1 Position in Galen’s Oeuvre

Where does περὶ ἀλυπίας sit within Galen’s writings on ethics and moral psychology, and what does it add to the picture? Galen’s contribution to moral psychology and ethics was previously known mainly from Affections and Errors (Aff. Pecc. Dig. 1 and 2). There is also highly relevant information in the admittedly problematic (because both abridged and to some extent distorted in the Arabic version) Character Traits (Mor.), and in some passages from The Soul’s Dependence on the Body (QAM) and De placitis Hippocratis et Platonis (PHP). The latter two, however (to simplify two highly complex texts), are concerned mainly with certain theoretical propositions, and in particular with aspects of the relationship of soul to body. Affections and Errors and Character Traits, meanwhile – both of which he lists, in his own account of his writings, alongside περὶ ἀλυπίας in the category of works giving his views on ethical philosophy¹ – bear a much clearer affinity to that work. All three belong within a genre of practical or popularizing works of moral philosophy intended for a non-specialist audience; they offer both theory and practical advice in the areas of ethics, education and personal development.

The closest similarity that περὶ ἀλυπίας has with another work in the Galenic corpus is, indeed, with Affections; and I should like now to spend a little time exploring both that similarity and what, specifically, Peri alupias adds to the other work. Both works are designed to help the reader or listener on the path to ethical improvement. According to ancient distinctions both of genre and of stage of personal development, works of ‘protreptic’ – encouraging the reader or listener to embark on the process of virtue acquisition in the first place – may precede a subsequent phase of instruction in which detailed guidance is given about the actual process.² Employing that broad categorization, one would have to situate Peri alupias in this subsequent phase too.

Affections and Errors has as its topic or aim the control of affections (pathē) and errors (hamartēmata) in general; περὶ ἀλυπίας has the specific focus of the elimination of distress (lupē). Some have linked περὶ ἀλυπίας to the genre of consolatio; and other recent work has explored both the philosophical and the literary relatives of the work, and aspects of Galen’s self-presentation within it. There are similarities between the text and others in the tradition of popular ethical writing; it has been suggested that Plutarch’s De tranquillitate animi provides the closest parallel.³

There is, I believe, more to be said, both about the interesting overlaps and differences between Affections and Errors and περὶ ἀλυπίας and about the distinctive understanding of lupē that arises from a consideration of both texts in conjunction. Such an approach is attractive both because Affections and Errors is a fascinating but in many ways frustrating text, unclear in a number of aspects of its organization and in particular giving a quite uneven discussion, and no clear typological categorization, of the different pathē of the soul;⁴ and also because it does, however, have quite a lot to say about lupē which may usefully be placed alongside the new material from περὶ ἀλυπίας.

2 Lupē in Affections and Errors

There are, in fact, passages in Affections and Errors which seem to present lupē as, not just as one pathos amongst many, but in some sense an overarching category. In chapter 7 of Affections, in what it is admittedly a not unproblematic passage textually, it is suggested that there are subspecies of distress, of which envy is one.

ὀνοµάζω δὲ φθόνον, ὅταν τις ἐπ᾽ἀλλοτρίοις ἀγαθοῖς λυπῆται. πάθος µέν ἐστι καὶ λύπη πᾶσα, χειρίστη δὲ ὁ φθόνος ἐστίν, εἴτε ἓν τῶν παθῶν εἴτε λυπῆς ἐστὶν εἶδος πλησιάζον δέ πως αὐτῇ,

By envy I mean becoming distressed at what others enjoy. All distress is an affection, but envy is the worst distress, whether it is an affection in itself or a subspecies of distress, somehow approximate to it …

Aff. Pecc. Dig. 1.7 (V.35 K. = 24,13–16 De Boer)⁵

Moreover, distress or grief assumes a central role for a major part of the text, chapters 7–9.

I digress for a moment to clarify a point of terminology. For the sake of consistency, I translate ania and cognates with ‘grief’ and cognates, and similarly lupē with ‘distress’; however, the two sets of terms seem, in the ethical context, to be regarded as virtual synonyms. Though it is arguable that the former gives, at times, a slightly intensified sense, it seems wrong to insist on a clear distinction. The verbal form ἀνιώµενος is used, for example, of the young man at Aff. Pecc. Dig. 1.7 (V.37 K. = 25,15 de Boer), and the verb is also used of Galen’s own mother who ‘would suffer grief at the smallest occurrence’ (ἀνιωµένην … ἐπὶ σµικροτάτοις, 1.8, V.41 K. = 28,6 de Boer). But the progress of the discussion, in 1.8 (from V.43 K. = 29,15 de Boer), makes it clear that lupē is regarded as the relevant overall heading.⁶ In what follows I shall therefore treat Galen’s discussion of cases of ania and lupē as referring to the same psychological–ethical phenomenon.

In chapters 7–9, then, we gain the impression that the eradication or lessening of lupē is an absolutely central strand in the fight against the affections. The discussion revolves around anecdotal reference to, quotation of and direct address to individuals amongst Galen’s most intimate circle of friends and family. First, at Aff. Pecc. Dig. 1.7 (V.37 K. = 25,15 de Boer), Galen introduces the character of a young man who came to him because of the excessive grief he suffered over small matters. The argument continues to be addressed to this individual’s problem up to the end of chapter 9 – albeit with some major digressions, in particular on the relationship between nature and nurture and on the ethical model offered by Galen’s own parents, and his own philosophical upbringing. But both the digressions and the material directly related to the young man serve to bring out the importance, and multifarious ramifications, of distress. First, Galen attributes to his father a contrast between universally admired virtues – justice, self-control, courage and discernment – on the one hand, and freedom from distress, on the other. The point is that people wish to appear to have the former virtues, but they want actually to be free from distress:

… φαίνεσθαί γε πειρῶνται τοῖς ἄλλοις ἀνδρεῖοι καὶ σώφρονες καὶ φρόνιµοι καὶ δίκαιοι, ἄλυποι µέντοι κατ᾽ ἀλήθειαν εἶναι, κἆν µὴ φαίνωνται τοῖς πέλας· ὥστε τοῦτο µέν σοι πρῶτον ἁπάντων ἀσκητέον ἐστὶ τὸ σπουδαζόµενον ἅπασιν ἀνθρώποις µᾶλλον τῶν ἀρετῶν.

… they wish to appear to others brave, self-controlled, discerning and just, while they actually want to be free from distress, even if it is not apparent to those around them. And this should therefore be what you cultivate first of all, since it is sought after by all people in preference to the virtues.

Aff. Pecc. Dig. 1.8 (V.43 K. = 29,8–12 de Boer)

The practice of freeing oneself from lupē is here presented as the practical, chronological starting-point of one’s ethical progress, on the commonsense grounds that this absence of suffering is something that all people actually seek. We shall see how this perception of Galen’s father’s surfaces again in Peri alupias. Galen also talks of the model his parents provided specifically in terms of their experience of distress. His father ‘never appeared distressed at any setback’, while his mother ‘would suffer grief at the smallest occurrence’. The terms ‘distress’ and ‘grief’ are here clearly being used to apply to one’s reactions to a very wide range of everyday events which are liable to upset one: thus, lupē (or ania) here can be seen as to some extent co-extensive with irritation or anger, even though this, in its more violent manifestations, was dealt with explicitly earlier in the work.

In the part of the text addressed more closely to the young man who wishes to be freed from distress, too, the term turns out to have a very broad reference. One may, for example, suffer lupē not just as a result of personal loss, but in the anxiety over possible future loss, including not just of possessions but of status. Although, as mentioned, Galen does not explicitly give us any categorization of the pathē or account of which are the most fundamental, we are reminded of the fact that lupē is, indeed, an overarching category, an ‘Über-pathos’, in some Stoic sources.⁷ Getting rid of lupē, then, begins to look rather like the Stoic drive for apatheia. Reference to the Stoic relatives of Galen’s thought in this area leads us to another relevant consideration: the absence of the tripartite soul at this point in the discussion. The earlier phase of discussion, based strongly on that Platonic distinction of the drives of the non-rational soul into those of spirited (thumoeides) and desiderative (epithumētikon), clearly implies that any pathos will be a pathos of one of these two – that this distinction will be of fundamental significance throughout. And, as already suggested, the examples that the text dwells on at length seem to be chosen as examples of the malfunction of the spirited – that is to say, of uncontrolled rage.

Yet the discussion of lupē which we have just been considering is interesting precisely because it seems to follow from this broader conception of lupē that it cuts across the spirited–desiderative distinction.⁸ This is supported both by the range of examples of lupē – distress at financial loss, distress caused by fear of loss, distress at perceived lack of status – and by the subsequent argument that the cause of all susceptibility to lupē is – an even more over-arching category – acquisitiveness (pleonexia). For such acquisitiveness or greed may be directed at personal possessions or luxury (surely, in Platonic terms, aims of the desiderative soul), but also at status and perceived position in society (those of the spirited).

3 Lupē and Its Control in περὶ ἀλυπίας

The discussion in περὶ ἀλυπίας contributes to the same picture. Here the Platonic tripartite soul does not, in fact, appear at all. Rather, removing or reducing one’s susceptibility to lupē appears as a procedure which is absolutely fundamental to ethical well-being. Much of the argument proceeds through models: the positive ones of Aristippus and of Galen’s own father, the negative ones of his mother and of the literary man whose distress led to his ultimate demise. Again, alupia seems to amount to something very similar to what a Stoic might call being unaffected by externals – or at least, to being affected by them as little as possible (we shall return to this point).

We might like to say that the two works are complementary: περὶ ἀλυπίας continues, and develops in more detail, particular themes outlined in Affections and Errors.⁹ But in drawing attention to this complementarity, it is important to re-emphasize the point: the more general work Affections and Errors is, to a considerable extent, itself a work about the reduction or elimination of distress. On the other hand, περὶ ἀλυπίας introduces perspectives on lupē that are not to be found in Affections and Errors, or found there much less clearly; and I turn to two of these now, before returning to a consideration of their complementarity and attempting to summarize the findings that accrue from considering the texts conjointly.

The first such ‘new’ feature of lupē in περὶ ἀλυπίας is its potentially severe physical consequences.

… Φιλίδης µὲν ὁ γραµµατικὸς ἀπολλυµένων αὐτῷ τῶν βιβλίων κατὰ πυρκαϊὰν ἀπὸ δυσθυµίας καὶ λύπης διέφθαρη συντακείς

… Philides the literary man, when his books were destroyed in the fire, wasted away as a result of low spirits and distress, and died.

Ind. 7 (4,6–8 BJP)¹⁰

In fact, this Galenic aspect of lupē is not by any means a finding new to Peri alupias, even if it is not mentioned in Affections and Errors. The medical, including potentially fatal, consequences of distress (as also of worry, phrontis and agōnia), as part of a disease pattern involving the connected phenomena of sleeplessness, dryness, heat and fever, is attested in a wide range of passages in Galen’s medical writings.¹¹ Indeed, the specific anecdote that Galen brings forward here about the literary man whose distress over losses similar to Galen’s did indeed prove fatal appears elsewhere, in Galen’s Commentary on Hippocrates’ ‘Epidemics VI’.

Such medical consequences are not the direct subject matter of Peri alupias, which is concerned rather with its prevention. The medical understanding of lupē, however, should be borne in mind as an important element in the intellectual background. For the medical context provides a framework within which lupē is, for Galen, a distinct and observable physical phenomenon. Whether someone is suffering from lupē is thus, in a sense, an objective fact – an affection of the psuchē which – just like those well-known affections of the psuchē in Galen’s anecdotes in Prognosis – is accessible to medical diagnosis.¹² It is not irrelevant here, either, to consider the criterion of a pathological state given in De sanitate tuenda: so long as the person is not distressed by an imperfect physical state, that state still counts as healthy.¹³ For Galen lupē is a concrete, distinct – and potentially a medical – state. It is not a vague characterization of the phenomenon of becoming slightly upset at events.

Another area in which περὶ ἀλυπίας seems to depart from Affections and Errors, or at least to give greater clarity, is in relation to the question – already touched on – of how complete an elimination of pathē is required or desirable. As a number of previous discussions have highlighted, Affections seems practically to align itself with a Stoic approach whereby pathē are in their nature purely negative, and something very close to their complete elimination is the aim. It has also been pointed out that this appears to conflict with his Platonism, or to be more precise with what one might expect at this period from an author indebted strongly to both Plato and Aristotle in his ethical thinking, and in particular that the term metriopatheia – the ‘moderation of the pathē’ – which appears in some ‘Middle Platonist’ authors is not mentioned by Galen.¹⁴

It is, of course, true that, within the Platonic tripartite model which is of such importance to Galen, including in Affections, anger – the righteous indignation of the thumoeides which checks the wild desires of the epithumētikon – has a positive, indeed an important, role, in a way which is quite contrary to Stoic thinking. For Galen, however, though this internal dynamic within the soul is important to his analysis, anger when functioning in this way apparently does not come under the heading of pathos. Pathos, for Galen, seems, in the ethical context, to be a purely negative term: that much he has taken over from Stoic usage, however bitterly he opposes the broader intellectual framework within which that usage arose. That is to say: there is, for Galen, a legitimate role for the non-rational parts of the soul, but pathos arises only in these non-rational parts and only when they are not behaving legitimately; for the Stoics, meanwhile, there is no such positive role for the non-rational parts of the soul, which indeed, properly speaking, do not exist: it is rather errors of rationality which lead to or constitute behaviour in pathos. Thus, Galen shares with Stoic thought the negative definition of pathos, while having a different understanding, not only of how pathos comes about and where it is located in the soul, but also of where pathos fits into the broader scheme of non-rational drives.

So we might say that even if Galen does acknowledge a positive role for some of (what the Stoics would call) pathē, this would not mean that he is advocating metriopatheia: for Galen, pathē are in their nature negative, and when anger (say) is functioning positively on behalf of the person that is not a case of pathos, not even of moderated pathos. Thus, an Aristotelian understanding of proper ethical/emotional response as consisting in some mean between opposites – that is, the exactly correct sort of pathos – seems to be absent from Galen’s thinking in his ethical writings.¹⁵

But there is a further question, or complication. Even if Galen (a) takes there to be a positive role for some emotions, but (b) does not refer to positive manifestations of emotions as pathē, and therefore (c) does not advocate the concept of metriopatheia, there remains a further question: is the total elimination of those emotions which are regarded as purely negative – those ones which Galen and the Stoics would both call pathē – required? Galen’s ‘official’ answer seems to be no – again, in keeping with a fundamentally Platonic–Aristotelian model (albeit one without the terminology of metriopatheia) – although, as discussed above, one can certainly gain the impression, throughout much of Affections and Peri alupias, that total elimination is indeed what is being advocated. In this context, it is also relevant to consider that Galen at least arguably (if one accepts a particular emendation of the text) allows also a moderate level of emotional attachment to societal status and political power, and even wealth. As elsewhere in Galen’s work, excessive preoccupation with status or reputation is, to be sure, considered a great evil. But on Garofalo’s emendation of sections 80–81 (see further below), freedom from distress is equated with the possession of only a moderate level of attachment to these aims.

The desirability of elimination of lupē again points to its special status. Unlike thumos or epithumia, for example, it has no positive role. One might indeed be tempted, following the analysis outlined above, to suggest that this is precisely what lupē is for Galen: the negative or pathological manifestation of non-rational drives which (as we have seen) are not in themselves necessarily pathē. Lupē, then, like pathos in general, has no positive role for Galen. Does that mean that we should aim for or require its complete removal? In spite of what I have referred to as hints of a Stoic-style apatheia, Peri alupias gives us something the other text does not, or at least gives us much less explicitly: a specific affirmation that one cannot always be unaffected by circumstances.

Addressing the issue directly (in sections 70–76), Galen explicitly denies the proposition that a person – or at least that he personally – can remain free from distress in every eventuality. In contradistinction to the extreme Stoic and Epicurean claims on unaffectedness, Galen prefers a more common-sense position. He knows his own limitations; he does not, like Musonius the Stoic, ask to be tested by every possible adversity; he does not accept that one can be happy inside that notorious philosophical example of torture, the bull of Phalaris; and he mentions specific circumstances that he knows would cause him distress (the destruction of his home city, the persecution of a friend by a tyrant). So, the text of περὶ ἀλυπίας makes it clearer and more explicit than that of Affections that, in spite of the desirability of freeing oneself from lupē as much as one can, total indifference to, or unaffectedness by, externals cannot in all circumstances be expected. It is just that we should aim for much higher expectations and achievement in this area than are normally the case. Quite how high a level of achievement he expects, or (for he is more explicit on this point) attributes to himself, is a somewhat complex question. A very high level of impassivity to the vicissitudes of fate will be termed megalopsuchia;¹⁶ and Galen does indeed attribute this quality to himself. At other points he emphasizes that his failure to succumb to distress, at least in response to most of his losses, was ‘no big thing’. The argument of the text functions, of course, by constantly emphasizing the enormity of his losses in order to highlight the distinctiveness of his reaction – a reaction of refusing to consider those losses to be enormous.¹⁷ He is ‘not at all moved’, ‘not now distressed, cheerfully carrying out my usual tasks as before’, ‘bearing without distress’, ‘not distressed, even with all such things touching me’; ‘I bore it very easily, not moved in the least’, ‘none of these things distressed me’, ‘I was not distressed as others, but bore the event very easily, after losing such a great variety of possessions, any one of which on its own would have been most distressing to others’.¹⁸

Before returning to a final consideration of the complementary nature of the text and of the overall picture of lupē that emerges, I consider one more specific area in which περὶ ἀλυπίας seems to diverge, or offer something distinct from, Affections and Errors, this time in the sphere of practical advice. First, the central policy recommended in the latter work – that of finding a neutral advisor to monitor and report to one about one’s faults – does not appear in Peri alupias. Conversely, the main technique which the latter work does prescribe, the praemeditatio malorum – that is to say, a sustained daily practice of anticipation of the worst, a practice which may include the internal or actual repetition of certain texts or propositions – does not appear, at least not explicitly, in Affections and Errors. A regular mental practice, involving recitation – specifically, of the Pythagorean Carmen aureum – is recommended in Affections, along with a process of self-interrogation whose rational force will affect one’s ethical behaviour. This practice, however, is based rather on the daily examination of one’s previous actions. περὶ ἀλυπίας gives us a further dimension of the use of text recitation for ethical or psychological purposes. A quotation from Euripides is central to the text’s message on the praemeditatio malorum. (Interestingly, the same text appears in De placitis Hippocratis et Platonis, there in support of a theoretical argument about how to conceptualize the process of praemeditatio within the soul.¹⁹) It is perhaps noteworthy that there is no appeal here to the use of philosophical texts in one’s daily exercise – a fact that one may relate to Galen’s insistence that his father’s successful moral education, described as similar to his own, was achieved ‘without arguments from philosophy’ and that he ‘did not frequent philosophers in youth’.²⁰ One may, indeed, connect this with Galen’s sceptical attitude towards the discipline of philosophy, certainly as generally practised in his own time.²¹

4 Practical Ethics and Life Aims in περὶ ἀλυπίας and Affections and Errors

But let us consider some further aspects of the complementary nature of περὶ ἀλυπίας and Affections and Errors. In Peri alupias, we again meet Galen’s father, and in a similar context. It is not just that his father was a model in his freedom from distress – that is, his ability not to be affected by adverse events. Rather, here too a specific perception is attributed to his father, one which matches that reported in Affections and Errors. Let us look at the passage, which may be compared with that cited above. (I follow the text of BJP, and excerpt what seem to me the most relevant phrases from a fairly long passage.)

οὐ γὰρ ἄλλος ἀνθρώπων τις οὕτως ἀκριβῶς ὡς καὶ οὗτος ἐτίµησε δικαιοσύνην τε καὶ σωφροσύνην… οἶδα δέ µου τὸν πατέρα καταφρονοῦντα τῶν ἀνθρωπίνων πραγµάτων ὡς µικρῶν… τοὺς ἥδιστα βεβιωκότας οὐδὲν ἔσχε πλείω τῶν οἰωνῶν τούτων οὓς κατὰ τὴν τῶν Ῥωµαίων πόλιν ὁρῶµεν ὑπὸ τῶν δεσποτῶν περιαγοµένους ἕνεκα τοῦ τὰς θηλείας ὀχεύειν ἐπὶ µισθῷ· τοὺς δὲ τῶν τοιούτων ἡδονῶν καταφρονοῦντας, ἀρκουµένους δὲ τῷ µήτε ἀλγεῖν µήτε λυπεῖσθαι τὴν ψυχήν, οὐδέποτε ἐπῄνεσεν ἀποµαντευόµενος µεῖζον τι καὶ κρεῖττον ὂν τὸ ἀγαθὸν ἰδίαν ἔχον φύσιν, οὔτε ἐν µόνῳ τῷ µήτε ἀλγεῖν µήτε λυπεῖσθαι περιγραφόµενον. ἀλλ᾽ ἐὰν καὶ τούτων τις ἀποχωρήσας ἐπιστήµην θείων καὶ ἀνθρωπίνων πραγµάτων ἡγήσηται τὸ ἀγαθὸν ὑπάρχειν, ἐλαχίστου µορίου τούτου ὁρῶ τοὺς ἀνθρώπους µετέχοντας… ὁ γὰρ ἐν τῷ καθόλου µὴ γινώσκων ὁποῖα τά τε θεῖα καὶ τὰ ἀνθρώπινα πράγµατά εἰσιν, οὐδ᾽ ἐν τῷ κατὰ µέρος οὐδ᾽ ἐπιστηµονικῶς τι ἑλέσθαι καὶ φυγεῖν δύναται.

For no other man esteemed justice and self-control as completely as he … I know that my father despised human affairs as trivial things … he valued those who live a life devoted to pleasure no more highly than those birds we see being taken round Rome by their masters to service females for a price. But those who despise such pleasures, and are content with neither experiencing pain nor distress in their souls, he never praised. He declared that the good was something bigger and more powerful than that, something which possessed its own nature rather than being defined only in terms of not suffering pain or distress. But if someone departs from these and holds that the good is a knowledge of matters both human and divine, I see that human beings possess only a very small part of this … For someone without even general knowledge of matters human and divine cannot choose scientifically in individual matters, either, what to choose and what to avoid.

Ind. 58–64 (18,22–20,10 BJP)

The refusal to praise those who are satisfied with being free from distress can surely be placed alongside the remark in Affections suggesting the attempt to attain alupia as a crucial, but not sufficient, stage in ethical progress.²² This feature of the Peri alupias argument – that it does not present a straightforward rejection of Epicurean aochlēsia and/or Stoic apatheia, but rather a statement of their insufficiency – seems to me a vital one. The two passages are, in fact, of a piece: both are suggesting the drive towards alupia as a practical starting-point in the attempt at ethical self-improvement; and both are asserting that alupia is necessary, but by no means sufficient, for virtue. If we wish to talk of ends or goals, we must mention ‘knowledge of things human and divine’ – however imperfectly we may attain to that.²³

As both this passage and that immediately following makes clear: (a) human affairs are to be despised; (b) freedom from distress is valuable; but (c) it is not sufficient, as there are higher human aims. These are reasserted a few lines later in the words ‘wishing to be actively engaged in both mind and body’, πάντα … ἐνεργεῖν … βουλόµενα καὶ κατὰ σῶµα καὶ κατὰ ψυχήν, 68, 21,7–8 BJP. (One must acknowledge, here, that this is presented as not merely a human aim, but in fact also that of ‘all animals’, the word πάντα here referring back to οὔτε ἐµαυτὸν οὔτε ἄλλον ἄνθρωπον οὔτε ζῷον τι – and in doing so one must also acknowledge that there an anti-Epicurean rhetoric at work here which has arguably taken Galen to a slightly unusual place in his argument, as made clear by the pejorative mention of aochlēsia and, indeed, by the explicit reference to other writings in which he attacks Epicurus.)²⁴ There is also a similarity between the two texts in the way in which these ‘higher-level’, or intellectual, aims are presented – and perhaps above all in the vagueness with which they are presented. If we turn to the relevant discussion in Errors – the part of that text devoted to the rational soul as opposed to the non-rational – two things seem striking in this context. One is Galen’s apparent slipperiness when it comes actually to defining ‘the goal of life’;²⁵ the other is that, whatever the precise answer on that, he is more interested in persuading one to engage in rational training and rational scientific activity than in any goal which would more obviously be defined as ethical.

One might even say that rational or intellectual activity of the correct kind, in that text, provides an answer that seems to stand in place of the answer to ‘the goal of life’; and that, perhaps, corresponds (at least as far as human beings are concerned) to what Galen describes here as ‘being actively engaged in both mind and body’. Galen (or his father) seems to have developed an interesting practical-ethics approach here. We might summarize the two-step approach as follows:

1. Ethical improvement must start with the identification of something that causes one actual distress, lupē. Once one has achieved that identification, the desire to make a change allows the possibility, at least, that one will make some progress. One is no longer in denial, at this stage, and may seek practical interventions to lessen one’s susceptibility. If one then succeeds in radically reducing one’s susceptibility to lupē, this is a necessary, but by no means sufficient, condition of virtue.

2. While it is difficult to define precisely the goal in this higher realm – that of the rational soul – some things are clear. In particular, (a) the aims pursued by most people – political ambition, accumulation of wealth – are to be despised; and (b) some kind of mental engagement, or directed activity, is essential to human life. Fairly clearly, too, the type of such activity strongly preferred by Galen (and his father) is that aimed at the acquisition of knowledge, especially (as far as possible) knowledge of a mathematically reliable kind.

In relation to step (ii), and in particular the definition of ‘the goal’, there is, as already suggested, some vagueness – though we should here acknowledge the limitations of our sources, since a considerable list of Galen’s ethical writings is lost to us. But it also seems at least possible that Galen is deliberately vague in this context, preferring a strong argument in favour of intellectually rigorous and mathematically-based mental activity to a conventional definition of virtue of the sort approved by any of the established philosophical schools. The above talk of a two-step approach should not, however, be taken to deny the interconnectedness of the phases. The removal, or reduction, of one’s liability to distress is for Galen intimately related to the adoption of appropriate life aims, or to which things in life we take to be valuable or not valuable. The early education mentioned above simultaneously instils appropriate notions of what counts as good and habituates one to appropriate reactions and behaviour: the rational (evaluative) and non-rational (habituated) responses go hand in hand. This connectedness is particularly reinforced, in περὶ ἀλυπίας, in sections 80–81, where the absence of distress is closely correlated with an appropriate assessment of the aims of honour, wealth, reputation and political power. (And especially so if we adopt the reading of this passage suggested by Garofalo, whereby ‘those who do not suffer distress as the many do’ are equated with ‘those who have a moderate attachment to honour, etc.’)²⁶ Here again, it seems, the taking of the two texts, Affections and Errors and περὶ ἀλυπίας, alongside each other, has helped to form a picture of Galenic thinking in this area.

A final point is worth our consideration: what range of emotional reactions is admissible within alupia? For Galen, as we have already suggested, lupē is a quasi-medical category. The usual context for its mention is in consideration of predisposing causes that can lead to physical ailments of various kinds. The example of the literary man dying of grief should not, perhaps, from this perspective, be seen as an extreme one. This, Galen seems to suggest, is within normal medical experience: it is the sort of thing that lupē can do, or rather lead to. Galen’s boast is that he was seen to be ‘not moved at all’, ‘not distressed’, that he ‘bore it easily’; and scholars have not been slow to point to both the boastfulness and the apparently unrealistic nature of the claim. But what exactly is meant here? It seems to me that Galen is not, in fact, presenting some other-wordly, saint-like behaviour. The point rather is that he is able to go about his daily business; he does not succumb; he does not allow his life to be ruined.

The terms used for what does not happen to Galen – kinēthenta, aniathēnai – seem to me perfectly consistent with the notion that one experiences some negative emotional reactions; what is crucial is that they are controlled, not allowed to dominate. And such control is a perfectly possible – Galen quite plausibly argues – as a result of the right kind of training in childhood in combination with ethical discipline, involving a consideration of how small such setbacks are in the scheme of things, in adulthood. The social aspect of one’s reaction, too, is relevant to this discourse. Galen uses a range of expressions to describe the visual or outward aspect of his behaviour: you saw that I was not moved, I was observed bearing it easily; the reaction is described as wonderful; it is compared with that of others; the term phaidron, too (literally ‘bright’, ‘radiant’), refers to an outward demeanour or impression. The observable, outward aspect must be considered; self-control includes a competitive element: one is judged by one’s ability publicly to rise above the normal reaction. But there must, surely, be a range of negative emotions which a person may experience without being defined as falling into lupē.

By lupē, in short, Galen means something more dramatic and more specific – and, in medical terms, far more dangerous – than a controllable feeling of sadness or annoyance. It is a negative emotion whose control is central to the ethical project of self-improvement, and which if uncontrolled can have disastrous medical consequences. In both contexts, the ethical and the medical, Galen develops the concept in a distinctive and original way. A way which, above all, attempts to do justice to the realities, the challenges and the dangers of lived experience.