I recently purchased, but then returned, Paul Edward’s Heidegger’s Confusions (Prometheus, 2004) when I found that it is nothing but an overpriced reprint of previously available materials. Twenty dollars for a thin (129 pp.) paperback is bad enough, especially given the mediocre production values of Prometheus Books; but the clincher was my discovery that there is nothing in this volume that has not appeared elsewhere. Edwards and his editors didn’t even bother to change the British quotation conventions in use in two of the reproduced articles to their Stateside counterparts.
There is also the question of the quality of Edward’s Heidegger-critique, a topic I plan to treat more fully in a separate post. But for now a comment on Edwards’ refutation-strategy in his second chapter, "Heidegger’s Quest for Being." (What follows summarizes, but also extends, the discussion in my article, "Do Individuals Exist?" Journal of Philosophical Research, vol. XX (1995), pp. 195-220, and my book A Paradigm Theory of Existence (Kluwer 2002), Chapter 4.)
In a nutshell, the Edwards strategy is this: Heidegger assumes something that Russell denies; therefore, Heidegger is wrong. But it is actually worse than that given Edwards’ liberal admixture of invective and insult. There is much to criticize in Heidegger, but hostile polemic, though it may serve some purpose in the political sphere, is out of place in philosophy. I’ll try not to pay back Edwards’ nastiness in his own coin, but I may not succeed. In any case, Edwards or his shade may profit from a taste of his own medicine.
In The Philosophy of Logical Atomism (1918), Bertrand Russell held that a vast amount of false philosophy has arisen from philosophers’ not knowing what existence is, namely, a property of propositional functions, rather than a property of the individual values of such functions. (Robert C. Marsh, ed. Logic and Knowledge, pp. 175-281, esp. pp. 232-235) A simple way to see Russell’s point is by considering the following two (putative) instances of the Fallacy of Division:
A. Humans are numerous; Socrates is human; ergo, Socrates is numerous.
B. Humans exist; Socrates is human; ergo, Socrates exists.
(A) is a clear example of the Fallacy of Division, the fallacy of assuming that whatever is true of a whole is true of each of its parts. Humans as a whole are numerous, in the sense that the class of humans has a very high cardinality indeed; but it doesn’t follow that each member of this class is numerous. Indeed, it is nonsense to say of an individual that it is numerous since numerousness can only be a property of classes and cognate items, and never a property of individuals.
Russell’s point is that (B) is ‘on all fours’ with (A), or, if you prefer, that (A) and (B) are in the same logical boat: both are instances of the Fallacy of Division. From the fact that humans exist, it does not follow that each member of this class exists. Indeed, for Russell, it is nonsense to say of an individual that it exists, since existence can only be a property of some such higher-order item as a set, a property, a concept, or a propositional function. Existence thus becomes a set’s having elements, a property’s being instantiated, a concept’s having objects falling under it (in Fregean jargon), or a propositional function’s being "sometimes true," as Russell puts it in the place above cited.
Technicalities aside, we can call this an instantiation account of existence and compress it into the slogan, ‘Existence is instantiation.’ It follows straightaway that existence cannot be ascribed to any individual: no individual can be meaningfully said to be instantiated.
Now if Russell is right about existence – if it cannot legitimately function as a first-order or first-level property no matter how broadly we take ‘property’ – then the metaphysics of existence, whether of Thomist, Heideggerian, or any other stripe, is an impossible enterprise, "rubbish," to use Lord Russell’s scholarly term of disapprobation. Inquiries into what it is for an individual to exist, or into the Being of that-which-is (das Sein des Seienden), are one and all predicated on a logical mistake.
Thus Edwards’ simple-minded point against Heidegger is simply this: the latter failed to grasp the elementary logical point that ‘. . .exists’ and cognates cannot legitimately function as first-level predicates. Russell was right, so Heidegger was wrong, and the quest for Being is a nonstarter. Of course, Edwards, being Edwards, adds insult to this injury: Heidegger is not merely wrong but lacks the intelligence to grasp Russell’s simple point, etc.
The truth, however, is that Heidegger’s assumption that existence is predicable of individuals is far more credible than Russell’s denial of this assumption. Edwards uncritically assumes the truth of the Russellian dictum in order to beat Heidegger over the head with it. It is Edwards, not Heidegger, who is the uncritical one. I will now explain what is wrong with the Russellian dictum.
Consider a sentence like ‘Cats exist.’ Sentences of this form are known in the trade as affirmative general existentials. Such sentences submit readily to the Russellian analysis. It is plausibly maintained that ‘Cats exist’ is not about cats but about the property of being a cat, and says of this property that it has instances. The virtue of this analysis emerges when we consider a negative general existential like ‘Unicorns do not exist.’ This true sentence cannot be about unicorns – after all, there aren’t any – it is about the property of being a unicorn, or the concept unicorn, and says of it that it has no instances. In this way one can deny the existence of unicorns without having to presuppose that there are unicorns. A well known paradox of reference -- dubbed by Quine Plato's Beard-- is avoided.
Now consider a negative singular existential, ‘Pegasus does not exist,’ to coin an example. Unless one is prepared to countenance Meinongian nonexistent objects, this sentence cannot be about what it appears to be about: its grammatical form hides, rather than reveals, its logical form. On a Russell-type analysis, this sentence is not about Pegasus, but about the concept winged horse of Greek mythology, and what it says about this concept is that it is not instantiated.
The Russell-type analysis works admirably well when it comes to negative existentials, whether general or singular. But it breaks down when applied to affirmative existentials, especially singular affirmative existentials. Consider ‘I exist.’ If Russell is right, and existence is in the same logical boat with numerousness, then ‘I exist,’ taken as predicating existence of an individual, is not false, but meaningless. So it has to be analyzed as predicating instantiation of a property. But which property? It has to be a property that individuates me, that distinguishes me from everything else, not merely in the actual world, but in all possible worlds in which I exist. Thus a property that only I have, but some other individual might have had, is not such that its instantiation could be identified with my existence. Let me explain.
Suppose I am the meanest man in the Superstition Mountains. Then I instantiate the property P of being the MMSM. Clearly, it is merely a contingent fact about me that I instantiate P. I might not have instantiated the property in question: my grizzled partner, Seldom Seen Slim, might have been the meanest man in the Superstition Mts., and when I die, Slim will no doubt be the meanest man in these parts. It follows that my existence cannot be identified with the being-instantiated of this property. For if anything is clear, it is that my existence is bound up with my identity: my existence is essential to me; my being the meanest man in the Superstition Mountains is not.
To make a long story short, haecceity-properties must be introduced to make the Russellian analysis work. Take the (putative) property of being identical to BV. This property individuates me in the actual world and across all possible worlds. If anyone has it, I have it; and no one distinct from me can have it. It captures my thisness, my haecceitas. If there is this property, then the following equivalence holds necessarily:
(E) BV exists if and only if the property of being identical to BV is instantiated.
But do biconditionals of this form show that existence cannot be ascribed to individuals? Note first that biconditionals, even if necessarily true, do not sanction reductions. It is necessarily true that triangularity is instantiated if and only if trilaterality is instantiated; but that scarcely shows that the the two properties are identical.
Note second that (E) is blatantly circular. For if the property of being identical to BV is instantiated, then it is instantiated by an individual, and this individual must exist to do any instantiating. Russell’s attempted elimination of singular existence (existence as ascribable to individuals) ends up presupposing singular existence.
Note third that if Russell and Co. are right, then the left-hand side of (E) is meaningless, as meaningless as ‘BV is numerous.’ It follows that (E) is meaningless. So (E), if true, is meaningless, and if meaningless, is meaningless. Therefore, (E) is necessarily meaningless, which entails that it necessarily lacks a truth-value.
Of course, if you say that the left-hand side of (E) expresses the same proposition as is expressed by the right-hand side, then (E) degenerates into a miserable tautology, in which case it cannot be used to effect a substantive metaphysical reduction/elimination.
But even if these these three unanswerable objections could be answered, there remains one that is truly insurmountable, namely, that there are no haecceity properties. But I will leave the proof of this for a separate post. (Cf. A Paradigm Theory of Existence, pp. 99-104.)
The upshot is that the Russellian theory cannot satisfactorily analyze affirmative singular existentials. It also cannot accommodate affirmative general existentials. It is undeniable that
(F) Cats exist if and only if the property of being a cat is instantiated.
This presupposes singular existence in two ways. First, if the property in question is instantiated, then (instantiation being a relation), it is instantiated by individuals, and they must exist if they are to do an instantiating. Second, the property itself must exist. Nothing can have properties unless it exists, so the property of being a cat cannot have the property of being instantiated unless it exists. But the property’s existence cannot consist in some other higher-order property’s being instantiated on pain of a vicious infinite regress. Therefore, the property must singularly exist – in which case singular existence has not be eliminated but is presupposed. Twist and shout as a Russellian might, he won’t shake off singular existence.
There is also this to consider. (F) is undoubtedly true. Why? The explanation is that the sentences on either side of this biconditional express one and the same proposition. General existence just is instantiation. Singular existence, however, cannot be instantiation, as was seen in the discussion of (E).
This brings me to what I shall term Russell’s Fallacy, to wit, the fallacy of confusing general existence (instantiation) with singular existence. Russell’s mistake is to think that because general existence is identifiable with instantiation, existence is so identifiable. Not so. Singular existence cannot be identified with the instantiation of any property.
Paul Edwards, thoughtless positivist, cantankerous curmudgeon, blinded by his animus against Heidegger and anything that transcends his prosaic anti-metaphysical pate, uncritically repeats Russell’s Fallacy to the point of denying the self-evident truth that existence in its fundamental singular sense belongs to individuals so that it does make sense, after all, to ask about the existence, and more generally, the Being, of the things that are.
The quest for Being easily survives the Russell-Edwards assault. Of course, this is not to say that Heidegger’s ‘theory’ of Being is worth much. I myself don’t think it is. But that’s another story.