I am grateful to Professor Miller for his thoughtful engagement with my “eye test” argument regarding evolutionary theory, and I appreciate the opportunity to clarify certain aspects of this argument that may have given some a mistaken impression of their inspiration or tendency. The liveliness of the debate over evolutionary theory since Darwin’s time undoubtedly attests, at root, to a passionate concern for the truth shared by the debate’s participants. This passionate concern is a credit to Professors Miller, Barr, and Kemp, regardless of whether they or I am more correct on the point under contention.

Turning to Miller’s recent critique in particular, I believe he has misunderstood my argument as dismissive of or even opposed to various modern scientific discoveries such as those in genetics or cell biology.

Using the example Miller provides in pursuance of this point, I entirely agree that the progressive understanding of elephants in terms of their DNA, cellular composition, etc. helps us to understand elephants better than we did before. And I entirely agree, moreover, that ordinary observation falls far short of disclosing a clear and complete knowledge of anything. In fact, as Miller does not quite say, it is evident that neither casual observation nor modern science will ever be capable of attaining perfect knowledge of anything. No matter how much we eventually come to know about elephants, we could never assert with any level of confidence that there is nothing more to be discovered. Scientists are in no danger of driving themselves out of business.

Scientists have, however, successfully driven the layman—and this includes the highly educated philosopher—out of the business of knowing something about the world. The primary point of disagreement between Miller and me lies not in acknowledging this state of affairs, but rather in how we evaluate it. Miller is not in the least troubled by his description of the ordinary layman’s given mode of understanding as a “current best guess about what the world is like,” a guess that must await the investigation of scientists for confirmation, clarification, or complete overturning.

Start your day with Public Discourse

Sign up and get our daily essays sent straight to your inbox.

And neither am I—in a sense and up to a point. At the risk of drawing Miller’s ire, the beginning of Aristotle’s Physics may be helpful in illustrating my position here. Aristotle agrees in one way with Miller’s depiction of the ordinary layman’s sort of understanding, describing this understanding as “confused” and “vague,” and asserting that “we do not think that we know a thing until we are acquainted with its primary conditions or first principles, and have carried our analysis as far as its simplest elements.” So far, so scientific.

At the same time, though, Aristotle holds that the admittedly confused and vague understanding with which we all start must be more certain, or better known to us, than what we come to understand through subsequent investigation. We can’t avoid starting the process of knowledge acquisition with what we know best. The certainty of later steps depends upon the certainty of prior steps, just as the stability of a structure depends upon the stability of its foundation.

Science is concerned with knowledge—this is, in fact, the very meaning of the word “science.” As the history of science abundantly attests, the process of knowledge acquisition is an incremental one. In the famous words of Isaac Newton, advances in science proceed by “standing on the shoulders” of previous science. We only know the most recent scientific advances as well as (or, more accurately, less well than) we know the various bodies of knowledge assumed to be valid by the scientist. This dependence of new knowledge on old knowledge stretches back through the history of science, with the new knowledge always only as certain as the old knowledge upon which it is built.

This long scientific chain can only be logically sustained if it has some beginning point: something we know without having to know anything else beforehand. Anything we subsequently come to know depends entirely upon the certainty of the knowledge with which we necessarily begin. The certitude of the scientist is, in other words, ultimately dependent upon and derivative from the certitude of the layman.

Given this basic epistemological point, it is misguided to characterize the layman’s understanding of what the world is like as mere guesswork. His understanding may be general, vague, or even confused, but it is much more than a guess; it must be, if the scientific enterprise is to get off the ground at all. There is no straightforward “guarantee” that our ordinary concepts “mirror the structure of reality” in a perfect or clear way, but if instead of a clear mirror we have not a cloudy mirror but a piece of plywood, this poses a serious problem for the scientific endeavor as a whole.

Take the example of the elephant. The layman—and even scientists were themselves laymen once—knows what an elephant is in a general and vague sort of way: it’s a big gray four-legged tusked and trunked herbivore. Scientists are able to tell the layman all sorts of things about elephants that he could never observe himself, things that add greatly to his knowledge of elephants. If, however, these scientists began to tell the layman that elephants weren’t actually four-legged, that they were purple instead of gray, or, most radically, that there wasn’t any such thing as an elephant to begin with, the layman would be justified in raising an objection. “What,” the layman may ask, “is the good of a scientist who leaves me with less knowledge than I had when I started?”

Miller’s example of the metamorphosis of the caterpillar into the butterfly illustrates my point perfectly. The ordinary observer would, indeed, think that a caterpillar and butterfly were two stable and distinct species, and he would be wrong. A naturalist does such an observer a valuable service by pointing out that, in fact, the caterpillar changes into the butterfly and consequently there is no distinction in species between the two. But what if the same naturalist, deciding to use this as a teaching moment, went on to point out that a properly situated observer would similarly see a fish turn into a bird, a monkey turn into a human being, and everything generally turn into everything else in a manner similar to the caterpillar and butterfly? At this point, the same observer who was fascinated by the naturalist’s account of the butterfly might be forgiven for shrinking back in disbelief.

The job of science is, or should be, to explain and account for, more and more accurately, our ordinary experience of the world—including experience in the broader sense outlined in my last treatment of this topic. While doing so will occasionally lead to a revision of certain aspects of our initial experience, it shouldn’t lead us to disavow the reliability of our starting points in general.

The meanings of names like “elephant” don’t need to reflect perfectly or capture entirely the “essence” of things in the world, but if there isn’t some substantial measure of correspondence between the meaning of the name and the thing—between “elephants” and elephants—we have a serious problem. While a disconnect between language and reality is a premise that makes for great movies (think of The Matrix or Inception), it isn’t one that makes for good science. We have to know what an elephant is, even if vaguely and confusedly, in order to learn anything about elephants, and not the other way around. What, after all, would be the use of studying elephants if they didn’t really exist? It is into such difficulties that Miller’s and Darwin’s radical disjunction between the understandings of the layman and the scientist carries us.

While Aristotle’s Posterior Analytics doesn’t refute evolutionary theory, it does indicate a piece of evidence that the layman, who has long been ignored by modern science and discounted as a legitimate knower of the world in which he finds himself, may bring to bear in testing the validity of evolutionary theory.