LECTURE #3: Lewis on Naturalism

I. "The Cardinal Difficulty of Naturalism" from Lewis's On Miracles

A. The distinction between the two senses of 'because': cause/effect, ground consequent.

Popularly, the two seem incompatible: if a belief is caused, then it was not deduced.

This was the original version of the argument, presented to the Socratic Society at Oxford. Philosopher Elizabeth Anscombe replied, arguing that caused does not imply irrational.

B. Lewis's reconstructed argument.

1. If naturalism (or any other thing) is known, then human inference must be (in general) reasonable.
2. If the presence or absence of rational grounds has no causal influence on the inferences we draw, then the our inferences are not in general reasonable.
3. If naturalism is true, then the presence (or absence) of rational grounds cannot have any causal influence on anything (including the inferences we draw).
4. Something can be known only if it is true.

Crucial premise is 3. Why can't some law of association, or the influence of natural selection or of trial-and-error learning, establish the needed causal connection between what is rational (what principles of reason are true) and our habits of inference?

C. Some possible arguments for 3, drawn from pages 17-22.

I. The "sole causation" argument.

1. An act of knowing must be determined, in a sense, solely by what is known.
2. If naturalism is true, then no act of knowing can, in any sense, be caused solely by what is known.

Key idea: naturalism cannot draw the required distinction between (1) those causal elements that constitute the object of knowledge, and (2) those causal elements that constitute the normal functioning of the mind in its normal environment. What naturalism lacks is an adequate account of normativity: how the mind is supposed to work.

II. An early version of Plantinga's against a Natural-Selection account of reason.

1. If naturalism is true, then the only thing that can have any causal influence on our habits of inference is natural selection, which favors those habits that are advantageous in the struggle for existence.
2. There is no conceivable connection between what is true and reasonable, on the one hand, and what is useful or advantageous.

III. The Unavailablity of the Axioms of Reason to the Naturalist

1. The inferences we draw are reasonable only if they are guided by rational axioms (like the principles of logic, mathematics, and metaphysics -- e.g., the principle of causality) that are known.
2. These rational axioms can be known only if the truth of the axioms is accessible to our minds.
3. If naturalism is true, then the truth of a proposition is accessible to our minds only if that truth can be learned from experience (broadly construed).
4. The truths of the axioms of reason cannot be learned from experience (since our learning anything from experience already presupposes the truths of those axioms).

IV. The Unavailablity of Speculative, Metaphysical Knowledge to the Naturalist

1. If naturalism is true, then the only truths that can possibly have an influence on our habits of inference are those concerned with the world of practical experience, i.e., not highly theoretical or speculative truths, like those of metaphysics.
   Consequently, if naturalism is true, no highly theoretical or speculative truths can be known (see argument above).
2. Naturalism is itself a metaphysical (highly theoretical and speculative) proposition.
   Consequently, if naturalism is true, it cannot be known.

D. Lewis's argument that the naturalist cannot defeat these arguments

1. Suppose the naturalist appeals to some argument linking biological usefulness and truth.
2. In appealing to such an argument, the naturalist is assuming the value of human reasoning.
3. But this is just what is at issue, so the naturalist would be begging the question (reasoning in a circle).

Because, unlike the case of naturalism, theism doesn't provide any basis for even a prima facie challenge to the value of our reasoning.