The notion of in the early Stoa and “moral duty.” reflections on Jack Visnjic’s book “The invention of duty. Stoicism as deontology.”

term of undesirable moral connotations suggested by more straightforward translations – such as “duty” or “Pflicht.”4 However, the recent book “The Invention of Duty” by Jack Visnjic called into question this scholarly consensus. Not only did Visnjic propose to get back to translating καθῆκον as “duty” (or at least as “prescribed” or “required action”),5 he also believes that the “duty” in question may and should be described as explicitly “moral.”6 To my mind, this last thesis is erroneous and here I would like to explain why this is the case. Accordingly, in the first part of this paper I examine Visnjic’s argumentation in favour of his revisionist conclusions while in the following two parts I offer some additional conceptual considerations against them which, in my view, were underestimated or completely overlooked by Visnjic.

In his book, Visnjic identifies four main reasons for general reluctance to translate the term καθῆκον as “duty.”7 The first of them “is the common belief that the ancient Greeks and Romans lacked a concept of moral duty as well as any deontological system of ethics.”8 Visnjic instead argues that the ancients did have some concept of moral duty, even if substantively it was not tantamount to some of its modern counterparts (e.g. the Kantian one).9 Although on the whole I do agree with this, I would like to notice that the mere presence of such a concept does not suffice to make a specific ethical theory “deontological.” Visnjic defines “a deontological (duty-based) ethics” simply as “a framework for deciding which choices are morally required, forbidden, or permitted.”10 However, it is not obvious that those who deny the “deontological” character of ancient ethics necessarily use this term in the same sense. The usual point they make is rather that ancient ethics was predominantly teleological and hence did not arrive at the concept of unconditional duty which may conflict with the agent’s own good or happiness.11 Accordingly, the decisive question one should ask to clarify this issue is rather as follows: did the ancient philosophers believe that sometimes it might be necessary to sacrifice one’s own happiness in order to fulfill one’s moral duty? It is only if this is the case that one could ascribe to them some kind of consistently “deontological” attitude. But I would side with those scholars who answer this question in the negative.12 In particular, Stoicism by insisting that moral goodness is the only factor which is of any importance for happiness13 makes such a conflict between fulfilling moral duty and being happy especially hardly conceivable.14 Still, by itself this is not a good reason to deny both that the Stoics could have some concept of moral duty and that they could use the term καθῆκον to express it.

The second reason mentioned by Visnjic is that “the literal meaning of the word kathêkon is thought to be something like ʻappropriate,’ without a sense of obliga-tion.”¹⁵ Contrary to this, Visnjic argues that in the pre-Stoic period both this word itself and the verb καθήκειν related to it “seem to express something that is obligatory , prescribed , or established by some authority.”¹⁶ Moreover, “even the examples that LSJ provides of the verb supposedly indicating what is fitting or proper… upon close inspection… seem to exhibit a sense of prescription.”¹⁷ These conclusions are corroborated by the review of relevant passages from various ancient authors¹⁸ and generally speaking look quite convincing. Does this amount to the decisive proof that within early Stoic ethics the term καθῆκον conveys the notion of specifically moral duty? First of all, when talking about philosophical terms, it seems natural to admit that the transition from the lexical sense of a word to its conceptual meaning is at least not necessarily straightforward. To establish the latter precisely one always needs to take into account the theoretical framework of the doctrines at issue. Thus, if one can point out serious conceptual reasons rooted in the very essence of Stoic ethical theory that run counter to a specifically moral understanding of the term καθῆκον, then, of course, they should be given more weight. Secondly, it is essential to specify at least tentatively, i.e. without going into complex discussions about the essence of “morality” in general, what qualifies as genuinely

“moral” duty at all. In this respect, I believe it would be fairly uncontroversial to say that at least two following conditions should be met: 1) an act may count as the fulfilment of “moral” duty if and only if it is performed by a rational agent; 2) an act that counts as the fulfilment of “moral” duty could also be legitimately described as “morally right” or “virtuous” (at least to some extent), whereas the breach of “moral” duty would constitute a “morally wrong” or “vicious” act. When it comes to Stoic ethics, this would imply that the terms καθῆκον and παρὰ τὸ καθῆκον express more or less the same kind of genuinely moral evaluation that is usually associated with such notions as καλόν/ honestum and ἀρετή/ virtus or, on the other hand, αἰσχρόν/ turpe and κακία/ vitium.

The first of these conditions has a direct bearing on the next reason in favour of the traditional translation of καθῆκον discussed by Visnjic: “...the Stoics often speak of kathêkonta as arising from nature, so that even plants and animals exhibit them.”¹⁹ Since plants and animals are not rational, they certainly cannot fulfill any genuinely moral duties. According to the standard interpretation, by using the term καθῆκον in this way the Stoics want to say that the activities in question are “appropriate” or “befitting” for these living beings in terms of their natural constitution and therefore are also in general agreement with the universal nature.²⁰ Visnjic’s reaction to this problem may be adequately summed up by the following quotation from his book: “I think it would be more accurate to say that plant and animal kathêkonta are activities prescribed to each living thing by their own nature and the providential nature of the universe. Likewise, human kathêkonta are also prescribed by nature. But in the case of rational beings, these prescriptions have an added moral dimension.”²¹ Now, I have no principal objections against the proposal to construe καθήκοντα as the activities “prescribed” by nature rather than “appropriate” to it. However, it does not obviously follow neither from this proposal itself nor from the fact that “in the case of humans, the kinds of kathêkonta that actually matter to the Stoics are rational actions, not biological functions”²² that human καθήκοντα necessarily have “an added moral dimension.” It may still be the case that they are prescribed in the same non-moral naturalistic sense as the καθήκοντα of plants and animals. Whether this is so or not, depends on the theoretical framework of Stoic ethics, not on the lexical sense of the word καθῆκον. For now, let us admit for the sake of the argument that Visnjic is right. First of all, if

καθῆκον best translates as “duty”, but the Stoics still use this term to describe purely biological functions of plants and animals, it is clear that in principle they can speak about “duties” in a figurative sense, which has nothing to do with morality. Secondly, Visnjic’s position implies that the term καθῆκον is applied by the Stoics both to the moral “duties” of human beings and to the morally neutral “duties” of plants and animals. Consequently, if this term covers both moral and non-moral range of activities, its general meaning cannot be limited to some exclusively moral concept.

I will skip any substantive discussion of “the fourth main reason against taking the Stoic concept to mean ʻduty’”, which, according to Visnjic, “is based on the etymological explanation for the word kathêkon offered by Zeno himself, the founder of Stoicism who coined the term.”²³ Zeno’s explanation comes down to the claim that this word is derived from the expression κατά τινας ἥκειν,²⁴ which is very rarely, if at all, used in other sources and is hard to interpret out of context.²⁵ Although historically some interpretations of this statement by Zeno may have contributed to the traditional understanding of καθῆκον as “appropriate action,”²⁶ they by no means constitute an important argument against the possibility of translating it as “duty.” On the other hand, I do not see how Visnjic’s own preferred hypothesis that Zeno’s etymology conveys the idea of something “having come right in front of you”²⁷ necessarily implies the notion of specifically moral duty.

The foregoing considerations, in my opinion, show that Visnjic’s argument in favour of construing the Stoic notion of καθῆκον as “moral duty” is far from being conclusive. However, the main drawback of Visnjic’s position is that he offers no serious discussion of numerous Stoic fragments that at least prima facie contradict his conclusions. These texts either explicitly declare or implicitly suggest that by applying to any activity such terms as καθῆκον and παρὰ τὸ καθῆκον one does not yet qualify it as “morally right”/ “virtuous” or as “morally wrong”/ “vicious”, respectively. Rather, these terms convey some peculiar distinction which, though undoubtedly normative, still remains essentially neutral with respect to genuine morality as conceived of within the early Stoic ethical discourse.

The first important piece of evidence to be mentioned here concerns the well-known fact that the Stoics distinguish between καθῆκον as such, which in this case may also be characterised as μέσον (i.e. “intermediate”, “middle” or “indifferent”), and τέλειον καθῆκον (“perfect duty”),28 which in its turn is identical with κατόρθωμα or “right action.”29 There is no doubt that this last term designates an activity which is truly virtuous or “right” in a specifically moral sense since κατόρθωμα is explicitly defined as an “activity in accordance with virtue.”30 Similarly, the opposite term ἁμάρτημα applies to an “activity in accordance with vice.”31 These definitions of κατόρθωμα and ἁμάρτημα reflect the general Stoic idea that the moral status of any activity depends on its being produced by either virtuous or vicious “disposition” (διάθεσις) of the agent’s soul.32 This has a further important implication for understanding the precise relationship between καθῆκον and κατόρθωμα. As several sources show, materially, i.e. in terms of their actual content, καθήκοντα and κατορθώματα can be identical. The only difference between them is the presence of a truly virtuous motivation proceeding from the corresponding state of the soul. For instance, according to Cicero’s testimony, returning a deposit is a καθῆκον (of-ficium), but it is only when one performs this action “justly” (iuste), i.e. exercising such a virtue as justice, that it becomes a κατόρθωμα (recte factum).33 Consequently, since μέσα καθήκοντα do not stem from the virtuous disposition, they cannot be considered morally right activities but rather, taken by themselves, remain morally neutral. Therefore, even if in some sense they may be called “duties” (as much as biological functions of plants and animals), they cannot have anything to do with “moral duties” sensu stricto. That καθήκοντα in this sense lie beyond the realm of genuine morality, which deals with the only real goods and evils, i.e. the moral ones, is expressly stated by Cicero, e.g. in SVF III, 49834 (= Cic. Fin. III, 58):

T1 “Now although we say that what is moral (honestum) is the only good (solum bonum), it is still consistent to perform appropriate actions (officio) despite the fact that we regard them as neither good nor evil (id officium nec in bonis ponamus, nec in malis)... Hence one can see that appropriate action is something intermediate, falling into the category neither of goods nor their opposite (officium medium quiddam esse, quod neque in bonis ponatur, neque in contrariis). Since there may yet be something useful about what is neither a virtue nor a vice (in iis rebus, quae neque in virtutibus sunt neque in vitiis), it should not be rejected… Hence appropriate action falls under the category of what is neither good nor the opposite (est igitur officium eius generis, quod nec in bonis putatur nec in contrariis)” (Transl. Woolf in Annas, Woolf 2001, 83).

There are two basic ways to interpret such statements. Some scholars believe that middle duties constitute a separate class of actions which are materially different from both κατορθώματα and ἁμαρτήματα.35 One immediately obvious problem with this interpretation is that it agrees rather badly with the idea that a middle duty can have the same material content as a perfect one. Both for this reason and because of other conceptual considerations expounded below, I prefer an alternative hypothesis which basically runs as follows: particular types of καθήκοντα (e.g. returning a deposit, taking care of one’s parents, etc.) can only be treated as morally neutral if considered without reference to the virtuous or vicious disposition of the person who performs them, but in practice such actions always become either morally right κατορθώματα or morally wrong ἁμαρτήματα.36 Still, on both of these interpretations, the notion of μέσον καθῆκον, taken by itself, has no genuinely moral dimension to it.

Surprisingly, Visnjic does not pay much attention to this evidence. Sporadically, he mentions that middle duties “occupy an intermediate position on the moral spectrum between perfect duties (sc. virtuous actions) and moral errors (hamartêmata)”37 or that “only perfect duties are acts of virtue for the Stoics,”38 but it remains unclear how such statements relate to his central claim that καθῆκον as such already amounts to “moral duty.” When discussing the difference between perfect and middle duties, Visnjic mostly focuses on another distinction found in the Stoic fragments – that between “the duties that always hold” and “the duties that do not always hold.”39 In his opinion, perfect duties coincide with the former of these categories, which comprises what he calls “vague and virtue-referencing injunctions”40 (e.g. “to be wise”, “to act justly”, etc.), whereas middle duties should be equated with the latter, which contains “concrete prescriptions that make no intrinsic reference to virtue”41 (e.g. “to marry”, “to obey one’s teacher”, etc.). Now, it is true that sometimes Stoic sources cite as examples both of κατορθώματα, i.e. perfect duties, and ἁμαρτήματα various abstract types of activity which generally presuppose that the agent acts in accordance with some virtue or vice.42 However, it would be wrong to infer from this, as Visnjic seems to imply, that κατορθώματα or perfect duties are limited to such “vague and virtue-referencing injunctions.” For one thing, κατορθώματα and ἁμαρτήματα can also include the so-called “good feelings” (εὐπάθειαι) and “passions” (πάθη), respectively, i.e. specific types of internal activity.43 For another, as we saw above, the Stoics regard as κατορθώματα even concrete and materially defined types of external actions, provided they are brought about by virtue (e.g. “to return a deposit justly”, “to walk around prudently”, etc.).44 Since this type of perfect duties coincides with middle duties in terms of their material content, the problem with Visnjic’s interpretation becomes evident: what is the point of distinguishing perfect duties as the only genuine form of morally right actions from materially identical middle duties if the latter should also be construed as duties in a specifically moral sense? Do we have to suppose that in Stoic view by performing such a moral duty one still does not commit an action which is morally right?

It may be, of course, that, in Visnjic’s view, perfect and middle duties, though both morally positive, still differ in the degree of their moral goodness. Unfortunately, he does not argue for such a view systematically, but, when comparing Stoic ethics with that of Kant, he actually says that “middle kathêkonta resemble actions in conformity with duty [i.e. in Kant’s theory – A.S.] in that an agent does something which duty prescribes, but not with the right mental state that would make that action have a special degree of moral worth.”45 This seems to imply that despite the lack of “the right mental state” they still have some degree of moral worth, which is presumably lower than that of perfect duties. To my mind, such an assumption has at least two essential flaws. First of all, it is obviously at variance with all the texts where middle duties are explicitly described as morally neutral.46 Sec- ondly, this supposition is simply too vague and, therefore, conceptually unsatisfactory. What is the exact degree of the alleged moral difference between middle and perfect duties? What theoretical reasons stand behind it? And, finally, what is the textual evidence for all this?

The second argument, which states that middle duties “lead the way towards happiness”, appeals to the following fragment taken from Stobaeus:

T2 “...Now, middle duties (τὸ μέσον καθῆκον) are determined by the comparative measurement of indifferent things (ἀδιαφόροις), which are classified as [either] “contrary to nature” [or] “according to nature” (παρὰ φύσιν καὶ κατὰ φύσιν). These indifferents offer us such a wellflowing life that if we do not persevere and without distraction keep selecting and rejecting them [according to nature], we shall not attain true happiness (ὥστ' εἰ μὴ λαμβάνοιμεν αὐτὰ ἢ διωθοίμεθα ἀπερισπάστως, μὴ ἂν εὐδαιμονεῖν)” (SVF III, 499; Stob. Ecl. 2, 7, 8a, 8–12 Wachsmuth-Hense).50

How does this prove that middle duties are “morally important”? Visnjic does not tell it clearly, but the implication must be somewhat as follows: since in Stoic view happiness is essentially identical with a morally good life,⁵¹ whatever is necessary for the former is eo ipso necessary for the latter and thus, needless to say, is “important” for it; according to T2 , the undistracted performance of middle duties is necessary for happiness and so, by implication, for a morally good life; hence, middle duties are “important” for a morally good life, and in this sense one is perhaps entitled to call them “morally important”. This argument may be valid, but obviously it hinges on the ambiguity of the expression “morally important” and by itself does not prove that middle duties are essentially moral, i.e. that they constitute genuinely moral activities. It is quite conceivable that something which is necessary for a morally good life does not have an intrinsically moral nature. That this is the case with middle duties seems to be clear from the textual evidence. First of all, although, according to T2 , if one does not fulfill middle duties without omissions, one cannot be happy, it does not follow from this statement that, if one does fulfill them all, one will become happy of necessity. Hence, the possibility that someone who unerringly performs all middle duties still does not achieve happiness is perfectly compatible with this passage. It is precisely this possibility that is envisaged in another fragment from Stobaeus:

T3 “Chrysippus says: The man who progresses to the furthest point performs all proper functions (ἅπαντα... τὰ καθήκοντα) without exception (πάντως) and omits none (οὐδὲν παραλείπει). Yet his life, he says, is not yet happy (οὐκ εἶναί πω... εὐδαίμονα), but happiness (τὴν εὐδαιμονίαν) supervenes on it when these intermediate actions (αἱ μέσαι πράξεις) acquire the additional properties of firmness and tenor and their own particular fixity (τὸ βέβαιον καὶ ἑκτικὸν καὶ ἰδίαν πῆξιν τινὰ)” (SVF III, 510; Stob. Ecl. 4, 39, 22, 2–7 Wachsmuth-Hense; transl. Long, Sedley 1987, 363).

The properties mentioned at the very end of this passage seem to represent what constitutes another indispensable characteristic of the agent’s virtuous motiva- tion, i.e. its absolute stability and consistency guaranteed by the infallible rationality of the sage.52 Accordingly, the point this text makes is that even by performing all middle duties with no omissions one does not achieve happiness if one still lacks perfect virtue. In other words, middle duties can actually contribute to a happy and morally good life if and only if they are performed by a perfectly virtuous sage. But in this case, as we saw, they would be in fact transformed into perfect duties or κατορθώματα. Nevertheless, in a sense middle duties are necessary for happiness because they provide material content for virtuous activity. As other sources show, in Stoic view virtue itself necessarily involves the practice of selection among preferred and dispreferred indifferents,53 which is basically tantamount to performing middle duties.54 Still, the Stoics make a point of stressing that what constitutes moral good and, consequently, happiness is not this practice itself, however successful it may be, but only executing it in a perfectly rational way, which is the essential property of the sage’s wisdom.55 To conclude, even if μέσα καθήκοντα “are prescribed by nature” and in some sense “lead the way towards happiness”, this does not prove that they are intrinsically moral themselves and therefore can be properly described as genuinely moral duties. On the whole, Visnjic’s interpretation of this notion seems to ignore that middle and perfect duties can be materially identical and makes the distinction between them conceptually vague and hard to grasp.

There is yet another important reason against Visnjic’s attempt to understand καθῆκον as moral duty. As was mentioned above, one possible interpretation of middle duties presupposes that in fact they can turn both into κατορθώματα and ἁμαρτήματα depending on either virtuous or vicious disposition of the agent’s soul. Now, if this is correct, then Visnjic’s claim about the moral status of middle duties becomes especially problematic since to accept it one would have to agree that in Stoic view one could adequately fulfill one’s moral duty by committing a morally wrong action. That would be a bizarre view because performing a moral duty is normally regarded as a morally positive activity, at least to some extent. By contrast, there is nothing particularly strange in a moral theory which admits that some actions, while being morally neutral by themselves, nevertheless acquire negative moral value whenever the agent’s motives are vicious.

The evidence in favour of the idea that some καθήκοντα should be included among ἁμαρτήματα appears to be somewhat circumstantial since, to my knowledge, there is no single Stoic source where this is stated with absolute clarity. Logically, however, such a conclusion is quite compelling because it follows from several important and abundantly attested Stoic doctrines. First of all, the Stoics believe that there can be no intermediate or neutral state of the soul between virtue and vice.56 They also hold the view that all the virtues are interconnected in such a way that anyone who has one of them necessarily has all the others,57 and the same is true about vices.58 Hence, no person can actually be partly virtuous and partly vicious, and the whole of humankind proves to be divided into virtuous sages and vicious non-sages.59 Given that the moral value of any activity depends on the agent’s disposition,60 it is clear that all the actions performed by the sages must be morally right (κατορθώματα),61 whereas all the actions by the non-sages are predetermined to be morally wrong (ἁμαρτήματα).62 This radically dualistic picture of the human condition is perfectly summarized in the following fragment taken from Stobaeus:

Т4 “ [a] It is the view of Zeno and his Stoic followers that there are two races of men, that of the worthwhile, and that of the worthless (δύο γένη τῶν ἀνθρώπων εἶναι, τὸ μὲν τῶν σπουδαίων, τὸ δὲ τῶν φαύλων). The race of the worthwhile employ the virtues through all their lives, while the race of the worthless employ the vices (καὶ τὸ μὲν τῶν σπουδαίων διὰ παντὸς τοῦ βίου χρῆσθαι ταῖς ἀρεταῖς, τὸ δὲ τῶν φαύλων ταῖς κακίαις). [b] Hence the worthwhile always do right in everything on which they embark (asi KaTopQouv ev anaaiv ofg npooriQsTai), and the worthless do wrong (a^apTavsiv). [c] The worthwhile man, using his practical experiences with regard to life in things done by him, does all things well (πάντ' εὖ ποιεῖν), just as he does them sensibly, with self-restraint, and in accord with the other virtues (καθάπερ φρονίμως καὶ σωφρόνως καὶ κατὰ τὰς ἄλλας ἀρετάς). The worthless man, conversely, does badly (τὸν δὲ φαῦλον κατὰ τοὐναντίον κακῶς)...” (SVF I, 216 = Stob. Ecl . 2. 7. 11g. 11–20 Wachsmuth-Hense; transl. Pomeroy 1999, 73).

This becomes especially evident if one takes into account the Stoic thesis that all ἁμαρτήματα or peccata are equal (ἴσα; paria ) to each other.⁶⁶ In some fragments, it is illustrated by the example of an agent who is already very close to attaining virtue but still remains no less vicious and unhappy than the worst criminals and evildo-ers.⁶⁷ As we saw in T3 , it is precisely a person of this kind who is able to perform all middle duties without any omissions but nevertheless does not achieve happiness. It is hardly conceivable that ἁμαρτήματα explicitly ascribed to such agents in other sources for some reason do not include all the καθήκοντα they perform. For one thing, such an assumption runs counter to the basic Stoic belief that it is virtue or vice that determines the moral status of all particular activities. But the very idea that all ἁμαρτήματα are equal is based precisely on this belief.⁶⁸ For another, T3 seems to imply that the agent in question simply does nothing else except fulfill all middle duties without omissions. For example, one cannot suppose that alongside with this activity, which presumably may be thought of as morally neutral or even positive, such a person sometimes happens to commit a παρὰ τὸ καθῆκον and, consequently, a ἁμάρτημα for committing a παρὰ τὸ καθῆκον is tantamount to failing to do a καθῆκον, but then one cannot say that this agent fulfills all καθήκοντα “without exception and omits none.” Thus, the conclusion that all these καθήκοντα are morally wrong actions appears to be the only logically possible.

One can judge about how Visnjic handles this problem from the following passage in his book, the only one where he actually addresses it: “...fulfilling one’s everyday duties is something that even a normal person can aspire to. This interpretation goes against the view of some scholars that the Stoics believed that ‘all the actions of the non-sage are vicious actions and errors’ (hamartêmata).69 Under such a view, a normal person can never hope to perform a kathêkon. But such a view cannot be right. The main passage that is cited in support of this claim does not actually say that all the actions of the nonsage are errors. Rather, it says that every action of the vicious person (φαῦλος) is an error. The figure of the vicious person functions conceptually like a mirror image of the sage. Between these extremes of perfect virtue and perennial viciousness lies a space inhabited by the vast majority of us, who undoubtedly make plenty of mistakes but may also aspire to perform our (middle) duties as much as possible.”70 I find this reaction puzzling on many levels. To start with a relatively minor point, there is no sense in stressing that “the main passage” mentioned here by Visnjic, i.e. SVF III, 560 referred to by Brennan, only speaks of “the vicious person” but not of “the non-sage” for it is perfectly clear from many sources mentioned above that in Stoic view all non-sages are vicious persons. In particular, T4 states that “there are two races of men, that of the worthwhile (τῶν σπουδαίων), and that of the worthless (τῶν φαύλων),” and it would be strange to believe that this statement somehow admits of the existence of a third “race,” which would consist of “the vast majority of us.” Importantly, both T4 and SVF III, 560 are taken from the same source, i.e. Arius Didymus’ “Epitome of Stoic Ethics” preserved by Stobaeus, so that one can legitimately infer that the exclusive disjunction between the virtuous sages and the vicious non-sages explicitly stated in T4 also holds for SVF III, 560 by implication. Generally speaking, both direct and indirect evidence in favour of this radically dualistic distinction in the Early Stoa is so ample and coherent that there is no point in trying to dismiss this dualism by cursorily discussing just one single fragment.

Conceptually, even more important is that Visnjic seems to think that this dualistic theory is logically incompatible with the idea that the non-sages are also able to fulfill middle duties. But we have seen that this is not so: one just has to admit that in this case middle duties turn into ἁμαρτήματα. Of course, this last suggestion is an hypothesis and, though, as I believe to have shown, it is very convincing and well-founded, it is still open to critical discussion.71 What is odd about Visnjic’s approach is that he deals with this topic, as if he is unaware that such an hypothesis is an option at all, and that despite its being espoused by many schol-ars72 including those he has undoubtedly read since he cites them in his book.73

To sum up, in my view, Visnjic’s claim that the Stoic term καθῆκον stands for moral duty is untenable because besides the four reasons against it discussed by him there are two important and rather well-known additional objections to it: on the one hand, in many texts middle duties are explicitly presented as morally neutral in contrast to materially identical perfect duties, the only ones which are truly morally good; on the other, it is highly plausible that middle duties can acquire even negative moral worth when performed out of vicious disposition. The evidence in favour of these conclusions is very ample, and anyone who wishes to call them into question should at the very least give this evidence a fair and detailed consideration, which Visnjic unfortunately fails to do. That said, I would like to emphasize that his book has many valuable insights in other respects, and his contention that καθῆκον is best translated as “duty” also seems quite reasonable from the philological point of view.