How Kant definitively proved the existence of God

Kant famously claims we know not the things in themselves, but only their appearances. “Causality” is an a priori category of thought that we can only apply to these appearances, and therefore we cannot apply them to the things in themselves; consequently we cannot go from our experiences of the world to posit a transcendent cause, such as God, for those appearances.

Of course, the obvious flaw, is that Kant held that the "things in themselves" are those which cause our perceptions. However, this view starkly contradicts another basic view of Kant's: that causality is a subjective category, and as such does not pertain at all to "things in themselves". 

Kantians wanted to conceive of things in themselves as the causes of our sensations, while their own accounts of causal inference ruled that claim out.

For theists we *can* know more than immanent phenomena both because phenomenal beings actually participate in God and because God is irreducible to what Kant deems noumena.

Against all sensation­alist and conceptualist theories the theist insists that the impressions pro­duced upon the senses and the concepts formed in the mind are not the objects of perception but are the media through which external objects are perceived ; not the quod but the quo of the cognitive act. And against all phenomenalist theories the theist makes the same assertion about the qualities which we observe. Extra-mental reality is the direct object of human perception; we know from experience that beings exist.

That we reach God right away can be shown by pointing to the fact that we know everything as finite and limited. But a limit can be known as limit only by him who is, in fact or in desire, beyond this limit. We are not beyond every limit in fact. But we are beyond every limit in desire, because we strive past it.

BUT in his essay ‘Theism’ for the Encyclopedia of Religion and Ethics, AE Taylor definitively shows how Kant, against his later intentions, proved the existence of God,

Since it’s hard to find, here I reproduce the most important excerpts, he begins speaking about Kant’s early, wonderful proof for the existence of God :

“Kant assumes a complex of simple substances or monads. His object is to show that such a complex must depend for its existence and character upon a supreme and ' necessary ’ extra-mundane substance,’ which is God. It runs as follows, the principles of interrelation between a plurality of substances cannot have their complete ground in the existence of these substances. Each substance is indebted for its mere subsistence only to its cause (if it has a cause). 

But the relation of effect to cause is not commercium (reciprocal interaction), but dependentia (one-sided dependence), and what we have to account for is the commercium of the substances which make up the universe. 

Not all these substances can be ‘necessary’, because, if they were, they would be absolutely without dependence on each other ; there would be no commercium between them, and they would not form a world at all.

The world, or totality of substances,' is, therefore, a totality of contingents; and the world, in virtue of its essence, consists of mere contingents. Moreover, no necessary substance is connected with the world (mundus) at all, unless as cause with effect, and therefore not as a part with a whole, since the parts of one and the same whole are connected by reciprocal dependence, a relation which does not affect a necessary being.  

The cause of the world, then, is an extra-mundane being and not a “ soul of the world.”  And the necessary being which is the extra-mundane cause of the world is one and not many. For the effects of different ‘ necessary beings’ would stand in no relations of reciprocal dependence, since their assumed causes are not reciprocally inter-related. Hence the unity of the substances composing the world in a single system is a consequence of the dependence of them all on 

one being, and it follows that this one being is not a mere ’architect of the universe’ but its Creator.

Thus down to 1770 Kant shows no doubt of the possibility of demonstrating theism. The argument on which he relies in all the essays examined Is one and the same— the Neo- Platonic argument a posteriori— and rests on the assumption that the world as given is an object for which we are bound by the principle of causality to seek an explanation. The proof, as with the Neo-Platonists, aims at establishing the existence of the One— the single, internally simple and perfect, extramundane source of all the existents which together make up the cosmos. 

The peculiarity of Kant’s special version of it is that, to escape the criticisms which had been directed against Descartes, he sets himself to deduce the existence of a  being which cannot be thought not to exist,' not from the logical concept of * ens realissimum  (* Kant’s term for God - the most real being, ideal of pure reason) , but from the consideration that, in the universe itself, some combinations of predicates of the same thing and some combinations of relations between the same things are possible, and others not. The existence of an actual extra-mundane being once established as a pre-condition of the difference in intra-mundane things between what is possible and what is impossible, the internal unity, simplicity, and perfection of the necessary being are then deduced as consequences of Its necessary existence. 

If this line of argument is not fallacious— i.e. if it really proves that something ‘ exists of necessity ’—it clearly has the double merit of being free from the objection to the ontological proof, and of being equally untouched by the considerations urged by Philo and Demen in Hume against the a posteriori proof. If the principle on which Kant relies— that the possible presupposes the actual— is sound, his argument seems to be a complete speculative demonstration of the * being of God * reduced to its most succinct expression. 

Kant is clearly right when he asserts that all existential propositions are synthetic, at any rate (to repeat the distinction of St. Thomas to which Kant himself pays no 

regard) quoad nos . And it follows at once from this single consideration that, if, as the Critique maintains, the synthesis in a synthetical proposition must always in the end be effected by an application of formal categories of the understanding to a material supplied in sensation or sensuous imagination, no synthetical proposition (and, by consequence, no existential proposition) can be affirmed of a subject which is purely ‘ intelligible’.

And Kant expressly makes this a main point in his criticism of the ontological proof. Unfortunately, however, this doctrine, if carried out to its full logical consequences, would lead to a result which Kant would nave been the first to reject. 

For it follows that there can be no such sciences as pure arithmetic and pure geometry. 

The subjects about which synthetical propositions are asserted in these sciences are one and all Objekte des reinen Denkens no less than the ens necessarium or the ens realissimum of speculative theology. 

No element whatever supplied by sense enters into the mathematician’s concept of a circle, a parabola, an integer, or a real number. Kant overlooks this all-important point because he assumes throughout his whole reasoning that, 

before I can demonstrate a proposition in geometry,I must draw the figure, and similarly that, before I can say what the sum of two integers is, I must count the units of which he supposes the integers to consist. The erroneous character of this view has been sufficiently demonstrated by the subsequent history of mathematical science, but ought to have been clear to Kant himself.

Even if all geometry, as he tacitly assumes, were metrical geometry, he ought to have seen that Descartes' invention of co-ordinates had already made the drawing of figures in principle superfluous in geometrical science. 

His conception of arithmetic is even more superficial-in fact, on a level with Aristotle’s.

As Couturat has correctly observed, Kant’s examples are all drawn from the demonstration of singular propositions (such as 7 + 5=12). 

If he had asked himself how any general truth in the theory of numbers is proved (how, e.g., we prove Fermat's theorem), he would have seen at once the inadequacy of his own theories. Indeed, mere consideration of a singular proposition which does not relate to integers (e.g., the proposition 2 x 5 + 3x6=6x16) might have taught him that arithmetic is not the same thing as counting, and even suggested to him that an integer is not a ‘ collection of units.” 

Our task becomes the comparatively simple one of considering; whether natural theology does or does not Involve (as Kant alleges that it does) an illegitimate use of the principles of pure logic. From Kant's point of view, to be sure, it does. 

But this is just because Kant assumes that the only legitimate use of logical principles is their employment to order a material given by sense.

If the doctrine of his 'Aesthetic' is rejected, and with it that part of the ‘ Dialectic ’ in which he absurdly tries to show that the mathematical doctrine of infinite series leads to antinomies, it is no longer obvious that what Kant calls a transcendent employment of the principles of logic— i.e. their employment independently of application to  the manifold of sense— need be Illegitimate.

In fact, it is not clear that the whole of the general theory of arithmetic is not just such an employment of logical principles os 'constitutive of a Denkobjekte.’ (It is certainly so if, as is probably the case, the series of natural integers can be defined wholly in terms of the primitive indefinables of logic.) One Kantian paralogism, in particular, may be noted here, it plays a prominent part in the assault on the theistic arguments. Kant complains that all the arguments for the ' necessary being' based on the causal principle depend on employing this principle, which is a mere rule for ordering the appearances of the sensible world, and has no meaning apart from these appearances, as a means of transcending the world of sense.

It might be a sufficient retort that the one form of causality with which we are intimately acquainted is our own volitional activity. In this activity, which is at once efficient and Intentional causality, what are connected as cause and effect are not an earlier and a later event in the ' world of sensible appearances,' but the self, which does not belong to that world at all, and an event which does belong to It. Kant could not deny the causal relation between the rational self and events in the Sinnontoelt without ruining the foundation of his own ethics, but the admission of such causality ought to debar him from attacking natural theology on the ground that it  uses the principle of causality as a means to transcend the world of the senses.

He only escapes open self-contradiction by his monstrous theory, which is not likely to find a defender at the present day, that the self with which we are acquainted is not the real self at all, but a phenomenal self apprehended by an inner sense. From the point of view of logic, the criticism is equivalent to a refusal to admit the validity of any logical inference from the terms of a series to a limit which is not itself a term of the series.

It Is not in Itself any more absurd to hold that examination of the things and events of the Sinnenwelt in the light of the causal principle reveals their dependence on something which does not belong to that world than it is to hold that a series of which every term is a rational fraction can be shown to have a limit which is not a rational fraction. (This is,the case when we represent a surd square root a as a recurrent continued fraction. Each of the convergents is a rational fraction, but the limit of the series Is not.) 

The general argument is thus Invalid. No a priori reason can be given why the causal principle should not enable us to transcend the world of sense, and the only real question which remains is whether the particular arguments of theists will stand scrutiny on their merits. There Is no general logical presumption against them of the kind Kant imagines. 

The typical form of the proof is stated in the Critique for examination is this: (a) If anything exists, an absolutely necessary being exists ; but at least one thing (vis. myself) exists ; ergo an absolutely necessary being exists ; (b) a necessary being must be completely determined by its concept ; the only concept which thus completely determines an object is the concept of the ens realissimum ; ergo the concept of the ens realissimum is the only one by means of a necessary being can be thought ; i.e. a Supreme Being necessarily exists. The argument thus consists of two stages : first, the proof that, because at least one thing exists, a necessary being exists, and, second, the proof that a necessary being can only be the Supreme Being. 

He begins by a charge of general fraudulence. The cosmological proof professes to appeal to experience, but it is really only the old discredited ontological argument dishonestly disguised. For it only uses the appeal to experience to establish the result: 'There Is a being which exists necessarily.' When we ask what this being is, we are referred bank to the ens realissimum as the only thing which meets the requirements of the case. Therefore it is only the ontological proof from mere concepts which contains the force of the demonstration and the alleged experience Is wholly superfluous.’  

The complaint is surely unfair. The objection to the ontological argument did not lie in the concept of the ens realissimum, but merely in the absence of an existential premise. 

If, then, the new argument supplies the missing existential premise, it is no objection to it to say that the necessary being of which it speaks turns out to be the ens realissimum. The only legitimate objection would be that the argument does not actually supply such an existential premiss as is really needed. This is what Kant next proceeds to urge. He complains that it ought to be shown that the necessary being is the ens realignmum. 

To prove this, we require to establish two propositions, of which one is the simple converse of the other : (a) every necessary being is an ens realissimum, ( b ) every ens realissimum is a necessary being. But this second proposition is 'determined merely by a priori concepts,' and therefore the mere concept of the most real being' must be the ground for ascribing to it necessary existence. Thus we commit the fallacy of the ontological proof, the establishment of a proposition by mere analysis of concepts. This criticism seems wholly verfehtt. The real objection to tne ontological proof was that it aimed at (or the) ens realissimum , it is ens necessarium or the simple converse, ' If anything is ens necessarium , it is also ens realissimum.

Both these propositions are Implications, not assertions of existents ; the existential import is brought into the cosmological argument entirely by the preceding proof, or 

attempted proof, that, if anything exists (as we know to be 

the case), a necessary being exists. Kant la entitled to contend that this has not been proved ; he is entitled to contend that the equivalence of ens necessarium and ens realissimum has not been made out. He is not entitled to treat the foot that the equivalence is an equivalence of concepts as proof of this second charge. Up to this point he is merely following the recommendation to give a dog a bad name and trust to its hanging him. We now come to the really relevant part of his onslaught.

This consists of the following allegations : the inference from the contingent to its cause has a meaning only in the sensible world, but the principle of causality is used in this proof to transcend the sensible world ; the argument from the Impossibility of an Infinite series of causes in the sensible world to a first cause Is illegitimate ; (the very notion of necessity presupposes conditions upon which the necessity in question depends, and it Is therefore impermissible to cut snort the regress from proximate to more ultimate conditions by the really empty concept of an unconditioned necessity ; the proof confuses the mere logical possibility of a concept (absence of internal contradiction) with its transcendental possibility, which 'requires a principle establishing the possibility of performing such a synthesis, but this latter can only be established 'in the field of possible experiences.

All these objections are valid only on the hypothesis that the Kantian theories about the limits of scientific knowledge are true, and it has already been contended that this hypothesis (Involving, as it does, the acceptance of the‘Transcendental and Aesthetic and the consequent recognition of the antinomy of the ‘Transcendental Dialectic' as inevitable) is certainly not so.

*In particular, It may be replied to (1) that all use of the principle of causality Involves transcending the sensible world ; consistent phenomenalism, as the work of each writers as Mach, Pearson, Avenarius, abundantly shows, is bound to eliminate the category of causality from science ; to (2) that the 1 cosmological argument ' is not an argument from the impossibility of an Infinite series of events at all. In fact it has been often maintained by thinkers who, like Aristotle, deny that the series of events has a first term, or, like Thomas, hold that it can only be known by revelation whether the series has a first term or not.

The real bearing of the argument cannot be seen at all,so long as we think of causality, as Kant does throughout the Critique, as a mere rule of uniform connexion between earlier and later events. Its real foundation is in the conception of efficient causality (activity or agency).

The point of the argument Is that, If there is not a First Mover or First Agent ( or agents' in the plural, as the case may be), the whole history of the world is a mere accident. Things are what they are because there happen to be such and such agents, And the reason why there are Just these agents and no others is that there happen to be (or to have been ) certain other by which the set first mentioned have been produced, and so on in indefinitum. 

This means that there might just as well have been no world at all, or one quite different from that which there is. An ultimate pluralistic realism, no doubt, might maintain this thesis, and we shall have to face it in the sequel. But it is an ignoratio elenehi to defend it by assuming phenomenalism and erroneous theory of the mathematical meaning of infinite series. The phenomenalism Is further In flat contradiction with the presuppositions of Kantian ethics, for which it is indispensable that every human self shall be a 'first cause’ of its own morally and legally imputable acts. 

If a first cause really means nothing intelligible, Kant's practical philosophy is no better than an idle sporting with insignificant words.

 One must add that the full force of the cosmological argument is only seen when it is combined with the argument from intentional causality. If the historical world-process has a meaning of any kind, whether Its meaning lies In the direction of events towards an end or result or in their internal, quasi-aesthetlo harmony, the explanation of it cannot, in the end, be merely that the constituents of the universe happen to be what they are. An ultimate pluralism must, to be consistent, deny that there is any meaning at all in the world-process. But, again, the Kantian phenomenalism affords no valid reason for entertaining this view. 

As to this last point, it does depend on the special principles of the critical philosophy, and, in particular, on the theory that an appeal to the senses lie at the root of all valid synthetic propositions.

 As has already been remarked, this theory is sufficiently disposed of by the simple consideration that every proposition in the theory of numbers is synthetic in Kant's sense of the word.

 In the present writer’s opinion, then, Kant’s attack on the cosmological argument (which he himself clearly regards os the central feature of his general assault on speculative theism) is a complete failure. He proves neither that the argument from the first of existence to the existence of a necessary being is fallacious nor that there is any sophism in the reasoning by which he himself had formerly established the equivalence of the necessary being with the perfect or ‘most real ’ being. This does not, of itself, prove that the cosmological argument is valid, but it does prove, if the remarks just made are in principle sound, that Kant’s objections to the argument are unfounded.

Even the complaint that we have no positive conception of either ‘necessary' or ‘most real’ being only amounts to the true assertion that we do not know what it would Be like to be God— a proposition which no reasonable theist least of all an orthodox Christian theologian, need lie concerned to deny. For the matter of that, I do not know what it would be like to be my cat, but that is surely an insufficient reason for denying the existence of my cat’s mind. 

The denial that God is an object of possible experience depends, of course, on taking a specific view of what is meant by experience. If It is indispensable to an experience that It should have an object into which sense-data outer os constituents (and this is what Kant always assumes), manifestly God cannot be experienced.

But it may he observed that it is no ground of objection to speculative theology in particular to say that it claims to give us knowledge about a being which is not an object of possible experience in this sense. The same thing Is equally true of arithmetic or any other port of pure mathematics.

The integers, e.a., are not objects of experience in this sense ; still less would it be possible to maintain that, when one utters the well-known proposition,  Every integer can be represented as the sum of four squares, of which— except in the case of the integer 0 — one at least is not 0,’ one is not transcending possible experience. It would be quite impossible to verify the proposition by examining its validity for each successive Integer (since there is an infinite number of them). 

Wherever I make a statement about a class with an infinity of members, I am dealing with an object which is not, in Kant’s sense of the words, 'an object of a possible experience.’ His doctrine reposes on the theory of his  Aesthetic  that, in the case of arithmetical propositions, one can Justify such an assertion by counting. But, though one could, e.g., prove the proposition quoted to hold good for a few case by actual counting, I manifestly cannot verify it or any other general proposition of life science by this method of appealing to intuition.

Again, we may fairly ask why experience should be assumed to be concerned only with objects which fall under the forms of Intuition.* Why are the saint’s moments of vision to be from the outset excluded from experience? If they are included, the statement that God Is not an object of invisible experience at once becomes questionable. It is a standing defect of the Critique that the concept of experience itself has never been subjected by Kant to careful and searching criticism.

The second observation which naturally suggests itself Is that the sharp opposition between speculation and practice might prove on a closer examination to be misleading. All that Kant can claim to have shown, even if every one of his charges against natural theology could be sustained, is that the facts of physical nature do not warrant the theistic hypothesis. But it is surely as much part of the task of speculative philosophy to explain The facts of the moral as it is to explain the facts of the physical order.

 The absoluteness of moral obligations Is an act of the moral order, and, if this fact is only intelligible from the theistic standpoint, then it may fairly he said that speculative philosophy is committed to theism. This was, indeed, Kant’s own conviction, and his hard and fast severance between speculation and practice does less than Justice to the view he intends to maintain.