MEDIEVAL AND

EARLY MODERN SCIENCE

JOHANNTES M. M. H. THIJSSEX University of Nijmegen

CHRISTOPH LUTHY University of Nijmegen

Editmzial Cmlfants

Joik BIARD, University of Tours Smo &uunuA, Univemity of Helsioki Jom- E. M m o c ~ , Harvard University

J~:RGEN h, hiax-Planck-Institute for the History of Science Tmo VERBEEK, University ofUtrecht

I THE METAPHYSICS I I AND NATURAL PHILOSOPHY i I OF JOHN BURIDAN

EDITED BY

I J.M.M.H. THIJSSEN AN^ JACK ZUPKO

BURIDAN'S THEORY OF DFTINITIONS IN HIS SCIENTIFIC PRACTICE

Just as the proof of the pudding is in the eating, so too, the proof of the viabili'y of a theory of scientific method Lies in whether it works in scientific practice. In this paper, I will address the question whether the "pudding" of Buriclan's theory of definitions he cooked up in his Summuloe proves "edible" in his scientific practice.'

In the next section, I will begin the discussion by presenting Buridan's the07 as it is laid down in the Sumulue. Then I will con- front the theorywith one of its particularly problematic applications in Buridan's scientific practice, namely, with Buridan's discussion of Aristotle's definition of the SOUL Finally, I willconclude the discussion with a somewhat tentative evaluation of the historical significance of thii application of Buridan's theory in his scientific practice.

2. Bwidon's Themy ofDefinitionr in the Summulae

In the eighth h-eatise of his Sum&, Buridan prowdes a systemat- ic account of his theory of definitions. The theory presents us with four kinds of definitions, namely, nominal, quidditative, causal, and

' Of co-, I mean 'sciennfic practicp" in the must <el-1 medieval sense, i.c., as referring 10 h e application of a cenain theoretical a-nn to a parlidar rdencc - in thb -.the adca t i on of the theorv of definitions to mrholoev. In , . L , u.

medieval science, &ere was no scientific p r i c e lo the modem sense, if by tbat we mean designing and conducbng experiments loing specialized insrrumcnw of various som, although even rhat should imply some thinking, when the hercienrist applie hk or her general theoretical knowledge to a particular situation. (Thtr ir. the process Buridan would refer to as cke "appropriation' of&e general rheoretid principles of a more general science 9- a more spcdfK field. Cf. for example. John Buridan. S~mmuLac. Dr &mmdmLM*bur [De Njk e.a.1. cap. 3, pars 6: cap. 8, pm 4.) Likwrise, later i~ncdie>-d thinkers could be described as 'empiridsT in their general approach to n a d philosophy, but cerrainly not in h e sense rhat later came lo be associated with the 6riri.h Empiricism. or rhe Logical Empiricisn. For discwion of !he latter point, see Zupko, ?+%at Is the Sdence of the Soul?"

descriptive definitions. In addition to these, Buridan also touches on "composite definitions" (defindiows complexae), that is, definitions

resulting from a combination of the aforementioned kinds. How- ever, since the particular problem I am going to consider here will primarily arise concerning quidditative and nominal definitions, I will concenhate on these two in this account of Buridan's theov. (Although later on I nil1 also refer to the other types, when they become r e l a a n t in the discussion of the problem.)

Buridan begins hi discussion of nominal definitions by the fol-

lowing characterization:

( I ) A nominal definition is an expression convertiblv explainingwhat thing or things the dqlniiurn signifies or connotes, and properly speak- ing it is called "interpretation." ( 2 ) It pemins to incomplex spoken t e r n to which there do not correspond simple concepe in the mind, but complex ones, w-hether these terms suppait for some thing or things or do not supposit for some thing or things.?

Thus, in accordance with this doctrine, a nominal definition pro- vides precisely the meaning of a simple term that is subordinated to a complex concept.' Indeed, a correct nominal definition of a term

is an expression that by its syntactical structure faithfully mirrom the semantic structure of the complex concept to which the term is subordinated. Therefore, the nominal definition and the term are

John Buridan, op.d.., z.3: '(I) Defiuitio dicens 'quid nominis' en oratio expri- mens convemb.iliter quid re1 quae definiaun significat aut connotat; et nomine p m prio rocahrr 'interpreiario'. (2) Et canrenit terminis incomplexis malibus quibus non correswndent in mente conceou simolices. red comolexi. sire illi termini om . allquo vcl altqu~l,u~siw nw pro ahquo ner pn,al>quih<u suppnant." Parsagesquor- cd horn thi5 work wr h m mvtmrladon, fonh<:oming in ThrY&Lhor)o/.Wrdvuoi I'hilaroph\wnes For demlls on the text m i k,r thr mnslation, xr the intrductorv essay of that mlume. Subsequent references to pasager from thk narkwillbe ro thi inninsic &ions of the teat

Cf.John Buridan, Smulac. D e w [Hubienl, cap. ,,pan 4:-Concerning the coruequencefmm the exponents towhar isqmunded, I sagthat that isa formal consequence, but what is infel~cd k not Merent from the premises in intention. but onlyverbalb.For m expound in this manner isjust to explain the meaning of the mmc (guidnmninir); but a dehnition tbatgivespredsely the meadngofa name and the name thus defined have to have enlirelv the same intention comswndinr IO .> thrtn in the rod. And the same g m f o r a propnuon that nqrurn some rrprvion un account of its syncatrpremau, for tlnr pn,psition and ie cxpouvna Iuve 81,

have cnnrrk $he same inrention comrp~nditbg lo rhrm in ihr soul.' Bundank here makes it entirely dear that as far hi theory is mncemed, synonymom expressions of spaken or witten Ian-- are not distinguished in mend language - a point which is still conuorerrial in the s e c o n d q litemure concerning Ockham'r theory. CT Panaccio, -Connotadre Tenm7 and Spade, "Synonymy and Equivoatiou.'

strictly synonymous: the term is to be regarded as a mere shorthand

for the complex phrase clearly explicating the structure of the cor-

responding concept? Consequently, monlg such simple tenns can have nominal definitions t o which there correspond complex concepts in the mind. As we can read in the Trcatie n Suppsitim:

. .. to some incomplex utterances there correspond complex conceprr, and to others incomplex concepu. Those to which there correspond complex concepe can, and should, be expounded as to their meaning (wid narinis) bv complex expressions that are equiralent to them in s&nification.~h'ose, iowever: to which there correspond incomplex concepts d o not have definitions precisely expressing their meaning (puidnornini~).~

In view of this doctrine, therefore, asimple term can have a nominal

definition if and only if it is subordinated to a complex concept

Now, since in accordance with Buridan's account of the ditference between absolute and connotative terms, all simple absolute t e r n are subordinated to simple concepts, this means that absolute terms

cannot have nominal definition^.^

' Kate bat when we are talking about the structure of a complex concept this need not be imagined to be the same kind of ~gomciic rrmcnve as that of the mr- responding expresioo: the complexjy of a complex concept is mere vrnmuic com- plexity a~ opposed to qntonic mrnplexig For this distinction, lee my introdunion to Buridan's Stmmh de d W a

j John Buridan. Suramulu. Dc ~ ~ f i m t i b u s [Vander Leql, ~ p . n, pan 4: "... v o w incomplexarum quibusdam mmspondent mnceptw mmplexi et qw%w dam concepm incomplexj. Et illae quibus corrrrpondent mnceptus mmplexi sunt et debent eawni a m - adamidnominis ~ ~ o r a t i o n e r i l k wuidentesin . . sipifimdo. lUae autem quibus correspondent conceptus inmmplui non habent cliffintiones praedse eqonenter quid nominir." Note that rmcdy speaking, despite existing mslational uaditions to the c o n q , quidu*lnanmisi not to be rendered ar "nominal definition," for the laner is +,Go r;rpmacnr quidnm'nis, i.e., =defmition expressing the quid Mnin$ therefore, since the phrase guid m i n i s alone is us* all" ennunded as mid rirmi&&mk cr i.i.e.. what is sienilied bv the name, arid

not h& nominal definitions, ie.. definitions preciselyexprening their quid MniniS but of c o r n they do haye quid mini.. i.e., thcy do have meancng or signification, given that they are meanin* in > + m e of being subordinated 10 a concept in the fust dace. Ne~enheless.for h e sake ofdmolicitu. in its rubseouent occurrences I d 1 , . rranrlare the wholrphmse dr/lnihompinolr@Mnrnisrtot m'dvfinttion rxpminp the mvmiuung ul tltr wane', Lug simpl) as .nom!nal drRniti<,n.' Funhermarc, rurcr evnb buradan IlimreH occ4onall) nars thc 5honrr ph- q u d m n u to refer lo a nominal definition, on such cari ions I %ill alro &"slate 9 u i d m . d ar -nominal definition.*

6Aconnotative term is a term matsigniSeswhatereritsignifie in relation tosome thingor ihing,itsmnnoarumormnnotae. Accordingb. such arerm, whenitoc-

To see this in more detail, suppose we have a simple absolute term A, which is not imposed to be a shorthand for a quidditative definition or some similar combination of other simple absolute terms.' This means that its signification and natural supposition are the same, taking these simply to be the set of its ultimate signifi~ata.~ Butsuppose Ahas auominaldefinition.Then this nominal definition has to be subordinated to a complex concept. However, since A was supposed to be absolute, the concepts making up this complex concept cannot be connotative, and cannot signify a connotation in respect of the significata of each other (as would be the case for instance in a genitive conshuction of two abaolute temu, such as hmiruS arinus); therefore, the total signification of the complex concept would have to be the union of the sets of the signiiicata of its component% But ifthis complex concept is made to be the subject of a m e n d proposition in which it has natural suppmition, then it can supposit only for the intersection of the sets of the signficata of in components. But then, since these components have to be diierenf they have to have different nonempty ses of significata. Thus, the intenection and the union of these sets will never be the same, which means that the total ultimate signification and the natural supposition of the same complex concept will never be the same. However, as was stated above, the total signification and the natural supposition of A have to he the same. Therefom, the simple term A cannot he subordinated to the same complex concept, and so A cannot have a nominal definition, contrary to our ass~mption.~

as the subject o r the predicate of a proposition, supporis forwhat itsignifies onbby means of an oblique reference ro what it connom, ie., using Buridan's terminolw, @ m- of appellating in appehnun or appeUaa. By conflart, absolute t e r n are thane that signi€y their significata not in relation to anything, aod, accordingly, in a propoeitional context they simply s u p p i t for their sbificam without appellating, i.e., obliquely refening w anything. Formore on this distinction and itr aipificance, see my introduction to Buridan's Sunnuloc L d M a ' I have to pro\& this restriction to tule out the a~cialcdseswhenforexample

sameone introducer. the rimple temBlrith the nominal definition=rational animal." In natural su~nmition a rimole abaolue term suomsin for all i n ulrimare , . . ,

signjfiicara (pa% pmenl hllum, 04 maytx nlerrly poasiblr), u, the nqnifial#on 2r.d natural ruppu.ioon of a s!rnplr ahoulucr arm u n ht, aadequatelv rhar~ctrnred in t m m of tlur set of in dgnihcam. tiotc abo that according ro lluridan in the m e of an absolute cerm this set is never empry, although the term m q actdfysupposit for nothing. (X.JohoBmidan. Soplunnnta [Scott), c. I. 6"conclu~ion. (Anewoanslation of the SophimwlnHill alro be included in my vanslation afthe S m m l o c d e d i d a N c 4 a the ninth menrise of that work)

Absolute terms, therefore, can have only other types of defini- tions, mart importantly, quidditative definitions. Buridan describes quidditative definitions in the following manner:

thing or rhings and in wbat *aps the name 'man' signifies, but is rather a &hition expressingvharthe thing is Tor which thename 'mao'supporis, this heingthe same as thar for which clr -&on ' r a t iod moral animal' suppoDi5. But ir is not nec- e-ru that those t e r n should oreciselv and adeouatelv suo-ir for lhe thines that ~~ ~, ~~ . . they $nqnih.Sn nnlya spoken term to which t h r x dcrr nor rorrepnnrl a n~uplc, but acomplrr concepr. boll* r*ich has a nominal drhnition in the 5mcl wnx, narneh l ad~ f in ,~ r i~~n~s l~~rhr i~n i f i r s preck lvrh tand bowlhat term ?aynntierFor the rigni- hcation of such a spoken term is explicated by means ofspokn &mw corresponding to the rimole conceotr of rhich the comvkx concevr commondine to rhat term is " composed. But when to some voken term there corresponds a simple concept, ar when m the term 'donkey' there cornponds the specific concept of donkqwhich n-e mume m be simple, it is n o t p m i l e far anotherspoken term tn r i d precisely and adequately ithe ~g or things that lhiu term signifies, unless i t is entirely sp onymous with iL Nor is it possible to positaspoken exprcsion consistingof terms of divene significations, without them cornpondin J to them otherconccpn which do not correspond to that term. But where a spoken term to which there corresponds a simple concept isconcerned it is possible to providea causd definition, ora d d p don declaring what the muses or propemes of the thing or things for whichthis term suomsits are. or ewnaauidd i lak definitionconsistineofthe eenusaod difference, . . to which them <omrpon& a complex contccp~ hut whch rdequately w p l m ~ n for thc ~ n b c thm@ *hbcharrruppos~rcd f o r b thc tr~cornplcxconre~ rhatcorresyondr to that spoken re-. Bur these w i n e had to be &lied in more derail in the treadse on &&tions, *ens and d&monsuations. And I am glad (hat I haw understood these issues." - *Re~poodetu quod illa non est diffinitio 'hominis' exprimens quid nominis, d c e t quid vel qquas res etquomodo boc nomen 'homo' sipificat, scd cst dithnirio exprimens quae res ert pro qua supportit iste termin- 'homo', guia lpsa est e a h pro qua ruppnit ism ontio 'animal rarionale mode ' . Sed Ron opr te r quad illi termini praecirc et adaequate supponant pro rebus quar signifi-L Code rolw terminusvocaliscui nonmrresmndet conceoaa simdexsed c o m d e m habet proprie ditlinitionem dicentem quid nominis,wilicet ptaecise significantem quid et quomado ille terminm sipificar Talis enim termini mdis significatio explicatu per reminos vocal- c o m p n d m t e s conceptibm simplicihm ex quibw cnmponi- ~r mncephx complrms correspondem ilii e m i n o . Sed cum slid termiuomxaii correspondeat concepnw simplex, ut huic Mi '&US' conceptus rpecilicus asini, m i t o ouod sit simolex non est oossibile alterurn terminum vocalem oraecke et . arbequrragniticare illuducl illaquae tile stplaificar. nirtit pm rvnonynw i l l6 Ncc en posviblle ihrr omticonern ~ o r d c m ex tcrnlmis diwrururn riq~ihrarlanum consri- & quin eis mmpondeaa t alii conceptus qui non correspondent illi termino. Sed d i s remini simplicis oi correspondet conceptus simplex potesr dari diffinitio a d i s vel descriptio ddedaraor quae runt c a m e vel pmprielates rei vel rerum pro aua >el nm ouibw ille terminus suomnit. vel etiam dillinitio o u i d i m h ex mnere . . . . ct dilTc.tcrrt~oa ct. ronesp,ndet ronrepnlr roml~lmu. supponeru a n a&?qtrate pro c&m rebus pro quihur rupponit cllc conreplus incotnplenu qui illi termlno *oral? <om.spndcl. Et haec dchcbant rnsgis drclarati in rrarlant dr clifintriobuhur,

BURIDAN'S THEORY OF DEFlNITlONS 35

(1) A ~uidditative definition is an e x ~ r a s i o n indigtine nrecisehr what -. thingis ( p i d uid em%) by- means of esKnlial predicam. (2) are

the genus of the ddmihmz and the essential difference. or diierences, {which are added j until the whole definition is convertible ~ i t h the dmniruin (s) This definition remonds ~reciselv, most ~ronerlv and . . , t& to thequestion "What is it:" iq) ~ n d it pre&poses the existence of the thing, ifone has to reply to the question M a t is it?"Io

As Buridan explains, this description entails that n o connotative terms can have quidditaiive definitions in this strict sense:

... let us asume that nothing is d i t e , except a stone. Then a white [thing] is a stone, and it is no thin^ other than a stone, nor is it a white- nes , o r an a ~ c g a t c o f a stone anda whitcnex~, if we e x p r m ourul\es propcrlv, hut it rs onlvlhat to whlrh a whitcncisbelon~s (mat), namely, . . . the stone, just as a wealthy man is not his wealth, nor the aggregate of a man and wealth, but only a man to whom this wealth b e l o n ~ (adjaunt Armlzar). R11t when I ask ;gcci<ely ?\Isat is thr white thing i&nA b h r n rsf);." I do not rrquire t h ~ t the reply should tndicate what a ~vhitcness . . is, or what an aggregate of a stone and a whiteness is, nor do I ask on account of what d i s~s i t ion a white thine is white; I only ask what the thing iswhich iswhiie, and that is nothingbut astone. Tkrefore,in the case asumed above. I eive asatisfactorvre~l~to the auestion if1 declare that the white thmiisa stone; and if ]'add &methi& else that s ip l i e s or connotes something other than the stone, then I provide more in my reply than what mas askedfor. Since, therefore, a purely quidditativc definition should nreciselv indicate what a thine is. if the term "white " . [thing]" (&urn) das a quidditative definition, then it is n e c e w v that it be the term 'stone." or its ouidditative definition. or an exoression consisting exclusively of substantial terms." But this is impossible, for

'@ Jolm Bundan, S m h . &drmonshahrmJnrr lk F2jk ea], n q : ' ( l ) De8nitio qutdirativ~ ~ J I "ratio indlrur prac:circ qud uden asc rci per praedicata crrential~a ( z ) (Luc wnr genw dcfituti et d~Rercnnn wl differenore nnennaladoncc wulir defroko sit m&ertibilis cum definito. ( 8 ) Et haec definitio rspondem practise, propriisiime et vere ad quaestionem 'quid st?'. (4) Et praeruppnit esse mi, si debeat responderi ad 'quid st".'

" The uiticd text h a here rhe following: Cum ngo dq%itiopurr quidimiuo dcbmt ind*mepaanWqu*l.&, neurrccrf si(lLrmmliw " o h " ~ d r j l n i ~ ~ m q u i d r t o t i ~ ( ~ m , quod ib, si sit iUjw &mini 'lapi%" d(4 riur piriirotiva Wnirio, iV loin] ycl quod sit om50 c& pmerin randnir Nbskmriolibur - 'Since, therefore, a purely ouiddirative definition should oreciselv indicate **at a lhinr is. i t is ne-arv that

e . 2 the term 'white [thing]' ( o h m ) h& a quiddiratiw definition, then that, kit is that of the term 'stone', should eirhcr be its quidditaliw definition, or that itshould be an expression consisting exclusively of subatantid tern." However, on the hvis of lhe appmm this can be mended as follom (providing the reading co-"ant with the Hubientext, which I w l a t e d in the main text): Cum ngo d+nirio pun quid&8 dm& hdicnrr pr- quid csf, wc&e Ln, n' iUc Unninw 'nhmr' hobrm d@nilMIm puidhtiumn, qrod illn sir is& t n i n m =lopir, '~el pied (4 e+m quiditoti~~

after removing u$iteness and retaining the substance subjected to it, the @nilurn namely,'white [thing]." would not suppositfor anything, for nothing would be awhite thing, and the definition would still s u p posit for something, namely, for the thing that it supposited for before. Thus it would not be cnnvertedwith the dmnifum, nor would it be hu- ly predicated of it, which is impossible; therefore, it b impassible that "white [thing]" (album) should have a purely and properly quidditative defi~lition."

So, if a simple term is absolute, then it has to be subordinated to a simple concept, a n d hence it cannot have a nominal definition. Thus, any simple term that has a nominal definition has to be subor- dinated to a complex connotative concept, and so the tern has to be connotative." Nevertheless, an absolute term can have a quidditab-e definition in the suict sense, namely, a definition which convertibly supposits for the same things as its @niturn does, but which always sigratfvs more, for it can contain only other absolute terns, namely, the strictly quidditatively predicable genera and differences of the definitum" However, since connotative terms cannot have quiddita- tive definitions i n this suict sense, it follows that a term can have a quidditative definition i n the suict sense if and only if i t is abso- lute.

&nil& VIOL"] davlino -John Buddan, W i o R n in Par tokwnd~l icmunr 6bm1 [Huhien], qunlsit om& r m u f i m o ~ a fninird$tmniol iba

I* John Buridan. Sumuloc. DL dnnrmcbntia26ur [De Rijk em.], 14. Cf. John Buri- dm, @IioRn in Partmimum A n a l y W m i i h [Hubien], Book 2, q. 8. (unpub lished edition of H. Huhien)

l3 Ob-<ously, th* much need not entail that all connowive temu have to have nominal definitions indeed, since Buddan allows the posgibiliry that some COMD

tative concepts are simple, the terms subordinated to them cannot hare nominal definitions in r h i s sect sense. (ff. John Buridan, S m m ~ l o r & -h&ibur [De Rijk e.z.1. z.4. seconddoubt)

I* John Buridan, qb.cir, ..+:=In respect of 1hi3 propeq the quidditak definition afso differs h m the nominal definition, for rhe quidditah definition should sig- nifi much m o m or somethiig other, than the dejnihmm And this decks h.om the other difference bewen them, namely, that to the dejkihrar ofa nominal definition there should cornspond a complex concept, whereas the species defined quiddita ti"+ has an incomplex co~~cept: for it is not the species, hut the definition ofthe species which is compaaed of genus and dderence, whether in rapect of otterance or conccpr" - lo hac etiam propriente diBen dehnitio quiditatim a dehitione dimnte 'quid nominis', quoniam illavel m~~lro pl- uel aliud debet SigIificare quam defininrm. Et hoc p m ~ n i t ex hac dia differentia qnia definiw definitione dicente 'quid nominis' oportct cornpondere concepturn mrnplexum, etspeciesquiditativc definita habct concepturn incomplexum; non enim species a t , wd definitio speciei quae compnitur ex genere ct differentia, sive secundurn Mcem rive secundum concept-.'

But this conclusion imposes a rather severe limitation upon the scope of quidditatively definable terms. And since the middle term of the most powerful scientific demonstrations has to be a quiddi- tative definition, this would also sewrely limit the scope of what is scientifically knowable. Therefore, Buridan immediately adds that i n a less strict sense even a connotative term can have a quidditative definition:

... a connotative term can have a quidditative definitionin a l e s proper sense, and one which is more broadlvstxalled. For it is bv means of the subjcra that their u atuibutrsarc definrd, a- issaid in bk. 1 of thr f i<rPnmAnn(r / ra and hk. :of the JldobhwuJ." lhrrefore. a definition

z ,

of a connotative term is called quidditative. heause it indicates %.hat it is (quid &) not only where the s u ~ ~ o s i t u m is concerned. but also the . . ronnotautm. For example. the definition of'pug' (simum) is -concave nose;" and by savinz 'nor' 1 -av what i t i ? and I likewise indicate what the term 'pug" supposits for (since it is a concave n w , and a pug is nothimabut a nose); however, bv addine "concave" I indicate what that

~ ~ . . " ~~~

term ak l l a tes , for that is the very same thing, and none other, that the term *concave" signifies, namely, concavity. Similarly, if I define 'pugnesr," then I say that it is the concavity of the nose; and when I say "concavity," then I indicate what it k that "pugness"suppsim for, since it is concaviy (because pugness is a concavity, and nothing eke); but when I add 'of the nose," I indicate what 'pugnesn appellates, for it is the nose, given that a conevity would not be a pngness, were it not in a nose.

Therefore, in this connection we have to note that in the ow of connotawe terms, the genus is not predicated of is species in pad in the sbictest sense, but broadly speakiny it i s predicated of it in mid Thus, when I say. "A white [thing] is [alcolored [thingl" I do ndr s;tv preciselywharawhite [thinz] is, hut I add what it is like. as I said earlier: ;hercfore, this is not quiddiinve prrdicaritm in tlle smctnt sense. ~ I I I ;

broadlvspeakiny it is adm~ttrd to bc quidditativr, became it irtclicateg what it isfor which 'white [thingj" supposiis, thisbeing the very same thing as that for which =colored [thincl' sup&ts, and at the same -~ . . timeiit indicates what 'white [thingl' appellates, for this is the same as what 'colored [thing]" appellate~.'~

l5 AriStode, PmnimAnolyfia, 1.3.73a8sqq.; % p 1 ~ , W.5. l o ~ o b e ~ ~ . l6 John Buridan, opciL, 2.4: -... termini connarativi pateft eae dehitio quidi-

t i n minus proprie et mag* cornmuniter dicta Nam per subiecta definiunw per se psiones eontm, sicut habetux prima Posteriorurn et septimo Yemphpiae. Defini- ti0 igiav termini mnnorativi dicinu q u i d i l t i ~ quia indiwt quid est non sohm de lupposito, sed erivn de connotato. Verbi mtia definitio 'rim? sit 'nitrus caus': oer h"c quod dm '~nasul ego dbro quid el h; iimilntcr lndico pro q ~ o illr tcrminlu'.rl- ntum'ruppnir.qwd h o ~ r s t nnrtnsirnw, rimurn mi", nti,mfi. rtuon ntalittd:wd c u m adda 'raws'. p r hor q u Indico quid e r clntod illc t r rnutu appllat, qutn hrx

So, clearly, what allows Buridan to say that even simple connota- tive terms can have quidditative definitions is that i t is possible to consuuct complex phrases *ich are convertibly predicable of such terms by means of quidditative or essential predication. That it is not only absolute terms that can be essentially predicated is dear hpm Buridan's general characterization of the diitinction between essential and denominative predication in his treatise OnBedicables:

everything that is predicated of something is either predicated es- sentia~i, so &at neither term adds some ema~eausconnotation to the simificatiou of the other, or it is predicated denominatively, so that - one term does add some exuinsic connotation to the signification of the other. This division k clearly exhaustiye, for it is given in terms of opposites."

Thus, whenever the subject o f a predication is aconnotative term, the predication is essentid if and only if the predicate does not connote

ert illud idem, el non aliud, quad isle t d u s ' m s ' signifiat sdlicetuvitatem. Et similiter si definio 'sirnitatem', ego dico quad rimitas ert cavita. nasi. et nun dico 'cavitas',ego indico quid est pro quo 'rimins' supni t , quiaata~itas (simimenim s t ~ d n s , et nihil aliud). Sed cum addo 'nasi', ego indico quid crt [Etol*] quod 'sirnitas' appellat, quiz est narm; mde adtas non esersirnitasai w n a t i n naso.

I h d e virndum drca hm auod in termini6 connoetivis mla non mediaNs .. ~ ~ ~ , m quid de wa s ~ r i e , pmpntuimc loquendo, wet commtullrer h~quendo pracd!ranu drcaIK~r,~*l noquid. L ~ r ~ d i n m i l h u m r ~ t c o l o r s ~ u m ' . ~ ~ n ~ ~ ~ ~ d U o p r a r r i ~ r q u i d album est~redetia& quale est, ricut prius di. ideo non est propriiirne p d a t i o quiditatin, sed comunius loquendo concedinu [VIOI"~ qucd sir quidimtiva, qtia indicat quid ert pm quo 'album' supponit, quia est illud idem pro quo 'coloraNm' aupponit, et indicai cum hoc quid est quod 'album' appellab quia est illud idem

'colorahun' appellar' Buridan. Sumnukw. Lk mdimbilibw [De Rijk], cap. 1, pan g: 'Omne

ergoh~od~raedicatur de aliquowi praedicam-ntialiter, r~limtiraquodneuter reminu5 rumr simihmtionem alteriu addat -earn connotationem, vel pcaed- ~~~ ~ ~~ , ~ ,v

icanu denonttnauvc, xilice, its qunl u n ~ ~ tcnnum addat mpcr sign6cxuru,ern nlreriu> nllqum connot;auonem: apparrr rnim quocl hacc dwicn sit rafiirn?, quia o c r isppmba.' Cf. I ~ h n RuriaL~n, Sunnuhe h pasdcommro (Rcsj,rap. q. (un t. ' ~ m : thar there p;edirations are essential is okous, far a relati- concept ir not onlya concept of something, butalsoaconcept of something [with mpect? la some thing; rhereforc a relati- term, in d m e of its proper signification and impmidon connotes something [with respect] to something, whence rhc addition 'Imth re -11 to something', ronsmedw<th a relative tern, amountr only lo the erpliarion of the connoration of that term, but it does not add some connotation exuimic to lhat tern, and so the predication is e55enrid."- "Qwd enim haec praedicatimer sinf esentialer apparec quia conceptus wlativw non rolum est alinjus conceptu, md alicujus ad aliquid concrps; idea temrinw &ti-, de propria dgnifiatione ct impositione, connoor ad aliquid et ideo istaadditio 'ad *quid'. super termtnum relatinun cadens, non est nisi e x w i o connorationh illius termini, et non addit connorationern alienam super illud lenninum. et idw n r praedicatio essentialis."

anything over and above the signification and connotation of the subject The reason for this should also be clear. A connotative term is only contingently true of a thiig it supposits for on account of the fact that it may become false of the thing in question simply because of a change in something else that it connotes. Thus, a connotatix~e tern is what in the modem parlance we would call a nowigid desig- nator of its supposita: it can cease to supposit for its supposita without their destruction, as a result of removing its connotata.'n Take a w q a wealthy man% wealth, and the man, while he will still go on existing, will cease to be supposited for by the term "wealthy man." Xeverthe- less, it is still possible to form an essential predication in which the subject tern is the connotative term "wealthy man." provided the predicate does not connote anything over and above the connota- tion of this term. Thus, the predication "A wealthy man is a man who possesses wealth" is an essential predication, iu which whenever the subject supposits for something ;he predicate also has to supposit for the same thing; and thus which is always true, assuming the natural supposition of its termsJg

-

All in all, we can summarize Bmidan's conclusions concerning what sons of terms can have H-hich of the two sons of definitions discussed so far in the following table:

Simplespoken SimpleSpoka SimpleSpokm Absolute Tenn, Crmnoialiue T?wn, Conmlolive T m , Sh$E Sjqb? ~Ont@1 cOl@kX CrmCqt

H a r m m i d No h'o Yes

Ha.s quidditative Yes Strictly no, but Strictly no, but definition improperly yes improperly yes

Ascan be seen from this table, only simple connotative terms subordi- nated to complex concepts can have both nominal and quidditative definitions, and they can only have quidditative definitions in the improper sense characterized above.

Is For the signb5cance of this point in the nominalist onrologial pmgram see K l i i , -Buridan's Logic and the Ontology of Modes- and KLima, *Ockham's Seman- tics."

For the example of %eatthy: seeJohn Buridan. S u d . Dc q p d i m h o Wandertecq], 5.n.Fornatural ruppmirion, W. 3.4;4.4.

Now, in view of the conclusions of the foregoing discussion, one may find rather strange the things Buridan has to say about the Aristotelian definitions of the soul. The most striking formulation in this regard comes horn Buridan's first redaction of hi questions on the De Alainra:

... the definition of the soul (dejnilio a n k ) is twofold: there is one definition of the soul expressing its quid m i n i s but there is another, expressing not only its quid nattnis, but also its quid rei. The definition of the soul expressing its quid nmntnis is that which Aristotle posits in Book 11, Chapter z of this treatise, and it is this: 'the soul is that by which we live, sense, are moved locally, and understand" (d. qlqal2). For this reason, this whole expression (molio) and the term 'soul" are equicllent in signification. But another definition, expressing the quid vi, is posited in Book Il, Chapter 1 of this treatise, and it is the afore- mentioned defmition [i.e., that 'soul is the first, substantial, act of a physical, organic body potentially having life" (d. 41zaz7-z8)1, and the present question is about this. Accordingly, this definition express es not onb the pid m i n i s , but also the quid vi, and not only the quid ni hut also the pmprcrqu*Lm

According to this passage, we have two distinct definitions of the term "soul." The first is a purely nominal definition:' which, in ac- cordance with the previous conclusions, establishes that this term is subordinated to a complex connotative concept, the structure of which is explicated by this definition. The second is also claimed to be a nominal definition, but also a quidditative definition, (and indeed, a causal definition as well). Again, if this is a nominal def- inition, then the term "soul" has to he subordinated to a complex

John Buridan, @astima i# A-De mimz [Patar], Book 2, q. 3, p 242.

11.714~: '. .. duplex est defimtio d e : quaedam eatdefinitio anitme exprimens quid nominis, quadam auvm ert non solurn exprimemquid nominis,vemm etiam exprim- quid x i . Definitio animae exprimens quid nominis en quam ponit Aris toleler in IIDcapirulo h a c t a ~ s 1' huius, et ert ha-: 'animaer quo t~rnus, sentimus et rccundum l o c m movemur ct intellieimm:' unde ista tog oratin et ism terminus ~ ~~~ ~ ~~ - anlnw aequiwlrnt in sipdicando. Alir autcm clrfini"~ mirnae expnmenr qrud rei porucur lo I"ap~tdo htuus If' rtcndrtini~io pracdicw. erdc illaesc pm.wsq lc rc t du. Unde sb dehnino nun ~olnm rxprilnar q u ~ d nonunis. scd c o r n exprinnlc quacl mi, et non solurn exprimit quid rei sed etiampmpter quid"

To besure, the definition 1 call here "the first" is the one that comes later in Ansrotlc's text in the ~econd chaoter of the second book. whereas thc other comes fint. ~n the fin, chnptcr. IIoue\er, d ~ e d~tinit ion ul .\riaurlc'rwcond cl~aplcrcomcv l i n t here in R!tridm'rtnr. nnrl it is ala, pnor iwohr s, i! can b uud tc, prow lllr

connotative concept explicated by this nominal definition. Indeed, since this nominal definition is obviously different fmm the fint, the term "soul" as defined by the second definition has to be s u b ordinated to a concept distinct from the one explicated by the first definition, so, apparently, this would establish that the term is equiv- ocal.= Furthermore, the second definition is also claimed to be a quidditative definition. However, being thc definition of a connw tative term; it can onty be a quidditaiive definition in the improper sense described in the previous section, namely, insofar as it supposiu for and connotes the same things as the term it defines does.

But this is still not everything Buridan claims concerning these two definitions. Both in his running commentary on Aristotle's text and in the Summtrlac he claims further that the first definition can be used to prove the second. As he writes in the Summuloe:

... one definition of "soul" assere that the soul is the principal intrinsic principle of living, xnsing and understanding, a~ b clear front bk. n of Onfb.%rl;2s hnt another is tlrat rhcsoul IS rhefinrsubstantial acr ofan organic plt~ical lndv, err. .And the second is demonstrated by mean5 ofdlr fint thus: evervprincipnl intrinsic ~rinci~leofliving. sensln~ and undenranding ir t h ~ . fimt subalanttal act of all orgamic ph~ical body p~vnti.~llv having life; ant1 errrv soul is ~uch a pnnc~plr: therefore, etc.?'

However, together with the previously quoted passage, this passage gives us serious reasons to believe that this demonstration cannot produce a quidditative definition of the soul, not even in the im-

~2 Noa that this equivocation is not the equivoetion Buridan alludes to with reference to Averroes (John Buridan. Qwur6m itiAlirtoldi~& anim [Pawl p. 43. n. ST), because that concern the alleeed eauivoetion between the conceot of rhe

u . ~~L~ ~~ ~~~

ioteiiective and that of the wgeetive and rensiti*~ souls. 4s Arinotle,Dranh Ui,~1zaz7-zR. ¶*John Buridan, S u m & . E. dmmmtn~iaibur [De Rijk e.al, 2.7: 'Lna ergo

dehitio 'animae' e ~ t quod anima est principium principalc inmimecum uiuendi, wntiendi et intelligendi, ut p t e t secundo De aoima; et alia est quod anima ert ac- nrr primus rubstantialis corporir phpici organid ctc Et demonstran~r secunda per primam sic: omne prindpium principle inuinwnrn vivendi dc. est actla primus subrantialis corporis physid organid habenGltam in potentia; et omnisanimaest huiusmodi principitun; ergo, dc- d John Buridan. W i a a inA&oUlirlk oninur [Pam], Book 2 , c. n, p. 5511. g61og:" ... illud quo & a m viui~ rentit. etc., est acrm mrpork physici orpnici vim habentis in potentia; red anirna ert qua anima- r u m +it, sentit, etc, sicut declaraturn est in isto capitulo: ergo condudim quod anirna ert artus corporis physici organici vim habentis in potenria. Et nomdum est quod bra demonrvauo procedit ex notioribus "obis. Irta enim definitio aoimae nnirnnestqun~rnw, sauimur, clc. esr notior quoad DO*, exqua damper operationes ct effect- quos experirnur in nobis, etc., qui effectus mt notiores suis camis."

proper sense characterized earlier. To see this more clearly, let us lay out this demonstration marking the concepts subordinated to iu terms by bracketing the terms in the following manner:

I. Every [principal intrinsic principle of hing, sensing and under- standing] is the {first substantial act of an organic physical body potentially having life)

n. Eve7 [soul] is a [principal intrinsic principle of living, sensing and understanding]

Therefore,

3. Every [soul] is [the first substantial act of an organic physical body potentially having life )

Now, since the expression "principal intrinsic principle of living,sens- ing and imdentanding" is the fimt definition of the term "soul," with which itis suictly synonymous, the terms of this argument enclosed in square brackets are subordinated to the same complex concept, the structure of which is explicated by this expression. But then the third term, enclosed in curl~brackets, being the other nominal, as well as quidditative, definition of the term "soul," has to be subordinated to a ditferent, complex connota.tive concept So, this demonstmtion demonstrates the predication of the correct quidditative definition of its d$niium only if the first premise is a case of quidditative or es- sential predication, at least in the sense in which connotative terms can be essentially predicated. As we can recall, such a predication is essential if and only if neither term adds any extraneous connotation over the other. But a closer lwk at these two terms dearly reveals that the second connotes the subject of the soul not mentioned in the first, whereas the fint connotes operations not mentioned in the second. Therefore, the predication clearly cannot be essential even in this wnse.

To be sure, one might still say that in the conclusion the term "soul" occurs as subordinated to the concept which is expressed by the second definition, and according to that concept the predication of this definition is quidditative. As Buridan remarks elsewhere:

"... if the nominal definition is truly predicated of the m u m then it is predicated of it quidditatively, and thus in the fint mode, for this is a predication of the same about the same in intention, although not in utterance, for the d+$nihrm signifies or connotesnolhing more nor less than does the definition ..."Pi

=John Buridan. S u m l o r . I)r dmtmshohrbur [De Rijk e.a], 6.3: "Tamen ad

However, this defense would invatidate the demonstration. For then the demonstration would dearly be acase of the fallacg ofequivac&~rm, since then the term "soul" would appear in the conclusion according to a different concept than that according to which it occurred in the minor premise. Therefore, if the demonstration is valid, then its conclusion cannot essentially predicate the quidditative definition of the term "soul," taking this term in the subject according to the concept expressed by its purely nominal definition. On the other hand, if the conclusion is interpreted as essentially predicating a quidditative definition of its dqiniium, then the demonstration is not d i d .

So, what went wmng here? Let us briefly survey the main points that have driven us into this predicament. These poinu can be sum- marized as follows.

1. The first definition of the soul is its nominal definition 2. The second definition is also its nominal and quidditaiive (and

causal) definition 3. The second definition is provable by means of the first, using the

first as the middle term 4. The conclusion of the proof has to be an essential predication

of the quidditative definition of its dq'inaum

The conclusion of the foregoing argument is that these points, ex- tracted from the pima l&w4 along with Buridan's theory of deiini- tions discussed in the previous section, are inconsistent. Therefore, whoever wishes to maintain the theory has to reject at least one of these poinrs. So, apparently, what went wrong was the combination of Butidan's theory of definitions in the Sumuloe and the strong claims of the prima &urn concerning the two definitions of the soul.

In fact, in the parallel passage of the tcrtia IrcIura, Buridan simply states the following:

Some definition is expressing the quid mminis, another is purely quid- ditative, and yet another is causal, explicating not only what the thing is (quad m 4, but also the reason why it is (PmpLn quid at). And [a definition ofl this kind is more perfect. And this definition [namely.

praeseru volo dicere quod dehitio dicem 'quid nominis', si \ere praedicatur de definite, paedicatur de eo quidiratiw,etic in prim0 modo, quiaestsimlpraedietio e i d m de eodem ~ecundurn intentionem, licet non secundum wocem, nihil plus aut minw sipilicet aut connoter defioitum quam ddnitio. pout alias dictum est.-

"the fint substantial act of an organic physical body potentially having life"] is of this kind?

In thii passage, the second definition is not claimed to be nominal, but quidditative and causal. The first definition, which in the pmM lecfura is daimed to be strictly nominal, is not even mentioned here.

Clearly, in this way it is easy to avoid the inconsistency derivable from the formulations of the prima ktwa But then the question nat- urally arises as to why Buridan provided that problematic formulation in the first place, when it so obviously gives rise to the inconsistency. assuming hi theory of definitions of the Summulac. One might offer several hypotheses here, sevelal of which may well be m e , but not particularly intriguing. For example, one may assume that Buridan was simply inadvertent, or that he did not reaq mean the fomula- tions of the pima Inhna as strictly as they sound, etc. However, in conclusion I will risk a different hypothesis. Not because I think I can pmre or disprove it here (so, it may well be false), but because it places the question of the viability of Buridan's theory in his scientif- ic practice in an interesting historical perspective, desening hrther exploration.

The hypothesis I would risk here is that in the problematic fomu- lations of the pinm lectura Buridan simply followed an older line of interpretation, which he later abandoned upon realizing its conflict with his 'official t h e o g of definitions. According to what I call this "older line" of interpretation, the two Aristotelian definitions of the soul are related to each other as a nominal and a quidditative defini- tion of the same definirum, the latter of which can be demonstrated in terms of the former, by means of a puiuqe, a p o s h - demonstra- tion. In any case, thii is how Aquinas interprets Aristotle's treatment of the two definitions of the soul provided in the De Anima

In his commentary on the De A n i w Aquinas states that when the Philosopher proves the quidditative definition of the sonI, he provides us with a proof from the better known effects to the lesser-

26 John Buridan, Quastiorm in Arimldir De aninn. tmlio bmno [Soboll, p. 35: 7rcm qucdam estdifihitio dicens quid nominis, alia pure quidditatim ali ou~alis, explicans non~olumquidestressedetiam pmpterquid est. Erralisescrnagisperfecta Et hu imod i a r km definitio."

BURIDAN'S THEORY OF DEFINITIONS 45

known cause. As such, this is a qwi~iqpe dem~nstration.~ However, we also know that according to Aquinas the middle term in such a demonstration has to be the nominal definition of the cause.=

Furthermore, Aquinas makes it dear that the definition thus demonstrated is a pmperly quidditative definition. Nevertheless, be- cause of the incomplete nature of the thimg defined, this definition has to contain an oblique reference to something extrinsic to the nature of the thing.= At the same time, however, he also insists that the correct quidditative definition should sign* the same essence as the term defined.' Therefore, the well-established quidditative defi- nition of the thing will always be essentially predicable of its d.z$niltlm Indeed, in away it can also function as a nominal definition, by spec- F n g what it is that anyone understanding both the term and the nature of the thing has to have in mind when using the term with understanding.)'

'' Aquinas, In &I fi rmim II ti Book I. Lc 2: '... deinde cum dicit diumur igihn indpit demonsvare definitionern animae superius p i t a m , modo praedicto. scilicetper effectum. Etutiturali demonsaatione. llludquodertprimumprimipim viwndi ert "?entiurn co'po~um a- et forma; sed anima est prim- principium -r:vendi his quae viwnc e r p ~ o t corporis vireotir actln et forma Manifaturn s t autern, quod haec dernonsaatio est ex posteriori. Ex eo enim quod anima est forma corpolir; viventis. n t prinapium operum vitae, et non e converso.-

ea Cf. Tho- Aquinas, Summa The&@ I , q. 2, a. P, ad zm. a. Tho- AqJAqr. Demtrrtamt&% c. 7: Tt quia, uldicaM ar, w n l i a e n

id quod per W i t i o n e m signilicatur, opomt ut eo rnodo habeant mentiam quo habeot diffinitionern. Diffinitionem autem babent incomplegm, quia non possunt diffiniri nisi ponatur subi~ctum in e o m diffiitione. Et hoc ideo at, quia non habent per se esre, absolurum a subiecto, sed sicut ex fa- ct materia relinquim esse substantiale, quando componunm, ita ex acidente etsubiecto relinquitur esse accidenrale, quando acidens subieem advenit Er idea etiam nec forma subrma dalh wmplclam esentiam habet nec materia, quia etiam in df i i t ione f o m e subsranrialis oponet quad ponatur iUud, cuim est forma; ct im diflinitio eiur est per additionern alicuiw. quad a t m polus eiw.sicut el difhnitio fo-e accidenglir. L nde et in diF.nrt!onr m w p,nltur nsipu, a namrala, qui roosiclcra~ annmam solurn in quantum a t fun!u phpio corpori5. Fed tarnen tntrr forms rubstanmalcs er accidentales canturn interest, quia s i n ~ r f o m substantialis "on habet per re esse abrolutum sine eo mi adwnit, ita nec illud cui adrzni< xilicet matcria. Et ideo ex coniunctione u n i q u e relinquitur iUud m e , in quo r a per se sulairtit, et ea cis efficitur unum per re: pmpter quod ex mniunctiane c o r n relinguitur arentia q u a e h . Unde forma, quarnvis in pe consideram non habeat completam rationern essentiae, mmen err pars aenriae cornplctae.' * See again in the quote above: hsrentia a t id quod per diliinitioncm significa-

nmr-

3' Cf e.g. lhomas Aquioas, Erpo* fihi Bodii & Hd&mdihs, lc. I: T t ~ o m Aquim, InLibmsPoNrionrmAn&tm% Bwkr. Ic. 6.

But this intetpretation will not cause the eouble for Aquinas we noticed in Buridan's prima lcclwa, because of the differences between their respective "background theories." In the k t place, since for Aquinas a nominal definition is not the anabis of an underlying complex concept, providing the nominal definition will not commit him to assigning a complex concept to the term defined On the con- lmry, in hihis conception we can obtain nominal definitions of terms subordinated to simple concepa by any sorts of indications somehow specifying what is meant by the corresponding term.)4 Therefore, in the argument laid out above, Aquinas is not committed to holding that the term "soul" and its nominal definition should be subordi- nated to the same concept For him, the term is always subordinated to the simple essential concept by means of which we conceive of the simple nature of the soul. The same nature is indicated with refer- ence to its d e c a by the nominal definition. The nominal definition is subordinated to a complex concept by which we have a rather inadequate grasp of the essence of the soul, not knowing what it is, but only what it causes. Fially, the same nature win be adequately grasped again bp means of the complex concept expressed by the quidditative definition, which allows ia essential classiication in the system of the categories.

As we could see, Buridan seems to follow quite faithfully the pattern of Aquinas's interpretation in his prima &ma He describes the first Aristotelian definition of the soul as a purely nominal def- inition, defining the soul in tern of its operations that are better known to us. He describes the second definition as being a nomi- nal and quidditative definition, prowble by means of the first in a gain-gpe demonstration. Finally, he states that this definition is also causal, insofar as it has to indicate the material cause, namely, the subject of the soul.

However, be soon must have realized that in view of his theory of definitions, he could not maintain the formulations of the pama &c- hmr Instead, he seems to have come to the conclusion that the best way to accommodate the two Mstotelian definitions in his 'official theory" is u, treat the first as a mere description,'s and the second

Far thebroader significanceof hiispointin latemedid metaphyicrweglima. 'Buridan's Logic and the Oolology of Mades.'

"John Buridan, Sumuloe. Drriclslonrhnlioni6~ [De E j k ea.]. 2.6: "(I ) A d-mp don is tuually defined thu: a dewripdm is an expression indicating what the thing is (quidaimem), the quiddig or mence of h e thing] in terms of inaccidenh oref- k t s tharare posteriorm it abrolutcbrpeaking (pormiom+ur). (2) Therefore, in a derniption rhe subject h defined in re- of i s artribute or atnibuces and rhe

q6 G W U KLIMA

as a causal definition." This solution does not place the suict log& cal requirement of synonymy between definition and dq?niLffm upon the fint definition, or the requirement of quiddiitive predicability upon the second definition. Therefore, with this adjusnnent no in- consistency of the sort derived h m the lectwa can arise, even assuming Buridan's "official theov.'."

Now if my hypothesis is correct, then its broader significance can be summarized as follows. As I haw argued elsewlvhrreP the new theory of signification and connotation utilized in the nomi- nalist ontological program was intimately tied up with a subtle rein- terpretation of the theory of definitions. From the point of view of this central programmatic concern, it was necessav to interpret nominal definitions as providing conceptual analyses, and qniddi- tative definitions as providing the characterizations of their @ni& in terms of the latter's quidditative predicates. But then, this new conception inevitably had to lead to clashes with the usual classi- fications of well-established Aristotelian definitions originally inter- preted within the semantic paradigm of the via anliqua Assuming the correctness of my assumption about Bnridan's motiwtions for revising the formulations of the pfimu led~wu, Buridan's handling of the Aristotelian definitions of the soul would illusmate precisely this phenomenon.

To be sure, the foregoing considerations assumed the authentic- ity of the prkma lechrrn. Thus, the inconsistency presented above may also be interpreted as providing evidence against the correctness of

cause in terms of its effect or effects. is) H-er.a dercrintion incamarata item that arc pnor and hener known lo u is@mlnud whmlk +ndr), but uhch arr not u, s h l e r ~ l v s p e a b n ~ (snnphur) (4) Therefore. In means of rlrscnpuc,nr one Mrnetimes pro"- quiddierive definitions or causal definitions to apply lo their definita, not pmpw @d hut only puin' Indeed, this squares very weU kth how the h t redaction of the W h g (Pam, 8ook z, r 2, pp. 47-48.11.95-99) meats the h e of the demonsnation of the quiddimtive/caddeRniti~n ofthe soul.

"John Buridan, q.&, e.5: =(I) -9 causal definition is an exp-ion that convertibly indicates what the thing h [yud ol arr ni the quiddit). or wence of the thing1 and the reason why ( p m p ~ c r p i d ) it iri. (2) Such a definition is provided by means of terms which in the nominative case xwuld s u p p i t for the cause or c a m of that thinear those thin= forwhich the term defined tdrtininml rruoomitr. - , , . ~ ., (3) These l r m r arc p!aced III those definitton* in an obl~qu~rrasc.''lhe last section and Hwidm'r rulneqlterl! comment $1" 11 clrarl! erubltrh dual ih,r npe 01 definmz,n cannor be quidditatively predicaccd of iv- &finihrmeven in the less proper sense, for it has to appelbte the calm of ~e thing defined.

' 5 In the papers referred to in n. 18, and in Klima. "Ontological Alrernadves."

BURIDAN'S THEORY OF DEFINITlONS 47

that assumption." But the issue of authenticity should be the subject of a different study.

"The single manuraipt on which the edition of Patar r g g ~ is based. B m p 477, b anonymous, and scholar. haw variously aruibued it to Blasiuc of P a m (Cf. Federici-Vescovini, La "Qnaufirma & ARinm") Buridan's student Dominic of Clavasio (Pactin, Pmrr I'himire dtl m u agent), an anonymous compiler of Buridan's reachingr ( . b h a l l . ~ ~ i a n Pn/chologv7), aswell= la Buridan himself (Michalslti, =La physique nouuelle;" John Buridam, QuoaMMinAI*lotrlir&onim [Patar]). .? extens-kc r u m m q of the controversy can be found in John Buridan, n

Arismlis Dc onimo Ihrar], 67*38*.