Bryan’s short book studies the term ἐοικώς/εἰκώς in three epistemically-significant passages: Xenophanes B35, Parmenides B8.60-1, and Plato Tim. 29b1-d3. It argues that each author intentionally modifies his predecessor’s use of the term, and that in these passages the term, meaning something like “likely” or “fitting”, never means simply “defective relative to truth, certainty, or knowledge”. The book’s prose and argumentation are completely transparent, the scope is well defined, and the evidence is always plausible and straightforward. The linguistic analysis is fair, and the inferences drawn to the philosophers’ respective theories of knowledge are persuasive. Bryan has not written a complete treatise on εἰκώς, but she has made the idea of writing a new one seem urgent; her work will provide an important point of departure. This review briefly discusses the four chapters.
“Xenophanes’ Fallibilism” (6-57) is structured by an investigation into the best translation of B35: ταῦτα δεδοξάσθαι μὲν ἐοικότα τοῖς ἐτύμοισι. The chapter settles on the following formulation (retaining the ambiguity between the “encouraging” and “classifying” readings of the infinitive): “these [teachings] have been believed to be like the truth.” Being like the truth is a good thing, since it is truth that it is like, and to which it could even be identical; but Xenophanes also wants to admit to the inherent uncertainty in any teachings. Bryan’s investigation involves appeal to Homeric uses of the term; the contrast between ὁμοῖος and ἐοικώς; and three possible readings of ἐοικώς. Bryan argues that Xenophanes writes in intentional contest with Homer and Hesiod. While the poets entreat the divine for their knowledge, Xenophanes believes that it is better, and more reliable, to investigate matters for oneself. All the same, Xenophanes recognizes that humans have perspectival, limited cognition; only the gods have perfect cognition. This means that, while our beliefs may be true, we can never be certain of it.
“Parmenides’ Allusive Ambiguity” (58-113) is structured like the previous chapter. This one aims to translate B8.60-1 (. . . διάκοσμον ἐοικότα πάντα φατίζω). The chapter ends with Bryan’s paraphrase of Parmenides: “I will present the following cosmology as a fitting example of the kind of account that mortals tend to regard as plausible. I do so in order that you may, by recognizing its speciousness, understand that and why such accounts should hold no appeal for you, a man who knows the truth.” This paraphrase depends on having first found that ἐοικώς has four meanings available to Parmenides—“similar”, “fitting/appropriate”, “specious”, “plausible”—and that Parmenides deploys them all simultaneously. This amalgam is best translated into technical English as “subjective plausibility”: what is εἰκώς seems true, but only to those unprepared to distinguish reality from appearance. Further evidence for this interpretation comes from the fact that five words that show up in both the Doxa and the Aletheia— δίκη, σήματα, κρίσις, ἔλεγχος, πίστις—have, Bryan argues, a forensic tone. The contrast with εἰκώς is πίστις ἀληθής, what Bryan translates as “genuine cogency”; this concept refers to the logical force of the valid argument, not to a cognitive state.
About ten pages of the chapter argue for the relationship between Xenophanes and Parmenides. Bryan believes that there is good reason to think the latter alludes to the former: doxographical tradition links them; the dating seems possible; and the two authors share terminology in their discussions of god. Bryan argues that Parmenides thinks, contrary to Xenophanes, that knowledge and certainty are possible, just as long as one investigates the right thing, namely being as opposed to becoming. A consequence of this view is that humans could in principle receive revelation, since god and human beings know the same thing. Bryan is not totally clear whether Parmenides also thinks, with Xenophanes, that engaging in reflection in a particular way in necessary for knowing the truth.
“Plato’s Timaeus” (114-60; the first of two chapters on the dialogue) argues that Timaeus’ qualification of his speech as an εἰκώς λόγος (29b1-d3) is not self-deprecating. Timaeus is instead “setting positive standards at which those who seek to describe the cosmos should aim”. Bryan defends this distinction by claiming that Timaeus invents a technical meaning of εἰκώς. He frequently sets aside the traditional acknowledgement-of-deficiency meaning “mere likeliness”, and replaces it with an original, punning meaning: “related to an εἰκών, a likeness.” For Timaeus, then, an εἰκώς account is an account of something that is itself similar to another more primary thing. Whereas we usually translate εἰκώς λόγος as “likely account”, Bryan would have us ignore the verisimilitude of the account and translate it “account of a likeness”. Timaeus’ proemium, then, claims to provide an εἰκώς λόγος of the cosmos that “will describe it as a representation of the stable paradigm and will, in some sense, consider the nature of the model as well as that of its likeness”. Bryan’s principal evidence is Timaeus’ claim that accounts are kin of (συγγενεῖς) their subject-matter. This avoids the counterintuitive position that accounts of likenesses are, because of their content, necessarily deficient in their accuracy. After all, some likenesses seem no more difficult to give an accurate account of than non-likenesses. Bryan also shows that her interpretation is not inconsistent with Timaeus’ other uses of εἰκώς in the Timeaus. Naturally, a reader might resist any resolution of an interpretative difficulty that depends on a invented word-meanings. (Bryan says that Socrates would not have used such “metaphysical jargon”, but that Timaeus would not have had the same scruples.) But the idea is very clever and presses hard on our assumptions about the nature of accounts and similitude.
“Imitation and Limitation in Timaeus’ Proemium” (161-95) continues the analysis of Timaeus’ description of his cosmological account. Bryan encourages reading Timaeus’ remarks as direct allusion to, and confrontation with, Parmenides (and not just allusion to Plato’s Republic, since the proportions discussed in the two dialogues are not perfect analogues). Πίστις is related to truth just as coming-to-be is related to being; since coming-to-be is modeled on being, and therefore shares somehow in actuality, πίστις is modeled on truth, and therefore shares somehow in truth. Thus Timaeus redeems “mortal cosmologies” against Parmenides’ denigration of them. It is true that humans cannot have perfect knowledge—they must rely on an εἰκώς μῦθος—but this weakness is a result of our inability to confirm the accounts on which we must rely, as Xenophanes has already claimed, not of the content of our beliefs. This weakness is marked by the term μῦθος, not the term εἰκώς.