Mind Association
Is Existence a Predicate? Author(s): Murray Kiteley Source: Mind, New Series, Vol. 73, No. 291 (Jul., 1964), pp. 364-373 Published by: Oxford University Press on behalf of the Mind Association Stable URL: http://www.jstor.org/stable/2251942 . Accessed: 12/12/2014 13:50 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact
[email protected].
.
Oxford University Press and Mind Association are collaborating with JSTOR to digitize, preserve and extend access to Mind.
http://www.jstor.org
This content downloaded from 193.63.81.241 on Fri, 12 Dec 2014 13:50:19 PM All use subject to JSTOR Terms and Conditions
IV.-IS
EXISTENCE
A PREDICATE?
BY MURRAYKITELEY KANT'Slaconic observation that existence is not a predicate has enjoyed an almost spotless reputation. Few philosophical dicta have been this fortunate. The fortunes of this dictum may rest on the incontestability of the arguments which have been given for it and there are several of them. I should like to look at four of these arguments. I think they all fail, some more seriously than others. Their failure is, however, instructive. 1. The First Argument Professor Malcolm has recently made this observation about existence again, being careful all the while to disengage simple from necessary existence, so that the latter might not be touched with the impredicative taint of the former. He cannot, he laments, find a rigorous proof of the doctrine that existence is not a predicate, but Kant's reasoning should, he hopes, be sufficient to convince.' The Kant-Malcolm argument is of the you'll-seeit-to-be-quite-obvious-if-you-just-look-at-it-from-this-angle type. Kant's angle is got by the comparison of 100 existing with 100 non-existent Thalers ; 2 Malcolm's by the king's specifications of the desired qualities to be sought in a chancellor, existence occurring last on the list (pp. 43 and 44). Kant wants us to ask, " Are there more Thalers in the first than in the second ? " and Malcolm wants us to ask, " Can qualifications for ministerial appointment include existence ? " (" No non-existent candidates need apply.") To both questions we can only answer " No ". But what do these cases show ? Consider the following two. Say that you were ordering a shipping box for books from a carpenter and among the specifications, size, wood, strength, you included non-emptiness. "I don't want a non-empty box delivered, is that clear ? " "But what," the inarticulate carpenter might ask, " has that got to do with making boxes ? Non-emptiness, he might have said, is not a real predicate. Consider a candy manufacturer. Does he, make two kinds of chocolates, packaged and unpackaged ? Packagedness, Kant1 Norman Malcolm, " Anselm's Ontological Arguments ", Philosophical Review, lxix (1960), 42-44. 2 The Critique of Pure Reason, trans. by Norman Kemp Smith (London, 1929), p. 505. 364
This content downloaded from 193.63.81.241 on Fri, 12 Dec 2014 13:50:19 PM All use subject to JSTOR Terms and Conditions
IS EXISTENCE
A PREDICATE
?
365
Malcolm might say, is not a real predicate. Only size, filling and topping make a difference among chocolates; packages are coverings not characteristics. If you look at existence from these angles, you are likely to agree with the orthodox opinion that existence is not a predicate. But if you look at it from other angles, you are liable to entertain doubts. You can imagine all sorts of cases where what is wanted is to know whether or not some thing does exist. There are many times where the important thing is to find out, not whether it has a short or long snout, but whether or not it is extinct or extant. You do not ask the White Hunter to lead you to the habitat of an existent wildebeest, but you might instruct a research assistant to track down Greek opinions on existent (rather than mythical) animals. Kant's 100 real Thalers add up to no more, of course, than 100 imaginary Thalers. Existence and non-existence make no difference in counting. But that does not mean that they make no difference. You cannot, e.g. deposit imaginary Thalers in your bank account. I do not wish to deny that there is something in this line of reasoning. It does have, as Malcolm indicates, a certain intuitive appeal, but this appeal rests on looking at the right examples and drawing the right conclusion from these examples. There are other examples which might have just the opposite effect. 2. The Second Argument This argument goes as follows: (1) If existence is a predicate, then all positive existential claims are analytic and all negative ones self-contradictory. This is so because the ascription of a predicate to a thing implies the existence of the thing. The ascription of the predicate existence would, thus, already imply the existence of the thing asserted to exist, so making the ascription analytic.1 And mutatis mutandis, one may show the selfcontradictory character of negative existential claims. (2) It is, however, false that all existential claims are either analytic or self-contradictory. So, (3) it is false that existence is a predicate. This efficient little argument can perhaps be traced back to Kant. Its recent versions, at any rate, have been sharply 1 Ayer, Wisdom and Broad, use respectively the three words " tautologous " " silly " and " platitudinous " to describe positive existential claims in their statements of this argument. See " 'Exists' as a Predicate " by George Nakhnikian and Wesley C. Salmon, PhilosophicalReview, lxvi (1957), 535, where their arguments are cited.
This content downloaded from 193.63.81.241 on Fri, 12 Dec 2014 13:50:19 PM All use subject to JSTOR Terms and Conditions
366
M. KITELEY:
criticized by Nakhnikian and Salmon; I shall discuss their criticism shortly. In the justification of the major premise (1) of this argument it is said that predicate-ascription implies existence. " For," says Ayer, " when we ascribe an attribute to a thing, we covertly assert that it exists (Language, Truth, and Logic, p. 43)." It is, no doubt, hard to see how one could ascribe anything to what is non-existent, how one could ever successfully apply a predicate to nothing at all. Even more generally, it is hard to see how one could, not only predicate of or ascribe to, but make mention of, talk about, refer to, speak of, or make statements about that which fails to exist. Yet that we do such unlikely things seems manifest. Examples are easy to find: " He spoke at length of the advanced characteristics of the building which Le Corbusier designed for the . . . competition." " The only right thing he said about Coriolanuswas that he was a Roman general." All the familiar examples from fiction, legend and myth come to mind. I should like to say just two things about the doctrine that predication entails existence. First, it must be restricted to successful, true or correct predications. A predication might fail just because of subject failure. To a non-Canadian who thought the city of Saskatchewan to be the home of the University of Saskatchewan, the helpful correction would point out the nonexistence of such a city not the actual location of the University -his mistake was one of misnaming rather than mislocating. Second, the doctrine, even thus restricted, is meretricious. Its false appeal comes from the inability to distinguish the truistic observation " If you are going to talk [predicate, refer, state], you have got to talk about something " from the quite erroneous statement " Whatever you talk about [predicate of, refer to, make statements about] must exist ". The first of these statements, when a statement, is genuinely truistic. It comes to little more than a gratuitous observation on the fundamentally subjectpredicate character of our talk. When not a statement but, e.g. a piece of advice to a pronoun-ridden conversationalist, it is a sound warning against dangling reference and topic-neutral chatter. The second of these statements is a false observation on what we can talk about, an observation which cannot, I should think, be corrected by the expedient of giving everything honorary existence, in intellectu or elsewhere. We sometimes gentlv reprove a child's fancy with something like, " That is just a makebelieve lion that chewed up your shirt, isn't it Brian? " But such gentleness is a mask for the hard truth: the child is a victim of fancy, his shirt is not a victim of a fanciful lion.
This content downloaded from 193.63.81.241 on Fri, 12 Dec 2014 13:50:19 PM All use subject to JSTOR Terms and Conditions
? IS EXISTENCEA PREDICATE
367
There may yet be one false charm of this doctrine unexposed. Ryle, while arguing the necessarily general character of statements about the future, says that particular, future episodes cannot be got to make statements about (Dilemmas, p. 27). I am not at all sure that Ryle is right, but there is this point I must concede: if what you are talking about is non-existent, it must at least be a conversation piece in some body of legend, fable or fiction, or some passe scientific hypothesis. You must be able to make, if challenged, identifying references-" you know, Bellerephon'shorse "-, and you must be able to say if challenged, why just those predicates apply-" see Bullfinch 7-, neither of which you could do without at least the existence of the Pegasus legend, if not the legendary Pegasus. The major premise of this argument is, then, faulty. There is no reason to believe that even were existence a predicate, all affirmative, existential assertions would have to be tautologous. It does not follow, of course, that existence is a predicate. Nakhnikian and Salmon reach the same conclusion by a different, and I think wrong, line of argument (pp. 536 ff). They argue that " Horses exist ", when elliptical for " Some horses exist ", can be logically transcribed with " E " the existential predicate as " (3x) (Hx . Ex) " which is not a tautologous form. They further argue that even when the predicate " E " is, as they claim, plausibly defined by the trivial property of selfidentity, the foregoing statement is not a tautology. The trivial property of existence just drops out. This is true enough for general existence claims, but what about singular ones ? " Pegasus exists ", by their definition of " E ", becomes " Pegasus is self-identical ", which is patently trivial. They have, thus, only undercut half of what the major premise claims. The rest stands; singular existential statements remain trivial, and, we would have to suppose, singular existence is not a predicate. 3. The Third Argument The third argument says that if existence is a predicate, then you should be able to affirm it universally and deny it particularly. You can, however, do neither of these. It is equally nonsensical to say either " All tame tigers exist " or " Some tame tigers do not exist". The square of opposition for existencestatements is fearfully truncated, indeed to the point of losing a dimension. Thus, existence cannot be a predicate.
This content downloaded from 193.63.81.241 on Fri, 12 Dec 2014 13:50:19 PM All use subject to JSTOR Terms and Conditions
368
M. KITELEY:
The " tame tigers " example is taken from G. E. Moore's discussion of this argument in his contribution to the Aristotelian Society symposium on whether existence is a predicate (1936). Moore compares " All tame tigers exist " with " All tame tigers growl"; and he also compares " Some tame tigers don't exist" with " Some tame tigers don't growl ". The twofold comparison shows, he thinks, an important difference between the usage of exists " in the one case, and the usage of " growls " in the other. "All tame tigers exist" and " Some tame tigers don't exist" are, says Moore, " queer and puzzling expressions ". He does not outright say that they are nonsensical, but they do not, he says, carry their meaning, if they have any, on the face of them. This argument turns on the queerness or nonsensicality of statements of the form " All ----s exist " and " Some ----s do not The queerness comes out well enough with " tame exist". tigers " in the blanks. But other fillings seem to give quite natural expression, e.g. " All the stamps in this issue still exist, but some in this one do not ". And even " All tame tigers exist " can be given a setting that makes it come to life. If a zoological survey team were compiling a directory of all tame tigers, they might first assign number-designationsto all those known to have been brought into the country after such and such a date. Say that they have fifty number-designations. They check these off to see which, if any, have survived. Then, at the end of their tally they say " Extraordinary, they all exist ". What do these examples show ? They seem to show that the verb " exists " does have uses, perhaps predicative uses, that go easily and naturally through all the quantifier changes from none to all in the schedule of generality. Moore was not unaware of this. He found a use of "not exists ", viz. being imaginary, that went through the schedule. The examples do not, however, show the argument to be faulty. They only show that there are some uses of " exists" which, since they do not make nonsense out of statements of the form " All ---s exist ", might be predicative; they do not rule out the possibility of there being other uses of this verb which, by making nonsense out of such statements, would be non-predicative. If then, one use of " exists " can be found which does make nonsense out of universal affirmative statements in which it appears, then the concept of existence associated with this use of the verb would not be a predicate. There is, I think, such a use of " exists ". It is the most exiguous use of it, the use which most closely corresponds to the non-locative use of the " there is " idiom. The oddness of " All
This content downloaded from 193.63.81.241 on Fri, 12 Dec 2014 13:50:19 PM All use subject to JSTOR Terms and Conditions
IS EXISTENCE
A PREDICATE
2
369
tame tigers exist" is the same, thus, as the oddness of " There are all tame tigers ".1 The oddness is somewhere in the same family with " The warmth of the temperature was . . . " and, more closely, " All the cars on the freeway were numerous ". " Exists," when employed exiguously, tells you something about tame tigers but nothing about each and every tame tiger; it tells you something about the membership, but nothing about the members. Existence, here, is something like full strength of a regiment: the regiment can be at full strength, but none of the members can be. The non-exiguous uses of " exists " (where one can say, e.g. that all tame tigers exist) might be called, following Hall, excluder uses. Excluders, he says, " serve to rule out something without adding anything, and ambiguously rule out different things according to the context." 2 Thus, when the tigercanvassers exclaimed " Extraordinary, they all exist ! " they did so because they had ruled out death in captivity, escape and shipment back to India. The force of their exclamation is not the absurd " None of them do not exist ", but rather the intelligible " None of them have died, nor escaped, nor been shipped back to India ". You can, then, say that all A's exist when by so saying you are denying for all the A's that thereare that they are, e.g. extinct, out of production, destroyed, hallucinatory, mythical, fabulous, or fictional. Each item in this list is an attribute and as an attribute can be affirmed universally or denied particularly. I am tempted to think that if you state existence using the formula " There are ---s " rather than " ---s exist ", the excluder business would never crop up. I cannot imagine, and here is the source of my temptation, " There are all ---s " ever making sense. I have italicized above, however, a form of this idiom which might be thought to be universal, viz. " all the A's that there are ". I am uncertain about it, but I suspect that it would demand the same kind of excluder-analysis as " exists ". Before drawing the moral from all this, there is an old puzzle which I should like to exorcise. If, so the puzzle goes, " Horses exist" and " There are horses" come to the same thing, then you would expect the statement " There are horses which exist " to be redundant, and the statement " There are non-existent horses" to be self-contradictory. The logical transcription of 1 Jesperson calls this use of " there " existential, as contrasted with its use as a " local adverb ", e.g. " There are all the tame tigers ", Modern Englishi Grammar, vii (Copenhagen, 1949), 107. 2 Roland Hall, " Excluders ", Analysis, xx (1959), 1.
This content downloaded from 193.63.81.241 on Fri, 12 Dec 2014 13:50:19 PM All use subject to JSTOR Terms and Conditions
370
M. KITELEY:
these two statements shows this clearly enough. The first is written " (3a) [Hxx. (3y) (y = x)] "; the second is written "(3x)[Hx. - -(3y) (y = x)] ", where " x exists " is defined by x) ". Since " (3y) (y = x) " is true of all x's, the "(3y) (y second conjunct of the first statement adds nothing to the conjunction, and the denial of the second conjunct of the second statement, being contradictory, makes the conjunction selfcontradictory. But so long as " exists " is used in a non-exiguous way neither redundancy nor self-contradiction will be incurred: " There are non-existent horses, e.g. Pegasus " is quite consistent because ccnon-existent " is used in the way of an excluder. The argument under consideration, then, can only be used to show that the verb to exist will fail, when used exiguously, to run through the complete schedule of generality. Only this use of the verb, then, is shown to be non-predicative. Other uses might well be predicative. 4. The Fourth Argument This argument is similar to the last one. It is as follows. If existence is a predicate, then there are certain kinds of inferences that should be valid. For example, the inference from " Donkeys exist and Eeyore is a donkey " to " Eeyore exists " should be valid. It is clearly not valid, so existence cannot be a predicate. Russell spoke of such inferences as " pseudo-syllogisms " and added that their fallaciousness was parallel to that of the argument " Men are numerous, Socrates is a man, therefore Socrates is numerous ". Russell did not conclude that the fallaciousness of these arguments rested on the predicative use of " exists " and " numerous ", but rather on the predicative misuse of them. They are predicates, but not predicates of particular things. Thus, to say that Socrates exists is to say something which is " a mere noise or shape, devoid of significance ". So to say is to misapply a predicate which only rightly applies to propositional functions. It is the conclusions of these pseudo-syllogisms which Russell marks as nonsensical. But if the conclusion of an inference is nonsense, one of the premises must be also, and the one which is, quite clearly, is the major. That premise says, in my example, that donkeys exist. For the inference to be valid it must be read " All donkeys exist ". But " All donkeys exist" may or may not be nonsensical depending on whether " exists" is used exiguously or as an excluder. If it is used as an excluder, then the statement might just be the false claim that no donkeys
This content downloaded from 193.63.81.241 on Fri, 12 Dec 2014 13:50:19 PM All use subject to JSTOR Terms and Conditions
IS EXISTENCE
A PREDICATE
?
371
occur in fiction, a statement from which you could infer, Eeyore being a donkey, that Eeyore does not occur in fiction, i.e. he exists. This would, then, be a perfectly good argument with one false premise. But for the exiguous use of " exists " this argument, like the last one, seems to be sound. Subsumption inferences, with " exists " in the place of the predicate, are invalid. Surely they would not be invalid unless it were improper for this verb to occupy the predicate position; and this impropriety may be what is heralded by the maxim that existence is not a real predicate. Only two of the four arguments have survived, and they not completely intact. What exactly do they show ? All four of the arguments advance in the same formation. They start with a claim about the way bona fide predicates behave; they continue with the observation that the putative predicate, existence, does not so behave; and they end with the conclusion that existence is not a real predicate. All the arguments claim to know what a real predicate is or does. Each of the arguments, then, states certain requisites of all predicate behaviour. Kant and Malcolmboth seem to think that you can only be using an expression in the predicative way if, by this use, you can add something to the subject, something new. Their slogan might have been: existence is not news. The other arguments have variously insisted on, as the tests of predicate behaviour: occasional occurrencein contingent assertions, appearance in predicate position in universal affirmative statements, and appearancein predicate position of the major premises of valid subsumption arguments. " Exists ", however, sometimes, meets all these tests. What, then, are we entitled to say ? At best we are entitled to say that some of the ways we use this verb are not predicative. And that modest entitlements rests solely on the assumption that these tests or requisites accurately bound the limits of predicate behaviour. But do they ? What are the marks of a genuine predicate ? Webster says that a logical predicate is that which can be affirmed or denied of a subject. But what kind of subject ? If the subject is a pseudo-subject, then the predicate affirmed or denied of it would be a pseudo-predicate. What, then, are the marks of a real or logical subject ? Webster says two things. He says, first, that the grammatical subject of a sentence need not be the same as the logical subject, which is the real subject of predication. For example, in " It is hard to do right ", " it)"
This content downloaded from 193.63.81.241 on Fri, 12 Dec 2014 13:50:19 PM All use subject to JSTOR Terms and Conditions
372
M. KITELEY:
is the grammatical while to " do right" is the logical subject. Then the logical or real subject, the entry goes on to say, " is the term a proposition is about; also, what such a term denotes: the topic of an affirmation or denial ". This is not too helpful. What seems to be given as the mark of a term which serves as the grammatical predicate is simply appearance in the predicate position. Thus, in the example above, " hard to do right " serves as the grammatical predicate. But the mark of a logical predicate, the entry seems to imply, is appearance in the predicate position of a standardsentence,i.e. a sentence in normal word order. " It is hard to do right " is not in normal word order; " To do right is hard " is. Thus the logical predicate in both these sentences is " hard ", even though it only occurs in the predicate position in one of them. Before applying this test to our verb, I should at least acknowledge the difficulties. What is the predicate position ? How do you know when a sentence is in normal word order ? The first question can be answered if you allow the adequacy of a grammar like Chomsky's. We need nothing more than his schematised set of parsing rules, the first of which, " Sentence NP + VP ", indicates the predicate position as that occupied by the verb phrase (" VP "), or some part of it (Syntactic Structures, pp. 26 ff.). The second question is less easy. For example, are active constructions in normal word order while passive are not ? We surely would not want to say this, implying as it does that "loves wisdom " is the logical predicate of " Socrates loves wisdom" in contrast to the merely grammatical predicate " by Socrates" in " Wisdom is loved by Socrates ". We want, that is, to make room for double-subject or relational constructions, so that the second, logical subject might be either the grammatical, direct object or the object of a prepositional phrase. The only sentences, then, with mixed-up word order, are those, like the example " It is hard to do right ", in which a pronoun serves as a dummy subject, what Jesperson calls the " preparatory it ". We must say, then, that the sentence " Lions exist " is in normal word order. And, since the verb " exists " occupies the predicate position in this standard sentence, we must conclude that this verb behaves predicatively, that it is the logical, not just the grammatical, predicate of the sentence. Note also that "lions " would be the logical as well as the grammatical subject. I argued above, however, for an assimilation of the exiguous pse of to exist with the use of the " there is " idiom, i.e. where
This content downloaded from 193.63.81.241 on Fri, 12 Dec 2014 13:50:19 PM All use subject to JSTOR Terms and Conditions
IS EXISTENCE
A PREDICATE
?
373
"there " is not a local adverb. For example, (1) Tame tigers exist, and (2) There are tame tigers come to the same thing. This assimilation must, of course, be sharply restricted; it is wrong for any of the excluder uses of " exists ". This wrongness comes out when it is recognized that both " There are non-existent tigers " and " There are tigers which exists " are neither contradictory nor redundant, as they would have to be if the assimilation were bonafide,when " exists" is used as an excluder. But, for the exiguous use of " exists " the assimilation is alright. And if the assimilation is alright, then clearly " tame tigers " cannot be the logical subject of either (1) or (2) above, and " exists " cannot be the logical predicate of (1). This, it seems to me, is the substance of the slogan that existence is not a real predicate. San Jose State College
25
This content downloaded from 193.63.81.241 on Fri, 12 Dec 2014 13:50:19 PM All use subject to JSTOR Terms and Conditions