This article was downloaded by: [Harvard College]On: 23 April 2013, At: 15:41Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Annals of SciencePublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/tasc20

The clock metaphor and probabilism:The impact of Descartes on Englishmethodological thought, 1650–65Laurens Laudan M.A. Ph.D. aa Department of History and Philosophy of Science, UniversityCollege, London, W.C.1Version of record first published: 02 Jun 2006.

To cite this article: Laurens Laudan M.A. Ph.D. (1966): The clock metaphor and probabilism: Theimpact of Descartes on English methodological thought, 1650–65, Annals of Science, 22:2, 73-104

To link to this article: http://dx.doi.org/10.1080/00033796600203065

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representationthat the contents will be complete or accurate or up to date. The accuracy of anyinstructions, formulae, and drug doses should be independently verified with primarysources. The publisher shall not be liable for any loss, actions, claims, proceedings,demand, or costs or damages whatsoever or howsoever caused arising directly orindirectly in connection with or arising out of the use of this material.

ANNALS OF SCIENCE A QUARTERLY REVIEW OF THE HISTORY OF

SCIENCE AND TECHNOLOGY SINCE THE RENAISSANCE

VoL. 22 June, 1966 No. 2

(Published November, 1966)

THE CLOCK METAPHOR AND PI%OBABILISM: THE IMPACT OF DESCAI%TES ON

ENGLISH METHODOLOGICAL THOUGHT, 1650-65

By LAURENS LAUI)AN, M.A., P~.D.*

H I S T O R I A N S have never been able to come to any very satisfactory conclusions about the influence of Descartes on seventeenth-century English thought. Until very recently, it was thought that his impact was slight, significant--if at Ml--only in theology. In the last several years, however, historians of science have detected Cartesian strains in English mechanics, optics and physiology dating from the 1650s. 1 Gradually, therefore, the real and substantial role of Descartes is coming to be more fully appreciated. However, there is still one aspect of English philosophico-scientific thought where Descartes' positive impact is thought to be negligible, viz., with respect to theories of scientific method. Indeed, most historians who have dealt with the development of scientific method in Britain have written as if the seventeenth century could be understood simply as a series of footnotes to, and commentaries on, Bacon's Novum Organum. Not only is seventeenth-century English philosophy of science said to be Baconian, it is equally thought to represent a violent reaction against the a priori Cartesian model of science, with its emphasis on

* D e p a r t m e n t of H i s t o ry a n d Ph i lo sophy of Science, Un ive r s i t y College, London , W.C.1. The a u t h o r is g ra te fu l to t he U.S. Na t iona l Science F o u n d a t i o n for f inancial suppor t , and to t h e Roya l Society o f L o n d o n for access to i ts a rch ives a n d for permiss ion to quote f rom ma te r i a l of wh ich copyr igh t r ema ins t he p rope r ty of t he Society.

1 A m o n g t he more i m p o r t a n t accoun t s of Car tes ian inf luences in Br i ta in , see: M. Nieolson, ' T he E a r l y Stage o f Car tes ian i sm in E n g l a n d ', Studies in Philology, 1929, 26, 356-374; J .Saveson, ' Descar tes ' Inf luence on J o h n Smi th , Cambr idge P la ton i s t ', J. Hist. Ideas, 1959, ~0, 255-263; S. L a m p r e c h t , ' The Role of Descar tes in 17th Cen tu ry E n g l a n d ', Studies in the History of Ideas, Boulder , Colorado, 1935; E. Bu r t t , Metaphysical Foundations of Modern Science, New York, 1932, passim; a n d Marie Boas [Hall], ' The E s t a b l i s h m e n t of the Mechanica l Ph i losophy ', Osiris, 1952, 10, 412-541. F r o m the po in t o f view of th is paper , t he ar t ic le b y Mrs. Hal l is pa r t i cu la r ly va luab le in exh ib i t ing Descar tes ' inf luence on Eng l i sh theor ies of m a t t e r , a t t r a c t i on a n d pneuma t i c s .

Ann. of S~i.~Vol. 22, :No. 2. e

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

74 Laurens Laudan on

all-embracing systems. These two factors, veneration for Bacon and scorn for Descartes, are allegedly the major stimuli for English writings on method from Hobbes to Newton. But apart from its already suspicious tidiness and simplicity, this account has some profoundly disquieting features, not least of which is the fact tha t many English scientists and methodologists of this period were as vocal in their esteem for Descartes as in their idolatry for Bacon; indeed, many praised Descartes more lavishly than Bacon. More significant, however, than such pompous deference, several natural philosophers suggested that their accounts of scientific method were derived from, and perfectly compatible with, Descartes' views on the subject. Unless such scientists were seriously misled, we must critically re-examine the view of modern scholars like R. F. Jones who insist that Descartes' methodological ideas had negligible impact compared with Bacon's. ~ There are, of course, well-established precedents for Jones's claim. The experimental tenor of the early Royal Society and its almost pathological aversion to hypothetical system- building seem to be symptoms of a latent, but well-entrenched, anti- Cartesianism. Furthermore, Thomas Sprat, in his influential History of the Royal Society (1667), extols the virtues of the experimental philosophy and barely mentions Descartes, except as an example of bad physics. This account is further reinforced by the lip-service which most British methodologists paid to Bacon, constantly speaking in exemplary tones of ' t h e noble Verulam ', ' ou r illustrious Lord Bacon ', etc. But despite such plausible precedents, this picture of Bacon as the sole guiding light

For example, Jones asserts t h a t ' Exper imentM philosophy remains a th ing distinct from the mechanical [and hypothetical] and Bacon, who was the chief sponsor of the former, far outweighs in importance Descartes, who lent his great influence to the l a t t e r . . . Needless to say, the scientific movemen t in England in the thi rd quar ter of the seventeenth century • . . was largely inspired by the great Chancellor [Bacon] . . .' (Ancients and Moderns, St. Louis, 1961, p. 169). Elsewhere he notes, ' it is a mistake to thilfl~ Cartesianism inspired the scientific movemen t in England ' (ibid., p. 185). Jones even goes so far as to suggest t ha t this period in English science should be called the ' Bacon-faced generat ion ' (ibid., pp. 237 ft.).

F. W. Westaway, another wri ter who denies Descartes ' influence on English method- ology, asserts tha t ' Cartesianism took bu t slight hold in England ' (Scientific Method: Its Philosophy and Practise, London, 1919, p. 127). Wi th Boyle in particular, historians have been too quick to apply the :Baeonian label. Thus, Butterfield, in a long discussion of Boyle's ideas, clings tenaciously to the view tha t Boyle was a devout follower of Bacon, withou~ ever hint ing about a possible debt Boyle might owe to Descartes (cf. H. Butterfield, Origins of Modern Science, London, 1957, pp. 130-38). Marie Boas [Hall], taking a similar line, argues tha t Boyle's c orpuscularism (and the methodology which sustains it) was no t derived f rom Descartes bu t was, rather, ' an independent development along lines suggested b y Bacon ' (Osiris, 1952, 10~ p. 461): Recently, however, Mrs. I-Iall has conceded tha t ' t hough it was Bacon who mainly inspired Boyle, he was influenced by Descartes as well ' (Robert Boyle on l~atural Philosophy, Bloomington, Indiana, 1965, 63).

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 75

of British philosophy of science is too one-sided and seriously over- simplifies the diversity of the origins of English methodology in this period. While it is certainly true to say that such writers generally endorsed Baconian experimentalism, 8 it is not correct to think that they all accepted his inductivism as well. Many thinkers were quite sceptical about the possibility of discovering indubitable scientific principles b y any quasi-inductive process. In opposition to Bacon, they freely and enthusiastically accepted Descartes' suggestion that the scientist must be content with hypothetical principles and conjectures rather than true and valid inductions. Descartes' hypotheticalism, when blended with Baconian experimentalism, became ~ cornerstone of the methodologies of several English philosophers, especially Boyle, Glanvill and Locke. In the general enthusiasm for Bacon, however, Descartes' contributions to English methodological thought have been neither documented nor carefully assessed. This paper is not an a t tempt to deny Bacon's real contributions bu t seeks rather to focus attention on another equally important stimulant to English philosophy of science, Descartes.

Thus, I will argue, in contrast to Jones and other writers, that many of the major British methodologists derived their philosophies of science as much from Descartes as from Bacon. I t follows, as a corollary to this, that they were neither so inductive nor so opposed to speculation as has often been suggested. I will claim that Descartes' methodology (especially that developed in the latter half of the Principles) was a fertile source for discussions of method among the English thinkers; and especially that his view of the universe as a ' mechanical engine ' or clock whoso internal parts can only be conjectured about served as an important stimulus for the English writers on method. In sum, I want to investigate the extent to which the hypothetical method of several seventeenth-century English scientists and philosophers is derived from Descartes' version of that method.

Before we can understand the debt of the English hypothetiealists to Descartes, we must clarify the sense in which his methodology can be characterized as ' hypothetical '. Such emphasis has been placed on the a prioristic method, which he espouses in the Discourse on Method, that it may seem strange to suggest that he believed hypotheses to be indispens-

3 More of ten t h a n not , it was n o t even Baconian expe r imen t a l i sm t h a t was a p p l a u d e d b u t s imp ly expe r imen ta l i sm. The work of scholars such as F . R. J o h n s o n (Astronomical Thought in Renaissance England, Bal t imore , 1937) m a k e s it h igh ly doub t fu l w h e t h e r t h e expe r imen ta l spi r i t o f Engl i sh science can be a t t r i b u t e d to Bacon a t all. M a n y o f B a c o n ' s predecessors a n d con tempora r i e s (e.g. H a r v e y a n d Gilbert) were accompl i shed experi- men ta l i s t s long before t he appea rance of t he Novum Organum.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

76 Laurens Laudan on

able to science. ~ The myth tha t Descartes was a rigid rationalist, consistently opposed to all scientific conjecture, is now so widespread that, to counter-balance it, we must begin by summarizing briefly the methodology which he develops in the Principles, and which is even formulated in embryonic form in the Regulae, the Discourse, the Meditations and the Dioptrique. We must read the Principles, as the Englishmen of his time did, without bringing to it the prejudices that come from excessive pre-occupation with the a priorism of the Meditations. (Equally, we must read it without forcing it into anachronous categories such as ' rationalism ' or ' empiricism '. Such pigeon-holes, whimsical inventions of Enlightenment historians, viciously undermine any attempt to understand seventeenth-century science and philosophy.) The figures I will be discussing knew Descartes primarily through the Principles 5 and it was thus natural for them to assume that he adopted a modest pose about the possibility o f certainty in science, rather than the vain and omniscient posture of the Discourse.

Towards the end of the Fourth Par t of the Principles (1644), Descartes makes a surprising confession. After trying to deduce the particular characteristics of chemical change from his first principles (i.e., matter and motion), he concedes failure. His programme for the derivation of the phenomena of chemistry and physics from a priori truths remains uncompleted. His first principles are, he admits, simply too general to permit him to deduce statements from them about the specific way particular chunks of matter behave under particular conditions. I t is not that matter behaves in violation of these first principles; Descartes was too confident, and his principles too vague, for him to be forced to admit that. But the very generality of his principles made them practically useless for explaining and predicting particular events. 6 Not content with leaving anything unexplained,

4 :Historians are g radua l ly beg inn ing to recognize the impor t ance of Descar tes ' hypo- the t ica l m e t h o d a n d t h e f u n d a m e n t a l role it p layed in his ph i losophy of science. Especia l ly useful in th i s regard are G. B u e h d a h l ' s d iscuss ions in ' Desca r t e s ' An t i c ipa t ion o f a " Logic of Scientific D i scove ry" ', Scientific Change (ed. A. C. Crombie), London , 1962, pp. 399-417, a n d ' The l%elevance of Descar tes ' Ph i lo sophy for Modern Ph i lo sophy of Science' , Brit. J. Hist. Sci., 1963, 1~ 227-249. See also R. Blake, ' T h e Role of Exper ience in Descar tes ' T h e o r y of M e t h o d ', in Theories of Scientific Method (ed. E. Madden) , Seatt le, 1960, pp. 75-103.

5 A l t h o u g h a n Eng l i sh t r ans l a t ion of t h e Discourse appea red in L o n d o n in 1649, i t s c i rculat ion seems to h a v e been qui te l imited. A p p a r e n t l y Descar tes ' Passions of the Soul was widely c i rcula ted in Br i ta in , b u t since i t h a s lit t le of methodologica l in teres t , we shall neglect i t in our discussion.

6 As he p u t it in t h e Discourse: ' B u t I m u s t confess also t h a t the power o f n a t u r e is so v a s t a n d ample , a n d these principles are so s imple a n d general , t h a t I obse rved h a r d l y a n y par t i cu la r effect concern ing wh ich I could n o t a t once recognize t h a t it m i g h t be deduced f rom t h e principles in m a n y different w a y s a n d m y grea tes t difficulty is u sua l l y to discover i n which of these w a y s the effect does depend on t h e m ' (l~; Descar tes , Philosophical Works, t r ans . , I-Ialdane a n d Ross , N e w York, 1931, col. i, p. 121).

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 77

Descartes departed from his usual devotion to clear and distinct ideas and advocated the use of intermediate theories (less general than the first principles, but more general than the phenomena), which were sufficiently explicit to permit the explanation of individual events and which were, at the same time, compatible with, but not deducible from, the first principles. Descartes recognized that all such intermediary theories were necessarily hypothetical. Because they were not clearly and distinctly perceived, it was possible tha t they were wrong. After all, nature is describable in a wide variety of ways and the fact tha t an explanation worked was no proof tha t it was true. He was a sufficiently sophisticated logician to realize that ' one may deduce some very true and certain conclusions from suppositions that are false or uncer ta in ' /

Descartes goes on to suggest tha t we do not need assurances of t ruth in such matters. I t will suffice if we can give an account of how nature might behave, not necessarily how nature does behave. After all, his was a corpuscular philosophy which sought to explain the macroscopic world in terms of sub-microscopic particles. By definition, such particles were unobservable and so any specific properties we attr ibute to them (e.g., such-and-such a size, shape, and motion) can only be done tentatively and with a clear appreciation of their hypothetical character. We can, of course, be sure that they have some size, shape and motion (our first principles guarantee tha t much), but we remain forever in doubt about the particular properties they are given, s Descartes justifies this excur- sion into the hypothetical by means of a metaphor which was widely exploited by later English writers who, as eorpuscularians like Descartes, wanted some rational apology for their use of hypotheses. He suggests tha t we imagine the world on the analogy of a watch, whose face is visible but whose internal construction is forever excluded from view. In such a case, the most we can say about the mechanisms of the watch is conjectural opinion, not infallible knowledge. We canpropose mechanisms for how the internal parts of the watch might be arranged, though we can never, ex hypothesis, get inside to see if we are right. Because the watch might be constructed in any number of ways, it is sufficient if we outline some possible arrangement which would account for its external behaviour (e.g., hands moving, cuckoos calling and bells chiming). In the same way, the physicist has honoured his commitments so long as the mechanisms he proposes are compatible with the phenomena at hand. To ask for more than this is to misunderstand the limitations on the physicist. The passage itself reads as follows:

' I t may be retorted to this that, although I may have imagined

R. Descartes, Oeuvres (ed. Adam and Tannery) , P a r i s 1897-1957, vol. ii, p. 199. s , I frankly confess tha t concerning corporeal things, I know only this: tha t they can

be divided, shaped and moved in all sorts of ways . . . ' (ibid., vol, ix, p. 102).

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

78 Laurens L a u d a n o n

causes capable of producing effects similar to those we see, we should not conclude for tha t reason tha t those we see are produced by these causes; for just as an industrious watch-maker may make two watches which keep time equally well and without any difference in their external appearance, yet without any similarity in the composition of their wheels, so it is certain tha t God works in an infinity of diverse ways [each of which enables Him to make everything appear in the world as it does, without making it possible for the human mind to know which of all these ways He has decided to use]. And I believe I shall have done enough if the causes that I have listed are such that the effects they may produce are similar to those we see in the world, without being informed whether there are other ways in which they are produced. '~

The clock ana logy is not mere ly an a f t e r t hough t which Descar tes t h r ew in to i l lus t ra te his a rgument . Ra the r , i t fo rmed an in tegra l p a r t of his w a y of looking a t the world and the role he assigned to the corpuscular ph i losophy in explaining t h a t world. H e tells us t h a t machines like the clock served as models for developing his mechanica l accoun t of nature:

' And in this, the example of certain things made by human art was of no little assistance to me; for I recognize no difference between these machines and natural b o d i e s . . . '1°

To unde r s t and the significance of the clock ana logy and w h y i t led Descar tes to advoca t e a hypo the t i ca l me thod , we m u s t look careful ly a t his accoun t of scientific knowledge. Though he f requen t ly speaks of deducing the facts of physics f r o m his first principles, n he never offers a n y deduct ion which does in fac t exhaus t ive ly or uniquely expla in some par t i cu la r in t e rms of these v e r y general principles. We need m a n y o ther a s sumpt ions to explain w h y observable bodies behave the w a y t h e y do, and these assumpt ions cannot all be der ived f rom the first principles. Wheneve r Descar tes ac tua l ly tr ies to deduce optical and mechanica l p h e n o m e n a f rom the first principles, he pers i s ten t ly fails and m u s t fall b a c k on a v a r i e t y of hypo the t i ca l assumpt ions . Nor should we be s ta r t led to find t h a t the ma t t e r - i n -mo t ion p a r a d i g m is too compressed to enable us to expla in par t i cu la r events . Af ter all, eve ry b o d y has m a t t e r and mot ion , bu t only some are luminous, or magnet ic , or dense, or abras ive . Clearly someth ing else is invo lved which gives m a t t e r these characteris t ics . W a n t i n g to avoid a n y not ion of occult forces (because only m a t t e r and mot ion t r u ly exist), Descar tes finds t h a t the only w a y to expla in such

9 Ibid. , vol. ix, p. 322. The passage in square brackets only occurs in the French edition of the Principles , not the Latin.

lo Ibid. , vol. viii, p. 326. 11 Recall Descartes' classic remark that ' As for physics I should believe myself to know

nothing of it if I were only able to say how things may b% without demonstrating that they cannot be otherwise ' (ibid., vol. iii, p. 39),

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Mefaphor and Probabilism 79

properties as light and magnetism is to assume tha t bodies exhibiting these properties have a different sort of motion than bodies which do not have these properties. 12 Despite Descartes' claim that he could deduce everything in his optics from his clear and distinct ideas, he is continually forced to employ assumptions tha t do not follow from any knowledge we have of the first principles. He is compelled to make additional assump- tions about the constitution of moving matter. Such assumptions form an essential part of every scientific explanation. This difficulty is even clearer in the Principles where Descartes tries to explain physical and chemical change. Again, he concludes tha t the first principles are too general to permit us to explain any event uniquely, and in saying as much, he finally comes to grips with the problem plaguing his a priori physics. In addition to the first principles, we need a set of principles of lower generality which will enable us to discover the specific mechanisms of nature. But these less general prineicles cannot be deduced from the first principles; and Descartes was certainly aware that these principles of intermediate generality were not deducible from his metaphysical strictures about nature. Buchdahl has correctly noted tha t ' I t is a scholar's legend tha t Descartes consistently believed that his physics was deducible from first principles. . . ,13.

Descartes' endorsement of the hypothetical method is most explicit in tha t section of the Princ@les where he develops the doctrine of the three elements, which he used extensively to explain chemical and physical change. Among the assumptions of this theory is the claim tha t matter is corpuscular and tha t these corpuscles have a certain size, shape, and velocity. He says of these assumptions:

' we cannot determine by reason how big these pieces of matter are, how quickly they move, or what circles they describe.. . [this] is a thing we must learn from observation. Therefore, we are free to make any assumptions we like about them, so long as all the consequences agree with experience.. . '. ~

i2 I n the Diop~ics, for example , DesearSes tries to exp la in different coloured r a y s o f l ight . He sugges t s t h a t l ight is composed of spher ica l corpuscles m o v i n g in s t r a igh t l ines a t infini te velocity, a n d wi th a ro ta t iona l m o t i o n a b o u t the i r centres . Different eoloured r ays of l ight are due to the differential speeds of ax ia t ro t a t i on wh ich the corpuscles can a s sume . A fas t sp in appear s to be red l ight , a m o d e r a t e sp in as yellow, and a slow sp in as blue. Now, Descar tes ha s succeeded in expla in ing t h e p h e n o m e n a of colour w i t h o u t recourse to ent i t ies excep t ma t t e r - i n -mo t i on . B u t a t t he s a m e t ime , he h a s been forced to go beyond the knowledge g iven b y the f irst pr inciples to hypothes ize , w i th ne i the r empir ical evidence no r a priori reasons , t h a t different a t o m s o f m a t t e r r e , a t e a t d i f ferent speeds a n d t h a t such ro ta t ion is the cause of colour. T h u s , he deduces t he p h e n o m e n a of colour f rom t he con junc t ion of t he first pr inciples ' m a t t e r ' a n d ' m o t i o n ' and a n a s s u m p t i o n a b o u t differential speeds of ro ta t ion .

i3 G. J3uchdahl , in Scientific Change (ed. A. C. Crombie) , London , 1962, p. 411. 14 1%. Descar tes , Oeuvres, 1897-1957, vol, ix, p, 325,

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

8O Laurens Laudan on

Here again, matter and motion are too general to explain the phenomena. We must, he insists, resort to less general hypotheses about the size and configuration of matter in order to explain the world. The scientist thus resembles the skilled watch-maker of the analogy who is given a watch but cannot see its internal mechanisms. Like the watch-maker, he knows the general principles which govern his subject matter, but he is uncertain about the way they exhibit themselves in any particular case. Equally like the watch-maker, the scientist can offer only conjectures about internal construction and mechanisms.

The role of the first principles in physics is thus to circumscribe the range of acceptable hypotheses by excluding certain entities. Our first principles tell us, for example, not to develop a science based on the hypothesis of a void; they warn us against hypotheses couched in the teleological language of final causality; and they forbid hypotheses postu- lating action at a distance. Viewed in this light, corpuscular metaphysics does not dictate which physical system we adopt, but only gives us certain regulative rules. Matter and motion thus function much as Occam's razor or the assumption of nature's uniformity operate in modern science. Physical hypotheses must be compatible with such regulative principles, but they are not deducible from them.

But though Descartes concedes that science is necessarily hypothetical and probabilistic, he is not willing to say tha t all hypotheses are equally good or that the scientist can never be confident about his principles. He does suggest that mere ad hoc hypotheses, invented to explain one particular phenomenon, are not very convincing. :But, he insists, when we put forward a hypothesis which accounts for a wide variety of phenomena successfully, we can be reasonably confident (though not certain) that it is true:

' Although there exist several individual effects to which it is easy to adjust diverse causes [i.e., hypotheses], one to each, it is however not so easy to adjust one and the same [hypothesis] to several different effects, unless it be the true one from which they proceed. ,15

While insisting on the conjectural character of scientific hypotheses, Descartes was careful not to succumb to the sceptic's temptation to grant all hypotheses equal status and improbability. He clearly declared the right of the scientist to believe those hypotheses which accounted for a wide cross-section of the facts at hand: ' it is not likely tha t tha t from which one may deduce all the phenomena is false ,.6 Thus, a sound hypothesis is one which is both compatible with the data and with the first principles, matter and motion. On Descartes' view, the logical gap separating the first principles from the phenomena can be bridged only

1~ 1bid . , vol. ii, p, 199. ~t~ 1bid , , vol. ix, p, 123.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 81

by hypotheses. Since compatibility rather than deducibility is the relation between the first principles and the hypotheses of physics, the first principles function in the same way, vis-a-vis the hypotheses, as the facts do. Our first principles, like the data, can inform us tha t certain hypotheses are wrong; but they cannot tell us which hypotheses are right. We can never get inside nature's clock to see if nature's mechanisms are what we think them to be. However, the clock analogy is important, not only for the considerable light it throws on Descartes' use of the method of hypothesis, but equally because it, or variants on it, were widely cited by subsequent writers who, as corpuscularians, were struggling with the same methodological problems. 17 In particular, it was used by many English writers (Boyle, Glanvill, Power and Locke) who, historians tell us, were Baconian experimentalists, uninfluenced by Descartes. I t thus provides a convenient motif in terms of which to explain the development of the method of hypothesis between Descartes and Newton

B O Y L E AND C A R T E S I A N :PROBABILISM

We now turn to consider the more general theme of this paper, namely, the impact of Descartes' hypothetical method on English writers of the following generation, is His influence can be seen most prominently in the work of Robert Boyle, who did much to fuse the Baeonian and Cartesian traditions into a coherent and sophisticated view of scientific method. Because Boyle's philosophy of science unifies major elements from both Descartes and Bacon, it must be understood in the context of the traditions which those two writers initiated. 19 In the mid-seventeenth century, Baconianism and Cartesianism signified quite different things

17 D. J . de Solla Price ( ' A u t o m a t a and t h e Origins of M e c h a n i s m and Mechanis t ic Ph i l o sophy ', Technology and Culture, 1964, 5, 9-23) h a s del ineated t h e f u n d a m e n t a l role t h a t clocks a n d o ther a u t o m a t a p layed as analogies for the mechan i ca l a n d corpuscula r scient is ts . However , Pr ice ha s no t d r a w n a t t en t i on to the methodological ramif ica t ions of t he c lock-anMogy which I i n t end to discuss in th i s paper .

is I m u s t m a k e i t as explici t as possible t h a t th i s paper is not a n a t t e m p t to da te precisely the in t roduc t ion of Car tes ian ideas~into Eng l i sh science and phi losophy. For t h a t , we shou ld p robab ly h a v e to look closely a t Hobbes , Digby , Char le ton, C u d w o r t h and More r a the r t h a n Boyle a n d t he o ther wri ters I deM with. My g e m is a r a the r different one, name ly , to sugges t t h a t there are cer ta in Car tes ian s t r a ins which loom large in Boy le ' s methodologica l wri t ings. W h e t h e r t h e y came di rect ly f rom Descar tes or t h r o u g h an i n t e r m e d i a r y source is a separa te ques t ion which I t ouch only incidental ly . (Cf. L. Gysi, Platonism and Cartesianism in the Philosophy of Ralph Cudworth, Bern , 1962.)

1~ p . p . Wiene r ' s art icle on ~Boyle's ph i lo sophy of science {' The expe r imen t a l ph i losophy of Robe r t Boyle ', Phil. Rev., I932, 41, 594-609), one of the earl iest on th i s topic, m u s t be r ead wi th m u c h caut ion . Wiene r m i squo t e s Boyle on two occasions (footnotes 6 a n d 22), a n d once draws inferences f r om the t e x t which it could no t poss ib ly suppor t (footnote 5). However , Wiene r does a capable job o f showing Boyle ' s c o n t e m p t for, a n d ignorance of, ancient and medievM science.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

82 Laurens Laudan o n

from what they do today. Bacon was not praised (or condemned) as an inductive philosopher so much as a n experimental one. Descartes, on the other hand, was not treated as an a priorist , but rather as an advocate of the corpuscular philosophy who encouraged the use of hypotheses in science. Boyle borrowed Bacon's experimentalism and Descartes' hypothetical corpuscularism, while prudently overlooking the inductive excesses of Bacon and the rationalistic strictures of Descartes. So far as we can judge from his published works, it never occurred to Boyle tha t the principles of science could be discovered either inductively or a priori . By neglecting Bacon's inductivism and Descartes' rationalism, Boyle viewed the methodologies of these two authors as healthy contrasts of emphasis within a commonly-held scientific world-view.

Nor should we be startled to find resemblances between the ' experi- mentalist ' Boyle and the ' ra t ional is t ' Descartes. After all and above all, they were both corpuscularians. Boyle began his philosophical education with heavy doses from Descartes; ~° and, as we have seen, the Descartes of the Pr inc ip le s was neither so a pr ior i nor so anti-experimental

~0 Boyle 's unpubl ished manuscr ip ts leave absolutely no doubt tha t he had read Descartbs, and more t h a n once. F r o m his earliest papers on na tura l phi losophy unti l his last ones, he made repeated references to Descartes and the Cartesians. There is one part icular ly interesting passage in which, discarding his normal humili ty, he candidly asesses the contr ibut ions of several impor tan t seventeenth-century scientific figures. His admira t ion for Descartes is certainly undisguised: ' t tobbes est obseur sans agr6ment, singulier e n s e s id6es, scavant , mais peu solide, inconstant dans sa doctrine: car il est t an res t Epicurien, t anres t Peripateticien. Boile est exact dans ses observations: il n ' y a personne on l 'Europe qui air enrichy la philosophic de ran t d'experiences que luy: il raissone assez consequemment sur ses experiences, lesquelles apr6s tou t ne sen t pas toujours indubitables: p arce que eesprincipes ne sent pas toujours e e r t a i n s . . . Gassendi, qui n ' a voulu passer que pour res taura teur de la philosophic de Demoerite et d 'Epieure, parle peu de son chef, il n ' a presque rien de luy, que la beaut6 du stile, pa r ou il peut passer pour un au teur admirable: pour le refuter dans sa physique, on n ' a besoin que des argumens d 'Aristote centre Democrito et ses disciples. Descartes est un genie des plus extraordinaires qul air pa ru darts ces de/'niers temps, d ' un esprit fertile, et d 'une medi ta t ion profound: L ' enchainement de sa doctrine va ~ son but , l 'ordre e n e s t bien imaging, selon ses prineipes: et son systeme, tou t reel6 qu'il est d 'ancien et de moderne, est bien arrang6. A la verit6 il enseigne t rop ~ douter: et ce n 'es t pas un

A m u c h more definitive account of Boyle's scientific method is to be found in the second chapter of M. Mandelbaum, Philosophy, Science, and Sense Perception, Baltimore, 1964, esp. pp. 88-112.

One should also ment ion R. Westfall 's useful ' Unpubl ished Boyle Papers Relating to Scientific Method ', Ann. Sci., 1956, 12, 63-73 and 103-117; Marie Boas [Hall], 'La M~thodologie Scicntifique de Rober t Boyle ', Rev. Hist. Sci., 1956, 9, 105-125; and A. 1%. and M. B. Hall, ' Phi losophy and :Natural Philosophy: Boyle and Spinoza ', in Mdlanges Alexandre Koyrd (ed. I . B. Cohen and ~ . Taton), Paris, 1964, vol. ii, pp. 241-256. Older, bu t still useful, sources are S. Mendelssohn, Robert Boyle als Philosoph, Wiirzburg, 1902, and G. Sprigg, ' The Honorable Rober t Boyle: A Chapter in the Phi losophy of Science ', Archeion, 1929, 11, 1-12.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 83

as we now judge him to be on the basis of his Discourse and Meditations. The Descartes of the Principles is not the proponent of systematic doubt so much as the modest inquirer after t ruth who admits, especially through- out the latter half of the Principles, tha t science is an hypothetical and conjectural enterprise which offers its followers only a probable story, not the revealed truth. Nor is it inconsistent for Boyle to draw from both Descartes and Bacon, for their methodologies, as Boyle construed them, were not contradictory. Indeed, Descartes' hypothetical me thod can be viewed (and was so viewed by Boyle) as an alternative formulation of Bacon's hypothetical Indulgence of the Understanding. 21

Any at tempt to explain Boyle's methodology must, of course, begin with the fact that he was an ardent adherent of the corpuscular or mechanical philosophy. ~2 Indeed, it is his corpuscularism which conditioned his whole approach to nature and which inclined him to adopt Descartes' method of hypothesis, while simultaneously taking his experimentMism from Bacon. As an advocate of the corpuscular philosophy, and as a writer firmly in the experimental tradition, Boyle

31 :Boyle was n o t t he first to sugges t t h e s imi lar i ty be tween B a c o n a n d Descar tes . T h e

a n o n y m o u s t r ans l a to r o f Desca r t e s ' Passions of the Soule, London , 1650, insis ts , in a n ' a d v e r t i s m e n t ' a p p e n d e d to t h a t work, t h a t t h o u g h ' m o s t m e n conceive n o t how necessa ry e x p e r i m e n t s are , ' Desca r t e s and B a c o n ' h a d the bes t no t ions , concern ing t he m e t h o d to be he ld to b r ing t he P h y s i c k s to the i r perfec t ion ' .

2~ The ve ry fac t t h a t Boyle chose to call h i s doct r ine the corpuscu la r ph i lo sophy is indica t ive of his filiations wi th Descar tes . Pr ior to Boyle, it was c o m m o n to d i s t ingu i sh th ree d i s t inc t theore t ica l s y s t e m s : the Aris to te l ian , t h e Car tes ian a n d the A tomic or E p i c u r e a n or Gassendian . Boyle r igh t ly po in ted out t h a t such a classif icat ion obscured t he considerable area of a g r e e m e n t be tween t h e Car tes ian a n d Atorais t ic p a r a d i g m s . R a t h e r t h a n call h i mse l f a n a tomis t , a n d t h e r e b y side wi th Gassend i aga ins t Descar tes , Boyle defines a more genera l pos i t ion (' t h e corpuscular ph i lo sophy ') wh ich pe rmi t s h i m to consider h imse l f in a Car tes ian t r ad i t ion while still s id ing w i th the A t o m i s t s on m a n y specific po in t s of in te rpre ta t ion . He p u t s i t th is way : ° . . . I considered t h a t t h e Atomica l a n d Car tes ian hypo theses , t h o u g h t h e y differed in some ma te r i a l po in t s f rom one ano the r , y e t in opposi t ion to t he Per ipa te t i c a n d o ther v u l g a r doct r ines t h e y m i g h t be looked u p o n as one ph i losophy . . . [for] b o t h the Car tes ians a n d t he A t o m i s t s expl ica te the s a m e p h e n o m e n a b y litt le bodies va r ious ly f igured a n d m o v e d . . . the i r hypo theses m i g h t b y a pe r son of reconcil ing disposi t ion be looked u p o n as one ph i losophy ' {Works, ed. Birch , London , 1772, vol. i, pp. 355-356).

Boyle un rese rved ly regards D e s c a r t e s - - m o r e even than Gassend i or Bacon--as the mechan ica l cam corpuscula r phi losopher par excellence : ' T h a t s t r i c t ph i losopher Descar tes who ha s wi th g rea t wi t a n d no less app lause a t t e m p t e d to ca r ry t he mechan ica l l powers h igher t h a n a n y of t he m o d e r n phi losophers a n d app ly i t to expl icate t h ings mechan i ca l l y ' (Royal Society, Boyle Papers, vol. ii, f. 137: cf. also Works, vol. iii, p. 558).

b e n raodele ~ des espr i t s n a t u r a l l e m e n t ineredules: m a i s enfin il es t p lu s original que l e s

au t r e s . . . En f in Galilei es t le p lus agreable des mode rnes , B a c o n le p lus subt i l , Gassend i le p lus s cavan t , H o b b e s le p lus resveur , Boyle le p lus eur ieux, Descar tes le p lus ingenieuc, V a n h e l m o n t le p lus na tu ra l i s t e : ma i s t rop a t t aeh6 a P a r a c e l s e ' {Royal Society, Boyle Papers, vol. xliv).

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

84 Laurens L a u d a n on

was acute ly aware of the immense gap separat ing the principles of corpuscularism from par t icular sciences such as chemis t ry and physiology. Though corpuscularians often scolded Aristotle for neglecting exper iment , t he y themselves were subject to the same charge for, in general, few a t t e mp t s had been made to use the mat te r - in -mot ion parad igm to explain the par t icu lar facts of physics or chemistry. Boyle perceived t ha t this was a serious weakness in the mechanical p rogramme:

' But I am sorry to see cause to add to what I have been saying, that as much as we [Corpuscular philosophers] magnify the necessity of experiments in our contacts with the Peripateticks about nature, we seem not yet to be sensible of this acknowledged necessity, when we contest with the particular difficulties that frequently occur, when we ourselves are to discover the cause of her phenomena, or to imploy her productions.' 2a

I f the corpuscular phi losophy is to be useful to na tura l philosophy, i t mus t do more t han propose a few vague principles which i t asserts to be compat ible wi th nature . I t mus t use those principles and others to explain wha t can be observed in the laboratory . I t s imply is not enough to say t h a t fire boils water because the rap id ly moving corpuscles of fire break up water clusters and send vapour to the surface. We must p a y careful a t t en t ion to describing the par t icu lar shape and veloci ty of fire corpuscles, to the mechanisms whereby t h ey break up liquid clusters and to the laws of boiling. More generally, the corpuscular phi losophy mus t cease to be mere ly a set of ambiguous metaphysica l principles which are so fluid t ha t t h e y are compat ible with an y phenomenon. Boyle wants to t rans form the corpuscular doctr ine into a sensible physical t heo ry which makes predict ions and provides explanat ions; in short, into a theory which approaches experience in order to learn f rom it and which stakes its fate, not on the philosopher 's abi l i ty to weave int r icate m y t h s and devise acl hoc adjus tments , bu t on the scientist 's abi l i ty to confirm those principles.

Boyle never seriously doubts t ha t na tu re is u l t imate ly mat ter- in- motion, ~4 bu t he insists t h a t we need to go beyond such crypt ic formulae if we are to have a science wor thy of the name:

' For it is one thing to be able to shaw it possible for such and such effects to proceed from the various magnitudes, shapes, motions, and concretions of atoms, and another to be able to declare what precise and determinate figures, sizes, and motions of atoms, will suffice to make out the proposed phenomena.' 25

38 Royal Society, Boyle Papers, vol. ix, f. 1. 34 As T. S. K u h n p u t s it: ' Ne i ther [Boyle 's] eclect ic ism nor his scept ic ism e x t e n d s to

doub t s t h a t some corpuscula r m e c h a n i s m under l ies each inorganic p h e n o m e n o n he invest i - gates . ' (' Robe r t :Boyle a n d S t ruc tu ra l C hemi s t ry in t he Seven t een th Cen tu ry , ' Isis, 1952, 48, 19).

za Robe r t Boyle, Works, 1772, vol. ii, p. 45.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 85

W e m u s t fo rmula te i n t e rmed ia ry theories which are a t once less general and more explicit t h a n m a t t e r and mot ion . L ike Descar tes in the Principles, Boyle real ized t h a t the typ ica l corpuscular ian doctr ine is too general to p e r m i t one to explain, in detail , the behaviour of ma t t e r . W e m u s t develop lower-level theories which, while compat ib le wi th the corpuscular phi losophy, are not s t r ic t ly deducible f rom it:

' There are a great many things w h i c h . . , cannot with any convenience be immediately deduced from the first and simplest principles; namely, mat ter and motion; but must be derived from subordinate principles; such as gravity, fermentation, springiness, magnetism, etc.' 26

I n t he ideal case, we should seek to derive eve ry th ing f rom the first principles. Unfo r tuna t e ly , however , there is a wide d iscrepancy be tween w h a t we hope for and w h a t we are p repa red to accept :

' That we may aspire to, but must not always require or expect, such a knowledge of things, as is immediately derived from their first principles.' ~

E v e n where our exp lana t ions are no t der ivable f rom mechanica l principles, t h e y m u s t be compa t ib le wi th t hem:

' the mechanical principles are so universal, and therefore applicable to so many things, that they are rather fitted to include, than necessitated to exclude, any other [subordinate] hypothesis, that is founded in nature, as far as it is so. And such hypotheses . . . will be found, as far as they have t ru th in them, to be legitimately (though perhaps not immediately) deducible from the mechanical principles, or fairly reconcileable to them . . .' as

E v e r y subord ina te hypothesis , in so far as it is t rue , is e i ther deducible from, or a t leas t compat ib le with, the corpuscular phi losophy. Descar tes m a d e precisely the same poin t in the Principles. So far as na tu ra l ph i losophy is concerned, these less general hypotheses are even more useful t h a n the concepts of m a t t e r and mot ion :

' T h e most useful notions we have in physicks . . . are not derived immediately from the first principles; but from intermediate theories, notions, and rules.' 29

~6 RoyaISoeiety, Boyle Papers, vol. ix, L 40. Elsewhere , he m a k e s t he po in t this way : i t would ' be backward to re ject or despise all expl ica t ions t h a t are no t i m m e d i a t e l y deduced f rom the shape , b igness a n d m o t i o n of a t o m s or o ther insensible par t ic les of m a t t e r . . . [for those who] p r e t end to expl icate every p h e n o m e n o n b y deduc ing it f rom the mechan ica l affect ions of a t o m s u n d e r t a k e a ha rder t a sk t h a n t h e y imagine ' (ibid., vol. viii, f. 166).

~7 Ibid., vol. viii, f. 184. Unde r l i ned in original. Cf. foo tnote 65 below. 2s 1%. Boyle, Works, 1772, vol. iv, p. 72. 2~ Royal Society, Boyle Papers, vol. ix, f. 40.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

86 Laurens Laudan o n

Boyle set himself the life-long task of enunciating such 'sub-ordinate principles' and ' intermediate theories ,a0 in order to provide scientific flesh for the metaphysical skeleton of corpuscularism:

' I thought it would be no slight service, not only to the Corpuscular hypothesis, but to natural philosophy itself, if I could by good experi- ments, and at least probable reasons, make out that almost all sorts of particular qualities may be mechanically originated or produced.' al

Having settled on such an undertaking, it was natural tha t Boyle should give some thought to the method whereby these subordinate principles could be discovered and confirmed. I t is at this stage that we see a remarkable blend of Baconian and Cartesian elements. With Bacon, Boyle emphasized that the proper foundation of physical know- ledge was experimentation; not merely casual observation of nature, but systematic and often artificial tinkering with the physical world so as to observe it under a wide variety of circumstances. Good natural philosophers, he writes,

' consult experience both more frequently and more heedfully [than the Aristotelians]; and, not content with the phenomena that nature spontaneously affords them, they are solicitous, when they find it needful, to enlarge their experience by trials purposely devised . . .' a2

Like Bacon, he envisaged the compilation of vast histories of nature which would summarize and codify the information gleaned from experiment, sa To this end, Boyle himself wrote experimental histories of fluidity, firmness, colours, cold, air, respiration, condensation, flames, human blood, porosity, liquors, t in and fire. s4 But what are we to do with such natural histories once they are compiled? Can we use them to induce, Baconian fashion, the principles and laws of science? Boyle's answer to this question is an unequivocal 'No '. Though a self- styled pupil of Bacon, Boyle never, to my knowledge, uses the term

so Boyle even sugges ts t h a t the subord ina te hypo theses m a y be the only ones which can

be f i rmly es tab l i shed: ' Though m e n be no t a r r ived a t such a p i t ch of knowledge as to be able to d iscover a nd so lemnely-es tab l i shed [sic] complea t and general hypotheses; ye t

subo rd ina t e axioms a nd hypotheses.., m a y be of v a s t use bo th in ph i losophy and to h u m a n life ' (ibid., vol. ix, £ 61).

81 Royal Society, Boyle Papers, vol. ix, f. 28. s2 1%. Boyle, WorIcs, 1772, vol. v, pp. 513-514.

3a :Boyle was as dogged as Bacon in p u t t i n g h is tor ies of n a t u r e h igh on his l is t of

pr ior i t ies : ' . . . we e v i d e n t l y w a n t t h a t upon wh ich a theory , to be solid a n d useful, m u s t be bu i l t ; I m e a n an e xpe r im e n t a l h i s t o ry . . . A n d th i s we so wan t , t h a t excep t pe rhaps

w h a t m a t h e m a t i c i a n s h a v e done concerning sounds, a n d the observa t ions ( ra ther t h a n exper iments ) t h a t our i l lus t r ious V e ru l a m h a t h (in some few pages) said of h e a t in his shor t

Essay de t"orma Calidi; I k n o w no t a n y one q u a l i t y of which a n y a u t h o r has g iven us a n a n y t h i n g c o m p e t e n t h i s to ry ' (ibid., vol. iii, p. 12).

3a See the t ab l e of con ten t s to Boyle ' s s ix -vo lume Works for references to these histories.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 87

' induct ion ,3~ nor does he ever seriously consider Bacon ' s v iew t h a t principles will emerge in a n y mechanica l way f rom a s t udy of nature , as So far is Boyle f rom Bacon ' s concept ion of a r igidly exper imenta l science t h a t he even suggests t h a t a good hypothes i s is more va luab le t h a n a well-conceived exper imen t :

' And tho' perhaps few have a greater love and value for experiments than I, yet for my part , I should think myself mere obliged to him that discovers to me some pregnant notion . . . than if he imparted some fine experiment. ' 87

F a r f rom following Bacon ' s induct ive line, Boyle takes a more Cartesian pose. The purpose of all this expe r imen ta t ion , he tells us, is to place us in a posi t ion to offer some hypothes is as a t e n t a t i v e account of the d a t a which we have so accumula ted .

The expe r imen ta l histories thus serve as the raw ma te r i a l for the theor is t who then proceeds to m a k e conjectures which are t es ted in t e rms of the i r abi l i ty to ' render an intelligible account of the causes of the effects, or p h e n o m e n a proposed [in the histories] ,.as Boyle does no t believe t h a t theories will arise r e a d y - m a d e f rom the data , or t h a t the d a t a will un ique ly de te rmine a n y single theory , s9 The d a t a are i m p o r t a n t

s5 Birch, t he editor o f Boy le ' s Works, records no usage o f ' i nduc t ion ' a m o n g the wr i t ings con ta ined in those volumes . U n f o r t u n a t e l y , Boyle ' s unpub l i shed MSS. in the l~oyal Society L ib ra ry are no t indexed.

as I do n o t m e a n to sugges t b y th i s t h a t Bacon ' s m e t h o d o l o g y left no r o o m for the hypo the t i ca l m e t h o d or t h a t it was t he exc lus ive i nven t ion of Descar tes . To t h e con t ra ry , I believe :Bacon's ' I ndu lgence of the U n d e r s t a n d i n g or Fi rs t V in t age ' is an unequ ivoca l s t a t e m e n t of t he hypo the t i ca l me t hod . See :N[. Hesse ' s chap te r on :Bacon in A Critical History of Western Philosophy (ed. O'Connor) , :New York , 1964, he r ' Hookc ' s D e v e l o p m e n t o f :Bacon's Me thod 'J Ithaca, (Actes d~ 10 e Cong~'~s International d'Itistoire des Sciences), Par i s , 1964, vol. i, pp . 265-268, and he r ' Hooke ' s :Philosophical Algebra ', Isis, for th- coming, for a d iscuss ion of t he hypo the t i ca l aspec t s of Bacon ' s me thodo logy . However , I t h i n k t he m y t h of B a c o n the an t i -hypo the t i ca l i s t has th i s m u c h t r u t h in it : t h a t Bacon ' s hypo the t i ca l I ndu l gence o f the U n d e r s t a n d i n g was conceived b y h i m as a s t r ic t ly t e m p o r a r y m e a s u r e unt i l t h e n a t u r a l h is tor ies became sufficiently comple te to pe rmi t t h e fool-proof mechan ica l i nduc t ion wh ich B a c o n out l ines in t h e Novunt Organum. B a c o n was a d a m a n t l y opposed to the v iew (suppor ted b y bo t h Descar tes and :Boyle) t h a t science m u s t be e te rna l ly conjec tura l a n d uncer ta in . F u r t h e r m o r e , Bacon leaves comple te ly u n m e n t i o n e d the par t i cu la r s of his F i r s t Vin tage a n d so I t h i n k i t is to Descar tes , r a t he r t h a n Bacon , t h a t Boyle t u r n e d for t he detai ls of his me thodo logy . B a c o n could neve r have accepted, as Boyle en thus ias t i ca l ly did, Descar tes ' c lock-maker ana logy a n d i ts impl ica t ions for t he severe l imi ta t ions on scientific knowledge.

a7 Royal Society, Boyle Papers, vol. ix, f. 105. Boyle ~ells u s t h a t one of t he p r i m a r y func t ions of e x p e r i m e n t a t i o n is ' to sugges~ h y p o t h e s e s ' (ibid., f. 30).

as 1~. Boyle, Works, 1772, vol. iv, p. 234. a~ Boyle is qui te b l u n t in his cr i t ic ism of t he view t h a t expe r imen t alone will lead u s to

t rue theories . ' . . . he, t h a t es tabl i shes a theory , wh ich he expec t s shal l be acquiesced in b y all succeeding t i m e s . . , m u s t no t only have a care, t h a t none of the p h e n o m e n a of na tu re , t h a t are a l r eady t a k e n not ice of, do con t rad ic t his h y p o t h e s i s a t t he p re sen t [time], bug t h a t no p h e n o m e n a , t h a t m a y be hereaf te r discovered, shall do i t for the fu ture . ' B u t

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

88 Laurens Laudan on

because without them we might accept a theory which would have been falsified if we had experimented more thoroughly. But it is the faculty of reason which constructs theories from the data; they do not spring full-blown from the histories. No matter how extensive our experimenta- tion, science remains fundamentally hypothetical. 4°

These ideas are similar, in expression as well as content, to Descartes' methodological position in the Principles and speak forcefully against those who minimize his impact on English thought in this period. But it might be argued tha t all I have said to this point merely suggests that Boyle and Descartes were both hypotheticalists, but not necessarily tha t Boyle derived his hypothetical method from Descartes. This perhaps seems just another of those coincidences that continually mislead historians of ideas. That Boyle had similar views twenty years after Descartes is, of itself, meagre proof of his debt to Descartes. In history as well as logic, post hoc is no guarantee of propter hoc. Fortunately, however, this is not the only evidence we can cite for the claim tha t Descartes exerted substantial influence on English methodological thought, especially Boyle's. There is a passage in Boyle's The Usefulness of Natural Philosophy (1663) which makes our case a good deal more cogent. For Boyle there takes Descartes' clock analogy and, by clumsily paraphrasing it, uses it to just i fy--much as Descartes d id--an avowedly hypothetical and corpuscular methodology. Boyle formulates the analogy thus:

' . . . many Atomists and other Naturalists, presume to know the true and genuine causes of the things they attempt to explicate; yet very often the utmost they can attain to, in their explications, is, that the explicated phenomena may be produced after such a manner, as they deliver, but not that they really are so. For as an artificer can set all the wheels of a clock a going, as well with springs as with weights . . . . so the same effects may be produced by divers causes different from one another; and it will oftentimes be very difficult, if not impossible, for our dim

40 Boyle writes: ' i t is some t i mes conducive to the d iscovery of t r u t h , to pe rmi t t he u n d e r s t a n d i n g to m a k e a n hypothes i s , in order to t he expl ica t ion of th i s or t h a t difficulty, t h a t b y e x a m i n i n g how far t he p h e n o m e n a are, or are not , capable of be ing so lved b y t h a t hypo thes i s , t he u n d e r s t a n d i n g m a y , even by i ts own errors, be i n s t ruc t ed ' (ibid., vol. i, p. 303). This s t a t e m e n t , a long wi th severa l others , could be ci ted as coun te r -ev idence to Jones ' asser t ion t h a t ' Boyle was inf luenced by t he comprehens ive charac ter i s t ic of t3aeon's ph i lo sophy in t h i n k i n g t h a t all t h e evidence m u s t be in before a genera l iza t ion shou ld be d r a w n ' (op. cir. foo tnote 2, p. 164). Boyle clearly, a n d even Bacon dimly , perce ived the necess i ty of m a k i n g hypo theses a n d genera l iza t ions before all t he evidence was collected.

consider ing ' how incomplea t t he h i s to ry of n a t u r e we ye t h a v e is, a n d how difficult it is to bui ld an accura te hypo t he s i s u p o n an incomple te h i s to ry of t he p h e n o m e n a ', we can neve r s ay w i th ce r t a in ty t h a t our theories are t rue (ibid., vol. iv, p. 59).

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabillsm 89

reasons to discern surely, which of those several ways, whereby it is possible for nature to produce the same phenomena, she [nature] has really made use of to exhibit them.' 41

I am not maintaining tha t Descartes was the first to liken na ture to a clock-like mechanism; on the contrary , this was a common me taphor among mechanist ic philosophers th roughou t the sixteenth and early seventeenth centuries. The fact t ha t English writers like Boyle also used the clock analogy is, of itself, no indication of their Cartesian leanings. However , Descartes was (so far as I can tell) the first to use the analogy to justify a hypothet ical view of knowledge and science: Tha t Boyle and others used the clock analogy in precisely the way Descartes did-- to buttress up a hypothet ico-deduct ive me thodo logy - - i s p robably indicative of their Cartesian leanings. This clock metaphor , borrowed f rom Descartes, par t icular ly s t ruck Boyle 's fancy; so m u c h so tha t he remained persuaded for the rest of his life tha t science could a t ta in only probable, not infallible, knowledge. The language in this passage is Boyle 's; bu t the t hough t is clearly Descartes ' . B o t h insist t h a t our theories only describe the mechanisms whereby nature might conceivably produce the effects we observe, not necessarily the mechanisms which nature in fact uses. I n another passage which reminds us of Descartes, Boyle puts the point this way:

' . . . it is a very easy mistake to conclude, that because an effect may be produced by such determinate causes, it must be so [produced], or actually is so.' 4~

He shared with Descartes the belief t ha t m a t t e r and mot ion are the ul t imate and t rue principles of physical science, and he insisted, again like Descartes, t h a t all subordinate principles, in terms of which we explain part icular events, are necessarily conjectural. While declaring himself a faithful atomist , Boyle was ve ry sceptical about the possibility of the 'Atomical Hypothes is ' ever becoming more t h a n a probable theory; and, for t ha t mat ter , he was equally pessimistic about ever discovering any of the t rue mechanisms of nature. 4a Like Descartes, Boyle is careful not to confuse verification with proof. I f a hypothesis has accounted for all the phenomena, then it has demons t ra ted its utility, bu t its validity is still an open question and forever remains so. We may , b y chance,

41 R. Boyle, Works, 1772, vol. ii, p. 4-5. Boyle actually formulates the clock analogy on several occasions, which suggests that it played a basic role in his thinking about science and method. This opinion is confirmed when Boyle says, in introducing the clock metaphor: ' To explain this a little, let us assume the often mentioned, and often to be mentioned, instance of a clock ' (Royal Society, Boyle Papers, vol. ii, £ 141).

42 R. ]~oyle, Works, 1772, vol. ii, p. 45. Compare this with Descartes' remark that ' one may deduce some very true and certain conclusions from suppositions that are false or uncertain ' (Oeuvres, 1897-1957, vol. ii, p. 199).

4a CL Works, 1772, vol. ii, pp. 46 ft.

Ann. of Sci . - -Vol . 22, No. 2. f

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

90 Laurens Laudan o n

s tumble on to a t rue hypothesis , bu t we can never prove it to be so. He likens science to a deciphering opera t ion in which

' men conjecturally frame several keys [i.e., hypotheses] to enable us co understand a letter written in ciphers [i.e., nature]. For though one may by his sagacity have fbund the right key, it will be very difficulty for him, to prove [it is the right one].' 44

Apar t f rom the evidence based on similari ty of language and theme, there are o ther factors which suggest t h a t Boyle drew his hypothe t ica l me thod from Descartes. Ju s t as Descartes justified his hypothe t ica l me thod b y a t t r ibu t ing it to Aristotle 's Meteors, 45 so does Boyle tu rn to the Stagiri te to show tha t the me thod of hypothesis had been ant ic ipa ted by ' the mas te r of those t h a t know ':

' Aristotle himself (whatever confidence he sometimes seems to express) does in his first book of Meteors ingenuously confess, that concerning many of nature's phenomena, he thinks it sufficient, that they may be so performed as he explicates them.' 46

Boyle's method then consists in this: The scientist conducts wide- scale experimentation to determine the ' divers effects of nature '. He ne x t suggests a hypothesis to explain wha t has been observed. The first

*a lbid., vol. i, p. 82. Cf. also Royal Society, Boyle Papers, vol. ix, f. 63. I n principle 205 of t h e f o u r t h book of t he Principles, Descar tes h a d s imilar ly compared scientific theor iz ing wi th decoding t echn iques : ' I f , for ins tance , a n y o n e w a n t i n g to r ead a le t ter wr i t t en in L a t i n charac te r s no t in the i r proper order, decides to r ead B wherever he f inds A a n d C where he f inds B . . . . a n d if he in th i s w a y finds there are cer ta in L a t i n words composed of these , he will n o t doub t t h a t t he t rue m e a n i n g is con ta ined in t he words, a l t h o u g h he d iscovered th is b y conjecture , a n d a l t h o u g h it is possible t h a t t he wr i ter did n o t a r r ange the le t ters in th i s order of succession, b u t in some o t h e r . . . (Oeuvres, 1897-1957, vol. ix, p. 323). Cf. also Ru le 10 o f t h e Regulaz.

4~ ' A n d lest i t be supposed t h a t Aris tot le did, or w a n t e d to do, more t h a n this , it m u s t be recalled t h a t he express ly s ays in t he i~rst book of t he Meteors, a t t he begimaing of t he s e v e n t h chap te r , t h a t w i t h regard to t h i ngs no t ev iden t to t he senses, he t h inks t h a t he offers sufficient expl ica t ions and d e m o n s t r a t i o n s of t h e m , if he mere ly shows t h a t t h e y m a y be as he expla ins t h e m ' (ibid).

adl%. Boyle, Works', 1772, vol. ii, p. 45. I t is s ignif icant t h a t ano the r Eng l i sh e x p o n e n t of a t o m i s m , W a l t e r Char le ton, who s imilar ly a d o p t e d a hypo the t i ca l t h e o r y of science (our conjec tures tell u s how t he world 'may be, r a the r t h a n how it is or m u s t be ', Physiologiea, London , 1654, p . 128), also appeals to Ar i s to t le ' s r e m a r k in t he Meteors, to s u p p o r t h is hypo the t i ca l i sm (ibid.). I m m e d i a t e l y thereaf te r , he quotes Descar tes ' Principles on t he s ame point . Th i s raises the possibi l i ty t h a t :Boyle borrowed t he reference to Ar is to t le ' s Meteors f r om Char le ton r a the r t h a n direct ly f rom Descar tes . l~owever, :Boyle s t a t e s t h a t t he por t ion o f his MSS. in wh ich the reference to Aris tot le occurs was w r i t t e n in 1651 or 1652, prior to t h e appea rance of Char le ton ' s Physiologiea. (Cf. no te 94.) A l t h o u g h Boyle m a y h a v e m i s r e m e m b e r e d the da te of composi t ion, the re is no evidence t t ia t he h a s done so. Indeed , where one can i ndependen t l y cheek :Boyle's m e m o r y on such ma t t e r s , he is u sua l ly qu i te accura te . (For a brief, b u t suggest ive , accoun t of t he re la t ions be tween :Boyle a n d Char le ton, see R . K a r g o n , ' Wa l t e r Char le ton, R o b e r t :Boyle a n d the Accep tance o f E p i c u r e a n A t o m i s m in E n g l a n d ', Isis, 1964, 55, 184-192.)

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 91

hypotheses should be fair ly low-level generalizat ions abou t t h e ' immedia te causes of the phenomena '. Then, ' ascending in the scale of causes ', he arrives u l t imate ly at t he most general hypotheses , which concern the ' m o r e catholick and p r imary causes of things ,.47 A t each level, the scientist checks to see i f the hypothesis conforms to the corpuseularian doctrine. I f so, he tes ts the hypothesis against the entries in all his tables and against the other known laws o f nature . I f it is falsified he rejects it, i f not, he continues to main ta in it. This is obviously similar to Descartes ' view t h a t the in te rmedia te hypotheses mus t be compatible wi th the first principles and wi th the phenomena. An hypothesis is not p roven t rue, of course, even if it is compat ible wi th all our evidence; bu t i t can be asserted with more confidence as it proves itself capable of explaining more and more phenomena:

' For, the use of an hypothesis being to render an intelligible account of the causes of the effects, or phenomena proposed, without crossing the laws of nature, or other phenomena; the more numerous, and the more various the particles are, whereof some are explicable by the assigned hypothesis, and some are agreeable to it, or, at least, not dissonant from it, the more valuable is the hypothesis, and the more likely to be true. For it is much more difficult to find an hypothesis, that is not true, which will suit with many phenomena, especially, if they be of various kinds, than but with a few.' 4s

Here again, comparisons with Descartes are in order. Descar tes had argued t ha t those principles are most likely which explain ' s eve ra l different e f fec t s ' r a the r t han one, and suggested tha t a well-confirmed hypothesis , which explains a cross-section of nature , is p robably t rue. 49 Boyle 's poin t is substant ia l ly the same.

There is ye t ano ther basic methodological pos tula te which Boyle and Descartes bo th accepted. This might be called the principle of the ~nulti-level identity of nature. Basically, this principle postula tes t h a t the laws of na tu re which apply to visible massive bodies also apply to objects which are ei ther too large or too small to be measured or observed.S° I t was by invoking such a principle t h a t seven teen th-cen tury scientists were able to assume t h a t the laws of visible-body mechanics applied to in teract ions between sub-microscopic corpuscles. I t was also in t e rms of this principle t ha t t hey re jec ted the Scholastic s t ra tagem of a t t r ibut ing

4v R. Boyle, Works, 1772, vol. ii, p. 37. 4s Ibid., vol. iv, p. 234. 49 'Although there exist several individual effects to which it is easy to adjust diverse

causes [i.e., hypotheses] one to each, it is however no t so easy to adjus t one and the stone [hypothesis] to several different effects, unless it be the t rue one f rom which they proceed. (R. Descartes, Oeuvres, 1897-1957, vol. ii, p. 198).

50 This principle received its definitive formulat ion at Newton ' s hands, in his th i rd Rule

of Philosophizing.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

92 L a u r e n s L a u d a n on

p r o p e r t i e s t o m i c r o - e n t i t i e s w h i c h do n o t d e s c r i b e o b s e r v a b l e en t i t i e s . D e s c a r t e s f o r m u l a t e s t h e p r i n c i p l e t h u s :

' . . . we do much be t t e r to judge of wha t t akes place in small bodies, which thei r minuteness alone p reven ts us f rom perceiving, by wha t we see occurr ing in those t h a t we do perceive, than , in order to expla in cer ta in given things, to invent all sor ts of novelt ies, t h a t have no re la t ion to those t h a t we perceive. ' 51

B o y l e , a d o p t i n g h o m e l i e r l a n g u a g e , p u t s t h e p r i n c i p l e t h i s w a y :

' bo th the mechanical affections of m a t t e r are to be found, and the laws of mot ion t ake place, no t only in the grea t masses, and the middle sized lumps, bu t in the smal les t f ragments of m a t t e r . . . A n d therefore to say, t h a t though in n a t u r a l bodies, whose bu lk is manifes t and the i r s t ruc ture visible, the mechanica l pr inciples m a y be useful ly admi t t ed , [but] t h a t [they] a rc no t to be ex tended to such por t ions of ma t t e r , whose pa r t s and t ex tu re are invisible; m a y perhaps look to some, as i f a m a n should allow, t h a t the laws of mechanism m a y t ake place in a town clock, bu t no t in a pocke t watch . ' 52

B y w a y o f r e c a p i t u l a t i o n o f t h e a r g u m e n t t h u s far , I h a v e e n u m e r a t e d b e l o w t h e m o s t i m p o r t a n t p a s s a g e s i l l u s t r a t i n g m e t h o d o l o g i c a l s i m i l a r i t i e s b e t w e e n D e s c a r t e s a n d B o y l e :

Descartes 1. ' A n d lest i t be supposed

t h a t Ar is to t le did, or wan ted to do, more t h a n this, i t mus t be recal led t h a t he express ly says in the first book of the Meteors, at the beginning of the seventh chapter , t h a t wi th regard to th ings no t ev ident to the senses, he th inks t h a t he offers sufficient expl icat ions and demons t ra t ions of them, if he mere ly shows t h a t t h e y m a y be as he explains them. '

2. ' F o r jus t as an indus t r ious wa tch -make r m a y make two watches which keep t ime equa l ly well and wi thou t a n y difference in thei r ex te rna l appearance , ye t wi thout a n y s imi la r i ty in the composi t ion of the i r wheels, so i t is cer ta in t h a t God works in an inf ini ty of diverse w a y s . . . A n d I

Boyle 1. ' Ar is to t le himself (whatever

confidence he somet imes seems to express) does in his first book of Meteors ingenuously confess, t h a t concerning m a n y of na tu re ' s phenomena , he th inks i t sufficient, t h a t t hey m a y be so per formed as he expl icates t hem ' .

2. ' F o r as an ar t i f icer can set all the wheels of a clock a going, as well wi th springs as wi th weights . . . . so the same effects m a y be p roduced b y d ivers causes different f rom one another ; and i t will of tent imes be ve ry difficult, i f no t impossible for our dim reasons to discern surely,

~1 1~. Descartes, Oeuvres, 1897-1957, vol. ix, p. 319. 52 1%. :Boyle, Wor]cs, 1772, vol. iv, p. 72. For a more extensive discussion of the principle

of multi-level identity in the seventeenth century, see my ' The Nature and Sources of Locke's Views on Hypotheses ', J. Itist. Ideas, forthcoming.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 93

believe I shall have done enough if the causes that I have listed are such that the effects they may produce are similar to those we see in the world, without being informed whether there are other ways in which they are produced'.

3. ' A l t h o u g h t h e r e ex i s t several individual effects to which it is easy to adjust diverse causes [i.e. hypotheses], one to each, it is however not so easy to adjust one and the same [hypothesis] to several different effects unless it be the true one from which they proceed.'

4. ' One may deduce some very true and certain conclusions from suppositions that are false or uncertain.'

5. ' We do much better to judge of what takes place in small bodies, which their minuteness alone prevents us from perceiving, by what we see occurring in those that we do perceive, than, in order to explain certain given things, to invent all sorts of novelties, that have no relation to those that we perceive.'

which of those severM ways, whereby it is possible for nature to produce the same phenomena, she has really made use of to exhibit them.'

3. ' F o r it is much more diffi- cult, to find an hypothesis, that is not true, which will suit many phenomena, especially if they be of various kinds, than but with a feW. ~

4. ' I t is a very easy mistake to conclude, that because an effect may be produced by such determinate causes, it must be so, or actually is so.'

5. ' And therefore to say, that though in natural bodies, whose bulk is manifest and their struc- ture visible, the mechanical principles may be usefully ad- mitted, [but] that [they] are not to be extended to such portions of matter, whose parts and texture are invisible; may perhaps look to some, as if a man should allow, that the laws of mechanism may take place in a town clock, but not in a pocket watch.'

The essence of Boyle ' s hypothet ical me thod is falsification. 5a Like both Bacon and Hooke, Boyle insists on the impor tance of refut ing hypotheses ra ther t han confirming them. Like them he compares scientific inquiry to the reductio ad absurdum methods of the mathemat i - cians. Scientists, he writes, should t ry :

'. diligently and industriously to make experiments and collect observa- tions, without being over-forward to establish principles and axioms, believing it uneasy to erect such theories, as are capable to explicate all the phenomena of nature, before they have been able to take notice of the tenth part of those phenomena, that are to be explicated. Not

5~ I have drawn, in part, from the unpublished Boyle material brought to light by R. Westfall (o2o. cit., see footnote 19). ttowever, Westfall gives no consideration to the Cartesian origins of :Boyle's hypotheticalism, suggesting rather the similarities between Bacon and Boyle.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

94 Laurens L a u d a n on

that I at all disallow the use of reasoning upon experiment . . . for such an absolute suspension of the exercise of reasoning were exceedingly troublesome, if not impossible. And, as in that rule of arithmetic, which is commonly called regula fals i [i.e., false position], by proceeding upon a conjecturally-supposed number, as if it were that, which we inquire after, we are wont to come to the knowledge of the true number sought for; so in physiology [i.e., natural philosophy] it is sometimes conducive to the discovery of truth, to permit the understanding to make an hypothesis, in order to the explication of this or that particular difficulty, that by examining how far the phenomena are, or are not, capable of being solved by that hypothesis, the understanding may, even by its own errors, be instructed. For it has been truly observed by a great philosopher [Bacon], that truth does more easily emerge out of error than confusion.' 54

Science, as he conceives it, is a dialectic between reason and sensation; the mind proposes theories which are then subjected to the careful scrut iny of the senses. Boyle does not share Bacon ' s con tempt for intellectual activities or his dis trust of 'phi losophical systems ', t hough he is quick to point out tha t :

' such kind of superstructures [should be] looked upon only as temporary ones; which . . . are not entirely to be acquiesced in, as absolutely perfect, or uncapable of improving alterations.' 55

Bu t so long as t hey are preceded by exper imentat ion and never enuncia ted with finality, theoretical systems and hypotheses p lay a central role in science. Boyle is the first to admit t ha t m a n y mechanist ic hypotheses appear, on first glance, unseemly and bizarre. Bu t he insists t h a t a deeper analysis will reveal their inherent value:

' Some hypotheses may be compared to those shops of drugsters and apothecaries,--where among the first things that the eye discovers, are serpents and crocodiles, and other monstrous and harmful things; whereas the inside is a repository or magazine of wholsome and useful medicines.' 56

Boyle 's a t t i tude towards the hypothet ical method comes out mos t clearly in his polemic against Hobbes ' s scathing critique of the Roya l Society. I tobbes, of course, believed hypotheses to be indispensable to natura l phi losophy and he had little s y m p a t h y for those in the Roya l Society who were so obsessed with the new experimental phi losophy t h a t t hey denied any role to speculations or conjectures. Boyle replies to Hobbes ' s critique by arguing (in a vein not quite faithful to the t r enchan t

54 R. Boyle, Works, 1772, vo]. i, pp. 302-303. The latter part of this passage confirms a point we suggested earlier, viz., that Boyle conceived his hypothetical method as perfectly compatible with Bacon's Indulgence of the Understanding. In making the point, Boyle even adopts Bacon's language (' it is sometimes conducive . . . to permit the understanding to make an hypothesis . . .').

~5 Ibid., p. 303. ~6 t~oyal Society, Boyle Papers, vol. ix, f. 113.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 95

empiricism of many members of the l~oyal Society) that the now organization is not opposed to the use of all hypotheses, but only to those that are not well grounded in the phenomena. Hobbes had proposed two requisites of a good hypothesis: that it be conceivable and that, if it is granted, ' the necessity of the phenomena may be inferred '.57 To these, Boyle adds a third and equally important condition: 'namely , tha t it [the hypothesis] not be inconsistent with any other t ruth or phenomenon of nature ,.ss Elsewhere, he formulates the same condition another way: it is sufficient to accept an hypothesis if we can demonstrate ' its fitness to solve the phenomena for which [it was] devised, without crossing any known observation or law of nature ,.59 Richard Westfall 6° and Marie Boas Hall 61 have pointed out that in Boyle's unpublished manuscripts, there is an even more explicit discussion of the characteristics of an acceptable hypothesis. Boyle writes that the requisites of a good hypothesis are (1) that it suppose nothing that is either impossible or absurd, (2) that it be self-consistent, (3) tha t it be sufficient to explicate the phenomena, ' a t least the chief among them ', and (4) that it be consistent with other known phenomena and ' manifest physical t ru th '. An excellent hypothesis, in addition to satisfying these four requirements, also (5) is the simplest hypothesis, (6) is the only hypothesis that explicates the phenomena or at least explicates them better than any other and (7) enables us to ' foretell future phenomena ,.62

What I have said here should not be interpreted to mean tha t Boyle's methodological ideas are identicM with Descartes' or that he uncritically mimicked his predecessor. There are important differences between Descartes and Boyle which should not be glossed over. Boyle's continual insistence on extensive experimentation finds no counterpart in Descartes and he had no patience with Descartes' exclusion of final causes from cosmology. Furthermore, Boyle often ehastizes Descartes, not for constructing hypothetical systems, but for constructing them with

s7 Hobbes ' s probabitistic account of scientific knowledge is very much llke Descartes, even in language. For instance, t tobbes asserts: ' . . . he tha t supposing some one or more motions, can derive f rom them the necessity of t ha t effect whose cause is required, has done all tha t is to be expected from natura l reason. And though he prove not t h a t the thing was thus produced, yet he proves tha t thus it m a y be p r o d u c e d . . , which is as useful as if the causes themselves were k n o w n ' (English Works, ed. W. Moleswor~h, London, 1845, vol. vii, pp. 3-4). Compare this with the latter half of the passage in which Descartes formulates the clock analogy.

5s R. :Boyle, Works, 1772, vol. i, p. 241. For an account of Hobbes ' s views, see E. Madden, ' Thomas Hobbes and the Rationalistic Ideal ', in Theories of Scientific Method cal. Madden, Seattle, 1960, pp. 104-118.

5s R. Boyle, Works, 1772, vol. iv, p. 77. 50 R. Westfall, op. sit. (footnote 53), pp. 113-114. 61 M. B. Hall, Robert Boyle on Natural Philosophy, Bloomington, Indiana, 1965, pp. 134-

135. n~ Royal Society Boyle Pa~ers, vol. ix, L 25,

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

96 Laurens Laudan o n

insufficient care. He also notes, rightly, that Descartes paid only lip- service to experimentation and spun his corpuscular philosophy largely out of his head. But just as the differences between the two should not be neglected, neither should their similarities, especially when they are so pronounced as they are in regard to the hypothetical method. Historians are certainly correct when they suggest tha t Boyle's experimentalism derives from Bacon, or at least from a tradition with which Bacon associated himself ; but I think they are mistaken when they suggest that Baeonian experimentalism was the whole of Boyle's methodological position and that Descartes' hypothetical and mechanical philosophy had negligible impact in comparision with Bacon's. 63 In the methodological writings of Boyle, Descartes' influence is every bit as pronounced as Bacon's.

Boyle was not only aware, as most of his compatriots were, that the Cartesian system was composed of hypotheses. He also appreciated, as too few of his countrymen did, Descartes' r e a s o n s for using hypotheses. In short, he understood Descartes' claim that hypotheses were essential to science, that they were the sign of a fertile mind aware of nature's inscrut- ability. Although Boyle often disagreed with the specific hypotheses which Descartes formulated, he never made the l~ewtonian mistake of thinking that Cartesian hypotheticalism must be rejected because Cartesian hypotheses were rejected.

Cartesian probabilism was the most reasonable methodology for a corpuscularian philosopher like Boyle to adopt, because the corpuscular programme entailed the reduction of visible phenomena to the behaviour of unobservable entities whose properties were obviously conjectural. Once one accepts the corpuscularian theory of matter, and with it the theory of knowledge which makes corpuscles unobservable in principle, then it is altogether natural to adopt a hypothetical methodology In short, the metaphysics and epistemology of the mechanical philosophy- led, by its own inner logic, to the acceptance of a certain methodology. 6a Boyle accepted the Cartesian theory of method, not because he was a Cartesian, but because he was a corpuscularian!

6a I n l ight of t he ana lys i s which I have given here, it is surpr i s ing to f ind M. Crans ton wr i t ing t h a t ' Boyle ' s i nduc t ive m e t h o d was essent ia l ly Bacon i an ' (John Locke, London , 1957, p. 75). Crans ton is wrong on two accoun ts ; Boyle ' s m e t h o d was ne i ther induc t ive nor Bacon ian .

64 I n a recent m o n o g r a p h , i~. Harr@ has m a d e a s imilar po in t a b o u t the eonnex ions be tween s e v e n t e e n t h - c e n t u r y m e t a p h y s i c s a n d methodo logy , focuss ing especially on t he l inks be tween the corpuscu la r ph i losophy and the doct r ine of p r i m a r y - s e c o n d a r y qual i t ies (Matter and Method, London , 1964). However , as I have po in ted ou t elsewhere (' Me thod and t he Mechanical P h i l o sophy ', History of Science, fo r thcoming) , Har r6 ' s ana lys i s is h a n d i c a p p e d b y his a s s u m p t i o n t h a t Descar tes bad n o t h i n g to do wi th e i ther t he m e t a p h y s i c s or t he me thodo l ogy of corpuscular ism,

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 97

In a sense, Boyle borrowed the best from both worlds. Visibly impressed with the clock analogy, he sensibly appropriated Descartes' hypotheticalism, but neatly dropped the doctrine of clear and distinct ideas which regularly plagued the development of Descartes' views on method. From Bacon, he inherited a vigorous experimentalism, but was careful not to import with it the inductive and rigidly 'empiricist' philosophy which cramped Bacon's acceptance of the hypothetical method and made his endorsement of it invariably half-hearted. 65

GLA:NVILL A:ND CARTESIAN HYPOTHETICALISM

A second English writer much impressed by Descartes' clock-metaphor was Joseph Glanvill (1638-1680) who, in his Vanity of Dogmatizing (1661) and Scepsis Scientifica (1665), has some illuminating, if depressing, things to say about scientific method. 66 Though both volumes are declarations of war against classical and medieval philosophy, Glanvill has kind words for at least two of his predecessors, Bacon and Descartes. Like Boyle, he drew from the two authors as if they were merely different sides of the same coin. He frequently alludes to ' those great men, the Lord Bacon and Descartes ,67 and his debt to each of them will become clear as we describe his methodology.

65 I t is thus highly unjustified to juxtapose Boyle and Descartes as if they personified, respectively, empiricism and rationalism. A part icularly misleading instance of such juxtaposi t ion is to be found in a recent essay by A. R. and M. B. Hall. They write: ' I s a scientific proposit ion demonst ra ted when it is shown to be a logical consequence of a set of intuit ively certain axioms? Descartes would have answered affirmatively, :Boyle negatively. I s a scientific propos i t ion sufficiently demons t ra ted only b y showing empirically tha t it holds? Descartes would have answered negatively, Boyle affirmatively ' (' Phi losophy and :Natural Philosophy: Boyle and Spinoza ', in Mdlangcs Alexandrc Koyrd, ed. I . B. Cohen and R. Taton, Paris, 1964, vol. ii, pp. 242-243). F rom the evidence presented here, one could more reasonably conclude tha t Boyle and Descartes would bo th answer each of the two questions affirmatively.

Boyle is so far f rom being the strict empiricist the Halls suggest he is t ha t he occasionally adopts a rat ional ism more rigorous t han Descartes ' . Fo r example, he observes tha t ' it is no t always necessary, though it be always desirable, t ha t he, t ha t p ropounds an hypothesis in as t ronomy, chemistry, ana tomy, or other pa r t of physics, be able a priori, to prove his hypothesis to be t r u e . . . ' ( Wor]cs, 1772, vol. iv, p. 77). Cf. footnote 27 above. Elsewhere, and again in a mos t ' unempirical ' vein, Boyle states t h a t ' where reason proceeds in a due manner . . . its conclusions are to be preferred to some testimonies of s e n s e ' (Royal Society, Boyle Papers, vol. ix, L 33).

Bu t it would equally be mis taken to th ink tha t Boyle was al together clear in his own mind about the precise relations between reason and experiment; for there certainly a r e

t imes when Boyle opts for the posit ion tradit ionally called empiricism. At one point, in a start l ing anticipation of Newton ' s F o u r t h Rule of Philosophizing, Boyle says ' tha t the well c i rcmnstanc 'd tes t imony of sense is to be preferred to any hypothesis . . . ' (ibid., vol. ix, f. 31).

66 For a general account of Glanvill 's scientific work, see M. Prior 's ' Joseph Glanvill, Witchcraf t and Seventeenth-Century Science ', Modern Philology, 1932, 80, pp. 167-193.

6v j . Glanvill, Sccpsis Scientifica (ed. Owen), London, 1885, p. 44.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

98 Laurens L a u d a n on

Nature , wri tes Glanvill , is v e r y subt le and i ts mechan i sms are not obvious to even the mos t as tu te observer . Only va in and p o m p o u s men believe t h a t the i r hypo the t i ca l scientific sys tems provide a fai thful image of the physica l world. Na t u r a l ph i losophy is seriously mi s t aken when it d ic ta tes how the physical world should operate . To deny a void or the ea r th ' s ro t a t ion because we conceive such things to be impossible is to m a k e the error of th ink ing t h a t na tu re and its Creator are confined within the bounds of m a n ' s d im reason; which is no t on ly absurd, b u t b lasphemous , to the pious Glanvill. At this point , Glanvill invokes Descar tes ' analogy:

' For Nature is set a going by the most subtil and hidden instruments; which it may have nothing obvious which resembles them. Hence judging by visible appearances, we are discouraged by supposed impossibilities which to nature are none, but within her spear [sic] of action. And therefore what shews only the outside, and sensible structure of nature; is not likely to help us in finding out the Magnalia [i.e., inner mechanisms]. 'Twere next to impossible for one, who never saw the inward wheels and motions, to make a watch upon the bare view of the circle of hours, and index: And 'tis as difficult to trace natural operations to any practical advantage, by the sight of the cortex of sensible appearances. ' 6s

Sensat ion alone can never give us insight into the t rue mechan i sms of na tu re ; the mos t we can realist ically hope for is p robab le and hypo the t i ca l knowledge. Thus, the ' w a y of inquiry for t rue phi losophers ' is

' t o seek t ru th in the great book of nature, and in that search to proceed with wariness and circumspection without too much forwardness in establishing maxims and positive doctrines: To propose their opinions as hypotheses, that may probably be true accounts, without peremptorily affirming that they are.' 60

Glanvfll leaves no doub t t h a t he bor rowed his hypo the t i ca l i sm f rom Descartes . Towards the end of the Scepsis Scientifica, he writes:

' And though the Grand Secretary of Nature, the miraculous Descartes hath here infinitely out-done all the philosophers tha t went before him . . . yet he intends his principles but for hypotheses, and never pretends that things are really or necessarily, as he hath supposed them: but he only claims that they may be admitted pertinently to solve the phenomena, and are convenient supposals for the use of life. Nor can any further account be expected from humanity, but how things possibly may have been made consonantly to sensible nature: but infallibly to determine how they truly were effeeted, is proper to him only that saw them in the chaos, and fashioned them out of that confused mass. For to say, the principles of nature needs be such as our philosophy makes them, is to set bounds to omnipotence, and to confine infinite power and wisdom to our shallow models.' vo

68 1bid., p. 155. This same passage occurs in his earl ier Vanity of Dogmatizing, London,

1661, p. 180. a0 j . Glanvfll , Scepsis Scientifica, London, 1885, p. 44. 7o Ibid., pp. 182-183.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

T h e C l o c k M e t a p h o r a n d P r o b a b i l l s m 99

Th i s s t a t e m e n t a lone is suff icient t o u n d e r m i n e the c la im t h a t Desca r t e s ' m e t h o d o l o g y h a d no i m p a c t on t he E n g l i s h wr i t e r s of th i s ge ne r a t i on . I t m a k e s clear t h a t Desca r t e s ' h y p o t h e t i c a l i s m , a n d the c lock -maker a n a l o g y wh ich s u p p o r t s it, were o b v i o u s l y a t t r a c t i v e to a s ign i f i can t class of E n g l i s h m e n who, for rel igious or o the r reasons , saw fit to p u t s t r i c t l i m i t a t i o n s on t he scienti,~t's c la im to t r u t h . 71 Glanvi l l , however , carries

his h y p o t h e t i c a l i s m f u r t h e r t h a n Desca r t e s ; he denies t h a t a n y scient i f ic pr inc ip les , e v e n those of m a t t e r - i n - m o t i o n , can be more t h a n hypo theses . I n this , he is closer to Boyle t h a n Descar tes , t h o u g h one is i nc l ined to d o u b t w h e t h e r even Boy le wou ld have gone so far d o w n the seep t ic ' s p r i m r o s e p a t h as G lanv i l l does w h e n he asser ts :

' For the best principles, excepting divine, and mathematical , are b u t hypotheses; wi thin which, we may conclude m a n y things with securi ty from error. Bu t yet the greatest certainty, advanced from supposal, is still bu t hypothetical. So tha t we m a y affirm, tha t things are thus and thus, according to the principles we have espoused: B u t we s t rangely forget ourselves, when we plead a necessity of their being so in nature , and an impossibil i ty of their being otherwise.' 72

G l a n v i l l t h e n is a n a v o w e d hypo the t i cMis t , more c o n s i s t e n t l y so t h a n e i t he r Desca r t e s or Boyle . B u t whi le ins i s t ing t h a t all scientif ic p r inc ip le s are b u t con jec tures , he be l i eved t h a t some hypo these s are b e t t e r t h a n o thers . T a k i n g Boy le ' s l ine, G lanv i l l a rgued t h a t s o u n d h y p o t h e s e s m u s t be b a s e d on e x p e r i m e n t a l h is tor ies of n a t u r e . 7a To t he R o y a l Society, he wro te :

' . . . f rom your promising and generous endeavors, we may hopefully expect a considerable enlargement of the history of nature , wi thout which our hypotheses are bu t dreams and romances, and our science meer conjecture and opinion. ' 7a

S h a r i n g t he genera l s e v e n t e e n t h - c e n t u r y c o n t e m p t for a n t i q u i t y , G lanv i l l sugges ted t h a t no t r u l y s o u n d t h e o r y had ever b e e n p roposed because ph i losophers were no t suff ic ient ly e x p e r i m e n t a l . 75 H e was

~1 Even in the early pages of the Scepsis Scientifica, Glanvill is pre-occupied with the clock analogy which he makes explicit in the passage cited above. Throughout his work, the model of nature as an observable but unknowable machine floats just beneath the surface: ' We cannot profound into the hidden things of nature, nor see the first springs and wheels that set the rest a going. We view but small pieces of the universal frame, and want phenomena to make intire and secure hypotheses ' (ibid., p. 75).

v2 Ibid. , pp. 170 I71. ~3 Glanvill indicates his sympathy for the English polemic against systems when he

writes : ' if such great and instructed spirits [i.e., the members of the t~oyal Society] think we have not as yet phenomena enough to make as much as an hypothesis; much less to fix certain [i.e., indubitable] laws and prescribe methods go nature in her actings: what insolence is it then in the lesser size of mortals, who possibly know nothing but what they have gleaned from some little systeme . . . to boast infallibility of knowledge . . .' (ibid., p. li).

~4 Ibid. , p. lxii. ~ 5 , . . . 'tis possible that all the hypotheses that have yet been contrived, were buil~

upon too narrow an inspection of things . . .'. (ibid., p. lxiii).

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

100 Laurens L a u d a n on

especially scornful of Aristotle 's excursions into na tura l science: ' t h e Aristotel ian hypotheses give us a ve ry d ry and jejune account of nature ' s phenomena ,.76 On the whole, Glanvill is satisfied with almost any hypothesis which saves the appearances ; a t least he criticizes only those hypotheses which are incompatible with experience. Bu t if two hypotheses bo th accord wi th na ture equally well, he suggests t h a t we should believe the simplest of the two:

' whichever doth with more ease and congruity solve the phenomena, that shall have my vote for the most philosophick hypothesis.' 77

Glanvill was no t only pessimistic abou t the possibili ty of finding t rue hypotheses about unobservable mechanisms, he even doubted t h a t we could confident ly discuss causal relat ions between observable objects. I n a passage t ha t perhaps ranks him as the seventeen th-cen tury Hume, he argues:

' So that we cannot conclude, anything to be the cause of another; but from its continual accompanying it: for the causality itself is insensible. But now to argue from a concomitance to a causality is not infallibly c o n c l u s i v e . ' 78

He goes on to insist t ha t because we cannot ' infallibly assure ourselves ' of the t r u t h of even the most obvious causal relation, ' the founda t ion of scientifical procedure is too w e a k ' to pe rmi t us to build an indubi table science upon it . ' 79

H E N R Y POWER

The final figure I want to consider is H e n r y Power (1623-1688), whose only claim to importance, and i t ought to be a modest claim, is based on microscopical observat ions in his Experimental Philosophy in three boo]cs 1 Containing new experiments--microscopical, mercurial, magnetical--with some deductions and probable hypotheses, raised from them in avouchment and illustration of the now famous Atomical Hypothesis (1664). s° Not unexpectedly , the pompos i ty of the t i t le is even surpassed b y the tex t . B u t though Power ranks as nei ther a great scientist nor a good philosopher, his volume is impor t an t in indicat ing jus t how wide-spread Descartes ' clock analogy became.

Like Boyle and Glanvill, Power pays homage equal ly to the ' ever-to- be-admired Descartes TM and ' t h e learned Verulam 's2. Conflating

7s I b i d . , p. 145. ~ I b i d , , p. 51. vs I b i d . , p. 166. T9 I b i d . , pp. 167-168. so Power was apparent ly the first English scientist to publish extensively on the micro-

scope. Immedia te ly after his book, numerous t rac ts were devoted to microscopical observations. Among these, Hooke 's lV I i c rograph ia (1665) was the mos t widely read.

81 H. Power, E x p e r i m e n t a l P h i l o s o p h y . . . , London, 1664, preface. s~ I b i d .

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 101

Descartes and Bacon much as Boyle and Glanvill did, Power is perfectly willing to allow hypotheses in science, so long as they 'save all the appearances ,sa and are ' confirmed and made good ,sa by several experi- ments, rather than a single one. But without experiments, he is convinced that ' ou r best philosophers will prove but empty conjecturalists, and their profoundest speculations herein, but glossed outside fallacies.' s5 Though not altogether opposed to hypotheses, Power is a good deal more optimistic than Descartes, Boyle and Glanvill were about the possibility of achieving indubitable knowledge about the physical world. Until Power's time, the hypothetical method had invariably accompanied corpuscularism~ for it seemed one could never do more than conjecture about the properties of the smallest bits of matter. But Power gives corpuscularism a strange turn. Impressed by the microscope's ability to penetrate into the inner structure of organisms, Power suggests that by the aid of the microscope we will eventually be able to see corpuscles and determine their precise mechanisms. 's6 He envisions the erection of a ' true and permanent philosophy ' based on the microscopical canvassing of nature and the ' infallible demonstrations of meehanieks.' s7 Power believes tha t the microscope thus provides the key for breaking into Descartes' universal clock and out of his hypothetiealism; for such an instrument offers us a glimpse at the internal mechanisms themselves. Power proceeds to turn the clock analogy against Descartes, suggesting tha t though the ' system-builders ' can never observe more than the outer appearances, the faithful experimenter, with the help of optical instruments, will eventually learn the truth:

' For the old dogmatists and notional speculators, that onely gazed at the visible effects and last resultances of things, understood no more of nature, than a rude countreyfellow does of the internal fabrick of a watch, that only sees the index and horary circle, and perchance hears the clock and alarum strike in it; But he that will give a satisfactory account of the phenomena, must be an artificer indeed, and one well- skilled in the wheel work and internal contrivance of such anatomical engines.' ss

sa Ib id . , p. 94. s4 Ib id . , p. 114. s5 Ib id . , preface.

8~ CL Ib id . , p. 82. 8~ ib id . , p. 192. 8s Ib id . , p. 193. I n his E s s a y , J o h n Locke likens our knowledge of sub-microscopic

mechanisms to a ' coun t ryman ' s idea ' of the ' inward contrivance of t h a t famous clock at Strasbourg, whereof he only sees the ou tward figures aud mot ions ' (Essay , London, 1929, iii, 6, §9). Could this be a gloss on Power ' s formulat ion of the clock-maker analogy? For a discussion of Locke's general methodological position, see R. Yost, ' Locke's Rejection of Hypotheses Abou t Sub-Microscopic Even t s ', J . H i s t . Ideas , 1951, 12, 111-130, and m y ' The Nature and Sources of Locke's Views on Hypotheses ', loc. cir. (footnote 52).

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

102 Laurens Laudan o n

After Power, writers like Hooke, s9 Newton 9° and Cowley 91 enlarged on this theme, suggesting quite confidently that improved instrumentation promised hope of discovering the true mechanisms of nature and conclusively establishing (or refuting) the corpuscular philosophy2 ~ As

s~ Hooke wri tes: ' . . . and b y the help of microscopes, the re is n o t h i n g so smaU as to escape our i nqu i ry . . . I t seems no t improbable b u t t h a t b y these helps [viz., optical i n s t r u m e n t s ] t he sub t i l t y of the compos i t ion of bodies, the s t ruc tu re of the i r pa r t s , t h e var ious t ex tu r e of thei r m a t t e r , t h e i n s t r u m e n t s and m a n n e r of the i r i nward mot ions , a n d all t he possible appea rances of th ings , m a y come to be more fully d i s c o v e r e d . . . ' (MicrograpMa, London , 1665, preface). W h a t we canno t discover i~bout corpuscles b y seeing t h e m , Hooke sugges t s we can learn b y l is tening to t h e m : ' There m a y also be a possibi l i ty of d iscover ing t h e in te rna l mo t i ons a n d ac t ions of bodies by t he sound t h e y make , who knows b u t t h a t as in a w a t c h we m a y heat ' t h e hea t ing of t he balance, a n d t he r u n n i n g of t he wheels, a n d the s t r ik ing of the h a m m e r s , and t h e g ra t ing of t he tee th , a n d t he m u l t i t u d e s of o ther noises; who knows , I say , b u t t h a t i t m a y be possible to discover the mo t ions of the in te rna l pa r t s of bodies, w h e t h e r an imal , vegetable , or n~ineral, b y the sound t h e y m a k e . . . ' (Posthumous Works (ed. Waller) , London , 1705, p. 39). Consider ing t h a t b o t h Hooke a n d Power looked to i n s t r u m e n t a t i o n to free science f rom its hypo the t i ca l i sm, i t is pe rhaps s ignif icant t h a t in t he preface to his 2/licrographia, Hooke m e n t i o n s Power ' s Experimental Philosophy. He goes on to r e m a r k t h a t he a n d Power e x a m i n e d one a n o t h e r ' s MSS. before t h e y wen t to press.

90 I n t he Opticks, N e w t o n exudes a s imilar o p t i m i s m w h e n he wri tes t h a t ' I t is no t impossible b u t t h a t microscopes m a y a t l eng th be i m p r o v e d to t he d iscovery of the par t ic les of bodies on wh ich the i r colours depend, if t h e y are n o t a l ready in some m e a s u r e a r r ived to t h a t degree of perfec t ion ' (Optieks, New York, 1952, f o u r t h edit ion, p. 261). Optics , on N e w t o n ' s view, is l ikely to become a comple te ly induct ive , observa t iona l a n d non -hypo the t i c a l science. Cf. ~ . K a r g o n , ' N e w t o n , ]~arrow a n d the Hypo the t i c a I Phys i c s ', ' Centaurus, 1965,11, 46-56, for an accoun t of N e w t o n ' s reac t ion to Car tes ian hypo the t i ca l i sm. I t is in te res t ing to compare N e w t o n ' s o p t i m i s m a b o u t seeing corpuscles wi th Char le ton ' s r e m a r k t h a t even t he a t o m s ' of the larges t size, or ra te , are m u c h below the percept ion a n d d i s ce rm nen t of t he aeu te s t optics, a n d r ema in c o m m e n s u r a b l e only b y t he finer digets of ra t iona l conjec ture ' (op. tit., p. 113).

91 I n his Ode to the Royal Society, Cowley gave poetic express ion to Power ' s opt imis t i6 hopes for the microscope g iv ing u s infallible knowledge of the inner pa r t s of n a t u r e ' s clock:

' N a t u r e ' s grea t works no d is tance can obscure, No smal lness he r nea r objec ts c an secure.

Y o u ' v e t a u g h t the cmdous s igh t to press I n t o t he p r iva t e s t recess

Of he r impercept ib le l i t t l e n e s s . . . Y o u ' v e learned to r ead her smal les t hand ,

A n d well b e g u n he r deepes t sense to u n d e r s t a n d . '

( In Sp ra t ' s History of the Royal Society, London , 1667, i m m e d i a t e l y following t he ' Epis t le Ded ica to ry '.)

9~ I f we s t ra in t he po in t a little, we can see inf luences of t he clock ana logy s t r e t ch ing even in to t he ear ly e igh t een th cen tu ry . I n his preface to t he second edi t ion of N e w t o n ' s Principia, Roger C o t e s - - i n a ve in ve ry m u c h like H e n r y P o w e r - - t u r n s t he ana logy aga ins t t he Cartesians. Af ter a v igorous a t t a c k on t he hypo the t i c a l i sm of t he Car tes ians , Cotes ins is ts t h a t ' T he bus iness o f t rue ph i losophy is to der ive t he n a t u r e s o f t h ings f rom causes t r u ly ex is ten t , a n d to inqui re af ter t hose laws on wh ich t he Grea t Creator ac tua l ly chose to

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

The Clock Metaphor and Probabilism 103

the faith in the unlimited magnifying powers of the microscope grew, English hypotheticalism waned2 3 After all, the clock anglogy is only persuasive when there is serious doubt about the possibility of perceiving nature's mechanisms. As such doubt faded, so did Descartes' influence and the method of hypothesis. Descartes thus fell victim to his own metaphor; victimized because the clock analogy, which he endorsed so enthusiastically, suggested to many later writers that the inner mechanisms of nature, like those of a clock, could be directly scrutinized by careful observation and instrumentation. The demand for a science free of all hypotheses, which was widely circulated after Newton, could never have gathered such enthusiastic adherents if the hypotheticalism of Descartes, Boyle and Glanvill had not died such a quick and needless death at the hands of those who thought nature's clock had no secrets which man's instruments could not seek out and know with certainty2 4

ss Al though m a n y scientists evidently believed the microscope pointed the way to complete observabili ty of micro-mechanisms, there were others who, seeing the real substance of Descartes ' a rgmnent , took precisely the opposite position. Jacques Rohaul t , for instance, argued in his Trait~ de Physique, Paris, 1671, t ha t the moral to be d rawn f rom the new worlds which the microscope unfolds is no t t ha t we are finally getting down to ul t imate reality, bu t ra ther , tha t na ture is infinitely c o m p l e x ~ h a t each new stage of magnificat ion will reveal still finer and more subtle processes. Though the microscope permits us to examine the fleas on a dog, it does not permit us to see the micro-organisms which infest the flea. For Rehaul t , there is a potential ly infinite regress to ever-smaller bi ts of ma t t e r which the microscope, however great its magnification, is powerless to exhaust .

9, Considering tha t :Boyle, Glanvill and Power all used the clock analogy in publicat ions between 1661 and 1664, one is inclined to ask whether they borrowed the analogy f rom Descartes or f rom one another. I have been able to find little evidence which would conclusively resolve this an t iquar ian puzzle. I n point of publication, Glanvilt 's Vanity of Dogmatizing (1661) was issued first and this might seem to consti tute a prima faeie case for Boyle and Power having plagiarized it f rom Glanvill. However, there is substant ia l evidence against such a view. Glanvill 's editor, Owen, notes tha t his Vanity of Dogmatizing was ' so scarce as to be hardly known at all exeep~ by name ' (Seepsis Seientifica, p. xvi). Fur thermore , Boyle's Usefulness of Natural Philosophy, in which he used the clock analogy, t hough published in 1663, was sent to press in ' 1660, or 1661~ and 1663 ' (Boyle, Works, vol. ii, p. 3) and Boyle claims tha t most of the first par t of tha t work in which the crucial passage occurs - -was wri t ten ten or twelve years earlier (1651-53). I n addition to the similarity of language between Boylc 's and Descartes ' formulat ions of the analogy (which would undermine the Deseartes-to-Glanvfll-to-:Boyle view), we might note tha t :Boyle never

found this mos t beautiful frame of the world, n o t those by which he might have done the same, had he so p l e a s e d . . . The same mot ion of the hour-hand in a clock may be occasioned either by a weight hung or a spr ing shu t up within. Bu t if a certain clock should be really moved wi th a weight, we should laugh at a m a n tha t would suppose it moved by a s p r i n g . . . for certainty the way he ought to have taken would have been actually to look into the inward par ts of the machine, t ha t he might find the t rue principle of the proposed motion ' (Principia, 4th ed., Berkeley, 1934, pp. xxvii-xxviii) . Descartes ' poin t t h a t the inner machinat ions of na ture the eonnexions between phe nomena - - a r c necessarily precluded f rom view is apparen t ly los~ on the optimistic Cotes.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013

104 The Cloclc Metaphor and Probabillsm

mentions Glanvill in his published works. As for Power, he never mentions Glanvill though he continually writes of Boyle and Descartes. Since Boyle's Usefulness predates Power's Experimental History (1664), it is possible that Power got the analogy from Boyle rather than Descartes. On the other hand, Power's book received the imprimatur in 1663 (when Boyle's Usefulness was barely off the press) and its preface is dated 1661. In light of such evidence (scanty though it is), I think it probable that all three writers got the analogy directly from Descartes rather than from one another.

Dow

nloa

ded

by [

Har

vard

Col

lege

] at

15:

41 2

3 A

pril

2013