Download as pdf or txt
Download as pdf or txt
You are on page 1of 38

Galen's Anatomy of the Soul

Author(s): R. J. Hankinson
Source: Phronesis, Vol. 36, No. 2 (1991), pp. 197-233
Published by: BRILL
Stable URL: http://www.jstor.org/stable/4182386 .
Accessed: 25/09/2013 11:20
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

BRILL is collaborating with JSTOR to digitize, preserve and extend access to Phronesis.

http://www.jstor.org

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

Galen's Anatomy of the Soul


R.J. HANKINSON

This paperpresentsthe firsthalfof a two-partinvestigationinto the nature


and coherence of Galen's philosophicalpsychology.In it, I shall consider
Galen's account of the structureof the soul, his vigorous defence of a
Platonictripartitionalpsychologyin contrastto the unifyingaccountof the
Stoics, paying particularattentionto his attempteddemonstrationsof the
distinctnessand variouslocations of the soul's parts, and to the epistemological and methodologicalprincipleshe bringsto bearin this effort. Part
of the purpose of this will be to demonstratein some detail how Galen's
twin professions of philosopher and physiologistcombine to inform his
particulartreatmentof these issues. In the second part,' I shall turnto the
moralaspectsof Galen'sphilosophicalpsychology,and in particularto his
accountof the passions,and the questionof whetherthey can be tamed, or
whether rather they require root-and-brancheradication;and I shall attempt to show how Galen's physicalismis to be made compatiblewith his
belief that, at least in their generalcondition,our passions,and the extent
to which we are the slaves of them, are up to us. Fromall of this, I hope,
there will emerge a picture of Galen engaged in what Fodor has called
'speculative psychology',2 that is the attempt to steer a middle course
between a purely empiricaland hence potentiallyimpoverishedapproach
to the science of psychologyon the one hand, and an overlyaprioristicand
rationalisticpsychology of the mind carriedthroughwith no regard for
empiricaladequacyon the other. And, I hold, Galen engaged in it in a
sophisticatedmanner- for, properlycarriedout, such a project need be
neitherbad science nor inadequatephilosophy,butindeedthe only fruitful
and intelligentway of going about either.3
1

To appear as 'Actions and passions:emotion, affection and moralself-managementin


Galen's philosophical psychology', forthcoming in Passion & Perception (the proceedings of the Fifth Symposion Hellenisticum), edd. J. Brunschwigand M.C. Nussbaum.
2 See J. Fodor, The Language of Thought(CambridgeMass: 1975), p. vii.
3 This is a strong, some may think lunatic, claim, which I cannot defend here. Of course,

Phronesis 1991. Vol. XXXV1/2 (Accepted February 1991)

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

197

1. A CautiousSyncretism
Galen is at least to some extent a Platonist,'althoughscholarshave sometimesallowedthatfactto obscurethe degreeto whichGalenwas, in the best
possiblesensesof the words,an eclecticanda syncretist;'he madeuse of the
wide variety of materials available to him in the tradition to weave a
compoundcloth of greatstrengthfroma varietyof strands.6It is important
to stress against its detractorsthat such an eclecticismis not necessarily
simplya type of sophisticatedplagiarism- on the contraryits resultscanbe,
and in Galen's case to my mind are, powerfullyoriginal.
One of the areas in which he both evinces and acknowledgesmost
generouslyhis debt to Platois in his psychology.Like Platoof the Republic
he is committedto tripartitionas a necessaryconditionfor explainingthe
phenomenaof psychicconflict;he followsthe Timaeusin his assignmentof
the three partsof the soul, rational,spirited,and appetitive,to the brain,
the heart, and the liver respectively;and he takes over Plato'sterminology
for the faculties.7Indeed, Galen wrote a massiveworkOn the Doctrinesof
as Fodor himself points out, the best form for such a defence consists simply in applying
the model, working it through, and showing that the results are fruitful.
See P.H. De Lacy, 'Galen's Platonism', AJP 1972.
See my article 'Galen's philosophicaleclecticism', forthcomingin Aufstiegund Niedergang der Romischen Welt,1I, 36 4; I couch the last remarkcarefullybecause there seems
to be a divergenceof opinion as to whichof 'syncretism'and 'eclectism'is the commendatory and which the pejorative term. Dillon and Long, in their recent collection The
Question of "Eclecticism"(California, 1988), tend to reserve the latter for reasoned,
principledselection of the best elements from a variety of traditions,leaving the former
for uncritical rag-baggery; my own intuitions regarding usage tend the other way,
although nothing really hangs on that - as Galen himself would have said, provided we
know what we mean the rest is futile quibbling over words. Perhaps we should have
'eclecticism'cover the cullingof diverse materialsfrom differentsources, while reserving
'syncretism' for the project of attempting to show that, surface differences notwithstanding, the various strands of the tradition in fact turn out to say the same thing.
Whatever the truth of the linguistic matter, Galen was certainly not merely an omnivorous intellectual magpie creating a pot-luck supper out of the offerings of his
predecessors.
6 In his philosophy of science, as well as in his general metaphysics,Galen owes more to
Aristotle than Plato: see my op.cit, n. 5 above.
see Plato, Rep. 434-441; and see also Galen's
7 logistikon, thumoeides, epithume'tikon:
On the Doctrines of Hippocrates and Plato (PHP = CMG V 412, De Lacy 1978-84)
V 307, 338, 363-5, 382, 516-18, etc. PHP was also edited for the Teubner series by 1.
Muller (Leipzig, 1874); Muller'stext is a huge advanceon Kuhn, but is no substitutefor
the magnificent CMG edition of P.H. De Lacy (3 vols, including English translation,
commentaryand indices): Berlin, 1978-84;'De Lacy, 1984'refers to the thirdvolume of
commentaryand indices.

198

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

Hippocratesand Plato the main purposeof whichwas to demonstratethe


harmonyof the views of its two principalsubjects,and to show, interalia,
the empiricalinadequacyand methodologicaldisreputabilityof the competing Stoic account.
But Galen's Platonisingis neither slavish nor unthinking;and in his
characterizationof the psychicfaculties, he deliberatelyavails himself of
Aristotelian and even some Stoic elements in addition to his Platonic
inheritance(althoughhe is to take issue with Aristotle'sassignmentof the
reasoningfacultyto the heart,as well as withmuchin the Stoictreatment):
for instance, he considers(at least in some contexts) nutrition,reproduction, and the threptikedunamis to be psychic in category.8Galen, in
commonwithmost Greekthoughton the subject,consideredthereto be no
radicaldistinctionof type between the physicaland the mental (or more
properly,the psychic).The bodycontainsorgans(organa), properlyconstituted, goal-directed,anhomoeomerousphysicalstructureswhichseverally
have functions (chreiai) that are contributory(and hence teleologically
posterior)to the overallfunctioningof the organismas a whole; and these
functionsare expressedin the characteristicactivities,or energeiai,which
these organsperform.A diseaseconsistsin the impairmentof one or more
of these activities;hence therapyinvolves the removalof the impediment
PHP V 521; cf. 533, 658; although characteristicallyhe doesn't want to make an issue
out of this: cf. MM X 635: 'it makes no difference whether you call it desiderative,
natural, or nutritive, nor whether you call it a soul or a power' (see also P.H. De Lacy,
'The third part of the soul', in P. Manuliand M. Vegetti (eds.) Le Opere Psicologiche di
Galeno (Naples, 1988), pp. 43-63, esp. pp. 52-5). On the general issue of Galen's
divisions of the soul, see De Lacy, 1988, who rightly stresses the flexibility of Galen's
thought behind the dogmaticfacade, and notes that in De Semine (Sem.) Galen alters his
account of the partition of the soul to allow the gonads a share as the fourth part of it
(Sem. IV 570, 572f., 622; cf. Ars MedicaI 319, and De MethodoMedendiad Glauconem
XI 97); and see also De NaturalibusFacultatibus(Nat.Fac.) 11 1-2, where Galen sides
with the Stoics against the Peripateticsin assigningthe vegetative functions not to soul
but to nature (cf. 6-7, 10, 12, 15, 17-18, etc.; however, at ib. 28-9, his languagesuggests a
general indifference to such terminological distinctions). Jaap Mansfeld, in an unpublishedmanuscriptentitled 'The idea of the will in Chrysippus,Posidoniusand Galen'
detects a further move too in Galen's De Moribus (a work lost in Greek but which
survivesin an Arabic epitome, translatedinto English by J.N. Matlock, in S.M. Stern et
al. (edd.) IslamicPhilosophyand the ClassicalTradition(Oxford, 1972) ), away from the
concept of a soul divided into partsand towardsa more Aristoteliannotion of a varietyof
psychic functions; however, I do not find the passages he quotes convincing as evidence
that Galen ever abandoned tripartition;nor do I share Mansfeld's view that such an
abandonmentwas more or less forced upon him by the fact that his 'physiology on the
one hand and his moral philosophy and philosophy of mind are not co-ordinate'.

199

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

and the strengtheningof the naturalfunctionalcapacityof the organ or


organs in question.9Many such functions and activitiesare assigned by
Galen to the soul (as I shall standardlybut with all the usual caveats
continue to renderpsuche), indeed to the rationalor hegemonicsoul;'0
Galen lists those of the imagination,memory, recollection, knowledge,
thought, consideration,voluntarymotionand sensation."Otherfaculties,
such as anger and states of passion in general, are to be assignedto the
thumoeides;`2while the thirdpartof the soul is that in chargeof
nutritionin the animal, the most importantfeature of which in us and all blooded
animals is the production of blood. To this power also belongs the enjoyment of
pleasures; and when it is moved by enjoyment more than is fitting, it produces
incontinence (akrasia) and licentiousness (akolasia). (PHP V 601; trans. here and
passim after De Lacy; cf. Symp.Diff. VII 55; MM X 636; note the Aristotelian
terminology for weakness of the will.)

9 Galen works this out in detail in the first two books of MM (see n. 5 above); but cf.
particularlyMM X 78-81. On the issue of the natureof psychologicalafflictions,and the
extent to which they are to be described as diseases, see PHP V 432-54; and see p. 203
below.
'? Galen uses both logistikon and hUgemonikoninterchangeably;he repeatedly emphasizes the fact that terminologyitself is unimportant,providedthat you providea clear
indication of the referents of the terms in your usage; hence this indifference as to
whether to adopt Platonicor Stoic terminologyis not simplycarelessnesson Galen's part
- it is part of his theory of scientific language; see my art.cit., n. 5 above.
"phantasia, mnemme,
anamnesis,episteme,noesis, dianoesis, PHP V 600; see also On the
Differencesof Symptoms (Symp.Diff.) VII 55f; and see MM X 636:
The thirdpart, the logicalsoul, is located in the brain,and it controlsthe activitiesin
accordancewith choice as well as perceptions, makinguse of the nervesas conduits,
and sending perception and movement to the whole animal by way of them.
and compare the remarks in On Habituations(Eth.: this does not appear in the Kuhn
edition; it is edited by Muller in SM 2 9-31):
As the hegemonic soul has capabilities(dunameis)directedtowardsall the technai,
it is necessary that there is one (sc. dunamis)with whichwe understandconsequence and conflict, and another with which we remember; and we are cleverer in
respect of the first mentioned, but more retentive in respect of the second. (Eth. 4,
= SM 2 25)

The distinction between intellectual sharpness and retentive ability is an ancient commonplace: see in particularthe Hippocratictext On Regimen 1 35.
12 PHP V 601; cf. Symp.Diff. VII 55-6; and see MM X 635-6:
a second partof the soul belongs to us not in virtueof our growingor being alive, but
because we are animals, it is located in the heartand is the sourceof the innate heat;
the arteries are the conduits for this source, which has many names: it is called the
living power (dunamisz6tike), the spiritedpower (dunamisthumoeides),the living
soul, and the spirited soul.

200

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

So Galendrawson a diverserangeof previoustheoriesin orderto construct


his own account;butit is importantto stressthatthe theorythatresultsis no
mere haphazardporridgeof badly-digestedandill-assortedscrapsfromthe
tableof his predecessors.He is not afraidto take issuewiththemon matters
of substantialimportance,andto take issuewiththemin his own characteristically,indeed uniquely,polemicalstyle.
For instance, he will not say, as Plato does, that any part of the soul is
demonstrablyimmortal.Indeed, in his pronouncementson the matterhe
exhibitsan admirablecaution,unwillingto commithimselfwith anydegree
of certaintyon matterswhichhe viewsas beingby theirverynatureresistant
to secure demonstration."3
Thus at PHP V 791-2, Galen writes:
Plato said that the cause who made us, the demiurge who fashioned the universe,
commanded his children to make the human race by taking . .. the substance
(ousia) of the immortalsoul from him and addingit to what was generated. But we
must realise that there is no formal similaritybetween provingand positing the fact
that we were made in accordance with the providence of some god . . ., and

knowing the substanceof the maker or even of our own soul ... [T]he statements
of the most divine Plato about the substanceof our soul . .. and still more all that
he says about our whole body, extend only to plausibilityand reasonableness (achri
tou pithanou kai eikotos).

That last remarkis important(indeed Galen, perhaps excessively charitably, takes Plato himselfto be committedto it by his remarksabout the
eikos muthos[Tim.29c-d]).Thereis a classof thingsaboutwhichwe can at
best speculate, and most particularlythese are the preserve of the
philosophers:
In philosophyit is not surprisingthat most disagreementshave not been resolved, as
the matters it deals with cannot be clearlyjudged by experiment (peira);thus some
say that the universe is uncreated, others that is created, some say that there is
nothing outside it, others that there is something, and of the latter, some say that
what surroundsit is a void that has no substancein it, others that it is surroundedby
incalculablymany other universes. Such a dispute (diaph6nia)cannot be settled by
evident perception (aisthesisenarges) (PHP V 766)14

The case is not the same as it is in medicine, where theoretical pronounce-

ments can, and indeed should, be made answerableto the tribunal of


experience.15
On these issues in general, see Frede, 1981.
On these questions, cf. De PeccatorumDignotione (Pecc.Dig., = SM 1 45-81 Marquardt [1884], = CMG V 411, De Boer [1937];) V 67, 98-9; On the Affected Parts
(Loc.Aff.) VIII 158-9; Pecc.Dig., on the diagnosis and cure of the soul's errors, is the
companion volume to Aff. Dig., and is also translatedby Harkins, 1964.
'" This view has a long and respectable pedigree in Greek medicine - indeed it is
traceable (in a stronger form) to the early Hippocratic treatise On Ancient Medicine
3

14

201

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

So Galen holdsthatthe questionsof the soul'ssubstanceandits mortality


are ones that at best admitof plausibility,andnot certainty(PHP V 791-3);
and that in any case they are extraneousto the concernsof the medical
practitioner (PHP

V 794-5).16 In On the Formation of Foetuses

(Foet.Form.)IV 700-2,apparentlyreferringto PHP (althoughhe mentions


anotherlost work, On the Formsof theSoul;cf. PHP V 803, and the notes
of De Lacy, 1984, p. 708), he writesin an even more aporeticvein:
in default of finding any scientificallydemonstratedopinion, I owned myself at a
loss with regardto the soul's substance, being unable to advanceeven as far as the
plausible . . . I made no attemptto assertanythingregardingthe soul's substancein
any of my work. For I was unable to find out by means of linear demonstrations'7
whether it was completely incorporeal, or if some part of it was corporeal, or
whether it was completely eternal, or if it was corruptible.

Similar avowals of ignorance are to be found in De Usu Partium(UP)


III 452, and De UtilitateRespirationisIV 472, 501. It appearsthat the real
natureof the soul is not merelynot susceptibleof precise,scientificdemonstration(I shall treat of Galen's notion of scientificdemonstrationa little
later on); we cannot even arriveat a plausible,if fallible,view aboutit.'8
Elsewhere, however, he allows himself to draw at least conditional
conclusions regardingthe soul's nature. In Thatthe Powers of the Soul
Depend upon the Temperamentof the Body (QAM)'9IV 774-5, he writes
that 'if the rationalpart is a form of the soul, then it is mortal:for it is a
temperamentof the brain'(cf. QAM IV 781-3), even thougha page or so
(VM) chs. 1 and 2. For Galen's own views on the relations between logos, reason, and
peira, experience, cf. MM X 33ff.; In Hippocratis de Victu Acutorum Commentaria
(HVA, = CMG V 91, Helmreich [1914]) XV 446ff.
ac FacultatibusXI 731; and
16 See also De SimpliciumMedicamentorumTemperamentis
In HippocratisEpidemicaXVIIB 247; on this issue, see L.G. Ballester, 'Soul and body,
disease of the soul and disease of the body', in Manuliand Vegetti (eds.) 1988,pp. 124-8;
but Ballester seems to me to overstressthe differencesbetween Galen's pronouncements
on the matter in different texts - they seem to me at least, taken together, to constitute a
coherent and interesting epistemological position.
'' 'linear demonstrations' are geometrical arguments: cf. Lib.Prop. XIX 40-1; PHP
V 656.
18 See also in this context the fragmentaryOn the Substanceof the Natural Faculties
(Subst.Nat.Fac.) IV 761-4; Subst.Nat.Fac. reproduces in part Galen's On His Own
Opinions (Sent.), which survives only in Latin, and is being edited for CMG by Vivian
Nutton.
SequunturIV 767-822:it appearsin SM
19 Quod Animi Mores Corporis Temperamenta
2 32-79, Muller (1891); there is an Italian translationof and interpretativeessay upon
QAM (as well as Aff. Dig. and Pecc.Dig.) in M. Menghiand M. Vegetti, Le Passionie gli
Erroridel' Anima (Venice, 1984).

202

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

earlierhe has said that in spite of Plato'sconvictionof the soul's immortality, he himselfhas no firmviews aboutit one way or the other (772-3:note
howeverthathe does thinkthatit is betterto considerthe rationalsoul to be
both mortalandcorporeal,on the groundsthatthe alternativewouldentail
the unappetisingconclusionthat somethingdivine and incorporealwould
be the slave of the body: 782, 787). Galen's caution on the issue is both
remarkableand admirable.However, he implicitlyrelies on the principles
that nothinghappenswithouta cause,20and that all causingis bodily,2'to
reject the possibility that the soul could be (at least in its entirety)
immaterial:
all of this creates a strong presumptionwith regardto the whole of the soul that it is
not incorporeal;for how could the soul be driveninto an unnaturalstate as a result
of its association with a body, unless it were some qualityof a body, or some form,
or some affection, or some power of a body? (QAM IV 788)

And Galen admits that even after much reflection he is unable to form any
clear conception of what incorporeal souls could really be, and how they
might be differentiated (776-7).
Whatever one thinks of those robustly anti-Cartesian intuitions, howev-

er, Galen once againreaffirmsthe pointthat, for the doctorandphysiologist at any rate, such enquiries are as superfluous as they are inconclusive

(QAM IV 788). He cannotsay anythingwith certaintyaboutthe natureor


essence of the soul:22but qua doctorhe does not need to (cf. in this context
Subst.Nat.Fac. IV 759, 763-4).23What is obvious, on his view, is that some

conditions which are clearly mental in nature (at least as regards their
effects or manifestphenomenology),such as delirium,depression,drunkenness,24 and insanity, are consequent upon physical alterationsin the
body, and consequentlysusceptibleof physicalcures:
so even those who postulate a special substancefor the soul will have to agree that it
is subordinate (douleuein) to the temperaments of the body (QAM IV 779; cf.
777-9, 788),
' This principle is stated frequently throughout his works: see particularly in our
contexts PHP V 389-90; 544, where it is described as 'one of the things known to
everybody'; cf. MM X 36, 50, where it is cited as a basic axiom.
21 For this Stoic commonplace: SVF 1 89, 2 336, 340, 341, etc.; cf. Sextus Empiricus
Adversus Mathematicos(M) 9 232ff; Galen himself subscribes to it explicitly at PHP
V 567; and the rejection of action at a distance forms the backdrop to what I call his
'general causal axiom': see below, p. 222.
2 See also On the Formationof the Foetus (Foet.Form.) IV 699.
3 See Ballester, 1988, pp. 124, 126.
24 He expends a great deal of time expounding Plato's views on the subject; QAM
IV 808-12.

203

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

althoughit shouldbe stressedthatGalennowhereexplainsthe natureof the


dependence.25Indeed, this is something that Plato agrees with (QAM
The sumIV 789-91),26 as does Aristotle,27and 'the divine Hippocrates'.28
phonia of the great authorities,so dear to Galen'sheart and the principal
probandumof QAM29(as indeed it is of PHP), is not threatened.
Yet, for all that he is concernedto point out the basicagreementof the
great men of the past, Galen is not blindto their genuinedifferences.His
syncretismis neither utterly uncriticalnor wholly desperate. Indeed, the
purposeof invokingthe 'divineHippocrates'is to reinforcethe claimthat
not merelythe lower partsbut the rationalsoul too is dependentupon the
constitutionof the body (QAM IV 804-5),whichis somethingthatPlato, at
least in the Phaedo (90-95), is at painsto deny (althoughhe appearsnot to
do this in Timaeus:see n. 26 above; and cf. Rep. 10, 610aff.), as well as to
emphasizethat not only are local disturbancesin the properregulationof
the soul attributableto environmentalinfluences,but its basic characteristics (ethe) are as well.A0
See Ballester, 1988, pp. 131-3.
2 Galen quotes Timaeus86c-e in this regard, perhaps a little tendentiously (so thinks
D.S. Hutchinson:see his 'Doctrines of the mean in fourthcenturymedicine, rhetoricand
ethics', in R.J. Hankinson [ed.J Method, Medicine, and Metaphysics[Edmonton 1988:
Apeiron Supp. Vol. XXI], on the grounds that Plato means only to emphasize that the
cases of the mind and the body areparallel(Tim. 87c-89dis at least neutralin this regard);
but Plato expressly says at ib. 86b: 'suchis the mannerin whichbodily diseases arise;the
disordersof the soul, which depend upon the body, originate as follows'; and so Galen's
readingof him seems justified (for Galen's Platonicinheritancein regardto the question
of responsibility,see my art. cit., n. 1). Galen quotes furtherfrom Tim. 24c and 80b (as
well as from Laws 5 747d), in order to show that the 'Platonists'are unfaithfulto their
master in regard to the question of whether the environment is responsible for our
characters. On Galen as a reader of Plato and Aristotle in QAM, see G.E.R. Lloyd,
'Scholarship, authority and argument in Galen's Quod animi mores', in Manuli and
Vegetti, 1988, pp. 11-42.
7 Galen proves this at length in QAM IV 791-8, with extended quotations from de
PartibusAnimalium 2 2-4, and the HistoriaAnimalium 1 8-10.
2 QAM IV 798-803, by way of extended quotationsfromAirs, Waters,Places; and from
the Epidemics, ib. 803-4.
9 Lloyd, 1988, has rightlyemphasized, however, the indeterminacyof the exact nature
of Galen's probandum in QAM: pp.33-9.
3 But once again Lloyd's remarks(see n. 26 above) are importanthere: '[f]or some of
his weaker claims that MB [i.e. the mixtures of the body] are signs of CS [i.e. the
capacitiesof the soul], that MB may influenceCS, he can quote supportingtexts and give
some evidence, though both texts and evidence are already interpretedin a distinctive
Galenic manner. At the same time the generalisationsabout CS following MB and MB
even constrainingCS allow the impressionto be createdthat causallinksbetween the two
might be established with some systematicityand in some detail.' (1988, pp. 38-9).

204

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

The soul, for Galen, is a locus of capacities:


Everyone knows that we possess souls: for we plainly see the things that are
activated through the body, walking, running, wrestling, the many varieties of
perception; and we know on the basisof an axiom that commends itself naturallyto
all of us that there is some cause for these activities - for we know that nothing
comes to be without a cause. But because of our ignorance as to exactly what the
cause of these things might be, we assign it a name on the basis of its capacity to do
what it does, a capacity productive of each of the things that comes to be.3
(Subst.Nat.Fac. IV 760)

It may turn out to be impossible to establish, on the basis of secure


demonstrations,what the physical, causal bases for these capacities are
(Galen once more runs throughthe gamutof competingviews about the
nature of the soul and its powers); and in those cases it is better not to
deludeyourselfthatyou havesecureknowledgeof somethingfor whichyou
don't possess a firm demonstration(Subst.Nat.Fac. IV 761), although it
may even so be possible to arriveat a more or less plausibleaccount. But
that there are causes, and separate causes for separate powers, can be
inferred a priori.32
Cf. in this context NatF.ac. II 9-10:
all capacities (dunameis) fall within the class of relative concepts; and they are
primarilythe cause of activities (energeiai),but also incidentallythe cause of their
(sc. the activities') effects; but if the cause is relational, since it is of what comes to
be from it and not of anythingelse, then it is clear that the capacitytoo is relational.
And so long as we are ignorant of the nature of the productivecause, we call it a
capacity. Thus we say that there exists . . . in each of the parts a special capacity
correspondingto the activityof the part. So if we are to investigatemethodicallythe
quantity and quality of the capacities, we need to begin from their effects.
32 QAM IV 769-72 spells this out; see in particular770-1:
whenever we say 'the rational soul which is located in the brain is able to perceive
throughthe mediumof the sense-organs,while it is able rememberin and of itself as
a result of seeing consequence and conflict in things, of performing analyses and
syntheses', we are pointing to nothing other than if we were to say compendiously
that 'the rational soul has many capacities (dunameis):perception, memory, intellect and each of the others' (cf. MM X 13-14, quoting Plato's Phdr. 270c-d, on the
individuation of dunameis; Eth. 4 [SM 2 25, quoted above, n. 19] and Rep. 5,
477d).
Galen goes on to make the interesting claim that each section of the soul has its own
desires (epithumiai): the rational part for truth, understanding, memory, knowledge
etc., and the spirited for liberty, victory, reputationand honour; we call the lowest part
'desiderative'simply on account of the greater variety and range of its desires, in much
the same way as 'the poet' means Homer, and 'the poetess' Sappho, 'callingthe principal
representativesof the genus by the name of the genus itself' (ib. 771). The basic source
for this (although not for the remarksabout namingand genera) is Plato once more: Rep.
9, 580d-e.

3'

205

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

In this vein, Galen rejectsthe straightforwardassimilationof the soul's


substanceto the pneuma (PHP V 605, 609; SMT XI 731; see Ballester,
1988, pp. 135-7),althoughhe agreesthat the pneumais the organonof the
soul, the meansby whichit communicatesits power:we shallreturnto this
later on. Galen clearlyowes somethinghere to Aristotle, who first raised
the question(in extremelygeneralterms)at the beginningof deAnima(1 1,
402bl-403a2) of whetherone shouldfirstexaminethe effects (erga)of the
soul, or whetherratherone shouldstartby investigatingits parts.
Epistemologicallyand scientifically,then, Galen steers a middlecourse
between excessive and unfounded dogmatismon the one hand, and a
completescepticismon the other. He is preparedto speakof matterseven
when knowledgeof them is not necessaryfor physicalor moralhealth(the
two are consideredin Subst.Nat.Fac. to be if not indissoluble,then at the
very least closely related), as long as they
bring adornmentto it as a result of knowingthem precisely(providedthat they are
indeed securely known), and bring both medicine and ethical philosophy to completion. (PHP V 762; cf. 794)

Such things are both useful, and accessible to anyone prepared to take
trouble to practise in it. And Galen proceeds to give a demonstration of his
epistemological stance:
that every body in our partof the universe [i.e. the sublunarypart]comes to be as a
result of the mixtureof elements I hold to be securely known . . . but whether as a
result of complete intermixtureof the elemental bodies themselves, or whether of
their qualities only, I do not think it necessaryto know, nor do I pronounceon the
matter (although I think it more plausible that the mixture are mixturesof qualities). But as for the soul, whether it is immortaland directsanimalsin conjunction
with bodily substances, and whether there is any substance of the soul as such, I
assert that these things cannot be securely known. (PHP V 762-3)

Furthermore, ignorance of 'ensoulment' and 'metempsychosis' is of no


consequence to medical practice. We can see how various sorts of physical
treatment affect the soul; but it is impossible to know whether the soul can
be separated from the body (PHP V 763-4).3
Thus a moderately clear picture of Galen's theoretical attitude to pronouncements about the soul (and indeed on all other controversial matters)
emerges; and it is one which is consistent with what he says elsewhere about
science and epistemological justification (notably in de Methodo Medendi
(MM) X 31-40). Broadly, we can know things either because

See also in this regardSent. 3, 7; cf. Ballester, 1988, 126-6, 135.

206

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

(I) they commend themselves indubitablyto the senses,'

or because
(II) they do so to the intellect,35

or because they are suitablyderivedfromthese two types of basicpremiss.


It is important,Galen thinks, to keep these two methods of coming to
know separate, and not to confound them - but they are both equally
serviceable.Now, this epistemologicaldualismis designedpartlyto fit into
Galen'ssyncreticprojectof derivingthe best featuresfromboth the Rationalist and the Empiricistschoolsof medicine;but there is nothingdisreputably ad hoc or indiscriminateabout the procedure. For Galen, certain
knowledgerestson evidencein at leastone of its senses- everythingthatwe
know will either be, or be derived from, a phainomenon (or a set of
phainomena)that comes to us enargos.But crucially,phainomenaare not
merelyperceptual- even axiomatictruthsof geometrycount as phainomena in this sense (MMX 36). It is vain to pretendthat anythingarrivedat by
any other route can meet the strictrequirementsfor knowledge.Put in this
light, Galen's position has both structuralclarityand methodologicalcautiousness to commend it - indeed, it sounds almost proto-Humean;and
Galen mightwell agree with Hume about the appropriatequestionsto ask
of any text of 'school metaphysics':
3' An example of this class in what follows is the claim that 'vocal sound comes out of the

windpipe' (256: see p. 215 below); at 766-7, Galen contrasts the case of medicine with
that of speculative cosmology, in which 'matters cannot be clearly adjudicated by
empirical test (peira)' and where 'such a dispute cannot be decided by clear senseperception'. Galen evidently means also to include in this class the proposition 'the ears
touch the brain' (240), and no doubt more controversiallythat 'irrationalanimals feel
desire and anger' (211).
3 Examples of this class include Plato's 'Principleof Non-Contradiction'(Rep. 4 436b):
797. Other examples given elsewhere include mathematicaltruths ('equals subtracted
from equals leave equals'), as well as metaphysicalclaims ('nothing comes to be from
nothing') and logical and semantic principles ('it is necessary that everything be either
affirmedor denied'): see MM X 36-7, and my notes ad loc. in my Galen on the Therapeutic Method, Books I and 2 (Oxford, 1991). At 240-1, in a passageto be considered below,
Galen rejects an opponents' premiss('all thingsthat are active have their source nearby')
on the grounds that it 'is neither evident to the senses nor to the mind, so as to be primary
and credible in itself (ex heautoupiston)'. At 358, he apparentlyallows a thirdcategory of
epistemologically respectable premiss: 'he [i.e. the genuine scientist] should inquire
which premisses . . . should be taken from simple sense-perception, which from experience (empeiria), either of life or the arts, and which from truths which are clearly
apparent to the intellect'; but note that the category of truthsderived from experience is
not an immediate one.

207

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

Does it contain any abstractreasoningconcerningquantityor number?No. Does it


contain any experimentalreasoning concerning matter of fact or existence? No.
Commit it then to the flames: for it can contain nothing but sophistryand illusion;'

None the less, the seeker after an authenticallyHumean, scientifically


scepticalGalen will be disappointed- for his actualpracticefails (as actual
practiceswill) to meet the severetheoreticalstandardshe sets himself;since
he claimsthat
while I am not so foolish as to make rash assertions about these things [i.e. the
substantial nature and possible immortalityof the soul], still I do claim to have
proofs that the forms of the soul are more than one, that they are located in three
different places, that one of them (the reasoningpart) is divine, while the other two
are concerned with affections, such that we are angrywith one of them, and desire
those pleasures that come through the body with the other (which we share with
plants)- and furtherthat one of these partsis situatedin the brain, one in the heart,
and one in the liver. These facts can be demonstratedscientifically. (PHP V 793)

We shalllook at those 'demonstrations'in a moment;but beforewe turnto


them, and to a detailed analysisof the types of argumentwhich Galen
bringsto bear in his discussionof the shortcomingsof Chrysippeanmoral
psychologyand their respectivestrengths,let us concludethis section with
some Galenic remarksin summary.Immediatelyafterthe passagequoted
above, Galen continues:
I made my case for this in the firstsix books of this treatise;but I said nothing about
the substance of the three forms of the soul, nor about their immortality. . . the
knowledge that the forms of the soul are situated in three places, of what their
powers are and how many they have, is useful for medicalscience and for that part
of philosophy called moral and political . . . but the further inquirywhether the
spirited and appetitive parts happen to be immortal .. . is of no use either to
medicine or to moral and political philosophy;and many philosophersand doctors
have passed over it, reasonablyenough; it belongs to the theoreticalratherthan the
practicalbranch of philosophy. (PHP V 793-4)

And for Galen psychology is, properly regarded, a part of practical


philosophy.37

Hume, EnquiryConcerningHuman Understanding,Sect. XII, PartIII (p. 165, SelbyBigge). Galen, no less than Hume is concerned with the uncovering of sophistry (see
below, pp. 209ff.). For his views on sophisms in general, and the general structureof
sophistry, see On Linguistic Sophisms (Soph.) XIV 582-98, edited by R.B. Edlow:
Galen on Language and Ambiguity (Leiden, 1977); and cf. Pecc.Dig. V 62ff.
" Again relevant is the discussion of Ballester, 1988, pp. 124ff.

208

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

2. Scienceand Sophistry:the Role of Argument


At the beginningof the second book of PHP, Galen writes:
I began with the doctrine that is of most importance ... namely their [i.e. Plato's
and Hippocrates'] teaching about the powers that govern us, their numbers, the
nature of each, and the place that each occupies in the body. And . . . I took
greatest care to avoid making assertions contrary to what is plainly apparent ...
Perhaps . . . it is better to distinguishthe different argumentspeople use when they
argue badly about absolutely anything. With references to their premisses ..
some are patently false, while others are inappropriate to the matter in hand.
Patently false are those such as . . . when one says that none of the irrational
animals feels desire or anger, as the Stoics do, or that the nerves have their origin in
the heart. Inappropriatepremisses were dealt with it length in my On Demonstration. (PHP V 211-13)

The Stoics in general (and Chrysippusin particular)are Galen's targetfor


muchof PHP; but Galenis not, as is sometimessupposed,uniformlyhostile
to orthodox Stoicism, even to orthodox Stoic psychology: at QAM
IV 783-4, for example,he is friendlyto theirelementalaccountof the soul's
composition, and Chrysippusis describedin this connection apparently
without irony for once as 'wise (sunetos)'. However Galen is implacably

hostile to the orthodoxStoic line on the position and numberof the soul's
parts, hymning Posidonius for having returned at least partiallyto the
Platonicway of truthon the issue: PHP V 390.38
38 Cf. also QAM IV 819-20, where Galen praises Posidonius for going against the

orthodox Stoic line on the uncorruptednature of children(see furthermy art. cit., n. 1).
Posidonius of Apamea (fl. 1st Century B.C.) was an 'Aristotelizing'Stoic accordingto
Strabo, II 3 8. Strabo, speaking as an orthodox Stoic, actuallysays: 'there is a great deal
of aetiologizing and Aristotelizing in him, to which our school objects on the groundsof
the non-evidence of the causes'; this has been taken to imply that 'Aristotelizing'means
'aetiologizing', i.e. supplying recondite causal explanations - and hence that the orthodox Stoics were opposed to such explanatorymanoeuvres:but those implicationsare
justified neither by the text, nor by what we know of orthodox Stoicism and of Posidonius; see J. Barnes, 'Ancient skepticism and causation', in M.F. Burnyeat (ed.) The
Skeptical Tradition(California, 1983). The fragmentsof Posidonius are collected in L.
Edelstein and 1.G. Kidd, Posidonius: The Fragments, (Cambridge, 1972: hereafter
'EK'); the references above feature respectively as T 83 (PHP), F 35 (QAM), T 85
(Strabo), EK. For an instance of Galen's partialitytowardsPosidoniusin other contexts,
see his Introductionto Logic (Inst.Log., ed. K. Kalbfleisch,Leipzig, 1896) 18 8. Here, as
elsewhere, Galen's enrollment of Posidonius into the camp of the Platonic eclectic may
not be entirely justifiable: see in the context of the psychology, I.G. Kidd, 'The Stoic
intermediates and the end for man', and 'Posidonius on emotions', both in A.A. Long
(ed.) Problems in Stoicism (London, 1971), pp. 150-72,200-15;and cf. M.C. Nussbaum,
'The Stoics on the extirpation of the passions', Apeiron XX 2, 1987, pp. 129-77
(p. 145-6).

209

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

Galen thinksthat the tnpartitionanddifferentiallocationof the soul can


be demonstratedscientifically(PHP V 793);the Stoicsby contrastheldout
for both the functionaland positionalunityof the soul (actually,this claim
requires care: the Stoics after all distinguishedseven distinct types of
psychicpneuma: SVF 2 826, 836, 879; but they held that there was no
distinctionbetween the rationaland the emotionalpartsof the soul). But
the crucialfeatureof Galen's attackis his sustaineddeploymentof formal
techniquesof logic and analysisin his attemptto show thatthe Stoicsfail to
offer scientificdemonstrationsof their position. It is not, of course, that
they don'toffer arguments- they do. But the argumentsthey deployare, in
Galen's view, unperspicuouslyformulated:they rely for their apparent
validityon hiddenambiguitiesandequivocationsin theirpremisses,andon
premissesthat fail to meet the standardof rigourappropriateto scientific
reasoning.Galen'sstrategyis to bringthis out by dissectingthe arguments
and showinghow they shouldbe rigorouslyformulated,and subjectingthe
alleged justificationsof their assumptionsto severe criticalscrutiny.This
strategyis employedin the serviceof a further,fundamentalcontentionof
Galen's:it is only by so doing, by clearlyexhibitingthe logicalstructureof
argument,that one may hope to purge science of plausiblebut fallacious
reasoning.Time and againthroughouthis works,Galenstressesthe importance of training in logic39 to anyone who wants to engage in scientific

reasoning:if one is not preparedto do this, or lacks the prequisiteinnate


intelligenceto do it, theyshouldconfinethemselvesto simpleEmpiricism.'
Let us then examinethe developmentof Galen'sattack.
First of all, he seeks to dissect and criticizethe types of considerations
they themselvesemploy in supportof their position. Their premisses,he
says, can either be inappropriate(ouk oikeia), or 'patentlyfalse' (antikrus
pseude:212, quoted above;cf. 286). I deal with the latterclassfirst. It will
be rememberedthat Galen invokestwo distinctcriteriafor the acceptance
of propositionsas being true (labelled (I) and (II) above), althoughboth
involve the notion of clear appearance,enargeia.The Stoics' view, that
'none of the irrational animals feels desire or anger' (212-13, quoted

3 See particularlyPecc.Dig. V 61-93, esp. 72-5; cf. PHP V 222, 732-3, 783; MM X 18,
39; On Hippocrates' 'Epidemics' (Hipp.Epid.) XVIIB 61-2; Soph. XIV 582-98; On
AntecedentCauses (CP) I 35, VIII 106, XI 142, XIII 170-2(CP exists only in a mediaeval Latin translation;it was edited for CMG [Supp. II] by Bardong [Berlin 19371;1 have
prepareda new edition, translationand commentary,Galenon AntecedentCauses,to be
published by Cambridge[forthcoming, 1992J:see my notes ad.locc.).
' See MM X 28-39.

210

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

above),4"clasheswithwhatis plainlyapparent,andhence cannotbe true. It


is supposed to be a datum of observation, a type (I) proposition, that
animalshave desires and emotions:just take a look at the familydog. But
such observationsare incompatiblewith the Stoic account- for they hold
that the rationaland irrationalsoul are one and the same (more precisely,
they hold that there is no irrationalsoul), and hence if animalslack reason
they must lack emotion and desire;thus are the Stoics led to contradictan
evidentdatumof experience.42But in whatsense it is 'patentlyfalse'thatthe
nerves have their origin in the heart? Surely that is not supposed to be
somethingthat we can simplyobserve.
No indeed: but on Galen's view it is, or at the very least ought to be,
directlyinferrablefrom propositionsthat meet one of the two criteriafor
certainty, that fall into either class (I) or class (II); and hence will itself
become patent. It is important that for Galen truths can become selfevident- in a methodologicallysignificantpassage(De DignoscendisPulsibus VIII 786-802), Galen describeshow he trainedhis own facultyof touch
so that he became able to detect the faint trace of the arterialsystole; but
cruciallyonce he had done this, it was to him enargosphainomenon.And
hence it is possible for a truth to be self-evident in this way even if few
(indeed, in the limitingcase, none) appreciateit as such. Muchof the restof
PHP is devoted to showinghow these truthsmay become self-evident.
But before we turn to an examinationof how that is to be done, let us
considerthe other classof inadequatepropositions,those thatare 'inappropriate'.In Galen'sview both argumentsandthe premissesof whichthey are
compoundedfall underfourgeneralheadings.In the firstplace, theycan be
'scientific'(epistemonikon),i.e. such as to be

41

Cf. 211, and 309: 'Chrysippus .. . holds that none of the irrationalanimals has the
spiritedor desiderative or rationalpart;as I said also in the firstbook [now lost], virtually
every Stoic deprives them of all these parts'(= SVFII 906). See also 338, 370-1, 32, 431,
459-60, 476, 484, 500.
42 Of course the extent to which it is an evident datum of experience may be disputed,
and certainly would have been by the Stoics themselves - they are perfectly willing to
allow perceptual and discriminatory faculties to animals, as well as according them
impulse (horne) and impression(phantasia):see SVFII 714, 83, III 169. What they deny
the aloga z6a is assent to the content of an impression,which they take to be characteristic of the rational soul; thus there need be no problem for them in accounting for how
irrationalanimalsbehave as if they had emotions and desires (and perhapseven as if they
could reason: this may have been the original point of Chrysippus'sdialectical dog:
Sextus, PH 1 69, cf. M 8 271). This is a case where Galen's own argument outruns its
evidential base.

211

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

found in the very essence (ousia) of the matterunder consideration. . . we should


firststate the essence and definitionof the thing underinvestigation,and then use it
as a standard (kan6n) and a target (skopos).43(PHP V 219, cf. 220)

Anythingelse is
superfluous and irrelevant; and this is how a premiss that is (a) scientific differs
from one that is either (b) rhetorical,or (c) used for training(gumnastikon), or (d)
sophistical. (PHP V 220; cf. 221-4)"

By 'stating the essence and definition', Galen means that we should attend

to what he elsewhere refers to (borrowinga Stoic term) as the 'common


conception'(koineennoia)of the matterin question,45andspellout exactly
whatthat conceptionamountsto. This is essentiallya matterof conceptual
analysis:
the governing part of the soul as even they Isc. the Stoics] allow is the source of
sensation and drive (horme). Therefore the demonstration that the heart is the
location of the governing part must not start from any other premisses than that
(1) it initiates every voluntarymotion in the other partsof the animal'sbody,

and
(2) every sensation is referred to it. (PHP V 219-20)

One begins with what everyone (more or less) would agree to be the
essentialfeaturesof the conceptunderinvestigation,whatis in other words
analyticallytrueof the concept.In ourcase, these turnout to be (1) and(2);
this is why Galen holds that, in proper scientific demonstrations,it is
frequentlynecessaryto replace names with definitions(MM X 50), real
definitionsthat explicateandmakepatentthe naturesof theirdefiniendain
" For Galen's use of the termskopos, see e.g. the firstsentence of On Sectsfor Beginners
(SI, = SM 3 1-19 Helmreich [1893]) I 64: 'the target (skopos) of medical science is
health, its end (telos) is the achievementof it'. Although Galen does not alwaysrigorously distinguishskopos from telos (cf. Stobaeus, Ecl. 2, = SVF 3 2-3), where he does so his
distinctioncorresponds to this characterization.Cf. MM X 217-23 for an accountof the
different skopoi in medicine.
4 The classes (a)-(d) form, in Galen's view, a descending sequence of increasing disreputability, although the proper order of the sequence is (a), (c), (b), (d) (in the
recapitulationof PHP V 221, and at ib. 222 where they are treated in reverse order, the
proper sequence is established); premisses that are 'dialectical'or 'gymnastic'are to be
preferredto those which are merely rhetorical:ib. 221. The typology is clearly Peripatetic in inspiration (see in particular Top. 1 1, 100a18-101a8; and cf. Rhet. 1 1-2, esp.
1356a33ff.): Galen views the Analytics as being concerned with (a), the Rhetoric (obviously enough) with (b), the Topicswith (c), and the SophisticiElenchiwith (d): see ib.
222. Aristotle himself calls type (b) 'contentious' (eristikoi: Top. 1 1, 100b24), and type
(d) 'paralogisms'(ib. 101a7). Cf. Albinus, Epit. 3 2; 6 4.
45
See MMX 43ff.

212

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

causalterms.46That muchis (at least in Galen'sview) uncontroversial.But


he proceeds:
what can this be shown from?Fromwhat else apartfrom anatomy?For if it supplies
the power of sensation and movement to all the parts of the body, then it is
necessary that there be some vessel growingout of it to performthis service. (PHP
V 220)

So detailed examinationby dissectionis essentialto the investigation.


Galen's argumenthere rests on some not perhapsaltogetherinnocent
assumptions;for instance, that the transmissionof any power must be a
physicalprocessmediatedby some organ(essentiallythat all causingis, on
analysis, causing by contact: I shall analyze this assumptionfurtherlater
on). But what matters is Galen's insistence that the premisses of the
argumentshouldbelong to class(a): thatthey shouldbe properlyscientific.
As a consequence, argumentbased on anecdote and authorityshould be
eschewed as falling underrhetoric,dialectic,or even sophistry:
all the premisses that are taken from men's opinions, whether those of non-experts,
class. (PHP V 227)
poets, or philosophers . . . belong to the third [i.e. rhetorical]47

One of the principalobjectionsGalen has to Chrysippus'methodof trying


to demonstratethat the heart is the seat of the rationalfaculty is that he
relies on quotationfromthe poets anddramatistsin supportof his position.
Galenpointsout thatsuch'authorities'tell just as often againstthe Chrysip4 On these issues, see my opp.cit., nn. 1 and 3 above; in the context of inquiryinto the
soul, see once again Aristotle, deAn. 1 1, 402bl2-403a2, esp. 21ff: 'andcontrariwisethe
attributescontribute greatly to an understandingof the essence (to ti einai): for when we
are able to give an accountof the appearanceof either all or most of the attributes,we will
be in a position to talk in the best possible way about the essence (ousia). For the essence
(to ti estin) is the starting-point of all demonstration, so that those definitions which don't

allow us to know the attributes, or at the very least make a reasonable estimation of
them, are all clearly enunciated for dialectical purposes and are without content.'
Malcolm Schofield has asked whether '(i) a premissof the type evident to sense perception is the same as (ii) a scientific premiss'(since Galen apparentlyurges us to make use
only of (ii) in demonstrations, and then spends some time dealingwith (i); and he further
wondered whether Galen was simply confusing (a) Stoic and (b) Aristotelian styles of
demonstration, where (a) is an inference from the phenomena to the best explanation
('there is sweating - so there are invisible pores in the skin'), (b) a deduction from
axiomatic first principles to a conclusion. I think Galen can be acquitted of this confusion, although admittedly his language is unhelpful. Basically, Galen will admit both
(a) and (b) into his science - (a) as a means of arrivingat the first principles, (b) as a
systematizationof the science that results from them (see particularlyMM X 37-50, and
my notes ad loc. in my op.cit. n. 35).
'4 See n. 40 above.

213

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

pean view as they do in favourof it (he gleefully seizes on the Homeric


accountof Tityus'spunishmentin Tartarus[Od. 11 576-811as supportfor
his belief that the liver is locus of the desires:PHP V 583-5- at any rate if
the punishmentis to fit the crime).Such'witnesses'are not only demonstrativelyuseless(althoughGalenallowsthatthey maybe broughtin to backup
a claim alreadydemonstrated[PHP V 585], or to drawattentionto some
obviousfact, e.g.about the phenomenologyof anger:PHP V 302-10,338);8
they also engenderthe famousprolixityof which Galen has the gall regularly(and apparentlyunselfconsciously)to complain.In fact
with regardto the types of premiss, Zeno and Chrysippustaughtus no method and
gave us no training,and consequentlythey are all jumbledtogether ... Often they
will begin with a rhetoricalargument,follow it with a gymnasticand dialecticalone,
follow that with something scientific, and finish with a sophistry;for they have no
idea that scientific premisses refer back to the essence of the matter under investigation, having it as their target (skopos: cf. 219, above, p. 210, n. 39); everything else is irrelevant. (PHP V 221)

The fundamentalcharge levelled by Galen againsthis Stoic opponentsis


that they pay insufficientattentionto the differingdemandsplaced upon
argumentby differentargumentativecontextsand purposes;and that they
foul thingsup as a resultof sloppinessboth in theirdeductionsand in their
deployment of logico-scientificvocabulary.In consequence, Chrysippus
adducesargumentswhichhave no probativeforce. For instance,he argues
that its centralpositionin the body impliesthat the heartis the controlling
organ (PHP V 228-9); Galen replies that first of all centralitydoesn't
necessarilyentailcontrol(thatis, it does not followsimplyfromconceptual
analysisof the notions 'centrality'and 'control'- at the very least, further
premissesare required);and secondlythat the heartis not in the middlein
any case: the navel is.49
48 Galen's use of authority is a subject in its own right, and is independently of great
interest: see Lloyd, 1988. A particularlyvivid case is that of the opening pages of MM
(X 1-20: see my remarksad loc. in my op.cit. (1), n. 5, forthcoming), in which Galen
imagines Thessalus arguing his claims to medical supremacy in front of a tribunal
comprisingthe great doctors and philosophersof the past; Galen claims that not one of
them could be found to agree with his position. But this is not simply an appeal to
authorityas such- ratherGalen's position here as elsewhere is that agreementamong the
great men of the past is a consequence of the truth of their views, and hence a weak
indication of it: but is no substitute for actual scientific investigation. Cf. Nat.Fac.
II 140-1, 178-80.
4 These 'arguments'are treated by Galen as being on a par with the claim that 'because
the brain, like the Great King, dwells in the head as in an acropolis, for that reason the
ruling part of the soul is in the brain', and others of that ilk: PHP V 230-1.

214

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

But Chrysippus'schampionshipof the claimof the heartto be the seat of


the hegemonic soul does not rest solely on such sandy foundations. He
offers an argument that at least purports, on the typology established
above, to be scientific in nature, although it too, on Galen's analysis,
rests on a faulty 'axiom' about the relationsbetween location and causal
influence(PHP V 240: see furtherbelow). The mainargumentin the Stoic
arsenal for the primacyof the heart is given in three distinctversionsby
Galen (PHP V 241-3), those ascribedto Zeno of Citium, to Diogenes of
Babylon, and to Chrysippushimself. Here is Zeno's argument:
[A] (i) voice (phonW)passes through (choreidia) the windpipe. (ii) If it were passing
from the brain, it would not pass through the windpipe. (iii) Voice passes from
(apo) the same region as discourse (logos). (iv) Discourse passes from the mind
(dianoia). Therefore (v) the mind is not in the brain. (PHP V 241)

That argument, however, is enthymematic;after a lengthy discussion


(some of which we shall returnto below), Galen concludesthat
Zeno's argumentlacks some of the premissesrequiredfor a complete formulation.
This will be more evident if we rephrase them for greater clarity, so that the
argument would be as follows: [B] (vi) voice is sent out through (ekpempetaidia)
the windpipe. (vii) If it were sent out of (ek) the brain it would not be sent out
through the windpipe. But (viii) voice is sent out of the same region as discourse.
(ix) Discourse is sent out of the mind. Therefore (x)[= (v)] the mind is not in the
brain. (PHP V 256)

The difference between [A] and [B] rests on the substitutionof the more
causally loaded 'is sent out' in [B] for the neutral 'passes'in [A]; and in
Galen's replacementof Zeno's preposition'apo' with 'ek'.52Galen's dissectionof [A]'s failureturnson the idea thatit confusesthe notionof causal
originwith thatof simpleimmediateplace of origin,andthatthis confusion
5 I translate 'phong' here and throughout as 'voice' rather than 'speech' (as DeLacy
rendersit), as it is the standardGreek for any animalnoise (even fish have it, accordingto
Aristotle: De An. II 8,), and as the Stoics clearly use the term thus: 'voice and speech
(lexis) differ, because even noise is voice, while speech is always articulate'(DL 7 57);
none the less, speech is sometimes clearly indicated.
S1 dianoia is usually rendered 'thought', and sometimes De Lacy will translate it thus
(e.g. PHP V 243, in a citation from Chrysippus,where the context virtuallydemands it);
but in general De Lacy's rendering of 'mind' seems to fit the sense better - what is
referredto in general in these passagesis not the operation, but what operates: cf. 257-8,
where Galen writes 'it is clearly possible to substitute the word "sovereign part (kurieuon)" for dianoia'.
52 This substitution is not entirely gratuitous: both Diogenes and Chrysippususe 'ek'
ratherthan 'apo' in their formulationsof the argumentPHP V 241-3; and Galen at least
claims that it is less prone to an ambiguousinterpretation.But see n. 56.

215

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

is fosteredby an ambiguityin the preposition'apo'. But this is not all there


is to the analysis.Galen continues:
Now the first premiss [(vi)] is of the type evident to sense-perception, so that it
requires no proof; for everything evident to the senses is credible in itself (pistaex
heautin). But the second premiss [(vii)] belongs neither to the class of things
evident to the senses nor to that of those evident to the intellect, since it is not one of
the primaryaxioms. 3 The argumentshould have had the followingform, if it was to
startfrom primaryand demonstrativepremisses: [Ci] (vi) voice is sent out through
the windpipe. (xi) All that is sent through something is sent out of the parts
continuous with it. (xii) The brain is not continuous with the windpipe. Therefore
(xiii) voice is not sent out of the brain. (PHP V 256-7)

(vi) and (xii) fall into class (I), as self-evident truths of observation. By
contrast (xi) is a truth of type (II) (or is at least supposed to be), and hence (I
suppose) knowable a priori. Together, (vi), (xi), and (xii) entail (xiii).

However, the argumentstill requiresthe proof of a furtherlemma 'using


the conclusion[sc. of the previousargument:(xiii)] as a premiss':
[Cii] (xiv) From the region from which voice is sent out, meaningfulvoice is sent
out; (xv) meaningfulvoice is discourse, (xvi) [= (ix)]. Discourse is sent out of the
mind; (xvii) discourseis not sent out of the brain;hence (xviii) [= (v), (x)] the mind
is not in the brain. (PHP V 257; there are, Galen notes, other ways of formulating
the argument:257-8).4

So where does this leave us? We need to go back a few stages into Galen's
refutation in order to find out. But first of all we should note the effect of the
reformulation on the argument: [A] was enthymematic, lacking the necessary axiom to articulate the inference. [C], by contrast, is fully spelled out;
and the difference is made by premiss (xi) (or (C2) in the formal version of

the Appendix), which purportsto be a universallytrue relationalaxiom;


Presumablythe 'primaryaxioms' are the same as the 'logical axioms' (logikai archai)
of MM X 37 (see my Galen on the TherapeuticMethod ad loc.); see also Thrasybulus
V 846-7; at Inst.Log. 17, Galen remarksthat virtuallyevery sound argumentderives its
soundness from an axiom, which he defines as 'ex hautoupistos logos' - on the force of
this see my 'Galen on the logic of relations', forthcomingin L. Schrenk(ed.) Aristotlein
Later Antiquity. Galen's objection to premiss (vii), the reconstructed premiss (ii) of
Zeno's original argument, is that it is unsupportedby the requisite general considerations, which is why the argumentrequiresthe additionof the 'axiom' (xi) - but of course
the 'axiom' fails to meet the appropriateepistemologicalcriteria.For similarmoves (in a
quite different argumentativecontext), see Inst.Log. 16, where Galen objects to both
Stoic and Peripateticformulationsof the argumentthat if Socratesis Lamprocles'father,
then the latter is his son, on the grounds that neither invokes the requisite axiom of
father-son relations at a sufficientlygeneral level. I discussthis case and its meta-logical
implicationsin my art.cit. above.
4 For a formalisationof these arguments,see Appendix.
S3

216

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

and it is that relationwhichis supposedto support(ii) (or (A2) ).55 But (xi)
is ambiguous- and it is on the detectionof that ambiguitythat Galen bases
his refutation.The damageis done by the apparentlyinnocentpreposition
'out of (apo).

Galen himself does not adopt the strategyof attacking(xi) in its full
generality;ratherhe fastens on (ii), Zeno's unsupportedderivativeof it,
reformulatingit as
(2a) if voice were sent out from (apo) the brain, it would not be sent out throughthe
windpipe. (PHP V 244-5) (= [B] vii)

in orderto 'replacethe word"passes"with a moreperspicuousterm'. (2a),


Galen claims, is not merely unscientific;
this premiss not only falls outside the first class [i.e. that of the scientific and
demonstrative premisses: see above] . . . but even outside the second and
third . . .; it belongs to the fourth class, the sophistical premisses, since it hides
behind a verbal form that has been given a fraudulentand sophistical ambiguityin
the hope of thus avoiding refutation. For the statement . .. 1(2a)] is unsound
because it contains the preposition 'from' (apo); in all such propositions, the
prepositions 'by' (hupo) and 'out of (ek) are umambiguous;. ... For voice sent out
through the windpipe is also sent out of something and by something; out of
something, namely the vessel which contains it, and by something, namely the
power which causes the container to move. (PHP V 244-5)

Thus (2a) might be glossed either as


(2ai) if voice were sent out by the brain, it would not be sent throughthe windpipe

or as
(2aii) if voice were sent out of the brain, it would not be sent throughthe windpipe.
(PHP V 245-6)
" (C9) of the formal version is also an importantaxiom, but I supply on
it Galen's behalf
in order to get rid of the identity-sign:note that the sense of 'in' which 'I' stands for needs
to be relatively strictly and tightly interpreted, otherwise the axiom turns out to be
obviously false.
5 One might be disposed to doubt Galen's claim here about 'ek', at least construed as a
claim about ordinary Greek - for 'ek' seems to have exactly the same indeterminacyas
Iapo', perhapsmore so. Aristotle wrote a chapterof his philosophicallexicon (Met. 5 22)
on its vagaries, and much of his metaphysicsof generation and change can be seen as an
attempt to purge Greek philosophy of the logical and metaphysicalshambles it had got
itself into as a result of too little appreciationof the logically Protean nature of 'ek' (it is
entailed by the logic of generation that everything that comes to be does so from its
opposite; but it cannot be the case that opposites cause opposites; hence generation is
impossible - that is, I think, Parmenides' argument: and it falls foul of the same
confusion). But for all that, Galen is surely substantially right about the danger of
confusion here; see n. 52 above.

217

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

(2ai) is false; (2aii) is true butirrelevant(it's not thatsense of 'outof thatis


at issue here). In general,(xi) is susceptibleof equivalentformulations,one
of them analogousto (2ai) andfalse, whilethe other, analogousto (2aii), is
true but logicallyimpotent. Consequentlythe argumentfails.
Galen reinforces the truth of this by asking the reader to consider
analoguesto (2ai) involvingurineandexcrement.If reasonwere located,as
the Stoicshave it, in the heart,andif it werethe casethatforA to controlB,
B would have to be sent out of A, then the excretoryfacultiescould not be
underrationalcontrol(since the heartis not adjacentto either the urethra
or the rectum).But such an argumentsimplyinvolvesthe same confusion:
By payinginsufficientattention
we are not, a priori at least, incontinent."7
to the properuse of languagethe Stoicsseek to drawabsurdcausalconclusions regardingthe limitsof agencyon the basisof a perfectlytruebutquite
irrelevantpropositionabouttransferandadjacency.Indeed, 'the answerto
our originalquestionremainsopen- nothingin the evidence[sc. of howthe
body is controlled]inclinesus to either view (PHP V 249)'.
3. EmpiricalInvestigation,theLocationof Functions,andthe Transmission
of CausalPower
So much for destructiveargument,Galen has, effectively, shown that the
Stoics have failed to make out a case for their contentionthat
(1) the heart is the locus of choice and control (LC);

the excretory argumentssuggest furtherthat they have to supporttheir


claim that it is impossiblethat
(2) LC lies in the brain.

It remains,however,for Galen to disprove(1) andto supporthis own view


(2).
He has establishedthatpositionallocationhasnothingdirectlyto do with
whetheror not an organcancontrolthe functionsof some other;you cannot
simplyinfer what moves what by seeing what is next to what. But then, as
'I need not provide a proof that this statement is not true: I need only ask that they
comment on the following piece of reasoning:
Urine passes throughthe genitals; if it were sent out by the heart it would not go out
throughthe genitals;but it is in fact sent out throughour choice; choice therefore is
not in the heart.
Once can construct an argumentabout excrement in the same way.' (PHP V 246). On
suchparabolai, see M. Schofield, 'The syllogismsof Zeno of Citium', Phronesisxxviii 1,
pp. 34ff.
S7

218

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

Galen notes, Chrysippushas a rejoinderto the advocatesof (2) - if they are


right, and it is possiblefor voice to be carriedfromthe windpipe,but at the
initiationof the head,58then equallyit mustbe possibleon theiraccountfor
(1) to be true. Galen representsthis concessionon Chrysippus'part as a
blunder, an admission that his own argumentwas not demonstratively
probative:59
this mayor maynot be a philosophicalmistakeon Galen'spart.
However, even if Galen's criticismof Chrysippusis justifiedad hominem,
he still needs, since he himselfrejects(xi), to be able to bluntthe objection,
even if it turns out (as I think ultimatelyit does not) that Chrysippusis
logicallybarredfrom framingit.
What he needs, then, is to show just what is requiredin the way of
connections between one part of the body and anotherfor control to be
exercised. Suchconnectionwill be physical;' hence we mustdiscoverwhat
is as a matterof fact the righttype of physicalconnection.The appropriate
route to such a discoveryis via anatomy;and anatomicalexperimentation
favoursthe nerves as being the appropriateconnectors.
The special organ of voice is the larynx,and the nerves that control its
musclesoriginatein the brain;by ligatingthese nerves,you can deprivethe
animalof vocal power(PHP V 235).61Galenreinforcesthisby pointingout
S8 apo

tes kephalhspoias tinos katarche'sgignomenes (255): Galen purports to quote


Chrysippus'sown words here.
59PHP V 255-6; cf. 261:

He conceded that it is possible that voice be sent out of the chest and through the
windpipe, while the head suppliesthe originof movement (archetes kineseos) to the
parts in that region; the argumentmust therefore not be considered demonstrative
as most Stoics supposed.
It could be objected that Galen's criticismis logically misguided: one may entertain a
position per impossibile for the purposes of making an ad hominem point, without
committing yourself in any real sense to its possibility(or alternativelyone may concede
that it is logically possible, yet deny that it is causallypossible; and one may even allow its
causal possibility, while simply denying its causal actuality). One might attempt to
defend Galen here; the Stoics rely implicitly,he thinks, on an assertionof the impossibility of there being causal arrangementsof the kind that Galen wants to defend - on the a
priori truth of the axiom (xi) (interpreted on the lines of (2ai): p. 217 above); but if it
were true a priori, then such counterfactualswould be even conceptually impossible.
That does not settle the matter- and more can be said in defence of the coherence of the
Chrysippeanposition even if (xi) is construedas an a priori axiom with immediatecausal
consequences.
60 Here Galen simplyadopts the virtuallyuniversalGreek view that causingis a matterof
transmissionby bodily contact (see n. 21 above): I examine this 'General CausalAxiom'
(GCA) further below.
61 Galen's famous experiments on the recurrent laryngeal nerve are described in AA
II 576-8: much of AA is devoted to variousexperimentsinvolvingligation and severingof
nerves.

219

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

that if the heart were responsiblefor muscularmovements,such ligations


would not affect the voice, and furtherthat as a matterof fact you can
damage an animal'sheart without immediatelyimpairingits voice (PHP
V 237).62Thusexperimentshowsus thatit is the nervesthatcarrythe causal
powerresponsiblefor voice; andas a matterof type (I) evidentobservation
we can see that the nerves in questionhave nothingto do with the heart.
Hence, by an applicationof a generalprincipleof causalmediation(GCA:
see furtherbelow, p. 222), the heartcan have nothing(at least directly)to
do with voice-production;and so (1) is false. The argumentalso provides
supportfor (2); but that supportneeds to be toughenedinto proof.
In order to do this, Galen needs to tease out the resultsof his ligation
experiments,and to reinforcethem with causal axioms. We have already
advertedto a generalaxiomof causalmediation(GCA); that will show, in
conjunctionwiththe anatomicalexperimentalevidence,thatthe nervesare
the media of transferencefor vocal power, and of consciouscontrol;63and
furtherthatwhateveris as a matterof fact the LCwill be locatedat one end
or the otherof the transferringmedium.But it won'tas yet tell us whichend
(cf. PHP V 563-40);hence we cannot yet determinewhere the LC is. In
orderto do that, Galen thinks,he mustdeterminethe origin(arche)of the
structuresin question;andit is to thatissuethathe turnshis attentionnext.
Galen wraps up his attack on the Stoic argumentwith the following
ringingparagraph:
I would perhaps say more about the fallacy of the argumentif Chrysippushad not
also recognised its absurdity . . . where he said that it is possible for discourseto be
sent out from the parts in the chest while the head providesthe origin of movement
(arche tMskinese6s), just as it is possible that the nerves all have their growthout of
the head but receive the origin of their power from the heart. Chrysippuswas
correct in saying this . . .; but what he said about argumentsfrom position, and
about those among them that rest for the most part on the evidence of poets, or of
the majorityof mankind,or etymology, or anythingof that kindwas not correct. He
would have done better to stick to the sort of premiss supplied by the scientific
method, and to examine them and judge them by means of perception (PHP
V 261-2; cf. 766, quoted above, p. 201)

The importantthinghere is thatGalenallowsthat, for all thathasbeen said


and demonstratedso far, it is possiblethat
I

Indeed you can remove it completelyand the animalwill continue to protestvocally, at


least for a while: 238-9.
'3 Galen's extraordinarilycomprehensivesequence of experimentson the nervous system and the spinal column are detailed in AA II 651-706; and see also the later books,
preserved only in Arabic (M. Simon, Sieben Bucher, Anatomie des Galen (Leipzig,
1906), pp. 13-31, 92-114, 261-73).

220

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

(3) the nerves might be the vehicles for the motor functions;
(4) the nerves have their origin in the brain;

and that
(5) the heart may in turn supply the brainwith the power of perception and choice
(dunamis aisthetikete kai proairetike:PHP V 262).

(3), (4) and (5) are not incompatible- hence, even if (3) and (4) have been
demonstratedit remainsto show that (5) is false, in orderto favour(2) over
(1). In the succeedingsentences, Galen spells out how one is to go about
doing that:
one must determine by dissection . . . the numberand natureof the structuresthat
connect the heart with the brain, and next one must cut or crush or ligate each of
them at the neck, and then observe what the effects are on the animal. The heart
and brain are connected by three kinds of vessels . . ; veins, arteries, and nerves.
(PHP V 263)

Then you mustproceedcarefullyto isolate each vessel in turnand see what


effect ligaturehas on each of them. This is difficultto do accurately,as the
vessels are very close together,andhence it is easy to trapmorethanone of
them at once (this fact, Galen thinks, is responsiblefor muchfaulty theorising about these matters:264; see p. 225 below). When you ligate the
nerves, the animalloses its vocal power(althoughthe involuntaryactivities
remain unaffected, as do some of the voluntaryones): but no such effect
happenswhen the arteriesare ligated;
hence one may readilyinfer that the heart needs no help from the brainto move the
pulse, and the brainneeds none from the heart for the animalto have sensation and
the power of voluntary action. (PHP V 264)4

Whatof those who thinkthat ligationof the arteriesproduces'aphasiaand


stupefaction'?Well, Galen says, they makea mistakeaboutthe phainomena (ligationof the arteriesdoesn'tcausevoicelessness),65
but they are right
about the conclusionsthey shoulddrawwere theiraccountof the phenomena correct:
'6 Galen wavers in quite what he assertsis lost when the nerves are ligated- sometimes it

appears only to be the voice, at other times it is the power of voluntarymovement and
sensation as well; and this vacillation blurs the outline of his argument. But it is not
radicallydamagingto it; and he was well awareof the distinctionbetween the sensory and
the motor nerve (see the passages of AA cited in nn. 56, 57 above; and in particular
Simon (1906), pp. 262-9, in which Galen describesin detail a sequence of neural ligatures
to produce a variety of different results:in no case, however, is it ligatureof the arteries
that produces the sensory and motor effects).
'5 For this reason Galen thinks that the carotid arteriesare wronglynamed: PHP V 195.

221

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

for if the animal really became stupefied (karodes)' . ., it would necessarily


follow that the heart sent out the primaryorigin of sensation and movement to the
brain, while the latter distributedit throughthe whole body via the nerves . .. For
as it was demonstratedearlier [Book One, V 187-210]that the heartis the source of
the arteries and the brain of the nerves, the conclusion would follow if their
statement were true, that the heart supplies the brain with psychic power via the
arteries (PHP V 265-6).

All of these argumentsmake use of the followinggeneralcausalaxiom:


GCA: If A transmitssome causal power P to B, then either (a) A is adjacentto B,
or (b) if A is remote from B, then there is some vessel V such that V links A
and B and P is transmittedthrough V.

The Stoic errorin the windpipeargumentcan now be construedas a failure


to see that (b) is a legitimateoption, and a correspondingassumptionthat
GCA amountedto its firstdisjunct(a).
But GCA on its own won't deliver Galen's conclusion; he requires
anothercausal axiom, presupposedas commongroundbetween him and
his opponents in the argumenthere, namely an axiom of causal priority
(ACP). This could be formulatedin a varietyof ways, but intuitively(and
semi-formally)it states that
ACP: Some organ 0 is only causallyprimaryin regardto some power P if it is not
the case that there is some further organ O' which is prior to 0 and
responsible for P.

ACP as it stands is both unclear and formallydeficient. The notion of


I
responsibilityneeds givingformalflesh, as does the conceptof priority.6"
shall not venturefurtherinto that particularminefield,and trustoptimistically that the general lines of the claim being made by ACP are clear
enough, even if its detailsare controvertedandobscure.But it mustbe the
case that 'necessary' and 'responsible'are to be construed in a sense
strongerthan that of simplybeing causalprerequisites- for, in that weak
sense, the heartwill satisfyACP (animalscan'tfeel or undertakevoluntary
movements unless they are alive; and the heart is necessaryfor life).!8
Indeed, failureexplicitlyto treatof these issuesseems to me to be perhaps
the most seriousflaw in Galen'scausalanalysis.He is of coursewrongthat
the carotidarteriesdo not affectconsciousness;buthe couldallowthatthey
Galen glosses karodes thus:
hoper autois onoma bouletaisOmaineinto anaisthetonte kai akineton (266).
67 Again informallyA is prior to B just in case A is necessary for B: but the difficulty
merely resurfacesin a new guise - 'necessary'in what sense?
I For a discussionof the difficultiesinvolved in determiningthe precise sense of 'causal'
and 'responsible' in related contexts, see my art.cit., n. 1 above.

222

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

did, and still preserveboth (1) andACP, if he allowsthat 0 can be primary


for P even thoughsome O' is (in some sense) responsiblefor the continued
functioningof 0; O' of coursecannotbe responsiblefor P (by ACP) - butit
does not seem that such a state of affairscould not obtain. However, Galen
appearscommittedto rejectingthis possibility.69
But howeverthat may be, we still need some wayof supporting(or at the
very least of understanding)Galen's claim to have demonstratedthat the
heart is the origin of the arteriesand the brainof the nerves, as he claims
alreadyto have done. For surely all he has shown so far (at best) by his
experimentsis that there is causal communicationvia the nerves and the
arteries- how can he claimto have shownits direction(cf. PHP V 563-4)?
Let us turnto a passagein Book Six of PHP, afterthe lengthyrefutationof
the Stoic accountof the passionswhichhasoccupiedmostof the intervening
section of the treatise.70
In Book Six, after establishing(at least to his satisfaction)the truth of
Plato's doctrine that the rationalpower is located in the brain, while the
spiritedpart residesin the heart,7"he turnsto the case of the liver, and the
appetitivepart:
this proof will not be from such clear evidence . ., nor are its premissestaken from
the very nature of the thing under investigation, (cf. 219, 227: and see above) but
from properties peculiar to it (ek ton tou6i subebekot6n idiai).72 For when the
nerves were stopped with ligatures or were cut, we could see that the parts
continuous with the brain retained their original powers, but those beyond the
ligature immediately lost both sensation and motion. And similarlywith the arteries: we saw the naturalpulse still remainedin the arteriescontinuous with the heart
but disappeared completely in those that were separatedoff by ligature." Again it

v'

See PHP V 532-4, a passage discussed below.


lhis treatment forms the subject-matterof my art.cit., n. 1.
At 333-4, Galen recapitulateshis argumentconcerning the brain and the heart:
one must begin from [their] attributes and properties . . . as they pertain to the
essence of the matter . . . [t]he main ones were that the brain is the source of . . .
nerves which transmitsensation and voluntarymotion . . . as the heart is the source
of the arteries . . . and when the . . . brain is pressed or wounded the whole animal

becomes stupefied, but neither the arterial motion nor that in the heart is destroyed . . . We also showed that neither supplies these powers to the other . . . but

each of them is as it were the fountain of its own power.


n The phrase is Aristotelian: see De An. I 1, 402a5; for some examples of 'properties',
see n. 71.
73 On Galen's experiments with arterial ligature, and the (false) conclusions he drew
thereby concerning the transmissionof the power of the pulse, see WhetherBlood is
NaturallyContainedin theArteries(Art.Sang.) IV 702-36;Art.Sang. is edited with useful
notes and introductoryessays by D.J. Furleyand J.S. Wilkie in Galenon Respirationand
the Arteries, (Princeton, 1984); see also M.P. Amacher, 'Galen's experiment on the

223

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

was clearly evident that the disturbancesof the soul that occur in anger and fear
cause the heart to depart from its naturalaction. We also mentioned in how many
ways the whole body is harmedby pressureon the brainor damageto its ventricles,
and that this too clearly indicatesthat it is the sourceof motion and sensation. But in
the case of the liver we are unable to make any such demonstration, whether by
exposing it and applyingpressure,or by ligatingthe veins. For it is not the source of
obvious motion as the heart is of pulsation and the brainof sensation and volition;
nor is it the cause of rapidinjury,as the others are, but it takes time for the weakness
of the liver to harm the animal's nutritionand colouring. (PHP V 519-21)

There are two importantfeatures of this passage - one is the eminently


plausibleidea thatit is easierto see thatsome interventionhashadan effect
if the effect in questionfollows immediatelyor rapidlyupon the intervention.74The secondis that, given thatcausalinfluenceis directional,flowing
one way only, if you interveneat some point in the vessel whichtransfers
the influence such as to interruptit, then all points downstreamof the
interruptionwill be affected,whilethose upstreamof it will not be affected.
That aqueousmetaphoris not accidental- Galen himselfmakesconsiderableuse of it, and it is arguablethatthe metaphorcoloursandconditions
his accountof the fluxionof bloodin the body:75at PHP V 572, he describes
the veins as ochetoi, conduits;76and in his defence of the role of the liveras
the source (arche)of the veins at 545-7, he writes:
if you wished to describe the distributionof water broughtinto a city, you would not
pass over its first entrance and find some other point from which to begin the
account; there is every necessity first to speak of that place in the city where the
water first arrivesfrom outside and from that beginningto proceed to describe the
rest. Therefore you should not look for one starting-point(arche')of the nature of
the thing, and another for instruction(didaskalia)about it. (PHP V 546)

Galen is makingthe ad hominempoint that even those who deny that the
liver is the archeof the veins none the less begin theiraccountof the veins
from the liver, which suggestsits primacy.
Nevertheless,the water analogyis not, as we shall see, entirelya happy
one; andindeedthe mostprominentmetaphorthatGalenemploysis thatof
the branchesof a tree - the veins split off fromtheirbasictrunk(555-7;cf.
arterialpulse, and the experiment repeated', Sudhoffs Archivfur Geschichteder Medizin, 48 (1964).
7 Cf. in a different context In Hippocratisde Natura Hominum (HNH) XV 161-2 (ad
ch. 13).
7 Cf. in this regardI.M. Lonie, 'Erasistratus,the Erasistratians,and Aristotle', Bulletin
of the History of Medicine 38, (1964), pp. 426-43.
76 Cf. Tim. 77c-e, where Plato employs the same languageand a similarmetaphor(cf. ib.
78c-79a).

224

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

530);thisimageis developedmostfullyat 522-7,whereit is interwovenwith


that of the water-supply;one short quotation from this passage should
suffice to give its flavour:
it is evident to anyone who is versed in the ways of naturethat largerthings are the
sources (archai) of smaller, just as the springis greater than the channels (ochetai)
into which it is divided. And yet some have attained the peak of unreason by
supposingthat what follows the archeis greaterthan the arche.7 They are misled by
rivers, which are very small at their sources, but increaseas they advance . . . Some
riversgrow larger . . . when tributariesflow into them, while some decrease in size
as channels branchoff. No riverthat comes froma single springis smallerat its head
than it is thereafter, but if it is collected from many springsit is reasonable that the
whole should become larger than any one of them. (PHP V 525-6)

So only certaintypesof watercoursewill serve as suitableanaloguesandno


doubt Galen is thinkingof irrigationsystems, or other sorts of artificial
water-supply.But what entitleshim to do this, in his view, is the absurdity
of supposingthat in animalsthe sources of the nutritivepower could be
various and at the extremities,flowing together to some confluence:the
point is made partlyby appealingto analogy, partlyby appealto reason:
[you will be] compelled to admit that the source of the . .. arteries, nerves, and
veins is in every partof the body, so that the heel perhaps,or the fingeris the source
of the largest artery . . . So perhapsthe person who introducesthis argumentis not
ashamed to say that the branchesof trees or the ends of roots are the archaiof the
plant . . . simply to say that the ends of the veins are archaiis absurd . . . every part
will be a source. If they reply that some of the ends are archai and others not they
will be putting forwardan undemonstratedassumption. (PHP V 526-7)

The model, in fact, is clearly that of irrigationsystems- the blood, like


waterin irrigationditches,flowsout froma centralsource, andis dispersed
to and absorbedby the extremities.78However, the assumptionthat it is
logicallyentailedby the conceptof a source,an arche',thatsucha thingmust
be unitaryand undivided(hence the remarksaboutthe impossibilityof the
root-endsbeing archaifor the tree) is clearlya very strongone, and on the
face of it unjustifiablyso. It is worth remarkingin this context that Galen
does hold that the root-ends are analogous to mouths (527: see n. 96
below), hence he cannot simplyhold that it is impossiblefor a varietyof
7 This sort of claim bears comparisonwith Descartes' famous idea that there is always at
least as much 'reality' in the cause as in the effect': MeditationsIII; although Galen's
employment of an analogous principle here is I think less philosophicallysuspect than
Descartes'.
78 For the pervasivenessof this model, and its possible effect of suppressingthe emergence of the hypothesis of the circulation of the blood, see Lonie, 1964 and see Timaeus
77c-e, 78c-79a.

225

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

separate parts to contribute to the same overall function. Rather, I think


the crucial notion is that of control: if control over some function is to be

exercised, then it must be done centrally- causalpower ramifiesout from


the original source.79 That is a powerful intuition, and I shall not pursue the

matterfurther.I am not convincedthatit is in all casesjustified,however.'


Galen'saccountis also guidedby analogywiththe case of the heart- for
in the case of that organ, he thinks you can just see directlythat it is the
source of arterialactivity,since
the heart alone of all the parts in the animalis seen to preserve its naturalactivity8'
for a very long time even after removal. (PHP V 531; cf. 185-7, 238-9; 561-3)

But of courseit is impossibledirectlyto see this in the case of the liver- and
hence its role in connectionwith the veins mustbe inferredby analogy,an
analogythat involvesappealsto the similarityof the branchingstructureof
the veins to that of the arteries, and to variousstructuresto be found in
plants.Galen'streatmentis muchmorecomplexanddetailedthanI can do
justice to here; but I shall brieflyconsiderone subsidiaryset of arguments
involvinganalogybefore turningto the actualaccountof the functioningof
the soul.
At PHP V 532ff., Galen considersa possibleobjectionto the claimthat
the liver is responsiblefor the nutritivepowersuppliedby the veins. Could
not the liversimplysupplythe material,whilethe heartinjectedthe power?
This is rejected partly on conceptual grounds;if an organ supplies the
matter, then it must be the source of the nutritivepower:for
it is reasonable that what provides the whole body with mattersuitablefor nourishment is the source of the power of nutritionand growth. (PHP V 533)

But that is not the whole story- if the heartpossessesthe nutritivepower,


then either the liver is superfluous(but nature,for Galen even more than
for Aristotle, does nothingin vain),82or the liverfunctions'like a servant',
7 Indeed, ramificationof this sort is sometimes taken to be the definingcharacteristicof
causal direction: see e.g. K. Popper, 'The arrowof time', Natureclxxvii (1956), 538; see
also J.L. Mackie, The Cementof the Universe(Oxford, 1974), ch. 7.
' Recent work on the nature of the immune system, and the proper way to describe it,
provides an example in point: see in this regard the paper by E. Levy 'Networks', in
Matthen and Linsky, 1988 (n. 82).
81 i.e. that of causing pulsation in the arteries: it is not quite clear what he has in mind
here; nor does it seem, on any interpretation,actually to be true.
I See my articles (1) 'Galen explains the elephant', in Matthen and Linsky (eds.)
Philosophy and Biology (CanadianJournalof PhilosophySupp. Vol. 14, 1988), pp. 13557, and (2) 'Galen and the best of all possible worlds', in Classical Quarterlyxxxix 1
(1989), pp. 206-277.

226

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

as Alexanderof Aphrodisiasbelieved.83Galen rejectsthison the basisof an


axiom:
(A): every organ that prepares material for another keeps it untouched for that
organ; (PHP V 534)

and (A) is taken to entail (in conjunctionwith a furtherunstatedaxiom,


which is nevertheless Galenic, that Nature will arrangethings economically)'"that there will be only one distributingchannelin the case of such
organs, leadingdirectlyto the organit supplies(Galen instancesthe route
fromthe lungsto the heart,fromthe stomachto the liver, andfromthe rete
mirabileto the brain); but the routes of distributionfrom the liver are
multiple, and hence do not matchthat model.85
But a crucial question remains to be addressed:Galen has admitted
(PHP V 519) thatin the case of the role of the liverhis 'proof'has not been
'from such clear evidence' as in the case of the heart and the brain;is he
therebycommittedto thinkingthat he has not genuinelyoffered a demonstration,at least not one whichmeetsthe strictandrigorousmethodological
requirementslaid down at the beginningof Book Two (PHP V 219-24;see
above, pp. 211-14)?Is he, willy-nilly,introducingpremissesthatare at best
dialectical(or even worse) into the heart of what ought to be a scientific
inquiry?The answerto that questiondepends, I think,on the natureof the
analogiesthat Galen uses, andthe role thatthey are supposedto play in his
demonstrations.He admits that he cannot offer an absolutely certain,
watertightdemonstration'from the nature [i.e. the essence] of the thing
under investigation';but that does not commit him necessarilyto mere
dialecticor rhetoric.The dialecticalpartof his exposition,the confutation
of his opponentson the basisof theirpremisses,is alreadyover - and he is
certainlynot' aimingat mere conviction,or worse still intellectualfraud.
I

DeAn.3, pp. 95-7 Bruns (Suppl. Arist. 11.1).

84 For a discussionof the axiom (or axioms) involved here, their Galenic provenance and

justification, see my artt.cit., n. 82 above (esp. (1), pp. 151-5). Briefly, the two principles I ascribe to Galen are what I call the Principleof Creative Economy (roughly that
Nature is not prodigal, and that the skill of the demiurgemanifests itself in the economy
with which he makes use of the materials available to him), and the No Redundancy
Assumption, namely the view that Naturewill not put more work into its creationsthan is
demanded by the exigencies of structureand function.
8' It is not clear what precisely is the status of (A), whether it is supposed to be an
empirical generalization, or an a priori truthof some kind: presumablyif it is the latter,
then it will have to be underpinnedby considerationsof the sort outlined in n. 84.
1 At least ostensibly - others, holding a lower opinion of Galen's intellectual honesty
might none the less suspect him of fraudulence here. But it is surely more interesting
philosophically, as well as more charitable, to succumbto such suspicions only as a last
resort.

227

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

Of coursehe mightstill be guiltyof it - Galenfrequentlyremarksthatthere


are two reasonswhypeople become'entrapped'by sophisms.87
Theycanbe
so either as a resultof ignorance,inexperience,and lack of trainingin the
logical methods, or as a resultof a deliberatedesire to cheat and mislead.
Perhaps,then, Galen is the sad, unwittingvictimof the very methodological failingshe discernsin others?
I thinkhe can, methodologicallyat least, be acquittedof this charge.To
rely on the 'propertiespeculiar'to something(see above, p. 223) is not to
resort to ad hominemargument,or mere rhetoric.Rather, in this case at
least, the propertiespeculiarto the liverare its structuralfeatures(thatit is
the originandsourceof the veins, giventhe constraintson whatcancountas
a sourcewhichhave alreadybeen discussed);one cannotdirectlyassessits
functionalrole, for reasonsthat PHP V 519-21(above, pp. 223-4)spelled
out. Now, for Galen the link between structureand function,while close
andperspicuous,is not a matterof immediateevidence,noris it in generala
simplelogicalrelation;it mayhaveto be inferredon the basisof a varietyof
Those considerationsare, in this case, precisely
complexconsiderations.88
those of n. 83 above.
Galen is wrong about most of this: but he goes wrongnot becauseof a
simplefailureto follow his own strictureson method.Ratherthe rootof his
errorlies in a too sanguine(but I have arguedelsewherewell motivated)
acceptanceof a particularlypowerfullyinterpretedteleology on the metaphysicalside, and an over optimisticepistemologicalbelief that the empiricalsciencescan be reducedto indubitableAristotelianaxioms.The upshot
of this is that the analogies he employs are not meant in themselvesto
constitutethe whole of the argument;that is, he does not feebly conclude
fromthatfactthatthe liverlooks like the heart(in some respect)thatit acts
like it too. Rather, the analogousstructureis a guide to the rightanswer,
because it serves to remind the investigatorof a basic teleologicaltruth
aboutthe arrangementof the world.
Of course, that in itself will not do all the work that Galen needs to do
here - nothing will. But it is relativelyeasy to see how the rest of the
argumentfor the primacyof the liverin relationto the veins, andof its key
role in the distributionof nutrientsvia the blood, mightbe plausiblyfilled
I

Cf. Pecc.Dig. V 72-5 for Galen's clearest accountof his notion of a sophism; see also
Nat.Fac. II 44ff., 51-3.
N See my artt.cit., n. 82 above for details.
9 artt.cit., n. 82.
9 For a discussionof the latterfeaturesof Galen's philosophyof science, see my 'Galen's
account of scientific knowledge', in J.A. Lopez Ferez (ed.).

228

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

in, on the basis of the (admittedlyquestionable) causal axioms earlier


discerned.
4. Desire and the Liver
Let us, then, allow that Galen has indeed establishedthat the liver is the
source of the veins and of the elaborationof the blood. But why imagine
that that has got anythingto do with the desires?Is he not committingan
ignoratioelenchion a grandscale by simplyassimilatingthe source of the
veins andthe appetitivesoul?Hippocratesis Galen'smodelon the nutritive
front,9 Plato on the philosophical(Tim. 70d-e); and Galen assertsthat
it makes no difference whether the liver is called the source of the veins, or of the
appetitive soul, but it is more appropriatefor a physicianto present his teaching in
terms of bodily organs, a philosopherin terms of powers of the soul; in either case;
the one follows from a proof of the other. (PHP V 577)

Galen then quotes from the Timaeus(70d-71b),on the constructionand


locationof the appetitivesoul; it is that partwhichis 'desirousof food and
drink, and all that it requires because of the nature of the body'; i.e.,
matters connected with nutrition.The argumentappearsto be that, because the liver is the elaboratorof productsnecessaryfor the nutritionof
the body, it mustalso be the locusof attractionfor the rawmaterialsthat go
to make up those products,andhence mustbe the sourceof the desiresthat
go along with that attraction.Galen, it must be remembered,is operating
with a broadconcept of 'soul' and the psychic;and he takes it from Plato
that even plants have a certain form of perception and hence of desire
(contra the general thesis of Aristotle in de Anima 3 10-11, and of the
Stoics:PHP V 521), namelya desirefor and attractivefacultyof whatever
is suitablefor theirnutrition,anda correspondingrepulsionfor whateveris
alien to their systems.92
But why imaginethe locus of desire for some set of substancesneed be
He quotes many times from de Alimentoon the subject of the rhizosis [literally,where
the rootgrowth of a plant becomes the trunk] of the veins: de Alim. 31 = CMG I 1,
82.13ff.; see PHP V 199-200, 531-2, 543, 577, 578.
'I Subst.Nat.Fac. IV 764-5; cf. PHP V 518, 521ff; and see Nat.Fac. II 159-60:'thus it is
confirmed by all the phenomena . . . that there must obtain in almost every part of an
animala certain inclination(ephesis) and, as it were, a desire (orexis) for their particular
quality, and an aversion, or as it were a hatred, for the alien quality.' Nat.Fac. consists in
large part of an attack on the physiology of Erasistratuswho denied the existence of
'naturalpowers' in the bodily of organsof attraction,repulsion, excretion and so on, and
who sought to explain the body's mechanics solely in terms of such purely physical
principles as horror vacui.

91

229

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

the same as the organwhichmakesuse of them?Whycould not one organ


desireon behalfof another?Well, perhapsto havea need for something(at
least in a suitablystrongsense), and to take steps to get it, just is to desire
that thing, other things being equal; Galen would be sympatheticto ElizabethAnscombe'sremarkthat 'the primitivesign of wantingis tryingto
get' (Intention[19571Section36). The liverclearlytriesto get things- hence
it wants them.
But even if we admitsuch an extensionalnotion of wantsand desires, it
will not, surely, suffice to ensure that 'it makes no differencewhetherthe
liver is called the source of the veins or the appetitivesoul', at least if by
'appetitive soul' we mean a locus of conscious (or at least potentially
conscious)desiresfor suchthingsas food, drinkandsex. It seemsas though
thereis simplya grossfallacyof equivocationgoingon here:mylivertriesto
get food; I try to get food; hence I (qua desire for nutrition)am my liver.
I think there are broadlytwo lines of defence open to Galen here. The
firstrelies on his Platonicinheritance.Thirstproperlyconstrued,for Plato,
is not the desire for a particularkind of drink, or for a good drink, but
simplyto drink(Rep. 437bff.);andthe samegoes for the otherdesires.The
level of articulation- the degreeof propositionalor cognitivecontentif you
like - is low, perhapsabsentaltogether.If thissortof nisusdoes carryalong
with it some sort of propositionalarticulation,then that is secondaryand
subsidiary,a consequence(perhapseven an epiphenomenon),andcertainly not a causeor component,of the desireitself:it is, in the languageof John
Searle, simply 'the froth on the wave'. When Galen remarksin his short
that one can never
treatise Thatthe Best Doctor Be also a Philosopher93
developthe habitsof hardworknecessaryto scientificachievementif one is
'a slave to the belly or the genitals' (Opt.Med. I 59), he employs that
metaphorof slaveryin a morethanusuallyliteralsense.' At thiselemental
level, it makes more sense to ascribethe sourceof the cravingto the part
which 'desires' (in a broad, non-intentionalsense) the sustenance,which
attractsthe raw-materialstowardsit, and elaboratesthem into something
else.95 Here again, Galen makes use of an analogy (it is not clear how

innocently):the rhizosisof plantsis like the liver- for the liversendsshoots


downwardsto the stomach to draw up nourishment,and upwards(the

Opt.Med. I 53-63, - MS 2 1-8, Maller (1891).


9 For Galen's use of the metaphorin moral contexts, see my art.cit., n. 1, forthcoming.
s These various processes of elaboration, in the heart, the stomach, the liver, and the
seminal vesicles, are central to Galen's metabolic system: see PHP V 565-74.
93

230

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

veins) to distributethe elaboratedproduct.'


The second defence appealsto teleologicalconsiderations(but is also, at
least in inspiration,Platonist).Supposeit reallyis the case that one organ
(the brain, let us say) desiressomethingon behalfof another.What could
be the point of such a duplicationof functions?A properlyeconomical
creatorsimplywouldnot toleratesucha wasteof resources.' Of coursethat
is not say that there is no role for the brainto play in regardto desire: its
function precisely is to weigh the raw data coming from the desiderative
partand to determinewhetherdesiresshouldor shouldnot be acted upon;
in orderto do thisthe desiresclearlymustbe givenpropositionalform- and
thatperhapsis sufficientto explainthe need for consciousdesires,at least in
rationalanimals.98
5. Conclusions
Let us briefly,then, take stock of whatGalen thinkshe has demonstrated.
Firstly,he thinkshe hasestablishedtripartition,alongPlatoniclines;but he
has done morethanthat- in demonstrating(at least to his own satisfaction)
the general agreementof Plato and Hippocrateson these issues, he has
shown how the physiologicalaccountof the functioningof the partsof the
body, of theirmutualinterdependence,andof the structuresthatjoin them
together,complementsthe philosophicalaccountof the separationof parts
of the soul. And he has elaborated and applied a method of scientific
investigationand argumentthat purportsto distinguishbetweencertainty,
mere plausibility,and outrightsophistry,and to assign authorityits due
weight. It must be admitted,I think, even by the most fervent admirerof
Galen's work, that his applicationof the method falls short of the high
idealshe sets himself.But it is the fact thathe sees the idealsso clearly,that
he recognizesso acutelythe need to marryconceptualrigourwith skill and
dedication in empirical investigation,that is impressive. However, just
wherethatleaves his attackon Chrysippeanmoralpsychology,andhow the
philosophicalandscientificresultsobtainedcanbe integratedwitha plausible accountof the phenomenologyof affection, and the theory of human
' PHP V 522-7, esp. 527: 'the ends of the roots can be called sources of the tree's
nutriment,just like the veins that descend into the stomachand the arteriesinto the lungs
[i.e. for the heart]'; and 556ff. See pp. 225-6 above.
7 For a detailed account of Galen's creatonism, see my artt.cit., n. 82 above.
I I examine in more detail Galen's pictureof the role and functionof the intellect in the
properly managed rational life in my art.cit., n. 1.

231

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

responsibility, is a question I leave for another occasion.99


University of Texas at Austin

Appendix: A Formalisationof the Stoic Arguments


It is perhaps worth setting out these argumentsin a formal manner, using the
techniques of modem symbolic logic. [A] translatesas follows:
[A]
1
2
2
2
1,2
6
6
1,2,6
9
10
9,10
1,2,6,9,10
1,2,6,9

(1) Tvw
(2) (x)(Sxb -TxW)
(3) (x)(Txw - Sxb)
(4)Tvw--Svb
(5) - Svb
(6) (x)(Svx * Sdx)
(7) Svb ++Sdb
(8)-Sdb
(9) Sdm
(10)m = b
(11) Sdb
(12) Sdb &-Sdb
(13) - m = b

A
A
2:contr.
3:UE
1,4: MPP
A
6: UE
5,7: MPP
A
A
9,10: = subst
8,11 &1
10,12: RAA

(i)
(ii)

(iii)

(iv)

(v)

(v = voice; w = windpipe; b = brain;d = discourse;m = mind; Txy = x passes


throughy; Sxy = x passes from y)
[C] makes use of two furtherabbreviations:'Cxy' = x is continuouswithy; 'Ixy'
= x is in y (the latter allows us to drop the use of the identity-signfrom the
argument):
[Ci]
1
2

2
1,2
1,2
6
1,2,6

A
(1) Tvw
(2) (x)(y)(Txy -. (z)(Sxz -+ Czy) ) A
2: UE (two steps)
(3) Tvw-- (z)(Svz-- Czw)
(4) (z)(Svz -- Czw)
1,3: MPP
(5) Svb - Cbw
4: UE
(6)- Cbw
A
(7) - Svb
5,6: MT-

(vi)
(xi)

(xii)
(xiii)

9 The other occasion being my art.cit., n. 1 above; I should like to thank Malcolm
Schofield for his extensive and acute editorial remarks.

232

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

But, given that discourse = df* meaningfulvoice (xv), and given that whatever
goes for voice goes afortiori for meaningfulvoice (whichentails (xiv) ), then we
can substitute 'd' for 'v' throughout [Ci], and hence obtain
(7*) - Sdb

(xvii)

Now we can formalize [Cii] as follows:


[Cii]
8
9
9
9
8,9
1,2,6,8,9

(8) Sdm
A
(xvi)
(9) (x)(y)(Ixy -- (z)(Szx -- Szy)) A
(10) Imb (z)(Szm-+ Szb)
9: UE (two steps)
10: UE
(11)Imb -(Sdm --Sdb)
8,11: MPP
(12) Imb Sdb
(13)-Imb
(xviii)
7*, 12: MTT
-

233

This content downloaded from 128.111.121.42 on Wed, 25 Sep 2013 11:20:17 AM


All use subject to JSTOR Terms and Conditions

You might also like