8

Sophia in the Eudemian Ethics¹

Christopher Rowe

By phronesis [wrote Werner Jaeger in 1923] the Eudemian Ethics [hereafter “EE”] understands, like Plato and the Protrepticus, the philosophical faculty that beholds the highest real value, God, in transcendental contemplation, and makes this contemplation the standard of will and action; it is still both theoretical knowledge of supersensible being and practical moral insight.²

Jaeger’s claim about phronêsis in the Eudemian Ethics (EE) provided the core of his wider thesis of a development in Aristotle’s thinking about ethics, from a “reformed Platonism” in the EE to the “late Aristotelianism”³ of the Nicomachean Ethics (NE). The claim was widely, and rightly, disputed, ultimately receiving the coup de grâce in Anthony Kenny’s 1978 book The Aristotelian Ethics (AE).⁴ As Kenny pointed out, “[Jaeger’s] account of the nature of phronêsis in the NE rested almost entirely on the description of that virtue in [the second of the ‘common books’ (hereafter, ‘CB 1–3’)].” Since Kenny had himself shown

that there are strong stylistic reasons for regarding [CB 1–3], as they stand, as belonging with the Eudemian Ethics, it would be possible to argue with as much justification as Jaeger that this Ethics, containing as it does the developed theory of phronêsis, must be the work of the later, mature Aristotle.⁵

In 1971, I had myself argued that despite the overwhelmingly negative reception of Jaeger’s thesis about phronêsis in the EE and the NE, in the terms in which it was stated, nevertheless that thesis was right: “Aristotle [in EE] envisages no essential distinction between the two kinds of wisdom, speculative and practical.”⁶ Like Jaeger, and indeed under his spell, I assumed in 1971 that the CB belonged to the NE, but with the difference that I proposed that two of them, i.e. what we habitually call NE V and VII, could be shown to be reworkings of original Eudemian material, as could other parts of the NE that overlapped with the EE⁷; and I duly set out to try to show it. “NE VI,” by contrast, including as it does the distinction between practical and theoretical reason, and so prominently, had on this account to be a new addition. But if Kenny’s statistical arguments had truly demonstrated that the CB originally belonged to the EE “as they stand,” little seemed left of my position. Once “NE VI” became EE V, the Eudemian account of phronêsis would be identical with what I, following Jaeger, had presupposed to be exclusively Nicomachean; and actually the “NE VI” account fitted at least as well, according to Kenny, with the undisputed books of the EE as it does with the undisputed books of the NE. This he showed under four heads. Phronêsis is

(1) an intellectual virtue concerned with the truth about mutable matters and the whole good of man. (2) … the virtue of a particular part of the rational soul, … distinguished from other intellectual virtues by being deliberative rather than intuitive and practical rather than theoretical. (3) … indissolubly wedded to moral virtue … (4) The union of phronêsis and moral virtue is dependent on the pre-existence of certain natural qualities, intellectual and affective.⁸

The net outcome is indisputable: so far as phronêsis is concerned, there is no incompatibility between the undisputed books of the EE and the second of the “common” books.

But if Kenny is right, then we also no longer have good enough reasons for calling the “common books” exactly “common,” since they will actually have been part of the EE.⁹ Quite apart from Kenny’s stylistic arguments, as presented in the AE, there was always the important but unanswered question why on earth in planning the NE Aristotle would have wanted to include two separate discussions of pleasure, in book VII and in book X – discussions which not only offer quite different treatments of their subject but fail to acknowledge the fact, containing as they do not a trace of a cross-reference between them. But then why not hand over one of those discussions to the EE, which is crying out for it? And if there is no reason against that, why not similarly hand over the rest of NE VII, too?

And so on. In short, if Kenny is right, there is now apparently reason enough for editors and translators to print the three “common” books as part of the EE, and – if both works are included in the same volume, and the EE comes first – no particular reason to print them again, as part of the NE. If, on the other hand, the NE is being printed separately, it will appear oddly depleted and lacunose if a gap is left between books IV and VIII, as it presumably did to whoever it was that first had the idea of importing three books of the EE into the NE.

But this is not the end of the story. One interesting new consideration is that there turns out to be a primary manuscript, i.e. one that does not derive from the three so far recognized as primary,¹⁰ that labels EE VII–VIII as IV and V, with no indication that anything is missing between III and IV apart from the problem of any unfulfilled promises in the first three books that we might have expected to be fulfilled before IV = VII. Since it is undoubtedly descended from the same single archetype as the other three, which in different ways treat the CB as an integral part of the EE, this may be no more than an aberration; but it does suggest that in the early 15th century, when the manuscript was written, there were at least some people who thought that “Aristotle’s Eudemian Ethics” consisted only of five books.¹¹ If this is, admittedly, merely circumstantial, a more substantial consideration is that while the three CB may in some important respects fit better in between EE III and VII than they do between NE IV and VIII, that is not to say that the snugness of the fit is quite complete. One may wonder, for example, about the fact that if NE VI is also EE V, then Aristotle will raise the question of the horos, and about the relationship between practical and theoretical thinking, twice over in the same work, without forward or backward reference,¹² if not quite in the same terms; in other words, the same sort of problem arises in relation to these subjects in an eight-book EE as arose in relation to pleasure in a ten-book NE.

This particular case can perhaps be reduced to a mere matter of untidiness, in the way that – in effect – Kenny suggests,¹³ and as such should worry us no more than the kinds of minor unevennesses, gaps, repetitions, or part-repetitions that we expect to find, and do find, in the Aristotelian treatises in general.¹⁴ These are not, mostly, intended or written as finished, literary works, and Aristotle’s style of argument is often one of starting and stopping, turning back on himself, beginning again, and so on. What we are used to treating as single treatises are in any case often assemblages of smaller, originally separate treatises, as notoriously in the case of the Metaphysics – and as in effect the NE would already be by virtue of its borrowing books from the EE; re-editing and revision, too (what I earlier called “reworking”) may also leave its mark.¹⁵ So, let us suppose that the EE, with its missing books back in place, is more or less of one piece. I say “more or less”: “book” VIII looks and is fragmentary: it is unusually short, and both starts and breaks off abruptly. For its part, book VII, on friendship, is or could be self-standing,¹⁶ like its counterpart in the NE, i.e. NE VIII–IX. Still, the EE with CB 1–3 feels reasonably complete, at any rate up to a point, and as Aristotelian treatises go; as complete, certainly, as the NE used to be, and indeed continues to be if it is allowed to keep its borrowed material. But this restoration to the EE of its original central books “as they stand” (to repeat Kenny’s phrase) is still not, I think, quite straightforward. I propose to spend the rest of this chapter discussing what I see as the remaining difficulties, and exploring a solution.

The issue on which I shall mainly focus is the treatment of sophia in the EE and the NE. While granting that, by and large, phronêsis is handled in the same way in the EE (with or without CB 1–3) and in the NE (ditto), I notice that the same is not obviously true of sophia. This becomes clear if we use the same sort of test as Kenny applied in the case of phronêsis,¹⁷ i.e. looking at the “doctrine … concerning” it first in the relevant “common” book, then in the NE without that book, then similarly in the EE without it.

The first part is easy enough. Sophia as described in CB 2 (i.e. EE V/NE VI) is (1) reserved for the highest kind of intellectual accomplishment or mastery, rather than the accomplishment of experts in particular skills (EE/NE 1141a9–16). (2) Sophia is the most precise of the kinds of knowledge, involving knowledge both of what follows from first principles and of first principles themselves; thus a combination of nous and of systematic knowledge, epistêmê, of the highest things (1141a16–20).¹⁸ (3) Sophia is distinct from politikê and phronêsis in not being concerned with things human and things one can deliberate about, or about bringing things into being (1141a20–b22). (4) Sophia and phronêsis are the excellences of two separate parts of to logon echon in us (1139a5–15, 1144a2). (5) Sophia produces happiness by being a part of it, as do ethical excellence and wisdom combined (1141a1–9). (6) Sophia is not prescribed to by phronêsis, but rather phronêsis prescribes on its behalf, seeing to it that it comes into existence (1145a6–11).

When we look at the NE minus the CB, we find nothing incompatible with any of these six features, and plenty that is either compatible with them or better than compatible. In book I, sophia tis appears as one of the things people have identified with eudaimonia, alongside aretê and phronêsis (and pleasure): “For some people think it is excellence (aretê), others that it is wisdom (phronêsis), others a kind of intellectual accomplishment (sophia tis: 1098b23–4).” Sophia without qualification is used as an example of an intellectual excellence, the other examples used being sunesis and phronêsis (1103a5–6); the justification for distinguishing intellectual from ethical excellences is then framed with sophia and sunesis alone, without phronêsis.¹⁹ Then, in book X, sophia is what is achieved by, the goal of, philosophia, the love of it (sophia); it is the highest aretê, the aretê of the highest part of us, activity in accordance with which is eudaimonia (1177a12–27). This passage can then be connected, retrospectively, and is surely intended to be so connected, with the parenthesis at I 7, 1098a16–18: “if all this so, the human good turns out to be activity of soul in accordance with excellence (and if there are more excellences than one, in accordance with the best and the most complete).” In 1177a29 and 33, the sophos is the person who lives the reflective or theoretical life successfully (and the more sophos he is, the more capable he will be of carrying on his reflections even on his own: a32–4); similarly at 1179a30 and 32.

In all these instances, with the possible exception of the second, sophia/sophos appears to be used in a way that conforms precisely with the treatment of sophia in CB 2. All six features of sophia listed above as deriving from that treatment are either actively or implicitly in play, especially in NE X but also, by virtue of the forward reference – assuming that that is what it is – at 1098a16–18, in NE I too. Indeed, sophia as defined in CB 2 comes to play the starring role in the discussion of the best life that (all but) rounds off the NE as a whole: see 1177a24, 29, 32, 33, 1179a30, 32. This is a special Aristotelian sophia, as book X confirms (i.e. in 1177a12–27). Sophia as used in ordinary speech – from which the specialized Aristotelian variety is distinguished at (CB 2) 1141a9–16 – is also in evidence here and there: see I 4, 1095a21; IV 7, 1127b20; IX 2, 1165a26; X 8, 1179a17. Maybe it is to distinguish Aristotelian sophia from this ordinary, common-or-garden sophia/sophos,²⁰ which is probably hardly distinguishable from ordinary, common-or-garden uses of phronêsis/phronimos, implying a quite unspecific – sometimes supposed, sometimes real – intellectual superiority, that Aristotle adds the tis in 1098b23–4: “For some people think it is excellence, others that it is wisdom (phronêsis), others a kind of intellectual accomplishment (sophia tis).” What he is hinting at here is, perhaps, already the technical sense of sophia introduced in (NE/EE) VI; it is hard to think of anyone else who identifies eudaimonia with sophia as opposed to phronêsis – which has already done duty on its own, a couple of Stephanus pages before, as a representative of intellect in a listing of goods (I 6, 1096b23–5).²¹

When we turn to the undisputed Eudemian books, by contrast, whether preceding or following the putative EE V, there appears to be hardly a trace of the special kind of sophia introduced in that book; the only sophia that is indisputably in evidence is that of ordinary speech. Sophia and sophos appear in total only six times: (1) 1215a23, (2) 1220a6, (3) 1220a12, (4) 1243b33, (5) 1243b34, and (6) 1248a35.²² Of these occurrences, none displays all of the features of sophia/sophos as defined in EE V/NE VI, and all but one display none. (1) refers to what we traditionally term the “older and wiser” (kathaper tines ôithêsan tôn sophôn kai presbuterôn). (2) is the Eudemian counterpart of NE I, 1103a5–6, giving examples of intellectual aretai; while the Nicomachean version gives us sophia, sunesis, and phronêsis, the Eudemian version has just sunesis and sophia (epainoumen gar ou monon tous dikaious alla kai tous sunetous kai tous sophous). In (3), sophia is again an example of an intellectual aretê, along with deinotês (ou gar legomen poios tis to ethos, hoti sophos ê deinos, all’hoti praos ê thrasus). In all of (1)–(3), sophia is used in a wholly untechnical and not the technical Aristotelian sense (the “we” in the last two cases surely refers to people in general). (4) and (5) are interesting insofar as they recall a familiar problem, about how to weigh wisdom/expertise (sophia) against wealth (“if one person complains that he has given wisdom, the other that he has given wealth”),²³ but the context is a discussion of justice in exchange, and sophia here need not be any particular kind of wisdom (or expertise). (6), for its part, is at least prima facie more interesting, as the only place in the undisputed books of the EE where phronêsis and sophia – actually, phronimos and sophos – appear together. I cite the whole sentence²⁴: alogoi gar ontes epitunchanousi²⁵ kai tou tôn²⁶ phronimôn kai sophôn tacheian einai tên mantikên, kai monon ou tên apo tou logou dei apolabein: “though lacking the reasoning they match even the speediness of divination achieved by [sc. practically] wise and intellectually accomplished people, so that one could almost take it for the divination engineered by reasoning.” The sentence describes those possessed of a special eutuchia that enables them to get things right over and over again despite their lack of reasoning: they can see or “divine” what they should do almost as if they were able to reason, whether like phronimoi or like sophoi. What Aristotle is describing looks not unlike the operation of nous, tôn eschatôn ep’amphotera, at CB 2, 1143a35–b5, as it grasps (“divines”?) the particular, whether in the practical or the theoretical sphere – only here without the phronêsis or sophia (/epistêmê).²⁷

So, here is one possible mention of sophia in the mode of CB 2 in the EE. But it would be hard to insist on it when the term nous, so closely associated with it in CB 2 (sophia being a combination of nous and epistêmê), is consistently used, in the same context, in a quite different way.²⁸ But in any case, the context requires just that sophos/sophia at 1248a35 be just any wisdom that is not specifically practical. I conclude that sophia and sophos are not used in the technical Aristotelian sense, as set up in CB 2, anywhere in the undisputed Eudemian books.

Does this matter for the issue in hand, that is, in relation to the question about the provenance of CB 2? Here is a possible reason why it might not matter. There is nothing in EE I–III or in EE VII–VIII that precludes the same sharp distinction that is conveyed in CB 2 through the opposition between phronêsis and sophia; phronêsis, it seems, is used according to the definition of it there, and given (a) that the intellectual excellences are said in EE II to have two functions, discovering “truth, either about how things are (pôs echei) or about how things come into being (peri geneseôs)”²⁹; (b) that the second of these functions is clearly assigned, in the undisputed Eudemian books in general, to phronêsis; and (c) that sophia is said in CB 2 “not to be of any coming into being” (oudemias … geneseôs),³⁰ maybe there was just no occasion to introduce this special use of sophia.

But this, I think, is an unsatisfactory response. Sophia, in its technical sense, is a key element – even the key element – in the NE’s answer to the fundamental question about the nature of eudaimonia. It is the jewel in the crown, and CB 2 provides the setting (or, better, the backdrop?) for that jewel. Since the EE obviously raises the same fundamental question about eudaimonia, one is bound to ask why there is barely a trace, in the undisputed books, of the centrepiece of the Nicomachean account (i.e. sophia). NE I–IV – I claim – prepares for CB 2 on sophia, of which NE X then makes triumphant use; the EE (minus CB) meanwhile neither looks forward to CB 2 on sophia³¹ nor makes any use of it.

Or is this just down to the fragmentary nature of EE VIII, and to the fact that Aristotle never gets round, in the EE, as he does in NE X 6–8, to giving a direct answer to the question he raises in EE I–II about the good life?

To answer this question, we need to look briefly at the structure and content of EE VIII as a whole. Aristotle does not go out of his way to link VIII 1–3 together, and indeed on the face of it the three chapters are on distinct topics: the difference between phronêsis and epistêmê (ch. 1); good fortune, eutuchia (ch. 2); the horos of kalokagathia (ch. 3). All three, however, can be seen as concerned – in effect, if not by design – with one overriding question, about the role of epistêmê in a successful life of practical activity, the question being approached by way of opposition to Socrates: that is, to Socrates in Plato, as the nature of the references makes clear.³² VIII 1 finds Aristotle returning (without acknowledgement of the fact) to the Socratic identification of the aretai as epistêmai with which he began the discussion of aretê and phronêsis in book I.³³ He concludes that the Socratic saying “nothing is stronger than phronêsis” is correct, only he shouldn’t have substituted epistêmê for phronêsis,³⁴ and the argument for that conclusion is a sort of commentary on Socrates’ claim³⁵ in the Hippias Minor that the person who goes wrong voluntarily is preferable to the person who does so involuntarily. The discussion in VIII 2 of good fortune starts from the premiss that doing well (eu prattein) comes from good fortune as well as from (the combination of) phronêsis and aretê. This is a premiss that Socrates in effect denies, in the Euthydemus, by identifying eutuchia with sophia; Aristotle adverts to the fact at 1247b14–15.³⁶ His question is: if there is such a thing as good fortune, and people really do get things right without apparently being intellectually equipped to do so, and regularly get them right, even continuously, not just every now and then, how does this come about? There then follows that notorious and extraordinary passage (1248a24–b3) in which, if we accept the text handed down by the manuscript tradition, as I think we should, the cause is said to be god, even a god that is somehow within us: an intellect, perhaps, that is indistinguishable – except by location – from the reason that operates in the universe at large, and that can function even unbeknownst to us.³⁷ Even though Aristotle starts in this chapter from an anti-Socratic position, his answer (after a winding, dialectical argument) to the initial question, about the cause of good fortune, seems to me to belong to the same general type as Socrates’ own evocation of a personal daimonion, an inner voice that tells him not to do things he was intending to do, and by implication is right to tell him not to do them.³⁸

VIII 3 has something of the same structure, to the extent that it begins with a set of ideas that contrast with Socrates’, but circles round and ends up using language, and apparently taking up a position, that is distinctly Socratic. The chapter makes what looks like a new start, on the subject of “what we were already calling kalokagathia,”³⁹ i.e. the ethical virtues taken together, explaining the relationship first between the two elements of the term, kalon and agathon, and then, more briefly, in 1249a17–21, between them and the pleasant. Aristotle thus explains the thesis that he announced at the very beginning of the EE, about the convergence in eudaimonia of the fine, the good, and the pleasant, and by doing so ties the end of the work to its grand opening, justifying with a splendid clarity his disagreement there with the inscription from Delos.⁴⁰

But now, having made the good man the measure, as he has, in action and in the choice of haplôs agatha, a.k.a. goods “in the abstract,”⁴¹ Aristotle needs to say just how the good man will do the measuring; it’s not sufficient to say “as reason dictates,” “as we said earlier.”⁴² He proceeds to give his answer: the horos is whatever “choice and possession (? acquisition: ktêsis) of natural goods” will most produce (? poiein) tou theou theôria, contemplation of/reflection on, god; any such choice that “either through deficiency or through excess prevents service to (therapeuein) and contemplation of/reflection on god” will be bad.⁴³ Theôrein “the god” is usually taken as doing Aristotelian metaphysics, and this may be right, but we should probably consider taking “the god” here more closely with what has been said about god and the divine in the preceding chapter,⁴⁴ that is, as something more like “god/reason as operating in the universe (and us).” The addition of therapeuein to theôrein seems to me to point in the same direction. I have argued elsewhere⁴⁵ that the expression therapeuein ton theon would have been recognizable to Aristotle’s audience as referring to Socrates, and that “Aristotle is here treating Socrates as a philosopher and theoretician like himself.”⁴⁶ On this interpretation, the addition of therapeuein will perhaps be for the purpose of clarifying theôrein: what Aristotle has in mind, more than anything, is doing philosophy, in a world that is supremely adapted as a subject for philosophy by the fact of its control by god/reason.

This gesture to Socrates, however, if such it is, is two-edged, insofar as it places Socratic philosophizing outside the very sphere to which he, Socrates, thought he was devoting himself: that is, the sphere of practical activity. A Socrates, like an Aristotle, may theorize about ethics, but the kalokagathos will apparently not require such theorizing in order to act in accordance with his kalokagathia. Socrates proposed that philosophy was itself the key to the good and happy life; not so, says Aristotle – the happy life is one lived according to the ethical aretai, accompanied by phronêsis, or according to kalokagathia, which unites all these into one. EE VIII 1 establishes that the ethical virtues are not matters of epistêmê; EE VIII 2 recognizes a sort of divine inspiration that enables successful lives for some even in the absence of reasoning altogether; EE VIII 3 then finishes off the Socratic position by prima facie freeing kalokagathia itself, and a fortiori activity in accordance with it, from any necessary involvement with theôria, even while endorsing Socrates’ view of the supreme importance of philosophy in itself. As its concluding sentence confirms, this last chapter has been about kalokagathia, and theôria has only entered the picture as providing the horos for it, and the skopos of natural goods/goods “in the abstract”: “so, as to the standard/limit of kalokagathia, and the aim/purpose of [our choice and possession/acquisition of] goods in the abstract, let this stand as our account.”⁴⁷

So where does this leave us, on the subject of eudaimonia, and of sophia, in EE VIII 3? Kalokagathia has just been identified as complete aretê, aretê teleios⁴⁸; and happiness, eudaimonia, was said in book II to be the “activity of a complete life in accordance with aretê teleia.”⁴⁹ But given how Aristotle defines the horos of kalokagathia, i.e. in terms of serving and reflecting on (theôrein) god, this “activity … in accordance with aretê teleia” ought also include reflective (theôrêtikê) activity: it is hard to see how, if one kind of activity is demarcated by its capacity to produce the conditions for another, and the first activity is essential to human eudaimonia, the second activity would not also be part of that eudaimonia. But then it would be odd to suppose that our first activity was meant to be defined by its capacity to produce the conditions for doing the second activity badly. So, if eudaimonia is activity in accordance with complete aretê, then that complete aretê will include the aretê relating to “service to and theôria of god” as well as the aretai that go to make up kalokagathia, and kalokagathia will be aretê teleios at 1249a16 strictly by comparison with doing fine (kala) things merely kata sumbebêkos (1249a14–16).⁵⁰ If the aretê corresponding to theoretical activity is to be given a name, it might as well be sophia. By this reckoning, even if the EE lacks an equivalent to NE X 6–8, and an explicit, final answer to the question about the content of eudaimonia, we can supply the one he perhaps would have given, had he got round to it (EE VIII 3 ends, after all, with the sort of sentence⁵¹ introduced by men oun that usually leads on to a new topic): namely that eudaimonia includes philosophical reflection as well as good practical activity, and that the first is a more important ingredient than the second.⁵²

But this is not the only way of reading the concluding paragraphs of what Kenny⁵³ calls the “cryptogram” that is EE VIII, and I myself have more than a suspicion that it is an overinterpretation, in one respect at least. A point that is often overlooked is that Aristotle sums up his discussion of the horos, i.e. as announced at 1249a21–b3, by saying (in the text as I reconstruct it) “And this the soul has [sc. a faculty for theôria], and this is the best horos, to be least aware of the remaining part of the soul as such” (1249b22–3). This, so far as I can presently see, ought somehow to be the counterpart of what has just been said about the promotion of theôria/therapeia as the criterion for choosing and possessing natural goods (we don’t need another criterion on top of that). If so, then the emphasis in the preceding context should be something to the effect of “and this the best horos for the soul, namely to be most aware of the rational part of the soul as such.” That, I propose, will be when it is philosophizing – serving and reflecting on god – as well as issuing the orders that will keep the non-rational soul in its place. CB 2 would lead us to expect Aristotle here to distinguish two rational parts of the soul, one concerned with reflection, the other with issuing orders, but I suggest that he does not. “And so it is with to theôrêtikon,” he says in 1249b13, referring back to “And this [sc. our archê],” he says, is twofold, for medicine is an archê in one way and health in another, and the former for the sake of the latter; and so it is with to theôrêtikon” – that is, I take it, to theôrêtikon includes both the counterpart of health in the analogy, which must be god (at the heart of the philosophizing self) and that of medicine, i.e. phronêsis. Even if, according to CB 2, theoretical and practical thinking belong to two different soul-parts, the EE is capable of treating them both as “theoretical”: see EE II 10, 1226b25–6 esti gar bouleutikon tês psuchês as to theôrêtikon aitias tinos. But this need not be taken as contradicting CB 2. Aristotle is not here comparing or contrasting different aspects of human rationality, or talking about the choice between the practical and the philosophical life; he is not talking directly about kinds of life at all. We would expect him to talk about lives again at some point, given that he raised the subject at some length in book I,⁵⁴ but he is not raising it here. Perhaps he answered it in some other part of the fragmentary book VIII, now lost; perhaps not.

In sum, what is uppermost by the end of EE VIII 3 is the idea that we need to lead rational lives, whether practical/political or both practical/political and theroretical. On this reading, that sophia of the CB 2 type does not figure is not because the opportunity does not arise, but rather because the Eudemian Aristotle does not show, or have, the same level of interest in sophia as he betrays in the NE – an interest that runs in a direct line from the beginning of the work, through CB 2, to X 6–8. With regard to sophia, then, CB 2 meshes more closely with the undisputed Nicomachean books – both the earlier and the later ones – than with the undisputed Eudemian ones. How to explain this state of affairs, if (as I am still minded to accept) the disputed books belonged originally to the EE? I have no definitive answer to this question. I am, however, struck by the fact that when we compare the text of EE I–II, the three references to sophia in NE I (i.e. references to sophia in Aristotle’s technical use) look like additions specifically designed to look forward both to CB 2 and to NE X, binding them into a whole that is rather more unified, in this single but absolutely central respect, than the corresponding sequence EE I–II–CB 2 EE-VIII. I also note that all three references in NE I can be removed without any damage to the syntax of the sentences to which they belong. I speculate that they were the work of whoever was mainly responsible for organizing what we know as the ten-book NE (Nicomachus?). I also speculate that what we know of as EE VIII is a fragment of an Ur-EE V, abandoned in favour of another version, including some of the same topics but dealing with the intellectual virtues more systematically,⁵⁵ and written not long after the original version⁵⁶; perhaps this Ur-EE fragment was installed in its present position, as EE VIII, by another editor, by false analogy⁵⁷ with NE X, after that was in place in the NE.

Both our ten-book NE and our eight- or five-book EE fall, and probably always fell, considerably short of the kind of unity suggested by labelling them “treatises.” No more, as I said earlier, do many Aristotelian works. But the EE and the NE prima facie look different, insofar as they both have or appear to have beginnings and endings, and links between those beginnings and endings. In between, however, in both, are parts that sometimes sit well with these beginnings and endings, and with each other, and sometimes fit less well, thus suggesting that that first appearance is at least to some degree deceptive. Aristotle is a specialist in highly focused discussions of particular topics, and it is, by and large, not his habit to set out the big picture, systematically, in any given area. He may have intended to make an exception for ethics, but even here the results often betray Aristotle’s preference for dealing with one sub-topic at a time (friendship; akrasia; pleasure). Maybe there was once a fully worked and fully organized EE, just as someone tried – and failed – to construct a fully coherent NE. But I would not bet on it. My own guess is that any original EE/collection of Eudemian books might have been hardly less uneven than the current version. I also hazard, on the basis both of the retention of the treatment of pleasure in what became NE VII, and of the apparent stylistic affinities to the EE of all three “common” books, that the degree to which EE IV–VII were reworked to fit them into their new home in the NE was relatively small. It may even be that the most extensive revision related to the style of the three books: not style as measured by Kenny’s stylometrics, which for the most part reduces to word counts, but rather style that can be, for example, golden, flowing, and full, or crabbed, awkward, or elliptical. EE I–III and VII–VIII, by and large, belong to the second category, NE I–IV and VIII–X to the first – as, in my long experience as translator of the ten-book NE, do CB 1–3.⁵⁸ Here is a final hypothesis. After writing whatever corresponded to what we may choose to call the original EE, that is, however rough and unfinished a collection of materials it might have been, Aristotle went on to write new ethical works – “books” – that nevertheless covered roughly the same territory as EE I–III and VII, in (what I claim to be) a more expansive style; he also wrote similarly expansive, flowing versions of the three intervening Eudemian books⁵⁹ that were much closer to the originals, one of the latter being the second EE V, i.e. the successor to the one that originally included the fragment we call EE VIII (or perhaps the fragment was always the whole?). But after that he also decided to write a completely new treatment of pleasure, along with the crowning chapters X 6–8, which Nicomachus could put together with the rest to make the (half-)finished article.⁶⁰

Notes

1. This chapter represents a belated corrective to Rowe, 1971a, which is a longer version of the identically titled chapter II 1 of Rowe, 1971b, published later in the same year; in what follows I shall quote from the later version (henceforth “The EE and NE”). Earlier versions of this chapter were delivered at the Centre Léon Robin, Paris, in November 2016, to the “B” Club in Cambridge in May 2017, and to the Dublin workshop on EE and NE in June 2018.

2. Jaeger, 1934, p. 239. Jaeger’s position here relies heavily on Kapp, 1912.

3. Jaeger, 1934, p. 231.

4. Kenny, 2016, including “Reconsiderations 1992” and “Reconsiderations 2016”; hereafter “AE.”

5. Kenny, AE, p. 161.

6. Rowe, The EE and NE, p. 69.

7. An idea that – though little has been added to it since 1971 – continues to resurface: thus, e.g. “the common books may have been somewhat revised for reuse in the Nicomachean Ethics,” say Rafael Woolf and Brad Inwood in the introduction to their translation of the Eudemian Ethics, see Inwood, Woolf, p. ix, while Hendrik Lorenz, reviewing Woolf and Inwood in Notre Dame Philosophical Reviews (1 October 2013) says “On stylistic grounds many scholars think that the common books were originally written to form part of the Eudemian Ethics, and so the recent trend to include them in translations of that work has something to be said for it. However, it is entirely possible, and in fact probable, that Aristotle himself revised and updated his discussions of such important topics as the nature of justice, the intellectual virtues, and lack of self-control for, and perhaps also after, inclusion in the Nicomachean Ethics. Furthermore, there are fairly strong historical reasons for thinking that these three books were included in a ten-book edition of the Nicomachean Ethics, long before someone decided to fill the large and rather conspicuous gap [caused by the removal of three books for modification and reuse in NE?] in the middle of an earlier five-book edition of the Eudemian Ethics by inserting the corresponding Nicomachean books, thus creating the appearance of a complete version of the Eudemian Ethics and at the same time presenting those three books as common to the two treatises.” Lorenz adds in a footnote “The reasons [for so presenting them] are presented with clarity and force in O. Primavesi, “Ein Blick in den Stollen von Skepsis: Vier Kapitel zur frühen Überlieferung des Corpus Aristotelicum,” Philologus 151.1 (2007), 51–77.” I shall return to these issues later.

8. Kenny, AE, p. 163.

9. I still leave open the possibility that they belong to the NE more closely than simply through being borrowed “as they stand” from the EE. I shall explore this possibility below. (See Dorothea Frede’s broadside in Frede, 2019, which is nter alia another useful reminder, alongside e.g. Lorenz’s review [n.7 above], that resistance to the view of the CB as the Eudemian is by no means dead.)

10. I.e. (1) P = Vat.1342, (2) C = Cant.I.i.5.44, and (3) L = Laur.81,15, to which can now be added (4) what was once Phillipps 3085 and is now Mon. 635, which I label as “B” (the Bayerische Staatsbibliothek having bought it at Sotheby’s in 1976).

11. It is indeed the shared view of Renaissance scholars, and of their Byzantine predecessors, that the CB belong to the NE: the evidence is complex, but see e.g. the note by Cardinal Bessarion on f. 270^v of Par.2042 (a collection of excepts by him from Aristotelian treatises).

12. See Rowe, The EE and NE, pp. 109–113.

13. Kenny, AE, pp. 181–182 (followed by direct criticism of the pages in “The EE and NE” referred to in the last note); cf. Kenny, 1992, p. 100.

14. In which case, I note, the argument from the double treatment of pleasure will also lose its force as an argument against the integrity of our ten-book NE.

15. Cf. e.g. Burnyeat, 2004, p. 179: “The surviving treatises, unlike the ‘exoteric’ works he sent to the booksellers, remained with him, always available for additions, subtractions, and other forms of revision.”

16. Cf. Kenny, AE, p. 42.

17. See above.

18. “Nous“Nous and systematic knowledge – systematic knowledge, as it were with its head in place, of the highest objects,” a19–20 (translations are my own, from Broadie, Rowe, 2002, except that I here leave nous untranslated for reasons that will become apparent).

19. 1103a7–8 “For when we talk about character, we do not say that someone is accomplished in a subject (sophos), or has a good sense of things (is sunetos), but rather that he is mild or moderate; but we do also praise someone accomplished in something for his disposition, and the dispositions we praise are the ones we call ‘excellences.’” (“In a subject” and “in something” are, of course, fillers, which are justified if, as I think, Aristotle here finds it convenient to separate off his specialized sophia by reference to sophia more generally – though, of course, even the first is about something.)

20. A model for the Nicomachean sophos might be the Thales of Plato’s Theaetetus, representing the “leaders of the [philosophical] chorus”: Theaetetus 173c–174c.

21. Kenny, in correspondence, says he thinks the uses of sophia in NE I are untechnical; “sophia tis means ‘some kind or other of intellectual accomplishment’ rather than ‘the special [sophia] that is exercised in contemplation.’” But I think that NE I, as it stands, shows a clear interest in distinguishing sophia from phronêsis, and it is hard to see why we should look for a basis for this distinction other than that offered by CB 2.

22. Thomas Case, in his seminal Encyclopaedia Britannica article on Aristotle (1910), claimed that the distinction between phronêsis and sophia familiar to us from NE  VI/EE  V was not well “prepared for” in EE I–III, but was present in EE VIII (he concludes on this basis that “probably therefore this part [of EE] was a separate discourse”), and cites “1246b4 seq, 1248a35, 1249b14.” (EE as a whole, Case thought, was a preliminary sketch for the NE; the treatment of phronêsis in I–III he calls “a chaos” [p. 31 of the reprint of Case’s article in Wians, 1996.) The first and third of these passages do not mention sophia, and it must therefore be a moot point whether they contain a reference to the distinction in question.

23. Cf. Protagoras’ practice of saying to his students “pay me what you think my teaching is worth”?

24. In the text as it will appear in my new Oxford Classical Text of EE, due for publication in 2023.

25. Susemihl, following the Latin version in the epitunchanousi Liber de bona fortuna, which is testimony to a Greek text earlier than any of our Greek manuscripts, i.e. P, C, B, and L: apotunchanousi PCBL.

26. tou tôn Sylburg: tôn PCBL ( tôn accounts for the following infinitive; tôn as it stands is little short of nonsense: why “these”? We would anyway expect tôn tôn).

27. Alternatively, or additionally, Aristotle’s sophôn at 1248a35 covertly refers to the thesis with which Socrates shocks the young Clinias in Plato’s Euthydemus, that eutuchia is actually sophia: 279d6. See further below.

28. I.e. for intellect or reason in general (1247a30, 1248a21, 29, 32).

29. II 4, 1221b30.

30. 1143b20.

31. Cf. Case as cited in n. 22 above.

32. According to W. Fitzgerald, Sôcratês preceded by the definite article refers to the Socrates of the Platonic dialogues, Sôcratês without the article to the historical Socrates (see Fitzgerald’s Selections from the Nicomachean Ethics of Aristotle, Dublin 1853); “Fitzgerald’s canon” is subsequently enshrined, but unattributed, in the 9th edition of Liddell–Scott–Jones’s Greek Lexicon, s.v. ho, hê, to). That does not hold for the EE: see n. 36. One is tempted to explain this by supposing that Bishop Fitzgerald thought the EE spurious; but then the rule probably does not hold consistently for the NE either. Burger, 2008, is inclined to see the NE as a whole as a conversation with Plato’s Socrates. A work or works on ethics written (as I suppose) not so long after Plato’s death could not avoid being in some sense a dialogue with the Platonic Socrates; here in EE VIII Socrates is more immediately the moving cause.

33. I 5, 1216b2 ff.

34. EE VIII 1, 1246b34–6. The outcome of the chapter is a somewhat less expansive version of CB 2, 13, 1144b17–30.

35. Or takes off from it; in any case the connection between VIII 1 and the Hippias Minor seems to me to be close and undeniable.

36. “As Sôkratês said”: the reference here to Euthydemus 279d6 is generally recognized.

37. So a nous after all: see text to n. 28.

38. Rowe, 2021.

39. Not in the undisputed books of the EE, or in the CB, but see NE IV 3, 1124a4, X 9, 1179b10: another illustration, perhaps, of the complex history of what we call the EE.

40. I 1, 1214a5–6.

41. Kenny’s translation of the phrase, the reference being to things that are good if one “abstracts” them from the sort of person who possesses them (the “natural” goods): they will be good only if possessed by the good person, while in the wrong hands they can be positively bad. The standard translation of haplôs as “without qualification” is – as I now see – to say the least unhelpful.

42. 1249b3–4; see CB 1, which says about the formula kata ton orthon logon almost exactly what EE VIII says about hôs ho logos here: alêthes men, outhen de saphes.

43. 1249b16–21.

44. Cf. Kenny, AE, p. 175.

45. Rowe, 2013.

46. Rowe, 2013, p. 322. I also bring in, in this context, Sarah Broadie’s important essay, Broadie, 2003.

47. 1249b23–5.

48. 1249a16.

49. 1219a38–9. (The adjective teleios appears as indiscriminately two- and three-termination in both the EE and the NE.)

50. And so will not refer back to book II after all.

51. A sentence that Donald Allan proposed to excise, for no reason that has appeared in print.

52. Cf. Verdenius, 1971, p. 297.

53. Kenny, AE, p. 178.

54. “There are three lives that everyone chooses to live who has the choice, the political life, the philosophical, the life of enjoyment,” 1215a35–b1; picked up again at 1216a27–9.

55. That is, rather than concentrating, as “EE VIII” does, on overturning the Socratic position and establishing that practical wisdom is not a matter of scientific knowledge. If the fragment was itself once a part of a wider discussion of the intellectual excellences, it is hard to think, given the terms it uses in dealing with the issues it does deal with, that the larger whole would have included discussion of sophia and nous in the specialized uses in which they are introduced in CB 2.

56. Given that, stylometrically, CB 2 appears inseparable from CB 1 and CB 3, themselves stylometrically Eudemian.

57. The analogy might perhaps have been encouraged by the link (see above) between the first part of VIII 3 and the rather grand opening in EE I; but it is a false analogy because, as I have argued, despite some overlap EE VIII and NE X 6–9 are ultimately about different things.

58. Kenny, in correspondence, is sceptical about this claim, and/or about its importance, on the grounds that it is unquantified (and perhaps unquantifiable?). I think nonetheless that the differences are real. Even – or perhaps especially – after Kenny’s stout defences of his statistical methods in “Reconsiderations 2016” in AE, I continue to worry about how exactly his measurements themselves relate to style. At AE, p. 280, Kenny cites G. R. Ledger’s criteria for variables to be used “in a statistical inquiry into style,” the third of which is that “they must have a fair chance of being linked with stylistic features”: what, I want to ask, is the precise nature of the link in this case? The statisticians’ reply is that this does not matter; something is being measured, and that something exhibits significant differences over given quantities of text. But if statistics do not lie, they do not tell the whole story. If some aspects of style go unseen by author and by reader, other aspects are surely entirely visible to both; despite the numbers, I still claim that EE I–III and VII–VIII are written differently from NE I–IV, VIII–X, and CB 1–3. (For criticism of Ledger’s methods, see Young, 1994.)

59. To make them accessible to a wider public? Given that on any analysis there is a huge overlap between the EE and the NE, that seems a decent enough explanation.

60. My thanks to Tony Kenny and Malcolm Schofield for their comments on versions of this chapter. The blame for surviving missteps remains entirely mine. The arguments of the present essay are developed further, and its conclusions refined and modified, in Rowe forthcoming 1 and Rowe forthcoming 2.

References

1. Broadie, S., ‘Aristotelian Piety’, Phronesis 48, 2003, pp. 54–70.

2. Broadie, S., and Rowe, C., Aristotle, Nicomachean Ethics: Translation, Introduction and Commentary. Oxford: Oxford University Press, 2002.

3. Burger, R., Aristotle’s Dialogue with Socrates: On the Nicomachean Ethics. Chicago: The University of Chicago Press, 2008.

4. Burnyeat, M., ‘Aristotelian Revisions: The Case of De Sensu’, Apeiron 37, 2004, pp. 177–180.

5. Frede, D., ‘On the So-called Common Books of the Eudemian and the Nicomachean Ethics’, Phronesis, 64, 2019, pp. 84–116.

6. Jaeger, W., Aristotle: Fundamentals of the History of his Development, English tr. of Aristoteles, Grundlegung einer Geschichte seiner Entwicklung, Berlin 1923, by Richard Robinson. Oxford: Oxford University Press, 1934.

7. Kapp, E., Das Verhältnis der eud. zur nik. Berlin: Ethik, 1912.

8. Kenny, A., Aristotle on the Perfect Life. Oxford: Oxford University Press, 1992.

9. ———, The Aristotelian Ethics (1st ed. 1978). Oxford: Oxford University Press, 2016.

10. Rowe, ‘The Meaning of phronêsis in the EE’, in Moraux, P. and Harlfinger, D. (eds.), Untersuchungen zur Eudemischen Ethik. Berlin: de Gruyter, 1971a, pp 73–92.

11. ———, ‘Aristotle and Socrates on the Naturalness of Goodness’, in Sattler, B. and Coope, U. (eds.), Ancient Ethics and the Natural World. Cambridge: Cambridge University Press, 2021, pp. 203–217.

12. ———, ‘A History of the Text of the Eudemian Ethics’, in Jimenez, M., Gartner, C., and Bobonich, C. (eds.), Aristotle’s Other Ethics. Forthcoming.

13. ———, The Eudemian and Nicomachean Ethics: A Study in the Development of Aristotle’s Thought, Proceedings of the Cambridge Philological Society, Supplement 3, 1971b.

14. ———, ‘Nous in Aristotle’s Eudemian and Nicomachean Ethics’, in Ramelli, I. (ed.), Human and Divine Nous from Ancient to Byzantine and Renaissance Philosophy and Religion: Key Themes, Intersections and Developments. Leiden: Brill. Forthcoming.

15. ———, ‘Socrates and His Gods: From the Euthyphro to the Eudemian Ethics’, in Lane, M. and Harte, V. (eds.). Cambridge: Politeia, 2013, pp. 313–328.

16. Verdenius, W. J., ‘Human Reason and God’, in Moraux, P. and Harlfinger, D. (eds.), Untersuchungen zur EE. Akten des 5. Symposium Aristotelicum, Berlin: de Gruyter, 1971.

17. Young, C. M., ‘Plato and Computer Dating’, Oxford Studies in Ancient Philosophy, 12, 1994, pp. 227–250.

18. Wians, William R. (ed.), Aristotle’s Philosophical Development: Problems and Prospects. Lanham: Rowman & Littlefield, 1996.