Maverick Philosopher

History of Philosophy Quarterly’s Referee’s Report

“Brentano on Existence” (#278) is a superb paper. It definitely merits publication in HPQ. The author distinguishes theories of existence as eliminativist versus identitarian, and then presents and critically assesses Brentano’s eliminativist approach to existence. The result is an incisive challenge to eliminativist approaches to existence, Brentano’s included. Since Brentano anticipates the eliminativist views of Frege and Russell on existence, as the author notes, the paper’s result bears on a significant tradition in twentieth-century philosophy. The paper will be of interest not only to Brentano scholars but also to all philosophers interested in existence, particularly whether existence belongs to individuals. The paper is lucid, cogent, and original. I highly recommend it.

Author’s Abstract

This article begins by distinguishing eliminativist from identitarian theories of singular existence. It is then argued that Brentano’s theory is most charitably construed as eliminativist, as denying that existence is attributable to individuals. Next, Brentano’s central argument is evaluated and rejected according to which (i) the copulative ‘is’ is syncategorematic; (ii) ‘exist(s)’ can be eliminated in favor of the copulative ‘is’; hence, the former is as syncategorematic as the latter. A positive argument against eliminativism is then presented. The penultimate section extends the critique to the eliminativism of Frege and Russell. The conclusion suggests that theories like those of Brentano, Frege, and Russell derive their plausibility from a waffling between identitarian and eliminativist approaches.

[HPQ 311]

BRENTANO ON EXISTENCE

Franz Brentano is an important transitional figure in the history of philosophy. Although he was steeped in Aristotle and the scholastics, his deflationary linguistic approach to metaphysical questions anticipates 20th century analytic treatments. Indeed, Gustav Bergmann calls him “the first linguistic philosopher.”[1] A good example of Brentano’s deflationism is his theory of existence, which in some ways anticipates the influential theories of Frege and Russell. My aim here is to present and evaluate Brentano’s theory of existence. Although I will be arguing that it is fatally flawed, there is much to learn from it.

1. Situating Brentano’s Theory: Eliminativism versus Identitarianism

The logically first question to ask about existence, and the question I will attempt to answer here, is whether or not there is any sense in which existence belongs to individuals.[2] In order to orient ourselves with respect to this question, it is useful to divide theories of existence into two main groups, eliminativist theories and identitarian theories. Since this is a distinction not confined to the theory of existence, I will begin by characterizing it in general terms.

An eliminativist about X, motivated by the puzzles to which X gives rise, denies that ‘there is any such thing’ as X. An identitarian about X, motivated by the same puzzles, aims to identify X with something theoretically more tractable, and less likely to engender perplexity. Clearly, to identify X with Y is to presuppose that X is there to be identified with Y. The distinction, then, is roughly this. Faced with a recalcitrant datum, the eliminativist, radical that she is, has no scruples about simply denying the datum and replacing it with something more congenial. The eliminativist is thus a replacement theorist. The identitarian, however, has a conservative nature:

[HPQ 312]

he aims to analyze or explicate the datum in question without denying it, distorting it, or changing the subject.

Hume’s regularity theory of causation is an example of an eliminativist theory.[3] On our ordinary conception of causation, causes produce or bring into existence their effects. But Hume was famously unable to find any production in the causal sequences he examined. So he threw out causation-as-production and replaced it with something that comported better with his empiricist strictures. He thus proposed a theory in terms of spatiotemporal contiguity, temporal precedence and constant conjunction. Taken as an analysis of our ordinary concept of cause, as an unpacking of what we ordinarily mean when we engage in causal talk, Hume’s theory is hopeless. When I say, with the vulgar, that an F-event caused a G-event, I mean that the former produced, and in producing necessitated, the G-event. I do not mean that it just happens to be the case that, hitherto, every F-event has been contiguously followed by a G-event. Thus Hume’s theory is more charitably interpreted as a replacement of our ordinary concept.

The philosophy of mind provides a second example of the eliminativist/identitarian contrast. An eliminativist about mental states simply denies their existence. He doesn’t identify pain, to coin an example, with delta A fiber stimulation (or whatever) as does the identitarian; he holds pain to be a bogus category of ‘folk psychology’ that in the fullness of time we will learn to do without. (I hope the manifest absurdity of eliminativism in the philosophy of mind does not prejudice the reader against the eliminativism/identitarian distinction as such.)

The contrast surfaces again, more plausibly this time, in the theory of existence, where the X in question is the putative existence of individuals. Observing a tree, I may say, ‘This tree exists, and it exists whether or not I am perceiving it.’ This is merely a claim, of course, and no such perceptual claim is self-validating. It may be that the object of my perception does not exist. But if the claim is to make sense, which it apparently does, then it must make sense to say of an individual, e.g., a tree, that it exists, where ‘exists’ means exists whether or not I or anyone am perceiving it. Now the eliminativist denies this: he denies that it makes sense to attribute existence to individuals themselves. He denies that ‘exists’ and cognates can ever be meaningfully employed as first-level predicates.

Bertrand Russell took an eliminativist line when he compared existence with being numerous. From ‘Human beings are numerous’

[HPQ 313]

and ‘Socrates is a human being,’ one cannot infer ‘Socrates is numerous.’[4] To think otherwise is to commit the Fallacy of Division: being numerous is attributable to classes and such-like, but obviously not to individuals. Russell thought that existence was in the same logical boat: existence is attributable to classes and propositional functions but not to individuals. Russell here followed in the footsteps of Gottlob Frege who held that existence is a property of those functions he calls concepts, but never a property of objects.[5] Meaningful attributions of existence are claims to the effect that a concept is instantiated, i.e., has something falling under it; since no individual can be instantiated, it follows straightaway that existence cannot be meaningfully attributed to individuals.

It is important to see that neither Frege nor Russell can be charitably interpreted as an identitarian, as saying something like, ‘For an individual to exist is for some concept to be instantiated,’ or ‘For an individual to exist is for some propositional function to be sometimes true.’[6] One reason for this is their plain denial that ‘exists’ is an admissible first-level predicate. If so, no schema of the form ‘x exists if an only if ___’ is admissible. A second reason is that these proposals are grossly circular. If a concept cannot be instantiated unless it is instantiated by something that exists, then it is hopelessly circular to identify the existence of the thing in question with its instantiation of any concept. Similarly, if a propositional function ‘Fx’ is true only if there is a substituend ‘a’ for ‘x’ which on substitution yields the true sentence ‘a is F,’ and if this sentence can be true only if ‘a’ denotes an existing individual, then it is shamelessly circular to identify the existence of an individual with a propositional function’s being sometimes true. Frege and Russell, then, are most charitably viewed as eliminativists.

For the eliminativist, then, there is no such thing as the existence of individuals. This is not to deny individuals; it is to deny the existence of individuals. So in reality there are not two items, Quine and Quine’s existence, and consequently no ‘problem’ of how Quine is related to his existence. Beyond mind and language, there is just Quine, who is neither distinct from his existence, nor identical to his existence. Compare: there is no question of how Quine is related to his numerousness; he is neither identical to it nor distinct from it since he is just not the sort of item that can be meaningfully said to be either numerous or not numerous.

Thus the eliminativist. The identitarian, on the other hand, claims that in reality (i.e., apart from us and our linguistic and conceptual operations) there is Quine, and Quine’s existence, and thus meaningful questions about how Quine is related to his existence, and

[HPQ 314]

what his existence consists in. One theory (theoretically possible even if metaphysically impossible) is that Quine is one and the same with his existence. Other theories will try to show that Quine is really distinct from his existence. For the identitarian then, individuals themselves exist. Existence in some sense belongs to them intrinsically, is attributable to them, even if it is not a property in any ordinary sense. (Note that an identitarian is not someone who identifies the existence of an individual with that individual, but someone who, presupposing that existence belongs to individuals themselves, attempts to analyze, explicate, or identify this existence with something, as one would do if one said that ‘x exists if and only if x occupies a spatiotemporal position’ or ‘x exists if and only if x is causally active/passive,’ etc.[7])

2. Brentano as an Eliminativist

Brentano, as I read him, is an eliminativist about the existence of individuals and to this extent anticipates the views of Frege and Russell about what existence is not. (Of course, Frege and Russell diverge from Brentano in their views about what existence is.) One of Brentano’s motives for being an eliminativist is precisely to avoid the ancient scholastic dispute over whether an individual and its existence are the same or different. If Brentano is right about existence, then “The fierce disputes in which the Medieval schools engaged concerning essentia and esse...”[8] were based on a false assumption, namely, that existence (esse) belongs to things themselves apart from minds and that therefore the existence of a thing is either identical with it or different from it. Brentano:

The question [concerning essentia and esse] always turns on whether the existence of a being is the same or a different reality than the being itself. Scotus, Ockham, Suarez rightly deny that it is a different reality. . . . But as a consequence they fall into the error of thinking that the existence of a thing belongs to the essence of the thing itself, and they regard it as the thing’s most general concept. Here the Thomists’ opposition was correct...How, they cried, could the existence of a thing be its most general concept? — This is impossible! — Then its existence would follow from its definition, and consequently the existence of a creature would be just as self-evident and antecedently necessary as the existence of the Creator Himself.[9]

For Brentano, both the Thomists and their opponents were wrong. It is neither the case that the essence and existence of a are identical, nor that they are different. For there is no such thing as the existence of a to be either identified with a or differentiated from it. Existence does not pertain to things themselves.

[HPQ 315]

But of course Brentano doesn’t merely deny that existence belongs to individuals themselves, he goes on to replace the existence of a tree, for example, with a judger’s acceptance (Anerkennung) of a tree, or else a judger’s acceptance of the presentation of a tree, and the nonexistence of a unicorn with a judger’s rejection (Verwerfung) of a unicorn or else a judger’s rejection of the presentation of a unicorn.[10] It is important to interpret this as a replacement, and not an identification or analysis or explication. If Brentano had said that the existence of a tree is (identically) some judger’s acceptance of a tree, then he would have ended up in a lunatic form of idealism according to which trees cannot exist unless human beings exist. Thus charity forbids us from interpreting him in this way. We must read him as an eliminativist, as one who denies that the existence/nonexistence contrast is applicable to individuals themselves. He is not saying that there is the existence of a, and that the existence of a is to be identified with a judger’s acceptance of a. He is denying that there is the existence of a, and replacing the existence of a with a judger’s acceptance of a. Compare Frege: he cannot be reasonably interpreted as identifying the existence of an individual with the instantiation of some concept; he must be interpreted as replacing the existence of individuals with the instantiation of concepts.

Thus charity demands that we give Brentano an eliminativist reading. But there is a wrinkle here that needs to be explored. It is surely absurd to say that the existence of a tree is identically some judger’s acceptance (affirmation, recognition) of a tree. It is part of the very sense of ‘exists’ (assuming that it is a first-level predicate) that if an individual exists, it exists whether or not any judger affirms it, indeed whether or not any judger exists. But what of the view that the existence of a tree is its correct acceptability by someone? Passages in Brentano and his students strongly suggest this view.[11] For example, Anton Marty holds that the existent is that which can be rightly recognized or affirmed.[12] If the correct acceptability line is free of idealistic taint, then perhaps we will not be forced to interpret Brentano as an eliminativist.

But the view that the existence of x is identically its correct acceptability by someone also implies an untenable idealism. For although what is acceptable need not be actually accepted, the acceptable (in any robust sense) requires the existence of beings capable of accepting. Surely the acceptability or affirmability of a tree is not an intrinsic, but a relational, property of it: a property the tree’s possession of which implies a relation to a judger, a being capable of accepting and rejecting. Hence the ability to be accepted

[HPQ 316]

is not an intrinsic capacity of the tree, but resides instead in the ability of a judger to accept. Thus to say that the existence of a tree consists in its correct acceptability is to imply that beings are on hand capable of accepting and rejecting. If so, the idealism remains: trees and the like could not exist except in worlds in which judgers exist. To avoid the idealism, one would have to say that an individual’s correct acceptability, and thus its existence, requires only that it be possible that there be a being capable of accepting it. But this is surely far too anemic. The existence of a tree cannot be identified with the mere possibility that there be a being capable of accepting it. For in possible worlds without trees, there is surely the possibility that there be beings capable of accepting trees. If the existence of a tree were identified with the mere possibility that there be a being capable of accepting it, then trees would exist in every possible world in which this possibility exists. But this possibility exists in every world – by the characteristic S5 axiom of modal propositional logic, to wit, ‘Possibly p –> Necessarily possibly p’ – so the proposal under consideration implies that trees are necessary beings! This avoids idealism all right, but only at the expense of something just as bad if not worse.

Since the identitarian reading issues in an unacceptable idealism, whether existence is acceptance or correct acceptability, I conclude that the eliminativist reading of Brentano’s doctrine is the only one consistent with charity. Thus when Brentano says, as in effect he does, that x exists if and only if whoever accepts or affirms x judges correctly, and x does not exist if and only if whoever rejects or denies x judges correctly,[13] we must not interpret this in an identitarian spirit. Note first that the truth of these biconditionals does not compel us to reach for an identitarian reading; we are not forced to read them as saying that (i) there is the intrinsic existence/nonexistence of x and that (ii) it consists in (i.e., is identifiable with) x’s correct acceptability/rejectability by someone. Generally speaking, the truth, even the necessary truth, of ‘Fx if and only if Gx’ does not logically require the identification of F-ness with G-ness. (To see this, substitute ‘triangular’ for ‘F’ and ‘trilateral’ for ‘G.’)

Note also that if the Brentano biconditionals are taken as requiring the identification of existence with correct acceptability, then they would be manifestly circular. For if I affirm x correctly, and individuals exist intrinsically, then the correctness of the affirmation would be grounded in the actual existence of the individual affirmed; what would make my judgment correct would be the actual existence of the individual affirmed. The judgment would be

[HPQ 317]

correct because x exists; it would not be the case that x exists because my judgment is correct. But if the judgment is correct because x exists, then, on pain of vicious circularity, the existence of x cannot consist in its correct acceptability. Thus the Brentano biconditionals do not support the identification of existence with correct acceptability.

The Brentano biconditionals must therefore be taken in an eliminativist spirit.[14] So taken, they have to be taken in the way we take ‘God is the ultimate anthropomorphic projection’ (Feuerbach). This seems to invite an identitarian reading on which (i) there is an entity named ‘God’ and (ii) this entity is identically the ultimate anthropomorphic projection. But of course, this is an impossible reading because self-contradictory: if God exists, he cannot be a projection, and if God is a projection, then he doesn’t exist. We must take the Feuerbachian dictum in an eliminativist spirit as implying that there is no such thing as God. Curiously, the Feuerbachian denies the existence of God using a sentence in which neither ‘exist’ nor ‘not’ occurs!

Similarly, we must take ‘X exists if and only if whoever accepts or affirms x judges correctly’ as implying that there is no such thing as the existence of x, as this is normally understood, namely, as the existence intrinsically had by individuals themselves independently of judgers. What the biconditional does, then, is to re-define the phrase ‘existence of x’ so as to mean the same as ‘correct acceptability of x,’ in the same way ‘God is the ultimate anthropomorphic projection’ re-defines the meaning of ‘God.’ The Brentano conditional amounts to a linguistic proposal that henceforth ‘existence of x’ shall mean just what ‘correct acceptability of x’ means, that the latter shall replace the former.

Brentano, then, is most charitably interpreted as an eliminativist, a replacement theorist. He replaces the existence/nonexistence of individuals with the judgmental acceptance/rejection of individuals, just as Hume (on one plausible interpretation, anyway) replaces causation-as-production with contiguous regular event succession. We should thus expect Brentano to reject as bogus the concept of existence. For if there is ‘no such thing’ as the existence of individuals apart from judgers, then there is ‘no such thing’ as existence apart from judgers, since existence cannot occur except as the existence of existents.[15] By denying that there are existing individuals apart from judgers, Brentano is committed to denying that there is existence apart from judgers. One should therefore be puzzled by Brentano’s assertion that the concept of existence is a

[HPQ 318]

concept we arrive at by reflection on the nature of judgment.[16] This is puzzling due to its identitarian flavor. Brentano seems to be implying that the concept of existence is legitimate after all, but just needs unpacking in terms of the concept of judgmental acceptance, or else in terms of the concept of correct acceptability. But if existence/nonexistence is explicated in terms of a judger’s acceptance/rejection of an individual, then the existence of this mountain before me is (identically) my (or someone’s) acceptance of it. This however returns us to an obnoxious form of idealism. A serious eliminativist who understands what he is about cannot hang onto the concept of existence. For again, existence is the existence of individuals (subjective genitive), the existence belonging to them intrinsically, the existence that determines them as existent. So if there is ‘no such thing’ as the existence of individuals, then there is no such thing as existence. Hence one cannot say that existence is a concept that arises from reflection on the nature of judgment. What arises from reflection on the nature of judgment is not existence, but at most a replacement for, or a conceptual successor of, existence, namely, the concept of judgmental acceptance. A serious eliminativist about existence cannot, without obfuscation, use ‘existence’ to refer to judgmental acceptance, just as a serious eliminativist about causation-as-production cannot, without obfuscation, use ‘causation-as-production’ to refer to contiguous regular event-succession. Intellectual hygiene forbids it.

A grammatical way to see my point is by comparing the phrases ‘existence of a’ and ‘acceptance of a.’ The first is a subjective genitive construction, a possessive: the existence of a is a’s existence, the existence that belongs to a. The second is an objective genitive: a is the object of acceptance (by a judger), not its subject.[17] This grammatical point suffices to show that the sentence ‘The existence of a = the acceptance (by a judger) of a’ either implies an unacceptable idealism, or involves an obfuscatory re-interpretation of ‘the existence of a’ as an objective genitive. But then we are no longer talking about plain old existence which is obviously enough the existence of individuals (subjective genitive).

What this suggests is that Brentano was not clear about the full implications of his view, and waffles between eliminativism and a form of identitarianism in which the existence of an individual is identified with a judger’s acceptance of it. Each, taken straight, is absurd (as I shall further support below); but by confusing them, the hybrid position seemed to Brentano to be plausible and indeed correct. Or at least that is my exegetical hypothesis. But before

[HPQ 319]

going further, we need to examine Brentano’s main argument for his eliminativism.

3. Brentano’s Argument

Having sketched Brentano’s position, we must now try to understand how he arrived at it.

The first step is to consider his argument for the negative claim that there is no such thing as the existence of an individual. This negative claim is equivalent to the thesis that ‘exists’ is a syncategorematic or synsemantic or co-meaning (mitbedeutend) expression, i.e., one that has no independent meaning, and that therefore neither denotes an individual nor expresses a property. A syncategorematic expression is one to which nothing in reality corresponds. Brentano gives ‘of’ and ‘but’ as examples of syncategorematic expressions,[18] to which one could add ‘or,’ ‘not,’ and numerous others. They cannot stand alone, semantically speaking. They are clearly not names, and it seems farfetched to think that they express properties. ‘Socrates’ and ‘wise,’ on the other hand, are examples of categorematic (autocategorematic) expressions. Everyone will agree that ‘Socrates’ denotes Socrates, whether directly or via the name’s sense. It is less clear that ‘wise’ has its own unique denotatum (the property of being wise), but it is clear that it has a non-logical content that bars it from being considered syncategorematic or synsemantic. Whatever the ultimate merit of the categorematic/syncategorematic distinction, it is a presupposition of Brentano’s argument, and one that I will not be questioning.[19]

Brentano’s deflationary argument is that (i) the copulative ‘is’ is synsemantic; (ii) ‘exists’ can be eliminated in favor of the copulative ‘is’; therefore (iii) the former is as synsemantic as the latter. Given that the copulative ‘is’ is synsemantic, i.e., does not denote anything or express a property, the same must also be true of ‘exists,’ in which case there is nothing in things themselves that is their existence. Existence is thus in no sense a property, feature, attribute, determination of individuals. Existence in no sense belongs to individuals, not even in the limit sense of being identical to them, of dividing into them without remainder. (If you say that, for any x, the existence of x = x, then you presuppose that existence belongs to individuals.) If so, there can be no legitimate problem of explicating what it is for an individual to exist, or of explaining how an individual or its essence is ‘related’ to its existence. A vast amount of Medieval controversy will have turned out to have rested on a mistake. Consign it then to the flames!

[HPQ 320]

Brentano’s argument for his negative thesis may be rendered as follows:

a. The copula is a syncategorematic or synsemantic expression.

b. Existential sentences and copulative (predicative) sentences are intertranslatable.

Therefore

c. ‘Exists’ is a synsemantic expression.

In support of the second premise, Brentano offers the following schedule of intertranslation:

Some man is sick.

A sick man exists.

There is a sick man.


No stone is living.

A living stone does not exist.

There is no living stone.


All men are mortal.

An immortal man does not exist.

There is no immortal man.



Some man is not learned.

An unlearned man exists.

There is an unlearned man.


Brentano concludes from this that “...the ‘is’ or ‘is not’ of the existential proposition is merely equivalent to the copula, so they are not predicates and have no meaning at all in and of themselves.”[20]

It seems clear that this analysis can be readily extended to singular sentences. Thus ‘Socrates is sick’ is equivalent to ‘Sick Socrates exists,’ and ‘Brentano was never a Catholic priest’ is equivalent to ‘The Catholic priest Brentano never existed.’ In any case, the logical equivalence of predications and existentials is not a thesis I will be questioning.

What I will be questioning is what Brentano infers from this equivalence. What are we to make of his deliciously seductive argument for the synsemanticity of ‘exist(s) and cognates? I submit that it trades on a confusion of the pure (existentially neutral) copula with the ‘is’ as it ordinarily functions in such sentences as ‘The Charles River is polluted.’ For premise (a) to be true, the copulae in question must be pure. But for premise (b) to be true, the copulae in question must be ‘existentially loaded.’ Thus the argument succumbs to a fallacy of equivocation.

To put it another way, Brentano faces a dilemma. Copulative sentences either feature a pure (existentially neutral) copula, which

[HPQ 321]

does not express existence, or they feature an existentially loaded copula, which does express existence, albeit while also discharging its copulative duties. If the former, the copulations cannot be rendered as existentials. If the latter, the copulative sentences can be rendered as existentials, but then this rendering has no tendency to show that ‘exists’ is synsemantic. For copulative sentences featuring existentially loaded copulae do not feature synsemantic copulae. A pure copula may well be synsemantic – let us grant this for the sake of the argument – but an existentially loaded copula is not.

The first question to ask is whether the ‘is’ in a sentence like ‘The Charles is polluted’ is a pure copula. Imagine the sentence asserted by someone upon emerging from an unpleasant swim in the river in question. Does the ‘is’ in this sentence express merely the connection between logical subject and logical predicate without expressing the existence of the logical subject? Or does it also express, or presuppose, the existence of the logical subject, i.e., the existence of that to which the grammatical subject of the sentence refers? Although it is obvious that the ‘is’ of existence is distinct from the ‘is’ of predication, this is consistent with one and the same concrete occurrence of ‘is’ exercising both a copulative and an existential function. The distinction between the ‘is’ of predication and the ‘is’ of existence is not like the distinction between the inclusive ‘or’ (vel) and the exclusive ‘or’ (aut). One and the same concrete occurrence of ‘or’ cannot be both inclusive and exclusive; but, I am claiming, one and the same occurrence of ‘is’ can express both existence and predication. One and the same concrete occurrence of ‘is’ cannot express both identity and predication, but it can express both existence and predication.

That our sentence, used in the circumstances we are imagining, does not feature a pure copula can be seen from the fact that it is analyzable as follows: ‘The Charles is (exists) & the Charles [is] polluted,’ where the brackets around ‘is’ signify that it is a pure copula. (With a nod toward Husserl, we might say that the pure copula ‘brackets existence.’) This analysis makes explicit the dual function of ‘is’ in our sentence. It functions both copulatively and existentially. When I say, emerging from its turbid waters, that the Charles is polluted, I am saying that it exists in a certain state, that of being polluted. I could just as well say that the Charles exists ‘pollutedly’ – turning an adjective into an adverb.[21] Emerging from its troubled waters, I am certainly not abstracting from its existence when I complain, ‘The Charles is polluted.’

[HPQ 322]

The correctness of this analysis is obvious from the proposition that an individual cannot have a property unless it exists – a proposition to which Brentano is no doubt committed given his opposition to Meinong’s doctrine of nonexistent objects according to which there are objects that actually[22] have properties despite their nonexistence. Now if the Charles cannot be polluted unless it exists, then one who understands this point and says, in normal circumstances, that it is polluted implies by his use of ‘is’ that it exists. I say ‘in normal circumstances,’ since I allow that one might for special purposes abstract from (i.e., leave out of consideration) the existence of the Charles in order to focus on its having of properties. So abstracting, one might say, ‘The Charles [is] polluted.’

My point, then, is that ‘The Charles is polluted,’ uttered in the imagined circumstances, features an existentially loaded copula. This, I take it, is simply a datum that any theory must be able to accommodate on pain of failing to satisfy elementary criteria of adequacy. Thus I don’t think I can be fairly accused of begging the question against Brentano. I am not simply assuming what he is out to deny, namely, that existence belongs to individuals; I am invoking a datum which, when properly understood, has the consequence that existence belongs to individuals.

What is more, it is precisely the fact that the ‘is’ in ‘The Charles is polluted’ has a double function, i.e., is existentially loaded, that allows the sentence to be rendered as ‘The polluted Charles exists.’ What happens in the translation from ‘The Charles is polluted’ to ‘The polluted Charles exists’ is that the copulative function of ‘is’ is expressed by the juxtaposition of ‘polluted’ and ‘Charles’ (with the adjective preceding the noun), while the existential function of ‘is’ is expressed by ‘exists.’ It is crucial to note that both sentences exhibit both functions. It is just that in the first sentence the copulative function is to the fore, while in the second, the existential function is to the fore. Since both (types of) sentences exhibit both functions, it is a mistake to think that the copulative function can be reduced to the existential, or vice versa. No reduction is possible in either direction. Brentano’s error, then, is to think that the intertranslatability or logical equivalence of predicative and existential sentences shows that the existential function can be reduced to the copulative. Intertranslatability no more shows this than it shows that the copulative function can be reduced to the existential.[23]

Indeed, Brentano’s argument is no better than the following obviously unsound argument:

[HPQ 323]

e. ‘Exists’ is an autocategorematic expression.

b. Existential and copulative sentences are intertranslatable.

Therefore

f. The copula is an autocategorematic expression.

Clearly, the truth of the premises is consistent with the falsity of the conclusion. The same holds for Brentano’s argument.

Thus Brentano appears to face a dilemma. If copulations and existentials are to be intertranslatable, the ‘is’ in a copulative sentence cannot express merely the connection between subject and predicate; it must also express, or else presuppose, the existence of the subject. Hence it cannot be a mere copula, and thus cannot be synsemantic. If, on the other hand, the ‘is’ in a copulative sentence is a mere copula, and is thus synsemantic, then copulative sentences cannot be expressed as existentials.

Brentano’s argument accordingly trades on a fatal confusion. It is plausible to maintain that the copula qua copula, or the ‘is’ in its purely copulative function, is synsemantic. But Brentano has no right to assume that the ‘is’ in such sentences as ‘Socrates is wise,’ ‘Some man is sick,’ etc. is a mere copula. The truth of the matter is that the ‘is’ in these sentences has both a copulative and an existential function in that it expresses both the connection of subject and predicate and the existence (not necessarily present existence) of the subject.

4. Brentano’s Eliminativism Eliminated

Having refuted, or at least neutralized, Brentano’s main argument for his eliminativism, I will now try to show directly that eliminativism is incoherent. At the end of section 2, I referred to a certain waffling on Brentano’s part as between eliminativism and a form of identitarianism according to which the existence of a is identified with a judger’s acceptance of a. The latter implies an unacceptable idealism, as does the more sophisticated view that existence is correct acceptability. But the former, strict eliminativism, seems not to be susceptible of coherent formulation.

What exactly is the (Brentanian) eliminativist claiming? That individuals themselves, individuals apart from us and our mental operations, neither exist nor do not exist, and that the phrase ‘the existence/nonexistence of a,’ if it is to be retained at all, is to be given the meaning ‘the correct acceptability/rejectability’ of a. Well then, supposing we acquiesce in this obfuscatory stipulation, does

[HPQ 324]

the eliminativist affirm that there are individuals that neither exist nor do not exist apart from judgers? That would be self-contradictory. For if there are (= exist) individuals apart from minds, then it cannot be the case that they neither exist nor do not exist. But on the other hand, if there are no individuals apart from judgers, what is it that judgers accept or reject? So there must be individuals apart from judgers. Obviously, it is they who set the standard for correct judgment. If so, how can they not exist? How can existence not belong to them? If they neither exist nor do not exist, then how can a judgment that a particular individual exists be correct? For if my affirmation of the tree before me is correct, this is so because the tree exists.

The absurdity of eliminativism consists in this, that it tries to affirm individuals apart from judgers while denying the existence of individuals (subjective genitive) apart from judgers. It hides this absurdity from itself by continuing to use the expression ‘existence of individuals’ but in an obfuscatory way that transforms this expression, which is naturally read as a subjective genitive or possessive, into an objective genitive. For if the existence of individuals is the correct acceptability of individuals, and ‘existence of individuals’ is taken as a subjective genitive, then the consequence is an untenable idealism. If, on the other hand, ‘the existence of individuals’ is read as an objective genitive, then idealism is avoided, but we are saddled with a wholly unnatural revision of the meaning of the phrase in question, and a brute stipulation that we are under no obligation to accept.

The truth of the matter is that one cannot correctly affirm an individual without affirming the existence of the individual (subjective genitive), and that it is the logically antecedent existence of the individual that makes any correct affirmation of it correct.

5. The Critique Extended to the ‘Fressellian’ Position

If Brentano’s eliminativism is untenable, how do matters stand with the eliminativism of Frege and Russell?

Lumping Frege and Russell together, we may say that on the ‘Fressellian’ theory, (i) existence is never a property of individuals (‘exists’ is never an admissible first-level predicate); (ii) existence is a property of properties (‘exists’ is a second-level predicate); (iii) existence is the property of being instantiated. It follows that if a first-level property is instantiated, it is instantiated by an individual that neither exists nor does not exist. (Recall my section 1 argument why the Fressellian doctrine cannot be interpreted along

[HPQ 325]

identitarian lines.) But this, I suggest, is simply unintelligible: we have no idea what it could mean for an individual to be real and available for ‘instantiation duty’ without existing. What could it mean to say that Socrates instantiates the property of being wise but neither exists nor does not exist? If properties (concepts = Begriffe) are functions as Frege held, what could it mean for Socrates to be the argument of the function x is wise if Socrates neither exists nor does not exist? How can a function have an argument that neither exists nor does not exist? Russell’s comparison of existence with numerousness is surely wrongheaded.

The ‘Fressellian’ doctrine, no less than the Brentanian, faces a dilemma. We must either interpret the doctrine in an identitarian or in an eliminativist way. The identitarian interpretation is impossible for the reasons given in section 1. But the eliminativist reading is equally impossible as we have just seen. If a first-level property is instantiated, then either (i) it is instantiated by an individual that exists, or (ii) it is instantiated by an individual that neither exists nor does not exist. But on (i), vicious circularity is the upshot: the existence of an individual cannot consist in the instantiation of a property if the existence of the individual in question is a logically prior condition of its instantiation of any property. On (ii), however, we are left with the unintelligible idea that a property is instantiated by an individual that neither exists not does not exist.

This dilemma comes into view only after a clear distinction between identitarianism and eliminativism is on the table. Thus the ‘Fressellian’ no less than the Brentanian doctrine may derive what credibility it has from a failure to draw this distinction.

6. Conclusion

Russell held that “an almost unbelievable amount of false philosophy has arisen through not realizing what ‘existence’ means.”[24] Russell, of course, directed this remark against the metaphysicians of existence, the Thomists for example. But if I am right, the remark can be directed with greater justice against such eliminativists as Russell himself. An unbelievable amount of false deflationism has arisen through not realizing that proposals like those of Brentano, Frege, and Russell derive what credibility they have from a failure to distinguish, as I have done here for the first time (as far as I know), between eliminativist and identitarian theories of existence. Once this distinction is clearly before the mind, and one cannot waffle between eliminativism and identitarianism garnering the benefits of both without the deficits of either, the theories

[HPQ 326]

of existence of Brentano, Frege, and Russell fall to the ground. Existence is restored to its rightful place in individuals, and the metaphysics of existence is cleared of the main objection that can be brought against it.

NOTES

[1]. Gustav Bergmann, Realism: A Critique of Brentano and Meinong (Madison: The University of Wisconsin Press, 1967), p. 234.

[2]. An individual may be defined as anything that has properties and stands in relations, but is not itself a property or a relation. Thus Socrates is an individual, while being taller than is not. By this definition, such abstract objects as sets are individuals. But what I have in mind primarily when I use ‘individual’ are concrete individuals, which may be defined as individuals capable of acting and being acted upon, or else capable of entering into events that stand in causal relations. Examples of concrete individuals include clouds, rocks, people, and black holes.

[3]. At least as standardly interpreted. Cf. J. L. Mackie, The Cement of the Universe (Oxford, 1980), p. 59. Mackie rightly asserts that Hume’s theory “aim[s] at reform rather than analysis of our ordinary concepts.”

[4]. Bertrand Russell, “The Philosophy of Logical Atomism” in Logic and Knowledge, ed. R. C. Marsh (New York: Capricorn Books, 1971), p. 232 ff.

[5]. Gottlob Frege, The Foundations of Arithmetic, trans. J. L. Austin (Evanston: Northwestern University Press, 1968), p. 65.

[6]. Talk of a propositional function being ‘sometimes true’ is Russell’s. See Bertrand Russell, op. cit. p. 232.

[7]. Of course, I am not suggesting that these are good theories.

[8]. Franz Brentano, Psychology from an Empirical Standpoint, trans. Rancurello, Terrell and McAlister (New York: Humanities Press, 1973), p. 229

[9]. Ibid., p. 229

[10]. Ibid., p. 208.

[11]. Cf. Franz Brentano, The True and the Evident, trans. Chisholm et al. (London: Routledge and Kegan Paul, 1966), p. 109 et passim.

[12]. “Seiend und existierend heisst, wie wir schon wiederholt betonten: was mit Recht anerkannt werden kann.” Anton Marty, Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie (Hildesheim: Georg Olms Verlag, 1976), p. 214. First published in 1908.


[HPQ 327]

[13]. Cf. Brentano, The True and the Evident, loc. cit.

[14]. If they are taken in neither an identitarian nor in an eliminativist way, they would not amount to a theory of existence. If I want to know what existence is, it is not enough to be told that a thing exists if and only if it is correctly acceptable by someone. For I can still ask: But what is it for an individual to exist? If I want to know what it is to be trilateral, it doesn’t help to be told that, necessarily, a thing is trilateral if and only if it is triangular.

[15]. If existence itself exists, then it is itself an existent. If existence is different from existents, then it can only exist or occur in existents.

[16]. Brentano, Psychology, op. cit., p. 210.

[17]. The difference between subjective and objective genitives is nicely illustrated by the Old Testament text, “The fear of the Lord is the beginning of wisdom.” Exercise for the reader: explain which is which, and why.

[18]. Ibid., p. 294.

[19]. For a thorough discussion of the distinction by a philosopher within the Brentano school, see Anton Marty, Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie (Hildesheim: Georg Olms Verlag, 1976), pp. 205-276.

[20]. Brentano, Psychology, op. cit., pp. 213-214.

[21]. In sentences like ‘The Charles exists pollutedly’ and (more idiomatically) ‘The Charles exists contingently,’ ‘exists’ is obviously functioning not only to express existence, but also as a sort of copula. We could say that such uses of ‘exists’ are ‘copulationally loaded.’

[22]. Actually, not merely possibly. The golden mountain for Meinong is actually golden, even though it neither exists, subsists, or has any mode of being whatsoever.

[23]. A mistake Fred Sommers makes. See Chapter 5 of my A Paradigm Theory of Existence: Onto-Theology Vindicated (Dordrecht: Kluwer Academic Publishers, 2002), pp. 127-151.

[24]. Bertrand Russell, op. cit., p. 234.