Curley, E - Spinoza´s Metaphysics. An Essay in Interpretation-Harvard University Press (1969)

8/10/2019 Curley, E - Spinoza´s Metaphysics. An Essay in Interpretation-Harvard University Press (1969)

http://slidepdf.com/reader/full/curley-e-spinozas-metaphysics-an-essay-in-interpretation-harvard-university 1/187

A philosopher's real power over mankind resides not in hismetaph ysical formu las, bu t in the spir it and tendencies w hichhav e led him to ado pt those formulas. Spinoza 's cr i tic has, the re-fore, rather to bring to light that spirit and those tendenciesof his author than to exhibi t his metaphysical formulas. Proposi-tions about substance pass by mankind at large like the idlewin d, which m an kin d a t large regards no t; i t will no t evenlisten to a word about those propositions, unless it first learns

what their author was driving at with them, and f inds thatthis ob ject of his is one w ith wh ich it sympathises, one, at anyrate, which commands i ts at tent ion.

-—Matthew Arnold , Essays in Criticism Spinoza and the Bible

Preface

Brought to you by | Brown University Rockefeller LibraryAuthenticated | 128 148 252 35

Download Date | 6/18/14 1:50 AM



Spinoza, in the three hundred years since he l ived and wrote, has beenpraised and damned for a most amazing variety of intellectual virtuesand vices. Damnation came first . The verdict of his Christian contem-porar ies was that in ident i fying God with Nature he was gui l ty ofatheism, and consequently his major work, the Ethics could not pru-dently be published d urin g his l ifetim e. O n the other hand , his co-religionists of the Jewish community in Amsterdam cast him out, notfor disbelief in God, but for the heterodox opinion that God is corporeal.

N or was it only the or thodox w ho cr ied "Heret ic " A s Hein r ich Hein eobserves,

The most opposite parties arrayed themselves against Spinoza. Theaspect of this army of adversaries is highly amusing. Near a swarmof black and white Capuchins bearing cross and censer, marchesthe phalanx of the Encyclopedists, who also take aim at this "daringth inker" by the side of the Rabbi of the synagogue of Amsterdam,who sounds the at tack with the sacred buck 's-horn, advances Arouet

de Voltaire, playing obligato on the shril l pipe of irony for thebenefit of deism.1

No doubt there was an ulterior motive in the Voltaire-Bayle dispar-agem ent of Spin oza as an atheist . Both clearly relished the notion th atthe most heretical of doctrines may be espoused by a man of unimpeach-able character. For their purpose, they needed a Spinoza who was ablank, frank atheist .

In a somewhat s imilar way, Spinoza 's "universal infamy" as an atheis tserved Hume's purpose when he at tempted to show an ident i ty betweenthe popular doctrine that the soul is an immaterial substance and the"hid eou s hypothesis" that there is only one substance. T o show thata thesis reduced to Spinozism was, in the eighteenth century, a veryserious criticism. But this only illustrates the esteem in which Spinozawas generally held.

vii





reface

Contrast this nearly unanimous condemnation of Spinoza prior to,say, 17 80 , with the treatment accorded him since. W e find Heine claim-ing that nothing but sheer unreason and malice could bestow on sucha doctrine the qualification of atheism— no one has ever spoken m oresublimely of Deity than Sp ino za, and Ernest Ren an crediting him withthe truest vision ever had of Go d . W e find the liberal Protestant

theologian Friedrich Schleiermacher writing of the holy and excom-municated Spinoza and the Catho lic poet Novalis coining the famous

phrase the God-intoxicated m an in his honor.I f, prior to 17 80 , the most opposite parties arrayed themselves against

Spinoza, since then the situation has reversed itself, and people of themost diverse interests and temperaments have seen in him a sympathizer.He is at once a pantheistic mystic, whose closest spiritual cousin is theBuddha, an d the last of the medieval Jew ish philosophers, whosephilosophic kinship is to Moses M aimon ides, Chasda i Crescas, and Aris-totle. While the Marxists have claimed him as a precursor of dialectical

materialism, H. H. Joachim and others have viewed him as expoundinga form of absolute idealism. Friedrich Nietzsche, when he discoveredSpinoza, immediately wrote an excited postcard to Franz Overbeck,proclaiming that at last he had found a predecessor Not only is hisover-all tendency like mine—making knowledge the most powerfulaffec t— but in five main points of his doctrine I recognize myself . . . hedenies the freedom of the will, teleology, the m oral world order, theunegoistic and ev il.2 C . S. Peirce regarded Sp inoza as one of the threeforerunners of pragmatism, though John Wild says confidently that thepragmatic theory of truth would have appeared absurd to Spinoza. Andmost recently he has been interpreted by Stuart Hampshire as a philoso-pher of science on the grand scale, giving metaphysical expression, withina fundam entally Cartesian fram ework, to the ideal of a unified science.

In the face of such diversity of opinion, one hesitates to add to theconfusion. Nevertheless, I feel that I have something to contribute tothe understanding of Spinoza, and for better or for worse I will havemy say. My method, contrary to Matthew Arnold's perfectly sensible

v





Preface

advice , will be to begin by trying to show ho w S pinoza s m etaphy sicalformulas are to be construed. I do not think we can profitably discusshis spirit and tend encies u ntil we have some idea of w ha t th e form ulasmean. No doubt we must approach them tentat ively, keeping in mindthe possibility of other interpretations, and no doubt our choice of inter-pretations to try out must be guided by some hypotheses about his spiritan d tendencies. But in the end we mu st come back to the text , an dit seems to me just as well to begin there. Those impatient souls who

mu st know w ha t my conclusions are be fore they learn th e reasons forthem may do what they would do if this were a detective story t u rnto the end of the book.

M y m ain concern is to provide the most coherent an d precise exp lana-t ion I ca n of the fun da m en tal proposi tions of Spinoza s metaphysics.To do this I will make use of a philosophical vocabulary which wasentirely un fam il iar to Spinoza. I do not contend th at Spinoza th ou gh tin these terms. I do contend that his philosophy can be expressed in

these terms with remarkable success. The measure of success here iswh ether or not w e can, in the langu age suggested, m ake connect ionsbetwee n the various things Spinoza says an d show him to hav e a co n-sistent, reasoned view of the wor ld. If this can be do ne, I shall considerthe use of these admittedly alien categories to be justified.

I have said that my primary aim has been understanding. I shouldperhaps add that , i f my understanding has been correct , many of thecriticisms usually directed against Spinoza are seriously misguided. Tothat extent, my interpretation of Spinoza will also be a defense ofSpinoza.

Tran slat ions are my ow n, unless otherwise indicated. T hose ofSpinoza s works are done fr om the cr it ical edit ion of Carl Ge bh ard t .I h ave quo ted freely fro m works other than th e Ethics an d this is per-haps the place to comment on the relative authority of these works.Briefly, m y position is th at no work o ther th an the Ethics and perhapsthe later letters in the Correspondence is fully satisfa ctory as a sourcefor the views of the mature Spinoza. The Metaphysical Thoughts and

x





Preface

Descartes Principles of Philosophy are bo th, in pa rt at any rate, exposi-tions of the views of Descartes, though I believe they contain importantindication s of Sp inoza s own views. Suspicion has bee n plausibly caston the Theological-Political Trea tise on the ground that there Spinozais acco m m oda ting himself to the opinions of the vulgar in order to m ak ehis message more acceptable. The Political Treatise is an unfinish edwork, as are the Short Treatise on God, Man and his Well-Being)and the Treatise on the Correc tion of the Intellect ). Both of these

latter works, m oreo ver, are early an d show significant differen ces onsome points from the final teaching of the Ethics. But after all reserva-tions have been made, these other works stil l remain valuable sourcesfor the study of Spinoza. I think there is no good reason not to appealto the m so long as th e passage cited is no t inconsistent with th e Ethics.W he n the passage has seemed to me inconsistent with the tea ching ofthe Ethics I have generally indic ated this.

There are a number of people and inst i tut ions to whom I should

here express my appreciat ion for the contr ibut ion they have made tomy writing of this book first , Professor Bernard Peach, who supervisedmy w ork when i t was a doctoral dissertat ion at D uke U niversi ty, an dwho gave of his t ime, energy, and wisdom to an extent that was surelybeyond the requirements of duty; Professor Romane Clark, whose com-ments on the various visions and revisions of the manuscript were alwaysincisive ; the G rad ua te S chool of Arts and Sciences at Duk e, for thegenerous fel lowship wh ich m ad e m y study ther e possible; my fel lowstudents in Professor Peach s Spinoza sem inar, wh o provided m uc h ofthe s t imulus for the ideas expressed h ere ; the Inst i tute of A dva ncedStudies of the Austral ian Na tiona l Universi ty, for giving m e the o ppo r-tuni ty to rethink an d rew ri te my earl ier work un de r condit ions as nearlyideal as one ma y ho pe for in an impe rfect world ; Professor Jo hn Pass-more, whose kind advice and cri t ic ism have been most helpful ; Mrs.Amel ie Ror ty, w ho read th is book in manuscr ip t for the Har va rd U ni-versity Press an d m ad e some very useful suggestions; a nd finally, my





reface

wife R uth witho ut whose moral support this book wou ld certainly neverha ve bee n finished as soon as it wa s and m ight never ha ve been finishedat all.

Canberra Austral ia E .M .C .February 1969

x i





ontents





Th e Definitions of Substance and M ode ιThe Bayle-Joachim InterpretationThe Wolfson InterpretationNotes Toward a New Interpretation of Spinoza s Definitions

2 The Causality of God 44First Thou ghts and Difficulties

The Model MetaphysicApplication of the Model Metaphysic to SpinozaThe Interpretation of the Definitionsof Substance and ModeObjections and Replies

3 Necessity 82Eve ry Truth Is a Necessary Truth

In What Sense Every Truth Is aNecessary TruthWhy Every Truth Is a Necessary TruthIs the Ac tua l World the Only PossibleWorld?Objections and Replies Again

4 Th e Divine Attributes 118

Thought and ExtensionOf the Unknown Attributes of God

Postscript 154

Notes 161

Index 173

xiii





Works ited in Shortened Form

SPINOZA S WO RK S

E Ethics, ed. Carl Gebhardt. Heidelberg: Carl Winter, 1925 (E IA2= part I , axiom 2; E IIP33C = part I I , proposition 33, corollary;E I V P 1 4 S = part I V , proposition 14, scholium; E ID3 = part I ,

definition 3; E IPI1D3 = part I, proposition 11, demonstration 3) .THE Tractatus de Intellectus Emendatione TdlE, 76, 11:29 =

Bruder edition, section 76, Gebhardt edition, volume II , page 29).KV Korte V erhandeling van God, de Mensch, en des zelfs Welstand

(KV I, il, 1:19 = part I, chapter ii, Gebhardt edition volume I,page 19).

PP Renati Des Cartes Principia PhilosophiaeCM Cogtata Metaphysica (CM I, iii, 1:243 = P ar t I chapter iii,

Gebhardt edition volume I, page 243 ).Ep. Epistolae (Ep. 21, IV : 127 = Epistle 21, Gebhardt edition volume IV,

page 127).TTP Tractatus Theologco-Politicus.TP Tractatus Politicus.NS Nagelate Schriften. Gebhardt s edition gives variant readings from

the Nagelate Schriften, the Dutch translation of Spinoza s works,which appeared in 1677 at the same time as the Opera Posthuma.The Nagelate Schriften were translated from Spinoza s ownmanuscripts, but apparently from manuscripts which represent,in some cases, an earlier draft. They therefore provide not onlya valuable check on the proofreading of the editors of the OperaPosthuma, but also an indication of Spinoza s final revisions.

OTHER WORKS

Adam and René Descartes. Oeuvres, ed. Charles Adam and PaulTannery Tannery. 12 vols. Paris: J. Vrin, 1897-1913.

Arnauld and Antoine Amauld and Pierre Nicole. La Logque ouNicole L Art de Penser, ed. Pierre Clair and François Girbal.

Paris: Presses Universitaires de France, 1965.

XIV





Works Cited in Shortened Form

Bayle

C a i r d

d e D e u g d

F r i e d m a n

G e r h a r d t

H a m p s h i r e

J o a c h i m , Study

Joach im,

Spinozas TractatusL o v e j o y

Po l lock

Russel l

Ve r s f e l d

Wi e n e r

Wol f

Wo l f s o n

P ie r re Bay le . Historical and Critical Dictionary Selections,t rans , and ed . Richard H. Popkin . Indianapol is :Bobbs-Merr i l l , 1965.

J o h n C a i r d . Spinoza. L o n d o n : W m . B l a c k w o o d a ndSons, 1907.

G . d e D e u g d . The Significance of Spinoza s First Kind

of Knowledge. Assen : V a n Go rcum , 1966 .

G e o rg e F r i e d m a n . Leibniz et Spinoza. Pa r i s : Ga l l imard ,

1962.

G . W. Le ibn iz . Die philosophischen Schriften, ed.G. J . Gerhardt . 7 vols . Hi ldesheim G e o rg O l m ,1960-1962 .

S tua r t Hampsh i re . Spinoza. Lo nd on : Fabe r and Fabe r,

1 9 56 ·

H . H . J o a c h i m . A Study of the Ethics of Spinoza.Oxfo rd : C la rendon P res s , 1901 .

H . H . J o a c h i m . Spinoza s Tractatus de IntellectusEmendatione. O xf o r d : C la re nd on P res s, 1940 .

A r t h u r O . L o v e j o y. The Great Chain of Being: A Study

of the History of an Idea. C a m b r i d g e , M a s s . : H a r v a r dUniversi ty Press, 1936.

F rede r i ck Po l lock . Spinoza: His Life and Philosophy.Lo nd on : C . K . Paul , 1880; second edit ion , 1899.References are to the f irst editon, unless otherwiseindica ted .

Ber t rand Russe l l . A History of Western Philosophy.Lo nd on : A l l en and Un win , 1946 .

M a r t h i n u s Ve r s f e l d . An Essay on the Metaphysics of

Descartes. London : Methuen , 1940 .Leibniz, Selections, ed . Ph i l ip P. Wien e r. N ew Y or k :

Scr ibner, 1951.

Spinoza s Short Treatise on God, Man an d his Well-being, t r ans , and ed . A . Wo l f . N ew Yo rk : R usse l lRussell , 1963.

H . A . Wol f son . The Philosophy of Spinoza: Unfoldingthe Latent Processes of His Reasoning. 2 vo l s. Cam br idg e ,Mass . : Harvard Univers i ty Press , 1934.

X V









Sp inoza s M etaphysics





I t is not t ru e tha t his followers hav e been very num erous .Very few persons are suspected of adhering to his theory andam on g those wh o are suspected of i t, there ar e few wh o hav estudied i t ; and among the lat ter group, there are few who haveunderstood it and have not been discouraged by the perplexitiesan d the impe netrable abstract ions tha t are fou nd in i t . . .of all the h ypotheses of atheism, Spino za's is the least c ap ab le

of m isleading anybo dy, fo r . . . i t opposes th e most distinct

not ions in the hu m an mind .

-—Pierre Bayle, Dictionary Spinoza

The Definit ions of

Substance and Mode





W e are about to plunge into waters tha t are very, very mu ddy. M an ypeople talk with a great deal of confidence about Spinoza's philosophy.But am ong those w ho ha ve studied it carefully, there is no general agree-ment on the meaning of even those doctrines that are the most central.What Bayle wrote, at the end of the seventeenth century, might bewritten with equal truth today. No philosopher should be easier to under-stand th an Spinoza, for none has taken so mu ch trouble to explainhimself, defining his key terms , listing his basic assumptions, and arg uing

carefully for each conclusion h e draw s fro m those assumptions. B ut fewhave proved harder.

Consider th e antithesis of substance an d m ode. T h a t this is centralto Spinoza's me taphysics is evident from the first axiom of the Ethics

which divides reality into substances an d m odes. Eve rything which existseither exists in itself or exists in som ething else (E I I ). What existsin itself, and is thu s conceived th roug h itself, is a substance (E ID3 ) .What exists in something else, through which it is conceived, is a modeor affection of substance (E ID 5). Th ere is noth ing tha t does not fallinto one of these two categories.

But w ha t Spinoza intends by this antithesis is not at all clear. T ru e,he gives us, as his geom etric meth od requires, definitions of these crucialterms. Th e definitions, however, are not particularly helpful What,precisely, is Spinoza saying when he says tha t to be a substance a thin gmust exist in itself and be conceived through itself? What are we toldwhen w e are told tha t modes exist in something else through which theyare conceived? Couc hed in a highly abstrac t, technical language, thesedefinitions are open to a wide variety of interpretations, as three h un dre dyears of commentary on Spinoza amply demonstrate.

In this chapter, I shall contend that two of the most widespreadand plausible interpretations of Spinoza's definitions are wrong. Andbecause the concepts of substance and mode are so central to his philoso-phy, this will involve arguing, as well, th at m an y of the things ofte nsaid about Spinoza—both by his critics and by people who thoughtthey agreed with him—are mistaken. In conclusion, I shall try to indi-





Spinoza s Metaphysics

cate, in a rough way, how I think we should understand the definit ions.Perhaps some of the abstractions are penetrable; perhaps some of theperplexities can be removed.

The Bayle-Joachim InterpretationThe first of the two interpretations with which I shall deal occurs asearly as Bayle's Dictionary and as late as Bertra nd Ru ssell's History

of Western Philosophy. Th is interpretation supposes that Spin oza under-s tood the terms substa nce and m od e in m uch the same w ay thephilosophers of the seventeenth century in general, and Descartes inparticu lar, u nderstood them . T o see i ts merits an d demerits, w e needto begin by examining the ordinary philosophical use of these termsat that time.

Undoubtedly the seventeenth-century account of substance best knownto people in the English-speaking world is Locke's, in the Essay Concern-

ing Hum an Understanding. It is so fam iliar that I shall m erely sum-marize its principal features in these five propositions :

( ι ) A substan ce, in the gen eral sense, is no thin g bu t the subje ct,or substratum, in which qualitites are said to inhere, or exist, or subsist,or rest.

( 2 A substance of a particular kind is a collection or combinationof qualit ies, together with the subject or substratum in which they allinhere.

( 3 The subject or substratum in which qualit ies inhere does notitself inhere in anything else, and hence may be said to exist or subsistby itself.

(4 ) T h e subject or substratum supports and unites into one thingthe qualit ies which inhere in i t and which cannot be conceived as sub-sisting by themselves.

( 5 The subject or substratum is not perceptible to the senses ; w h a t -ever we may be said to perceive through our senses is a quality. Theexistence of the subject, however, must be inferred from the existence





The D efinitions of Substance and Mode

of the qualities we do perceive, since the qualities cannot be conceivedto subsist by themselves.1

Though this notion of substance as the imperceptible subject in whichqualities inhere, w ith all the difficulties it involves for empiricism, isbest know n to us throu gh Locke, it is by no mea ns peculiar to him .It w as the comm on prop erty of the philosophers of the seventeenth cen-tury. Loo king at the most famo us logic text of tha t era , the Port-RoyalLogic we find a very similar account :

W hatev er we conceive is represented to our m ind either as a thing,or as a man ner of a thing, or as a mod ified thing.

W h a t is conceived as subsisting by itself an d as the sub jec t of allth at is conceived w ith rega rd to it, I call a thing. Th is is wh atis also called a substance.

I call a manner of a thing, or mode, or attribute, or quality,w hat , being conceived as in the thing an d as not being able to

subsist without it, determines it to exist in a certain way andcauses it to be called such.

I call it a modified thing, w hen w e consider the substance as deter-mined by a certain mann er or mode.2

T he Arnauld-Nicole terminology is different from Locke's. Th eir thingor substa nce is Locke's substance in the general sense ; their m odi-fied thi ng is, roughly, Locke's substance of a particu lar kin d and

where Arn auld and Nicole usually speak of a m an ne r or mo de,Locke usually speaks of a qu ality . But the doctrine is the same, astheir exam ple shows :

W he n I consider a body, the idea wh ich I hav e of it represents tome a thing or a substance, because I think of it as a thing whichsubsists by itself, and which has no need of any subject in orderto exist.







The efinitions of Substance and M ode

why they are right to say in the School th at the na m e of sub-stanc e is not univocal with respect to God and his creatures,th at is, th at there is no signification of this wo rd w hich w e con-ceive distinctly which applies to him and to them. But since,among created things, some are of such a nature that theycannot exist without others, we distinguish them from thosewhich need only the ordinary concurrence of God, calling thelatter substances, an d th e for m er qualities or attribu tes of these

substances. A nd th e notion we thus have of created substance cor-responds in the sam e way to all, th at is, to those wh ich are im m a-terial as well as to those wh ich are mate rial or co rporeal. Fo r inorder to understand that these are substances, it is necessary onlyto perceive that they can exist without the aid of any created thing.4

Most of Descartes' other definitions of substance are variations on thistheme.® T he re is in th em no explicit reference to an impe rceptible sub ject

or substra tum in which qualities inhere, just the notion of a capacityfor independent existence, more or less strictly construed. If we hadonly these definitions, there m ight b e some tem ptatio n to suppose t ha tDescartes' concept of substance is differen t fro m Locke's, Arn auld 's, a ndNicole's.

In other places, however, we find Descartes giving a more clearlyLockean account of the notion of substance. When he puts into geometricfor m his argum ents fo r the existence of G od an d th e distinction of them ind fr om the bod y, he does define substance as the subject in wh ichqualities inh ere : Ev ery thin g in wh ich there exists imm ediately, as ina subject, or through which there exists, anything which we perceive,th at is, any p roperty or quality or attrib ute of which th ere is a realidea in us, is called a substance (A da m and Tan nery , V I I : 6 ) .A nd h e goes on to explain th at W e have no other idea of substanceitself, precisely taken, than that it is a thing in which there exists, for-mally or eminently, what we perceive, or what exists objectively in someone of ou r ideas, because it is know n by the na tur al light that noth ing






has no real attributes ibid. ). H ere it is clear eno ugh tha t a substan ceis a su bject in w hic h qualities inhere, that it is distinct fro m its qualities,and that, though it is imperceptible in itself, its existence must be inferredfrom the existence of what we do perceive, its qualities, because theycannot exist of themselves (that is, inhere in nothing).

Indeed, Descartes is quite anxious to insist that we have no immediateknowledge of substance. This thesis is reiterated both in his reply tothe objections of Hobbes and also in his reply to Arnauld.® And of courseit is implicit in the Second Meditation, in the discussion of the pieceof wax (Adam and Tannery, VI I : 3 0 - 3 4 ) .

Thus Pierre Gassendi's empiricist criticism of Descartes on this pointcomes as no surprise. W e m ay be said to have a perfect idea of a m an ,Gassendi remarks, if we have observed him carefully, and often, andfrom every angle. To the extent that we have observed him only cur-sorily, or not more than once, or in part, our idea will be imperfect.

However, if we have not examined the man himself, but a maskwhich covers his face and clothes which conceal his body com-pletely, then it must be said that either we have no idea of thisman or, if we have any, it is most imperfect and extremely confused.Hence, I say that of accidents we have, indeed, a distinct andtrue idea, but of the substance lying hidden under them only aconfused and extremely fictit ious one. (Adam and Tannery,V I I : 2 8 5 - 2 8 6 )

Descartes himself supplied the materials for this analogical argumentthrough his use of the clothing metaphor to illustrate his conceptionof the relation between substance and its qualities ( A d am and Ta nn ery ,V I I : 3 2 , I X : 3 5 ) .

The distinction between substance and its qualities needs to be empha-sized, because commentators on Descartes have sometimes blurred it.For ex am ple, M arthin us V ersfeld , noting that there is no real distinction





The D efinitions of Substance and Mode

between substance and attribute, but only a distinction of reason, saysthat substance is simply the principal attribute substantialized . . .there is nothing occult in substances, since the principal attribute, whichis transparent to the understanding, constitutes their nature or essence(Versfeld, p. 96). This seems to me most misleading. There is nothingoccult in substances, except the substances themselves.

No real distinction exists between substance and attribute, becausea real distinction, in Descartes' use of the term, can obtain only between

two entities each of which is capable of existing apart from the other,that is, between two substances (Adam and Tann ery, V I I I 28-30) .Where one or both of the entities is incapable of independent existence,the distinction between them must be either a distinction of reason ora modal distinction. The distinction between substance and its principiliattribute, then, is only a distinction of reason because the attribute cannotbe conceived as existing by itself. But it is no less a real distinctionin the sense of that term in which re al rules out illuso ry.

Nowhere does Descartes indicate this more clearly than in a conversa-tion w ith Frans B urm an, where he comments on the definition of sub-stance given in his geometrical presentation of the arguments for Godand for the real distinction between mind and body

Beyond the attribute which specifies [specificat a mediaevalism]substance, there must further be conceived substance itself whichlies under [substernitur] that attribute as, since mind is a thinking

thing, there is beyond thought the substance which thinks, and soforth. (Adam and Tannery, V : 156 )

Substance cannot, then, be said to be simply the principal attributesubstantialized.

That Descartes defines substance in two such d ifferent ways has natu-rally provided scholars with something of a puzzle. Why does he dothis? Brunschvicg suggests that the definition in terms of existence in







The Definitions of Substance an d Mode

and body are both substances, then, by that definit ion, they can existindependently of one another, and so are really distinct. This, as I say,seems to me correct. I would only add that one other reason why Des-cartes generally preferred to define sub stan ce in terms of a capac ityfor independent existence was probably that such a definit ion wouldlead more readily to a proof of the immortali ty of the soul (see Adamand Tanne ry, IX : ι o ).

In any case, substances have these two features for Descartes : they

exist independently of everything but God, and they are the subjectsin which properties inhere. Either one of these features may be takenas defining. And it is worth noting that Spinoza gives both definit ionswhen he presents Descartes ' Principles in geometr ic form .8

Be fore leav ing Descartes, i t should be observed that the terms m o d eand at t r ib ute hav e a ra ther special m ean ing in his though t . Bothmodes and attributes are properties of substance and cannot be conceivedas existing apart from their subject. An attribute, however, is a very

general, hence invariant, hence essential property of a substance, forexample, thought, or extension. A mode, on the other hand, is a moreor less specific property, hence liable to variation, hence nonessential,for example, my thinking that I exist , or roundness. Since any variable-ness is impossible in God, we predicate of him only attributes, not modes( A d a m a n d Ta n n e r y, V I I I : 26 ) .

I have dwelt at such length on seventeenth-century uses of the termssubsta nce an d m od e par t ly because I wish to emphasize the natural-

ness of the interpretation Bayle puts on them in the long and hostilearticle he devotes to Spinoza in his D ictionary Bayle has had his crit icsin recent years, but many figures of the Enlightenment accepted andrepeated Bayle's cri t icisms of Spinoza, and it is not surprising that theyshould have done so.10

Bayle bases his objections on the attribution to Spinoza of the follow-ing doctrines :1 1 ( 1 ) there is only one substance, 2 ) this substance isGod , and 3 ) al l f inite beings are modifications of God. That Spinozaheld these views, no one who is familiar with the Ethics can doubt .






Ba yle then infers from them that G od is the real subject of allpropositions :

If it we re true, then, as Spinoza claims, that men aremodalities of Go d, one wo uld speak falsely wh en one said, Pe terdenies this, he wants that, he affirms such an d such a th ing ; foractually, according to this theory, it is G od wh o denies,, wan ts,affirm s; and consequently all the denominations that result fromthe thoughts of all men are properly and physically to be ascribedto God.12

A n d Bayle has no trouble showing that this view has consequences th atare indeed astonishing.

Spinoza's God, he maintains, will be as changeable as the Proteusof the ancient poets, for a mutable being is merely one which, its sub-stance remaining the same, acquires new modifications and loses thoseit had. S pinoza's G od will be the subject of contradictory properties,fo r different men love and d o not love, affirm and do not affirm thesame thing at the same time. These contradictory properties m ust beascribed to God in violation of the laws of logic.

M oreo ver, all kinds of m oral enormities must be predicated directlyof God. We can no longer say that Brutus killed Caesar, we must saythat God killed Caesar. Nor, for that matter, can we say that Caesarwas killed by Brutus. We must have it that God was killed by Brutus,

or rather that G od killed himself. A nd so on, throughout a w hole listof odious actions:

All the phrases by which what men do to one another are ex-pressed . . . ha ve no other true sense than this, G o d hates him self,he asks fav ors of himself and refuses them , he persecutes him self,he kills himself, he eats himself, he slanders himself, he executeshimself. (Bayle, p. 3 1 2 )

12





The efinitions of Substance and Mo de

It is small wonder that Bayle charged Spinoza's system with being con-trary to the most distinct notions of our minds.

N ow it is perfectly clear that these consequences wo uld not be ac-ceptable to Spinoza. The immutability of God is a corollary of E IP20.Spinoza's respect for the principle of contradiction seems amply indicatedby his frequent use of the reductio ad absurdum proof technique (EI P 6 C ; E I P I I ; E I P I 3 ). A s for the attribution of h um an sins to Go d,Spinoza would clearly not allow this. Evil, he writes to William de

Blyenbergh, is merely a negation (E p. 23 ), and n othing whic h involvesnegation may be attributed to God ( E I P 8 S ) .

The question is Do these absurd and unacceptable consequences infact follow from the Spinozistic doctrines Bayle says they follow from?Clearly, this depends on how we interpret the terms subs tance andm od e and on w ha t we conceive the relation between substance and

mode to be. Bayle himself is quite insistent that this is the crux of them atter. If there is any m istake in his objections, it lies here. Bu t ifhe is mistaken, it is Spinoza's fault, not his :

I took these terms in the sense that all the new philosophers [forexample, the Cartesians and Gassendists] understand them. . . the general doctrine of the philosophers is tha t the idea ofbeing contains, immediately under it, two species—substance andac cid en t— an d that substance subsists by itself . . . a nd an accidentsubsists in some other bein g . . . T h ey ad d that to subsist by

itself signifies only no t bein g depen dent on any sub ject ofinhesion. 13

Bayle then distinguishes three different views of the nature of accidentsheld by the philosophers of his time. Some, with the doctrine of transub-stantiation in m ind, are force d to recognize a re al distinction betw eensubstance and accident, and to admit that an accident can subsist with-out a sub ject, tho ugh this is contrary to the definition of ac cid en t.

13





Spinoza's Metaphysics

Other philosophers say that, while some accidents are really distinctfrom their subject, some are not. The accidents which cannot exist apartfro m their subjec t they call m od es. A n d finally,

Descartes, Gassendi, and in general, all those who haveabandoned scholastic philosophy, have denied that an accident isseparable from its subject in such a way that i t could subsist afterseparation, and have ascribed to all accidents the nature of those

that are cal led m od es an d hav e employed the term m od e. . . ra ther than that of acc ide nt .

Since Spino za wa s a great Ca rtes ian i t seems only reasonable to sup-pose that he used these terms in the same way the rest of the newphilosophers did, tha t his m odes depend on substance as on a subjectof inhesion . Ind eed , Ba yle is right, i t does seem re aso na ble— par ticula rlyon a superficial comparison of Spinoza's definit ions with those in

Descartes.Nevertheless, I am persuaded that this view, for all its initial plausibil-

i ty, is a mistake. Sin ce B ayle rests his case so heavily on S pino za's C ar -tesianism, I shall try to show tha t he is m istaken by ex am inin g som eof the differences betw een S pino za's philosoph ical term inology an dDescartes'.

W e m ay b egin by c all ing attention to the difficulties that Spin oza'sdefinitions of sub stan ce an d m od e caused one of his most acu tecontemporaries. Leibniz received a copy of Spinoza's Opera Posthumashortly after publication, and the reading notes he made on the Ethicsare very instructive. H e com plains first tha t the definit ion of sub stan ceis obscure (Ger ha rdt , 1 : 1 3 9 ) . F or wha t is i t to exis t in se? A goodquestion, indeed.

Then he wonders about the second def ining phrase. Does Spinozam ean tha t substa nce is w ha t exists in i tself and also what is conceivedthrough itself? Or does he mean that substance is that in which bothoccur together? I f Spino za means the la t ter — as he cer ta inly doe s— then

14





The Definitions of Substance and Mode

It will be necessary for him to prove that what has the one also hasthe other, since, on the contrary, it seems rather that there are somethings wh ich are in themselves, tho ug h they are not conceived thro ug hthemselves. And so men comm only conceive of substance. Thus wehave an added feature in Spinoza's definit ion of substance which makesit puzzling to someone who is thinking in Cartesian terms. For Descartes,tho ug h substan ce is suppo sed to exist in itself, it is conceived thr ou ghano ther. T h e existence of substance, which w e do not kno w imme diately,

is inferred from the existence of its attribute, which we do know imme-diately. Descartes' doctrine is that we cannot know the existence of sub-stance m erely fro m its being a thin g th at needs only itself in o rde r toexist. Th us Spinoza very prob ably h as Descartes in min d w hen he writestha t those wh o con fuse substances an d mo difications will find the dem on-stration th at it pe rtain s to the na tu re of substan ce to exist ( E I p7 )hard to follow:

But if men were to consider the nature of substance, they wouldnot entertain the slightest doubts about the truth of νη indeedthis proposit ion would be numbered among the common notions.For they would understand by substance what is in itself and isconceived through itself, that is, that whose knowledge does notrequire the knowledge of any other thing.1*

Spino za has ad de d a clause to his definition of sub stan ce tha t dis-qualifies most of the things Descartes would call substances from beingso classified. An entity whose existence m ust be infer red fr om th e ex-istence of some thing else, whose k now ledge re quires the know ledge ofano ther thing , such as an attr ibute, would not be a substance in S pinoza'ssense of the term.

This is not, I think, a mere exclusion by definition. Spinoza's viewis that anything which really exists in itself must be conceived throughitself. He does not, as Leibniz would have him do, undertake to provethis. But he could, by arg uin g: (i) tha t knowledg e of an effect dep end s

l





Spinoza s Metap hysics

on and involves knowledge of its cause (E IA4) ; (2) that what existsin itself is not the effect of any external cause;15 (3 ) th at, therefore,knowledge of what exists in itself does not depend on or involve knowl-edge of anything else; and (4) since what is not conceived throughany thin g else must be conce ived thr ough itself ( E IA2 ), w ha t existsin itself must also be conceived through itself.16 Thus the notion ofsomething existing in itself, but conceived through something else—theusual, Cartesian notion of substance—is self-contradictory.

For his part, Leibniz does not merely find the notion that substancemust be conceived through itself as well as exist in itself an unconven-tional one. He argues that it involves an inconsistency with the definitionof attribu te. For Spinoza says in that definition

that an attribute is perceived by the intellect as constituting theessence of substance. Therefore, the concept of the attribute isnecessary to form the concept of substance. If you say that anattribute is not a thing and that you really require only that asubstance does not need the concept of any other thing, I answerthat you must explain what may be called a thing, so that wemay understand the definition and understand in what way anattribute is not a thing.

The objection Leibniz makes here is natural enough, from a Cartesianstandpoint, but the hypothetical reply he is led to put in Spinoza's mouthis wrongheaded. It is not that an attribute is not a thing, but that

it is not another thing. Spinoza, unlike Descartes, does identify substancewith its attribute, or rather, with the totality of its attributes.This point is not immediately obvious from Spinoza's definition of

attrib ute , but it is suggested by his definition of G o d as a substanceconsisting of infinite attributes and is confirm ed as the Ethicsprogresses :

There is nothing outside the intellect by which several thingscan be distinguished from one another, except substances, or

16






w h at is the same thin g (0 4 ) , thei r a t tr ibutes , an d thei r affect ions .

E Ip4D)G o d is e ternal , or a ll of h is a t t r ibutes are e ternal . ( E I p i g )God is immutable , or a l l of h is a t t r ibutes are immutable .

( E I P 2 O C

T h e Correspondence is decis ive on th is point . T h er e w e find that Sp ino zahad o r ig ina l ly de f ined a t t r ibu te in m uc h the same w ay he de f ined

s u b s t a n c e .

By a t t r ibute I unders tand whatever i s conceived through i tse l fand in i tself , so that the conception of i t does not involve theconcep t ion of ano ther th ing . (E p . 2 , I V : 7 )

Ev en af ter he ha d cha nge d the def ini t ion of a t t r ibu te to w h at i t isin the Ethics, he regard ed these terms as tw o nam es for the same thing :

B y substance I und ers tand w ha t is in i tse lf an d is con ceive dthrough i tse l f , that i s , that whose concept does not involve theconce pt of anoth er th ing. I und ers tand the same by a t t r ibute ,except that i t is called attr ibute with respect to the intel lect , whichat t r ibutes such an d such a natur e to substance. (E p. 9 , I V : 46 )

As an example o f how one and the same th ing can be ca l l ed by twonam es, he offers the th i rd pat r iarch, w h o was cal led both Israel an dJa co b. S o substa nce is not , as w ith Desc artes, a dist inct enti ty. I t isthe at tr ibutes themselves.

Because Spinoza does not d is t inguish a t t r ibute f rom substance, hewill al low that his at tr ibutes possess the defining characterist ics of sub-s tance. T h a t ev ery a t t r ibute . . . mu st be conc eived thro ug h i tse l fis expl ic i t ly a theorem of Spinoza 's sys tem (E IP I O) . A n d i t is notdi ff icul t to const ruct a proof that what i s conceived through i tse l f mustexist in i tself . For if i t existed in something else, knowledge of i t would

17Brought to you by | Brown University Rockefeller Library

Authenticated | 128 148 252 35Download Date | 6/18/14 1:50 AM




depend on knowledge of that in which i t exis ted (by E IA4), and i twould not be conceived through itself .17 Leibniz , therefore , is wrongw he n he says that mode s dif fer fro m attributes in tha t a n attributeexis ts . . . in a substance but is conceived through i tself , whereasmodes exist in and are conceived through substance. In Spinoza's schemeof things, each attribute exists in itself and is conceived through itself.

This example alone is enough to show that Spinoza's use of theseterms is highly idiosyncratic. But further differences appear when weexamine his notion of a mode. Spinoza's modes are, prima facie, ofthe wrong logical type to be related to substance in the same way Des-cartes ' modes are related to substance, for they are particular things(E Ip25c) , not qual i t ies . And i t i s diff icul t to know what i t wouldm ean to say that particula r things inhere in substance. W h e n qualit iesare said to inhere in substance, this may be viewed as a way of sayingthat they are predicated of i t . W h a t i t w ou ld m ean to say that onething is predicated of another is a mystery that needs solving.

No doubt we can give a sense to the statement that one thing ispredicated of another. Bayle, in effect, does this by interpreting Spinoza'sdoctrine that every finite thing is a mode of God to mean that everyproperty of every finite thing is predicable of God. But once we seethat som ething of this sort must be d one, the assimilation of Spin ozato Descartes loses a good deal of its initial plausibility.

These two features of Spinoza 's phi losophical terminology—that hismo des are particu lar things a nd th at his substance is no t distinguishedfrom i ts a t t r ibutes—may help to explain the fact , pointed out by H. A.W olfson, that the term su bje ct , wh ich figures so prominent ly intraditional definit ions of substance is absent from Spinoza's. Neitherhis modes nor his attributes can exist in substance as in a subject, theydo not inhere in substance. For the attributes do not exist in anythingother than themselves, and the modes are themselves the sort of thingthat would properly be the subject of predicat ion. I t may be too muchto say, as W olfson does, that the term su bje ct has no m eanin g atall for S pino za (I , 72 ) . H e d oes, after all, use the notion of existence

18





The D efinitions of Substance and Mod e

in a subject on occasion (Ε IIIP5). But we can accurately say thatthe term would have been of no use to him in explaining the relationshipof either attributes or modes to substance.

When Spinoza does undertake to characterize the relation betweenmodes and substance—beyond saying that they exist in i t and are con-ceived through it—the terms he uses are quite different. He says thatGod is the cause of the things that are in him E IPI8D ) , that hehas produced them (E IP24), that he has determined them E IP26) ,

or that they follow from him (E IP28D) . A n d since he identifies sub-stance and attribute, he is just as likely to say that the modes are deter-mined by, or produced by, or follow from an attribute of God, or theabsolute nature of an attribute of God (E IP2 1- 23 E IP28D) . Thisterminology may be vague and confusing. Indeed, i ts clarification willtake all of the next chapter. But it does not in the least suggest inherencein a subject. W h a t it does suggest is merely som e kind of causa ldependence.

It may be, of course, that Bayle himself did not take this interpretationseriously. Bayle's intentions are a paradigm of obscurity. According toone very popular view, his professed fideism is merely a way of concealinghis true purpose, which is to destroy orthodoxy by showing its absurdity.Because he claims to believe what is absurd in spite of its absurdity,he is protected from those in authority. But he gets his real messageacross to the free spirits who know how to read him.

So, it might be argued, Bayle's real target is not the alleged atheismof Spinoza, but the doctrines of transubstantiation and the Trinity.These, he suggests, are quite similar to his version of Spinozism andapp ear contra ry to the most distinct notions of our m ind s in justthe same w ay. Bay le is very ca reful to point out where the flaw inhis interpretation of Spinoza must lie, if there is any. He acknowledgesthat the usual reading of Spinoza is different on this point. And hesupports his own treatment of Spinoza with arguments that are some-times very peculiar.18

Moreover, the charge that Spinoza's system is false because it is con-

19







The Definitions of Substance an d Mode

and in this Descar tes and Locke and Arnauld represent a t radi t ion goingback ultimately to Aristotle. But i t is one of Spinoza's principal noveltiesthat he breaks with this long tradition.

I have already indicated my main reasons for this content ion, andthere is no need to repeat them here. W h a t I w ish to do now is toconsider some of the conclusions Joachim comes to concerning the statusof finite modes in Spinoza's system as a result of his interpretation ofthe definit ions of substance and mode.

Roughly, his posit ion is this particular things may be viewed in eitherone of two ways—as par ts of natura naturata or as parts of the commonorder of nature. In so far as they are viewed as parts of natura naturatathey are seen in the context of a modal system which expresses God'snature and are seen to be t imelessly and necessarily actual. And thisis the way they really are. But so viewed, their particularity vanishes.They are completely real, but not individually distinct. Individualitycan properly be attributed only to God.

O n th e other han d, in so fa r as partic ular things are view ed as partof the common order of nature, they are seen in abstraction from themodal system on which they depend and are considered to be indi-vidually distinct but finite, transitory, and not completely real. This isthe world of presentat ion, the phenomenal world, in which things occurat a particular t ime and place, interact with one another, result frominfinitely long chains of causes, initiate infinitely long chains of effects,change, come into being, and pass away. These occurrences are unintell i-gible and fortuitous, although ascribed to necessary and universal laws.And this world is never more than an i l lusion.

Thus, Spinoza's metaphysic turns out to be monistic in a very strictEle atic sense. Spin oza's substance, Jo ach im says, is one (no t as a unityof diverse but related elements, but) as a unity which has overcomeand taken into itself the distinctness of its diverse elements, and thisabsorption is so complete that in i t there remain no 'elements, ' no dis-t inctness, no articula tion (p. 10 8) . For all things are m erely statesof the one substance, God. It is only by an abstraction that they can

21







T h e D ef i n i t i o n s o f Su b st a n c e a n d M o d e

a finite body; it owes its nature to its place in the whole systemof bodies which together constitute the corporeal universe. Anyattempt to explain it—to understand its essential nature—wouldcarry you outside it , or would force you to regard it as havingno essential natu re or individuality; for it is through and throughconstituted by its relations, and if you include the m in its na tur e,

it will have becom e merged in the whole attribute of extension.

P· 23)

Th is is the familiar flower-in-the-crann ied-wa ll motif. An d no doubta good part of what Joachim says here is correct. Spinoza does, ofcourse, hold tha t finite bodies depend fo r their existence on othe r finitebodies ( E I P 2 8 ). I think I should have said that they depend for theirnature , not on their pla ce in the whole system of bod ies, but on theattributes and the infinite modes. But that is a difficult point, and inany case the natures of particu lar things do depend on something outsidethose things. And since knowledge of an effect depends on and involvesknowledge of its cause, understanding the essential nature of a bodywould require going outside it.

Ho wev er, Spinoza never infers from these facts that bodies have n oessential nature or individuality. He denies explicitly that their essentialnature is constituted by their relations.21 And in the later parts of theE t h i c s he seems to identify the essence of particu lar things with th eforce by which they persevere in existence E I IP45S E IIIP 6, 7 ).Indeed, this notion is so important for Spinoza's ethical theory thatit is difficult to see how h e cou ld deny t ha t things have an essentialnature.22

Moreover, the view that the individuality of particular things is swal-lowed up in that of the whole whose parts they a re seems quite con traryto the discussion that follows E HP 13. There Spinoza is concerned toshow how th e whole of nature c an be conceived as one individual,whose parts, that is, all bodies, differ in infinite ways without any

change in the whole individual E IIL7S). He begins by considering

3






the simplest bodies, which are distinguished from one another by motionand rest or speed and slowness. This he supposes to be self-evident.He then posits that a composite body will constitute one body or indi-vidual if i ts parts communicate their motion to one another in a certainfixed proportion. So long as the parts preserve toward one another thesame kind of motion and rest , the individual will retain the same na-ture—even i f the par ts are replaced by others . Composi te bodies maybe of varying orders of complexity; some are composed only of simple

bodies, others are composed of parts which are themselves composite.But the principle in virtue of which a composite is regarded as oneind ividu al is the sam e, regardless of its com plexity. A n d it is in thisway that the whole of nature can be conceived as one individual .

Now the interesting thing about this discussion is that Spinoza seemsto regard th e indiv idua lity of these simplest bodies as som ething tha tis immediately intell igible, whereas he regards the individuality of thewh ole of nature as something requir ing a long and elaborate explanat ion .

But on Joachim's version of Spinoza, i t is the other way around.2 3

O n l ythe whole of nature is truly an individual. The individuality of finitethin gs is an i l lusion of the finite inte llec t— an il lusion w hi ch is very

difficult to understand because this finite intellect is itself supposed tobe an illusion.

The discussion which fol lows E II is , of course, nothing but anelabo ration of E IID7 : B y singu lar things I und erstand things w hi chare finite and have a determinate existence; and if several individuals[N S or singular things] so concur in one action that they are all simul-taneously the cause of one effect, I consider them all , so far [my italics],as one individua l thin g. T h e qual i f icat ion so fa r should be enou ghby itself to assure us that Spinoza does not think that particular thingslose their individuality by being considered as parts of a compositeindividual .

T h e second matter whic h requires comm ent is the at tr ibut ion toSpinoza of i l lusionism. Joachim is not clear how far he wants to goin this. In one place , he says that the world of pre sen tatio n— the wo rld

24





The efinitions of Substance and Mode

of things in time and place . . . is not, it seems, ever more than amere illusion ( pp. 11 2- 11 3 ) . But earlier he says that this world ofisolated and perishable things, with its apparently arbitrary se-quences . . . this world of the unscientific experience is largely il-lusory . . . Bu t it is not through and through illusion or, if it is, theremust at least be a real basis of the illusion (pp. 7 9 - 8 0 ) . Th us, thephenomenal world may, perhaps, be something more than an illusion.The reason for this hesitation is that Joachim 's Spinoza must explain

the illusion as due to the defects of the finite human intellect, therebyimplicitly attributing some reality to finite things. Spinoza, Joachim says,is fa r too ready to dismiss things as 'mere illusions' (p . 1 14 ), forhe needs them to explain the fact of the illusion.

In part, again, I think this is a deduction from Joachim 's interpreta-tion of the definitions of substance and mode. For him, Spinoza's sub-stance is a unity in which there are no real distinctions. The phenomenalworld is a p lurality of d istinct entities, so the phenom enal world must

be an illusion. But Joachim also apppeals to passages where Spinozasays such things as T h e hum an m ind, whenever it perceives thingsfrom the common order of nature, has adequate knowledge neither ofitself, nor of its own body, nor of external bodies, but only a confusedand mutilated knowledge (E IIP29C). So to evaluate Joachim's inter-pretation, we need to understand the meaning of phrases like perceptionof things from the common order of natu re and adequa te knowledge.For Spinoza never says tha t the world of the finite is an illusion, andwe must know w hether what he does say entails that it is an illusion,a deceptive appearance of a reality which is in fact quite different fromwhat it seems to be.

In the scholium to the corollary quoted above, Spinoza explains whathe intends by the phrase perception from the common order of naturein the following w ay :

I say expressly that the m ind has adequate knowledge neither ofitself, nor of its body, nor of external bodies, but only a confused and

5






mutilated knowledge, whenever it perceives things from the com-mo n order of nature , that is, wh enever it is determined to con-temp lating this or that externally, from w hat is indeed a chancemeeting with things—and not when it is determined to understand-ing their agreements, differences and oppositions internally, fromthe fact that it contemplates several things at once. Fo r w hene verit is ordered internally, in this or some other way, then it contem-plates things clearly and distinctly.

Th is exp lanation is not as help ful as we might wish, but it is at leastclear that percep tion from the comm on order of na tu re is anothernam e for what is elsewhere called kno wled ge of the first kind, or opin-ion, or imagination (E I IP4 S2 TdlE, 19, I I : o). Of this, Spinozadistinguishes two subclasses : knowledge from report or signs, such asthe knowledge I have of my birthday or my parentage, and knowledgefrom vagrant experience (expe rien tia vaga), such as my know ledge

that I , being a m an like other men , will some day die or that w aterextinguishes fire. These are all things I know because I hap pen to h avebeen so placed in the world that I have ha d experiences of one kindand not those of a contrary kind. Al l the wate r I hav e seen has, wh enput to the test, extinguished fire. Thus, I am determined to these percep-tions externally and fortuitously. Fo r I hav e experienced only a verysmall part of what might have been experienced, and what I have seenmay be misleading. My ideas are very inadequate, they do not haveall the properties or internal marks of a true idea ( E I ID 4 ), they arenot clear and distinct.

On the other hand, when my perceptions are determined internally,then my knowledge is either of the second kind, reason, which perceivesthe thing in question through hav ing the comm on notions and adequ ateideas of the properties of things, or of the third kind, intuition, whichproceeds from an adequate idea of the essence of a divine attributeto an adequate knowledge of the essences of things. Spinoza commonlyillustrates the differences between these different kinds of knowledge

26





The Definitions o f Substance and Mod e

by appeal ing to a mathematical example. Suppose we are given threenumbers , a b a n d c and told to f ind a fourth which bears the sameratio to c as b does to a. We may have been taught in school to mult iplyb a n d c dividing their product by a to find the correct answer. Thiswould be knowledge from report , one form of opinion. Or we maydiscover this rule for ourselves by experimenting with cases where theanswer is obvious an d gen eral izing from them . This inductive proc edu rewou ld yield knowledge fro m va gra nt experience, anothe r form of opin-

ion. But if we have studied Euclid, we will know that this is the proce-dure to fol low because we understand the common property of propor-tionals. This would be knowledge of the second kind, reason. And if thenum ber s are very s imple ones, we may see at a glance w ha t the rat iois betwee n a an d b and infer fro m this the correct answer. Th is wo uldbe intuition, or knowledge of the third kind.

N ow ther e is m uc h t ha t is difficult in this classification of the kind sof knowledge. But this much can be said. Spinoza is distinguishing be-

tween the various ways in which we may at tain knowledge. And theprincipa l thing is tha t some of the m (reason, intui t ion) give adeq ua teideas, which bear al l the internal marks of t ruth, and therefore arecertainly t rue, whereas the others (repo rt , va gra nt experience) give in-adequate ideas, which do not bear al l the internal marks of t ruth, arenot clear and dist inct , and therefore are not known with certainty tobe true.

But i t does not fol low tha t percep tions s temm ing fr om these lat tersources are false. A fte r all, Spinoza s exam ple indicates tha t even theeternal t ruths of mathematics may be, and perhaps in most cases are,know n th rou gh one of these defective mod es of cognit ion. So a perce p-t ion ar is ing from the com m on orde r of na tur e is merely uncertain an dneed n ot misrepresent real ity. N o d oub t such perceptions often do, an daccording to Spinoza knowledge of the first kind is the only cause offals ity ( E IIP 41 ) . But perceptions fro m the com mo n orde r of n atu reare no t— as Jo ach im s talk of i llusions un fortun ately suggests—invariablyfalse. And Spinoza evidently reposed enough faith in them to be able

7






to assert, as a corollary to E P 3, that the human body does existas we perceive it, in spite of the fact that our perception of i t fromthe common order of nature is inadequate .24 So the talk of inadequateknowledge gives us no warrant at all for dismissing the phenomenalworld as a mere i l lusion. And an interpretation of Spinoza's definit ionsof substance and mode which is reduced to such desperate expedientsneeds more supporting evidence than it has ever been given.

The Wolfson InterpretationIn 1934, there appeared on the scene an alternative interpretation

of Spinoza 's def ini t ions, put forward with overwhelming erudi t ion byH. A. Wolfson. Though diametr ical ly opposed to the l ine of interpreta-tion followed by Bayle and Joachim, i t too has a very high degree ofinit ial plausibil ity an d has achieved sufficiently w ide ac cep tan ce tha ti t is already beginning to appear in textbooks on the history of philoso-phy.2 5 Wolfson takes us on an exci t ing phi losophical journey, beyondDescartes and the rest of Spinoza's contemporaries, back through themedievale to Aristotle, in an attempt to reconstruct the Ethica MoreScholastico Rabbinicoque Dem onstrata w hic h must hav e preceded, inSpinoza 's mind, the Ethica Ordine Geom etrico Dem onstrata wh ich wasleft us at Spinoza's death. I t is the evidence for this reconstruction thatwe must now examine.

O n this interpretation, the term sub stan ce denotes the summumgenus, Substan ce; the term m od e denotes individuals , or individua lessences; and the relation of mode to substance is a relation of individualor species to genus. When Spinoza says that the modes exist in substance,he is using this phrase in the same sense Aristotle uses when he saysthat m an is ' in ' anim al, and generally species ' in ' gen us (Physics,21 oa, 18 ). B ut Substa nce, since it is the summum genus, does not existin anything else in that sense, it is not a species of any genus, andso it must be said to exist in itself.

This view has i ts own, very considerable historical precedent and,i t would seem, great explanatory power. Consider, for example, the

28








M a n is thus conceived through his genus an im al and his species[difference?] rat ion al, and he is thus also defined by the combina-tion of these two terms. And so everything that is in somethingelse, as an in dividu al in its genus, may be thus said to be conceivedby that something else. (I, 76)

Substance, on the other hand, as it is not a species of any genus, isnot conceived through anything else. Hence, by E I 2, it must be con-

ceived throug h itself. W olfson , how ever, regards the phrase con ceiv edthrough itself as hav ing a purely negative imp ort, so that the axiomWhatever is not conceived through something else must be conceived

through its elf is a disguised tautology , an d the im plication of sayingthat substance is conceived through itself is that Sp ino za's substanceis inconceivable, and its essence undefinable, and hence, unknowable.And this agrees with the medieval doctrine according to which thatunique substance, God, is undefinable and in his essence unknowable.

Wolfson's interpretation has other fruitful consequences. For instance,it provides an apparent explanation of Spinoza's view that substanceis prior in nature to its modes ( E IP I ), since Wolfson can point tothe Aristotelian principle th at the genus is prior to the ind ividua l( Topics, 14 1a , 2 6 ff ) . Ag ain, when Spinoza says that Go d is the imm a-nent cause of all things (E I P I 8 ), W olfson explains this by adducingM aim onid ean and Aristotelian texts according to which the genus isthe internal cause of the species or individual essence (I , 3 1 9 - 3 2 4 ) .

And so it goes. Wolfson begins his work by claiming that if we hadall the philosophical literature available to Spinoza, we could reconstructnearly the whole of the Ethics just by making appropriate selectionsand arrangements of the doctrines of his predecessors. This is an extremethesis—one which Wolfson is forced, later, to qualify in curious ways,26

and one which makes Spinoza's originality quite problematic. ButWolfson defends it very persuasively.

Nevertheless, I shall contend that this interpretation of the definitionsis fundamentally incoherent and that some, at least, of the consequences

3






W olfson draws fro m his interpretation are in clear conflict with Spinoza'sexplicit doctrine.

First, let us reflect that like m ost of the philosophers of his time,but unlike most of the medievals, Spinoza has, in general, little usefor arguments from authority. The knowledge we have of things fromreport only is hardly worthy of the name knowledge. Indeed, in theShort Treatise KV II, i, 1 :5 4 ) Spinoza says that the man wh o appliesthe rule of three to find the fourth proportional merely because this

is what he has been told to do has no more knowledge of the rulethan a blind m an has of color. An d, in p articular, he has little use forthe authority of Aristotle. Writing to Hugo Boxel, who had tried topersuade him of the existence of ghosts, he says that

To me the authority of Plato, Aristotle, and Socrates is not worthmuch. Had you cited Epicurus, Democritus, Lucretius, or anyof the atomists and defenders of atoms, I would have been aston-ished. But it is no cause for wonder that those who invented occultqualities, intentional species, substantial forms, and a thousand othertrifles, devised specters and ghosts and trusted old women in orderto weaken the authority of Democritus. They were so envious ofhis good reputation th at they burned all his books, wh ich he hadpublished with great praise. If you are disposed to put faith inthem, what reason have you for denying the miracles of the HolyVirgin and all the Saints? These have been described by so manyillustrious philosophers, theologians, and historians, that I could

produce a hundred of them against hardly one of the other. Ep.56, IV: 261-262)

This does not sound much like a man whose work would first havebeen composed more scholastico rabbinicoque.

A ga in, consider the identification of Spinoza's one substance withthe summum genus of traditional philosophy. This is to say, in effect,that the only thin g w hic h satisfies the definition of sub stan ce is the

31






genus Substance itself. T o b e a substance, for W olfson's Spinoza, isto be not-in-anything-as-a-species-is-in-its-genus. The genus Substanceis not-in-anything-as-a-species-is-in-its-genus, for it is the summumgenus. T he genus Substance is, therefore, a substance, a m em ber ofthe genus Substance. A nd since, for anybody's Spinoza, there is onlyone thing that satisfies the definition of substance, the genus Substancemust be the only thing that is not-in-anything-as-a-species-is-in-its-genus.Everything else must be a species of some genus, including individuals.

Now this, if you stop to think about it, is a very strange doctrineindeed. First of all, it requires us to treat a class, the genus Substance,as a m em ber of itself. A nd m any logicians now adays, w ith R ussell'sparadoxes in mind, would reject the notion of a class being a memberof itself as unintelligible. N ot every logician today w ould re ject it, a ndeven if every logician did, this fa ct by itself w ould do n oth ing to showw ha t Sp inoza's views are. But anyon e who feels unco mfo rtable withclasses that are members of themselves will naturally suspect that an

interpre tation which requires the m ma y ru n into othe r difficulties aswell.A nd so it does. Consider the identification of Spinoza's modes w ith

individuals or individual essences. It ma y have been noticed th atWolfson, with a nice disregard for logical proprieties, uses these twonotions interchangeably. But surely they c anno t be so used.27 If themodes are individuals, then they cann ot b e said to exist in Substanceas a species exists in its genus. W e mu st say th at they exist in S ubstan ceas an individual in its genus. And Wolfson does say this in variousplaces. But this cannot be a correct account of Spinoza's philosophy.For if the modes exist in Substance as individuals exist in a genus,then the modes are, in plain language, substances. For a genus mustalways be predicable of the individuals that exist in it.

So the modes must be individual essences an d their relation to Sub-stance must be that of species to genus, in spite of the fact that Spinozacalls them individual things. It is something of a mystery what the statusof individual things would be on this interpretation. They are evidently

32





The Definitions o f Substance and Mode

not modes, they cannot be substances, and these are presumably theonly two categories available. Perhaps, appearances to the contrary, thereare no individual things, only genera and species, though this wouldhardly seem Aristotelian, much less Spinozistic.

These are, of course, not decisive objections. They raise questionsabout the coherence of Wolfson's interpretation or about its compatibilitywith Spinoza's presumed spirit and tendencies, but do not show it tobe wrong. There are, however, other objections, which are of a textual

nature and are, it seems to me, decisive.In the first place, it is not at all clear that Spinoza would regard

the genus Substance as the summum genus. Bay le, it will be recalled,said that the general doctrine of the philosophers of his time was thatBeing is the summum genus, and that Substan ce is a species of thegenus Being. There may be good reason for denying this, on the tradi-tional ground that B e in g is not predicated univocally of the categories,but there are very strong indications that this was also Spinoza's view.In his Metaphysical Thoughts, he begins by defining Bei ng, and goeson to remark that from his definition, it is easily seen that : Beingis to be divided into being wh ich exists necessarily, or whose essenceinvolves existence, and into being whose essence involves only pos-sible existence CM I, i, 1 : 2 3 6 ). This last category, he says, isdivided into substance and modes. Here he is in part expounding theviews of Descartes. If he were speaking for himself, he would makethe division between substance and mode correspond to that be-tween necessary and contingent being; it would not be a subdivi-

sion of contingent being. Bu t there is every reason to suppose that hewould regard the genus Substance as a subaltern genus of the genusBeing. In the Ethics he treats Being explicitly as the summum genus:W e are accustomed to refer all individuals in nature to one genus which

is called the most general, that is, to the notion of Being, which embracesabsolutely all the individuals in nature E I V pref., II : 207). In thetraditional Aristotelian philosophy, it could be said of Substance thatit embraces all individuals in nature because entities in the other cate-






gories are not individuals but qualities, relations, and so forth. But inSpinoza's philosophy, this could not be said, since the category of modeincludes individuals. N or would the traditional ground for denying thatBeing is the summ um genus hold, for there is a definition of be ingwhich can be predicated univocally of both substances and modes.

But quite apart from the question whether Spinoza would regardSubstance as the summum genus, it seems highly unlikely that he wouldbe willing to treat any genus as the cause of all things. In the Treatise

on the Correction of the Intellect, he writes that

We should inquire whether there is any being, and if so, what sortof being there is, which is the cause of all things. In this waythe being's subjective essence essentia objectiva, used in the Ca r-tesian sense, as a representation in the intellect of an external reality]will be the cause of all our ideas, and our mind , as we have said,will reflect Natu re as fa r as possible. For the mind w ill have sub-

jectively Nature's essence and order and union. From this we cansee that, above all, it is necessary for us always to deduce all ourideas from physical things, that is, from real entities, proceeding,as fa r as possible, according to the series of causes, fro m one rea lentity to another, so that we do not pass over to abstractions anduniversals, lest we either infer something real from them, or inferthem from something real. Both inferences break the true progressof the intellect. (TdlE, 99, II ¡36)

From this it seems clear enough that universals and abstractions, underwhich heading genera and species fall, are not regarded by Spinozaas real beings, and hence not possible candidates for the cause of allthings, substance.28

These facts—that Spinoza did not regard Substance as the summumgenus and that he denied the reality and causal efficacy of un iversals—weigh very heavily against Wolfson's contention that the modes existin substance as a species in its gen us, and also against his related

4






interpretation of the doctrine that God is the immanent cause of allthings.

There is a similar objection to his explanation of the doctrine thatthe modes are conceived through substance in terms of the Aristotelianview that individual essences are conceived through the genera and spe-cies which form their definit ions. For in rejecting the notion of universaleas causes, Spinoza also rejects the theory of definition associated withit . This may be seen in the Short Treatise wh ere he takes up the spe-

cious arg um ents by w hic h traditiona l philosophers seek to excuse theirlack of know ledge of G o d (K V I, vii , I ¡46). One of these arguments isthat God, because he is not a species of any genus, cannot be properlydefined, for a correct definit ion must be by genus and difference.

Spi no za replies that alth ou gh all logicians adm it this, I do not kno wwh ere they get i t fr om . Fo r if i t is true, w e can hav e no know ledgeat a l l . Knowing a thing perfect ly, on the t radi t ional view, involves know-ing it through a definit ion by genus and difference. From this i t follows

that we cannot know perfect ly the summum genus because i t has nogenus above it and cannot be given a definit ion by genus and difference.But on this traditional view, the summum genus is supposed to be thecause of our knowledge of all other things. If i t is not known, thennone of the things that are explained by that genus can be understoodor known—that is , nothing at a l l can be known. Hence, Spinoza proposesa new theory of def init ion, fol lo w ing our divis ion of na ture , accord ingto which things will not be defined by genus and difference.

Definit ions, he says, must be of two kinds. Those of things whichsubsist throu gh themselves, the attributes, ne ed no genus, or any thing ,throug h wh ich they m ight be bet ter understood or expla ined . Th oseof things which do not exist through themselves, the modes, requirea reference to their attribute, thr ou gh w hic h, as their genus they mustbe un dersto od. Spin oza does not elabora te furth er on this ne w theoryof definition in the Short Treatise.

In the Treatise on the Correction of the Intellect TdlE 96-97 ,I I 35 -36 ), how ever, he gives a fuller acco un t. T he re he mak es it clear

5






that the phrase as their gen us must not be pressed too far. T h e at-tributes are not related to modes as genus to individual. The phrase

as their gen us indicates a com parison only. For the first rule for thedefinit ion of created things is that the definit ion must mention theirproximate causes . And Spinoza la ter explains that the proximate causesare his fixed and eternal things, which, though they are singulars, onaccount of their presence everywhere and their most extensive power,will be to us like universale, or genera of the definitions of singular

m u ta b le th in g s ( T d l E , ι ο ί , 11 : 3 7 ) .It is clear, I think, that Spinoza could make precisely the same objec-

tion to Wolfson's interpretation of Spinozism that he makes to the tradi-t ional philosophers who seek to excuse their ignorance in theology. Byident i fying God or substance with the summum genus, th rough whichall things are supposed to be conceived, and then declaring, as on thetradi t ional theory he must , that Spinoza 's substance is inconceivable ,Wolfson commits Spinoza to the same absurdi ty Spinoza is complainingof in the traditional philosophers nothing is knowable.

A n d of course, as w e migh t expect f rom al l this , Spin oza 's G od isnot—though Wolfson says he is—-inconceivable . His essence is not unde-f inable and unknowable. For qui te apar t f rom the fact that Spinozagives us a definit ion of God (E ID6) , he also proves, as a th eore m ,that th e hu m an min d possesses an ade qua te kno wled ge of the eternaland inf inite essence of G o d (E I IP 4 7) . I t wo uld be diff icult , I suppose,to find so bold a claim in any prior philosophical l i terature, even ifwe were to include what was not avai lable to Spinoza. But perhaps

that indicates the defects of an excessively historical approach.

Notes T ow ar d a N ew Interpretation of Spinoza'sDefinitionsT h e m oral I a m inclined to dr aw from this discussion is tha t if weattempt to interpret Spinoza very much along the l ines of some oneor another of his predecessors, we will be hard put to make sense of

his philosophy and are l ikely to find ourselves attributing to him doctrines





The Definitions of Substance and Mod e

that are foreign to his thoug ht. W e m ight say of Spinoza's m etaphysicswhat Artemus Ward is reputed to have said of the early English T h eresearches of many eminent antiquarians have already thrown muchdarkness on the subject ; and it is probab le, if they con tinue their labors,that we shall soon kn ow n othin g at all about it. T hi s is not to suggestthat Spinoza's concept of substance bears no relation at all to traditionalconcepts of substance. It does. But the precise nature of that relationshipneeds some examining.

It is often said that Spinoza arrived at the main doctrines of hisphilosophy by following out the logical consequences of the Cartesianconcept of substance. And in a sense, that is certainly true; but onlyin a sense. The Cartesian distinction between substance and mode, aswe have seen, involved two elements a distinction between independentand dependent being and a distinction between subject and predicate.And these two distinctions were usually thought to coincide the subjectsof predication were thought to be independent entities and the qualitiespredicated of them were regarded as dependent. If I understand himcorrectly, Spinoza, in classifying particular things as modes, was intenton emphasizing the fact that the two distinctions do not coincide, thatwhat is the subject of predication is not an independent entity. He didnot intend to say that the relation of particular things to God was inany way like the relation of a predicate to its subject.

This does not mean merely that he recognized and insisted upon theScholastic and Cartesian commonplace that the finite things which aresubjects of predication are not strictly substances because of their de-pendence on God. It means also that he would not accord them eventhat relative independence which they possess in Descartes. Thus in thepreface to his Principles of Cartesian Philosophy, wh en he wishes togive examples of points on which he disagrees with the author whomhe is expounding, he mentions the Cartesian view that minds and bodiesare substances. And he gives as his reason for disagreement, not thefact that minds and bodies depend on God—his version of the Cartesiandefinition of substance as what requires only the concurrence of God

in order to exist PP IID2, 1:181 ) would preclude that object ion—but

7





Spinoza s Metaph ysics

the fact that they depend on other minds and bodies. T he imm ediateissue is whether the will is distinct from the intellect and free. Spinozaobjects that

in these assertions . . . Descartes only stated and did not provetha t the h um an mind is absolute thinking substance. Although ourAuthor [the preface was written by Lewis Meyer, at Spinoza's direc-tion] admits that there is a thinking substance in nature he deniesthat it constitutes the essence of the human mind. He holds that,in the sam e way extension is determined by no limits, thou ght isdeterm ined by no limits. A nd as the h um an b ody is no t absolute,but only extension determined in a certain way, according to thelaws of extended nature, by motion and rest, so also the mind,or the hu m an soul, is not absolute, bu t only thou ght determine din a certain w ay, according to the laws of thinking na ture , by ideas.

The distinction between the independent and the dependent does not

correspond to the distinction between su bject an d pred icate, even ifwe qua lify the no tion of inde pend ent existence in the way Descartesdoes.

Of course, the distinctions could be made to correspond, trivially,by saying w ha t B urgersdijck says that in speaking of independencein this connection we intend to exclude only th at k ind of depen dencewhich a predicate has on its subject and not any other kind of depen-dence. But Spinoza does not choose to do this. He simply defines sub-stance in terms of independent existence, taken in the broadest sense.An d he makes the contrast between independen t an d de pendent being,not the contrast between subject and predicate, the funda m enta l antithe-sis of his philosophy. Insofar as Spinoza retains the notion of independentexistence, his concept of substance is a traditional one. But insofar ashe makes no use of notions arising from the logical analysis of proposi-tions, it is radically new . It m ay be, as Russell once said, th at allsound philosophy should begin w ith an analysis of propos itions. Butif that is so, then Spinoza's philosophy is unsound.






One consequence of Spinoza's radicalism in regard to his conceptionof substance is that certain criticisms that have been directed againsthim are really irrelevant to his philosophy. To take one instance,A. E. Taylor has argued {Mind 55 :1 03 -1 0 4 [1946] ) that Spinozaassumes, falsely, that every true proposition is one which asserts or deniesa predicate of a subject. That this is one of Spinoza's assumptions hethinks is clear from the definitions of substance, attribute, and mode, andfrom th e fact tha t Sp inoza does not recognize any other type of existent.

That it is false he thinks is clear from the fact that an assertion ofa relation between tw o terms c annot b e reduced to the ascription ofa predicate to a subject. But as H. F. Hallett has quite correctly replied{Mind 55:286 [1946]), the fact that relations are not thus reducibleis quite beside the point. This kind of objection depends on assimilatingSpinoza's use of the terms subs tance, attrib ute, and m od e to thatof his contemporaries but Spinoza's use is very different.29

Again, Russell has argued that because Spinoza's metaphysic requires

the concept of substance, and because that concept is one w hich neitherscience nor philosophy can now adays acc ept (p. 601 ), the metaphysicmust be rejected. Russell does not explain, in the passage where hesays this, w hy science and ph ilosophy can no longer accep t the con ceptof substance but it is clear from his discussion of Berkeley and Humethat he has in mind the empiricist criticisms of the concept of substanceheld by such people as Descartes and Locke, criticisms which turn pri-marily on the notion of substance as an imperceptible substratum.30

If, as I have argued, Spinoza does not conceive of substance in thatway, then his concept of substance is not, on the face of it, open tosuch criticism.

It may be asked, of course, whether the concept of a being inde-pendent of all causes is any more intelligible than the concept of animperceptible substratum. If it is not, then Spinoza is no better off thanif he held views similar to those of Locke and Descartes.

The brief answer to this, I think, is that the concept of an independentbeing is just as intelligible as the concept of a dependent one. Since

9






we kno w w ha t i t mean s to say that one thing depends on another,we must know what i t means to say that i t does not depend on another.And the idea of an independent being is simply the idea of a beingthat does not depend on any other. W e m ay think, on empir ical ground s,that there is no such independent being. Everything, as far as our experi-ence goes, is depen dent in na ture. A n d presum ably if there is no inde-pendent being, then the concept of an independent being is a uselessone. But to say this is not to say that it is unintelligible.

Th is answer, however, is not adequ ate . Fo r one thing, though w e m ayk n o w w h a t we mean when we say, in ordinary circumstances, that onething depends on another, i t is not at all clear what i t means to say,in Spinoza's system, that one thing depends on another. I have arguedthat the relation of mode to substance is a relation of causal dependencewhich is unlike the relation of predicate to subject and unlike the relationof species to genus. But I ha ve no t said wh at this relation is l ike. T othat extent , then, the def ini t ions of substan ce and m od e have beenlef t very vague.

Ag ai n , som eone migh t argue that, by de fining substanc e as w h at existsin i tself and is conceived through itself , Spinoza does not mean to saymerely that i t is independent of all external causes. He means also thatit is the cause of its own existence, or that its essence involves existence(E IDI ) . I t is not, or not merely, the negative notion of independentexistence, but also the positive notion of an existence which is logicallynecessary and consequently self-explanatory. It would be better to speakof a self-dependent or self-sufficient being . Fo r Spin oza, the con ceptof substance involves the ontological argument and is only as acceptableas that argument .

There is a good deal of force to this objection. However, i t is notquite accurate, as I see it, to say that the definition of substance initself involves the ontological argument. Spinoza is not defining substanceas a being whose essence involves existence. He is defining it as a beingthat h as no ex ternal cause. T o be sure, i t follow s from this definitionthat substance is also a being whose essence involves existence but that

is because Spinoza accepts the principle of sufficient reason in a particu-

40






larly strong form. Every existing thing has a reason or cause for itsexistence; if that reason or cause is not external to the thing, then itmu st lie in the natu re of the thing itself (E IP I ID2 ). Since substancecannot, by definition, have an external cause, it must be its own cause( E I P 7 D and E I P 6 C D 2 ) . H ence , the objection is not so muc h to theconcept of substance per se as to the strong form in wh ich Spino zaaccepts the principle of sufficient reason. This is a small point, but itis just as well to be accurate about such matters. Nevertheless, the fact

remains that Spinoza is committed to the ontological argument, andthis seems as good a reason for rejecting his metaphysic as any.

It is tempting to try to defend the ontological argument in Spinozaalong the follow ing lines. W e m ight argue that in Spinoza the func tionof the ontological argu men t is different than in such writers as Anselmand Descartes. Its purpose is not to prove the existence of a being whoseexistence is questionable, namely, God, but to justify the deificationof a being whose existence no one doubts, namely, Nature.3 1 If Naturecan b e said to hav e some, at least, of the properties traditionally ascribedto God, then the identification of Nature with God has some warrant.And of the properties traditionally ascribed to God, necessary existenceis one of the most important. Spinoza's ontological argument is a crucialstep in w ha t Frederick Pollock calls his eutha nasia of theo logy .

W e m ay well wo nde r whether anythin g is explained by saying thatNa ture is i ts ow n cause. Con tem pora ry philosophers are l ikely to con tendthat this is another case where a philosopher has. pushed a legitimate kindof inqu iry beyo nd th e limits of its me aning fulness. W e ca n ask of an yparticular, finite thing W h a t is the reason or cause of its existence?But we cannot ask this of the whole of Nature, of the totality of things.The question does not make sense, for there is nothing that could, logi-cally, count as an answer to it.

Th is, I supppose, is the probab le reaction that science and philosophywould nowadays have to Spinoza's principle that everything which existsmust have a cause or reason for its existence. And I suppose that acontemporary defender of Spinoza might reply on his behalf that saying

the question does not ma ke sense, cann ot logically m ake sense, w he n

41






asked of the whole of Nature is not very different from saying thatNature is i ts own cause. I t seems to be just another way of dissolvingthe sense of the mystical which attaches to the fact that the world is.

This defense is tempting, perhaps, but will i t do justice toSpin oza? I am n ot sure. W ha t t roubles m e is the assumption that w ha tis the cause of i ts own existence, for Spinoza, is the whole of Nature,tha t the term sub stan ce, wh ile i t connotes indep end ent existence,denotes the totality of things. This is, of course, an assumption that

is very of ten made.82 A n d there is some (tho ug h surprisingly l i t t le)textual support for i t . For example, in the Treatise on the Correction

the Intellect Spino za wri tes that

Since the origin of nature, as we shall see afterwards, can beconceived neither abstractly nor universally, cannot extend morewidely in the understanding than it does in fact, and has no similar-i ty to mutable things, no confusion ought to be feared concerning itsidea, provided that we have the s tandard of t ruth already men-tioned for this is, in fact, a being unique and infinite, that is,i t is all being, beyond which there is no being. (T d l E , 7 6 , I I : 2 g )

This seems definite enough. But in the same passage just one paragraphearlier, Spin oza refers to the source and origin of N at u re (/orai etorigo Naturae) as the pr im ary elements of the wh ole of na ture (pr imaelementa totius Naturae ) , which certainly makes the origin of natureappear to be something less than the totali ty of things. And in the Short

Treatise (KV I, vii i , 1 :4 7 ), Spino za speaks of dividing the whole ofNature into Natura Naturans (a bein g conc eived clearly an d distinctlythrough itself, without the need of anything else, that is, all the attributesof Go d, that is , substance or G o d ) and Natura Naturata (the mo desof substance) .

The weight of these passages seems to be in favor of saying thatsub stan ce denotes, not the wh ole of N atu re, bu t only its active part ,

i ts primary elements.83 And if that is correct, then before we can evaluate

the ontological argument, or the concept of substance, or any other

42






aspect of Spinoza's metaphysics, we must ask what in Nature mightanswer to this description. This is a serious gap in our interpretationof Spinoza.

Another gap occurs in connection with Spinoza's modes. Here thedenotation of the term is, in part at least, fairl y clear. M o d e denotes,am ong other things, particular finite things, individu al m inds and bodies,for example. But it also denotes certain things such as motion and rest,wh ich Spino za describes as an infinite and eternal mode of the attribute

of extension. And these infinite modes are rather mysterious entities.M oreo ver, as noted above, the connotation of the term m od e has

been left very va gue , in that the nature of its dependence on substancehas been specified only neg atively, by saying th at it is not the dependen ceof a predicate on its subject, and not the dependence of a species on itsgenus. Something mo re definite and positive must be said about that.

We have spent a long time covering rough ground, and it may wellseem that w e do not ha ve very m any positive results to show fo r our

effor t. I hav e argu ed, along the wa y, that for Spinoza substance is de-fined simply as that which is independent of external causes, that Spinozaidentifies substance a nd attribute, that he prob ably does not iden tifysubstance with the whole of Nature, and that on his view whateveris in itself must be conceived through itself (and vice versa). But mymain concern has been a negative one to show, not only that thereis serious disagreement am ong S pinoz a scholars even abo ut fund am ent almatters of interpretation, but also that two of the most prominent andplausible lines of interpretation rest on very shaky foundations.

These polemics were necessary, I think, to pave the way for whatis to come. In the remainder of this book I shall be striking out ina differen t direction, to propou nd a n interpretation w hich in someway s is quite radical. I shall not often look back ove r my shoulderto see wh at others ha ve done. A n d I shall not be able to prove th atmy w ay is the only wa y of reading Spinoz a. Bu t it does seem to methat any responsible critic must take very seriously the difficulties alterna-tive interpretations face and ask himself how he would deal differently

with the texts.

4





T o begin with, then, it seems to me th at the m ost imp ortantand interesting thing w hich philosophers have tried to do isno less tha n this, namely To give a general description of thewhole of the Universe, mentioning all the most important kindsof things which we know to be in it, considering how far itis likely tha t there a re in it imp ortan t kinds of things w hichwe do not absolutely know to be in it, and also consideringthe m ost im porta nt w ays in which these various kinds of things

are related to one another.— G. E. Moore, Some Main Problems of Philosophy WhatIs Philosophy?

The Causal i ty of God





First Thoughts and Difficulties

T h e problem of determining wh at Spinoza mean s wh en he says that G odis the cause of all things is as difficult as it is important. I will beginby stating certain things that are rather vague, and perhaps obvious,but I hope undeniable. I shall then construct a theory which I thinkfits these facts and explains a number of others as well. The result willbe an interpretat ion of Spinoza which, though nei ther obvious nor un-

deniable, is nonetheless fairly precise and supported by considerableevidence.

There is perhaps no catchword of Spinozistic crit icism more familiarthan the contention that Spinoza assimilates the relation of causalityto the relation of logical implication. And no doubt anything said sooften must have some justification. In this case, the justification is two-fold. First , as we have noted earlier, when Spinoza wants to characterizethe relation between substance and mode, beyond saying that modes

exist in and are conceived through substance, he says that God is thecause of the things that are in him, that he has produced them ordetermined them, or that they follow from him. This last form of expres-sion, which is the most definite, is strongly suggestive of the relationof logical implication.

Second, Spinoza often elaborates these statements by appealing toa geometr ical analogy. Thus, in the schol ium to P 7 he writes thatfrom the supreme power of God, or from his infinite nature, infinite

things in infinite ways, that is, all things, have necessarily flowed outor always follow by the same necessity, and in the same way as fromthe nature of a tr iangle i t follows, from eternity and to eternity, thati ts three angles equ al tw o right an gles II : 62 ) . H ere , as in man y otherplaces, Spinoza seems to be saying, as clearly as we could possibly wish,that the causal relation between God and the things that are in himis not merely a necessary relation, but a relation whose necessity is logical.

Statements of this sort have caused no end of trouble. Confrontedwith the claim that things proceed from G od as the properties of a

5






tr iangle proceed fro m its na tur e, Leibn iz com plained that ther e isno analogy between essences and existing thin gs. W h a t this suggestsis that the relation of logical implication is one which obtains only be-tween essences, that it makes no sense to speak of its obtaining betweenexisting things. And this is a cri t icism with which we may well sympa-thize, though we might now prefer to say that relations of logical implica-tion hold between propositions, or statements, rather than essences.

Moreover, as Spinoza insists in the passage quoted above, the relation

of logical implication is a timeless one. God's causality, therefore, mustalso be t imeless—or as Spin oza puts i t , G o d 's om nipotence has beenactual from eternity and will remain in the same actuality to eternity( E I p i 7s , I I : 62 ). T h is b eing the case, it is extremely d ifficult to seehow Spinoza could ever allow for the possibility of change. Indeed,this is one reason why some of his interpreters have been so bold asto assert that for Spinoza the phenomenal world, in which things comeinto being and pass away is an illusion.

Again, many of Spinoza's interpreters have felt that, if all things dofollow from God in this way, then it ought to be possible, in principle,to deduce the existence of particular finite beings from the existenceof the Infinite Being. Spinoza does not carry out any such deduction,which has been variously interpreted as a flaw in the system or as anatural defect of human knowledge. 1

These cri t icisms may or may not be well founded. I t would be rashat this point to say very much about a matter that has such far-reachingimplications. But this much may be said with certainty Spinoza con-ceives the causal relation between substance and its modes to be insome way analogous to the logical relation between ground andconsequent.

Now it may be noticed that I have spoken of this causal relationas a relation between substance and its modes, but that Spinoza oftenspeaks of it as a relation between God's nature or essence and the thingswhich follow from that nature or essence. Thus, in E IP I 6 infinite nu m -bers of things are said to follow fr o m the necessity of the divine n at ur e.

6





The Causality of God

In the demonstration of E IP33, Spinoza paraphrases this by sayingthat al l things have necessarily followed from the given nature of G od .W e a re given another paraphrase of the same proposition in the dem-onstration of E IP 17 when Spinoza writes that infinite numbers of thingsfollow fro m the laws of the divine natu re. In the passage quotedabove, things were said to follow from God 's infinite natu re or fromhis suprem e power —p hrases that are made explicitly equivalent byE I P 3 4 , where Spinoza declares that the pow er of God is his essence

itse lf. H e also speaks of things as following from the divine attributesE I P 2 9 S and from the absolute nature of the divine attributes E

IP2 ι ), though this last fo rm of expression seems to be applied onlyto the infinite modes and not to all things. All of these phrases, exceptthe last, appear to be used equivalently, with the result that we cansay indifferently of things that they follow from G od , or from his nature,or fro m his pow er, or from the laws of his nature, or from his attributes.

Th e next thing to be observed is that in many places Spinoza expresses

a very strong belief in the possibility of exp laining all events in termsof scientific laws. Writing O f M irac les in his Theological-PoliticalTreatise, he contends that every event in nature is determined by naturallaws. No thing happens in nature which would contradict its universallaws, nor does anything occur w hich does not agree with them or followfrom them TTΡ vi , I II : 83 ). By a law, here, he appears to havein mind a regularity in nature. Th e term 'law,' taken absolutely, signi-fies that according to which each individual, or all or some members ofthe same species, act in one and the same fixed and definite manne rTTPiv, 1 1 1 : 5 7 .

Because he thinks all events are determined by laws of this sort, Spi-noza denies that there can be such a thing as a miracle—if by m iracleis meant an event which cannot, in principle, be referred to some regu-larity in nature. Where an event cannot be referred to any known regu-larity, it will no doubt seem very strange and mysterious to us. An dif we are superstitious, we may say that it defies natural explanation.But this is both an intellectual mistake, since all things have natural

7






causes, and a mark of impiety rather than piety, for reasons we shallsoon discover.

First , however, we should point out, in view of the things that haveoften been said about Spinoza by his commentators , that when Spinozaspeaks of things as having natural causes, he does not appear to haveanything very esoteric in mind. Discussing the crossing of the Red Seaby the Jews escaping from Eg yp t , for exam ple, he suggests that a w aywa s opened for them by a n east win d wh ich blew very s t rongly all

n igh t , addu cing Exodus 14 :2 1 in ev idence. Aga in , in the case whe reGod is said to have changed the hearts of the Egyptians, so that theyhated the Israeli tes (P salm 1 0 5 :2 4 ) , Sp inoza argues that this too wa sa natural change, since we find in Exodus 1 that the Egyptians hadno slight reason for . . . red ucin g the Israelites to slave ry. Ex od us

tells us, what the Psalm does not, that the Egyptians feared the growingnum bers and po we r of the Jew s. T h a t Scrip ture refers such events di-rect ly to God, not a lways mentioning the at tendant c i rcumstances whichwould provide a natural explanation, results from the desire for edifica-tion. Scripture 's aim is on ly to na rrate w ha t seizes the po pu lar im agin a-tion, and to do so in that method and style which best serve to arousewonder and consequently to impress the minds of the masses with devo-t i on ( T T Ρ vi , I I I 90) . But though i t may not a lways be mentionedby Scripture, there is always a natural cause of the sort indicated.

Note that Spinoza speaks of the laws of nature in the same logicallanguage that he uses of God's power or essence. Nothing happens innature that does not follow from her universal laws. This suggests thatthere may be some very close connection between the laws of natureand God's power or essence—a suggest ion which is confirmed by passagesboth in the Theolog ical-Political Treatise and in the Ethics. In theform er work , Spin oza writes :

By the guidance of God I understand the fixed and immutableorder of nature, or concatenation of natural things ; for the universallaws of nature, according to which all things happen and are deter-

48






mined, are nothing but the eternili decrees of God, which alwaysinvolve eternal truth and necessity. Therefore, whether we say thatall things happen according to the laws of nature, or whether wesay that they are governed by the decree and guidance of God,we say the same thing. Because the power of all natural things isnothing but the power i tself of God, through which alone all thingshappen and are determined. TTP ii i, II I : 45—46)

Again, in the Ethics Spin oza writes that :

Nothing happens in nature which could be at t r ibuted to any defectin i t , for nature is always and everywhere one and the same. Itsvirtue and its power of acting are the same—that is , the laws andrules of nature, according to which all things happen and arechanged from one form to another, are a lways and everywherethe same. And therefore there must also be one and the same groundfor understanding the nature of things of any kind whatever,namely, the universal laws and rules of nature. (E III pref. , II : 138)

In these passages i t seems clear enough that Spinoza is, somehow, identify-ing the power or essence of God or Nature with the scientific laws thatgovern phenomena. This is why it is impious to believe in miracles.T o deny that things have a natu ral explan ation is to deny the pow erof God. Belief in miracles, paradoxically, is atheism. Conversely, so faras we achieve an understanding of how N ature wo rks according to

eternal laws, we increase our knowledge of God.Here, I contend, we have a key to understanding those passages in

w hic h S pinoza speaks of things as fol lowing fro m G od s nature . Fo rthings to follow from the necessity of the divine nature, for them toexist in and be conceived through the one substance, is for them tobe determined by and intell igible in terms of scientific laws. This much,I think , mu st be ad m itted on any interpretation of Spino za s m etaphysics.But i t may be that we can go a good deal further and say something

9






much more defini te about the relat ion between substance and i ts modes,something which will , if i t is correct, clarify a number of very obscurepassages in Spinoza's writings.

The odel etaphysicLet me begin by sketching a metaphysic of a type which will be familiarto all who have studied the British philosophy of the early years of

this century. Perhaps no one has ever advanced this part icular meta-physic, b u t ma ny p eople have held views abou t the natu re of the universewhich were, at least, expressed in these categories. I shall then explorethe possibility of understanding Spinoza's system in terms of this modelmetaphysic, although it is obviously described by him in a very differentway.

Suppose we have a set of propositions—call i t A-—which constitutesa co mplete an d accu rate description of the world of extend ed objects ,

a description that includes propositions stating whatever is, or has been,or will be the case with respect to such objects. It may be that a reallycomplete description is impossible, not merely because there are indefi-nitely many objects to describe, but because objects may be classified,and hence described, in indefinitely many ways. Nevertheless, we canat least say what logical kinds of proposition would have to be containedin any description of this world that professed to be complete.

T he re wo uld be s ingular proposit ions, like Th is table is three feetlong an d T he Ea rth revolves in an el liptical orbi t aro un d the Su n.These at t r ibute a property to an object or s tate that some relat ion holdsbetwe en two or mo re objects . A nd th ere wo uld be general proposit ions,like All plane ts ha ve elliptical orb its an d So m e bodies are thre e feetlong . The se at t r ibute a prop erty or a relat ion to all or some mem bersof some class of individual objects.

General proposi t ions are a mixed but interest ing lot . The fundamentaldistinction to be draw n a m on g them is not the distinction between uni-versal generalizations, wh ich are of the fo rm All F is G, and part icular

5 0






generalizations, which are of the form So m e F is G, but rather thedistinction between nomological and accidental generalizations.

The distinguishing characteristic of nomological propositions is theirstrict universality they are universal propositions, stating somethingabout all members of some class of objects, and that class is not definedwith reference to some particular time and place. For examp le, Eve rybody not acted on by an external force continues in its state of restor uniform motion in a straight line. Nom ological propositions are

commonly known as law s of na ture. Bu t this latter designation isambiguous and has misleading associations, so that it is perhaps bestavoided.

Accidental generalizations are all those which are not nomological,either because they are not universal, like Som e bodies are three feetlon g, or because, though un iversal, they are not strictly universal, inthat the class of which they are true is defined with reference to someparticular time and place. For exam ple, A ll of the objects in my room

on the ist of Ma y weighed less than 15 0 pou nds . Th e designationof such propositions as accidental is, as we shall see, also misleading;but it is well entrenched, there is no handy substitute, and there is,as we shall also see, some justice in the designation.

The defining characteristic of nomological propositions is their strictuniversality. But there is another characteristic which follows from this :that they are necessarily true, that they could not have been otherwise.Only necessary propositions are true without spatial or temporal limita-tion. Necessity and strict universality are indissolubly linked. So nomo-logical generalizations are, and accidental generalizations are not, neces-sary truths.

As a result, nomological generalizations do, though accidental gen-erlizations do not, support counterfactual inference. The principle ofinertia, given above as an example of a nomological generalization—if,indeed, it is a genuine law of nature, a true nom ological proposition—could not have been otherwise. Consequently, if my pen, which in factis now acted on by m y hand, were not acted on by any external force,

5






it would continue in a state of rest or uniform motion in a straightline.

By con trast, the (tr ue) generalization that A ll of the objects in m yroom on the ist of M ay w eighed less than 15 0 pou nd s could hav ebeen otherwise and does not support counterfactual claims of the kindthat laws do. I was not in fact in my room on the ist of May. Butsuppose I had been. It would then follow, not that I would have weighedless than 15 0 poun ds, but that not all of the objects in my room on

the ist of May weighed less than 150 pounds. Weight-losing, alas, isnot that easy.

For our present purposes, however, the most important characteristicof nomological propositions is that they are not reducible to conjunctionsor disjunctions of singular propositions. Accidental generalizations are.When I say that all of the objects in my room on the ist of Mayweighed less than 150 pounds, my statement may be viewed as equivalentto: T hi s object, which was in my room on the ist of M ay , weighed

less than 150 pounds, and that object, which was in my room on theist of M ay , weigh ed less than 1 5 0 pounds and . . . and so on, througha complete enumeration of the objects in my room on that day. OnceI ha ve said, of each of the objects in my room on that day, that itweig hed less than 15 0 pou nds, I do not add anything furth er by goingon to say that they all weigh ed less than 15 0 pound s.

With nomological propositions it is different. There the notion ofa complete enumeration is out of place. A strictly universal generalizationis, in principle, an assertion about an infinite number of individual ob-jects. It cannot be replaced without loss by any finite conjunction ofsingular statements. Hence, if we were giving a complete description ofthe world, we would need to include nomological generalizations asseparate items. But we would not need to include accidental generaliza-tions. They would be redundant.

Because there are general propositions in our description, the proposi-tions making up the set A are not logically independent of one another.Some follow logically from others, with the result that you could not

5






consistently deny them without denying some one of the propositionsfrom w hich they follow. Fo r exam ple, suppose it is true that

( ι ) Any body wh ich is near another body with a mass of5-975 X IC>24 kgm and a radius of 6. 37 1 X i o m and which startsfrom rest and falls freely toward the other body for one second falls adistance of 4.9 m.and that

( 2 ) Body a was near another body with a mass of 5.9 75 X 1 o24 kgmand a radius of 6. 37 1 X 1 oe m and it started from rest and fell freelytoward the other body for one second,and that

(3) Body a fell 4.9 m.

(3) follows logically from 1 ) and (2)—you could not consistentlydeny 3 ) w ithout denying either ( 1 ) or 2 ). But if there wereno general propositions in A and hence no propositions like ( 1 ), then

presumably 2 ) and 3 ) would be logically independent of one another.Either one could be consistently denied without denying the other.Not only is it the case that some propositions in A follow from others,

but every singular proposition in A will follow from other propositionsin A in the way that 3 ) follows from 1 ) and 2 ). It w ill followfrom a law, or set of laws, together with a statement of antecedentconditions. Among the nomological propositions, there is another kindof logical interdependence. They can be organized into a deductive sys-tem, such tha t, given a few of them as axioms, the rest can be derivedas theorems.

Th us , in our complete description of the world, the ideal of a unifiedscience is realized. Every true singular proposition about extended objectshas its scientific explanation, and every nomological proposition has itsplace in a system analogous to New ton s Principia.

The world, of course, mirrors our description of it. As the Philosophersays, the world is everything that is the case, it is the totality of facts,not of things. For it is in virtue of the existence of facts that propositions

5






are true. If the proposition tha t this table we ighs fifty po un ds istrue, then it is true because there is, in the world, the fact this propositiondescribes. If it is not true, that is because the fact this proposition pur-ports to describe does not exist. No doubt the table exists, too, but afterwe have mentioned the fact of which i t is a consti tuent i t would besuperfluous to mention the table as well .

For every distinct kind of proposition, there is a distinct kind of fact.As there are singular propositions, so there are singular facts. As, among

general propositions, the accidental ones are, and the nomological onesare not , equivalent to conjunctions or disjunctions of singular proposi-tions, so, among facts, those to which true accidental generalizationscorrespond are simply sets of singular facts, whereas those to whichtrue nomological generalizations correspond are general nomologicalfacts . And as, among nomological proposit ions, some are axioms andothers theorems, so some nomological facts are basic and othersderivative.

Needless to say, facts are not independent of one another. It is notthe case that any one could be different and everything else remainthe same. This is a consequence of the existence of general nomologicalfacts . They provide the causal nexus which just if ies the inference fromthe existence of one singular fact to the existence of another. The singularfact that body a fel l 4.9 m. is causally dependent on the singular factthat i t s tarted from rest and fel l freely for one second toward a bodyhaving such-and-such mass and radius and the general nomological factthat this is what bodies do under those conditions.

In general , wherever there is a relat ion of logical dependence amongpropositions, there is a relation of causal dependence among the cor-responding facts . Th u s, the world is thorough ly determinist ic. E ve rysingular fact depends on other singular facts and on nomological facts .And the derivative nomological facts , l ike that described by Gali leo 'slaw for freely fal l ing bodies, depend on the basic nomological facts .The only facts that are independent of al l others are these basic nomo-logical facts.

5







Spinoza s M etaphysics

structural property of the world —is not always strictly maintained,even am ong contemporary philosophers. Witness Arthu r L ove joy, w hospeaks in The Great Chain of Being of those generalized facts wh ichwe call emp irical law s (p. 14 8 ). But I think the distinction will turnout to be a useful one. Wh ere Sp inoza says tha t the attributes of G odexpress eternal and infinite essence (E ID6) we will read this as roughlyequivalent to his statement that the laws of nature acc ord ing to whichall things are changed from one form to another, are everywhere and

always the sam e (E II I pref., II : 138 ). T h e laws of nature, beingstrictly universal propositions, describe facts that are present throughoutall of space and time.

Part of the justification for talking about facts, rather than things,lies in Spinoza's doctrine that every idea involves an element of affirma-tion (E IIP49). If, for example, we have an idea of a triangle, thenthat idea involves the affirmation that the sum of its interior anglesis equal to 180 degrees. Thus Hampshire comments in his Spinoza that

T h e w ord ide a is used so widely in Spino za as to include w ha twe wo uld normally call an assertion or prop osition ; deliber-ately, and in opposition to Descartes, he makes no distinction be-tween hav ing an idea and asserting or m aking a statement.So an idea in his sense may be qualified as true or false, and oneidea may be said to follow logically from another; in normal usagewe speak only of propositions or assertions as true or false, or asfollowing logically from each other, (p. 66)

W heth er Spinoza's use of the term ide a is farther than Descartes' fromthe norm al use of that highly amb iguous term we need not say. Bu t

this does suggest that the correlate, in the attribute of extension, ofa Spinozistic idea, that with which a true idea must agree, its ideatum(Ε ΙΑ6) ought to be a fact involving an extended object, rather thanan extended object itself.

Suppose w e app ly this doctrine to the attribute of extension. W e havean idea of extension. That idea will involve certain affirmations about

extended things, for there are certain respects in which all bodies agree






(E I IL2, I I 98 ). T h e affirmations involved in our idea of extensionattribute these common properties to extended things and so describethe nature of extension. Such affirmations are general propositions ofthe form everything which has the attribute of extension has also com-mon property X What I am contending is that these propositions, in-cluded by Spinoza am ong the com m on notions, are for him the fund a-mental laws of nature, and that the facts they describe constitute thenature of the attribute of extension.

This conception of things has a fruitful application to the interpreta-tion of Spinoza's statements concerning our knowledge of God. If weregard the nature of the divine attribute of extension as made up ofa set of facts described by general propositions about all things havingthat attribute, then I think we are in a better position to understandhow Spinoza can claim so confidently that the human mind has anadequate knowledge of the eternal and infinite essence of God(E IIP47 ). This proposition must surely have shocked Spinoza's contem-poraries. Even today, S pinoza's readers ought to be taken aback byit. For viewed as a doctrine in the realm of natural theology, ratherthan as a thesis about our knowledge of the foundations of science,it is one of Spinoza's most radical departures from his predecessors.

But reflect. Spinoza tells us that each man has adequate knowledgeof the common notions (E IIP38-39). We are assuming that the factswhich the common notions describe constitute the nature of the attributeof extension. The attribute of extension is one of the attributes thatconstitute the essence of God. Hence, in knowing the common notions,man knows the nature of extension. In having adequate knowledge ofthe nature of extension, he has adequate knowledge of the essence ofGod.

This view will be confirmed, I think, if we look at the rather briefproof Sp ino za gives of this proposition :

E IIP47 : The human mind has an adequate knowledge of theeternal and infinite essence of God.

D em : T h e hum an mind possesses ideas (II P2 2) from which

7





Spinoza s Metaphy sics

(IIP23) it perceives itself and its own body (Ilpig), and externalbodies as actually existing. Therefore (IIP45-46) it has an adequateknowledge of the eternal and infinite essence of God. Q.E.D.

Anyone who is at all surprised by the proposition itself should beastounded by the casual way it is proved. Nevertheless, on this interpreta-tion it is intelligible enough. The key proposition is IIP45, accordingto wh ich eve ry idea of every bo dy , or of every singular thing actua lly

existing, necessarily involves the eternal and infinite essence of God.I think we can easily see why this is so. The common properties whichappear on an analysis of the nature of extension are called commonbecause they are present in every body. Hence, every idea of every bodywill involve them. That is, every idea of every body will involve theaffirmation that that body has comm on property X. A n d this affirmationis simply an instance of the general proposition that every extended thingis an X, which is one of the propositions describing the nature of the

attribute of extension.So far I have been attempting to justify taking the attribute of exten-sion to be a set of facts described by certain fundamental laws of natureknown as common notions. What of the infinite modes? I said abovethat they were also to be regarded as sets of nomological facts, thoughderivative ones. What evidence is there for this?

Let us take first Ε IP2 I Th ere Sp inoza says that all things wh ichfollow from the absolute nature of any attribute of God must alwaysexist and be infinite. T h e things which follo w are generally spokenof as the infinite modes. Later, in E Ip22 and E IP23, Spinoza dis-tinguishes between infinite modes which follow immediately from theabsolute nature of a divine attribute and those which follow mediately,through some other infinite mode. Hence we have two sorts of infinitemode : the immediate and the mediate. Spinoza does not give us, inth e Ethics, any examples of infinite modes, nor does he tell us howmany there are supposed to be under any given attribute. But he doesgive examples in the Short Treatise and in one of the letters, and it






is usually thought that there will be one immediate infinite mode andone mediate infinite mode un der each attribute.

One way of approaching Spinoza's doctrine of infinite modes is toask in w ha t sense they m ight be thoug ht to fo llo w from their respectiveattributes. Their relation to the attributes cannot be one of temporalsuccession because both terms of the relation are eternal. Nor can it,on the face of things, be a relation of logical consequence, for thatis a relation which holds between propositions, and not between things

(or facts). Nevertheless, the relation of logical consequence, being onewhich holds timelessly, is clearly closer to what Spinoza has in mindthan the relation of temporal successsion. Let us say tentatively thatit is a relation of causal dependence between general facts, which isthe counterpart in the order of things (more accurately, the order offacts) of a relation of logical consequence in the order of ideas (moreaccurately, the order of propositions). On this interpretation, Spinoza'sthesis that every infinite and eternal mode of the attribute of extensionfollows either directly from the absolute nature of the attribute of exten-sion or indirectly from some other infinite mode w hich follows fromthe nature of extension (E IP23)—put in logical terms—amounts tothe thesis that every scientific law relating to extended objects can bederived either directly from the fundamental laws governing extendedobjects or from a finite series of nomological propositions which termi-nates ultimately in the fundamental laws.

Spinoza does not give us much information about what the infinitemodes are, but the information we are given provides an interestingconfirmation of the claim made here. We want to say that all otherscientific laws follow logically from some set of propositions stating prop-erties common to all bodies. Now in Ep 64, when Spinoza is askedto give an example of something that is produced immediately by God,he mentions, under the attribute of extension, motion-and-rest. As aresult, motion-and-rest has acquired the title of immediate infinite andeternal mode under the attribute of extension, which is to say that itis identified with w ha t follows imm ediately from the absolute nature

9










Of the general modes, Spinoza says that we know only two mot ionin matter and the understanding in the thinking thing. The face of thew ho le universe is not nam ed here. A n d th ou gh m otion is said to beboth infinite and eternal, these characteristics are put forward as oneswhich it is not Spinoza's business as a philosopher to exp lain :

As for what particularly concerns motion, since i t belongs moreproperly in a treatise on natural science, rather than here, to show

that i t has been from all eternity and will remain immutable forall eternity, that it is infinite in its kind, that through itself it canneither be nor be understood, but only by means of extension, allof these things I shall not treat here. I shall merely say this aboutit , that i t is a Son, product, or effect, created immediately by God.(K V I , ix , I :4 8 )

In a footnote Spinoza explains that what is said here about motion

is no t said seriously. Fo r the A u th o r still intend s to disco ver the ca useof i t , as he has already done to some extent a posteriori. This suggeststhat at the t ime of writing the Short Treatise Spinoza nevertheless hopedto give an a priori foundation to those properties of motion whose ex-plan ation is said to belo ng mo re properly to a treatise on natur al sci-ence . W hen he came to wr i te the Ethics, he evidently felt that hehad succeeded in doing this, for the properties asserted without proofin the Short Treatise are there proved more geometrico and m ade thebasis for a new classification.

If my interpretation is correct, Spinoza has roughly the same justifica-tion for call ing these general modes infinite and eternal that he hasfor describing the attributes in this way. The general modes are factsdescribed by general, nomological propositions—necessary propositionswhich are a lways and everywhere val id . Their inf ini tude and eterni tyis a function of their strict universali ty. Thus we have the picture pre-sented on page 63.

W ith this picture as back grou nd, I think we m ay m ake some hea dw ay

on the problem of the finite modes. In E IP26 Spinoza says that any-

62






to ta l i tyoffacts

gene ra l = nomo-fac ts logi -

ca lfac ts

bas ic

de r iva -tive,b u tp r i m a r y

de r iva -tivea n ds e c o n d a r y

a t t r ibu te = na tu ran a t u r a n s

i m m e d i a t einfini tem o d e

= med ia t einfini tem o d e

gene ra l— n a t u r a

n a t u r a t a

s ingularfacts

finitemodes

= pa r t i cu la rn a t u r an a t u r a t a

thing whic h has been determined to any action has necessarily beenso determined by G od . Bu t in E IP 28, he says that

any singular thing, or any thing which is finite and has determinateexistence, can neither exist nor be determined to action un-less . . . by another cause w hich is also finite and has a determinateexistence; and again, this cause can neither exist nor be determinedto action unless . . . by another cause w hic h is also finite anddetermined to existence and action, and so on, ad infinitum

Now this is difficult. God is not something finite. How is it that finitethings are conditioned to act both by God and by something finite?From the proof of E IP28, the answer appears to be that finite things

are conditioned to exist and act by God only insofar as he is modified






by som e m od ifica tion w hi ch is f inite. T h a t is, the som eth ing finitement ioned in E Ip 28 is a modif icat ion of God, and in being condi t ionedto act by something finite, a f inite mode is conditioned to act by God.

But this will not do. In E IP33, when Spinoza a t tempts to provetha t th ings cou ld have been produc ed by Go d in no o ther ma nnerand in no o ther o rder than tha t in which they have been produced ,he argues that , i f th ings had happened at a l l d i fferent ly, God 's naturewould have had to be different , which is impossible . The order of events

in nature must depend, in some way, on God's essence, that i s , on theattr ibutes, and not simply on God as modified in finite modes. So weare back where we s tar ted.

This di ff icul ty was wel l put by Leibniz . Not ing that , according toSpinoza , fin ite and tempora l th ings cann ot be p rodu ced imm edia te lyby an infinite cause (Ε Ip2 i ) , and tha t they are pro du ced b y othercauses, individual and finite (E IP28 , he asks:

How wil l they f inal ly then spr ing f rom God? For they cannotcome from him mediate ly, in th is case , s ince we could never reach inthis way things which are not similarly produced by another f initething. I t cannot , therefore , be sa id that God acts by mediat ingsecondary causes, unless he produces secondary causes.3

W h a t Sp ino za's philoso phy seems to requ ire, fo r finite things, is bo than infinite series of finite causes and a finite series of infinite causesterminating in God. I t is difficult to see how these requirements are

to be jointly satisfied.I th ink that we have the mater ia ls a t hand for an answer to th is .

Let me cast my approach in the form of a l ine by l ine commentary onthe proof of E Ip28. I shall quote the individual sentences of the proofalternately with the translation into the logical form of my interpretation :

ι . W ha tev er is determ ined toexistence and action is thusde te rmined by God (E IP26 andI P 2 4 C .

ι a . A n y proposi t ion in A (exc eptthose which are axioms of scientif ictheory) must fol low from other

proposit ions in A.

64






2. But what is finite and has adeterminate existence could not beproduced by the absolute nature ofany at t r ibute of God, for whateverfollows from the absolute nature ofany attribute of God is infinite ande t e r n a l ( E I P 2 I ) .

3. The finite and determinatemust therefore fol low from God,or from some attribute of God,insofar as the latter is consideredto be affected by some mode, forbesides substance and modes,nothing exists E IAI, ID3, ID5,and IP25C) .

4. But the finite and determinatealso could not follow from God orfrom any one of his attributes, sofar as that attribute is affected bya modif icat ion which is e ternal andinf ini te (E IP22) .

5. I t must, therefore, follow or bedetermined to existence and actionby God, or by some attribute ofGod, insofar as the attribute ismodif ied by a modif icat ion whichis finite and has a determinateexistence.

2a. Bu t a singular propositioncanno t be dedu ced (solely) f ro mthe axioms of the scientific theoryof A for these are generalpropositions, and from generalpropositions alone only generalpropositions can be deduced.

3 a. A singular proposition m ust

then follow from propositionswhich are not axioms of thescientific theory of A that is, fromsome theorem of the scientifictheory of A or from some singularproposition, for these are all thepropositions there are in A.

4a. Bu t a singular prop ositioncann ot be deduced (solely) f ro mtheorems of the scientific theory ofA for these too are generalpropositions.

5 a. T h e singular propositions ofA must then follow from othersingular propositions of A (take ntogether with some generalpropositions, for it is no morepossible to deduce a singularconclusion from singular premisesalone, than it is to deduce onefrom general premises a lone) .

N o w , as the paren thetical ad ditions show , the fit betw een Spin oza s

words and my interpretation is not a perfect one. Nevertheless, the fit






is close enough to be interesting, and if we read Spinoza as I thinkwe shou ld, the d ifference s will tu rn out to be n egligible.

W ha t I propose is tha t we unders tand Spinoza as ma inta in ing tha tfinite things de pe nd up on G od bo th ins ofar as he is m odifie d by finitemodifications and insofar as he is modified by infinite modifications.T o use a dis tinction that S pinoza introduces later (E I I I D I ) , nei therthe infinite m odes no r the finite modes a re by themselves ade qu ate causesof finite modes. Taken separately, they are only partial causes; the ex-

istence and actions of a particular finite mode cannot be understoodeither by reference to other finite modes alone or by reference to infinitemodes alone, but only by reference to both infinite and finite modes.

P ut in the lang uag e which I introdu ced earlier, this wo uld be to saytha t the s ingular facts wh ich exist a t any given mo m en t are determ inedby the previously existing singular facts and by certain general facts butthat neither the previously existing singular facts nor the general factsalone suffice to de term ine w ha t facts now exist. T h e previously existing

sing ular fa cts give us th e infin ite series of finite causes. T h e gen era lfac ts give us the finite series of infinite causes, term ina ting in G od . T hi sis the m etaph ysical equ ivalen t of the logical or epistemologica claimthat propositions describing the existence and actions of particular thingscan be deduced from the laws of nature, if and only if the laws aretaken toge ther with a s tatement of anteceden t condit ions.

Spinoza never says, in so many words, that neither the infinite modesnor the finite modes are, by themselves, adequate causes of finite modes.Bu t he does say a nu m ber of things which pers uade m e tha t this washis view. Consider first the passage in the Treatise on the Correctionof the Intellect in w hich Sp inoza discusses the relation betw een the fixedand eternal things, on the one hand, and the part icular and mutablethings, on the other. The re Spinoza says tha t

Ab ove all i t is necessary for us always to ded uce all ou r ideas fro mphysical things—that is, from real entities, proceeding, as far aswe may, according to the series of causes, from one real entity






to another real entity . . . by the series of causes and real entities,I understand here not the series of singular mutable things, butonly the series of fixed and eternal things. It would be impossiblefor human weakness to grasp the series of singular mutablethings . . . nor is there any real need for us to understand th eirseries, since the essences of singular mutable things are not to bedr aw n from their series or order of existence . . . the essence isreally sought solely from the fixed and eternal things, and from

the laws, inscribed in those things as in their true codes, accordingto wh ich all singular things occur an d are ordered indeed, thesesingular mutable things depend so intimately . . . on the fixedthings that they can neither be nor be conceived without them.( TdlE, 9 9 - 1 0 1 , 11 : 3 6 - 3 7 )

This is an interesting and difficult passage. The singular mutable thingsare naturally enough identified with the finite modes of the Ethics; but

there has been considerable speculation as to what the fixed and eternalthings might be.

It seems to me that Pollock is basically right in identifying themwith the infinite modes of the Ethics (thoug h I suspect that we shouldadd the attributes as well). Note that the fixed and eternal things aredescribed, in a consciously m etap horica l phrase, as ha vin g laws in-scribed in them . The se laws appe ar to be the laws of nature. W e ha ve,then, an apparently finite series of eternal entities, somehow associatedwith the laws of nature, on which the infinite series of particular thingsdepends very intim ately.

If we follow Pollock in identifying these things with the infinite modes,and the interpretation I have been proposing in identifying the infinitemodes with the facts described by derivative nomological propositions,then we have a fairly clear answer to the questions of what these eternalentities are (they are nomological facts), and how they are associatedwith the laws of nature (they are described by them), and how particularthings depend on them (for these nomological facts, in conjunction with

67






o the r s ingu la r f ac t s , de t e rmine wha t s ingu la r f ac t s ex i s t a t any g iven

m o m e n t ) . T h e p o i n t is n o t t h a t S p i n o z a t h o u g h t e x p l ic i t l y i n t h e s e c a t e -

go r i es . H e d id no t . T h e po in t i s t ha t t he con cep t o f a no m olo g ica l f a c t

is t he k ind o f con cep t t ha t S p in oza ' s sys t em seems to r equ i r e , t he k in d

of concep t t ha t he seems to be g rop ing fo r o r t ry ing to expres s .

Po l lo ck ' s c om m en t i s i n t e re s t ing in t h i s con nec t ion . H e a sks :

W h a t can these e t e rna l t h ings be? T h e in t e rp re t a t ion tha t l ie s

nea res t a t h an d fo r a m od er n r eade r i s t o i den t i fy th em wi th the con-s t a n t r e l a t i o n s a m o n g p h e n o m e n a w h i c h w e n o w c a l l l a w s o f n a t u r e .Bu t t h i s is ev iden t ly no t admiss ib l e . Sp inoz a , t he dow n- r ig h t en em yof abs t r ac t ions and un ive r sa l e , kn ew the d i f f e r enc e be tw een r e l a t ionsand th ings f a r t oo we l l t o confuse them in th i s way. Bes ides , hew a n t e d n o a r t i f i c i a l w a y o f d e s c r i b i n g l a w s o f n a t u r e t h e n a m ewas a l r eady f ami l i a r i n h i s t ime , and he cou ld speak o f t hem, whenhe tho u gh t fit , jus t as w e do . In fac t , he does spe ak of the e t e r na l

t h i n g s a s h a v i n g l a w s o f t h e i r o w n in s o m e w a y a t t a c h e d t o o ri n v o l v e d i n t h e m , w h i c h p e r v a d e t h e w h o l e w o r l d o f p h e n o m e n a .C lea r ly, t he re fo re , t he th ings in ques t ion a re no t t hemse lves l aws .( P o l lo c k , p p . 1 5 0 - 1 5 1 )

H e then goes on to a rgu e tha t t hey a re t he in f in i t e m ode s . N o w , t ha tthe e t e rna l t h ings a re no t t he l aws themse lves , I ag ree . Po l lock , however,p rov ides no r eason fo r t h ink ing tha t t hey a re no t t hose gene ra l f ea tu re so f r e a l i t y w h i c h t h e l a w s d e s c r i b e . W h i l e S p i n o z a h a s a n a t u r a l w a yof r e fe r r ing to the l aws , he does no t have a na tu ra l way o f r e f e r r ingto wha t t he l aws desc r ibe .

B u t — t o r e t u r n t o o u r p r o b l e m — w h a t is o f s p e c i a l i n te r e st h e r e i sthe dep end enc e o f t he s ingu la r m u t ab le t h ings on the fixed and e t e rna lth ings . Th i s s eems to mean tha t t he f in i t e modes a re de t e rmined toexis t and ac t so le ly by the f in i te ser ies of inf in i te modes , and to cont radic tE IP28 o f t he Ethics, w h i ch seems to say tha t fin ite m od es a re de t e rm inedto exist and act solely by an infini te series of other f ini te modes.

8






My view, however, is that these two passages are complementary,rather than contradictory. Neither the finite modes nor the infinitemodes, by themselves, are adequate to determine the existence and actionsof a particular finite mode. The difference between the Ethics and theTreatise is merely one of emphasis. In the Ethics Spinoza is emphasizingthe dependence of finite modes on other finite modes, whereas in theTreatise he is emp hasizing their dependence on the infinite modes. B utin any causal explanation of one sort, there is an implicit reference

to a causal explanation of the other sort. Whenever we explain a particu-lar event by mentioning a law of nature under which it may be sub-sumed, we are implying the prior existence of appropriate antecedentconditions. A nd conversely, wh enever w e give as the explanation ofsome event its antecedent conditions, we are implying the existence ofa law linking those conditions with that event. This, I think, may explainthe otherwise rather puzzling remark that occurs in Ep. 40 :

For example, if someone asks by what caíase a body determined ina certain way is moved, we may reply that it is determined tosuch a motion by another body, and this one again by another,and so on to infinity. This, I say, is a possible reply because thequestion is only about motion, and by continually positing anotherbody, we assign a sufficient and eternal cause of its motion.( I V: 1 9 8 )

It is difficult to see how Spinoza could regard such a series of finitecauses— even tho ugh it extends to infinity— as an eternal cause of sucha motion. Neither the series nor its individual members are eternalentities in Spinoza's usage of that term, for according to E ID8, byetern ity Spinoza understands

existence itself, so far as it is conceived to follow necessarily from thedefinition alone of the eternal thing.

69






Explanation For such existence, like the essence of the thing, isconceived as an eternal truth. It cannot, therefore, be explained byduration or time, even if the duration be conceived without begin-ning or end.

Since the essences of these finite things do not involve existence, thethings are not eternal. A n d jud ging fro m the final remark in the explan a-tion of the d efinition, the fa ct that their series extends to infinity wou ld

be irrelevant. The only thing I can see that could be regarded as aneternal cause here would be the law governing the series—presumablythat would be the principle of inertia— or rather, the nomological factdescribed by that law.

To say that the infinite and finite modes are separately necessaryand only jointly sufficient conditions of finite modes does, I think, giveus a plausible way of reconciling what looks like an outright contradictionin Spinoza s metaphysics. This interpretation is supported by other things

which Spinoza says. Let us begin by considering the note he appendsto E Ip 28:

Since certain things must have been produced by Go d imm ediately,that is, those which necessarily follow from his absolute nature,these primary products being the mediating cause for those [N S

other] things which, nevertheless, without God can neither be norbe conceived, it follows first, that of the things immed iately p ro-duced by him, God is the proximate cause absolutely. [N S I saythe proximate cause absolutely and] not in their own kind, as theysay ; fo r the effects of G od can neither be nor be conceived w ithouttheir cause. (E IP15 and IP 4.C)

It follows, second, that God cannot properly be called the remotecause of singular things, unless to distinguish them f ro m the thingswhich he has immediately produced, or rather which follow fromhis absolute nature. Fo r by a remote cause we understand one whichis in no way conjoined to its effect. But all things which are, are

70






in God, and so depend on him that without him they can neitherbe nor be conceived. [My italics]

This is a passage concerning which there may be some difference ofopinion, but whose general drift seems clear enough. God is assertedto be the proxim ate cause of the things he has produced imm ediately.It is denied that he is the remote cause of singular things, in the peculiarsense wh ere a remote cause wo uld in no way be conjoined to its effe ct.

Bu t it is implied that G od is the remote cause of individu al things ina m ore natural sense, nam ely, that h e is their cause through th e m edia-tion of the things which he has produced immediately.

At first glance, we might suppose that the things which God has imme-diately produced would be only the immediate infinite modes. Thatis what the phrase, taken by itself, naturally suggests, and some inter-preters have th ough t th at they are obviously referred to. But Sp inoza alsoidentifies the things in question as those wh ich follow fr om the absolutenature of G od , a description whic h clearly has a broader denotation,since it is used by Spinoza to refer to all the infinite modes and not merelyto those wh ich follow im mediately from his absolute nature (see E IP23and its dem.). Moreover, it appears from the clause italicized that Spi-noza regards the first description as likely to mislead. So I think wemay take the things immediately produced by God to be all the infinitemodes.

O n this reading, Spinoza's doctrine is that ( 1 ) God is the proximatecause of the infinite modes, 2 ) G od is the remote cause of the finitemodes, in the sense that he is their cause through the mediation ofthe infinite modes. Th is w ill agree with the Short Treatise accordingto which

God is the proximate cause of the things that are infinite and im-mutable, and which we say have been created immediately by him,bu t in a sense, he is the remote cause of all particular things. (K VI, iii , 1:36)






The genera l [natura naturata ] consists of all the modes whichdepend imm ediately on G od . . . the par t icular [natura naturata]consists of all the particular things which are produced by the generalmode . ( K V I , v i ii , 1 :4 7 )

T h e Short Treatise, i t should be noted, knows no distinction betweenimmediate and mediate inf ini te modes.

This confirms my contention that the infinite modes are at least a

partial cause of the finite modes—and also, incidentally, strengthens theidentification of the infinite modes with the fixed and eternal thingsof the Treatise on the Correction of the Intellect, which ,

thoug h they are s ingular . . . on accoun t of their presence andpower everywhere, will be to us l ike universale, or the genera of defi-nit ions of singular mutable things. They will be the proximate causesof all things. (TdlE, 101,11:37)

But the infinite modes cannot be an adequate cause of finite modes.For if they were, the finite modes, l ike the infinite modes, would followfrom the absolute nature of God, and we know that nothing finite canfollow from the absolute nature of God (E IP2 I ).

In the Short Treatise Spinoza expresses this view more clearly thanin any other place. There he distinguishes between a principal cause,wh ich produ ces an effect by vir tue of i ts ow n powers a lone, withou tthe aid of an yth ing else an d a subsidiary cause causa minus princi-palis), w hi ch is m erely one cond ition or fac tor w hi ch is necessary butnot adequate to produce a cer ta in effect . 4 God, Spinoza says,

is a principal cause of his works which he has created immediately,such as motion in matter, in which a subsidiary cause can haveno place, since i t is always confined to particular things, as whenhe dries the sea through a strong wind, and similarly with all par-ticular things in nature. {K V I , i ii , 1 : 3 5 -3 6 )

72






That is, God is the principal (or in the terminology of the Ethics ade-quate) cause of the infinite modes. But of the finite modes, he is asubsidiary (or partial) cause—at least , in so far as he is infinite. Inthe realm of the finite, other finite modes are also required as subsidiarycauses.

W e m ay tak e this, then, as our answ er to Leibn iz ' question H o wdo things f inal ly spr ing from G o d ? W e wil l not get bac k to G odby following up the infinite chain of finite causes. Even if we could

do that, i t would not lead us in the right direction. But we will getback to God by following the equally necessary finite chain of infinitecauses.

What is perhaps of more interest is that in recognizing that Godis not, in so far as he is infinite, the adequate cause of finite things,we have also an answer to that very bothersome puzzle about the relationof God's causal i ty to t ime and change, an answer which we can le tSpinoza give in his own words. In the second dialogue of the Short

Treatise Spinoza has one of the characters, Erasmus, raise an objectionto the author 's spokesman, Theophilus, of a sort which almost makesone think that Spinoza had foreknowledge of what historians of philoso-ph y wo uld later write abou t him . Erasmu s speaks :

Y o u said . . . tha t the effec t of an inner cause can not perish solong as its cause lasts. This, indeed, I see to be certainly true. But ifthis is so, then how can God be an inner cause of all things, sincemany things perish? (K V I, i i , 1 : 3 3 )

But Erasmus is more alert than most historians of philosophy, and heis able to answer this one himself :

But you will say, according to your previous distinction [betweenprincipal and subsidiary causes], that God is properly a cause of theeffects w hic h he has prod uce d imm ediately by his attributes alon e,without any further circumstances, and that these cannot cease

7







The ausality of God

of ca usal depende nce used here mus t be clarified before we can be saidto hav e a n interp retation of Spinoza's langu age w hich is at a ll definite.U ntil w e hav e some analysis of Go d's causality, the connotations ofthe terms substance and m od e are very uncertain.

Toward the end of the chapter, moreover, I suggested that theremig ht be some difficulty abou t the denotations of these terms. M od edenotes, in pa rt a t an y rate, such fam iliar individual things as me, you,the desk in my office and so forth . But its deno tation also includes

some rather more obscure things, such as motion and rest, and the faceof the whole universe. What these things are, and how they might berelated to substance in a way like the way you a nd I a nd th e deskin my office are, are puzzles of th e first ord er.

Ag ain, the deno tation of the very term subs tance itself is uncle ar.It is comm only taken to deno te the whole of N atu re, th e totality ofthings. But common as this assumption is, there is little in Spinoza'swork to justify it, and a good deal to support th e contrary assertionth at subs tance denotes, not the whole of N atu re, but only its activepart— those prima ry elements of the whole of N atu re which Spinozadescribes in the Treatise on the orrection of the Understanding as

the source and origin of natu re. W ha t, I asked, might these primaryelements be?

These deficiencies can now be m ade good. T he term m ode is tobe un derstood as denoting certain sorts of fac t, not individua l things.Finite modes ar e singular facts. T he infinite modes a re derivative, nomo·logical, general facts, those to which the lower level generalizations ina unified science would correspond. A nd th e prim ary elements of thewhole of Nature, which constitute substance or its attributes, are thatset of basic, nomological general facts to which would correspond thehighest level generalizations in such a science.

The relation of causal dependence that modes bear to substance isthe ontological c oun terpa rt of the relation of logical dependence wh ichwould obtain between the corresponding propositions in a complete andaccu rate description of the world. Substance is causally inde pend ent of

all othe r entities beca use it is th at set of facts to w hich the axioms of

75






our unified science correspond. These propositions are not the logicalconsequence of any others. All other propositions are, in one way oranother, logical consequences of them. Th e derivative nom ologicalpropositions fo llow logically fr om the axiom s, without any other proposi-tion being necessary fo r their derivation. Th e fac ts to which they cor-respond, the infinite modes, thus depend on the absolute nature of theattributes of substance. Th e singular propositions follow fro m the nomo-logical ones—ultimately, that is, from the axioms—but only with the

aid of other singular propositions. Th e fac ts to which they correspond,the finite modes, thus depend not on the absolute nature of the attributesalone, but also on other finite modes.

The notion of causal dependence which I am employing here is boundto be somewhat unfam iliar. Let me try, briefly, to put it in historicalcontext. In the Tractatus, Wittgenstein denied that there are any causalrelations. Atomic facts are independent of one another (2.061 ) ; fromthe existence or nonexistence of one atomic fact we cannot infer the

existence or nonexistence of another (2.062, cf. 5.134, 5-135)· Butatomic facts are all the facts there are cf. 2 ). As a result, any onefact can either be the case or not be the case, and everything else remainsthe same 1. 2 1 ) . Th ere is no causal nexus which wou ld justify an infer-ence from the existence of one fact to the existence of another (cf.5.136). Belief in a causal nexus is superstition 5.1361 ). A necessityfo r one thing to happen because another has happened does not exist;there is only logical necessity (6 .3 7).

To interpret Wittgenstein on any matter is perilous, but I take itthat in such aphorisms as these we have a genuine denial of causality.The reason fo r this denial appears to be that Wittgenstein makes twoassumptions : ( 1 ) it is a necessary condition of one fact s causing anotherthat the proposition describing the cause entail the proposition describingthe effect; 2 ) all the propositions in a complete description of theworld would be logically independent of one another, for the worldis the totality of atomic facts. If the second assumption is true, thenthe necessary condition of causality can never be satisfied.

76






T h e similarity of this view to H um e's is striking and W ittgensteinclearly has Hume in mind. But I think it would be wrong to say thatHume rejected causality. Rather he presents a reductive analysis ofcausality, definin g ca us e in such a w ay that i t is a necessary con ditionof one thing's causing another only that the cause precede the effectin t ime and that the cause be the sort of thing which is regularly followedby things of the kind to wh ich the effec t belongs. If I say ( to ad ap tMark Twain 's example) that the man's decis ion to carry the cat home

by the tail caused him some discomfort, I mean that the decision wasfollowed by discomfort and that decisions of this sort generally are. Humerejects Wittgenstein's first assumption, though not his second. As a result,even though he denies necessary connections,5 he does not deny causality.Wit tgenstein seems at any rate, to want to take this further step.

It is a curious fact of contemporary history that most philosophersin our century have sided with Russell in rejecting Wittgenstein's attemptto reduce general propositions to truth-functions of singular ones, but

have tended to accept , with some misgivings, Hume's reduct ion ofcausality to regular sequence. Curious, because unnecessary. For if wehave both singular facts and general facts at our disposal, we can locateam on g the furnitu re of the universe entit ies wh ose existence will en ta ilthe existence of other entities.6

The conception of causality which I attribute to Spinoza is unusualat least in the sense that i t is not a reductive Humean one. Spinozaaccepts Wittgenstein's first assumption, though not his second. Spinoza'sconception of causality is also more general in that it does not restrictthe causal relation to events (or rather singular facts). We may alsospeak of causal relationships holding between general facts where thenotion of a temporal sequence has no application. And this seems tome quite in order. For surely one of the most important and interestingfeatures of science is that it seeks explanations, not only for individualfacts and sequences of facts, but also for the regularit ies which thoseindividual facts and sequences of facts exemplify. This is one reasonfor preferring an ontology of facts to one of events.






bjections and Replies

Such, in outline, is the inte rpreta tion of the concepts of substance andmode in Spinoza which I am proposing. It ha s the advantage of beinga fairly definite interpretation. It has, also, the advantage of presentingSpinoza's metaphysic as one which, in my judgm ent, is coherent andplausible. And it has the advan tage, as I have been a t pains to pointout along the way, of rendering intelligible many things which are other-

wise very obscure. But as any interpre tation must, it has its disadvantagesas well, and it is now time to take some accoun t of the objections thatmight be raised against it.

O ne difficulty is that, although this way of taking Spinoza may makesense of a great many passages in his work, it will not deal equallywell with all of them. For example, when Spinoza is arguing that sub-stance, as such , is indivisible (Ε ρι2, IP13, IP15S), he appears to bethinking of substance in a way fa r more C artesian th an I have been

willing to allow. And though there may not be anything in the passagesmentioned which is flatly inconsistent with m y interpreta tion, it is noteasy to see how they could be treated in m y terms.

Th is is true, and in reply I can only point out tha t the adequacyof any interpretation must always be a matter of degree. I feel thatmy way of reading Spinoza does greater justice to his system as a wholethan any other I have seen. But admittedly it falls short of the ideal,which would be to translate each of Spinoza's metaphysical doctrinesinto its terms.

Another objection, which must inevitably arise, is that I have pro-vided, not an interpretation, bu t an anachronistic reinterpretation inthe light of subsequent thought, of the sort Iris Murdoch was complain-ing of when she wrote there has been of late something of a tendencyto read back into the great metaphysicians our own logical formulae,and to treat them as if they were trying ineptly to do what we havedone successfully.7 Just as Joachim saw Spinoza through lenses groundby Hegel and Bradley, so I have seen him through those ground by

Russell, Moore, and Wittgenstein.

8






Th is is a charg e to which I m ust concede substant ial me ri t . N o do ubtmy prefere nce for t ranslat ing Spinoza s langu age into the langua ge oflogical atomism is due, in part at least, to my philosophical upbringing.But aside fro m point ing ou t how str ikingly mu ch of Spinoza c an survivethe t rans la t ion—and I am not through yet—there are a number ofthings to be said in my defense.

First , the seven teenth centu ry is no t, afte r all , such a very rem ote ances-tor of our ow n century . In m an y ways i t set the problem s that hav e

dominated philosophy—or at least English-language philosophy—sincethen : problems about our knowledge of the external world, about therelat ions of mind and body, about the nature of space and t ime andso on. In part icu lar, i t gave promin ence to the idea tha t laws playan essential role in explanation. Even if this did not occur in Spinoza,we should have to recognize it in Descartes cf. t h e P rinciples par tI I ) . M oreov er, we ou ght n ot to unde rest im ate the intel ligence of thesemen by supposing that , because they lacked the means we have forexpressing ourselves, they could not see what we see. I have attributedto Spinoz a the insight th at laws alone a re not suff icient for the exp lana-tion of any particular fact. It does not strike me as implausible thatan intelligent man, well-acquainted with the science of his t ime,8 shouldsee this. A nd I take com fort in the fac t th at some interp reters of Des-cartes have found it necessary to credit him with precisely the sameinsight.9

Ag ain, to some extent w e can divorce w ha t I say fr om how I sayi t. T h e conclusion th at s ingular facts do not de pen d on nomologicalfacts alone can be t ranslated back into Spinoza s langu age by sayingth at Go d is no t, insofar as he is infinite, th e ad eq ua te cause of finitemodes, or that what is finite and has a determinate existence does notfol low fro m the absolute nature of any at t r ibute of God. This conclusionis an important one. Though i t fol lows from my interpretat ion of thedefinitions of substance and mode, i t is also supported by considerabletextual evidence. It is a conclusion which has been anticipated, butnot a rgu ed for in detai l, by other wri ters on Spinoza, who h ave had

backgro unds qui te d i fferent f rom m ine .1 0

And it should, I think, survive,

9







The ausality of God

embodying a perfection which no finite object can approach. The Godof the New Testament, and of Judaeo-Christian theologians since theNew Testament, is generally an unhappy compromise between thesetwo.

I subm it that Spin oza called his one substance G o d for the perfectlygood reason that i t had much in common with the God of the philoso-phers. That i t had li t t le in common with the God of the Old Testamentcould hardly have mattered to him. The aim of Scripture, on his view,

is not to present metaphysical truth, bu t to inculcate mo rality. To thatend it must accommodate itself to the understanding of the people forw ho m it is intended. T he re is as m uc h justification for calling the onesubstance G o d as there is for dignifyin g with that nam e the beingto who m Luth er offered up his pray ers— an d as l it tle.

T o this the reade r ma y reply that there is one feature of crucia l impo r-tance both in the traditional philosophical conception of God and inSpinoza's conception of his one substance, which I have so far failed

to account for, namely, necessary existence. My basic nomological factsmay be appropriately described as an uncaused first cause, providedthat w e understand first logically rather than tempo rally, and theym ay be present through out all of t ime and space. Bu t, i t will beargued, that in no way justifies characterizing them as necessary existents.Unless they can b e so characterized, Spinoza's substance will turn outto be merely sempiternal, not eternal, and its identification even withthe God of the philosophers will fail . The basic nomological facts aremere brute facts, contingencies for whic h there is and can b e no explana-tion. And therefore they cannot serve to give an ultimate explanationof the existence of any finite being. Moreover, by limiting God's poweras I have, by talking of antecedent conditions stretching back into theinfinite past, I have introduced a further element of contingency intothe system. But Spinoza's rigorous determinism can no more admitcontingency than the traditional philosophic conception of God canadmit a l imitation on God's power.

This is a very cogent and important objection. I shall do my best

to meet it in the next chapter.





Spinoza begins an d ends with theological term s; and yet whe nwe translate his doctr ines into m od ern lang uage we f ind a viewof the world s tanding wholly apart from those which have beenpropounded or seriously influenced by theology. His earlier writ-ings help us to understand the seeming riddle. He started withthe intentio n of m ak ing theology philosoph ical bu t with th e de-terminat ion to fol low reason to the ut termost . Reason led himbeyond the atmosphere of theology al together but his advance

was so continuous that the full extent of i t was hardly perceivedby himself.

—Pol lock

Necessity





very Tru th Is a Necessary Truth

No one who has even the s l ightest acquaintance with the thics willdeny that Spinoza's system is rigorously deterministic. That in naturenothing is contingent, all things being determined by the necessity ofthe divine nature to exist and act in a fixed way (E IP29), that thingscould have been produced by God in no other way and in no otherorder than they have been (E IP33 ) , are am on g the best kn ow n and

most important propositions in that work. But to ascertain their precisemeaning is a task more difficult than might at f irst appear.

W e m ay seek to capture their import in a ma nne r congenial to modernears by var ious paraphrases. W e m ay say, with H am pshire , that forSp ino za the actu al wo rld is the only possible w orld (p. 42 ) , or w ithC. de Deugd, that all possibil i ty is actuality to an omniscient being(p. 70). Or we may say, as I should prefer to, that for Spinoza everytruth is a necessary truth. But each of these formulas will be foundto harbor ambiguities, and it is no easy matter to decide which inter-pretation of them comes closest to expressing Spinoza's intent.

Let us take the last one first. One of the theses I wish to defendin this chapter is that although Spinoza is committed to the view thatevery truth is a necessary truth, he is also committed, quite consistently,to the view that some are not. This, however, is a needlessly paradoxicalw ay of puttin g it . T o rem ove the air of con tradictio n, w e need to ma kea distinction all truths are necessary, but not all truths are necessaryin the same sense ; som e are necessary in a sense w hi ch allows , indee dwhich requires, that in another equally legitimate sense they are notnecessary but contingent.

Stil l, before explorin g the senses of nec essar y invo lved here, i t willbe best to first consider the thesis that every truth is a necessary oneas naively as we can, without raising any question as to its possiblemeanings.

Prima facie, there would seem to be very strong objections to thethesis. For one thing, there are many propositions which we should






wish to call true, but which we should not wish to call necessarily true,on the ground that, for all we can tell, they could have been false.It is true that S pinoza had dark hair, but tha t truth does not strikeus as a necessary one. To use a test of necessary truth made popularby Hume, its falsity is conceivable; to suppose that Spinoza did nothave dark hair does not appear to involve us in any contradiction.

Moreover, it is natural to wonder what sort of justification Spinozacould possibly give for saying that a proposition like Sp ino za had d ark

ha ir is necessarily true. If Sp inoza had been Leibniz and held a Leib-nizian view of individual concepts, then we might be able to constructa justification. According to Leibniz,1 all true propositions about indi-viduals—except existential ones—are in principle capable of being shownto be identities, that is, of the form A is A or AB is A . The reasonfor this is that the concept of an individual is supposed by him to includethe concept of every predicate that may be truly ascribed to the indi-vidu al at any time in his life, excep t the predicate . . exists. H en ce,on Leibniz' view of individual concepts, any true nonexistential proposi-tion about Spinoza would be, or would seem to be, what we shouldcall today an analytic truth.2 Just as w e m ight show all bachelors areunm arried to be analytically true by substituting for the subject term,bac he lor, its definition, unm arried m ale, so we could show Sp inoz a

had dark ha ir to be analytically true by substituting for Sp ino zahis definition or individual concept. Presumably the correct account ofSp inoz a had dark ha ir, in Leibn izian terms, wou ld be something like

this:

The individual who was born in Amsterdam, on 24 November1632, to M ichael and Ha nna h Spinoza . . . w as educated underRabbi Saul Morteira and Rabbi Manasseh ben Israel . . . haddark hair . . . wrote the Ethics . . . g round lenses for a living. . . and died at the H agu e, on 21 Febru ary 167 7, had dark hair.

Since this has at least the appearance of being analytically true, and

since few philosophers would question the necessity of analytic truths,

84





ecessity

a Leibn izian acco unt w ould give Spino za some sort of justification forholding the proposition Sp ino za had dark ha ir, and others like it,to be necessarily true.

There would be some difficulty about this. First, and obviously, refer-ence to other individuals has been ma de in spelling out the individualconcept of Spinoza. Since each of these, in turn, has an individual con-cept capable of analysis, the account presented here could not be anultimate one even if we were able to fill in all the blanks. Giving a

complete analysis of a singular proposition is going to be difficult atbest. Leibniz holds that it requires going through an infinite processand is possible only for God. Statements involving essential referenceto individuals may thus be characterized as infinitely analytic. By contrast,those dealing with abstractions, such as T h e wise m an is ha pp y, arefinitely analytic. A full analysis of the concept of the subject is possiblewithout going through an infinite process, and consequently the reduc-tion to an identity can be accomplished by man.

But second, and not so obviously, the analysis of the concept of anindividual involves a regress which is not merely infinite, but viciouslyso.3 Suppose we set out to give an analysis of the concept of Spinozaalong the lines suggested above. At some stage we shall wish to replacethe proper nam e M anas seh ben Israel by a definite description whichgives its analysis. If that analysis is to be complete, then the definitedescription will have the form T h e individual w ho was born in Lisbonin 1604 . . . was R ab bi of the second Am sterdam Synag ogue . . .educated Sp inoza . . . went to En gland on a mission to Olive r Cro m -

well . . . and so on through a comp lete enumeration of M anasseh'sproperties. But that w ill not be enough. W e shall also have to replaceall the proper names in this definite description by definite descriptionsof their ow n— an d am ong them, of course, is the nam e Sp inoz a. Inshort, we cannot give a complete analysis of the individual concept ofSpinoza until we have a complete analysis of the individual conceptof Manasseh ben Israel but we cannot give a complete analysis ofthe concept of Manasseh ben Israel unless we already have at handa complete analysis of the concept of Spinoza. Under these circum-






stances, even God might find it hard to give a complete analysis ofthe concept of an individual.

Still, whatever difficulties there might be in a Leibnizian accountof singular propositions, if Spinoza had held views like those of Leibniz,then we might be able to understand his thinking that every truth isa necessary truth. But Spinoza manifestly did not hold such views.

First of all, Le ibn iz doctrine of indiv idua l concepts, as he himselfpoints out, involves him in bringing b ack the substantial forms of the

m edieval philosophers. But substantial forms were very m uch out offashion in Leib niz time, and no one rejected them m ore em phaticallythan Spinoza (Ep. 56, I V : 26 1-26 2 ). Ag ain , Leibniz inferred from hisdoctrine of individual concepts that every individual substance was aworld apart, dependent on and affected by no other substance but God.Had Spinoza held the same doctrine, he would have inferred similarlythat men are always active, never passive E III DI-3), and conse-quently that they are always virtuous and free (E IVPI8S) That he

could not have accepted such a conclusion is too obvious to need proof.So the Leibnizian route can provide no explanation for Spinoza s holdingthat every truth is a necessary one.

In W hat Sense Every Tru th Is a Necessary T ruthTo understand how Spinoza might try to justify his doctrine that everytruth is necessarily so, w e sh all need to con sider wh at sort of analysisSp inoza suggests for the con cept of necessary truth. W e shall find itto be different in important respects from that of Leibniz. Accordingto the latter, true propositions divide into truths of reason (necessarytruths) and truths of fact (contingent ones). Truths of reason are truein virtue of the principle of contradiction their denials involve a con-tradiction, either explicitly or implicitly. Examples would be the proposi-tion that God exists and the laws of logic and mathematics.

Truths of fact are true in virtue of the principle of sufficient reason,which says, on one version at least,4 that nothing happens without a

8





ecessity

reason why it should be so rather than otherwise. Examples would beexistential propositions (except the proposition that God exists), thelaws of nature, and singular propositions.

Since singular propositions are reducible through definitions to identi-ties, it might appear that they ought to be classed as truths of reasonrather than as truths of fact. But Leibniz resists this conclusion. Some-times his ground for denying that they are strictly necessary seems tobe that they are infinitely rather than finitely analytic, though this seems

inadequate. It suggests that the contingency of these propositions is anillusion of the finite intellect. Sometimes, more plausibly, his groundseems to be that, although all nonexistential propositions about individ-uals are analytic, existential ones are not. Sp ino za had dark ha ir,if true, is analytically true; but it is not necessarily true without qualifica-tion, fo r it is true only if Sp ino za exists is true. Th is latter propositionis not necessary, but contingent, as it depends on God's free decisionto create this world, a decision motivated but not necessitated by thefact that this world, if created, would be the best of all possible worlds.

If I am correct in giving ( ι ) T h e individua l w h o . . . h ad darkhair . . . had dark ha ir as the proper Leibn izian accoun t of 2 ) Spi -noza had dark ha ir, then I suppose most contemp orary philosopherswould agree with Leibniz that ( 1 ), despite its appearance of analyticity,is not necessarily true. On a Russellian account of definite descriptions( ι ) is false if there is no such individual. O n a Strawson ian accoun tit is neither true nor false.6

Leibniz will allow that truths of fact are necessary in one possiblesense of the term. Though they may fail to be true and are thereforenot absolutely necessary, nevertheless they are hypothetically necessary.Given God's decision to create this world, rather than some other possibleworld, things could not have been otherwise. But God need not haveso decided. His decision had a cause in his knowledge that this worldwould be the best of all possible worlds, and it could have been predictedwith certainty, but it was not absolutely necessary. And therefore, neitheris the world which follows from it.






T h e similar it ies an d differences betwee n L eibniz a nd Spinoza o n thispoint are instructive. Consider, for example, the following passage inth e Ethics:

Since I have shown by the preceding, more clearly than by noonlight, that there is absolutely nothing in things on account of whichthey may be called contingent, I wish now to explain briefly whatwe must understand by contingent; but f i rs t , necessary andimpossible.

A thing is called necessary either by reason of its essence or byreason of its cause. For the existence of any thing follows neces-sarily either from its essence and definition, or from a givenefficient cause.

For the same reasons are things said to be impossible, that is,either beca use th e essence or definition of the th ing involves acontradict ion, or because there is no external cause determ inate

for the production of such a thing.But a thing is not called contingent except on account of a defect

in our know ledge. Fo r if we do no t know tha t the thing s essenceinvolves a contradiction, or if , knowing that i t involves no contra-diction, we nevertheless can affirm no thin g w ith certainty ab ou tits existence, because the order of causes is hidden from us, thenthe thing can never appear either necessary or impossible to us.And hence we call i t either contingent or possible. (E IP33S)

Now Spinoza is talking here about the necessity or impossibility of things,rather than of truths, and he speaks only of the necessity or impossibilityof the existence of things, and not of the necessity or impossibility oftheir possessing certain properties or entering into certain relations. Butthis need no t preven t us fr om translat ing w ha t he says about thingsinto talk abo ut t ruth s and developing a general accou nt of necessarytru th tha t will accord w ith Spinoz a s inten tions.

I t m ight be ob jected at this point that to interpret Sp inoza in this





Necessity

way presupposes that he did not have the vocabulary to talk aboutnecessary truths instead of necessary things. But this distinction and vo-cabulary were, it might be argued, available to him. If he did not usethe m , it was either fr o m carelessness, which seems unlikely, or byintention.

Q uite so. T h e voca bulary was available an d Sp inoza deliberatelyavo ided using it. H is reasons fo r avoiding it are of some interest. I na letter to Simon de Vries Ep. ι o ), Sp inoza writes :

Yo u ask w he the r things or the affec tions of things a re also eter naltruths? I say, certainly. If you ask why I do not call them eternaltruths, I reply to distinguish them, as everyone customarily does,fr om those things which do no t explain any thing or any affectionof a thing. F or exam ple, nothing comes into existence fro m nothing .This, I say, and similar propositions are called absolutely eternaltruths, by wh ich m en wish to signify only tha t such things hav e

no place outside the mind.

I take it that Spinoza avoided talk of necessary truths because he wantedit to be clear tha t he was talking ab out the w orld, abou t so methingthat had existence extra mentem. Lacking the convenient term fac tto serve as an ontological correlate of the term tru th , he ha d to mak edo with the inconvenient term thin g.

Bu t thou gh th e ontological imp ort of Spinoza's doctrine mu st alwaysbe kept in mind, I think we can most easily come to terms with Spinozaif we represent him as holding the following view. All propositions areeither necessarily true or necessarily false. This will hold both for existen-tial and for nonexistential propositions. But, restricting ourselves totruths, not all truths are necessary in the same sense. Some are absolutelynecessary, in the Leib nizian sense th at th eir den ial is explicitly or im -plicitly self-contradictory their truth follows from the essence or defini-tion of th e subjec t. But others ar e only relatively necessary. T he ir den ialdoes not involve a contradiction, either explicitly or implicitly. Their






truth, rather, is grounded in the fact that they follow logically fromother propositions which are true, propositions which given an efficientcausal explanation of them.

Relative necessity is like Leibniz' hypothetical necessity. It has theform given p, q is necessary. But i t is also, and more important, unlikehypothetical necessity. The proposition (or set of propositions) relativeto which q is necessary provides an explanation in terms of efficientcausation. Leibniz ' hypothetically necessary propositions are explained,

ultimately, in terms of a final cause, God's decision to create the bestof all possible worlds.

Again, Leibniz ' hypothetically necessary propositions do have an ult i-mate ground in the will of God. Spinoza's relatively necessary proposi-t ions are grounded both in the will , or intellect, of God (the laws ofnature) and in an infinite series of prior finite causes. Hence, in a sensethey have no ultimate cause.

Finally, and most important, Leibniz ' hypothetically necessary proposi-

t ions are ult imately grounded in something which is i tself supposed tobe cont ingent . Assuming a good argument for the cont ingency of God'swill , L eibn iz is righ t to describe his hypo thetically necessary propositionsas in themselves contingent. But a relatively necessary proposition inSpinoza's system is grounded in a series of other propositions each ofwhich is either absolutely or relatively necessary. Therefore, though itwould be correct to say that they are contingent in the sense of notbeing absolutely necessary, this would be misleading, as i t might suggestthe Leibnizian picture of things.

Since some may find this conclusion difficult to accept, let me adduceone further passage in support of i t . Consider the case of true propositionsof the form x exists wh ere the values of the varia ble χ are singularreferr ing expressions for exam ple, G o d , Sp in oza , this table ,etc. I t is clear that, l ike Leibniz, Spinoza would say that only one suchproposition is absolutely necessary, nam ely, G o d exists.8 Otherwisehe w ou ld not say tha t T h e essence of things pro duc ed by G od doesnot involve exis tence ( E IP 24 ) . But for Spino za, this does not preclude

90





Necessity

saying that true existential propositions are necessary in some other sense.W itness the follo wi ng passage :

For each thing there must be assigned a caíase or reason why itexists or why it does not exist. For example, if a triangle exists,there must be a reason or cause why it exists; and if it does not exist,there must also be a reason or cause which prevents or denies itsexistence.

This reason or cause must be contained either in the nature of thething or outside i t . For example, i ts own nature proclaims the reasonwhy a square circle does not exist , because i t involves a contradic-tion. The reason why substance, on the other hand, exists, alsofollow s fro m its natu re alon e, because i t involves existence (P7 ).

But the reason why a circle or a triangle exists or does not existdoes not follow from the nature of these things, but from the orderof corporeal nature as a whole. For from this i t must follow, either

that a tr iangle does now exist, or that it does not now exist.(E IPI ID2, my italics)

Those things whose reason for existence is contained in their natureare clearly the things called necessary by reason of their essence inE IP3 3S . T o say tha t the reason for a thing s existence is contain ed inits nature is equivalent to saying that the proposition asserting i ts ex-istence is absolutely necessary. W e m ay wish to say that there is onlyone such thing existing, namely substance. Or we may wish to say thatthere are many, namely all the attributes of substance. For substancesimply is the sum of i ts attributes. But whichever we say, substance,or all of the attributes of substance, are eternal ( IP2O) . I t would bemisleading, though correct, to say that they exist now, since this wouldsuggest that the nonexistence of the attributes at some other time isa logical possibility, which it is not.

But to say that the reason of a thing lies outside its nature, in the

91






order of corporeal nature as a whole (E IPI I ), or in the given efficientcause (E IP33S), is equivalent to saying that the proposition assertingits existence is not absolutely, but only relatively necessary. This willbe true of the modes of substance generally (with the possible, but awk-ward exception of the infinite modes). Since, for any mode now existing,its nonexistence at some other time is logically possible, there is no inap-propriateness in saying that it now exists. And Spinoza does speak inthis way of the modes.

On this interpretation, Spinoza can offer either of two explanationsfor the apparent contingency of some true propositions. Either we donot know the essence of the thing sufficiently clearly to perceive theabsolute necessity of the proposition, or we lack sufficient scientific infor-mation to perceive its relative necessity. This lack of information maybe an ignorance of the relevant laws, or of the antecedent conditions,or both. The reason why an omniscient being cannot feign anything,in the sense of supposing something to be the case which is not, is thatneither th e order of essences nor th e order of causes is concealed fromhim. Relative to his perfect knowledge of the definitions of things,of the laws of nature, and of the antecedent conditions, the hypothesistha t Spinoza did no t have dark hair would possess the same kind ofimpossibility fo r him w hich th e hypothesis that there is a round squarepossesses for us.7

For Leibniz, on the other hand, all and only absolutely necessarypropositions are necessary. Hence, he gives only the first of Spinoza stwo alternatives as an explanation of the apparent contingency of somenecessary propositions that we do not have adequate knowledge of theessence of the thing.

W ith regard to truths which are not absolutely necessary, he willsay tha t we sometimes fail to perceive their reason, or cause, but notthat we fail to perceive their necessity. For they are not, strictly speaking,necessary. Th e explanation for our failure to perceive their reason willbe, not that we do not know the efficient causal explanation of thefacts but that we do not know their explanation in terms of final causes.

9 2





ecessity

W e do no t see how the existence of this or that state of affa irs servesto realize some value.

W hy Every Tru th Is a Necessary Tru thSo far I have been concerned to show that Spinoza s universal denialof contingency is compatible with the admission that some truths arecontingent, in one important sense of that ambiguous term. His deter-

minism has the form :( ι ) Ev ery tru th is either absolutely or relatively n ecessary,

and not the form :( 2 ) Ev ery truth is absolutely necessary.

Now I want to consider the problem of how Spinoza might justify thisclaim. I think that if we grant Spinoza certain assumptions, assumptionswh ich I am confident that he w ould ma ke, his thesis must be true. Solet me list those assumptions, and consider briefly the grounds he hasfor them.

( ι ) T h e proposition that G od exists is absolutely necessary.2 ) A ll other singular truths, tho ugh ab solutely conting ent, ha ve a

scientific explanation, in the sense that they follow from a statementof certain antecedent conditions and nomological propositions. Hencethey are relatively necessary.

( 3 ) A ll accidental general truths, though absolutely contingent, havea scientific explanation in the same sense, and hence are also relativelynecessary.

(4 ) A ll nomo logical general truths are absolutely necessary.These four theses assign each of four classes of truth to one or theother of the two categories of necessary truth. If the classification isexhau stive, as it seems to be, an d if e ach thesis is true, then Spin oza sdeterminism is established. So we may proceed to ask whether Spinozawould make these claims and how he would support them.

T h at he wo uld m ake the first is certain. It is E IP I 1. An d the variousproofs adduced for that proposition make clear that it is absolute neces-







Necessity

A less theologically oriented passage occurs later in the chapter, whenSpinoza is arguing, in effect, that reason must be the judge of whatis truly scriptural:

Wherefore we may here conclude absolutely that whatever is trulynarrated in Scripture as having happened must have happened,like all things, according to the laws of nature. And if anything isfound which can be demonstrated in a clear proof either to be

inconsistent with the laws of nature, or to have been unable tofollow logically from them, surely we must believe that it has beenadded to the Sacred Writings by sacreligious men. For whateveris contrary to nature is contrary to reason, and wh at is contrary toreason is absurd, and accordingly to be rejected. TTΡ vi, III : 91

Here he seems to be taking the view that for a singular truth not tohave a scientific explanation would involve a contradiction.

The third thesis—that all accidental general truths are relatively neces-sary—is not so easy to document. My reason for attributing it to Spinozalies in E I P 8 S 2 . There Spinoza, in the course of arguing that therecan only be one substance of the same nature, has occasion to discussthe proposition that there exist exactly twenty men, which Spinoza sup-poses for the sake of the discussion to be true. Spinoza contends that :

If a certain number of individuals exist in nature, there must nec-essarily be a cause why those individuals, and neither more nor fewer,exist. For example, if there are twenty men in existence . . . itwill not be sufficient in order to provide a reason why twenty menexist to show the cause of human nature in general. It will benecessary as well to show the cause why neither more nor fewerthan twenty exist. But this cause cannot be contained in humannature itself, since the true definition of m an does not involve thenumber twenty. Therefore the cause why these twenty men exist,and consequently why each exists, must be outside each. I I : 50-51

9






Spinoza goes on to conclude that wherever a thing is of a kind whichcan have several individuals it must have an external cause. Since itpertains to the nature of substance to exist (that is, it does not havean external cause), there can only be one substance of the same kind.

T h e thing that is interesting about this passage for our purposes, how -ever, is that the proposition Th er e are [exactly] twenty m en wo uldbe an exam ple of an acc iden tal generalization and that Spinoza seemsto want to assimilate this proposition to singular propositions and to

say that it is necessary in precisely the same way they are. This is whyI suggested in the preceding chapter that we might try to understandSpinoza in terms of the (Poppeiian) view that accidental generalizationsare equivalent to singular propositions or truth-functions of singularpropositions.8

Th is is a view to w hich there are, of course, familiar objections. I tis commonly held that no set of singular propositions ever entails ageneral proposition. Suppose that Brown, Jones, and Smith are the only

men in the tearoom at 4:00 and that Brown, Jones, and Smith arewe aring shoes. Fro m the premises that

Brown is in the tearoom at 4:00 and Brown is wearing shoes.Jones is in the tearoom at 4:00 and Jones is wearing shoes.

Smith is in the tearoom at 4:00 and Smith is wearing shoes.

we cannot infer that

All the men in the tearoom at 4:00 are wearing shoes,

unless we have the additional premise that

Brown, Jones, and Smith are all the men in the tearoom at 4:00.

And this is a general proposition. If no set of singular propositions everentails a general proposition, then general propositions, whether strictly





ecessity

universal or not, cann ot be reduc ible tha t is, equ ivalen t) to singularpropositions or truth-functions of singular propositions.9

This objection is widely regarded as conclusive. One might, however,undertake to rebut it in the following way. Though the additional propo-sition which is necessary for the derivation of our general conclusion,that is,

Brown, Jones, and Smith are all the men in the tearoom at 4:00

might ordinarily be regarded as a general proposition, nevertheless, ifwe consider the way it would be established we shall see that this isa mistake. The proposition would be established by examining the roomas a whole to see whether or not it is occupied by men other thanBrown, Jones, and Smith. The additional premise required, therefore,can be is properly) regarded as a singular proposition abou t the roomas a whole, to the effect that

The tearoom at 4:00 is entirely free of other men.

T h is certainly satisfies the u sual test for singular propositions its subjectterm is a singular referring expression. Moreover, like the typical singularproposition, but unlike the typical universal proposition, it does not needto be established by inference. It can be known by direct observation.For that matter, our original conclusion might have been transformedimmediately into a similar proposition about the whole of this particular

region of space and time

The tearoom at 4:00 is entirely free of shoeless men.

or, if our original conclusion is thought to imply that there are menin the room at that time

The tearoom at 4:00 is not entirely free of men, but it is entirelyfree of shoeless men.

9






In general, wherever a universal generalization is restricted to a particu-lar region of space and time, it is equivalent to a singular propositionabout the whole of that region.

Whether or not this is a satisfactory rebuttal, I shall not undertaketo say here. But it does seem to me to be the sort of thing some logiciansmight wish to say,10 and it also seems to me an appropriate way of lookingat Spinoza. For by redrawing the usual distinction between universaland singular propositions, it makes the fundamental logical antithesis the

contrast between the infinite and the finite. And this, I think, is verymuch as it should be.

Th is brings us to the fourth and m ost interesting claim, that all no mo -logical propositions are absolutely necessary. Here we may appeal toSpinoza's own account of the notion of a law in the Theological-PoliticalTreatise. Af te r defining a law generally as tha t according to whicheach individual, or all or some members of the same species, act inone and the same fixed and definite m ann er, Spinoza goes on to dis-

tinguish between two kinds of law, those which depend on natural neces-sity and those which depend on hu m an conventions :

A law which depends on natural necessity is one which followsnecessarily from the nature or definition of the thing, whereas onewhich depends on human convention, and is more properly calledan ordinance, is one which men prescribe for themselves and othersin order to live m ore safely and c om fortably, or for some otherreasons. For example, that all bodies, when they strike other, smallerones, lose as m uch of their motion as they comm unicate to theothers, is a universal law of all bodies, which follows from thenecessity of nature. Similarly, that a man, when he recalls onething, immediately recalls another like it, or one which he has per-ceived simultaneously with it, is a law which follows necessarilyfrom human nature. But that men should yield, or be forced toyield, something of the right which they have by nature, or thatthey should commit themselves to live in a certain way, these things

depend on human convention. T T P i v, 111 : 5 7 - 5 8 )





ecessity

Here we find Spinoza making quite clearly the now familiar distinctionbetween descriptive and prescriptive laws. But what is important forour purposes is that Sp ino za wishes to regard the laws of nature (tho ug hno t the law s of m en ) as possessing the sam e kind of necessity the prop osi-tion G o d exists is suppo sed by him to possess; they follo w necessarilyfrom definit ions, and are hence absolutely necessary. Given our identifica-t ion of the essence of God with the fundamental nomological facts ,this is just what we should expect.

The view that nomological propositions are absolutely necessary isnot one l ikely to f ind much favor among contemporary phi losophers .They would argue that the laws of nature must be established empiri-cal ly—by examining nature , not by drawing out the consequences ofdefinit ions. And therefore, they would conclude, scientific laws cannotbe absolutely necessary.

What Spinzoa 's reply to this object ion would have been is hard todetermine. But without going very deeply into his theory of knowledge,which deserves a separate and thorough treatment, let me suggest thathe might well have rejected the assumption, implicit in this objection,that what is established empirically cannot also be established a priori .It is, after all, implied in his discussion of the Rule of Three that eventhe necessary truths of mathematics may be arrived at by any one ofthe three kinds of knowledge.

There was, moreover, much in the science of his t ime which mighthave encouraged him to take a s imilar view of physics . The unprejudicedand attentive reader of Galileo's Tw o ew Sciences cann ot fail to bestruck by the frequency with which that paradigm of the empir icalscientist appeals to a priori arguments as well as to a posteriori ones.For instance, when Salviati proposes to define uniformly accelerated mo-tion as that motion which, starting from rest , acquires equal incrementsof speed in equal intervals of t ime, however small the t ime interval,Sagredo raises a doubt as to whether this definit ion

corresponds to and describes that kind of accelerated motion which

we meet in nature in the case of freely fall ing bodies,11






for if it did

the n there is no degree of speed however small . . . w ith whichwe ma y not find a body travelling after starting from rest . . .[but] our senses show us tha t a heavy falling body suddenly acquiresa great speed.

Salviati replies first tha t th e experience w hich seems to show this is

misleading. He next adduces empirical suppo rt for his assumption ofa grad ual increase of speed an d contends that

the sam e experime nt w hich at first glance seemed to show one thingwh en more carefully exam ined assures us of the contra ry.

And then he concludes with an a priori argument

But without depending upon the above experiment which is doubt-

less very conclusive it seems to me th a t it oug ht no t to be difficultto establish such a fact by reasoning alone.

T he argu m ent th at follows is not a very good one. Ne ither for thatm atter is his empirical argum ent. But the procedure w hich this passageillustrates is a common one in Galileo.12

T he traditional notion of Galileo as the experime ntal scientist p arexcellence has of late come und er heavy attack fro m historians of sci-

ence. Rec ent research strongly suggests th at m any of the experimentshe described were never perform ed. H ad the y been they would nothave yielded the righ t results. But in trusting to reason or rathe r tow ha t we migh t call following Bacon a union of the experimental a ndratio nal facu lties Galileo gave voice to an ideal of science wh ich d om i-nate d his age. As the science of na ture developed A. R. H all writes

it becam e clear tha t its starting point the Laws of Motion fro mwh ich everything else could be derived with the aid of suitable

iooBrought to you by | Brown University Rockefeller Library




Necessity

definitions, according to mathematical strictness, did not have torest solely on the dubious validity of such experimental tests ascould be devised by men upon the surface of the Earth. I t seemedthat the Laws of Motion could not be false that is, that i t wasinconceivable that any alternative propositions could be valid. . . I t was superf luous to a t temp t to tes t the laws by exper imentsunder condit ions that in any case could never correspond to thoseun de r which th e laws would be r igorously true . Such experiments

did not prove the laws, but i l lustrated their applicability. Accordinglythe science of motion did not depend al together on inference andthe accuracy of experiments for i ts application to those bodies be-yond the reach of experim ent , unless i t should be t ru e that N atu rewas not everywhere uni form and confo rmab le to herse lf — whichno 17th Century philosopher or scientist was prepared to suppose.13

T his is no t to say, of course, th at Spino za wa s right in think ing thelaws of nat ur e to be absolutely necessary. But it does suggest th at Spi-noza's view was an intelligible one for a seventeenth-century philosopherto hold. In the final section of this chapter, I shall consider whetheror not i t is an intelligible view for a twentieth-century philosopher tohold.

Is the Actua l W orld the On ly Possible WorldSo far I have treated Spinoza's determinism in relation to the thesisthat every truth is a necessary one, arguing that, correctly understood,this will serve to express Spinoza's meaning and that, so understood,it permits a kind of contingency to creep into his system. Then I exploredthe rationale for the thesis. Now I wish to ask whether, on this interpreta-tion, the thesis does entail that the actual world is the only possibleworld.

Before we can answ er that q uest ion, how ever, we must decide w ha tmeaning we are to attach to the latter proposition. It is not at all obvioustha t i t has any clear mea ning. A fter Leibniz, we tend to picture an

1 1






infinite number of possible worlds, of which the actual world is theone that happens, by grace of G o d or Cha nc e, to exis t. T o say thatthe actual world is the only possible world seems to be to say that,of all the infinity of possible worlds, only one is possible. There arepossible worlds that are not possible. If the proposition does not saythis, then it is not clear what the actual world is being contrasted with.

The diff icul ty is surmountable . The intent of our formula can becap tured if we rephrase it as of all prim a facie possible worlds, only

the actual world is really possible . W ha t this mean s m ight be explainedin the fol lowing way. Suppose we conceive, as we have in the precedingchapter, of a complete description of the world, a set of propositionsA, containing every truth about the past, present, and future. Let usthen define a prima facie possible world as the descriptum of any setof propositions which is either identical with A or formed from A bynegating one, some, or all of the propositions in A, and conjoining thesenegations with every proposition in A which has not been negated. The

actual world wil l be the descr iptum of A.Using a device analogous to a truth table, we might first l ist the

propositions of A horizontally, and then, in vertical columns under eachproposition in A, l ist either tha t proposition or its neg ation , in the follow -ing manne r :

I . Ρ I r2 . Np q r

3 Ρ Nq r

4 · Np Nq r5· Ρ 1 Nr6 . Np q Nr7· Ρ Nq Nr8. Np Nq Nr

In this way all possible combinations of propositions and their negationswill be covered. If A contained only three propositions, p, q, and r,our table would have only three columns and eight rows. There would

102





Necessity

be only eight prima facie possible worlds. Since A contains infinitelym any propositions, the tab le will extend indefinitely to the right andindefinitely downward.

T h e claim th at th e ac tual world is the only possible world can nowbe made more clear. It will mean that there is only one row in thetable that constitutes a self-consistent set of propositions, namely, the toprow. Every other set involves a contradiction.

Given that every truth is either absolutely or relatively necessary, it

is easy to see wh at the ground s fo r this wou ld h ave to be . Any propositionin A which is absolutely necessary, say p cannot be consistently denied.H enc e any set of propositions containing Np will fail to be self-consistent.This will eliminate many rows on the table. Any proposition in A whichis not absolutely necessary, say q will be relatively necessary. T h a t is,q will be entailed by some other proposition r wh ich is also in A. Soany set containing Nq and r will be inconsistent. And again, many rowswill be eliminated.

Clearly, if every truth is either absolutely or relatively necessary, thena great m any rows of the tab le will turn out to be inconsistent. Butequally clearly, in the end we will have to admit a plurality of self-con-sistent rows. For consider the case of our relatively necessary propositionq. In itself it is contingent, so no row containing Nq is ruled out onthat ground alone. Relative to another proposition r, it is necessary,so any row containing Nq and r is impossible. But what of rows contain-in g Nq and Nr? The proposition r will be a conjunction of an absolutelynecessary proposition, s and a relatively necessary proposition, f. The

conjunction can be consistently denied if one of its conjuncts, s andt can. The former, being absolutely necessary, cannot. Can the latter?Well, it is only necessary relative to another proposition, u. Th e conjunc-tion of Nt and u is inconsistent, but the conjunction of Nt and Numay be all right. So the same question arises about t that arose aboutq and it gets the same answer. We are off and running now on thephilosophers favorite horse, the infinite regress. But this is a very gentle,well-mannered regress. Not vicious at all.

1 3






In short , had we but world enough and t ime, we could form a self-consistent set of propositions by taking any relatively necessary proposi-t ion in A, negating it , and then negating any other relatively necessaryproposition we ha ve to in order to m ainta in consistency. W e will beforced to deny infinitely many relatively necessary propositions, but withpatience and perseverance we can sti l l arrive at a complete set , con-sistent in itself, but inconsistent with A. If every truth is necessary, butsome are only relatively necessary, then the actual world is not the only

possible world. Only if every truth were absolutely necessary would theLeibnizian formula fol low.

The result is interesting and, I venture to say, conclusively demon-strated by the preceding argument . Whether or not i t would have beenacceptable to Spinoza is not so easy to decide. But i t is worth pointingout that the Leibnizian formula may be interpreted in other ways, accord-ing to which Spinoza would be able to say that the actual world isthe only possible world.

Suppose we take one of the other rows in the table which gives aself-consistent set of propositions and describes another prima facie possi-ble world that is really possible, though not actual. Call the set Β andthe world it describes the fi-world. Can it exist? In a sense, of course,i t can. It is , as C. I . Lewis might say, a consistently thinkable world.But in another sense, i t cannot, merely because the A-world exists andthe existence of the 5-world is incompatible with the existence of theA-world. To say that both of these worlds existed would be to say thata proposition a nd its nega tion m ight bo th be true. In this sense, perhapsa trivial one, the actua l world is the only possible world . Gi ven thew ay the term w o rl d is bein g used here, there could not be m ore thanone prima facie possible world which was actual.

T his w ould app ear to be the sort of s i tuat ion ruled out by E I P 5 :there canno t be two or mo re substances ha vin g the same na ture . I f

the worlds described by A and Β both existed, there would be twosubstances having the same nature . They would have the same nature ,because Β must contain exactly the same nomological propositions as

104





Necessity

A Oth erw ise since the no m olog ical propositions axe absolutely necessaryit would not be a self-consistent set. So the 5-world must be distinguishedfro m the A-w or ld on ly by a difference in i ts m odification s in that Βwo uld contain different non nom ological proposit ions. In saying thatthese tw o worlds cann ot coexist w e are saying that Spin oza is rightto hold that there cannot be two or more substances having the samenature .

Th is l ine of thoug ht can b e pushed a bi t fur ther. W e c an argue n ot

only that there cannot be tw o substances ha vin g the same nature bu talso that there cannot be two substances having different natures. Ifthe ß-world has a different nature than the A-world, then i t wi l l containthe negation of one or more of A's nomological propositions. But thenomological propositions are absolutely necessary. So the ß-world simplycould not exist . There could not be a world, or substance, having anature different f rom that of the actual world. And s ince we have alreadyargued that there could not be another substance of the same nature,

there could not be two or more substances, period. Such is the rationalewe might give, on this interpretat ion, for that fundamental theoremof the Ethics w hic h says that Besides G od , no substance can be orbe conce ived ( I P1 4) .

If there is some other self-consistent row, the world described by thatset of propositions cannot exist in addition to the world described byA In tha t sense, surely, the a ctua l wo rld is the on ly possible w orld .But perhaps Spinoza would wish to hold that the actual world is theonly possible world in the stronger sense that the actual world, in allits detail, must exist. If the ß-world is really possible, though not actual,i t seems a fair question to ask Spi no za W h y no t? It hard ly seemssufficient to say that it can 't becau se the A- w or ld does. W h a t w e m eanis W h y the ¿4-world rather than the 5- w or ld ? If Spin oza is goingto rule this question out of order, then he is going to accept a l imitationon the principle of sufficient reason. He is going to admit that thereis something whose existence is without a reason or cause. He cannotanswer that the ,4-world exists because its existence is logically necessary

105






given the necessary trut h of its nomological propositions. Th e ß -w orldpossesses exactly the sam e reason for existence since Β contains exactlythe sam e nomological propositions. If it were no t for a d ifference inthe nonnomological propositions they contain, A would be identical withB. O ur question is this: W hy do the nomological propositions, whoseabsolute necessity we are conceding for the sake of argum ent, have theinstances they have, rather than some other possible set of instances?It m ay be tha t Spinoza would reject the question. It certainly seems

tha t h e should. But if we say tha t he would, we must remem ber tha tour ground for this is merely that he has to, to be consistent with hisaccount of necessity. We should remember also that it seems to entaila qualification on th e principle of sufficient reason tha t Spinoza neverexplicitly makes.

Still, even if Sp inoza does not, on the most na tur al interp retation ,hold the actua l world to be the only possible world, his determinismis quite strong enough to be philosophically interesting. Any given event

is to be seen as the necessary outcome of prior events. Certainly thisform of determinism raises all the fam iliar problems ab out mo ralresponsibility.

Objections and Rep lies gainAt the end of the last chap ter I posed two questions. First, have Ino t, by my talk of antecedent conditions stretching back into the infinitepast, introduced an alien element of contingency into the system? Second,are not the basic nomological facts contingent, and hence merely sempi-terna l, or is there some basis for holding them to be necessary an d eternal?To the first of these questions I answer, firmly, yes and no. There is,on my interpretation, an element of contingency in the system. Howfar it is an alien element introduced into the system I cannot say.There certainly are substantial grounds for thinking that it is not.

T o tu rn now to the second question, it would appe ar tha t it hasalready been answered in dealing with the first question. For Spinozathe basic nomological propositions, indeed, all nomological propositions,

1 6





Necessity

and consequently the facts they describe, are logically necessary in themost strict and absolute sense. But this answer is the source of muchdifficulty, both for the internal consistency of the system and for itsplausibility.

In the first place, if all nomological propositions are absolutely neces-sary, then the infinite modes, the facts described by the derivative nomo-logical propositions, possess the same reason for existence that substancedoes. Their essence, to use Spinoza's terminology, involves existence,

and they are their own cause E IDI ), independent of everything else.This is implied in Spinoza's description of them as infinite and eternal,for eternity, by definition, involves necessary existence and not just ex-istence without beginning or end (E ID8) But this conflicts withthe description of them as modes. A mode, by definition, is an effect,something dependent in nature, something whose existence is not self-explanatory. An eternal mode is a contradiction in terms.

This difficulty does not arise merely because of my somewhat idiosyn-

cratic way of interpreting Spinoza. It has been noted and dealt withby Wolfson, who argues that Spinoza simply does not mean what hesays, that when he speaks of the infinite modes as eternal, what he reallymeans is that they are sempiternal (I, pp. 376-377). His evidence isthat in Ε Ip2 1, where Spino za m entions the infinite modes for the firsttime in the Ethics, he says : W ha teve r follows from the absolute natureof any attribute of God must lw ys exist and be infinite, or [sive]through that same attribute is eternal and infinite. T h e wo rd alw ay s[semper] suggests temporality, omnitemporality, to be sure, but tempo-rality nonetheless.

Su ch solutions may be left to those w ho m they satisfy. T h e pointremains that whatever follows logically from something which is logicallynecessary must itself be logically necessary. If Spinoza regards the exis-tence and causality of God as being strictly logically necessary, thenhe has to say that the infinite modes are eternal, whether he wantsto or not.

Secondly, and this affects the plausibility of the system, not its con-sistency, many contemporary philosophers would hold that the laws of

107






nature cannot be logically necessary on the grounds that they are informa-tive. They tell us how things are, whereas logically necessary proposi-tions give us no information about the nature of things. We know noth-ing about the weather when we know that it either is or is not raining.W e know n othing about bodies when w e know that they are eitherin motion or at rest. When Spinoza does give us an example of a lawthat seems to ha ve some content, as wh en he cites the law th at al lbodies, when they strike other smaller bodies, lose as much of their

mo tion as they comm unicate to the others, it is difficult to see howthe law could be regarded as logically necessary. We would like to seethe proof.

Moreover, it will be contended, the reason why logical truths, andanalytic truths generally, are uninformative is that they are true simplyin virtue of the meanings of their terms. Their truth is guaranteed bythe conventional rules of the language in which they are expressed.Since analytic truths depend for their truth only on those rules, and

not on the existence of an y pa rticu lar state of aff air s in the wo rld, theycan say nothing about the world. They do not describe the nature ofthings, they reflect the prescriptive rules of lang uag e.

This, doubtless, is a very common view today. And though it hasits critics, I shall not join them now. What interests me here is thatthe objection which arises from this view of analytic truth forces usto consider how Spinoza conceived of definitions and what he thoughttheir status was. Once we do this, I think we will see that Spinoza'snotion of definition is very far removed from that which most contempo-rary philosophers have. As a result, what is implied when he says thatsomething follows from the definition of a thing is very different fromwh at wou ld be imp lied on a more conventional account.

The question of the nature of definition is raised in Spinoza's corre-spondence by Simon de Vries, who invites Spinoza to choose betweentwo conceptions (Ep. 8, IV:39-40) : the first, that of Giovanni Borelli,a more or less orthodox Aristotelian, according to whom definitions mustbe true, primary, clearly known to us, and essential, since they are needed

as premises in demonstrations the second, that of Christopher Clavius,according to whom definitions are formulas of human construction (artis





Necessity

vocabula , w hic h need not be justified so lon g as we do not affirmthe thing defined of anything which we have not first shown to possessthe defining characteristics. On this second view, a definit ion need notbe true, primary, or best known to us.

Spi no za replies tha t there are tw o kinds of definit ion E p . 9,I V : 4 2 - 4 4 ) : ι ) those wh ich explain a thing insofar as i t exis ts outsidethe intellect; these must give a true description of their object and donot differ from propositions or axioms, except in that definit ions are

concerned with the essences of things or of affections of things, whereaspropositions or axioms are more general and include eternal truths;14

2 ) those w hic h expla in a thing insofar as w e conc eive i t or can co nceiveit ; they are not appropriately characterized as true or false, nor arethey susceptible of proof. Spinoza does not say that definit ions that ex-plain a thing as it exists outside the intellect must be primary or bestknown to us. Here he insists only that they be true.

W h a t Sp ino za is givin g us is a version of the tradition al distinction

between real and nominal definit ions. And I think it may be helpfulhere if we pause for a moment to compare his classification of the differ-ent kinds of definition with that in the Port-Royal Logic pp . 86 -99 ;164-167) . Arnauld and Nicole recognize th ree fundamenta l ly d i ffe ren tkinds of definit ion. Two are species of nominal definit ion définitiondu nom ; one is real definition é inition de la chose :

ι ) T h e first sort of nom inal definit ion, w hi ch corresponds rough lyto what would today be cal led stipulative definition, occurs w he n awriter assigns a meaning to a word as the meaning it will have in hisdiscourse. Words in themselves are indifferent and may be made tosignify anything. So writers have freedom of stipulation. This sort ofdefinit ion is arbitrary, cannot be contested, and may be taken withoutany further justification as a first principle, provided that the writermakes his usage clear and sticks to it.

2 ) T h e second sort of no m ina l definit ion, w hic h corresponds roughlyto what would today be cal led a lexical definition, occurs wh en a wri terexplains what a word means according to the ordinary usage of some

language. Such definit ions are not arbitrary. However, they are requiredto represent not the truth of things, but the truth of usage. Their job

109






is to connect the word with the idea people commonly connect i t wi th ,and their truth or falsity hangs on whether or not they do this.

( 3 The third kind of definit ion is real definition. Here the not ionseems to be that one treats the word defined as having its ordinarymeaning, as being linked with the idea i t is commonly linked with,and then asserts that certain other ideas are contained in (implied by)that idea. This kind of definit ion is not arbitrary, for i t does not dependon m an s wil l that ideas imp ly wh at they do. Hen ce, def init ions wh ich

are proposed as real definit ions can be contested and cannot be takenas first principles without some justification. Unless they are as clearas axioms, they require proof . Arnauld and Nicole say much that isinteresting about these different types of definit ion. They point out, forinstance, that though we are free to stipulate any meaning we like fora wo rd it is no t alwa ys advisa ble to exercise this freed om . In genera lthey take the posit ion that an established usage should not be changedunless there is something unclear about i t , and that when we do intro-

duce a new usage, i t should not be too far removed either fromthe existing usage or from the etymology of the word. They also havesome very suggestive things to say about the difficulty of giving adequatelexical definitions.

But the point which is of the greatest importance for our purposesis that they insist quite strongly on the need for distinguishing clearlybetween stipulative definit ions and real ones. The former never needjustification, the latter often do. It then becomes crucial to see whatthe criteria are for a good real definit ion.

And it is at this stage that Arnauld and Nicole are neither clearnor helpful . They are Aris totel ian and Cartesian. The requirements theylay do wn for a go od real definit ion are as follows :

( ι ) A good def init ion should be both universal and pr op er— by wh ichthey mean that it should give the essential attributes of the thing, thatis, state necessary and sufficient conditions for being that thing, or athing of that kind.

They seem to think of this as being done in traditional genus and

difference fashion, though they give as examples of successful real defini-tions a very mixed lot of propositions : man is a ra t ional animal , mind

ι io








differ for different kinds of def ini tion. Simp lifying somew hat , we cansay that the definit ion of a created thing (that is , a thing whose essencedoes not involve existence, a m od e) mu st

( ι ) give the prox im ate cau se of the thing , a nd( 2 ) suffice fo r the ded uc tion of all the properties of the thin g. TdlE,

9 6 - 9 7 . n 3 5 - 3 6 The definit ion of an uncreated thing (that is , a thing whose essencedoes involve existence, an attribute) mu st

( ι ) exclu de every cause by ma kin g it clear that the thing requiresnothing for i ts explanat ion beyond i ts own being and,

( 2 ) leave no room fo r the question whe the r the thin g exists an d,( 3 ) suffice fo r the ded uc tion of all the thin g's properties.

Spinoza i l lustrates this conception of definit ion with the following exam-ple. A circle is badly defined as any figure in which lines drawn fromthe center to the circumference are equal. For this does not give theessence of a circle but only one of its properties. A circle is well defined

as that figure which is described by any line of which one end is fixedand the other moving. For this definit ion gives the proximate causean d suffices for the de du ction of such prop erties of circles as the on egiven in the bad definit ion.

It is interesting that Spinoza's example should come from mathe-matics. Circles, and mathematical entit ies generally, are classed by Spi-noza as entit ies of reason. They are not real things. They are the same,he says, however they are defined, and it does not much matter howthey are defined. W ith real things i t is otherwise. W h y no t, then, g iveas an example a definit ion of a real thing? Spinoza's ostensible reasonis that he does not wish to lay bare the errors of others. This is verym uc h in cha racter a nd is prob ably correct. Bu t i t is na tura l to suspectthat examp les w ou ld be difficult to find outside of m athem atics. T h esort of definit ion Spinoza requires is going to be available, at best , onlyin the advanced sciences. If we accept the view of Professor H. H. Dubsthat the nuc lear equ at ion of a chem ical e lement is the beginningof a real definition,15 then there will be real definit ions outside of mathe-

matics. But there will not be very many. And they will not be discoveredby examining linguistic usage.

11 2





Necessity

It may be objected that if this is what Spinoza has in mind whenhe speaks of definition—the search for a cause which w ill explain athing's properties—then definition is a legitimate and valuab le activity,but one which is ill-named. For to call this activity de finitio n leadsus to think that we need not make the empirical tests appropriateto all theories about causes.16

The point is well taken. I am not certain, however, that Spinoza him-self was misled in this way. He speaks, for instance, in the Theological-

Political Treatise of the interpretation of nature as consisting in thesystematic study concinnanda) of the history of nature, fr om wh ichwe infe r the definitions of natu ral things . . . nature does not give usdefinitions of natural things . . . they must be inferred from the diverseactions of nature TTΡ 7 , I I I : 98 -99 ). It is difficult to know quitehow to relate such statements to Spinoza's over-all theory of knowledge,but they certainly must be taken into account when we are consideringhis theory of definition.

I suggested at the beginning of this discussion that once we saw howSpinoza conceived of definitions we would see why he thought theydescribed the nature of things. The real definitions which his sciencewould employ would not be true simply in virtue of linguistic usage.An d so the charge that his nomological propositions are vacuous mustfail.

But only at a price. Once we see that definitions, real ones, at anyrate, are not merely the expression of human conventions, we can seehow they may be informative. They depend for their truth not onlyon our rules for the use of language, but also on the existence of thestates of affairs they purport to describe. Nevertheless, now there nolonger seems any reason to regard them as necessary. To question themis not just to show ignorance of the way words are used, it is to raisesubstantive doubts about the nature of things. To call them definitionsat all is probably a mistake, tending only to confusion.

Sha ll we conclude, then, that the basic nomological facts are contin-gent, not necessary, sempiternal only, and not eternal? I am not sure.

It seems to me that Spinoza could argue—or that we could argue onhis beha lf— tha t the very independence which characterizes these facts

3





Spinozas etaphysics

is i tself a sufficient reason for call ing them necessary. When we saythat something is contingent, etymology suggests, and ordinary use con-firms, that w e ordinarily u nderstand this to be a relative notion. W emean that there is another thing on which our something is contingentor dependent . T o speak of something whic h is cont ingent , but w hic his not contingent on anything, is , in this everyday use of the word, unintel-l igible. If so, then the basic nomological facts can hardly be calledcontingent.

Th is point ma y be ap proach ed in a s lightly different w ay b y consider-ing what , in one common case, we mean when we say that somethingcould have been otherwise. Suppose there has been an accident. A carhas smashed into a tree and the people inside have been inju red . W em ay say, consolingly, A h we ll , i t could ha ve been w or se — at leastthey were wea rin g their seat-belts. By this I take i t tha t w e m ea n

If they had not been wearing their seat-belts, i t would have beenwo rse. U nd er the circumstances, that is , g iven the exist ing b ackg roun dconditions, such as the velocity of the car, the fact that the dashboardwas not padded, and so on, and given that the laws of nature are whatthey are, the fact that they were wearing seat-belts was a sufficient condi-tion of their not being injured more seriously. And under those circum-stances, the fact that they were not wearing seat-belts—had it beena fact—would have been a suff ic ient condi t ion of their being injuredm ore seriously. T o say that something co uld ha ve been otherwise, inthis sense, is to say that under certain roughly specifiable conditionsit would have been otherwise.

In this very natural sense, the basic nomological facts could not havebeen otherwise. For there is nothing by which we may explain theirbeing the way they are , nothing which, had i t been otherwise, wouldhave led to different results.

It might be objected that, on the account I have given of Spinoza,this is not true. There is something which, had it been otherwise, wouldhave led to different results. The attributes (the basic nomological facts)are a sufficient cond ition of their infinite m odes (the deriva tive n om ologi-

cal facts), even if they are not a sufficient condition of their finite modes

114





Necessity

( the s ingular facts) . But wherever p is a sufficient condition of q q isa necessary condition of p. So we shall hav e to say th at th e derivativenomological facts are a necessary condition of the basic nomologicalfacts. That is, if the derivative nomological facts had been different,the basic nomo logical facts wou ld ha ve been differ ent .

I reject the prem ise tha t wherev er p is a , suff icient condit ion of qq is a necessary condition of p. This is, I realize, a very common view.Bu t it is an abs urd one. Con sider the case of th e accide nt. W e said

that , under the circumstances, the fact that they were wearing a seat-bel twas a sufficient condition of the fact that they were not injured moreseriously. Shall we say that the fact that they were not injured more seri-ously was, therefore, a necessary condition of the fact that they woretheir seat-belts? But it was not a condition of their wearing their seat-beltsat all . T hi s is simply a case w he re the c om mo n philosoph ical usag eis perverse. One fact cannot be a condition of another, in the senserequ ired here, unless it is in som e wa y (logically or tem po rally) prior

to the other.There i s a more fundamenta l objec t ion to my argument . I have ex-plained the notion of necessity, in the sense in which I wish to applyit to th e basic nom ological facts, in term s of cou ld not have been o ther -wise. A nd I have suggested tha t som ething could not have been other-wise, in this sense, if there are no circumstances under which it wouldhave been otherwise. It will no doubt be objected that the conceptswh ich I h av e used to explain necessity are at least as obscu re as theconcept of necessity itself. In particular, I have used counterfactual con-ditionals, and these are widely thought to be very troublesome. I havenot, to use the Port-Royal terminology, given a definition of necessitythat provides us with a clearer and more distinct idea of the thingdefined.

This general l ine of argument will be familiar to any reader ofW . V . O . Qu ine ' s T w o Dogm as of Em pir ic ism, and I am af ra id thereis no way out of the kind of di lemma Quine presents . When we under-take to explain a modal concept, l ike necessity, we must use either terms

which are unobject ionable, but not logical ly s trong enough to do the

1 1 5






job, or terms which are logical ly s t rong enough, but equal ly di ff icul t .(Or near ly so . Counterfactual condi t ionals are not qui te as obscure ascou ld have been o therwise . In Engl i sh , cou ld hav e be en is ve ry

a m b i g u o u s . I am not sure what the moral of th is s tory is . Perhapsw e shou ld give up us ing m od al terms. O r perhap s w e shou ld revise ourexpecta t ions as to what a phi losopher ' s explanat ion can achieve.

A t any ra te , if the indepe nde nce of the bas ic nom olog ical facts , andnot thei r suppose d analyt ic t ru th , is take n as the gro und for ca l l ing

them necessary, then the object ion made ear l ier about the necess i ty ofthe in f in i t e modes can be dea l t wi th . The p rob lem was tha t wha t fo l lowsfrom something which is logical ly necessary must i t se l f be logical lynecessary. But what fo l lows f rom, or depends on, something which isnecessary in the sense of being absolutely independent is not i tselfnecessary in that sense . Qui te the contrary.

There will , on this view, be three species of necessity. The first isthat possessed by the most general nomological facts , those const i tu t ing

the nature of the divine a t t r ibutes . These , s ince they do not and cannotowe their existence to anything else, possess a necessity which is absolute,unco ndi t ion al , an d a tem pora l . T h e necess ity of the less gene ral n om o-logical facts, by contrast , is relat ive and condit ional but st i l l atemporal .They do owe thei r exis tence to something e lse , but that to which theyowe their existence, the most general facts, is something which is i tselfabsolutely necessary. The necessity of the singular facts is both relat iveor condi t ional and temporal . For they a lways owe thei r exis tence bothto nomological facts , whose exis tence is not subject to temporal var ia-t ion, and to other s ingular facts , which come into being and pass awayas the condit ions of their existence are and are not realized. As Spinozaputs it,

Some things , then, are t ransient in themselves ; others, indeed,are not transient because of their cause. But there is a third which,through i ts own force and power a lone, i s e ternal and imper ishable .The f i rs t are a l l the par t icular th ings which have not exis ted f rom

al l t ime, or which have had a beginning; the second are a l l the

11 6





ecessity

[general ] modes which we have sa id are the cause of the par-t i cu la r modes ; and the th i rd is God, or what we take to be onea n d t h e s a m e , t h e Tr u t h . K V , I I , ν , 1 : 6 2 - 6 3 )

In the l ight of what has been said earl ier, this requires l i t t le comment.I t wi l l , perhaps , be objected that the exis tence of the bas ic nomological

facts i s s t i l l hardly se l f -explanatory. But th is complaint s tems f rom afai lure to apprecia te suff ic ient ly the di fference between the two ser ies

of causes the infinite series of finite causes and the finite series of infinitecauses . A s even St . T h o m a s adm it ted, there is no reason, apar t f ro mthe wo rd of sacred Scr ipture , to suppose that for m er series should ha vea f irst m em ber. Bu t it is most implausible to suppose that the exp lana t ionof laws could proceed ad inf ini tum, pass ing a lways to more generalp r inc ip les under which the lower l eve l ones may be subsumed . And i fwe mus t s top somewhere , then when we reach tha t po in t , the demandfor fur ther explanat ion is unreasonable .

T o say th is is to accept the view th at u l t imate ex plan at ions of asort are in princip le possible. I t is no t to say tha t any e xp lan atio n w enow have is u l t imate , or that an ul t imate explanat ion must be in tui t ivelyev iden t . W e ma y wel l agree wi th Po pp er tha t

i t is not truisms which science unveils . Rather, i t is part of thegreatness and beauty of sc ience that we can learn , through ourown cr i t ica l invest igat ions , that the world is u t ter ly di fferent f romwhat we ever imagined—unt i l our imagina t ion was f i r ed by the

refutation of our earl ier theories,17

without accept ing his conclus ion that the process of conjecture and refu-ta t ion has no natural end. Thus we might , wi th the pragmat is ts , th inkof sc ience as a progress ive act iv i ty, whose explanatory theor ies converge,as toward an ideal l imi t , upon that sys tem of proposi t ions which const i -tu tes the t ru th about the world in which we l ive .

I conclude, then, that there is , wi thin the f ramework of th is in terpreta-t ion, a way of making sense of the eternity, or necessary existence, of

G o d .

117





Jus t w ha t Spinoza m ean t by these ideas an d how he conceivedthe mental correlates of bodies and of their motions and changesis a question th at ha s provo ked d ifferen ce of opinion . Since the reare ideas of inanimate as well as animate objects and processesideas can scarcely be regarded as individual psychical entities. . . N or since fo r Spino za th e infinite intellect of G od is im-person al can they be the thou gh ts of a self-conscious m ind re-flecting upon the order and connection of physical events. One

is tem pted rath er to think th at the idea m eant for h im w ha twe should call the t rut h a bou t each par t icular physical event .

—Ful le r istory of Philosophy

The Div ine At t r ibu tes





T h e inte rpretation presented so fa r was designed originally to answ ercertain questions about substance and its modes w hat is substance?what are its modes? how are they related? And I think I can givea tolerably clear and plausible answer to these questions. But if thereis anything in this app roac h to Spinoza, it ought to be ca pable of beingapplied to other puzzling areas in his system, and in particular to theproblem of the relation between the attributes. In what follows I shalltry to show how I believe this can be done.

There are, of course, really two problems about the relation betweenthe a ttributes of Spinoza s substance. The re is the problem of un de r-standing how Spinoza conceived of the relation between thought andextension, and there is the problem of understanding how he conceivedthe two known attributes of thought and extension to be related tothe infinitely many unknown attributes.

Thought and ExtensionSince, as we might exp ect, Spinoza gives us more inform ation abo utthe two known attributes than about the infinitely ma ny unkno wn at-tributes, it will be best to begin by considering their relation. First Ishall sketch what looks like a promising way of extending the line ofthou ght so fa r developed. T he n I shall try to show why it w on t do.A nd finally I shall try to work out an alternative which takes betteracc oun t of the peculiarities of Spinoza s philosophy.

Consider the following proposal. Spinoza, I have suggested, built onthe possibility of a unified science of extended objects. There are certainfundamental laws of nature governing the behavior of all extended ob-jects. Th e various changes in the objects can all ultimately be understoodin terms of these laws. Analogously, we might suppose that he recognizedas well the existence of nonextended objects, thoughts, volitions, sensa-tions, and the like, and that he envisaged the possibility of a unifiedscience of nonextended objects, a universal psychology, strictly parallelto the universal physics we have been imagining, with fundamental laws

9






of psychic events, derived laws, and nomological explanations of indi-vidual psychical phenomena.

We might be s truck by the fact that when, in the Theological-PoliticalTreatise, Spinoza wan ts to give examples of laws wh ich depe nd on na tu -ral necessity, he gives one exa m ple fr o m physics, a law of im pa ct, an done example from psychology, a law of association. And we might arguetha t par t I I I of the Ethics, which discusses the origin and nature ofthe human emotions is an at tempt to work out the rudiments of this

universal psychology.Co nfro nted with Sp inoza s not ion that th e mod e of extension and

the idea of that mode are one and the same thing, expressed in twodifferen t ways ( E I I P 7 C S ) , we might appeal to the currently p opu larview that mental events are contingently identical with brain events,that a description of what goes on in us in psychological terms hasa d ifferent sense than , but th e same reference as, the correspondingdescription in physiological terms. And clearly, there are many passages

in Spinoza—particularly those in which he connects the vitality of themind with the vi tal i ty of the body—which would tend to support thissort of approach.

But al though the approach described might take us a certain distance,I d on t believe th a t we can get very fa r with it . It a ttribu tes to S pinoz aa view wh ich is, perha ps, at t ract ive an d plausible, probab ly mo re p lausi-ble than the one I shall argue for, but i t is in many respects a misleadingmod el for interpre t ing him. T o m ention only one of i ts most pro m inen tdefects, nothing in this l ine of interpretation would suggest that forSpinoza there should be a mo de of tho ug ht fo r each mo de of extension.In fact, this l ine would suggest rather the opposite.

T h e view tha t sensat ions, thoug hts , an d so for th are brain processesimplies that where you do not find the appropriate physiological struc-tures, you cannot expect to find any mental processes either. Even thoughmy watch is a fairly complicated piece of machinery, you would notexpect i t to have any sensations. But plainly, Spinoza does think thatevery mode of extension has its corresponding mode of thought, or

ra ther, is i ts correspo nding m od e of thou gh t :

1 2





The Divine Attributes

T h e ord er an d connection of ideas is the same as the order an dconnection of things ( E IIP 7) . . . L et us here recall w ha t ha sbeen proved above, viz., that whatever can be perceived by the in-finite intellect as constituting the essence of substance pertains toone substance only. Therefore, thinking substance and extended sub-stance are one and the same substance, which is now grasped underthe one at tr ibute and now un der the o ther. So also a mod e ofextension and the idea of that mode are one and the same thing,

bu t expressed in two ways . . . Fo r exa m ple, a circle existing inna tur e an d the idea wh ich is in God of an existing circle are onean d the same thing, but explained th roug h different at tr ibutes.

E I Ip7S)

If my watch, or as I should prefer to say, the fact that my watch hassuch and such a character, is a mode of extension, then there is a modeof tho ug ht wh ich is a d ifferent expression of the sam e th ing.

This is, to say the least, puzzling. It is almost enough to make onedespair of making sense of Spinoza. But before we despair, let us tryano ther tack. W e hav e to face the fact tha t Spinoza intends his terms

ide a and mo de of tho ug ht to be understood in a way for whichDescartes and Locke have not prepared us.

T h e first thing to do is to recall wha t we noted earlier abo ut Spinoza'sideas involving an elem ent of affirmation . T his is som ething th at Spinozagoes to great pains to emp hasize. In defining the term id ea as a con -cept of the mind, which the mind forms because it is a thinking thing( E IID 3), he explains tha t he uses the term con cep t rath er tha n per-cep tion because the latter term suggests wrongly tha t the mind is passivein relation to its object, not active. Later he enlarges on this when hew arns his readers to distinguish carefully between an idea an d an imag e.Those, he says,

who think that ideas consist in images which are formed in usthrough contact with [N S external] bodies, persuade themselves

that those ideas of things [N S which can m ake no trace on our

1 2 1






brains, or] of which we can form no similar images, are not ideas,bu t only inventions, w hi ch w e feign from a free choice of the will ;so they think of ideas as of mute pictures on a tablet , and preoccu-pied with this prejudice, they do not see that an idea, insofar asit is an idea, involves affirmation or negation. (E IP49CS, I I : 132

So ideas, for Spinoza, involve an activity of the mind, an activity ofaff i rmat ion or negat ion. Hence, i t would be appropriate in most contexts

in Spin oza to substitute the term pro po sition or assertio n for theterm id ea . By contrast , this wo uld not be app ropr iate in most contextsin Descar tes or Locke.1 Spinoza's ideas are the sort of thing that canfollow from and entail one another, that can be true or false in thetraditional sense of agreeing with their object (E Ia6 .

It is customary, of course, to attribute to Spinoza a coherence theoryof truth. But the reader who feels that I am placing more weight onE Ia6 than that axiom can comfortably bear may f ind i t worthwhile

to reflect on the passage in the Metaphysical Thoughts in wh ich Spinozadiscusses the m ean ing of the terms tru e and fal se. T h e passagenot only supports a correspondence theory of truth, but also is curiouslycontem porary in its m ethodology. S pinoza begins by rem arking th at

Since the multi tude first invent words, which are afterward takenover {usurpantur) by the philosophers, i t seems app ropr iate forone who seeks the original signification of any word to ask whati t f i rs t denoted among the mult i tude—part icular ly where there isa lack of any other explanations which, from the nature of the lan-guage, could be brought forward for investigating the signification.{C M I , v i , 1: 2 4 6

W e m ight note in passing that Spino za works on the same m ethodologi-cal principle in the preface to part IV of the Ethics, where he undertakesto give an accou nt of the m ean ing of the terms pe rfec t and im per-fec t and go od and ev i l . Sp inoza cont inues :

122








as a proposit ion or idea, bearing logical relations to other proposit ions,is to conceive the m od e unde r the a t t r ibute of thou ght . T o ta lk a bo utit as a fact , hav ing causa l relations with oth er facts, is to conc eive i tunder the a t t r ibute of extension. But the proposi t ions which make upthe set that gives a complete and accurate description of the worldare identical with the facts they describe, and the causal relations be-tween the facts have their counterpar t in the logical re la t ions betweenproposit ions.

T hi s is ho w I propose to read Spinoza 's doctr ine that the o rderan d con nec tion of ideas is the same as the order an d co nn ection ofth ings (E I IP 7 ) . Spinoza thinks that th is proposi t ion is obvious f romE I a4 , accord in g to wh ich kno wled ge of an e ffec t depends on andinvolves kno wled ge of it s cau se. For, he says, th e idea of an ythin gwhich is caused depends on the knowledge of the cause whose effecti t i s ( E I IP D ) . This is to say that wherever you have two facts s tandingin a causal relation you also have two proposit ions standing in a logicalrelation, the proposit ion describing the cause entail ing the one describingthe effect . T h e set of true proposi t ions is the wo rld con ceived un derthe a t t r ibute of thought ; the se t of facts , the wor ld co nce ived un derthe attr ibute of extension. But the two worlds are not two, they are one.

So far, I suspect , my explanat ion of what Spinoza means must seemquite as mysterious as the doctrine i t is designed to explain. But I thinkI can enlarge on what I have said in a way that wi l l be helpful . AsI am here using the notions of fact and proposit ion, proposit ions arerelated to facts in roughly the way that , in Aristotle, form is relatedto mat ter. The analogy runs as fol lows for Aris tot le , a concrete objectis a un i ty wh ich is separable f or purposes of tho ugh t in to tw o elements ,the form, or universal e lement , which is capable of character iz ing manyobjects , and the mat ter, or par t icular e lement , which makes the objectthis par t icular object . Let us say that a concrete s i tuat ion is a lso ana-lyzable in to two elements , an abst ract pat tern which can character izemany s i tuat ions and a par t icular iz ing e lement which makes i t this pa r-t icular s i tuat ion. Th us , i f i t is a fact that M ar y is m ak ing pies , then

this fact is a particular instance of a type of si tuation that could have

124





The Divine ttributes

many ins tances . There are , or could be , many s i tuat ions of which thesen tence M a ry i s m ak ing p ies m igh t be t ru ly a ff i rmed . Ca l l the abs trac tpat tern common to a l l these the proposi t ion. Cal l the par t icular iz inge lement the fac t .3

There is a d i ff icul ty about th is analogy. The re la t ion of form to mat terin Aris tot le i s a many-one re la t ion. A plural i ty of forms may character izeone and the same object . The same might be sa id of the re la t ion betweenproposi t ions and facts . For normal ly we should wish to say that one

and the same fact might be t ru ly descr ibed by a number of d is t inctproposi t ions . For example , a t one s tage dur ing the process of Mary 'smaking her p ies , i t was t rue that

( ι ) M a r y is rota ting the dial on the stove an d,( 2 ) Sh e is tur nin g on the electrici ty in the ov en an d,( 3 ) Sh e is start in g to he at the o ve n.

These sentences are not logical ly equivalent , and so express di fferentproposi t ions . But many people would have a s t rong temptat ion to say

that they are al l t rue in virtue of one and the same fact .4 W h a t i s

di ff icul t abo ut th is is that S pino za 's sys tem requires the re la t ion betw eenmodes of thought and modes of extension to be a one-one re la t ion.

In this case I am not sure that our instinct to speak of three differentdescr ipt ions of the same fact i s correct . Mary might have rota ted thedia l wi thout turning on the e lect r ic i ty ( i f there had been a faul ty connec-t ion) ; and she might have turned on the e lect r ic i ty wi thout rota t ingthe dia l ( i f the c i rcui t had been c losed, say, by her act ivat ing a photoelec-t r ic ce l l as she approached the s tove) . So i t i s not c lear to me thatthe truth -con dit ion s of ( ι ) ar e the sam e as the truth-c ond it ions of 2 ) .( ι ) is rendered true by something she did to the dial . 2 ) is re nd eredtrue by the fact that when she did th is to the dia l , something e lse hap-pened—namely, the c i rcu i t was c losed .

T h e quest ion W h a t are the ident i ty cond i t ions for fac ts? is a ver yobscure one . I th ink a g rea t many peop le who have he ld cor respondencetheor ies o f t ru th have supposed—righ t ly o r wrongly—tha t the re a re jus tas many facts as there are dis t inct t rue proposi t ions . Whether or not

this is so, i t is the sort of vie w w e ne ed fo r interpre ting Sp ino za . If

125






it does violence to our ordinary notion of what a fact is , so be i t .In interpreting Spinoza's doctrine that the mode of extension and

its idea are one as equivalent to the claim that facts and true propositionsare one, in the sense described, I am saying, in effect, that on thispoint Wolfson is r ight . Wolfson recommends that we understand Spi-noza's theory of the mind-body relationship as differing only in terminol-ogy from the Aristotelian view that the soul is the form of the body( I I , 36-64 ). A n d I think th at this is substan tially co rrect. It is no t

quite right, because we have to extend the notion of form and matterin the way done above in order to allow for Spinoza's emphasis onthe assertional element in ideas. And we have to make further adjust-ments to allow for the relation's being one-to-one. But only in a waylike this can we account for Spinoza's notion that there is a mode ofthou ght for every m ode of extension. Spin oza's statement tha t all thingsare anim ate, thoug h in different degrees , that there must be in G odan idea of everyth ing (E IIP13CS , does not imply that my watchhas thoughts and sensations, any more than Aristotle 's doctrine thatplants have souls implies that flowers feel pain.

Suppose that the fact that my watch is made of gold is a mode ofthe attribute of extension. Then the true proposition that my watchis made of gold is a mode of the attribute of thought. The true proposi-tion is the idea in thought of the mode of extension, it is its form,or, if you like, i ts mind. I think Spinoza would be will ing enough tosay this. When he says that the human mind is the idea of the humanbody (E I I P 1 3 , he is not credi t ing the human body with anythingthat any other mode of extension does not have.

But to say that for every mode of extension there is a correspondingmode of thought is not to attribute consciousness to all things. Spinozaimplies quite plainly that such things as stones do not possess conscious-ness (see Ep. 58, I V 66 ) . M y wa tch does not know tha t it i s m adeof gold . Ap pa ren tly i t is not a sufficient cond ition of consciousness th atthere exist in thought an idea of a mode of extension ; it is necessaryalso that there exist in thought an idea of that idea. Knowledge or

consciousness arises only where the mind contains not merely ideas ofthe affections of the body, but also ideas of those ideas.

126






This point is not stated very clearly in Spinoza, but i t does comeout, I think, on a close reading of E IIPI 1-30. For example, inE H p 12, Spino za says:

Whatever happens in the object of the idea constituting thehuman mind must be perceived by the human mind, that is , anidea of the thing will necessarily exist in the mind. In other words,if the object of the idea constituting the human mind is a body,

nothing can happen in the body which is not perceived by themind [Ν S : or of which there is not a thought in the soul].

W e are told by the follow ing proposition that the objec t of the ideaconst i tut ing a human mind is a bod y— an d, m oreover, i t appears f romE IIPI6C2 that the object of the idea const i tut ing a human mind isthe body which is uni ted with that mind : T h e ideas w e have of externalbodies indicate the constitution of our own body more than the natureof extern al bo die s. In term s of the current interpreta tion, this seriesof doctrines could be stated in the following way : my body is a setof facts, my mind a set of propositions describing those facts; my mindmust contain a proposition corresponding to every fact that constitutesmy body, for the propositions simply are the facts, considered in a differ-ent way.

One might infer from this the surprising doctrine that the mindis aware or conscious of everything that goes on in the body. Thatseems, on the fac e of i t , to be w ha t E H P 2 is saying. If so, doctors'

jobs would be a good deal easier than they are . Your doctor wouldnot have to take a blood test to see if your white corpuscle count wasd o w n ; he could simply ask you. This is absurd, and we should requirevery good evidence before we attribute such a doctrine to Spinoza.

In the first place, let me say that I think some significance may beattached to the fact that in the wording of E H P 2 Spinoza uses theverb pe rce ive rather than k n o w . T hi s suggests the possibili ty tha t,a l though nothing can happen in the body which the mind does not

perceive, there may be things happening in the body of which the mindis not aware or conscious.

127






This suggestion is confirmed, I think, in E H P 19 when Spinoza saysthat the hum an mind does not know the hu m an bod y itself, nor doesit know that the body exists, except thro ugh ideas of affections by wh ichthe bod y is af fec te d. I take this as m ean ing tha t it is not simply inhaving an idea of an affection of the body that the mind knows thebody—rather i t is by means of that idea that the mind knows the body.

When we say that the human mind perceives this or that, we saynothing but that God has this or that idea, not insofar as he is infinite,

but insofar as he is explained through the nature of the human mind,or insofar as he constitutes the essence of the hu m an m in d (E I I P I I C )

To say that God has this or that idea is simply to say there isin God a mode of extension which may also be regarded, under theattribute of thought, as a mode of thought, that is, a fact which mayalso be regarded as a proposition. It is only when something else isad ded to the m ind's pe rce ptio n of the bo dy tha t consciousness arises.

The nature of this something else is explained in E IIP20-23, where

Spinoza introduces the concept of an idea of an idea. E IIP22, forexam ple, tells us that the hu m an m ind perceives not only the affectionsof the bo dy , bu t also the ideas of these affe ctio ns , tha t is, tha t itcontains not only ideas of bodily modifications, but also ideas of theseideas. A n d E IIP 23 says that the mind does not know itself exceptinsofar as it perceives the ideas of the affection s of the bo d y, tha tis, excep t insofar as it has ideas of ideas. W e can e qua te hav in g a nidea of a n idea with bein g conscious. E H P 19 and E IIP23 together,then, say that the mind does not know the body except insofar asit perceives itself, or has ideas of ideas, or is conscious. It is worth notingin this connection that, wh ile every individu al thing has a m in d con-taining ideas of the affections of i ts body (E IIP13S), the existenceof ideas of ideas is proven only for hum an m inds (E IIP2O) . I inferfrom this that, although Spinoza is willing to assert that everything isanimate (in a very odd sense of the term), he is not prepared to saythat anything except a human being is conscious.

It may be helpful to think of an idea of an idea as a proposition

128






about another proposition. Since I am identifying the possession of anidea of an idea with consciousness, it seems natural to say that an ideaof an idea is a special kind of proposition about a proposition, namely,one expressing what is sometimes called a propositional attitude (forexample, A knows that p ).

As soon as we formulate it this way, however, we seem to run intodifficulty. The idea that is the object of my idea of an idea will haveas its ideatum a modification of my body and it will agree with that

modification. This amounts, on my interpretation, to saying that whatI know when I know that something is the case is a proposition whichdescribes some fact about my body. And this would certainly be a verystrange doctrine. For surely we would want to say that relatively fewof the propositions I know are propositions describing facts about mybody. How is it then that I can know propositions which describe factsabout bodies other than my own?

Again, this probem does not arise simply because of my peculiar

w ay of form ulatin g Spino za s doctrines. It is there in Spino za,5

as thefollowing passage shows:

W e understand clearly wh at is the differen ce between the ideaof Peter which constitutes the esssence of the mind of Peter andthe idea of Peter which is in another man, say Paul. For the formerexplains directly the essence of the body of Peter himself, and doesnot involve existence except as long as Peter exists the latter, onthe other hand, indicates the constitution of the body of Paul morethan the nature of Peter. (E I IP17CS, I I : 1 0 5 - 1 0 6

We can frame essentially the same question in Spinoza s own language :How is it that the idea of Peter which is in Paul is appropriately de-scribed as an idea of Peter when what it agrees with is a modification ofPau l s body?

Readers who come to Spinoza from a study of the British empiricistswill naturally feel tempted to answer this query by saying that Paul s

129






idea of Peter, though it may in some sense correspond to a modificationof Paul 's body, is properly an idea of Peter because i t is a representationof Peter, a mental image of him. Spinoza quite explicit ly rules this outin a number of places. Ideas are not images. Indeed, Spinoza is will ingto app ly the term im a ge only to certain m odifications of the bo dy ,an d ev en these he does not think of as pictures of som ethin g exte rna l :

T o keep the customary voca bulary , we shall cal l the affect ions

of the human body whose ideas represent external bodies as presentto us images of things, although they do not reproduce the formsof things. (E I I P 1 7 S )

By ideas I understand, not images of the sort formed at thebase of the eye, and if you please, in the middle of the brain,but concepts of thought [N S or the objective essence of a thing,insofar as it exists only in thought]. (E IIP48S, cf . E I IP49S)

Clearly we cannot think of Paul 's idea of Peter as a mental image ofPeter, or even as the mental correlate of a bodily image of Peter. Buthow, then, is it an idea of Peter?

Our question has an answer, if not a satisfactory one. E I I P 1 6 - 1 7gives us the justification for saying that an idea that ind ica tes theconstitution of Paul 's body is an idea of Peter the idea in Paul ' s mindhas as an object a m odif icat ion of Paul ' s bod y that invo lves the natureof Peter. But what is i t for a modification of Paul 's body to involvePeter 's nature? W e get a c lue in E II P1 7D :

As long as the human body is affected [ in a way which involvesthe nature of some external body] , so long wil l the human mindcontemplate this affection of the body, that is , i t will have an ideaof an actually existing mode which involves the nature of the ex-ternal body, an idea which does not exclude the existence orpresence of the nature of the external body, but asserts it.

From this we can say that when a modif icat ion of the human bodyis caused by an external body (or one of the modifications of that body),

130






the modification of the human body involves the nature of the externalbody, and the idea of this modification asserts the existence of the natureof the external body (in addition to agreeing with i ts own immediateo bj ec t) . Th is is a bit vagu e, to be sure, bu t there is furthe r help later on :

The ideas of the affections of the human body involve the natureof external bodies as well as the nature of i ts own body E I I P I 6 ) ,and they must involve not only the nature of the human body,

but also the nature of i ts parts, for the affections are ways in whichthe parts of the human body, and consequently, the body as awhole, are affected (E II post . 3) . But (E IIP24-25) an adequateknowledge of external bodies and of the parts composing the humanbody does not exist in God insofar as he is considered as affectedby the human mind [alone] but [only] insofar as he is affectedby other ideas [as wel l ] [N S: that is (E II P 1 3 ) , this knowledgeis not in God insofar as he constitutes the nature of the human

mind]. Therefore, these ideas of affections, insofar as they are re-ferred to the human mind alone, are l ike conclusions withou tpremises. T h a t is, as is self-eviden t, they are co nfu sed ideas.( E I I P 2 8 D )

This doctr ine may be read in the fol lowing way : among the propositionsin A, there are some which describe the body of Paul ; at least someof these propositions describing facts about Paul will not be deduciblesolely from other propositions describing facts about Paul (taken together

w ith the laws of na ture ) ; these propositions will require for their d edu c-tion at least some propositions describing facts about bodies externalto Paul 's . Suppose that one such proposition, say p, requires as a premisefor i ts deduction from the laws of nature the proposition q, where qdescribes some fact about Peter 's body. In such a case we shall say thatp involves q, which describes Peter 's nature8 and that therefore p'"asserts" (poni t the existence of the nature of Peter 's body. That is ,it affirms the existence of som ethin g of th at na ture . In this instan ce,

p would be an idea which primarily indicates the constitution of Paul 'sbody, but is also an idea of Peter.

3






If this is a correct account of how Spinoza thought that we haveknowledge of the external world, then we can begin to see why heregarded that knowledge as necessarily inadequate. We have ideas ofthe external world in having ideas of modifications of our bodies whichcannot be understood solely through other modifications of our bodies,but only through the modifications of other bodies. And to say thisis ipso facto to say that these ideas a re inad equa te :

The human mind is a part of the infinite intellect of God :and therefore, when we say that the human mind perceives thisor that, we say nothing but that God has this or that idea, notinsofar as he is infinite, but insofar as he is explained through thenatu re of the hu m an m ind, o r insofar as he constitutes the essenceof the human mind. And when we say that God has this or thatidea, not only in sofar as he constitutes the nature of the h um anmind, but also insofar as he has, with the human mind, the ideaof some other thing, then we say that the human mind perceivespartially or inadequately. (E I I P I I C

To put this in my terms the hum an m ind is a proper subset, M , ofthe totality of true propositions, A ; to say that the mind perceives pis to say that p is a mem ber of M ; w henever p cannot be deducedfrom other members of M alone, but only from them in conjunctionwith mem bers of A w hich are not mem bers of M , then p is an inadequateidea. Since it is necessarily the case, whenever p is an idea of an external

body, that p cannot be deduced from other members of M alone, senseperception must give inadequate knowledge. But in God, all ideas areadequate. That is, the totality of true propositions, A, forms a system.Ev er y proposition in it except the axiom s of scientific theo ry) followsfrom other propositions in the system.

Thus when Spinoza wishes to show that our knowledge of the durationof our body can only be very inadequate E I IP 3 0 ) , he argues in thefollowing way:

132






ι . T h e duration of our body doesnot depend on its essence

E HA I ) , nor on the absolutenature of G od Ε IP2 ) .

2. R at h er it is determined to existand act by [N S other] causes, ofa kind which are also determinedby other things to exist and act ina fixed and definite way, and thesethings by others, and so on, toinfinity E IP28) .

3. Therefore, the duration of ourbody depends on the commonorder of nature, and on theconstitution of things.

4. Moreover, of the reason whythings are so constituted, there isan adequate knowledge in God,ins ofar as he has the idea of allthings, and not only insofar as hehas the idea of the hum an body

. E U rg e ) .

5. Therefore, the knowledge ofthe duration of our body is quiteinadequate in God insofar as heis considered as constituting onlythe nature of the human mind,that is, the knowledge of theduration of our body is quiteinadequate in our mind

E I I P I I C ) .

1 a The fact that a exists, where ais a human body, is not absolutelynecessary, nor does it follow fromthe laws of nature alone.

2 a. R at he r the fa ct that a exists,is only relatively necessary, that is,necessary given the nom ologicalfacts in conjunction with) othersingular facts, each of which isitself only relatively necessary.

3a. The fact that a given humanbody exists, therefore, depends

partly) on the existence andnature of other finite things.

4a. A complete explanation of anysingular proposition in A is possiblefor one who knows all other

prior) singular propositions in A,but not for one wh o knows onlysome proper subset of them.

5 a. Th ere for e, the p ropositionthat a exists does not follow fromthe subset of A which constitutesthe mind of a.

3 3






I n t h e f o l l o w i n g p r o p o s it i o n ( E I I P 3 1 ) , S p i n o z a d e m o n s t r a t e s , b y t h esame l ine o f r ea son ing , t he same r e su l t fo r ou r knowledge o f t he du ra t ionof a l l ex terna l s ingular th ings .

N o t a l l k n o w l e d g e , of c o u r se , is i n a d e q u a t e i n t h is w a y . W h e n e v e rw e a r e d e t e r m i n e d e x t e r n a l l y t o p e r c e i v e s o m e t h i n g , o u r p e r c e p t i o n o fi t i s i na de qu a te . B u t some t im es , Sp ino za th inks , w e a re de t e rm ined in -t e rna l ly to have a ce r t a in idea , and then ou r i dea wi l l be an adequa teo n e ( E I I P 2 g s ) . T h e r e a s o n w h y th i s is p o s s ib l e, a p p a r e n t l y, is t h a t

t h e c o m m o n n o t io n s , w h i c h a r e e q u a l l y in t h e p a r t a n d i n t h e w h o l e ,c a n n o t b e c o n c e i v e d e x c e p t a d e q u a t e l y ( E I I P 3 8 ) . T h e s e id e a s m u s tb e a d e q u a t e i n t h e h u m a n m i n d , n o t j u s t i n G o d . A n d w h a t e v e r f o l l o w sf r o m i d e a s t h a t a r e a d e q u a t e i n t h e h u m a n m i n d m u s t a l s o b e a d e q u a t ei n t h e h u m a n m i n d ( E I I P 4 0 ) . S o , w h e n e v e r a n i d e a i n t h e h u m a nm i n d f o l l o w s f r o m t h e c o m m o n n o t i o n s a n d f r o m o t h e r p r e v i o u s i d e a si n t h e h u m a n m i n d , t h a t i d e a w i l l b e i n t e r n a l l y d e t e r m i n e d , a n d c o n s e -q u e n t l y, a d e q u a t e .

T h i s a c c o u n t o f w h a t is r e q u i r e d f o r a n id e a t o b e a d e q u a t e m a yh e l p t o e x p l a i n w h y S p i n o z a a t t a c h e d t h e e t h i c a l i m p o r t a n c e h e d i dto the possess ion of adequate ideas . For to the ex tent tha t our ideasa re in t h i s way in t e rna l ly de t e rmined , we a re t he adequa te cause o fo u r a f f e c t i o n s E I I I d i , a n d h e n c e , a c t i v e r a t h e r t h a n p a s s i v e E

I I I d 2 . A n d i t i s the ac t ive l i fe w h ic h cons t i tu tes v i r tu e i t se l f ( EI V P 2 2 - 2 4 ) .

Bu t t he accoun t i s no t w i thou t i t s d i f f i cu l t i e s . Fo r one th ing , i t i sn o t e as y t o s e e w h y S p i n o z a s h o u l d h a v e s u p p o s e d t h a t i d e a s w h i c ha re in th i s s ense ade qu a te shou ld a l so be ad eq ua te in t he sense o fpossess ing a l l the proper t ies , or in t r ins ic denominat ions , of a t rue idea

E I ID4) , t ha t i s , t ha t t hey shou ld be c l ea r and d i s t i nc t . Sp inoza seemst o b e w o r k i n g w i t h t w o d e f i n i t i o n s o f a d e q u a c y, a n d I k n o w o f n oa rg u m e n t t o s h o w t h a t o n e m a y b e r e d u c e d t o t h e o t h e r.

M o re im po r t an t , pe rh aps , i t is no t ea sy to see ho w , on the in t e rp re t a -t i on I have g iven o f t he r e l a t ion be tween though t and ex tens ion , t hesys tem can a l low for the ex is tence of fa l se ideas . I f a t rue idea s imply is

134






the fact it correctly describes, a false idea must be one which describesno fact, a mode of thought which has no corresponding mode of exten-sion. But like form without matter, that is an impossibility.

I am not sure whether this is a difficulty in the interpretation orsimply a difficulty in the system. There is at least one sense in whichSpinoza does deny the existence of false ideas. In E IIP32D, he arguesthat Ev ery idea which is in God agrees entirely with its [N S objectand] ideatum (E I IP 7C ), and therefore, every idea which is in God

is true (E I A 6 ). Since whatever is is in God E I P 1 5 ) , and thereforeevery idea is in G od , this would appear to entail that every idea, withoutqualification, is true.

Th ere is some problem in knowing wh at to make of the phrase whichis in God . I suspect that what Spinoza has in mind is something likethis. Every idea, or at least every idea which is an idea of an externalobject, may be considered in either of two w ays— either in relation towh at w e might call its primary object, the mode which is its correlate

in the attribute of extension, or in relation to its secondary object, themode wh ich caused its extensional correlate. Th us Pau l's idea of Peterhas as its primary object a modification of Paul's body and as its secon-dary object a modification of Peter's body. Considered in relation toits prim ary object, that is, as an idea in G od , it is, necessarily, true.But considered in relation to its secondary object, that is, as an ideain us, it need not be true. Th is interpretation of the phrase ide a whichis in G o d seems to be confirmed by a passage in the Short Treatise:

Between the idea and the object, there must necessarily be a union,since the one cannot exist without the other; for there is nothingof which there is not an idea in the thinking thing, and no ideacan exist unless the thing exists . . . Bu t it is necessary to notethat we are here speaking of the ideas which arise necessarily inGod of the existence of things, together with their essences, notof the ideas which the things now actually present to us andproduce in us. The two differ greatly. For in God the ideas do not

3 5






arise as they do in us, from one or more of the senses, throughwh ich we a re nearly always very imp erfectly affec ted by the things,but from their existence and essence, that is, from all that they are.( K F , I I , X X , I : 9 7 .

But even if we grant that Spinoza does want to make a distinctionof this sort betw een ideas in G o d an d ideas in us , i t is sti ll notclear wh at a cco unt we are to give of false ideas. I suggested abo ve t h at

Paul 's idea of Peter was an idea of Peter in virtue of the fact thatthe modification of Paul 's body which is i ts primary object was causedby the modification of Peter which is i ts secondary object; Paul 's ideanot only characterizes a modification of his body, but also affirms theexistence of som ething else of a sort sufficient to pr od uc e tha t mo difica-t ion. Th is does not leave mu ch room for error. I f it is an idea of Peter ,that is, if the corresponding modification of Paul 's body was causedby a mo dification of Peter 's b ody , the n th e thing whose existence it affirms

must exist. It is, therefore, a true idea both with respect to its primaryobject and with respect to its secondary object. If , on the other hand,there is noth ing external wh ich has pro duce d this pa rt icula r modif icat ionof P au l 's bod y, then it just is no t an ide a of an yth ing ex ternal to Pa ul.

This, again, seems to be a conclusion which Spinoza cheerfully ac-cepted. T h e imaginat ions of the m ind, considered in themselves, con-ta in no er ro r (E I I P1 7S ). T he re is noth ing posit ive in ideas on acco untof wh ich they are called false (E I IP33). Falsi ty, or error, for Spinoza,consists no t in o ur h av ing an id ea wh ich fails to agree w ith its obje ct,but solely in our not having as well other ideas of the object whichwe m ight hav e ha d. A character is tic exam ple which he gives is tha tof the error some are supposed to mak e in thinkin g the sun to be abo uttwo hund red fee t away f rom us (E I IP 35 S) . T he bas is for th is opin ionlies partly in the fact that our bodies are affected in a certain wayby the sun, that is, in the way in which the body typically is affectedby large, bright, distant objects, and partly in the fact that our bodiesare no t affe cted by the sun in such a way as to exclude the possibility

3 6






of certain distances. The error lies in the partiality of our idea of theobject, in our not being affected by it in as many ways as are abstractlypossible. Thus Spinoza writes, in the Short Treatise

When someone acquires a form or mode of thinking as a resultof an action on him by the whole of an object, i t is clear thathe has an altogether different perception of the form and natureof this object than someone else, who has not had so many causes

and so is m oved to aff i rm or deny som ething abou t the o bjectby a different and lesser action, since he has perceived the objectby means of fewer or s lighter affections. ( K V , II , xv , 1:7 9 ; c f .KV I I , xv i , 1 :83-84)

It is solely in terms of a difference in degree of awareness of the objectthat Spinoza undertakes to account for the difference between t rue andfalse belief about the object.

What seems to have motivated this view—apart f rom the internallogic of the Spinozistic system—is the observation that even after wehave corrected an erroneous judgment about a perceptual object wedo not alter our original perception of it. The sun still looks the sameto us even when we know that i ts distance from us is equal to severalhun dred times the diam eter of the earth. But Sp ino za s theory of errorstill does not strike me as a very satisfactory one. For it seems to obliteratethe distinction we should normally wish to make between error andignorance. Spinoza identifies believing that the sun is only two hundred

feet away with not believing that i t is much further away. But surelythe two states of mind are very different.

There remains to be discussed (at least) one major problem in inter-pret ing Spinoza s accou nt of the rela tion between m ind an d body, andthat is the notoriously difficult question of what Spinoza intended byhis doctrine of the eternity of the mind. This is ground on which theprud ent com m entator will hesti tate to tread , for there is some reasonto think that here Spinoza himself was conscious that he was not ex-

137






pressing his views as clearly as might be desired. Nevertheless, I thinkit may be possible to make some headway on the problem.

O ne wou ld no t have expec ted Spinoza to ho ld tha t the hum an m indcanno t be absolutely destroyed with the bo dy (E VP 2 3 ). Th e wholespirit of his theory of the relation of mind to body seems to be to estab-lish the closest possible connection between them, and one would natu-rally expect that when the body is destroyed the mind would perishwith i t. But Spinoza clearly though t otherwise. Som ethin g of the m ind

remains wh ich is e ternal (E V P 2 3 ). W e get a c lue to Spinoza 's inten-tions, I think, from the proof which follows this proposition :

In God there necessarily exists a concept or idea which expressesthe essence of the hum an b ody ( V P2 2 ), an d wh ich is, therefore,necessarily something which pertains to the essence of the humanmind ( I IP13) . But we a t t r ibu te to the human mind no dura t ionwhich can be defined by time, except insofar as i t expresses the

actua l existence of the bo dy , whi ch is expla ined thro ug h dura tion,and wh ich c an be def ined by t ime, that is ( I IP8 C) , w e cannotascribe duration to the mind except while the body has duration.Nevertheless, since this something is that which is conceived bya certain eternal necessity through the essence itself of God VP22 ) ,this something, which pertains to the essence of the mind, willnecessarily be eternal. (E VP23D)

The key here seems to me to be the third sentence which says, roughly,that the mind only possesses durational existence to the extent that itsbody does. Th is does not te l l us w ha t we wa nt to know . W e w an t toknow how the mind can possess nondurational existence apart fromthe body. But the citation of IIP8C is more helpful. IIP8 says that

the ideas of singular things, or of nonexistent modes, must be com-prehended in the infinite idea of God, in the same way that theessences of singular things or modes are contained in the attributes

of God.

138






The corollary goes on to add that

when individual things do not exist unless insofar as they are com-prehended in the attributes of God, their objective being or ideasdo not exist except insofar as the infinite idea of God exists; andwhen individual things are said to exist, not only insofar as theyare comprehended in God's attributes, but also insofar as they aresaid to have duration, their ideas also involve existence through

which they are said to have duration.Now the mind, considered as something which continues to exist afterthe destruction of the body, would be an idea of a nonexistent singularthing or mode. So it wo uld exist as something wh ich is, in Spinoza'sterminology, com prehen ded in the infinite idea of G od . A n d the ques-tion now is W h a t is it for an idea to exist as com prehen ded in theinfinite idea of G od ? In E IIP8 S, S pinoza tries to clarify this notionby appealing to an analogy which he confesses is not exact. But, he

says, I cannot give an exam ple wh ich will adequately explain the thing

I speak of here, since it is un iqu e. Still, the illustration he does giveseems to be of some help. Imagine a circle, ABCD in which the twostraight lines AC and BD intersect at a point, E. It is a general propertyof circles that all rectangles formed from the segments of such intersect-ing straight lines are equal, that is, the rectangle with base AE andheight EC is eq ual in area to the rectangle w ith base BE and height

ED.1 Sp inoza contends that in any circle

139






there are contained infinitely many rectangles equal to one another,though none can be said to exist except insofar as the circle exists;nor can the idea of any of these rectangles exist except insofar as it iscomprehended in the idea of the circle. Of these infinitely manyrectangles, let two only [that constructed out of AE and EC andthat constructed out of BE and ED] be conceived to exist. Theideas of these two rectangles also now exist not merely insofar asthey are comprehended in the idea of the circle, but also insofar

as they involve the existence of those rectangles. By this they aredistinguished from the remaining ideas of the remaining rectangles.

T h e view here seems to be this. O nc e a particular circle is given, itdefines a certain class of rectangles which could be constructed out ofthe segments of lines intersecting within it. For example, the squareof its radius places an upper limit on the area of such rectangles. Thesewill not all be equal in area. But if we take any one point within thecircumference of the circle as the point of intersection, this point willdefine a still narrower class of possible rectangles, all of which will beequal in area. The circle, we may say, contains these rectangles im-plicitly, in that the possibility of their construction is implied by thenature of the circle. This possibility is actualized whenever someonedraws two intersecting lines within the circle, though, of course, thereare infinitely many different ways in which it might be actualized. Thatis, there are infinitely many lines which could be drawn through thepoint of intersection, each of which would generate a rectangle equal in

area to, but of different dimensions than, the existing rectangles. Theexistence of the undrawn rectangles as implicit in the existence of thecircle is analogous to the nondurational existence of the ideas of non-existent modes as comprehended in the infinite idea of God. Theexistence of the drawn rectangles is analogous to the durational existenceof the ideas of things actually existing.

Such is the example Spinoza gives. And though its point may notbe readily apparent, still, if we translate it into the language I have

been using, it seems to me that we arrive at an intelligible view. Supposethat w e understand by the infinite idea of G o d the laws of nature.

140








I would suggest that Spinoza speaks here of a kind of actual existence,and not merely of possible existence, because he thinks of nonexistent-but-possible entities as being contained implicitly in something whichitself actually and eternally exists. (This is one respect in which theanalogy with the circle is inadequate—the circle does not exist eternally.

A n d he will, therefore, call the ideas of such things tru e, even tho ughthe things themselves never exist at any particular time or in any particu-lar place:

If an architect conceives a building properly [ordine], thoughsuch a building may never have existed, and may never exist,nevertheless the thought is true; and the thought remains the same,whether the building exists or not. (T d l E , 69 ,11 :26)

T h e use of the term tru e here strikes us as par ado xical, for it seems

to be cut off from the usual requirement of correspondence with reality.But I suspect Spinoza might insist that, just as the thought remainsthe same, so, in a sense, the building remains the same, whether it existsat a particular time and place or not. For just as the idea of the buildingis comprehended in the infinite idea of God, so the essence of thebuilding is contained eternally in the attributes of God (E I Ip 8c) .

T o say this is to imply th at even the hu m an bo dy is not absolutelydestroyed in death, that something of it too remains which is eternal,that the parallelism of the attributes which seemed to be threatenedby the doctrine of the eternity of the mind is, after all, maintained.A n d m any w ill feel this to be an ob jection, on the ground that, inSpinoza, only the mind is explicitly stated to be eternal. But this isa rather dubious argument from silence. For to the extent that Spinozajustifies his doctrine of the eternity of the mind by an appeal to EIIp8, he does commit himself to saying that what holds of the mindholds equally well of the body. I fail to see why he should have thoughthimself required to be any more explicit than in fact he is.

A m ore serious objection , I think, is this : Spinoza speaks only of

142






a part of the mind as being eternal if he were thinking of the ideasof things as bei ng co ntain ed in the idea of G o d in the ma nn er th atI have suggested, then there would be no reason for him to state hisdoctrine in this l imited way. For if the description of a particular build-ing, say, is consistent with the laws of nature, so that the infinite ideaof God contains the idea of that building, then the whole idea of thebuilding is contained in the idea of God, not just a part of i t . More-over, later in part V of the Ethics the doctrine of the eternity of the

mind is connected with knowledge of the second and third kinds andwith the apt i tude of the body to do many things. There (E VP38 and39) Spinoza says that the more the mind understands things by thesecond and third kind of knowledge and the more the body is fi t formany things, the greater will be the part of the mind which is eternal.The apparent implication is that a person could increase the portion ofhis min d w hich is e ternal by com ing to understand m ore an d m orethings sub specie aeternitatis and (Heaven he lp us ) by improving thefitness of his body.

These passages would not present a problem if Spinoza had saidmerely that, as our knowledge of our bodies sub specie aeternitatis in -creases, so also our consciousness of the eternity of the mind increases.To conceive a thing sub specie aeternitatis is to conceive it insofar asit is conceived through the essence of God {E VP30D), that is , to seeit as contained in an attribute of God and explicable in terms of thelaws of that attribute. To the extent, then, that we come to understanda particular state of affairs as an actualization of a possibil i ty perma-nent ly embodied in the "f ixed and eternal things," we view the ideaof that state of affairs as contained in the idea of God. If the state ofaffairs is one involving a part of our body, the idea of that stateof affa irs will be one of the ideas constituting ou r m ind , an d w e shallhave become conscious of a portion of our mind as something eternal.But Spinoza's language suggests that an increase in our understandinginvolves, not merely an increase in our consciousness of the eternityof the mind, but an increase in the eternity of the mind itself . And

this is a doctrine which, at present, I find completely unintelligible.






Of the Unknown Attributes of GodSpinoza tells us a fair amount about the attribute of extension, ratherless about the attribute of thought, and next to nothing about the in-finitely many other attributes of God. No doubt this is as it shouldbe, since he holds that they are at least unknown, and probably un-knowable .8 But i t is frustrating. For i t means that whatever we may sayin an attempt to understand what Spinoza intends by this doctrine,and why he believes i t , must be more than usually speculative. Nonethe-less, I think it may prove possible to say something useful on the subject.

W e begin by consider ing an analogy put forw ard by Bru nschvicg.The relation between substance and its attributes, he says, may bestbe understood in the following way :

A thought, in i tself one and indivisible, may be expressed withcomplete exactness in an infinity of languages. One could say withequal justice that no one of the translations contains the thought

and that they all manifest it in its entirety. Just as there is a perfectparallelism between the different texts which express one and thesame thought, so also there is an intimate and perpetual corre-spondence between the different attributes which proceed from oneunique activity. Ultimately, from the point of view of absolutereality, all these attributes are only one and the same thing.9

This is, as an analogy should be, very suggestive.10 But perhaps i t canbe improved upon. I have been supposing, throughout this book, thatit makes sense to talk about a complete description of the w orld consid-ered as a system of extended objects. Now suppose we had an alternativedescription of the world, equally complete, and different not merelyin i ts terminology, but in i ts conceptual framework.1 1 What I have inm ind is the kind of rearr ang em ent of fa cts that Erik Stenius describesin his book on Wittgenstein's Tractatus:

It holds goo d even of our kno wled ge of the wo rld that the

facts as a who le can be analysed in different ways. Revo lut ionaryadv anc es in science hav e, in effec t, been con nected with a re-

144







Sp i n o z a s M e t a p h y si c s

will generate, not one, but a pair of attributes. For every time we re-construct the fact s which make up the totality of fact s, we get a newset of propositions corresponding to those facts.

Th e result is that we shall have to say either ι ) that the attributeof thought is coextensive with all of the other attributes combined,or 2 ) that just as there is an attribute of thought corresponding tothe attribute of extension, so there are infinitely many other attributesof thought corresponding to some one of the other unknown attributes.

These two possibilities may be represented in the following way :

extension \

X > thoughtY \Z , etc. /

extension

thoug h t

X

thou gh tz

Y

thought^, etc.

Neither of these pictures is the one we should expect, which looks more

like this:

extensionthought

Ζ, etc.

1) substance

2) substance

146







Spinoza?s Metaphysics

Now I think that Tschirnhaus must have understood Spinoza ratherwel l on this point . In his reply (E p. 64, IV :2 77 ~ 27 8 ) to the questionwhether there are not as many worlds as there are attributes13 Sp inozamerely refers Tsch irnhau s to E I IP7 S, accord ing to w hic h extend edsubstance and thinking substance are one and the same substance, nowcom prehended under one at t ribute , no w under ano ther. So we m aytake i t that there is only one world, which is now understood in oneway, now in another. But Spinoza does not reject anything else Tschirn-

haus says, even though there is already a suggestion here that thoughtmay have a special posit ion among the attributes. The greater part ofthe letter is taken up with providing the requested proof that the humanmind can know no other a t t r ibutes of God beyond thought andextension.

Tschirnhaus is not satisfied. In his next letter, he returns to the attack,asking aga in for a proof that the min d c an perce ive no other attributes ;for, though he sees clearly enough that this is so, he also thinks thathe can demonstrate the contrary from E IIP7S. He then purports toset up the deduction, and asks Spinoza to show him where he has mis-und erstood the sense of the scholiu m :

Although I gather f rom the schol ium that the world is cer ta inlyunique, nevertheless it is no less clear from that scholium that itis expressed in infinite ways. From this it seems to follow that themodif icat ion which const i tutes my mind, and the modif icat ionwhich expresses my body, although it is one and the same modifica-

tion, is nevertheless expressed in infinite ways—in one waythrough thought, in another through extension, in a third throughan attribute of God unknown to me, so on to infinity. For thereare infinite attributes of God, and the order and connection ofmodifications seems to be the same in all . {Ep 6 5 , I V: 2 7 g )

So far there does not seem to be much that Spinoza could take exceptionto. But n ow Tsc hirn ha us goes on to ask :

W h y the mind, w hic h represents a cer ta in mod if icat ion, the sam e

148






modification being expressed not only in extension but in infiniteother ways, perceives only that modification expressed through ex-tension, that is, the human body, and no other expression throughother attributes.

Historians of philosophy have generally felt that Tschirnhaus has a goodquestion here. Granted that Spinoza's proof of the mind's confinementto thought and extension seems to run smoothly enough, and granted

that Tschirnhaus has hardly demonstrated the contrary, nevertheless,his final query seems a pertinent one. Spinoza's answer is maddeninglybrief, and has widely been regarded as unsatisfactory :

I say that although each thing is expressed in infinite ways inthe infinite intellect of G od , nevertheless, those infinite ideas by w hic hit is expressed cannot constitute one and the same mind of a singularthing, but infinitely many minds. For each of the infinitely manyideas has no connection with any other, as I have explained inE IIP7S and is obvious from E Ipio. If you give your attentionto these briefly, you will see that there is no longer a ny difficulty.

Ep. 66 , IV:2 8 o )

It is unfortunate that Spinoza took such an optimistic view of his cor-respondent's acuteness. Later Tschirnhaus complains, through Schuller,that b y this argu m ent the attribute of thou ght is m ad e to extend itselfm uch mo re widely than the rest of the attributes, thou gh this seems

inconsistent with the doctrine that each of the attributes constitutes theessence of substance Ep. 70, I V : 302 ). B ut Sp inoza has said his lastword on the subject.

Now what are we to conclude from all of this? First, I think wemay say that the attribute of thought does extend more widely thanany of the others. There is evidently a mode of the attribute of thought,or a mode of an attribute of thought, for every mode of every one ofthe other attributes. This is stated quite explicitly in the Short Treatise:

Therefore the essence of the soul consists only in the existence

149






of an idea, or objective essence in the attribute of thought, originat-ing from the essence of an object which in fact really exists innatu re. I say of an ob ject w hi ch really exists wi tho ut furt herspecification in order to include here not only the modes of exten-sion, but also the modes of all the infinite attributes, which havea soul just as extension does. ( K V , appendix , 1 :1 1 9 )

The same view that Spinoza put forward in answer to the cr i t ic isms

of Tschirnhaus had been expressed clear ly in his ear l ies t work—thoughit would take a perceptive reader to see i t in the Ethics.

Secondly, since Spinoza says there is no connection between the ideaof a thing in the attribute of extension and the idea of the same thingconsidered under some one of the unkown attributes, i t might be betterto represent his thought here by diagram 2 ) rather than ( 1 ) . H e seemsto think that there are an infinite number of self-contained systems ofideas.

And finally, it might be suggested14 that the peculiar status thus givento thou ght is imp lied in the very definit ion of attribute as th at w hi chthe intellect perceives as constituting the essence of sub stan ce ( E IA 4) .T h e intellect men tioned here is later referred to as an infinite intellect(E II P 7 S ). If i t is to perceive all the attributes of substance an d thereare infinitely many of those attributes, then it must be what Spinozaw ou ld c all absolutely infinite, and n ot merely in finite in its kin d ( E ID2 ).And since an attribute is defined as something which this intellectperceives, there could hardly be an attribute which the intellect failed

to perceive.So I think we may say that the special status which our interpretation

acco rds to the a ttribute of though t is not a difficulty. O n the con trary,i t is a requirement of any adequate interpretation that i t account insome way for the undeniable fact that thought does have a special posi-t ion among the attributes.

Still, I should not wish my hypothesis of the possibility of alternativecomplete descriptions of the world to be taken as anything more than

150






a very tentative suggestion. It is, necessarily, founded on very little evi-dence. Some of i ts implications are right, but others are not. For example,there is nothing in the hypothesis to explain why the infinite unknownattributes should be unknowable. The possibil i ty of alternative scientificaccounts of the same phenomena was a l ive one for Spinoza and hiscontemporar ies—Gal i leo ' s Dialogues on the Two Chief Systems of theWorld we re published in 16 32, the year of Spin oza's birth. M y hypo the-sis would accord better with the contrary view, that the unknown at-

tributes are knowable. But i t is clear that the mature Spinoza wouldnot accept this.

Putting aside the problem of what is involved in Spinoza's assertionof infinitely man y u nk no wa ble attributes, there remains the question :

W h y did he define G o d as a substance consisting of infinite at-tributes? -—a question which Seligman has posed as a difficult one forany naturalist ic interpretation.

In a way, the answer to this is easy. Spinoza's definit ion of God,

like most of his definitions of key terms, is presented as a stipulativedefinit ion, an explanation of the meaning he attaches to the term inhis w ork . B y G od , I und erstand . . . Bu t Spin oza's st ipulations arerarely arbitrary, they generally preserve some continuity with existingphilosophical usage. Since there is clearly ample precedent for the defini-t ion of God (cf . Wolfson, I , 118) , the fact that Spinoza def ines Godin this way poses no real problem, for a naturalist ic interpretation orfor any other.

The real problem—and i t is a problem that is par t icular ly acute fora natural is t ic interpretat ion—is to understand why Spinoza should havethought that Nature must have infinitely many attributes. To conceiveGod as a being absolutely infinite is quite traditional; to conceive Natureas a being absolutely infinite is revolutionary. The naturalist ic interpreter,w ho wishes to read the equat ion G o d = N at ur e with the emphasison N at ur e , has some explaining to do.

W e can , of course, turn to the argum ent S pin oza o ffers us in variousplaces. The major premise is that






( ι ) T h e m ore bein g or reali ty a thin g has,the more at t r ibutes i t has (E Ipg) .

In the demonstration Spinoza claims that this is evident from the defini-t ion of a t t r ibu te . Elsewhere he is mo re enl ightening. Fol low ing Des-cartes, he rejects the possibil i ty of a vacuum on the ground that nothing,that is , nonbeing, what does not exist , can have no properties(Ep. 13 , I V 65 ). C onv ersely, to be, or to be real, is to ha ve pro pe rties —as Spinoza puts i t , each being must be conceived under some attribute(Ep. 9, E I P I O S ) . From this Spinoza thinks i t follows that the morebeing or reali ty a thing has, the more attributes i t has, and that

(2) The most real being, the ens realissimum,must have infinitely many attributes.

This is nowhere asserted explicit ly, but i t appears to be presupposed.T o g et fro m this to the conclusion that Na tur e has infinitely m an y a t-tributes, we would need the further premise that

(3 ) N atur e is the ens realissimum.

T h e question then w ou ld be w ha t reason there is for asserting this.It is not too difficult to find a possible reason. Remembering that

Spinoza identifies reali ty with perfection ( E I I D 6 ) and perfect ion withpower (E I V pref . ) , w e m ight argue that the om nipotence of N ature ,the status of Natura naturans as the necessarily existing, uncaused firstcause of all things is a sufficient ground both for calling it the ens real-issimum and for concluding that i t must have infinitely many attributes.

But while this argument may, after a fashion, be intell igible, i t doesnot seem to me to provide any real understanding. It smacks too muchof verb al man ipula tion. A n d I suggest tha t the difficulty in un derstand-ing Spinoza on this point arises from the fact that here he is tryingto articulate an intuitive feeling which resists clear expression. In theShort Treatise he writes:

After the preceding considerat ion of Nature , we have so far been

able to find in i t only two attributes which belong to this all-perfectbeing. And these attributes give us nothing to satisfy us that they

152






are the only ones wh ich constitute this perfect bein g— on the con-trary, we find in ourselves something which openly proclaims tous the existence, not merely of a great m an y, but of infinitely man yperfect attributes, which must belong to this perfect being beforeit can be said to be perfect. And where does this idea of perfectioncom e from ? Th is something . . . can only come fro m the infiniteattributes themselves, wh ich tell us that they are, but not wh atthey are . (KV, I , i , 1 :17)

It is possible that we owe this form less vision of unseen w or ld s15 tothe scientific developments of Spinoza's time. The discoveries of thetelescope and the microscope had a very considerable effect on the seven-teenth-century imag ination. Fo r they dem onstrated forcef ully that, tobeings operating w ith different perceptual equipment, our fam iliar worldwould look altogether different. Something of this fascination with theeffects of changed conditions of perception is reflected in much of theliterature of the time, Swift's Gulliver s Travels being the best kno wnexample.16 Spinoza himself displays it when he writes to Oldenburgabout how the world might seem to a tiny worm living in the blood(Ep. 32 ). I shou ld not wish to press this sort of speculation too fa r.The worm, after all, no matter how small, is still a mode of extension.The point is simply that considerations of this kind may have inspiredin Spinoza the feeling that the world must be capable of being under-stood in infinitely m an y d ifferen t wa ys, an d that this feeling may then havefound philosophical expression in the doctrine of the infinite attributes

and dubious support in the unconvincing argument outlined earlier.At any rate, whatever may have been the philosophical motivation

for thinking that Nature must have infinite attributes, this doctrine cer-tainly did much to justify Spinoza's identification of God and Nature.For, as he was to point out himself to Oldenburg {Ep. 7 3 , I V : 3 0 7 ) ,that identification did not involve the equation of G od w ith a certainmass or corporeal ma tter . O n the contrar y, it assumed a conceptionof Nature as being, in its own right, absolutely infinite. Provided that

we follow Spinoza in his conception of Nature, I think there is no reasonto object to the equation.

5 3





Spin oza s philosophy consists m ainly in the nega tion of the d oub ledual ism between God and the world and between soul and bodywhich his teacher Descartes had set up.

— S c h o p e n h a u e r , The ourfold Roo t of the Principle of Sufficient Reaso n

Postscr ipt





F. H. Jacobi once said that no one to whom a single line of the thics

remained dark could claim to have understood Spinoza. By that s tan-dard, i t may be doubted whether any one has ever understood Spinoza,including Spinoza himself . But by any reasonable s tandard, I think wecan claim to have made some progress toward an understanding.

Ad mittedly, those wh o hav e gone to Spinoza expect ing to f ind thestrife of co nflicting systems ha ve fou nd , generally, the strife of conflict-ing systems. Those who have sought a grand synthesis of the Greek,

Jewish, and Christian traditions have discovered, not too surprisingly,a Spin oza rightly called th e last of the me dieva ls. It was, per hap s,inevitable that I , who have looked in Spinoza for a system that wouldap pe ar co herent an d plausible to me, should, on the whole, hav e suc-ceeded in locating one.

Su ch reflections m ust h ave a sobering effect on any wo uld-b e inter-preter of Spinoza. But without wishing to maintain that only my portrai tgives a good likeness, I would still insist that it is a good likeness, thatmy emphasis on explanat ion in terms of natural law is fundamental lyrighthe ade d, and tha t anyon e disposed to differ with me should askhimself how he would deal with the passages I have cited.

Before surrendering this work finally to the critics ' knives, however,there is one quite general objection I should like to anticipate. My ac-coun t of Spinoza tends to suggest th at he was prim ari ly a m etaphysicianwith an interest in philosophy of science, whose views happened to haveconsequences for ethics, whereas Spinoza's writings and the ti t les of hism aj or works suggest rather th at his concern with ethics was prim aryand his interest in metaphysics secondary. To ignore the moral convic-t ions tha t un derlay the metaphysics is to leave out of acco unt w ha tmattered most to Spinoza.

This note is of ten sounded in the l i terature on Spinoza—and no doubtrightly. It is certainly true, as the autobiographical passages at thebeginning of the Treatise show, that Spinoza's interest in philosophywas very largely the result of ethical c oncerns. I n this early wo rk hedescribes movingly how his disillusionment with the worldly goods of

r5 5Brought to you by | Brown University Rockefeller Library





Fame, Weal th , and Sensual Pleasure led him to search for a more worthyobject of his love, to look beyond the world of change for somethingeternal. Plainly, this philosophical mood accounts for a great deal ofwhat we find in the Ethics. A boo k on Sp ino za — ev en a book restr ic tedto his metaphysics—which ignores altogether this aspect of his thoughtis bound to give a somewhat misleading impression.

Stil l , there are two things that we need to remember. The first isthat Spinoza himself would surely have rejected the suggestion that his

metaphysical views were dependent on or determined by his ethicalviews. The order in which he wrote the Ethics supports the contraryconclusion that he regarded the doctrines about God in part I as capableof standing on their own and that he took them to be logically priorto the ethical theses developed in the later parts of that work. Whateverthe motives may have been that led him to engage in his curious brandof theology, his phi losophical temperament was very far f rom that whichpostulates the eternal order to satisfy the demands of practical reason.And the god whose exis tence he thought demonstrable would have givenlit t le consolation to poor old Lampe.

Second—and more impor tan t—for the h i s tor ian t ry ing to unders tandwhy Spinoza arr ived at this ontology rather than some other one, i tis not very helpful to be told that he had a lust for the eternal. Thiswill explain, perhaps, why he did not come up with an ontology ofthe Humean sort , in which the furniture of the world is constructedout of fleeting impressions and ideas. But it will not explain why hisPlatonism took the peculiar form it did. Nor will i t add a great dealto invoke the scientific spiri t of the age. Descartes and Leibniz, Lockeand Berkeley, all worked in much the same intellectual climate. Eachcame to terms with the scientific revolution in one way or another. Eachmade room in his ontology for the eternal in one way or another. Butno two of these men d id this in the same wa y. A n d Spinoza s wa y wasradical ly different f rom that of any of them. Why?

I think we m ay best answer this question by seeing Spin oza s philosophyas a reaction against Descartes. For all that Spinoza owed Descartes—

156





Postscript

and he owed h im m uc h— th e poin ts of d i ffe rence be tween the two phi -losophers offer the best opportun ity fo r reconstructing the develop m entof the Spinozistic system. There are, of course, many points of difference.But the most fundamental one, in my view, was that for Spinoza theworld was thoroughly intel l igible . As Spinoza had Meyer point out inhis preface to the Principles of Desca rtes Philosophy , wh en we readin that work that

this or that exceeds hu m an gras p, w e mu st not suppose thatthe author [Spinoza] is putting this forward as his own opinion.For he thinks that all these things, and many others even moresublime and subtle, not only can be conceived by us clearly anddistinctly, but can be explained very easily, provided that, in thesearch for truth and the knowledge of things, the human intellectis led by a path different from that opened up by Descartes. 1 : 132 )

If rationalism consists in having this optimistic view of man's abili tyto comprehend the world around him, then Spinoza was plainly andunequivocally a rationalist .

But Descartes was not. In at least two crucial respects he made theworld a mystery—firs t by exempting man from the domain of law andsett ing him up as a kingd om within a kin gd om (E II I pref . ) w hoseactions could only be given a verbal explanation by appeal to the obscurenot ion of the hu m an w il l ( E V pref . ) , and second by ma intaining thatthe eternal truths, the laws of nature, had been set up by God as aking establ ishes the laws in his kingdom (Adam and Tannery, 1:145,cf. E IIP 3S ). Desc artes had felt com pelled to deny that these eternaltruths cou ld hold independen tly of the will of Go d. T o do so wo uldbe to cast G od as a Jupiter or Satu rn, m ak in g him subje ct to Styxand the Fates .

W ith this denial Spinoza agreed (E IP33S2 ) . Bu t he could not accep tthe anthropomorphic Cartesian alternative for quite a number of reasons.To begin with, since acts of will are the sort of thing that take place

157






at a particular t ime, i t applied a temporal concept to the explanationof something supposed to be eternal. But more important, it did notoffer a genuine explan ation. It postulated G od s will as a cause in onebreath, and in the next admitted that this cause could just as easilyha ve had a contrary effect. T o d o this is to cloak ignora nce in speciousverbiage. A cause from w hich the effect cann ot be inferred is not acause. Furthermore, since it was conceded that God could have willeda different set of laws, there was no good reason to think that he would

not change his mind and alter the eternal order. In the letter to MarinMersenne cited above, Descartes argued that this could not happen, onthe ground that God s will was immutable. But as Descartes also deniedthat there was any necessary connection between God s nature or essenceand his will as exemplified in the nature of the world he created, itwa s difficult to see w hy his will should not change. A s Spino za rema rked,

if it is permitted to attribute to God another intellect and anotherwill without any change in his essence and perfection, what is thereason why he cannot now change his decrees about created thingsand nevertheless remain equally perfect? His intellect and will re-garding created things and their order would remain the same inrelation to his essence and perfection no matter how it be con-ceived. (EIP33S 11:75)

The specter of Hume rises before us and we contemplate the possibilityof a world in which nothing is stable and enduring.

Spinoza s wa y out of this dilemm a w as to identify the will or intellectof God, the universal laws of nature, with the essence of God. Godis neither the slave nor the master of eternal truth, he is eternal truth.The laws of nature neither require nor admit explanation in terms ofanything more ultimate, they could not have been otherwise. This isthe first of the two main tenets of Spinozism and the second, thatman is a part of nature, following her laws in all his actions, is pro-pounded in the same spirit . Both doctrines arise from the fundamental

demand for absolute intelligibility.

158





Notes Index









Notes

P R E FA C E

1. Heinrich Heine, Religion and Philosophy in Germanytrans. John Snodgrass(Boston: Beacon Press, 1959), p. 80.

2. The Portable Nietzscheed. Walter Kaufmann (New York: Viking Press,1954). P· 92.

ι TH E DEF INITION S OF SUBST ANC E AND MOD E

1. An Essay Concerning Hum an Understandinged. A. S. Pringle-Pattison(Oxford: Clarendon Press, 1924), II, 23.

2. Amauld and Nicole, p. 47.3. Amauld and Nicole, p. 47. Their language in this passage may suggest

that modes are to be regarded as nonsubstantial particulars which cannot existapart from their own proper subject, rather than as uni versais, which, thoughthey must exist in some subject or other, need not exist in any particular subjectin order to exist. For they say that they conceive the roundness as unable tosubsist naturally without the body whose roundness it is. But this statementis clarified a bit further on. I t is not that one cannot conceive the mode withoutpaying distinct and explicit attention to its subject; rather what shows thatthe idea of the relation to the substance is involved at least confusedly in thatof the mode is that we cannot deny this relation without destroying the ideawe have of the mode . . . for example, I can easily conceive prudence without

paying distinct attention to a man who is prudent, but I cannot conceive prudencewhile denying the relation it has to a man or to another intelligent nature whichhas that virtue (p. 48 ). The locus cl ssicus for the notion of inherence is, ofcourse, Aristotle's Categories chap. ii. T he same difficulty, less easily resolved,arises there. See Categories and de Interpretationetrans, with notes by J L. Ackrill(Oxford: Clarendon Press, 196 3) and G. E. L. Owen, Inherence ,Phronesis10:97-105 (1965).

4. Adam and Tannery, IX:47. I have rendered the French version, which isfuller and, it seems to me, clearer than the Latin. The French version in generalmakes a great many additions to the Latin. Adam thinks that some of these





Notes to Pages J—ii

must have been made by Descartes, others by the French translator; but howmany and which ones were made by Descartes is obscure. In any case, Descartesauthorized the French version.

5. E.g., in the synopsis of the Meditations, Ad am and Tannery, I X : 10, substanceis what cannot exist without being created by God; in the Third Meditation,Adam and Tannery IX = 35, substance is that which of itself is capable of existing;in the reply to the Fourth Objections, Adam and Tannery IX: 175, substanceis wh at ca n exist by itself, i.e., with out the aid of any other substance.

6. A da m and Tann ery, V I I : 176, 222. In the latter passage, Descartes makes

a distinction analogous to Arnauld's own later distinction between a thing and aqualified thing: Th us to the first I reply that by a complete thing I understandnothing other than a substance clothed with those forms or attributes whichsuffice for me to recognize from them that it is a substance. For we do not knowsubstances immediately, as is noted elsewhere, but only from this, that we perceivecertain forms or attributes, which must exist in something in order to exist, wecall that thing in which they exist substance.

7. Brunschvicg, L a Ré volution Cartésienne et la notion Spinoziste de la sub-stance, Revue de Métaphysique et de Morale, 12:764 (1904).

8. Adam and Tannery, IX.-47 (still following the French version). In the Latin,

the crucial sentence is obscure : Ver um tam en non potest substantia primu manimadverti ex hoc solo, quod sit res existens, quia hoc solum per se nos nonaffìc it . Adam and Tannery, V i l i : 25. Ha ldane and Ross render this: Ye t sub-stance cannot be first discovered merely from the fact that it is a thing thatexists, for that fact alone does not affect us. The P hilosophical W orks of Descartes,2 vols. (Cambridge, Eng.: Cambridge University Press, 1911), I, 240.

9. The definition in terms of independent existence ( By substance we under-stand wh at requires the concurrence of God alone to exist is given in PPI ID2,1:181. The definition of substance as the subject of attributes ( Everything inwhich there exists immediately, as in a subject, or through which there existsanything which we perceive, that is, any property or quality or attribute, ofwhich there is a real idea in us, is called substance ) occurs in PPID5, 1:150.

10. Among the critics are R. H. M. Elwes, who in his introduction to theChief W orks of Spinoza, 2 vols. (N ew Y or k: Dov er, 1951 ) dismisses it as grosslyinaccurate, and Paul Vernière, who in Spinoza et la pensée française avantla Révolution, 2 vols. (Pa ris: Presses Universitaires de France, 19 54) , I, 304describes it as a caric ature . Am on g the Enlightenment followers of Bayle areHume, in the Treatise of Human Nature, book I, part IV, chap. 2, and Voltaire,in Le philosophe ignorant. It should be pointed out, however, that Hume's useof the Bayle critique had an obvious ulterior motive, and that Voltaire's attitude

162







Notes to Pages 19-23

it has in Descartes' system? I believe the latter is correct, for in the other sense,Spinoza would have acknowledged creatures distinct from divine substance whohave been made either out of nothing or from a matter distinct from God.Now it would be easy to prove from a great number of passages that he admitsneither of these two things. Extension, according to him, is an attribute of God.It follows that God is essentially, eternally and necessarily an extended thing, andthat extension is as proper to him as existence. From which it follows that theparticular varieties of extension, which make up the sun, earth . . . and so on,are in God in the way in which the School philosophers suppose they are in prime

m att er. I cannot see that this follows. In any case, if it did the argument couldend here, for it would have been shown that the particular varieties of extensionexist in God as in a subject, which was the point to be proven. Bayle, however,continues: If these philosophers supposed that prim e ma tter is a simple andperfectly unique substance, they would conclude that the sun and the earth arereally the same substance. It must be the case that Spinoza came to the sameconclusion. If he did not say that the sun is composed of divine extension, he wouldhave to adm it that the sun's extension has been m ad e from nothin g. Th e alter-native mentioned originally, that creatures distinct from God might be made froma matter distinct from G od, has apparen tly been forgotten. Bay le goes on : Bu t

he denies creation, and he is therefore obliged to say that the substance of Godis the material cause of the sun, is what composes the sun, [subjectum ex quo ]and consequently that it is not distinct from Go d . (Bay le, p. 335. The bracketedphrase is omitted by Popkin. Spinoza never speaks of God as the material causeof his modes. To the extent that he uses the Aristotelian classification of causes,God is described only as the efficient cause of his modes. See £ I P I 6 C I ; KVI iii,1 : 3 5 ; CA/11, χ, 1:268.

19 . Cf . the article Py rrh o. Th e ground for holding them false is that theyare inconsistent with what is known by religious authority. This is strange itselfbecause one of the self-evident truths thus known to be false is the principle

of contradiction.20. Cf. remark O. Bayle's reply is crudely pragmatic. If Spinozism is no less

open to objection than Christianity, then Christianity ought to be preferred becauseit promises us infinite happiness in the next life.

21. TdlE 101 11:36-37. T h e essences of singular m utable things are notto be drawn from their series or order of existence. That would provide us withnothing more than their extrinsic denominations, relations, or at most, circum-stances, all of which are far from the inmost essence of things. The essence isreally only to be sought solely from fixed and eternal things.

22 . There is an interesting passage in Joachim's posthumously published com-

64





Notes to Pages 23—3 g

mentary, Spinoza's Tractatus, in which he seems to take back much of whathe had said in the Study. See the discussion of Spinoza's doctrine of essence,pp. 36-44. How far this represents a change of view would be difficult to say.

23. Joachim recognizes this in Spinoza's Tractatus. See pp. 60-88, esp. pp. 80-81.24. Again Joachim's Spinoza's Tractatus takes what looks to be a much less

exterme position. See pp. 34-35. The same example suggests what Joachim denies,that the content of knowledge need not vary from one kind of knowledge toanother. Only the way in which knowledge is arrived at, and consequently thedegree of certainty, vary.

25. See e.g. , P. H. Nidditch's chapter on Spinoza, in D. J. O'Connor's CriticalHistory of Western Philosophy (Ne w Yo rk : Free Press, 1964 ).

26. As Passmore has pointed out in T h e Ide a of a History of Ph ilosophy ,History and Theory, Beih eft 5 :2 -3 (19 65 ), i t is necessary, according to W olfson,to take into account Spinoza's supposed criticisms of his predecessors. This makesthe task of finding precedents much easier.

27. There is some warrant in Aristotle for identifying an individual with itsessence. In the Metaphysics Aristotle raises the question W he ther each thingand its essence are the same or different (103 131 5) and concludes that eac hprim ary an d self-subsistent thing is one and the same as its essence (10 32 35 ),

principally on the ground that to know a thing is to know its essence. Later,however, this is clarified We have stated that the essence and the thing itselfare in some cases the same, i.e., in the case of primary substances, e.g., curvatureand the essence of curvature, if this is primary. (By a primary substance I meanone which does not imply the presence of something in something else, i.e.,in something that underlies it which acts as matter). But things which are ofthe nature of matter, or of wholes that include matter, are not the same as theiressences (i0 37 a3 2- i0 37 b7 ). Aristotle 's ground for this is perhaps given in bookVII, chap. XV, where he argues that concrete things are indefinable.

28. In one passage (I, 325-328), Wolfson attempts to undercut this objection

by arguing that God is an exception to the general rule about universale, sinceit is an ens rationis only in the sense that its real existence can be discoveredonly by the mind, by the ontological proofs based upon the adequacy of theidea of God in our mind. In truth, God is an ens reale. But he does not reallyprovide any evidence for this claim. The passages he does cite, from the ShortTreatise and from Aristotle, are quite irrelevant, since they merely point outcertain differences between the whole-part relation and the relation of universalto particular.

29. Hallett does not attempt to argue this, except by pointing out that allof Taylor's evidence for equating the relation of substance and mode with that

165






of subject and predicate comes from Leibniz, not Spinoza.30. Pag e 687. H um e banished the conce pt of substance from psycholo gy, as

Berkeley had banished it from physics.31 . If this is correct, then Spinoza does have a reply to the criticism of

G. Dawes Hicks (Proceedings of the Aristotelian Society n.s., 18:337-338 [1917—18]) that the argument for E IP7 (that it pertains to the nature of substance toexist) is a petitio principii: T h e argument assumes . . . wh at it purports to estab-ish. For if substance has no cause outside itself, it follows that it must havean inner cause only on the assumption that it is existent, and that such existence

needs a cause, in accordance with the dictum . . . that 'of any existing thingthere must necessarily be some cause on account of which it exists. '

32. See, e.g., Ha mp hsire, Th er e ca n only be one substance so defined [ascausa «ti], and nothing can exist independently of, or distinct from, this singlesubstance; everything which exists must be conceived as an attribute or modificationof, or as in some way inherent in, this single substance; this substance is thereforeto be identified with Nature conceived as a whole or as the totality of things

(p · 3O·33. Since arriving at this conclusion I find that Paul Seligman has reached

it by a differen t route. See So m e Aspe cts of Spinozism ,Proceedings of the

Aristotelian Society n.s. , 61:121 1960-61 ). He cites a concurring passage fromWolfson (I , 3243) .

2 T H E C A U S A L I T Y O F G O D

ι. See, e.g., Joachim, Study pp. ggff and 120, and Caird, passim.2. I think Albert Rivaud is right in supposing that in the axioms and lemmas

following E IIP13 we have a sketch of a work on physics which Spinoza projectedbut neve r finished. See L a Physique de Spin oza, Chronicon Spinozanum 4:24-57(1924-26) .

3. Animad versiones ad Joh. Georg e W achte ri l ibrum de recondita Hebraeorumphilo sop hia, W iener, p. 497. O n the history of this wo rk see Fried ma n, pp .155-178. Cf . Gerhardt , 1:148.

4. These definitions are not given by Spinoza himself, but are derived fromBurgersdijck's Logic whose classification of causes Spinoza followed. See Wolf,pp. 190-195.

5. There is, as Professor Passmore has reminded me, a sense in which Humedoes not deny necessary connections either, but merely explains them awaypsychologically.

166






6. Of course most people who have rejected the reduction of general propositionsto truth-functions of singular ones have not followed Russell in admitting generalfacts, but have sought a way out through some form of instrumentalist conceptionof general propositions.

7. Me taphy sics and Ethics, in The Nature of Metaphysics, ed. D. F. Pears(London: Macmil lan 1957).

8. Spinoza's first published work was an axiomatic presentation of the firsttwo parts of Descartes' Principles of Philosophy. It is well known that he groundlenses for a living and that he had acquired a sufficient reputation in opticsthat people were frequently referred to him (for example, Leibniz). One fruitof this work was his Treatise on the Rainbow. His correspondence shows himkeeping in touch with the activities of the newly formed Royal Society throughits first secretary, Henry Oldenburg. He was particularly interested in RobertBoyle's work in chemistry.

9. I have in mind Ralp h M. Blake's excellent study, T h e Role of Experiencein Descartes' Th eor y of M eth od , wh ich originally appeared in the PhilosophicalReview, 38:125-143, 201-218 (1929), and has since been reprinted in Theoriesof Scientific Method: The Renaissance through the Nineteenth Century, ed. Ed-ward H. Madden (Seattle: University of Washington Press, i960).

10. See, e.g., Mo rris Coh en, Am or Dei Intellectualis,Chronicon Spinozanum,11 1: 3 -1 9 (192 3). Sim ilar remarks apply to my emphasis on explanation in termsof causal laws (see Ha rald Hö ffding, Da s Erste Buch der Eth ica , ChroniconSpinozanum, 11:20-53 [1922]) and my emphasis on the ideal of the unity ofscience (see Hampshire's Spinoza .

11. Som e Aspects of Spinozism,Proceedings of the Aristotelian Society,61:109-128 (1960-61) .

3 N E C E S S I T Y

ι. T h e interpretation of Leibn iz on the topic of necessary truth and on relatedtopics has considerable difficulties of its own that I do not wish to get too deeplyinvolved in here. It will suffice for my purposes if the views I attribute to Leibnizrepresent a plausible reading of the texts. My account is drawn mainly fromthe Discourse on Metaphysics Gerhardt IV :42 7-4 63 and two short papers inLouis Couturat Opuscules et fragments inédits de Leibniz (Paris : Presses Un i-versitaires de France, 1903), pp. 16-22 and 518-523.

2. By ana lytic I understand here a proposition whose truth or falsity maybe determined by reduction to a logical truth or logical falsehood through the

167





Notes to Pages 8 -go

substi tut ion of definit ions for definable terms. For a discussion of definit ion, see

be low.3 . I ow e th is poin t to M r. Lor en Lo ma sky w ho mak es i t , not agains t L eibniz ,

but agains t P. F. S tawso n 's vers ion of Leibniz , in an unp ubl ished paper, S t ra wso nand the Ident i ty of Indiscern ib les . H e wo uld not charge Leibn iz wi th a v ic iousregress, s ince he does not think that , for Leibniz, relat ional predicates are includedin the conce pt of an individ ual . I am not cer tain that he is r ight on this poin t ,but i t is an obscure one. Leibniz certainly mentions relat ional predicates in someof his exposit ions of the theory of individual concepts (cf . the Discourse on

Metaphysics, sec t ion 8 , wh ere he inc lude s w i l l conq uer Dar ius and Po rus inthe indiv idual concept of Alexander) . No doubt he had hopes of reducingthese relat ional p red icates to non relat iona l ones, but i t is diff icu lt to say prec iselywhat h is in tent ions were , or how far he was or could be successful in car ry ingthem out. I think i t w ou ld be very d iff icult to m ake an y sense of the do ctrin ethat each indiv idual subs tance expresses the whole universe except on the supposi -t ion tha t some sor t of re la t ional predica tes are inc luded in the indiv idual concept .

4 . C f . Ge rhar dt , V I I 1356. T o f ind a cons is tent accou nt of the pr inc ip le inLeib niz is notor ious ly d i ff icul t . See Lov ejo y, pp . 14 5- 146 , and Ge org e Parkinson ,Logic and Reality in Leibniz's Metaphysics (Lo nd on : Ox fo r d Un ive r s i ty P res s,

1965) . pp · 56 -75-5 . Ho we ver, the c la im tha t exis tent ia l proposi t ions are not genera l ly analy t ic

looks rather arbitrary. Leibniz just if ies his doctrine of individual concepts by anappeal to the nature or definit ion of truth. A true proposit ion simply is onein which the concept of the subjec t inc ludes the concept of the predica te . I fso , why are most exis tent ia l proposi t ions an except ion? The only answer tha tsuggests i tself is that existence is not a predicate. But Leibniz wil l hardly wishto say that because he treats existence as a predicate in the ontological argument.

6 . Lovejoy appears to deny th is Sp ino za . . . thus expressed the princip leof plenitude [roughly, whatever is logically possible exists] in i ts most uncompromis-ing for m and . . . represen ted i t as necessary in the str ict logic al sense. Ev ery thin gshared in the same completely sufficient reason for being that the existence ofG od w as by most phi losophers conc eived to possess (p . 15 5) . Th is sugges tsthat al l t rue existential proposit ions are absolutely necessary.

In o ther p laces Lo ve joy is mo re caut ious : T h e o ther argum ent for the exis tenceof God is the argument f rom the necess i ty of the exis tence of anything whoseexis tence i s not prec luded by some logica l impo ss ib i l i ty (p . 153, m y i ta l ics) .Lovejoy seems here to recognize tha t the nonexis tence of th ings tha t exis t (Godexcepted) must be logica l ly imposs ib le in some sense o ther than absolute logica limpossibil i ty.

168






7. C f. the discussion of fictions in the Treatise, TdlE, 52-58, I I : 19-22.8. I should point out, however, that the notion of an accidental generalization,

as I am using it, is broader than Popper's notion of a numerically universalstatement. My accidental generalizations include existential generalizations.

9. I write gen eral in this parag raph, though accuracy would require uni-ve rsa l. M y reason is that the questions 1 ) a re universal generalizations reducibleto truth-functions of singular propositions? and (2) are existential generalizationsreducible to truth-functions of singular propositions? are commonly treated asthough they would necessarily have the same answer. But it is not clear tome that they would. For though it may be the case that no singular propositionor truth-function of singular propositions ever entails a universal generalization, itis surely not the case that no singular proposition ever entails an existentialgeneralization. If existential generalizations are not red uc ibl e to singular proposi-tions it must be because they are logically weaker, not logically stronger, thansingular propositions. Jon es is a m an does entail Th ere are m en, though

Th er e are men does not entail any singular proposition or truth-function thereof.This is a bit curious and shows just how obscure the notion of reducibility is.I have been proceeding on the assumption that logical equivalence is necessaryand probably sufficient fo r reducibility. I do not think red uc ible has alwaysbeen used that way. Russell, when he wrote The Philosophy of Logical Atom ism,thought there were no disjunctive facts, though he thought we had to admitexistence facts. But surely disjunctive propositions stand in the same relationto singular ones that existential propositions do to singular propositions and truth-functions thereof— they are logically w eaker, not logically equivalent.

10. I have borrowed the example from W illam K neale, Probability andInduction (Lon don : O xfor d University Press, 194 9) , pp. 29-30 . I take it that I amalso expounding his view of the matter and that he, in turn, is providing arationale for the cryptic passage (in section 23 of The Logic of Scientific Discovery,London: Hutchinson, 1959, originally published in Vienna in 1934 as Logik

der Forschung in which Popper puts forward the thesis here discussed. Poppertakes no account of the Russellian criticism of the reducibility thesis.In answer to the objection that, when the region is large, it will be impossible

to take in the whole at once, Kn eale replies that we can wo rk out a schemefor dividing the whole into an exhaustive set of parts each of which can beinspected in turn. Th e statement about the whole would be equivalent to a con-junction of statements about its parts. Presumably there would be no need fora furthe r conjunct to the effect that a, b, c, d, etc. are al l the parts of thewhole. For if the division is systematic, this statement could be argued to beanalytic.

169






M y d ifficulty with Kn eale lies in his claim that T h e use of the wo rd 'al l'in restricted universal statements of the type we have been considering can thereforebe explained by reference to its use in the singular in such totality statementsas 'The building is all stone' or 'Africa is all hot,' but the use of the wordin what I call totality statements is primitive and does not admit or requireexplanation by reference to any other use (my ita lics). I can see why someonemight wish to say that T he buildng is all stone does not have to be transformedinto A ll parts of the building are stone. But I cannot see why it may notbe. Hen ce my hesitation in the text over whether to write ca n be reg ard edor is pròperly rega rded .

i t . Two New Sciences, trans. H. Crew and A. de Salvio (New York: Dover,1914) , pp. 162-164.

12. See Gerald J. Holton, Introduction to Concepts an d Theories in PhysicalScience (Reading, Ma ss.: Addison-Wesley, 195 2), pp. 31 8 -3 19 . Cf. Two NewSciences, pp. 62-63, 276, 290-291. There is an extremely interesting discussionof Galileo's methodology in Ludovico Geymonat, Galileo Galilei (New York :McGraw-Hil l , 1965) .

13. See A. Rupert Hall, From Galileo to Newton, 1630-1720 (New York :Harper and Row, 1963), p. 38.

14. Th is was the remark which prompted de Vrie s to ask whether things ortheir affections were eternal truths. Cf. above, p. 89 and Ep. 10, I V : 47.15 . Definition and Its ProblemsPhilosophical Review, 52 :566 (1943) .16. Cf. Richard Robinson, Definition (London : O xford University Press, 1950 ),

pp. 161-165. I am very much indebted to Robinson's whole discussion of definition.17 . The Logic of Scientific Discovery, p. 431.

4 T H E D I V I N E AT T R I B U T E S

1. I had originally add ed the nam e of Hume to those of Descartes and L ocke,but Professor Passmore has reminded me that, in view of Hume's theory ofbelief, this would not be correct. The similarity between Hume and Spinozaon this point is not doubt connected with the fact that they both adopt no-substancetheories of the mind. Just as, for Hume, the mind is nothing but a bundleor collection of perceptions, so for Spinoza the mind is entirely constituted byits ideas.

2. F a c t also causes some embarrassment for Britan when he gets down tonarratio falsa, quae erat facti, quod nullibi contigerat. C f . The Principles ofDescartes Philosophy (Chica go: Open Court, 190 5), p. 132.

3. My debt here is to C. A. Baylis ( Facts, Propositions, Exemplification,

170





Notes to Pages 125 144

and Truth , Mind, 67: 459 -47 9 [1948] and G. I. Lewis, An Analysis of Know ledgeand Valuation (Lasalle, 111.: Op en C ourt, 1946). But they are not to be heldresponsible for the use I make of their ideas.

4. My example is constructed on the analogy of one of Donald Davidson's inActions, Reasons and Causes, Journal of Philosophy, 60:686 (1963). Davidson

writes I flip the switch, turn on the light and illum inate the room . Unbekno wnstto me I also alert a prowler to the fact that I am home. Here I do not dofour things, but only one, of wh ich four descriptions have been given. I supposethat what applies to actions applies, mutatis mutandis, to facts. For a varietyof cases, some of which seem to me more difficult, see Ro m ane C lark, Fa cts,Southern Journal of Philosophy, 21:123-136 (1966).

5. Cf. Joachim, Spinoza s Tractatus, p. 90 : there are two notoriously difficultquestions in Spinoza's epistemology— 1 ) in what precise sense, according toSpinoza, we know our body, and (2) how, on the other hand, a mind, whichis the essentia objectiva of our body, can be (or have) knowledge of anythingbesides.

6. Or that the fact which p describes involves Peter's nature. Spinoza speaksboth of bodily mo difications and of their ideas as involving the na ture ofexternal bodies.

7. This is theorem 35 in book III of Euclid's Elements. I have given a some whatdifferent exposition of the exa mp le tha n Spinoza does, but I th ink it is fai thfu lto his intentions. I should mention that in the interpretation of this scholiumI am very much indebted to M r. A. J. Wa tt, whose Th e Divine Intellect andWill in the Ethics of Spinoz a, an M . A. thesis subm itted to the University ofWestern Australia, came to my attention as I was preparing the final draft ofthis chapter.

8. There is a passage in the Short Treatise in whic h Spinoza says tha t upto now only two of the attributes are known to us through their essence (K VI, vii, 1:44), implying that the unknown attributes are not, in principle, unknow-

able. But his mature view seems to be that they are. See below.9. Brunschvicg, Spinoza et ses contemporains (Pa ris: Felix Alean, 1923), pp.

67-68.10. How ever, the reference in the last sentence to Th e point of view of absolute

reality seems to imply some thing with which I should not agree, nam ely, tha tthere is no real difference between the attributes. This view of the attributesis not, I think, sustained by the evidence. On the whole, much discussed questionsee Francis Haserot, Spinoza's Definition of Attrib ute,Philosophical Review,62:499-513 (1953). Haserot's arguments seem to me quite conclusive and shouldput an end to the subje ctive interpreta tion of the attributes.

11. This would have to be the case because the complete description I speak

1 7 1






of is a set of propositions, where propositions are regarded not as linguistic entitiesbut as a kind of universal in rebus.

12. Erik Stenius, W ittgenstein s T ractatus (Ithaca: Cornell University Press,i960), p. 27.

13. Those who are struck by Tschirnhaus' use of the word mundus in thisconnection will be intrigued to know that he and Leibniz were very close friends.

14. This was argued by Pollock in Spinoza: His Life and Philosophy (London:C. K. Paul, 1880), p. 173. In the second edition (1899) Pollock is able to citea letter of Leibniz that confirms his interpretation, p. 161.

15. The phrase is Pollock's, but he denies that this is all the doctrine of theinfinite attributes amounts to, on the ground tha t it is an extension of the paral-lelism already fixed in Spinoza's view of the world of human experience as anecessity of scientific thou ght (2d ed., p. 1 58 ).

16. The classic study of this is Marjorie Nicolson's The Microscope and theEnglish Imagination (Northampton, Mass., 1935). The bulk of this monographhas been reprinted in the author's Science and Imagination (Ithaca: Great SealBooks, 1956).

172





ndex

Accidenta l genera l iza t ions , 5 1 - 5 2 ,9 5 - 9 8

Adequa te ideas , 2 5 - 2 8 , 1 3 2 - 1 3 4Aristotle, 28 , 2 9 - 3 0 , 3 1 , 1 2 4 - 1 2 6Arnau ld , An to ine , 5 - 6 , 2 0 - 2 1 ,

1 0 9 - 1 1 1Attr ibutes , 8 - 9 , 11 , 1 5 , 1 6 - 1 8 , 3 5 - 3 6 ,

4 7 , 5 5 - 5 8 . 75, 150

Bayle, Pierre, 2, 1 1 -14 , 19 -20 , 33Being, 1 3 , 3 3 - 3 4Borell i , Giovanni, 108Boxel , Hugo, 3 1Bri tan, H . H ., 123Brunschvicg , 9 - 1 0Burgersdi jck , Franco, 38, 163Burman , F rans , 9

Cause , 30 , 34 , 39-42 , 45 -5° . 54 , 59 ,7 5 - 7 7 , 1 5 8 ; p roximate cause , 36,7 0 - 7 2 ; adequate cause , 6 6, 7 2 - 7 4

Change , 11- 1 2 , 4 6 , 7 3 - 7 4 , 9 1 - 9 2Clavius , Chr is topher, 108

De Deugd , C . , 83Deduction of f ini te from infinite, 46,

6 2 - 7 4

Defini t ion , 2 9 - 3 0 , 3 5 - 3 6 , 1 0 8 - 1 1 3 , 151

Descar tes , 6 - 11 , 1 3 - 1 4 , 1 5 , 2 0 - 2 1 , 3 7 - 3 8 ,56 , 79 , 122 , 1 5 6 -1 5 8

De Vries , S imon, 108Dubs , H . H . , 11 2

E r r o r, 1 3 4 - 1 3 7Ete rn i ty, 46 , 69 -70 , 73 -7 4 , 107 . ee also

Proposit ions, necessaryEterni ty of mind, 1 3 7 - 1 4 3Exper i ence , se e Percept ion; Theory of

knowledge

Facts , 5 3 - 5 4 , 5 5 - 5 6 , 8 9, 1 2 2 - 1 2 6 ,1 4 4 - 1 4 5

Fixed and e ternal th ings , se e Modes ,infinite

Gali leo, 9 9 - 1 0 0 , 151Gassendi, Pierre, 8, 13-14Ge nus an d species, 2 8- 3 6G o d , 6 - 7 , 10 , 12 -1 3 , 30 , 35 , 36 , 41 ,

4 5 - 5 0 , 5 8 - 5 9 , 8 0 - 8 1 , 11 7 , 1 5 1 ,157-158

Hal l , A. R. , 10 0 -101Hal l e« , H . F. , 39Hampsh i re , S tua r t , 56, 83Hobbes, 8H u m e , 77 , 156 , 158

I d e a , 56 , 1 2 1- 1 2 6 , 1 3 5 - 1 3 6I l lusionism, 2 4-28 , 46Individ ual i ty of par t ic ular th ings , 2 2 - 2 4

Jacobi , F. H. , 155Joach im H. H . , 20 -28 , 74

Laws of na ture , 4 7 - 4 9 , 5 ' , 5 5 - 5 6 ,9 8 - 1 0 1 , 1 0 6 - 1 1 7 , 5 7 - 1 5 8

Leibniz , 14-18, 46 , 64 , 84-93Locke, 4 - 5 , 20-21, 122Love joy, Ar thu r O . , 56, 168

Mersenne , Mar in , 158Miracles , 3 1 , 4 7 - 4 9Modes , 5 - 6 , i l , 1 3 - 1 4 , 1 8, 1 9 , 2 0, 2 1 - 2 8 ,

2 8 - 2 9 , 32 -33 , 35 -36 , 37 -40 , 43 , 75 ;finite, 18, 6 2 - 7 4 , 7 5 ; infini te, 36, 43,47 , 58 -62 , 67 -68 , 7 1 - 7 2 , 75 , 107 ,1 1 6 - 1 1 7

Natu re , 41 -43 , 47~49 , 6 1 , 1 5 1 - 1 5 3 ,«57-158

Nicole, Pierre, se e Arnau ld

1 7 3





n d ex

Nomological propositions, s Laws ofnature

Oldenburg, Henry, 1 5 3Ontologica argument, 4 0 - 4 2 8 0 - 8 1

1 0 6 - 1 1 7

Perception, 1 2 6 - 1 3 7Pollock, Frederick, 41, 82Popper, Karl, 5 5 9 6 11 7Porphyry's tree, 29

Propositions, 5 0 -5 3 56 1 2 1 - 1 2 6 ; neces-sary, 5 1 - 5 2 8 3 -9 3 1 0 6 - 1 1 7 157-158

Quine, W. V. O., 11 5

Russell, Bertrand, 4 39 55

Schuller, G. H., 147Seligman, Paul, 80 15 1

Stenius, Erik, 1 4 4 - 1 4 5Substance, 4 - 11 , 1 3 - 1 6 1 9 2 0 - 2 2 2 4 - 2 5

2 8 - 3 0 31-32, 33-34, 37-43Sufficient reason, 4 0 - 4 1 8 6 - 8 7Summum genus, 2 8 - 2 9 3 1 - 3 2 3 3 - 3 4Swift, Jonathan, 1 5 3

Taylor, A. E., 39Theory of knowledge, 2 5 -2 8 9 9 - 1 0 1 1 1 3Time, 46 73-74, 107Truth , 5 3 - 5 4 1 2 2 - 1 2 6 1 3 4 - 1 3 7 14 2

Tschirnhaus, E. W., 1 4 7 - 1 5 0

Universale, s Genus and species

Versfeld, Marthinus, 8 - 9 1 0

Wittgenstein, Ludwig, 5 3 5 5 7 6 - 7 7 1 4 4Wolfson, Η. Α., 18, 2 8 - 3 6 1 0 7 1 2 6

1 7 4









Date post:	02-Jun-2018
Category:	Documents
Upload:	anonymous-hrfqjxsnv
View:	215 times
Download:	0 times

Curley, E - Spinoza´s Metaphysics. An Essay in Interpretation-Harvard University Press (1969)

Documents