Democritus: An Early Realist Philosopher

The Colorado Freedom Report:  A libertarian journal of politics and culture.

The Colorado Freedom

Democritus: An Early Realist Philosopher

by Ari Armstrong

Part I: Perception


Let us first review some traditional interpretations of Democritus' theory of perception. Robinson titles his section on Democritus' perception "The Illusions of Sense,"(1) indicating that perceptions may be inherently misguiding ("illusion" is "something that deceives by producing a false impression"(2). Some properties "really" belong to bodies; others "only appear to belong to them" (R202). Robinson supports his claims by quoting Democritus: "Sweet exists by convention, bitter by convention, color by convention; but in reality atoms and the void alone exist" (R202). Perception itself contains the illusion, Robinson concludes; it conveys the false "appearance." Robinson claims that Galileo restates Democritus' view by saying, "I think that these tastes, odors, colors, etc., on the side of the object in which they seem to exist are nothing else than mere names, but hold their residence solely in the sensitive body" (R203).

Robinson extends his interpretation by quoting a key passage from Democritus: "There are two forms of knowledge: one legitimate, one bastard. To the bastard sort belong all of the following: sight, hearing, smell, taste, touch. The legitimate is quite distinct from this." "We need not take" this quote, Robinson tells us, "to mean that the senses have no place whatever in knowledge" (R204). The senses can provide inaccurate data, but they provide the only data we have, and must be the starting point of reason, the legitimate form of knowledge.

Robinson claims that Democritus' view of perception is similar to Parmenides':

Parmenides had expressed the same thought [as Democritus and Galileo] in very nearly the same words: "Those things which mortals have established, believing them to be true, will be mere names: 'coming into being and passing away,' ... 'change of place and alteration of bright color'." (R203)

Democritus does not believe that time and space are illusory, says Robinson (and this is the common evaluation of Democritus), but Democritus does make "concessions" about color and taste; he makes a more "radical... break with the world of colors, sounds, and odors" than thinkers such as Empedocles and Anaxagoras (R203).

What is the significance of Robinson's comparison of Democritus with Parmenides? For Parmenides, the truth is that being is "whole and immovable" (R116); so any belief that entertains the possibility of motion (or any changing quality) is necessarily a false belief. A false belief is generated by the person, since no person could possibly obtain such a belief from the world (which is a changeless unity). The very realm of awareness is self-generated. The parallel Robinson makes is that, for Democritus, perceptions can also be false, and contain data that is generated by the person rather than by the external world. Perception can add things to awareness which have nothing to do with reality.

Robinson's interpretation of Democritus is rooted in that of the ancient commentators. Theophrastus (a pupil of Aristotle) says of Democritus' writings,

It seems strange to insist that the same thing appears to all who perceive the same things and to question the truth of these appearances, when he has already said that to people in different states different things appear, and that neither is closer to the truth than the other. (R205)

Strange, indeed. Even contradictory; in the first case all people have the same appearances which are false; in the second case all people have different appearances which are neither true nor false. In Guthrie's words, "Theophrastus claims to see a glaring contradiction in a theory which calls sensations affections of the sense-organ and at the same time accounts for them by atomic shape..."(3)

Sextus Empiricus also finds contradiction in Democritus' thought:

And in his Confirmations, although he had promised to assign the power of conviction to the senses, he is none the less found condemning them, for his says, 'In reality we know nothing for certain, but what shifts according to the condition of the body and of the things which enter it and press upon it'. (G458)

Sextus' criticism is the same as Theophrastus'; on one hand, Democritus seems to support the notion that the senses somehow relay a knowledge of reality to the perceiver (even if we have to get past the false appearances first), and on the other hand he seems to be saying that perceptions are necessarily a product of the perceiver and therefore arbitrary.

Sextus, in his discussion of Democritus, seems to emphasize the strain in the atomist's thought in which perceptions yield false knowledge.(4) Thus, he says that both "Democritus and Plato throw things into confusion by rejecting the senses, annihilating their objects, and holding only to the intelligibles" (G457). In Guthrie's words, Sextus holds that Democritus has a "conviction that the truth is not in sense-impressions themselves..." (G459). Sextus, in reaching this conclusion, relies on the key passage about "bastard" knowledge, which we've seen is also crucial to Robinson's interpretation. It is very helpful in understanding Sextus' views if we take a look at his commentary surrounding that passage:

[I]n his Canons he says that there are two kinds of cognition, one through the senses and the other through the intellect. Of these he calls that through the intellect 'legitimate', and attests its trustworthiness for the judgment of truth, and that through the senses 'bastard', denying that it is free from error in the discernment of truth. To quote his actual words: 'There are two forms of cognition, one legitimate, one bastard. To the bastard belong all these: sight, hearing, smell, taste, touch. The other is legitimate, and separate from these.' (G459)

Though Guthrie says that "Sextus makes a genuine effort to understand Democritus" (G456), we must recognize that Sextus is explicitly unsympathetic to Democritus' ideas. Thus, we might say that Sextus "makes a genuine effort to understand Democritus" only within the context of Sextus' own theories. And what theories would those be? Robinson tells us that

Sextus Empiricus, who lived during the late second century A.D., was a physician of the empirical school... and an admirer of the skeptic Aenesidemus. He produced an exposition of the skeptical philosophy entitled Outlines of Pyrrhonism, followed by two works of criticism entitled Against the Dogmatists and Against the Schoolmasters, in which he quotes the opinions of earlier philosophers on cognition and the reliability of the senses. (R309)

What we should therefore keep in mind is that, 1) Sextus, as a skeptic, is going to distrust any claim that perceptual knowledge is reliable or "truthful," and 2) Sextus, as an empiricist, is going to oppose such appeals to "higher" cognition as Plato and Democritus make. We should not be surprised to find that his criticisms "Against" Democritus is colored by Sextus' own views. Again, the problem as Sextus sees it is that Democritus tries to squirm out of his own conclusions that perceptions can be false.

It seems that it is Sextus' interpretations that have been most influential in the modern interpretation of Democritus' theory of perception, and the notion that the senses can lead to false (unreal) knowledge persists. Guthrie relies heavily on Sextus' interpretations, saying that the senses "give a false picture" (G461). W.T. Jones says that the "subjective component" of perception -- the body -- leads to "illusion." "Illusion" is Jones' term, by which he means that the perception can convey the unreal; it can be false. Jones relies on the "convention" passage for his interpretation.(5) (The other passage commonly relied upon is the "bastard knowledge" quote). Robinson adopts the term "illusion" for his own interpretation (as we've seen) -- retaining the Sextus point of view. Even more recently, Richard Tarnas (in his book published just two years ago) interprets Democritus in the same way. Interpreting the same "convention" passage from Democritus, Tarnas writes:

Much of human experience, however, such as hot and cold or bitter or sweet, derived not from the atoms' inherent qualities but from human 'convention'.... [T]he immediate evidence of one's own senses might be an illusion.(6)


Such is the common interpretation of Democritus. I believe the common interpretation is inaccurate. I believe Democritus' theory of perception has nothing to do with (and is antithetical to) the notion that awareness can be false or illusory.

I have a lot of explaining to do, it seems. But why do I even think that a different interpretation is called for? When I was first reviewing the ideas of the Atomists (from Robinson's text), I, like Theophrastus, thought it quite odd that Democritus accounts for perception "in terms of the constituent atoms" (for instance, "sour consists of atoms that are bulky, jagged, and many-angled, without curves;" R200), but then turns around and claims that perception is mere "convention" and "bastard knowledge."

But then I realized that Democritus is perhaps not quite so odd as everyone has been making him out to be (unless we consider "revolutionary" to be odd). All we must do to see the coherency in Democritus' thought is to strip away Sextus' commentary (I pick on Sextus, though others have obviously contributed) and start over with a more sympathetic interpretation of the troublesome passages Democritus offers us.

Let us start with the "bastard knowledge" quote -- "There are two forms of knowledge: one legitimate, one bastard." Does Democritus mean, as is commonly believed, that "bastard" perception is untrue and illusory? Not necessarily. To take the metaphor back to its roots -- what is a bastard child? It is a real child, presumably, not a "false" child or an illusory one. What is meant by "bastard" is that we do not know from which parents it comes. This is, I think, the meaning of Democritus' comparison -- bastard knowledge is simply knowledge the origins of which we do not know. The legitimate, the finer, form of cognition recognizes perception for what it is at root -- a certain relationship among a group of atoms. Legitimate cognition, so to speak, knows its parents.

Now I proceed to the more difficult passage: "Sweet exists by convention, bitter by convention, color by convention; but in reality atoms and the void alone exist" (R202). What could this mean, other than that perceptions are inherently illusory? Democritus is not (I suggest) saying that the perception itself is "convention," meaning arbitrary and unreal; he is saying that the concept "sweet" is often incorrectly reified. "Sweetness" is not some object out there in reality that hits us with its force during perception. It is only by "convention" that we, say, consider a lemon to be sour in and of itself; really it is just a collection of atoms, many of which are jagged and angular. "Sour," really, denotes nothing other than an experience we have when we eat a lemon.

But our experience of "sour" is not in any way arbitrary or illusory. Rather, our experience is determined by the nature of the stimulating object and by the nature of the perceiver's body. If I am sick, the atoms of my body are in a different state and must therefore react differently with external objects. That doesn't mean the experience is "false," just different.

My interpretation is supported if we take a look at the completed "bastard knowledge" passage; Robinson cuts the quote too short in his text. W.T. Jones includes the entire passage, which continues, "That is, the objects of sense are supposed to be real and it is customary to regard them as such, but in truth they are not. Only the atoms and the void are real" (J92).

Once we understand what Democritus means by "bastard" and "legitimate" knowledge, we are able to grasp other aspects of his thought. Perceptions are real and determined (not arbitrary), but we can EVALUATE perceptions in an incorrect way. That is, we can mistakenly believe that "sweet" and "sour" and "red" are objects in the world rather than perceptual experiences determined by the interaction of atoms.

Misevaluated perception comes neither from "bastard" cognition nor from "legitimate" cognition, but from the misuse of the legitimate. Bastard cognition is perception, which is NOT convention, and legitimate cognition is correct evaluation (which is also not convention); illegitimate cognition is incorrect evaluation, which DOES lead to those false conventions. Or: Bastard cognition is the experience of eating a spicy chili dog; legitimate cognition is recognizing the shapes of the atoms which, when they interact with the atoms of our body, generally create the sensation of "spicy;" illegitimate cognition would be incorrectly assuming that "spiciness" is an entity in itself which exists outside of our experience of it.

So the two levels of Democritus' cognition are 1) perception and 2) evaluation (or, bastard and legitimate). We as atomists must "necessarily say that what appears in sensation is true," as Aristotle says (G457). It is incorrect evaluations that leads to false knowledge (not perception. It is the failure to recognize this distinction which has lead to the misinterpretation of Democritus).

Why is it that evaluations may be wrong but perceptions never are? Perception is just an experience -- and as such antecedes any judgment of "rightness" or "wrongness;" "truth" or "falsehood." Perception just is. (In a similar way, a moss-covered rock is neither true nor false; it is only what it is -- a moss-covered rock.) Evaluating something is technically an experience, too, in that it is the movement of atoms. But it is a very different kind of experience than perception. Perception involves an experience between the body and an external object; evaluation is a self-generative experience which occurs solely within the body. In one sense, an evaluation "just is" the same way that a perception "just is;" both consist of atoms moving in a particular way. But in another sense, we evaluate ABOUT something else, and thus can reach correct or incorrect evaluations, about that other thing. (I can correctly evaluate "sour-ness" to be an event between my body and a lemon, or I can incorrectly evaluate "sour-ness" to itself be an existing entity.) The standard for my evaluations is the actual nature of things.

Before moving on, it may be helpful to place Democritus briefly within his wider historical context. We might view pre-Socratic Greek philosophy as the attempt to reconcile human thought with the principle, "ex nihilo, nihil fit." Before the age of philosophy, the gods created the world. Thales made water his "first principle," which the gods only organized and controlled but did not create (R293). The Ionians expanded this notion, positing an "infinite" substance and explaining motion with the vortex.

The Eleatics pointed out, merely, that if everything is a unitary "stuff" then no motion is possible -- if there is One, then there cannot be many. In other words, not only must matter not "come into existence or pass away," but motion itself cannot come into being or pass away. Thus, Parmenides has eliminated the "ex nihilo." Or has he? Why do men believe that there is change and motion? "[T]hose things which mortals have established, believing them to be true, will be mere names..." (R116). I point out that Parmenides negates the "ex nihilo" on the cosmological level, but at the same time transfers it to the personal level. How can people have false opinions unless those opinions have been generated by the individual -- out of nothing? It is only under Parmenides' assumptions that it is appropriate to speak of perception as "illusory" and "untrue," because it literally "comes of nothing."

Empedocles and Anaxagoras attempted to preserve motion with "Love and Strife" and "Nous," respectively, but they did not solve the Parmenidian problem.

The atomists solved the Parmenidean problem. Leucippus shattered the unitary one with the void and made motion, as well as matter, eternal -- without inception (G397). Democritus purged the "ex nihilo" from the microcosm (humans) as well as the macrocosm -- thus offering a philosophy far superior to that of Parmenides, even by Parmenides' standards. Perception, too, "comes of something."

Part II: Ethics

Perception is not illusory; it is the interaction of two groups of atoms which precedes evaluation. Only evaluations can be illusory, if the evaluations are incorrect, say, about the nature of a perception. But what about morality? What about feelings about what is good and bad, and what about ethical theory? Is morality arbitrary?

The experience of happiness -- the aim of morality for Democritus -- is a very different kind of experience than perception. Perception is an experience which involves the external world, but "happiness and unhappiness belong to the soul" (R222) and not in any way to an external object; happiness is not the product of an experience with an external object. (Both evaluative activity and the experience of happiness are, in this way, generated within the body.)

Because happiness and perceptions are not comparable experiences, Jones is wrong in comparing them. Jones, recall, thinks that Democritus' perception can be illusory, and he argues that, in a similar way, morality can be illusory:

The Atomists made a place for knowledge in the field of ethics only by reducing values -- that is, the felt, appreciated, enjoyed experiences themselves -- to an illusion. A necessary illusion, it is true, but still an illusion.... The goodness of moderation that Democritus experienced is as much an illusion (a necessary illusion) as the greenness of cabbage he saw. (J103)

We clearly (within the atomist framework as I interpret it) may not link illusory moral values to illusory senses (which is an oxymoron, anyway), as Jones tries to do.

But it may still be the case that morality is arbitrary, and illusory in this sense (and thus different from perception). Robinson makes this argument. Robinson contends that, while Democritus himself would not support any unjust action, his theory leaves room for an arbitrary morality: "self-interest may dictate a line of action directly contrary to what is permitted by usage or law," and even "without regard to consequences" (R236).

Does Democritus' theory contain this weakness, as Robinson claims? If we hold "happiness," or, to express the idea in other words, "cheerfulness or contentment" or "well-being" (G492), to be a standard of value and therefore of right action, then we might be able to escape arbitrary values and moral feelings. Democritus clearly does believe that morality is not necessarily arbitrary but may be based on a standard. He compares the state of the soul to the state of the body; the body can be damaged (done an objective harm, according to a standard) by sickness, and the soul can be damaged by "sick ethics." "[D]isease occurs in a household, or in a life, just as it does in a body," he says (R221). We may deplore bad ethics as we do a disease, and, just as we call a physician to cure a body, so we ought to call a moral philosopher to cure a "life."

But we are far from freeing Democritus from moral arbitrariness. What is happiness for me and what is happiness for you are probably two very different things. And what do we do with the person who seeks happiness by hurting others? If happiness is the standard of ethics, but every person has a different feeling about what makes them happy, then the "standard" of ethics really has no purpose other than to tell people to do whatever they feel like doing. Perhaps Robinson is correct in saying that "self-interest" (the pursuit of happiness) might lead us to any act. Democritus speaks of curing the soul with wisdom (R222), but isn't Democritus' moral "wisdom" really just describing what Democritus likes to do?

The answer to the rhetorical question posed above is "no." Democritus admits that morality is arbitrary for some people, but, he would add, wisdom can lead to an objectively better morality. Recall that, while perception cannot be discussed in terms of "falsehood," evaluations may be. The reason morality is arbitrary for some people is that their feelings of happiness are in part dependent upon false evaluations. In other words, if I merely believe that some activity will make my life better off -- say, robbing a bank -- then I will derive some satisfaction from successfully completing the activity. What is the problem with succumbing to this sort of "self-interest"? The problem is that the evaluation -- "robbing a bank will make my life better" -- is false and illusory. In the long run, robbing a bank will probably get one shot, sent to jail, or placed in permanent hiding. A sick body self-destructs, and so does a sick soul. Wisdom, by deciding which acts are in one's ACTUAL self-interest, also modifies that in which one finds happiness. Notice the hierarchy: in perception, the experience always precedes evaluation. In morality, the evaluation may precede the experience.

There is one loose end I must yet tie up. Here is the dilemma for ethical theory: since every event is determined by previous events, and since all human actions are events, then what use is it talking about ethical theory? Aren't a person's actions the product of necessity? If I may answer for the atomist point of view, I point out that ethics can itself become a part of the causal sequence. As I develop an ethical theory, and teach this theory to others, I am helping to determine those acts which people do (again, by modifying that in which people find happiness). "The cause of going wrong is ignorance of the better" (R225). In this respect, Democritus' ideas are very much the same as Socrates' and Plato's. In the apology, Socrates says, "[N]othing else [than ignorance] would make me commit this grave offense [of spoiling the character of one of my companions] intentionally." As Edith Hamilton sums up, "No man does evil voluntarily[, because] virtue is one with wisdom [and] wickedness has its roots in ignorance."(7)

I must note concerning those writers whom I've criticized: though I have disagreed with Robinson, Guthrie, Jones, and others, each made a valuable contribution to my understanding of Democritus. And all praised Democritus for his accomplishments, a sentiment that has worn off on me -- Guthrie, for instance, says that we may admire Democritus' "boldness and consistency" (G454). If my interpretation of Democritus is correct (which seems quite likely to me), then there is all the more reason to praise the atomist and learn from his teachings. Democritus in many ways paved the way for modern thought; he even foreshadows David Kelley, a modern epistemologist, who claims that "perception by its nature is the awareness of external objects", but "that the forms in which we perceive both primary and secondary qualities are relational phenomena; there are no entities of which they are intrinsic qualities"(8) (a notion that affected my interpretations of Democritus). While Robinson claims that Democritus' "conclusions were insupportable" and left room for Plato's corrections (R289), perhaps Walter Kaufmann says it better: Democritus' "philosophy may be viewed as an important alternative to Platonism."(9)


(1)John Mansley Robinson, An Introduction to Early Greek Philosophy, Houghton Mifflin 1968, p201. I shall hereafter cite passages from this work with an "R"; e.g., (R201).

(2)The Random House Dictionary of the English Language, Random House 1969.

(3)W.K.C. Guthrie, A History of Greek Philosophy II: The Presocratic Tradition from Parmenides to Democritus, Cambridge University Press, 1965, p439. I shall hereafter cite passages from this work with a "G"; e.g., (G439).

(4)If I use "perception" and "sensation" interchangeably, it is because there is no need to distinguish between the two for our purposes. I recognize that in other contexts it is necessary to differentiate the concepts.

(5)W.T. Jones, A History of Western Philosophy: The Classical Mind, 2nd ed., Harcourt Brace Jovanovich 1970, p92. I shall hereafter cite passages from this work with a "J"; e.g., (J92).

(6)Richard Tarnas, The Passion of the Western Mind, Harmony Books, 1991, p22.

(7)Ed. Edith Hamilton and Huntington Cairns, The Collected Dialogues of Plato, Princeton University Press 1989, p12, 308.

(8)David Kelley, The Evidence of the Senses: A Realist Theory of Perception, Louisiana State University Press 1986, p143, 118.

(9)Walter Kaufmann, Philosophic Classics: Thales to Ockham, 2nd ed., Prentice-Hall 1968, p43.

Back to Main Page

The Colorado Freedom