Among the three prototypical clinical presentations within the spectrum of consciousness, delirium (that is, wakeful impaired consciousness) is undoubtedly the most controversial. Unlike sleep and fainting, which the ancients presumably designated in a similar manner as we currently do, this condition has been less clearly defined and has always prompted more theoretical elaboration. Its changeable nature and ubiquity make the characterisation of delirium less stable, especially considering that it appears in multiple forms and can be present in (and caused by) many other mental or physical, acute or chronic conditions.Footnote ¹

First and foremost, it should be emphasised that delirium is nowadays regarded as a cluster of symptoms rather than a diseaseFootnote ² in its own right. In other words, we should not confuse the features of wakeful impaired consciousness with some specific underlying diseases that often trigger it (currently included within the psychiatric domain). It is in this regard that I will challenge several classicists’ simplistic solution of equating all descriptions of diseases involving cognitive compromise with mental illness. Far from attempting an extrapolation of modern categories into ancient texts, by addressing our current framing of these conditions I aim to make the reader aware of the bias with which we approach the sources. Even assuming the low likelihood of finding equivalences between contemporary and ancient theorisations, it is worth bearing in mind that nowadays mental illness and delirium describe utterly different medical realities. (I shall later discuss how this shift in understanding is not a mere change of names, but enables novel insights into terminological and philosophical issues crucial to ancient medical – and also legalFootnote ³ – thought.)

Current medical understanding and definitions

Our current construction of consciousness as a spectrum has already been mentioned in Chapter 1. It is the intermediate areas of this spectrum, particularly those where patients are awake and hyper- or hypoactive but unaware of or disengaged from their surroundings that can be easily confused with madness. It is, therefore, at these points that definitions need to distinguish these different kinds of affections.

My understanding of delirium/wakeful impaired consciousness follows the DSM 5, where it is defined as an acute and fluctuating syndrome characterised by inattention, disturbed awareness and cognition, and in more serious cases, sleep disturbances and hallucinations.Footnote ⁴ Delirium tends to be episodic, self-limited and changeable.

Mental illness, madness or insanity, on the other hand, are considered as an heterogeneous group of chronic conditions (that is, extended in time), which can present various clinical signs, including extreme moods, hallucinations, delusions, thought disorders and negative symptoms (such as apathy, inexpressiveness and lack of motivation).Footnote ⁵ Only the episodes of ‘acute psychosis’ – which sometimes occur in the course of these longer processes – can be equated with delirium, and hence considered as a specific form of impaired consciousness. Acutely psychotic patients seem out of touch with reality, confused, and any of the above-mentioned symptoms can be present.

It cannot be emphasised enough that delirium is neither equivalent nor specific to psychiatric disorders and can occur in many other conditions. Actually, more often than not it happens in the context of severe systemic diseases (infections, head trauma, intoxication, burns) and several neurological conditions. In other words, an episode of delirium can be a symptom of madness, but it is much more often a manifestation of other conditions, and – more importantly – mental illness cannot be diagnosed only through a single episode of delirium.

Another condition that compromises ‘the mind’ in a related and easily mistakable manner is dementia. Patients suffering from dementia are predominantly elderly, and their main characteristic is a chronic and progressive deterioration of the neurocognitive abilities, which interferes with their activities of daily living. Memory, language, visual-spatial skills, judgement and problem-solving capacities tend to be compromised. Serious cases can present depression, apathy, hallucinations, delusions, agitation, insomnia or sleep disturbances. In other words, dementia can also present specific episodes of impaired consciousness in certain moments during its prolonged course.

It seems quite evident from the above that the common feature shared by all these disorders, which sometimes makes differential diagnosis difficult, is the compromising – in one way or another – of the so-called higher order functions (HOFs). This notion subsumes numerous brain capacities including memory, verbal communication, perception, attention (which together allow judgement, decision-making and planning), as well as the bodily manifestation of the former; namely, behaviours, which are the ways in which human beings conduct themselves under certain stimuli, in determined situations or circumstances.Footnote ⁶

Invariably, delirium, mental illness and dementia affect one or several of these HOFs, hence (to use Aitchinson’s terms) their fuzzy edges, which have caused confusion between them, since antiquity up to the present day. As we shall see below, the Hippocratic doctors also had an intuition of these constructs, but contrary to what is implied in recent scholarship, they mostly commented on the compromise of HOFs in acute, changeable and short-lived situations, which we can easily identify as delirium or wakeful impaired consciousness. The later authors that we will analyse, conversely, did pay attention to more persistent and stereotyped disturbances of these HOFs. Nevertheless, they made important efforts to distinguish those chronic conditions from sudden episodic disorders and strived to recognise both types in order to provide a specific treatment for each of them.

Delirium as evidence of changing ideas about disease

Like several other concepts that we have been discussing, a distinct and universally accepted definition of disease can also be problematic, particularly because it acquires different nuances in different specialties. Nevertheless, any modern textbook that addresses medical conditions is, roughly speaking, divided into (at least) four main sections that shape our current idea of illness: clinical presentation, bodily location, aetiology – often related to abnormal functioning, damaged mechanisms or structures – and treatment. It is precisely from the interaction of these elements that the notion of illness or disease emerges. Our construct of any such distinct nosological entity presupposes a unity underlying all these components. Namely, doctors are always – explicitly or implicitly – searching for a cause that alters a certain physiological mechanism in the body, which triggers determined symptoms and which needs to be reverted through certain therapeutic measures to achieve a cure. On the contrary, when that unity fails to exist (or science has not yet discovered a logical explanation to link symptoms, organs, mechanisms and treatment), practitioners talk of syndromes, which are conceived as mere groups of signs and symptoms but not as diseases in their own right.

In order to illuminate the ancient writers’ understanding of the construct ‘disease’, the analysis will be scrutinising the extent to which they organised such elements into a coherent whole. In other words, by looking at the way in which ancient doctors articulated signs and symptoms with the affected parts of the body, the pathophysiological mechanisms and their therapeutic approaches, we will be able to gauge how the concept of disease changed in the medical discourse.

Impaired consciousness in the Hippocratic corpus: delirium case studies

Over half a millennium separates the literary emergence of Hippocratic medicine from the authors – such as Rufus of Ephesus and Aretaeus the Cappadocian – who usually play starring roles in scholarship on ‘mental disorders’ in antiquity. Similarly, the appearance of affections such as mania or melancholiaFootnote ¹ is a product of Hellenistic medical developments. In contrast, Hippocratic texts generally include forms of mental derangement and disturbance – delirium, coma, incoherent speech, hallucinations – as symptoms within a larger array of signs, but there are very few allusions to mental derangement as affections in themselves. This pattern reflects a general Hippocratic looseness in the notion of disease, a preference for listing signs rather than providing abstract definitions. These circumstances notwithstanding, scholars have tended to take these notes about episodes of delirium as proof of a larger category of madness underlying these glimpses, thereby drawing generalising conclusions about ancient psychology and body–mind relationships.

As we shall see below, even if the distinction between chronic mental diseases, on the one hand, and acute and changeable episodes of delirium, on the other, can be found in the corpus, the bulk of cases fall into the latter category. Let us examine three groups of disorders or case studies: two forms of delirium affecting young girls, two cases of alcohol-associated delirium and several discussions about phrenitis.

Symptoms, location, cause and treatment of delirium: the loose notion of disease in the Hippocratic Corpus

As highlighted above, impaired (or abnormal) perceptions and speech disorders – amid several other symptoms – are very characteristic of our current idea of wakeful impaired consciousness. When chasing these two major clinical manifestations across the subsequent authors and periods, it will become clear that not only ideas about impaired consciousness but also the understanding of the notion of disease among ancient doctors changed over time.

Undoubtedly, in the previous examples hallucinations were important for the Hippocratic doctors. The young girls ὁρῆν δοκέειν δαίμονάς τινας ἐφ’ ἑωυτῶν δυσμενέας (‘think that they can see spirits hostile to them’);Footnote ¹⁵ a phrenitic patient δείματα ὁρᾷ καὶ ὀνείρατα φοβερὰ καὶ τοὺς τεθνηκότας ἐνίοτε (‘sees terrible things, frightening dreams and sometimes, even the dead’);Footnote ¹⁶ and the author of On the sacred disease explains that when the brain is ill, μήτε τὴν ὄψιν ἀτρεμίζειν μήτε τὴν ἀκοὴν, ἀλλ᾽ ἄλλοτε ἄλλα ὁρᾷν καὶ ἀκούειν (‘neither sight, nor hearing remain still; instead sometimes we see or hear certain things, whereas at other times others’).Footnote ¹⁷ It cannot be emphasised enough that these abnormal perceptions are framed in the texts as only one clinical finding among various others, such as fever, pain, abnormal movements in the hands, photophobia, etc. We shall later see that their status and relevance in post-Hellenistic and Galenic accounts will increase.

On the other hand, speech disorders will suffer the exact opposite process: namely, they were consistently related to delirium in the Hippocratic collection, but their importance waned in later periods. Above is the example of Euryanax’s young daughter, whose hypoactive delirium – described as parelêreiFootnote ¹⁸ – caused her to not talk. I have elsewhere mentioned several other cases where the symptom was so strongly related to this condition that the terminology denoting senseless talking was used interchangeably with delirium (as opposed to normal speech, which was utilised to convey the idea of lucidity).Footnote ¹⁹

In terms of the bodily parts, although the locus affectus is a post-classical notion,Footnote ²⁰ Padel has remarked that among the tragic poets, and also in the Hippocratic corpus, there are several attempts at associating states of altered consciousness with some splachna or ‘innards’. In her insightful analysis, these were often identified with specific organs (heart, brain, liver, lungs), certain liquids (blood, bile) or airy substances (including the breath); they could also be assimilated to specific tissues (blood vessels, the diaphragm) or even to more abstract concepts (thumos, menos, nous, psuchê).Footnote ²¹ Regardless of their anatomic placement, which was different in different texts, they tended to be involved in or enabled feeling and thinking.

Accordingly, the author of On the sacred disease (Morb. Sacr. 14.5: 26, 14–15) attributes hallucinations to the brain, whereas in On diseases of girls the heart and the diaphragm as well as the poor state of the blood are responsible for similar phenomena.

What is interesting about this passage is the fact that the author relates specific symptoms to specific innards. Although he does not explain how the heart, the thumos or the state of the blood cause those hallucinations, there is a link between anatomical location and clinical manifestation, which approaches a description of our way of thinking about diseases. Similarly, according to On breaths, drunkenness is associated with changes in the blood:

This excerpt is only mentioning mood changes. Nevertheless, if alcohol mainly affects the volume of blood, thereby compromising the psuchai – which seems to be the core implication of this passage – we can presume that the alterations of consciousness described in the examples of the young drinkers should also be attributed to similar occurrences.

Concerning phrenitis, finally, some Hippocratic texts consider it to be caused by bile. The discrepancy, though, emerges regarding where it causes the damage. The author of the nosological treatise On affections considers that ἡ δὲ νοῦσος γίνεται ὑπὸ χολῆς, ὅταν κινηθεῖσα πρὸς τὰ σπλάγχνα καὶ τὰς φρένας προσίζῃ (‘the disease occurs due to bile, when it sets in motion and settles in the organs and the diaphragm (ta splachna kai tas phrenas)’),Footnote ²³ whereas for the author of On diseases I it needs to enter the blood:

Ultimately, the debate that all these Hippocratic authors are having concerns their idea of mind and its localisation (which – as Padel points out – is not alien to the philosophical questions that the tragic poets were debating at the same time).Footnote ²⁴

As far as mechanisms and treatments are concerned, we have already caught small glimpses in the previous discussion: delirium in young virgins stemmed from an abnormal accumulation of blood that affected the heart and diaphragm due to obstructed drainage from the uterus. Not surprisingly, the author acknowledged that the release came when the discharge of blood stopped being obstructed.Footnote ²⁵ The recommended treatment to achieve this, however, seems to be more related to sociological needs – he advised marriage and pregnancy – than to the biological mechanism that caused the disease.

In cases of drunkenness the author of On breaths mentioned the increase in the volume of blood, but neither blood-letting nor any other treatments are suggested.Footnote ²⁶ Regarding phrenitis, as McDonald has pointed out, the therapeutic approach is fairly non-specific.Footnote ²⁷ Authors tend to group the condition with other diseases,Footnote ²⁸ particularly with peripneumonia and pleuritis,Footnote ²⁹ and recommend general treatments for specific symptoms. Purgation, (upwards and downwards),Footnote ³⁰ warming (especially in the painful site) and drinks except for wine are recommended.Footnote ³¹ Interestingly pain is the doctors’ main concern and the main target of their treatment. Unlike what we will see in the post-Hellenistic sources, there are almost no remedies specifically aimed at delirium. Only the author of On affections posits that λούειν δὲ πολλῷ καὶ θερμῷ κατὰ κεφαλῆς ἐν ταύτῃ τῇ νούσῳ συμφέρει· … αὐτὸς αὑτοῦ ἐγκρατέστερος γίνεται (‘it is convenient to bathe those suffering of this disease in abundant hot water from the head downwards … for individuals become more in control of themselves’).Footnote ³²

As we have seen, the articulation between symptoms, affected parts, pathophysiological mechanism and treatment is rather loose at best, or non-existent in most cases. There are very few accounts that link symptoms to affected parts, and there seems to be a complete divorce between therapeutic methods and physio-pathological mechanisms.Footnote ³³ Consequently, it is safe to suggest that these doctors’ notion of illness – compared to ours and to their successors – is rather loose. In other words, impaired consciousness in general and phrenitis in particular are framed as clusters of related symptoms not clearly bound to their triggering mechanism and need to be treated like other acute conditions, not too far from our own ideas about impaired consciousness, but quite different from our notion of disease.

Among the post-Hellenistic sources the opposition impaired consciousness–mental illness becomes more explicit. Modern scholars, however, have persevered in ignoring impaired consciousness. As far as the notion of disease is concerned, the different components that we have seen as barely theoretically related in the Hippocratic corpus will be more strongly linked among these authors.

Celsus

Celsus’ On medicine offers a good example of the above. It has triggered abundant discussion about madness where scholars have completely ignored consciousness.Footnote ¹ The reason for this omission is that the modern debate tends to focus on insania as a whole but disregards the context in which it appears, especially the relations that the author establishes with the conditions addressed in the subsequent chapters. Indeed, Celsus presents delirium within a system of oppositions that aims at distinguishing it from mental illness, on the one hand, and from other forms of impaired consciousness, on the other.

The whole discussion is presented in the first part of the nosology section of On medicine (book 3). In it, chapter 18 tackles insania, an umbrella-term comprised of three nosological entities. Although a superficial reading might suggest that they are equally relevant, the first of them – phrenesis – has a different status. Not only does it take up most of the explanation, but it is also the only one explicitly referred to by its name, and whose relations to other conditions are clearly established.

Phrenesis is initially contrasted with the second and third types of insania – possibly melancholia and mania, which we would nowadays classify as mental illnessFootnote ² – and subsequently, it is the only one of the three emphatically opposed to the conditions addressed in the following chapters, namely, cardiacumFootnote ³ and lethargy.Footnote ⁴ siquidem mens in illis labat, in hoc constat (‘While in [phrenesis] the mensFootnote ⁵ gives in, in [cardiacum] it endures’),Footnote ⁶ and in eo difficilior somnus, prompta ad omnem audaciam mens est: in hoc marcor et inexpugnabilis paene dormiendi necessitas (‘whereas in [phrenesis] sleeping is more difficult, and the mens is prone to any kind of insolence; in [lethargy] there is torpor and an almost overpowering need of sleep’).Footnote ⁷ It should be highlighted that the latter two diseases are the closest ancient equivalents to the other prototypes of impaired consciousness that I have chosen. As a matter of fact, among the post-Hellenistic writers phrenitis became the prototypical example of wakeful impaired consciousness, lethargy of drowsy impaired consciousness and cardiacum of total loss of consciousness.

The fact that Celsus feels the need to contrast only the first type of insania – that is, phrenesis – to these other forms of impaired consciousness constitutes a hint that such conditions were somehow related, or at least easily confused (whereas the second and third types of insania were perceived as more distinct entities).

The strongest hint to relate phrenesis to delirium, however, comes from the distinction among the different kinds of insania. Differential diagnosis between these three entities is based on the presence of fever, the length of the condition and some specific symptoms, particularly hallucinations. Indeed, the first kind of insania, phrenesis, et acuta et in febre est (‘both acute and with fever’), and causes deceptive apparitions (vanas imagines) to which the mens surrenders.Footnote ⁸ These patients can be alii tristes sunt; alii hilares; alii facilius continentur … alii consurgunt et violenter quaedam manu faciunt … (‘sad, others cheerful, some are easily controlled … and some resist and act violently’).Footnote ⁹ The second genus of insania oppresses for a ‘longer period because it usually begins without fever’ and ‘consists of sadness that black bile seems to bring about’.Footnote ¹⁰ The third kind, finally, is the ‘longest of all. So much so that it does not even compromise life, for it tends to occur in strong bodies’.Footnote ¹¹ Some of the affected patients are ‘deceived not by their mens, but by imagines like those perceived – according to the poets – by Ajax or Orestes, when they were insane’, whereas in others ‘the animus becomes void of understanding’.Footnote ¹² In this description (that is, in the characterisation of the second type of the third form of insania), mens, animus and consilium are used interchangeably.

The acute nature of phrenesis, the presence of fever and the various kinds of hallucinations (vanas imagines) that cause all sorts of disturbed behaviour – that is, impaired responsiveness – once the mens gives in to them, make it easy for us to identify it with the Hippocratic phrenitis, and therefore with a condition where (wakeful) loss of consciousness predominates. These coincidences notwithstanding, it should be noted that Celsus, probably influenced by the ongoing debates, feels the need to distinguish between this condition and others that no Hippocratic writer mentions (at least not as such, even accepting the leading hypothesis in scholarship that the second and third kinds of insania correspond to mania and melancholia, respectively).Footnote ¹³ Beyond the lack of fever and the more chronic course of the second and third subtypes, Celsus finds the cognitive disturbances themselves to be distinct in each condition. Whilst phrenesis can trigger virtually any kind of hallucination (thereby affecting connectedness to the environment) and compromise responsiveness (as any episode of delirium does), the other two are more stereotypical. The variant caused by black bile can only produce pathological sadness, whereas patients with the third type of insania either have their animus compromised or suffer from imagines. Yet, the latter are different from those caused by phrenesis because they do not affect the mens. Judging by what we know about Ajax and Orestes, we can suppose that those apparitions resemble what we would nowadays define as a structured delusion, that is, a condition where reasoning is preserved but the patient holds elaborated beliefs and ideas, which are stable in time and not shared by the others (quite different from the more chaotic thought disorders of acute deliriums).

According to Stok, On medicine is the first extant medical text where these three ‘psychiatric’ conditions appear related.Footnote ¹⁴ Beyond the fact that phrenesis is stricto sensu not a psychiatric condition, it is likely that the discussion of these three illnesses and their distinction was very much part of the post-Hellenistic debate. A hint to support this hypothesis can be found in the pseudo-Galenic Introduction (Eisagogê). Although the compiler of this introductory handbook discusses phrenitis among acute diseases, and melancholia and mania among the chronic ones, he defines phrenitis as ἔκστασις διανοίας μετὰ παρακοπῆς σφοδρᾶς (‘disruption (ekstasis) of the dianoia with strong delirium (parakopê)’),Footnote ¹⁵ and the other two – which are different forms (eidê) of the same illness – as περὶ τὴν διάνοιαν ἐκστάσεως (‘dianoia disrupted (ekstaseôs)’).Footnote ¹⁶ Considering the coincidence in terminology, it is not unlikely that these diseases were perceived as related entities, and it was the dichotomous classification (acute versus chronic) which prevented the author from discussing them together.Footnote ¹⁷ In other words, the difficulty of separating acute delirium from chronic mental illness is evidenced in other post-Hellenistic works.

In summary, according to Celsus, wakeful impaired consciousness is described in opposition to other forms of impaired consciousness (cardiacum and lethargy); it is contrasted to mental illness (second and third types of insania), and presented as an array of acute symptoms that change substantially during their course. As I will argue, the notion of disease that emerges from Celsus’ approach to delirium is strongly influenced by this system of oppositions.

Aretaeus

Like Celsus, Aretaeus’ work supports the post-Hellenistic view of impaired consciousness and mental illness as easily confusable phenomena that needed to be separated through opposing features.Footnote ³⁰ His attempts at distinguishing phrenitis from melancholia and mania are proof of this.Footnote ³¹ Regrettably, only the chapter on the cure of phrenitis is extant (unlike the one on causes and symptoms); therefore, most conclusions will be based on passages drawn from it, as well as on various scattered allusions that Aretaeus made when he addressed other conditions.

This passage offers a clear distinction between the cognitive impairment that characterises phrenitis, as opposed to the one that occurs among maniacs (and also melancholics, whom – like the author of the pseudo-Galenic Introduction – Aretaeus considered to be suffering from a different form of the same disease, SD I.3. CMG (H).III: 39, 28). In phrenitis, the abnormal behaviour is a consequence of impaired perceptions or disturbed connectedness (namely the hallucinations) but alertness is preserved, whereas in the others connectedness to the environment is intact, and the primary problem is their altered responsiveness due to impairment in their judgement. Generally speaking, both melancholia and mania present an altered responsiveness that manifests as abnormal behaviours, extreme emotions and delusions. Unlike the phrenitic delirium, these delusions are structured and persistent beliefs, which are not prompted by wrong perceptions, but by bad judgement (alertness is damaged).Footnote ³² Examples abound: those who mistrust remedies (SD I.5. CMG (H).III: 40, 1–2); the person who believed himself to be a brick and avoided water for fear of dissolving (SD I.6. CMG (H).III: 42, 19–20); the builder who could not be away from the building site (SD I.6. CMG (H).III: 42, 20–9); mystic delusions involving self-mutilation ordered by the gods (SD I.6. CMG (H).III: 43, 40–1; 44, 1), etc. It should be noted, however, that within these chronic conditions there are some specific moments where the author seems to be describing an acute delirium: μετεξέτεροι δὲ καὶ παραισθάνονται, παραφορῇ τῆς αἰσθήσιος (‘some suffer from illusory perceptions and disturbances in their senses’).Footnote ³³ Nevertheless, this is still consistent with our understanding of mental illness, where some phases of acute psychosis can occur.

There is yet another clear delimitation between the domain of wakeful impaired consciousness/delirium and delusion when Aretaeus defines mania:

In this passage, beyond the longer duration of mania – which is implicit in its classification as a chronic disease – Aretaeus uses the same delirium vocabulary that he had used in phrenitis to refer to intoxications with wine and psychoactive herbs (paraphorê, ekmainei).Footnote ³⁴ Again, like the young lads from the Hippocratic corpus, drunkenness seems to be – at least terminologically – close to phrenitis. Moreover, not only should mania be distinguished from these acute forms of wakeful impaired consciousness, but it also differs from lêrêsis. This condition (which we would nowadays probably consider as akin to dementia) has become a nosological entity in its own right, where there is αἰσθήσιος γάρ ἐστι νάρκη καὶ γνώμης νάρκωσις ἠδὲ. τοῦ νοῦ ὑπὸ ψύξιος (‘numbing of perception and altered gnômê or nous due to coldness’).Footnote ³⁵

As stated in Chapter 1, Galen’s system was centred in the intersection between an anatomical axis, which determined the site to which the treatment needed to be applied, and a humoural-physiological axis that guided the quality that needed to be allopathically counterbalanced, as well as the quantity of the correction required. This shift of focus, from the Hippocratic emphasis on clinical descriptions, and the post-Hellenistic stress on nosologic taxonomy, conditioned a different understanding of impaired consciousness and mental illness.Footnote ¹ Although Galen did not disregard diseases and symptoms, in his approach these elements are subordinated to his main concerns: the organ affected, the type of duskrasia and the degree of impairment in the qualities (required by his system to choose an adequate treatment). Such a change of emphasis had powerful consequences, for it often blurred the boundaries between impaired consciousness and madness.

The efforts that we have seen in On medicine and in Aretaeus’ work (also in the Introduction and the Medical definitions) to distinguish phrenitis from other forms of mental conditions – particularly from melancholia and mania – will change.

Although I consider that Galen’s approach strongly distances him from his forerunners’, this passage suggests that he still engaged in the debate around boundaries between different forms of impaired consciousness (phrenitis–lethargy), and the fuzzy edges that separated the former from mental illness (mania and melancholia). Additionally, the passage highlights how the distinction between chronic and acute diseases, which was a key organising feature in the post-Hellenistic authors to differentiate these conditions, is irrelevant for Galen. The only time-reference included is the continuous nature of the symptoms in primary diseases of the brain.

Actually, Galen’s main concern here seems to be the localisation of the disease. It is crucial for him to know whether the damage was produced in the brain or somewhere else in order to decide where to apply the treatment. In a very similar passage from On the causes of symptoms he adds:

In this case the aim is to explain the kind of humours involved in each condition. Because both phrenitis and mania happen in the brain, and are caused by similar kinds of humours, they will very likely receive a similar treatment. As a result, for Galen it is not as important as it was for the other post-Hellenistic doctors to distinguish one from the other, and therefore, he will not put as much effort as they did into trying to expose the difference. As a matter of fact, allusions to mania are very scarce in the works under scrutiny.Footnote ² Only the presence or absence of fever is enough to put a label to either condition, which – in all probability – will anyway be treated in a similar manner. On the other hand, melancholia is caused by the opposite mixture and therefore will require the opposite treatment. Hence, as we will see later, it is worth distinguishing this condition from the other two.

Going back to the dichotomy phrenitis–mania, Galen classifies as the former several conditions that other authors had associated with the latter.

The extreme types are illustrated by a corresponding example: the first one involves a patient whose perceptions were sound enough to recognise the objects that he threw through the window, but whose judgement did not allow him to acknowledge the senselessness and danger of what he was doing. For the second kind the author uses his own experience with phrenitis, when he started to have visual hallucinations but was still intellectually sound enough to tell his friends what to do to cure him of the disease. The similarities between these two typologies of phrenitis, and the other authors’ contrasts between mania and phrenitis, respectively, are not difficult to spot, particularly if we apply the analysis of connectedness and alertness that was used in Aretaeus (indeed the first variety of Galenic phrenitis is conceptually equivalent to Aretaeus’ definition of mania in SD I.6. CMG (H).III: 43, 1–4).Footnote ³ Clearly, Galen was not particularly interested in clinically distinguishing two diseases that required the same treatment, even if he was able to clinically differentiate the diverse nature of the symptoms.

On the other hand, in diseases that warranted an alternative therapeutic approach due to their different pathophysiology or bodily localisation, the distinct clinical presentations are more explicitly stated. Such are the cases of melancholia and various conditions with fever. In the case of melancholia, there is a thorough description of the kind of mental disturbance that one should expect.

Naturally, if mania, phrenitis and melancholia all occur in the brain, and they all involve hallucinations or impaired judgement, but the former two are caused by hot and dry humours, whereas the latter is due to cold and moist ones, it is crucial to differentiate them, for they will have opposite treatments. This might be the reason why Galen’s description of melancholia is so thorough, while beyond the fever, mania and phrenitis are barely distinguished. Concerning the specific symptoms, it is interesting to highlight that – like with Aretaeus – the kind of mental disturbances that Galen is describing would be now considered as delusions rather than hallucinations or delirium, for they seem to be structured, stable in time, and apparently, reasoning is not affected (judgement is). Furthermore, the story about the man who believed himself to be a pot and was afraid of being crushed reminds us of Aretaeus’ patient who thought he had become a brick and avoided water, for fear of dissolving.

Apart from melancholia, the other diseases that could be easily confused with phrenitis – and therefore needed to be more accurately distinguished – were several other conditions with fever and delirium, which did not originate in the brain.

Galen’s clinical description in these passages is thorough. He goes through all the phrenitic symptoms,Footnote ⁴ which mostly coincide with the descriptions of the disease that we have found in the other surviving post-Hellenistic medical writers. The nuance comes in Galen’s practical approach. As stated above, he only goes to great lengths to distinguish between conditions whenever such a distinction impacts on the treatment. In this case, the locus affectus is at stake: by being able to tell phrenitis from pleuritis, peripneumonia or ‘diaphragmatic near-phrenitis’, he will know the correct part of the body to which the treatment needs to be applied.

Finally, this focus on anatomical location and physiologic mechanism is even more evident when Galen contrasts phrenitis with drowsy impaired consciousness (lethargy), because they were construed as perfectly opposite conditions in terms of symptoms, humours and qualities involved. Actually, they are conceived as the two prototypical extremes of a spectrum, in the middle of which one can find a mixture of both (which is designated by some authors as tuphomania).Footnote ⁵ Its characterisation is the only example that reminds us of the Hippocratic description of vigil coma:

So far, we can suggest that, on the one hand, Galen’s system obscured the distinction between impaired consciousness and a specific form of mental illness (mania), which had reached a rather sophisticated degree of characterisation during the post-Hellenistic period. On the other, it preserved clear boundaries between phrenitis and certain conditions that – according to his system – required different therapeutic approaches (melancholia and lethargy). The subordination of every finding to his tripartite theoretical framework becomes even more evident when trying to independently explore the different components of the notion of disease that we have been chasing in the other sources. The system is so closely knit that it is virtually impossible to isolate each element: we cannot choose to only explore symptoms, or affected organs, or mechanisms or treatments without mentioning the other three.

Galen’s notion of disease: the perfect meshing of symptoms, loci affecti, mechanisms and treatments

Undoubtedly, hallucinations still constitute an important element of phrenitis: ἀγρυπνίας ἢ καὶ τινας ὕπνους θορυβώδεις ἐπὶ φαντάσμασιν ἐναργέσιν, ὡς καὶ κράξαι ποτὲ καὶ ἀναπηδῆσαι (‘insomnia, or some turbulent dreams with such vivid apparitions that [affected individuals] sometimes scream and jump up from bed’).Footnote ⁷ However, there is a coherent theoretical framework that underpins the strong links between such a symptom (abnormal perceptions), a certain part of the body and the mechanism that explains it.

This passage relates the abnormal perceptions to a primary problem in the brain, a secondary problem in the eyes and a mechanism involving humours. Furthermore, Galen elsewhere clarifies that

In a nutshell, paraphrosunê (which invariably presents hallucinations) is associated with a primary dysfunction of the phantastikon, dependent on the hêgemonikon. It is in the seat of this ruling part of the psuchê (the brain) that the problem is generated, and from there it travels to the eyes – the visual organ of the aisthêtikon. In other words, according to Galen hallucinations are generated in the brain, and impaired perceptions are a consequence of this phenomenon.

Regarding the other symptom associated with delirium in the HC – speech disturbances – the waning tendency that we have remarked among post-Hellenistic authors persists in Galen. One of the very few occasions where it appears unequivocally associated with a state of impaired consciousness is the intermediate state between phrenitis and lethargy.

Because Galen’s theoretical framework is so consistent, we have already discussed several examples where the locus affectus plays a defining role in the definition of disease. What I would like to emphasise now is how – through his particular approach to this concept – Galen succeeded in coherently integrating most of the organs or affected parts, which had been appearing in the medical tradition since the HC onwards (without resorting to Celsus’ uncommitted stance or to Aretaeus’ inconsistencies).

Undoubtedly, the primary organ involved in all psychic activities for this author was the brain. Its relevance is revealed in The art of medicine: after defining the four archai or principles of the body, namely, the brain, the heart, the liver and the testicles (Ars. Med. CUF 5.2: 287, 1–2. K.I: 319), Galen places the hêgemonikon in the brain and associates certain psychic conditions with the impairment in its qualities.

The localisation of the hêgemonikon in the brain (enkephalon), and its possible alterations as a direct result of the characteristics of its substance (dense–thin, malleable–immalleable, stable–fluid, hot–cold) are explicit.Footnote ⁸ Moreover, there is a surprising degree of sophistication when describing the exact location of the damage:

The intersection between the anatomical and the humoural axes is clearly revealed here.Footnote ⁹ Psychic diseases can be differentiated according to the exact location in the brain (the ventricles versus the parenchyma) and the nature and quality of the humour involved.Footnote ¹⁰

Other examples illustrate how mapping the locus affectus onto the actual body determined where the treatment needed to be applied:

Apart from illustrating Galen’s self-advertising technique of discrediting his opponents to highlight his own skills, the passage very clearly shows the way in which his anatomical understanding serves the crucial purpose of providing him with a rationale to guide the treatment of each condition.Footnote ¹¹ Furthermore, even the specific location within the brain (for example, the body as opposed to the ventricles, Loc. Aff. 3.12. K.VIII: 200, 12–19) can have therapeutic relevance:

An impressive awareness, indeed, of the pharmacokinetic notion that we currently define as the ‘bioavailability’ of a drug is shown in this passage. Galen seems to understand – as we currently do – that some fractions of any drug or treatment are lost before they reach the site of action, and therefore, those losses need to be compensated for when calculating the dose. Additionally, there is an unambiguous illustration of the interplay between anatomy and humoural theory. In this sense, I agree with Devinant that l’humuralisme hippocratique is reinterpreted in qualitative terms, but I would also add the quantitative dimension.Footnote ¹²

The system so far described might appear straightforward and simple: psychic diseases are to be found in the brain, and therefore treated in the head. However, there are nuances. The concept of sympathy (sympatheia) complicates the anatomical location of the illnesses.

It is this idea of sympathy that will allow Galen to integrate all the different organs that were considered to be affected in phrenitis throughout the tradition. This notion explains how certain conditions that affect different parts of the body (and not primarily the head) can also compromise the hêgemonikon and cause psychic disturbances. As Holmes has accurately described, the brain is enmeshed in ‘networks crisscrossing the body’, thereby making it vulnerable to conditions that originate in distant locations.Footnote ¹³ Several examples of sympathy can be found in On the affected parts.

This passage supports Devinant’s remark that Galen went one step further than Archigenes: not only did he diagnose the locus affectus, not only did he suggest a sympathetic link between the organs, but he also described the mechanism that drove the problem from one organ to the other.Footnote ¹⁴ The practical importance of the whole explanation is, again, to establish where the treatment should be applied. In this respect, Galen offers one of his several self-aggrandising anecdotes. He tells the story of a young man who had fallen from a certain height and hit his upper back. The patient developed speechlessness and paralysis in both legs, which other unskilled doctors were trying to treat:

Beyond the marketing purposes, the story illustrates how the locus affectus is not always obvious, and the extent to which it is key in order to provide an effective cure. Other examples of sympathetic versus protopathetic delirium were presented above, where Galen explains how to distinguish pleuritis, peripneumonia and diaphragmatic near-phrenitis from actual phrenitis (Loc. Aff. 5.4. K.VIII: 329, 5–19; 331, 15–18). All of them illustrate how he was able to embrace the long tradition of organs affected by phrenitis without renouncing his encephalocentric position.

As exemplified throughout this section, allusions to pathophysiological mechanisms are ubiquitous in most Galenic discussions about symptoms and affected organs. I would now like to place the focus on therapies, in order to show how all four elements articulate in an integral and rational notion of disease. Galen’s construction of impaired consciousness as a spectrum that ranges between hot and cold offers a good starting point to illustrate the above: ἀγρυπνητικαὶ μὲν αἱ ἐγκαύσεις, καταφορικαὶ δ’ αἱ ψύξεις γινόμεναι (‘heat triggers insomnia, whereas coldness triggers drowsiness (kataphora)’).Footnote ¹⁵ Naturally, the treatment is aimed at counterbalancing such disturbances:

We have already seen several of these elements in Aretaeus’ discussion: the thickness, the heat or coldness of the humour according to the affection, and the intervention of atmos. However, in his eclectic approach it was not so clear how they interacted, nor was the way in which the treatment affected the body so elaborately explained. Galen justifies carefully and specifically the mechanism of his therapies, as well as their impact on the disturbance, through his theoretical model. He even explains through it the route of administration. Drugs are to be applied to the nostrils because they can easily reach the brain – the affected part – when the patient breathes in. These explanations highlight Galen’s constant interest in rationality, particularly when contrasted with Celsus’ or the author of the Anonymus Parisinus. Even when they recommend similar products (for example, poppy), they only consider the effects, with no attention whatsoever on the kind of interaction that the drug produces.Footnote ¹⁶

Galen’s use of opium as an analgesic is also useful in contrasting his understanding of drowsiness and anaesthesia as separate – even if often simultaneous – phenomena. Unlike Celsus, who considered them both as one and the same thing (therefore he considered that papaver cured pain through sleep),Footnote ¹⁷ Galen distinguishes different mechanisms through which opium produces sleep and relieves pain that is resistant to treatment.

While in the discussion on delirium the drug seems to be useful because it cools the overheated brain, the relief of pain needs, apart from its cooling capacity, its drying and thickening powers too. Once again humoural physiology is at the centre of the explanation both of the disease and its treatment.

The basics of this theory are outlined in On the causes of diseases, where Galen establishes the correspondence between the humours and their qualities.Footnote ¹⁸ In On the difference of diseases Galen explains that the imbalance (duskrasia) of such qualities is the necessary cause of an illness.Footnote ¹⁹ However, in his personal reading of Hippocrates neither all humours, nor all qualities have the same status.Footnote ²⁰ There seems to be a hierarchy:

For certain, bile and phlegm are more frequently mentioned in these kinds of diseases, and the pair hot–cold seems to have pre-eminence over dry–moist. Moreover, the antithetic character of the humours and their qualities enables Galen to explain with a persuasive logic the opposite nature of certain symptoms, such as insomnia–drowsiness. In other words, what we nowadays consider as hyperactive impaired consciousness is associated by Galen with heat, whereas hypoactive impaired consciousness is associated with coldness. By distinguishing two hierarchies in the status of the contrasting pairs of qualities, he adds complexity to the theory, thereby increasing the level of expertise required to treat the conditions.

The humoural theory is further complicated by Galen’s gradual understanding of it. In this sense, I disagree with Jouanna’s hypothesis of the dual nature of madness that he finds to be similar to the Hippocratic corpus.Footnote ²¹ On the contrary, Galen conceives the qualities of humours as a continuum where properties are dependent on the gradation of a certain quality:

In this way, humours have simultaneously a qualitative and quantitative dimension that determines the specific characteristics of diseases. The degree of their violence is, in this case, directly related to the darkness of the bile.Footnote ²² Accordingly, the treatment should compensate for the degree of distortion. In Galen’s own words: δύο εἰσὶν οὗτοι σκοποὶ περί τε τὸ ὑγιεινòν καὶ τὸ νοσερόν, ἡ ποιότης τε καὶ ἡ ποσότης τοῦ προσφερομένου (‘there are two aims concerning health and disease: the quality and the quantity of what needs to be provided’).Footnote ²³ Further down he expands:

The actual interaction of all these ideas can be clearly seen in Galen’s self-promoting anecdote regarding his discovery of the treatment for memory loss:

This criticism of Archigenes is another case where anatomy indicates the site to apply the treatment (the specific location in the brain and meninges as opposed to the skull probably suggests a stronger remedy, for it needs to reach deeper). More importantly, though, the nature of the quality imbalance is crucial for deciding the kind of treatment. The cupping-glass, with its heating and drying effects, can only be justified for cold and moist conditions, in a clear example of opposites cure opposites.

Once again, Galen’s impressive achievement was articulating into a coherent whole the allegedly Hippocratic views concerning humours and qualities, with some Platonic and Aristotelian theorisations about the four elements, and with the latest discoveries of the Alexandrian anatomical revival.Footnote ²⁵ In this way, he found a logical rationale for several treatments that were widely accepted because of their long historical tradition, and used this advantage to criticise his competitors, who could offer no coherent explanation, even if they chose those same therapies. In other words, this solid system, which conceived a tight and self-contained notion of illness, enabled Galen to build authority in the highly competitive and cultivated milieu that he inhabited.

This analysis has challenged a rather unanimous assumption of current scholarship, which frames ancient descriptions of delirium as mental illness.Footnote ¹ It proposes instead to delimit a specific clinical presentation – hyperactive impaired consciousness – as an alternative category that better describes – with fewer anachronistic theoretical assumptions – most of the cases that scholars have so far perceived as madness. This finding might seem particularly controversial when applied to the Hippocratic corpus, for contrary to an important scholarly tradition,Footnote ² I propose that the vast majority of descriptions are addressing cases of delirium, whereas madness was rarely discussed. Similarly, from the post-Hellenistic sources onwards, although I agree with most scholars that medical texts did discuss madness,Footnote ³ I consider mental illness and wakeful impaired consciousness as two different types of entities that needed to be distinguished.

In light of these findings, the tendency in contemporary scholarship to force a modern construct such as mental illness onto ancient narratives seems to be rather misleading. Indeed, although it is possible to find some coincidences and draw certain parallels, our conception of mental illness is so deeply theory-laden that it can hardly be extrapolated to the ancient medical texts. As a result, when researchers assume the validity of these categories they end up grouping and classifying the ancient material in ways that do not necessarily reflect their original conceptions. Impaired consciousness, conversely, which is clinically closer to our definition of a syndrome rather than to a fully fledged disease, is easier to recognise and presupposes fewer theoretical assumptions.

In this sense, the clear distinction between impaired consciousness and madness presents us with new challenges. Considering the strong emphasis put by the authors (particularly the post-Hellenistic ones) on the differential diagnosis between these two conditions, questions should be raised concerning the relevance of such a distinction. Beyond the evident impact that it had in terms of therapy within medical discourse, contrasting delirium with madness might also have been important in other disciplines.Footnote ⁴

In terms of chronological changes, our analysis of wakeful impaired consciousness in general and phrenitis in particular shows how they were conceived, how their understanding evolved in the studied sources and how the notion of disease mutated in parallel with those changes. Overall, we can see a persistent tension between theory and clinic, and a progressive trend away from concrete symptomatic descriptions towards more abstract conceptualisations.

In the Hippocratic corpus the tension leaned towards the clinical end. Most descriptions of impaired consciousness show a certain laxity in the notion of illness, by giving pre-eminence to thorough and detailed clinical accounts over more abstract association of symptoms with distinct diseases. This is particularly evident in the books of Epidemics, where most cases are comprised of catastases and detailed descriptions (most of them without a specific name). But even in the nosological treatises – where allegedly individual diseases such as phrenitis (and also lethargy) are tackled – the boundaries between the different conditions tend to be less emphasised than in later works. Hippocratic authors grouped phrenitis with other related diseases that required a similar kind of careFootnote ⁵ (particularly with peripneumonia and pleuritis).Footnote ⁶ In this way, the notion of disease that emerges from these authors is loose, non-specific, and it mostly consists of collections of manifestations rather than abstract nosological entities.Footnote ⁷

In post-Hellenistic treatises this emphasis slightly shifts towards theory; we could claim that there was a growing focalisation on certain symptoms, paralleled by a progressive development towards their conceptual abstraction. Of the large Hippocratic list of symptoms, some gained increasing attention, whereas there is growing indifference towards several others. Such is the case of speech disorders and coma vigil, which paled into insignificance (especially in Aretaeus’ work, where the latter is not even described, while in Celsus it appears only once).Footnote ⁸ On the contrary, the relevance of insomnia grew remarkably in close association with the presentation of phrenitis,Footnote ⁹ and so did hallucinations.

As discussed, impaired consciousness in On medicine was directly related to the vanas imagines (hallucinations); however, they were construed as a crucial symptom rather than its main cause. In Aretaeus, perception (aisthêsis) becomes a totally independent principle,Footnote ¹⁰ and it is its impairment (manifested through hallucinations) that triggers diseases with a compromised consciousness. In other words, the growth in relevance of this symptom within the medical discourse is accompanied by a progressively abstract reflection on perceptions.Footnote ¹¹

This process of privileging certain features to the detriment of certain others also reveals the way in which illnesses became better defined notions with stricter boundaries. As Pigeaud has remarked, post-Hellenistic works placed a strong focus on nosological taxonomy. In this way, contrary to the Hippocratic corpus, these treatises put their emphasis on diseases, which were organised into comprehensive classificatory systems based on a restricted number of symptoms.Footnote ¹² Undoubtedly, this novel theoretical framework required adapting the vague and broad notion of impaired consciousness into smaller categories, and encouraged doctors to funnel the complexities of delirium into stricter classifications that often used opposing symptoms as dichotomous sorting criteria:Footnote ¹³ acute versus chronic, with or without fever, wakeful versus somnolent, affecting the head or the heart, etc. This tendency to compartmentalise diseases is present – to a greater or lesser extent – in all the post-Hellenistic authors discussed, and the challenge for these doctors was to adapt the vast Hippocratic clinical diversity to the available classifications and possible explanations, where symptoms, affected organs, physio-pathological mechanisms and treatments were more cohesively grouped. In this sense, Celsus’ encyclopaedism and Aretaeus’ lax eclecticism are different approaches with the same aim.

Galen will take matters further. In his systematisation of the psuchê, perceptions are also an independent activity (or HOF to use our contemporary terminology), but they are not the primary problem. They are vicariously compromised by an affection of the ruling part, a concept with an even higher level of abstraction: it is in the hêgemonikon that the problem actually occurs. Furthermore, the symptom-focused definition of disease turned into an activity-focused one. This author conceived illnesses as the impairment of certain activities (energeiai), which in turn depended on the conditions of the body (diatheseis) that enabled such activities.

The relevance of concrete symptoms as the key elements to define and classify diseases has waned and been replaced by a more abstract concept, namely, the activity. Unlike the post-Hellenistic therapeutic approach, Galen’s method did not target the symptoms but the abnormal diathesis that was hindering a specific energeia. Accordingly, τό γέ τοι τῆς θεραπείας δεόμενον οὐδὲν ἄλλο ἐστὶ πλὴν τῆς βλαπτούσης τὴν ἐνέργειαν διαθέσεως (‘nothing else but the condition (diathesis) that damages the activity (energeia) needs treatment’).Footnote ¹⁴ Thus, diseases were conceptually reframed. The complexities of the diagnosis of phrenitis still required there to be fever and delirium (among other symptoms).Footnote ¹⁵ However, for Galen that implied a combination of θερμότης ὡς ἤδη βλάπτειν ἐνέργειαν (‘heat that was such as to damage an activity (energeia)’)Footnote ¹⁶ – i.e. Galen´s definition of fever – accompanied by πλημμελεῖς … κινήσεις τῆς ἡγεμονικῆς δυνάμεως, ἐπὶ μοχθηροῖς συνίστανται χυμοῖς ἢ δυσκρασίᾳ τῶν κατὰ τὸν ἐγκέφαλον (‘a defective movement of the hegemonic capacities secondary to pernicious humours or duskrasia in the brain’)Footnote ¹⁷ – that is, Galen’s definition of delirium. Thus, the symptoms were only relevant as long as they provided information about the diathesis of the body that was impairing the energeia, whereas the name of the diseases was merely a simplified way to group relevant symptoms. Conveniently, both the excessive heat and the abnormal humours could be targeted by Galenic medicine.

As a matter of fact, up until Galen, treatments were conceived as lists of actions or drugs (getting married, purgation, beatings, changes in illumination, etc.) aimed at palliating symptoms and mainly based on previous experience or common sense. Depending on the notion of disease, such catalogues of instructions were more or less specific for a kind of condition. Only seldom (occasionally with Aretaeus) can one see a physio-pathological rationale behind a specific therapy. It is with Galen that combating the cause becomes the main (and often the only) objective of the treatment, which was strictly informed by his tripartite understanding of disease, where the locus affectus dictated the site of application, and the type and magnitude of the duskrasia indicated the kind and potency of the remedies.Footnote ¹⁸

In other words, Galen – despite his thorough symptomatic accounts – took the tension between theory and clinic to the more abstract end. The Hippocratic writers’ detailed clinical descriptions, and the post-Hellenistic attempts to delimit distinct nosological entities shared a common interest in the patient and their clinical presentations, even if they put a different emphasis on each component. Galen’s approach, on the other hand, regarded the individual as a collection of organs and humours aimed at fulfilling certain activities (energeiai), and most of his attention was focused on finding correlations between impaired activities and alterations in these components.Footnote ¹⁹ In his coherent and cohesive theoretical system, diseases presented symptoms emerging from altered functions in specific bodily parts, which could be explained by mechanisms (mostly humoural), the correction of which required corresponding treatments. (It is in this sense that the focus is placed on theory, and the clinical presentations are only relevant as long as they can reveal the more abstract mechanisms underlying the problem.)

In contrast to the changes in the notion of disease that emerge from the analysis of wakeful impaired consciousness throughout the different periods, there are also features that remained stable. We could argue that what we nowadays designate as delirium could be identified – to a greater or lesser degree – by all the ancient sources under analysis as a variable and intermittent state of abnormal behaviour, often accompanied by altered perceptions, which could sometimes be triggered by wine (and other substances), fever or certain acute conditions. Although it could appear in the midst of longer infirmities (such as diseases of young virgins in the HC or melancholia from the post-Hellenistic sources onwards), it was construed as a recognisable entity that often needed to be distinguished from conditions nowadays identified as mental illness.