Hostname: page-component-74656f9884-skzmx Total loading time: 0 Render date: 2026-11-06T00:00:15.739Z Has data issue: false hasContentIssue false
  • English
  • Français

THE SPECTRUM OF HEALTH IN ANCIENT GREEK MEDICINE AND PHILOSOPHY

Published online by Cambridge University Press:  21 September 2026

Arthur Harris
Arthur Harris*
Affiliation:
University of Cambridge
*
ah954@cantab.ac.uk
Rights & Permissions [Opens in a new window]

Abstract

This article examines a fundamental problem in ancient Greek medical thought, namely, how to conceptualize the variety of human health conditions. Beginning in the Classical period, the notion that health gradually shades into illness stood in tension both with the assumption that these categories are dichotomous and with the definition of health as an exact proportion. After examining how these matters were framed by Aristotle, this article explores two later attempts at solving the tensions. It is shown that Galen of Pergamon used the resources of Peripatetic philosophy to reconcile two of these conflicting intuitions while following certain Herophileans in rejecting the idea that health and illness are dichotomous. This article also offers an interpretation of the largely overlooked position of Asclepiades of Bithynia, that health does not admit of degrees. It is argued that this was a reasonable view within a corpuscular physiology and did not, pace Galen, lead to the ‘doctrine of perpetual suffering’. Asclepiades could have effectively deployed his account in arguments against the Empiricist sect and against humoural theories.

Keywords

AsclepiadesGalenhealthillnessancient medicineAristotle

Information

Type
Research Article
Information
The Classical Quarterly , First View , pp. 1 - 17
Check for updates
Copyright
© The Author(s), 2026. Published by Cambridge University Press on behalf of The Classical Association

1. INTRODUCTION

‘Is health uniform, or varying since it increases and diminishes?’Footnote 1 So the late Roman medical writer of the Methodist sect, Caelius Aurelianus (fl. c. fifth century c.e.), raised an issue of wide relevance, still with us today, about how to conceptualize the variety of human health conditions. Ancient Greek attempts to grapple with this problem, however, have been incompletely understood because the specific disagreement mentioned by Caelius has often been treated one-sidedly. The view that health is uniform, an all-or-nothing affair, which Caelius attributes to Asclepiades of Bithynia (before 91 b.c.e.), is liable to seem perversely crude.Footnote 2 This article argues, to the contrary, that the problem touched upon by Caelius pervaded Greek discussions of health and sickness throughout antiquity. The notion of health and illness as contrary poles of a continuous spectrum was only one possible model. The contrasting view that health does not admit of degrees was a live and credible alternative, one that might even hold value today.

Both views can be understood as attempts to resolve the uneasy relations between three common assumptions. The notion (i) that health and illness come by degrees conflicted with (ii) the definition of health as an exact proportion of some kind, and (iii) the assumption that there is a fundamental rift between health and illness, that they are mutually exclusive and dichotomous.Footnote 3 These three assumptions are in some form detectable as early as the mid fifth-century b.c.e. Herodotus and the Hippocratics made comparative judgements that some healthy people are healthier than others, thereby implying that health comes by degrees.Footnote 4 At the same time, Alcmaeon defined health as ἰσονομία and some Hippocratic writers claimed that health was an exact proportion of one kind or another.Footnote 5 The notion that there is a fundamental divide between health and illness, and nothing in between, can be traced back to Aeschylus’ ominous description of sickness as health’s ‘neighbour with a common wall’ in the Agamemnon, first performed in 458 b.c.e. Footnote 6 It was later reflected in the Hippocratic juxtaposition of the stability of health and the flux of disease.Footnote 7

These tensions in early Greek medicine have been noted by earlier scholars, and numerous studies have explored Galen’s conception of the spectrum of health.Footnote 8 But Asclepiades’ view that health is uniform has not been closely examined, nor have Galen’s criticisms of this doctrine been appropriately scrutinized. Asclepiades’ position was more plausible than has been recognized and arguably has advantages over Galen’s apparently more nuanced account. This article takes a diachronic perspective, beginning with Aristotle, to better contextualize the debate and question Galen’s self-portrait as the torchbearer of classical tradition. In truth, Galen and Asclepiades both rejected traditional assumptions in their accounts of health.

Section 2 examines how Aristotle brought the tensions out in the open by clearly articulating assumptions (i) and (iii). Section 3 argues that Asclepiades’ position was more reasonable than the discussions in Galen and Caelius Aurelianus suggest. The doctrine that health does not admit of degrees did not lead to the ‘doctrine of perpetual suffering’ and may have served in arguments against the Empiricists. Section 4 shows that Galen’s account of health defuses the tension between (i) and (ii) by a similar strategy to that used in earlier Peripatetic accounts of virtue such as Aspasius’. Meanwhile Galen’s commitment to (iii) was weak, since he used the Herophilean notion of a neutral bodily state. Growing theoretical interest in chronic conditions may have undermined the traditional dichotomy.

Both Asclepiades and Galen would be sceptical of attempts to measure health. The conception of the human body as a composite of parts which may be in different conditions poses a problem for measurement, but it also sits uneasily with Galen’s model of latitude. It was not Asclepiades’ view that oversimplified, but rather Galen’s view of health states as approximations to the ideal humoural blend, εὐκρασία. Being healthy is rather, as Asclepiades might have put it, like being a triangle. The particular instances are infinitely varied yet cannot be reduced to a simple linear scale.

2. ARISTOTLE

Like the Hippocratics, Aristotle thought that some people are healthier than othersFootnote 9 and that health is or depends on a proportionate balance of basic qualities or humours.Footnote 10 His work on health and disease is lost and the extant works do not present a systematic theory of health.Footnote 11 Nevertheless, three points of discussion in the Categories and the Physics are crucial for understanding later developments: (1) whether there is an intermediate or neutral condition between health and sickness; (2) whether comparisons in respect of a quality such as ‘A is healthier than B’ imply that that the quality in question admits of degrees; (3) whether there is a qualitative continuum between health and sickness.

On the first issue, Aristotle says that there is no intermediate condition between health and sickness.Footnote 12 By contrast, there are intermediates between good and bad, and between just and unjust, which are neither good nor bad and neither just nor unjust.Footnote 13 As we have seen, an implicit notion that health and illness are dichotomous can be traced back to the mid fifth century b.c.e. Aristotle’s claim made this earlier conception explicit, in a way that complicated any attempt to think of recovering health as gradual progression along a scale from sickness to health. At what point does the transition from one category to the other occur? One could punctuate the scale with a clear division between degrees of health and degrees of illness, but this is bound to seem arbitrary. Although the problem of the boundary is likely to strike us as a problem of vagueness, it does not seem to have been connected to the bald man (φαλακρός) paradox (if a man loses hairs one by one, when does he become bald?) or the heap (σωρείτης) paradox (if grains are piled up one by one, when do they form a heap?) in antiquity.Footnote 14 The ancient heap and bald man paradoxes were understood as individual paradoxes involving discrete counting, not puzzles about vagueness in general. In any case, we cannot be sure of how Aristotle would have approached these paradoxes.Footnote 15

On the second issue, Aristotle says that qualities admit of more and less (Categories 8, 10b26), ‘for one thing is called more pale or less pale than another, and more just than another’, but there are exceptions.Footnote 16 A clear exception is shape (11a5–9), ‘for things which admit the definition of triangle or circle are all equally triangles or circles, while of things which do not admit it none will be called more that than another.’ The cases of conditions (διαθέσεις) such as justice, health and grammar are more complex (10b30–11a5):

It might be questioned whether one justice is called more a justice than another, and similarly for the other conditions. For some people dispute about such cases. They utterly deny that one justice is called more or less a justice than another, or one health more or less a health, though they say that one person has health less than another, justice less than another, and similarly with grammar and the other conditions. At any rate things spoken of in virtue of these unquestionably admit of a more and a less: one man is called more grammatical than another, juster, healthier, and so on.

Aristotle’s own view is unclear in this passage. Does he agree that there are no degrees of justice, health and grammar? Or does he distance himself from this view by attributing it to ‘some people’? In their commentaries, Porphyry and Simplicius attributed to Aristotle the view that conditions themselves do not admit of the more and the less, while the things which are qualified by conditions admit of remission and intensification.Footnote 17 The second-century c.e. Peripatetic Aspasius wrote in his commentary on the Nicomachean Ethics that, according to Aristotle, virtue does not admit of degrees: if it did, then virtue, a mean, would also have a mean, an excess, and a deficiency, which is impossible.Footnote 18 He explained that when Aristotle used the comparatives ‘more liberal’ (ἐλευθεριώτερος) and ‘more courageous’ (ἀνδρειότερος), he meant only to contrast those who are liberal or courageous with those who absolutely are not, and that he emphatically did not mean that these virtues admit of degrees.Footnote 19

The evidence internal to Aristotle’s writings is not definitive. There are passages which suggest that virtues do not admit of degrees. In Nicomachean Ethics 2.6, Aristotle says that there is only one point at which one can hit the mark in virtue, while there are many ways of missing it (1106b28–33). In the same chapter, he also says that virtue is a mean and that there are no extremes in the mean (1107a18–27).Footnote 20 On the other hand, one might expect Aristotle not to deny that slight deviations from the mean of justice are justice, but rather to count such attributes as justice in a secondary sense. That is in effect what Aspasius argues. He claims that only imperfect (ἀτελεῖς) or natural (φυσικαί) virtues come by degrees, not virtue in the strict sense. Justice proper does not admit of degrees, but imperfect or natural justice does. The argument that virtue as a mean cannot admit of degrees would presumably transfer to the case of health, which is or depends on a due proportion (συμμετρία), and perhaps a similar account of imperfect health could be constructed.Footnote 21

A third point of importance for later conceptions of the spectrum of health was made in Aristotle’s discussions of qualitative change or alteration in the Physics. Briefly, Aristotle argues that every change, including those in respect of quality, is continuous and in principle measurable.Footnote 22 Just as we can compare two lines and say one is twice as long as another, so we can, in principle, compare the ‘amount’ of alterations of the same kind. For instance, Aristotle says that twice as much alteration takes twice as much time, and that in half the time there will be half as much alteration.Footnote 23

But Aristotle does not think that a continuous alteration necessarily has a ‘path’ in a qualitative continuum, just as a motion occurs along a continuous spatial path.Footnote 24 Rather, he says that every change is continuous because magnitude is continuous, and by this he seems to mean spatial magnitude.Footnote 25 In several passages, he proposes a model of qualitative change as consisting in the gradual spread of a quality through an object, thereby making the continuity of qualitative change similarly dependent on the continuity of spatial magnitude.Footnote 26 The spreading model of alterations is not applicable to all alterations. For instance, Aristotle thinks a whole body of water freezes all at once (ἀθρόον), without gradual progression through intermediate states between liquid and solid.Footnote 27 Solidity does not spread gradually over the lake’s surface. It is also difficult to see how the spreading model could apply to changes like becoming musical.Footnote 28

There is a final complexity to consider. Physics 7.3 argues that every bodily or psychic state (ἕξις)—health is one of Aristotle’s examples—is either a perfection (τελείωσις) or a deficiency of human nature, and neither the achievement nor the loss of such a perfection is an alteration.Footnote 29 At first glance, this suggests that the loss or achievement of health is not an alteration after all. This is puzzling, because becoming healthy is one of Aristotle’s standard examples of alteration. This tension may be mitigated by a distinction in the Categories between two ways in which one may be healthy. Aristotle says that health and sickness are conditions (διαθέσεις) and not states (ἕξεις), but also that they might count as states if over time they become ingrained and difficult to change.Footnote 30 Perhaps achieving a condition of health is an alteration, while the acquisition of a more stable state of health is a perfection and hence not an alteration.Footnote 31 Although Aristotle does not discuss whether the achievement of a perfection is continuous, he does argue that it essentially involves the alteration of affective qualities. In the case of health, a mixture and due proportion of hot and cold must be achieved.Footnote 32 Thus one might say that the achievement of the perfection of health is continuous in a derivative sense, since the alterations of the affective qualities involved are continuous.

The Categories and the Physics provide the essential background to later developments, introducing the first clear conception of a spectrum between the contrary poles of health and illness. The claim that there is no intermediate state between health and sickness accentuated the problem of identifying where health ends and sickness begins. Any sharp division will be arbitrary. And although falling sick and recovering are continuous, this does not imply that the spectrum of health is continuous rather than discrete. Later Peripatetics such as Aspasius argued, with reference to the mean of virtue, that a mean cannot admit of more and less. A parallel argument could be made for health, which would distinguish perfect health as the mean from imperfect health as a range of deviations. Although the means of health and virtue and the perfections of Physics 7.3 represent improved versions of desirable states, Aristotle does not assimilate them. One reason may be that it is implausible that a precise proportion or mean of affective qualities can remain stable in a human body.

3. ASCLEPIADES

Returning to the text with which we began, Caelius Aurelianus reports that Erasistratus (c. 315–240 b.c.e.) and Asclepiades of Bithynia denied that health varies:

Is health uniform, or varying since it increases and diminishes? According to Asclepiades and Erasistratus it is uniform, for they deny that there is a peak and latitude of health. According to Herophilus and the Methodists it is varying, since it increases and diminishes. We also agree with this, for if we at one time feel strong and at another time feel tired, health evidently varies.Footnote 33

This seems to be related to a debate that Galen refers to in his De sanitate tuenda:

I am overcome by amazement at the opinion of men who think that both health and due mixture (εὐκρασία) are without latitude and one (ἀπλατῆ τε καὶ μίαν). And if something is contrary to this, they say it is not health. For they do not notice that they are introducing perpetual suffering (ἀειπάθεια) in making their account about something that has either never occurred in the body of an animal or not persisted for the least [time], if it ever occurred.Footnote 34

The view that health does not admit of more and less has puzzled commentators. It would of course be possible to reply to Caelius that a tired person’s health has not been affected, even if they are more susceptible to illness, but it is not easy to see what could motivate opposition to the claim that some healthy people are healthier than others. Caelius’ testimony is so perplexing that its reliability might be questioned. Yet in the case of Asclepiades, further evidence from Galen suggests some contexts in which he might have argued that health is uniform and without latitude. It is possible that Erasistratus has been assigned to a doxographical category on the basis of perceived similarities to Asclepiades rather than on anything he explicitly asserted.Footnote 35 For now, let us set him aside and focus on the Galenic evidence for Asclepiades’ view on health and disease. This will require careful unravelling of Galen’s distortions. In particular, Galen suggests that Asclepiades unwittingly committed himself to the absurd claim that everyone is ill, the so-called ‘doctrine of perpetual suffering’. Galen reaches this conclusion, however, only by arguing from the assumptions of his own physiological theory. On Asclepiades’ very different corpuscular theory, where the key determinant of health is the flow through the body of particles that have no affective qualities, the doctrine of perpetual suffering does not follow.

Galen’s De experientia medica stages a dialogue between an Empiricist and a Rationalist whose speeches, we are told, are based on the arguments of Asclepiades.Footnote 36 One of the Rationalist’s main arguments concerns the diversity of individual cases within the same disease category.Footnote 37 The Rationalist can unify cases of an illness, despite their many differences, by referring to a theory of the underlying, unobservable causes that they have in common. Since the Empiricists lack such an account, it is argued, they have no principled way to unify cases of an illness. For them, every disease category is fractured into an infinity of possible variations.

In making this argument, the Rationalist interlocutor draws an analogy to triangles.Footnote 38 Triangles are infinite in variety and so cannot be comprehended by the memory of particular instances alone. Nevertheless, general understanding of triangles is possible through the use of reason.Footnote 39 For instance, their common properties can be explained by the single, unvarying account or definition of the triangle.Footnote 40 Crucially, despite their infinite variety, no triangle is more or less of a triangle than any other.Footnote 41 Perhaps there is analogously a single, unvarying account or definition of each disease which explains the common properties of individual cases, although these cases may differ in all kinds of accidental respects. While phrenetics may suffer to a greater or lesser degree, and may be easier or more difficult to cure, none of them is any more or less phrenetic than another.Footnote 42 The Asclepiadean conception of disease has two sides: on the one hand, the endless diversity of individual cases, on the other, the uniformity of what makes them instances of the type of disease that they are.

Asclepiades may have argued for the uniformity of health in a similar way. Health has an exact definition which an individual either satisfies or fails to satisfy. Healthy people vary in all sorts of ways, but not in their being healthy.Footnote 43 In fact, none of them is any more or less healthy. Asclepiades’ doctrine may therefore have been less eccentric than the testimonia suggest. It did not mean that all healthy people are the same. They are as varied as all the triangles, yet like them they have a unifying definition. And it was precisely the variety of particular instances that gave Asclepiades leverage over the Empiricists.

The Empiricist in Exp. med. does not directly challenge the claim that a uniform theory would be required to give principled unity to diagnostic categories. Instead, he responds with incredulity, claiming that none of the Rationalists can in practice relate the multiplicity of phrenitis to a single definition.Footnote 44 Thus the Rationalist interlocutor’s position is no better off than the Empiricist’s. Later in the text, Galen’s Empiricist criticizes Erasistratus for claiming that ‘the logos discovered the phrenetic to be uniform’:

When he speaks thus, we answer and say: what kind of logos is this? The logos of Erasistratos who operates with three kinds of vessels, or the logos of Asclepiades who operates with atoms and pores? Further, the circumstance that (the case of) the phrenetic is not found to be uniform is evident to everybody, in view of the fact that the mixtures and humours which are in our bodies are four or three or innumerable.Footnote 45

The final sentence is perhaps ironic. The Empiricist does not believe in humours and mixtures, but cannot miss an opportunity to draw attention to the lack of consensus among Rationalists on what the correct logos is.Footnote 46 Presumably the Empiricist could use similar tactics in replying to Asclepiades. However, the content of the logos makes a difference. By the lights of a humoural theory, the Empiricist claims, the case of the phrenetic cannot be uniform.Footnote 47 It may be implied that uniformity would be possible on Asclepiades’ theory of ‘atoms and pores’ (Walzer’s transl.), a reference to the particles (divisible and so not exactly atoms) that Asclepiades claimed flowed through the pores or channels of the body.Footnote 48

A connection between Asclepiades’ physiological theory and the uniformity of health is also suggested by another Galenic text. In De temperamentis 3.3, I.676 K, Galen alludes to an argument that humoural physiology and pathology lead to the ‘doctrine of perpetual suffering’ because, on such a theory, things in contact with each other act on each other, even if this mutual affection is not apparent to perception. Galen’s response is that imbalance at the imperceptible level of the humours and primary qualities is necessary but insufficient for a departure from health. There must also be perceptible impairment of the body’s natural activities. He notes that some took another path, denying that things in contact affect each other in order to avoid the doctrine of perpetual suffering. This is likely a reference to Asclepiades’ theory of corpuscles (ὄγκοι), which have only primary qualities and would not alter when in contact.Footnote 49

Galen’s rhetorical manoeuvres sometimes mask the true nature of his disagreement with Asclepiades. In San. tu. 1, Galen argues that everyone must agree that health is a proportionality (συμμετρία), whatever physiological theories they endorse, and hence that everyone should agree that health has breadth or latitude (πλάτος). Galen suggests that rejecting either claim is a failure to realize the consequences of one’s own theory. Clearly this does not mean Galen’s rivals actually agreed that health is a proportionality and has breadth. Asclepiades held that corpuscles flow freely through the channels in health, while in illness the channels are blocked.Footnote 50 Since the basic parameters are continuous, Galen could argue that this implies that health is a proportionality.Footnote 51 In De morborum differentiis 2.4, Galen insists that, although Asclepiades’ particles are unalterable, the human body made of them is a complex structure, compounded from many things, and so the quality of this structure can become better and worse by degrees.Footnote 52 This argument is not compelling, however.Footnote 53

Galen insinuates in De sanitate tuenda 1 that the view that health is uniform entails that everyone is ill, the ‘doctrine of perpetual suffering’, but this argument should be handled carefully. When Galen makes this claim, here and in other passages, he is arguing from the perspective of his own physiological theory. If the body’s parts are ultimately composed of humours and are constantly in contact with each other and the external environment, they will constantly be affected and hence it will be practically impossible for anyone to attain a state of perfect proportionality for any significant amount of time. If, on the other hand, the body consists of particles that possess shape, motion, and perhaps weight, but no affective qualities, then there is no compelling argument that a state satisfying the exact definition of health must be nearly impossible to achieve and maintain. Galen was well aware of this possibility, as we have seen in De temperamentis 3, but he chose to ignore it in De sanitate tuenda 1 to suit his agenda.Footnote 54

At De sanitate tuenda 1.5, VI.16–17K, Galen proposes to argue for degrees of health independently of assumptions about the elements of physiology. After arguing that there are many different healthy constitutions, he claims that these constitutions must differ either in the common form in virtue of which they are called healthy (τὸ κοινὸν ἐν ἁπάσαις εἶδος, ἀφ’ οὗπερ ὑγεῖαι λέγονται), or in terms of the more and less. The structure of this argument, combining the variety of individual conditions with the stability of the common form, a framework that is not repeated in these terms in Galen’s writings on hygiene, suggests engagement with a position similar to the one I have reconstructed for Asclepiades. But the attempt to meet Asclepiades on his own terms was half-hearted, since the dichotomy Galen poses begs the question.Footnote 55

In summary, Galen associates what he calls the ‘doctrine of perpetual suffering’ with the claims that health is uniform and that illness has a strict definition, and he says that a desire to avoid the doctrine motivated the denial that things in contact affect each other. Although Galen does not explicitly link all these points to Asclepiades, the connections can be made thanks to Caelius’ testimony and the fact that Asclepiades was the leading medical theorist to claim the body was made of elements without affective qualities. Perhaps Asclepiades’ promotion of the category of chronic disease, to be discussed in the next section, handed his opponents further material for developing this line of attack. Yet although Galen tried to argue that Asclepiades’ definitions of health and illness led to the rebarbative doctrine of perpetual suffering, the accusation is unconvincing.

Asclepiades insisted on strict definitions of health and illness and so treated each condition as uniform, and it was his corpuscular physiology that enabled him to take this approach without collapsing into the doctrine of perpetual suffering.Footnote 56 Galen’s Exp. med. shows how this insistence on strict definition could be deployed against the Empiricists. Presumably it could also have been deployed against humoural theorists, charging them with the doctrine of perpetual suffering since, on a humoural theory, almost no one can be healthy in the strict sense. They constructed their theory of health about a fiction, the practically unattainable state of proportionate mixture. The eucratic and euectic person does not in fact exist, and a commonsensical objection along these lines might lie behind Caelius’ nolunt enim summum sanitatis. Finally, it is unsurprising that Caelius and other Methodists rejected the Asclepiadean claim that health is uniform, since this was based on a definition of health in terms of unobservable theoretical entities and the Methodists had no interest in such irrelevant speculations.Footnote 57

4. GALEN

Galen’s account of the latitude of health, as set out in works such as Thrasybulus, De sanitate tuenda, De temperamentis and Ars medica, was not radically innovative but a careful reformulation and elaboration of earlier ideas. Galen’s discussions comprise three notable features, which I discuss by turn: (1) a reconciliation of health’s latitude with its definition as an exact proportion; (2) the delimitation of the border of health and illness by pain or clearly impaired natural activity; (3) the explicit use in the Ars medica of the Herophilean notion that some patients are neither healthy nor ill.

Galen’s theory of the latitude of health has a similar basic structure to Aspasius’ theory of virtue. For Galen, health is an exact proportion. This proportion cannot admit of degrees for the reason given by Aspasius with regard to the mean of virtue. Rather than conclude that health does not admit of degrees, Galen calls the indivisible mean ‘peak condition’ (εὐεξία) and introduces a notion of imperfect, sub-peak health, which admits of degrees. Galen sometimes calls imperfect health ‘healthy δυσκρασία’.Footnote 58 By this manoeuvre, analogous to Aspasius’ distinction between perfect mean virtue and imperfect virtue, Galen converts what might look like an argument for the indivisibility of health, that health is a συμμετρία, into an argument for its latitude. Perfect ‘peak condition’ is surrounded by a penumbra of states of imperfect health.Footnote 59

Galen usually demarcates the limits of this penumbra by the presence of pain or clear impairment of natural activity.Footnote 60 Imperfect balances of humours that have not produced pain or clearly impaired natural activity are not cases of illness, but rather are within the domain of healthy δυσκρασία. As Galen puts it, we are healthy when ‘we are neither distressed nor hindered in the activities of life’.Footnote 61 Weakness of the activities does not always indicate disease, but only when it is contrary to the nature of the individual.Footnote 62 For example, someone who was naturally able to read a book from four cubits away and then became incapable of doing so would be diseased, unless the eyesight had weakened due to old age.Footnote 63 The inability to read from four cubits away, however, is not a disease in people who never had such natural gifts.Footnote 64 Furthermore, not just any discomfort or awkwardness in natural activity is indicative of illness. In the Ars medica, Galen lists various conditions that are contrary to nature, like a biting sensation in the stomach or ‘some other moderate difficulty’ (τις ἕτερος μέτριος πόνος) in defecation, vomiting or urination. People with these conditions are not diseased, ‘so long as they are not hindered in their customary actions’ (ὅταν γε μηδέπω διὰ ταῦτα πρὸς τὰ συνήθη τῶν ἔργων ἐμποδίζωνται).Footnote 65 A brief bit of πόνος such as digestive discomfort or constipation is not enough of an impairment to count as illness.

The absence and presence of pain or suffering was included in some Hippocratic definitions of health and its contrary. Most significantly, the text On Human Nature, which Galen judged to be an authentic work of Hippocrates, claimed that a person enjoys perfect health when the elements are mixed proportionately, but suffers pain when one is in defect or excess and separated.Footnote 66 The Hippocratic author in this passage first contrasts being healthy (ὑγιαίνει) not with being ill (νοσεῖ) but with being in pain (ἀλγέει).Footnote 67 Yet where that author directly connects an imbalanced humoural mixture or the separation of a humour with pain, Galen is clear that imperfect mixtures are not always painful.Footnote 68 The role of natural activity (κατὰ φύσιν ἐνέργεια) in Galen’s definitions of health and illness is probably not an innovation. Although it is absent, for example, in the Hippocratic Corpus and in the definition of πάθος attributed to the Herophileans in the pseudo-Galenic Definitions,Footnote 69 the fact that Galen makes no claim to originality and that similar vocabulary is found in Erotian and Soranus suggests that this may have been a common feature of accounts of health and illness by the early Roman period.Footnote 70

Galen’s approach to the boundary problem is seemingly straightforward and empirical. A person is ill when they have clearly taken a turn for the worse, clearly, that is, to someone familiar with their individual nature and usual way of life.Footnote 71 Galen has little time for the notion of hidden or imperceptible illnesses and ridicules the notion of unobservable ‘theoretical fevers’.Footnote 72 Yet at the same time he claims to be able to detect in a patient’s pulse subtle warnings of the onset of illness.Footnote 73 It is far from clear what difference there was in practice between Galen and his rivals at this point.

The notion of a neutral state between health and disease is not a consistent feature of Galen’s writings, yet it plays a prominent role in the Ars medica, where Galen claims that it was introduced by οἱ περὶ τὸν Ἡρόφιλον and encompassed those who are about to be diseased, those who are recovering and the elderly.Footnote 74 The neutral either participates in both extremes equally or in one more than the other,Footnote 75 and so each of health, neutrality and illness has latitude (πλάτος).Footnote 76 This adoption of the neutral state might seem in tension with the notion that health comes in a wide range of degrees, since the neutral presumably includes part of the spectrum that would otherwise belong to health. For Galen, however, it serves as another viable expression; the insight that there is no sharp dichotomy between two invariable attributes.Footnote 77 Moreover, the neutral would only threaten the domain of health if ‘neutral’ were understood as ‘neither healthy nor sick’, but for Galen in this context the neutral condition that has latitude is neutral in the sense of participating in both extremes and not necessarily participating equally.Footnote 78 So understood, there is no contradiction between being imperfectly healthy and being in the neutral condition.

Galen writes as if the notion of a neutral bodily state had never been controversial and does not explicitly explain his abandonment of the traditional assumption that health and disease are dichotomous.Footnote 79 Given the vagueness of his reports that the neutral bodily state was introduced by certain Herophileans, their motives are likely to prove elusive. The fragmentary evidence does not allow us to pinpoint this development in the history of the Herophilean school from the mid third century b.c.e. and the mid first century c.e. Footnote 80 Nevertheless, it is possible to note that the neutral bodily state was a concept well suited to the expansion of concern for medical conditions other than acute disease. Medical writers of the fifth and fourth centuries tended to prioritize acute disease, and their theorizations of the tumultuous course of disease contrasted sharply with the notion of health as a stable proportion based on controlled regimen. In the last two centuries b.c.e., increased attention was given to chronic disease, especially among the Asclepiadeans and Methodists.Footnote 81 According to Caelius Aurelianus, Asclepiades distinguished chronic and acute forms of various ailments,Footnote 82 while his student Themison of Laodicea, the founding figure of Methodism, was the first to structure a work around the categories of acute and chronic diseases, probably in the late second or early first century b.c.e. Footnote 83 Perhaps also relevant was the development in the Hellenistic period of old age as a special topic in medicine. Galen is the first medical writer whose books specifically addressing old age have survived.Footnote 84 He attributes the term for regimen for the elderly, γηροκομικόν, to the νεώτεροι, a term that could mean anyone writing after the death of Alexander the Great.Footnote 85 Plausibly, with such shifts in the interests and priorities of medical writers, the supposed rupture between health and the disorderly flux of illness became less compelling.

In summary, Galen’s account of health likely draws on a range of post-Classical writers whom he does not name. The structural similarity between Galen’s account of health and Aspasius’ account of virtue is likely not a chance resemblance. Galen tells us that he was taught by a student of Aspasius the Peripatetic when he was fifteen years old, in 144 c.e., and he advises that his own commentary on Aristotle’s Categories should be read only by students with the guidance of a teacher or the commentaries of Adrastus or Aspasius.Footnote 86

5. CONCLUSION

Ancient Greek doctors and philosophers faced two problems in conceptualizing the variability of human health. The notion (i) that degrees of health and illness can be compared stood in tension with (ii) an exact definition of health, and (iii) the assumption that health and illness are dichotomous. The seeds of both problems were sown as early as the mid fifth century, when assumptions (i)–(iii) were expressed by Aeschylus, Alcmaeon and Herodotus. This article has explored two different solutions from the late Hellenistic and Imperial periods. Asclepiades gave up the assumption that health comes by degrees. Galen instead weakened his commitment to (iii) by borrowing the Herophileans’ notion of a bodily state that is neither healthy nor diseased, and meanwhile accommodated (ii) by offering a theory of imperfect health akin to the Peripatetic Aspasius’ account of imperfect virtue.

Asclepiades’ account of health has been overlooked and underestimated, in part thanks to Galen’s misrepresentations. Nevertheless, the inconsistencies in Galen’s criticisms in different contexts provide opportunities to glimpse beyond the fog of polemic. Asclepiades likely argued that health was uniform due to its exact theoretical definition, and he may have deployed this account against the Empiricists and perhaps humoural theorists. Galen’s argument that the doctrine of the uniformity of health leads to the doctrine of perpetual suffering was far from compelling. It assumed a theory of qualitatively altering and affective body parts, which does not match Asclepiades’ corpuscular theory. Galen could point to continuous variables in Asclepiades’ physiology and suggest that this should mean that health is practically unattainable if it is uniform, but such an argument is unconvincing. Although Galen and Caelius depict Asclepiades’ doctrine of the uniformity of health as a bizarre outlier, perhaps joined only by Erasistratus, it was likely not such a radical position. It involved rejecting the common assumption that there are degrees of health, but it also developed the familiar idea that health has an exact physiological definition. From an Asclepiadean perspective, it would presumably be the Herophileans and Galen who were aberrations, with their unprecedented idea that some people are neither healthy nor ill.

Mathematical terminology (equality, proportion, mean, latitude) was a common feature of these discussions, yet the modern questions of whether attributes such as health or intelligence are continuous or discrete, quantitative (and thus in principle measurable) or merely ordinal were not seriously entertained.Footnote 87 Aristotle and Galen sometimes seem to have a geometrical model in mind, where the extremes of health and illness are represented by opposite ends of a line segment with a range of degrees of health and illness in between. However, as we have seen, they were not fully committed to the existence of a scale of health with all the properties of this model. For Aristotle, the fact that falling sick and recovering are continuous alterations does not imply that there is a continuous scale from health to sickness, and Galen explicitly said that the extent of the domain of health cannot be measured.Footnote 88

The notion of a single scale on which all individuals’ health conditions could be compared may not have seemed plausible in the ancient world. Galen’s insistence that the observable activities that constitute health vary from person to person indicated one kind of incommensurability. Yet even outside Galen’s theory, the possibility of quantifying health was ruled out by other considerations. First, the ambiguity of prognostic and diagnostic signs presented a practical obstacle from the Hippocratics onwards.Footnote 89 Secondly, the conception of the human body as a composite of parts which can be in different conditions simultaneously may have suggested that there was no straightforward way to compare different bodies taken as wholes.Footnote 90

This further vindicates Asclepiades’ position. For Asclepiades, people who are not ill are healthy and it is meaningless to say that some of them are healthier than others. Such comparisons are typically based on observations such as that one person has stronger eyesight, or is less frequently tired, or is less susceptible to certain ailments, or that one group has a longer average lifespan. From an Asclepiadean perspective, ‘healthier’ is at best a shorthand for such attributes. It may be convenient for some purposes and contexts, but it is a mistake to think that it indicates, in addition to such attributes, a real difference of degree in health for which these are proxies. Galen straddled on the one hand the model of ‘latitude’ and on the other the notion that health is multidimensional and relativized to individuals whose bodies are intricate, composite wholes. From an Asclepiadean point of view, the notion of latitude was an empty abstraction that oversimplified the complex ways in which healthy people differ. Both sides had a point and the dilemma is still with us today. We think of attributes like health in terms of the more and less, or even as measurable, yet the assumption of a simple ordinal scale is difficult to justify.

Footnotes

*

I am grateful to James Allen, David Leith, Geoffrey Lloyd and Vivian Nutton for comments on earlier versions, to Chiara Thumiger for kindly supplying a copy of an article, and to audiences in Cambridge and Heidelberg for valuable discussion. I thank the Needham Research Institute and Darwin College, Cambridge, where I conducted the research for this article as Lloyd-Dan David Research Fellow. I am grateful for subsequent support through appointments at the University of Toronto, the Center for Hellenic Studies, and University College London.

References

1 De salutaribus praeceptis 10: uniformis est sanitas, an uaria cum extenditur et minuitur?

2 Asclepiades’ floruit was once located in the 50s b.c.e., but E. Rawson, ‘The life and death of Asclepiades of Bithynia’, CQ 32 (1982), 358–70 argued that he was dead by 91 b.c.e. See also R. Flemming, ‘Antiochus and Asclepiades: medical and philosophical sectarianism at the end of the Hellenistic era’, in D. Sedley (ed.), The Philosophy of Antiochus (Cambridge, 2012), 55–79.

3 These assumptions differed in their origins. Assumption (ii) came from medical theory, while (iii) is more basic to the human experience of health and illness, not relying on systematic theory or inquiry. In an intermediate position was (i), where a modicum of inquiry and reflection is involved in making a comparison.

4 Herodotus (2.77, 4.187) judged that the Libyans were the healthiest people, followed by the Egyptians. Among the Hippocratic writers, comparative judgements of health are made in Art. 56; Aph. 6.2; Nat. hom. 4; VM 12.

5 Alcmaeon fr. 4, Hippocratic Nat. hom. 1.4, Vict. 1.2, Aer. 12, Aph. 5.62.

6 Agamemnon 1002–3 νόσος | γείτων ὁμότοιχος.

7 See also Plato, Charmides 170e–171a, Gorgias 495e, Timaeus 82a–b.

8 J. Jouanna (transl. M.B. DeBevoise), Hippocrates (Baltimore, 1999), 328–35 outlines the tension between continuity and rupture in some Hippocratic texts. Among studies of later Greek conceptions of the spectrum of health see N. Almberg, ‘Zu dem Begriff “Breite der Gesundheit” bei Galen’, Theoria 15 (1949), 17–28; H. von Staden, Herophilus: The Art of Medicine in Early Alexandria (Cambridge, 1989), ch. 4; V. Boudon, ‘Les definitions tripartites de la médecine chez Galien’, in W. Haase (ed.), Aufstieg und Niedergang der Römischen Welt 37.2 (Berlin, 1994), 1468–90; S. Grimaudo, Difendere la salute: igiene e disciplina del soggetto nel De sanitate tuenda di Galeno (Naples, 2008), ch. 3; P.N. Singer, ‘The fight for health: tradition, competition, subdivision and philosophy in Galen’s hygienic writings’, BJHP 22 (2014), 974–95; O. Lewis, C. Thumiger and P. van der Eijk, ‘Mental and physical gradualism in Graeco-Roman medicine’, in G. Keil, L. Keuck and R. Hauswald (edd.), Vagueness in Psychiatry (Oxford, 2016), 27–45; P.N. Singer, Galen: Writings on Health (Cambridge, 2023), 27–31.

9 Cat. 8, 10b26–11a5; Eth. Nic. 10.3, 1173a23–25.

10 One of the two extant versions of Ph. 7.3, β, says health is the proportionate mixture of affective qualities (246b21), while version α says health is in the mixture (246b5–6).

11 Aristotle promises to discuss health and disease at Long. 464b32–33, cf. Resp. 480b22–4, Sens. 436a17–18.

12 Cat. 10, 12a4–11.

13 Cat. 10, 12a24–25.

14 This is despite the fact that the sorites played a role in Hellenistic medical debate (Galen, Exp. med. 16, 20).

15 See S. Bobzien, ‘Chrysippus and the epistemic theory of vagueness’, PAS 102 (2002), 217–38 for further commentary on the differences between the ancient sôritês and modern discussions of vagueness. Aspasius and the anonymous commentator on the Nicomachean Ethics discussed the sôritês paradox as a possible objection to the idea of virtue as a mean (J. Moline, ‘Aristotle, Eubulides and the Sorites’, Mind 78 [1969], 393–407). But this use of the sorites involved counting particular actions, and in the case of health there is no obviously analogous thing to be counted.

16 Here and below I quote from Ackrill’s translation.

17 Nevertheless, according to Simplicius ad loc., Aristotle did not fully agree with the people he refers to, since they maintained, Simplicius says, that qualified things admit of degrees of participation in qualities.

18 In Eth. Nic. 50.15–51.1, on which see K. Ierodiakonou, ‘Aspasius on perfect and imperfect virtues’, in A. Alberti and R.W. Sharples (edd.), Aspasius: The Earliest Extant Commentary on Aristotle’s Ethics (Berlin, 1999), 142–61. J. Barnes, ‘An introduction to Aspasius’, ibid., 1–50, dates the commentary to 131 or a little after.

19 In Eth. Nic. 86.17–22, 99.27–100.4.

20 D. Reed, ‘Degrees of virtue in the Nicomachean Ethics’, AncPhil 37 (2017), 91–112 argues that there are degrees of perfect virtue, with heroic virtue as the highest degree.

21 See T. Tracy, Physiological Theory and the Doctrine of the Mean in Plato and Aristotle (Chicago, 1969) on the notion of the mean in medicine and ethics. Also worth mentioning in this connection is the difficult argument against the harmonia (‘attunement’) theory of the soul in Phaedo 93a11–94b3, which raises the issue of whether an attunement can be more or less tuned. The relevance to discussions of health would probably be clear. Note the equivalence of harmonia and krasis at Phaedo 86d2, and that Aristotle (De anima 1.4, 408a1–3), after making the objection that it is hard to see how the soul could move the body on the harmonia theory, adds that it would be more appropriate to apply harmonia to health or one of the other bodily excellences. For discussions of the Phaedo argument see D. Gallop, Plato: Phaedo (Oxford, 1975), 157–66 and C.C.W. Taylor, ‘The arguments in the Phaedo concerning the thesis that the Soul is a harmonia’, in J.P. Anton and A. Preus (edd.), Essays in Ancient Greek Philosophy, vol. 2 (Albany, NY, 1983), 217–31.

22 Despite the differences between Aristotle’s account of continuity, on which the continuum itself is fundamental and is not composed from an infinite of actual elements, and the modern definition of continuity by Dedekind cuts, on both accounts a continuum is quantitative in the sense that distances can be defined on it. Aristotle’s idea is roughly that something is continuous if it is potentially infinitely divisible into divisible parts without gaps.

23 Ph. 7.5, 250a31–b2. He also says that in comparing alterations one cannot talk of equality but only of similarity, however (Ph. 7.4, 249b2–3; cf. Gen. corr. 2.6, 333a29–30); cf. Cael. 1.7, 274b34–275a13.

24 The fact that becoming healthy is continuous does not necessarily imply that there is a continuum containing a potential infinity of conditions between health and illness, on which not only ordinal relations (‘A is healthier than B’) but quantitative ones (‘A is healthier than B by the same amount that B is healthier than C’) could be defined. For the connection between continuity and measurability, see Ph. 4.12, 220b24–32 and E. Hussey, Aristotle’s Physics Books III and IV (Oxford, 1983), ad loc.

25 Ph. 4.11, 219a10–21, 220b24–32; cf. 3.7, 207b21–4.

26 See especially Ph. 4.9, 240a23–6. At Gen. corr. 2.10, 337a25–33, Aristotle states that change is continuous because the changing thing is continuous, not because that in which it changes is continuous. Cf. Ph. 7.4, 249b11–19.

27 Sens. 6, 447a1–3. Ph. 8.3, 253b14–30 gives the example of plants or dripping water which wear down a stone for a long time before a whole mass of the stone suddenly collapses, an ‘all at once’ change in quantity rather than quality; cf. Ph. 1.3, 186a15–16.

28 Example from U. Coope, Time for Aristotle: Physics IV.10–14 (Oxford, 2005), 53. One reason for favouring the spreading model may have been that, if every alteration were continuous because it involved passage through a qualitative continuum, this would have the consequence that something changing from one qualitative state to another must pass through all the intermediate states.

29 Aristotle also says the perfection itself is not an alteration.

30 Cat. 9a8–10.

31 This suggestion is due to U. Coope, ‘Commentary on 246a10–246b3’, in S. Maso, C. Natali and G. Seel (edd.), Reading Aristotle Physics VII.3 (Las Vegas, 2012).

32 Elsewhere Aristotle speaks of health in different terms: bile and phlegm (Ph. 2.2, 194a23–24), moisture (Ph. 3.1, 201a34–b3) and blood (Hist. an. 2.19, 520b29–31).

33 De salutaribus praeceptis 10: uniformis est sanitas, an uaria cum extenditur et minuitur? secundum Asclepiadem et Erasistratum uniformis: nolunt enim summum sanitatis et †eanditudinem† esse. secundum Herophilum et methodicos uaria, cum extenditur et minuitur. id etiam nos adserimus. si enim modo nos fortes modo fatigatos esse sentimus, manifeste uariatur sanitas. There are several textual issues, including whether to read nolunt or uolunt, and what lies behind the corrupt eanditudinem. On these questions, see I. Garofalo, Erasistrati fragmenta (Pisa, 1988), 115 (fr. 154); J.T. Vallance, The Lost Theory of Asclepiades of Bithynia (Oxford, 1990), 124–5; M. Tecusan, The Fragments of the Methodists, vol. 1 (Leiden, 2004), 202 (fr. 49). More generally on the De salutaribus praeceptis, a fragment of Caelius’ Medicinales responsiones, see V. Rose, Anecdota graeca et graecolatina: Mitteilungen aus Handschriften zur Geschichte der griechischen Wissenschaft (Berlin, 1864/70); A. Roselli, ‘Le Medicinales responsiones di Celio Aureliano’, in G. Sabbah (ed.), Le latin médical: la constitution d’un langage scientifique (Saint-Étienne, 1991), 75–86; G. Flammini, ‘Celio Aureliano e le prefazioni ai Gynaecia e ai frammenti delle Medicinales responsiones’, in C. Santini, N. Scivoletto and L. Zurli (edd.), Prefazioni, prologhi, proemi di opere tecnico-scientifiche latine (Rome, 1998), 146–76; K.-D. Fischer, ‘Bisher unberücksichtigte Handschriftenfunde zur Überlieferung der Werke des Caelius Aurelianus’, in P. Mudry (ed.), Le traité des Maladies aiguës et des Maladies chroniques de Caelius Aurelianus: Nouvelles approches (Nantes, 1999), 141–72.

34 1.5, VI.28.14–29.3K; cf. VI.14.1–5K.

35 Vallance (n. 33), 123–30 discusses some apparent similarities. For instance, both held that many illnesses are due to blockage of the body’s channels. D. Leith, ‘Erasistratus’ triplokia of arteries, veins and nerves’, Apeiron 48 (2015), 251–62 has shown that ancient doxographers sometimes misleadingly conflated the two. See also W.D. Smith, ‘Erasistratus’ dietetic medicine’, BHM 56 (1982), 398–409 on Galen’s misrepresentations of Erasistratus. For Erasistratus’ definition of health as εὐταξία and αὐτάρκεια see fr. 168 Garofalo (n. 33).

36 Exp. med. 2.

37 Exp. med. 3.

38 ‘Rationalist’ was a blanket term for anyone who insisted that therapeutics should be based on an understanding of the causes of disease. On this point see H. von Staden, ‘Hairesis and heresy: the case of the haireseis iatrikai’, in B.F. Meyer and E.P. Sanders (edd.), Jewish and Christian Self-Definition (London, 1982), 76–100.

39 Exp. med. 3 asserts that ‘it was discovered by means of the logos that the sides of the triangle are three kinds in all’, meaning that every triangle is either equilateral, isosceles or scalene.

40 Euclid defines a triangle as a plane figure contained by three straight lines (Elements 1, def. 19). Then either all three lines are equal, or only two are, or none are. Exp. med. 3 also mentions another general feature of triangles, that every internal angle is either right, acute or obtuse. This intuitive classification can be proven from Elements 1, prop. 32.

41 A point we saw earlier in Arist. Cat. 8, 11a5–9.

42 Similarly, fevers may be intense or easily resolved (Caelius, Morb. ac. 1.praef.8, I.14.107), but neither case is more or less of a fever than the other. For Asclepiades distinguishing different forms of phrenitis see Morb. ac. 1.15, with C. Thumiger, ‘Asclepiades on phrenitis and the problem of consistency (Caelius Aurelianus, Ac.Dis. 1.1, 14–15)’, Technai 10 (2019), 23–43.

43 Galen insinuates that there was an Asclepiadean definition of health as a proportion of pores, or onkoi and pores, or anarma (Meth. med. 4.4, X.267–268K and San. tu. 1.5, VI.15K). But as Singer (n. 8 [2014], 978–9; [2023], 164) emphasizes, Galen had an interest in arguing that everyone agreed that health is a proportion of some kind. See also Vallance (n. 33), 142 and Galen’s similar strategy in Meth. med. 1.5, X.39–42K. Caelius Aurelianus, De salutaribus praeceptis 1 defines health as temperies, which Roselli (n. 33) suggests is a Latin rendering of εὐκρασία.

44 Exp. med. 22–3. Phrenitis was mentioned in the Rationalist’s speech in a different context at Exp. med. 5; it is a recurring example through 22–8. The extraordinary variety of symptoms encompassed by phrenitis made it a useful example. Compare Meth. med. 2.7, X.140–5, 149–54K where phrenitis is Galen’s example in discussing the relation between general disease concepts and individual cases. For further discussion on phrenitis as an exemplary disease in Galen see C. Thumiger, Phrenitis and the Pathology of the Mind in Western Medical Thought (Cambridge, 2023), ch. 4.

45 Exp. med. 27, transl. Walzer. The term logos in Walzer’s translation is sometimes used of the faculty of reason (Exp. med. 1) and sometimes for a theory or account (Exp. med. 9).

46 According to Anonymus Parisinus, De morbis acutis et chroniis 1.1.1, Erasistratus maintained that phrenitis was an affection of the meninx in the brain. Asclepiades’ own account of phrenitis was apparently not so very different. See Exp. med. 25 and 28 and Caelius Aurelianus Morb. ac. 1.1, 14–15, 20 with Thumiger (n. 44).

47 The reason is not explicit, but we shall see a possible explanation shortly.

48 On Asclepiades’ theory of ὄγκοι and πόροι see D. Leith, ‘The qualitative status of the onkoi in Asclepiades’ theory of matter’, OSAPh 36 (2009), 283–320 and id., ‘Pores and void in Asclepiades’ physical theory’, Phronesis 57 (2012), 164–91. See also Thumiger’s (n. 44), 27 notes on Walzer’s translation of the Arabic.

49 See the summary of Asclepiades’ view at Caelius Aurelianus, Morb. ac. I.14.105–15 with Leith (n. 48 [2009]). One could take Galen’s reference to certain things (ἔνια) that are unaffectable as an allusion to a specific material thought to be unaffectable, e.g. diamond (Galen, Hipp. elem. 2.34, I.422K; Hero, Pneumatica 1.praef.58), but this would not avoid the doctrine of perpetual suffering unless human bodies were made of diamond.

50 See e.g. Celsus, De medicina 1.praef.14–15. That there are membranes in the spine in quibus cum fuerit obtrusio non ingerit phrenitim (Caelius Aurelianus, Morb. ac. 1.praef.7) does not necessarily mean that obtrusio there is healthy.

51 Cf. Morb. diff. 2.3.

52 VI.839–40K.

53 It could be, for instance, that what matters is whether there is blockage somewhere in the body or not, and that what Galen calls the overall quality of the structure is irrelevant.

54 The Asclepiadeans’ strict definitions of health and illness in theoretical terms may be the target of Galen’s criticism in other contexts. In De diebus decretoriis, Galen writes that a fever is reckoned to begin when there is an observable manifestation. ‘If you wish to consider the causes theorized by logos, you will quickly reach the doctrine of perpetual suffering. But the medicine of the Asclepiads is not of this kind, since they do not think it worth diagnosing or treating fevers theorized by logos’ (IX.798K). The opponents’ logos commits them to unobservable, ‘theorized’ fevers. Galen’s own side, the Asclepiads, hold that these cannot be diagnosed or treated precisely because they are unobservable. However, Galen himself used pulse diagnosis to detect highly theorized conditions of the onset of disease and his disagreement with Asclepiades, if Asclepiades is the target in this passage, is likely exaggerated. Asclepiades also appears to have discussed signs of impending illness (e.g. Caelius, Morb. ac. 1.1.24–6). At most perhaps they differed over whether to call such patients ‘about to be diseased’ or ‘diseased’. Cf. San. tu. 1.5, VI.19K.

55 Asclepiades might say that, like triangles, the various healthy constitutions differ in their per se properties but not in their specific properties.

56 There is circumstantial evidence that Asclepiades sometimes used a relatively ontological notion of disease, which might bolster to his doctrine of uniformity. See Plutarch, Quaest. conv. 8.9 with V. Nutton, ‘The seeds of disease: an explanation of contagion and infection from the Greeks to the Renaissance’, MedHist 27 (1983), 1–34 and A. Harris, ‘New diseases and sectarian debate in Hellenistic and Roman medicine’, Apeiron 55 (2022), 167–91. On ‘ontological’ disease concepts see O. Temkin, ‘The scientific approach to disease: specific entity and individual illness’, in A.C. Crombie (ed.), Scientific Change (London, 1963), 629–47.

57 I have not found specific evidence to support the obvious hypothesis that the doctrine of the uniformity of health may have been influenced by the Stoic view that virtue does not come by degrees (Diog. Laert. 7.127, Plut. Comm. not. 1063A–B, Sen. Ep. 118.15–17). One difference is that the Stoics claimed that in practice everyone falls short of virtue, while the physicians who claimed that health is an all-or-nothing affair apparently assumed that it is easy to obtain.

58 E.g. Temp. 2.4, I.609.8–12K.

59 Thras. 6–7, I.814–16K further distinguishes stable (καθ’ ἕξιν) and unstable (κατὰ σχέσιν) forms of imperfect health, roughly comparable to the distinction found in Arist. Cat. 9a8–10. Galen’s example of unstable imperfect health is someone who has recently recovered from a severe illness, who needs building up to return to his normal activities. His natural activities are no longer impaired, but they are still weak.

60 San. tu. 1.5, I.18–19K; Thras. 7, I.814–17K; Ars med. 27, I.381K; Temp. II.4. See Grimaudo (n. 8), ch. 4 and D. King, Experiencing Pain in Imperial Greek Culture (Oxford, 2017), ch. 3 for further discussions of the relation of pain and disease in Galen. Sometimes Galen lists additional markers of disease. Ars med. argues that the internal parts must be diagnosed by impaired activity, excretions, pains, unnatural swellings, some of these, or all of them (20, I.355K), suggesting that neither impairment of natural activity nor pain is strictly necessary, and that Galen in Thras. and San. tu. simplifies the demarcation. On Galen’s notion of natural activity see Grimaudo (n. 8), ch. 2, Singer (n. 8 [2014]) and M.J. Schiefsky, ‘Galen’s teleology and functional explanation’, OSAPh 33 (2007), 369–400.

61 San. tu. 1.5, VI.18K.

62 San. tu. 1.5, VI.19K.

63 See San. tu. 1.5, VI.20–21K for Galen’s argument that old age is not a disease (in contrast to e.g. Cicero, De senectute 11.35) because ageing is a natural process. San. tu. 5.4, VI.330–32K, however, argues that issue of whether old age is a disease, neutral condition or unstable health is merely verbal.

64 San. tu. 1.5, VI.20K.

65 Ars med. 21, I.362K.

66 Nat. hom. 4.

67 The author may still be thinking of Melissus (30 B7 DK) who is criticized in Nat. hom. 1.

68 See especially Nat. hom. 4, VI.40.6–9L where the author claims the separation of a humour necessarily leads to pain and suffering. Galen’s commentary (HNH XV.60–62K) interprets Nat. hom.’s καὶ μάλιστα ἢν μεμιγμένα ᾖ as picking out the perfect, eucratic state. See J. Jouanna, ‘Galen’s reading of the Hippocratic treatise The Nature of Man: the foundations of Hippocratism in Galen’, in J. Jouanna (transl. N. Allies, ed. P. van der Eijk), Greek Medicine from Hippocrates to Galen: Selected Papers (Leiden, 2012), 313–33 on Galen’s reading of Nat. hom. Pain in this passage of Nat. hom. might include imperceptible pain, which B. Holmes (The Emergence of the Physical Body in Ancient Greece [Princeton and Oxford 2010], 108–20) calls ‘bodily sensing’ in contrast to conscious sensing.

69 134, XIX.386–87K.

70 Erotian under ἔμπηροι notes ‘Bacchius says πηρόν means ‘blind’, but it can be taken to mean anyone hindered in their natural activities’. Sor. Gyn. 2.28 writes that if a disease has caused a mother’s milk to stop, then it should be removed ‘for when the disease is removed the impediment to activity is also removed (καὶ ὁ παραποδισμὸς λύεται τῆς ἐνεργείας)’.

71 More generally, Galen defines health at the physiological level as συμμετρία and at a higher, more observable level as natural activity. Although in some contexts Galen says it makes no difference which of these two aspects is taken as fundamental and definitive (Thras. 11, V.822–823K), his considered view seems to be that health is natural activity (San. tu. 1.5, VI.19K; cf. I.4 VI.12K).

72 See references in n. 54.

73 The sign of an approaching paroxysm might be extremely difficult to perceive. Galen tells us he correctly noted the absence of such a sign in the emperor’s pulse when two other doctors thought it was present (Praen. XIV.658–9K). Prognostication of the onset of illness was also important for some Hippocratics. The author of Vict. 1.2 even claimed that this was his own discovery.

74 Ars med. 21, I.358K and San. tu. 6.2, VI.388K. The testimonia (frr. 42–8 von Staden) suggest that Herophilus of Chalcedon (c. 330–260 b.c.e.) introduced the notion of the ‘neutral’ primarily in his division of medicine into three parts, one of which was ‘knowledge of the neutrals’, which may have covered the study of aids applied in disease such as drugs, surgery and therapeutic dietetics: von Staden (n. 8), 89–114; cf. Singer (n. 8 [2023]), 404–5. In Ars med. 1, I.308K, Galen not only distinguished the healthy, neutral and diseased, but also distinguished between these categories as applied to bodies, signs and causes, and between what is neutral ‘generally’ (ἁπλῶς) and ‘at the present moment’ (ἐν τῷ νῦν). No sources attribute these finer distinctions to Herophilus himself and it is likely that the explicit framework of bodies, signs and causes was developed by later Herophileans (von Staden [n. 8], 103–8). Neutral signs and therapies are presumably those which are productive or indicative of health in some cases and not in others. It is only the notion of a neutral bodily state ‘at the present moment’ which contradicts Aristotle’s claim that health and illness are dichotomous. Galen is clearly operating with such a notion. ‘Generally’ neutral bodily states would presumably be humoural mixtures that are healthy in some circumstances and not others (Boudon [n. 8], 1489–90).

75 Ars med. 1, I.309K.

76 Ars med. 4, I.316K.

77 Compare Singer (n. 8 [2023]), 28–9.

78 See above on the distinction of three senses of ‘neutral’.

79 This is despite his familiarity with Aristotle’s Categories, on which he wrote a commentary (Lib. prop. 19), now lost.

80 See von Staden (n. 8), 457–62 on the final stages of the school.

81 While a strong notion of acute disease features in several Hippocratic texts (e.g. Aff. 6–11, Prog. 2, Acut. passim), there are fewer references to chronic disease (Art. 41, 49; Aph. 2.39, Gland. 14; Morb. 1.3 on ‘long’ [μακρά] ailments).

82 E.g. Morb. chron. 3.8.100 on acute and chronic kinds of dropsy and other special diseases. Asclepiades wrote a book on acute diseases (Morb. ac. 2.1.6).

83 Morb. chron. 1.praef.3. As with Asclepiades, Themison’s dates are uncertain. The key sources for their relationship are Celsus, De medicina 1.praef.12, Caelius Aurelianus Morb. chron. 2.215, Pliny, HN 29.5.6. The acute-chronic distinction provided the structure for several nosographic handbooks of the Roman period such as those of Aretaeus, the Anonymus Parisinus and Soranus. See also Celsus’ reservations about a simple acute-chronic distinction at De medicina 3.1 and the more complex set of categories in the pseudo-Galenic Definitions 138–42, XIX.388–99K.

84 See especially the theoretical discussion of De marcore and more practical advice on regimen for the elderly in San. tu. 5. As we have seen, after attributing the notion of neutral bodily condition to the Herophileans, Galen says that the elderly are neutral, although he also thinks that an elderly person can be called healthy for old age if their capacities for seeing, hearing, walking and so on are adequate for their needs (San. tu. 6.2, 338K). Temp. 2.2 states that old age is cold and dry but notes that some doctors think it is cold and wet (perhaps tempting given the common association of old age with winter). In the first century b.c.e., Octavian’s Asclepiadean physician, M. Artorius Asclepiades, wrote a work on longevity (Clement of Alexandria, Paedagogus 2.23). Old age and chronic conditions were already connected in Aphorisms 2.39; cf. the ailments of old age at Aphorisms 3.31 and at Coac. 30.502.

85 San. tu. 5.4, 6.330K. Galen explains that he means writers after Alexander at HNH II.44, XV.105K. Note Galen’s comments at San. tu. 5.7–8, VI.348–9K and 5.9, VI.356K that comprehensive discussion of therapeutics does not belong to γηροκομικόν.

86 Aff. pecc. dig. 8, V.42K; Lib. prop. XIX.42K.

87 On an ordinal scale, relations of relative rank (‘A is healthier than B’) can be defined. On a quantitative scale, distances can also be defined (e.g. ‘A is healthier than B by the same amount that B is healthier than C’).

88 Temp. 2.4, I.609.10–11K ὁρίσαι δ’ οὐδ’ ἐνταῦθα μέτρῳ καὶ σταθμῷ τὸ ποσὸν ἐγχωρεῖ. Presumably Asclepiades would have agreed that health cannot be measured, since on his view it has no latitude. Caelius’ De salutaribus praeceptis is corrupt at a key point (nolunt enim summum sanitatis et †eanditudinem† esse), but the conjectures altitudinem, quantitudinem, magnitudinem and latitudinem suggest that Asclepiades might have explicitly expressed scepticism about quantifying health.

89 Relevant here is the limited role in antiquity for quantitative measurement in examining patients.

90 Similar considerations continue to shape discussions of health. For example, D.M. Hausman, Valuing Health: Well-Being, Freedom, and Suffering (Oxford, 2015), ch. 4 argues that health cannot be measured because it is multi-dimensional, which is to say that there is no way to derive the overall functional efficiency of an organism’s many parts and processes. The question of how to draw the boundary between health and sickness is also as alive today as it ever has been. See for example A.S. Schroeder, ‘Rethinking health: healthy or healthier than?’, British Journal for the Philosophy of Science 64 (2013), 131–59; W.A. Rogers and M.J. Walker, ‘The line-drawing problem in disease definition’, Journal of Medicine and Philosophy 42 (2017), 405–23.