In his criticism of the so-called ‘ontological argument’, Kant develops in an original way the main objections that have been raised against this proof of God’s existence throughout its history. Among these objections, Kant considers that which is based on ‘a precise determination of the concept of existence’ (A598/B626)Footnote 1 as the most definitive. Certainly, if existence is not a perfection or reality containable within the concept of a thing, as Kant famously claims, then from the concept of a being involving all perfections or realities, i.e. from the concept of the ens realissimum, there is no way to infer the existence of such a being. The ‘proof of the existence of a highest being from concepts’ (A602/B630), as Kant describes the ontological argument, is therefore completely annihilated.
As Leibniz, a firm supporter of this proof, explained in a letter to Eckhardt, if it is assumed that existence is a perfection, then existence increases reality in such a way that ‘when A is thought of as existing, more reality is thought of than when A is conceived as possible’.Footnote 2 Consequently, the opposite thesis, namely, that existence is not a reality containable within the concept of a thing, could also be stated by Kant as follows: ‘The actual contains nothing more than the merely possible’ (A599/B627).
To illustrate this new version of his negative thesis on existence, Kant initially used some comparisons which he did not use again later. In a handwritten note, most likely drafted prior to the publication of his 1763 The Only Possible Argument, Kant wrote: ‘[N]o more is posited in God’s goodness, insofar as it is considered as existing, than in the goodness which lies in His possibility’ (Refl 3706, 17: 241). In the mentioned pre-critical work, Kant used, instead, the example of Julius Caesar’s individual concept to explain this very conception of existence. Thus he wrote, ‘Draw up a list of all the predicates which may be thought to belong to him, not excepting even those of space and time. You will quickly see that he can either exist with all these determinations, or not exist at all’ (BDG, 2: 72).
However, as is well known, in the Critique of Pure Reason Kant changed these examples for the celebrated comparison between a hundred actual thalers and a hundred possible ones: ‘A hundred actual thalers do not contain the least bit more than a hundred possible ones’ (A599/B627).Footnote 3 In his 1763 essay on the concept of negative magnitudes, Kant had already used the example of ‘a hundred thalers’, as a sum of money of which one is a creditor or a debtor, to explain the concept of negative magnitudes (see NG, 2: 172). It is easy to understand that, given this particular context and for this specific purpose, the quantitative determination of the thalers was essential. But in the face of Kant’s comparison in the first Critique between a hundred actual and possible thalers, the reader may wonder with some surprise: why exactly a hundred thalers, whether actual or possible? Would not the example work if we compared a single actual with a single possible thaler, in the same way that Kant compared God’s sole actual goodness with God’s sole possible goodness, or the individual concept of Julius Caesar with the existence of this unique historical figure?
It is perhaps less known that Kant seemingly took the example of the hundred thalers as a basis for a criticism of the ontological proof from Johann Bering’s booklet Examination of the Proof for the Existence of God from the Concepts of a Most Perfect and Necessary Being (Bering Reference Bering1780; hereafter, PdB).Footnote 4 Bering (1748–1825), professor of philosophy in Marburg, who would eventually became a disciple of Kant, published his short book (88 pages) one year before the appearance of the Critique of Pure Reason. His investigation puts forward an examination of the reasons for and against the traditionally called argumentum a priori, as well as a refutation of its claims to validity. When referring to his comparison between a hundred actual and possible thalers – in which the quantitative determination of the thalers is essential – Bering writes, ‘This example should, I hope, put beyond doubt what I said earlier, existence itself is not a reality, and therefore no other relationship takes place between what is possible and what is actual than that between nothing and something’ (PdB,79–80). Although there is no indisputable direct evidence that Kant ever read Bering’s treatise, it is not surprising that Dieter Henrich, in his book on the ontological argument in the modern age, has written, in commenting on the last part of Bering’s work:
This section of Bering’s writing deserves a place of honour in the historiography of philosophy. For in it the line of argument of the Critique of Pure Reason against Anselm’s conclusion is completely and clearly preformed. The agreement is so great that one can claim with certainty that Kant knew and appreciated this writing, which he never quoted. (Henrich Reference Henrich1967: 119)Footnote 5
A reading of Bering’s opuscule shows that, although it is reasonable to assume that his example of the thalers inspired Kant, the arguments in which both philosophers use the same example are vastly different. Consequently, I intend to carry out three tasks. The first is to expound on the similarities and disparities between Bering’s and Kant’s comparisons of the actual and possible thalers, to show Kant’s originality in the use of the example.Footnote 6 The second task is to assess Bering’s and Kant’s reasons for employing the case as a refutation of the ontological argument. The third task is to point out that the debate regarding the validity of this employment requires a discussion of the metaphysical problem of universals. I will conclude by suggesting that Kant’s (and Bering’s) position on this problem decisively determines the value and scope of his critique of the ontological argument.
1. Comparative Analysis of Bering’s and Kant’s Versions of the Example of the Hundred Thalers
In order to show that the actual does not contain more than the possible, ‘Bering was the first’, as Henrich claims, ‘to make use of the famous example of the hundred thalers’ (Henrich Reference Henrich1967: 120). As Bering puts it:
If existence is a reality, then a thing that exists has only one more reality than can be found in the very same thing insofar as it is just possible. In this way, in a hundred actual thalers there would be one more reality than in a hundred possible thalers, and from this it would follow further that a hundred possible thalers would be just as good as ninety-nine actual ones, since one would have to take the extra thaler there for actuality. But whoever has the choice between the two, which one will he take? (PdB,79)
The reasoning can be stated in the following form:
Conditional syllogism (of the figure tollendo tollens):
It is clear that, in Bering’s syllogism, the quantitative determination is essential. The immediate aim is indeed to show that comparison between a given number of possible and a smaller number of actual thalers leads to an absurdity: taking a possible thaler, ‘the extra thaler there’, for actuality or existence.
Certainly, the claim that existence is a reality – and consequently, that an existing thing has at least one more reality than a merely possible thing – invites the comparison in question, in order to ‘calculate’, so to speak, their respective quantity of realities. Bering illustrates this ‘comparative calculus’ with a noteworthy example: ‘a hundred possible thalers would be just as good (eben so gut) as ninety-nine actual’. With this example, Bering seems to mean, first, that, regarding their number of perfections or realities, a hundred possible thalers is equivalent to ninety-nine actual ones. Second, that, if the quantity of realities or perfections is equivalent, then a hundred possible thalers would be as useful or as ‘perfect’ as ninety-nine actual thalers. Third, and lastly, that thus no one would have any reason to prefer the ninety-nine actual thalers to the hundred possible ones.
Obviously, this is not the case at all. Bering’s rhetorical question is proof enough: ‘whoever has the choice between the two, which one will he take?’ The actual thalers, of course. This shows that the comparison between quantities of both ‘kinds’ of thalers is absurd. It contravenes an evident arithmetical law: heterogeneous quantities cannot be compared. It is instructive to quote at length the passage in which Bering explicitly states this reason:
When I relate realities or perfections and compare them with each other, they have to be homogeneous or of the same kind. Indeed, I can affirm that a subject that has ten realities is more perfect than a subject that has only five of them, whether I consider both metaphysically or in abstracto, or as actually existing subjects. But this comparison disappears as soon as I treat one in abstracto, as something whose actuality does not matter to me at all, and I consider, instead, the other as existing. I cannot say that in five actual realities there is less perfection than in ten, if these are only possible. Everything possible, as soon as it is put into relation with the actual, is a nothing, and therefore in this case no other comparison takes place than that between nothing and something. (PdB, 74)
Thus Bering’s justification of the falsehood, or better, of the absurd character of the consequent of his conditional syllogism can be formulated as follows:
Categorical syllogism (of the first figure):
Bering’s criticism of the ontological argument recalls that of Gassendi. Without using the appeal to quantities, in his objections to Descartes’ ontological proof, Gassendi wrote that ‘It is quite right for you to compare essence with essence, but instead of going on to compare existence with existence or a property with a property, you compare existence with a property’; and he adds, ‘if a thing lacks existence, we do not say it is imperfect, or deprived of a perfection, but say instead that is nothing at all’.Footnote 7
Bering goes one step further and explicitly points out that the truth disclosed by the example of the thalers is a universal principle not admitting any exception. Since the actual contains nothing more than the possible, existence is not a perfection in the case of the concept of the ens realissimum either:
The kind of existence cannot change anything here. Whether it is called necessary or contingent, it does not make any difference in this general principle, for just as necessary existence is based on the most supreme extant (vorhandene) perfections from eternity to eternity, so contingent existence is precisely contingent because the realities required for it are contingent. (PdB, 80)
The reasoning justifying the application of the negative thesis to the case of God can be formulated (and completed) as follows:
Corollary: Since existence is not a reality containable within the concept of a thing, neither are the modalities of existence, namely, contingent existence and necessary existence. Thus, from the concept of a being containing limited perfections it is only inferred that, if it exists, it exists contingently. Similarly, from the concept of the ens realissimum it is only inferred that, if it exists, it necessarily exists.
In addition, Bering teaches that speaking of existence in abstracto can give rise to the error in question. But the misunderstanding disappears if, instead of the abstract expression, we refer to the existing being in concreto: ‘So I would say, instead of the necessary existence of God, rather the necessarily existing Deity, or the necessarily existing divine perfections and properties’ (PdB, 81). Obviously, that there is a Deity or an ens realissimum is precisely what must be proved. This will certainly never be achieved from the mere concept of the ‘divine perfections and properties’, among which one cannot find the equivocally called ‘necessary existence’.
Without quoting Bering, in his main work and in some university lectures (see V-Phil-Th/Pölitz, 28: 1028; V-Th/Volckmann, 28: 1177), Kant uses the same example with the same aim. However, by contrast with Bering, he puts forward the following comparison:
Thus the actual contains nothing more than the merely possible. A hundred actual thalers do not contain the least bit more than a hundred possible ones. For since the latter signifies the concept and the former its object and its positing in itself, then, in case the former contained more than the latter, my concept would not express the entire object and thus would not be the suitable concept of it. (A599/B627)
The argument can be structured in this form:
Conditional syllogism (of the figure tollendo tollens):
It goes without saying that Kant’s comparison between the content of the actual thalers and that of the possible ones does not require any determination of the number of thalers. The comparison would also work in the case of a single actual in regard to a single possible thaler. This fact alone would suffice to clearly distinguish Bering’s reasoning from Kant’s. Moreover, Kant’s argument differs from Bering’s (and even from Gassendi’s) in that the reason supporting the falsehood or absurdity of the consequent of the conditional syllogism is not the heterogeneity between the actual and the possible, i.e. between existence and essence, or between a something and a nothing. Rather, Kant’s argument takes the heterogeneity between the actual and the possible for granted, and it goes beyond this obvious fact. Certainly, the actual and the possible, however heterogeneous they may be, must be in a certain relationship of mutual correspondence. They must have, therefore, something in common. What they have in common can only be their respective ‘content’: ‘A hundred actual thalers do not contain the least bit more than a hundred possible ones.’ If, to show the truth of this claim, Bering looks at what separates the actual from the possible, Kant, by contrast, focuses above all on what unites them.
Kant’s originality in the use of the example is fully appreciated when one remembers the way in which he conceives the nature of existential judgements. Strictly speaking, an existential judgement is not the representation of a relation of concepts to each other, but rather the cognition of an object.Footnote 8 To affirm the existence of something is tantamount, so to speak, to putting the possible in relation to the actual. The possible, Kant says in the text quoted below, ‘signifies the concept’. Kant understands the term ‘possible’ in two senses, as the logically possible and as the really possible. Logical possibility refers to the thinkability or conceivability of the concept, and only requires the absence of contradiction among the marks of the concept (see Bxxvi, note). Real possibility, instead, refers to the cognizability of the object thought in the concept, and requires, in addition to logical possibility, the agreement of the object ‘with the formal conditions of experience’ (A218/B265). The cognizability of an object is also called the ‘objective reality’ (objektive Realität) of the concept, i.e. the relation of the concept to an object (see A109, A155/B194). Therefore, that the possible ‘signifies the concept’ means both that the concept is not contradictory in itself and that its object is somehow cognizable. The actual, by contrast, signifies ‘its object’, i.e. the set of properties pertaining to the thing in reality, outside the concept. Certainly, actuality or existence requires, in addition to real possibility, the connection of the object ‘with the material conditions of experience’, i.e. with sensation (A218/B266). Consequently, Kant completes his claim by adding the object ‘and its positing in itself’. Thus, to cognize or to judge something as actual or existing is not, according to Kant, to take the subject concept ‘together with all [its] predicates’ and to add a ‘new predicate to the concept’. To judge something as existing is, as Kant himself writes, to ‘posit (setzen) the object in relation to my concept’. It is not, then, to posit a concept relatively to another concept as its mark, but to ‘posit the subject in itself with all its predicates’, to posit it absolutely, or – to avoid a possible misunderstanding – to posit what is perceived in relation to what is thought. Being, as Kant famously claims, ‘is merely the positing of a thing or of certain determinations in themselves’ (A599/B627).
It is noteworthy that, for positing the object in relation to the concept, object and concept must have the same ‘content’. Both object and concept, Kant expressly recognizes, ‘must contain exactly the same’ (A599/B627). Certainly, the word ‘content’ does not have the same meaning when referring to the concept and when applied to the object. Using a scholastic distinction also adopted by Descartes, we can say that the ‘content of a concept’ is its objective reality (realitas objectiva, or modus essendi objectivus). However, ‘objective reality’ in this sense is of course not to be understood in the Kantian sense mentioned above, but as the ‘reality’ or set of ‘perfections’ that the object has insofar as it is an ‘object of thought’, an object merely represented in the concept. Analogously, we can say that the ‘content of the object’ is its actual or formal reality (realitas actualis sive formalis, or modus essendi formalis), i.e. the ‘reality’ corresponding to the object insofar as it is ‘an existing thing’, the reality pertaining to the object in virtue of its ‘actuality’. More recently, Alvin Plantinga, in a celebrated essay, claims that what Kant means by the ‘content of a concept’ is perhaps ‘the set of properties a thing must have to fall under or be an instance of that concept’; by contrast, the ‘content of an object’ is ‘the set of properties that object has’ (Plantinga Reference Plantinga1966: 539). We truly cognize when the content of our concept (i.e. its objective reality, or the set of properties a thing must have to be an instance of a certain concept) coincides with the content of the object (i.e. its formal reality, or the set of properties that the object possesses in itself). Thus, concept and object must have exactly the same ‘content’. Otherwise, as Kant claims in the above passage, ‘my concept would not express the entire object and thus would not be the suitable concept of it’ (A599/B627).
The conclusion of Kant’s conditional syllogism states, then, that affirming that a hundred actual thalers have one more reality than a hundred possible ones is false, because this assertion would be tantamount to the absurdity that a concept could never express its entire object and, therefore, that no true cognition is possible at all:
Thus when I think a thing, through whichever and however many predicates I like (even in its thoroughgoing determination), not the least bit gets added to the thing when I posit in addition that this thing is. For otherwise what would exist would not be the same as what I had thought in my concept, but more than that, and I could not say that the very object of my concept exists. (A600/B628)
In this light, the reason supporting this conclusion or, put in equivalent terms, the reason for claiming that being or existence is not a reality, ‘i.e. a concept of something that could add to the concept of a thing’ (A598/B626), can be set forth as follows:
disjunctive syllogism (of the figure tollendo ponens):
The minor premise is expressly stated at several points in the Critique of Pure Reason. Neither in the logical possibility of the concept nor in its real possibility can we find existence:
In the mere concept of a thing no characteristic of its existence can be encountered at all. For even if this concept is so complete that it lacks nothing required for thinking of a thing with all of its inner determinations, still existence has nothing in the least to do with all of this, but only with the question of whether such a thing is given to us in such a way that the perception of it could in any case precede the concept. For that the concept precedes the perception signifies its mere possibility; but perception, which yields the material for the concept, is the sole characteristic of actuality. (A225/B272–3; see also A219/B266, A599/B627, A639/B667)
Thus to cognize the existence of an object, it is necessary to go beyond the concept and verify that the object is given absolutely, independently of thought. It is not strange, then, that this conception of existence supposes a radical criticism of the proof that seeks to cognize the existence of the highest being from the sole concept of such a being:
Thus whatever and however much our concept of an object may contain, we have to go out beyond it in order to provide it with existence. With objects of sense this happens through the connection with some perception of mine in accordance with empirical laws; but for objects of pure thinking there is no means whatever for cognizing their existence, because it would have to be cognized entirely a priori, but our consciousness of all existence (whether immediately through perception or through inferences connecting something with perception) belongs entirely and without exception to the unity of experience, and though an existence outside this field cannot be declared absolutely impossible, it is a presupposition that we cannot justify through anything. (A601/B629)
The objection against the ontological argument can therefore be formulated as follows:
Corollary: Since existence is not a reality containable within the concept of a thing, we have to go beyond the concept in order to ascribe existence to its object. In the case of empirical concepts, the contingent existence of their objects can only be cognized in a possible experience. In the case of the pure concept of the ens realissimum, the necessary existence of its object cannot be cognized in any way.
Let us conclude the comparison between Bering’s and Kant’s versions of the example of the hundred thalers by simply pointing out an apparent coincidence. Both thinkers agree on the obvious answer to an unavoidable question. If the actual thalers and the possible thalers are not distinguished in their ‘content’, what then is the difference between them? The answer is, of course, that the actual thalers are ‘coins jangling in our pockets’, so to speak, while the possible thalers are not. Bering writes, indeed, that there is a ‘tangible difference’ between actual and possible thalers: ‘no one gives me anything’ in exchange for a hundred thalers ‘that I do not have but can possess’ (PdB, 80). Kant, in turn, famously responds that ‘in my financial condition there is more with a hundred actual thalers than with the mere concept of them (i.e. their possibility)’ (A599/B627).
However, Bering and Kant’s discovery of the truths involved in the example of the hundred thalers cannot circumvent a thorny question. Is the case really adequate for rebutting the ontological proof?
2. Assessment of Bering’s and Kant’s Examples as a Basis for an Objection to the Ontological Proof
The issue under discussion requires comparing the concept of a hundred thalers with that of a supremely perfect being. To properly contrast them, some elementary remarks concerning concepts are needed.
Kant teaches that, in every concept – taken as that which is conceived in the act of understanding – matter and form must be distinguished. ‘The matter of concepts is the object, their form, universality’ (Log, 9: 91). By its form, a concept is then, as Kant expressly claims, ‘a universal representation, or a representation of what is common to several objects’. Thus all concepts, without exception, are universal, or general (allgemein), to the point that Kant claims that ‘[i]t is a mere tautology to speak of universal or common concepts’.
Presumably, the concept of a hundred thalers and the concept of an ens realissimum cannot be distinguished as regards their universality or generality. Both are universal representations produced by the understanding. However, following the traditional doctrine, Kant teaches that, due to its universality, every concept is predicable of a multitude of representations and, ultimately, of a multitude of things. ‘The more the things that can be represented through a concept’, Kant claims, ‘the greater is its sphere’ (Log, 9: 96), i.e. the greater is its extension, on which the universality of the concept rests (see Log, 9: 95). It is easy to recognize the universality or generality of the concept of a hundred thalers. A multitude of a hundred thalers, whether possible or actual, can be represented through it. However, through the concept of a highest being, only a unique being can be represented. Is this concept really universal?
Kant expressly recognizes that ‘the concept of an ens realissimum is the concept of an individual being, because of all possible opposed predicates, one, namely that which belongs absolutely to being, is encountered in its determination’ (A576/B604). Nevertheless, he also admits that the representation of such a being, as a concept, is necessarily universal. ‘[O]nly in this one single case is an – in itself universal – concept of one thing thoroughly determined through itself, and cognized as the representation of an individual’ (A576/B604). Here it is not necessary to deal with the reasons given by Kant as proof of this statement (see A571/B599ff.).Footnote 9 It is enough to note that, according to Kant, the concept of a supremely real being, despite the fact that through it a unique being is represented, is as universal or general, by its form, as the concept of a hundred thalers. Consequently, both concepts only differ with respect to their matter.
Certainly, in the concept of a hundred thalers we think of a multitude of beings that differ only numerically (solo numero), that are endowed with certain limited perfections, and whose non-existence is perfectly thinkable. In the concept of an ens realissimum, by contrast, we conceive of a unique and unrepeatable being provided with all perfections to the maximum degree and whose non-existence is supposed to be not thinkable at all. The ontological argument relies precisely on the peculiarity of the (matter of the) concept of such a being, as Kant himself recognized:
[Y]ou challenge me with one case that you set up as a proof through the fact that there is one and indeed only this one concept where the non-being or the cancelling of its object is contradictory within itself, and this is the concept of a most real being. (A596/B624)
Thus, as regards a critique of the ontological proof using the example of the hundred thalers, either in Bering’s or in Kant’s version, do we not have a kind of ignoratio elenchi?
It seems difficult for a supporter of the ontological argument not to answer this question in the affirmative. Remember the way in which Aristotle describes this fallacy: ‘Those [paralogisms] that depend upon whether something is said in a certain respect only or said without qualification occur because the affirmation and the denial are not concerned with the same point.’ And again: ‘Those that depend on whether an expression is used absolutely or in a certain respect and not strictly, occur whenever an expression used in a particular sense is taken as though it were used absolutely.’Footnote 10 The criticism based on the example of the thalers seems indeed to be a fallacy of secundum quid ad simpliciter: from the claim that existence is not a reality in the case of limited and contingent beings (‘something said in a certain respect’) it is concluded that existence is not a reality in any case, including the case of the ens realissimum (‘something said without qualification’). What is valid for beings of a certain kind, no matter how numerous they may be, is not simply valid for the absolute being as such. Let us then examine Bering’s and Kant’s arguments, understood as a basis for a rebuttal of the ontological proof, from a merely logical point of view.
As we have seen, Bering’s and Kant’s respective conditional syllogisms, both illustrated by the case of the hundred thalers, try to oppose the ontological proof by demonstrating that existence is not a reality that can be added to the concept of a thing. Bering’s main reason is that a merely possible thing and an actual one are heterogeneous things and, therefore, cannot be compared to each other.
But the ontological argument is nothing more than the statement that this evident truth cannot be applied to the case of the concept of God. In fact, the proof can be rephrased as the discovery of the logical impossibility of thinking of the object ‘God’ and at the same time thinking of this object as ‘a merely possible thing’. It is a flat contradiction, Anselm puts in his Proslogion, to think of God as ‘something than which nothing greater can be thought’ and to think that the object of this concept exists only in thought, but not in reality; or to think that this ‘something than which nothing greater can be thought’ is something which can be thought not to exist in reality.Footnote 11 A being which only exists in thought and not in reality, or a being whose non-existence can be thought of, is in no way ‘something than which nothing greater can be thought’. No wonder Descartes could sum up his Fifth Meditation proof of God’s existence in this formula in a letter to Mersenne: ‘[I]t is almost the same thing to conceive of God and to conceive that he exists.’Footnote 12
Faced with the exceptionality of the case of the concept of God, in which we have to consider existence as really identical to essence, to start the rebuttal by simply affirming, as Bering does, that essence and existence are heterogeneous things is either an ignoratio elenchi or, not much better, begging the question. The same logical errors appear again in Bering’s ‘corollary’. If a criticism of the ontological argument is intended, it is necessary to prove, without neglecting or taking it for granted, that we can really think at the same time, with no contradiction, of ‘the most supreme extant (vorhandene) perfections from eternity to eternity’ and their non-existence in reality.
Kant’s argument does not seem to be sounder.Footnote 13 The same error of the ignoratio elenchi, albeit in a different way than in Bering, is barely more hidden. Recall the main reason in Kant’s disjunctive syllogism for the conclusion that existence is not a real predicate: object and concept must have the same ‘content’. What does ‘concept’ mean, however?
It can be said, without further ado, that a concept is the complex of marks or predicates that can be attributed to an object. However, Kant famously distinguishes two kinds of predicates: logical and real. Unfortunately, this apparently clear distinction contains an ambiguity as presented in the Critique of Pure Reason. This ambiguity has given rise to an interpretation that claims that the distinction is not mutually exclusive.Footnote 14 However, as Kant usually opposes ‘logical’ to ‘real’ as an exhaustive dichotomy (see, for instance, NG, 2: 171, 202; MSI, 2: 393; A299/B355; ÜE, 8: 198), the distinction between logical and real predicates can be presented as an exclusive distinction as follows.
Kant distinguishes, first, between predicates from whose content abstraction has been made and predicates from whose content abstraction has not been made. The former deserve the name ‘logical’ because ‘logic abstracts from every content’ (A598/B626). The latter, by contrast, can be called ‘real’. However, to avoid possible confusion, let us call the former ‘formal’ predicates and the latter ‘material’. Among ‘material’ predicates, Kant further distinguishes ‘analytical’ from ‘synthetic’ ones (see ÜE, 8: 223, 230, 242; Log, 9: 59). Analytic predicates, despite having material content, may well also be called ‘logical’ predicates. These predicates serve indeed to express the ‘logical essence’ of an object, or the definition that serves merely to distinguish an object from others (see Log, 9: 143), not to mention that ‘analysis’, unlike synthesis, is a task of formal ‘logic alone’ (see Log, 9: 64). Consequently, Kant also calls synthetic predicates ‘real’ because they represent realities or determinations of an object, which go ‘beyond the concept of the subject’ and enlarge it (A598/B626). It becomes clear then that the distinction between logical and real predicates, understood in either sense, is mutually exclusive and, therefore, exhaustive. ‘Formal’ opposes ‘material’ and, in turn, ‘analytical’ opposes ‘synthetic’.
With this clarification in mind, ‘concept’ can then be taken to mean the complex of all logical and real predicates of a thing, i.e. the complex of all analytical and synthetic predicates. Kant thus speaks of the ‘complete concept’ (A510/B538, A572/B600) or, equivalently, of the ‘thoroughgoing determination’ of a thing through all the predicates with which such a thing may be thought of (A600/B628; cf. BDG, 2: 72). In this sense, the concept of ‘a hundred thalers’ contains not only the concepts of ‘a certain quantity of silver coins used at Kant’s time in Europe’, but also all the other real predicates concerning their weight, shape, dimension, colour, economic value, etc. This meaning of ‘concept’ makes plausible the major premise of Kant’s disjunctive syllogism. Indeed, for a concept ‘to express its entire object’ or, in equivalent terms, to be the ‘suitable concept’ of its object, the realities thought in the concept have to be exactly the same as those that the object actually has.
But ‘concept’ can also mean the complex of logical predicates, i.e. of analytical predicates which suffice to distinguish an object from all others. The complex of logical or analytical predicates, ‘a certain quantity of silver coins formerly used in Europe’, arguably suffices to distinguish ‘a hundred thalers’ not only from any other object, but from any equal amount of another coined currency. This meaning of ‘concept’ – call it ‘mere concept’ – makes plausible the minor premise of Kant’s disjunctive syllogism. Certainly, the nominal definition of a concept ‘expresses merely the possibility’ – indeed the logical possibility – of its object, since only the absence of contradiction among its marks (or logical or analytical predicates) is required.
Thus, if in the major premise of Kant’s disjunctive syllogism ‘concept’ is taken in the sense of ‘complete concept’ and in the minor in the sense of ‘mere concept’, as the truth of the premises demands, then the argument contains a patent quaternio terminorum that definitively invalidates it. If, on the contrary, in both premises ‘concept’ is understood in the same sense, a dilemma appears. Either we have to assume that a ‘mere concept’ has exactly the same content as its object, or we must take for granted beforehand that existence is not part of the content of the ‘complete concept’ of any object at all. The first claim is a patent falsehood, since it is obvious that a concept that suffices to distinguish its object from all others does not need to express its entire object. So understood, the major premise of the syllogism is, therefore, unacceptable. But the last assertion is a petitio principii, because it amounts to asserting without any proof just what should be substantiated by the conclusion. Thus, understood this way, the minor premise begs the question.
Similar logical errors appear in the ‘corollary’ to Kant’s argument. The ontological proof consists in pointing out the unique case of the concept of ‘something than which nothing greater can be thought’, or of the ens perfectissimum, in the entire field of human knowledge. Anselm enunciates this finding by arguing that anyone who rightly understands that God is ‘something than which nothing greater can be thought’ cannot think that such a being does not exist in reality. The impossibility of thinking of the non-existence is not found in any other case. That is why Anselm challenges anyone to find for him, ‘besides that than which a greater cannot be thought’, ‘[anything else] (whether existing in reality or only in thought) to which he can apply the logic of my argument’.Footnote 15 Descartes expressed the same idea by suggesting, in his Replies to Caterus’ Objections, that since necessary existence is contained in the nature or concept of God, ‘it may be truly affirmed of God that necessary existence belongs to him or that he exists’.Footnote 16 This conclusion is only valid in the case of the ens perfectissimum, because, as he replied to Gassendi, ‘in the case of God necessary existence is in fact a property in the strictest sense of the term, since it applies to him alone and forms a part of his essence as it does of no other thing’.Footnote 17 Anselm and Descartes agree, then, in claiming that, in the unique case of God, the existence can be cognized ‘entirely a priori’, from the concept alone.
In the face of these reasons, to argue, as Kant does, that in the case of the ens realissimum its necessary existence cannot be cognized in any way is either an ignoratio elenchi or a petitio principii. Kant’s claim, in fact, either does not take into account the impossibility of considering the non-existence of God, or takes for granted what has to be proved.
However, against all these reasons, Bering and Kant might well answer that they are not committing an ignoratio elenchi. As critics of the ontological argument, they refuse this proof of God’s existence from the concept of God by relying on a fact reputed to be indisputable. Any concept, however it may be constituted, whatever the predicates or marks it may include, is of exactly the same nature as any other. Just as actual existence is not included in the concept of a hundred thalers, neither is it in the concept of the ens realissimum, even though in this concept one thinks of the impossibility that its object does not exist. Both the hundred thalers and the ens realissimum may really exist or may not exist. The only difference is that in the concept of the hundred thalers we think of a contingent existence, while in the concept of the ens realissimum we think of a necessary one. This only means that, if there are thalers, we can say that there might not have been; and if there are no thalers, we can say that there might have been. By contrast, the ens realissimum exists or does not exist, but we cannot say that, in the first case, there might not have been or that, in the second, there might have been.Footnote 18 We have indeed to think of such a being as an existing one, but we are not committed in any way to affirming its actual existence outside our thought.
Consequently, the critics of the ontological proof could reargue that to claim that there is a concept, and only one, that includes actual existence among its marks, is precisely either to ignore the very nature of any concept (an ignoratio elenchi) or to accept in advance, without any reason whatsoever, the exceptionality of the concept of the ens realissimum (a petitio principii). Is it possible to escape this state of deadlock in the discussion?
3. From the Example of the Hundred Thalers to the Problem of Universals
Among the various possibilities for escaping the impasse into which the previous discussion has led us, it may be useful to examine one of them, even if only very briefly. Considered from the sole point of view of their nature as concepts – certainly not from the standpoint of the diverse cognizability of their objects – neither Bering nor Kant recognizes a substantive difference between the concept of the hundred thalers and that of the ens realissimum. By contrast, Anselm and Descartes insist on the peculiarity of the case of the supremely real being. What does this discrepancy indicate?
This disagreement reveals that Anselm and Descartes, on the one hand, and Kant and Bering, on the other, would give a different answer to the following question. When we predicate of a thing what it is and include it in the collection of things that are of the same kind (even if there can only be a unique thing of such a kind) – i.e. when we use a universal designation to refer to a thing – do we in each case only use a concept or do we also refer to something real that is, in some way, in the thing itself? This question is an elementary way of approaching what is traditionally called the ‘metaphysical problem of universals’. The critics of the ontological argument Kant and Bering, and the proponents of the proof Anselm and Descartes, seem to disagree deeply on the ontological status of universals. With regard to the nature of the predicates that one thing can in principle share with others – its universal features, or universalia – the former could be regarded as accepting a kind of ‘conceptualism’, i.e. the claim that universals do not exist independently of human understanding because they are produced by it. The latter, on the contrary, could be included among the supporters of a kind of ‘realism’, i.e. the claim that universals are not only products of the understanding, but also exist independently of our understanding, with an extramental basis in the nature or essence of things. For the former, all concepts, as mere products of the understanding, are equivalent. For the latter, the very essence of things makes a difference regarding the concepts through which we conceive them.Footnote 19
Let us go one step further. Realists about universals claim that there are essences that are somehow independent of the mind. They also state that essence is really distinct from existence. But is this true in all cases? Do we not cognize a case, and indeed only one, in which essence is identified with existence? Certainly, by a common definition, God is his own existence. Can we not then assert that the highest being exists from the mere cognition of the divine essence? Some realists about universals – Anselm and Descartes – answer in the affirmative: as God’s essence includes existence, to cognize ‘what God is’ is to cognize ‘that God is’. The conceptualists – at least Bering and Kant – respond negatively: no concept whatsoever – no essence as it is in the mind – includes existence.
The following question then arises. Ultimately, is not this difference in answer to the metaphysical problem of universals a fundamental reason behind disagreement over use of the example of the thalers as a criticism of the ontological argument?
It will be rightly said that the metaphysical problem of universals is not such a problem for Kant. He not only devotes to it very few observations, made in passing,Footnote 20 but also thinks that the traditional ‘realist’ response can lead to serious misunderstandings. However, the problem, far from having disappeared, remains to this day central to philosophical debate. Its discussion then seems inescapable. Kant, moreover, certainly adopts his own posture towards this fundamental issue – as stated, a kind of ‘conceptualism’. It is simply obvious, he seems to think, that universality is an exclusive product of our understanding. ‘Only the conceptus is universalis’, Kant adamantly claims before his university students, thus putting an end to the scholastics polemic between Occamists – to whom he attributes the contention that universals are ‘mere names’, and Scotists – to whom he assigns the opinion that universals are ‘real things’ (V-Met/Volckmann, 28: 422). To some extent, Bering’s position is analogous to that of Kant. Accordingly, it seems possible to see the position that transcendental idealism – perhaps tacitly – adopts in the face of this problem as one of the fundamental reasons for the rejection of the ontological argument and the use of the example of the thalers to criticize it.Footnote 21 Let us very briefly go through this possible line of thought.
The characterization of concepts as universal representations produced by the spontaneity of the understanding (see A50/B74; Log, 9: 91) leads Kant to pose two problems. The first is: how can concepts refer to singular things, thus making cognition of them possible? We can call this the epistemological problem of universals. Kant expressly subdivides this general problem into two main questions. The first emerges from the nature of concepts as mere representations, i.e. as ‘inner determinations of our mind’: ‘[H]ow do we come to posit an object for these representations, or ascribe to their subjective reality, as modifications, some sort of objective reality?’ (A197/B242). The second is raised by the intellectual nature of concepts, as representations that are ‘entirely unhomogeneous’ in comparison with sensible intuitions: ‘[H]ow is the subsumption of the latter under the former … possible?’ (A138/B177). Although the first question is raised by Kant expressly with regard to the case of representations of appearances, and the second with regard to the case of ‘pure concepts of understanding’, both questions could be understood in a broader sense as referring to concepts in general.
Largely, Kant’s transcendental philosophy can be seen as a response to these questions. However, in the development of the various and difficult enquiries required by the epistemological problem of universals, Kant ineluctably stumbles upon a new problem not expressly posed by him. This problem arises from the universal nature of our concepts as representations that are ‘common to several objects’ (Log, 9: 91). Where does the universality of concepts come from? What ontological status does the universal have? These issues belong to what we may call the metaphysical problem of universals. One of the very few passages that Kant devoted to this problem runs: ‘The scholastic dispute between the realists and the nominalists was over the question of whether universals (universalia) were actual things or only names’ (V-Met-L2/Pölitz, 28: 560).
Kant does not even seem to accept this approach to the question. To speak of universalia in essendo, of ‘real essences’, as medieval philosophers do, or to use, as Eberhard and many metaphysical compendiasts do, this ‘thoroughly reprehensible scholastic expression’ universal things (allgemeine Dinge) can lead to a serious misunderstanding. These expressions in no way allude to ‘a special class of objects’, Kant says in his response to Eberhard, since they do ‘not designate any difference in the nature of things, but only in the use of concepts’ (ÜE, 8: 217). Nor does the other possibility suggested by the dilemma seem acceptable to Kant. Just as things are not universal, names are not universal either. This at least follows from a concise sentence in one of his lectures on metaphysics, in which he arguably rejects universalia in significando: ‘Between names and things lie concepts and these are the universalia’ (V-Met/Dohna, 28: 636, my translation).
Whatever these phrases may mean, the broad outlines of Kant’s position seem clear. To put it in scholastic terms, there are only universalia in repraesentando. The universality proper to the concept is, then, merely representative. This universality is the capability that every concept has of being a representation common to a plurality of objects. But to state that only concepts are universal in this precise sense is tantamount to claiming that universality comes exclusively from the understanding or, in other words, that universals cannot exist independently of our mind. Nothing in the reality of things is universal; or, if preferred, individual things do not share universal essences.
To be sure, for Kant, what we have called the metaphysical problem of universals is not a metaphysical problem at all. In his reply to Eberhard he recognizes that ‘the conflict between nominalists and realists … still belongs merely to logic and certainly not to transcendental philosophy’ (ÜE, 8: 217). Nevertheless, Kant’s position on the problem clearly involves a metaphysical stance regarding the reality of things. Commenting on a passage in Baumgarten’s Metaphysica, Kant himself cannot help but admit this ontological thesis: ‘Only the conceptus is universalis, but not the thing itself, and an ens omnimode determinatum is singulare or individuum’ (V-Met/Volckmann, 28: 422, my translation). A basic tenet of Kant’s transcendental idealism is indeed that universality in phaenomena is due merely to our concepts, whereas noumena are of themselves neither universal nor singular.
Certain indications invite us to attribute to Bering a position on this issue that is in some way analogous to that of Kant. In criticizing the Leibnizian formulation of the ontological proof, Bering distinguishes what he calls the ‘physical essence’ from the ‘metaphysical essence’ of a thing. He writes: ‘The former consists of the totality of powers and effective capacities found in a substance’ (PdB, 25–6). By the latter, by contrast, ‘is meant all the essential properties of the same thing, but without regard to its substance, its powers and effectiveness, which are only seen from the side of how far they can be thought of in such a connection’ (PdB, 26). Thus Bering seems to hold a metaphysical position according to which in reality there are only individuals, namely, substances or physical essences. Universality occurs only in the understanding, i.e. in the concept expressing the metaphysical essence of a thing as the totality of its logically compatible essential properties.Footnote 22
This conception of the universality of concepts as the exclusive product of the understanding determines Bering’s and Kant’s use of the example of the hundred thalers in their criticisms of the ontological proof. Both philosophers in fact do not recognize a substantive difference between the concept of a hundred thalers and that of the ens realissimum – regardless of the difference in cognizability of their objects, both concepts are no more than entities produced exclusively by our understanding. But an opposing view on the nature of concepts would justify, in turn, rejection of the example of the thalers as a basis for objection to the argument, as can be inferred from the claims of Anselm and Descartes, to cite only the originator of the proof and its renovator in the modern age.
Bering undoubtedly bases his criticism of the ontological argument on the assumption of the essential parity of all concepts, whether they represent God or the beings created by Him, the ens realissimum or a hundred thalers. Bering says, ‘The essence of things consists of their inner possibility.’ From this truth ‘nothing follows concerning their real existence’. And he concludes, ‘In terms of this essence, the Creator and the creature are thus completely equal in the fact that each one consists of the inner possibility, so that what cannot be deduced therefrom for one, we cannot look for in the other either’ (PdB, 25).
A similar assumption underlies Kant’s criticism of the proof. Both the concept of a hundred thalers and that of the ens realissimum are, as concepts, mere mental entities. The former, as an empirical concept, arises from three acts of the understanding, as Kant explains in his Logic: the comparison of many individual objects, reflection on what they have in common and abstraction from their differences (see Log, 9: 94–5). The latter has its origin in reason, when it applies the principle of thoroughgoing determination as ‘the principle of the synthesis of all predicates which are to make up the complete concept of a thing’ (A572/B600) to the ‘idea of an All of reality (omnitudo realitatis)’ (A575–6/B803–4). In both cases, we have to go beyond these concepts if we are to ascribe existence to their objects. This is only possible in the first case, but not in the second, since its object is beyond any possible experience.
Anselm expressly denies this supposition. In his Proslogion, he first distinguishes two kinds of cogitare, or thinking: cogitare secundum voces, i.e. thinking a thing ‘when the word signifying it is thought’, and cogitare secundum res, i.e. thinking a thing ‘when that which the thing is, is understood’.Footnote 23 This distinction is tantamount to affirming that the universality of concepts is not the exclusive product of the understanding, but can also be a reflection or expression, i.e. a cognition, of the real essence of things. There are not only universalia in repraesentando, but also universalia in essendo. Then Anselm establishes a second distinction, this time among real essences or universals in reality: the distinction, we could say, between ‘essences greater than which something can be thought’, such as a hundred thalers, and ‘the essence greater than which nothing can be thought’, such as the unique case of the ens realissimum. The former can be thought not to actually exist. The latter decidedly cannot be so, if one really understands ‘that which the thing is’. For it is impossible to think, or to understand, without contradiction that ‘that than which a greater cannot be thought’ really is such a being and, at the same time, that it exists only in the understanding, as a mere ens rationis that could even contain contradictory marks. Thus, Anselm claims, ‘no one who understands that which God is can think that God does not exist, even though he says these words in his heart either without any signification or with some strange signification’.Footnote 24 This affirmation of God’s existence in no way relies on a mere logical necessity imposed by the pure mental construct of a being containing all perfections, including existence. On the contrary, it is based on a true cognition of the essence of such a being understood as ‘that than which a greater cannot be thought’. It is precisely this cognition of the true essence of God, albeit indirectly obtained and imperfect and not exhaustive, that prevents us from thinking that the Supreme Being does not really exist. In this realist position on the problem of universals, the case of the genuine concept of God is simply incomparable to any other.
Analogous distinctions, mutatis mutandis, can be found in Descartes’ Meditations. The French philosopher famously distinguishes between ‘adventitious ideas’ (ideae adventitiae) that come ‘from things which are located outside me’, ‘factitious ideas’ (ideae a me ipso facta) that ‘are my own invention’, and ‘innate ideas’ (ideae innatae) that seem ‘to derive simply from my own nature’.Footnote 25 Factitious ideas, representing fictitious and mutable natures, are partly similar to what Anselm calls cogitare secundum voces. In contrast, both adventitious and innate ideas, representing true natures, are closer to the Anselmian cogitare secundum res. Innate ideas, however, are distinguished from the adventitious in that they do not derive from sensible experience and do not represent contingent natures, but ‘have their own true and immutable natures’.Footnote 26 Despite this difference, the essences represented by adventitious ideas and by all but one of the innate ideas have something in common. All these essences, whether contingent or necessary, can be perfectly conceived without existence, though in the reality of things ‘existence is merely existing essence’.Footnote 27 The case of the ens perfectissimum is the only exception. Not even in thought can such a being be thought not to exist, for it is a plain contradiction to think of a supremely perfect being and to think of it as lacking existence. This impossibility of thinking is based, not on a merely logical necessity imposed by the factitious idea of a being involving all perfections, but on a cognition of the true and immutable essence of God. ‘So we shall come to understand’, Descartes expressly recognizes when responding to certain objections, ‘that necessary existence is contained in the idea of a supremely powerful being, not by any fiction of the intellect, but because it belongs to the true and immutable nature of such a being that it exists.’Footnote 28
To sum up, use of the example of the hundred thalers to rebut the ontological proof is justified by a certain ontological assumption concerning the universality of concepts, whereas rejection of the pertinence of the example is grounded in denial of such an assumption.
4. Conclusion
The metaphysical problem of universals as formulated in these pages is far from being solved. But even apart from the question of the validity of Kant’s main objection to the ontological argument employing the example of the hundred thalers, our discussion of the latter in the light of Kant’s and Bering’s reasoning leads us to two main outcomes. The first is that both Bering and Kant characterize the ontological proof in a manner that is contrary to the intentions of its proponents, or at least Anselm and Descartes. As the title of his booklet shows, Bering describes the proof as constructed ‘from the concepts of a most perfect and necessary being’. As is well known, Kant describes the argument as ‘proof of the existence of a highest being from concepts’ (A602/B632), or as a proof in which ‘the existence of a highest cause’ is inferred ‘entirely a priori from mere concepts’ (A590/B618). Kant calls the Anselmian or Cartesian proof ‘ontological’, precisely because he considers it as a paradigmatic case of ontology, i.e. of the ‘proud’ science that presumes to offer synthetic a priori cognitions of things from mere concepts (see A247/B303). But Bering’s and Kant’s characterization of the ontological proof directly contradicts what Anselm and Descartes expressly claimed regarding the nature of their reasoning. As stated in the previous section, neither Anselm nor Descartes conceive their argument to be grounded on the sheer concept of God, whether arbitrarily produced or built in view of logical necessities imposed by our reason. Rather, both thinkers assert that the starting point of their reasoning is the exceptional essence of God inasmuch as it is correctly – though not directly and fully – cognized by us.
The second main outcome of the above discussion is that neither Bering nor Kant can put forward their criticisms as evidence of the intrinsic invalidity of the ontological proof. Bering apparently fails in his stated purpose of discovering the ‘unground’ (Ungrund) of the proof (PdB, 3) – at least its last and intrinsic unground. Kant, in turn, seems not to have succeeded in proving ‘the impossibility of an ontological proof of God’s existence’ (A592/B620) – at least the intrinsic impossibility of such a proof. Both thinkers have succeeded, however, in showing the incompatibility of the ontological proof with a certain conception of the nature of universals and, in the case of Kant, with his related conception of the ‘unity of experience’.
