The Laws of Physics and the Laws of Logic

I recently conducted a typically inconclusive discussion with Victor Reppert in the comments section of my post “Can Brains Think?” While I doubt that we will ever agree (each of us in in his mid-sixties), I think I can identify one issue that keeps coming up again and again. Repeatedly Victor pointed to the difference between the laws of physics and the laws of logic. Physical things, like the brain, operate in accordance with the laws of physics. Rational thinking draws inferences in accordance with the laws of logic. As Victor sees it, if, as physicalists (like me) view it, thinking is an activity of the brain, then the thoughts, as physically determined, will only have such content as physical processes determine, which processes are, like all other physical processes, ultimately the interaction of subatomic particles interacting via the fundamental forces of nature.

If, then, the physical is a closed system, with all effects ultimately determined by microphysical interactions, then there is no room for a causal role for reasons, which, Victor holds, there must be if inferences are to be due to reason and not merely in accordance with reason. The latter is a crucial distinction. Victor realizes, of course, that a machine can be programmed to match inputs with outputs in accordance with the rules of logical inference. Indeed, we might be able to train rats to run through mazes in some manner that physically realizes logical rules. That, however, does not mean that machines or rats are rational beings. To qualify as rational, it is not sufficient that our thoughts merely accord with logical rules. Rather, the conclusions inferred must be inferred because they are recognized as rational, and this is possible only if our thoughts have rational causes (e.g. the laws of logic) and not merely physical ones. In short, for Victor, the laws of logic must have causal powers.

I hope that the above accurately summarizes Victor’s views.

To adumbrate a reply (a full reply would require a book), I consider two senses of “rational,” one in the context of an externalist/reliabilist account of warrant and another in the sense, due to W.K. Clifford, of epistemic rights and duties. With respect to the former sense, I argue that accordance with logical rules is sufficient for rationality; with respect to the latter, I argue that an ethic of belief, a lá Clifford, is independent of how reasons are caused to appear to us. Instead, rationality is a matter of living in accordance with our fundamental perceptions of what is reasonable and true.

According to an externalist/reliabilist epistemology, rationality of beliefs is a matter of warrant, and warrant is a matter of having beliefs that are caused in a reliably-truth tracking way. Warrant on this account has nothing to do with subjective awareness of “reasons” or an ability to adduce such reasons in defense of beliefs. What matters is that the beliefs are generated in an objectively reliable manner, as, for instance, produced by a cognitive faculty operating normally in an appropriate environment with respect to appropriate sorts of inputs. For instance, my if my visual faculties are operating normally in broad daylight and with trees present, then my belief that trees are present is warranted if produced by the operation of my visual faculties. Likewise, if my internal logical faculty standardly operates so as to reliably conclude “p” when and only when p is logically inferred from premises that entail p, then my conclusion that p is warranted. On this account, my ability to cite the rules of inference that led me to conclude p are neither here not there. Surely, this is intuitive as well. Someone can think logically that has never had a course in logic and thinks that “Modus Tollens” is the name of the back-up catcher for the Mets.

The inference does not even have to be conscious. Immediately upon noting that the dog did not bark in the nighttime, Sherlock Holmes concluded that the abduction of Silver Blaze was an inside job. He did not have to consciously think “If the dog did not bark, he must have recognized the abductor. The dog did not bark, and so the abductor was known to him. Therefore, the abduction must have been an inside job.” No, the inference was immediate and unconscious, but was surely fully rational (in the sense of warrant) for that. The upshot is that with respect to warrant accounts of rationality, the inference may be warranted, and therefore rational, if it is accomplished by a causal process that reliably produces beliefs in accordance with logical rules. Whether the believer is aware of the rules or can adduce them is irrelevant. Victor’s requirement that the logical rules themselves have causal efficacy does not apply to a warrant account of rationality.

Consider then the account of rationality developed by W.K. Clifford in his famous essay “The Ethics of Belief.” Clifford holds that rationality is a matter of obeying our epistemic duties and committing to beliefs in accordance with our epistemic rights. Epistemic duties are analogous to ethical duties; indeed, I think that they are a kind of ethical duty. We have a responsibility not to commit ourselves to beliefs, such as belief in God, unless we have conscientiously attempted to vet that belief in accordance with the highest standards of rationality. If we commit ourselves to beliefs that are backed by no or insufficient evidence, then surely we have committed an epistemic (and possibly moral sin).

Clifford gives the example of a ship owner who sells berths to poor immigrants desperate find a better life across the ocean. The ship is old and has not been inspected in a long time. However, inspections are costly and would eat into the owner’s profits. Besides, the ship has made the journey many times without mishap, and there is no reason to think that this time will be any different, right? Of course, the ship breaks up in a storm in mid-ocean and the poor immigrants meet a watery doom. The owner collects the insurance and tells no tales. Surely, Clifford says, by failing in his epistemic duty to ascertain the actual condition of his ship, and to insouciantly form the belief that his ship is sound, the owner is as guilty of the immigrants’ deaths as if he had intentionally killed them.

For Clifford, then, and his account of the ethics of belief, rationality is a matter of exercising due diligence in making sure that we have met our epistemic duties, that is, making sure that we do not commit ourselves to beliefs unless we have fairly vetted them with respect to the evidence.

In agreement with Clifford, I accept that that we have epistemic duties just as we have ethical duties (indeed, I would say that epistemic duties are a species of ethical duty). We perform our ethical duty by attempting, insofar as we can, to follow the ethical norms that seem right to us. However, which ethical or epistemic norms appear right to us is not a matter over which we have control. If something seems (morally or epistemically) right to us, then that is how it seems and we have no control over that. It seems to me that torturing sentient creatures for fun is morally wrong. I cannot help having it seem that way to me, and I am glad that I cannot help it. Similarly (having had the benefit of training in logic and critical thinking), inferring “p” from “If p, then q” and “q” seems wrong to me, and I cannot help that it seems wrong to me.

At epistemic or ethical “rock bottom” what seems right or wrong to us is not a matter of our control. Since ethics deals with what we can control (“ought” implies “can”), then we have no duties with respect to how things fundamentally seem to us. Our epistemic or ethical duties, therefore, boil down to conforming our behavior (including belief-behavior) to what fundamentally and involuntarily seems right to us. What ethically or epistemically appears right to me is therefore caused by factors over which I have no (direct) control, and therefore rationality in Clifford’s sense does not apply to those fundamental perceptions. Clifford’s account of rationality therefore can say nothing about the ultimate causes of my beliefs, i.e. my fundamental perceptions of what is true and reasonable. Rationality is a matter of living in accordance with those fundamental perceptions of rational norms.

I have, then, considered two prominent senses of rationality—the warrant version and the ethics of belief version. On neither account do the “laws of logic” cause my belief. Rather, on each account rationality consists of believing or acting in ways that are in accordance with such laws. Maybe Victor has some completely different sense of rationality in mind which he claims requires that the laws of logic have causal powers. I think I have shown that two prominent theories of rationality do not require that we invest abstractions like “the laws of logic” with occult causal powers.