chiaradonna universals medicine-libre

8/10/2019 Chiaradonna Universals Medicine-libre

1/91


2/91


3/91

SEMINARIE CONVEGNI

33


4/91


5/91

Universals inAncient Philosophy

edited byRiccardo ChiaradonnaGabriele Galluzzo


6/91

Scuola Normale Superiore Pisa

----


7/91

Table of contents

Introduction

R C, G G

Universals before Universals: Some Remarks on Platoin His ContextM B

Platos Conception of the Forms: Some RemarksF A

Platos Five Worlds Hypothesis (Ti. cd),Mathematics and UniversalsM R

Plato and the One-over-Many PrincipleD S

Universals, Particulars and Aristotles Criticism of Platos FormsL M. C

Universals in Aristotles Logical WorksM M

Universals in AristotlesMetaphysicsG G

Epicureans and Stoics on UniversalsA B

Alexander, Boethus and the Other Peripatetics: The Theory ofUniversals in the Aristotelian CommentatorsR C


8/91

One of a Kind: Plotinus and Porphyry on Unique Instantiation

P A Universals, Education, and Philosophical Methodologyin Later NeoplatonismM G

Universals in Ancient MedicineR C

Universals in the Greek Church FathersJ Z

Bibliography

Index locorum

Index of names


9/91

Universals in Ancient Medicine

. Universals and particulars in Aristotles account of medicine

Plato and Aristotle famously use medicine as the standard exampleof an art (techn). Both Plato and Aristotle compare medicine andrhetoric. In Platos GorgiasSocrates argues that arts, such as medicine,can give a rational account (logos) of both their subject matter and thecause of the things they do. Unlike arts (but just like pastry baking),rhetoric cannot provide any such rational account, for it is merelybased on experience or the rule of thumb (a-b; see also b-

c; a-; etc.)1. e same parallel between medicine and rhetoricalso comes up in Aristotle, who compares the two in the opening chap-ters of his Rhetoric. Aristotle draws attention to some analogies be-tween medicine and rhetoric, and this is obviously of great importancefor assessing his views on rhetoric and how they differ from those ofPlato (especially in the Gorgias). I will only recall three main features ofmedicine that emerge in Aristotles Rhetoric. (i) Unlike rhetoric, medi-cine can instruct and persuade about its own particular subject mat-

ter (i.e. what is healthy or unhealthy: see Rh., , b). (ii) Likerhetoric, medicine allows for imprecision, so that complete masteryover the art is no guarantee of reaching any successful result. Still, ac-cording to Aristotle, this should not prevent us from regarding medi-cine as an art, since the competent practitioner does everything in hispower to achieve a good result, even if he may fail in his goal (Rh., ,

Here I will not focus on Plato, for his numerous discussions of medicine do notreally consider the role played by universals in it (unless indeed one reads universals

into Platos famous account of Hippocrates method in P., Phdr. cd, but this would

be controversial to say the least). is specific issue is first tackled by Aristotle and, as I

aim to show in this contribution, Aristotles discussion provides the framework for the

subsequent debates on universals and medicine. For a full account of Platos views on

medicine and its epistemic status, see H ; A J. .


10/91

Riccardo Chiaradonna

b-; see also Top., , b-)2. (iii) Like rhetoric, and all

other arts, medicine does not focus on individual cases as such, but onuniversals:

[] None of the arts theorizes about what is individual [

]. Medicine, for instance, does not theorize about what will help to

cure Socrates or Callias, but only about what will help to cure a patient of a

certain kind or patients of a certain kind [ ]: this alone

is subject to art what is individual is indeterminate and cannot be known [

] (A., Rh., , b-,trans. Rhys Roberts, with some changes).

Medicine is an art and as such includes a body of theoretical knowl-edge. is knowledge, however, does not focus on the treatment ofeach individual patient qua individual, since according to Aristotleparticular cases are indeterminate and cannot rationally be knownin their singularity (more on this below). Rather, medicine theorizesabout what helps to cure patients of a certain kind, patients whohappen to be in given conditions that are knowable and definableuniversally (e.g., as Aristotle argues in theMetaphysics, doctors theo-rize about what can treat a certain illness in all patients with a certainphysical disposition). e last-mentioned feature is extremely interest-ing for our discussion. at arts focus on universals is famously statedin the opening chapter of Aristotles Metaphysics, where medicine istaken as a case example illustrating the distinction between experienceand art; however, Aristotles view in theMetaphysicsis more nuancedthan what we find in the Rhetoric3.Aristotle presents experience as aknowledge of particulars, which originates from repeated perceptionand memory and may be seen as an organized set of data derived fromperception and retained in memory (Met., , b-a;APo., , a-). Since experience does not involve reason as a cognitivepower distinct from perception and memory, experience is not exclu-sively proper to human beings: some irrational animals also partake init, albeit in a limited way (Met., , b-). Art is different, since

it involves reason as a cognitive power distinct from perception and

See on this S , p. . I will come back to this issue below. On medicine inMet. , see F M. ; S , pp. -; H-

; on the epistemic status of medicine according to Aristotle, see also F

M. .


11/91


memory that is only proper to human beings. Reason has the specific

function of grasping universals, i.e. recurrent items that are such asto be predicated of many4, furthermore, rational knowledge not onlyknows that something is the case (to hoti, a), but also why it isso (dioti). Universal items are the proper object of rational knowledgeand cannot be grasped as such by experience: experience is knowl-edge [gnsis] of particulars [tn kath hekaston], whereas art is knowl-edge of universals [tn katholou] (Met., , a-, trans. Ross).

is famous schematic distinction, however, is subject to further

qualification. First, experience and rational technical knowledge arenot simply opposed to one another (as was the case in Platos Gorgias).Aristotle rather suggests that technical knowledge originates from ex-perience, while not being identical to it. Consequently, at a-, hecites Polus words experience made art, but inexperience luck with-out rejecting his view5. Of course Aristotles position does not coincidewith that of Polus, since he regards art as the result of experience in be-ings who, in addition to that, also possess intellectual or rational cog-nitive power. Experience, however, has a crucial (though somewhatdifficult to determine) position in Aristotles account of the formationof general concepts, both in Met. and APo. , and both textssuggest that we could not rationally grasp universals without experi-ence and memory.

Experience and art are not mutually connected simply because artcannot arise without experience. What Aristotle also suggests is thatexperience and technical knowledge may be equally effective in prac-tice. His example is significantly taken from medicine. An empirical

practitioner can successfully heal his patients through mere associativelearning based on experience, without grasping universals and withouthaving any rational explanation to offer for treatments administered:For to have a judgement [echein hupolpsin] that when Callias was illof this disease this did him good, and similarly in the case of Socratesand in many individual cases, is a matter of experience (Met., ,a-, trans. Ross). It has been noted (rightly in my opinion) thatthis view of medical empirical practice involves some power of gener-

is is indeed a very sketchy characterisation and I will not dwell on the

definition(s) of universal in Aristotle: see the remarks in M, C and

G, this volume. On Aristotles reference to Polus and its anti-Platonic character, see A

.


12/91


alization, even if this empirical generalization differs from generaliza-

tion proper, which implies the rational grasp of explanatory univer-sals. As R.J. Hankinson puts it, one may well assume that the empiricalhealer does not recognize the universal as such (he is not committed tothe view that a certain treatment is beneficial to everyone in a particu-lar condition); nonetheless, the empirical healer too acts in a certainway because the universal is true6.

Aristotle acknowledges that a treatment merely based on memoryand experience is as successfully repeatable as a treatment based on

rational knowledge, although empirical practice is not able to provideany universal and rational explanation for this fact. ese remarks sug-gest that the relation between experience and technical knowledge is,so to say, a foundational one: both experience and technical knowledgecan ensure successful practice, but technical knowledge alone providesan adequate understanding of the reasons for success, whereas experi-ence is unable to provide anything of the sort. Aristotle himself seemsto suggest this conclusion atMet., , a-:

[] to judge that it has done good to all persons of a certain constitution,

marked off in one kind [ ], when

they were ill of this disease, e.g. to phlegmatic or bilious people when burning

with fever, this is a matter of art []. For men of experience know that the

thing is so, but do not know why, while the others know the why and the

cause (trans. Ross, with some slight changes).

e above distinction could not be any clearer; yet the lines replaced

by [] provide some further remarks, which at least partially mean-ing, as far as practice is concerned qualify the hierarchy subsistingbetween experience and art. Again, medicine provides Aristotles caseexample for illustrating this complex situation:

[] we even see men of experience succeeding more than those who have

theory without experience. e reason is that experience is knowledge of indi-

viduals, art of universals, and actions and productions are all concerned with

the individual; for the physician does not cure a man, except in an incidentalway, but Callias or Socrates or some other called by some such individual

name, who happens to be a man. If, then, one has theory without experience,

and knows the universal but does not know the individual included in this, he

See H , p. .


13/91


will oen fail to cure; for it is the individual that is to be cured (A.,Met.,

, a-, trans. Ross).

ese remarks are immediately followed at a by the words [] (and yet), which introduce Aristotles emphatic state-ment that knowledge and understanding belong to art rather thanexperience. e line of this argument is complicated to say the leastand may reflect a certain tension in Aristotles views on the status oftechnical knowledge (and of medicine in particular). On the one side,Aristotle argues that experience and art are hierarchically ordered insuch a way that art alone belongs to knowledge or understanding inthe proper sense, which involves generalization and causal reasoning.Nonetheless, medicine provides a powerful case example for illustrat-ing the potential weaknesses of technical knowledge when confrontedwith individual situations. From this perspective, experience may notonly be as effective as art, but even more effective. As Aristotle argues,the reason for this is that doctors do not cure the general kind hu-man being, but Callias or Socrates or some other individual, who hap-

pens to be a human being ( , Met., ,a). is is apparently disconcerting: how can the species humanbeing be an accident of Socrates or Callias? Strange as it may seem,Aristotles position can actually be explained with reasonable clarity.He suggests that each individual human being is not cured insofar ashe/she is a mere instantiation of a general kind, but insofar as he/she isthat single individual, in his/her irreducible particularity.

Both in the Rhetoricand theMetaphysicsAristotle argues that medi-

cine includes a body of technical knowledge that as such deals withuniversal items. For example, a trained physician will know that asubstance of a certain kind (say, camomile) can heal in virtue of itsdefining properties human beings who instantiate a certain constitu-tion from a disease that is definable (say, stomach-ache). Unlike theempirical healer, the rational doctor knows more than merely the fact based on previous individual observations that administering aparticular remedy to a particular patient heals particular symptoms. In

his case, proper generalization rationally accounts for the repeatabilityof therapy in all particulars of the same kind. us, one could concludethat the trained doctor heals individual patients just like the empiricalhealer, but attains this result in a different way: for the rational doctordoes not heal the individual patient as an individual, but insofar ashe/she is the individual bearer of a disease that is universally definedand may also be found among other individuals of the same kind. An


14/91


argument such as this, however, does not hold without qualification.

Certainly, Aristotle links medical art to universal explanatory knowl-edge: medicine is an art and arts as such do not theorize on individu-als. Yet individuals cannot be removed from the practice of a givenart, and this holds paradigmatically for medicine, since as Aristotleargues inMet. doctors do not cure the universal species humanbeing, represented by its individual instantiations; rather, they cure theindividual Socrates or Callias, who happens to be a human being. In itspractice medicine should then be set in relation to individual patientsand situations; and relating the knowledge of generalities to particularsituations is not a straightforward exercise.

Technical knowledge alone thus appears incapable of ensuring un-varyingly successful practice in medicine, since practice entails beingconfronted with particular situations that are indeterminate and out-side the domain of technical knowledge (seeMet. ). A well-traineddoctor, for example, may know the universal definition of a diseaseand be capable of explaining its symptoms in causal terms, but stillfail his diagnosis when treating an individual patient. Hence Aristotles

crucial remark that experience (i.e.a kind of knowledge intrinsicallydirected to individuals) is essential in the practice of medicine, sinceif someone has the theory without the experience, and recognizes theuniversal but does not know the individual it includes, he will oenfail to cure his patient (Met., , a-). To sum up: experience isnecessary in order to relate and successfully adapt a body of technicaluniversal knowledge to the particular situations that are the object ofclinical practice.

A further crucial problem subsists, which has even deeper conse-quences on the generalization problem in medicine. Let us assumethat a combined use of reason and experience ensures the correct ap-plication of a given medical theory that good doctors will reach adiagnosis and administer the appropriate treatment for a disease ina given case. Even so, no certainty exists that the therapy chosen willprove successful: different individual human beings affected by thesame disease may react in different ways to the same treatment, so that

a given treatment will sometimes prove effective and sometimes fail.erapy is not repeatable for individuals of the same kind without fur-ther qualification, since no one patient is perfectly similar to othersand this lack of precise similarity can affect the outcome of a therapy.Furthermore, a large number of factors related to an individual patientcan affect the outcome of a therapy, and at least some of these factorslie outside the domain of universal technical knowledge in the proper


15/91


sense. Medicine cannot in any way remove such factors from its focus,

for it is intrinsically directed to the treatment of individuals.Problems such as these were tackled from the age of Hippocratesonwards, and the ancient physicians emphasized that ideally not onlythe nature of the human being in general should be considered, butthe peculiar nature of each individual (see e.g. Vict.). is overall ap-proach is set out in the famous methodological chapter in HippocratesEpidemics(Epid. I.), where we find the view that doctors should takeaccount of the individualnature of each person and of a large number

of additional factors in order to do justice to particular cases:

e following were the circumstances attending the diseases, from which

I framed my judgments, learning from the common nature of all and the

particular nature of the individual [

], from the disease, the patient, the regimen prescribed and the

prescriber for these make a diagnosis more favourable or less from the

constitution, both as a whole and with respect to the parts, of the weather

and of each region; from the custom, mode of life, practices and ages of each

patient [] (trans. Jones).

is passage with great clarity displays a line of thought that repeat-edly comes up in the Hippocratic treatises, where the doctor is sup-posed to understand the individuality of the patient in order to givehim advice and heal him (e.g. VM)7. is view is sometimes con-nected to the idea that individuals contain blends (krsis) derivingfrom the humoural composition of the body, so that each human be-ing may be seen to embody one distinctive mixture (see VM.)8. Anapproach such as this is obviously at odds with what Aristotle arguesin the Rhetoric, where he asserts that medicine is an art that does nottheorize on individuals as such, but only on what heals individuals of acertain kind. Instead, the remarks in the Hippocratic corpus are closertoMet. , where Aristotle argues that it is the individual that is to be

For further details, see S , pp. - and pp. -. On this, see the overview in S , pp. -; S , pp. f.

and pp. f. e view set out in VM. is so described by S , p. :

[E]ach individual contains a distinctive blend of humors present in different amounts

and degrees of concentration; the particular amounts and concentrations of the vari-

ous humors determine the distinctive capacities of the individual to assimilate certain

foods and to be affected by others.


16/91


cured, e.g. Socrates or Callias. We might be led to infer from this that

Aristotles views in the Rhetoricand theMetaphysicsare mutually con-tradictory, yet this would certainly be the wrong conclusion. In bothtreatises, Aristotle maintains that medical technical knowledge dealswith universals. His remarks on individuals in Met. concern notthe understanding, but the practice of medicine: as noted above, it isexperience, not technical knowledge, that is responsible for this prac-tice. e separation between the theory and the practice of medicineclearly emerges in a passage from Aristotles On Sensation:

It also belongs to the natural philosopher to obtain a clear view of the first

principles of health and disease, inasmuch as neither health nor disease can

exist in lifeless things. Indeed we may say of most physical inquirers and of

those physicians who pursue their art more philosophically, that while the

former end by discussing medical matters, the latter start from a discussion

of nature (A., Sens. a-b, trans. Beare, with slight modifications).

is passage reveals at least two facts. First, that for Aristotle medi-cine and natural philosophy are two separate and well-defined fields:his problem is to explain how they are related9. Second, that this rela-tion is close, as far as the theoretical or philosophical aspect of medi-cine is concerned, to that which subsists between two hierarchically or-dered sciences according to Aristotles view of subalternation (seeAPo. ), since philosophical medicine draws its principles from naturalphilosophy and applies them to a more specific and well-defined field(the knowledge of health and disease). e relation between medicineand natural philosophy is thus similar to that between harmonics andarithmetic or between optics and geometry. It has been noted that thisposition, which implies a strict subordination of medicine to naturalphilosophy, is similar to that rejected in the Hippocratic treatise OnAncient Medicine (De vetere medicina)10.is remark is certainly cor-rect, but needs qualification. Aristotles overall view seems to be basedon the distinction between philosophically oriented and practically orempirically oriented physicians (see also Platos similar distinction in

Lg., IV, ac and IX, cd). While what he says about philosophi-cal medicine in On Sensationactually recalls the position rejected inOn Ancient Medicine, his remarks about the practice of medicine in

See the remarks in S , p. f. S , p. .


17/91


Met. are much closer to what can be found in treatises such as the

Epidemicsor On Ancient Medicine, since Aristotle shares the idea thattreatments should be geared towards individual patients. However, acrucial difference subsists: according to the authors of EpidemicsandOn Ancient Medicine, it is medicine as such that is concerned withindividuals there is no distinction drawn here between a theoreticalaspect of medicine (dealing with generalities) and a practical or em-pirical one (aimed at treating individuals). is, by contrast, is the viewwhich can be attributed to Aristotle, although Aristotle never explicitly

presents it.e above distinction between the theory and practice of medicineis open to several objections. In a sense, it paved the way for the laterradical distinction, in the Hellenistic Age, between the conjecturalparts of medicine (e.g. diagnosis and therapy) on the one hand and itsscientific ones on the other (e.g. aetiology and physiology: see Erasis-tratus ap.[G.], Int. XIV. K.)11. However, it would no doubt begrossly misleading to ascribe an anti-empiricist view to Aristotle, evenif Jaegers celebrated idea that Aristotle should be regarded as the phil-osophical inspirer of Diocles of Carystus methodological empiricismin medicine has repeatedly been rejected12. Aristotles position is ratherthat experience is necessary to adapt a corpus of technical universalknowledge to individual given situations, as far as this is possible. isempirical adaptation, however, is doomed to be imperfect and remainoutside the boundaries of science proper.

Regularity devoid of all exceptions is nowhere to be found in thesublunary region, for here nature displays no complete regularity, but

only regularity of the sort that allows for exceptions and hence holdsfor the most part (hs epi to polu). As a matter of fact, according toAristotle this is the status of all rational knowledge focusing on thesublunary physical region (Met. ). ings vary from case to case,however, and the epistemic status of medicine cannot straightforward-ly be compared to that of sciences such as zoology or botany, sincemedicine involves a practical aspect that is unavoidably confronted

References to Galens works are given in Roman (volume) and Arabic (page)

numerals according to Khns edition (with the exception of course of those works

not included in Khn). ore recent editions, such as those of CMG and Les Belles Let-

tres, also indicate Khns pagination. For the list of the abbreviations used for Galens

works, see H a, pp. -. See E and F M. ..


18/91


not only with what is for the most part, but with what is individual

and accidental. Under such premises, the completely successful andscientific practice of medicine is de iureimpossible to attain. Indeed,as Aristotle argues in his Rhetoric, this should not prevent us fromregarding medicine as an art, even if its practice does not allow forrepeatability without exceptions and thus cannot escape occasionalfailures: the competent practitioner will do everything in his powerto attain a successful result, even if he may fail in his goal (Rh., ,b-). Remarks such as these, however, confirm that the theory

and practice of medicine remain somewhat removed from one anoth-er: medical theory shares the epistemic status of the natural sciences(i.e. sciences which focus on what is for the most part), while medi-cal practice is doomed to be at least partly empirical and removed fromscience proper; at the same time, it seems somewhat difficult to isolatetheory from practice in medicine (nor do Aristotles remarks in Met. invite us to do so). All this helps explain why Aristotle sometimesregards the scientific status of medicine as intrinsically feeble. Signifi-cantly, he makes extensive use of medical analogies in his ethical writ-ings: his comparisons rest on the fact that both the art of the physicianand that of the ethical philosopher deal with individual situations andpractical actions that contain accidental features and thus exceed theboundaries of science in its proper and true sense; hence Aristotlesemphasis on the unavoidably imprecise character of medical knowl-edge (see EN, ,b ff.)13.

e potential separation between the theory and the practice ofmedicine is deeply rooted in Aristotles views on knowledge. A cur-

sory reference toMet. may be appropriate here. is chapter is thefocus of an in-depth discussion by Gabriele Galluzzo in this volumeand I will not dwell on it. I will limit myself to following Galluzzosanalysis and recall the overall conclusion which emerges from Aristo-tles text namely, that particulars can well be objects of definition, buttheir definition is always de iureapplicable to multiple objects, evenwhen there is de factoonly one particular which satisfies it. No defini-tion exclusively picks out a particular object to the exclusion of others

of the same kind, since each definition is a conjunction of predicatesand predicates are always (at least de iure) applicable to a plurality ofobjects (see Met., , a- and a-b). Particulars are situ-ated outside the domain of definitions and there is no room for de iure

See the classical article by J .


19/91


non-recurrent individual natures in Aristotles world of knowledge14.

Particulars can only rationally be known insofar as they representsome general kind. As noted above, what we find in the Hippocraticcorpusis instead the thesis according to which cures should be indi-vidualized in order to effectively treat particular human beings, whoare never exactly similar to one another. EpidemicsI. obviously doesnot dwell on the ontological and epistemological aspects of this posi-tion, but the authors view that doctors should consider the nature ofeach individual is potentially laden with consequences. A view such as

this suggests possible philosophical developments, whereby individualunrepeatable natures would be seen as objects of rational knowledge.It is more than plausible that this medical approach to particularsblended with the later Hellenistic (and in particular Stoic) theses onontology and epistemology. As we shall see below, Galens views onthe knowledge of particulars are radically different from those of Ar-istotle and provide a full philosophical explanation of the Hippocraticnotion of individual nature.

. Empiricist generalizations and Methodist generalities

is long preamble on Aristotle was necessary to set the later medicaltheories against their philosophical background. As I aimed to show,Aristotles remarks are significant in that they opened up a range ofpossible approaches to the status of universals in medicine; each ofthese approaches was actually pursued by Hellenistic and post-Hel-

lenistic doctors (obviously I do not intend to suggest that later doc-tors always referred to Aristotle, but simply wish to draw a doctrinalparallel). Here I will not provide any overall account of the history ofmedical epistemology in the five centuries dividing Aristotle and Ga-

Indeed, according to the reading developed by F M., P , Aris-

totles theory of individual substantial forms in Met. might be potentially at odds

with this conclusion. e issue is however very controversial and Frede and Patzigsinterpretation faces a number of difficulties: see G, this volume. Furthermore,

even according to Frede and Patzigs reading individual forms should not be con-

ceived of as individual unrepeatable quasi-Leibnizean natures (the individual essence

of Socrates as Socrates), for they are co-specific and do not differ in nature from each

other precisely as forms: see F M., P , , pp. f.; F M., P

,, p. .


20/91


len15. Instead, I will focus on a very limited set of problems or theories

that are particularly important for any attempt to assess the views onuniversals and particulars.Aristotles complex view of the roles played by reason and experience

paved the way for three different developments, which coincidegrossomodowith the epistemological positions held (i) by Rationalist doctors,(ii) Empiricist doctors, and (iii) Galen. Aristotles idea that medicinehas a theoretical aspect the science of what is healthy and unhealthy which aims to rationally explain investigated objects in causal terms

foreshadows the distinction between the scientific and conjecturalparts of medicine drawn by Hellenistic Rationalist doctors. Rational-ist doctors, however, tended to marginalize experience in a way thatis alien to Aristotle. Despite significant differences among their views(there were actually several distinct types of Rationalist doctors), theygenerally argued that reason should be sufficient to determine the na-ture of a disease, ascertain its internal causes and, consequently, dis-cover the appropriate treatment capable of removing these causes (seee.g. G., Sect. Int.I.- K.). It is theoretical knowledge, then, thataccording to these doctors should enable the physician to account forhis practice. Indeed, this approach runs the risk of not doing justice toactual clinical practice (with all its failures), and de factoending in ab-stract speculation (significantly, Galen reports that Erasistratus stoppedpractising medicine to entirely devote himself to the study of the art:see PHPV. K.). It was probably this impasseof rational medicinethat prompted the reaction of Empiricist doctors from the third cen-tury BCE onwards16. As noted above, Aristotle can in no way be re-

garded as a mere forerunner of Rationalist medicine, and his views onthe cognitive value of experience actually point to a different possibledevelopment. Aristotle regards experience as something necessary toadapt and qualify medical theory in its actual practice, where technicalknowledge must be applied to individual and variable situations. Fur-thermore, he claims that experience, and experience alone, is capable ofaccounting for successful (but non-technical) medical practice, withoutin any way referring to reasoning about causes or universal entities. A

is task largely exceeds the limits of the present discussion, so I will simply refer

here to some excellent studies devoted to the subject: F M. ; F M. ,

pp. ix-xxxvi; F M. c; F M. ; V ; A J. ; A J.

, pp. ff.; F M. . e classical work on the Empiricist school remains D .


21/91


much more radical version of this view can actually lead to the position

of Empiricist doctors, who famously criticized the Rationalist approachto medicine: rejecting all talk of hidden causes, they argued that expe-rience is a sufficient basis for the art of medicine, without ever refer-ring to any faculty of reasoning distinct from perception and memory(see e.g. Sect. Int. I.- K.). However, I do not intend to suggest thatAristotle and the Empiricist doctors held the same views about experi-ence: rather, the Empiricist view is close to that of Polus, as reported inMet. , according to which experience is a sufficient basis to estab-

lish an art

17

. e Empiricists, therefore, rejected the overall Rationalistidea that doctors should grasp the basic nature of both the human bodyand unhealthy affections in order to decide on the appropriate treat-ment. All that doctors need to know, they argued, is what is harmfuland what is beneficial to a patient: on their view, there is no need forany theory to attain this knowledge, which can instead be groundedon pure observation. While Rationalist medicine grounded therapy inphysiology and pathology, according to Empiricist doctors experiencebased on a physicians own direct observation (autopsia) and drawingon the previous observations of earlier reliable practitioners (histo-ria) is fully sufficient in itself to establish medical knowledge18. As weshall see below, Galens position may in a way be seen as a synthesisbetween these two approaches and thus be compared to the view heldby Aristotle, who regarded medicine as a rational art that in practicerelies on experience. In fact, although Galen is certainly a Rationalistand maintains that causal knowledge provides the basis for appropriatetherapy, he vehemently criticizes bad Rationalist doctors and specula-

tive philosophers for neglecting experience and indulging in ground-less theories. Nonetheless, Galens views on experience differ consider-ably from those of the Empiricist doctors (but also Aristotle), since hesuggests that experience can be treated rationally and so to speak reduced to reason. Furthermore, he argues that reason can at leastapproach to the knowledge of particulars as such (see below, Part ).

See F M. . is is just a very sketchy account. For further details (with numerous refer-

ences), see the studies by M. Frede, Vegetti and Allen mentioned above, note . e

controversial issue of whether there was an evolution in the ancient Empiricist school

need not occupy us here: for further details, see the contrasting discussions in F

M. c; M . On the Empiricist kind of reasoning (epilogismos) and its

difference from the Rationalist one (analogismos), see A J. , p. ff.


22/91


is sketchy account of medical views on knowledge aer Aristo-

tle would be a very partial one indeed if no reference were made toat least two additional factors. First, the Hellenistic and post-Hellen-istic philosophical traditions, which profoundly modified the doctri-nal background of the fourth century and interacted extensively withmedicine. Specialists have repeatedly focused on the relation betweenStoicism and Rationalist doctors (in particular with respect to the the-ory of inferences from signs), as well as that between Empiricist doc-tors and Neopyrrhonism. In the present paper, instead, I will especially

focus on the theory of individuals as formulated by some Hellenisticand post-Hellenistic doctors most notably Galen. ese physiciansdeveloped the ancient Hippocratic view that medicine should consid-er individual natures by taking account of Stoic ideas on the natureand knowledge of individuals (see below, Part ). Another essentialaspect of post-Hellenistic medicine is the epistemology of the so-calledMethodist school19. Medicine, according to Methodists, is nothing buta knowledge of manifest generalities, or as M. Frede has put it ofcertain general, recurrent features whose presence or absence can bedetermined by inspection (see G., Sect. Int. I. K. and I. K.;MMX. K.; [G.], Opt.Sect. I. K. and K)20. e Methodist theoryof generalities (koinottes) may be seen as a radical overthrowing ofthe Hippocratic approach to individuality; this theory is both of philo-sophical interest in itself and crucial to an understanding of GalensPlatonic-Aristotelian account of division and universals, which hechiefly developed as a critical reaction against Methodist medicine.

Both Empiricist and Methodist doctors held distinctive views on

universals and particulars. eir views are different toto caelo fromeach other and an account of them is necessary to understand Galensapproach to the universal generalization problem21. As noted above,Aristotle qualifies experience as the knowledge of individuals and thusseparates experience from art, which is the knowledge of universals.is position is not exempt from possible objections, since experiencetoo seems to involve some power of generalization. Aristotles answerto this objection would probably be that empirical generalizations can-

not be seen as generalizations in the true and full sense. Indeed, the

See the collection of sources in T . See F M. , p. . is will be a cursory account, since the issues in question have already been

made the focus of a series of important contributions: see above, note .


23/91


empirical healer acts in a certain way because universals are true, but

he does not recognize universals as such he is not committed to theview that a certain treatment heals all human beings in a given condi-tion from a certain disease. Reason, and reason alone, can grasp uni-versals as such. e Empiricist doctors, however, did not assign anyposition to reason in establishing medical knowledge. Unlike Aristotleand Rationalist doctors, they rejected all talk of causes or naturesthat can only be grasped through reason; accordingly, they replacedreasoning about illnesses and their causes with the observation and

recording of manifest symptoms or clusters (sundromai) of symptoms(see G., Subf. Emp., , ff. Deichgrber), the recommended treat-ment of which is always the same22. Significantly, Empiricists regardedapparent instances of inferential reasoning in medicine (e.g. the tran-sition from symptoms to a suggested therapy) as cases of being in-duced to recollect23; thus they treated technical knowledge as basicallya matter of acquiring the disposition to be reminded of certain thingsby certain observations. is view may appear suspect and indeed becriticized for obscuring the difference between being reminded andcoming to know. A defence of it could also be provided, but I will notdwell on the matter24.

What I will focus on is instead a different aspect of the question.Unlike Aristotle, the Empiricists overtly ascribe a capacity of gener-alization to experience. Yet this capacity cannot be grounded on theintellectual grasping of any universal recurrent feature, nor can reasonprovide guidance for experience. e Empiricist view is rather thatknowledge of medical theorems is merely based on repeated unas-

sisted observation, either direct (autopsia) or recorded by previousreliable practitioners (historia). e way in which repeated experiencecan account for the formation of general knowledge is obviously radi-cally different from the way in which reason can account for it:

By experience we mean the knowledge of those things which have become

apparent so oen that they already can be formulated as theorems, i.e., when

it is known whether they always have turned out this way, or only for the most

part, or half of the time, or rarely (G., Subf. Emp., , - Deichgrber,trans. M. Frede).

See A J. , pp. f. A J. , p. . A J. .


24/91


Let us return for a moment to Aristotles Rationalist account of tech-

nical knowledge. According to Aristotle, the trained doctor is capableof finding the correct treatment since he knows that a certain remedyheals all individuals of a certain kind from a certain disease (Met., , a-). Since, then, the doctor grasps the properties that de-fine a remedy, a certain illness and all individuals of a certain kind, heuniversally knows that administering the remedy in question will healthose individuals from their illness. e Empiricist account of gener-alization can best be understood in opposition to this Rationalist ac-

count of technical generalization.According to the Empiricists, general propositions of the kind AllAs are F are actually nothing but the result of the repeated observa-tion of individual cases. e Empiricists avoid all reference to non-observable natures to be grasped intellectually; hence their refusal toadopt the Hippocratic humoural theory (and the consequent criticismaddressed by Galen against them: see MMX. K.). eir emphasison direct observation and on the careful recording of individual cases,however, can still be seen as being connected to the Hippocratic meth-od of the Epidemics; significantly, it was favourably regarded by Galen.Galen reproaches Empiricist physicians for focusing only on observ-able characters (in order to discover what the correct treatment mightbe, as criteria they adopt the patients age and gender, the observablequalities of his/her flesh, etc.), while neglecting the true criterion fordetermining the individual nature of each patient, namely the balanceof his/her elemental constitution. Nonetheless, their practice does jus-tice to the crucial fact that medicine aims to heal individuals; conse-

quently Galen is moderately favourable to the Empiricist view, at leastinsofar as it is opposed to that of the Methodists, which he notoriouslyregards as hopelessly false and misleading, for it subverts the practiceof the art (Sect. Int. I. K.).

As noted above, the Empiricist practice of medicine depends on thecareful observation and recording of individual cases, whereby gen-eral medical theorems are merely based on the relative frequency ofthe observed facts. Generalization is thus intrinsically connected to

frequency of observation: it is precisely in this context that a recog-nizable (albeit rudimentary and non-mathemathized) notion of prob-ability and degrees of probability emerged25. For example, a generaltheorem concerning the therapeutic power of a remedy will merely

See F M. , p. ; A J. , pp. f.


25/91


result from the repeated experience that a substance effectively treats a

certain pattern of symptoms in patients with certain observable char-acters (see for example Galens remarks on their treatment of woundsin MM X.- K.). Experience, however, allows for exceptions.e Empiricist distinguished four levels of frequency in the connec-tion between phenomena: always, for the most part, half the time andrarely (see Subf. Emp. , -; , ff. Deichgrber; Exp. Med. , Walzer; [G.], Def. Med. XIX. K.). Accordingly medical theo-rems will include an explicit specification of the frequency of the con-

nections they report. As noted by J. Allen, this view on generalizationmay well be linked to the idea that the theorems that comprise medicalknowledge are themselves stochastic and thus cannot attain true uni-versality and stability (see [G.] Opt.Sect. I. K.)26. Alexander ofAphrodisias also held this view while arguing that medicine falls shortof the criteria that qualify true sciences, whose theorems are universaland necessary (more on this below).

As we shall see, generality is no unqualified good according to Galen.Many of his polemical remarks in the treatise On the Method of Heal-ing(De methodo medendi) are addressed against a view of medicinethat allows for indiscriminate generalization, i.e. the Methodist theoryof generalities or common conditions (koinottes) as probably de-veloped by the Methodist doctor essalus of Tralles, who was activein the age of Nero and is Galens favourite polemical target in MM.Here I will only recall the fact that the Methodist school was tradition-ally taken to be inspired by the corpuscular theory held by the Ration-alist doctor (and strenuous opponent of the Empiricist school) Ascle-

piades of Bythinia (first century BCE) another of Galens pet hates27.While probably not endorsing Asclepiades Rationalist physiology,according to which the body is formed by atoms and invisible pores(with illnesses depending on either the constriction of these invisiblepores or an excessive flow through them), the Methodists developedhis general ideas in a distinctive way. ey assumed (i) that all diseasesare just a matter of constriction, relaxation (stegnsis; rhusis) or a com-bination of both; and (ii) that constriction and relaxation are not hid-

den states, but manifest phenomena and common conditions. It is bygrasping these manifest general conditions, then, that the Methodistsclaimed they could find indications as to the appropriate treatment to

See A J. , p. . See A J. , pp. - and p. .


26/91


be adopted in each case. All training, in their view, was simply geared

to make common medical conditions evident to physicians with suffi-cient clarity; hence the Methodist claim that six months were sufficientto apprehend medicine (see Sect. Int. I. K.;MMX. K.). ere wasactually some debate in antiquity as to whether the Methodist believedthat koinottescould be perceived or not; their attitude to reason is alsoa matter of debate28. Certainly, their generalities were not meant to bemade the object of inferential reasoning like the non-manifest stateswhose knowledge, according to the Rationalist doctors, accounted for

the choice of the correct treatment (see Sect. Int. I.- K.). However,neither were koinottsmeant to be grasped through repeated experi-ence, nor, according to the Methodists views, was the indication ofthe appropriate treatment to be grasped through observation and ex-perience. As M. Frede has put it, that a state of constriction requiresrelaxation and a state of relaxation requires replenishment is seen bythe Methodists as truths of reason. Unlike the Empiricists, they thusgrant that reason has a constitutive position in medical knowledge;however, their conception of reason is a non-committal one and assuch is radically different from that of the Rationalists. It is worthquoting M. Fredes account of the Methodist position in full:

ey refuse to attribute to reason any obscure powers which we would have not

dreamed of in ordinary life. ey are just noting, in this and in other contexts,

[] that there certain things that are obvious to rational creatures, though

it does not seem to be by observation or experience that they are obvious29.

Accordingly, the Methodist notion of indication differs consider-ably from that of the Rationalist doctors. Methodist indication doesnot refer to any knowledge of hidden pathological states; rather, theMethodists claim that each disease is indicative of its treatment, since,once one is aware of the disease in the appropriate way (i.e. once thecommon condition of constriction or relaxation has become mani-fest to a physician), it will also be obvious how the disease should betreated.

e Methodists famously adopted an outrageously critical attitudeto Hippocratic medicine (essalus wrote a letter to Nero against theharmful precepts of Hippocrates and proclaiming the virtues of the

See F M. , p. . F M. , p. .


27/91


Methodist sect: seeMM X.- K.)30. In fact, their theory of common

conditions may be seen as a radical overthrowing of the Hippocraticprinciple according to which individual patients are the proper objectof therapy. While the Empiricist doctors did not endorse the Hippo-cratic humoural theory, their method was nonetheless based on thedirect observation and careful recording of individual cases; and as aconsequence of this, they could still be seen as following the overall pat-tern of Hippocratic medicine. e attitude of the Methodists is com-pletely different, since they emphatically claimed that individualizing

features (such as gender, causes, the knowledge of affected parts, theage and constitution of the patient, etc.) are irrelevant for any attemptto discover the appropriate treatment (seeSect. Int.I. K.)31. On theirview, grasping the common condition was perfectly sufficient for at-taining this purpose; we should not worry, then, about individuals andhow to know them. Hence, Galens polemical remark (MMX. K.):the Methodists talk as if they were applying their therapies to the ge-neric human being instead of individuals. In a sense, the Methodistsmay be taken to have developed one of the criteria set out by Aristo-tle for assessing artistic knowledge, i.e. generality. Indeed, their viewon generalities is not based on any ontological theory about causesand essences: the author of Opt. Sect. (I.- K.) informs us thatthe Methodists talk about generalities was based on our ordinary talkabout similarities (homoiots tis en pleiosin). us they compared theirgeneralities to humanity, a feature (eidos) that we grasp in all humanbeings on the basis of their mutual similarity. It would probably bemisleading to search for a precise ontology of generalities here. Rather,

it seems that the Methodists (here as elsewhere see what has beennoted above about their conception of reason and indication) usedlogical or ontological notions in a distinctively non-committal way.As we shall see below, Galens criticism of the Methodists aims to re-verse their position. Galen is perfectly happy to admit that we shouldtake the ordinary meaning of a term as the starting point for scien-tific research. is is the case because ordinary language mirrors ourpre-scientific knowledge of the world, that of our common concep-

tions; the job of scientific investigation is to analyse these notions,thus unveiling their underlying essences. Without an agreement oncommon conceptions, it is impossible to discover the substance of the

See T , p. . F M. , p. .


28/91


matter at issue (see MMX. K.). Yet, according to Galen, adopting

this procedure correctly shows that Methodist generalities in no wayshare the status of the species human being, since koinottesare justarbitrary and artificial constructions not based on the real structure ofthe world. Hence, as we shall see below, Galens emphasis on diairesisand his idea that differentiaemust be appropriate to each genus andnot transgress its limits.

As I aimed to show, Aristotles remarks inMet.A set out with para-digmatic clarity what I would call the structural dilemma of medical

knowledge. On the one hand, medicine is an art and as such is basedon a body of general knowledge; on the other hand, medicine is suchthat generalization cannot hold without substantial qualifications andthe experience of individual unrepeatable cases plays a fundamentalrole within it. e Methodist and the Empiricist approaches may beseen as radicalizations of the two poles of this dichotomy. e Meth-odist view on generalities at least as described by Galen developsthe quest for universality to such an extent that it regards individualsas irrelevant. e Empiricists, by contrast, regard generalization as themere result of individual repeated observations, so that medical theo-rems should include an explicit specification of the relative frequencyof observed facts.

. Galen on universals and definitions

Galens monumental treatise On the Method of Healingcontains a

detailed critical discussion of both the Empiricist and the Methodistview. e first two books of this work are particularly interesting, sincethey make up a sort of general methodological premise to therapeu-tics, in which the theory of universals has a prominent position. Galenoen refers to his (now lost) treatise On Demonstrationand it is morethan plausible that in this work he fully developed those theories whichhe somewhat cursorily mentions inMM32.Galens philosophical train-

On the chronology of GalensMM, see H b, p. . e literature

on this work is rather abundant (though unfortunately a critical edition is still miss-

ing). I will especially refer to H ; B ; E .

A French and an English translation of this work have recently been published: see

B ; J, H . On Galens On Demonstration,see the

seminal work by M ; more recently, C a; H ,


29/91


ing was famously very extensive and rather unique for its day: he was

extremely well acquainted with Plato, Aristotle and eophrastus,with the Hellenistic traditions, and with works by Platonist and Ar-istotelian philosophers and commentators. is fact explains the dis-tinctive character of Galens approach. As I aim to show, the relationbetween medicine and philosophy is a reciprocal one in Galen since(i) he discusses and recasts the distinctive epistemological problems ofthe medical tradition by making extensive use of technical and philo-sophical theories (in particular, but not exclusively, Platonic and Aris-

totelian ones); (ii) in doing so, he comes to develop a highly distinctiveversion of these philosophical doctrines, which can only properly beunderstood by taking account of the specific medical purposes of Ga-lens approach to logic and philosophy (this is paradigmatically thecase with the theory of definition and specific differentia).

In the second book of MMGalen repeatedly argues in favour of atheory of immanent recurrent natures whose Aristotelian character isas evident as it is striking33. It is worth quoting some lines in full:

It is necessary for all diseases to be called diseases because they share in one

and the same thing [ ], in the same way as do human

being, cow, and each other living being. For there is some thing unique in

all human beings [ ]. For this

reason all human beings are in fact called by the same name. Similarly there

is some one thing unique to all dogs, which we attend to when we wish to

have an understanding of dogs. Equally, in horses there is some single unique

thing in virtue of which they are called horses (G., MMX. K., trans.

Hankinson, with slight modifications).

is view on universal immanent things involves a rudimentary real-ist ontology, of the kind that Galen (to the best of my knowledge) neverdeveloped in any detail (for example, he does not explain what the onticstatus of immanent recurrent features is, or the way in which they arerelated to particulars, etc.). is may appear disappointing, especiallyif we compare Galens approach to that of professional philosophers

such as Alexander of Aphrodisias or Porphyry, whose discussion of im-

who interestingly suggests that Galens DDis in the background of C. A., Strom.

VIII. See H , p. .


30/91


manent natures is highly sophisticated34. Yet an adequate assessment

of Galens view should take account of his specific purpose, which is amedical one. Here as elsewhere, Galen draws from logic and philoso-phy only inasmuch as this is necessary for him to address medical mat-ters appropriately; on his view, logical and philosophical technicalitiesshould not be pursued in themselves. In fact, Galens account of divisioninMMis part of his Rationalist account of therapy: in brief, what he ar-gues is that the principal indication of the appropriate therapy comesfrom the scientific understanding of the essence (ousia) of each disease

(MMX.; X.- K.; Fac. Nat. II. K.). Diseases are organizedin genera and species and each specific disease further determines itssummum genus: generally speaking, disease involves the impairment ofsome natural function or activity and can appropriately be defined asthe disposition (diathesis) that impedes this activity: seeMM X.; X.K.; Sympt. Diff. VII.; - K. Galens division of diseases is actuallyrather problematic and its details should not keep us here35. Here it issufficient to remark that Galen regards the division of the genus dis-ease down to its infimae species(seeMM X. K.), as well as the defini-tion of each of these species, as the rational basis of treatment. An un-derstanding of the specific essence of a disease will provide the principalindication for its appropriate treatment, whose aim is to remove thepathological disposition of the patient, thus restoring the affected bodyto its healthy and natural condition. All individual instances of disease,then, share in a unique specific thing, just as all particular human be-ings or dogs or horses do: an appropriate knowledge of this specific realnature is the first basis of therapy according to Galens method.

Galen oen points to the philosophical background of this overalldoctrine and presents Plato, Aristotle and eophrastus as his chiefauthorities regarding logical methods of division and definition (seeMMX. and X. K.). Galen mentions Platos Philebus, SophistandStatesman, as well as Aristotles On the Parts of Animals, since Aris-totle tries in that book to enumerate all the differentiaeof animals; aquotation from Platos Phaedrus is predictably added some lines below(Phdr. bc: seeMMX. K.). Galens list of auctoritatesmay indeed

strike us as somewhat surprising, since Platos dialogues on divisionare followed by Aristotles PA, which famously includes in its firstbook a scathing criticism of dichotomic division. Yet things become

See for further details C a. See B , pp. -; H , p. .


31/91


clearer as soon as we realize two facts, namely: (i) that for all their dif-

ferences, Plato, Aristotle and eophrastus, according to Galen, forma unique philosophical front, that of the logical method, to be ad-dressed against essalus theory of koinottes; (ii) that while drawinghis overall inspiration from Plato, Aristotle and eophrastus, Galendoes not simply report their theories; rather, his theory of division anddefinition presents some peculiar features, which can only properly beunderstood against the background of Galens own epistemology.

Just aer mentioning his philosophical authorities, Galen overtly

opposes them to essalus:

Yet the outrageous essalus thinks he is worthy of credence when he simply

asserts that there are only two kinds of disease in the sphere of regimen. []

And if you have discovered these things by some method, as you boast, why

dont you reveal it to us? (G.,MMX.- K., trans. Hankinson).

Basically, Galen builds on a general principle which he could eas-ily draw from both Plato (see Phdr.e) and Aristotle (see PA,A ,b ff.): the division of natural kinds should correspond to the ap-propriate joints of reality. Dividing correctly, then, is no arbitrary orstipulative procedure: for it entails that one conjoin the genus to thespecies-forming (eidopoios) differentia (see MMX. K.). What is aspecies-forming differentia? As R.J. Hankinson puts it, A differentiaDis species-forming with respect to some genus Gif and only if either(a) the conjunction of Gand Dis sufficient to identify a species, or (b)the conjunction of G, D, and some further set of differentiae is suf-

ficient non-redundantly to identify a species36. For example, as Galenargues (MMX.- K.), one should not divide the genus animal onthe basis of differentiaesuch as so and hard, heavy and light, whichare appropriate not for animal, but for substance; the appropriate dif-ferentiae (oikeiai diaphorai) of the genus animal are instead mortaland immortal, rational and irrational, tame and wild, etc.

is view finds a couple of interesting parallels in writings attributedto Alexander of Aphrodisias. InMant., , , - Bruns, we find

a sketchy account of division and differentiaewhich is similar to thatof Galen and has rightly been compared to it37:

See H , p. . See B , p. .


32/91


For the proper differentiaethat divide something [

] do not extend beyond that which they divide; for example, noneof the differentiae that divide animal occurs outside animal or belongs to

anything which is not an animal. For thedifferentiaethat properly dissect

something must be contained within what is dissected by them [

[] ] (A. A., Mant., , -, trans. Sharples, with

slight modifications).

is quaestiois designed to show that male and female are not dif-

ferent in species: the topic is closely connected to Met.I and it ispossible (but not provable with any certainty) that this text from theMantissa ultimately derives from Alexanders lost commentary onMet. I (the term oikeiosoccurs in this chapter from the Metaphysicstoo, where it designates male and female as oikeia pathof the genusanimal: see Met., I , b)38. ere are actually several compli-cated questions surrounding these lines fromMantissa, in particularregarding the view that (a) dividing (diairetikai) differentiaeshouldbe contained in the genus they divide (if X is a differentiathat dividesY, Y is predicated of X)39and (b) dividing differentiaeshould not ex-tend wider than the divided genus. Certainly, these problems were ex-tensively debated among commentators and divergent solutions wereproposed40. Alexanders texts On Differentia, preserved in Arabic,contain an extremely complex set of discussions of these issues, whoserelation with the passage from the Mantissais somewhat difficult todetermine41. Actually, the Arabic Diff. I criticizes the view accordingto which dividing differentiaeshould not extend beyond the divided

genus; however, the contradiction withMant. is perhaps not tobe overemphasized42. In Diff. I [] Alexander also deals with the ap-

See S b, p. . See B , p. . See the discussion in L , pp. -; B , pp. -. Further

evidence is now provided by the rediscovered commentary on Aristotles Categories

preserved in the Archimedes Palimpsest most probably, a part of Porphyrys bigcommentaryAd Gedalium: see C, R, S .

Diff.I and II according to Rasheds classification: see R , pp. ff. and

pp. ff. e peculiar dialectical context ofMant. should probably be taken into ac-

count (here Alexander cursorily discusses the theses about genus and differentiaonly

inasmuch as it serves to explain why male and female are not dividing differentiae).


33/91


propriate genera which should be taken into account while defining a

species43

. Here we find the remark that thedifferentiawhich makes upthe defined species is a differentiathat belongs to the genus (unediffrence appartenant ce genre, trans. Rashed).

ere are close family resemblances between MMX. ff. K. andthese passages from Alexander. Grosso modo, all of these texts empha-size that differentiaeshould not be arbitrary or stipulatively attachedto the genus in order to make up the species; the connection betweengenus and differentiain definitions should instead be an intrinsic one

and reflect the real essence that we aim to define; hence the emphasison the appropriate character of the differentia. Both Alexander andGalen reject the idea that definition is a merely stipulative or formalprocedure that combines concepts without doing justice to the realstructure of the world. A division of species and differentiae shouldcut them at the joints (MMX. K., clearly echoing Phdr. e).So far so good; but things become much more complex when we cometo examine the details. Let us quote some lines fromMMX.- K.in full:

For in the first place not every differentiathat is conjoined with the genus con-

tributes something towards the creation of the species, but only that from the

appropriate division of the genus [ [] ].

Only these are species-forming differentiae: all the others are superfluous [].

Hence it is impossible to discover the species-forming differentiaeof anything

without first having accurately circumscribed its definition, or the formula of

its substance [

] (trans. Hankinson, with slight changes).

Species-forming differentiaeare here regarded as resulting from theappropriate division of the genus. e difference from the view held inMant. is significant, for according to Alexander a genus should bedivided by its appropriate dividing differentiae; instead, Galen arguesthat differentiaecome from the appropriate division of the genre. isoverall view is even more strikingly asserted at the end of the passage,

where Galen argues that circumscribing the definition of a substanceis a preliminary requirement for discovering its specific differentiae.

For variations in Alexanders views on differentia, see R , p. note , p.

and pp. f. See R , p. .


34/91


e use of Aristotelian notions here is very evident (logos ts ousias

is an obvious reminiscence of Cat., , a- and Top., , b);however, Galen thesis that the grasping of essential definitions shouldprecede the discovery of specific differentiaappears somewhat peculiarand is repeatedly asserted inMM(see X.; X.; X. K.). Hence, di-vision seems to be regarded by Galen as a mean of systematically pre-senting something the essence of which one has already come to know.In his classic work on Galens On Demonstration, Iwan von Mllerrightly talks about die Notwendigkeit, erst eine vollstndig Definition

zu gewinnen, ehe an die Division gegangen werden kann44

.Before proceeding any further, it is worth discussing a possible in-terpretation of the lines quoted above. In his IsagogePorphyry codi-fies a well known distinction, which was certainly familiar to previouscommentators and may be thought to be of some help for interpretingthese passages, namely the distinction between dividing and species-forming differentiae or, rather, between the dividing and the spe-cies-forming or constitutive function that differentiaecan play (sincethe same differentiae can under different respects be both dividingand species-forming: see Isag., , ff. Busse). Without focusing toomuch on the details45, we may simply recall that differentiaeare takento be dividing insofar as they divide a genus into its subordinate spe-cies (hence rational and irrational are dividing differentiaeof the ge-nus animal), since exactly one of the dividing differentiaeis predicateddifferentially of everything which the genus is predicated of generally;instead, differentiaeare taken to be species-forming or constitutive in-sofar as together with the genus they make up the definition of the

divided species (rational is therefore a constitutive differentia of thespecies man, whose definition is rational animal). Alexander (Mant. ) talks about dividing differentiae, whereas Galen talks about spe-cies-forming differentiae: hence, one may conclude, the distinction oftheir views. is explanation is tempting, but some remarks may beaddressed against it.

To the best of my knowledge, the distinction between dividing andspecifies-forming differentiaedoes not come up in Galen and certainly

is not hinted at in the lines quoted above. e Greek text runs as fol-lows:

M , p. . See the in-depth discussion in B , pp. ff.


35/91


, , .

.

Galen is focusing here on the definition of species: this definitionis made up of a (set of) differentia(e) conjoined with a genus. Galenremarks that in order to really be constituents of the species, these dif-ferentiaeshould come from the appropriate division of the genus.

How can this division be accomplished? One might say: viathe samedifferentiae, insofar as they are appropriate dividingdifferentiaeof thegenus (and not constitutive differentiaeof the species). is is well pos-sible, but Galen does not suggest anything of the sort. Rather, he mere-ly argues that, say, rational and mortal are appropriate differentiaeofthe genus animal since they can make up a species (that of man) whenconjoined with that genus. One may well reach the same result (thedefinition of man) by conjoining footed and biped with rational ani-mal (MM X. K.). Such divisions are indeed different and the orderof cuts may change (as a matter of fact, the cuts rational/irrational andmortal/immortal can come in either order, whereas footed has clearlya greater extension than biped)46. However, the actual end result doesnot change, since in either way we have a definition of the species hu-man being and inMMX. K. Galen overtly states that there is morethan one way of arriving at the species in question. e order of cutsdoes not really matter and at PHPV. K. Galen does not considermaking the right number of cuts in the wrong places a way in which

division can go wrong: what really matters is that the cuts be neithertoo few nor too many (see also MMGXI. K.); the reason is simple,since cuts in the diairesiscorrespond to species-forming differentiaeand a wrong number of cuts entails that the definition of the specieshas not been circumscribed correctly. To sum up: the lack of mentionof dividing differentiaeatMM X.- K. may not be haphazard, sinceGalen does not claim that we come to define the species by dividingthe genus through its appropriate (dividing) differentiae. What Galens

discussion rather implies is that the definition of the species in ques-

See on this and what follows H , p. . R , p.

argues that the choice between rational and biped as species-forming differentiaeof

human being raises some questions concerning Alexanders hylomorphic ontology.

As far as I can judge, Galens overall approach does not share this kind of concerns.


36/91


tion should somehow already be known from the start and act as a

guide for finding the species-forming differentiaewhich are appropri-ate to the genus (i.e. which make up its subordinate species). But howcan this be possible?

Before attempting to answer this question, it is worth discussinga further parallel with the commentators. We know from S., InCat., , ff. Kalbfleisch that Herminus, while interpreting Cat., ,b-, argued that differentiaethat occur in different parallel gen-era which fall not one under the other, but rather all under the same

genus (e.g. the differentiae biped and quadruped that occur both inthe genus terrestrial and in the genus winged, including mythologicalcreatures such as the sphinx or the gryphon, which both fall underthe genus animal) are primarily differentiaeof the superior genus thatincludes the parallel genera (i.e. of the genus animal)47. is view isobviously open to the objection that animal would thus be both bipedand quadruped: we find a remark of this sort in Alexanders text Diff.I, preserved in Arabic (see Diff. I [i])48. In this work, Alexander reactsagainst an adversary whose position is actually extremely close to thatof Herminus49. Galens view that rational/irrational, mortal/immortal,tame/wild, are all differentiaeof the genus animal may actually recallHerminus view that biped and quadruped are primarily differentiaeof the genus animal. However, Herminus view is certainly connectedto his refusal to admit species-forming differentiae: on his view, dif-ferentiae are just dividing differentiae (see S. In Cat., , -Kalbfleisch). is remark does not solve all problems, but can at leastserve to dismiss the objection that the genus animal would be equally

determined by differentiae such as biped and quadruped: Herminusmerely argues that these differentiaeare primarily dividing differentiaeof the genus animal. Furthermore, what we know about his views onhow major and minor terms should be determined in syllogistic showsthat Herminus was very interested in fixing the correct order of cuts indivisions that start from the highest genus (see A. A., In APr.,

Further crucial evidence on Herminus theory of the differentia is given by the

newly rediscovered part of PorphyrysAd Gedalium (see above, note ). I will not

dwell on this. Porphyrys account supplements that of Simplicius, but is not at odds

with it. For an overall account of Herminus, see M , pp. -. See R , p. See R , p. and pp. ff.


37/91


38/91


acquires full meaning when interpreted as a reaction against Method-

ist therapeutical principles. What Galen is doing is reproaching es-salus for his practice of drawing therapeutic indications from hyper-general differentiaethat lack any real informative value for treatment.Galen does not deny that the Methodist costive/fluid differentiaemaybe attached to diseases (just like rare/dense, hard/so, taut/relaxed: seeMMX. K.); what he denies is that enumerating differentiaesuch asthese may in any way suffice to adequately determine the species ofdiseases (such as inflammation, tumour or oedema), thus establishing

a pathology, which (in his view) is the only adequate ground for ther-apy. Hence Galens irony (MMX.- K.) directed against es-salus view that a concave wound in a fleshy part of the body shouldbe treated by administering a remedy that generates flesh (embalontes[] to sarktikon pharmakon), thus replenishing the wound. isis obviously true, but one should determine precisely how this rem-edy should be prepared, and this, according to Galen, requires (MMX. ff. K.) detailed particular indications (about pharmacology, theelemental composition of the wounded flesh, etc.: see MMX.-K.), as well as rationality (logos) and logical methods.

Yet, while Galens polemical strategy clearly emerges from the text,his own views on division and definition still appear somewhat vague.As I aimed to show, his emphasis on appropriate differentiae and thenecessity of adequately determining species can be seen as a reactionagainst the Methodists. Galens peculiar view that one should first graspa species in order to discover its differentiaecan also be understood aspart of his anti-Methodist approach. us we return to the problem

mentioned above: how can one be capable of grasping the essentialdefinition of a species accurately, if this awareness is a necessary condi-tion for finding the differentiaethat account for appropriate division?We may suppose that one comes to grasp species by inductive reason-ing: by division, each species will subsequently be ordered under its ge-nus. Yet Galen dismisses induction as inappropriate for demonstration(see ras. V. K.)52, so another explanation should be found. Actu-ally, Galen himself provides an answer to this question atMM X. .:

[] we must now I suppose explain what a disease actually is in its definition,

so that we may thus attempt a proper division of it. How then do we find this

out correctly and methodically? How else than by the means specified in On

Further references in B , p. .


39/91


Demonstration? First of all the common conception must be agreed upon [

]: without it it is impossible to discover thesubstance of the matter at issue (trans. Hankinson).

e Stoicizing reference to the common conception points to adistinctive theory by Galen. He sees the ordinary use of (the Greek)language as mirroring a pre-theoretical knowledge of the world, whichscientific enquiry should take as its basis and criterion for discover-ing essential or scientific definitions by scientific enquiry accordingto logical methods (see MM X. K.)53. Scientifically understanding

the essence of something is thus presented, in a famous passage fromDiff. Puls. (VIII. K.; see ff. K. for the whole account), as a tran-sition from notional to essential definition (ennomatikos horos;ousids horos) (e.g. that of pulse). It is crucial to note that notionalaccounts are by no means merely conceptual; nor are they the objectof merely linguistic analysis: rather, they reflect a pre-scientific, notadequately unfolded, and yet perfectly real, acquaintance with theworld, which should be the basis of all scientific accounts54. Cutting

at the joints, then, entails respecting the distinctions expressed byour linguistic awareness of the world (hence Galens criticism againstthose physicians who do not respect these distinctions and, accord-ingly, provide arbitrary definitions: see Diff. Puls. VIII. K.; see alsoGalens polemic against Archigenes artificial use of language in Loc.Aff. VIII.- K.)55. e question of how to correctly divide can

For further details, see B ; C b. e status of notional accounts is closely parallel to that of Galens dialectical

premises in PHP: see C b, p. . According to Galen, language is a necessary condition for establishing correct

classifications, but not a sufficient one, for Galen recognizes the existence of inef-

fable differentiae, i.e. differentiaethat can be perceived and are relevant for medical

knowledge, but for which there is no corresponding word. is happens, e.g., with cer-

tain types of pain (Loc. Aff.VIII. K.), with certain types of pulse (Diff. Puls. VIII.

K.) or discolourations of the skin (Loc. Aff. VIII.K.). Galen acknowledges thatmany perceptual experiences cannot be expressed in words (Dign. Puls. VIII.-

K.). Furthermore, he also seems to recognize the existence of a subset of unsayable

properties that can only be perceived in a way which does not make them fully and

consciously available to us (Loc. Aff.VIII.- K.). On this, see the enlightening

discussion in R . It is however extremely important that Galens em-

phasis on unsayable properties does not lead him to disqualify language as a mean for


40/91


41/91


Certainly, an accurate scientific account can in no way be compared

to either the awareness of specific unity that (according to Galen) canbe found in donkeys, or the awareness reflected in ordinary language.In Diff. Puls. VII.- K. Galen shows how we can reach a scien-tific definition (that of pulse) by starting from the ordinary accountof the thing in question, i.e. the conceptual account that reflectsour elementary awareness of the perceptible accidents of the investi-gated phenomenon. According to the present reconstruction of Ga-lens views, a full division which starts from the summum genusand

ends with the species by enumerating all of its constitutive differentiaeshould then be seen as the final result of the enquiry which unfoldsfrom our preliminary awareness of the investigated thing. is ordi-nary and preliminary awareness provides a sketchy but in no way ar-bitrary or stipulative map of reality, thus acting as a basic criterion fordivision. Accordingly, we do not accomplish division by starting froma summum genusand then dividing it according to differentiae; rather,we set off from a preliminary grasping of the investigated thing andproceed by unfolding this account according to a methodical use ofour cognitive capacities (reason and experience). It is this procedure(and not a formal procedure of division) which allows us to discoverspecies-forming differentiae. As R.J. Hankinson aptly remarks, whatthe scientific taxonomist does is to make clear and rigorously defineddistinctions which everybody (indeed, every animal) already knows insome sense of know58.

. Galen on particulars

In the first two books ofMMGalen emphasizes the scientific char-acter of therapeutics and links division to the kind of rational under-standing pursued by medicine. All this obviously tends to presentmedicine as a body of general knowledge which (according to Galen)can attain the same status that we find in hard sciences such as math-ematics. at medicine as such deals with universals is actually also

asserted at the beginning of theArs medica. e author of this treatise(probably Galen, although the issue is debated) follows Herophilus indefining medical science (epistm) as the knowledge of health-re-

H , p. .


42/91


lated, disease-related, and neutral things (I. K.)59. e author fur-

ther explains (I. K.) that this definition may be taken to mean thatmedicine is a science (a) of all of these things taken as particulars, (b) ofsome of these things taken as particulars, and (c) of things of such andsuch a kind60. Both (a) and (b) are rejected: medicine cannot focus onall individuals, since these are infinitely many; but on the other handit cannot focus only on some individuals, since in this case it wouldbe incomplete and would not be an art. Instead, medicine focuses onkinds of individuals. Focusing on kinds of individuals both belongs

to the science and is adequate to all the particulars of the science(trans. von Staden). J. Barnes rightly qualifies this view as thoroughlyAristotelian61and, indeed, Galens use of hopoinis similar to Aris-totles use of toiide/toioisdeat Rh., , a- and of toioisdeatMet., , a. In all of these passages, medicine is regarded as anart to the extent that it is based on a corpus of general knowledge 62. AsI aim to show, however, this is not Galens last word on this issue.

As noted above, this view is potentially aporetic, since general medi-cal knowledge cannot account for clinical practice, which is unavoid-ably confronted with individual and variable situations. Still, a body ofgeneral scientific knowledge should in principle account for unquali-fied repeatability (for example, a geometrical demonstration can un-qualifiedly be repeated for all particular geometrical objects that satisfycertain conditions established ex hypothesi) and medicine falls short ofthis criterion63. Just aer Galen, Alexander of Aphrodisias (see In APr.,, -, Wallies) drew all consequence from the aporetic statusof medical knowledge and overtly denied that medicine (like all sto-

chastic arts) could be regarded as being rigorously scientific, since itdeals with contingent objects and its syllogisms (unlike those of truesciences) are not apodeictic but problematic. Certainly, Galen did nothold anything of the sort: he repeatedly argued that medicine is a fullydemonstrative and certain form of knowledge comparable to that pos-sessed by arithmeticians and geometers (see PHPV. K.;MMX.

See S , pp. ff. , ,

. Text aer Boudon-Millots edition: see B-M

, pp. , -, . See the contribution on Barnes in B-M (Discussion). For further parallels, see G-B , pp. f. I develop this point in C b.


43/91


K.). If this is the case, however, one should address the familiar prob-

lem of how this certain general body of knowledge is to be appliedto clinical practice, which does not allow for unqualified repeatability.A possible solution is sketchily presented in a famous passage fromthe pseudo-Galenic On the Best Sect(De optima secta). Unlike Alexan-der of Aphrodisias, the author of this treatise regards general medicaltheorems as being certain and precise in nature; what is neither certainnor precise he argues but merely conjectural, is their practical andempirical application to individual cases (Opt. Sect.I.- K.). is

treatise is spurious; yet some scholars are inclined to assume that itrepresents Galens final answer to the problem of the scientific statusof medicine64. I agree that the view voiced in the treatises finds signifi-cant parallels in Galen: a passage such as Ars Med. I. K., for ex-ample, equates genuine knowledge with the knowledge of universals.is Aristotelizing idea certainly found a prominent place in Galensepistemology, but I would be hesitant to claim that this was Galensfinal answer to the problem of the scientific status of medicine. Asnoted above, this answer is only apparently convincing: in itself it isactually aporetic, unless one adopts the radical strategy of insulatingscientific theoretical medicine from clinical conjectural practice thatwas familiar to some Hellenistic doctors. Otherwise so long as clini-cal practice is taken to be an integral part of medicine claiming thatmedicine is a science since its general theorems are necessary, whiletheir particular applications are merely conjectural or empirical, doesnot really solve any problem. Certainly, Galens intention in MM isnot to separate a body of general theoretical knowledge from clinical

practice based on experience. Quite on the contrary, his work aims toshow that clinical practice can be treated scientifically and accordingto logical methods65.

roughoutMM, Galen argues that demonstrative medical skill assuch is capable of curing individual human beings. Galen does not con-fine medical practice which involves the treatment of each individual outside the domain of medical knowledge in its proper and full sense.At the very beginning of his short therapeutical work To Glaucoon

the erapeutic Method (De methodo medendi ad Glauconen) (MMGXI. K.), Galen argues that doctors should know (epistasthai) not onlythe common nature of all human beings, but also the nature proper

See I , pp. -. See on this B , pp. f.


44/91


(idian) to each one. is is obviously consistent with Galens overall

Hippocratic programme, which he emphatically reaffirms (e.g.) in thepassage mentioned above (MMX. K.), where he criticizes Method-ists for talking as if they were applying their therapies to the generichuman being rather than individuals. As Galen emphatically argues,it is not the generic human being that is cured, but each one of us(hmn hekastos). True medicine, according to Galen, should take ac-count (conjecturally, as we shall see below) of the nature of (each in-dividual) patient (MM X. K.):

. e words [] assuch may not necessarily refer to an individual unrepeatable nature.Galen could simply be claiming that medicine should consider recur-rent natures instantiated by individual patients. Yet this is certainlynot the case, since Galen immediately goes on to explain that mostdoctors call this nature idiosyncrasy and that they all agree that itcannot be grasped (akatalpton). e term idiosyncrasy (idiosunk-rasia) occurs sev

chiaradonna universals medicine-libre

Documents