Existence Exists: Analytic or Synthetic?

 Recently over the transom:

I am A. Kashfi, Professor of philosophy from Tehran University, Iran.

I am currently engaged in studying your esteemed book A PARADIGM THEORY OF EXISTENCE. In this book, you argue that “existence exists”. Regarding this proposition, a question has arisen for me. I would be grateful to have your response.

Is this proposition analytical or synthetic?

If this proposition is analytical, its equivalent would be: "Existence is existence" or "Existent is existent," which, as it is evident, doesn't contain particularly useful information.

If this proposition is synthetic, it requires that the concept of “existence” be distinct from the concept of “existent”. I want to know what the distinction between these two concepts is. Here, which concept "other than existence" (note: distinct from existence), in accordance with the synthetic nature of the mentioned proposition, are we attributing to existence?

 

In other words, I understand the proposition "A tree exists", but what does the proposition "existence exists", (given the synthetic nature of this proposition), mean?

Yours sincerely, A. Kashfi

Thank you for writing, Professor Kashfi. Nothing in philosophy fascinates me more than the topic of Existence, and so it is with pleasure that I think through your questions.

To understand what I mean when I say that Existence exists you have to understand that I distinguish between existing items (existents) and Existence.  Thus I do not use 'existence' as some philosophers do to refer to existents collectively.  Nor do I mean by 'Existence exists' what Ayn Rand means by it. The distinction between Existence and existents, as I construe it, is motivated by (i) the apparent fact, evident to the senses,  that there are many existents but that (ii) these existents all have something in common, namely, Existence. Existence is one to their many as that in virtue of which the many existents exist. The distinction gives rise to several questions. Here are four. First, what is it for an individual existent to exist? Second, what is Existence itself in its difference from individual existents? Third, how is Existence common to existents? Fourth, does Existence itself exist?

Reinhardt Grossmann proffers a quick answer to the fourth question: How could it fail to? "If existence did not exist, then nothing would exist." (The Categorial Structure of the World, Indiana UP, 1983, p. 405) Of course, he is not talking about my theory, but his own. He goes on to say that it does not matter to which category you assign Existence.  Whatever Existence is, it must exist if anything is to exist. I argue in my book that Existence cannot be a first-level property and thus that it cannot be that existents exist by instantiating Existence, not that this is what Grossmann maintains. Suppose that I am wrong and that Existence is a first-level property, a property of individual existents, and that the latter exist by instantiating Existence.  Then surely that property would have to exist if anything exists. Here then is an example of a meaningful use of 'Existence exists.' If Existence is a first-level property, a property of individuals, then Existence must exist if anything is to instantiate it.

The distinction between existents and Existence is nothing new. In Aquinas it is the distinction between ens/entia and esse.  In Heidegger it is the distinction (ontologische Differenz) between das Seiende and das Sein. We also find it it Islamic philosophy. Fazlur Rahman, glossing Mulla Sadra, writes, "Existence is that primordial reality thanks to which things exist . . ." (The Philosophy of Mulla Sadra, SUNY Press, 1975, p. 28. Diacriticals omitted.)

Sadra is clearly distinguishing between the things that exist (existents) and that in virtue of which they exist, Existence. There are of course very important differences between the three thinkers mentioned, and between their views and mine. But there is a close affinity between my view and that of Aquinas, and a somewhat close affinity between my view and that of Mulla Sadra.   

For Aquinas, Existence itself exists as God: Deus est ipsum esse subsistens. I am using 'Being' and 'Existence' interchangeably. For Aquinas, then, God is (identical to) self-subsisting Being.  God is both Being (esse) and the supreme being (ens).  In my jargon, the God of Aquinas is the Paradigm Existent. God does have have esse; he is (identical to) esse. So the Paradigm Existent  is both Being (esse) and being (ens). That is equivalent to saying that Existence exists. 

Aquinas is saying that Being itself is. On my reading, he is making three interconnected claims. (1) Being is not other than every being, as it is for Heidegger. His is not an 'alterity' theory of Being.  (2) Being does not divide without remainder into beings. He rejects what I call radical ontological pluralism.  (3) God (self-subsistent Being) is not a being among beings; God is the being, where 'the' connotes uniqueness. See God: A Being among Beings or Being itself? Aquinas thus rejects an ontic conception of Being/God. Everything other than God is in a dependent and derivative way or mode. It is important to note that God for Aquinas is not only unique, but uniquely unique: unique in his very mode of uniqueness. If you understand what Aquinas is saying, then you will understand what I am saying when I say that Existence itself exists. Existence itself is Existence in its difference from the phenomenal existents which derive their existence from the Paradigm Existent.

'Ens' (being) is the present participle of the infinitive 'esse' (to be). This linguistic fact points us in a Platonic direction: phenomenal existents (you, me, my cats, the Moon, Trafalgar Square, my bicycle, its parts such as the chain, and its parts, the links . . .) participate in noumenal Existence. In virtue of this participation, phenomenal existents exist and form a unified plurality of existents. This plurality is no illusion. It is real, but derivatively real. What is derivatively real, however, is not ultimately real. I agree on this point and others with Plato, Aquinas, and Sadra.

Taking a further step in the Platonic direction, I will note that instead of 'Paradigm Existent' I could have used 'Exemplary Existent.' Both exemplars (paradigms, standards) and universals are ones-over-many, but an exemplar is not a universal. Universals have instances, but Existence has no instances.  Exemplification is not instantiation.

As for Sadra, if Existence is the "primordial reality," then this is tantamount to saying that Existence itself exists. For if Existence is real, then it cannot have a merely conceptual or mental status, as it would be if it were a product of abstraction, and if it is the primordial reality, then everything other than it is real in virtue of being dependent on it. As for Heidegger, while he too distinguishes Being from beings, he denies that Being itself is. Das Sein ist kein Seiendes! The overly triumphalistic subtitle of my book, "Onto-theology Vindicated" was meant to signal my opposition to Heidegger, whose critique of what he calls metaphysics is in part a critique of onto-theology. An onto-theological approach to Being avoids both the alterity view and the ontic view. But to explain this in any depth is beyond the scope of this response. See my Heidegger category for more on Schwarzwaldontologie. See also Three Theisms: Ontic, Alterity, and Onto-Theological and their Liabilities.

Analytic or Synthetic?

I will now respond to Professor Kashfi directly. He asks whether 'Existence exists' is analytic or synthetic and finds difficulties either way. My short answer is that Kashfi's question is not relevant to my broadly Platonic view. His question, couched in Kantian terms, is modern; my theory, harking back to Platonic exemplarism, is ancient. His question presupposes that Being is a being among beings. But that I deny. Now to the details.

Immanuel Kant applies the terms 'analytic' and 'synthetic' to judgments (Urteile). In the simple categorical case, a judgment involves a relation between a subject-concept and a predicate-concept. Thus the judgment expressed by 'Bodies are heavy,' (Kant's example of a synthetic judgment a posteriori) relates the concept body to the concept heavy via the copula 'are.'  But there has to be more to it than this, Kant insists, since we need to know "in what the asserted relation consists." (CPR B 141) His answer is that the relation of subject-concept to predicate-concept in a judgment is grounded in the bringing-together of concepts in the objective unity of apperception.  It is this objective unity of apperception (self-consciousness) that is "intended by the copula 'is.'" (B 142) In  Prolegomena to Any Future Metaphysics, section 22, Kant writes, "The union of representations (Vorstellungen) in one consciousness is judgment." If these representations are united in the consciousness of a particular person, then the judgment is "accidental and subjective." If, however, they are united in a "consciousness in general," then the judgment is "necessary and objective."  This consciousness in general is what in Critique of Pure Reason he calls the objective unity of apperception.

Kant's central problem, as explained in his letter to Marcus Herz, is this: On what ground rests the relation of that in us which we call representation to the object? A judgment is a representation composed of concepts which are themselves representations. Judgments purport to be true or objectively valid. Suppose I see a green tree and judge that the tree is green. The judgment purports to be true  whether I or anyone make it. The purport is that the tree is green in reality apart from us  and our subjective mental states. I have an empirical representation of  green and an empirical representation of tree.  What I don't have is an empirical representation of what the 'is' denotes.  I have no empirical representation of the copulative tie or, equivalently, I have no empirical representation of the existence of the green tree. (The tree is green if and only if the green tree exists.) So how do I know that the tree is green? How do I secure the objective validity of the judgmental representation? What is the ultimate ground of the synthesis of subject and predicate in the object?  What makes it the case that the judgment expressed by 'This body is heavy' is true independently of my particular mental state and thus true for all actual and possible finite cognizers? What assures me that the judgental purport is satisfied? This is Kant's problem. To put it oxymoronically, it is a classically  modern problem. (The modern period in the West begins with Descartes, 1596-1650.)

Kant's solution is a transcendentally idealist one. The ultimate ground of the synthesis of subject and predicate in the object is is supplied by the objective unity of apperception which is also the transcendental unity of apperception. This solution is fraught with difficulties. For me, the central difficulty is the one I tackled in my doctoral dissertation: what exactly is the status of this transcendental unity of apperception? But that is basically what Kant is maintaining: we, in our transcendental capacity,  constitute objects in their objectivity. For we are the source of the the objective synthesis that lends objectivity to judgments.

Whether a judgment is analytic (e.g., 'All bodies are extended') or synthetic ('This body is heavy'), all such judgments are about phenomenal particulars in space and time.  But neither Kant's transcendental unity of apperception nor my Paradigm Existent is a phenomenal particular in space and time. For Kant, the ultimate transcendental condition of anything's being an object is not itself an object among objects. Similarly, the The Paradigm Existent is not an existent among existents. It is no more such than the God of Aquinas is a being among beings. 

And so I say that the question 'Analytic or Synthetic?' is inappropriately asked of 'Existence itself exists.'

Kashfi writes, "If this proposition ['Existence exists'] is synthetic, it requires that the concept of 'existence' be distinct from the concept of 'existent'."  Kashfi thereby assumes something that I explicitly deny early on in my book, namely, that Existence is a concept. Concepts track essences. The concept triangle, for example, 'captures' the essence TRIANGLE. The existence of an existing thing, however, cannot be captured, grasped, 'made present to the mind'  by any concept.  Existence is trans-conceptual.

One reason is that existence is not essence. Another reason is that each existing thing has its own existence: Socrates' existence is his and not Plato's. The two philosophers differ numerically in their very existence. They differ numerically as existents. Thus their numerical difference is numerical-existential difference. But as Aristotle said (in Greek, not in Latin): Individuum qua individuum ineffabile est. Individuals as such are ineffable. That strikes me as obvious given that (i) there is and can be no concept that captures or grasps the haecceity (non-qualitative thisness) of an individual, and (ii) there are no haecceities except those of existing things.  (Pace Plantinga, there are no such metaphysical monstrosities as  uninstantiated haecceities.)  There are, in other words, no individual concepts. Definitions and arguments here and in the surrounding entries in the identity and individuation category.

Neither the existence of Socrates, the existence that is his alone and not possibly shared with any other existent, nor Existence itself in its difference from existents is a concept. My point is that Existence either in finite existents or in itself cannot be reduced to a concept. I am not saying that we have no concept of Existence; we do. It is just that the concept of Existence is the concept of something that is not and cannot be a concept.  Existence is in this respect like God. We have various concepts of God, but God is not a concept. Or do you think that a mere concept in a mortal's mind created the world? Similarly, do you think that that in virtue of which finite existents exist is a concept in a mortal's mind?  See The Concept GOD as a Limit Concept.

To understand what I mean by 'Existence itself exists,' you have to understand that Existence itself is like the Platonic Form/Idea, Humanity. The former, like the latter, is not self-predicable or self-instantiating: Humanity is quite obviously not human, nor a human. And because Humanity is not self-predicable, one cannot sensibly ask whether the predicate-concept human is analytically contained in the subject-concept Humanity or synthetically  attached to it. The Form/Idea Humanity exists by being (identical to) itself. Its Being is (identically) its self-identity unlike a particular human such as Socrates whose Being is not its self-identity. (If Socrates' Being were his self-identity, then he would be a necessary being, when in fact he is a contingent being.) The same goes for Existence in its different from existents: its Being is its self-identity, which implies that the Paradigm Existent is metaphysically necessary.

If Humanity were self-predicable (self-instantiating), then the Third Man Regress would be up and running. For if  'is human' is univocally predicable of both the Form Humanity and its phenomenal instance Socrates, then a second Form -- call it Humanity II -- would have to be introduced to explain what is common to both Humanity and Socrates. And so on into an infinite explanatory regress which, as explanatory, is vicious. (Sone infinite regresses are benign, e.g. the truth regress.)

The Form/Idea Humanity is a CASE of itself, but not an INSTANCE of itself. A case because Humanity  is not a universal what-determination abstractly common to particular phenomenal beings, but a paradigm or exemplar.  The standard meter bar in Paris might prove to be a useful analogy if you take it the right way (which is of course that way I want you to take it.) The standard meter bar is obviously not an instance of itself because it is a material particular, and such things do not have instances.  For the same reason, you cannot predicate the standard meter bar of itself.  The standard meter bar is nonetheless a case of itself in that it is a metal bar exactly one meter in length: it sets the standard by being (identical to) the standard.

Now the standard meter bar is a phenomenal particular relative to which other phenomenal particulars either measure up or fall short, whereas the Paradigm Existent is not a phenomenal particular.  This is a point of disanalogy. But if you understand how the standard meter bars functions as a paradigm, you should also be able to understand how Existence could so function, mutatis mutandis.

Comments

[Dominik Kowalski]

The posts on existence are always the best ones on this blog.

And well I guess now it's as good of a time as any other to ask this: Bill, in which way do you differ from Barry Miller? If I understand Miller correctly, for him the existence of an individual is a first-order predicate, which you seemingly deny. Is the disagreement any deeper and if so, what does it amount to?

One remark from Gaven Kerr, defending an explicitly Thomistic version of the argument, I always found peculiar was that his argument wouldn't work on a nominalist framework. Sadly I never got an answer as to what exactly was meant by nominalism. Having your argument in mind, it seems like the general idea can still be applied on a nominalism that affirms properties. Aren't tropes in need of unification as well? The real distinction doesn't seem to be too concerned with the status of abstract objects and universals, although I believe prior commitments and arguments about existence will lead us to a realist position, at least when it comes to universals. Nonetheless, it seems unnecessary when we just present the argument in isolation.

I never had the opportunity to read "From Existence to God". Getting my hands even on a library copy is difficult enough here. I reached out to Routledge 2 years ago, together with Editiones Scholasticae, whether we could get the license to put the book back into print, however there seemed to be no interest, there's no other explanation for the outrageous licensing and royalty fees that were demanded. I hope it will be possible at some point though.

[BV]

Thanks for your comments, Dominik, especially since you and I appear to think along similar lines. One response now, more later.

>>If I understand Miller correctly, for him the existence of an individual is a first-order predicate, which you seemingly deny. Is the disagreement any deeper and if so, what does it amount to?<<

First of all, Miller and I both distinguish between predicates and properties. Here's Miller, From Existence to God, p. 118, n. 6: "Properties are what a predicate stands for, or what are attributable to something by a predicate. The level of a property is the level of its correlative predicate." So it is not the case, as you say, that for Miller existence is a first-order (first-level) predicate.

Kant famously wrote, in the context of his critique of the Cartesian ontological argument, Sein ist offenbar kein reales Praedikat . . . ("Being is obviously not a real predicate . . ." KdrV A598 B626) What Kant means is that Being is not real property of individuals given Miller's distinction between predicates and properties.

As for the difference between Miller and me, it is terminological. In my book I define properties as instantiable entities. (If memory serves, I got this definition from Roderick Chisholm whose NEH seminar I attended at Brown University in the summer of '81.) I probably say in my book (I am too lazy to get out of my chair and check): P is a property =df P is possibly such that it is instantiated.

Max is one of my cats. It is evident that Max exists. I take it to be a datum that the declarative sentence 'Max exists,' when assertively uttered by someone, attributes existence to Max. The same goes for Max's brother Manny. Both of my 'tuxies' (Tuxedo cats) exist. They have something in common, namely existence. Having defined 'property' as just did, I go on to argue that existence cannot be a property of existing individuals. For it cannot be that Max, Manny, the bobcat that just walked by . . . exists by instantiating existence. (See PTE for the arguments.) So existence cannot be a first-order/first-level property, as I define these terms.

And yet Miller and I agree that, in my way of putting it, not his, existence "belongs to individuals" such as Max in a way that it would not belong to individuals if existence were a second-level property of such items as concepts, properties, propositional functions, and some other items I discuss in my book.

This sets up one of the problems my book tackles: What is it for an individual to exist if existence is not a property at any level? (It seems to me that I utterly demolish the still-popular 'thin theory of existence')

Now Dominik, tell me what you agree with and what you disagree with supra, and we will take it from there.

[Dominik Kowalski]

Thank you for your answer, Bill. I don't see us disagreeing on anything actually. Another argument against the thin theory that I like to defend is the incompatibility of it with the PSR, since it's not possible to give an account of how the necessary being can be necessary. I haven't found C. J. F. Williams tackling that issue, and neither Quineans like Peter van Inwagen, despite the latter having an article called "God's Being and Ours". If you have a reference that I have missed, please let me know.

The only difference may be that I'm not comfortable defending any definitive theory of properties. Though I'm inclined towards a constituent ontology, nonetheless I'm interested in how the argument is affected by different views on properties. Prima facie there seems to be no problem with combining the idea of external unification with a trope theory and even a mad dog nominalist like William Lane Craig, who toys with ideas like denial of properties, still has to keep a difference between individual and existence, however he might want to cash that out.

As I've said, I think there's an entailment to be argued for, meaning that the nature of existence of an individual in combination with the existence of the paradigm existent yields an argument for a particular view of properties. Arguably you have done something similar in ch. 3, arguing that a non-constituent ontologist has to deny the existence of individuals. But I'll freely admit that the arguments for the constituent ontology need to get carefully read again by myself at some point. The last time it went over my head.

Thanks for the correction about the distinction of property and predicate. This cleans up a lot.