O R G A N O N 4(1967) L ’ANTIQUITÉ CLASSIQUE
ET LES DÉBUTS
DE LA SCIENCE MODERNE
Jürgen M ittelstrass (German Federal Republic)
REMARKS ON NOMINALISTIC ROOTS OF MODERN SCIENCE
In discussions dealing w ith the beginning of m odern science, i.e. of Galileo’s physics, the question continues to be raised—especially since P ierre D uhem ’s far-fetched antedating of this b eg in n in g 1— as to the p a rt played in it by nominalism. S tarting from the assum ption, usually form ulated none too precisely, th a t up to this very day m odern physics is essentially nominalistic, an effort is being m ade to establish a link betw een th e nom inalism of late Scholastic tim es and modern physics. In the following, I shall attem p t to show how such a con nection, although very lim ited in scope, can indeed be observed.
To sta rt w ith, here is a brief com m ent on the historical p a tte rn of nominalism as far as this now concerns us. As a philosophical theory of language, nominalism contests the assertion of the “realists” th a t the order of signum and res, as applied in the use of proper names, is valid also in the case of w h at is called universalia; th a t is— in m odern w ording—th a t to predicates (or to concepts, if the phonem ic realization of predicates m ay be ignored) th ere corresponds som ething still “distinguishable” from existing things. According to nom inalistic in terpretation these universals are (in a ra th e r u n fo rtunate term ino logy, dictated by the adverse party) nothing but “m ere nam es,” whose correlates in reality are again supposed to be nothing b u t existing
1 Duhem started w ith an attem pt to present Leonardo da Vinci as a d ecisive precursor of Galileo (Les Origines de la S tatique, I—II, Paris 1905— 1906). A fter wards he forwarded the claim that even before Galileo th e Parisian Term inists headed by N icole Oresm e laid th e foundation of modern physics in th e 14th century (Etudes sur Léonard de Vinci, III, Paris 1913). W hile in volu m e 7 of his Le S ystè m e du Monde (1956) h e again tries to substantiate th e thesis of th e authorship o f the Parisian Term inists, volum e 10 contains som e remarks w hich reveal a more m oderate opinion (Le S ystèm e du Monde. Histoire des Doctrines Cosm ologiques de Platon à Copernic, X, Paris 1959, p. 45). For a presentation and a critical evaluation of this thesis see: E. Rosen, “R enaissance Science as Seen by Burckhardt and His Successors,” in The Renaissance: a Reconsideration of the Theories and Interpretations of the Age, ed. by T insley H elton, Madison 1961, pp. 77— 103; J. H. Randall Jr., The Career of Philosophy. From the Middle A ges to the Enlightenment, I, N ew York 1962, pp. 267—283.
40 J. M ittelstrass
things. 2 It seems th a t the earlier extrem e nominalism, as in terpreted by Roscelin de Compiegne (about 1050— 1125), adm its solely proper nam es—which is obviously absurd—while the la ter nominalism of W illiam Ockham (d. 1349/50), m ore im portant in our context, persis ten tly upholds the differentiation between p roper nam es and predicates, already treatin g the latter, to some extent, as synsem antic expressions and calling them universalia in praedicando. 3 O ckham ’s fu rth e r thesis, th a t of these universals taken by them selves an esse in m ente h o ld s4 is—due to its obscure form ulation—far from suitable to solve the real problem in this controversy on universals. The relevant question is, w h eth er and how “ab stract” objects can reasonably be discussed. Even so, O ckham ’s suggestion does p rev en t any naive assum ption w ith out, how ever—as falsely claimed by “realists”—renouncing a “know ledge of the general” and consequently of science. To be sure, general sentences—and nothing else is implied by the custom ary talk about the general—are allowed for in Ockham ’s theory. 5 In these or other sentences, any expressions, sounding like proper nam es y et not naming any object existing in this world, are to be elim inated. It is, however, left to the historical nominalism to prove th a t all these expressions can actually be elim inated w ithout making the common world smaller. Ockham him self suggests to speak m erely of “moving”, or of “moving bodies,” instead of using words like “motion”, thereby attem pting to avoid from the beginning the misleading “realistic” concept “motion exists.” 6 In other words, to pu t it in a nutshell, Ockham ’s world
2 It is w ell known that the dispute over “u niversals” dates back to the futile discussion on genera and species, started by Porphyry in his preface to A ristotle’s Categories (“Isagoge sive quinque voces,” ed. A. Busse [Comm, in Arist. Graeca, IV, 1], Berlin 1887, p. 1; furthermore: Boethius, “In Porph. Comm.” I [Migne, Patrologia Latina 64] 83A). S ystem atically, this dispute represents a variant of the argum entations, by w hich Plato, in th e Cratylus, developed his programm e of a philosophy of language in an antithetic form. The fact that the m edieval dispute over universals does not follow this particular pattern but rather •— on the part of the “realists” — follow s P lato’s supposed realism concerning Forms, is under standable in v iew of th e w ay how Plato’s w ritings have been transmitted; still, it is an unfortunate continuation of the dispute as introduced by Plato him self. Indeed, it can be shown, that Plato considered to be im possible a “reasonable” philosophy of language w ithin such a type of antithesis, as that between nominalism and realism , and that for this reason he pointed out som e acceptable “solution” of the problem s under discussion; cf. K. Lorenz, J. M ittelstrass, “On Rational Philosophy of Language. The Program m e in Plato’s Cratylu s Reconsidered,” Mind, LXXVI (1967), pp. 1—20.
S um m a Totius Logicae, I, 15 (Franciscan Institute Publications, Text Series No. 2, ed. Ph. Boehner, St. Bonaventure 1951, pp. 45ff).
/* Cf. S um m a Totius Logicae, I, 14 (ibid. pp. 43ff) and I, 12 (ibid. pp. 38ff), w ith direct reference to the “psychological” theory of language in De interpretatione (1. 16a 3—4).
5 Cf. W. Stegm iiller, „Das U niversalienproblem einst und jetzt”, A r ch iv fiir Philosophie, 6 (1956), pp. 192—225 and 7 (1957), pp. 45— 81; especially 6 (1956), pp. 202ff.
6 “Sum m ulae in Libros Physicorum ”, III, 7 (Philosophia Naturalis, Rome 1637, pp. 54f.); cf. Tractatus de Successivis, I (Tractatus de Motu) (Franciscan Institute Publications, Philosophy Series No. 1, ed. Ph. Boehner, St. Bonaventure 1944, p. 45).
N om in alistic Roots of M odern Science 41
consists, on the one hand, of “single things” whose behaviour, in the m anner of moving bodies, is th e object of a scientia realis and, on the other hand, of “linguistic sym bols” by m eans of w hich this scientia
realis operates “nom inalistically” ; these symbols, in turn, are the object
of a fu rth e r science, a scientia rationalis, w hich in itself establishes the rules for the use of these symbols.
L et us add here, th a t this scientia rationalis concentrates p rin ci pally on problem s of scientific m ethodology and, being itself an exam ple of the logica moderna, still follows the preceding logica nova which, contrary to the logica vetus, also knew and discussed A ristotle’s
Second A n a ly tic s.7 Considering this w e m ay claim th a t this p attern
of nominalism indeed drew very near to the tendencies of the new science in its evolution then. 8 A fter all, G alileo’s m echanics rested on the very fact th a t th e question as to the “essence” of motion was deliberately abandoned in favour of th e study of moving bodies and that, for the first time, m ethodical m eans were granted by w hich exact descriptions of the behaviour of moving bodies became possible. The scholastic controversy w hether motion is a flu x u s form ae, th a t is, an “independent” state still distinguishable from a moving body, or ra th e r a form a fluens, m eaning a form identical w ith th e trajecto ry of the moving body, 9 is in no w ay of in terest to Galileo, the au th o r of the Discorsi—although it may be held th a t actually he is subscribing to the form a-fluens-theory which, significantly, had been cham pioned by Ockham, to o .10 In conform ity w ith O ckham ’s advice, Galileo thus
7 Among A ristotle’s w ritings, logica v e tu s comprised th e Categories and De i n t e r p r e t a t i o n . These are supplem ented by th e Eisagoge by Porphyry, the Commentaries by Boethius, as w ell as the book De sex Principiis ascribed to Gilbert de la Porree. Upon publication of the rem aining w ritin gs of the Organon in the m iddle of the 12th century, studies of th e Second Analytics, the Topics and th e Sophistic Refutations were, by contrast, distinguished as logica nova. Later on, both w ere covered by the common term logica antiqua (or logica antiquorum). This took place, after interests in a philosophy of language had been renew ed under “nom inalistic” leadership in the m iddle of th e 13th century and brought about the evolution of a logica moderna (or logica modernorum), particularly as the result of a discussion on proprieta tes termin orum. These differentiations w ere never upheld very accurately: thus th e difference betw een antiqui and m oderni appears as early as the 10th century (C. S. Barach, Zur Geschichte des Nominalism us vor Roscellin, Wien 1866, p. 14); nor w as this d ifference strictly upheld in later tim es. Cf. L. M. de Rijk, Logica Modernorum. A Contribution to the H istory of Early Terminist Logic, I, A ssen 1962, pp. 14ff.
8 Of recent literature cf. O. Fleckenstein, Natu rwissenschaft und Politik. Von Galilei bis Einstein, München 1965, p. 17.
u A detailed account of this term inologically at tim es rather obscure discussion (this explains th e numerous contradictions in its modern recapitulations!) is given by A. Maier, Z wis chen Philosophie und Mechanik. S tudien zu r Naturphilosophie der Spätscholastik, Rome 1958, pp. 59ff.
,n “Ideo dicendum est, quod motus non est talis res distincta secundum se totam a re perm anente, quia frustra fit per plura, quod potest fieri per pauciora . . . Ponendo enim, quod corpus sit in uno loco et postea in alio loco, sic procedendo sine omni quiete et omni re m edia alia ab ipso corpore et ipso agente, quod m ovet, vere habemus m otum localem; igitur frustra ponitur talis res alia” (Tractatus de Successivis I, Boehner, p. 45). Cf. “Sum m ulae in Libros Physicorum ”, III, 6—7 [Philosophia Naturalis pp. 53ff). A. Maier, op. cit., pp. lOOff.
42 J. M ittelstra ss
ceased reasoning about “motion” ; instead, he establishes rules about bodies in so far as they are in motion. To be sure, in his methodical reflexions Galileo m akes use of a clearly nom inalistic vocabulary, as seen p articu larly well in the fam ous quotation from the Dialogo: he claims th a t “g rav ity ”, which Simplicio, the representative of a schola stic A ristotelianism , suggests to be the cause of th e motion of falling bodies, is b u t a “nam e” for certain observed re g u la ritie s.11 In his works, e.g. in connection w ith th e problem of secondary qualities, 12 Galileo repeatedly protests th a t references to causative substances, as commonly used in scholastic discourses, constitute nothing but “pure nam es” (puri nomi) and represent by no means w ell-founded and justifiable explanations. This m ay be considered to be evidence of G alileo’s deliberately accepting the nom inalistic parlance of argum en tation. Ockham ’s nominalism had reached N orthern Italy as early as about 1400, by w ay of the P arisian Ockhamism; among other places it was in Padua th a t Averroistic circles arden tly discussed it ever since. There is little doubt, therefore, th a t Galileo m ust have been w ell acquainted w ith nominalistic reasoning. He may also have known such works as those w ritten by A lbert of Saxony (d. 1390) and M ar- silius of Inghen (d. 1396)—because in the Juvenilia, considered to be lecture-notes from his time a t Pisa, m ention is made of Ockham, A lbert of Saxony, as well as of P aulus V enetus (d. 1429), a man of great im portance in this line of tradition, and of the Doctores P ari- sienses.
Galileo’s aversion to the assum ption th a t there are any substances to be substituted for phenomena, combined w ith his mechanics which obviously m ade such substances superfluous, had an extraordinary effect upon fu tu re thinking. There seemed to be—as was still the case w ith Leibniz in his youth 13—a necessity of choosing betw een the ac ceptance of “substantial form s” and the “m echanical th eo ry ”—a de cision w hich was m ade almost unanim ously in favour of th e “m echani cal th eo ry” and which, at the same time, adopted its nominalistic pathos. The nom inalism of the new science w as rightfully looked upon as an attem p t of vindicating one’s own scientific theory philosophically; and at the rate a t which this new science, startin g from Galileo’s mechanics, progressed successfully, nom inalism itself was being ac cepted as a “philosophical” theory of now established value. This connection is probably most distinctly discernible in Hobbes’ works. On th e one hand, Hobbes, like Descartes, intended to w rite a M eta
11 “Dialogo sopra i due m assim i sistem i del m ondo,” II, in Le Opere di Galileo Galilei, Edizione N azionale, VII, pp. 260f.
12 Cf. “II Saggiatore,” ibid., VI, p. 348.
13 Cf. the autobiographical note in Leibniz’s letter to Remond, dated Jan. 10, 1714, Die philosophischen Schriften von G. W. Leibniz, ed. C. J. Gerhardt, III, p. 606.
N om in alistic Roots of M odern Science 43
physics corresponding to the new science and, on the other, he holds a strictly nominalistic view on language which strongly relies on O ckham ism .14 In Hobbes’ w orks aind, more clearly, in the so—called B ritish Empiricism, it once more becomes ap p aren t how much nom i nalism and the new science are in agreem ent w ith each other. Even so, it m ust be stressed th a t n eith er nominalism, tu rn in g psychological at a steadily increasing rate, nor the new science w hich more and more abandoned its “sy nth etic” character, represent theories which are soundly based by themselves.
If one bears in m ind th a t Hobbes and the B ritish Em piricism testi fied to the com patibility — how ever problem atical in this historical form—of nominalism and the new science, and if one adds th e fact th a t even in Galileo’s tim es the new science made use of a nom ina listic language for its “philosophical” justification, one m ight be tem p ted to assume, th a t th e new science could not possibly come into existence w ithout the historical nominalism of O ckham ’s version. In consequence, it m ight then be claimed th a t only w ith the new concept of “reality ”, as advanced by Ockham, the creation of a new type of physics w as made possible. And one m ight continue in this vein and argue th a t this new type of physics had m erely to supply a b etter “theory” about a w orld w ith which th e m etaphysicians had long ago come to term s. However, this allegation would vastly exaggerate the p a rt played generally by philosophy in the history of science; in p a rti cular, it is m anifestly false as fa r as the origin of m odem physics is concerned.
The version given so far, i.e. th a t Galileo him self sought to find in nominalism a “philosophical” justification of his own physical theory, does by no means im ply th a t nom inalism is in an y w ay already p re supposed in the actual system of G alileo’s physics. This system should ra th e r be im agined to follow Euclid’s geom etry in so fa r as G alileo’s mechanics is also based on axioms. The validity of these axioms, however, is not to be considered self-evident as those in geom etry are, b u t it is subject to an experim ental exam ination of th e ir logical implications. This “geom etric” or, as previously labelled, “sy n th etic” character m akes Galileo’s mechanics appear indifferent to the histori cal altern ativ e betw een nominalism and realism, because at th a t tim e n eith er realism n o r nominalism could have supplied a satisfactory explanation of “Euclidic” connections of sentences and the concepts used w ithin these connections. Thus the tradition, w hich Galileo follows in the “Euclidic” build-up of his mechanics, had from the very begin ning not been affected in any way by the realism -nom inalism contro versy. As a rule, deductive procedures had been freely applied by
14 Cf. “D e corpore,” I, 2 (Opera philosophica quae latine sc ripsit omnia, ed. G. M olesworth, I, pp. llff.).
4 4 J. M ittelstra ss
both realists and nominalists, while then one was facing a m a tter of methodology, m anifestly quite irrelevant to one’s own controversial problems. As far as it may be asserted th a t the greatness of G alileo’s historical achievem ent lies in the field of methodological reasoning—he being the first to develop w ithin the fram ew ork of his Euclidic model a m ethodically fully convincing approach to the explanation of physical phenom ena—the allegation m entioned above seems inappropriate. In fact, Galileo did not have to depend on historical nominalism, even if this philosophical theory proved convenient for his scientific purpose, by reason of its opposition to the A ristotelian-Scholastic “substances” and because of its persistence in pointing out a w orld of “single things”.
Now, the objection m ight be raised that, w hile Galileo’s mechanics does not, in its theoretical core (i.e. in its “Euclidic” structure), contain anything specifically nominalistic, it still envisages from the very beginning a world th a t can only be com prehended by a nominalistic mind. U ndoubtedly this assumption fits very well w hat previously has been said about com parable intentions of both Ockham ’s nominalism and the new science; b u t even here some reservations are to the point. This assum ption is acceptable if it is m eant to say that, to some extent, the domain of physics had become sm aller than it had been before, when “substantial form s” had been p a rt of it; again, this assumption is erroneous, if it should be used to chain the new physics to a definite philosophical doctrine. Because once more Galileo surpasses his nomi nalistic predecessors in methodological sagacity. His rem ark—previously m entioned as evidence of a nominalistic influence—th a t “g rav ity ” is bu t a name, not an acceptable explanation of certain observable regu larities, is undoubtedly proof of his acquaintance w ith the nominalistic vocabulary. But, at th e same time, this rem ark m ust be understood on the background of a differentiation which methodologically had become of great im portance for his mechanics. We refer to his dif ferentiation between a kinem atic and a dynamic approach. It is well known, th a t Galileo pays no attention to the dependence of the mo tions, as dealt w ith in his mechanics, on the forces causing these motions; and it is his clear-cut distinction betw een the cause of motion and the form of motion, which for the first tim e made it possible to form ulate precise statem ents on the phenom enon of motion. This means, however, th a t henceforth the dynamic approach does not be come superfluous, b u t th a t it is now, in a m ethodically convincing order, even required by th e mechanics which has been carefully lim ited to a kinem atic approach. 13
15 A detailed account of the m ethodological aspect of G alileo’s physics, w hich in the context of this article could be sketched m erely in its general outline, may
N om in alistic Roots of M odern Science 4 5
Accordingly, Galileo, in the passage quoted above, does not attem p t to banish the discussion on “g rav ity ” from physics, as any ard en t nom inalist w ould have done; he ra th e r prefers to postpone such a discussion and m akes it quite clear th a t a t this point nothing could be gained by introducing some sort of m ysterious “substance”. H ere he is, of course, in full agreem ent w ith historical nominalism, the term inology of w hich he adopts.
Thus, in every attem p t of linking late Scholastic nom inalism w ith the rise of the new science, one must, w ith reg ard to Galileo, carefully distinguish prim ary methodological insights from his in terp retativ e m eans of expression. W hen Galileo discusses his own physics, eith er for com bating adverse notions or for the purpose of prom oting his own opinion, he does so mostly in order to em phasize th e reliability of m athem atical dem onstration and the liability to e rro r of m etaphysical allegations w ithin th e domain of his science. In doing this he p refers to use the nominalistic parlance, because it seems to be least involved w ith metaphysics. It should be added, th a t in fact Galileo had no interest w hatsoever in philosophical problem s as trad itio nally discussed. In favour of frag m entary research he dispenses w ith any “philosophy” of his own which, most certainly, w ould have overreached this type of research. And it is characteristic th a t a m etaphysicist like Descartes p rom ptly called this attitu d e of Galileo a philosophical deficiency. 16 There w as in G alileo’s mind no room for anything like th e question, how to explain the fact th a t the theorem s of his m echanics hold tru e for th e w orld — a question th a t occupied m odern philosophy up to K a n t’s provoking form ulation how experience is possible a t all. In this context a rem ark like Galileo’s fam ed com m ent on the Book of N ature, as w ritten in m athem atical wording, 17 m ust be looked upon ra th e r as the resum ption and successful continuation of a w ell-know n m eta phor. 18 This rem ark also shows how easily even some C hristian platonism in Galileo’s reasoning could agree w ith the nominalistic beliefs w hich he explicitly professed elsewhere.
be found in E. J. Dijksterhuis, Die Mechanisierung des W e ltbildes (in German by H. Habicht), Berlin—Göttingen—H eidelberg 1956, pp. 371ff.
16 Letter dated Oct. 11, 1638 to Mersenne, in O eu vres de Descartes, ed. Ch. Adam and P. Tannery, II, p. 380. H ere Descartes refers to h is perusal o f th e recently published Discorsi. W ith regard to this “philosophical” protest against the “unphilo- sophical” character of the n ew science, com pare th e present author’s work: Die Rettung der Phänomene. Ursprung und Geschichte eines an tiken Forschungsprinzips, Berlin 1962, pp. 255ff.
17 “II Saggiatore,” loc. cit., VI, p. 232; letter dated January 1641 to Liceti, Ed. Naz. X V III, pp. 293ff.
18 On the history o f th e so-called “book m etaphor” cf. E. R. Curtius, Europäi sche L iteratu r und lateinisches Mittelalter, 3rd ed., Bern 1961, pp. 323ff.
4 6 J. M ittelstra ss
In conclusion one m ight com ment on the in terrelation between nom inalism and th e new science as follows: occasionally the new science tries to find its philosophical justification in nominalism. In spite of this, science in its actual stru c tu re is by no m eans dependent upon historical nominalism. However, it is the use of nominalism that, in line w ith Galileo’s attitu d e, the new science could afford to do w ith o u t any “philosophical” justification of its own. There is y et no definite ex planation on hand as to how such a justification can be established.
