Jonathan Barnes argues that Democritus’ atomistic beliefs are challenged by his skeptical beliefs concerning the possibility of human knowledge. But more precisely, Democritus’ skepticism is concerned with the validity of sense-perception—and more broadly, the validity of a posteriori judgments. It is widely held that propositions which are known a priori are necessarily true, arguably at least. But propositions that are known a posteriori are contingent because they depend on external conditions (external to the mind and pure reason), which may change in time, for example, rendering the proposition false. ‘Necessary’ in this sense means that something is true in every possible world, whereas by ‘contingent’ I mean that something is true only in some possible worlds. By arguing that Democritus’ epistemic justification for atomism is a priori, I may appear to advocate that atomism must be a necessary truth. My argument rather is that Barnes’ begs the question of whether atomism should be embraced by presupposing a posteriori as the only valid epistemology. Then I proceed to explain under what conditions might atomism be embraced by ordinary, rational persons.

We know that if atomism is true, perception is illusory. This is easily understood, for if the world we see W is composed of elementary particles that cannot be perceived by the senses, then perception is not a valid means to the atomic view of the universe. But suppose what would be the case if atomism were false. The mutually exclusive alternative would be that we live in the world W’ in which matter is infinitely divisible. Unfortunately, this cannot be perceived either. So it should be said that if ‘infinitely divisible matter’ is true, perception is therefore also illusory, for how is perception able to perceive this fact? Democritus obviously did not “perceive” or “experience” an indestructible magnitude. Democritus, after all, was not able to purchase an electron microscope at the Agora, and even if he could the epistemological uncertainty remains the same. This means that the classical argument for atomism was an a priori construction, since neither W nor W’ can be known by experience, which is an important point because it means that Democritus would have maintained that atomism is known without the aid of experience, thus without the aid of perception. But Barnes’ overlooks this point.

Barnes says in one passage (p. xlvii) that if we “distrust our senses, how can we say anything about the structure of reality?” This should send up a red flag, for Barnes presupposes that the structure of reality can only be known through sense-perception, through a posteriori judgments. Aside from the question of solipsism—which is also resolved a priori, since illusions presuppose reality—perhaps the next question rational animals ask concerns what reality is, and what reality is made of—hence the question of atomism. It is important to see that our inquiry has not yet required us to call upon our a posteriori faculties. We do not need to. Barnes’ assumptions are false. Since Democritus’ epistemology is a priori, we can distrust our senses and say a lot about the structure of reality as well.

But does a priori reasoning solve for atomism? Consider this analogy. If I am a physicist and I believe that my colleagues and I have arrived at an indestructibly smallest particle P, it would still be true that, with sufficient government grants, this particle could be smashed into something smaller, (P-1) for example. It is not unlike the mathematician who proclaimed he had arrived at the largest number, for it is still true that an even larger number exists, (x+1). But no sum of experience will arrive at infinity. No one could possibly experience infinity, or with regard to the structure of matter, experience zero. In fact, only induction, which is a kind of epistemological shortcut, has something to say on the subject. Thus speaks the physicist, “Since we have been smashing larger particles into smaller and smaller particles, inductive logic suggests that particles are most likely infinitely divisible.” But even induction is not conclusive. On the other hand, if we accept atomism as an a priori truth like the fact that there is no largest number, then a priori alone is sufficient.

According to Barnes, however, I am stuck in the unhopeful position of not being able to trust my senses, and therefore unqualified to expound on the structure of reality. My poor perception cannot arrive at the answer to these matters. But alas, if perception is illusory, why embrace anything at all? This is a logical extension of Barnes’ obstruction that, “If perception is illusory, why embrace atomism?” It is my point that if Barnes’ obstruction is warranted for atomism, it is also warranted for any alternative to atomism. Why embrace anything, for that matter? Having been reduced to an absurdity, why embrace Barnes’ own argument?

This is, after all, Barnes’ Waterloo. He assumes one can only pass a posteriori judgments on the nature of matter. I also have my rational mind to work with, the atomist rightfully contends, where perception does not impede my judgment. In fact, if my perception is so flawed, then I might as well embrace my a priori judgments—and if the a priori judgments lead me to an atomic theory of the universe, then I should actually embrace it.

Atomism, of course, is either true or untrue. Moreover, to embrace a theory one must to some extent be convinced that it is valid. At the very least, I am prepared to embrace a theory on the basis of pragmatism. That is, if I believe embracing atomism is useful in explaining other phenomena, for example, if I work in the pharmaceutical industry or perchance am a particle physicist at Fermilab, and for certain reasons need to suppose that matter is indestructible at some level, I will. Atomism, then, can serve as a means to practical ends. But, pragmatism aside, how might I be convinced that to some extent atomism actually is valid?

Just as I have stated that a posteriori cannot solve for atomism, it may be objected that a priori cannot solve this problem either. Perhaps a more thorough argument would lead us to the world W’ in which matter is infinitely divisible. The atomism of W, after all, seems to have been superceded by W’ in the realm of the inductive sciences. At this level of argument I am prepared rather to agree with the objector. An a priori argument would seem to suggest that W’ is the actual world in which we live, although this is hardly conclusive. But Barnes does not seem to be concerned whether atomism is true as a scientific theory, and for that matter, neither am I. Still, we want to answer the question of why we should embrace atomism.

Perhaps it should not be embraced because atomism is thought to presuppose the validity of sense-perception and “to gain support from its capacity to explain the phenomena of perception,” as Barnes puts it. But in what sense, does atomism presuppose the validity of sense-perception? That reality exists, as I have shown, does not presuppose sense-perception; Descartes’ brain in a vat is able to determine whether reality exists. That sense-perception as a phenomenon exists, it is explained by atomism only insofar as atomism is true and explains all phenomena. That sense-perception especially is explained by atomism is a non-unique argument. Moreover, to emphasize my point, if atomism is true, a posteriori judgments are misleading. So why ought an atomist be required to presuppose the validity of something a posteriori to prove an a priori construct?

That atomism may or may not be true is an open question. That atomism may be defended by Barnes’ assumptions has been shown to be true through my own argument. Since Barnes’ believes that ordinary, rational people ought to believe his arguments, and since his arguments do not stand, then this entails that ordinary, rational people ought to rather believe atomism provides a rational alternative to the idea that matter must be infinitely divisible. Of course there may be other reasons for supposing that atomism ought to be embraced, in the scientific, linguistic, philosophic communities or elsewhere. There may be a priori arguments in favor of the necessary truth of atomism, although it is certainly difficult to see what the reasons might be. Hence my tentative conclusion: if Barnes’ skepticism about the atomists’ view on perception is rational, so is Democritus’ skepticism about the possibility of human knowledge.

According to Barnes, however, I am stuck in the unhopeful position of not being able to trust my senses, and therefore unqualified to expound on the structure of reality. My poor perception cannot arrive at the answer to these matters. But alas, if perception is illusory, why embrace anything at all? This is a logical extension of Barnes’ obstruction that, “If perception is illusory, why embrace atomism?” It is my point that if Barnes’ obstruction is warranted for atomism, it is also warranted for any alternative to atomism. Why embrace anything, for that matter? Having been reduced to an absurdity, why embrace Barnes’ own argument?

This is, after all, Barnes’ Waterloo. He assumes one can only pass a posteriori judgments on the nature of matter. I also have my rational mind to work with, the atomist rightfully contends, where perception does not impede my judgment. In fact, if my perception is so flawed, then I might as well embrace my a priori judgments—and if the a priori judgments lead me to an atomic theory of the universe, then I should actually embrace it.

Atomism, of course, is either true or untrue. Moreover, to embrace a theory one must to some extent be convinced that it is valid. At the very least, I am prepared to embrace a theory on the basis of pragmatism. That is, if I believe embracing atomism is useful in explaining other phenomena, for example, if I work in the pharmaceutical industry or perchance am a particle physicist at Fermilab, and for certain reasons need to suppose that matter is indestructible at some level, I will. Atomism, then, can serve as a means to practical ends. But, pragmatism aside, how might I be convinced that to some extent atomism actually is valid?

Just as I have stated that a posteriori cannot solve for atomism, it may be objected that a priori cannot solve this problem either. Perhaps a more thorough argument would lead us to the world W’ in which matter is infinitely divisible. The atomism of W, after all, seems to have been superceded by W’ in the realm of the inductive sciences. At this level of argument I am prepared rather to agree with the objector. An a priori argument would seem to suggest that W’ is the actual world in which we live, although this is hardly conclusive. But Barnes does not seem to be concerned whether atomism is true as a scientific theory, and for that matter, neither am I. Still, we want to answer the question of why we should embrace atomism.

Perhaps it should not be embraced because atomism is thought to presuppose the validity of sense-perception and “to gain support from its capacity to explain the phenomena of perception,” as Barnes puts it. But in what sense, does atomism presuppose the validity of sense-perception? That reality exists, as I have shown, does not presuppose sense-perception; Descartes’ brain in a vat is able to determine whether reality exists. That sense-perception as a phenomenon exists, it is explained by atomism only insofar as atomism is true and explains all phenomena. That sense-perception especially is explained by atomism is a non-unique argument. Moreover, to emphasize my point, if atomism is true, a posteriori judgments are misleading. So why ought an atomist be required to presuppose the validity of something a posteriori to prove an a priori construct?

That atomism may or may not be true is an open question. That atomism may be defended by Barnes’ assumptions has been shown to be true through my own argument. Since Barnes’ believes that ordinary, rational people ought to believe his arguments, and since his arguments do not stand, then this entails that ordinary, rational people ought to rather believe atomism provides a rational alternative to the idea that matter must be infinitely divisible. Of course there may be other reasons for supposing that atomism ought to be embraced, in the scientific, linguistic, philosophic communities or elsewhere. There may be a priori arguments in favor of the necessary truth of atomism, although it is certainly difficult to see what the reasons might be. Hence my tentative conclusion: if Barnes’ skepticism about the atomists’ view on perception is rational, so is Democritus’ skepticism about the possibility of human knowledge.