2. THE CASE FOR BRACKETING THE THIRD COLUMN

The text printed by Rowe can be traced back to a brief remark made by D.J. Allan in his 1966 review of Dirlmeier’s commentary on the EE.Footnote ⁴ Allan first suggested that the third column was a later interpolation modelled on Nicomachean Ethics II 7, where a first succinct characterization of the virtues and vices is provided by reviewing the triads, including the means, one by one (1107a33–1108b10). In effect, in EE II 3 Aristotle proceeds differently: after the table the text of the EE disregards the items in the third column and focusses instead on the extremes in the first two columns. Here is an excerpt from this section of the text of EE II 3 (henceforth ‘the review of extremes’) for the first line of the table: ‘For the irascible person (ὀργίλος) is the one who gets angry more than one ought to, and more quickly, and at more persons than one ought to, the impassive one (ἀνάλγητος) is the one who is deficient in whom he is angry at, and when, and how’ (1221a15–17). Allan conjectured that ‘a later systematiser’ had been responsible for the insertion of the third column in the table of EE II 3 in an attempt to assimilate Aristotle’s EE to the Eth. Nic. and the latter’s approach to the investigation of character virtue. This hypothesis would also explain the appearance of suspicious items in the third column, above all δίκαιον and φρόνησις, for the latter would have been attempts on the later editor’s part to find a suitable match for the two lines in question.Footnote ⁵ Although this is not explicitly mentioned in his brief remark, Allan may have thought that the interpolation would have been facilitated by the space remained available beside the first two columns, whether in papyri or in codices of the text of the EE; it is worth noting, however, that at least in our medieval MSS of the EE the text of 1220b38–1221a12 (viz. the above table) is presented in continuous prose, not in tabular form.Footnote ⁶ Be that as it may, Rowe, unlike the editors of the previous OCT text of the EE,Footnote ⁷ has adopted Allan’s proposal in his new edition.

Another consideration which prima facie speaks in favour of the excision of the third column, not countenanced by either Allan or Rowe, is this: we find back-references to the table of EE II 3 later in EE III, where Aristotle goes on to apply his general account of character virtue to the individual virtues of character (courage, temperance, mildness, etc.); significantly, the back-references inserted in EE III speak only of the extremes focussed on after the table (1228a28; 1230b12;Footnote ⁸ 1231b8; 1233a9). Here are the first and the third of these back-references:

No mention is made here of the virtues of courage and mildness, the first and second virtues—in reverse order—of the table of EE II 3.Footnote ¹¹ On the whole, there seems to be a clear division of tasks between EE II and III: throughout EE II, Aristotle works towards an adequate definition of character virtue by singling out components of the would-be definition and assembling them in a step-by-step fashion; only later in EE III does he derive the virtues from the two opposite vices reviewed in the brief survey of EE II 3. Correspondingly, his procedure in EE III consists in recalling the extremes previously inventoried in EE II 3, as in the two exhibits of evidence quoted above, and then deriving the correspondent virtues from them.Footnote ¹² That being said, we should consider the possibility that in the back-references Aristotle might be referring to the review of extremes rather than the table: if that were to be the case, the back-references would turn out to be of no bearing for the question of the third column; we shall come back to this question in due course.

3. THE ARGUMENT OF EE II 3

From now on, it will be useful to take a closer look at the surroundings of the table to understand the argumentative path Aristotle is travelling alongside the intended function of the table of EE II 3 in its context. As mentioned in the introduction, the opening section of the chapter (1220b21–35) is devoted to introducing the doctrine of the mean. In its bare-bones version, the doctrine states that in every continuous and divisible domain there will be excess and deficiency as well as a mean. As Aristotle’s examples reveal (gymnastics, medicine, housebuilding, steersmanship), the source domain of the doctrine is craft expertise, whereas its target domain in the context of EE II 3 is character virtue; in broad outline, the project undertaken here consists in transferring the apparatus of the doctrine from craft expertise to character virtue. The general presentation of the doctrine leads to the first provisional definition of character virtue, according to which character virtue is concerned with certain means and is itself a certain intermediate (1220b34–35, περὶ μέσ’ ἄττα εἶναι καὶ μεσότητα τινά).Footnote ¹³ Next Aristotle formulates two questions: (i) what kind of intermediate character virtue is (ἡ ποία μεσότης ἀρετή) and (ii) which means it is concerned with (περὶ ποῖα μέσα) (1220b35–36), which are evidently meant to specify further the content of the provisional definition just arrived at. Finally, the next two lines tell us how Aristotle intends to proceed in the ensuing investigation with the introduction of the table:

The table of EE II 3 should thus allow us to make progress with either question (i) or (ii) or both by studying each of the items in the table; the execution of this plan is presumably what we get in the review of extremes after the table. Now, of the two above questions, (i) what kind of intermediate character virtue is and (ii) which means it is concerned with, it is the latter, I submit, which is centre stage in the rest of EE II 3.Footnote ¹⁴ This emerges, on the one hand, in the ensuing section of the EE: in the second interim definition of character virtue in EE II 5 Aristotle states that the latter is concerned with means in pleasures and pains and in pleasant and painful things, whereas he confines himself to repeating that it is an intermediate (1222a10–12). But it is also clear, on reflection, in the light of the lead-up to this first definition in EE II 3: as we have seen, the doctrine of the mean has it that in every continuous and divisible domain there is excess, deficiency as well as a mean. It is natural at this point to expect Aristotle to indicate a specific domain of something continuous and divisible to which the doctrine of the mean will apply in the case of character virtue: this domain will amount to what character virtue is concerned with. Two viable candidates present themselves, namely action (πρᾶξις) and the affections (πάθη): although upon introducing the doctrine of the mean Aristotle points out that change is continuous and that action is change (1220b26–27), thereby giving the impression of intending to focus on that domain, in the remainder of EE II 3 it is clearly affections which are selected as the primary domain of application for the doctrine of the mean (see for example 1221a14–15; 1221b10; b22).Footnote ¹⁵

We will gather evidence to this effect shortly, but first it is worth noting that there are good reasons why Aristotle should turn to affections: first, in the previous sections of the treatise he has stressed that the bearer of character virtue is the non-rational part of the soul (1220a10–11; restated at 1221b30–31) and the latter has been said to be responsible for desire and the affections (1220a1–2).Footnote ¹⁶ Second, by turning to the affections he will be in a position to distinguish the several character virtues depending on the specific affection they are concerned with. Third, if he were to focus on action, it would be harder for him to set apart the character state of the virtuous person from the one with self-control; in general, affections (or at least those relevant to ethical enquiry) provide motivations for actions and, as such, represent a more ‘upstream’ phenomenon than the actions which stem from them.

Some twenty-five Bekker lines before the table, Aristotle has taken pains to itemize what he means by ‘πάθη’ together with, and in contrast to, the terms ‘capacities’ and ‘states’ in a passage worth citing in extenso.

There are important parallels between the way in which Aristotle characterizes πάθη in this excerpt from EE II 2 and the table of EE II 3 together with the appendix: first, both seem to be open-ended lists (compare ‘by affections I mean such things’, followed by examples, and ‘these and similar ones are the affections that happen to the soul’). Second, the list of πάθη in the excerpt from EE II 2 fixes the order of the first four items on the table in EE II 3: πραότης is the virtue concerned with θυμός;Footnote ¹⁷ ἀνδρεία with φόβος;Footnote ¹⁸ σωφροσύνη with ἐπιθυμία; αἰδώς is skipped in the review after the table, possibly having fallen out of the text.Footnote ¹⁹ After the first four items, the order becomes admittedly more haphazard. Third, which might easily go unnoticed in the appendix to the table, πάθη are there said to be psychic occurrences that happen to the soul: this is a re-statement of the idea affirmed in the text just quoted, namely that affections fail to qualify their bearers, hence a fortiori cannot amount to virtues and vices.

We should now be in a position to make sense of Aristotle’s otherwise puzzling appendix to the table of EE II 3: ‘these and the like are πάθη that happen to the soul, and all of them are said either by excess or deficiency’ (1221a13–14). These lines are one of the concerns which prompts Rowe to bracket the third column of the table, (a) of the introduction. The problem subsides by supposing that Aristotle intends to refer to the πάθη underlying the individual virtues of character.Footnote ²⁰ On this reading, each individual line of the table is meant to identify the περὶ ὅ (‘that concerning which’) of a virtue of character or other character trait, thereby providing an answer to question (ii) above, namely which means it is concerned with (περὶ ποῖα μέσα). As for the emphasis placed on the extremes both in the appendix and the review of extremes, we can explain this by bringing out one of the presuppositions of the doctrine of the mean: Aristotle is assuming that it is only in domains where an excess and a deficiency can be located which admit of a mean; by surveying the items in the first two columns he is thus showing that since in the affective domains in question there is both an excess and a deficiency, there will also have to be a mean. In view of this, it is readily explicable why Aristotle expresses himself as he does and no longer necessary to impugn the third column.

A later passage should be brought in at this stage, for the following reasons: first, it provides confirmation for the view so far advocated; second, it speaks pretty clearly against the excision of the third column of EE II 3. In EE II 5, Aristotle concludes the first half of the whole book with a summary of his argument up to the present point, the first half of which is of particular relevance to us:

Aristotle speaks of having catalogued states, both vices and virtues, ‘in accordance with each affection’, which confirms the interpretative proposal articulated so far: his primary concern in these sections has been to assign specific affections like anger, fear, appetite, etc. to the several virtues of character. Further, and equally importantly, by ‘states contrary to these’, he intends to refer to the means which are opposed to both extremes, as the aside in the text makes clear, for these are the states which are in accordance with right reason.Footnote ²² The text contains a clear indication that in the foregoing Aristotle takes himself to have catalogued not only the vicious extremes but also the virtuous means. Since the virtuous means have appeared nowhere else but in the third column of the table,Footnote ²³ this tells in favour of retaining the third column of the table.Footnote ²⁴ As we have seen, one of the motivations behind the excision of the third column of EE II 3 concerned the prematurity of listing the virtues at this stage of Aristotle’s enquiry given that virtue has not been defined yet, Rowe’s worry (b) of the introduction; nevertheless, Aristotle may well have included the means in the table of EE II 3, if only for orientation and as a preliminary sketch of what is to come in EE III (cf. also 1221b7–9). In sum, although the concern underlying Rowe’s (b) reflects a correct reading of the argument of EE II and III, it is unnecessary and misguided to implement it through editorial intervention by deleting the third column of EE II 3.

The suggestion that the table of EE II 3 is a tool to scout the πάθη underlying the individual virtues of character faces two difficulties, worth dealing with briefly here. First, there are some items in the table which, instead of being connected to πάθη, are linked up to pleasure (1221a25–28) or pain (1221a28–31; 1221a37–b3) in the sketch. This mismatch should not worry us, however, for, as we have seen, Aristotle conceives of πάθη as inherently bound up with pleasures and pain in the sense that an adequate definition of a πάθος, say θυμός, will have to mention pleasure and pain (or either of them).Footnote ²⁵ On Aristotle’s view in the EE, character virtue is concerned with pleasure and pain (1220a39–40; 1221b38–39; 1227b1–5): this tenet expresses the general truth equally valid for all virtues of character, while affections are indirect ways of relating to pleasure and pain in so far as they are psychic phenomena inherently bound up with them; in other words, character virtue is concerned with affections and, through these, with pleasure and pain. As for the second difficulty, this consists in the fact that in some cases, say, generosity (the seventh line on the table), it seems obvious that the character virtue in question concerns primarily actions related to certain external goods. Aristotle is not oblivious to this (1221a32–34; 1231b28–29), still in EE III 4 he takes pains to characterize the virtuous and the vicious persons in terms of the pleasures and pains they experience (1231b29–32). He can thus still claim that generosity is concerned with one’s affective reactions to the acquisition and expenditure of wealth. The difference would be that in this case the underlying affection is not standardly lexicalized as in the case of other character virtues.Footnote ²⁶

4. PARONOMY

I have so far said that the table features the same items subsequently surveyed in the review of extremes. We should register, however, a slight shift which occurs between the table and the review: the table, on the one hand, features abstract terms like ὀργιλότης, ἀναλγησία and πραότης; the review, on the other, cognates like ὀργίλος and ἀνάλγητος. The linguistic relation between the former and the latter should remind us of what Aristotle in the Categories calls ‘paronymy’.

As Ackrill puts it in his Categories commentary, ‘a thing is paronymous if its name is in a certain way derivative’.Footnote ²⁷ This corresponds closely to the relation between the items on the table and the subsequent review of extremes: ὀργίλος is derived from ὀργιλότης, ἀνάλγητος from ἀναλγησία; θρασύς from θρασύτης, δειλός from δειλία; and so on and so forth;Footnote ²⁸ with the caveat that the sketch skips two lines in the table, viz. the third (the αἰδώς trio) and the tenth (the σεμνότης trio). Importantly, though the test for paronymy is linguistic in nature, Aristotle’s point is not about etymology: the idea is rather that ‘courageous’ means ‘having courage’ and is thus a way to introduce a quality by predicating it of someone. Aristotle’s interest in paronymy is attested in EE III 1, where he notes that the rash person is said paronymously from boldness (1228a36–37, ὁ γὰρ θρασὺς παρὰ τὸ θράσος λέγεται παρωνύμως). It bears emphasizing that as a matter of linguistic happenstance in some cases words are missing in natural languages: the most striking example, remarked upon in the Categories, is that of ἀρετή whose paronym is lacking and for which σπουδαῖος is used in its stead (10a27–b11). Likewise, in the EE Aristotle occasionally points out that some of the character traits under examination are nameless (1221a3; 1221a28–31; 1231a38–b2; 1233b19–23).

All this may at first look like a curiosity about the EE, but in reality it plays an important role in the argumentative strategy of the treatise: first, in so far as virtue and character traits more generally are qualities, Aristotle is keen to point out that individual ascriptions of character traits introduce corresponding qualities, on the model of ὀργίλος and ὀργιλότης. This is simply a particular instance of the more general principle that whatever is φ-like is said to be so according to φ. Significantly, it is the quality signified by the abstract term which admits of definition: courage not the courageous. Second, and notwithstanding this, the relation of paronomy allows Aristotle to enquire on two levels, namely on the level of the quality as well as on that of the individual qualified accordingly.Footnote ²⁹ Perhaps Aristotle thinks one can select the level of investigation somewhat freely depending on one’s specific objective. Most importantly, while abstract terms like ὀργιλότης may seem abstruse and somehow unfathomable, by referring to an individual correspondingly qualified the paronym remains close to ordinary experience.

The shift from the abstract terms to their paronyms sheds light on a peculiar feature of the text of EE II 3. After the table and the appendix, the review of extremes begins with an explanatory particle: ‘for (γάρ) the irascible person is the one who gets angry more than one ought to […], the impassive one is the one who is deficient in whom he is angry at […]’ (1221a15–17). Aristotle thus intends the review of extremes to provide support for the items on the list in so far as pointing out that some people are irascible, others impassive, provides evidence for there being by paronomy correspondent qualities and more generally a correspondent πάθος and a correspondent affective range. Indeed, in the above text Aristotle notes that the irascible person is someone who gets angry (ὀργιζόμενος) more than one ought to, the central point being the connection between the epithet in question and the underlying πάθος. Further, by counterposing the irascible to the impassive he will show that the affective range in question admits of an excess and a deficiency and, thus, of a mean. Generalizing, in his review of extremes after the table Aristotle applies the following pattern: person X is active with regard to πάθος Φ in excess, while person Y in deficiency. In some cases, the result of applying this scheme to a particular case will be obvious, as in the case of the underlying affection of anger, where it needs little to no argument to show that some people exhibit a tendency towards excess, others towards deficiency: this is likely to be the reason why here in EE II 3 anger-related character traits have pride of place, for in this case both the underlying πάθος and the fact that there are people who exhibit it either towards excess or deficiency are readily apparent. In other cases, however, the outcome of the application of this scheme can be anything but obvious. Aristotle will apply it to cases where either the excess or the deficiency is rare and thus likely to escape notice, to the point of positing in some cases that in a given place there should be a correspondent character trait and, by paronymy, a correspondent quality.Footnote ³⁰

Finally, the issue of paronymy is relevant to a question previously left in abeyance, namely whether the back-references of EE III (1228a28; 1230b12; 1231b8; 1233a9) hark back to the table or the review of extremes. As we saw earlier, if the former were the case, it could be used in support of the excision of the third column, given that the back-references mention extremes rather than the coordinate means. Crucially, however, the back-references of EE III mention paronyms, not abstract nouns, which suggests that they are meant to refer to the sketch of extremes; thus, they cannot be used to support the excision of the third column of the table after all.

5. JUSTICE

The shift which occurs between the table and the review of extremes, from qualities to their paronyms, mitigates problems concerning one of the odd lines on the table, namely the trio κέρδος, ζημία and δίκαιον. We saw in the introduction that this was Rowe’s justification (c) for the editorial bracketing of the third column of the table. Here is what he observes in his note: ‘δίκαιον in a4 is surely a mistake by someone: δίκαιον may be by far the commonest term for justice in EE, but it does not refer to the inner state of a person, rather to the state of affairs between persons.’Footnote ³¹ Rowe is assuming, not implausibly, that the table of EE II 3 features ‘inner states of a person’ and on this assumption it is legitimate to be at a loss at the appearance of δίκαιον on the table. The problem with this is that if δίκαιον is not ‘an inner state of a person’, nor are by the same token κέρδος and ζημία, thus on that ground the whole trio would have to be bracketed. This suggests that getting rid of one of them will not do and that the items of this trio stand or fall together. Besides Rowe, scholars who have looked closely at the table of EE II 3 have taken different views of the δίκαιον trio, ranging from the suggestion that this deliberately anticipates the special status Aristotle assigns to justice vis-à-vis the doctrine of the mean in Eth. Nic. V to the dismissal of it as a mere blunder.Footnote ³² If we look at the way in which Aristotle goes about spelling out the extremes after the table, it becomes clear, however, that he is searching for a correspondent πάθος in the domain of justice.

κερδαλέος is the paronym of κέρδος, as ζημιώδης is that of ζημία, both of which are associated here to πλεονεξία. Aristotle has thus in mind πλεονεξία as the characteristic motive of injustice—which shows that ‘profit’ in the table means ‘unjust profit’ or ‘profit at someone else’s expense’ and similarly ‘loss’ ‘unjust loss’ or ‘loss to someone else’s advantage’. This corresponds quite closely to, and arguably anticipates, the way in which in Eth. Nic. V 2, after distinguishing justice as a whole from particular justice, he goes on to stake out a specific domain for the latter over and above the former.Footnote ³³ To recall, what is distinctive about particular justice is that one acts unjustly because of πλεονεξία, aiming at gain at someone else’s expenses (1130a16–32); from this, Aristotle concludes that the characteristic motive of particular justice is the pleasure derived from profit (1130b1–4).Footnote ³⁴ Admittedly, πλεονεξία does not apply to justice as a whole; since, however, the latter coincides with complete virtue (τελεία ἀρετή) in so far as it is other-directed, and the components of complete virtue (viz. courage, temperance, mildness, etc.) have already been examined and proven to conform to the doctrine of the mean in the previous sections of the ethics, it is sensible of Aristotle in Eth. Nic. V to concentrate his efforts on showing that there is an affective range correlated with particular justice, which will provide the requisite domain for the doctrine to apply.Footnote ³⁵ From this vantage point, it is natural to take the δίκαιον trio in the table of EE II 3 and the way it is articulated in the review of extremes as preparatory work for the conclusions he will arrive at in Eth. Nic. V.

To be sure, in Eth. Nic. V Aristotle himself admits that justice is not an intermediate in the same way as the other virtues of character: for one thing, just action is a mean between acting unjustly and suffering injustice (1133b30–32), which does not hold of the other virtues and, strictly speaking, is no part of the mean doctrine; second, and more to the point, in the case of justice there is only one opposite which is responsible for both excess and deficiency (1133b32–1134a1). Worst of all perhaps, there does not seem to be a disposition corresponding to what in EE II 3 Aristotle calls ζημιώδης, namely the person who is deficient in matters of gain and loss: although it does seem in principle possible to conceive of someone who is insufficiently self-assertive in matters of distribution, in the remainder of Eth. Nic. V Aristotle will take the view that no-one can suffer injustice voluntarily (especially 1138a4–35); hence the situation envisaged in the case of ζημιώδης seems to be eventually ruled out on principle. But perhaps this is a conclusion Aristotle reached at some point in his investigation of justice. Unless we have reason to think that Eth. Nic. V was written before the EE, this should not lead us to conclude that the two texts are fundamentally incompatible.

I hope to have shown that the δίκαιον trio together with the relevant section of the review of extremes represents Aristotle’s attempt in the EE to analyse the affective domain of particular justice with a view to applying the mean doctrine to it. Before concluding this section, it is worth bringing out a general conclusion of our examination: based on what we have seen, it seems sensible to treat the table of EE II 3 as a heuristic tool employed within Aristotle’s Lyceum with a view to investigating the affective domains of the several character virtues and other related character traits. The attribution of this heuristic function to the table goes some way towards explaining why it is not intended to be exhaustive (consider what is stated in the appendix ‘these and similar ones are the affections that happen to the soul’), and likewise why not all the items therein contained need be virtues. Indeed, some of them will turn out to be affective means, not virtues (the quasi-virtues of EE III 7);Footnote ³⁶ others will be demoted even further to something intermediate between virtue and vice (καρτερία, see Eth. Nic. VII 7, 1150a9–b19);Footnote ³⁷ thus, I suggest that we treat this part of the text of the EE as a cross-section of an ongoing enquiry into character, carried out by Aristotle at the time of writing this treatise. This supposition regarding the nature and function of the table of EE II 3 should reinforce our confidence in our defence of the third column against Rowe’s seclusion.

6. ΦΡΟΝΗΣΙΣ

We are left with the issues posed by the last line in the table, the φρόνησις trio, Rowe’s worry (d) of the introduction.Footnote ³⁸ There are three, increasingly trickier issues with this line and the way it is treated in the review of extremes. First, one might wonder whether the terms πανουργία, εὐήθεια and φρόνησις form an intelligible triad: a well-known passage from Eth. Nic. VI 13 (1144a23–28) links φρόνησις to πανουργία by arguing that δεινότης is common to both while serving a praiseworthy goal in the former and a shameful one in the latter, but there is no mention of εὐήθεια there; a passage from the biological works, meanwhile, contrasts φρόνησις with εὐήθεια together with other opposite pairs of character traits (Hist. an. IX 1, 608a13–17; cf. also EE VIII 2, 1247a15–20). One way to understand this is to recall that pursuit of one’s own interest is distinctive of φρόνησις (Eth. Nic. VI 5, 1140a25–28; 7, 1141b33–11) and further suppose that the πανοῦργος goes too far with it, the εὐήθης not far enough. This explanation suggests a plausible, if narrow understanding of the trio in question; perhaps it is a specific function of φρόνησις which is singled out in the dedicated line of the table.Footnote ³⁹ The second problem is that the extremes that are co-ordinate with φρόνησις in the table are articulated in terms of πλεονεξία in the corresponding section of the sketch of extremes.

Aristotle is applying here the same pattern found throughout the review of extremes. To recall: person X is active with regard to πάθος Φ in excess, while person Y is active in deficiency. Since πλεονεξία makes a second appearance here after the δίκαιον trio (see 1221a22–24), the worry is that the φρόνησις trio might reduplicate the δίκαιον one. The problem can be assuaged by finding a sense of πλεονεξία different from, and broader than, the one germane to justice. Pearson has plausibly suggested that the trio in question concerns people’s attitude towards a broader category of core goods than those relevant to justiceFootnote ⁴⁰ —a suggestion compatible with the aforementioned explication of the φρόνησις triad. However, and we come here to the third problem, if this is along the right lines, it is difficult to avoid the conclusion that in this way φρόνησις is being presented as an intermediate between the two extremes that are co-ordinate with it, which conflicts with Aristotle’s standard view whereby the mean doctrine is distinctive of the character virtues as opposed to those of the intellect (EE II 1, 1220a14–16 and 3, 1220b21–35; 5, 1222a6–12; cf. also Eth. Nic. II 6, 1106b16–28). The reply attempted here is that φρόνησις appears in the table not qua intellectual virtue but in so far as there is a non-rational basis underlying it together with the terms co-ordinate with it.Footnote ⁴¹ The resulting thought, then, would be that the relevant continuum picked out by the trio in question is some form of non-rational attachment to a broad category of core goods liable to serve one’s own interest.

Having said that, let me conclude on a cautious note. In his accompanying note Rowe states that, despite any reasonable misgivings one might have about the φρόνησις trio, it would be ‘unsafe to start ridding ourselves of particular items or groups of items’.Footnote ⁴² I concur with his judgement here, and I hope to have shown in the foregoing that careful consideration of Aristotle’s argument in EE II 3 makes it unsafe to rid ourselves of any item whatsoever in the text. The most one can say about the φρόνησις trio, charitably and in line with the conclusion drawn in the last section concerning the heuristic function of the table of EE II 3, is that at the time of writing the EE Aristotle at least attempted to apply to it the apparatus of the doctrine of the mean; the fact that he does not seem to have developed this idea anywhere else in the corpus suggests that the attempt remained a one-off.