Teaching Theory of Knowledge

1. Theory of Knowledge in Ancient Philosophy

      The following materials would be suitable for an intermediate or advanced course (for students who have had some experience reading and discussing philosophy, and who know some logic -- preferably, predicate calculus). The suggested readings contain enough to introduce several major themes in epistemology (some of which relate also to philosophy of science), and to introduce the student to the ways in which the ancient Greeks went about doing philosophy in one broad area of their major concerns.

      Some of the topics sketched below could also be introduced as modules into standard courses on theory of knowledge, partly to provide helpful formulations of contemporary problems, and partly to provide interesting alternatives to present day ways of conceptualizing the issues.

A note on translations: Except where noted below, we recommend:

Plato. Collected Dialogues. Hamilton and Cairns (eds.). Princeton. 1971.

Aristotle. The Complete Works. Barnes (ed.). Princeton, 1984.

Sextus Empiricus: Vol. 1. Bury (tr.). Leob Classical Library, Harvard, 1961.

1.1 Knowledge and Extreme Relativism.
1.2 Perception and Objects of Perception.
1.3 Scientific Explanation.
1.4 The Problem of the Starting Point.

1.1 Knowledge and Extreme Relativism. As represented by Plato, the Sophists equated truth with appearance, and argument with persuasion: for any subject, S, and any belief, p, the Sophists held that p is true for S just in case things appear to S' as p represents them. But some beliefs may be pragmatically better than others. If p is true for S, and not-p is true for another subject, S', they are both right, but acting on one of these beliefs will further an agent's interests better than acting on the other. The purpose of education and argument is to persuade people to hold the beliefs that are best for them by making things appear to them in ways that best serve their interests. In Theaetetus I, Socrates argues that if truth is relative in this way, no belief can be of any more use than its denial as a guide to action, or as a premise in theoretical reasoning. If this is so, and if some beliefs are in fact better for theoretical and practical purposes than their denials, then there must be beliefs whose truth or falsity depends on facts which are external to, and independent of the way things appear to the subject, or to anyone else. And, Socrates holds, the purpose of reasoning and inquiry is to find out whether a given belief is true by uncovering good reasons for holding or rejecting it --not, as the Sophists held, merely to persuade.

      Theaetetus I is the earliest, concerted argument for realism, and for distinguishing between well and ill justified beliefs. It is highly sophisticated, and it differs interestingly from present day explanationist arguments for realism. Rather than arguing that realism is the best explanation for the pragmatic virtues of certain beliefs, Socrates tries to show that the Sophist position is internally incoherent.


Plato. Theaetetus I. (Theory of perception, and realism). With commentary by John McDowell (tr.). Oxford. 1972.

Plato. Gorgias. (The Sophist program for anti-realism).

Lee, E. "Hoist on His Own Petard." In Lee, Mourelatos, and Rorty (eds.), Exegesis and Argument. Assen, 1974. (On the Theaetetus).

      If Socrates' argument succeeds, it is natural to ask what are the kinds of things which can be known, what it is to understand them, and how claims about, and putative explanations of them can be assessed for adequacy and accuracy. Except for skepticism, the most characteristic issues in ancient Greek theory of knowledge arise from these questions. The following is an incomplete, but representative sampling.

1.2 Perception and Objects of Perception. Plato's theory of perception is set out in the Theaetetus and the Timaeus. His view, a harbinger of the distinction between primary and secondary qualities, is that the senses provide nothing more than appearances of things which cannot themselves be perceived, that the objects of perception have no fixed natures, and therefore, that what perception gives us (i.e., appearances) cannot be known. The objects of knowledge are the forms. Perceptibles can be partially understood on the supposition that they are the products of interactions between geometrical particles which constitute the body and the physical things it confronts. To the extent that these particles resemble the geometrical forms, they can be understood as approximations of facts which can be inferred from pure geometry.


Plato. Republic V: 474483. (Knowledge and belief; forms and perceptible things). The Grube translation (Hackett, 1974) is more readable than Shorey (in Hamilton and Cairns).

Plato. Republic VI: 506 to the end. (The divided line).

Plato. Republic VII: 523-533. (The cave, mathematics and dialectic).

Plato. Timaeus.

      The opposing view is Aristotle's: the objects of knowledge are perceptibles, and abstractions from perceptible objects. The best introduction to this view, and to Aristotle's philosophy in general, is John Ackrill, Aristotle: The Philosopher, an accessible, reliable, and remarkably readable survey which mcludes good translations of a number of crucial passages, and a lucid, well-organized commentary.

      Aristotle's theory of perception, which develops further the prototypical Greek version of the primary/secondary quality distinction, and which greatly influenced Locke and the others who originated the versions with which we are most familiar, is to be found in his De Anima. He objects to the Platonic idea that only the forms qualify as objects of knowledge are included in the Physics and Metaphysics.


Aristotle. Physics I, II. (On scientific explanation).

Aristotle. De Anima. (On perception).

Aristotle. Metaphysics. (Objections to Plato on the objects of knowledge, and scientific knowledge).

Ackrill, J. Aristotle, the Philosopher. Oxford, 1981.

1.3 Scientific Explanation. Many ancient Greek epistemologists and philosophers of science believed that to understand nature is to be able to explain the features, and motions of things, and what they do to other things by appeal to their hidden constituents. Thus both Plato and Aristotle held that the heat in a thing explains its ability to warm up other things. This kind of explanation, and its early history, are discussed by Moravcsik in "Herakleitean Explanation". An early account of it, along with a discussion of teleological and other sorts of explanation, is developed in the Phaedo. Aristotle's account of explanation (the four "causes") is found in the Physics I, II. Parts of Animals illustrates Aristotle's program for using teleological explanation in biology. Posterior Analytics is Aristotle's most systematic treatment of adequacy conditions for scientific knowledge. In effect, what the Greeks mean by 'knowledge' is not justified true belief (with or without a fourth condition). It is understanding constituted by adequate explanation. Hence the relevance of ancient Greek philosophy of explanation to theory of knowledge.


Plato. Phaedo: 97-107. (On scientific explanation).

Plato. Timaeus. (Plato's atomism).

Moravcsik, J. "Herakleitean Explanation." In Robb (ed.), Language and Thought in Early Greek Philosophy. La Salle, 1983.

Aristotle. Parts of Animals. (On teleological explanation). De Partibus Animalium I and De Generatione Animalium 1. With notes by D. M. Balme (tr.). Oxford, 1985.

Aristotle. Posterior Analytics. (On scientific knowledge and the problem of the starting point).

1.4 The Problem of the Starting Point. Plato's Meno version: In order to discover the correct answer to a question, the subject must be able to recognize the truth of the correct answer (if s/he finds it) and the incorrectness of alternative answers. But recognition of correctness would seem to require comparing candidate answers to the correct answer, and if that is what it takes, no one could discover the correct answer unless s/he already had it. How then can a correct answer be discovered? Plato's answer is the doctrine of recollection. A close relative of this problem is posed by Aristotle in the opening sections of Posterior Analytics.

      Aristotle's version: Posterior Analytics II argues that the establishment of the correctness of an explanation requires deductions from first principles which are both true and known with certainty. But if our knowledge of a first principle required us to be able to deduce it, we should have to already know true principles to use as premises in the deduction. And if the knowledge of these latter principles required their deduction, it looks as though we are on our way to a vicious regress or circle. Aristotle's solution to this, and his version of the Meno problem, is developed in the closing chapters of Posterior Analytics II. For commentary, see Robert Turnbull.


Plato. Meno. (On the problem of the starting point).

Aristotle. Posterior Analytics. (On scientific knowledge and the problem of the starting point).

Turnbull, R. G. "Physics I: Sense Universals, Principles, Multiplicity, and Motion." In Machamer and Turnbull (eds.), Motion and Time, Space and Matter. Ohio State University Press, 1976. (On 'Posterior Analytics').

Cross Reference

For a discussion of contemporary arguments for realism to contrast with the Theaetetus account, see "Realism" in the "Contemporary Sources" section.