Karbowski (2015) Complexity and Progression in Aristotle’s Treatment of Endoxa in the Topics

Karbowski, Joseph (2015). Complexity and Progression in Aristotle’s Treatment of Endoxa in the Topics. Ancient Philosophy 35 (1):75-96.

Traditionally, endoxa were thought to be objectively probable or subjectively plausible views, and the adjective ‘endoxos’ was translated accordingly. 1 More recently, however, scholars have suggested that endoxa are ‘reputable’ beliefs (Barnes 1981, 498-502), ‘common’ beliefs (Irwin 1988, 36-40 and Owen 1961), ‘noted or accredited’ beliefs (Bolton 1999, 59-61), and ‘accepted’ or ‘acceptable’ beliefs (Smith 1997, xxiii-xxiv; Brunschwig 1984/5, 33 and cf. ‘idées admises’ in Brunschwig 1967, 113-114 n3). These are very different proposals. What it is for a belief to be plausible is not the same as what it is for it to be probable or commonly believed; and although these epistemic features sometimes coincide, it remains a substantive question which of them, if any, Aristotle intended to single out with the adjective ‘endoxos’. In what follows I aim to sort through this quagmire of interpretive possibilities and clarify what epistemic feature is central to Aristotle’s conception of endoxa.

The interpretation for which I shall argue is a revised version of Barnes’ interpretation. On this view, endoxa are essentially reputable, respectable, or noteworthy beliefs. That is to say, what it is for a belief to be endoxon is for it to be worthy of attention and respect. However, unlike Barnes, I am keen to stress that beliefs can earn this reputable or respectable (endoxon) status in multiple ways: not only by being accepted by the majority of human beings or the wise, but also by being probable or plausible. If this is right, then the alternative interpretations mentioned above are partially correct. The epistemic features they champion (being commonly held, probable, and plausible) do ultimately factor into Aristo tle’s conception of endoxa; but they do so indirectly, by being features whose possession qualifies a belief as endoxon, i.e., reputable, respectable, or noteworthy, not by being directly denoted by the term ‘endoxos’.

I. Endoxa and Acceptable Beliefs

Aristotle’s most explicit description of endoxa appears right at the beginning of the Topics. In the course of distinguishing dialectical arguments from other types of arguments he says the following about endoxa:

Endoxa, on the other hand, are things that seem so (τὰ δοκοῦντα) to all, or most, or the wise, and of them either all, most, or the most famous and reputable. 2 (i 1.100b21-3)

According to Aristotle, endoxa are beliefs held by all or most human beings (‘the majority’ for short) or all, most, or the most famous and reputable of the wise (‘the wise’ for short). 3 Though seemingly straightforward, it is debatable whether this is meant to be a description of a single special class of beliefs, viz., endoxa, or a more loosely unified list of different types or groups of endoxa. Most scholars construe the passage in the former way, but that interpretation has been forcefully contested by Robin Smith. 4 He believes that the traditional reading makes what Aristotle is up to in the passage rather mysterious and challenges its proponents to explain what unifies this apparent ‘jumble’ consisting of the views of the majority and the wise (see Smith 1993, 345-346). Instead, Smith prefers to view the beliefs mentioned in the passage as referring to entirely different lists or types of beliefs that are unified only insofar as they are acceptable to different types of dialectical interlocutors (all, most, the wise, etc.). For this reason Smith translates ‘endoxos’ as ‘acceptable’ and accordingly construes endoxa as beliefs acceptable to one’s interlocutor.

II. Endoxicality and De Facto Acceptance

Since endoxa are beliefs actually accepted by epistemic authorities, Brunschwig stresses that endoxicality is a feature that belongs to a belief de facto (de fait), i.e., in virtue of its actual acceptance by an epistemic authority. This observation rules out interpretations that construe that feature as a property that belongs to a belief de jure (de droit) or in virtue of facts about its intrinsic content, e.g., the view that endoxa are essentially plausible or probable beliefs. Brunschwig’s idea seems to be that while we consider a belief probable or plausible by gauging whether the content of the belief is true or likely to be so, in determining whether a belief is endoxon its actual endorsement by some epistemic authority is all that matters. Such beliefs may of course have content that strikes one as being true or plausible, but those are inessential features that do not enter into the determination of a belief’s status as an endoxon (Brunschwig 1967, xxxv). In Brunschwig’s view, what it is that makes something endoxon is the fact that it is actually held by an epistemic authority.

III. Core vs. Derivative Endoxa in Topics i 10

Aristotle first [A] justifies that beliefs that bear certain similarities to endoxa are dialectical premises and then [B] proceeds to do the same for beliefs that are negations of the contraries of endoxa. His justification has the same general format in both cases: beliefs similar to or the negations of the contraries of endoxa are legitimate dialectical premises because they would/will themselves appear (ἂν φανείη, φανεῖται) to be endoxa. In order to determine whether or not these beliefs are endoxa we must understand the force of this appearance vocabulary.

A non-veridical interpretation of this appearance language has been defended by Reeve 1998, 239 that views these as ‘apparent endoxa’, distinct from the other ‘genuine’ endoxa mentioned in the chapter. This interpretation runs the risk, however, of conflating these beliefs, which are considered to be legitimate dialectical premises, with the merely apparent endoxa (ta phainomena endoxa) that function as premises of contentious arguments (Top. i 1.100b23-25; Soph. Ref. 2.165b7-8). Since Aristotle repeatedly insists that the premises of genuine dialectical arguments are endoxa, it is most plausible to view these two sets of beliefs [2-3] as genuine endoxa as well, albeit endoxa whose status as such is dependent upon their relation to other beliefs, e.g., those held by the majority and the wise, whose status as endoxa has already been established or is being taken for granted. Such ‘derivative endoxa’ are beliefs that would/will appear (ἂν φανείη, φανεῖται) to be endoxa—and really be endoxa—once it has been established that they are similar to or the negations of the contraries of other ‘core endoxa’. 11

If the foregoing is correct, Topics i 10 contains evidence of some expansion of the scope of endoxa beyond the set of beliefs mentioned at Topics i 1.100b21-23. Not only beliefs that are actually held by the majority of human beings or the wise are endoxa; beliefs similar in content to those views and the negations of their contraries are also endoxa. Importantly, the latter beliefs are de jure endoxa in terms of Brunschwig’s scheme (§2). For they appear to the sensible person to be endoxa, and so count as genuine endoxa, in virtue of facts about their content, viz., because they are similar to or the negations of the contraries of other endoxic claims, not because they are actually held by epistemic authorities. 12

In the light of this observation Brunschwig must modify his initial proposal, but he need not completely abandon it. What emerges from Topics i 10 is, in effect, a distinction between core endoxa and derivative endoxa. 13 Core endoxa at least include those beliefs mentioned at Top. i 1.100b21-23: beliefs held by all or most human beings and all, most, or the most famous and reputable of the wise. Derivative endoxa are or include the negations of the contraries of these beliefs and the beliefs similar in content to them. Derivative endoxa ultimately qualify as endoxa by virtue of their relations (or the relations of their contents) to core endoxa, and for that reason they are de jure endoxa. By contrast, the core endoxa qualify as endoxa in some other way. For all we know at this point they may do so because they are actually held by epistemic authorities. Thus, while Brunschwig’s de facto interpretation is an ill fit for derivative endoxa, it remains a live candidate for the core cases, at least through Topics i 10. However, when we examine Topics viii, we will find evidence of the possibility of de jure core endoxa, and this development constitutes a more serious threat to Brunschwig’s interpretation.

IV. Endoxa and Dokounta in Topics viii 5

V. Interim Conclusions from Topics viii 5

These observations have important consequences for the sorts of beliefs that Aristotle counts as endoxa in the Topics. It seems now that he is inclined to treat things that seem to be true or are plausible to the individual setting the dialectical standard as endoxa. 23 This constitutes an expansion of the scope of endoxa because earlier the beliefs that primarily counted as endoxa were beliefs of epistemic sources whose authority was recognized by the dialectical standard. Thus, in Topics i 10 the dialectical standard was the sensible person, and the beliefs that counted as endoxa there were beliefs held by the majority of human beings or the wise (or things similar to them or the negations of their contraries). There was no indication in that chapter, or anywhere else in Topics i, that these were beliefs actually held or accepted by the sensible person, only that he would grant them (ἂν θείη) in a dialectical discussion. 24 In Topics viii 5, however, Aristotle opens the door for the beliefs that seem true or are plausible to the relevant dialectical standard—things that dokei to him—also to count as endoxa.

VI. Adoxa and Implausible Beliefs: Topics viii 9

Beware when defending an adoxos thesis. A thesis can be adoxos in two ways. For both those things that cause one to say absurd things (ἄτοπα) [sc. are adoxa], e.g., if one were to say that all things move or that nothing does; and all those that a bad character would choose and that are contrary to our wishes [are adoxa], e.g., that pleasure is the good or that it is better to do injustice than to suffer it. For people may not be aware that you are defending the thesis for the sake of argument but will think you believe what you say and hate you. (160b17-22)

In this passage Aristotle points out two ways in which a thesis can be adoxos: either because it is absurd (ἄτοπα) (or has absurd consequences) or because it is something that a bad character would choose contrary to our wishes. The second sort of belief seems to be the main motivator of the advice in the chapter, because holding ethically corrupt or suspect beliefs can be a legitimate ground of hate or disgust, whereas holding absurd or silly (ethically indifferent) views by itself cannot. But the former class of adoxa are presumably most on Aristotle’s mind throughout Topics viii, as is evidenced by the close connection between adoxa and silly or simple-minded beliefs in 160a3 and a6-8, and the lack of concern for ethically suspect beliefs elsewhere in the book

VII. The Two Strands of Endoxa in the Topics

In the light of the foregoing, it is natural for us now to pursue the proposal that endoxa are fundamentally plausible beliefs. One might even urge that this conception was in the background of Topics i. Although Aristotle did not play up the persuasive or convincing nature of the beliefs of the majority or the wise in that book, he surely had to be assuming that they were persuasive, at least to the majority or the wise (see 104a8-9). Therefore, the stage for the later developments concerning endoxa was already set at the beginning of the treatise.

In light of this observation one might propose that Aristotle fundamentally views endoxa as different sorts of beliefs in different poles of the treatise. In Topics i endoxa are essentially beliefs accepted by epistemic authorities, while in Topics viii they are essentially plausible beliefs. This proposal has the virtue of duly acknowledging the different strands mentioned above, but it comes at a great cost. For, unless one resorts to speculative developmental hypotheses to underwrite this difference, one must maintain either that the term ‘endoxos’ in its various permutations is ambiguous in the Topics or that is has a disjunctive meaning (‘accepted or plausible’). 31 However, the former (ambiguity) proposal is implausible because the text contains none of the typical Aristotelian sign-posts that would indicate that the term is ambiguous, and the latter (disjunctive) one is too, because then the class of endoxa (relative to any single dialectical standard) would be an artificial, non-natural unity, like any other purely disjunctive set. Therefore, before resorting to ambiguity or disjunction to explain the different strands in Aristotle’s treatment of endoxa in the Topics, we should consider whether there is any way to reconcile those strands while salvaging the univocity of the relevant term and the unity of the set of beliefs to which it applies. I undertake these tasks in the next section.

VIII. Endoxa as Reputable or Noteworthy Beliefs

It is worth stressing that to deny that endoxa are fundamentally plausible or accepted is not to deny that those features play an important role in Aristotle’s thinking about endoxa. The scholars who champion those features have certainly picked up on genuine phenomena in the text. But they seem to enter the story as criteria of endoxicality or features whose possession qualifies a belief as endoxon, instead of being constitutive of endoxicality itself. 34 That is to say, although the previous observations suggest that what it is for a belief to be endoxon is for it to be respectable, reputable, or noteworthy, beliefs do not earn that status for no reason. Just as people become famous and reputable (γνωρίμοι καὶ ἐνδόξοι) for certain reasons, e.g., because they are exceptionally wise, wealthy, athletic, etc. so too beliefs become reputable or respectable (endoxa) for certain reasons, e.g., by being accepted by epistemic authorities or by being independently plausible. Emphatically, this does not entail that being a reputable/endoxic belief is identical or reducible to being an accepted or plausible belief, any more than being a reputable/endoxic person is identical or reducible to being wise, rich, or athletic. A belief qualifies for an endoxic status in virtue of its being accepted or plausible, but what it becomes, when it becomes endoxon, is reputable, noteworthy, or respectable.