Kant, Spinoza, and the Metaphysics of the Ontological Proof
This paper provides an interpretation and evaluation of Spinoza’s highly original version of the ontological proof in terms of the concept of substance instead of the concept of perfection in the first book of his Ethics. Taking the lead from Kant’s critique of ontological arguments in the Critique of Pure Reason, the paper explores the underlying ontological and epistemological presuppositions of Spinoza’s proof. The main topics of consideration are the nature of Spinoza’s definitions, the way he conceives of the relation between a substance and its essence, and his conception of existence. Once clarity is shed upon these fundamental issues, it becomes possible to address the proof in its own terms. It is then easy to see that Kant’s objections miss their target and that the same is true of those advanced by another of the ontological argument’s most famous critics, Bertrand Russell. Finally, several interpretations of Spinoza’s proof are proposed and critically evaluated; on all of them, the argument turns out to be either invalid or question-begging.
KeywordsKant Russell Spinoza Ontological argument Ideas Possibilities
“...conceive, if you can, that God does not exist.”—Spinoza, Ethics
1 Kant’s Critique of the Ontological Argument
“God does not exist” is a self-contradictory statement, and therefore necessarily false.
The negation of a necessarily false statement is a necessarily true statement.
Therefore, “God exists” is necessarily true.
If, in an identical proposition, I reject the predicate while retaining the subject, contradiction results; and I therefore say that the former belongs necessarily to the latter. But if we reject subject and predicate alike, there is no contradiction; for nothing then is left that can be contradicted. To posit a triangle, and yet to reject its three angles, is self-contradictory; but there is no contradiction in rejecting the triangle together with its three angles. The same holds true of the concept of an absolutely necessary being. If its existence is rejected, we reject the thing itself with all its predicates; and no question of contradiction can then arise. (Kant 1961: 502 [A595/B623])
“A triangle has three sides”
“God is almighty”
“There are triangles”
“There are no triangles”
“God does not exist”
The logical counterpart of “A triangle has three sides” is “God is almighty.” Kant argues that they are both necessarily true; the reason why this is so is that they are analytic: the predicate (“having-three-sides,” “being-almighty”) is included in the concept of the subject (“triangle,” “God”). This is why, in the passage quoted above, Kant says that “to posit a triangle, and yet to reject its three angles is self-contradictory.” A statement such as “There are triangles,” however, belongs to a different logical type. According to Kant, what it says is that there are objects (Gegenstände) that instantiate the concept (Begriff) of a triangle. Analogously, “God exists” should be read as “The concept of God finds exemplification in reality.” Grammatical appearances notwithstanding, “God exists” is not a statement about God, but about the concept of God, and it is not analytical. On this interpretation, it becomes possible for Kant to say that “there is no contradiction in rejecting the triangle with its three angles”. Mutatis mutandis, there is no contradiction in “rejecting” God with all of its properties (omniscience, omnipotence, etc.).
‘Being’ is obviously not a real predicate; that is, it is not a concept of something which could be added to the concept of a thing. It is merely the positing of a thing, or of certain determinations, as existing in themselves...If, now, we take the subject (God) with all its predicates (among which is omnipotence), and say ‘God is’, or ‘There is a God’, we attach no new predicate to the concept of God, but only posit the subject in itself with all its predicates, and indeed posit it as being an object that stands in relation to my concept. The content of both must be one and the same... the real contains no more than the merely possible. A hundred real thalers do not contain the least coin more than a hundred possible thalers. (Kant 1961: 505 [A 600/B 628])
Otherwise stated, if we read existential statements as statements about things and not about concepts, then they are empty of content. What are we ascribing to a thing when we say that it exists? Like Hume before him, Kant contends here that there is no real property of things that answers to the predicate “existence.” The real function of the grammatical predicate is that of “positing the thing,” rather than to say something about the thing’s actual constitution. Misled by superficial grammatical appearances, those who advance the ontological argument fail to see that existence is not a genuine property; eventually, this failure leads them to overlook the difference in logical kind between the statements “God is almighty” and “God exists.”
It is important to notice that these criticisms are independent of Kant’s overall epistemological project. Rather than as a consequence of the theory of the limits of human knowledge exposed in the first Critique, the refutation of the ontological argument is meant to provide independent support and confirmation for the thesis that knowledge out of pure reason is impossible. One might not endorse Kant’s critical philosophy, but still reject the ontological argument on Kantian grounds. And this is, arguably, what most contemporary philosophers actually do.
2 The Ontological Proof and the Metaphysics of Perfection
God has all perfections.
Existence is a perfection.
Therefore, God necessarily exists.
...there is nothing we can be more certain than we are of the existence of an absolutely infinite, or perfect, Being—that is, God. For since his essence excludes all imperfection, and involves absolute perfection, by that very fact it takes away every cause of doubting his existence, and gives the greatest certainty concerning it. (p11s)1
This passage provides some prima facie evidence for ascribing to Spinoza the ontological argument in its prototypical form (although the evidence is perhaps not as compelling as one would like to have it, since Spinoza does not say here either that existence is a property or that it is a perfection). This is not, however, the only version of the argument one finds in Spinoza, nor is it the main one. There is a widespread agreement among interpreters that in arguing for the existence of God Spinoza appeals to the concept of substance, rather than to that of perfection.2 Hence, two questions arise: do Kant’s critical remarks have any validity against Spinoza’s original formulation of the ontological argument? And, granted there is a way of saving Spinoza’s proof from Kant’s condemnation, does he succeed in providing, if not a proof, at least a plausible argument for the thesis that God exists?
3 Spinoza’s Ontological Argument: a Preliminary Survey
There cannot be a substance with no attribute. (implicit premise)
There cannot be two substances of the same attribute. (p5)
God, the substance that possesses all attributes, necessarily exists. (p11)
Therefore, God is the only substance. (p14, Spinoza’s substance-monism)
God is the only substance. (p14)
Except for substances and modes there is nothing. (axiom 1)
Therefore, whatever is, is in God. (p15, pantheism)
The logic of the first part of the argument is clear enough: if substances need not-shareable attributes in order to exist, and if one substance possesses all available attributes, then that substance will be the only existing one. It is God’s very magnitude that makes it impossible for other substances to exist. In this part of the argument, serious difficulties of interpretations arise, especially in connection with the notion of attribute. Some of Spinoza’s basic ontological assumptions will be discussed in what follows; however, attention will be focused upon those concepts that are needed to understand Spinoza’s justification for p11, the conclusion of his own version of the ontological argument. In this context, the pivotal notion is not that of attribute, but of essence. The outline of Spinoza’s argument for pantheistic monism also illustrates the vital role the ontological proof plays in his thought; if the proof fails, his whole metaphysics remains unsupported.
d6: By God I understand a being absolutely infinite, that is, a substance consisting of an infinity of attributes, of which each one expresses an eternal and infinite essence.
d3: By a substance I understand what is in itself and is conceived through itself...
d1: By cause of itself I understand that whose essence involves existence, or that whose nature cannot be conceived except as existing.
p7: It pertains to the nature of a substance to exist.
God is a substance. (d6)
A substance necessarily exists. (p7)
Therefore, God necessarily exists. (p11)
Apparently, Spinoza is here attempting to derive the existence of God (p11) from his definition (d6). What is, however, a definition? This is a crucial question to be considered later. One natural answer is to say that d6 formulates Spinoza’s idea of God. Hence, there is prima facie justification for classifying Spinoza’s argument as “ontological” in the Kantian sense—that is, as one that tries to deduce the existence of God from his bare idea.
If something is God, then it is a substance.
A substance necessarily exists.
Therefore, if something is God, then it necessarily exists.
There is one and only one x that is God, and this x is a substance.
Since this incorporates the statement that God exists, it can’t be used as a premise in the ontological argument without introducing a vicious circularity. Thus, here is the problem: Kant’s interpretation of the first premise yields a non sequitur, with Russell’s we are guilty of a petitio principii. Is there a third way of analyzing the first premise, “God is a substance,” so that Spinoza’s argument can be construed as a valid argument?
4 Are Spinoza’s Definitions Stipulative Definitions?
the common usage of the word “God”;
a novel usage of that word, or the concept or notion that should be associated with it;
the thing (whatever that is) that that word/concept supposedly refers to.
If the definitions of Ethics I are intelligible, they must be nominal (i.e. of words), not real (i.e. of things). But Spinoza declares elsewhere in the Ethics (I, 8S2) that true definitions are of things; and the form of words he adopts at the outset of the First Part indicates that it is substance, not ‘substance’, which is being defined.
The question whether Spinoza’s definitions there are real or nominal must be distinguished from the question whether they are truth valued...There are good reasons to think that the definitions of Ethics I are stipulative; and if they are stipulative, then they are non-truth-valued. (Delahunty 1985: 91)
...the true definition of each thing neither involves nor expresses anything except the nature of the thing defined.
Although Spinoza is explicitly talking of his definition of substance when he makes this claim, it is only natural to think that what he says applies to the other definitions at the beginning of the Ethics as well. Delahunty has two reasons to deny that Spinoza’s interpretations were meant to be real. One is the form in which the definitions are cast (“By God I understand a being absolutely infinite...”). By itself, this does not suffice to show that Spinoza’s definitions are stipulative. Since Spinoza wants to make sure that we think of God in the way he does, it is just natural that he should use the locution “by...I understand...” in his definition. Spinoza’s correspondence also contains some important methodological clues. His second epistle to Oldenburg makes it clear that it would not be inconsistent for him to use this turn of phrase while also believing to have provided a real definition. In this letter, he uses a formulation quite similar to the one provided in the Ethics, yet he also goes on to say that there is little doubt that his definition of God is the true one: Verùm, quòd haec sit vera Dei definitio, constat ex hoc, quòd per Deum intelligamus Ens summè perfectum, & absolutè infinitum (Spinoza 1925b: 8). Interestingly, this passage suggests that the definition of God as a being having infinite attributes does not stand in contrast with what we usually mean by God (intelligamus). Based on the scanty evidence provided by the second epistle, Spinoza’s definition would be an analysis of the ordinary conception of God as an infinite being, i.e., d6 expresses clearly what we already, yet confusedly think when we think of God.
If Spinoza had advanced real definitions purporting to be true at the outset of the Ethics, they would indeed have been combated by many of his readers. (Delahunty 1985: 95)
Yes, they would have. But why should this show that Spinoza did not intend his definitions to be real? He would have been equally open to attack if he had advanced his definitions as stipulative ones. For a reader dissatisfied with Spinoza’s conclusions, it is as easy to ask “How do you know what God truly is?” as it is to ask “Why do you want to define God in that way?” (As will be shown shortly, moreover, Spinoza does even answer the first question in the Ethics.) Upon the whole, Delahunty’s explanation of why Spinoza could not have meant his definition to be real is unconvincing. Most importantly, Spinoza’s definitions can perform the logical role they are supposed to perform, only if they are taken to be real. Consider what would happen to Spinoza’s ontological argument if d6 were only nominal. What kind of information would d6 convey? It would not be a definition of a real thing; hence, it would have to specify logical connections between concepts. But this would mean analyzing d6, “God is... a substance,” as “If something is God, then it is a substance.” And it has already been shown in the previous section that, on this reading, Spinoza’s argument turns out not to imply the existence of God.
But there is a further possibility that needs to be considered. In his Tractatus de intellectus emendatione Spinoza explains that it is possible to teach the concept of a sphere to a pupil by showing how it can be geometrically constructed by imagining that a semi-circle rotates around a fixed point. (Spinoza 1925c: 27) Such a definition could be called a constructive definition; although no real sphere has ever been constructed in this way, imagining that it could be generated thus helps us to form the right mathematical conception of a sphere. Analogously, Spinoza’s d6 could be a constructive definition that shows us which concepts we must put together to achieve a correct metaphysical notion of the Deity, one that is wholly purged of misleading associations inherited from tradition, prejudices, or false metaphysics. This might explain Spinoza’s turn of phrase (“by... I understand...”) and would not be inconsistent with the main tenet of the present interpretation, namely that to understand d6 is to have formed a clear notion of something real. A constructive definition leads us to see what something is, in Spinoza’s example, a sphere, but does not generate the geometrical entity whose nature we come to grasp by means of it.
5 Things and Their Essences
But if d6 is a real definition (or at least a constructive definition that leads us to form a clear conception of a real something) of what real thing is it a definition? The answer is given in a passage from the Ethics that has been already quoted: “...the true definition of each thing neither involves nor expresses anything except the nature of the thing defined” (p8s2); since “nature” is a synonym of “essence” (as it is made clear by d1), d6 is a real definition of God’s essence. Interestingly, in the first book Spinoza defines God, mode, substance, and attribute. He does not define essence. Why is this so? One answer is that he simply assumes that there are essences, that they are required in order for there being real definitions in the first place.
What is, however, a Spinozian essence? Since Spinoza distinguishes between “essence” and “existence,” one might conclude that these are just two distinguishable aspects of an existing thing. There is no reason to think, however, that Spinoza’s essences are “abstractions” from more concrete realities. On the contrary, the most fruitful way of thinking of an essence is to think of it as being itself a real thing, something which truly is. But just how does an essence relate to the thing it is an essence of, if it is not just an abstraction from it? One answer is that the essence of a thing is the thing itself, considered as a mere possibility.4 In the present case, the essence of the individual substance God is God itself (“is” meaning here nothing less than numerical identity). This point must be considered carefully. It is not as if there were two Gods, a possible God or detached essence described by the definition and a real one whose existence is yet to be ascertained. The idea is that there is only one God, but that this can be real in two different ways, either as a possible existent or as an actual existent.
Consider, as a way of illustration, a work of art such as Michelangelo’s statue of David in Florence. On Spinoza’s understanding of essences, that statue would have occupied some place in reality even prior to its actual realization by Michelangelo. If, for some unfortunate circumstance, the David should be destroyed, it would not disappear from reality altogether, but would go on existing as David’s unactualized essence. This is pretty much how Spinoza himself explains the distinction between essences and existing things in his Cogitata Metaphysica (Spinoza 1925a: 239). Here, he asks the reader who can’t grasp the distinction to visit a sculptor. The sculptor will be able to describe the work he wants to realize in all its details before actually realizing it. According to Spinoza, one could not ask for a more concrete illustration of the distinction between essence and existence, of the essence’s independency from actual things, and of the essence’s availability to the human mind. Spinoza’s fourth letter to Oldenburg is also illuminating in this respect. At one point, he asks his friend to consider that human bodies are not created, but generated; they were, in a different guise, even before coming into existence.5
The view that an essence is the thing it is an essence of might seem strange. Surely, on this interpretation, the phrase “essence of” is a highly misleading one, as it suggests that an essence is something qualifying the individual it is an essence of, and hence in some way distinct from it. However, this interpretation does nothing more than take Spinoza at his word, since d6 is a definition of God and at the same time he says that definitions are definitions of essences; hence, there must be a sense in which God and his essence are one and the same thing. One recent interpreter that rejects the identification of an essence with the thing it is an essence of raises precisely this doubt: “The essence of a thing cannot be just the thing, otherwise there would be absolutely no distinction between the two. Hence, there would be no reason to differentiate between them or to employ the term ‘essence’ when it would simply mean ‘the thing itself’.” (Devaux 2007: 106) The answer to this is that, if one recognizes a distinction between possible and actual existence, then there is a reason to differentiate between “essence” and “the thing itself.” Only the former term refers to the thing insofar as it is a possible being; the latter leaves its existential mode wholly undetermined.
2d2: I say that to the essence of any thing belongs that which, being given, the thing is [NS: also] necessarily posited and which, being taken away, the thing is necessarily [NS: also] taken away...
Spinoza would seem to claim here that a thing’s essence is a thing’s sufficient and necessary condition, which again suggests that the essence is not identical with the thing it is the essence of. However, 2d2 is surely not inconsistent with the thesis that the essence and the thing are one and the same, since, trivially, if they are identical, then if the essence is given, the thing must be given as well, and if the essence is taken away, then the thing is taken away too. It should be admitted that this reading would seem to make 2d2 empty; but if the thing and its essence are two, what accounts for their necessary connection? One advantage of the identity-thesis is that the reason for this becomes evident at once. (Note that we cannot model the relation between God and His essence as a part/whole relation, since Spinoza is explicit as to the fact that an infinite substance is “indivisibile” (1p13), that is, that it has no parts. Hence, God is not its essence plus something else.)
Another advantage of acknowledging that God and His essence are one and the same, more generally, that an essence is the thing it is an essence of, is that this identification helps to dispel another perplexity some readers might feel while reading the Ethics. Having introduced a whole series of ontological categories in his definitions—the notions of substance, mode, attribute, and essence—Spinoza goes on to say in axiom 1 that whatever is, is either a substance (in se) or a mode (in alio). What happened to the attributes and the essences? Were they not supposed to be “real” after all? On the present interpretation, the explanation of why essences are not mentioned in axiom 1 is that, if an essence is an essence of a substance, then it is that substance itself. Since the same applies to modes (modes must have essences as well), there is no reason to make an explicit reference to essences in axiom 1; they do not constitute an additional, independent category of being besides substances and their modes: they are these substances and modes themselves. (Alternatively, the perplexity could be dispelled by limiting the validity of the axiom to actual beings; limiting the axiom in this way, however, also requires that one acknowledges a distinction between actualities and possibilities.)
6 The Task and Structure of the Proof Clarified
God exists (as a possible being),
God exists (as an actual being).
There is a real definition/true idea of God.
If there is a real definition/true idea of X, then X is real (as a possibility).
Therefore, God exists (in the existential mode of possibility).
One might be puzzled by the fact that Spinoza should begin by assuming the truth of his definition of God. As an answer to this, however, one might quote 2p43, where Spinoza says: “He who has a true idea at the same time knows that he has a true idea, and cannot doubt the truth of the thing.” In other words, d6 stands in need of no justification for someone that really understands it: to ask for a justification would just be to display one’s failure in grasping the idea expressed by the definition. Admittedly, this sounds question-begging; in case of a disagreement, Spinoza’s claim that one who clearly apprehends the true idea of God can’t fail to know that the idea is true would seem to make rational debate about the divine nature impossible. But the situation is not as hopelessly question-begging as it might look at first sight. The opening definitions could be validated in the eyes of the skeptic by the fact that they make it possible to construct a metaphysical view that nicely explains the overall nature of the universe and our place in it; considering that the book’s aims are not merely theoretical, but practical as well (the book is called Ethica after all, and not Metaphysica), the definitions might also be validated by the fact that they lead to what Spinoza took to be a satisfactory answer to a question that greatly troubled him, namely the ancient question “How should I live?” All this, however, is conjectural. The safest thing to say is that Spinoza is not much worried by the epistemological problems posed by his definitions (as also recognized by Nadler 2006: 48).
But I think the most important consideration at this point is that I find within me countless ideas of things which even though they may not exist anywhere outside me still cannot be called nothing; for although in a sense they can be thought at will, they are not my invention but have their true and immutable natures. When, for example, I imagine a triangle, even if perhaps no such figure exists, or has ever existed, anywhere outside my thought, there is still a determinate nature, or essence, or form of the triangle which is immutable and eternal, and not invented by me or dependent on my mind. (Descartes 1986: 45-46)
Mathematical truths are not created, but discovered. But the objectivity of essences and of eternal truths is not recognized as the conclusion of a philosophical argument; rather, it is immediately grasped in the very act of apprehending them.6
7 The Atheist’s Challenge
If some were to say that he had a clear and distinct, that is, true, idea of a substance, and nevertheless doubted whether such a substance existed, that would be indeed the same as if he were to say that he had a true idea, and nevertheless doubted whether it was false.
Spinoza’s response to this kind of challenge will be discussed at the end of this paper, after we have a better hold upon his ontological proof.
8 Can We Turn a Paralogism into a Valid Argument?
God exists/does not exist. Kant argues that “God exist” means “The concept of God is instantiated.” As against this, Spinoza could object that his argument presupposes a very different semantic analysis, since “God exists” means “God exists in the mode of actuality.” (Is this interpretation as bizarre as it looks at first sight? It is more in tune with our linguistic intuitions than Kant’s; it seems natural to say that “God exists” is about God, not about our concept of God.)
Existential statements. Kant argues that existential statements are empty if interpreted as statements about things, because existence is not a property things might or might not possess. As against this, Spinoza could reply (a) that the predicate “existence” is ambiguous between “existence in the mode of possibility” and “existence in the mode of actuality”; (b) that these are not special properties of things, but rather ways of being; (c) that since a thing could be merely possible, it is not empty or uninformative to say that it is actual.
The crucial consideration here is that Spinoza does not think of existence as a yes-or-no matter, while Kant (and modern logic) apparently does. Kant’s critique can be reformulated in Russellian terms as the charge that Spinoza fails to distinguish between two possible meanings of “is,” namely the copula and the existential quantifier. But there is no evidence that Spinoza is guilty of such a mistake. The conclusion to be drawn is that Kant’s critique fails to engage Spinoza’s ontological argument, as this was meant to be understood by Spinoza. Interestingly, Kant’s failure to come to terms with Spinoza’s argument comes clearly to the fore in a remark he makes, almost casually, at the beginning of his discussion of the cosmological proof: “The attempt to extract from a purely arbitrary idea the existence of an object corresponding to it is a quite unnatural procedure and a mere innovation of scholastic sublety” (Kant 1961: 507 [A 604/B 623]). Surely, if there is something Spinoza denies, is that his definition of God expresses a “purely arbitrary idea.” Incidentally, this raises the question whether there is still much justification for referring to Spinoza’s proof as “ontological”; strictly speaking, the relevant logical transition is not from concepts to real things, but from a mode of existence to a mode of existence of a different sort. It might be of interest in this respect to notice that the founder of this genre, Anselm of Canterbury, begins his own proof in the second chapter of his Proslogion with the assumption that something than which nothing greater can be conceived [i.e., God] exists in the intellect. Does that something also exist outside the intellect? (Anselm 1995: 84) This suggests that Spinoza’s argument might be called “ontological” by appealing to tradition rather than to Kant’s definition. But whether, or in what sense, Spinoza’s argument can be called “ontological” is a relatively uninteresting question; it still remains to be ascertained whether it is valid.
God is a substance. (d6)
A substance necessarily exists. (p7)
Therefore, God necessarily exists. (p11)
Undoubtedly, once it is clear that “God” has a reference, this begins to look like a valid reasoning.
9 Spinoza’s Explanatory Rationalism and the Concept of Substance
A substance cannot be produced by anything else...therefore it will be the cause of itself, that is, its essence necessarily involves existence.
Spinoza is here appealing to what, in Leibniz’s terminology, has come to be known as the principle of sufficient reason. The principle is at work throughout the Ethics. In the second scholium to p7, Spinoza writes that “there must be, for each existing thing, a certain cause on account of which it exists”; unexplained facts are simply not an option. Axiom 2 also approximates to an explicit statement: “What cannot be conceived through another, must be conceived through itself.” If “being conceivable” is read as “can be made intelligible,” then Spinoza is ruling out that some things may lack an explanation.
There must be a reason why a substance exists. (principle of sufficient reason)
A substance cannot be produced by another substance (so we can’t appeal to other substances for an explanation). (p6)
Therefore, a substance is a cause of itself (i.e., its existence is self-explanatory). (p7)
There must be a reason why there is an actual substance. (principle of sufficient reason)
A substance cannot be produced by another substance. (p6)
Therefore, an actual substance is the cause of itself (it’s self-explanatory). (p7)
God is a possible substance. (d6)
An actual substance necessarily exists. (p7)
Therefore, God necessarily exists. (p11)
This argument is invalid: since the middle term “substance” occurs with two different meanings in the two premises, the reasoning is vitiated by the fallacy of equivocation. Thus, Spinoza’s ontological argument fails—although for very different reasons than those urged by Kant. As against this, Spinoza might protest that the proposed reading of p7 is too restrictive and that his claim was meant to include possible substances as well. The rejoinder is that the thesis that a possible substance necessarily exists is not entailed by the argument he provides, nor is it self-evidently true.7
God is an actual substance. (d6)
An actual substance necessarily exists. (p7)
Therefore, God necessarily exists. (p11)
This interpretation has the significant advantage that the argument can be reconstructed as formally valid, i.e., it does not charge Spinoza with a fairly obvious logical blunder. It must also be admitted that it does square well with his own formulation of p11.8 Nevertheless, this reading has the major disadvantage that God’s actual existence would be assumed right from the start; there are then no arguments for God’s existence in Spinoza’s philosophy?
A further question needs to be asked before concluding this section. As it has been argued, Spinoza’s commitment to the principle of sufficient reason leads him to demand an explanation of an actual substance’s existence. But what accounts for the existence of a possible substance such as God is supposed to be? Spinoza’s own formulations of the principle of sufficient reason (axiom 2 and p7s) strongly suggest that its validity is to be restricted to actual existents, yet a consistent rationalist would want to push this demand for complete intelligibility to the outmost limit—so, why is there even such a thing as God’s essence? The answer might turn out to be perplexingly simple. It seems natural to think that the only constrain on the realm of the possible is the principle of non-contradiction. Hence, granted the assumption that possible beings are real things, what explains an essence’s existence is that it can be conceived without contradiction. It might not be immediately obvious that this is an answer; we tend to expect something more concrete. But since essences are not actual things, ordinary concepts of causation or of creation do not apply at this ontological level. If we accept this explanation, then we see that Leibniz was only being a consistent explanatory rationalist when he urged that all ontological arguments should be prefaced by a proof that the notion of God is a coherent one.
...the very nature of a square circle indicates the reason why it does not exist, namely because it involves a contradiction (rationem, cur circulus quadratus non existat, ipsa ejus natura indicat). On the other hand, the reason why a substance exists also follows from its nature alone, because it involves existence. But the reason why a circle or triangle exists, or why it does not exist, does not follow from the nature of these things... (p11d)
Spinoza distinguishes here between: (a) essences that necessitate the actual existence of the things they are the essences of, such as the essence of God; (b) essences that necessarily preclude the actual existence of the things they are the essences of, such as the essence of a square-circle; (c) essences that neither necessitate nor preclude the existence of the things they are the essences of, such as the essence of a triangle. Category (b) is a puzzling one: Spinoza seems to be admitting the reality of self-contradictory essences. This generates a Quinean problem: apparently, the realm of essences is larger than the realm of possibilia—but how could there be self-contradictory beings?9 This is a “Meinongian” self-refuting position and it is surprising that Spinoza should have expressed himself in this way.
10 Spinoza’s Refutation of Atheism
We are now in a position to deal with an exegetical problem that has been left untouched so far. The standardized form of the ontological argument that has been discussed in the previous sections is not explicitly formulated by Spinoza in the Ethics, although it is obviously there. Instead, Spinoza gives his explicit argument for p11 the form of a reductio ad absurdum. Some critics have wondered why this is so. In this connection, D. Garrett writes: “The form of the argument [for p11] is dictated by two considerations: Spinoza’s expressed preference for reductio arguments, and his desire to utilize both of the alternative definitions of ‘cause of itself’ given in d1.” (Garrett 2001: 8) But there is an alternative explanation: Spinoza’s choice might be dictated by the polemical need to reject the atheist’s claim, identified in section 7, that God’s non-existence can be distinctly conceived.
If you deny this [the conclusion of the ontological argument, i.e. that God necessarily exists], conceive, if you can, that God does not exist. Therefore (by a7) his essence does not involve existence. But this (by p7) is absurd. Therefore, God necessarily exists.
a7: If a thing can be conceived as not existing, its essence does not involve existence
The notion of conceivability is a notoriously slippery one and Spinoza does not say much to make it clear. If the Cartesian interpretation of Spinoza’s theory of the mind that has been adopted in this paper is correct, it is to be understood in terms of a subject’s ability to form clear and distinct ideas. On this theory, if we could form a clear and distinct idea of God as non-existent, we would know that God does not exist. On any analysis of the concept of knowledge, a person wouldn’t know that p, if p weren’t true. Hence, trivially, if someone could conceive that God does not exist, then God would not exist. Spinoza must therefore deny that the atheist can have what he claims to have—a clear grasp of his definition and at the same time a transparent idea of God’s non-existence.
God is a substance. (d6, premise)
I can form a clear and distinct idea of God as non-existent. (hp to be rejected)
If a thing can be conceived as non-existing, then it does not necessarily exist. (axiom 7, premise)
God does not necessarily exist. (2,3, MP)
What does not necessarily exist is not a substance. (converse of p7, premise)
God is not a substance. (4,5, MP)
Therefore, God is and is not a substance. (1,6, &).
Since Spinoza obviously takes all other premises either to be self-evident (d6, a7) or to have been previously proved (p7), he concludes that thesis (2) must be rejected; it is therefore inconceivable that God should not exist. It could be said that this reconstruction is not faithful to the text and that the premise to be rejected should not be the above proposition (2), but simply “God does not exist.” Contrary to what some have argued (Doney 1980: 41-42), however, this is not the most natural reading of the text. Spinoza is not asking us simply to assume that God does not exist, but to conceive of his non-existence (si negas, concipe, si fieri potest, Deum non existere). In the context of a theory of clear and distinct ideas, this is a more demanding task than the mere entertaining of a hypothesis.
It is inconceivable that God should not exist. (rejected hp)
What is inconceivable is logically impossible. (suppressed premise)
It is logically impossible that God should not exist. (1,2)
If a statement is logically impossible, then its negation is necessary. (suppressed premise)
Therefore, God necessarily exists. (3,4)
This line of reasoning is perfectly valid, but it has the disadvantage of having to ascribe to Spinoza a suppressed premise, (2), that he does not explicitly endorse in the Ethics. In ascribing this premise to Spinoza, however, no injustice is done to his ability as a philosopher since, if correctly understood, it is neither implausible nor controversial. Considering that “conceivability” is to be explicated in terms of clear and distinct ideas, the premise does not state that human thought can set a limit to logic. This would indeed be a difficult position to defend, as there are obviously many things human beings might fail to conceive and which are nonetheless possible. What the suppressed premise says is that the limitations of human thought—when it satisfies the requirement of an ideal, perfected rationality—express genuine logical impossibilities. And this is a quite different claim.10 (If there is something controversial here, this is not the premise, but the whole theory of clear and distinct ideas.)
But is the proposed integration really necessary? On the basis of the foregoing analyses, it seems reasonable to maintain that Spinoza’s main argument for p11 is the more straightforward one that takes d6 and p7 as its premises. The argument now under discussion need not be viewed as Spinoza’s main proof for God’s necessary existence, but rather as a way of reinforcing a conclusion already achieved by rebutting the atheist’s alternative suggestion that it is possible to have a clear and distinct conception of God’s non-existence. What the reductio really shows is that claiming to be able to conceive God’s non-existence is claiming to be able to conceive that God is and is not a substance—which is absurd. Indeed, a philosopher who would advance this objection would be, in Spinoza’s own words, like someone who claims “to have a true idea [i.e., an idea of God as a substance, and therefore as necessarily existent], and nevertheless doubted whether it was false [i.e., claimed to be able to conceive that God might not exist, hence that He is not a substance].”
God is a possible substance. (d6, premise)
I can form a clear and distinct idea of God as non-existent. (hp to be rejected)
If a thing can be conceived as non-existing, then it does not necessarily exist. (axiom 7, premise)
God does not necessarily exist. (2,3, MP)
What does not necessarily exist is not an actual substance. (converse of p7, premise)
God is not an actual substance. (4,5, MP)
Therefore, God is a possible substance but not an actual substance. (1,6, &)
The conclusion of this argument, proposition (7), is not a contradiction anymore: it is just a statement of the atheist’s position.
One final remark: Spinoza’s second demonstration for p11, the last of his a priori proofs, is also best understood as a reply to the skeptic rather than as an independent argument for that thesis. Assume that God does not exist. Consistently with Spinoza’s explanatory rationalism, there must be a reason why God does not exist; this cannot be God’s essence, for this is not incoherent like that of a square-circle. The reason of God’s non-existence must therefore be searched in another substance. But another substance, if there were one, would have to be radically different from God, for according to p5 no two substances share a common attribute. Hence, since substances with no common attributes have nothing in common (this is Spinoza’s p2) and cannot therefore act on one another, such a hypothetical substance would be unable to prevent God from achieving actualization. Since there is no internal or external reason that could prevent God from existing, Spinoza concludes that God necessarily exists.
This argument works only on the assumption that God’s essence is not incoherent, an important point that is left unproved. Moreover, the step from the mere absence of an obstacle to God’s actualization is problematic. By itself, the fact that nothing prevents an essence from being actualized does not provide a sufficient reason for concluding that the essence is actualized; some positive ground is needed in order to draw this conclusion. The argument becomes intelligible, however, if it is interpreted as a means to dispel residual skeptical doubts. Granted that we already have a proof of God’s necessary existence, why should we hesitate in accepting this conclusion? If nothing could prevent God from coming into existence and if God’s nature as a substance suffices to explain his existence, then there is really no reason to deny that God exists.
In this paper, we have analyzed Spinoza’s ontological argument and defended it from Kant’s general condemnation of this kind of reasoning by showing that the two philosophers work with different ontological categories. This leads them to interpret the argument in radically different ways, which generates a problem of incommensurability. Kantians will have to provide different objections than those of the first Critique to show what is wrong with Spinoza’s reasoning. Does this mean that Spinoza has succeeded in providing a plausible argument after all? A closer examination of Spinoza’s proof shows that it is loaded with controversial ontological and epistemological assumptions; most importantly, it can be rejected as invalid even if these assumptions are granted. The overall philosophical moral to be drawn is that one needs to understand the metaphysics to get the semantic right; to think otherwise is to be cut off from understanding—and from criticizing fairly—the great masters of the past.
All references to the Ethics (Spinoza 1966) adopt the standard conventions: “p” stands for “proposition,” “s” for “scholium,” “d” for “demonstration.” Thus, 2p11d refers to the demonstration of the eleventh proposition in Book II; for the sake of simplicity, the book’s number is not specified in case of Book I.
The peculiarity of Spinoza’s version of the ontological argument is emphasized by J. Bennett 1984: 72.
The term “pantheism” is here used for the sake of convenience; the questions whether and in what sense Spinoza’s infinite substance deserves to be called “God” and, if yes, whether his metaphysics is a form of pantheism lie outside the range of the present paper.
“quaeso, mi amice, ut consideres homines non creari; sed tantùm generari, & quòd eorum corpora jam antea existebant, quamvis alio modo formata” (Spinoza 1925b: 14).
It could be objected that there is another class of ideas in Spinoza’s Ethics that has a claim to be identified with Cartesian clear and distinct ideas, namely what Spinoza calls “adequate ideas.” An adequate idea is such as to possess all the marks of true ones, but all internally: “By adequate idea I understand an idea which, insofar as it is considered in itself, without relation to an object, has all the properties, or intrinsic denominations of a true idea” (2d4; my emphasis). As Spinoza also explains immediately after having provided the definition, talk of “intrinsic denominations” is meant to exclude “the agreement of the idea with its object.” This is tantamount to saying that adequate ideas do not refer to their ideata, whereas true ideas do. Two questions need to be distinguished here: Should Cartesian clear and distinct ideas be identified with Spinoza’s true ideas or with adequate ideas? Is the idea conveyed by d6 adequate or true? The correspondence with Oldenburg suggests that, at least in the eyes of Spinoza, a clear and distinct idea is identical with a true one. What truly matters here, however, is to recognize that, for Spinoza, d6 expresses God’s true idea.
Spinoza writes at one point that p7 is obviously true: “if men would attend to the nature of substance, they would have no doubt at all of the truth of p7. Indeed, this proposition would be an axiom for everyone, and would be numbered among the common notions” (p7s2).
This reading is provided by Marcus 1993: 173: “What is asserted in Proposition XI is that he exists in a certain way—necessarily. His existence simpliciter is never in question.”
This worry is expressed in Doney 1980: 42: “The principle… that what is not conceivable is not possible, seems to me to be by no means an unexceptional, or trivial or obvious principle. Interpreted in certain ways, it is a substantive principle which can be taken to have rather far-reaching and disturbing consequences.” If “conceivable” is interpreted in the right way, that is, by reference to clear and distinct ideas and therefore rationality per se, the principle is unproblematic.
A previous version of this paper was read at the ETH, Zürich, in November 2009; I am indebted to the participants and especially to Prof. Michael Hampe, for the critical discussion and for important suggestions, and to Pauline Phemister for comments on an earlier draft.
- Anselm (1995) Proslogion. froommann-holzboog, Stuttgart-Bad CannstattGoogle Scholar
- Bennett J (1984) A Study of Spinoza’s Ethics. Hackett, IndianapolisGoogle Scholar
- Delahunty R.J. (1985) Spinoza. Routledge & Keegan Paul, LondonGoogle Scholar
- Descartes R. (1986), Meditations on First Philosophy, trans. by J. Cottingham. Cambridge University Press, CambridgeGoogle Scholar
- Devaux, S. (2007) The Role of God in Spinoza’s Metaphysics, Continuum, London.Google Scholar
- Doney W (1980) Spinoza’s Ontological Proof. In: Kennigton R (ed) The Philosophy of Baruch Spinoza. The Catholic University of America Press, Washington, D.C.Google Scholar
- Garrett D. (2001) Spinoza’s “Ontological” Argument. In: Loyd G. (ed) Spinoza. Critical Assessments, Vol. II. Routledge, London and New YorkGoogle Scholar
- Hartshorne C (1962). Anselm’s Discovery. Open Court, La Salle, Ill.Google Scholar
- Kant I. (1961) Critique of Pure Reason, trans. by N.K. Smith. MacMilllan, LondonGoogle Scholar
- Marcus R.B. (1993) Spinoza and the Ontological Proof. In: Modalities: Philosophical Essays. Oxford University Press, Oxford.Google Scholar
- Nadler, S. (2006) Spinoza’s Ethics: An Introduction. Cambridge University Press, Cambridge.Google Scholar
- Plantinga A (1974) The Nature of Necessity. Oxford University Press, Oxford, 1974.Google Scholar
- Quine W. (1953) On what there is. In: From a Logical Point of View. Harvard University Press, Cambridge and LondonGoogle Scholar
- Spinoza B (1925a) Cogitata Metaphysica. In: Gebhardt C (ed) Opera, Vol. I. Carle Winters, Heidelberg.Google Scholar
- Spinoza B (1925b) Epistolae. In: Gebhardt C (ed) Opera, Vol. IV. Carle Winters, Heidelberg.Google Scholar
- Spinoza B. (1925c) Tractatus de intellectus emendatione. In: Gebhardt C (ed) Opera, Vol. II, Carl Winters, Heidelberg.Google Scholar
- Spinoza B (1966). Ethics, translated by E. Curley. Penguin, LondonGoogle Scholar
- Sprigge T.L.S (2007) The God of Metaphysics. Clarendon Press, Oxford.Google Scholar