This was once revealed to me in a dream.
Nikolai BerdyaevFootnote 1What epistemic status did Galen of Pergamum (c. 129 – c. 215 c.e.) afford dreams, and what can this tell us about the relation between rationalism and religiosity in his thought? This is a perennial question, which has regularly exercised Galen’s scholarly readers. And for good reason. Galen derides the use of charms, prayers and divination in medicine while avowing himself a devotee of Asclepius. He mocks those who disallow reason in conjunction with experience but happily advertises his own cures learned from dreams. Dreams appear an unstable, even self-contradictory category for Galen the physician and philosopher, indicating other, countervailing commitments.Footnote 2
Scholars have regularly depicted dreams as demarking a frontier between ‘rational’ and ‘irrational’, between the philosopher and the devotee, between naturalistic explanation and superstitious enthusiasm. But, I propose, this is not so. The connection between Galen’s attitude toward dreams and his religious devotion is much less direct than we might expect and their epistemic status both more contextual and less apparent. In this article I take the role of dreams in medical diagnosis, therapy and autobiography in turn, to build up a coherent picture of Galen on dreams which resituates the oneiric within the rational without devaluing the significance that Galen does sometimes afford dreams. By shifting the conversation, I propose to ask a new question: what do dreams do for Galen? Animated by this question, this article concludes that Galen is uninterested in the origins of dreams and his infrequent reliance on them says little or nothing about his religiosity, but instead he is impressed by the vividness (ἐνάργεια) of some dreams. Vivacity invokes philosophical traditions of epistemological criteria and rhetorical discourses of persuasion, which Galen subtly leverages to make vivid dreams a confirmation of experience, a source of surprise and a technique of self-fashioning.
THE STORY SO FAR
A full literature review being prohibitive, I focus on how several prominent studies have made dreams a transition point between binaries of ‘rational’ and ‘irrational,’ or philosophical and religious, in Galen, to highlight the perceived epistemic instability of dreams.
On one side, dreams had a venerable, scientifically respectable, role in diagnosis. Oberhelman, who has written extensively on dreams in ancient medicine, argues regularly that Galen followed suit, as can be summed up thus: ‘Dreams are fully incorporated into Galen’s medical science and play an active role in his treatment of illnesses.’Footnote 3 Oberhelman’s claims rest, at least in part, on his acceptance of Diagnosis from Dreams as authentically Galenic.Footnote 4 This short treatise, which we will discuss at some length below, presents a semiotics of medically significant dreams, a theory of their causation and a complex taxonomy of dream-types, all of which suggest a role for dreams in diagnosis. As Holowchak says, in this text, Galen ‘is encouraging practicing physicians not to neglect dreams as diagnostic tools’.Footnote 5 Hulskamp cautiously allows the authenticity of Diagnosis or, at least, its compatibility with the rest of Galen’s corpus, but mutes the ‘active role’ played by dreams in diagnosis.Footnote 6 Oberhelman, Holowchak, Hulskamp and others all see rational reason for Galen to rely on oneiric indicators in diagnosis—but leave open the question of why he seems not to do so in practice.
Galen’s own dreams and those of his father, Nicon Aurelius, however, prove less amenable to this assessment, and here scholars have frequently detected cracks in Galen’s apparent rationalism. Certainly, Diagnosis includes ‘prophetic’ dreams in its taxonomy, but more important are moments when Galen talks about therapies he learned from dreams, or when patients had prescriptive dreams, especially in a cultic context. Kudlien began his study of ‘Galen’s religious beliefs’ with Asclepian dreams, arguing that ‘Galen believed that in dreams, the god did give advice, prescriptions, orders, concerning actual cases or questions.’Footnote 7 Following Kudlien, von Staden and now Brockmann,Footnote 8 van der Eijk and Tieleman have claimed a degree of ‘compatibilism’ between rationality and piety in Galen. Van der Eijk argues for a ‘compatibilism’ in Galen’s religious thought that makes Asclepian activity continuous with and operative within the same natural laws as rational medicine.Footnote 9 Tieleman limits this ‘compatibilism’, arguing that Galen ‘accepts dreams, premonitions and signs as means of foretelling the future: he does not deny their occasional effectiveness … What he does deny is an exclusive reliance upon them and their use for the wrong purposes.’Footnote 10 Thus, Van Nuffelen has also concluded, ‘as Galen makes clear, divine dreams do not offer cures that go beyond what a doctor can prescribe and do. They do not transcend the science of medicine, for the knowledge of the gods is quantitatively but not qualitatively different.’Footnote 11
Compatibilism as thus described, amounts to a Gebildetenreligion (an intellectual piety), which accords with Galen’s theological manifesto in On His Own Opinions:
ἄρ’ οὖν καὶ περὶ θεῶν ἀπορεῖν φημι καθάπερ Πρωταγόρας ἔλεγεν ἢ καὶ περὶ τούτων ὁποῖοι μέν εἰσι τὴν οὐσίαν ἀγνοεῖν, ὅτι δ’ εἰσὶν ἐκ τῶν ἔργων γιγνώσκειν· ἐκείνων γὰρ εἶναι τήν τε τῶν ζῴων κατασκευὴν ἔργον, ὅσα τε διὰ κληδόνων ἢ συμβόλων ἢ ὀνειράτων προσημαίνουσιν. ὁ δὲ παρ’ ἐμοὶ τιμώμενος ἐν Περγάμῳ θεὸς ἐπ’ ἄλλων τε πολλῶν τὴν ἑαυτοῦ δύναμιν καὶ πρόνοιαν ἐνεδείξατο ἐμέ τε θεραπεύσας ποτέ· κατὰ θάλατταν δὲ Διοσκούρων ἔχω πεῖραν οὐ μόνον τῆς προνοίας ἀλλὰ καὶ τῆς δυνάμεως.Footnote 12
So then, I say I am uncertain concerning gods—just as Protagoras said—and ignorant about their substance, but I know from their works that they are. For I know the formation of living beings to be their work, and whatever they indicate in advance through signs, symbols or dreams. The god honoured by me in Pergamum has shown his power and providence in many different ways and especially when he healed me. I have experienced at sea not only the Dioscuri’s providence but also their power.
Galen demarcates the range of possible knowledge about the divine, which comes through natural channels. While excluding a knowledge of divine ‘substance’, he readily affirms both existence and providence (πρόνοια), which he pins first to the structure (παρασκεύη) of living beings, including humans, animals and plants. Galen undoubtedly has in mind the teleological orientations of bodies and their parts, on which he expatiates at length in On the Usefulness of Parts. This sounds like a perfectly rational compatibilism.
But Galen pushes providence’s range to the individual by adding the personal action of gods like Asclepius and the Dioscuri (the twin gods Castor and Pollux), who were frequently invoked by sailors and other travellers. By ‘experience’ of their power and providence at sea, Galen could well mean something as mundane as favourable winds, or he may intend that curious phenomenon later known as St Elmo’s Fire and, perhaps, a belief that this indicated his rescue from shipwreck.Footnote 13 As to Asclepius, Galen claims here and in On His Own Books that the god healed him,Footnote 14 and once mentions another man healed by the god.Footnote 15 He does not, however, offer any details of how either healing occurred. Then there is Galen’s mention of divine foretelling in dreams and symbols, which presents a more metaphysically challenging indication of providence. Divine foreknowledge (προσημαίνουσιν) raises questions of how much the gods direct and control events in the world—which Galen does not attempt to answer. Van der Eijk ends his study on a justifiably uncertain note, wondering whether ‘Galen’s attitude to Asclepius and his belief in dreams’ abrogates his compatibilism and whether it results from a ‘personal religious experience’.Footnote 16
Absent fuller explanation, it seems, dreams lead not to a neat compatibilism but to tension, even unacknowledged inconsistency, in Galen’s thought. Boudon-Millot, for example, traces Galen’s dismissal of religious healing as ‘outside the medical art’.Footnote 17 But, she says, ‘Let us not too hastily conclude that he entirely excluded religious practice from within medicine.’Footnote 18 Why? Asclepius and, more specifically, the dreams he sends: ‘Within rational medicine itself, there exist, then, two complementary and non-antagonistic approaches to the same phenomenon, according to which the dream could be attributed to the divinity or the patient’s physical state.’Footnote 19 It is not just that Asclepian priests might conform to Hippocratic methods, but that dreams point to a direct and irreducibly other intervention by the god himself. Von Staden has argued that generally Galen uses personal deity (whether Asclepius or his δαίμων) as cypher for innate, but carefully cultivated, rationality. And yet, Galen’s affirmation of dream-remedies force von Staden to conclude that Galen is either inconsistent or had a change of heart later in life.Footnote 20 More recently, Pietrobelli has pursued this latter possibility and asked if Galen had a conversion experience that led him to new religiosity, in which dreams played a key role.Footnote 21
All these scholars except Hulskamp share a central assumption: that when Galen speaks approvingly about acting on his own and his father’s dreams, he is attributing them to Asclepius.Footnote 22 Oberhelman puts it bluntly: ‘The respect which Galen accorded to such dreams from Asclepius was due to his genuine conviction of the god’s power to heal through dreams.’Footnote 23 Others are more cautious, but none question the direct connection between Galen and the god which in turn complicates and destabilizes Galen’s rationalism and blurs the edges of his medical art.
I will show that this basic assumption is flawed, that Galen is probably more compatibilist than even has been allowed, and, more importantly, that we learn more by asking what Galen does with dreams than by wondering where they come from.
DREAMS AS SYMPTOMS AND SIGNS
Dreams had a place in the ‘scientific medicine’ of the Hippocratics among the indications to which a physician must attend, and Thumiger has carefully traced the multiple lines along which dreams become symptomatic of disease in the Corpus, whether dietary, febrile or more specifically mental (e.g. epilepsy and phrenitis).Footnote 24 Within the Corpus, there are multiple, even opposing, explanations of why dreams might be significant indicators of health and disease, though most texts hold to some sort of naturalistic view. Dreams may be significant (1) as part of a pattern of sleep/sleeplessness, (2) because they are unusual or bothersome or (3) as images of what is happening or will happen in the body. As to (1), texts encourage physicians to inquire about dreams among other nocturnal experiences, like insomnia, excessive sleep, restless sleep and so on. Thus, in Epidemics, dreams are classed with sleep patterns as matters of ‘habit’.Footnote 25 A physician would be interested in both a patient’s normal patterns and changes to those that might indicate illness. As to (2), unusual, vivid or terrifying dreams feature among symptoms and signs but without pathological explanation. Now I would note that in the Corpus and later medicine, vivid (ἐναργής) dreams were associated with feverish mental states and, therefore, seen as indicators of acute phrenitis,Footnote 26 chronic melancholyFootnote 27 and other diseases which impact psychic as well as somatic capacities. Galen followed this tradition, and we will return to what oneiric vivacity meant for him.Footnote 28
It is really (3), the diagnostic and prognostic value of dream images, which requires explication, because this requires interpreting the content, rather than the mere fact, of dreams. Dream diagnosis assumes theories of dream causation or, at least, a semiotics of the oneiric. Galen’s only extended discussion of dream diagnosis is found in a fragmentary and possibly spurious text, On Diagnosis from Dreams (Diagnosis).
Diagnosis survives in apparently incomplete form (only 503 words and beginning with καί) and its opening lines are shared nearly verbatim with Galen’s Commentary on the First Book of the Hippocratic Epidemics. Footnote 29 Much of the text could be genuine because it accords with Galen’s attitudes on display elsewhere—the variable origins of dreams, the value of dream imagery for diagnosis and so on. It references works otherwise unknown, however, and includes passages that seem to contradict what Galen does say elsewhere. Consequently, Diagnosis is as problematic as it is precious a source. Demuth believed it apocryphal, but Oberhelman, following Guidorizzi, has argued that Diagnosis is a fragment of an authentic work, which included a lost commentary on the Hippocratic Regimen IV.Footnote 30 Boudon-Millot, however, has pushed back, saying that Diagnosis is a compilation by a follower of Galen at some remove, stitched together from his commentaries on Epidemics and on Plato’s Timaeus.Footnote 31 Hulskamp has made the most extensive analysis of Diagnosis, in relation to both the Hippocratic and Aristotelian traditions, as well as Galen’s larger corpus, and concluded that it is probably ‘spurious’—but not inconsistent with authentic Galenic works.Footnote 32
At stake in this debate are two sections of Diagnosis, the first a description of a mechanism of dream production, the second a reflection on the multiplicity of oneiric origins. Both passages would, if authentic, be invaluable in piecing together Galen’s appraisal of the oneiric. We will deal with both but begin with the production of dreams. Diagnosis says ‘In sleep the soul descends into the body’s depths and withdraws from external perceptions. There it perceives instead the body’s disposition, and it receives an image of everything it desires, as though already present’ (ἔοικε γὰρ ἐν τοῖς ὕπνοις εἰς τὸ βάθος τοῦ σώματος ἡ ψυχὴ εἰσδῦσα καὶ τῶν ἐκτὸς ἀποχωρήσασα αἰσθητῶν τῆς κατὰ τὸ σῶμα διαθέσεως αἰσθάνεσθαι, καὶ πάντων, ὧν ὀρέγεται, τούτων ὡς ἤδη παρόντων λαμβάνειν φαντασίαν).Footnote 33 Boudon-Millot argues that this line could well be inspired by Galen’s Commentary on the Timaeus. One fragment says, ‘This pneuma was shown by Galen to be generated in the brain’s ventricles and spread to all the sensoria during waking hours; but when it becomes drier than it should be, growing weak it descends to the depths (εἰς τὸ βάθος), being in need of rest and refreshment.’Footnote 34 The fragment, however, describes a mechanism not of dreams but of sleep both indebted to Aristotle and in line with Galen’s encephalocentric physiology.Footnote 35 As von Staden and others have shown, in other works Galen explicitly and at multiple points states that intellect and perception—each of which requires pneuma—are both inactive during sleep.Footnote 36 How, then, can a weakened pneuma produce images? Hulskamp has suggested an intriguing possibility on the basis of passages in Galen’s On the Movement of the Muscles and On the Causes of Symptoms, both of which soften the language of perceptual inactivity either to ‘difficult’ or ‘weak’ perception.Footnote 37 Thus, Hulskamp argues that, while the pneuma needed for perception is weakened or largely incapacitated (κεκμηκός), requiring rest and refreshment (ἀναπαύσεώς τε ἅμα καὶ θρέψεως), it may still be at least barely capable of supporting perception. The pneuma performs within the brain what the soul does more generally, which is to say, digesting and sleeping.Footnote 38 This would explain some oneiric activity, but does not yet explain how images are produced in the mind during sleep, let alone dreams not directly related to current perception—which is to say, the majority of them.
In the section taken from the Commentary Galen explores how differently coloured dream imagery denotes excess of correlated humours.Footnote 39 Dreams of conflagrations show that the dream ‘is hindered (ἐνοχλεῖται) by yellow bile’; dreams of smoke, gloom and deep darkness show black bile. Streams and rivers point to cold moisture generally, while snow and ice connote phlegm.Footnote 40 Later, he adds even smells—the stench either of putrifying humour or excess faeces seep into the dream.Footnote 41 Galen’s humoural theory provides a simple and straightforward framework for correlating oneiric imagery via colour, temperature and smell to the predominance of a given humour or material in the body. Unlike Artemidorus or Philo, Galen is entirely uninterested in the narrative unfolding in dreams; he is simply interested in the qualities of the images. The dreams project the humours through the kind of free association that Aristotle warned of in his appraisal of dreams.Footnote 42 While making dreams useless for prediction, this does make them useful for diagnosis and, perhaps, prognosis. Galen’s approach to prognosis is context- and, indeed, patient-specific, and this text is no exception. Thus, he says that the same imagery of snowfall may indicate hunger or the onset of chills, depending on whether the patient had eaten shortly before sleeping.Footnote 43 The text later considers also the course of disease, as well as further effects of bodily desire, which dreams often show being fulfilled: ‘Those about to sweat out a crisis seem to bathe and swim in tubs of hot water; the thirsty often have an image of drinking without being slaked, just as the hungry eat insatiably, and those full of semen dream of sex’ (τινὲς δὲ ἱδροῦν κριτικῶς μέλλοντες λούεσθαι καὶ κολυμβᾷν ἔδοξαν ἐν θερμῶν ὑδάτων δεξαμεναῖς. οὕτω δὲ καὶ ἡ τοῦ πίνειν φαντασία χωρὶς τοῦ πίμπλασθαι διψῶσι πάνυ συμβαίνει, καθάπερ καὶ τοῦ ἐσθίειν ἀπλήστως τοῖς πεινῶσι, τοῦ δὲ ἀφροδισιάζειν τοῖς σπέρματος πλήρεσιν).Footnote 44 Through these means, as Boudon-Millot argues, Galen would be able to incorporate dream imagery, via humoural theory and commonsense physiology, into diagnostic practice.Footnote 45
Shortly after, though, the text raises a more pernicious problem of oneiric epistemology, which hinges on the varied origins of dreams:
ἐπεὶ δὲ ἐν τοῖς ὕπνοις οὐκ ἐπὶ ταῖς τοῦ σώματος ἡ ψυχὴ διαθέσεσι φαντάζεται μόνον, ἀλλὰ κᾀκ τῶν συνήθως ἡμῖν πραττομένων ὁσημέραι, ἔνια δὲ ἐξ ὧν πεφροντίκαμεν, καὶ δή τινα μαντικῶς ὑπ’ αὐτῆς προδηλοῦνται (καὶ γὰρ τοῦτο τῇ πείρᾳ μαρτυρεῖται), δύσκολος ἡ διάγνωσις τοῦ σώματος γίγνεται ἐκ τῶν ἀπὸ τοῦ σώματος ὁρμωμένων ἐνυπνίων.Footnote 46
Since in dreams the soul forms images not only about the body’s dispositions but also from our constant daily practices and sometimes from things we are worried about, but some things are even foreshown by it prophetically (for experience testifies to this), and so the diagnosis of the body from dreams seen concerning it becomes difficult.
Dream imagery is not simply context-specific, but fundamentally ambiguous because the same image may depict a present bodily state, a persistent worry, a desire or even a prophetic indication. The first three categories fit easily within Galen’s theories of sleep and sit well with his psychology and physiology more broadly. Οur text includes a category of predictive dreams (μαντικῶς), however, which precludes a simple diagnostic sortation. This last category takes us beyond the limits of medicine, as Galen uses the same term for divinatory practices as opposed to prognostic ones.Footnote 47 While tempting to equate divinatory dreams with divine ones, we have good reason to think that these, too, emerge from some capacity or activity of soul. Aristotle, for example, admitted that dreams can come true, but denied the possibility of divine origins because ‘really worthless people are visionaries who have true dreams’ (πάνυ γὰρ εὐτελεῖς ἄνθρωποι προορατικοί εἰσι καὶ εὐθυόνειροι).Footnote 48 We will return to the physiology of such dreamers later, but suffice to say that Aristotle believed predictive dreams to be a side effect of the purposive structure of beings within a teleologically ordered cosmos, not the work of divine beings or psychic processes.Footnote 49 Likewise, the anatomist Herophilos of Alexandria and the oneirocritic Artemidorus held to oneiric schemata that disentangled divinatory from divine dreams.Footnote 50 There is a long-standing tradition, then, of accepting the predictive potential of dreams without presuming or even allowing a divine or external origin. So too in Diagnosis, given the psychological mechanism averred in the same text and absent any mention of divine origins (such as we saw above in On his Own Books), predictive dreams likely emerge from the operations of a dreamer’s own naturally ‘prophetic soul’.Footnote 51 Without invoking divine origins, this multiplicity nevertheless raises the stakes for dream interpreters, including physicians, since the content may signify different things depending on its origins.
Diagnosis underscores the problems this raises for diagnosis in two frustratingly inconclusive examples. First, a wealthy man’s dream is interpreted prophetically as regarding his servants, but it actually presaged paralysis. Second, a dream was deemed to show an excess of blood, for which the patient was bled.Footnote 52 We are not told with what success. These two examples recall those given in Questioning Patients by Rufus of Ephesus, one of the rare physicians whom Galen expressly admired.Footnote 53 But unlike Rufus’ examples, neither gives us any indication of how the physician ought to parse such dreams. What emerges from these two stories is diagnostic ambiguity, both in the meaning of dream imagery and in the value of treatment following it.
This ambiguity and the implied unreliability of dream diagnosis may explain why, despite clearly having a robust semiotics of oneiric imagery, Galen never actually records dream content in his case studies. Why not? Hulskamp offers a couple of reasons. First, there is the hermeneutical challenge presented by oneiric imagery, which is not only confused, but also filtered through the haze of waking memory. Then there is Galen’s aggregating approach to diagnosis; as she puts it, ‘Galen valued concurrence of various diagnostic indicators, which might or might not include dreams.’Footnote 54 Hulskamp’s point may be extended. Galen almost never makes a diagnosis or prognosis based on a single consideration (σκοπός), instead multiplying them in his quest for a complete understanding of the patient and their health or illness—age, circumstance, normal humoural mix for that individual, past and present habits of diet, sleep, sex and so on; not to mention reported pain, measured pulse, consistency of blood and, of course, urine.Footnote 55 His approach to therapy couples this specific knowledge of patients with equally specific understanding of the courses of diseases and powers of natural substances in relation to bodily parts to facilitate a precise calibration of the Hippocratic principle of opposites curing opposites.Footnote 56 In this process Galen prefers immediate indicators like blood and urine to those that require patient mediation and interpretation, like pain.Footnote 57 Dreams are, by nature, difficult to recall or describeFootnote 58 and so, while Galen has a semiotics of dream imagery at hand, and a place for it within his humoural physiology, it would have been difficult to implement in practice. Dreams’ ambiguity and dreamers’ unreliability preclude the oneiric from playing much, if any, actual role in diagnosis or prognosis. Moreover, apart from the brief discussion in Diagnosis—if it is even authentic—Galen is otherwise silent on the question of dream production, which suggests that he was not particularly interested in it.Footnote 59 Insofar as dreams do feature in diagnosis, they are caused and determined by psycho-physiological factors and, therefore, confined within the limits of rational medicine.
VIPER VENOM: DREAMS AS EXPERIENTIAL KNOWLEDGE
Turning to therapy, dream-learning is both a serious asset and a pernicious problem for Galen. In this section I show that for much the same reasons that dreams may be unreliable or at least ambiguous indicators of bodily dispositions, they provide an irreplaceable mode of experience which can lead to innovation and success if handled appropriately, or stultification if abused. Oneiric epistemological ambiguity, as we shall see, is matched by its didactic potential.
Accidental experience and its limits
Galen discusses dream-learning as a mode of experience in context of his various accounts of the Empirics, for whom the collection of observations and the aggregation of individual cases was the essence of medical art.Footnote 60 In On the Sects for Beginners, Galen reports the Empirics to divide experience into two kinds: from observing what happens without human intervention (ἀπὸ ταὐτομάτου), which they call a ‘natural form’ of experience (εἶδος … φυσικόν); from noticing accidents and chance events (falls, wounds, etc.), which they call a ‘chance form’ (εἶδος … τυχικόν). The initial observation of either of these they call ‘an accidental form of experience’ (περιπτωτικὸν εἶδος τῆς ἐμπειρίας). There is another mode of experience, ‘when people willingly reach out to try something, either being encouraged by dreams or imagining it in some other way’ (ὅταν ἑκόντες ἐπὶ τὸ πειράζειν ἀφίκωνται ἢ ὑπ’ ὀνειράτων προτραπέντες ἢ ἄλλως πως δοξάζοντες). This form of experience they call ‘improvised’ (αὐτοσχέδιον).Footnote 61 Galen then adds a third, the ‘imitative’ (τὸ μιμητικόν), which he believes undergirds Empiric (and, really, any medical) practice: repeated observations in diagnosis and therapy, accumulated until one knows that an observation holds true or is effective most of the time, if not always. Empiric epistemology works through remembering accumulated observations, sorted by memory into general and often quite specific plans of action.
Often hailing the irreplaceability of experience in medicine, Galen is sympathetic to the Empiric position, defending their approaches against critiques from Dogmatic (or Rationalist) physicians. Galen’s sympathy with Empirics, however, ends where experience stands in need of technical reason (λόγος), since ‘neither experience nor reason alone are sufficient to discover everything’.Footnote 62 Noting that Galen refuses to be identified with any of the ‘schools’ of his day, Frede argued that ‘Galen is a rationalist, he regards himself as a rationalist, but he rejects the particular kind of rationalism adopted by the rationalists.’Footnote 63 That is, Galen does not oppose technical reason and experience. Rather, knowledge is both built up from the observation of numerous cases and deduced from principles through methodical inquiry, and both must be subjected to the criteria of demonstration. Thus, as van der Eijk shows, Galen is after experience ‘qualified’ by rational inquiry and susceptible to demonstration.Footnote 64 And this is where dream-learning becomes problematic for him.
Galen expects that someone can and should explain how they arrived at the recipe for a complex drug. He imagines, then, an opponent ‘steeped in experience alone’. For this extreme Empiric, the drug’s components were learned ‘by coincidence’ (κατὰ σύμβασιν), which we have seen is how Empirics refer to initial observations. Galen continues, ‘They say one could sometimes discover such things from dreams (ἐξ ὀνειράτων), or sometimes if by chance something was poured into something else’. Dreams and spills can well inspire new substances and interventions. Without quibbling over its origins or even its ultimate efficacy, Galen seizes on the real issue: his imagined opponents praise ‘one who dares to use the mixture, but do not say what his daring hopes for’ (εἶτά τινα τῷ μικτῷ τολμῆσαι χρήσασθαι, τῆς τόλμης δ’ οὐ λέγειν τὴν ἐλπίδα).Footnote 65 Dreams and spills are both ‘accidental’ or ‘coincidental’ experiences—mere initial observations—and, because of their infrequency or even unrepeatability, not susceptible to the discrimination of ‘imitative experience’. Explanation and rational demonstration are not yet at hand, since these can only come through repeated observation or through methodical inquiry to which the coincidental should conform.
Galen’s account of Empiric epistemology allots a curious place to dreams. In On the Sects for Beginners, dreams are a sidenote, only the initial impetus to make observations, and not a full experience. Yet as in On the Method of Medicine above, Galen suggests that some Empirics rely on dreams as experience, even at the expense of reason and nature. He does so again in On the Composition of Drugs by Class.Footnote 66 The difficulty seems to lie with the nature of oneiric experience: it is real and undeniable, but it is not repeatable or imitable. In the polemic against one ‘steeped in experience alone’, Galen’s criticism seems to be against relying solely on this. No wonder that the otherwise sympathetic Galen concludes: ‘Such claims, then, are obviously rubbish.’Footnote 67 Either experience has been accumulated and efficacy demonstrated, or else one stupidly praises unreflective audacity.
What can we make of this rejection, when Galen elsewhere approves of dream-taught therapies? We turn now to those, to parse how the self-proclaimed rationalist justifies his own reliance on dreams.
Dream experimentation
Galen several times describes successful applications of therapies learned from dreams, both his own and those of others, from which it is immediately obvious that he can and does accept therapies suggested by dreams.Footnote 68 It is less obvious how the dream functions within his medical epistemology and demand that experience be qualified by reason. What, if anything, exempts his own ‘daring’ from that of the Empirics he mocks?
In a famous example from On Healing through Phlebotomy, Galen says,
ἔγωγ’ οὖν ὅθεν ὁρμηθεὶς ἐπὶ τὸ διαιρεῖν ἀρτηρίας ἧκον ἤδη σοι φράσω. προτραπεὶς ὑπό τινων ὀνειράτων δυοῖν ἐναργῶς μοι γενομένων ἧκον ἐπὶ τὴν ἐν τῷ μεταξὺ λιχανοῦ τε καὶ μεγάλου δακτύλου τῆς δεξιᾶς χειρὸς ἀρτηρίαν, ἐπέτρεψά τε ῥεῖν ἄχρις ἂν αὐτομάτως παύσηται τὸ αἷμα, κελεύσαντος οὕτω τοῦ ὀνείρατος.Footnote 69
How I, at least, was impelled to come to cutting veins, I have already told you. Having been encouraged vividly by two dreams I cut the artery between the forefinger and the middle one of the right hand and allowed it to bleed until it stopped on its own, as the dream commanded.
Galen offers no further explanation of the dream except that this happened when he was still young.Footnote 70 The issue is the site of the cut, the vessel cut and the cessation of blood-flow—all were at least somewhat unexpected, but ultimately quite successful. The dream had, in some way or another, suggested both the experimental action and its outcome.
Galen then says that a ‘devotee of the god in Pergamum was delivered from chronic side pain’ (χρονίου πλευρᾶς ἀλγήματος) through venesection of his little finger, ‘as he too came upon this on account of a dream’ (ἐξ ὀνείρατος ἐπὶ τοῦτο ἐλθὼν καὶ αὐτός).Footnote 71 Without claiming divine origins for either dream, by naming Asclepius Galen allows the reader to believe, if they choose, that the god was behind them. The proximity of Asclepius provides a plausibility structure for Galen’s reliance on a dream which is further confirmed by another’s successful use of the same treatment. Encouraged by this, scholars have either claimed outright that this dream was Asclepian,Footnote 72 or have connected this incident with Galen’s mention in On his Own Books of how ‘Asclepius saved me when I had a morbid condition of abscess’ (με θανατικὴν διάθεσιν ἀποστήματος ἔχοντα διέσῳσε).Footnote 73 But neither conclusion is warranted. First, as already noted, Galen does not say that his dream was in any way Asclepian, only that it was vivid—a point to which we will return. Second, in Phlebotomy, Galen describes the illness as ‘chronic pain’ (χρόνιον ἄλγημα), without a hint of its being life-threatening—which detail Galen would certainly include if it were. This would, therefore, seem to be a separate incident from that recorded in On his Own Books, and the Asclepian connection cannot be made on those terms or on the basis of the story itself.
The purported Asclepian connection obscures the real import of the passage, which is to demonstrate by his and others’ experience that arteries can be safe and effective sites for blood-letting, despite a general reluctance to use them lest patients bleed out.Footnote 74 The dreams offer a rationale for Galen’s initial experimentation, but he spends much of the chapter offering a physiological explanation for the relative safety of some arteriosection. The unnamed devotee of Asclepius serves as demonstration that this method worked in multiple instances and can, therefore, be admitted as a possible therapeutic practice. The dreams are merely good excuses for experimentation, while their ‘teaching’ accords with Galen’s science. In short, they invigorate and legitimatize, but do not delimit his practice.Footnote 75
Sometimes, though, Galen credits dreams with more radical therapies. In his compendious guide to simple medicaments, Galen discusses vipers and, more specifically, their venom, as materia medica. Galen relates several of what Mattern calls his ‘case studies’, as experiential bases for rational treatment.Footnote 76 In these stories, patients unknowingly or unexpectedly ingest viper flesh or venom and are cured. Galen couches the tales as knowledge derived from experience (πεῖρα).Footnote 77 The conditions under which viper venom is effective must be gradually deduced by its different effects at different points in the progress of disease: in some cases disastrous, in some healing and in some just odd. It seems that one effect of viper venom is to turn elephantiasis into leprosy which Galen says is far more treatable. A third case study therefore concerns a diviner (ἔμπειρος οἰωνῶν) who learns of viper venom—and its dangers—from Galen, but uses it in conjunction with his own augury to ‘imitate what we had learned through experience’ and thus became leprous. Whether this was in itself an improvement is difficult to say, but Galen was then able ‘to cure his leprosy using the customary drugs’ (συνήθεσι φαρμάκοις).Footnote 78 This diviner’s art merely confirmed Galen’s medical one.
The fifth case concerns a wealthy man from Thrace who, ‘with a dream prompting him’ (ὀνείρατος προτρέψαντος αὐτόν), came to Pergamum. There ‘a dream of the god commanding him to drink the drug made from viper daily and to anoint his body, the disease turned within a few days to leprosy, and this disease was then healed by the drugs which the god commanded’.Footnote 79 This last case is the only unambiguously Asclepian connection in the lot—another man’s dream, reported second-hand, accepted by Galen but without committing the physician to its divine origin, only its effective outcome. After all, Galen criticizes dream learning because one cannot explain what they hoped to accomplish, let alone how it might take place. There must be some balance between trying new, even surprising ideas and relying on untestable means. In Galen’s tale of the wealthy Thracian, this balance seems to hold. First, the divine dream takes place in Pergamum, almost certainly within the cultic context of the Asclepieion there. Second, Galen does not claim the dream for himself and does not build a practice from it, although he, whatever his own opinion of the dream, would certainly find the god’s command a useful tool in persuading readers of the value of such a counter-intuitive remedy. Third, this divinely given dream-command perfectly mimics what Galen says of that diviner who imitated what he himself had learned ‘through experience’, and the similar outcomes—elephantiasis turning to leprosy, then cured by drugs—serve to confirm a framework within which to use viper venom. When it comes to viper venom, Mattern argues that ‘Galen is giving an example of a therapy he arrived at by Empiric methods, in which case histories play the leading role.’Footnote 80 True, but while the idea might come from a dream, its application and incorporation into the physician’s repertoire required careful study of patients, circumstances and other details. As in his neutral discussion of Empiric divisions of experience, there is nothing wrong with observing things which happen by chance or in seeking out experiences, whatever the motivation. This drives innovation through experimentation. What matters is how that experience is fitted to a rational framework and subjected to demonstrative proofs before being claimed as part of the medical art. It is this entire constellation of sources that guides Galen’s claims to knowledge, by qualifying his and others’ experience, even if the means of acquiring experience are themselves less explicable.
VIVID, NOT ASCLEPIAN
In Galen’s descriptions of significant dreams one quality stands out: their vividness. With that in mind, let us turn to Nicon Aurelius’ dreams. Scholars have almost unanimously accepted that these were Asclepian. In addition to the scholars discussed above, we may add Strohmaier,Footnote 81 Nutton,Footnote 82 Schlange-Schöningen,Footnote 83 GarofaloFootnote 84 and Singer.Footnote 85 Mattern strikes a more cautious note: ‘It was perhaps Asclepius himself who guided Galen’s father to this decision in dreams.’Footnote 86 This reading is reasonable, given the intellectual climate of Galen’s day and his own avowed worship of Asclepius, but it is incorrect. The dreams are not Asclepian either in their origin or in their imagery, and once that connection is severed, they can be plausibly accounted for by Galen’s psychophysiology. Galen describes these dreams in three passages, worth quoting in full, because they demonstrate the utility of dreams rather than the origin.
First, there is On the Method of Healing:
εὐθὺς ἐκ μειρακίου φιλοσοφίας ἐρασθέντες ἐπ’ ἐκείνην ἥξαμεν πρῶτον. εἶθ’ ὕστερον τοῦ πατρὸς ὀνείρασιν ἐναργέσι προτραπέντος ἐπὶ τὴν τῆς ἰατρικῆς ἄσκησιν ἀφικόμεθα καὶ δι’ ὅλου τοῦ βίου τὰς ἐπιστήμας ἑκατέρας ἔργοις μᾶλλον ἢ λόγοις ἐσπουδάσαμεν.Footnote 87
I was already smitten with philosophy when I was young and came first to that. Later, my father having been urged in vivid dreams, I came to the discipline of the medical art and my whole life I have laboured diligently in both sciences, in actions more than in words.
While it is hard to imagine Galen labouring in anything more than in words, his father’s ‘vivid dreams’ (ὀνείρασιν ἐναργέσι) were decisive in his choice of career. Galen repeats this story verbatim in two other places. First, there is a passage in On the Order of his Own Books, in which Galen, with typical humility, presents his own education as the set curriculum for any aspiring physician:
εἶναι δὲ χρὴ τοῦτον ἀγχίνουν θ’ ἅμα καὶ μνήμονα καὶ φιλόπονον, ἔτι δὲ πρὸς τούτοις εὐτυχηκότα τοιαύτην εὐτυχίαν οἵαν ἡμεῖς εὐτυχήσαμεν ὑπὸ πατρὶ παιδευθέντες, <ὃς> ἀριθμητικῆς τε καὶ λογιστικῆς καὶ γραμματικῆς θεωρίας ἐπιστήμων ἡμᾶς ἐν τούτοις τε κἀν τοῖς ἄλλοις, ὅσα παιδείας μαθήματα, θρέψας, ἡνίκα πεντεκαιδέκατον ἔτος ἤγομεν, ἐπὶ τὴν διαλεκτικὴν θεωρίαν ἦγεν ὡς μόνῃ φιλοσοφίᾳ προσέξοντας τὸν νοῦν, εἶτ’ ἐξ ὀνειράτων ἐναργῶν προτραπεὶς ἑπτακαιδέκατον ἔτος ἄγοντας καὶ τὴν ἰατρικὴν ἐποίησεν ἀσκεῖν ἅμα τῇ φιλοσοφίᾳ.Footnote 88
He must be shrewd, a good rememberer and a hard worker. Yet beside these qualities he needs to have been fortunate in such good fortune as we ourselves were blessed with, having been taught by our father. He, being knowledgeable in theories of arithmetic, logic and grammar, reared us in these and whatever other subjects pertain to education until we were fifteen, and then he guided us to dialectic, as we were going to turn our mind to philosophy alone. Then, having been urged by vivid dreams as we were turning seventeen, he caused us to start training in medicine as well as philosophy.
In this fuller account of his early education, Galen assures the reader that Nicon had him study all the liberal arts. He assures us, again, that his first love was philosophy, and that medicine came to join it (rather than replace it). Finally, there is a brief notice in On Prognosis: ‘For he [Eudemus] had also learned that while my father was guiding me toward philosophy, it had been commanded to him through vivid dreams to instruct me thoroughly in matters of medicine as well—and not as some trivial subject’ (καὶ γὰρ ἐπέπειστο τοῦ πατρὸς ἐπὶ φιλοσοφίαν ἄγοντός με, προσταχθὲν αὐτῷ δι’ ὀνείρων ἐναργῶν ἐκδιδάξαι καὶ τὰ τῆς ἰατρικῆς οὐχ ὡς πάρεργόν τι μάθημα).Footnote 89 In this story, Galen presents medicine as a complement rather than replacement for philosophical studies. In Method of Healing, above, he described them as the ‘two sciences’ to which he has devoted his life, while here he suggests that medicine joined philosophy. Of particular importance, as we shall see, is the negative affirmation that medicine is ‘not some trivial subject.’
In all three passages, the critical moment for Galen’s career comes at seventeen, when his father was ‘urged’ (προτραπείς, προτραπέντος) or ‘commanded’ (προσταχθέν) by ‘vivid dreams’. Galen never says that these dreams are divine, let alone Asclepian. Nor are they prophetic. They in some way exhort or suggest a course of action, which his father was free to ignore as well as take. They do not ‘indicate in advance’ (προσημαίνουσιν), as Galen says of divinely sent dreams.Footnote 90 The dreams themselves are simply ‘vivid’, just as were the dreams that inspired Galen to new therapies. This is their defining quality. So too, Galen says of an unnamed patient, that ‘Because of an incredibly vivid nocturnal dream that befell him, he both praised my counsel and ordered the material for the drug, requesting a decoction with lettuce juice’ (ἀλλὰ διά γε τῆς νυκτὸς ἐναργέστατον ὄναρ αὐτῷ γενόμενον ἐπῄνεσέ τε τὴν ἐμὴν συμβουλήν, ὥρισέ τε τοῦ φαρμάκου τὴν ὕλην, θριδακίνης χυλῷ διακλύζεσθαι κελεῦσαν).Footnote 91 The dream is important insofar as it confirms the advice that Galen had already given! Indeed, probably the most important point for Galen is the experiential confirmation that follows: ‘The man, having used this decoction alone, was so completely helped by it that he needed no other.’Footnote 92
In each case, Galen emphasizes the ‘vividness’ or ‘vivacity’ (ἐνάργεια) of the dreams. Why? I propose several interconnected reasons.
First, language of vivacity taps a long history of literary accounts of divine dreams, from the Odyssey to Plutarch.Footnote 93 Galen may be playing on that tradition, a possibility further suggested by his language of ‘command’ and ‘urge’.Footnote 94 But while he may be happy for readers to assume one, he never names a divine agent. Rather, in these passages, Galen insists only on the quality of dreams’ imagery, which can have far less exalted causes. While literary myths connect vivacity with the gods, philosophers and physicians connect it to an abundance of black bile. To begin with the last, we have seen that Aristotle actually used the predictive truth of dreams as an argument against divine origins. ‘Worthless’ people can have predictive dreams, he suggests, for two psycho-physiological reasons. First, those whose ‘nature is garrulous and melancholic’ see all kinds of things because of this. Second, just as people have particularly vivid or correct dreams of people they are deeply concerned about, melancholic people (οἱ δὲ μελαγχολικοί) are particularly likely to ‘guess correctly’ since their imagination moves so quickly from image to image, making links and forming conjectures which often turn out right.Footnote 95 This ability to link images makes the melancholic person into what Struck has called a ‘sensitive instrument’, prone to correct guesses and true dreams, not because of their effort but because of the purposive structures of nature which govern motion both in the heavens and the human body and which turn even dreams to good effect.Footnote 96 Aristotle is joined by the Hippocratic tradition and Galen too, in associating vivid dreams with melancholy. In a commentary on the Hippocratic Prognosis, Galen says that ‘images in sleep seem totally vivid to melancholics, because of their dryness’ (οὕτω γοῦν καὶ τοῖς μελαγχολικοῖς διὰ τὴν ξηρότητα πάντως ἐναργῆ φαίνεται τὰ κατὰ τοὺς ὕπνους φαντάσματα).Footnote 97 Such oneiric activity is activated by excessive dryness which, left unchecked, can also induce phrenitis, for which vivid dream activity is an indication. Vivid dreams are, therefore, a symptom of developing imbalance and mental disease, and yet may well turn out to predict future events—that is, to be ‘true’. This is not to say that Galen or his father were melancholic, let alone phrenitic, only that physiology provides sufficient explanation for mantic vivacity, without the operation of a divine agent. Galen’s referring to these particular dreams being ‘vivid’ is intended to mark them as unusual and provocative.
Vividness (ἐνάργεια) constituted a criterion of the truthfulness of representations in Hellenistic philosophy and an integral means of persuasion in ancient rhetoric. Galen follows both discourses, which together explain his positivity toward those few dreams he calls ‘vivid’ and his inclination to rely on them for inspiration and confirmation.
Galen was confident in the data gained from the senses, and in the power of deductive reasoning to complete what senses could not. Chiaradonna puts it well: ‘Galen’s epistemology is based on firm (if not excessive) confidence about the reliability of our basic cognitive capacities.’Footnote 98 This was a source, as discussed already, of his affinity for the Empiric school as well as his quarrel with its single-minded reliance on sensation and memory.Footnote 99 While the Empiric school—as Galen describes it—made a specific and highly reductive use of sensation, the importance of clarity, of an idea being self-evident, was well accepted across the schools of Hellenistic philosophy, each of which tied vividness to truthfulness in representations and axioms.Footnote 100 Likewise, ancient rhetoric afforded ἐνάργεια a key role in sparking listeners’ imagination as a means to persuasion. In her magisterial study of ancient rhetoric, Webb argues that vividness is not merely a trick of the trade, but a valid means of argument.Footnote 101 Its rhetorical value, claims O’Connell, lies in simulating sense-perception for listeners, resulting in them having the same mental perception (φαντασία) as the speaker without needing first-hand experience.Footnote 102 This account makes sense in light of the place of ἐνάργεια in philosophical epistemologies: a vivid account is also an evident one, the details that create a mental perception being the same ones that make it plausibly true.
Galen recognizes dreams as dreams. The danger of vivid dreaming for melancholics and phrenitics lies in their inability to distinguish the dream from reality, an issue that Galen understood both medically and, more broadly, as an epistemological issue generally solved by denying that dreams possess ἐνάργεια.Footnote 103 However, every time Galen claims to have relied on or been guided by a dream, he calls it both vivid and a dream—something more than a passing phantasm but nothing mistakeable for waking reality. Thus by language of ‘command’ and ‘exhortation’, Galen intends that the dream itself, rather than a divine agent, was a persuasive experience, but should not be deemed ‘true’ on its own.
Galen understood sensation to be most reliable when its impressions are clear and distinct (ἐναργής), as are fundamental axioms, and what reason needs to add to sensation should be made clear (δῆλα) by demonstration (ἐπίδειξις).Footnote 104 It is this last element that is so important, because demonstration allows Galen to step from ‘plausible’ (πιθανόν) or ‘probable’ (εἰκός) to ‘true’. It is one reason he gives for staying out of speculative philosophy and theology—being neither self-evident nor demonstrable, claims about the nature of the gods can only ever be plausible, never firmly true.Footnote 105 Within his own work, Galen prized this combination of sensation, deduction and demonstration as ‘qualified experience’, the basis of philosophical inquiries and medical treatments alike,Footnote 106 because by demonstration one could reason from the plausible to the necessary and so confirm through experiment what had first been inspired only by a vivid or persuasive idea.Footnote 107 Thus, a vivid dream is 1) more persuasive, 2) has sufficient evidentiary quality to provide a plausible basis for action, but 3) its truth can only be confirmed through demonstration, which is possible because 4) vivid dreams are more memorable upon waking.Footnote 108 The dream needs no god behind it when a mixture of memories, desires and prior knowledge, catalyzed perhaps by the accelerant of black bile, naturally percolates into a plausibly useful, crucially testable, idea or plan of action.
CONCLUSION: THE UTILITY OF DREAMS
We have long read Galen’s accounts of his own and, especially, his father’s, dreams, exactly as he intended and, by succumbing to his rhetoric, somewhat missed the strategy in play. While never stating outright that these were ‘Asclepian’, Galen knew that readers would assume as much. A strategy of suggestion, with plausible deniability, perhaps? Dreams, in these instances, are an apologetic for Galen’s actually rather odd career choice. Here we need to remember the relative standing of a physician in Roman society apart from the elevation lent it by Galen. It seems to have been a respectable enough job—though frequently performed by enslaved persons and competing in a marketplace of humbler arts—but, at its best, could not hold a candle, intellectually speaking, to philosophy. Neither had the social status or political clout of oratory. But Galen was studying philosophy and intended to be an orator and remained proud of his efforts in both. His medical career could be seen as a derailment from more glorious tracks, his standing lowered and his talents questioned. But divinely sent dreams brook no objection. Galen knew the truth of that when he declined to join Marcus Aurelius on pretext, he would later tell us, of Asclepian dreams.Footnote 109 If desertion is no crime, then medicine is no demotion, provided the god himself commands it or is assumed to have commanded it. And so Nicon Aurelius’ dreams function apologetically for Galen, as an unassailable explanation of a questionable career move. In these instances Galen, always conscious of his reputation and status, deploys paternal dreams as tools of self-fashioning and apology, rather than indicators of his religiosity or statements on the actions of Asclepius.
Galen’s references to Nicon’s dreams contrast sharply in tone with those passages where he speculates on diagnosis through dream-imagery or includes dreams among the symptoms of diseases. In those moments, dreams are one of the considerations to which a physician must attend. So too the dreams that suggest diagnoses or therapies. These may well be tried, and may well have unexpectedly good results, but the dream motivates only in concert with experience in medicine generally and the patient specifically, together with other signs and indications that allow the physician to explicate, if not how it operates, then at least what they can expect to happen. Without regard to their origin, Galen weaves dreams into the epistemological fabric of fundamentally rationalist medical knowing and practice. So too, dreams indicate divine existence and providence as bookmarks in the tome of nature which Galen reads at length. Galen is sincere in his devotion to his ‘ancestral God, Asclepius’, and adamant in his belief that the god had, indeed, saved him when younger, though by what means we will never know. But this sincerity requires nothing from dreams, and Galen can rely on them in some circumstances and, with total intellectual consistency, reject them in others; he can deploy them when useful while maintaining his own reservations about their epistemic value. Vivid dreams may be divinely sent or, at least, allowed to appear such, or they may be the result of too much black bile—and yet no less accurate in their indications and suggestions. Galen is happy to record dreams when they confirm to a patient what he had advised—that is, when they are persuasive. Dreams are not cracks in Galen’s rationality but, rather, rhetorical tools that he deploys in service of his self-fashioning as a leading intellectual, of his superiority as medical author and, in the more intimate moments of his clinical practice, of his competence in treatment.
Open access
