The Limits of Reason

Dear Dr Craig

I am an atheist but still a big fan of yours. I always defend you against dumb internet atheists who never bother to read anything yet think they can ridicule a man with two PhD's and two dozen books.

You defend the classic God proofs so well. But I think you are relying on commonsense and intuition too much in this day and age. We are not in an age where we can be confident that the laws of reason are the same as the laws of reality, like people in the time of Aristotle believed. If that was the case, we would never have had to abandon Aristotelian physics. It sounded perfectly intuitive but turned out to be false even on the simple idea of inertia, which is a principle that our brains will just not accept because of how we are wired apparently.

So we can just see how reason is limited in understanding physics, then how much more would it be limited in understanding the creation and God. When thinking about the beginning of time and about creation and God our reason actually generates contradictory ideas. It is not satisfied with the idea that the past should be infinite, yet at the same time not with the idea that time has a beginning either. They both sound absurd and we are forced to believe the opposite, yet its opposite is also equally absurd. Furthermore reason demands that the causal chain to the past should not go on forever, but it cannot really make sense of the very idea of a "first cause" either. And also it demands that contingent things must ultimately be explained by a necessary being, but it finds the very idea of a "necessary being" incoherent at the same time. It wants to have God as the creator of time, yet it cannot comprehend the idea that there can be an agent that acts like create, yet has no time dimension of his own, while at the same time in our own experience we can act precisely because we are in time; it is what makes any action possible in the first place.

These examples should be an indicator that we shouldn't really pursue our intuitions to their logical conclusions beyond the limits of the natural world. Because reason wants to follow the train of thought to the end, but apparently it is trying to deal with a realm that doesn't work in human logic after a point. We may feel we are onto something, but that is just an illusion, and we shouldn't take such feelings seriously.

I will admit that atheism comes with its own problems. It is obvious from how we atheists have to either accept positivism or postmodernism and they both have fatal problems. Postmodernism is self-refuting as you explain in your great book Philosophical Foundations for a Christian Worldview. And positivism apparently is just logically immature and on its way to postmodernism if one has to be consistent. Like Wittgenstein matured and abandoned his positivism to become a postmodernist. And that is the end of the road.

So it seems the debate between atheism and theism  is a stalemate. But if you still say that I must reject atheism because it ends up in the absurdity of postmodernism and I must therefore adopt its negation that is theism, well, then I will have to remind you of fatal problems in your worldview such as JEDP theories for the origin of Torah and the academic success of Darwinism which demands acceptance.

Best Regards



I really appreciate your interesting question, KS, and especially your sticking up for me in your atheist forums!

As I read your letter, I thought, Wow, this sounds just like a good Kantian!  (Have you read Kant’s Critique of Pure Reason [1781]?  Your arguments are echoes of his!)  Now scarcely any philosophers today are Kantians with respect to reason’s ability to deliver to us important truths about reality.  Since the demise of Verificationism in the mid-twentieth century, metaphysics, despite Kant’s strictures, has been booming once again.  That suggests that there must be something wrong with your argument.  So let’s talk about it.

First, it seems to me that we have no choice but to take common sense and intuition as our starting points.  I very strongly suspect that even those who claim to place no stock in common sense and intuition in fact rely on them all the time with respect to unconscious metaphysical assumptions.  So when a philosophical viewpoint flies in the face of common sense and intuition (e.g., that the external world does not exist), then we may justly demand a very powerful argument in favor of that viewpoint.  In the absence of some defeater of what common sense and intuition tell us, we are rightly sceptical of that viewpoint and perfectly rational to reject it.  So while the deliverances of common sense and intuition are certainly defeasible and may on occasion need revision, still they are an indispensable starting point which should not be lightly abandoned.

Are the laws of reason and the laws of reality the same, as people in Aristotle’s time believed?  Nothing has happened since the time of Aristotle that has undermined the truths of logic or logic’s applicability to the world.  Aristotle’s logic is called syllogistic logic.  He identified valid argument forms which are still recognized today, e.g.,  All As are Bs; no Bs are Cs; therefore, no As are Cs.  This is an undeniably valid pattern of reasoning.  The principal advance of modern logic over Aristotle’s is that modern logicians came to realize that the premises of syllogistic reasoning like “All As are Bs” have themselves a logical structure which Aristotle’s logic failed to disclose.  A statement like “All As are Bs” has in modern sentential logic (the logic of sentences) the structure of a conditional:  “For any item x, if  x is an A, then x is a B.”  This allows us to make inferences that Aristotle’s syllogistic logic cannot express, e.g., “Whatever begins to exist has a cause; the universe began to exist; therefore, the universe has a cause.”

Formal logic has become a discipline of incredible technical precision and rigor, akin to mathematics.  Indeed, formal logic often goes by the name mathematical logic.  There is nothing in the advance of this discipline that should lead us to doubt reason’s ability to make valid inferences about reality.  Indeed, the development of subdisciplines like modal logic (the logic dealing with the necessary and the possible) and counterfactual logic (dealing with subjunctive conditional statements) has been a great asset in our being able to reason more carefully and rigorously when doing metaphysics.

Don’t confuse Aristotelian logic with Aristotle’s physics!  Aristotle was not only a great philosopher but a natural scientist as well.  As you might expect, his scientific work has been superseded by subsequent science, as more sophisticated instruments for probing the physical world have developed.  As science advanced in our understanding of nature’s laws, Aristotelian physics was replaced by Newtonian physics, which was in turn replaced by Einstein’s physics, which will soon, we expect, be superseded by a quantum gravitational unified physics.  In each successive scientific revolution, the earlier science is not simply abandoned; rather its truths are recast and preserved in the theory that supersedes it and its inaccuracies abandoned.

I hope you can see that none of this gives any cause to doubt the efficacy of human reason in knowing reality; quite the contrary, this is testimony to the incredible power of human reason!

The lesson here for the natural theologian is that he needs to be scientifically literate and to keep abreast of current discoveries and new theories in science.  For that reason I have striven to be responsible in this regard.   I want to have a theology that is scientifically informed and so to present an integrated perspective on reality.

Now you remind us quite rightly that when it comes to subjects like God and creation, we are doing metaphysics, not physics (though physics may provide evidence in support of premises in a metaphysical argument leading logically to a conclusion which is of theological significance).  So if we have plausibly true premises which imply by the standard rules of logic a conclusion of theological significance, why should we resist that conclusion?

Here’s where your Kantianism enters the picture.  You assert, “When thinking about the beginning of time and about creation and God our reason actually generates contradictory ideas.”  You’re claiming that reason leads us into antinomies and so cannot be trusted.  I have responded to this Kantian claim in The Kalam Cosmological Argument (1979), Appendix 2:  “The Kalam Cosmological Argument and the Thesis of Kant’s First Antinomy.”  KS, if you’re serious about getting your reservations resolved, please read that section.  I show that there is no antinomy because there is nothing incoherent about a beginning of time.  Kant thought that in order for time to have a beginning, there had to be a time before time during which nothing existed.  That is a mistake.  All that is required is that there was a time which was not preceded by any prior time.  Far from being incomprehensible, this is precisely the concept of a beginning to time that is used in contemporary astrophysics.  For example, the agnostic cosmologist Sean Carroll characterizes cosmological models which feature a beginning of the universe by saying, "there was a time such that there was no earlier time.”1

Similarly, there’s no problem about postulating a Creator or first cause who exists timelessly sans the universe.  Again, Carroll uses precisely this notion with respect to a boundary condition on spacetime:  “There is no logical or metaphysical obstacle to completing the conventional temporal history of the universe by including an atemporal boundary condition at the beginning.”2  God’s eternal, atemporal state is, as it were, such a boundary condition to time.  God’s act of creating the universe is simultaneous with universe’s coming into being.  So God is atemporal sans creation and temporal since creation.  So where’s the problem?

As for the argument from contingent being to a metaphysically necessary being, what is the difficulty supposed to be?  Many philosophers think that abstract objects like numbers and other mathematical objects exist necessarily.  So where is the incoherence in the idea of a necessary being?  It’s a being which exists in every broadly logically possible world.  (Here the advances in modal logic that I spoke of earlier actually help us to better understand this notion of a metaphysically necessary being.)  So what’s the objection?

These pseudo-antinomies thus do not support the radical conclusion that “we shouldn't really pursue our intuitions to their logical conclusions beyond the limits of the natural world.”  Indeed, when you assert, “reason is trying to deal with a realm that doesn't work in human logic after a point. We may feel we are onto something, but that is just an illusion,” we may justifiably turn the tables and ask you, “How do you know that?  How, on your view, can you know anything about what that realm is like?  How do you know human logic doesn’t work there?  Indeed, how can logic ‘not work’?”  KS, you, like Kant before you, are in the self-refuting position of making metaphysical claims yourself!

The lesson here is not that we should just quit thinking but that we should think even harder.  Listen, KS, you’re not at the end of the road by a long shot.  Even for an atheist, your choices are not limited to Positivism and Post-Modernism.  But why stick with atheism?  Theism offers an intellectually expansive and richly rewarding view, not to speak of its spiritual benefits.

And, KS, what shall I say in response to your final paragraph?  “C’mon Man!”  You know better than that.  You can be a theist and a Christian and accept the documentary hypothesis of the Pentateuch as well as a Darwinian theory of evolution, if you think that’s where the evidence leads (see QoW ## 253, 269).


1 “Does the Universe Need God?” in The Blackwell Companion to Science and Christianity, ed. James B. Stump and Alan G. Padgett (Blackwell Wiley, 2012).  For a discussion of Carroll’s article listen to our three forthcoming Reasonable Faith podcasts.

2 Ibid.