In the medical practice of Asclepian dream incubation, dreams offered a conduit through which the divine power of the healing god could be visited upon an ailing suppliant. This practice was enough of a part of everyday life in fifth-century Athens that it achieved the dubious honor of an extended parody in Aristophanes’ Plutus. An extensive inscriptional record suggests that it continued to flourish for many centuries.Footnote 1 But there was another type of dream employed in ancient Greek and Roman medical practice, with a much scanter trail of evidence. These dreams had endogenous, physiological origins and provided information about the internal disposition of the body not by divine intervention, but by some manner of inward perception on the part of the patient. With the rising interest in observational methodology in the fith century, opsis, and ideally autopsy, became the basis on which scientific knowledge was produced and elaborated. Taboos against physically opening the human body, in life as well as in death, prevented physicians from directly observing their patients’ interiors.Footnote 2 The visions of dreams, then, could potentially provide doctors with a uniquely valuable diagnostic tool: genuine access to the observation of a body's internal condition, albeit in a strange, mediated form.
The earliest extant articulation of this diagnostic method (as well as the fullest and most confident one) occurs in the fourth book of the Hippocratic treatise On Regimen, or De Victu (ca 410 BCE). The book opens with an assertion by the author that ‘whoever has learned correctly about the proofs that come in sleep will find that they have great power in regards to all things’ (Hp. Vict. 4.86). In general, modern scholarship on the ancient medical tradition has accepted that this treatise reflects an early instantiation of a commonly used method, a ‘standard prognostic and diagnostic tool of the Greek and Roman physician’.Footnote 3 But while there are indeed references to the event of dreaming being recognized as a symptom,Footnote 4 there are in fact no real references to diagnosis or prognosis based on the interpretation of physiological dream content between On Regimen from ca 410 BCE and the short Galenic treatise On Diagnosis from Dreams from the second century CE, which itself contains just as much warning about potential misdiagnosis as it does successful diagnosis.Footnote 5 Rather than a standard tool, diagnosis and prognosis from dream analysis appears to have been a promising but ultimately unrealizable method for most physicians and medical theorists.
This article has two main goals, which amount to two sides of the same coin: first, to consider what obstacles may have prevented the promising method of diagnosis from dreams from being adopted as a regular part of ancient medical practice and, second, to turn around and determine how these obstacles were overcome or sidestepped by the author of On Regimen. On the one hand, I argue that although dreams have been heralded as a productive nexus between secular and divine healing, the available evidence suggests that difficulties surrounding dream categorization may have actually further polarized secular and divine healing professionals, while raising obstacles to effective interpretation. On the other hand, I argue that even when physiologically significant dreams were correctly categorized and in the hands of the appropriate professionals, the fact that most extant medical theories of dream production were analogous to theories of artistic creativity made reliable interpretation difficult, due to issues of symbolic ambiguity. But the theory of dream production presented in On Regimen eschews this creative model, suggesting instead that physiological dreams were the result of the soul's direct perception of the internal disposition of the body, thereby avoiding such interpretive pitfalls. This reading requires a careful reconsideration of the Hippocratic treatise's theory of perception and the nature of its micro-macrocosmic vision of the relationship between the human body and the rest of the cosmos. It also requires a sensitivity to the conceptual intersections between medicine, philosophy, and literary criticism. Analogies and disanalogies between poēsis and the production of dream content provide valuable context for the unique dream theory of On Regimen. But more broadly speaking, they also draw attention to the way that ideas of artistic production were a part of a larger discourse about mankind’s material and embodied inhabitation of the world.
The Problem of Dream Categorization
An understanding that dreams fall into different categories, as well as a recognition of the difficulty in determining into which category a particular dream might fall, already had a long background by the fifth century. Traces of an ambivalence in ancient Greek thought concerning the simultaneous power and unreliability of dreams extends at least as far back as the Iliad and the Odyssey. Assuming that Homer is the earliest of our poets and assuming that the Iliad is earlier than the Odyssey, the first extant dream in the Greek tradition is the lying one Zeus sends to Agamemnon in the form of Nestor, assuring him that the gods had resolved their disputes, and that if the Greeks were to lay siege to Troy it would fall that very day (Il. 2.5–34). Agamemnon correctly recognizes the dream as divine, and therefore worthy of notice, but fails to consider that it may be intentionally deceptive. Dreams in Homer are important enough not to be ignored, but slippery enough not to be trusted. This ambivalence creates an epistemological pressure, acknowledged in the Odyssey when Penelope refers to dreams as ἀμήχανοι ἀκριτόμυθοι, ‘impossible, indiscernible’ things. Famously, she proceeds to categorize them into two types: those that pass through the gates of horn (κέρας) and fulfil (κραίνω) true things, and those that pass through the gates of ivory (ἐλέφας) and deceive (ἐλεφαίρομαι; see Od. 19.560–7). Her wordplay suggests a recourse to etymology in an attempt to reify a material taxonomy for what is, in practice, impossible to distinguish.Footnote 6 In the Homeric system, to the extent that we can characterize it as such, dreams all have an external, divine origin, but they are intrinsically difficult to trust and interpret because sometimes the gods intentionally mislead. No empirically reliable method could be used to discern one from another until they are confirmed or denied by events.
The belief that some dreams have an origin outside of the subject remains a strong and persistent one throughout Greek literary and philosophical thought,Footnote 7 and as notions about dreaming shifted categories tended to multiply rather than to replace this early concept. When the Hippocratic author of On Regimen posited that certain dreams arise from the internal condition of the body, this did not supplant the more traditional notion, but rather rested uncomfortably beside it. He relegated those dreams that have origins external to the subject and bear information from the gods to the domain of other specialists:
ὁκόσα μὲν οὖν τῶν ἐνυπνίων θεῖά ἐστι καὶ προσημαίνει ἢ πολέσι ἢ ἰδιώτῃσι ἢ κακὰ ἢ ἀγαθὰ, εἰσὶ οἳ κρίνουσι περὶ τῶν τοιούτων τέχνην ἔχοντες.
(Vict. 4.87)Regarding those dreams which are divine, and foretell to cities or to private people things that are bad or good, there are interpreters who possess the expertise to deal with such things.
The dream types in this system are so different from one another that each requires practitioners of an entirely different technē to properly interpret them, but the author provides no standards by which one might distinguish one kind of dream from another.
A decade later, Herodotus relates a story that illustrates the difficulty of differentiating dream origins. When considering whether or not to invade Greece, Xerxes has a recurring dream of an apparition urging him to carry out the expedition, and he is convinced that it is a message from the gods. When he tells his uncle Artabanus, who has been trying to dissuade him from the campaign, Artabanus replies:
ἀλλ᾽ οὐδὲ ταῦτά ἐστι, ὦ παῖ, θεῖα. ἐνύπνια γὰρ τὰ ἐς ἀνθρώπους πεπλανημένα τοιαῦτά ἐστι, οἷά σε ἐγὼ διδάξω, ἔτεσι σεῦ πολλοῖσι πρεσβύτερος ἐών· πεπλανῆσθαι αὗται μάλιστα ἐώθασι αἱ ὄψιες τῶν ὀνειράτων, τά τις ἡμέρης φροντίζει. ἡμεῖς δὲ τὰς πρὸ τοῦ ήμέρας ταύτην τὴν στρατηλασίην καὶ τὸ κάρτα εἴχομεν μετὰ χεῖρας.
(Hdt. 7.16)But these things are not divine, my son. For the dreams that wander upon men are of such a sort as I will teach you, being older than you by many years. The visions of dreams that are most accustomed to visit are those things one thinks about during the day. And we have recently been very much engaged with this campaign.
The difficulty in immediately distinguishing the quotidian and the divine dream is revealed when Artabanus is proven wrong. When he dresses in Xerxes’ robes and sleeps in his bed, the recurring vision appears to him as well, proving that its origins are in fact external and therefore divine. Arguably this story does suggest a method for distinguishing between dream types, but with obvious limitations, including the fact that the dream must be a reliably recurring one.
A fragment of Herophilus, the anatomist of the Hellenistic period, arguably presents the earliest truly systematic taxonomy of dream origins, a tripartite system that will go on to be influential in Stoic and early Christian dream theory.Footnote 8 His three categories were dreams inspired by the gods (θεοπνεύστους), ‘natural’ dreams that arise from the disposition of the body (φυσικούς), as well as a third category, ‘compound’ dreams that serve as wish fulfillment and that have external, but not divine, origins (συγκραματικούς) (Ps.-Plu. Placit. 904f6–8).Footnote 9 This is a brief fragment, more concerned with aetiology than with potential therapeutic applications, and it does not address methods of distinguishing dreams; the theory of dream production implied by the fragment will be discussed below. But it is worth noting here that, as this system was adopted by Christian thinkers, the third category transforms from benign interaction between bodily composition and sense impressions into malignant demonic dreams, sent to deceive the subject, suggesting continued concern about the perils of misidentifying a dream's origins.Footnote 10
A direct reference to the difficulty of dream classification in an explicitly medical context occurs in the brief Galenic treatise on dreams.Footnote 11 Here, the difficulties concerning proper dream taxonomy, particularly in regard to the ability to distinguish physiological dreams from prophetic dreams, and therefore to know what type of interpretive tools to apply to them, are not just implied but clearly stated:
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. εἰ μὲν γὰρ ἀπὸ τῶν ἐφ’ ἡμέρας πραττομένων ἢ φροντιζομένων ἔδει διακρίνειν αὐτὸ μόνον, οὐδὲν ἂν ἦν χαλεπὸν, ὅσα μηδὲ πέπρακται, μηδὲ φροντίζεται, ταῦτα ἐκ τοῦ σώματος ὁρμᾶσθαι δοκεῖν· ἐπεὶ δὲ καὶ μαντικά τινα συγχωροῦμεν εἶναι, πῶς ταῦτ’ ἂν διακριθείη τῶν ἀπὸ τοῦ σώματος ὁρμωμένων, οὐ ῥᾴδιον εἰπεῖν.
(Gal. Insomn. 833)But since in dreams the soul produces impressions based not only on the dispositions of the body, but also from those things customarily done by us every day, and some from things we have thought—and indeed some things are revealed by it prophetically (for this also has been witnessed in experience)—the diagnosis of the body from the dream visions arising from the body becomes difficult. For if it were necessary only to distinguish this cause from the things done or thought day by day, it would not be at all difficult to conclude that whatever has not been done or thought is arising from the body; but since we agree that there are also some prophetic dreams, it is not easy to say how these might be distinguished from the ones arising from the body.Footnote 12
Galen goes on to give an example of the sort of error that can arise from dream misclassification, and entrusting the wrong type of dream to the wrong type of interpreter:
ἐθεάσατο γοῦν τις τὸ ἕτερον τῶν σκελῶν λίθινον γεγονέναι, καὶ τοῦτο ἔκριναν πολλοὶ τῶν περὶ τὰ τοιαῦτα δεινῶν ὡς πρὸς τοὺς δούλους τείνειν τὸ ὄναρ, ἀλλὰ παρελύθη τὸ σκέλος ἐκεῖνο ὁ ἄνθρωπος, οὐδενὸς ἡμῶν προσδοκήσαντος τοῦτο.
(Insomn. 833f.)Someone dreamed that one of his legs had turned to stone, and many of those clever about such matters interpreted that this dream pertained to his slaves—but in fact he became paralyzed in that leg, although none of us had expected that.
‘Those clever about such matters’ likely refers to popular interpreters of prophetic dreams, as the correspondence between legs and slaves can be found in the symbology of non-medical Byzantine divinatory Oneirocritica.Footnote 13 The source of their mistake was that they misidentified the dream's origin. They applied a set of hermeneutics that would have been appropriate in the case of a prophetic dream, but, as this dream was in fact conveying physiological information, their efforts failed. As was the case in On Regimen, distinguishing between types of dreams was crucial not only because different types call for different interpretive methods, but because they call for interpreters of different professions. Physiological and divine dream interpretation of medically significant dreams coexisted, but the domains they occupied were so different that they called for completely different specialists.
With G.E.R. Lloyd leading the charge, the last few decades have seen a salutary pushback against the anachronistic tendency to mark a sharp divide between ‘religious’ and ‘secular’ medicine, and many have shown that in reality the systems are much more dynamic and interactive. Dreams represent a conceptual nexus between these worlds. Oberhelman summarizes this potential, when he writes:
In sum, ancient physicians held dreams—divine, mantic and physical—to be valid sources of information regarding an individual's health. Dreams were, to them, natural phenomena, even when they were divine, and, as such, they belonged to natural science. It was not happenstance, therefore, that dreams were the special province of Asclepius, the god of doctors, himself the epitome of the rational theology of ancient medicine, for dreams were the synthesis of nature and divinity and, accordingly, they became a standard prognostic and diagnostic tool of the Greek and Roman physician.Footnote 14
However, our scant evidence in fact suggests that this may not have been such an easy synthesis. Rather, it would appear that physiological dreams potentially served as a field in which the similarities and differences between mantic and medical professions had to be openly defined, and the medical differentiated from the mantic. The fact that this was shared territory, and that it was very difficult to know which specialist was suited to evaluate a particular dream, may have in fact contributed to preventing dreams from becoming ‘a standard diagnostic and prodiagnostic tool of the Greek and Roman physician.’ The power and potential was acknowledged, but the taxonomic difficulties, and the concerns that this provoked about specialization and division of labor, was perhaps one reason that the method apparently failed to be widely adopted.Footnote 15
The Problem of Interpretation
Another layer of difficulty in medical dream diagnosis was that, even if the type of dream could be determined, the correspondence between dream content and physiological affliction was rarely as straightforward as the Galenic example of a pre-paralytic leg turned to stone. Consider the following example from Rufus of Ephesus:
ἄλλῳ δέ τινι ἐν πυρετῷ ὀξεῖ πολλάκις ἀνὴρ Αἰθίοψ ἐπιφοιτῶν κατὰ τοὺς ὕπνους παλαίειν ἐδόκει καὶ ἄγχειν αὐτόν. καὶ οὗτος εἶπε πρὸς τὸν ἰατρὸν τὸ ἐνύπνιον. ὁ δὲ οὐδ’ αὐτὸς ἐνεθυμήθη οἷόν τι ἦν, πρὶν αἱμορραγίᾳ λάβρῳ ἐκ ῥινῶν ἐκρίθη ἡ νόσος.
(QM 31)An Ethiopian man often seemed to invade the dreams of a man afflicted with an acute fever, and to wrestle and choke him. This man even told the dream to his physician, but the physician was unable to understand what the dream was about until the illness came to a crisis with a violent hemorrhage through the patient's nostrils.
Rufus goes on to assert his confidence in the diagnostic potential of dream analysis:
πάνυ δὲ ἐμαυτὸν πείθω κατὰ τοὺς χυμοὺς τοὺς ἐν τῷ σώματι δόξας ἐνυπνίων ἐγγίγνεσθαι σημαινούσας καὶ ἀγαθὰ καὶ κακὰ τῷ ἀνθρώπῳ, ὧν κατάληψις ἄλλη οὐκ ἔστι μὴ ἀκούσαντι.
(QM 33)I am completely convinced that it is in accordance with the humors in the body that the visions of dreams arise, indicating things both good and bad for a person, things of which there is no other comprehension except to one hearing [of these visions].
But it requires of the physician a great deal of conjecture to determine to what, precisely, these ‘visions’ might refer. In this example, both the patient and the doctor clearly thought that the dream was a potentially relevant diagnostic tool, but it was not until a physically observable symptom emerged that its true meaning retrospectively became clear. In order to properly interpret the content of dreams, one needs to have an understanding of the mechanism by which they are produced, how the dreaming faculty of the subject translates, say, an excess of blood into an aggressive Ethiopian.
This leads us to an examination of the theoretical descriptions of the production of physiological dreams, of which there are only three extant articulations in the Greco-Roman medical corpus: the two more fully elaborated theories in On Regimen and On Diagnosis from Dreams, and the theory suggested by the terminology of Herophilus’ fragmentary dream taxonomy.Footnote 16 I will first address the theories of Herophilus and Galen. The theory of On Regimen is a special case, and this will be easier to elucidate after demonstrating the common ground between the other two.
In the Herophilus fragment already referenced above, the vocabulary used to describe the production of different types of dreams is strikingly similar to that used by literary critics to describe methods of poetic production:
̔Ηρόφιλος τῶν ὀνείρων τοὺς μὲν θεοπνεύστους κατ’ ἀνάγκην γίνεσθαι, τοὺς δὲ φυσικοὺς ἀνειδωλοποιουμένης τῆς ψυχῆς τὸ συμφέρον αὑτῇ καὶ τὸ πρὸς τούτοις ἐσόμενον.
(Ps-Plut. Placit. 904f6–8)Herophilus says that some dreams are god-inspired and arise by necessity, while others are natural and arise when the soul forms for itself an image of what is to its own advantage and of what will happen next.Footnote 17
The word θεόπνευστος (‘god-inspired’) is evocative of the long Greek tradition of external divine poetic inspiration, which is characterized by minimal creative input from the poet himself. When the Muses breathe a divine voice into Hesiod on Mt. Helicon, they simultaneously emphasize his own debased status as a rustic shepherd (Th. 26–32). He serves as an empty vessel to be filled. In Plato, breath is specifically associated with the phenomenon of poetic enthusiasm (R. 499c, Lg. 811c, Men. 99d).Footnote 18 In the Ion, Socrates uses the analogy of a chain of metal rings that serve as a conduit for magnetic force to describe the way that poetry is transmitted from the Muses to the poet to the rhapsode, in a model that similarly emphasizes the passivity of the poet. Another such analogy is given in the Laws:
ὅτι ποιητής, ὁπόταν ἐν τῷ τρίποδι τῆς Μούσης καθίζηται, τότε οὐκ ἔμφρων ἐστίν, οἷον δὲ κρήνη τις τὸ ἐπιὸν ῥεῖν ἑτοίμως ἐᾷ.
(Lg. 719c)Whenever a poet sits upon the tripod of the Muse, he is not in his senses. He is like a fountain where the water is allowed to gush forth unchecked.
In the Herophilus fragment, the passivity of the subject in the case of divine, prophetic dreams is similarly emphasized by the language of force, the assertion that the experience of these dreams is not optional, but κατ’ ἀνάγκην, ‘by necessity’.
The emphasis on the dreamer's passivity in the case of divine dreams is in contrast with the process involved in the production of a physiological dream, which requires much more active, interpretive participation of the subject. The word used to describe the activity of the soul in this process is ἀνειδωλοποιουμένη, (or, in another version of the Herophilus fragment, the essentially synonymous εἰδωλοποιουμένη) referring to the process of ‘forming an image’ in one's mind. Both of these words typically appear in literary critical contexts, where they describe a process of visualization that a poet or prose writer can use to imbue their work with vivid imagery.Footnote 19 It is an active and intentional part of the authorial creative process.
Galen's description of the creative production of physiological dreams shares much with Herophilus’ account:
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. καὶ εἴπερ ἔχει ταῦθ’ οὕτως, οὐδὲν ἂν εἴη θαυμαστὸν, ὁπότε μὲν ὑπὸ πλήθους χυμῶν ἡ ψυχικὴ δύναμις ἐνοχλεῖται βαρυνομένη, μόγις μὲν κινουμένους ἑαυτοὺς κατ’ ὄναρ φαντάζεσθαι καὶ βαστάζοντας ἄχθη τινά· τοὐναντίον δὲ, ὅταν ᾖ κούφη τε καὶ ἀπέριττος ἡ τοῦ σώματος διάθεσις, ἤτοι πετομένους, ἢ θέοντας ὠκύτατα τοὺς οὕτω διακειμένους ὁρᾷν ὄναρ…
(Insomn. 833)For it seems that in sleep the soul dives into the depths of the body and, having separated itself from external perceptions, perceives the disposition throughout the body, and forms an impression of all the things it touches upon, as though they were actually present. If this does indeed hold true, it should be of no surprise that when the psychic faculty is weighted down by a plethora of humors, in their dreams they see themselves barely able to move and to carry some burden; and also the opposite, whenever the disposition of the body is light and without excess, people of such dispositions will dream that they are flying or running very fast.
The activity of the soul in the production of physiological dreams has mimetic qualities that are similar to the description in Herophilus; but it will be helpful to digress briefly to consider the larger context of Galen's theories of perception, and how this account differs from his model of waking perception. The question of whether ‘appearance’ was a matter of the direct transference of external phenomena to the subject's hegemonic faculties, or whether judgment (δόξα) interceded in the process, dates back to Plato and Aristotle, and was central to the epistemological debates between Pyrrhonian Skepticism and Stoicism.Footnote 20 A strictly mechanical perceptive system, which relied on the interplay between psychic pneuma in the lens of the eyeball and the air outside of it, was a cornerstone in Galen's defense of empiricism.Footnote 21 By disallowing any sort of interpretive intervention in the process of perception, he rendered the results of his observational methodology more reliable. But for whatever reason, this tight mechanical system breaks down in the account of dreaming given here, where Galen explicitly describes an interpretive step between the soul's direct encounter with the phenomena and how they ‘appear’ to the dreaming subject.Footnote 22 First the soul ῾perceives᾽ (αἰσθάνεσθαι) the disposition of the body, and then it ῾forms an impression᾽ (λαμβάνειν φαντασίαν) of the things it has touched upon ‘as though they were actually present’ (ὡς ἤδη παρόντων). The examples he goes on to give, of the man weighted down by humors dreaming that he is carrying something heavy and the man with a light body dreaming of flying or running swiftly, show a process in which a phenomenon inside of the body is translated into the perception of a phenomenon or experience that takes place in waking life. Physiological dreams, therefore, do not represent the soul's direct perception of the body's disposition, but are the result of some creative production based on initial perceptions. It is worth noting that the word phantasia has uses not only in the realm of psychology but also in the realm of literary criticism, and the word's more creative valence may be operational here.Footnote 23
The idea that physiological dreams are the result of a creative process of the soul has tended to dictate modern scholarship's characterization of how the dreams would have been decoded by doctors to reveal the inside of the patient's body. Oberhelman, noticing the relationship between dream production and literary creativity described above, asserts that ‘[medical dream] interpretation was derived from one of several methodologies. These methods involve wordplay, metaphor, conformity of the dream-content to everyday life, and analogy (metonymy).’Footnote 24
Staden essentially agrees when he notes that ‘semiotic dreams were of course widely accepted by Greek physicians from the Hippocratics to late antiquity as providing useful “signs” (semeia) or “indications” (endeixis) of a given patient's condition or “dispositions” (diathesis).’Footnote 25 But the problems a physician might encounter with such a system are obvious: the ‘applicability of a dream-symbol is infinite’,Footnote 26 just as a literary symbol can be approached from an endless number of perspectives under the rubric of an endless number of methodologies. As we saw in the case described by Rufus of Ephesus, it seems that often these interpretations were effective only in hindsight, after they had already been proven or disproven by the emergence of physically observable symptoms. Unlike other symptoms that were directly observable to either the physician or the patient, the relationship between the physical situation and its perception was greatly complicated by the mediation of the creative faculty of the soul, this creativity here manifesting in the form of wordplay: the word Αἰθίοψ (‘Ethiopian’) descends etymologically from αἴθω and ὄψ (‘burnt face’). In his dream, then, the patient was being overpowered by something fiery, his body's way of warning him about an excess of hot blood. But there was so much room for error in the interpretation of this creative process that it was useless for successful diagnosis. While a diviner could claim a special knowledge of or relationship to a divinity that might explain his unique ability to interpret a god-sent dream,Footnote 27 the physician had no similar authoritative recourse in his interpretation of a dream of physiological origins.
In many cases, a proliferation of possible interpretive outcomes was not an obstacle to the perceived power of a diagnostic tool. Sphygmology, for example, gave Galen the opportunity to demonstrate his virtuosic abilities of observation and discernment. By endlessly refining his taxonomy of the pulse over hundreds of pages, creating an interpretive system that could not be replicated by aspiring rivals, he positioned himself as an absolute medical authority; the slipperiness of the pulse's signification was an aide rather than an obstacle to Galen's program of self-fashioning.Footnote 28 The difference in the case of physiological dreams is, I argue, that the site of interpretation is displaced. In the field of sphygmology, subtle variations in the pulse's speed, magnitude and frequency were thought to directly reflect the ‘condition of the innate heat and the body's residues’, and it was the task of the physician to observe and interpret this condition.Footnote 29 But in the case of dreams, the physician's usual role in the act of diagnosis, to encounter and interpret a patient's bodily disposition, has been supplanted by the patient's soul, which on its own encounters and interprets the disposition of the body in which it resides. The physician is left subordinate to this creative faculty's idiosyncracies, like the hapless physicians trying to understand the pulse by reading Galen's endless books rather than by directly observing and interpreting themselves.Footnote 30 It is relevant, then, that the author of the only extant and extensive catalogue of medical dream interpretation put forth a theory of dream production that eschews the creative model.
On Regimen IV: A Special Case
At this point in our discussion, it might be useful to turn the fundamental question around. Instead of asking why Galen and other doctors did not enthusiastically embrace diagnosis from dreams, perhaps we should ask how, in the world of the Hippocratic text On Regimen, it was espoused as an unambiguously powerful tool. I argue that though the theory of dream production put forth in that treatise in many ways resembles other Greco-Roman medical accounts, a small but fundamental distinction may have rendered the method more reliable than it appears to have been elsewhere. It will be helpful here to give a brief overview of the text's overall structure and formation. It is billed as a treatise on dietetics for the maintenance of health, but goes far beyond prescriptions for diet and exercise. The author asserts that, in order to properly treat human regimen, one must first understand the nature of man in general (Vict. 1.2). Furthermore, the author posits a highly elaborated micro-macrocosmic system in which various aspects and components of the human body correspond to aspects and components of the natural world, meaning that the text in fact gives an account not only of the nature of mankind but also of the forces and functions of the earth and the cosmos. Joly calls it ‘formule la plus précise et la plus claire pour exprimer la doctrine macro-microcosmique.’Footnote 31
The fourth book is devoted entirely to medical dream interpretation, and contains a catalogue of things a person might see or experience in their dreams, followed by a diagnosis of balance or imbalance in the body along with a dietetic prescription. This book was for a long time suspected to have been spurious, a haphazard collocation of Near Eastern and Greek folk dream symbology incorrectly appended to the rest of the text. But Jouanna has shown multiple threads of theoretical correspondence between Book IV and the rest of the treatise, including the fact that the circuits of the sun, moon and stars in Book I are directly reflected in the circuits of the body described in Book IV, and the work as transmitted is now generally considered to be genuinely unified.Footnote 32
Dambska summarizes an understanding of the mechanism of dream production in On Regimen that is attractive to many of its readers, particularly those of us living after the age of Freud:
Le Traité souligne le caractère symbolique de nos rêves—thèse chère aussi à la psychanalyse moderne. L'hypothèse fondamentale du Traité que l’âme, en se dérobant pendant le sommeil aux influences des sensations exterieures, prend connaissance à travers les symboles des rêves de l’état de son propre organisme, a été souvent reprise plus tard.Footnote 33
On this reading, the treatise faces the same potential semiotic problem as the other systems we have discussed but to a lesser extent, the difference between this treaty and the other ancient accounts of medical dream interpretation being a matter of degree rather than of quality. With more fully elaborated analogical relationships at its disposal, the soul creates easily interpretable symbols for the benefit of the diagnostician. Drawing this relationship between On Regimen and the other accounts, Oberhelman avers:
Analogy (metonymy, or literal association) was the most prevalent method of dream-interpretation among the medical writers… Here the dream was interpreted by drawing an analogy between the dream-contents (which represented the external world, or macrocosm) and the internal workings of the dreamer's body (microcosm)… The author of the Regimen 4 sets forth this principle clearly.Footnote 34
However, I will suggest a stronger thesis, namely that the relationship between the condition of the body and the content of the dreams in On Regimen is not in fact one of metaphorical correspondences, but one of identity. The micro-macrocosmic conception of the world is not, as it is often characterized, merely an analogy. It is a reality: that is, the correspondence between the circuits of the cosmos and the circuits of the body is not a symbolic or conceptual, but a material one. As Cambiano has noted, there is no suggestion that any ‘psychic work’ is taking place during dreaming in On Regimen, and to suppose so requires unnecessarily multiplying hypotheses.Footnote 35 The condition of the body as seen in dreams is not a representation but a direct perception of its disposition. This hypothesis is supported by the psychic dream theory expressed in the book's opening paragraph, which provides a different sort of description for the production of physiological dreams than the other two texts treated above:
ὅταν δὲ τὸ σῶμα ἡσυχάσῃ, ἡ ψυχὴ κινεομένη καὶ ἐγρηγορέουσα διοικεῖ τὸν ἑωυτῆς οἶκον, καὶ τὰς τοῦ σώματος πρήξιας ἁπάσας αὐτὴ διαπρήσσεται. τὸ μὲν γὰρ σῶμα καθεῦδον οὐκ αἰσθάνεται ἡ δὲ ἐγρηγορέουσα γινώσκει πάντα, καὶ ὁρῇ τε τὰ ὁρατὰ καὶ ἀκούει τὰ ἀκουστά, βαδίζει, ψαύει, λυπεῖται, ἐνθυμεῖται, ἑνὶ λόγῳ, ὁκόσαι τοῦ σώματος ὑπηρεσίαι ἢ τῆς ψυχῆς, πάντα ταῦτα ἡ ψυχὴ ἐν τῷ ὕπνῳ διαπρήσσεται.
(Vict. 4.86)When the body is at rest, the soul, being set in motion and awake, administers her own household, and of herself performs all the acts of the body. For the body when asleep has no perception; but the soul when awake has cognizance of all things—sees what is visible, hears what is audible, walks, touches, feels pain, ponders. In a word, all the function of the body and of the soul are performed by the soul during sleep.
Prima facie, the passage has much in common with Galen's theory described above. The soul, closed off from external perception, turns these faculties inwards. But note that there is no similar reference to ‘taking up a phantasia’, and nothing like Herophilus’ ‘making an eidos’. The soul is perceiving and performing the same way that it does while the body is awake. The soul travels through and perceives the body as the body travels through and perceives the outside world.Footnote 36 This is still a mediated activity in the way that any sense perception might be, but it is not a creative or interpretive one.Footnote 37
Indeed, Eijk notes some curious moments within the dream catalogue of On Regimen of ‘physical correlations between sign and significance’, including the debated line, ὅ τι δ ̓ἄν ἐν τῷ σώματι καθαρὸν ἐνεὸν ἐκκρίνηται ἐκ τῆς περιόδου κατὰ φύσιν ἀφ’ ἔσπέρας πρὸς ἠῶ, ὀρθῶς ἔχει (Vict. 4.89), that is, ‘If something in the body, being pure, is being secreted from its natural circuit from the west toward the east, this is good.᾽ He comments on the passage,
The words ‘in the body’, in combination with ‘from the west toward the east’ appear out of place here, unless they are taken to mean ‘something that represents something in the body’. Yet it seems more natural to interpret the text as saying that the dreamer is actually dreaming of something being secreted from his body, or that something is actually being secreted from his body while he is asleep.Footnote 38
He does not push the observation any further, the central concern of his paper being the relationship of the treatise to Near Eastern dream books, but the moment he points to indicates a striking collapse of the supposed semiological relationship between dream content and patient physiology. I argue that this is not a strange slip, but in fact characteristic of the entire system. The highly elaborated micro-macrocosmic vision of On Regimen is not an analogy, but an actual, literal correspondence. The three circuits of the body are not simply like the circuits of the cosmos, but are the same: the central circuit has the dynamis of the moon, the one in the middle the dynamis of the sun and the outer the dynamis of the stars (Vict. 1.10). And if, as has been affirmed by Jouanna, perception in this treatise is the result of a material mixing of the soul with ‘perceptive particles’ that are released from the objects of the outside world, there is no reason to assume that perception functions differently when the soul is turned inward.Footnote 39 As the author writes earlier in the treatise, ὁποῖα γάρ τινα πάσχει τὸ σῶμα τοιαῦτα ὁρῇ ἡ ψυχὴ κρυπτομένης τῆς ὄψεως (‘whatever things the body suffers, the soul sees these things when vision has been obscured’, Vict. 3.71).
Opsis here must refer to vision in its outward-facing, waking mode, while horaō refers to its inward-facing corollary, the soul's material encounter with the body's disposition. When a dreamer sees cosmic disorder, he is directly perceiving his disordered body.
This conflation of the body's experience in the world with the soul's experience in the body is a logical consequence of the materiality of the soul. There was an earlier communis opinio that On Regimen parroted Orphic-Pythagorean conceptions of mind–body dualism.Footnote 40 It has also been noted that its description of the relationship between the body and the soul, especially the assertion that ‘the soul is a servant (ὑπηρετέουσα) to the waking body’ (Vict. 4.86), parallels the image of the soul as a prisoner trapped inside the body in Plato's Phaedo. This has been convincingly refuted by many scholars, most recently Roberto Lo Presti.Footnote 41 In the first book of On Regimen, the author describes the soul as a part of the body: ἐσέρπει δὲ ἐς ἄνθρωπον ψυχὴ πυρὸς καὶ ὕδατος σύγκρισιν ἔχουσα, μοίρην σώματος ἀνθρώπου (Vict. 1.7), that is, ‘A soul enters into a human, having a mixture of fire and water, a part of the body of the human.’ Later the author goes on to say that in different combinations this mixture produces different levels of intelligence or perceptive capacity, that is, φρόνησις (Vict. 1.35). Some characteristics of the soul can be altered by a change in regimen. Different modulations of food, drink and exercise can affect the mixture (σύγκρισις) of the soul and therefore change its fundamental quality. There are also immutable aspects of a person's disposition, dictated by the swiftness or slowness of the circuit along which the soul travels.Footnote 42
Both the mutable and immutable characteristics of the soul are dictated by strictly concrete, material contingencies. These contingencies dictate how the soul encounters things that lie in its path, which results in perception. As Lo Presti succinctly puts it in his analysis of this passage, ‘il corpo conosce.’Footnote 43 In a system in which the soul is the seat of intelligence and perception in the body, and perceives by encountering the perceptive particles that are released by various mixtures of fire and water in the outside world, it is logical that, when the soul is deprived of the externally oriented perceptions of the body and is turned inward, its interaction with the body's internal mixtures would produce the same kinds of perceptions as those that are produced when its faculties are turned outward. That is, since the cosmos and the body are materially the same, and the soul's sensate faculties are similarly material, the perception of the cosmos outside the body and the perception of the cosmos within the body should also be the same.
When the soul is turned outwards, during waking hours, it perceives the world around it: rivers, weather, other people, the celestial bodies. All of these entities are, as we know from the first book of On Regimen, complex mixtures of water and fire, the two fundamental elements that operate in a productive tension to create everything around us. An intrinsic aspect of this ontology is that the human body is no different. It, too, consists of a complex mixture of fire and water. And so when the soul turns inward in sleep, it makes sense that it would perceive these same elements: rivers, weather, other people, the celestial bodies. This is not an analogical model, or an interpretive symbolic one, but rather a single, coherent system. In Book I, the author describes the relationship between the cosmos (‘the whole’) and the human body:
ἑνὶ δὲ λόγῳ πάντα διεκοσμήσατο κατὰ τρόπον αὐτὸ ἑωυτῷ τὰ ἐν τῷ σώματι τὸ πῦρ, ἀπομίμησιν τοῦ ὅλου, μικρὰ πρὸς μεγάλα καὶ μεγάλα πρὸς μικρά· κοιλίην μὲν τὴν μεγίστην, ξηρῷ καὶ ὑγρῷταμεῖον, δοῦναι πᾶσι καὶ λαβεῖν παρὰ πάντων, θαλάσσης δύναμιν, ζώων ἐντρόφων τροφὸν, ἀσυμφόρων δὲ φθορόν· περὶ δὲ ταύτην ὕδατος ψυχροῦ καὶ ὑγροῦ σύστασιν· διέξοδον πνεύματος ψυχροῦ καὶ θερμοῦ· ἀπομίμησιν τῆς γῆς, τὰ ἐπεισπίπτοντα πάντα ἀλλοιούσης.
(Vict. 1.10)In a word, fire arranged all things in the body in the same way as an imitation of the whole, small things with respect to large things and large things with respect to small things; the greatest cavity is a storeroom for the dry and the cold, to give to all things and to take from all things, the power of the sea, the nourishment of [nourished] animals, the destruction of those ill suited, and around this the composition of water both cold and wet; a passageway for cold and hot breath; an imitation of the earth, altering all things that fall into it.
This could easily be read as the articulation of another analogical model, complete with another suggestion (elicited by the term ἀπομίμησις) of a creative process somehow mediating between the world and the body. On this reading, the cosmos and the human body would exist in an ontological hierarchy; the human body would be some sort of copy or image of the rest of the world. Bartoš, following Burkert, has compellingly argued that such a reading relies on a Platonic (and therefore anachronistic) understanding of mim ē sis:
Modern Hippocratic scholars…still tend to read μίμησις in the Hippocratic treatises in the ‘Platonic’ sense (as I will call it for the sake of convenience): that is, as a relation of ontological priority, where one thing is imitated as a pattern or paradigm and the other imitates it as a copy (or a duplicate) derived from it.Footnote 44
But, as Burkert asserts, mimēsis appears to have a different valence in the Hippocratic texts, one that indicates similarity or identity between things or concepts but that does not require positing that one of these things is derived from another:
One may just as well say that the human body ‘imitates’ the cosmos as that the parts of the cosmos ‘imitate’ human organs. In the same way, either the arts imitate nature or nature imitates the arts. Imitation is a two-sided correspondence, which makes it possible to interpret separate things following the same pattern, but without implying differences of rank or a relationship of ontological priority.Footnote 45
Bartoš’s analysis of apomimēsis in On Regimen is less interested in the mechanism of the dreaming faculty per se, but he does note that this refinement of our interpretation of the Hippocratic use of the concept of mimēsis is helpful in understanding the discussion of diagnostic dreams in the treatise:
In order to interpret dream visions in relation to the bodily conditions, all we need to know is that the microcosmic and macrocosmic structures correspond to each other simply because they share certain universal principles immanently present at all levels of reality. The dreaming soul apparently follows these universal principles and visualizes the bodily states in the form of dream visions. We can remember our own visions, but, unless we are trained in dream interpretation, we cannot understand that they actually reflect our bodily state of health.Footnote 46
While Bartoš recognizes the continuity between what lies within the body and the world outside of it, he posits that the soul in On Regimen performs a similar sort of mediating ‘visualization’ process as it does in the theories of Herophilus and Galen. I argue that we should instead follow the text as written, and understand dream experiences as the result of direct perception. That is, a particular mixture of water and fire outside of the body is perceived as a planet in its proper orbit. The same mixture of water and fire within the body is perceived as a planet in its proper orbit.
The lack of an ontological hierarchy between the cosmos and the body, as well as the body and the soul, prevents the author of On Regimen from ever encountering the semiotic obstacles that may have made accurate diagnoses from physiological dreams difficult for the other ancient Greek and Roman physicians who bother to mention it. This is characteristic not only of the author's approach to dreams, but also to other kinds of understanding throughout the treatise. Interspersed between advice on how and when to eat, bathe, sleep and have sex are tautological, gnomic statements, equating not only the universe and the human body, as discussed above, but also everything with everything else. For example:
γενέσθαι καὶ ἀπολέσθαι τωὐτό, συμμιγῆναι καὶ διακριθῆναι τωὐτό, αὐξηθῆναι καὶ μειωθῆναι τωὐτό, γενέσθαι, συμμιγῆναι τωὐτό, ἀπολέσθαι, μειωθῆναι, διακριθῆναι τωὐτό, ἕκαστον πρὸς πάντα καὶ πάντα πρὸς ἕκαστον τωὐτό…
(Vict. 1.4)‘Becoming’ and ‘perishing’ are the same thing; ‘mixture’ and ‘separation’ are the same thing; ‘increase’ and ‘dimunition’ are the same thing; ‘becoming’ and ‘mixture’ are the same thing; ‘perishing’, ‘dimunition’ and ‘separation’ are the same thing, and so is the relation of the individual to all things, and that of all things to the individual…
The ontology elucidated in this text does not consist of a matrix of arbitrary or creative semiological relationships, but is rather a single system, a single reality that can either be known or not known, predicated on the single element of the productive interaction between fire and water (Vict. 1.3).
If the physician understands how the material cosmos functions—and he will, if he has read Book I—he will know how the body works, not by analogy but by extension. For this reason, the physician posited by On Regimen can be absolutely confident in his dream analysis. The author writes in Book I that ‘the characteristics of seercraft and of human nature are the same: for those who know, always rightly interpreted; for those who know not, sometimes rightly and sometimes not’ (Vict. 1.12). Similarly, in Book IV, when talking about interpreters of divine dreams, the author says that when they turn their attention to dreams that are a result of physical symptoms, they do so ‘sometimes with, sometimes without success; but in neither case do they know the cause, either of their success or of their failure’ (Vict. 4.87). Knowledge is not the result of a process of interpretation but an immediate and absolute understanding, predicated on our bodies’ metabolization of the material world.
The author of On Regimen’s resolutely materialist views, and his commitment to an understanding of the human being as a microcosm of the world exterior to it, allowed him to see the visions of dreams as an unparalleled window into the body's inner workings. Recognizing the exceptionalism of the dream system of On Regimen troubles the claim that diagnosis by dream was a standard element in the Greco-Roman physician's medical tool kit and brings to the fore the philosophical diversity that characterizes the Greco-Roman medical tradition. Furthermore, we find that a phenomenon long considered a point of unification for diverse approaches to caring for a patient's body may have instead provided a field upon which distinctions between methodologies were magnified.
