Bryn Mawr Classical Review

Bryn Mawr Classical Review 2010.04.05

Christopher Bobonich, Pierre Destrée (ed.), Akrasia in Greek Philosophy. From Socrates to Plotinus. Philosophia Antiqua, 106.   Leiden/Boston:  Brill, 2007.  Pp. xxvii, 307.  ISBN 9789004156708.  $167.00.  

Reviewed by Peter Lautner, Pázmány Péter Catholic University (

The book is a collection of papers, most of which were originally delivered at a conference in Louvain-la-Neuve. It also contains an extensive bibliography, and an index of ancient sources and modern authors.

The problem of akrasia touches upon both psychology and ethics since the term can refer both to a certain behavior and to a feature of character. Furthermore, even when discussing the states of soul, it can mean both weakness of knowledge and what later was to be called "weakness of will". The authors of the collection focus on these issues to different degrees, some concentrating on the theory of action, some concentrating on the nature of the soul in the akratic person.

Thomas Brickhouse and Nicholas Smith critique the view that the Protagoras does not allow for non-rational desires. They argue that the expression δύναμις τοῦ φαινομένου refers to the non-rational appetites and passions. Their case is supported by Laches 191C8-E2, where Socrates claims that the desires to have pleasure and to avoid pain are motivational forces that are to be resisted by the courageous person. Moreover, the discussion of akrasia in the Protagoras does not suggest that it is simply a result of new information that one acts contrary to what he previously believed was the right thing to do. Instead, what most people would say is that the person who was convinced previously that a certain action was not good for him but acts contrary to that belief is akratic only if he does so in the absence of new information. This can best be explained with reference to motivational forces concerning pleasure and pain, which are not rational. This is possible even within the context of the Protagoras itself. The authors criticize the view held by D. Devereux that Socratic intellectualism means that moral knowledge is necessary and sufficient for virtue, and the occasional presence of unruly non-rational desires is irrelevant.1 They reject it on the basis of the Laches and Gorgias 507C8-E3. The difference between Socrates on the one hand and Plato and Aristotle on the other, according to Brickhouse and Smith, is that for the latter passion and appetite can pull the person to pursue the object of passion, whereas for Socrates they cause him to believe that the object of passion is a good. Socrates in the Laches and Gorgias believed that we always act for the sake of what we consider good. Yet one might wonder whether the assumption of a non-rational motive force in humans in the Socratic theory brings these texts quite close to Plato's theory in the Republic, insofar as it gives room for the independent force of appetite.

Christopher Rowe discusses the notion of punishment in the Gorgias, which helps him corroborate his general thesis that the main gap between the so-called earlier and middle dialogues is not the theory of ideas, whatever form it may take, but the different attitudes towards moral psychology. The core of the Socratic view on human motivation can be summarized in three points: /1/ all human agents always and only desire the good, /2/ they desire the real good, not the apparent one, and /3/ our deeds on any occasion are determined by this desire along with our belief concerning what will in fact contribute to our real good. This has the consequence that we only do what we think will be good for us. The Socratic view was abandoned in the Republic. The Gorgias presents a special problem, for it contains the full Socratic position, but it also seems to have features that are characteristic of the Republic. For, in Rowe's view, primarily it is a dialogue on political theory. Punishment is the paradigmatic case, for in this dialogue it amounts to philosophical discourse and education. As Rowe does intend to distinguish phases in Plato's thought with reference to moral issues (20-21), one might say that division in such terms might not have much to say about dialogues such as the Parmenides, where we cannot find arguments pertaining to moral issues.

Chris Bobonich examines the Protagoras and claims that in this dialogue the soul is portrayed as opaque to the person himself. In an alleged case of akrasia, although the person thinks that he is overcome by pleasure, as a matter of fact he is pursuing his overall good. His judgment is simply mistaken because he has overvalued short-term goods. Thus the agent is mistaken both in that he makes a wrong choice and in that he thinks himself to be subject to akrasia, though he is not. This phenomenon relates to a lack of self-knowledge, which in turn threatens to undermine the idea that the akratic person is an agent in the strict sense. One of the crucial texts to support the interpretation is 352D6-E2, a passage taken to imply that people are giving a description from the inside of the alleged case of akrasia. Of course, the crucial question is how to distinguish between real and apparent beliefs in cases where individuals hold a belief but are unaware that they hold it.

As a preliminary to the discussion of the notion of akrasia in the Republic, Christopher Shields stresses that tripartite division does not imply that the soul consists of three different agents, independent from one another. In Book IV, just as in Book X, we are compelled to regard the soul as a metaphysically simple entity.2 Furthermore, Socrates' purpose was not to deny a manifest experience, but rather to reject a positive account of how this experience is to be conceptualized and explained (64-65). Having this in mind, he makes a distinction between two concepts of akrasia. The first refers to an implementation failure. A subject (S) (i) "avowedly and sincerely prefers in an all-things-considered way some action b to a; (ii) S supposes that both a and b are equally available alternative actions; and (iii) S intentionally performs a rather than b" (65). This is perfectly possible. The second, narrow concept relates to an agent who is ruled by a single calculative faculty (think of the Protagoras) and has a single highest goal. We can consider him in full possession of his wits and always trying to maximize pleasure. Faced with a choice between a and b and knowing that b is the pleasure-maximizing option, he may be imagined as opting for a to the exclusion of b. On Shields' view, although it would not be logically impossible, it is impossible on the hypotheses concerning his agency, epistemic status, and governing preference structures. This has nothing to do with the unity of the soul. Thus the division of the soul in the Republic may not be viewed as an attempt to accommodate the theory of the soul to the phenomenon of akrasia. This leads to the suggestion that the three parts are autonomous agents with no access relations to one another's states (85). Here Plato also denies the possibility of akrasia in the face of highly unified intentional agency. Still, one might insist on the divergence of the two dialogues in this matter by asking if the conflict between appetite and spirit can be made fit in with the scheme of the Protagoras.

Roslyn Weiss's paper is a modified version of Chapter 6 of her book.3 She interprets the crucial passage in Republic 438A as referring not to positions Socrates occupied in earlier dialogues and not discrediting the Socratic thesis that "everyone desires good things". It is not the thesis "everyone desires good things" to which Socrates objects in the Republic, but the inference from this thesis to the conclusion that thirst or hunger is for good drink or good food (90). It seems that it is not Republic IV but the Protagoras that represents a reversal of Socrates' views. But her distinction between ἀγαθός and χρηστός might be difficult to trace back in the text (cf. 438A4, 439A5-6).

Starting with an analysis of Timaeus 86D5-E3, Gabriela Roxana Carone's aim is to show that ἀκράτεια, which may be related to akrasia, is an intellectual failure insofar as it happens when reason fails to control desires. On the other hand, the Timaeus does not advocate suppression of the lower parts of the soul, but rather the proportionate interaction between all parts. Appetite is capable of pictorial representation and of phantasia. She also stresses that the existence of the three parts of the soul is not to be taken as aggregational since the whole resulting from this composition is more than the mere sum of the parts. Each part is redefined according to its function. One might raise the doubt whether the World Soul is qualitatively exactly the same as the human reason (it is definitely not the same as the whole human soul) since the passage on the mixing bowl says that human reason is not so pure as the World Soul (41D6-7). One might also ask in what sense appetite is capable of phantasia if it is defined, in the Sophist, in terms of belief and perception.

Louis-André Dorion approaches Plato's notion of self-control through a comparison with Xenophon's Memorabilia. As a result, it seems that Plato, though apparently borrowing the term from Socrates, never uses it in the 'Socratic' dialogues, which can be explained in part by the denial of akrasia in these works. If we insist on the cognitive nature of virtue, we reject not only akrasia, but also self-control, insofar as it is a disposition different from virtue and knowledge, a mastery of the passions which allows the agent to be faithful to his intention to act in accordance with his knowledge of the good. The question is why Plato uses the term in later dialogues. It can be linked to σωφροσύνη, which is characterized in the Phaedo as devoid of thought. Furthermore, the rehabilitation of 'self-control' goes hand in hand with the new account of akrasia, depicted not only in the Republic but in the Gorgias as well. One might remark that the reason that Plato did not discuss self-control as one possibile definition of σωφροσύνη in the Charmides might not be the paradoxical nature of the concept of self-control, as Dorion claims (121-122), since knowledge of itself, another candidate for σωφροσύνη, is shown to be just as paradoxical, but is discussed in that dialogue.

Pierre Destrée's discussion of the causes of akrasia in Aristotle concentrates on Nicomachean Ethics 7.3, and defends the view that due to his appetite the agent subject to akrasia fails to recognize a consideration blocking the deed in question. Thus it is an intellectual failure. Destrée tries to reconcile this interpretation with the more common sensical claim, found in other texts (e.g., NE 7.2), that the akratic person acts against his own decision because his desire overcomes his intellect. Destrée shows that the view endorsed in those other texts can be accommodated to an intellectualist interpretation. Aristotle rejects only Socrates' conclusion that there is no such thing as akrasia; he accepts Socrates' explanation of it. The Aristotelian view is best analyzed with reference to De motu animalium. As a result, we can be clear that the final cause of the action is a good which is desirable, and therefore we somehow must know this good. This happens at the level of phantasia: the akratic person loses the ability to use phantasia logistikê and can only use phantasia aisthêtikê, with the latter capacity overwhelming the former. Aristotle's account differs from the Socratic position on crucial points: for Aristotle does not consider the akratic's ignorance as the final cause of his bad choice, holding that the cause of ignorance is the appetite, of which the cause is the person's bad habit (164). One might ask for a clarification of the relation between phantasia logistikê and deliberation since the latter plays a considerable role in the decision-making procedure but might not be made subservient to phantasia logistikê.

Marco Zingano argues that Aristotle adopts a Socratic perspective even in the formulation of the issue to be discussed. Action takes its origin at the intersection of two distinct capacities, practical reason and emotion. Consquently, akrasia is a result of a conflict between what practical reason takes to be the object of pursuit or avoidance, and what desire strives for as pleasurable. The akratic state can also be described as based on false opinion (this is with reference to 1147a25, although the text does not necessitate such an explanation). Attention has been paid to the different descriptions of the nature of practical syllogism. While in De motu animalium and De anima the conclusion is an action, in NE it is a proposition which may not necessarily be followed by action. The difference in the unity of practical syllogism reveals a difference in doctrine. One might point to a problem in this account concerning the status of this reason and opinion in the practical syllogism. If it is a false opinion, is it distorted by desire? If it is somehow in the conclusion, then what is the precise explanation of the mental conflict?

David Charles' analysis of the weak akrates (NE 7.3) focuses on the failure the akratic person commits in using knowledge. He labels as weak akrates the one who knows the relevant major and minor premises of the practical syllogism but fails to act accordingly, and says his state is like that of a person "who does not know but merely says but does not know it". After criticizing two approaches, the cognitive and the Humean (which accepts the existence of two independent sources of action, cognition and desire) he relies on De anima 3.7 to argue that there is a kind of perception that necessarily includes pursuit. Unlike perception of a colour or a shape, perception of the object as pleasant essentially involves taking pleasure in and being drawn towards the object. It is impossible to characterize what it is to see something as pleasant without mentioning this distinctive form of response to it. Aristotle extends this account of desire for pleasure to the case of intellectual desire for the good. If the object imagined/thought is good to the thinker, the soul as it were asserts that it is good and thus pursues it. Desire is not an independent component supplied when practical intellect has done its work. Applying this scheme to the view expounded in NE 6-7, we see that προαίρεσις is a distinctive form of both cognition and desire, and it will be faulty if the agent is not properly attracted to doing what is best (205). The characteristic feature of the weak akrates is a failure in the distinctive form of rational sensitivity to value which leads to action.

After alerting us that the terms akrasia and ἐγκράτεια hardly turn up in early Stoic material, due to the wholesale rejection of the phenomenon, Jean-Baptiste Gourinat contrasts the term with its antonym, ἐκράτεια. Both are included in the two main lists of virtues and vices. In Diogenes' list, both akrasia and ἐγκράτεια are subordinate, although it is only the latter that has been defined -- though Diogenes and Stobaeus define it differently -- and subordinated to σωφροσύνη, of which the object is our impulses. This implies that akrasia is subsumed under ἀκολασία. But both akrasia and εγκράτεια are concerned with pleasure, the one dealing with it as a vice, the other as a virtue. There seems to be a difference in the way Cleanthes and Chrysippus explained these phenomena, insofar as the former treats ἐγκράτεια as a cardinal virtue, whereas the latter claims only that, along with akrasia, it is simply related to pleasures. One might ask for a further clarification of the difference between impulse in the strict sense (223) and impulse in general. The distinction plays an important role in the argument that courage does not deal with impulse in the strict sense, though it deals with repulsion.

Ricardo Salles examines Epictetus' views on moral responsibility for precipitate actions. As precipitancy is one form of akrasia in Aristotle's ethics, it is important to know that the Roman Stoic used the same term and seems to give more attention to the problem. On the "Normative Argument" (as Salles calls it), I am responsible for my precipitancy and so can be blamed for it. In Epictetus, the argument seems to rely on the assumption that only those persons can be blamed for precipitancy who are capable of being reflective. They are reflective in the sense that they always examine their first impressions before assenting to them. Nothing rules out that someone forms a cataleptic impression (an impression that cannot be false), but finds it unconvincing given the external circumstances in which the impression is formed (254). Epictetus stresses that there are exercises by the aid of which we can avoid automatic acceptance of the first impression (Diss. 2.18-8-9 and 2.12-14, 23-24 respectively). These exercises, however, are not totally independent from factors beyond our control. One might register another interesting point in Epictetus' account, since on the criteria of some Stoics we cannot have cataleptic impressions unless they prove to be unimpeded. The standard example is Admetus who refused to accept that the one he sees is the resuscitated wife Alcestis. He assumed that there is a better explanation for this impression than that it was actually caused by his wife (Sextus Empiricus AM 7.254, PH 1.228).

Lloyd Gerson examines the scarce evidence for akrasia in Plotinus.4 After surveying the preliminaries, Plato, Aristotle and the Stoics, and claiming that "the distinction between a dualistic and a monistic psychology a propos an account of akrasia is, I believe, something of a red herring" (273), he points out that Plotinus was fairly ready to appropriate Aristotelian and Stoic insights into a Platonic framework. The supposed distinction between first-order and second-order desire in Plotinus is explicitly taken over from Frankfurt's analysis with the aim to show that in the Plotinian context first-order desire is always rational insofar as it has to be conceptualized as good. It can be an involuntary desire insofar as it is not endorsed by a second order-desire. First-order desire is tied to embodiment and therefore is inseparable from evil. One might remark that this conceptualization may fulfill the Stoic criteria for rationality, but might differ from Plotinus' criteria.

There are two indices, of ancient sources and modern authors. The Bibliography is both extensive and informative, but on p. 258 the two items following M. Forschner's Die stoische Ethik are by W. W. Fortenbaugh. I have noticed only a few slips, e.g., on p. 223 πάθη should be not "reasonable" but "unreasonable" impulses. The volume is a welcome addition to the literature on ancient ethics. Furthermore, as some of the papers deal with their subject with a clear eye to contemporary problems in ethics or moral psychology, modern moral philosophers may also find much food for thought in it.


1.   D. Devereux, 'Socrates' Kantian conception of virtue', Journal of the History of Philosophy 33 (1995), 381-408.
2.   He argued for this possibility in 'Simple souls', E. Wagner (ed.), Essays on Plato's Psychology. Lexington, 2001, 137-156.
3.   The Socratic Paradox and Its Enemies. Chicago: University of Chicago Press, 2006, reviewed by Rachel Singpurwalla in BMCR 2007.09.26.
4.   As a matter of fact, the term and its cognates are far from being characteristic of Plotinus' discussion of ethical issues.

Comment on this review in the BMCR blog
Read Latest
Index for 2010
Change Greek Display
Books Available for Review
Bryn Mawr Classical Commentaries

BMCR, Bryn Mawr College, 101 N. Merion Ave., Bryn Mawr, PA 19010

HTML generated at 09:28:06, Friday, 02 April 2010