Bryn Mawr Classical Review 94.06.09

Julia Annas, The Morality of Happiness. New York/Oxford: Oxford University Press, 1993. Pp. x + 502. $55. ISBN 0-19-507999-X.

Reviewed by Dirk T. D. Held, Connecticut College.

Philosophers seeking alternatives to the principles dominant in modern ethical thought have recently looked to the ancient ethics of virtue and character. The wide response to MacIntyre's After Virtue, the continuing debate over virtue ethics, and the writings of communitarian philosophers such as Charles Taylor reflect dissatisfaction over the inability of contemporary ethics, despite technical sophistication and logical rigor, to make clear what it takes to live a moral life. Ancient ethical theories in contrast, as Julia Annas reminds us, originate from reflection over final ends and a person's life as a whole. Because of this perspective and because ancient ethics is 'agent centered' (not 'act centered' like modern ethics), ancient ethical theorists can readily address the demands of living 'the developed moral life.' (65)

Modern ethical theories are generally typified by some form or combination of two principles. Theorists invoking the principle of duty as the foundation of morals strive to articulate obligations that hold universally and absolutely. Kant is the most famous proponent of this type of theory, known as deontological (classical scholars will be among the few to recognize the etymology). The appeal of deontological theory is its impartial application to all. Yet its very universality, suitable for Kantian noumenal selves, isolates it from the contingency bound practices of human beings. The other principle informing modern ethics is consequentialism, of which utilitarianism is a well known version. Consequentialist theories measure ethical value by the consequences of actions, aiming to ensure a positive balance of good over bad -- however defined -- either through following rules to maximize good (rule utilitarianism) or calculating likely outcomes of acts in terms of good and bad (act utilitarianism). Both consequentialist and deontological approaches have a tendency to set moral problems in isolation from each other and as a result to fall into abstraction. Nonetheless they are regarded as moral theories because they orientate right and wrong from a person's effects on others. Ancient theories are said to contrast in this respect since they set forth constitutive conditions for the good life and the agent's happiness (eudaimonia). Gregory Vlastos accordingly claimed that all Greek ethical systems were based on a 'Eudaimonist Axiom,'1 which holds in various articulations that everyone desires happiness and that it is the ultimate goal of all rational activity. Under this description, ancient eudaimonism appears antagonistic to the main focus of modern morality.

We are therefore confronted with a choice between distinct normative standards, Right and Good. But is the choice absolute? Must the ethics of the good life be incompatible with the morality of right conduct? In short, is there anything 'moral' about ' happiness'? These are the questions Julia Annas takes up in the challenging and important book under review. In The Morality of Happiness, Annas presents both a brilliant analysis of ancient ethical theory and a powerfully argued defense of the priorities and approaches taken by ancient philosophers to ethical questions. Her mastery of a wide range of ancient (and modern) texts along with her skilled exposition of philosophical argument make this an outstanding book. Annas' analysis of ancient ethical theory is complex and subtle, aided by her facility at fleshing out abstruse argument with clear examples and lucid prose. The reader also benefits from the clear architectonic structure Annas applies to the whole enterprise.

Annas not only argues that is there no incompatibility between the good life and right conduct, but that ancient theories of virtue and modern notions of what is moral are conceptually alike. Indeed in the introduction she states the thesis that 'ancient ethical thought is a recognizable form of moralthought.' (14) This is not the only way to address the reconciliation of right and good. Hans Krämer for example in his Integrative Ethik (Frankfurt 1992) moves to integrate the ethics of the good life with the moral aims of modern philosophy but steadfastly rejects collapsing one normative field into the other. It is illustrative that both Krämer and Annas rely on the formal dimensions of ethical argument to build their respective cases. Krämer does so because the plurality of goods in the non-teleological modern world precludes unequivocal privileging of any single end. Annas on the other hand strategically exploits the 'thinness' of happiness as a final end to argue that upon philosophical reflection the ancients grasped the inseparability of an agent's happiness from the interests of others.

The Morality of Happiness is not a history of the traditional doxographic sort, nor did Annas set out to write such a work, but an analysis of the formal structure of ancient ethical theory. A significant consequence is the near total exclusion of Plato on the grounds that the dialogue form renders Plato's own position on the structure of ethical theory opaque. For this reason Annas begins with Aristotle, the first to 'lay out the framework of Greek ethics' as a 'distinct subject' (17), and continues through the Stoics, Sceptics, Cyrenaics, Epicureans, and writers of the later Aristotelian school. Here her familiarity with some of the less traveled domains of ancient philosophy pays handsome dividends.

Part I of the book explores the nature of a final end, the structure and status of virtue, and the relation of virtue to a final end. As noted above, ancient ethical discourse takes the agent's life as a whole for its focus and starting point. Annas agrees with Aristotle that we do things for a reason and regularly subordinate ends to other ends. She further argues that what Aristotle says agrees with our own practice, and that careful reflection will reveal the cumulative movement of end-directed actions towards a single final end (33). Such a unified view of our goals, Annas admits, does not sit well with moderns. One needn't be a dues paying post-modernist to protest that fragmentation and disjunction seem more typical of life in the electronic age than any kind of unity. This is not to suggest that Annas follows the path of those promoting any kind pre-modern world view. She simply contends that rational reflection bears out the truth of the fact that we do look at our lives as a whole, and indeed that to be rational, we must live our lives in an organized way. Even the ancient thinkers she cites do not agree on which specific ends should be pursued. They do, however, agree that 'the agent's life has a unitary structure' (33, n.18) provided by the formal constraints required of a final end. Aristotle notes these as completeness and self-sufficiency. The latter remained a concern for philosophers after Aristotle but was dropped as a formal requirement. For all ancient philosophers completeness was the principal formal constraint on final ends and retained the meaning Aristotle gave it: the end must be both final and comprehensive, indeed final because comprehensive. 'The good that gives shape to the aims of a human life ... includ[es] and organiz[es] into a whole the ends of smaller scale desires.' (40) That good is generally called happiness (eudaimonia). However, to say that our final good is happiness, Annas points out, is to say very little. Although its formal role in ancient theories is clear, it is so thinly specified that ongoing disagreement over what happiness is made this 'the most important and central question in ancient ethics.' (46)

Annas holds that a grasp of the structure and status of the virtues is necessary for an agent adequately to understand his final end. Virtues have a complex nature: they are dispositional, habituative, and intellectual. Modern theories of morality which stress decision making privilege reasoning and assume it sufficient for moral action. This ignores the lived life, the development over time of the disposition to make right judgements, and the habituation of such dispositions. This is what is required to 'have a virtue,' and will constitute phronesis a term that connotes not only the reliable regularity of virtuous judgements but the unity of the virtues as well. 'Fully to have a virtue involves grasping what it contributes to the good of one's life as a whole; and when we think this through we see that all the virtues are unified in an agent's deliberations, so that if you have one you have them all.' (77) After laying this largely Aristotelian groundwork, Annas examines the views of Stoics, Epicurus, and Sceptics on these matters. She concludes that 'there is no one favoured paradigm for moral reasoning' although the schools are united in stressing qualitative differences between the novice and practiced moral reasoner. An agent will be one or the other, and the ethically mature agent will better grasp the point of why he acts virtuously.

The scope of ancient ethics is larger than that of modern ethics. The reader is reminded that Aristotle includes as virtues many things that strike us as peculiar and unrelated to virtue. Yet these virtues emanate from the whole fabric of ancient social life which is the necessary context for an agent's understanding of his final end. Annas rejects the suggestion that such inclusiveness is not relevant to our moral life, for to bring 'the whole of our lives into reflective focus' cannot be done in isolation. This additionally answers the claim that an ethics of virtue must be egoistic, devoid of concern for others: since the virtues as a whole have to be developed, and these include justice, courage, and the like, the good of others is immediately implicated. Annas allows that 'An ethics of virtue is at most formally self-centered or egoistic; its content can be as fully other-regarding as that of other systems of ethics.' (127) Later, in Part III, Annas presents evidence to the effect that concern for others not only can be accommodated by ancient theories but that the Stoics in particular move from an original concern for oneself and one's happiness to impartiality in moral judgements between one's own interests and the interests of others even unknown to the agent. Annas concludes this section confident in the truth of her intuition that ancient theories of virtue will prove to be also theories of morality.

Having secured ancient ethical judgement within the framework of a final end Annas turns in Part II to a question critical for any ethical system, namely how ethical judgements are to be grounded or justified. Ancient philosophers regularly sought justification in nature. They did not however look for it in the constricted notion of what modern ethical naturalists might consider 'brute' nature but sought it in the realities of human nature, a realm with a distinct normative value. Annas traces the appeal to nature starting with Aristotle and proceeding through the various philosophical schools which succeeded him. Her intention in this part of the book is to show that though 'the appeal to nature takes different forms in different theories it forms a common element in ancient eudaimonistic ethical theories.' (142) In Aristotle's teleological view, human nature is 'not just a pre-ethical starting point but something which provides ethical goals.' (146) This perspective is eventually refined by the Hellenistic schools, most notably the Stoics, into a coherent ethical argument that equates the agent's final end to a life in accordance with nature. The Stoics, for all their determinist talk of 'the nature of things,' do not see nature as an independent foundation of ethical judgements. Annas relies on Cicero for evidence of the interdependence between nature and virtue, and concludes that 'Virtue and [human] nature both function as parts of a theory which is built up as a whole; neither has priority.' (173) Chapters on Antics of Ascalon, the Epicureans (who in removing our false beliefs about pleasures 'rethink what is natural' until it becomes an ethical ideal), and the Sceptics round out the discussion. Her view is that nature in ancient ethics has been misunderstood. Rather than a pattern to which humans should conform, nature has a limiting role: negatively, it sets parameters on what is possible; positively, it endows human rationality with developmental schemata leading to ethical behavior. Finally, nature functions as an ethical ideal attainable by the agent possessing the necessary skills in practical decision making.

In Part III, titled 'The Good Life and the Good Lives of Others,' Annas faces head on the challenge of incorporating the needs of others into eudaimonism. She separates the agent's final good from simple self-interest by relying on two facts. One is that acting virtuously is the way to successful attainment of the final good, the other that the virtues are dispositions to do what is right as established independently of the agent's interests. Therefore even though ancient ethical theories formally are agent-centered, they are not necessarily self-centered in respect to content and if so, not egoistic. She might strengthen the grounds for this approach by noting that one's own good and the good of others are construed differently by moderns than by the ancients. In ancient societies a person's moral identity is far intricately embedded in the social context than is the case for moderns who only begrudgingly turn their post-Kantian eyes away from the ideal of the morally autonomous individual. An ethical system based on the perspective of a whole life would be hard pressed to remove social inter-relations as an area of primary concern. In fact, only one group of ancient thinkers did not think one's life as a whole was important for a person's ethical choices. They were the Cyrenaics, whose hedonistic philosophy valued the interests of others only as they were useful to the agent's own interests. But as Annas notes, they were the exception that makes the rule. She uses them to make the point that crude hedonism does not fit well with the assumptions that ethical theory starts from reflection on one's whole life and needs to include some concern for others. Even Epicurus is rehabilitated to fit Annas' requirements, since he stipulates in K.D. 5 that a life of virtue is a necessary condition for pleasure (the right kind of stable pleasure), and the virtues are determined independently of the agent.

Aristotle first raises the question of whether limits should be put on the concern for others, and concludes that the ethical importance of philia does not entail unlimited concern for such, in Aristotle's words, as 'the furthest Mysian.' We have to wait until the Stoic doctrine of oikeiosis before impartiality towards all others, including 'the furthest Mysian,' is seen as an important component of the ethical point of view. Such an inclusive doctrine extends concern for those close to us, as family, indefinitely outwards. There is no way ethically or rationally to break off the commitment at any point, and this fact becomes the basis of Stoic ideas about justice and community life. After this, things were never the same, and the Aristotelians were forced to accommodate ethical impartiality into their systems. An issue related to impartiality is justice, the virtue most obviously other directed. It is complicated by its application to societies and institutions as well as to individuals, and in a survey of the topic Annas makes note of the difficulty of giving a definitive account of justice as it is related to eudaimonism.

Part III concludes that charges of ethical egoism against the ancient theories are not valid. Annas determines 'that there are no structural barriers in a eudaimonist theory against the acceptance even of demands for impartiality' towards others (323), and for the most part morality is necessary for happiness. The ancient theories differ among themselves in the attention allotted the interests of others, ranging from the anomalous Cyrenaics who assign these interests only instrumental value to Stoic demand for impartiality between these interests and one's own. Nonetheless nothing structural in eudaimonistic theories separates morality and self-interest. In the ancient theories only 'superficial' philosophical tensions exist between my own interests and the interests of others.

At this point, the reader may ask whether Annas has reconstructed happiness to the extent that it no longer fits our intuitive understanding. Modern interpreters have had perhaps equal difficulty as the ancients in defining 'eudaimonia.' Annas acknowledges this dubiety and singles out the refractory nature of happiness as a positive impetus for philosophical reflection. The success of ancient theories, she claims, was that they were all 'more or less revisionary.' (331) The impetus to 'Revising Your Priorities,' the title of Part IV, is an impetus to abandon unreflective acquiescence to intuitions about our happiness and final end. It turns out on Annas' reading that the ancient debate about virtue and happiness is a debate, with various outcomes, over what parts of our intuitions regarding happiness should be retained and what parts abandoned. It is the success, not the failure, of ancient ethical theory that it leads away from an intuitive understanding of happiness.

This approach is applied to the subjects of Annas' study in separate chapters. Epicurus is shown to have a surprisingly subtle view, making ataraxia dependent upon careful reflection and monitoring of desires, with the resultant internalization of the final end and indifference towards external goods. The Sceptics likewise are after ataraxia, for whom happiness comes by realizing the futility of holding anything, external or internal, good in itself; to do so only creates anxiety. Based on Aristotle's inability to provide a coherent account of the relation of external goods to happiness, Annas finds in him a deep tension between the relation of virtue to happiness. This unresolved tension presented a problem for his Peripatetic successors and for the Stoics. The latter found that external goods were not required for happiness, and that hence virtue was sufficient for happiness. Aristotle, giving considerable credence to common endoxai that valued worldly goods and success, holds that while virtue may be necessary for happiness it is not sufficient. Theoretically the Stoic position is neater, but Annas grants that Aristotle's position on this may indicate 'the limits of ethical theory, at least of theory that aims to stand in a realistic relation to people's ethical views.' (424)

These debates, in Annas' view, are really about the role morality has in happiness, which can be seen as the role morality has in the good life. The choice between Aristotle's insistence that virtue was necessary for happiness and the good life, or Stoic arguments for the sufficiency of virtue (even when one is on the rack) for happiness can not be decisively made. Annas finds a parallel to modern debates over ethical theories which have varying claims on our intuitions and reason. The ancient debate about the role of virtue in happiness is conceptually akin, argues Annas, to the role of morality in our own lives. The predicament of the ancients is our own.

The book concludes with a comparison of the structure of ancient and modern ethics, including 'virtue ethics' (the ancient form is inherently revisionist, the modern morally conservative). There are brief biographical notices of ancient philosophers mentioned in the text, a list of primary sources, and two indices in addition to an extensive bibliography.

The Morality of Happiness will generously repay thoughtful reading. It is distinguished both by the comprehensive scope of its general thesis and the subtlety of its details. Annas offers an abundance of provocative arguments on individual philosophers and individual passages, certainly far more than can be adequately scrutinized even in a lengthy review. It will be required reading for any scholar of ancient ethical thought, and contemporary moral philosophers will find rich insights to bring back to their work as well.


  • [1] Gregory Vlastos, Socrates. Ironist and Moral Philosopher, Cambridge, 1991, p. 203 ff.