Immanuel Kant criticised what he first termed the Ontological Argument at the beginning of his Critique of Pure Reason (1781). Focussing on the argument as presented by Rene Descartes, which suggested that existence is a perfection and thus a necessary attribute of God, who is a supremely perfect being, in the way that having three sides is a necessary property of a triangle or having valleys is a necessary property of being a hill – Kant concluded that the argument was “so much labour and effort lost”. For Kant, existence is not a perfection and is wrongly used as a predicate. He used the example of a sum of money – the difference between a real and imaginary sum is not that the real sum is worth more, just that the real sum might be in my pocket. Existence is not a predicate and does not describe the properties of an object, it just informs me whether there is such an object in the real world. Bertrand Russell developed this point, using the example of the claim “the present King of France is bald”. Russell pointed out that although the claim seems sensible, as if it is referring to the properties of the King of France’s head and might be either true or false, in actual fact, the claim is meaningless because there is no present King of France for the claim to refer to and thus no way that the claim is either true or false. Existence is not a predicate, it is not just another property that the present King of France does or does not have, it is the ground of meaning on which all sensible claims must be made. Michael Palmer used another example to explain this; that of two candidates applying for a job. If a panel is faced with two CVs listing the “perfections” of the candidate A and candidate B, it would be ridiculous to list “exists” as one of them – existence is neither a perfection nor properly used as a predicate, rather it is what makes the analysis of the CV and the contest between the candidates meaningful. Kant’s criticisms of the Ontological Argument were highly influential and following the publication of the Critique of Pure Reason in 1781, scholarly interest in the Ontological Argument declined steeply. Nevertheless, developments in the second half of the 20th Century showed that Kant’s criticisms are far from conclusive and reawakened scholarly interest in the Ontological Argument for God’s existence.
While Kant’s criticisms were directed at Descartes’ Ontological Argument, they are often applied to the arguments of St Anselm of Canterbury, presented in his Proslogion (1078). Anselm argued that if God is “that than which nothing greater can be conceived of” then God must, a priori, exist because it is greater to exist in reality than just in the mind, so if God existed only as an idea in the mind (as he must, if Anselm’s definition was accepted) then something greater could conceivably exist… something that existed in reality as well as in the mind. If God is, therefore, the greatest conceivable being then God must exist in reality, because existence is a perfection which makes something greater. Clearly, Kant’s arguments that existence is not a perfection and that existence is wrongly used as a predicate seem to undermine Anselm’s argument fatally. As Gaunilo of Marmoutiers had observed in his “On behalf of the fool”, the idea that the perfect island has to exist just by virtue of being the perfect island is absurd; nobody is going to book a ticket to go there on the basis of an argument like that. Nevertheless, that ignores how Anselm developed his argument in the next chapter of the Proslogion, a point that he made in response to Gaunilo’s attack by restating this part of the argument in his Responsio. In Proslogion Chapter 3 Anselm reasons that it is better to exist necessarily than to exist only contingently, therefore necessary existence – not being able to not exist – must be an attribute of that than which nothing greater can be conceived of. This development of the Argument could defeat Kant’s standard criticisms, in that while existence is not a perfection or rightly used as a predicate, that does not necessarily apply to necessary existence, which is a total state of existence either possible or impossible, not a property which might or might not be added to a object that could only ever contingently exist.
Norman Malcolm argued that Anselm’s argument in Proslogion 3 can be presented in terms of modal logic. Either God’s necessary existence is impossible – as in it contains a formal contradiction – or possible. If God’s necessary existence is possible, then it is necessary. Remember, necessary existence is not existence in the sense that we could encounter through our senses. The world of sense is a world in which things exist contingently and might or might not exist, as St Thomas Aquinas observed in his Third Way to God (Summa Theologica I.II QIII). To exist necessarily is to exist in a different way, a way that is by definition beyond anything that we could experience through our senses. For Kant, because necessary existence is beyond possible experience, then it can only be speculative to even speak of it. Kant called necessary existence a “cupola of judgement” meaning that it strays so far beyond possible knowledge to be a flight of fancy or a castle in the air. Nevertheless, this assumes Kant’s world view and the primacy of sense-experience. For rationalist philosophers like Descartes, Leibniz and later Malcolm, what is real cannot be limited to what can be experienced through the senses. The world of sense is faulty, partial, subjective and limited; empirical knowledge is contingent and ever-changing. For Descartes and Leibniz, rational knowledge should be primary because it is none of these things. A clear and distinct idea, an idea which contains no contradictions, is certain, complete, objective and constant. Just as we know that 1+1=2 without resorting to a posteriori reasoning based on experiences with apples and oranges, we know that God necessarily exists a priori because he is supremely perfect. Alvin Plantinga and William Lane Craig have developed this line of reasoning, using the device of possible worlds. A concept is possible if it could be instantiated in any possible world. A unicorn is a possible concept; although unicorns (at least in the sense of being live horses with single horns!) don’t exist in this world, it is not inconceivable that they might exist in a multiverse. The concept of a horse with a horn is not contradictory, it is possible. On the other hand, a square circle is impossible and could not exist in any possible world because the definition of a square is to have four straight sides, something which directly contradicts the definition of a circle. God’s “maximal greatness” – which must include Omnipotence, Omniscience and Omnibenevolence – is possible not impossible and, because maximal greatness precludes the possibility that God might or might not exist in any one universe, God must necessarily exist in every possible world, including this one. In essence, Plantinga and Craig show that Kant’s criticisms of the Ontological Argument fail because they rely on an impoverished epistemology.
To explain this, for Kant, all existential claims have to be synthetic, they have to refer to something in the world of sense-experience and therefore contain the possibility of being either true or false. If I say “unicorns exist” I am making the claim that there are such things as unicorns in the world – if I saw a unicorn I would know that the claim was true and if no evidence of unicorns has ever been found it is probably fair to say that the claim is false. Making the claim “God exists” does not refer to anything in the world of sense experience and it is not possible, therefore, for the claim to be either true or false… in the terms of the Logical Positivists, it is a meaningless claim. In 1951 American Philosopher WV Quine attacked the basis of Kant’s objections to the Ontological Argument in his essay “Two Dogmas of Empiricism”. Like Karl Popper and AJ Ayer, as a young man, Quine had spent time with the Vienna Circle of Logical Positivists, but by 1950 Quine came to reject their approach, not only to establishing meaning in language but also to bigger questions about epistemology – “what can we know?” and “what does it mean to say that something exists?” Logical Positivists were Positivists, that is to say that they approached philosophy on the basis that scientific, empirical observation is the only source of knowledge and that metaphysics is a waste of time. Positivism formally began with the work of French philosopher Auguste Comte, but looked back to Kant and before that to Hume and Locke. Locke rejected continental rationalism and argued that human beings have no innate ideas, being born as tabula rasa and gaining all their knowledge and understanding from experience. Hume agreed to a large extent, although he acknowledged the limitations of sense-experience as well. Kant said that Hume “awoke me from my dogmatic slumbers and gave a new direction to my philosophical enquiries” and adopted Hume’s fork, the categorisation of possible knowledge into either what is known analytically or what is known from sense-experience, dismissing any other claimed knowledge – including most metaphysical and religious claims – as speculative. Scientific method drew on the work of Locke, Hume and Kant in that it came to focus on sense-experience as the source of all new knowledge and limiting the role of reason to one of analysis and clarification. It is fair to say, therefore, that the Logical Positivists were empiricists. Quine rejected empiricism, and the Logical Positivism of his youth, arguing that in embracing Hume’s fork Kant had awoken from one set of dogmatic slumbers only to fall into another set of dogmatic slumbers. On what basis, Quine asked, did the Logical Positivists claim that sense-experience is the only source of new knowledge? The Logical Positivists failed to provide an adequate explanation of why meaningfulness should depend on either sense-experience or logic and on nothing else. This point was later developed by Alvin Plantinga in his “God and Other Minds” when he pointed out that the Verification Principle is itself unverifiable and therefore self-defeating. Quine also questioned the lack of any adequate explanation for the authority of logic, pointing out that you have to accept logic in order to defend why you should accept logic, which is circular. The same applies to the authority of the empirical senses, Quine argued. On what basis do we say that the sense-experience are the only source of new knowledge without just appealing to sense-experience? This is pure reductionism and again, the justification for Positivism is circular. The answer is that the Logical Positivists adopted Kant’s world-view without much thought, ignoring the fact that Hume (not to mention Descartes) had already outlined the serious problems with relying on the senses in that they are faulty, limited in scope and that data always needs to be interpreted through reason anyway. If Quine is correct and Kant’s epistemology is more dogmatic than critical, then his criticisms of the Ontological Argument start to collapse. On what basis did Kant claim that all existential claims have to be synthetic? For Quine, he had no adequate justification for assuming the authority or primacy of sense-experience, other than by appealing to that same sense experience. Without any proper justification for his epistemology, it seems that Kant’s criticism of the Ontological Argument is on shaky ground indeed.
Further, to say that sense-experience is the only source of new knowledge is to relegate whole fields of study, discussion and indeed human experience to junk status. The fact that the Verification Principle proposed by the Logical Positivists as the gold standard of meaning had to be liberalised through the 1930s and 1940s shows that the claim that discussions about topics such as Ethics, History and Aesthetics (let alone Religion) are meaningless is unworkable and runs against what most people believe and experience. Quine proposed an alternative holistic approach called Ontological Naturalism, which moved away from the attempt to define the meaning of individual statements in terms of their reference and towards assessing their meaning in terms of cohering with and contributing to the whole field of science as an explanatory framework. Popper also rejected Verificationism, proposing another, more generous and inclusive, approach to meaning in scientific terms in the Falsification Principle. Both were influential and contributed to a decline in Logical Positivism to the extent whereby by 1960 it was declared “dead, or as dead as any philosophical movement can be.” The decline of Logical Positivism demonstrates the inadequacy of Kant’s world-view for the modern world. What place has a system which claims that sense-experience is the only source of new knowledge in a world of Quantum science and particle-Physics, in which the very act of observing particles changes their state? Long gone is the idea that the senses offer human beings a transparent window on the external world, even the world of matter and energy. Today, the whole field of theoretical Physics would have to be declared “meaningless” by Schlick, Ayer and Carnap… and yet the insights it yields offer humanity unthought of technological advances… they work. Further, Physics suggests that what appears “real” to our senses is far from being solid and as it appears. On the Planck scale no matter exists… if I hit the table the contact I experience is really the interaction of charges in the fields which make up the vast majority of each atom in the wood and in my hand. The universe, which appeared like a vast machine to Kant and which still appears eternal to the amateur start-gazer has been revealed to be infinite while still having edges, a shape and a colour and while expanding at an increasing rate… into nothing. All of this suggests that reason – mathematics – can yield new knowledge and confidently move past anything we can hope to observe through our senses, to a much greater degree than either the Logical Positivists, or Kant, allowed.
On the other hand, even theoretical Physicists admit the need to test their theories through experiments. Very recently, different theories on black holes were tested when radio-telescopes were linked together to take a photograph of a black hole. The photograph – an observation – was necessary to check analytical calculations and prevent them from being purely speculative. This supports Kant’s claim that all existential claims must be synthetic. Physicists theorise about black holes, but it is not possible to say that or how they exist unless and until we take a photograph or make some other observation to verify (or falsify) the theories. Nevertheless, there are aspects of Physics – as there are aspects of Theology – which resist any possible observation. By definition, it is not possible to observe God’s necessary existence, because by definition it must be outside of the matrix of time and space in which our senses operate. Similarly, it is not possible to observe what caused the Big Bang which created the space-time continuum, or to experience conditions in a multiverse. The extent to which cosmological theories like cosmic inflation and string theory are pseudo-scientific (to use Popper’s phrase) because they are not falsifiable or subject to normal scientific method has been a matter of controversy on the letters’ page of Scientific American since 2017. Nevertheless, this should not stop Physicists (or Theologians) from using reason, the other source of knowledge available to them, to push forward the boundaries of knowledge. While Kant was right to be cautious and to warn against metaphysical speculation – because after all, the greatest obstacle to finding something is being convinced that you already have it – his world-view with its focus on the senses as the arbiter of possible knowledge is too restrictive for the 21st Century in the way that the world-view of the Logical Positivists became too restrictive for the 20th Century. It also sits ill with both developing insights about the way in which our senses work and rely on our brains and pre-existing ideas and with insights about the different reality beyond how things appear to our senses, on the Planck scale. While Popper’s Falsification Principle is more flexible than the Verification Principle, it still limits what can be said scientifically to that which can be falsified in relation to observations, it still assumes a Kantian world-view, and herein lies the problem for particle Physics and Cosmology with Scientific Method as it is conceived today. It seems that Kantian epistemology and the assumed world-view within which science has operated since the 1790s is on the verge of being rejected; to use Thomas Kuhn’s phrase, the Positivist scientific (and philosophical) paradigm is shifting and giving way to a paradigm which is more open to reason providing new knowledge which cannot be checked by observation. Given this, it seems that Kant’s criticisms of the Ontological Argument will lose a great deal of their power. As William Lane Craig has pointed out, the Ontological Argument (as he presents it) is valid. If it is accepted that an argument can also be sound – its propositions can be said to be true – even when they cannot be verified or falsified empirically, then the Ontological Argument is much more persuasive.
Nevertheless, even if Kant was too cautious about using reason as a source of new knowledge, it could be fair to say that the Ontological Argument pushes things too far. St Thomas Aquinas made just this point in his Summa Theologica I.II question 1 (1264) when he wrote “because we do not know the essence of God, the truth of God’s existence is not self-evident to us.” Aquinas dismissed the Ontological Argument because to suggest that any human being can have a “clear and distinct idea” of God to use Descartes phrase from the Meditations, sufficient to analyse that idea and find necessary existence – a unique property of God – within it, is arrogant. God is, to use Augustine’s words “other, completely other”, outside time and space, so even if we have innate ideas – as Descartes and Plato argued – it stretches credulity to rely on those ideas being so complete in relation to God so as to make the Ontological Argument plausible. Clearly, what it means for God to be perfect or great is not what it means for human beings to have a perfection or be called great. Aquinas even rejected the idea that it could just be a matter of scale, with God at the top of the scale and people (and other created things) at the other. If God is timeless then God’s perfection and greatness must also be timeless and cannot include any potential for God to be other than he is, to change or to choose. Aquinas’ saw claims about God’s nature as necessarily analogical. Yes, our greatness depends on God’s greatness, but in the way that the healthiness of a yoghurt depends on the healthiness of somebody who eats it. Healthy people are slim, muscly and bounce around… healthy yoghurts are none of these things! For Aquinas, God’s nature can be described (and known) positively, but only in a very limited sense, not completely enough to support an ontological demonstration of God’s necessary existence from an a priori definition of His nature. Nevertheless, St Anselm and Descartes on one hand and Karl Barth and Iris Murdoch on the other would all reject this argument, with reasons that also serve as criticisms of Kant’s approach to the Ontological Argument.
For Anselm it is true to say that God’s attributes are not the same as human attributes, because God is timelessly perfect, nevertheless it is possible to understand enough about God’s greatness to deduce that necessary existence is a necessary property of it. This is because, according to the a posteriori argument in Anselm’s Monologion, God is the best explanation for our ability to judge things in this world as more or less perfect. God creates us with an innate conception of perfection, the top of a scale which we use to measure things in this world every day. Indeed, Aquinas – although he still denied that this would give people a clear enough concept of God to analyse and find necessary existence within – included a similar argument as his fourth way to God in Summa Theologica I.II question 3, arguing that claims about God’s nature can be understood as analogies of proportion as well as as analogies of attribution (as above). This suggests that Aquinas’ rejection of Anselm’s approach was more about the degree to which we can conceive of the essence of God and a matter of interpretation, rather than about Anselm’s whole methodology in the Monologion, which he later used to support his reasoning in the Proslogion. Descartes “trademark” argument in the Meditations supports Anselm’s belief that God creates us with an innate idea of God’s existence and there have been other, similar arguments in the work of scholars from St Augustine to CS Lewis, arguing from the experience of believers, their desire for and innate awareness of God, a posteriori to the conclusion of His existence. It seems that many believers could accept the idea that we have an understanding of God’s supreme greatness even if we cannot completely conceive of what that might entail. It seems, therefore, that Aquinas’ dismissal of a priori attempts to argue that God’s existence is self-evident is rather hasty. Just as Aquinas’ rejection of the idea that God’s existence is self-evident and exclusive focus on arguments from observation as the only approach to defending God’s existence rationally does not fit with believers’ experience of God as an innate idea, Kant’s rejection of both a priori and a posteriori arguments for God in the Critique of Pure reason does not fit with the imperative to believe in God which he himself expressed in strong terms, most completely in “Religion within the bounds of Reason alone” (1794). While Kant dismisses all the rational arguments for God’s existence, including the Ontological Argument, he argues that God is a necessary postulate, an assumption that it is our rational duty to make, to explain the existence of the moral law which we know as a synthetic a priori. Kant reasons, therefore, that we know the moral law a priori, before experience, although it is supported in all respects by experience as well. The moral law is, furthermore, necessarily explained by God. On what basis can Kant argue that Descartes is wrong to claim that we know God’s existence a priori, something which is supported in all respects by experience as well, but proceed to make a similar claim about the moral law. It is worth asking, is Descartes concept of God actually different from Kant’s concept of God? All that Descartes attempts to demonstrate through his Ontological Argument is that Supreme Perfection necessarily exists… he makes no claim about the Ontological Argument proving anything about the nature of that Supreme Perfection, stopping short of listing attributes like being male, being the father of Jesus etc. Is it fair to claim that the moral law can be known a priori – as Kant does – and reject the idea that the necessary existence of Supreme Perfection cannot be known a priori, when both seem equally borne out by experience.
Further, Karl Barth (and later Iris Murdoch) pointed out that the Ontological Argument, although not conclusive as a proof of God’s existence for the non-believer, it is deeply persuasive for the person with faith. In Barth’s “Faith Seeking Understanding” (1930), far from pre-empting Norman Malcolm’s claim that the Proslogion was an exercise in modal logic (a claim that has been accepted by Hartshorne, Plantinga and Craig), Barth argued that Anselm’s work had been wrongly characterised as a philosophical treatise and argued that it was really a prayer, a deeply spiritual meditation on contingency and necessity and the nature of reality. If Barth is correct then Kant’s criticisms of the argument, at least as they are applied to Anselm rather than their intended target, Descartes, are misplaced. To say that Anselm’s work is “so much labour and effort lost” misses the point that he may not have intended to formulate a deductive proof such as to convince a non-believer at all. Further, while the philosophical intent behind Descartes’ arguments in the Meditations is difficult to deny, it is worth mentioning that this work too was written from the perspective of pre-existing faith. Descartes too had “faith seeking understanding” and was not engaged in a work of Christian apologetics. Kant’s criticisms approach Descartes’ ontological argument for God in isolation and attempt to dismantle it from the perspective of a radically different world-view and epistemological framework. Given the doubt that has been cast on the possibility of attaining objective Truth and a grip on ultimate reality – a doubt that was adopted by Kant from Hume but which was often subsequently ignored – the idea that Kant’s world-view and epistemological framework, which as has been explained rests on as many assumptions as does Descartes, should have the authority to dismiss Descartes’ Ontological Argument is unconvincing. Today, most philosophers have to look to coherence rather than to correspondence to find meaning, and thus have to be open to the possibility that an Ontological argument might be both valid and sound within one form of life and simultaneously valid but not sound in another form of life.
In conclusion, Kant’s criticisms of the Ontological Argument are effective only if his epistemology and world-view are accepted.