PHAINOMENA AS WITNESSES AND EXAMPLES: THE METHODOLOGY

Transcription

1 Created on 20 May 2015 at hours page 193 PHAINOMENA AS WITNESSES AND EXAMPLES: THE METHODOLOGY OF E U D E M I A N E T H I C S 1. 6 JOSEPH KARBOWSKI A s philosophical methodology has received a great deal of attention recently. Scholars have begun to question some of the commonplaces about Aristotelian methodology that were inspired by Owen s influential work on the topic. 1 One of these concerns the scope of the famous endoxic method described at NE 7. 1, 1045 b 2 7. Owen maintained that that passage gives a representative description of Aristotle s philosophical methodology, whose data are for the most part the materials... of dialectic, and its problems are accordingly... conceptual puzzles. 2 Importantly, this proposal is not meant to imply that Aristotle employs the endoxic method in every single one of his treatises. Owen carefully restricts that method to philosophical treatises, like the Nicomachean Ethics and Physics, and distinguishes it from the empirical method employed in, for example, the De caelo. 3 However, even this more Joseph Karbowski 2015 A version of this paper was presented at the Junior Faculty Ancient Philosophy Workshop at Northwestern University organized by David Ebrey. I thank the participants of the workshop for their constructive feedback and encouragement, especially my commentator, Agnes Callard. I would also like to thank Dorothea Frede, Brad Inwood, Sean Kelsey, and an anonymous referee for their penetrating written comments. 1 See G. E. L. Owen, Tithenai ta phainomena [ Tithenai ], in S. Mansion (ed.), Aristote et les problèmes de méthode: communications présentées au Symposium Aristotelicum tenu à Louvain du 24 août au 1 er septembre 1960 (Louvain, 1961), ; repr. in G. E. L. Owen, Logic, Science and Dialectic: Collected Papers in Greek Philosophy (Ithaca, NY, 1986), Ibid. 88. Owen s characterization of the NE 7. 1 method as dialectical is controversial. Although that method proceeds from ἔνδοξα, Aristotle himself never calls it dialectical, and it is unclear whether it employs any of the strategies for dialectical discussions described in the Topics. For these reasons, I will continue to refer to the NE 7. 1 method as the endoxic method and forgo any reference to it as dialectical ; cf. D. Frede, The Endoxon Mystique: What Endoxa Are and What They Are Not [ Mystique ], Oxford Studies in Ancient Philosophy, 43 (2012), at Owen, Tithenai, 91. Owen identifies philosophical treatises as those in which Aristotle is doing conceptual analysis, but this problematically attributes to Aris-

2 Created on 20 May 2015 at hours page Joseph Karbowski modest thesis has come under fire from recent work on the methodology of the textbook philosophical treatises. For instance, careful studies of the enquiries into happiness and character virtue in the Nicomachean Ethics (NE) have persuasively shown that the endoxic method is not their governing methodology; 4 and these conclusions have been reinforced by Dorothea Frede s recent contribution to the debate, which argues, more generally, that the endoxic method is a rara avis in the Aristotelian corpus, whose use is more or less confined to NE Frede s wide-ranging and impressive study says very little about the Eudemian Ethics (EE). 6 This is understandable, as the literature to which she is reacting focuses primarily on the NE. However, the methodology of the EE is a topic well worthy of examination for at least two reasons. First, the EE is a serious treatise of moral philosophy by a rare philosophical talent. 7 Therefore, both its doctrines and its methodology deserve consideration in their own right, quite apart from developmental questions about the treatise s relation to the NE and compositional questions about the origin of the so-called common books. 8 Second, many scholars maintain that the totle an anachronistic conception of philosophy; cf. J. M. Cooper, Nicomachean Ethics VII. 1 2: Introduction, Method, Puzzles [ Introduction ], in C. Natali (ed.), Aristotle s Nicomachean Ethics, Book VII (Oxford, 2009), 9 39 at 29 n See C. Natali, Rhetorical and Scientific Aspects of the Nicomachean Ethics, Phronesis, 52 (2007), ; id., Posterior Analytics and the Definition of Happiness in NE I, Phronesis, 55 (2010), ; G. Salmieri, Aristotle s Non- Dialectical Methodology in the Nicomachean Ethics, Ancient Philosophy, 29 (2009), Marco Zingano argues that the earlier books of the NE are nondialectical, while those of the EE are dialectical: see M. Zingano, Aristotle and the Problems of Method in Ethics [ Problems ], Oxford Studies in Ancient Philosophy, 32 (2007), For a reservation about Zingano s characterization of the EE, see n. 27 below. 5 See Frede, Mystique. 6 Frede briefly argues against the application of the endoxic method in EE 3. 1 and 7. 1 ( Mystique, 190 n. 10). However, she does not discuss at any length the rich set of methodological remarks in EE Compare the apt remarks of Inwood and Woolf: [The EE] clearly demands our attention as a discussion of fundamental human values written by one of the great philosophers of the western tradition (B. Inwood and R. Woolf, Aristotle: Eudemian Ethics (Cambridge, 2013), viii). 8 The current paper neither makes nor depends upon any substantive views about the developmental relation between the NE and the EE or the original home of the common books. For convenience I will refer to the common books by their place in the Nicomachean Ethics. Emphatically, this is not to deny the difficulty of these questions or the value of the work done on them by scholars such as Harlfinger, Rowe, Kenny, et al.; see D. Harlfinger, Die Überlieferungsgeschichte der Eudemischen Ethik, in P. Moraux and D. Harlfinger (eds.), Untersuchungen zur Eudemischen Ethik

3 Created on 20 May 2015 at hours page 195 Phainomena as Witnesses and Examples 195 endoxic method, or a close approximation of it, is described in EE 1. 6, the treatise s central methodological chapter: 9 There are passages elsewhere in the EE that present and discuss the method for ethics... in ways linked closely in language and substance to what we find in NE 7. 1: see EE 1. 6, (Cooper, Introduction, 27 n. 39) The methodology described here [at EE 6. 1 =NE 7. 1, 1145 b 2 7] is the one announced in Book I, 1216 b 26 35, and these early chapters of Book VI offer the most substantial implementation of it. (A. Kenny, Aristotle: The Eudemian Ethics (Oxford, 2011), 172; cf. 151) Thus, in addition to being of interest in its own right, the EE s central methodological chapter is worth examining for the light it promises to shine upon the scope of the endoxic method. In particular, a careful study of that chapter will either confirm the dominant scholarly interpretation of EE 1. 6, which considers it another locus of the endoxic method, or it will lend further support to recent sceptics about Owen s view, who believe that that method is more or less confined to NE In this paper I undertake a detailed examination of the philosophical method described in EE For convenience I will refer to this method as the Eudemian method, but this locution should not be taken to imply that it is the only method used in the EE. The focus of the current paper is the method of EE I aim to understand the structure of that method and to determine whether it is indeed as similar to the endoxic method as scholars suppose. I will argue that close scrutiny of Aristotle s description of the Eudemian method and his employment of it in the enquiry into happiness in [Untersuchungen] (Berlin, 1971), 1 50; A. Kenny, The Aristotelian Ethics: A Study of the Relation between the Eudemian and Nicomachean Ethics of Aristotle (Oxford, 1978); C. J. Rowe, The Eudemian and Nicomachean Ethics: A Study in the Development of Aristotle s Thought [Development] (Cambridge, 1971); and, most recently, O. Primavesi, Ein Blick in den Stollen von Skepsis: Vier Kapitel zur frühen Überlieferung des Corpus Aristotelicum, Philologus, 151 (2007), However, whether the EE was composed before or after the NE, and whether or not it originally housed the common books, the method described in EE 1. 6 merits close attention. 9 See also F. Dirlmeier, Aristoteles, Eudemische Ethik, übersetzt und kommentiert [Eudemische Ethik] (Berlin, 1962), 184; L. Jost, Eudemian Ethical Method, in J. P. Anton and A. Preus (eds.), Essays in Ancient Greek Philosophy, iv. Aristotle s Ethics [Ethics] (Albany, NY, 1999), 29 40; P. Simpson, The Eudemian Ethics of Aristotle (New Brunswick, 2013), 111 n. 6; Zingano, Problems, Inwood and Woolf compare ( cf. ) the EE 1. 6 and endoxic methods in their recent translation, which suggests that they view them as at least comparable: see Inwood and Woolf, Aristotle: Eudemian Ethics, xxiii, 9.

4 Created on 20 May 2015 at hours page Joseph Karbowski EE reveals that that method is in fact substantially different from the endoxic method of NE Briefly, whereas the Eudemian method aims to discover explanatory definitions of ethical topics by constructing complex deductive arguments which draw on the phainomena for support and illustration, the endoxic method seeks truth about ethical matters by purging endoxa of their epistemic shortcomings via aporetic investigation. D. J. Allan famously compared the Eudemian method to the mathematical pattern of deduction exhibited by Euclid s Elements. 10 However, I will conclude by suggesting that it is more accurately viewed as an application of the rational (kata ton logon) mode of argumentation found throughout the Aristotelian corpus. 1. The enquiry into happiness in EE The first six chapters of the EE constitute an introduction to the treatise and its main topic, viz. happiness. In the last of these, EE 1. 6, Aristotle describes the methodology that he plans to observe in the subsequent investigation. We will examine that important chapter in the next section. This section offers an overview of Aristotle s enquiry into happiness in EE Those chapters are of special interest to the current project, because they can help illuminate the method described in EE Aristotle explicitly indicates that his enquiry into happiness is governed by the Eudemian method (EE 1. 6, 1216 b 26; 1. 7, 1217 a 18 21). Consequently, EE can serve both as a source of information about the Eudemian method and also as a constraint upon our interpretation of that method. A major virtue of the interpretation developed below is that it offers an account of the Eudemian method that conforms to Aristotle s actual procedure in EE Aristotle begins EE 1. 7 with a statement of his aim: to discover more clearly what happiness is, starting from initially unclear claims about it (EE 1. 7, 1217 a 18 21). He observes that it is generally agreed that happiness is the greatest and best of human goods (1217 a 21 2), and this claim serves as his point of departure. In the rest of the chapter Aristotle proceeds to refine this remark. Since happiness may be achievable by beings superior to humans, 10 See D. J. Allan, Quasi-Mathematical Method in the Eudemian Ethics [ Quasi- Mathematical ], in S. Mansion (ed.), Aristote et les problèmes de méthode (Louvain, 1961), I discuss his proposal in sect. 4.

5 Created on 20 May 2015 at hours page 197 Phainomena as Witnesses and Examples 197 he stresses that he is concerned with human happiness, not that pertaining to the gods (1217 a 22 9). This leads him to distinguish what can be achieved by human action from what cannot (and what can be achieved by superhuman action) (1217 a 30 8). The sort of happiness he has in mind falls under the former category. Consequently, Aristotle concludes that human happiness is the best of the goods achievable by human action (1217 a 39 40). In EE 1. 8 Aristotle subsequently examines what the best is and in how many ways it is said (EE 1. 8, 1217 b 1). 11 He canvasses three main views of the best good in the chapter: (i) the best good is the Form of the Good; (ii) it is the common good; 12 and (iii) it is the goal of all that is achievable by human action. The details of the chapter and of the various criticisms Aristotle levels against the first two views are subject to dispute. 13 However, its general thrust is relatively straightforward. In addition to being first among goods and the cause of the goodness of other goods, Aristotle presumes that the best good must also be achievable in action (cf. EE 1. 7, 1217 a 39 40). The Form of the Good garners attention because, as the Platonists understand it, it meets the first two criteria: it is first among things and is the cause of the goodness by virtue of its presence (παρουσία: 1217 b 4 5, 6 8). Nonetheless, as Aristotle laments repeatedly in the chapter, it is unattainable in action (1217 b 23 5; 1218 a 38; 1218 b 7 9), and so cannot be the best good. The candidate championed by the second view, the common good, suffers the same shortcoming. According to Aristotle, it is not achievable in action (1218 b 1 7, 8 10), and so it is not the best good either. 11 At EE 1. 8, 1217 b 1, λέγεται ποσαχῶς either anticipates (a) the three different views about the best good in the chapter (the Form of the Good, the common good, or the good qua end of action) or (b) the homonymy of goodness/being mentioned at 1217 b 25 6 and again at 1218 b 4 6. Interpretation (b) is more natural, given Aristotle s tendency to use the phrase λέγεται ποσαχῶς/πολλαχῶς to signal homonymy. However, against it is the fact that the remark at 1217 b 1 is referring to how many ways the best good (τὸ ἄριστον), not the good (τὸ ἀγαθόν), is said. For this reason I prefer interpretation (a), though nothing major in this paper hangs on it. 12 The common good (τὸ κοινὸν ἀγαθόν) is similar to the Form of the Good in that it is a univocal property from which all good things derive their goodness. However, unlike the Form of the Good, which is separate from the things that participate in it (EE 1. 8, 1217 b 14 6), it inheres in them, which explains why it is changing (1218 b 9). 13 For detailed discussion of EE 1. 8 see D. J. Allan, Aristotle s Criticisms of Platonic Doctrine concerning Goodness and the Good, Proceedings of the Aristotelian Society, 64 (1963 4), ; Simpson, The Eudemian Ethics of Aristotle, ; M. J. Woods, Aristotle s Eudemian Ethics: Books I, II, and VIII, 2nd edn. (Oxford, 1992),

6 Created on 20 May 2015 at hours page Joseph Karbowski The only view whose candidate meets all of the relevant criteria for the best good is the third one (1218 b 10 12). The best good qua goal of all achievable goods is itself achievable; goals are among the items achievable by human action (cf. EE 1. 7, 1217 a 36 8). It is also the first among achievable goods and the cause of the goodness of the things that are pursued for its sake (1218 b 16 24). 14 Thus, Aristotle concludes that happiness is the best human good in so far as it is the end of all achievable goods (1218 b 24 5). His next order of business is to determine what this best good qua ultimate end is (1218 b 25 6). 15 Making a fresh start on the investigation, Aristotle begins EE 2. 1 by dividing human goods into goods of the soul and external goods (EE 2. 1, 1218 b 32). 16 He argues that the former are more choiceworthy than the latter, because wisdom, pleasure, and virtue the three main candidates for happiness mentioned at EE 1. 1, 1214 a 30 b 6 are found in the soul, and everyone considers the goal to be one or some combination of these (1218 b 33 6). This move permits Aristotle to focus upon psychic goods in his search for the best human good/happiness. After distinguishing psychic goods into states/capacities and activities/processes (1218 b 36 7), he assumes that a virtue is the best disposition, state, or capacity of anything that has a use or function (1218 b 37 9). Aristotle supports 14 The point of the observation at EE 1. 8, 1218 b 22 4, that no one explains why health or any other starting-point is good is unclear, though it may be intended to confirm that the best good qua end is first among goods; see Simpson, The Eudemian Ethics of Aristotle, The final line of EE 1. 8 is notoriously obscure. Instead of directly claiming that the next order of business is to examine what the best good qua end of human action is, Aristotle says that we must examine how many the best good (sc. is) (σκεπτέον ποσαχῶς τὸ ἄριστον πάντων: EE 1. 8, 1218 b 26). Scholars have proposed various emendations to make this remark more intelligible, e.g. reading πῶς for ποσαχῶς, adding καί after ποσαχῶς, adding λέγεται after πάντων, etc. Woods even goes so far as to suggest that the line is spurious (Aristotle s Eudemian Ethics: Books I, II, and VIII, 84). Nonetheless, we can make sense of Aristotle s point without resorting to any of these emendations. By asking how many things constitute the best good, Aristotle is essentially asking what the best good is (cf. the connection between what the virtues are and how many they are at NE 3. 5, 1115 a 4 5). However, this form of the question ( How many...? ) leaves it open that multiple things may have a claim to be the best good, or that it is constituted by a number of different first-order goods. This presumably reflects the fact that, at this intermediate stage in the investigation, Aristotle does not want to beg the question against the reputable sources who identify happiness with virtue, wisdom, and pleasure, or some combination of these items (EE 1. 1, 1214 a 30 b 6). 16 Aristotle points out that this division can also be found in the exoteric works (EE 2. 1, 1218 b 33 4), but the reference is uncertain.

7 Created on 20 May 2015 at hours page 199 Phainomena as Witnesses and Examples 199 the previous account of virtue inductively by appeal to well-known functional items such as cloaks, boats, and houses, and then applies the relevant thesis to the soul, since it has a function (1219 a 16). 17 This implies that psychic virtue is the best state or capacity of the soul. However, as Aristotle apparently realizes, he cannot yet conclude that it is happiness or the best good of the soul tout court, since in principle the activity of psychic virtue may be even better than psychic virtue itself. Indeed, the subsequent axiological remarks in the chapter support this very conclusion. Aristotle assumes (ἐχέτω) that the value of a state s function correlates with the value of the state itself, i.e. that a better state has a better function (1219 a 6 8), and that a thing s function is its end (1219 a 8). 18 These assumptions entail that the function of a state is better than the state itself (1219 a 9). For, by definition, the function is an end, and an end in a domain is the best thing for the sake of which everything else is chosen (1219 a 8 13). But, Aristotle points out, functions come in two main types: products over and above the employment of the state, e.g. houses and health, and the employment of the state itself, e.g. seeing and contemplating (1219 a 13 17). Thus, when a thing s employment is its function the employment of the state is better than the state itself (or its mere possession) (1219 a 17 8). This conclusion proves relevant to the soul. For, Aristotle assumes (ἔστω), its function is to make something alive, 19 which is an employment (1219 a 24 5). And since the function of a thing and its virtue are similar, except that the function of a thing s virtue is to perform the relevant function well (1219 a 18 23), the function of the virtue of the soul must be living well. This, i.e. the function/employment of psychic virtue, Aristotle concludes, is happiness, the best of human goods (1219 a 25 8). Immediately after drawing this conclusion, Aristotle offers a summary of the preceding argument: This strategy is an instance of Aristotle s technical mode of argument by example. 18 This claim is stipulated without any support, but other theses introduced by similar posit vocabulary are argued for in the chapter; see sect. 4 below. 19 Kenny translates ἔστω... ποιεῖν as let us postulate (Aristotle: The Eudemian Ethics, 15; cf. 155). However, postulate is a rather strained translation of ποιεῖν, and in any case is unnecessary because ἔστω by itself suffices to introduce assumptions or posits. The more natural reading of ποιεῖν is also preferable because Aristotle believes that the soul itself is not the subject of psychological attributes (including living) but is rather the causal principle that makes the body alive: see DA 1. 4, 408 a 34 b My translations of the Eudemian Ethics, including NE 7 (=EE 6), follow Inwood

8 Created on 20 May 2015 at hours page Joseph Karbowski This conclusion is clear from what we have laid down, namely that happiness is the best thing, the ends are in the soul and the best of goods, and the things in the soul are either a state or an activity. So since the activity is better than the disposition, and the best activity belongs to the best state, it is clear from what has been laid down that the activity of the soul s virtue is the best thing. And the best thing is also happiness. Happiness, then, is the activity of the good soul. (EE 2. 1, 1219 a 28 35) This summary can be represented as follows: (1) Happiness is the best human good and an end achievable in action. (2) Goods/ends in the soul are best among human goods. (3) Therefore, happiness must be a good (the best good) in the soul. (4) Goods in the soul are either states or activities. (5) Activities are better than states, and the best activity is correlated with the best state. (6) Therefore, happiness must be the best activity in the soul, the one correlated with the best state. (7) The best state of the soul is its excellence. (8) Therefore, happiness is the activity of the good, i.e. excellent, soul. Immediately following this summary Aristotle adds an important qualification. Since happiness is complete, it must be the activity of a complete life in accordance with complete virtue (1219 a 35 9). In what is essentially an appendix to the investigation Aristotle confirms his definition of happiness, citing beliefs held by all of us (τὰ δοκοῦντα πᾶσιν ἡμῖν) as his witnesses (marturia) (1219 a 39 40). 21 He mentions three such beliefs with which his definition harmonizes: (a) that happiness is the same as living well and doing well; (b) that both life and action are employments and activities; and (c) that one cannot be happy for a single day, as a child, or at every stage of life (1219 b 1 6). After mentioning Solon s remark with approval, Aristotle then turns to a brief discussion of praise and encomia. He observes that praise and encomia are given for deeds and Woolf, Aristotle: Eudemian Ethics, though I also consult Dirlmeier, Eudemische Ethik; Kenny, Aristotle: The Eudemian Ethics; and Woods, Aristotle s Eudemian Ethics: Books I, II, and VIII. 21 For discussion of the nature of the witnesses and examples in EE see sect. 2(c) below.

9 Created on 20 May 2015 at hours page 201 Phainomena as Witnesses and Examples 201 and that we tend to judge the quality of a person from their deeds (1219 b 8 11). 22 Aristotle then rather abruptly asks why happiness is praised and answers that it is because other things are praised because of it (by promoting it or being part of it) (1219 b 11 16). 23 His final order of business, before turning to discussion of the soul and its virtue(s), involves showing how the foregoing helps to resolve a puzzle about why good people are no better off than bad ones for half of their lives (1219 a 16 18). The reason, he explains, is that sleep involves the soul s idleness (1219 b 18 20). 2. The Eudemian method In the portion of the text previously summarized, Aristotle is avowedly implementing the methodology described in EE I shall now examine that chapter with a view to gaining deeper insight into the structure of the relevant method. EE 1. 6 contains a number of methodological remarks, but we may fruitfully begin by examining its opening lines: [1] In all these matters we must try to seek conviction through arguments, using the phainomena as witnesses and examples. [2] The best situation is that everyone be in manifest agreement with what we are going to say; failing that, that everyone should in some fashion agree, as they will do when they have had their minds changed. [3] Each person has some affinity with the truth, and it is from this that one must prove one s case on these issues in one way or another. [4] If we start from what is truly but not clearly spoken, clarity will be won as we make progress, continually substituting what is more intelligible [sc. by nature] for what is usually spoken of confusedly. (EE 1. 6, 1216 b 26 35) The first sentence [1] is a preliminary description of the scope ( all these matters ), aim ( conviction ), and argumentative strategy ( argument using the phainomena as witnesses and examples ) of the method. The second sentence [2] explains why (γάρ) Aristotle is seeking conviction via arguments that appeal to the phainomena, viz. because doing so will help him secure the widest (manifest or 22 These remarks presumably provide additional confirmation that happiness involves the use or employment of virtue, not merely the possession of it; cf. NE 1. 8, 1098 b a 7; 1. 12, 1101 b This remark highlights and reinforces happiness s status as the best human good and the cause of the goodness of the other things; cf. NE 1. 12, 1101 b a 4.

10 Created on 20 May 2015 at hours page Joseph Karbowski qualified) agreement possible; 24 and the third sentence [3] in turn explains why (γάρ) people will in fact agree when they have had their minds changed, viz. because each person has some natural affinity for the truth. 25 The fourth sentence [4] offers information about the epistemic status of the relevant starting-points (they are true but unclear claims) and an additional remark about procedure ( substituting what is more intelligible for what is usually spoken of confusedly ). These are by no means the last words about methodology in EE 1. 6, but they are important. For they indicate that Aristotle has particular views about the proper aims, argumentative strategy, and starting-points of ethical enquiry, and they inform us about what those views are. I believe that the norms about proper procedure for ethical enquiry articulated in this passage (and the rest of EE 1. 6) comprise a methodology which differs substantively from the endoxic method of NE The remainder of this section paints a more detailed picture of the Eudemian method, drawing both upon Aristotle s description of it in EE 1. 6 and upon his procedure in EE Below I shall discuss the Eudemian method s aim(s), argumentative strategy, and starting-points/phainomena. (a) The Eudemian method: cognitive aim(s) In EE 1. 6 Aristotle claims to be seeking rational conviction (πίστις) about ethical matters (EE 1. 6, 1216 b 26 7). 26 His account of these matters will most certainly have to be convincing to him, but he is also clear that it will need to be something with which all human beings either manifestly or qualifiedly ( in some fashion ) 24 Cf. J. Barnes, Aristotle s Methods of Ethics [ Methods ], Revue internationale de la philosophie, 34 (1981), at 507 8; Woods, Aristotle s Eudemian Ethics: Books I, II, and VIII, Walzer/Mingay have suggested that we replace ὧν at EE 1. 6, 1216 b 31, with οὗ, but the emendation lacks manuscript support. Barnes suggests that ὧν refers all the way back to τοῖς φαινομένοις at 1216 b 27 8 ( Methods, 506). This construal is not impossible, but a closer and more natural antecedent of ὧν is οἰκεῖόν τι ( some affinity... ) in the preceding clause: see Dirlmeier, Eudemische Ethik, 11; Inwood and Woolf, Aristotle: Eudemian Ethics, 9; Kenny, Aristotle: The Eudemian Ethics, 9. This interpretation is not precluded by the fact that ὧν is plural. The affinity for truth naturally invites the plural because Aristotle believes it to be possessed by every human being. 26 In the Rhetoric πίστις designates the rhetorical modes of persuasion, but in philosophical contexts it often signifies rational conviction intimately associated with belief (δόξα): see DA 3. 3, 428 a 19 23; cf. the entry on πίστις in H. Bonitz, Index Aristotelicus (Berlin, 1870), 595 a 61 b 59.

11 Created on 20 May 2015 at hours page 203 Phainomena as Witnesses and Examples 203 agree (1216 b 26 30). Aristotle ostensibly cares about general agreement because (γάρ) he believes that humans have a natural affinity for the truth (1216 b 30 1; cf. Rhet. 1. 1, 1355 a 15 18). According to him, humans by nature are rational animals with cognitive faculties whose function (ergon) is discerning the truth (NE 6.2, 1139 a 29); and while this does not entail that everything we believe will be true, it does imply that we at least have an innate disposition for the truth and that we will rarely be entirely or radically mistaken (Metaph. α 1, 993 a 30 b 11). Consequently, Aristotle supposes that an account that fails to secure at least qualified agreement from the bulk of mankind must be mistaken. 27 This last observation reveals that Aristotle s concern with general agreement in EE 1. 6 is subordinate to a concern for truth. That is to say, he is seeking a generally convincing account of ethical matters because he thinks that proposals that do not have that feature are unlikely to be true. Truth, however, is not Aristotle s ultimate cognitive aim in the EE. 28 In the sequel to the previously quoted passage he clarifies that he has even greater cognitive ambitions: In every field of enquiry, arguments made philosophically differ from those made non-philosophically. Hence one should not, even when it comes to 27 Qualified (τρόπον τινα) agreement is a form of agreement which, Aristotle says, will be produced or achieved after their minds have been changed (μεταβιβαζόμενοι) (EE 1. 6, 1216 b 29 30). The relatively rare verb μεταβιβάζειν tends to pick out a process of mental conversion in which some of an interlocutor s beliefs, those which do not seem to us to have been well said, are replaced by other, better beliefs (Top. 1. 2, 101 a 30 4; cf. R. Smith, Dialectic and Method in Aristotle, in M. Sim (ed.), From Puzzles to Principles? Essays on Aristotle s Dialectic [Puzzles] (Lanham, Md., 1999), at 42 3). In order to achieve this conversion the speaker argues dialectically with the interlocutor (Top , 161 a 30 6), showing how the novel, less faulty beliefs follow from other beliefs accepted by the latter. This observation suggests that qualified agreement involves someone being committed to believing something, or at least finding it plausible, because it follows from other beliefs held by the person in question, albeit in a non-obvious way. The process of mental conversion aims to make that commitment clear by showing that the relevant belief follows from the individual s other beliefs. This explains why Aristotle appeals to majority beliefs as some of his evidence in the EE and why he stresses that his definition of happiness harmonizes with things we all believe at EE 2. 1, 1219 a 39 b 8. However, we must not exaggerate the significance of this observation and conclude, with Zingano, that the Eudemian method is dialectical (see Zingano, Problems, 304). For Aristotle is clear that agreement with generally held beliefs is only a necessary condition of adequacy for his ethical principles. In keeping with the avowed philosophical (or scientific ) nature of the investigation, Aristotle additionally expects his principles to have an explanatory status (EE 1. 6, 1216 b 36 9), which dialectical definitions do not have. 28 I say cognitive aim, because the ultimate aim of the EE, just like the NE, is action (πρᾶξις), i.e. to help us achieve virtue and happiness, not merely to know what they are (EE 1. 1, 1214 a 8 14; 1. 5, 1216 b 20 5).

12 Created on 20 May 2015 at hours page Joseph Karbowski politics, regard as superfluous the kind of study that makes clear not only what something is [τὸ τί] but also its cause [τὸ διὰ τί]. For this is the philosophical approach in every field of enquiry. (EE 1. 6, 1216 b 35 9) In this passage Aristotle indicates that his approach in the treatise is robustly philosophical. This sort of study, he says, is one that has causal or explanatory aspirations: it seeks to know the why (τὸ διὰ τί). 29 However, in so far as it also aims to know the what (τὸ τί) it has definitional ambitions too, 30 and these are intimately connected to its explanatory ones. For, according to Aristotle, the what it is [τί ἐστι] and the why it is [τὸ διά τί ἔστι] are the same (Post. An. 2. 2, 90 a 14 15, 31 2): the essence of an item consists in the causally fundamental features that underwrite and explain its other (derivative) necessary features. 31 Accordingly, properly philosophical definitions have an explanatory or causal status and differ from dialectical ones that do not (DA 1. 1, 402 b a 2; cf. Top. 6. 4, 141 a 29 33; 142 a 7 ff.). 32 Consequently, the definition of happiness sought by Aristotle must likewise have an explanatory status, and in particular be capable of explaining the various nonexplanatory features associated with it, e.g. why happiness is the best, noblest, and most pleasant thing Contrast this remark with the claim at NE 1. 4, 1095 b 6 8, that there is no need for the why (διότι) if the facts (τὸ ὅτι) are already clear. This seems to be an important methodological difference, but the matter needs further scrutiny; cf. Inwood and Woolf, Aristotle: Eudemian Ethics, xxiii. 30 Margueritte s emendation of τι to ὅτι is unnecessary: see Dirlmeier, Eudemische Ethik, From the beginning, the EE is pitched as a treatise that seeks to define or say what happiness, virtue, wisdom, and pleasure are: see EE 1. 1, 1214 a 14 15; 1. 5, 1216 a 37 b 2. The remark at EE 1. 6, 1216 b 35 9, adds a further constraint that adequate philosophical definitions must satisfy, viz. that they must be causal or explanatory. 31 Aristotle s view of the interconnection between definition and demonstration/ explanation is illuminatingly discussed in D. Charles, Aristotle on Meaning and Essence [Essence] (Oxford, 2000). 32 For further discussion of the difference between scientific or philosophical definitions and dialectical definitions see Charles, Essence, Cf. Simpson, The Eudemian Ethics of Aristotle, 214. Gigon endorses a more deflationary interpretation of the cause mentioned at 1216 b 39, emphasizing that the EE is explanatory only in so far as it aims to describe the ultimate goal of action, viz. happiness, and the means to achieving it: see O. Gigon, Das Prooimion der Eudemischen Ethik [ Prooimion ], in Moraux and Harlfinger (eds.), Untersuchungen, at 129. Those are undoubtedly aims of the treatise, and they explain why ethics is a practical discipline. However, this deflationary interpretation ultimately fails to do justice to the philosophical nature of the EE. For, as Aristotle acknowledges at EE 1. 6, 1216 b 35 6, it is possible to give philosophical and non-philosophical treatment of the same issues. Consequently, philosophical and non-philosophical treatments

13 Created on 20 May 2015 at hours page 205 Phainomena as Witnesses and Examples 205 These considerations strongly suggest that Aristotle conceives the EE as an attempt to discover the first principles of a demonstrative science of ethics/politics along the lines of the Posterior Analytics. These first principles are substantive definitions of happiness, character virtue, etc., which explain why these items have their derivative necessary features. Importantly, in the EE Aristotle does not explicitly present ethical demonstrations (ἀποδείξεις) that use his preferred definitions as premisses; he is only working towards the first principles, not proceeding demonstratively from them in the treatise (cf. EE 1. 6, 1216 b 32 5). Nonetheless, since he believes that properly philosophical and scientific explanations take the form of demonstrations from first principles, it is apropos to describe the explanations to which he is alluding at EE 1. 6 (1216 b 36 9) as demonstrations. 34 Let us now consider the argumentative strategy that the Eudemian method prescribes for the discovery of the relevant principles. (b) The Eudemian method: argumentative strategy Since the EE is a philosophical treatise, it is understandable that Aristotle would aim to secure conviction through rational argumentation (διὰ τῶν λόγων: EE 1. 6, 1216 b 26 7; cf a 1 2). 35 However, he is adamant that not just any kind of arguments have a cannot be distinguished by their topics alone. Instead, they must be distinguished by the manner in which they treat of the relevant topic (causally or non-causally), which favours the interpretation defended in the body of the paper. 34 Importantly, this scientific interpretation of the ethical theory is not incompatible with the ultimate practical orientation of the treatise. Unlike the NE, the EE does not contain any remarks about ethical (im)precision or the fluctuation of ethical subjects: see C. Bobonich, Aristotle s Ethical Treatises, in R. Kraut (ed.), The Blackwell Guide to Aristotle s Nicomachean Ethics (Oxford, 2006), at 26 7; Rowe, Development, 70 1; and in any case, the imprecision seemingly impacts upon the application of the principles to particular cases/actions, not the search for the principles itself. Moreover, Aristotle explicitly admits that practical wisdom (φρόνησις) involves knowledge of both ethical universals and particulars (NE 6. 7, 1141 b 14 23; cf. NE 10. 9, 1180 b 11 28). Consequently, a demonstrative science of ethics can contribute to moral education by offering causal insight into the relevant ethical universals, even if this general insight must be supplemented with practical perception (and a state of character) in order to promote good action reliably; cf. C. D. C. Reeve, Action, Contemplation, and Happiness: An Essay on Aristotle (Cambridge, Mass., 2012), In Aristotle an appeal to logos/logoi is often contrasted with an appeal to perception or sense experience: see MA 1, 698 a 11 13; Meteor. 4. 1, 378 b 13 26; De iuv. 2, 468 a 20 3; Pol. 7. 4, 1326 a 26, 29; cf. Dirlmeier, Eudemische Ethik, 183; D. Henry, Optimality Reasoning in Aristotle s Natural Teleology [ Optimality ], Oxford

14 Created on 20 May 2015 at hours page Joseph Karbowski place in philosophical enquiry. In the second half of EE 1. 6 Aristotle indicates that proper philosophical arguments should be appropriate (οἰκεῖοι) to the subject-matter (1216 b a 10); clearly distinguish the premisses of the arguments from what they purport to establish (1217 a 10 11); attend appropriately to the phainomena (1217 a 11 14); and of course have true premisses (1217 a 14 17). These conditions are intended to prevent fallacious, sophistical arguments from infecting wholesome philosophical enquiry, 36 and Aristotle goes to great pains to meet them in his enquiry into happiness. Consider again the argument by which Aristotle arrives at his substantive definition of happiness in EE 2. 1: (1) Happiness is the best human good and an end achievable in action. (2) Goods/ends in the soul are best among human goods. (3) Therefore, happiness must be a good (the best good) in the soul. (4) Goods in the soul are either states or activities. (5) Activities are better than states, and the best activity is correlated with the best state. (6) Therefore, happiness must be the best activity in the soul, the one correlated with the best state. (7) The best state of the soul is its excellence. (8) Therefore, happiness is the activity of the good, i.e. excellent, soul. This is not a demonstration (ἀπόδειξις), which uses the fundamental explanatory definition of happiness to explain one of its derivative features. Instead, it is an argument that proceeds towards such a causal principle, i.e. has it as its conclusion. 37 Notice that this Studies in Ancient Philosophy, 45 (2013), at 252. Bolton interprets this mode of argument as dialectical (R. Bolton, Two Standards for Inquiry in Aristotle s De caelo [ Standards ], in A. C. Bowen and C. Wildberg (eds.), New Perspectives on Aristotle s De caelo (Leiden, 2009), 51 82). However, I dispute this interpretation in J. Karbowski, Empirical Eulogos Argumentation in Aristotle s Generation of Animals III. 10, British Journal for the History of Philosophy, 22 (2014), I discuss this distinction and what it suggests about the Eudemian method in sect This is nicely emphasized in Dirlmeier, Eudemische Ethik, 186 7; cf. Simpson, The Eudemian Ethics of Aristotle, For this reason, it is similar in its structure to the sort of arguments called syllogisms of the that [τὸ ὅτι] by Aristotle in Post. An However, this argument is not a syllogism in the technical sense, because it has more than two premisses.

15 Created on 20 May 2015 at hours page 207 Phainomena as Witnesses and Examples 207 argument meets all of the conditions of good philosophical arguments mentioned above. It is a valid argument, whose premisses are clearly distinguished from the conclusion that follows from (ἐκ) them. Its premisses are appropriate to the subject-matter. For they include claims about happiness, the soul, its virtue and constituents, which squarely fall within the domain of ethics. Moreover, its premisses are also true claims in agreement with the relevant phainomena. This is ensured by Aristotle s use of the phainomena as the witnesses and examples of these premisses. I will discuss this aspect of his procedure shortly. First, let us reflect more generally upon the structure of the previous argument. Aristotle s central argument for happiness in EE has a complex deductive structure. Its premisses include a preliminary unclear definition of happiness and additional theses about the soul, virtue, etc., and its conclusion is a clearer definition of happiness. The movement in this argument is depicted by Aristotle s claim that we must continually substitute or replace what is customarily spoken of confusingly with what is more intelligible [sc. by nature] (EE 1. 6, 1216 b 33 5). In accordance with standard Aristotelian doctrine, philosophical enquiry proceeds from what is familiar to us to what is familiar by nature, and it aims to make what is familiar by nature familiar to us (Top. 6. 4, 141 b a 11; Phys. 1. 1, 184 a 16 21; Metaph. Ζ 3, 1029 b 1 12). Aristotle follows this general path in his enquiry into happiness in EE He begins from an initially agreed-upon definition of happiness (as the best human good) in EE 1. 7; proceeds to refine it in EE 1. 8; and then, from this refined definition and additional theses about the soul and its virtue, deduces a further definition of happiness that is clearer by nature. 38 A number of argumentative strategies are invoked to establish the premisses of the aforementioned argument. Aristotle arrives at the first premiss by reflective clarification of a generally accepted definition of happiness (EE ). He supports the second by appeal 38 Aristotle s advancement through a variety of definitional stages, beginning from a preliminary unclear (by nature) definition of happiness and ultimately discovering a clearer, causal definition, bears a striking similarity to the scientific procedure described in the second book of the Posterior Analytics. For discussion of the latter method, which duly stresses its progression through a variety of definitional stages, see R. Bolton, Definition and Scientific Method in Aristotle s Posterior Analytics and Generation of Animals, in A. Gotthelf and J. G. Lennox (eds.), Philosophical Issues in Aristotle s Biology (Cambridge, 1987), at ; Charles, Essence, chs. 2 3.

16 Created on 20 May 2015 at hours page Joseph Karbowski to the popular works and universal agreement (EE 2. 1, 1218 b 32 6). The fourth premiss is an assumption which he thinks is obvious or unobjectionable (EE 2. 1, 1218 b 36 7), and the fifth premiss is established by an argument that appeals to the connection between a thing s function (ergon) and end (telos) (EE 2. 1, 1219 a 6 11). Finally, the seventh premiss of the argument is established by induction (ἐπαγωγή) from particular craft examples (EE 2. 1, 1218 b a 5). This last argumentative strategy deserves a closer look, because induction from craft examples is a recurring argumentative strategy in the EE:... let it be assumed further, concerning virtue, that it is the best disposition or state or capacity of each of the things that have some use or function. This is clear from induction, since we consider things to be this way in all cases. For example, a cloak has a virtue, since it has a function and use, and its best state is its virtue. The same applies to a boat and a house, and so on, and hence to the soul, since it has a function. (EE 2. 1, 1218 b a 5) In this passage Aristotle inductively establishes a general thesis about virtue that it is the best state or disposition of a thing that has a use or function by appeal to cloaks, boats, and houses, and then applies it to the soul because it has a function too. This argument is what Aristotle elsewhere calls an example (paradeigma): it draws a novel conclusion about a particular target subject (the soul) via the application of a generalization that has antecedently been established by appeal to other items similar to the target (Pr. An , 68 b a 16; Rhet. 1. 2, 1357 b 26 36). 39 Aristotle sometimes describes argument by example as a kind of induction (e.g. at Post. An. 1. 1, 71 a 9 10), because its first step involves inductively supporting a generalization; but in his more precise moments he is careful to distinguish it from induction, because examples proceed from the particular to the particular, while induction proceeds from the particular to the universal (Rhet. 1. 2, 1357 b 26 9; Pr. An , 69 a 16 19). This passage is the only occurrence of the technical mode of argument by example in the first two books of the EE, whereby Aristotle both establishes a generalization and applies it to a novel target case. More frequently, he cites particular examples from the 39 For further discussion of this mode of argumentation see J. Allen, Inference from Signs: Ancient Debates about the Nature of Evidence (Oxford, 2001),

17 Created on 20 May 2015 at hours page 209 Phainomena as Witnesses and Examples 209 crafts as inductive support for certain generalizations without subsequently applying them to other particulars (EE 2. 1, 1220 a 28; 2. 3, 1220 b 30). Although this strategy only corresponds to the first inductive stage of the technical mode of argument by example, we may nonetheless consider the particular items that he invokes as examples (paradeigmata). For they are intended to be familiar items that constitute evidence for certain generalizations by being obvious or uncontroversial instances of them, and Aristotle typically introduces them with for example (οἷον). In fact, the occurrence of οἷον is so frequent in EE that it is hard to doubt that Aristotle conceives of examples as a key part of his method. 40 This presumption is confirmed by the first lines of EE For in them Aristotle claims to be seeking conviction through arguments using the phainomena as witnesses and examples (1216 b 26 8). Scholars who discuss the Eudemian method seldom draw attention to the witness and example qualifications here, 41 but their presence suggests that they are important for Aristotle. In particular, by labelling the phainomena witnesses he is most likely indicating that they will have an evidential role in the subsequent enquiry; 42 and by calling them examples (paradeigmata) he is ostensibly specifying another important role (or pair of roles) 43 the phainomena will play. 40 See EE 1. 7, 1217 a 37; 1. 8, 1217 b 36, 38; 1218 a 35; 1218 b 19; 2. 1, 1219 a 2, 14, 16, 20; 1220 a Dirlmeier offers some brief remarks about the terms μαρτύριον and παράδειγμα in his commentary (Eudemische Ethik, 184). However, Kullmann is the only scholar who pays more than lip service to the use of witnesses and examples in the EE: W. Kullmann, Wissenschaft und Methode: Interpretationen zur aristotelischen Theorie der Naturwissenschaft (Berlin, 1974), He rightly identifies the subsequent confirmation of the definition of happiness via harmonization with ἔνδοξα at EE 2. 1, 1219 a 39 b 8, as an instance of this strategy. But even he seems to underestimate how often Aristotle uses φαινόμενα as witnesses and examples in EE For Kullmann mentions only EE 2. 1, 1219 a 39 b 8, as the corresponding passage where this strategy is employed. By contrast, if the current interpretation is correct, Aristotle uses φαινόμενα as witnesses and/or examples whenever he provides evidence for or illustrates the premisses of his central argument in EE The testimony of witnesses was one of the main types of evidence used in legal contexts: see D. C. Mirhady, Athens Democratic Witnesses, Phoenix, 56 (2002), But the term was appropriated by philosophers to refer to their central evidence for a thesis or conclusion; cf. Plato, Gorg We can distinguish two roles that examples play in EE : (1) an evidential role in which they serve to support or establish important theses (EE 1. 8, 1217 b 36; 1218 b 19; 2. 1, 1219 a 2, 20; 1220 a 24) and (2) a non-evidential role in which they merely serve to illustrate antecedently plausible theses (EE 1. 7, 1217 a 37; 1217 b 38; 1218 a 35; 2. 1, 1219 a 14, 16). Admittedly, the line between these two uses is sometimes blurred, and it is unclear how sharply Aristotle himself distinguished them. However, what