy) could hardly est without rules 2 he lked for the rules and found Aristotle. For hm the rules of Arstotle were not a parcular way of viewng the theae, they were the reason for excellence where excellence was found and guides to mprovement where improvement seemed necessa. Gd theae was an ebodient of the rules of Arstotle. Lessing gradualy prepared a dierent view. First he restored what he thought to be the real 2 Vorrede um Sterbenden Cato" ' quoted fromJ Chr. Gohed, zur itatur, Stuga, 1972, p. 2
220
AGAIN ST METHOD
Aristotle s opposed to the Aristotle of Coeille nd Gottched Next he permitted violons of the letter of Aristotles rules provided such violons did not lose sight of their im. And, nll he suggested a dierent paradigm and emphasized that a mnd invenve enough to consuct it need not be rescted by rules. f such a mind succeeds in his eorts then let us forget the textbk! 3 n a dierent (and much less interesng) domain we have the opposion between those who suggest that languages be consucted and reconsucted in accordance wth simple and clear rules and who favourably compare such ia/ lanag with the sloppy and opaque natural idioms and other phiosophers who assert that natural languages, being adapted to a wide variety of circumstances, could never be adequately replaced by their anaemc logical competors. This tendency to view dierences in the sucture of adions (complex and opaque vs simple and clear) as dierences in kind (real vs iperfect realizaon of it) is reinforced by the fact that the crics of a pracce take an obseers posion with respect to it but reman parcipants of the pracce that provides them with their obecons. Speaking the language and using the standards of ths pracce they discover liitaons, faults, errors when all that really happens is that the two pracces - the one that is being criczed and the one that does the crczing- dont t each other. Many a against an outandout atal are of ths kind. They noce that materialism changes the use of mental terms, they illusate the consequences of the change with amusing absurdies (thoughts having weight and the like) and then they stop The absurdies show that materialism clashes with our usual ways of speaking about mnds, they do not show what is better - materialism or these ways But taking the parcipants point of view with respect to common sense tus the absurdies into arguments against materialism t is as if Americans were to obect to forei currencies because they cannot be brought into simple relaons (1 1 or 1 1 0 or 1 1 00) to the dollar. The tendency to adopt a parcipants view with respect to the posion that does the udging and so to create an Archiedian point for cricism is reinforced by certain disncons that are the pride and oy of amchair philosophers. refer to the disncon between an evaluaon and the fact that an evaluaon has been made,
Hambu Dramatue, Stuck 48 Cf, however, Lessng's crtcism of te
claims of the orgnal genuses' of hs tme in Stck Lessing's account of te relaon between reason' and pracce is qute complex and n agreement with the developed further below
SEVENTEEN
221
proposal and the fact that the proposal has been accepted, and the related disncon between subjecve wishes and objecve standards of excelence. When speang as obseers we often say that ce groups accept certain standards, or think hghly of these standards Speang as parcipants we equally often use the standards without any reference to their origin or to the wishes of those usig them We say theories ought to be falsiable and conadicon free and not want theories to be falsiable and conadicon free or sciensts become ve uhappy unless their theories are falsiable and conadicon free. No it is quite correct that statements of the rst knd (proposals, rules, standards) (a) conta no reference to the wishes of individual human beings or to the habits of a ibe and (b) cannot be derived from, or conadicted by, statements conceg such wishes or habits, or any other facts. But that does not make them objecve and independent of adions To infer from the absence of terms conceg subjects or groups in there ought to be . . that the demand ade is objecve would be just as erroneous as to clai 'objecvity i.e. independence from personal or group idiosyncrasies, for opcal ilusions and mass hallucinaons on the grounds that the subject, or the group, nowhere curs in them There are many stateents that are ulated objecvely, i.e wthout refence to adions or pracces, but are s eant to be untood in relaon to a pracce Examples are dates, coordinates, statements conceg the value of a currency, stateents of logic (after the discove of alteave logics), statements of geomey (after the discove of NonEuclidean geomees) and so on. The fact that the retort to you ought to do X can be thats what you thnk! shows that the same is ue of value statements. And those cases where the reply is not allowed can be easiy reced by using discoveries n value theo that correspond to the discove of alteave geomees, or alteave logical systems: we confront objecve value judgement from dierent cultures or dierent pracces and ask the objecvist how he is going to resolve the conlict.4 Reducon to shared principles is not always possible and so we must admit that the demands or the formulae expressing them 4 In the play e Rulng s ater tued into a somewhat vapid lm wth eter
O'Toole) two madmen claing to be God are confronted wth each other Ths maellous idea so confuses the playwrght that he uses re and brmstone nstead of dialogue to get over the problm His na solution, however, is quite nteresng The one madman ts into a good, upright, noa Brsh Cien who pays Jac the Ripper on the sde Dd the playwrght mean to y that our mode obectivists' who have been through the re of relavism can retu to noacy ony f they are permed to annhlate al disturbing elements?
222
AGAINS T METHOD
are incomplete as used and have to be revised. Connued insistence on the objecvity ofvalue judgements however would be as illiterate as connued insistence on the absolute use of the par up-down after discove of the spherical shape of the ea. Ad n argument such as it is one thng to utter a demand and quite a dierent thing to assert that a demand has been made - therefore a mulplicity of cultures does not mean relavism has much in comon with the argument that anpodes cannot est because they would fall down. Both cases rest on antediuvian concepts (and inadequate disnc ons) Small wonder our raonalists are fascinated by them With this we have also our answer to (b) t is ue that stang a demand and describing a pracce may be two dierent things and that logical connecons cannot be established btween them T does not mean that the interacon between demands and pracces caot be eated and evaluated as an interacon of pracces For the dierence is due, rst, to a dierence between obseerattude and parcipantattude one side, the side defending the objecvity of its values, u its adion instead of aining it - whch ds no tu the adon into an obecve measure of validity Ad second, the dierence is due to concepts that have been adapted to such one sidedness. The colonial ocial who prlaims new laws and a new order n the ame of the king has a much better grasp of the situaon than the raonalist who recites the mere letter of the law without any reference to the circumstances ofits applicaon and who rerds th fatal incompleteness as proof of the objecvity of the laws recited. After this preparaon let us now look at what has been caled the relaon between reason and pracce. Siplig matters somewhat we can say that there est three views on the matter. A Reason guides pracce. ts authority is independent of the authority of pracces and adions and it shapes the pracce accordance with its demands. This we may call the ialstic ion of the relaon. B Reason receives both its content and its authority from pracce t describes the way in whch pracce works and formulates i underlying prnciples This version has been called natural and it has casionally been atibuted to Hegl (though erroneously so) Both ideaism and naturalism have dicules. The dicules of idealism are tha the idealist does not only wt to act raonaly he also wants hs raonal acons to have resul And he wants these results to cur not only among the deaao he uses but in the real world he nhabits For example, he wants re
SVNN
223
human beings to build up and maintain the siety of his dreams, he wants to understand the moons and the nature of real stars and real stones Though he may advise us to put aside all obseaon o) the heavens' 5 and to concentrate on ideas only he eventually retus to nature in order to see to what extent he has grasped its laws 6 It then often tus out and it often has tued out that acng raonally in the sense preferred y him does not produce the expected results This conlict between raonality and expectaons was one of the main reasons for the constant reform of the canons of raonality and much encouraged naturalism But naturalism is not sasfacto either Having chosen a popular and successful pracce the naturalist as the advantage of being on the right side', at least for the me being But a pracce may deteriorate; or it may be popular for the wrong asons Much of the popularity of mode scienc medicine is due to the fact that sick people have nowhere else to go and that television, rumours, the technical circus of well equipped hospitals convince them that they could not possibly do better) Basing standards on a pracce and leaving it at that may forever peetuate the shortcomngs of this pracce The dicules of naturalism and idealism have certain elements in common The inadequacy of standards often becomes clear from the barrenness of the pracce they engender, the shortcomings of pracces often are ve obvious when pracces based on dierent standards lourish Ths suggests that reason and pracce are not two different kinds of enes but pa ofa single dia/eicalprocs The suggeson can be illusated by the relaon between a map and th adventures of a person using it or by the relaon between an arsan and his insuments Orignally maps were consucted as images of and guides to reality and so, presumably, was reason But maps like reason contain idealizaons Hecataeus of Miletus, for example, imposed the general outlines of Anamander' s cosmolo on s account of the occupied world and represented connents by geometrical gures) Te wanderer uses the map to nd his way but he also corrects it as he proceeds, removing old idealzaons and inoducing new ones Using the map no matter what will soon get him into ouble But it is better to have maps than to proceed without em In the same way, the example says, reason without the idance of a pracce will lead us asay while a pracce is vastly improved by the addion of reason 5. Plato, Rublc,
6
pno
530bf
224
AGANS THOD
Ths account, though better than naturalsm and dealsm and much more realsc, s sll not enrely sasfacto. t replaces a one sded acon (of reason upon pracce or pracce upon reason) by an nteracon ut t retans (certan aspects o the old vews of the nteracng agences reason and pracce are sll rerded as enes of derent knds. They are both needed but reason can est wthout a pracce and pracce can est wthout reason. Shall we accept ths account of the matter? To answer the queson we need only remember that the derence between reason and somethng unreasonable that must be formed by t or can be used to put t n ts place arose from tug suctural derences of pracces nto derences of knd. Even the most perfec standards or rules are not ndependent of the materal on whch they act (how else could they nd a pont of attack n t?) and we would hardly understand them or know how to use them were they not wellntegrated parts of a rather complex and n places qute opaque pracce or adon, vz. the language n whch the sor rationis expresses hs ste commands 7 n the other hand even the most dsorderly pracce s not wthout ts regulares, as emerges from our attude towards nonparcpants. 8 at is caed reon ' and raice are therre two dert p ofpraice, the derence beng that the one clearly exhbts some smple and easly producbe formal aspects, thus makng us forget the complex and hard understood pperes that guarantee the smplcty and producblty, whle the other drowns the formal aspects under a great varety of accdental properes. But complex and mplct reason sll reason and a pracce wth smple formal features hoverng above a peasve but unnoced background of lngusc habts s sll a pracce. Dsregardng (or, rather, not even nocng) the sensegvng and applcaonguaranteeng mechansm n the rst case and the mplct regulares n the second a raonalst perceves law and order here and materal yet n need of beng shped there. Th habt, also commented upon n an earler part of ths secon, to take a parcpants pont of vew wth respect to the former and 7 Ths nt has been made with great force and wth the help of many examl by Wgensten (cf my essay Wgensten's Phlohl Itgato', Phl R
1955) What have raonalsts repled? Russell (coldly) I don't understand' Sr Popper (breathessly) He s rght, he s rght- I don't understand t eher!' In a wo he nt s rrelevant because leading ronasts don't understand t I, on he o hand, would st doubng he ntellgence (and perhaps aso he ntelectua hone of rationalsts who don't understand (or pretend not to understand) such a sml pont 8 Cf my sho comments on cove classcaons' n Chapter 1
SVNTEEN
225
obseers attude towards the latter futher separates what s so nmately connected n realty And so we have nally two agences, ste and orderly reason on the one sde, a malleable bu not enrely eldng maeral on the other, and wth ths all the roblems of raonalty that have provded phlosophers wth ntellectual (and, let us not forget, also wth nancal) nourshment ever snce the Rse of aonalsm n the West ne cannot help nocng that the arguments that are sll used to support ths macent result are ndsngushable from those of the theologan who nfers a creator wherever he sees some knd of order obvously order s not nherent n matter and so must have been mposed from the outsde The nteracon vew must therefore be supplemented wth a sasfacto account of the nteracng agences Presented n hs way t becomes a valty For there s no adon no matter how hard headed ts scholars and how hardlmbed ts warrors that wll reman unaected by what occurs around t At any rate -what changes, and how, s now a matter ether for hstocal rearch or for poltil aion carred out by those who parcpate n the nteracng adons shall now state the mplcaons of these results n a seres of theses wth correspondng explanaons We have seen that raonal standards and the arguments supporng them are vsble parts of specal adons conssng of clear and explct prncples and an unnoced and largely unknown but asolutely necessa background of dsposons for acon and j 6ement The standards become objecve measures of excel lence when adopted by parcpants ofadons ofths knd We have then objecve raonal standards and arguments for ther valdty We have fuher seen that there are other adons tha also lead to udgements though not on the bass of explct standards and prncples These value judgements have a more mmedate character, but they are sll evaluaons, just lke those of the raonalst n both cases judgements are made by ndvduals who parcpate n adons and use them to separate Good from E We can therefore state Tradtons are neither good nor bad th smp are bjecvely speakng , e ndependently of parcpaon n a adon, there s not uch to choose between humantaransm and anSemsm orlla raonalty s not an arbter of adons, t s tself a adon or an aspect of a adon t s therefore nether good nor bad, it simply is.
226
AGA NS MTHOD
A tradton sumes rable or unrable prees on wh mpared wth some tradton, e only when vewed by parcpants who see th world n terms of ts values The projecons of these parcpant appear obee and statement descrbng them sound obee because the parcpants and the tradon they project are nowhere menoned n them They aresubee because they depend on the adon chosen and on the use the parcpants make of t The subecvty s noced as soon as parcpants realze that dfferent adons gve rse to derent udgements They wll then have to revs the content of ther value statements just as physcsts revsed the content of even the smplest statement conceng length when t was dscovered that length depends on reference systems and just as evebody revsed the content of down when t was dscovered that the eath s sphercal Those who dont car out the revson cannot prd themselves on formng a specal schl of especally astute phlosophers who have overcome moral relavsm, ust as those who sll clng to absolute lengths cannot prd themselves on formng a specal school of especally astute physcsts who have overcome relavty They are ust pghaded, or badly nformed, or both and mp relats ofprese the nd that se to he be fd Ptagor Protagorean relavsm s reonable because t pays attenon to the plurasm of tradons and valus And t s lzed for t does not assume that ones own vlage and th sange customs t contans are the navel of the world 9 v ve tradton h speal was of ganng ow. Som adons relect about these ways and change them from one group to the next thers take t for granted that there s only one way of makng people accept ther ews Dependng on the adon adopted ths way wll look acceptable, laughable, raonal, foolsh, or wll be pushed asde as mere propaganda Argumnt s propaganda for one obseer, the essence of human dscourse for another v We have seen that dvduals or groups partcpang n th nteracon of tradons may adopt a pragmac phlosophy when judgng the events and sctures that arse The prncples of ther phlosophy often emerge only durng the nteracon peopl change whle obserng change or partcpang n t and the adons they use may change wth them) Ths means thatudgng a hstocal pcs one may use an as yet unspeed and unspeableprace ne may ba 9 Protagoras s dscussed in detail n Chapter I , secons 3 ofFarewe to Re
SEVENTEEN
227
udgements and acons on standards that cannot be specied in advance but are inoduced by the ve udgements (acons) they are supposed to guide and one may even act without any standards, simply following some natural inclinaon. The erce warrior who cures his wounded enemy instead of killing him has no idea why he acts as he does and gives an enrely erroneous account of his reasons. But his acon inoduces an age of collaboraon and peaceful compeon instead of permanent hoslity and so creates a new adion of cmerce between naons The ueson how will you decide what path to choose? How will you know what pleases you and what you want to reject? has therefore at least two answers, . (1) there is no decision but a natural development leading to adions which in reospect give reasons for the acon had it been a decision in accordance with standards or (2) to ask how one wll udge and choose in as yet unknown surroundings makes as much sense as to ask what measuring insuments one will use in as yet uneplored domains Stndards which are intellectual measuring insuments often have to be ted to make sense of new historical situaons just as measuring insuments have constantly to be invented to make sense of new physical situaons. There are therefore at least two dert ways of olee dng an ssue which I shall call a d change and an change respecvely. In the rst case some or all parcipants adopt a well-specied adion and accept only those responses that correspond to its standards. If one party has not yet become a parcipant of the chosen adion he will be badgered, persuaded, 'educated unl he does and then the echange begins. Educaon is separated from decisive debates, it curs at an early stage and guarantees that the grown-ups will behave properly. A ratonal bate is a special case of a guided echange If the parcipants are raonalists then all is well and the debate can start right away If only some parcipants are raonalists and if they have power (an important consideraon!) then they will not take their collaborators seriously unl they have also become raonalists a siety based on raonality is not enrely free; one has to la the game of the intellectuals. 0 open echange, on the other hand, is guided by a pragmac philosophy The adion adopted by the pares is unspecied in the 10 It perhaps hardy necessa to y', says john Stu M, that ths dne (pluralsm of deas and nstuons) i meant to apply only to human bengs n the aturty of ther facules' - e to feow ntellectuas and ther pupls ' Lberty', m ThePh ilos ofJohn St Mi, ed M Cohen, ew ork, 1 96 , p 1 97
228
A G A I N S T M E T H OD
beginnig and develops as the exchange proceeds. The parcipants get iersed into each others ways of thinking feeling perceiving to such an extent that their ideas percepons world-views ay be enrely changed they becoe dierent people parcipang in a new and dierent adion. An open exchange respects the parer whether he is an indivdual or an enre culture whie a raonal exchange proises respect only the fraework of a raonal debate. An open exchange has no organon though it ay invent on there is no logic though new fors oflogic ay eerge in its cours. An open exchange establishes connecons between diernt adions and anscends the relavis of points and iv Howevr it anscends it in a way that cannot be ade obecve but depends in an unforeseeable anner on the (historical psychological aterial) condions in which it curs. (Cf. also the last paragraph of Chapter 16.) . A free soe s a soe n whch a tradtons are eql ghts equal s to edcaton and other postons ofpower Ths is an obvious consequence of i and If adions have advantages oy
fro the point of view of other adions then chsing one adion as a basis of a free society is an arbia act that can be used o b resorng to power. A free society thus cannot be based on a parcular creed; for exaple it cannot be based on raonalis or on huatarian consideraons. The basic sucture ofa free siety s ptee sture, not an ideolo it ncons lke an iron railing not lke a convicon. But how is ths sucture to be conceived? Is it not necessa to bate the atter or should the sucture be si mpose And if it is necessa to debate the atter then should debate not be kept free fro subecve inluences and based on 'obecve consideraons onl? This is how intelectuals y to convince the fellow ciens that the oney paid to the is wel spent and that their ideolo should conue to assue the cen posion it now has. I have already exposed the errors-cum decepons behind the phrase of the 'obecvity of a raonal debate the standards of such a debate are not 'obecve they only aear to be 'obecve because reference to the group that prots fro e use has been oitted ey are lke the invitaons of a clever tyat who instead of saying ' I want you to do . . . or ' I and y wife want you to do . says 'What all of us want is . . or 'what the gods want of us is . . or even better 'it is raonal to do . . . and so sees to leave ut hs own person enrely. It is soewhat depressing to see how an intelligent people have fallen for such a shallow ick. We reove it b obsering
SEVETEEN
229
that a free soe wil not be imposed but wil ee on where
pele gang in an change cf. v above) intdu protee stur of the d aud to Czen naves on a small scae, collaboraon between naons on a larger scale are the developments I have n mnd. e Unted States are not a free set n the sense descrbed here. e ba sling the sture ofafree so are ba t d bat T does not mean that the concrete developments descrbed under the last thes alrea use opn debates, t means that the could use them and that raonasm s not a necesa ngredent of the basc sucture of a free ety. The resuts for scence are obvious Here we have a parcular don, obecvely' on par wth al other adon theses and ) Its results w appear macent to some adons, execrable to others, barely woth a yawn to sl futher adons Of coue, our well-condoned materac contemporares are ble to burst wth exctement over events such as the mnhots, the double he, non-equbrum thermodyncs But let us lk at the mtter from a derent pont of view, and t becomes a rculou exerce n fulty It needed bons of dolars, thouds of wel-aned asssts years of hard work to enable some culae and rther lmted contemporares to perform a few graceless hops n a place nobody n hs rght mnd would thnk of vig a dred out, arless, hot stone But myscs, usng only ther mnds, aveled across th celesal spheres to God hmself, whom they viewed n splendour, recevng sength for conung the lves and enlghtenment for themselves and ther felow men It s only the llterac of the general publc and of ther ste aners, the ntelectuals and ther amazng lack of magnaon that makes them reect such comparsons wthout futher ado A free sety does not obect to such an attude but t wll not permt t to become a basc deolo ether x. A free so insists on the saraton ofs and so. More about ths topc n Chapter 19
1 1 Cf Noan Mer, F the M, London, 1970
8 Yet t s possble to aluate stanr ofratonal and to mpre th e pnpl of mprt are nether abe tradition nor bond change and t s mpossble to nal th .
I shall now illusate some of these results by showing how standards are and have been cricized in physics and asonomy and how ths predure can be extended to other elds Chapter 7 started with the general problem of the relaon between reason and pracce In the ilusaon reason becomes scienc raonality, pracce the pracce of scienc research, and the problem is the relaon between scienc raonality and research I shall discuss the answers given by idealism, naturalism and by a thrd posion, not yet menoned, which I shall call naive anarchism According to iasm it is raonal (proper, in accordance with the will of the gods or whatever other encouraging words are being used to befuddle the naves) to do certai thigs me what m t raonal (proper, etc.) to kill the enemies of the faith, to avoid ad ho hypotheses, to despise the desires of the body, to remove inconsistencies, to support progressive research programmes and so on Raonality usce, the Divine Law) are universal, ndependent of mood, context, historical circumstances and give rise to equ universal rules and standards. There is a version of idealism that seems to be somewhat moe sophiscated but actually is not Raonality (the law, etc) is no longe said to be universal, but there are universally valid condional statements asserng what is raonal in what context and there ae correspondig condional rules. Some revewers have classied me as an idealist in the sense just described with the proviso that I y to replace familiar rules and standards by more revoluona' rules such as proliferaon and counterinducon and almost eveone has ascribed to me a methodolo' with anything goes' as its one basic principle'. But in 20
EIGHTEEN
231
Chapter 2 I say quite explicitly that my intenon is not to replace one set of general rules by another such set: my intenon is, rather, to convnce the reader that, al metholo, the most oous ones, he ther lmts or, to express it in terms ust explained, my intenon is to show that idealism, whether of the simple or of the context dependent knd, is the wrong soluon for the problems of scienc raonality hese problems are not solved by a change of standards but by taking a dierent vew of standards altogether Idealism can be dogmac and it can be crical. In the rst case the rules proposed are rerded as nal and unchangeable; in the second case there s the possibity of discussion and change. But the discussion does not take pracces into account it remans rescted to an absact domain of standards, rules and logic he limtaon of all rules and standards is recozed by nae anarchsm. A naive anarchist says (a) that both absolute rules and context-dependent rules have their limits and infers (b) that all rules and standards are wothless and should be given up Most reviewers rerd me as a naive anarchist in this sense, overlooking the many passages where I show how certain procedures ad sciensts in their research For in my studies of Galileo, of Brownian oon, of the Presracs I not only demonsate the lur of famliar standards, I also to show what not so famliar predures did actually succeed. hus while I agree wth (a) I do not agree wth (b) I argue that all rules have their limits and that there is no comprehensive raonality', I do not argue that we should preed without rules and standards. I also argue for a context ual account but again the contextual rules are not to rce the absolute rules, they are to sult them Moreover I suggest a new relaton between rules and pracces It is this relaon and not any parcular rule-content that characterzes the posion I wsh to defend his posion adopts some elements of naturalsm but it reects the naturalist phiosophy. According to naturalism rules and standards are obtained by an analysis oftradions. As we have seen the problem is which tradion to choose. Philosophers of science will of course opt for science as their basic adion. But science is not one adion, it is many, and so it gives rise to many and partly incompable standards (I have explained this diculty in my discussion of Lakatos). Besides the predure makes it impossble for the philosopher to give reasons for hs choice of science over myth Philoshil P, Vol 2, Chapter 0 Cf aso Chapter 9
232
AGANS MTO
or Aristotle Naturalism cannot solve the problem of scienc raonali As in hapter 1 7 e can now compare the drawbacks of naturalism and idealism and arrive at a more sasfacto view . Naturalism says that reason is completely tenned by research. Of this we retain the idea that research can change reason Idealism says that reason completely g research Of this e retain the idea that reason can change research Combining the two elements we arrive at the idea of a who s of the d and s changed t. This corresponds to the interaconist view of reason and pracce formulated in Chapter 17 and illusated by the example of the map. Now the interaconist view assumes two dierent enes, a disembodied guide on the one side and a well-endowed pracce on the other. But the guide seems dis embodied only because its body', i.e. the ve substanal pracce that underlies it, is not noced and the pracce' seems crude and in need of a guide only because one is not aware of the complex and rather sophiscated laws it contains. Thus the problem is not the interacon of a pracce with something dierent and exteal, but the elmt ofone tradton unr the mp ofoth A look at the way in which science eats its problems and revises its standards' conrms this picture In physics theories are used both as descripons of facts and a standards of speculaon and factual accuracy. Meung nstmts are consucted in accordance with laws and their readings are tested under the assumpon that these laws are correct In a similar way theories giving rise to physical principles provide standards to judge other theo by: theories that are relaviscally invariant are better theories that are not Such standards are of course not untouchable. The standard of relavisc invariance, for example, may be removed when one discovers that the theo of relavity has serious shortcomings Shortcomings are casionally found by a direct examinaon of the theo, for example by an examinaon ofits mathemacs, or its predicve success. They may also be found by the development of alteaves (cf Chapter 3) ie by research that violates the standards to be examined The idea that nature is innitely rich both qualitavely and quantavely leads to the desire to make ne discoveries and thus to a principle of content increase which gives us another standard to judge theores by theories that have excess content over what is already known are preferable to theories that have not. Ain the standard is not untouchable It is in trouble the moment we discover that we inhabit a nite world The discove is prepared by the
GHTN
233
development of Aristotelian' theories which refrain from going beyond a given set of properes it is again prepared by research that violates the standard The procdure used in both cases contains a variety of elements and so there are dierent ways of describing it, or reacng to it. One element and to my mind the most important one is cosmolocal The standard ·we use and the rules we recommend make sense only in a world that has a certain sucture They become inapplicable or start running idle in a domain that does not exhibit this sucture. When people heard of the new discoveries of Columbus, Magellan, Diaz they realized that there were connents, climates, races not enumerated in the ancient accounts and they conectured there might be new connents of knowledge as well, that there might be an America of Knowledge' just as there was a new geographica enty called America', and they ied to discover it by venturng beyond the limits of the received ideas The demand for content increase now became ve plausible It arose from the wish to discover more and more of a nature that seemed to be innitely rich in extent and quality. The demand has no point in a nie world that is composed of a nite number of basic qualies How do we nd the cosmolo that supports or suspends our standards The reply inoduces the second element that enters the revision of standards, theoing in a general sense, including myth and metaphysical speculaon The idea of a nite world becomes acceptable when we have theories describing such a world and when these theories tu out to be better than their innist rivals The world is not directly given to us, we have to catch i through the medium of tradions which means that even the cosmological argument refers to a certain stage of compeon between world-views, theores of raonality included. Now when sciensts become accustomed to eang theories in a certain way, when they forget the reasons for this eaent but simply rerd it as the essence of science' or as an important part of what it means to be scienc', when philosophers aid them in their forgetfulness by systemazing the familiar procedures and showing how they low from an absct theo of raonality then the theories needed to show the shortcomngs ofthe underlying standards will not be inoduced or, if they are inoduced, will not be aken seriously They will not be taken seriously because they clash with customa habits ad systemazaons thereof. For example, a gd way of examning the idea tat the world is nite both qualitavely and quantavely is to develop an Aristotelian cosmolo Such a cosmolo provides means of
234
AGANST METHOD
descrpon adapted to a nte world whle the correspondng methodolo replaces the demand for content ncrease by the demand for adequate descrpons of ths knd Assume we noduce theores that correspond to the cosmolo and deveop them n accordance wth the new rules What wll happen? Scensts wll be unhappy for th theores have unfamlar propertes. Phlosophers of scence wll be unhappy because they noduce standards unheard of n ther professon Beng fond of surroundng ther unhappness wth aras called reasons' they wll go a lttle futher They wll say that they are not merely uhappy, but have arguments' for ther unhappness. The arguments n most cases are elaborate repeons and varaons of the standards they grew up wth and so ther cove content s that of But the theo s ad h!' or But the theores are developed wthout content ncrease!' And all one hears when askng the futher queson why that s so bad s ether that scence has preeded derently for at least 200 years or that content ncrease solves some problems of conrmaon theo Yet the queson was not what scence does but how t can be mproved and whether adopng some conrmaon theores s a gd way of leang about the world. No answer s fothcomg. And so nteresng possbles are removed by rmly nssng on the status quo It s amusng to see that such nsstence becomes the more determied the more crcal' the phosophy that s faced wth the problem We, on the other hand, retan the lesson that the vald, uslns, aqua ofular stanr can be checked on rearch that vot th. A futher example, to llusate the pont. The dea that nformaon conceng the exteal world vels undsturbed via the senses nto the mnd leads to the standard that al knowledge must be checked by obseaon theores that agree wth obeaon are preferabe to theores that do not. Ths smple standard s n need of replacement the moment we dscover that senso nformaon s dstorted n many ways We make the dscove when developng theores tha conlct wth obseaon and ndng tat hey excel n many other respects (Chapters to 1 1 descrbe how Galleo conbuted to the dscove) Fnally, the dea tat thngs are well dened and that we do not lve n a paradocal world leads to the standard that our knowledge must be self-consstent Theores that contan conadcons cannot be part of scence Ths apparently qute fundamental standard whch many phlosophers accept as unhestangly as Catholcs once accepted the dogma of the mmaculate concepon of the Vrgn loses ts authorty the moment we nd that there are facts whose only
EGHTEN
235
adequate descripon is inconsistent and that inconsisent theories may be fruiul and easy to handle whle the attempt to make them conform to the demands of consistency creates useless and unwieldy monsters 2 The last example raises futher queons which are uualy formulated s objecons against it (and againt the cricism of other standards as well, standards of content increase included One objecon is that nonconadicon is a necessa condion of research A procedure not in agreement with this stadard is not research i is chaos It is therefore not possible to examine non condicon n the manner described in the last example The main part of the objecon is the second statement and it is usually supported by the remark that a conadicon implies eve statement This it does but only in rather simple logical systems Now it is clear that changng standards or basic theories has repercussions that must be taken care of Admitng velies larger than the velity of light into relavity and leaving eveg else unchanged gives us some rather puling results such as imagna masses and velocies Admitng welldened posions and momenta ito the quantum theo and leaving everything else unchanged creates hav with the laws of interference Admitng conadicons into a system of ideas aegedly conected by the laws of standard logic and leaving eveg else unchanged makes us assert eve statement Obviously we shal have to make some futher changes, for example we shall have to change some rules of derivaon in the last case Carrying out the change removes the problems and research can preed as planned (Scienc pracce containig inconsistencies is already arranged in the rigt way But says an objecon that is frequently raised at th point how will the results of the research be evaluated if fundamenl standards have been removed? For example, what standards show that research in violaon of content increase produces theories whch are b than their innist rivals as I said a few paragraphs ago? Or what standards show that theories in conict with obseaons have somethig to oer while their obseaonally impeccable rivals have not? Does not a decision to accept unusual theories and to reject familiar ones assume standards and is it not clear, therefore, that cosmological invesons cannot y to provde alteaves to all standards? These are some of the quesons one hears with rng reularity in the discussion of 'fundamentl principles such as 2 . Cf C hapter 16, text to foootes 91
236
AGAIS OD
consistency, content increase, obseaional adequacy, falsiability, and so on It is not difcult to answer them It is asked how research leading o the reision of standards is to be evaluated. For example, when ad on what grounds shall we be sased tha research containng iconsistencies has revealed a fatal shortcoming of the standard of on-conadicon? The queson makes as ltte sense as the queso what measuring insuments will help us to explore an as yet unspecied region of the universe We don't know the region, we cannot sy what wll work in it To advance we must either enter the region, or start makng conjectures about it We enter the region by arculang unusual intellectual, sial, emoonal tendencies, no matter how snge they may seem when vewed through the spectacles of established theories or standards It would certainly be sily to disrerd physicalatur that do not agree with deeply ngraned spiritual noons But it is equally shortsighted to cural fanti that do not seem to t into the physical universe Fantasies and, in fact, the enre subecvity of human beings are just as much a part of the world as leas, stones and quarks and there is no reason why we should change them to protect the latter Simlar consideraons apply to the standards that are supposed to guide our thoughts and acons Tey are not stable and they cannot be stabilized by tying them to a parcular point of view For Aristotle knowledge was qualitave and obseaonal Today knowledge is quantave and theorecal, at least as far as our leading natural sciensts are conceed Who is riht? That depends on what knd of informaon has privileged status and this tu depends on the culture, or the cultural leaders' who use the informaon Many people, without much thought, prefer technolo to harmony with Nature hence, quantave and theorecal informaon is rerded as real' and qualies as apparent' and seconda But a culture that cenes on humans, prefers pesonal acquaintance to absact relaons (intelligence quoents eciency stascs) and a naturalists' approach to that of molecular biologists will say that knowledge is qualitave and will interet quantave laws as bkkeeping devices, not as elements of reality Combining the consideraons f the last two paragraphs we see that even the apparently hardest sienc fact' can be dissolved by decisions undermining the values that make it a fact anor by research that replaces it by facts of a dierent knd This is not a new predure Philosophers fom Parmenides to 20th-century (undialeccal) materialists and scinsts from Galileo and Descartes to Monod used it to devalue, and to declare as mere appearance, the qualitave features of human life But what can be used to support
G
237
cence can also be used anst t The (cultural) measurng nstments that separate real' from appearance' change and must change when we move from one culture to another and from one hstorcal stage to the next, ust as our phscal measurng nstments change and must change when we leave one phscal regon (one hstorcal perod) and enter another
9 S neither a single tradition nor the bt tradition there is ctr pele who he become stomed to its prce its bts and i disantag In a moa it should be saratedfrom the state just a church are now saratedom the state.
I shall now summarze the arguments of the precedng chapters by ng to anser the followng three quesons. 1 . at is sce? How do censts proceed, how do ther standards der from the standards of other enterses? 2. at s so eat about sce? What are the reasons that mght compel us to prefer the scences to other forms of lfe and ways of therng knowledge? 3 How are we to use the sc and who the matter? My answer to the rst queson s that the wde dvergence of ndvduals, schls, hstorcal perods, enre scences makes t exemely dcult to den comprehensve prncpes ether of method, or of fact The word scence' may be a sngle word but there s no sngle enty that corresponds to that word In the doman of method we have scensts lke Salvador Lura who want to e research to events permtng song nferences', predcons that wll be son ly supported and shaly reected by a clear-cut expermental step' Accordng to Lura the experments (Lura and Delbueck, 1943) whch showed that the resstance of bactera to phage nvason s a result of envronment-ndependet mutaons and not of an adaptaon to the envronment had precsely ths character. There was a smple predcon luctuaons, from one culture to the next, of survng colones of bactera on an ar contanng an excess of bacterophages would be small n the rst case, but would contan
f
I SE Lua, A Slot Mhne B Tt Te, ew York, 1985 p 1 1 5
238
IETEE
239
avalanches in the second The predicon could be tested in a simple and sighoard way and there was a decisive result (The result refuted Lamarckism, which was popular among bacteriologists but praccally exnct elsewhere - a rst ndicaon of the complety of science) Sciensts inclined in the manner of Luria show a considerable lack of enthusiasm in the "big problems of the Unverse or of the early Eath or in the concenaon of carbon diode n the upper aosphere/ all subjects that are loaded with weak nferences 3 n a way they are conuing the Aristotelian approach which demands close contact with eerience and objects to following a plausible idea to the bitter end However, this was precisely the predure adopted by Einstein, by researchers in celesal mechanics between Newton and Poincar, by the proponents of atomism and, later, the kinec theo, by Heisenberg during the inial stages of max mechanics and by almost all cosmologists Einsteins rst cosmological paper is a purely theoreical exercise containng not a single asonomical constant The subject of cosmolo itself for a long me found few supporters among physicists Hubble the obseer was respected, the rest had a hard me Jouals accepted papers from obseers giving tem only te most curso refereeing wereas our own papers always ad a s passage to a point were one became quite wo out wit explaining points of matemacs, pysics, fact and logic to te obtuse minds wo constute te mysterious anonymous class of referees, doing teir work, like owls, in te darkness of te nigt5
s it not realy sange, asks Einstein, that human beings are normally deaf to the songest argument whie they are always inclined to overesmate measuring accuracies? - but just such an over esmang of measuring accuracies is the rule in epidemiolo, demography, genecs, specoscopy and in other subects The variety increases when we move into sciences lke cultural 2 ibid , p 1 9 ibid 4 De C 29a24 5 F. Hoyle n Y. Teian and E.. Bilson (eds), Cosmolo and Asthys, Ithaca and London, 1 982, p. 2 1 6 Leer to Bo, quoted fm the BoEnstn Le, ew York, 1971, p 1 92
240
AGAI ST E THO
anthropolo where a compromse has to be found between the eects of personal contact and the idea of an objecve approach on the one side and the praccal needs for quick acon and theorecal thorogess on the other To hear a semnar at a uversit about modes of producon in the moing', writes Robert Chambers, and ten aend a meeng in a eent oce aut acultual eenion in te aften leaes a chzoid feeling One mit not know tat t referred to te same small faers and gt doubt weter eter cussion as anyng to conbute to te oter7
But it not ue that sciensts preed in a methodical wa avod accidents and pay attenon to obseaon and experimen Not aas Some sciensts propose theories and calculate cases which have little or no connecon wth reaity The great gro in teccal achievements which began in the nneteenth cent', we read in L Prandtl's lectures Funmta of Hyd- and Amechan, left cienc knledge far beind Te multudnous proble of pracce could not be anwered by te ydrynamics of Euler; t could not een be iscussed Ts was ciey because sg from Eulers equaon of moon te cience ad become more and ore a purely academic analysis of te ypotecal friconless ideal uid Ts teorecal deelopment is assiated wt te naes of Helmol Kelin Lamb and Rayleig Te analycal result obtained by means of tis called clasical ydrodynamics ually do not agree at all wt te praccal penomena Terefore te engineers put teir ut n a mass of empcal data collecely known a te cence of ydraulcs a branc of knowledge wic grew more and more unlke ydro dynamcs8
According to Prandtl we have a disorderly collecon of facts on th one side, sets of theories starng from smple but counterfactu assumpons on the other and no connecon between the two Mor recently the aomac approach in quanum mechacs an especialy in quanum eld theo was compared by cycal obse to the shakers, a religious sect of New England who buit sold b 7 Rur�, Lonon, 1983, p 29 8 OG Teens, ew Yor 1954, p 3
IETEE
241
and led celibate lives, a non-scienc equivalent of proving rgorous theorems and calculang no cross secons' 9 et in quantum mechanics this apparently useless acvity has led to a more coerent and far more sasfacto codicaon of th facts than had been achieved before, while in hydrodynamics physical commonsense' casionally tued out to be less accurate than the results of rigorous prfs based on wldly unrealisc assumpons early exampe is Maell's calculaon of the viscosity of gases For Maell this was an exercise in theorecal mechanics, an extension ofhis work on the rings of Satu Neither he nor hs contemporaries believed the outcome that viscosity remans constant over a wide range of density and there was cona evidence et more precise measurements conrmed the predicon 10 Few people were pre pared for such a tu of events Mathemacal curiosity had started the work, cross-ferizaon, not general principles, had brought it to a conclusion Meanwhile the situaon has changed in favour of theo In the ses and sevenes, when science was sl n public favour, theo got the upper hand, at universies, where it ncreasngly replaced professional skills, even in medicne, and in special subects such as biolo or chemisy where earlier mohological and substance related research was replaced by a study of molecules In cosmolo a rm belief n the Big Bang now tends to devalue obseaons that clash with it. Such obseaons', writes C Burbidge, are delayed at te rfereeng stage as long as possble wt te ope tat te autor will ge up If tis does not cur and tey are publsed e second lne of defence is to iore tem Iftey ge rse to some comment te best approac s to argue simply tat tey are opelessly wrong and ten if all else fails an obseer may be treatened wit loss of telescope me unl e canges s program 1 1 9 RF Seater and AS Wighan, PC Sn Statt and All ew Yo, 1964, p 10 For quantum mechancs cf secons 4 1 and 42 of ans rmas, Cht Quantum Mechani and Redunm, Berln-ew York, 1 98 1 Maxwell's cacuaons are reproduced in e Stc P ofJam Clk Mell, ed W ven, ew York, 1 965 (rst published in 1 8), pp 377 The conclusion s stated on p 39 1 A remarkable result here presented to us is that fths explanaon ofseous frcon e ue, the coecient of frcon is independent ofthe dens Such a consequence of a mathemacal heory s very staing, and he only experment I have met wh on he sect does not seem to con it' For examples from hydrodnamics cf G rkho, Hydnam, ew York, 1 955, econ 20 and 2 1 Poblem of Cosmogony and Cosmolo', in F Beola, jW Sulenc and DF adore (ed), N I n Astno, Cambridge, 1988, p 229
242
A G A I S T E T O
Thus all we can say s that scensts preed n many derent ways, that rules of method, f menoned explctly, are ether not obeyed at all, or funcon at most lke rules of thumb and that mportant results come from te conluence ofachevements produced by separate an often conlcng ends The dea that ' "scenc knowledge s n some way pecularly posve and free from derences of opnon12 s nothng but a chmaera The stuaon n the arts s qute smlar as a mater of fact t curs n all areas of human acvty Cennno Cennns L e of 130 contans praccal advce based on a rch experence and complex sklls Leon Batsta Albers Del Pura of 143/6 a theorecal ease closely ed to cenl perspecve ad academc opcal theo Perspecve soon became a maa among arts Leonardo and Raphael then ponted out, the one n words, the other praccally cf the sphere on the rght hand sde of hs School ofAth n the Stanza della Seatura of the Vacan), that a pcture that s to be vewed under normal crcumstances, from a comforable but not well-dened dstance and wth both eyes wde open cnnot obey th rules of cenal perspecve They thereby clared the derence between physologcal opcs and geomecal opcs whch Kepler more than a centu later, sll ed to brdge by an easly refuted hypothess cf C 9, text to fooote 0) But cena perspecve remaned a bass on whch varous changes were superposed So far I have been talkng about redure, or method Now methods that are not used as a matter of habt, wthout any thought about the resons behnd them, are often ed to metaphyscal belefs For exple, a radcal form of emprcsm assumes ether that humans are the measure of thngs or that they are n harmony wth them Appled consstently methodologcal rules may produce results whch agree wth the correspondng metaphyscs Luras predure s an exple It dd not fal; t helped to buld a subect whch today at the forefront of research Enstens approach dd not end n dsaster t led to one of the most fascnang mode theores general relavty But methods are not rescted to the area where they scored ther rst umphs Luras requrements, for exaple also tued up n cosmolo; they had been used by Heber Curs n hs 'grand debate wth Harlow Shapely by Ambarzuman, who opposed emprcsm to absact prncples and they are now beng appled by Halton Arp, Marret Geller and ther collaborato Whatever the results, a world bult up n the manner of Lura 12. .R Camp, Fi ofSe, ew Yo, 1 957, p. 21 .
IETEE
243
lttle n common wth the world of Ensten and ths world an ders consderably from the world of Bohr ohann Theodore Merz descrbes n detal how absct world-vews usng cor respondng methods produced results whch slowly ld them wth emprcal content 1 3 He dscusses the followng vews Frst, the ronomcal v, whch rested on mathemacal renements of acon at a dstance laws and was extended (by Coulomb, Neumann, mpre and others) to eleccty and maesm. Laplace's theo of capllarty was an outstandng achevement of ths approach. Secondly, the atomic v, whch played an mportant role n chemcal research (example sterehemsy) but was also opposed by chemsts Trdly, the netc and mechanl v, whch employed atoms n the area ofheat and elecc phenomena For some scensts atomsm was the foundaon of everythng Fourthly, thephysl v, whch ed to acheve unversalty n a derent way on the bass of general noons such as the noon of ener. It could be connected wth the knec vew, but often was not Physcans, physologsts and chemsts lke Mayer, Helmholtz, du Bos Reymond and, n the praccal area, Lebg were outstandng representaves ofths vew n the second half of the 19th centu whle Ostwald, Mach and Duhem extended t nto the 2 Starng hs descrpon of the moholol v, Merz wrtes Te derent aspects of nature wc I ae reewed n te foreng capters and te arous scences wc ae been elaborated by ter ad comprse wat may approprately be teed te absact study of natural objects and penomena Toug all te metods o reasonng wt wc we ae so far become acquanted orgnated pmariy troug obseaon and te reecon oer tngs natural tey ae ts n common tat tey - for te puose of examnaon - remoe ter objects out of te poson and surroundngs wc nature as assed to tem: tat tey str tem Ts press of absacon s eter lterally a press of remoal from one place to noter from te great work - and storeouse of nature erself to te smll workrm te laborato of te expementer or - were suc remoal s not possble - te press s carred out merely in te relm of contemplaon one or two specal properes are noted and desrbed wlst te number of collaterl data are for te moment dsregarded [A trd metod not deeloped at te me s te cre aon of unnaturl condons and tereby te producon of unnatu ral penomena] 13 A to ofEuean ought n the th Ctu rst published 112)
244
AGAIT ETHOD
Tere s moreoer in addon to te aspect of conenence one e powerful nducement for scenc workers to perseere n ter press of absacon Ts s te praccal usefulness of suc researces n te arts and ndusres Te wans and creaons of arcial life ae tus proed te greatest incenes to te absact and arcial eaent of naturl obects and presses for wc e cemcal and electrcal laboratores wit te calculang rm of te matemacian on te one sde and te worksop and facto of te oter ae n te course of te centu become so renowned . Tere s oweer in te uman mind an opposte interest wc founately countercts to a considerable eent te onesided workng of te spirit of absacon scence Ts s the genune loe of nature te consciousness tat we lose all power f to any great eent we seer or weaken tat connecon wc es us to te world as it s - to tings real and natural t nds ts expresson e ancent legend of te mgty gnt wo deried all is sengt from s moter eart and collapsed f seered from er In te study of natural obects we meet [terefore] wt a class of students wo are atacted by tings as tey [Ter] scences are te uly descrpe sciences in opposion to te absact ones 1 4
I have quoted this descripon at length for it shows how dierent predures rest on and provide evidence for, dieren world-vews Finally, Merz menons the getc v, the pchhysl v, the vtalstc v, the statstl v together wth their predures and their ndings What can a single comprehensive world-view of science' or a single comprehensive idea of sce oer under such circumstances? It can oer a suey, a list simlar to the list given by Merz, enumerang the achievements and drawbacks of the various approaches as well as the clashes between them and it can ien science with this complex and somewhat scattered wars on many fronts Alteavely it can put one view on top and subordinate the others to it, either by pseudo-derivaons, or by declaring them to be meaningless Reduconists love to play that me Or it can disregard the dierences and present a paste ob where each parcular vew and the results it has achieved is smoothy connected with the rest thus producing an impressive and coherent edice the' scienc worl-view 1 4. Ibd., Vol. 2, ew York, 1 965, pp. 2f.
IETEE
245
Expressng t derently we may say that the assumpon of a sngle coherent world-vew that underles all of scence s ether a metaphyscal hypothess yng to ancpate a future unty, or a pedagogcal fake or t s an attempt to show, by a udcous up- and downgradng of dscplnes, that a synthess has already been acheved Ths s how fans of unformty preeded n the past (cf Plato's lst of subects n Chapter of hs Rubc), these are the ways that are sll beng used today A more reasc account, however, would pont out that [ t]here s no smple "scenc map of realty or f there were, t would be much t complcated and unweldy to be grasped or used by anyone But there are many derent maps of realty, from a varety of scenc vewonts' 15 It may be obected that we lve n the 20th centu, not n the 19th, and that may uncaons whch seemed mpossble then have been acheved by now Eples are stascal thermodynamcs, molecu lar bolo, quantum chemsy and superstrngs These are ndeed ourshng subects, but they have not produced the unty the phase the' scenc vew of the world nsnuates Actualy, the stuaon s not ve derent from tat whch Merz had noced n the 19th centu Tuesdel and others connue the physcal approach Prandtl maled Euler, Truesdell prases h for havng provded rgorous concepts for research Moholo, though gven a low status by some and declared to be dead by others, has been revved by ecologsts and by Lorenz's study of anmal behavour (whch added s of moton to the older statc s) and t has always been of mportance n gaacc research (Hubble's classcaon) avng been n the doghouse, cosmolo s now beng courted by hgh ener physcsts but clashes wth the phosophy ofcomplementarty accepted by the same group Commenng on the problm M Kafatos and R Nadeu wrte: Te essenal requirement of te Copenagen nteretaon tat te experimental setup must be taken into account wen making obseaons is seldom met in obseaons wt cosmological import [toug suc obseaons rely on ligt te pardi case of complementarty] 1 6
oreover, the obseaons of Arp, M Geller and others have 15 John Zan, Teing and Leaing Aut S and Se , Cmbrdge, 1 980, p. 9 16 Compemenr nd Cosmoo', n M. Kafatos (ed.), Be Quantu m and the Cntins ofthe Une rdrecht, 1980, p. 263.
246
AGAI NST METHO
thrown considerable doubt on the homogeneity assumpon which plays a cenal role in it Extended to 1 ,000 meparsec, Geller's research may blow up the enre subect We have a rabid materialism in some pars molecular biolo, for example), a modst to radcal subecvsm in others some versions of quantum measurement, anthropic principle) There are many fascnag results, specula ons, attempts at interetaon and it is certainly woth knowing them But pasng them together nto a single coherent scienc' world-view, a procedure which has the blessngs even of the Pope this is going too far After all, who can say that the world which so senuously resists unicaon really is as educators and meta physicians want it to be dy, uniform, the same evehere? Besides, as was shown in Chapters 3 , a paste ob eliminates precisely those conicts that kept science going in the past and wil conue nspiring its praconers if preseed At this pont some defenders of uniformity rise to a higher level Science may be complex, they say, but it is sl raonal' Now the word raonal' can either be used as a collecg bag for a variety of predures this would be its nominalist interetaon or it describes a general feature found in eve sngle scienc acon I accept the st denion, but I reect the second In the second case raonality is either dened in a narrow way that excludes, say, the arts; then it also excludes large secons of the sciences Or it is dened in a way that lets all of science surve; then it also applies to love-making, comedy and dogghts There is no way of deimg science' by something songer and more coherent than a list I come to the second queson what's so great about science There are various measures of greaess Pula, ie familiarity with some results and the belief that they are impoat, is one of them Now it is ue that despite periodic swings towards the sciences and away from them they are sll n high repute with the general public or, rather, not the scien, but a mythcal monster science' n th singular n German it sounds even more impressive De Wssscha) For what the general public seems to assume is that th achievements they read about n the educaonal pages of their newspapers and the threats they seem to perceive come rom a sngle source and are produced by a uniform predure They know that biolo is dierent from physics which is dierent from geolo But 17. Cf hs message on he cason of he 3 annvesay of ewon's Pnia, publshed n John Paul on Sce and Relion, oe ame, 1990, esp M6.
IETEE
247
these dscpnes, t s assumed, arse when the scenc way' s appled to derent topcs; the scenc way tself, however, remans the same I have ed to argue that scenc pracce s much more dverse Addng that scensts keep complang about he scenc llteracy of he general publc and that by the general publc' they mean the Weste mddle class, not Bolvan peasants (for example), we have to conclude that the popularty of scence s a ve doubtful matter ndeed What about prail antag The answer s that scence' somemes works and somemes doesn't Some scences (economc theo, for exmple) are n a pret sor shape Others are sucently mobe to tu dsaster nto umph The n so beuse th are not ted to a palar mehod or world-vi The fact that an approach s scenc' accordng to some clearly formulated crteron therefore s no guarantee that t wl succeed Eh e must beudged sarate, especally today, when the fear ofndusal esponage, the wsh to overtake competors on the way to a Nobel Prze, the uneven dsbuon of funds, naonal rvalres, fear of accusaons (of mapracce, plagarsm, waste of funds, etc) put rescons on what some dreamers, many phloso hers among them, sll regard as a free ntellectual adventure' 8 The queson of tth, nally, remans unresolved Love of uth s one of the songest moves for replacng what really happens by a seamlned account, or, to express t n a less polte manner, love of uth s one of the songest moves for lyng to oneself and to others Besdes, the quantum theo seems to show, n the precse manner so much beloved by the admrers of scence, that realty s ether one, whch means there are no obseers and no thigs obseed, or t s many, n which case what s found does not est n tself but depends on the approach chosen
�
What are the vews that are beng compared wth scence when t s declared to be superor? EO Wlson, the father' of sobolo, wrtes relgon endure for a long me as a tal force n sety ke te mytcal gant Antaeus wo dw ener from s moter te eart relgon cannot be defeated by tose wo may cast t down Te srtual weakness of scenc naturalsm s due to te fact tat t as no suc rmal source of power So te me as come to ask does 18 Ths was reaized by govement advsers after the stwar euphoria had wo o See joeph Benavd, Stc Gwth, Berkeley, 1991 , p. 525, quoted above
248
AGAIS T MTHO
a way est to dert te power of religon nto te seces of te great new enterise?
For Wson the man feature of the alteaves s that the have pow I regard ths as a somewhat narrow characterzaon World vews also answer quesons about orgns and puoses whch sner or later arse n almost eve human beng Answers to thse quesons were avaable to Kepler and Newton and were used b them n ther esearch the are no longer avalable toda, at least not wthn the scences The are part of non-scenc world-vews whch therefore have much to oer, also to scensts When West Cvzaon nvaded what s now called the Thrd World t mpoed ts own deas of a proper envronment and a rewardng lfe It thereb dsrupted delcate pattes of adaptaon and created problems that had not esed before Both human decenc and some apprecaon of the man was n whch humans can lve wth nature prompted agents of development and publc health to thnk n more complex or, as some would sa, more relavsc' was There est approaches, the approach called Prma envronmental Care' among them, whch oer lega, polcal and scenc nformaon bu moded n accordance wth the needs, the wshes and, what s most mpoant, the sl and the nowledge of lal populaons 20 Smlarl, he movement called lberaon theolo has moded Church de to brng t closer to the sprtual needs of the pr and dsadvantaged, especall n South Amerca Let me pont out, ncdental, that not al deas whch seem repulsve to the prophets of a New Age come from scence The dea of a world chne and the related dea that nature s materal to be shaped b man should not be blamed on mode, e post-Cartesan, scence It s older and songer than a purel phosophcal de could ever be The expresson world machne' s found n Psedo Donsus Areopagta, a msc of unknown dent who wrote about and had emendous nuence Oresme, who ded n 1382 as bshop of Lseux, compares the unverse to a vast mechan cal clk set runnng b God so that all th wheels move as harmonousl as possble' The senment can be eas understd: 19 uman Nature, Cambridge, Mass, 1972, pp. 192f. 20 Lsns Leaed n Cmmun-Bed Envnm Mant, reedings ofhe 1 rmary Environment Care Workshop, ed Grazia Boni, Inteaona Course for rmary Heah Care Managers at si Level in eveloping Counies, Istuto Superore di Snit, Rome, 199 1 . For a more pular presenaon cf. G Boni, r Environment Care For Enironment Adaes and oli Maers', C Cu, forthcoming
NNTEEN
24
is was the time when mechanical clks of astounding intricacy and elaboraon' were consucted all over Europe eve town was supposed to have one Lynn White Jr, from whose book I have taken is informaon, also describes the change of atude that curred in the Carolingian Age The old Roman Calendars had casonally shown genre scenes of human acty but the dominant adion (which connued n Byantum was to depict the months as passive personcaons ing smbols of attributes The new Carolngan calendar whch set the patte fo the Middle Ages shows a coeive attude towards natural soues The pictures [are about] scenes of ploughing haesng wdchopping ople knking down acos for the pigs pig slaughterng Man and Natu a now o things and man is the master To sum up there is no scienc world-view' just as there is no uniform enterise science' except in the minds of metaphysicians, schoolmasters and policians ying to make their naon compe ve Sll, there are many things we can lea from the sciences But we can also lea from the humanies, from religion and from the remnants of ancent tradions that surived the onslaught ofWeste Civilizaon No area is unied and perfect, few areas are repulsive and compleely without merit There is no obecve principle that could direct us away from the supermarket religion' or the supermarket art' towards the more mode, and much more expensive supermarket science' Besides, there are large areas of nowledge and acon in which we use procedures without any idea as to their comparave excellence A example is medicine, which, ough not a science, has increasingly been connected with scienc research There are many fashions and schools in medicine just as there are many fashions and schools in psycholo It follows, rst, the idea of a comparison of Weste medicine' with other medical edures does not make sense Secondly, such a comparison is often aginst the law, even if there should be volunteers a test is legally impossible Adding to this that health and sicness are culture-dependent concepts we see that there are domains, such as medicine with no scienc answer to queson 2 This is not really a drawback The search for objecve guidance is in conict with the idea of individual responsibility which allegedly is an important 2 1 Medial Tenolo a Sol Change, Oord, 1960, pp. 56f.
250
AGAIS T METHO
ingredient ofa raonal' or scienc age It shows fear, indecision, a yeg for authority and a disrerd for the new opportunies that now est we can build world-views on the basis of a personal choice and thus uite, for ourselves and for our friends, what was once separated by a series of historical accidents 22 On the other hand, we can agree that in a world full of scienc products sciensts may be given a special status just as henchmen had a special status at mes of sial disorder or priests had when being a cizen coincided with being the member of a single universal Church We can also agree that appealing to a chimaera (such as that of a uniform and coherent scienc world-view') can have important polical consequences In 184 Commander Per, using force, opened the ports of Hakodate and Shimoda to American ships for supply and de This event demonsted the milita inferiority of Japan The members of the Japanese enlightenment of the early 1870s, Fukzawa among them, now reasoned as folows Japan can keep its independence on if it becomes songer It can become songer only wth the help of science It will use science eecvely ony if it ds not ust pracse science but also believes in the underlying ideolo To many dional Japanese ths ideolo the' scienc world-view was baaric But, so the folowers of Fuwa argued, it was necessa to adopt barbaric ways, to rerd them as advanced, to inoduce the whole of Weste Civiizaon in order to suive 23 Having been thus prepared Japanese sciens sn branched out as their Weste coleagues had done before and falsied the form ideolo that had sed the development Th lesson I dra from ths sequence of events that a ufom scienc view of the orld' may be useful r i s it gves them movaon wthout tyng them down It is ie a ag Though preseng a single patte it makes people do ny dierent thigs However, it dist r ou (philosophers, y-by-ngh myscs, prophets of a New Age) It suggests to them the most naowminded religious comment and encourages a similar narrowindedness on ther part 22 Wolfgng auli, who was deepy conceed about the inteecua stuaon of the me, demanded that sience and religion again be united: leer to M Fe 8 Aut 1948. I agree ut woud add, enrey n the spit of au, tha the unican shoud a peona maer; it should not preped b phiohicasien acheists ofthe ind nd ied by their inio n educaon. t s derent in the Third World where a song fith s sues.) 23 eis in Cen Blker, eJaneEnlght, Cambdge, 1969. F the a backound cf Chapters 3 nd 4 of ihd Sto, t ofM Jan, Hnworth, 1982.
NINN
251
What I have sad so far already contans my answer to queson 3 a communty wi use scence and scensts n a way that agres wth ts alues and ams and t wll correct the scenc nstuons n ts mdst to brng them closer to these ams he obecon that scence s self-correcng and thus needs no outsde nterference overlooks, rst, that eve enterse s self-correcng (look at what happened to the Catholc Church after Vacan II) and, secondly, that n a democracy he self-correcon of the whole whch es to acheve more humane ways oflvng overules the self-correcon of the parts whch has a more narrow am unless the parts are gven tempora ndeendence Hence n a demracy local populaons not only wll, but also should, use the scences n ays most sutable to them he objecon that czens do not have the experse to judge scenc matters overlks that mportant problems often le across the boundares ofvarous scences so that scensts thn these scences don't have the needed experse ether Moreover, doubtful cases lays produce experts for the one sde, experts for the other sde, and experts n between But the competence of the general publc could be vastly mproved by an educaon that exposes expert fallblty nstead of acng as f t dd not est
20 Thepont ofv unrng ths book not the rult ofa weplanned tran ofthought but ofamts pmpted ntal counte Ang at the wanton ton of ltural h om whch we al could he leaed at the conceted surance wth whch some nteeua ntee wth the l ofpele and contpt r the tre phr th use to belsh ther mse w and st s the moterce behnd my work
The problem ofknoledge and educaon n a free socety rst suck me durng my tenure of a state felloshp at the Wemar nstut zur Methodologschen Eeuerung des Deutschen Theaters (19), whch as a conuaon of the Deutsches Theater Moskau under the drectorshp ofMam Vallen Sta and students ofthe nstut perodcally vsted theaes Easte Germany 1 A specal an . Lke many people ofmy generon I was nvolved n he Second World War. Ths event had lle nuence on my hnkng For me he war was a nusance, not a mor problem. Before he war I had ntended to study asonomy, acng and sngng and to pracse hese professons smulaneously. I had excellent teachers (Adolf Vogel, my sngng teacher had an nteaonal repuaon and aught outsandng oper sngers such as oan Bayley) and had ust overcome some maor val dcules when I receved my drft noce (I was eghteen at he me). How nconvenent I hought. Why he hell should parcpate in he war games ofa bunch ofdots How do I get out of Varous aemps misred and I became a solder. I appled for ocers' anng toavod bullets as long as possble. The aempt was not enrely successl; I was a leutennt before he war had come to an end and found myselfin he mddle ofhe German reeat in oland and hen n East Geany, surrounded by eeng clans, nfan units, anks, olsh aulares whom I suddenly commanded (he higher ocers quc dsappeared when maers became sc). The whole colourl chaos hen appeared to me lke a sage and I became careless. A bulet ht me on my rght hand a second but graed my face, a hrd got stuck n my spne, fell to he ground, unable to rse, but wh he happy hought he war s over for me, now at last can retu to sngng and my beloved asonomy bks' t was ony much later hat became aware of h mor problems ofhe entre age. t seems to me hat hese problems are sl wh us They whenever an ndvdua or a group obeces ts own peona concepo of a lfe and acts accordingy Cf Fare to Re, pp 3. Ths xplains he caso olence of my arguments
252
TWNTY
253
rought us from cty to cty We arrved, dned, talked to the actors, atchd to or ree plays After each performance the publc as sed to reman seated hle e started a dscusson of hat e had us seen. There ere classcal plays, but there ere also ne plays hch tred o analys recent events. Most of the me they dealt th th ork of the ressance n Naz Germany. They ere ndsn gushale from earler Naz plays eulogzng the acvty of the Naz unerground n demrac counes. In both cases there ere deologcal speeches, outbursts of sncerty and dangerous stuaons n the cops and robbers adon. Ths puzzled me and I cmmented on t n the debates ho should a play be structured so that one recozes as presenng the good sde'? What has to be added to the acon to make the struggle of the resstance ghter appear morally superor to the suggle of an llel Naz n Ausa before 1938? It s not sucent to gve hm the rght slons' for then e take hs superorty for granted, e do not sho heren t conssts. Nor can hs noblty, hs humanty' be the dsngushng mark eve movement has scoundrels as ell as noble people among ts folloers. A plarght may of course decde that sophscaon s lux n moral battles and gve a blackhte account. He may lead hs folloers to cto but at the expense of tung them nto barbarans. What, then, s the soluon? At the me I opted for Esensten and ruthless propaganda for the rght cause'. I don't kno hether ths as because ofany deep convcon ofmne, or because I as carred along by events, or because of the macent art of Esensten. Today I ould say that the choce must be left to the audence. The plarght presents characters and tells a sto. If he rrs t shoud be on the sde of sympathy for hs scoundrels, for rumstances and suerng play as large a role n the creaon of evl and vl ntenons as do those ntenons themselves, and the general tendency s to emphasze the latter. The plarght (and hs olleague, the teacher) must not t to ancpate the decson of the audene (of the pupls) or replace t by a decson of hs own f they should tu out to be ncapable of makng up ther o mnds Unr no rmtanc mut he t to be a moral rce. A moral force, whether for od or for evl, tus people nto slaves nd slave, even slave n the serce of The Gd, or of God mself, s the most abect condon of all. Ths s how I see the son today. Hoever t tk me a long me before I arrved at s ve . After a year n Wemar I anted to add the ences and the humanes to the arts, and the theae. I left Wemar and became a sdent (hsto, aula scences) at the famous nstut fur
254
G I N S O
Osterrechsche eschchtforschung hch s part of the Unversty of Venna. Later on I added physcs and astronomy and so nally retued to he subjects I had decded to pursue before the nterrupons of the Second World War. There ere the followng nuences' (1 ) The Kra ire. Many of us scence and engneerng students were nterested n the foundaons of scence and n broader phlosophcal problems. We vsted phlosophy lectures. The lectures bored us and e ere soon thron out because e asked quesons and made sarcasc remarks I sll remember Professor Hentel advsng me th rased arms Herr Feyerabend, enteder se hale das Maul, oder se verlassen den Vorlesungsaal' We dd not gve up and founded a phlosophy club of our on. Vctor Kraft, one of my teachers, became our charman. The members of the club ere mostly students, but there ere also vsts by faculty members and fore dtaries. Juhos, Hentel, Holltscher, von Wrght, Anscombe, Wttgensten came to our meetngs and debated th us. Wttgensten, ho took a long me to make up hs mnd and then appeared over an hour late, ve a sprted performance and seemed to prefer our dsrespectful attude to the fanng admraon he encountered elsehere. Our dscussons started n 1949 and proceeded th nterupons up to 1952 (or 1953). Amost the hole of my thess as presented and analysed at the meengs and some of my early papers are a drect outcome of these debates. (2) The Kraft Crcle as part ofan organzaon called the Atan oege Soe. The Sety had been founded n 1945 by Ausan resstance ghters to provde a forum for the exchange of scholars and deas and so to prepare the polcal uncaon of Europe. There ere semnars, lke the Kraft Crcle, durng the academc year and nteaonal meengs durng the summer. The meengs took place (and sll take place) n Alpbach, a small mountan vllage n Trol. Here I met outstandng scholars, arsts, polcans and I oe my academc career to the frendly help of some of them. I also ben suspecng that hat counts n a publc debate are not arguments but 2 any of them have now become scientsts or engineers Johnny Sagan s rofessor of athematcs at the Unversity of Illnois, Henrch Eichho director of New Haven obseatory, Goldbergerde Buda adviser to eleconic s, whle Erch jantsch, who ded much too soon, met members of our circle at the asonomic a obseatory and later became a of dissident or pseudodissdent scientists, yng to use old tradtons for new puoses 3 Oo olden, brother of Fr olden of the olden publshng house, was fr many years the dynamc leader and organzer
WENY
255
certan ways of prsenng one's case. To test the suspcon I nteened n the debates defendng absurd vews wth great assurance I was consumed by fear after all, I was just a student surrounded y bgshots but havng once attended an ag school I proved the case to my sasfacon. The dcules of tc raonalty were made ve clear by (3) Fel hrha, who arrved n Venna n 1 947. We, the students of physcs, mathemacs, asonomy, had heard a lot about h . We knew that he was an excellent expermenter and that hs lectures were performances on a grand scale whch hs assstan had to prepare for hours n advance. We knew that he had taught theorecal physcs whch was as exceponal for an expermentalst then as t s now. We were also famlar wth the persstent rumours that denounced hm as a charlatan. Rerdng ourselves as defenders of the purty of physcs we looked foard to exposng hm n publc. At any rate our curosty was aroused and we were not dsapponted. Ehrenhaft was a mountan of a man, full of vtalty and unusul deas. Hs lectures compared favourably (or unfavourably, depend ng on the pont of vew) wth the more rened performances of hs colleagues. Are you dumb? Are you stupd? Do you really agree wth eveg I say?' he shouted at us who had ntended to expose h but sat n sent astonshment at hs performance. The queson was more than jused for there were large chunks to swallow Relavty and quantum theo were rejected at once, and almost as a matter of course, for beng dle speculaon In ths respect Ehreha's attude was ve close to that of Stark and Lenard both of whom he menoned wth approval. But he went futher and crczed the foundaons of classcal physcs as well. The rst thg to be removed was the law of nera undsturbed objects nstead of gong n a saght lne were supposed to move n a helx. Then came a sustaned attack on the prncples of elecomaec theo and especally on the equaon dv B = 0. Many years before the fundamental debate he produced convncng evdence for mesoscopc maec monopoles. Then new and sursng properes o lght were demonsated and so on and so foth Each demonsaon was accompaned by a few gently roncal remarks on schl physcs' and the theorecans' who bult castles n the ar wthout consderng the experments whch Ehrenhaft devsed and conued devsng n all elds and whch produced a plethora of nexplcable results. We had soon an opportunty to wess the attude of othodox physcsts. In 1949 Ehrenhaft came to Apbach. In that year Popper conducted a semnar on phlosophy, Rosenfeld and M.H.L. Pce taught physcs and phlosophy of physcs (manly from Bohr's
256
A G A I N S M H O D
comments on Ensten whch had ust appeared), M Harann bolo, Duncan Sandys talked on problems of Brsh polcs, Hayek on economcs and so on. There was Hans Thrrng, the senor theorecal physcst from Venna, a superb teacher who constantly tred to mpress on us that there were more mportant thngs than scence, who had taught physcs to Fegl, Popper as ell as the present author and was an early and ve acve member of the peace movement. Hs son Walter Thrrng, now Professor of Theorecal Physcs n Venna, was also present a ve dsngushed audence and a ve crcal one. Ehrenhaf came well prepared. He set up a fe of hs smple experments n one of the count houses of Apbach and nvted eveone he could lay hands on to have a look. Eve day from two or three n the afteoon parcpants went by n an atude of wonder and left the buldng (f they were theorecal physcsts, that s) as f they had seen somethng obscene. Apart from these physcal preparaons Ehrenhaft also carred out, as was hs hab, a beauful pece of adversng. The day before hs lecture he attended a farly techncal talk by von Hayek on The Senso Order' (now avalable, expanded form, as a book). Durng the dscusson he rose, bewlderment and respect n hs face, and started n a most nnocent voce: Dear Professor Hayek. Ths was a maellous, an admrable, a most leaed lecture. I dd not understand a sngle word. . . .' Next day hs lecture had an overow audence. In ths lecture Ehrenhaft ve a bref account of hs dscoveres, addng general obseaons on the state of physcs. Now, gentlemen,' he concluded umphantly, tung to Rosenfeld and Pce who sat n the front row, hat can you say?' And he answered mmedately. There s nothng at all you can say wth all your ne theores. Stzen mssen se bleben! Sll mssen se se!' The dscusson, as was to be expected, was qute turbulent and t was connued for days wth Thrrng and Popper takng Ehrenhaft's sde aganst Rosenfeld and Pce. Confronted wth the experments the latter occasonally acted as some of Galleo's opponents must have acted when confronted th the telescope. They ponted out that no conclusons could be drawn from complex phenomena and that a detaled anayss was needed. In short, the phenomena were a Dreck a word that as heard qute frequently n the arguments What was our atude n the face of all ths commoon? None of us was prepared to gve up theo or to deny ts excellence We founded a Club for the Salvaon of Theorecal Physcs and started dscussng smple experments. It tued out tha the relaon between theo and experment as much more complex than
WENY
257
shown n textbks and even n research papers. There are a few paradgmac cases where the theo can be appled wthout maor adusents but the rest must be dealt wth by casonally rather doubtful appromaons and aula assumpons 4 I nd t nteresg to remember how lttle eect all hs had on us at the me. We connued to prefer absacons as f the dcues we had found had not been an expresson of the nature of hngs but could be removed by some ngenous devce, yet to be dscovered. Only much later dd Ehrenhaft's lesson snk n and our attude at the me as well as the attude of the enre professon provded me then wth an excellent llusaon of the nature of scenc raonalty. (4) Phl Frank came to Apbach a few years after Ehrenhaft. He undermned common deas of raonalty n a derent way by showng tha the arguments anst Copecus had been perfectly sound and n agreement wth experence whle Galeo's predures were unscenc' when vewed from a mode standpont. Hs obseaons fascnated me and I examned the matter futher. Chapters 8 to 1 1 are a late result of ths study (I am a sow worker). Frank's work has been eated qute unfarly by phlosophers e Puam who prefer smlsc models to the analyss of complex hstorcal events. Aso hs deas are now commonplace. But t was he who announced them when almost veone thought derently (5) In Venna I became acquanted wth some of the foremost Marst ntelectuals. Ths was the result of an ngenous PR ob by Marst students. They tued up as dd we at all maor dscussons whether the subect was scence, relgon, polcs, the theae, or free love. They talked to those of us who used scence to rdcule the rest whch was then my favourte cupaon nvted us to dscussons of ther own and noduced us to Marst thnkers from all elds. I came to know Bethold Vertel, the drector of the Burgtheater Hanns Esler, the composer and musc theorecan, ad alt Hoch, who became a teacher and, later on, one of my best frends. When starng to dscuss wth Holltscher I was a ravng posvst, I favoured sct les of research and had only a ptyng smle for the three basc prncples of daleccs whch I had read n Staln's lttle pamphlet on daleccal and hstorcal materalsm. I was nterested the realst poson, I had ed to read eve book on realsm I could lay hands on (ncludng Klpe's excellen and, of course, Matalm and Empotm) but I found that the arguments for reasm worked only when the realst assumpon had 4. Cf Chapter 5 on h appromaons
258
G I NS OD
already been introduced. Klpe, for example, emphasized the disncon betwen impression and the thing the impression is about. The disncon gives us realism only if it characterizes real features of the world which is the point at issue Nor was I convinced by the emark that science is an essenally realisc enterise Why should science be chosen as an authority? And ere there not posivisc interetaons of science? The so-called paradoxes' of posivism, however, hich Lenin exposed ith such consummate skll, did not impress me at all They rose only if the posivist and the realist mode of speech were mied and they exposed their dierence. They did not show that realism was better though the fact that realism came ith common sense ve the impression that it as. Hollitscher never presented an argument that ould ead, step by step, from posivism into realism and he would have rerded the attempt to produce such an argument as philosophical folly. He rather developed the realist posion itself, illustrated it by examples from science and common sense, shoed how closely it w as connected with scienc research and eveday acon and so revealed its sength. It was of course always possibe to tu a realisc predure into a posivisc procedure by a judcious use of ad hoc hypotheses and ad hoc meaning changes and I did ths frequently, and without shame in the Kraft Circle we had developed such evasions into a ne art) Hollitscher did not raise semanc points, or points of method, as a crical raonalist might have done, he connued to discuss concrete cases unl I felt rather foolish wth my absact objecons. For I sa no ho closely realism was connected wth facts, procedures, principles I valued and that it had helped to bng th about hile positivism merely bed the results a rather complicated way after they had been found: realism had fruits, posivism had none. This at least is ho I would speak today, long aer my realist conversion. At the me I became a realist not because I as convinced by any parcular argument, but because the sum total of realism plus the arguments in favour of it plus the ease wth which t could be applied to science and many other thngs I vaguely felt but could not lay a nr on nally looked bett to me 5 I remember that Rechenbach's answer to ngler's account of elatvty playe an mant pa Dngler extralated frm what could be acheved by smp e mechancal operatons (manufacture of a Eucldean plan suface, for example) whle Rechenbach nted out how the actua sucture of the world would mod th e results of these operations n the large. It s ofcourse ue that Rechenbach's account can be ntereted as a more ecent predctive machne and that t seemed mpressve to me only because I dd not slde nto such an nteretation. ch shows to wht extent the force ofarments depends on ational changes of attude
WNY
25
than the sum total of posivism plus the arguments one could oer for it plus . etc., etc The comparison and the nal decision had much in common with the comparson of life in dierent counies (weather, character of people, melodiousness of language, fd, laws, instuons, weather, e, etc) and the nal decision to take a job and to sart life in one of them Experiences such as these have played a decisive role in my attude towards raonalism Whie I accepted realism I did not accept dialeccs and historcal materialism my predilecon for absact arguments (another posivist hangover) was sll t song for that Today Stalins rules seem to me preferable by far to the complicated and epicyclerdden standards of our mode friends of reason rom the ve beginning of our diussion ollitscher made it clear that he was a communist and that he would y to convince me of the intellectual and sial advantages of dialeccal and historical materialism There was none of the mealymouthed I may b wrong, you may be rightbut together we shall nd the uth ta with which 'crical raonaists embroidr their attepts at indinaon but which they forget the moment their posion is serously endangered Nor did ollitscher use unfair emoonal or intellectual pressures f course, he cricized my attude but our personal relaons have not suered from my reluctance to follow him in eve respect Thi is why Walter ollitscher is a teacher while Popper, whom I also came to know quite well, is a mre propandist At some point of our acquaintance ollitscher asked me whether I would like to become a producon assistant of Brecht apparently there was a posion avaiable and I was being considered for it I declined or a whie I thought that this was one of the biggest mstakes of my life Enriching and changing knowledge, emoons, attudes through the arts now seems to me a much more fruitful enterse and also much more humane than the attemp to inuence minds (and nothing else) by words (and nothig else) Reading about the tensions inside the Brecht Circle, the almost religious attude of some of its members, I now think that I eapd just in me (6) During a lecture (on Descartes) I ve at the Ausan Colege Society I met Elbeth Acombe, a powerl and, to some people, forbidding Brish philosopher who had come to Vienna to lea erman for her anslaon of Wittgensteins works. She ve me manuscripts ofWittgensteins later wrgs and discussed them with me The discussions extended over months and casionally preeded from moing over lunch unl late into the eening They had a profound inuence upon me though it is not at al easy to speci parculars n one casion which I remembr vividly
26
A A N ME I D
Anscombe, by a seres of sklful quesons, made me see ho our concepon (and even our percepons) ofell-dened and apparently self-contaned facts may depend on crcumstances not apparent n them. There are enes such as physcal objects hch obey a conseaon prncple' n the sense that they retan ther denty through a varety ofmanfestaons and even hen they are not present at all hle other enes such as pans and aftermages are annhlated th ther dsappearance. The conseaon prncples may chan from one developmental stage of the human organsm to another and they may be derent for derentlanguages (cf. Whors covert classcaons' as descrbed n Chapter 1 6). I conectured that such prncples ould play an mportant role n scence, that they mght change durng revoluons and that deducve relaons beteen pre revoluona and post-revoluona theores mght be broken o as a result I explaned ths earlyverson ofncommensurablty n Popper's semnar (1 952) and to a small group of people n Anscombe's at n ord (also n 1 952 wth Geach, von Wrght and L.L. Hart present) but I as not able to aruse much enthusasm on ether cason. Wttgensten's emphass on the need for concrete research and hs objecons to absact reasonng (Look, don't thnk!) somehat clashed wth my on tendency toards abstracess and the papers n hch hs nuence s noceable are therefore mxtures of concrete examples and sweepng prncples. 7 Wttgensten as prepared to take me on as a student n Cambrdge but he ded before I arrved. Popper became my supersor nstead. (7) I had met n Apbach n 1948. I admred hs freedom of manners, hs cheek, hs dsrespectful atude toards the German phosophers ho ve the proceedngs eght n more senses than one, hs sense of humour (yes, the relavely unknon Karl Popper of 1948 as ve derent from the establshed Sr Karl of later years) and I also admred hs ablty to restate ponderous problems n smple and joualsc language Here as a free mnd, joyully pung forth hs deas, unconceed about the reacon of the professonals.' Thngs ere derent as rerds these deas themselves The members of our crcle kne deducvsm from Kraft ho had rtten about t before Popper, 8 the falscaonst phlosophy as taken for 6 C Chaper 1 6, tet to fotes 12 7 For detals c my comments on these papers n D Wissschathretche Realm und deA utott Wsscha, Veweg Wesaden, 1978 8 Cf my reew of Kraft's Erkntnlehre in JPS, Vol 1 3, 1963 pp 3 1 9 and esp p 321, second pararaph C also the references n opper, Loc oStc De Mll's Syst oLoc, Vol 2, London, 1879, Chapter 14, ives a dealed account of the predure
WENTY
261
ranted in the physics seminar of the conferenc under the chairmanship of Arhur March and so we did not understand what all the fuss was about. 'Phiosophy must be in a desprate state, we said, 'if ivialie such as thee can cot as major dicoveies Poppr himslf did not seem to thk too much of his phiosophy of scienc at the me for when asked to send us a lst of pubicaons he included the So but not the Loc ofStc D Whil in London I read Wittgensteins Phh tg in detail Being of a rather pedanc tu of mind I rerote the bk so that it looked more ie a ease ith a conuous argent P of this ease was anslated by Anscombe into English and publshed as a review by Phh R in 1955 I also visited Poppers seminar at the LSE Poppr idas were sar to those of Wittgentin but they were more abact and anamc This did not deter m but increasd my o tndncie to absacon and doasm At the end of my stay in ondon Popr invited me to become his assistant I declined despite the fact tat I as broke and did not kno here my next meal as going to come from My decision as not based on any clear rco able an of thought but I gue that aving no xed phiosophy I preferred smbling around in the orld of ideas at my o sed to being guided by the ritual of a raonal debate Agan I as luc. Joseph Agassi ho got the ob did not have much privacy To year later Popper, Schringer and my o big mouth got me a ob in Bstol here I started lectuing on the phiosophy of ience (8) I had studied theae, to, mathemacs, physics and asonomy; I had never studied phiosophy The prpect of having to address a lae audience of eager young people did not exac l my heart ith oy One eek before the lectures sed I sat don and wrote eveg I knew on a piece of paper. It hary led a page Agassi came up ith some excelent advice 'k, Paul, he said, 'the rst line, hs i you rst lecture; the second line, this i your scond lecture - and so on I tk his advice and fared rather el except that my lectures becam a stale colecon of crac from Wittgenstei, Bohr, Popper, Diner, Edon and othe While in Bristol I conued my studies of the quanum theo I found tat important pysical principles rested on methodolocal pons that are violated whenever physics advances physics gets authority from ideas it propates but never obeys in actual research, methodolosts play the role of publity agents hom physicits hire to prae the results but hom th would not pet acce to the enterise itself at falscaom i not a soluon became ve
262
I N S T O
clear n dscussons th Davd Bohm ho gave a Hegelan account of the relaon beteen theores, ther evdence and ther successors 9 The materal of Chapter 3 s the result of these dscussons I rst publshed t n 1961) ° Kuhn's remarks on the omnpresence of anomales tted these dcules rather ncely but I sll ed to nd general rules that ould cover al cases and non-scenc developments as ell To events made me realze the fulty of such attempts. One as a dscusson wth Professor CF von Wezscker n Hamburg 1965) on the foundaons of the quantum theo Von Wezscker shoed ho quantum mechancs arose from concrete research hle I complaned, on general methodologcal grounds, that mportant alteaves had been omtted. The arguments supporng my complant ere qute od they are the arguments summarzed n Chapte 3 but t as suddenly clear to me that mposed thout rerd to crcumstances they ere a hndrance rather than a help a person yng to solve a problem hether n scence or elsehere mt be complete eem and cannot be rescted by any demands, norms, hoever plausble they may seem to the logcan or the phlosopher ho has thought them out n the prvacy of hs study. Norms and demands must be checked research, not by appeal to theores of raonalty. In a lengthy arcle 4 I explaned ho Bohr had used ths phlosophy and ho t ders from more abstract predures. Thus Professor von Wezscker has prme responsblty for my change to anarchsm' though he as not at all pleased hen I told hm so n 1977.
hi
9 have explaned the Hegeliansm ofBom n the essay Aganst ethod' whch appeared in Vol 4 of the Minnota Studir the Philoshy oSce 1970 0 Popper once remarked (in a discssion at the Minnesota Center for the Philosophy of Science n the year 1962) that the example of Brwnan moon is st another version of Dhems example (conict between specc laws sch as Keplers laws and general theores sch as Newtons theory) Bt there is a most imant dierence he deviations from Keplers laws are in prncple obseable (in prnciple meaning given the known laws of nare) while the microscopic deviaons frm the second law of thermodynamics are not (measrng inments are sbeted to the same caons as the thngs the are spposed to measre) Here we nnot do withot an alteatve theory Cf. Chapter 4 fn. 2 I read Khns bk in manscrpt n 1 960 and discssed it extensively with Khn 12 Cf the accont n Repl to Crticism oston Studi Vol 2 1965 13. Cf On the Improvement of the Sciences and the Arts and the Possible Identy of the wo n oston Studi Vol 3 1 967 14 On a Recent Crqe of Complementarty Philoshy o Sce 1968/69 (two parts)
WNY
263
(9) The second event that prompted me to move away from raonalism and to become suspicious of all intellectual pretensions was quite derent To explain it, let me start with some general obseaons. The way in whch sial problems, problems of ener disbuon, ecolo, educaon, care for the old and so on are solved' in First World siees can be roughly described in the following way A problem arises Nothig is done about it People get conceed. Policians broadcast this conce. Experts are called in. They develop theories and plans based on them. Powergroups with experts of their own eectvarious modicaons unl a watered down version is accepted and realized. The role of experts in ths press has gradually increased We have now a situaon where sial and psychological theo of human thought and acon have taken the place of this thought and acon itself. Instead of asking the people involved in a problemac situaon, developers, educators, tch nologists and siologists get their informaon about what these people really want and need' from theorecal studies carried out by their esteemed colleagues in what they think are the relevant elds Not le human beings, but absact models are consulted; not the target populaon decides, but the producers of the models Intellectuals all over the world take it for granted that their models will be more intelligent, make better suggesons, have a better grasp of the reality of humans than these humans themselves. What has ths situaon got to do with me? From 198 to 1990 I was a Professor of Philosophy at the niversity of Califoia in Berkeley My funcon was to car out the educaonal policies of the State of Califoa which means I had to teach people what a small group of white intellectuals had decided was knowledge I hardly ever thought about this funcon and I would ot have taken it ve seriously had I been informed I told the students what I had leaed, I arranged the material in a way that seemed plausible and interesng o me and that was all I did Of ourse, I had also some ideas of my own' but these ideas moved in a fairly narrow domain (though some ofmy friends said even then that I was going bat) In the years around 1964 Mecans, blacks, Indians entered the university as a result of new educaonal policies. There they sat, parly curious, parly disdainful, parly simply confused hoping to get an educaon' What an opporunity for a prophet in search of a followng! What an opporunity, my raonalist friends told me, to onibute to the spreading of reason and the improvement of manknd What a maellous opporunity for a new wave of enlightenment! I felt ve dierently. For it now dawned on me that
264
AGAINST HO
the inicate arguments and the wonderful stories I had so far told to my more or less sophiscated audience might just be dreams, reecons of the conceit of a small group who had succeeded in enslaving eveone else with their ideas Who was I to tell these people what and how to think? I did not know their problems though I knew they had many I was not familiar with their interests, their feelings, ther fears though I knew that they were eager to lea Were the arid sophiscaons which phlosophers had managed to accumulate over the ages and which liberals had surrounded with schmaly phrases to make them palatable the right thig to oer to people who had been rbbed of their land, their culture, their diity and who were now supposed rst to absorb and then to repeat the anaemic ideas of the mouthpieces of their oh so human captors? They wanted to know, they wanted to lea, they wanted to understand the sange world around them did they not desee better nourishment? Their ancestors had developed cultures of their own, colourul languages, harmonious views of the relaon between people, and between people and nature whose remnants are a living cricism of the tendencies of separaon, analysis, self-cenedness iherent in Weste thought These cultures have important achevemens in what is today called sociolo, psycholo, medicine, they express ideals of life and possibiies of human estence. Yet th were ner amined with the rpe th eed except by a small number of outsiders; they were ridiculed and replaced as a matter of course rst by the religion of brotherly love and then by the religion of science or else they were defused by a variety ofinteretaons' Now there was much talk ofliberaon, of racial equality but what did it mean Did it mean the equality of these adions and the adions of the white man It did not Equality meant that the members of dierent races and cultures now had the wonderful chance to parcipate in the white man's manias, they had the chance to parcipate in his science, his technolo, his medicine, his polics These were the thoughts that went through my head as I looked at my audience and they made me recoil in reulsion and terror from the task I was supposed to perform For the task this now became clear to me was that of a ve rened, ve sophiscated slavedriver And a slavedriver I did not want to be Experiences such as these convinced me that intellectua predures which approach a problem through conceps are on the wrong ack and I became interested in the reasons for the emendous power this error has now over minds I started examining the rise of intellectualism in Ancient Greece and the causes that brought it about I wanted to know what it is that makes people who
WNY
265
have a rich and complex culture fall for d absacons and mullate their adions, their thought, their language so that they can accommodate the absacons. I wanted to know how intellectuals manage to get away with murder - for it is murder, murder of mnds and cultures that is commtted year in year out at schls, universies, educaonal mssions in forei counes The end must be reversed, I thought, we must start leaing from those we have enslaved for they have much to oer and, at any rate, they have the right to lve as they see t even if they are not as pushy about their rights and their views as their Weste conquerors have always been. In 19-5 when these ideas rst curred to me I ed to nd an inteeu soluon to my misgivings, that is, I took it for granted that it was up to me and the ies of me to devise educaonal policies for other people I envisaged a new kind of educaon that would live from a rich reseor of dierent points of view permitng the choice of adions most advantageous to the individual. The teacher's task would consist in facilitang the choice, not in replacing it by some uth' of hs own Such a reseoir, I thought, would have much in common with a theatre of ideas as imagned by Pscator and Brecht and it would lead to the development of a great variety of means of presentaon. The objecve' scienc account would be one way of presenng a case, a play another way (remember that for Arstotle agedy is more philosophical' than hsto because i reveals the sture of the historical press and not oy its accidenal details), a novel sll another way Why should knowledge be shown in the garment of academic prose and reasoning? Had not Plato obseed that written sentences in a bk are but ansito stages of a complex press of growth that contains gestures, jokes, asides, emoons and had he not ed to catch ths press by means of the dialogue? And were there not dierent forms of knowledge, some much more detaied and realisc than what arose as raonalism' in the 7th and 6th centu in Greece? Then there was Dism I had studied Dadaism after the Second World War. What attracted me to ths movement was the style its inventors used when not enged in Dadaisc acvies It was clear, lumnous, simple wthout being banal, precise without being narrow; it was a style adapted to the expressio of thought as well as of emoon I connected ths style wth the Dadaisc exercises themselves. Assume you tear language apart, you live for days and weeks in a world of cacophonic sounds, j umbled words, nonsensical events Then, after th preparaon, you sit down and write: the cat is on the mat'. This simple sentence hch we usually utter without thought, lie talng machnes (and much of our is indeed roune), now seems lie the creaon of an
266
AGAINS EHOD
enre world God said let there be light, and there was light Nobody in mode mes has understood the mracle oflanguage and thought as well as the Dadaists for nobody has been able to imagne, let alone create, a world in which they play no role Having dicovered the nature of a lng orr, of a reason that is not merely mechanical, the Dadaists sn noced the deterioraon of such an order into roune . They diaosed the deterioraon oflanguage that preceded the First World War and created the mentality that made it possible. Aer the diaosis their exercises assumed another, more siniser meaning They revealed the frightening similarity between the laguage of the foremost commercial avellers in importance', the language of phiosophers, polcians, theologians, and brute inarculaon The praise of honour, paosm, uth, raonality, honesty that lls our schls, pulpits, polical meengs imperctb me nto nai ton no matter how much it has been wrapped into itera language and no matter how hard its authors y to copy the style o the classics, and the authors themselves are in the end hardly disnguishable from a pack of gunng pigs. Is there a way to prevent such deterioraon? I thought there was I thought that rerding achevements as ansito, rescted andpeonal and eve uth as eated by our love for it and not as found' would prevent the deteroraon of once promising faitales and I also thought that it was necessa to develop a new phiosophy or a new relgion to ve substance to ths unsystemac conjecture I now realize that these consideraons were just another example of intellectuaisc conceit and folly. It is conceited to assume that one has soluons for people whose lives one does not share and whose problems one does not know. It is flish to assume hat such an exercise in distant humanitarianism will have eects pleasing to the people conceed From the ve begnning of Weste Raonalism intelectuals have rerded themselves as teachers, the world as a schl and people' as obedient pupils In Plato ths is ve clear. The same phenomenon curs among Chrisans, Raonaists, Fascts, Marsts. Marsts did not y to lea from those they wanted to liberate; they attacked each other about interetaons viewpoints, evidence and tk it for granted that the resulng intelectual hash would make ne food for the naves (Bakunin was aware of the doctrinarian tendencies of contempora Marsm and he intended to retu all power - power over ideas included - to the people immediately conceed) My own view diered from those just menoned but it was sll a vi, an absact fancy I had invented and now ed to sell without having shared even an ounce of the lives of
WNY
267
he receivers. This I now regard as insuerable conceit. So - what remains Two hings remain. I could follow my own advice to address and t to inuence only those people whom I think I undrstand on a pe rsonal basis. This includes some of my friends; it may include philosophers I have not met but who seem to be interested in simlar problems and who are not too upset by my style and my general approach. It may also include people from dierent cultures who are attracted, even fascinated by Weste science and Weste intellectual ife, who have stared parcipang in it but who sll remember, in hought as well as in feeling he life of he culture hey left behind. My account mght lessen he emoonal tension hey are liable to feel and make hem see a way of uning, raher han pposing to each oher, he various stages of heir lives. noher possibility is a change of subjec. I stared my career as a student of acng, heae producon and singing at he Instute for he Mehodological Reformaon of he German Theae in he German Democrac Republic This appealed to my inellectualism and my dramac propensies. My intellectualism told me hat prblems had to be solved by hought. My dramac propensies made me hink hat hamming it up was better han going hrough an absact argument. There is of course no conict here for argument wihout illusaon leads away from he human elements which aect he most abstract problems The ars, as I see hem today, are not a domain separated from absact hought, but complementa to it and needed to fully realize its potenal. Examning his funcon of he ars and yng to establish a mode of research hat unites heir power wih hat of science and religion seems to be a fascinang enterise and one to which I might devote a year (or two, or hree . )
Postscri on Relaivism
In a crical noce of my bk Fare to Reon drew ugg suggests 'that Feyerabend and lieminded sial crics should eat relavism wth the disdan that they normaly resee for raonal ism 1 T I have now done, in ree Dlo of nowdge, where I say at relavism gives an excellent account o the relao between doac worldviws but is oly a rst step towards an uderstandig of live adions, and in Bo Reon on th Phih ofPaul K Fa, where I wte that 'relaism is as much of a chera as absolusm [the idea that there ests an objcve uth], its cantankerous 3 In the same bok I cal my earlier adce to keep hands o adions an 'idiy 4 In both cases I se objecons aganst relaism, indicate why I changed my and menon some of the reg dcules. Andrew ugg adds that my 'comient to relaism as a geneal theo (or pnpled outlk) is considerably less than total and [that I] can plausib be read as argug that the ouble with adional versions of relasm is that they are pitched at t high a level of absaon 5 This is cy ue of what I say in Farel but ancipaon (which I noce only now, as a result of u comments) cur alrea in Sce n a Free Soe Thre I dnguish between parcipants and exteal obseers of adion deribe objecism as an illusion created by the special posion of the foer and summae my arents in a series of theses, all them pinted in itcs Thesis i reads: Tradions are neither go nor bad, they simply are. Thesis : A adion umes desable or I C Ju ofPh Vo 2 1 p 1 6 - received 989 2 ord 1 5 1 (MS nished 989/) 3 h p 515 (MS nished 989) 4 ibd p. 59 5 . c. 6 ndon 978 pa 1 secon 2 pp 27- reped who chan in Chape 7 o he nd edon ogt Mh ndon 988 and wh added coen Cher 17 225 o he psen edon
268
S S C R I ON R A I I S
269
undesirable properes only when compared with some radion, i.e. only when vewed by participants who see the world in terms of their wn values And so on This sounds like Protagoras, and I say so, in thesis iii Hwever, I then describe (theses v and vi) hw adions interact I dscuss two possibilies, a guided exchange and an open xchange. A guided exchange adopts a wellspecied adion and accept[ s] ony those responses that correspond to its standards. If one party has not yet become a parcipant . he will be badgered, persuaded, educated' unl he does - and then the exchange begins.' A ratonal bate, I connue, is a special cas ofa guided exchange. ' In the case ofan open exchange the parcipants get immersed into each other's ways of thinking, feeling, perceiving to such an extent that their ideas, percepons, worldviews may be enrey changed they become dierent people parcipang in a new and dierent tradion. A open exchange respects the parer whether he is an individual or an enre culture, while a raonal exchange promses respect only within the framework of a raonal debate. An open exchange has no organon though it may invent one; there is no logic though new forms of logic may emerge in its course.' In sum, an open exchange is part of an as yet unspecied and unspeciable pracce These comments imply, rst, that tradions are rarely well dened (open exchanges are going on all the me) and, secondly, that their interacons cannot be understood in general terms. Keeping adions alive the face of exteal inuences we act an only partly conscious way. We can describe results after they have occurred, we cannot incoorate them into a lasng theorecal sucture (such as relavism). In other words, there cannot be any h of knowledge (except as part of a special and fairly stable tradion), there can at most be a (rather incomplete) ht of the ays in which knowledge has changed in the past In my next bk I shall discuss some episodes of such a histo In the meanme I have started using the term relavism' ain, t in a new sense In th second edion of the present bk I xplained ths sense by saying that Sciensts [and, for that matter, all mmbers of relavely uniform cultures] are sculptors of reality.' That sounds like the song programme of the sociolo of science xcpt that sculptors are rescted by the properes of the material they s. Smilarly individuals, professional groups, cultures can reat a wide variety of surroundings, or realies' - but not all 7 op ci p 270 Cf aso he more deaied accon in Reaism and he Hisorici of Knowed 1 989
270
AGAI N H OD
approaches succeed some culturs thrive, others linger for a hile and then decay. Even an obecve' enterise like science which apparently reveals Nature As She Is In Herself inteenes, eliminates, nlares, produces and coies the results in a severely standardzed way - but ain there is no guarantee that the results will congea into a unied world. Thus all we arehd when experimeng, or interferng in less systemac ways, or simply living as part of a welldeveloped culture is how what surrounds us rpon to our acons (thoughts, obseaons, etc.) we not arehd the suundn thse Culture and Nature (or Being, to use a more general term) are always entangled in a fashion that can be explored only by entering into futher and even more complicated entanglements Now, considering that sciensts use dierent and often conadic to methods of research (I describe some of them in Chapter 1 9 of the present edion), that most of these methods are successful and that numerous nonienc ways of life not only surved but proteed and enriched their inhabitants we have to conclude that Being responds dierently, and poste, to many dierent approaches Being i like a person who shows a friendly face to a friendly visitor, becomes ang at an ang gesture, remains unmoved by a re wthout givng any hint as to the prnciples that make Him (er? It? Them?) act the way they do in the dierent circumstances What we nd when living, experimenng, doing research is therefore not a single scenario called the world' or being' or reality' but a variety of responses, each of them constung a special (and not aways welldened) reality for those who have called it foth This is relavism because the type of reality encountered depends on the approach taken However, it diers from the phiosophcal dtrine by admitng faiure not eve approach succeeds In my reply to crics I caled this form of relavism cosmological' relavism, in an arcle publhed in 9 I spoke of an ontological' relavism, in Nature as a Work of Art' 10 I argued that the world of mode science (and not only the descripon of this world) is an artwork consucted by generaons of arsasciensts while in Realism and the Hisoricity of nowledge' 1 1 I indicated how such views are related to the ideas of Niels Bohr. In the last arcle I also menoned 8 In ono Mnevar (ed) d Re Drech-Boson-London 11 p 570 9 No 8 nsJan-Apr 12 10 C Kwledge Vo No 3 13 c oe 7 above
POSSCRIP ON RELAIVISM
271
that ontologcal relavism might be simlar to Thomas Kuhn's more recent philosophy Having before me a copy of Kuhn's Robert and Maurine Rothschild Disnguished Lecture of 1 9 November 1991 I can now describe the similaries and the dierences in greater detai We both oppose the song progrme in the siolo of science As a matter of fact I would say, exactly as Kuhn does, that the clas of the song programme' are absurd an example of deconsucon gone mad' I also agree that it is not enough to underme the authority of the sciences by hstorical arguments why should the authority of hsto be greater than that of, say, physics? we can show historically is that a appeal to scienc authority s into conadicons That undermines any such appeal however, it does not tell us how science should now be intereted or used (Such quesons, I would say, have to be answered by the interested pares theselves, according to their standards, concepons, cultural coients) Kuhn says that the dicules that have seemed to underme the authority of science should not be simply seen as obseed facts about its pracce Rather they are necessa characteriscs of any developmenal or evoluona press' But how do we know that science is an evoluona process rather than a stac way of ng more facts and better laws? Either from obseed facts about its pracce' or from interetaons that are imposed from the outside In the rst case we are back at the situaon Kuhn wants to overcome whie the second case means that science is being incoorated into a wider (cultural) context - a context that values developments - and is intereted accordingly (the predure I menoned in parentheses above). It seems that is what Kuhn really wants, ie he wants to settle the queson philosophcally, not by appealing to facts I would agree if I knew that for him ths is one way among many and not the only possible predure Summaing hs argument Kuhn makes three asserons First, the Archimedian plarm, outside hsto, outside ofme and space, is gone beyond recall.' Yes, and no It is gone as a sucure that can be described and yet shown to be independent ofany descripon. It is ot gone as an unknown background of our estence which aects s but in a way which forever hdes its essence Nor is Archimedianism gone as a possible approach. It would be the polically correct approach in a theocracy, for example. Secondly, Kuhn says that in the absence of an Archimedian laorm comparave evaluaon is all there is' That is of course ue and valy so. Thirdly, he challenges the adional noon of uth
272
AGA NS EHOD
as correspondence to realty. I am not suggesng, let me emphasze, that there is a reality which science fails to get at. My pont s, rather, that no sense can be made of the noon of a reality as it has ordinarily funconed n the phlosophy of science.' Here I aee with the prviso that more metaphysical noons of reality (such as those proposed by Pseudo Dionysus Areopagita) have not yet been disposd of Let me repeat that the cultures that cal foth a cetin realty and these realies themselves are never well dened Cultures change, they interact with other cultures and the indenteness resulng therefrom reected in ther worlds. T hat es intercultural understang and scienc chan possible poten ay eve culture is al cultures We can of course ne a world where cultures wel dened and sctly sepated and where scienc tes have naly been naed down such a world ony acles or revelaon could reform our cmolo.
Index
Alber Leon Batsta 242 nmns 71-3 105 anarchism 49 anyhing gs 14 23 1 epistemological 9 methodolo and 9-10 12-13 19262 naive 47-9, 23 1 polical 12-13 Anamander 95, 185 Anscombe, Ezabeth 25-60 anthropolo 1890, 197 case study of quantum be 191 anythng gs anarchism appearnces 1 95n, 197-8, 203 o naturl interetaons argument emoonsand 117 incommensurabilty and 150 value of 1 5-17, 64 o anarchism; incommensurabilty Aristotle Copecan theo and 72-3 79 cosmolo , 109-12, 135, 141 , 233-4 dynamcs and moon 34n, 121, 135 intuive view ofhumans 124 knowledge & percepon 89 , Halton 242, 245 art archac style and percepon 169-186
perspecve 185n 87 199-201 203215 (Galleo) 39-40n 80 asonomy ancient 36n 135 136 Ptolemac 135-6 mediaeval 135 136 o Copecus Galileo; Newton telescope aula scences 5 1-2, 1 10, 113-14 Bacon, Francis 32n, 60, 1 12-13n, 117-18 Barrow, Isaac 45-6 Bohm, David 262 Bohr, Niels 15n, 3 1-2, 40 Boltzmann, Ludwig Brae, Tycho 55, 135, 142 Brecht, Bertolt 9, 259 Brownan moon (Dr Robert Brown) 27-8, 28n, 262n Bruno, Giordano 127 Burbidge, C. 241 Caap, Rudolph 1 19 Chambers, Robert 20 chldren development of 15-17 perceptual stages 1 50, 1 67-8 Chalmers, Alan 76n China medicne 37 Church & Chrisanity 248 accepts proven scence 132-3 attudes 125-6, 131 Galleo and 124, 125-34
273
274
INEX
clashng with accepted theories see counternducon classical mechncs �7, 72-3, 113 classicaons 166 common sense Copecansm and 103-4, 1 10, 112-13 see o natural nteretaons; obseaon communcaon guided en 45-6, 229 concepts archpatacc 176, 1 7785, 199-203, 21 1 9 chansof 155 integral nature of �1 , 205 totarnsm of 165-9, 199-207 see o dea; counterinducon; naturl nteretaons consstency condon 27, 230, 50,235 Copcus, Nchola acceptance of 96, 145 dees evidence 17, 35, 3940, 51 , 60, 61 eect ofRevoluon' 135-6 Galileo's defence of 72, 77 metholo 1342 syl ofpoess 1 121 theo detached from experence 103-4 1 10, 1 12-13 see o Galilei, Galileo cosmolo 238, 241, 242 archc 186, 1 88, 199, 239 Arstotle and , 109-12, 135, 141 cultul percepons and 1 75-6 ite 233-4 need for reevaluaon 1 1315, 120 seeo onomy; concepts counternducon Copc theo nd 51-3, 7780 tl for reserch 23, 61-4
criss, theo of 145-6 crical onalism 147, 1 5 1-8 see o Popper, Karl; onalism culture 3-4 Dadsm 265-6 Descartes, Ren 54
DueoMethod 49n
dialecc 18 Dogenes ofSnope 62 dscove 1-2, �7, 1 , 1 1 7 seeo scence; theories ofscence Duhem, Perre 24 educaon 2635 scence 2, 1 1-1 2, 161-2 Ehrenaft, Felix 28, 4 7 Enstein, Albert Brownian moon d 28 metholo 11 1 , 42-3n, 138n, 63, 239 theo 18, , 107,242 elecyncs 47 empircism 1 8, 20, 29, 72, 10-0, 1 1 7-18, 147 Aristotle and 109-0 autonomy principle 29 epstemological llusion 73 epstemolo 9, 1 1, 1 2, 136 illusons 73, 138, 156 prejudces 197 essenalsm see patacc aggregtes evaluaon of theores see metholo EvansPrtchard, EE. 189 events, eec on science 1 1 7, 138 experence see nturl interetaons; obseaon experments see empiicim; facts facs autonomy principle 27 collecon & dscove 77, 145 slecon & suppression 21 51,155
INEX
theorecal nature 1 1, 22, 3940, 63 o counterinducon; emprcsm; natul nteretaons falesmyhs & fales fth relgon falsicaon 23, 145, 154, 221 science suated by 51, 154 o crcal onalsm; Popper, Karl Favaro, Antonio 127 Fegl, Heer 1478, 212 Feller, Maret 242, 245-6 Frank, Philipp 257 formlism oc freedom 1 7, 229 Gle, Galileo 17-18, 3940 49, 54, 1 23 39-40n, 80 at arment 657 counternducon nd 7780 dncs & mechancs 25 mn nd 91-9 propanda 65,73,76n, 15,118 Nu 81-2 telescope and opcs 81-5, 8, 99-102 tower rent 5460, 68, 76, 107 als of 125-34 o huch; opeicus, Nicholas Giedymn,J 192-3 Gosched,ohnn 219 goveent and science 163 grammar lngua & linguscs Greeks, ncent 265 art & culture shap views 1 7, 17787 98-202 asonom & cosmolo 36n, 1 87, 1 97-202 o Arstotle; Plato; Paendes
275
Hnsn, NR. 212 Hanfmnn, G.M.S 1 75 Hecltus 95, 203-4 heecn 35 141 Hesse, Ma 36 histo conceptual chnges 1 72, 201 educaon ofence 11-12 evaluaon oftheories 1-2,32, 337, 1078, 21 2 events of, eects on scence 117 mterialism 259 metholo 9-10 51-3, 1478 adons 225, 228 world views 2435 Hollher, Walter 257-9 Homer 1 7782, 203, 2 humntarinsm 3-4 162, 228 Hubble, E.P. 239 Hume, David 50 hypnagc llunaon 1 82n hythess 1 h 115 , 44, 49, 75 conado 115, 23 o deas; theores idealsm 222-3, 231 deas and aon 1 7, 149, 216 232 comparsn to other ideas 21 o concepts; htheses deolo 17-18, 62, 1 84 Il (Homer) 17782, 203, 2 incommensubilty 1 50, 165, 1 author's arval at thess 21 1-13, 262-3 suspenson ofunves 2057 inconssten consten theo; facts ndivdual, delopment of 12 o eedom; huntarinsm induce lc loic nsuments 1 7, 1 10, 232 o telescope ntuon 1 1, 1 50 onalsm 1 58, 2
276
INDEX
Japan 250 uscaon 14 7-9 Kant, Immanuel 51, 55-6, 56n Kaufmann, W 4 Kepler,Johann 24, 94n, 242 t 45, 82, 88, 91n, 98n, 991 polyopa of 88n erkegard, Sren 17, 1 54 knowledge 21 Americao 216 enumeraon vs understanding 186 see o epsemolo Kraft Circle 255 Kropotkn, Peter A 13 Kuhn, Thomas 31, 21 1, 21 2-13 Lagella,Julius Caesar 84 Lakatos, Imre 34n, 1 58n, 162-3 language & lnguiscs classicaon 165 deterioron of 266 eect of 1 64, 1 72, 199 invenon of scienc 1 1 , 18, 63-4, 71 , 150, 193-4 obseaon 57, 59, 64, 71, 1 12 phlosophy 207, 220 relavity prnciple 209 nslang deas & concepts 189,207,20910 law & order see un metholo leag see educaon Lessing, Gotthold 21 920 Lenn, Vladir I 9n, 10 liberty see freedom lnguiscs see anguage & linguscs Loewy, Emmanuel 171 logc anthropolo and 19 conadcon of 15n, 196 flsicao 51 nducve 5, 152 limits speculaon 1 1, 1 06, 192, 195 Lorentz, Hendrk 47
Lorenz, Konrad 131n, 245 Luria, S.E. 239, 22 Lysenko, TrommD. 37, 160 McMulln, Ean 69n, 91 n Maestlin, Mchael 13 5-6, 142-4 Ma, Karl, & Marsm 1 07n mathemacs 50, 136, 195 Maell, ames Clerk 241 Medawar, Peter 131n medcine 38, 249 Melanchthon, Phlip 135 mental sets 199, 202 Me,ohann T 2434 metaphysics, science as 76, 78, 1 14, 12 1, 154, 245 methodolo as the acvity of science 2342 anarchisc 910, 1 2-13, 19, 23 1 counterinducon 23, 51-3, 614 crcal ronalism 14 7, 1 5 18 empiricsm 18, 20, 29, 72, 10910, 1 17-18 falsicaon 23, 51, 145, 155 law & order 13, 19 logc and 1 1 , 15n, 106, 152, 197,221 paence needed with status quo 1115 pluralisc 23, 28,32,33, 108 preconcepons in 1-2, 53 problems of 187-97 ronal 12-13, 1 7, 147 scenc progress and 119, 157-8 see o hypotheses; ronalism scence theories of science Mill,John Suart 1 19 194 describes acceptance of theories 2930,31 L 34n, 38 mnbody problem 64, 1234 model theo see theores of science morlity 16, 253 see o humntaransm
INEX
moon Galileo and 55-60, 65-76, 121 peetual 278, 28n relavty of 69-70n, 71 myh & faitales 21, 36n, 53 natural nteetaons 22, 59, 65, 74 see bo obseaon natulism 222-3, 231-2 Neumn,John von 49n Neuth, o 149 Newton, Isc 35, 4 168 acceptance of theo 25 opcs 46 Nuer poples 189- objecvity 2 1 7, 221-2, 225-6, 228 seeo facts; menl sets obseaon asonomcl vs terresal 85, 892 ocu moves cence om 1035, 1 10, 1 12 historcal chns 52 language 57, 59, 64, 71 , 1 12 natur interetaons 22, 59, 65, 74 senso 57, 234 theories and 39, 149, 1 5 1 see o metholo; percepon; telescope Oss (Homer) 178 ontolo 1 55 1 75-6 ti (epler) 45, 82, 88, 91n, 98n,99-1 opcs 24, 46, 1 378, 242 natur obseaon 87 telescope and 82, 898, 15 Oresme, Nicole 248 patacc aetes 169-75 17786, 1 9202 Paends 43-4, 62, 2 arcipant vs obser 215-21 pasion (eregardn) 7
277
percepon 95, 109 1 14, 152 ofarchc cultures 1 69-7 5 conceptual frwor and 168, 1 75-7 seeo natur interetaon; obseaon; prspecve peetual moon 278, 28n, 262n Pe, Conderatthew 250 perspecve 1 85n, 1 87, 192, 202-3,215,242 Phoshtgs
(Wttnstein) 261 philoph of science 14, 9, 148, 15 1, 154, 197 prac' 217-18, 2278 standards 233 seeo Bacon, Fncis; K, Thoma; theories ofscience physcs 47, 73, 1 13 see o moon; opcs Paget,Jen 167 Plato 71, 151,245 plulsm 23, 29, 32, 33 Pluch 96, 98, 205n Popper, Sr arl R. 34n, 42n, 14 7, 1518 author's acqunce wth 261 Prndtl, L 240 prejudce, recoon 22-3, 58 presupons 58 see o facts: theorecal nature propagnda 11 7, 1 4 Galileo nd 65, 73, 76n, 1 5, 1 1 8 Protago 226 Pseudo Donysus Apropagta 248 psychol see menl sets; percepon Ptolemy (Claudus Ptolmaus) 135, 137 public pacipaon in cinc 2, 21 Parns 35 quantum theo 18, ca stud of 11 Quin, Ward n 22
278
INEX
raonalsm 2, 62, 64, 1 14, 161,246,265 crical 147, 151-8 falsicaon 23, 51 , 145, 154 LawofReason' 12-13 nature of 1 5, 1 1 5-16, 225 obecvity 217, 221-2, 225-6, 228 reconsucon 191 vs scence 2115 standards wthin 230 realism 141-2, 144 naive 59-60, 71 reality assumpons 141-2 relavity Einsten's 119, 1 , 49 Galileo's 73-4, 1 18 religon 1 1 , 53, 137, 218 coestence wth scence seeo Church & Chrsanity research see facts empiricsm methodolo Sacrobosco,Johannes 136 Schwachild soluon 48 scence anarchism 9-10, 1 2-13, 14, 19 aulia 51-2, 110, 113-14 chauvinism n 2-4, 37 coherence vs subdvison 138, 144 conseavsm in 1 1 , 106, 192 cultural inuence 3-4 discove & success 2-4, 47, 106, 1 16, 148 educaon and 2, 1 1- 12, 161-2 moneyand 37 standards 1, 12, 16 1-3, 214, 227-8, 231 , 232-3, 2342 world views 243-5 see o hsto methodolo philosophy of science; theories of scence senses see un natural interetaons; obseaon
s Nun (Galieo) 81-2 speculaon see hypotheses deas;
metaphysics Stalin,Joseph 257 state see veent status quo 1 11 5
Te ture oftc Rolutons
(uhn) 31,212-13 Swss cheese theo 262
technolo 263 telescope celesal dicules 891 Galileo and 81-5, 20 senses and 93-1 02, 103-4 terresal success 83-4 see o opcs tests see empricism metholo Thales 185n theolo see Church & Chrsanity; religon theories ofscence acceptance of 17-18, 32, 63, 77 changng of percepons 23, 52-3, 123-4 consistency condions 27, 230, 50, 235 historical vew 1-2 32, 33-7 incommensurabilty 150, 165, 190,205-7,211-13 uscaon 147-9 as metaphysics 76, 78, 1 14, 1 1 , 154 numercal dsagreement 39-42 qualitave disagreement 39, 42-50 philosophical 1-4, 9, 31, 148, 151 , 154, 196, 212-13 separaon from obseaon 147, 151-8 termnolo 1 1 , 1 8, 63-4, 71, 150202-3 world views 243-5 seeo concepts; histo hypotheses; metholo science
