Linguistics by Jean Aitchison – Chapters 1, 2 & 3
WHAT IS LINGUISTICS Most people spend an immense amount of their life talking, listening, and in advanced societies. Reading and writing. Normal conversation uses 4,000 or 5,000 words an hour. A radio talk, where there are fewer fewe r pauses, uses as many as ,000 or !,000 words per hour. hour. A person reading at a normal speed covers "4,000 or "5,000 words per hour. hour. #o someone who chats for an hour, listens to a radio talk for an hour and reads for an hour possi$ly comes into contact with %5.000 words in that time. &er day, the total could $e as high as "00,000. 'he use of language is an integral part of $eing human. (hildren all over the world start putting words w ords together at appro)imately the same age, and follow remarka$ly similar paths in their speech development. All languages are surprisingly similar in their $asic structure, whether they are found in #outh America, Australia or near the North &ole. *anguage and a$stract thought are closely connected, and many people think that these two characteristics a$ove all distinguish human $eings from animals. . An ina$ility to use language l anguage ade+uately can aect someone-s status in society, and may even alter their personality. personality. ecause of its crucial importance in human life, every year an increasing num$er of psychologists, sociologists, anthropologists, teachers, speech literatures, computer scientists and copywriters /to name $ut a few professional groups reali1e that they need to study language more deeply. deeply. #o it is not surprising that in recent years one of the fastest2e)panding $ranches of knowledge has $een linguistics, the systematic study of " language. *inguistics tries to answer the $asic +uestions -3hat is language- and -how does language work. 6t pro$es into various aspects of these pro$lems such as -3hat do all languages have in common--, -3hat range of variation is found among languages--. -how does human language dier from animal communication-, -7ow does a child learn to speak-, -7ow does one write and analy1e an unwritten language-. -3hy do languages change-. -'o -'o what e)tent are social class dierences re8ected in languageand so on.
What is Language? A person who studies linguistics is usually referred to as a linguist. 'he more accurate term -linguistician - is too much of a tongue2twister to $ecome generally accepted. 'he word -linguist- is unsatisfactory9 unsatisfactory9 it causes confusion. #ince it also al so refers to someone who speaks a large num$er of languages. *inguists in the sense of linguistics e)perts need not $e 8uent in languages though they must have a wide e)perience of dierent types of languages. 6t is more important for f or them to analy1e and e)plain linguistic phenomena such as the 'urkish 'urkish vowel system, or o r :erman ver$s than to make themselves understood in 6stan$ul or erlin. 'hey are a re skilled o$;ective o$servers rather than participants consumers of languages rather than producers as one social scientist 8ippantly commented.
How does linguistics difer rom traditional grammar?
6t is a common fallacy that there is some a$solute standard of correctness which it is the duty of linguists schoolmasters, grammars and dictionaries to maintain. 'here was an uproar in America when in "!>" 3riters 'hird New 6nternational ?ictionary of the @nglish- *anguage included words such as ain't and phrases such as ants is in one's pants. 'he editors were deli$erately corrupting the language or else they were incompetent, argued the critics. -3e$ster 666 has thrust upon us a dismaying assortment of the +uestiona$le, the perverse, the unworthy and the downright outrageous- said one angry reviewer. ut if people say ain't and ants in one's pants , linguists consider it important to record the fact they are o$servers and recorders, not ;udges. -6 am irritated $y the fre+uent use of the words dierent to on radio and other programmers ran a letter to a daily paper. -6n my schooldays of fty years ago we were taught that things were alike to and dierent from, were our teachers so terri$ly ignorant- 'his correspondent has not reali1ed that languages are constantly changing. And the fact that he comments on the frequent use of dierent to indicates that it has as much right to $e classied as -correct- as dierent from. 'he notion of a$solute and unchanging -correctness- is +uite foreign to linguists. 'hey might recogni1e that one type of speech appears through the whim of fashion, to $e more socially accepta$le than others. ut this does not make the socially accepta$le variety any more interesting for them than the other varieties, or the old words any $etter than new ones. 'o linguists the language of a pop singer is not intrinsically worse /or $etter than that of a duke. 'hey would disagree strongly with the Daily Telegraph writer who complained that Ba disc ;ockey talking to the latest Neanderthal pop idol is a truly shocking e)perience of ver$al s+ualorC. Nor do linguists condemn the coining of new words. 'his is a natural and continuous process, not a sign of decadence and decay. A linguist would note with interest, rather than horror, the fact that you can have your hair washed and set in a glamorama in North (arolina, or your car oiled at a lubritorium in #ydney, or that you can $uy apples at a fruitique in a trendy su$ur$ of *ondon. A second important way in which linguistics diers from traditional school grammar is that linguists regard the spoken language as primary, not the written. 6n the past, grammarians have over2stressed the importance of the written word, partly $ecause of its permanence. 6t was diDcult to cope with 8eeting utterances $efore the invention of sound recording. 'he traditional classical education was also partly to $lame. &eople insisted on moulding language in accordance with the usage of the -$est authors- of classical times, and these authors e)isted only in written form. 'his attitude $egan as far $ack as the %nd century .(. when scholars in Ale)andria look the authors of 5th century :reece as their models. 'his $elief in the superiority of the written word has continued for over two millennia. ut linguists look rst at the spoken word, which preceded the write language ten everywhere in the world, as far as we know. Moreover, most writing systems are derived from the vocal sounds. Although spoken utterances and written sentences share many common features, they also e)hi$it considera$le dierences. *inguists therefore regard spoken and written forms as $elonging to dierent, though overlapping systemsE which must $e analysed separately9 the spoken rst, then the written. A third way in which linguistics diers from traditional grammar studies is that it does not force languages into a *atin2$ased frame work. 6n the past, many traditional te)t$ooks have assumed un+uestioningly that *atin provides a universal framework into which all languages t, and countless schoolchildren have $een confused $y meaningless attempts to force @nglish into foreign patterns. 6t is sometimes claime d, for e)ample, that a phrase such as for John is in the -dative case-. ut this is $latantly untrue, since @nglish does not have a *atin2type case system. Al other times, the in8uence of the *atin framework is more su$tle, and so more misleading. Many people
have wrongly come to regard certain *atin categories as $eing -natural- ones. =or e)ample, it is commonly assumed that the *atin tense divisions of past, present and future are inevita$le. Fet one fre+uently meets languages which do not make this near threefold distinction. 6n some languages, it is more important to e)press the duration of an action 2 whether it is a single act or a continuing process 2 than to locate the action in time. 6n addition, ;udgements on certain constructions often turn out to have a *atin origin. =or e)ample, people fre+uently argue that -good @nglish- avoids -split innitives- as in the phrase to humbly apologize , where the innitive to apologize is -split- $y humbly . A letter to the *ondon @vening #tandard is typical of many9 -?o split innitives madden your readers as much as they do me- asks the correspondent. B(an 6 perhaps ask that, at least, ;udges and editors make an eort to maintain the form of our languageC 'he idea that a split innitive is wrong is $ased on *atin. &urists insist that, $ecause a *atin innitive is only one word, its @nglish e+uivalent must $e as near to one word as possi$le. 'o linguists, it is unthinka$le to ;udge one language $y the standards of another. #ince split innitives occur fre+uently in @nglish, they are as -correct- as unsplit ones. 6n $rief, linguists are opposed to the notion that any one language can provide an ade+uate framework for all the others. 'hey are trying to set up a universal framework. And there is no reason why this should resem$le the grammar of *atin, or the grammar of any other language ar$itrarily selected from the thousand spoken $y humans.
The scope of linguistics
*inguistics covers a wide range of topics and its $oundaries are diDcult to dene A diagram in the shape of a wheel gives a rough impression of the range covered /=igure " 6n the center is phonetics, the study of human speech sounds. A good knowledge of phonetics is useful for a linguist. Fet it is a $asic $ackground knowledge, rather than part of linguistics itself. &honeticians are concerned with the actual physical sounds, the raw material out of which language is made. 'hey study the position of the tongue, teeth and vocal cords during the production of sounds, and record and analyse sound waves. *inguists on the other hand are more interested in the way in which language is patterned. 'hey analyse the shape or form of these patterns rather than the physical su$stance out of which the units of language are made. 'he famous #wiss linguist. =erdinand de #aussure e)pressed the dierence well when he compared language with a game of chess. 'he linguist is interested in the various moves which the chessmen make and how they are aligned on the $oard. 6t does not matter whether the chessmen are made of wood or ivory. 'heir su$stance does not alter the rules of the game. Although phonetics and linguistics are sometimes referred to together as -the linguistic sciences-, phonetics is not as central to general linguistics as the study of language patterning. =or this reason, information a$out phonetics has $een placed in an appendi) at the end of the $ook.
6n =igure ", phonetics is surrounded $y phonology /sound patterning, then phonology is surrounded $y synta). 'he term -synta)-, used in its $roadest sense, refers to $oth the arrangement and the form of words. 6t is that part of language which links together the sound patterns and the meaning. #emantics. /meaning is placed outside synta). &honology, synta) and semantics are the -$read and $utter- of linguistics, and are a central concern of this $ook. 'ogether they constitute the grammar of a language /=igure %. ut a word of warning a$out dierences in terminology must $e . 6n some /usually older te)t$ooks, the word -grammar- has a more restricted use. 6t refers only to what we have called the synta). 6n these $ooks, the term -synta)- is restricted to the arrangement of words and the standard term morphology is used for their make2up. 'here is not a case of one group of linguistics $eing right in their use of terminology and the other wrong, $ut of words gradually shifting their meaning with the terms -synta)- and -grammar- e)tending their range. Around the central grammatical hu$ comes pragmatics, which deals with how speakers use language in ways which cannot $e predicted from linguistic knowledge alone. 'his relatively new and fast e)panding topic has connections $oth with semantics, and with the various $ranches of linguistics which link language with the e)ternal world psycholinguistics /the study of language and mind9, socio2linguistics /the study of language and society applied linguistics /the application of linguistics to language teaching, computational linguistics /the use of computers to simulate language and its workings, stylistics /the study of language and literature anthropological linguistics /the study of language in cross2cultural settings philosophical linguistics /the link $etween language and logical thought. 'hese various $ranches overlap to some e)tent, so are hard to dene clearly. &sycholinguistics and sociolinguistics are perhaps the ones which have e)panded fastest in recent years. =or this reason they are given chapters to themselves in this $ook. =inally, there are two important aspects of linguistics which have $een omitted from the diagram. 'he rst is historical linguistics the study of language change. 'his omission was inevita$le in a two dimensional diagram. ut if the wheel diagram is regarded as three dimensional as if it were the cross2section of a tree, then we can include this topic. 3e can either look at a grammar at one particular point in time /a single cut across the tree, or we can study its development over a num$er of years, $y comparing a num$er of dierent cuts made across the tree trunk at dierent places /=igure G. ecause it is normally necessary to know how a system works at any time $efore one can hope to understand changes, the analysis of language at a single point in time, or synchronic linguistics, is usually dealt with $efore historical or diachronic linguistics. 'he second omission is linguistic typology, the study of dierent language types. 'his could not $e tted in $ecause it spreads over several layers of the diagram, covering phonology, synta), and semantics. 'his chapter has e)plained how linguistics diers from traditional grammar studies, and has outlined the main su$divisions within the su$;ect. 'he ne)t chapter will look at the phenomenon studied $y linguistics, language.
WHAT IS LANGUAG *inguistics can $e dened as the systematic study of a language a discipline which descri$es languages in all its aspects and formulates theories as to how it work. ut what e)actly is languages people often use the word in a very wide sense9 -the language of 8owers- the language of music, $ody language and so on. 'his $ook in common with most linguistic $ooks, uses the word to mean the speciali1ed sound signalling system which seems to $e genetically programmed to develop in humans. 7umans can, of course, communicate in numerous other ways they can wink, wave, smile, tap someone on the shoulder, and so on. 'his wider study, which overlaps with a is not the concern of this $ook. 6t is usually known as -the psychology of communication-, and reading on the topic is suggested on p. %"". 6t is also clear that humans can transfer language to various other media9 written sym$ols, $raille, sign language, and so on. #ign language in particular has interesting characteristics which are not all predicta$le from the spoken word. 7owever, language $ased on sound is more widespread, and perhaps more $asic, and so has $een given priority in this $ook. ut can language $e dened And how can it $e distinguished from other systems of animal communication A useful approach was pioneered $y the American linguist (harles 7ockett. 'his is to make a list of design features , and to consider whether they are shared $y other animals. #ome important ones will $e discussed in the ne)t few pages.
Use o sound signals 3hen animals communicate with one another, they may do so $y a variety of means. (ra$s, for e)ample, communicate $y waving their claws al one another and $ees have a complicated series of -dances- which signify the wherea$outs of a source of nectar. ut such methods are not as widespread as the use of sounds which are employed $y humans, grasshoppers, $irds, dolphins, cows, monkeys and many other species. #o our use of sound is in no way uni+ue. #ound signals have several advantages. 'hey can $e used in the dark, and at some distance, they allow a wide variety of messages to $e sent, and they leave the $ody free for other activities. 7umans pro$a$ly ac+uired their sound signaling system at a fairly late stage in their evolution. 'his seems likely $ecause all the organs used in speech have some more $asic functions. 'he lungs are primarily used for $reathing. 'eeth, lips and tongue are primarily for eating. 'he vocal cords /thin strips of mem$rane deep in the throat were used primarily for closing o the lungs in order to make the ri$ cage rigid for actions re+uiring a great eort. 3hen people lift something heavy they automatically hold their $reath. 'his is caused $y the closing of the vocal cords. 'he grunt when the heavy o$;ect is dropped is caused $y the air $eing e)pelled as the vocal cords open. Millions of years ago the possi$ly needed a rigid ri$ cage for swinging in the trees 2$ut humans still need this mechanism today for such actions as weightlifting, defecation and child $irth.
Ar!itrariness 'here is often a strong recogni1a$le link $etween the actual signal and the message an animal wishes to convey. An animal who wishes to warn o an opponent may simulate an attacking attitude. A cat, for e)ample, will arch its $ack spit an appear ready to pounce.
6n human language, the reverse is true in the great ma;ority of cases there is no link whatsoever $etween the signal and the message. 'he sym$ols used are ar$itrary. 'here is no intrinsic connections for e)amples, $etween the word elephant and the animal its sym$oli1es. Nor is the phrase Bthis $ananas are $adC intrinsically connected with food.
The need or learning Most animals automatically know how to communicate without learning. 'heir systems of communication are genetically in$uilt. ee dancing, for e)ample, is su$stantially the same in $ee colonies in dierent parts of the world, with only small variations. @ven in cases where an element of learning is involved, this is usually minor. 6n one e)periment a chaDnch reared in a soundproof room away from other chaDnches developed an a$normal type of song. Fet when the $ird was e)posed to only occasional tape recordings of other chaDnches, its song developed normally. . 'his is +uite dierent from the long learning process needed to ac+uire human language, which is culturally transmitted. A human $eing $rought up in isolation simply does not ac+uire language, as is shown $y the rare studies of children $rought up $y animals without human contact. 7uman language is $y no means totally conditioned $y the environment, and there is almost certainly some type of innate predisposition towards language in a new2$orn child. ut this latent potentially can $e activated only $y long e)posure to language, which re+uires careful learning.
"ualit# Animals who use vocal signals have a stock of $asic sounds which vary according to species. A cow has under ten, a chicken has around twenty and a fo) over thirty. ?olphins have $etween twenty and thirty and so do gorillas and chimpan1ees. Most animals can use each $asic sound only once. 'hat is the num$er of messages, the animal can send -is restricted to the num$er of $asic sounds, or occasionally the $asic solI nd.E plus a few simple com$inations. 7uman language works rather dierently each language has a stock of sound units or phonemes which are similar in num$er to the $asics sounds possessed $y animalsE the average num$er is $etween thirty and forty. ut each phoneme is normally meaningless in isolation. 6t $ecomes meaningful only when it is com$ined with other phonemes. 'hat is, sounds such as f, g, d, o, mean nothing separately. 'hey normally take on meaning only when they are com$ined together in various ways, as in fog, dog, god. 'his organi1ation of language into two layers 2a layer of sounds which com$ine into a second layer of larger units 2is known as duality or dou$le articulation. A communication system with duality is considera$ly more 8e)i$le than one without it, $ecause a far greater num$er of messages can $e sent. At one time, it was thought that duality was a characte"istic uni+ue to human language. ut now some people claim that it e)ists also in $ird2song, where each individual note is meaningless. 6t is the com$ination of notes into longer se+uences which constitutes a meaningful melody.
"is$lacement Most animals can communicate a$out things in the immediate environment only. A $ird utters its danger cry only when danger is present. 6t cannot give information a$out a peril which is removed in time and place. 'his type of spontaneous utterance is nearer to a human $a$y-s emotional cries of pain, hunger or contentment than it is to fully developed language. . 7uman language, $y contrast, can communicate a$out things that are a$sent as easily as a$out things that are present. 'his apparently rare phenomenon, known as displacement, does occasionally occur in the animal world, fore)ample, in the communication of honey $ees. 6f a worker $ee nds a new source of nectar, it returns to the hive and performs a comple) dance in order to inform the other $ees of the e)act location of the nectar, which may $e several miles away. ut even $ees are limited in this a$ility. 'hey can inform each other only a$out nectar. 7uman language can cope with any su$;ect whatever, and it does not matter how far away the topic of conversation is in time and space.
Creati%it# &'roducti%it#( Most animals have a very limited num$er of messages they can send or receive. 'he male of a certain species of grasshopper, for e)ample, has a choice of si), which might $e translated as follows9 ". %. G. 4. 5. >.
6 am happy, life is good. 6 would like to make love. Fou are trespassing on my territory. #he-s mine. *et-s make love.
Not only is the num$er or messages )ed for the grasshopper $ut so are the circumstances under which each can $e communicated. All animals, as far as we know are limited in a similar way. ees can communicate only a$out nectar. ?olphins, in spite of their intelligence and large num$er of clicks, whistles and s+uawks seem to $e restricted lo communicating a$out the same things again and again. And even the clever vervet monkey, who is claimed to make thirty2si) dierent vocal sounds, is o$liged to repeat these over and over. 'his type of restriction is not found in human language which is essentially creative /or productive. 7umans can produce novel utterances whenever they want to.
'atterning Many animal communication systems consist of a simple list of elements. 'here is no internal organi1ation within the system. 7uman language on the other hand, is most denitely not a hap2ha1ard heap of individual items. 7umans do not ;u)taposed sounds and words in a random way instead they ring the changes on a few well dened patterns. 'ake the sounds a, $, s, t. 6n @nglish there are on four possi$le ways in which these sounds could $e arranged $ats, ta$s, sta$ or $ast -inner $ark of lime-. All other possi$ilities, such as Js$ai, a$tts, -st$a are e)cluded /an asterisk indicates an impossi$le word or sentence. 'he starred words are not e)cluded in this case $ecause such se+uences are unpronouncea$le $ut $ecause the rules su$consciously followed $y people who knows @nglish do not allow these com$inations even for new words. A new washing powder called #$ai would $e unlikely to catch on since @nglish does not permit the initial se+uence s$, even though in some other languages /for e)ample ancient :reek this com$ination is not unusual. #imilarly, consider the words $urglar, loudly, snee1ed, the. 7ere again only three
com$inations are possi$le. 'he snee1ed loudly, the ourglar and /perhaps 'he $urglar loudly snee1ed. All others are impossi$le, such as J 'he loudly $urglar snee1ed or J#nee1ed $urglar loudly the. Note also that had the four words $een $urglars, a, snee1es, loudly there is no way in which these could $e com$ined to make a well2 formed sentence. JA $urglars is an impossi$le com$ination, and so is J$urglars snee1es. 6n $rief. @nglish places rm restrictions on which items can occur together and the order in which they come. =rom this it follows that there is also a )ed set of possi$ilities for the su$stitution of items. 6n the word $ats. =or e)ample, a could $e replaced $y e or i, $ut not $y h or 1. which would give J$hts or J$1ts. ln 'he sentence K'he $urglar snee1ed loudly, the word $urglar could $e replaced $y cat, $utcher ro$$er, or even /in a children-s story $y engine or shoe 2 $ut it could not $e replaced $y into, or ama1ingly, or they, which would give ill2formed se+uences such as J'he into snee1ed loudly or J'he ama1ingly snee1ed loudly. @very item in language then has its own characteristic place in the total pattern. 6t can com$ine with certain specied items, and $e replaced $y others /=igure 4.
=ig. 4
*anguage can therefore $e regarded as an intricate network of interlinked elements in which every item is held in its place and given its identity $y all the other items. Noilem /apart from the names of some o$;ects has an independent validity or e)istence outside that pattern. 'he elements of language can $e likened to the players in a game of soccer. A striker, or a goal2keeper, has no use or value outside the game. ut placed among the other players, a striker ac+uires an identity and value. 6n the same way linguistic items such as the, $een, ac+uire signicance part of a total language network.
Structure de$endence now look again at the network of interlocking items which can constitutes language. A closer inspection reveals another, more $asic way in which language diers from animal communication. *ook at the sentences9 'he penguin s+uawked, 6t s+uawked, 'he penguin which slipped on the ice s+uawked. @ach of these sentences has a similar $asic structure consisting of a su$;ect and a ver$ /=igure 5 'he penguin 6t 'he penguin which slipped on the ice
s+uawked
'he num$er of words in each sentence is no guide whatsoever to its $asic structure. #imple counting operations are +uite irrelevant to language. =or e)ample, suppose someone was trying to work out how to e)press the past in @nglish. 'hey would have no success at all if they tried out a strategy such as -Add 2ed to the end of the third word-. 'hey might, accidentally produce a few good sentences such a s9 Lncle 7er$ert toasted seventeen crumpets.
ut more often the results would $e +uite a$surd. (larissa hate frogs2ed. 'he girl who2ed hate frogs scream. 6n fact, it is +uite impossi$le for any$ody to form sentences and understand them unless they reali1e that each one has an inaudi$le, invisi$le structure which cannot $e discovered $y mechanical means such as counting.
And in the sentence, illy swims faster than 7enrietta. 6t is generally agreed that the sentence means -illy swims faster than 7enrietta swims-, and that the second occurrence of swims is -understood-. #uch sophistication is mind2$oggling compared with the thirty2si) cries of the vervet monkey, or even the relatively comple) dances $y which $ees indicate the wherea$outs of honey to their colleagues.
Human language %ersus animal communication #o far, the main similarities and dierences $etween human and animal communication can $e summed up as follows9 7uman language is a signaling system which uses sounds, a characteristic shared $y a large num$er of animal systems. 6n animal communication, there is fre+uently a connection $etween the signal and the message sent, and the system is mainly genetically in$uilt. 6n human language, the sym$ols are mostly ar$itrary, and the system has to $e painstakingly transmitted from one generation to another. ?uality and displacement 2 the organi1ation of language into two layers, and the a$ility to talk a$out a$sent o$;ects and events 2 are e)tremely rare in the animal world. No animal communication system has $oth these features. (reativity, the a$ility to produce novel utterances, seems not to $e present in any natural communication system possessed $y animals. =inally, patterning and structure dependence may also $e uni+ue language features.
'o summari1e language is a patterned system of ar$itrary sound signals characteri1ed $y structure dependence, creativity, displacement, duality and cultural transmission. 'his is true of all languages in the world, which are remarka$ly similar in their main design features. 'here is no evidence that any language is more -primitive- than any other. 'here are certainly primitive cultures. A primitive culture is re8ected in the voca$ulary of a language, which might lack words common in advanced societies. ut even the most primitive tri$es have languages whose underlying structure is every $it as comple) as @nglish or Russian or (hinese.
)rigin and unctions o language *anguage, as we have seen, seems to $e a highly developed form of animal signaling. ut there is a missing link in the chain. 7ow, and when, did we start to talk 'his is a pro$lem of interest mainly to ethologists /students of animal $ehaviour, and one which has not yet $een solved. Many linguists regard this fascinating topic as $eing outside the realm of linguistics proper. 'hey are more interested in studying actual language than in speculating a$out its remote origins. ut although how language $egan is a pu11le, why language $egan seems rather clearer. &ossi$ly it $egan $ecause humans needed a greater degree of cooperation with each other in order to survive, and this cooperation re+uired eDcient communication. (onse+uently, the primary function of language is *o impart factual information and to convey essential commands. ut language9 can also $e used to communicate feelings and emotions. 'his aspect of language is not as well developed is -information talking-, $ecause humans, like other primates, can convey emotions $y screams, grunts, so$s, gestures and so on. #o they need language only to conrm and ela$orate these more primitive signals. 6n addition, there is the language of social chit2chat, the meaningless small talk of everyday life. -7ello, how nice to see you. 7ow are you 6sn-t the weather terri$le- 'his social patter has $een called phatic communion and is primarily a device to maintain social contact on a friendly level. #ome ethologists call it -grooming talking- and suggest that it is a su$stitute for the friendly grooming indulged in $y monkeys. 'here are other $iologically less important functions of language. 7umans may use language for purely aesthetic reasons. 6n writing poe try, for e)ample, people manipulate words in the same way as they might model, clay or paint a picture.
TH STU"* )+ LANGUAG 'he discipline of linguistic can $e linked to a pathway which is $eing cut through the dark and mysterious forest of language. ?ierent parts of the forests have $een e)plored at dierent times so we can depict the past as a winding one.
As =igure > shows, they had $een three ma;or directions in linguistics in the past two centuries. *et us discuss each of these in more detail.
,-th centur# historical linguistics efore the "!th century, language in the western world was of interest mainly to philosophers. 6t is signicant that the :reek philosophers &lato and Aristotle made ma;or contri$utions to the study of language. &lato for e)ample comma is said to have $een the rst person to distinguish $etween nouns and ver$s. "> is the year which many people regard as the $irthdate of linguistics. 6n that year, an @nglishman, #ir 3illiam ones, read a paper to the Royal a Asiatic #ociety in (alcutta. Now. #anskrit /the old 6ndian language, :reek, *atin, (eltic and :ermanic all had striking structural similarities. #o impressive were this likenesses that these languages must spring from one common source, he concluded. Although ones has the credit of making this discovery, it was an idea that was occurring independently to several scholars at the same time. #ir 3illiam ones discovery red the imagination of a scholar. =or the ne)t hundred years, all other linguistic work was eclipsed $y the general preoccupation with writing comparative grammar, grammar which rst compared the dierent linguistic forms found in the various mem$ers of the 6ndo2 @uropean language family, and second, attempted to set up a hypothetical ancestor, &roto26ndo2 @uropean, from which all these languages were descended. /=igure e)clude 7ittie and 'ocharian, which were not recogni1ed as 6ndo2 @uropean languages until the %0th century. 'he "! century concern with reconstructing &roto26ndo2@uropean, in making hypothesis a$out the way it split into the various modern languages, was encouraged $y the general intellectual climate of the times. 6n the mid2"!th century, ?arwin pu$lished his famous origin of species, putting forward the theory of evolution. 6t seemed natural to attempt to chart the evolution of language alongside the evolution of #pecies. 'his emphasis on the language change eventually lead to a ma;or theoretical advance. 6n the last +uarter of the century, a group of scholars centered around *eip1ig, nicknamed B'he grammariansC, claimed that the language change is Kregular. 'hey Argued that is, in any word of a given dialect, one sound changes into another, the change we also take all the other occurrences of the same sound is similar for phonetic surroundings. =or e)ample, in old @nglish the word Kchin pronounced Kkin /spelt cinn. 'his change from k2sound to ch aected all the other k2sounds which occurred at the $eginning of a word $efore e or i. #o we also get chicken, child, chide, chip, chill, cheese, cheek, chest, chew and so on 2 all of which originally had a k2sound at the $eginning. Although, today, the claims made $y the Foung grammarians had $een modied to some e)tent /as will $e discussed in chapter "%. 6t was an important step forward for linguistics to reali1e that language changes were not ;ust optional tendencies, $ut denite and clearly stata$le the laws /at the young grammarians, perhaps, misleadingly called them. 'he in8uence of the "! century scholars was strong. @ven today, one still meets mem$ers of the general pu$lic who e)pect the cataloguing of linguistic changes and the reconstruction of &roto26ndo2@uropean to $e the central concern of modern linguistics.
arl# to mid./0th centur# descri$ti%e linguistics
6n the "0th century, the emphasis from language change to language description. 6nstead of looking have a selection of items changed in a num$er of dierent languages, linguists $egan to concentrate on descri$ing single languages at one particular point in time. 6f any one person could held responsi$le for this change of emphasis, it was the #wiss scholar =erdinand de #aussure /"52"!"G, who is sometimes la$elled Kthe father of modern linguistics. Ama1ingly he died without having writing any ma;or work on general linguistics. ut his a student collected together his lecture notes after his death and pu$lished them under the title Kcourse in general linguistics "!"5, which e)erted a ma;or in8uence on the course of linguistics, &articularly in @urope. ?e #aussures crucial (ontri$ution what is e)plicit and reiterated statement that all language items are essentially interlinked. 'his was an aspect of language which have not $een stressed $efore. No$ody had seriously e)amined the relationship of each element to all the others. As Note earlier, it was #aussure who rst suggested that language is like a game of chess, a system in which each item is dened $y its relationship to all the others. 7is insistence that language is a carefully $uild structure of interwoven elements initiated the era 0. *et us Now Return to this. 6n America, linguistics $egan as an oshoot of anthropology. Around the $eginning of the %0th century, anthropologist were eager to record the culture of the fast2dying in American26ndian tri$es, and the American 6ndian languages with one aspect of this. Adopt an interesting, the work of those early #cholar was, for the most part, hapha1ard and lacking cohesion. 'here were no rm guidelines for linguists to follow when they attempted to descri$e e)otic languages. 'his state of aairs changed with the pu$lication in "!GG of *eonard loomelds comprehensive work entitle simple language. loomeld considered the linguistics should deal o$;ectively and systematically with o$serva$le data. #o he was more interested in that way items were arranged than in meaning. 'he study of meaning 3as Not amena$le to rigorous methods of analysis and was therefore, he concluded, Kthe weak part in language study, and will remain so until human knowledge advance is very far $eyond its present state. loomeld had immense in8uence 2 far more than the @uropean linguists working during this period 2 in the so2called lommeldian era lasted for more than %0 years. ?uring this time, large num$ers of linguists concentrated on writing descriptive grammar of unwritten languages. 'his involved rst nding native speakers of the language concerned and collecting sets of utterances from them. #econd it involved analy1ing the corpus of collected utterances $y studying the phonological and syntactic patterns of the language concerned, as far as possi$le without recourse to meaning. 6tems were /in theory identied and classied solely on the $asis of their distri$ution within the corpus. 6n the (urse of writing such grammars, a num$ers of pro$lems arose which cannot $e solved $y the method proposed $y loomeld. #o an enormous amount of attention was paid to the renement analytical techni+ues. =or many, the ultimate goal of linguistics with the perfection of ?iscovery procedures 2 a set of principles which would ena$le a linguist to discover /or perhaps more accurately, Kuncover in a fool proof way the linguistic units of an unwritten language. ecause of their overriding interest in the
internal patterns of the structure of the language, search linguists are sometimes la$elled Kstructuralists.
1id to late /0th centur# linguistics and the search o uni%ersals 6n "!5, linguistics took a new turning. Noam (homsky, then aged %!, a teacher at the Massachusetts 6nstitute of technology, pu$lished a $ook called syntactic structures. Although containing fewer than "%0 pages, this little $ook started a revolution in linguistics. (homsky is, argua$ly, the most in8uential language of the century. (ertainly, he-s the linguistic whose reputation has spread farthest outside linguistics. 7e has, in the opinion of many, transformed linguistics from a relatively o$scure discipline of 6nterest mainly to &h? students in future Missionaries into a ma;or social science of direct relevance to psychology, sociology, anthropology, philosophers and others. (homsky has shifted attention away from detailed description of actual utterances, and started asking +uestions a$out the nature of the system which produces the output. According to (homsky, loomeldian linguistics was $oth are too am$itious and far limited in scope. 6t was too am$itious in that it was unrealistic to e)pect to $e a$le to lay down fullproof rules for e)tracting a perfect description of a language from a mass of data. 6t was too limited $ecause it concentrated on descri$ing set of utterances which happened to have $een spoken. A grammar, he claimed, should $e more than a description of old utterances. 6n short, the traditional viewpoint that the main task of linguists is simple to descri$e the corpus of actual utterances cannot account for the characteristics of productivity, or creativity, as (homsky prefers to call it. 'hese, as we noted i n chapter %, is the a$ility of human $eing to produce and (omprehend an indenite num$er of novel utterances. (homsky points out that anyone who knows the language must have internali1ed a set of Rules which specify the se+uence is permitted in their language. 6n his opinion, a linguists task is to discover this rules, which constitutes the grammar of language in +uestion. (homsky therefore uses the word Kgrammars interchangea$ly to mean, on the one hand, a person-s internali1ed rules, and on the other hand, a linguist guess as to these rules. 'his is confusing, as the actual rules in a person-s mind are unlikely to $e the same as a linguists hypothesis, even though they would pro$a$ly $e some overlap. A grammar which consists of a set of statements or rules which specify which se+uences of a language are possi$le and which impossi$le, is a generative grammar. (homsky, therefore, initiated the era of generative linguistics. 6n his words, a grammar would $e Ka device which generates all the grammatical se+uences of a language and none of the ungrammatical ones. #uch a :rammar is perfectly e)plicit, in that nothing is left to the imagination. 'he rules must $e precisely formulated in such a way that anyone would $e a$le to separate the well2formed sentences from the ill2formed ones, even if they did not know a word of the language concerned. 'he particular type of generative grammar favored $y (homsky is a #o (alled transformational one. 'he $asic characteristics of 'ransformational :enerative :rammar /':: are outlined in chapters "4 to ">.
