IGNORED INTELLECT .Duliem - University of Arizonaaversa/scholastic/Duhem/Ignored... · IGNORED...

7
IGNORED INTELLECT .Duliem Historical and personal circumstances hlocked the career of this right-tying, royalist, religious extremist despite J a variety of important contributions. His work is now J"~~ gaming greater rccog)uUon. by Donald G. Miller THIS YEAR MARKS the 50th anniversary of the death of Pierre Duhem, one of France's greatest intellects. His work encompasses thermodynamics, physical chemistry, hydrodynamics, elasticity, electricity and magnetism as well as the history and philosophy of science. In his 32 years of scientific activity, he published without collaborators more than 350 papers and 21 books for which a nearly complete bibliog- raphy exists. 1 (Some of the work is in many volumes, some in several edi- tions, and some in translation.) His work in the philosophy of science -• 3 has received continuing attention. His work in the history of science in the middle ages 4 « 5 is well known to his- torians of science. Until recently, however, his purely scientific work was almost completely ignored or forgotten by working scien- tists, with the exception of the Gibbs- Duhem and Duhem-Margules equa- tions, well known to physical chem- ists. In view of his enormous scientific output, particularly valuable in hydro- dynamics, elasticity and thermodynam- ics, it is curious that this work is so little known. Why, for example, did Duhem spend his life in academic exile, denied pro- motions and salary increases, and ostracized by most French scientists? Why in fact did he never get a position in Paris, with its better libraries, bet- ter students and all the accompanying prestige, which would have been ex- pected as an appropriate award for his accomplishments? The reasons-sci- entific, political, and religious—make an interesting story, not without paral- lels today. PHILOSOPHICAL OUTLOOK Duhem belonged to the Energeticist school along with Ernst Mach, Georg Helm, and Wilhelm Ostwald. Duhem's views on the nature of physical theories were similar to theirs and to Henri Poincare's. In common with them, he believed that physical theories serve as descriptions that condense and classify the results of experiments, rather than as explanations or inter- pretations of experiment. However, he differed from the others in believing that as physical theories evolve by successive adjustments to conform to experiment, they approach asymptot- ically a sort of "natural classification" that somehow reflects an underlying reality. 2 ' 3 This view removes him from the ranks of the positivists. Another point emphasized by Du- hem was the impossibility of a truly "crucial experiment." An experimental test of one proposition necessarily in- volves all the other propositions of the theory. Therefore the contradictions of a theory by experiment can be re- moved not only by changes in the one proposition believed to be "crucially" tested, but by changes in the other propositions leaving the "crucial" one unchanged.- 1 3 Duhem arrived at the description position more or less in the following way. In the period 1850-1910, there were two opposing viewpoints: one that tiieories should be based on ex- periment and description, the other that all of physics could be reduced to classical mechanics and explained by atomic and molecular models. Du- hem's scientific formation took place in the period 1880-1890, well before the discovery of radioactivity and the experiments of Jean Perrin or Henry G. J. Moseley. At this time, the group using atomic and molecular models for their "mechanical explanation of the universe" changed the properties of their "billiard balls" at will depend- ing on the phenomenon. These prop- erties, often contradictory, were as hard to believe in then as they are now. Duhem, with his passion for logic, could not accept such contra- dictory models. He became convinced that on the contrary, instead of hav- ing all of physics reduced to classical mechanics, classical mechanics itself was a special case of a more general continuum theory. Therefore along with the other Energeticists, Duhem believed a gen- Donald G. Miller is a physical chemist with a PhD from Illinois. He has been at the Lawrence Ra- diation Laboratory, Livermore, since 1956 working in thermo- dynamics of irrever- sible processes, elec- trolyte solutions and die history of science—primarily thermodynamics. PHYSICS TODAY DECEMBER 1966 47

Transcript of IGNORED INTELLECT .Duliem - University of Arizonaaversa/scholastic/Duhem/Ignored... · IGNORED...

IGNORED INTELLECT

.DuliemHistorical and personal circumstances hlocked the career

of this right-tying, royalist, religious extremist despite Ja variety of important contributions. His work is now J"~~

gaming greater rccog)uUon.

by Donald G. Miller

THIS YEAR MARKS the 50th anniversaryof the death of Pierre Duhem, one ofFrance's greatest intellects. His workencompasses thermodynamics, physicalchemistry, hydrodynamics, elasticity,electricity and magnetism as well asthe history and philosophy of science.In his 32 years of scientific activity,he published without collaboratorsmore than 350 papers and 21 booksfor which a nearly complete bibliog-raphy exists. 1 (Some of the work is inmany volumes, some in several edi-tions, and some in translation.) Hiswork in the philosophy of science -• 3

has received continuing attention. Hiswork in the history of science in themiddle ages 4« 5 is well known to his-torians of science.

Until recently, however, his purelyscientific work was almost completelyignored or forgotten by working scien-tists, with the exception of the Gibbs-Duhem and Duhem-Margules equa-tions, well known to physical chem-ists. In view of his enormous scientificoutput, particularly valuable in hydro-dynamics, elasticity and thermodynam-ics, it is curious that this work is solittle known.

Why, for example, did Duhem spendhis life in academic exile, denied pro-motions and salary increases, andostracized by most French scientists?Why in fact did he never get a positionin Paris, with its better libraries, bet-ter students and all the accompanyingprestige, which would have been ex-pected as an appropriate award forhis accomplishments? The reasons-sci-

entific, political, and religious—makean interesting story, not without paral-lels today.

PHILOSOPHICAL OUTLOOK

Duhem belonged to the Energeticistschool along with Ernst Mach, GeorgHelm, and Wilhelm Ostwald. Duhem'sviews on the nature of physical theorieswere similar to theirs and to HenriPoincare's. In common with them, hebelieved that physical theories serve asdescriptions that condense andclassify the results of experiments,rather than as explanations or inter-pretations of experiment. However, hediffered from the others in believingthat as physical theories evolve bysuccessive adjustments to conform toexperiment, they approach asymptot-ically a sort of "natural classification"that somehow reflects an underlyingreality.2' 3 This view removes him fromthe ranks of the positivists.

Another point emphasized by Du-hem was the impossibility of a truly"crucial experiment." An experimentaltest of one proposition necessarily in-volves all the other propositions of thetheory. Therefore the contradictions ofa theory by experiment can be re-moved not only by changes in the oneproposition believed to be "crucially"tested, but by changes in the otherpropositions leaving the "crucial" oneunchanged.-1 3

Duhem arrived at the descriptionposition more or less in the followingway. In the period 1850-1910, therewere two opposing viewpoints: one

that tiieories should be based on ex-periment and description, the otherthat all of physics could be reducedto classical mechanics and explainedby atomic and molecular models. Du-hem's scientific formation took placein the period 1880-1890, well beforethe discovery of radioactivity and theexperiments of Jean Perrin or HenryG. J. Moseley. At this time, the groupusing atomic and molecular modelsfor their "mechanical explanation ofthe universe" changed the propertiesof their "billiard balls" at will depend-ing on the phenomenon. These prop-erties, often contradictory, were ashard to believe in then as they arenow. Duhem, with his passion forlogic, could not accept such contra-dictory models. He became convincedthat on the contrary, instead of hav-ing all of physics reduced to classicalmechanics, classical mechanics itselfwas a special case of a more generalcontinuum theory.

Therefore along with the otherEnergeticists, Duhem believed a gen-

Donald G. Miller isa physical chemistwith a PhD fromIllinois. He has beenat the Lawrence Ra-diation Laboratory,Livermore, since 1956working in thermo-dynamics of irrever-sible processes, elec-trolyte solutions and die history ofscience—primarily thermodynamics.

PHYSICS TODAY • DECEMBER 1966 • 47

eralized thermodynamics was the un-derlying descriptive theory for all ofphysics and chemistry. He spent hisscientific life trying to build up sucha generalized thermodynamics, whichwould include mechanics and electrici-ty and magnetism as well. These at-tempts culminated in the workEnergetique, ° in which however hewas unable to include electricity andmagnetism. This valuable book, whichDuhem himself believed to be hismost important and lasting contribu-tion to science, is interesting in thatthere is not a single word aboutatoms and molecules.

HISTORICAL STUDIES

That Duhem is known to historians ofscience is not surprising. He discovered,essentially single-handed, the sciencedone in the middle ages. Besides LeSysteme du Monde1' and Leonard deVinci, 5 he wrote on the developmentof mechanics7- 8 and of theories ofheat.0 He began his historical studieshoping to find how gropings andchanges in a theory would lead to the"natural classification" mentioned ear-lier.

Duhem's principal contributions tothe history of science were his find-ings that (a) the middle ages werenot in fact dark ages for science; (b)an unbroken thread of physics wascarried on from about 1200 to therenaissance, principally within theChurch since there was little learningoutside of it; and particularly impor-tant (c) this science was known to daVinci and Galileo. Duhem hit uponthis by recognizing in da Vinci's note-books statements by earlier writersand references to manuscripts fortu-nately available in the BibliothcqueNationale. These discoveries, support-ed by quotations from the originaltexts given in detail in Lc Systeme duMonde, completely revolutionized thestudy of medieval science. Recent stud-ies have resulted in revision of cer-tain details, but Duhem's main find-ings remain unimpaired.

SCIENTIFIC WORK

Duhem began his scientific work withthe generalization and application ofthermodynamics. Exploiting theanalogy between the potential of classi-

cal mechanics and the more general"thermodynamic" potentials of F. J. D.Massieu and J. Willard Gibbs (that is,the Gibbs and Helmholtz free energiesG and A), Duhem treated a number ofsubjects systematically; for example:thermoelectricity, pyroelectricity, capil-larity and surface tension, mixtures ofperfect gases, solutions in gravitationaland magnetic fields, osmotic pressure,freezing points, stability of equi-librium, and the generalization of LeChatelier's principle. His success withthese problems in the period 1884-190010a rank him with J. H. Van'tHoff, Ostwald, Svante Arrhenius, andHenri Le Chatelier as one of thefounders of modern physical chemis-try.1013

Contained in his thermodynamicwork is a long critical study of thefirst and second laws, which for thatperiod includes an excellent set ofaxioms for thermodynamics.11 An-other paper contains an early attemptto treat a system equivalent to a liquidcrystal.12

The same continuum-theory out-look permeated his work on hydro-dynamics13 and elasticity.14 One ofhis principal results was that a trueshock wave (that is, with a discon-tinuity in velocity) cannot be propa-gated through a viscous fluid, althoughthere exist quasi shock waves withrapid changes of velocity across a thinlayer if the fluid is only slightlyviscous. Such questions treated by Du-hem 60 years ago have become im-portant recently and we find his hydro-dynamics book13 reprinted by theMinistere de l'Air of France.

A large portion of Duhem's writingsare on electricity and magnetism andhis attempts to bring them into theframework of Energetics. If the sys-tem's electrodynamic energy is con-stant, Energetics can be successfullyapplied. If not, then there are compli-cations that require choosing someelectromagnetic theory. Although Du-hem accepted Maxwell's genius, hecould not accept Maxwell's theory be-cause of the contradictions containedin its unrigorous development, itssign mistakes, and its lack of experi-mental foundation. Duhem preferreda more general electromagnetic theorydue to Hermann von Helmholtz,which could be constructed in a logical

way from the classical fundamental ex-periments of electricity and magnetism.This theory, which Duhem helpedelaborate, can also describe the electro-magnetic theory of light and the ex-periments of Heinrich Hertz at theexpense of some complication. More-over, it can be shown that Maxwell'stheory is a special case of Helmholtz'sby choosing appropriate values of twoconstants. Duhem was a pitiless criticof Maxwell's theory, claiming not onlythat it lacked rigorous foundation butthat it was not sufficiently general toexplain the existence of permanentmagnets. 15 Duhem later admittedthat not only had his criticisms notbeen accepted; they had not evenbeen read or discussed, and of courseMaxwell's theory has triumphed.

Duhem attached great importanceto his thermodynamics of false equilib-ria.16 False equilibrium correspondsto Gibbs's "passive resistance" and to"metastable" equilibrium. It char-acterizes substances such as diamondor petroleum that are unstable ther-modynamically with respect to othersubstances, yet have remained un-changed for geological periods of time,but which will transform into thestable products if the perturbations arelarge enough. False equilibrium canalso be interpreted, however, as a caseof extremely slow reaction rate, anda violent polemic over this issue tookplace between 1896 and 1910. Most,but by no means all, of those inter-ested in such questions prefer the in-finitely-slow-reaction-rate view today.Since essentially the same results comefrom either viewpoint, the choice is apersonal one.

Until recently, almost all of thiswork had been completely ignored.However, in the last few years Du-hem's contributions to hydrodynamicsparticularly have been receiving in-creasing attention, and a number ofpeople now publishing in the Archivefor Rational Mechanics and Analysiscite Duhem quite regularly (for ex-ample, C. Truesdell17).

LIFE AND CAREER

Pierre-Marie-Maurice Duhem was born10 June 1861 in Paris, and died 14September 1916 in Cabrespine.

His father, a commercial representa-tive, was originally from Roubaix in

48 • DECEMBER 1966 • PHYSICS TODAY

French Flanders. His mother, a Pari-sienne, was of a bourgeois familyoriginally from Cabrespine, a smallvillage in the Montagnes Noires notfar from Carcassonne. None of hisclose relatives had been extraordinarilydistinguished. He was the oldest offour children, with a brother andtwin sisters; the brother and one ofthe twins died at an early age.

Duhem was educated privately untilhe was 11. He then entered the Col-lege Stanislas, a Catholic lycee, to pre-pare for the entrance examinationsof the Ecole Normale Superieure andthe Ecole Poly technique, the trulyelite schools of France. He was abrilliant student, interested in every-thing, as good in history, languagesand letters as he was in science andmathematics. At Stanislas he was in-fluenced towards science, and in par-ticular towards thermodynamics, byJules Moutier. His ability in lan-guages, particularly in Greek and Lat-in, served him well later in his studyof medieval scientific manuscripts.

At the conclusion of his studies atStanislas, it seemed very probable thathe could pass the examinations forcither of the "grandes ecoles." Hisfather hoped Duhem would choose theEcole Polytechnique because his ca-reer would be assured. His mother,being very religious, wanted him toenter the Ecole Normale in humani-ties because she was afraid that engi-neering or science would destroy hisfaith. Duhem's choice was scienceat the Ecole Normale, indicating adesire for an academic career in sci-

ence. When the extremely rigorousentrance examinations were over, hewas ranked first and considerablyabove the other students. Howeverowing to a recurrent stomach dis-order, he did not enter until 1882.

The intellectual stimulation and thepossibilities for research at the EcoleNormale made Duhem's years thereamong his happiest. At the licence ex-aminations in 1885 he was first inphysics and in chemistry although onlyseventh in mathematics (out of 30) ,but his superiority was so obvious thathe was ranked first in his class. Atthe physics agregation examination in1885 he was again first.

Doctoral theses

The usual classwork of a Normalienhad not been sufficient for him. Hisfirst published paper appeared in 1884on the subject of chemical potentials.Moreover he presented a doctoral the-sis in physics (1884) before he hadreceived his licence (an undergrad-uate degree) .

This thesis was on the applicationof thermodynamic potentials to physi-cal and chemical problems. It also con-tained an attack on Berthelot's "maxi-mum-work principle"—which says thatthe criterion for a chemical reactionto occur is based on the heat ofreaction. This "principle" is of coursefalse, since the true criterion is basedon the thermodynamic potential.(However, the fact that the maximum-

work principle becomes more correctas the temperature goes to absolutezero led Walther Nernst to the third

DUHEM at three, at 12 andas a student at College Stanis-las where he prepared for hisentrance into the Ecole Nor-male Superieure in 1882.

law of thermodynamics.) Berthelot,who had already spent 20 years andwas to spend ten more trying to sustainhis principle, was extremely influentialin official circles and was able to getDuhem's thesis refused. To presentsuch a thesis called for unusual per-sonal courage on Duhem's part. How-ever Duhem believed it had to be doneto avoid the spreading of an incorrectthermochemistry, even if it risked sti-fling his advancement. Therefore af-ter the thesis was refused, Duhemhad it published separately as abook,18 which is important historical-ly for its systematic use of thermo-dynamic potentials in an era wheneveryone else was using artificial cyclicprocesses. The enmity of Berthelotand his friends on this issue aloneblocked Duhem's career for manyyears. Berthelot (who was Minister ofPublic Instruction in 1886-7) is re-ported to have said that Duhem wouldnever teach in Paris, and was in a posi-tion to make that statement true.At that period, and to a lesser extenteven now, the best and most influen-tial posts were in Paris.

After the refusal of this thesis, hestayed on at the Ecole Normale foranother two years until he had pre-pared a second thesis, this time for adoctorate in mathematics on the theoryof magnetism. This thesis was ac-cepted in 1888. During his five yearsat the Ecole Normale, he published30 papers and one book. In his lasttwo years, Duhem was invited by LouisPasteur to join his famous group. ButDuhem was too committed to physics

PHYSICS TODAY • DECEMBER 1966 • 49

PRESENTING A THESIS,that of Paul Saurel, a gradu-ate of the College of the Cityof New York, in 1900. Leftto right are Marchis, Du-hem, Brunei (dean), Saurel,Chevallier, Malus, Lenoble.

to accept this opportunity for an as-sured career.

In 1887, Duhem was appointedmaitre des conferences in the Faculte"des Sciences at Lille. It was a stimulat-ing atmosphere and his ideas onEnergetics began to take shape. Hisextraordinary ability as a teacher be-came evident, and students flocked tohis courses, impressed by a professorwho cared about them. While at Lille,he met Adele Chayet and they weremarried in October, 1890. Their firstdaughter Helene was born in Septem-ber 1891, but Mme Duhem and theirsecond daughter died in 1892. Du-hem's married life had been very hap-py but tragically short. He never remar-ried, thereafter devoting the rest ofhis life to his work and his daughter.

Shortly after the death of his wife,he fell into disagreement with thedean over how severely a laboratoryassistant should be reprimanded. Dur-ing the altercation, the two formerfriends came to blows. Thus in 1893,Duhem requested a transfer. In thesix years he was at Lille, he had pub-lished three books and 50 papers.

By this time, his brilliance, abilityand further promise were clearly evi-dent, and in ordinary circumstancesthese would have earned for him aposition in Paris. But the opposition,among others, of Berthelot and LeChatelier (with whom Duhem hadhad a priority fight) prevented suchan appointment, and Duhem wasnamed maitre des conferences atRennes. This situation was so badthat after one year Duhem requesteda change.

In exileOnce again a position in Paris wasexpected, but instead he was appointedcharge de cours at Bordeaux. Thisterribly disappointing setback causedDuhem to consider refusing the post.His friends, after discreet inquiries,told him that he must realize thatBordeaux was really the road to Paris.Therefore he accepted the position in1894. In 1895 at the age of 34 hewas appointed professor de physiquein a chair changed to theoretical phys-ics for him. He was to stay in exileat Bordeaux the remaining 21 yearsof his life, cruelly disappointed, neverto attain his ambition of teaching inParis.

Duhem's ability as a teacher, evi-dent at Lille, became almost legendary.His courses and public lectures weremodels of organization and lucidity,exhibiting the precision, logic and at-tention to detail that characterized hiswritings. He worked on university af-fairs for a while until personal at-tacks by anticlerical members of theuniversity administration caused himto resign from the Conseil de l'LTni-versite. He had been primarily respon-sible for the regulation permittingprovincial universities to grant doc-toral degrees for theses prepared inthe provinces, which had formerlybeen sent to the Sorbonne. His isola-tion in Bordeaux limited the num-ber of good students. The most prom-inent of them was Lucien Marchis,and the most prominent of his disci-ples and admirers were Emile Jou-guet and Louis Roy. Duhem also in-fluenced Jacques Hadamard, and in

particular introduced Hadamard toHogoniot's work on waves.

In 1900, he became a correspondantto the Academie des Sciences (Sectionde Mecanique) and in the course ofyears became a member of severalforeign academies of science and re-ceived two honorary degrees. Althoughit was clear that his work was ofvery high quality, and had been hon-ored abroad, Duhem was not recog-nized by the French Academie becauseat that time its members had to be inParis. However when a class of sixnonresident members was created in1913, Duhem was among the first sixelected. All earlier proposals to bringhim to Paris and elect him to theAcademie had met with uncompromis-ing opposition.

Duhem's interests fell roughly intoperiods. His work in thermodynamicsand electromagnetism was predomi-nantly in the period 1884-1900; hereturned to the latter between 1913and 1916. His interest in the philoso-phy of science was mostly between1893 and 1906, and the history ofscience primarily between 1904 and1916, although his earliest papers datefrom 1895. Finally he concentratedon hydrodynamics and elasticity be-tween 1900 and 1906.

His tremendous output was in partdue to regular, methodical work. Herelaxed by walking and reflecting.Once his ideas were well in mind, hesat down and wrote steadily, fillingpage after page without stopping, with-out erasures or words crossed out.At his death, he left complete manu-scripts ready to be printed, but no

50 DECEMBER 1966 PHYSICS TODAY

f notes whatsoever. (These manuscriptswere the last five volumes of LeSysteme du Monde, which were finallypublished in 1954-59 after 40 yearsof unceasing efforts by his daughter,Helene Pierre-Duhem. The only re-maining manuscript is 17 chapters onthe history and applications of capil-larity theories, written several yearsbefore his death.)

After his daughter grew older andleft for Paris to finish her education,Duhem lived alone in Bordeaux, work-ing steadily. He saw his daughter dur-ing vacations, which he spent eitherat the house inherited from his motheror in hiking over various parts ofFrance by himself or with an oldfriend. On these trips he would sketchthe countryside, inking or charcoalingthem in later when he wanted to re-

• lax from his writings. These land-f scapes are extraordinarily well done.

He showed this pronounced artisticability as a boy, and gained a greatreputation at College Stanislas andthe Ecole Normale as a caricaturist.

His life passed in this way: a mount-ing pile of publications, perusal ofmedieval manuscripts, vacations in themountains or at Cabrespine, and somecharitable works. As a result of the al-most universal scientific and some-times personal ostracism he suffered,Duhem had withdrawn from most so-cial contacts. The lack of students evenfor the licence and his personal dif-ficulties with the university adminis-tration caused him to reduce his con-tact with the university to his coursehours only. In 1913, however, when hewas elected a nonresident member ofthe Academie, this belated recognition,supported finally even by some of hisold enemies, pleased him enormously.Thereafter his personality seemed tobloom again, and his social contactsincreased. He became interested in theCatholic students' association and wasactive in a widows' and orphans' aidsociety after World War I started.

In the summer of 1916, Duhem andhis daughter went to Cabrespine asusual for their summer vacation withthe usual program of work, hiking,and sketching. During the eveningof 2 September, after a rather strenu-ous hike, Duhem suffered a painfulheart attack. The heart condition hadundoubtedly originated in an attack

of rheumatic fever he had had as achild. The doctor diagnosed it as angi-na, the symptoms of which had beenpresent for six or seven years, buthad been mistaken for the lifelongstomach cramps Duhem had suffered.Duhem resigned himself to the pre-scribed rest and limited activity. Onthe morning of 14 September, he feltsomewhat better. However, while ink-ing in a drawing and talking to hisdaughter about the war he had an-other attack and died within a fewminutes, prematurely, aged 55, leavingunfinished the erudite Systeme duMonde. He is buried in Cabrespine.

REASONS FOR OPPOSITIONCertainly Duhem's promise was evidentwhile he was a student, and under or-dinary circumstances, in view of hishigh-quality work and foreign honors,he would have been called to Paris andbeen elected to the Academie in 15to 20 years. Why was he not?

The reasons for Duhem's difficultiesinvolve certain aspects of Duhem's per-sonality as well as differences betweenscientific schools.

Personality

Duhem was absolutely honest and hadfirmly held opinions. He prided him-self on his independence, as evidencedby the publication of his refused thesis.He always protested vigorously thingshe believed unjust. He never feareda polemic, was a savage critic, andchose his adversaries without regardto rank or reputation; an example wasthe fight with the dean at Lille. An-other example can be found in oneof his attacks on the maximum-workprinciple where he pictures Berthelotstanding among the ruins of hisfavorite but now discarded scientificideas, listening to faint murmurs inthe background whispering about thesemen who "seemed during their livesto be something, and now no onespeaks of them any more." ld Thosewhose writings or teachings Duhembelieved to be inimical to science as heconceived it became practically per-sonal enemies.

Ordinarily he was a likable manof perfect courtesy and charm, withmany close friends. His students prac-tically adored him. On a person-to-

person level, he was conscientious,helpful, and deeply committed tocharitable works. However when hesaw someone else unjustly treated orhe felt his honor or integrity ques-tioned, he would unhesitatingly riseto attack and fight with the vervecharacteristic of his battle withBerthelot. In such circumstances, evenhis friends admitted that he some-times went too far. Such attitudes didnot "win friends and influence peo-ple," and it has been said that he andhis adversaries "ont fait la science avecl'epee." It is possible that the painfulwriter's cramp he developed at age40 and the very painful stomachcramps he suffered from all his lifecontributed to his sharpness andasperity.

Moreover Duhem was a sincere, be-lieving Catholic who deeply resentedthe liberal, anticlerical nature of theThird Republic. His political viewswere to the far right. He was a royalist,antidemocratic and anti-Republic, op-posed to universal suffrage, and evento the Christian Democratic move-ment. He had all the prejudices of thehigher bourgeoisie, an inheritancefrom his mother and her family, be-lieving completely in the "naturalsuperiority" of the upper classes, theChurch, and the army. Since the Drey-fus case was fundamentally a struggleof these groups versus the Republicand liberal thought, it is not surprisingto find Duhem an anti-Dreyfusard al-most to the bitter end. Along with theother inherited prejudices of his fam-ily he was antisemitic. Neverthelesshe maintained friendly relations withhis Jewish comrades from the EcoleNormale, such as Jacques Hadamard,all his life, and many of his closefriends were Dreyfusards.

Political and religious convictions ofthis kind were in complete contrastwith those held by most of his col-leagues and were especially offensiveto the free thinkers and liberals whodominated French science and educa-tion at that time. Thus a publicconflict over the Dreyfus case resultedin a bitter personal hatred betweenDuhem and certain members of theUniversity of Bordeaux administration.It is also very likely that the scientificantipathy between Berthelot and Du-hem was accentuated by such political

PHYSICS TODAY • DECEMBER 1966 • 51

and religious differences. (For exampleBerthelot was a great mechanistchampion in the mechanism-vitalismcontroversy, a controversy with im-portant philosophical and religiousovertones.) There is no doubt thatDuhem's advancement was retarded inpart by opposition to his politicaland religious views, a phenomenonreminiscent of the treatment of left-wing American scientists in the recentpast. In addition to keeping him inexile in the provinces and limiting hissalary, the opposition also took theform of a partial conspiracy of silence.His work, though popular abroad,was ignored or else taken over withoutcitation in France. It is curious thateven after 50 years there is much moreinterest now in other countries in Du-hem's work, particularly in the UnitedStates, than there is in France.

In the face of all this opposition,Duhem was characteristically self-ef-facing, never complaining of his trials.He did nothing to aid his advance-ment. He refused to engage in aca-demic politics, and he would not makethe expected visits to the influentialmembers of the Academie. He wasclearly a man of principle; he refusedto be nominated for the Legion ofHonor in 1908 because it would havebeen given by the Republic he hatedand signed by a man he detested.Similarly, when later in his life he wasurged to be a candidate for a Historyof Science Chair at the College deFrance in Paris, he refused, sayingthat he could not enter Paris by aback door; he was a physicist and onlyas a physicist would he go to Paris.When he was nominated as a non-resident member of the Academie in1913, characteristically he tried to with-draw so that the next closest candi-date, a 90 year old naturalist, couldbe elected before dying. Even hisenemies conceded that he wanted tobe in Paris not for the usual reasonsof prestige, but because his scientificinfluence would have been greaterthere than in Bordeaux.

Scientific opposition

Apart from the opposition based onpersonal antipathies, there was alsothe opposition of different scientificschools. In the case of the maximum-work principle, Duhem was r ight-

but the struggle partially ruined hiscareer. However the neglect of por-tions of his scientific work resultedfrom the triumphs of theories he bit-terly opposed.

For example, the atomic and mo-lecular theories he despised play amuch more predominant role in mod-ern science than does generalizedthermodynamics, thereby reducing theoverall significance of Duhem's con-tributions. Duhem of course was notalone in his distaste for atomic andmolecular models. The Energeticistsand many other eminent physicists be-fore 1905-10 did not believe in atomseither. Moreover until the discoveryof radioactivity and the experimentsof Perrin and Moseley, their beliefwas not as unreasonable as it mayseem now. (One of the most im-portant pieces of evidence for anatomic theory in 1900 was the Lawof Multiple Proportions. Hence it isinteresting that Ostwald using thework of F. Wald, was able to deriveit from the phase rule alone.20)Our understanding of atoms is verydifferent today than it was then. Du-hem held out against atoms longerthan most, because of the characteris-tic rigidity of his personality, in thehope that a more general theory likeEnergetics would arise from the ruinsof atomic theories.

For a second example, the completetriumph of Maxwell's theory of elec-tromagnetism over the competingHelmholtz theory has naturally left inobscurity Duhem's contributions to thelatter. Furthermore Duhem could notaccept relativity because he believedit violated "common sense" (a viewthen shared by many others). More-over it mutilated classical mechanicsin order to leave completely unalteredMaxwell's theory and atomic theoriesof electrons, both of which Duhemwas convinced were erroneous.

The polemical nature of some ofhis writings in these controversies suchavs energetics vs atomism, Maxwell'stheory vs Helmholtz's, false equilib-rium, and the maximum-work prin-ciple, made personal enemies of manyof his scientific contemporaries. Aswe have seen, Duhem's work coveredmany fields, and each one seemed toprovide him with a new set of enemies.Thus opposition to him was due not

only to Berthelot, who died in 1907.(Berthelot and his friend Lippmanwere clearly originally responsible forkeeping Duhem from a post in Paris.By 1900, however, Berthelot recog-nized his unfairness and thereaftervoted for Duhem's promotions. Thischange of attitude was appreciated byDuhem who, when ref. 19 was re-printed later without his permission,wrote Berthelot that his permissionwould have been refused.) Among hisother influential enemies according totradition were Lippmann, le Chatelierand among the atomists, Jean Perrin.(Perrin and Duhem were in agree-

ment however on the falsity of themaximum—work principle.) Theseand many more tried to block Du-hem both on personal and scientificgrounds. Although many of his sup-porters were prominent scientists, theydid not hold positions sufficiently in-fluential to help him.

AN ASSESSMENT

With hindsight, it would seem thatDuhem should not have so bitterlyopposed theories involving models.However his passion for logic andclarity caused him to be repelled bythe illogical and self-contradictorymodels in vogue during his formativeyears, and he built his whole scientificphilosophy on the deliberate avoid-ance of such aids. Because of therigidity of his personality, he couldnot change his views when the modelsbecame more plausible and more help-ful in new discoveries. The intuitiveapproach using models or other mentalaids are often, but not always, moreappropriate in the search for newphenomena. Efforts such as Duhem'stowards logical description and unifica-tion are usually best for continuumtheories that describe phenomena al-ready known. Nevertheless continuumtheories have been used recently, asin the past, for the discovery of newphemomena. Today science is con-cerned with both types of theory, aswell as combinations of them.

Although the value of Duhem's his-torical and philosophical researcheshave never been in doubt, our descrip-tion of triumphs of scientific theoriesthat Duhem opposed may leave a nega-tive view of the lasting value of hisscientific work. Indeed P. Humbert

52 • DECEMBER 7966 • PHYSICS TODAY

said21 in 1932: "Au fond Duhem apasse sa vie a etablir avec un soinextreme la grammaire d 'une langueqUe Ton ne parle plus." ("Funda-mentally Duhem spent his life estab-lishing with great care the grammar

of a language no longer spoken.")In my opinion this negative view

is not admissible. Even if some of hisfavorite ideas have been eclipsed bythe ascendance of atomic theories, hispurely scientific investigations inthermodynamics, physical chemistry,hydrodynamics, and elasticity are im-portant, useful and significant today.Moreover many of his works, particu-larly in thermodynamics, are of con-siderable historical importance. In hisrejected thesis18 he showed the greatutility of the Gibbs and Massieuthermodynamic potentials in the analy-sis of practical problems. He was theprincipal expositor of Gibbs's ideas inFrance and perfected them in de-tail.6' 1Oa- n He helped banish the er-roneous maximum-work principle fromscience, and he gave or anticipated anumber of important concepts ofthermodynamics. (For examples, we

References

1. Mem. Soc. Sci. Phys. Nat. (Bordeaux)(7), 1, part 2 (1927).

2. P. Duhem, La Theorie Physique, SonOb jet et sa Structure (Chevalier etRiviere, Paris, 1906) 450 pp.; 2nded., 1914; 3rd ed., 1933.

3. P. Duhem, The Aim and Structure ofPhysical Theory (Princeton UniversityPress, Princeton, 1954), 344 pp.,translated by P. Wiener; (paper, Ath-eneum, New York, 1963).

4. P. Duhem, Le Systeme du Monde:Histoire des Doctrines Cosmologiquesde Platon a Copernic (Hermann,Paris) I (1913) 512 pp., II (1914) 522pp., Ill (1915) 549 pp., IV (1916) 597pp., V (1917) 596 pp., VI (1954) 739pp., VII (1956) 644 pp., VIII (1958)512 pp., IX (1958) 442 pp., X (1959)528 pp.

5. P. Duhem, Etudes sur Leonard deVinci, Ceux qu'il a I us et ceux quiI'ont lu (Hermann, Paris) I (1906) 355pp., II (1909) 474 pp., I l l (1913) 610pp.

6. P. Duhem, Traite d'Energetique,(Gauthier-Villars, Paris, 1911), 2volumes of 528 pp. and 504 pp.

7. P. Duhem, L'Evolution de la Mecani-que (Joanin, Paris, 1903) 348 pp;taken from Rev. Gen. Sci. 14, 63, 119,171, 247, 299, 352, 416 (1903).

8. P. Duhem, Les Origines de la Stati-que (Hermann, Paris, 1905-06) 2volumes of 360 and 364 P P ; taken

note the first precise definition of areversible process, the differentiationbetween quasistatic and reversibleprocesses, the first serious axiomatiza-tion of thermodynamics, the definitionof heat in terms of work and energy15 years before Caratheodory, the useof Euler's theorem for partial molalquantities, and a generalized phaserule including masses.) His work onquasi shock waves13 and systems likeliquid crystals1- is also important.

Pierre Duhem is a fascinating ex-ample of a brilliant scientist fearlesslyadhering to his views, who was caughiup in historical and personal circum-stances that blocked his career andpartially suppressed his scientific work.Practically a right-wing, royalist, anti-semitic, anti-Republic, and religiousextremist, he was exiled to the prov-inces and his scientific work almostsystematically ignored in France ow-ing to opposition to his unpopularpolitical and religious views, as wellas to his scientific positions.

In my opinion Duhem's scientificwork, although ignored until recentlyfor interesting but invalid historical

from Rev. Quest. Scient., 4, 462 (1903);5, 560 (1904); 6, 9, 394 (1904); 7, 96,462 (1905); 8, 115, 508 (1905), 9,115, 383 (1906); 10, 65 (1906).

9. P. Duhem, "Les Theories de la Cha-lcur," Rev. Deux Mondes 129, 869(1895); 130, 380, 851 (1895).

10. (a) P. Duhem, Traite Ele'mentaire dcMe'canique Chimique (Hermann,Paris, 1897-99), 4 volumes of 299, 378,380, and 381 pp.; (b) J. R. Partington,A History of Chemistry (Matmillan,London, 1964), vol. 4, p. 616.

11. P. Duhem, J. de Math, pures ct ap-pliq., 8, 269 (1892); 9, 293 (1893); 10,203 (1894).

12. P. Duhem, Ann. Ecole Normale, 10,183 (1893).

13. P. Duhem, Recherches sur VHydrody-namiquc (Gauthier-Villars, Paris) lc

Serie (1903), 212 pp.; 2C Serie, (1904),153 pp.; new edition (including sevennotes from Comptcs rendus, Publica-tions Scientifiques et Techniques duMinistere de l'Air, Paris, 1961), 396pp.

14. P. Duhem, Recherches sur VElasticite,(Gauthier-Villars, Paris, 1906) 218pp.

15. P. Duhem, Les Theories Electriquesde J. C. Maxwell: Etude Historique etCritique (Hermann, Paris, 1902) 228pp.; taken from Ann. Soc. Sci. (Brux-elles), part II, M£m., 24, 339 (1900);25, 1, 293 (1901).

reasons, now deserves more recog-nition. There is of course no questionabout the importance of his work inthe philosophy and history of science.Since his contributions to any one ofthe fields of pure science, philosophyor history would have done credit toone person, the ensemble from the penof a single man marks Duhem as oneof the greatest intellects of hisperiod.

Existing biographical literature-1-^ hasof course been drawn upon heavily. Inaddition considerable useful informationhas been obtained from a number ofpeople, including J. Hadamard, A. Mas-sou Her, iV. Dufourcq, D. Cot, R. Taton,and E. Bauer. I am especially grateful toHe'lene Pierre-Duhem for the invaluableinformation about her father's life andcareer so graciously given to me in severalpersonal intemiews. This work was donein part during the tenure of a EulbrightFellows/iip i)i France 1960-61, in partunder the auspices of the U.S. AtomicEnergy Commission, and in part with theaid of a grant from the Research Com-mittee, University of California at SanDiego. This paper Jias been adapted froma more detailed manuscript being pre-pared for publication elsewhere. Q

16. P. Duhem, Theorie Thermodynami-que de la Viscosile, du Frottement, etdes Faux Equilibres Chimiques (Her-mann, Paris, 1896), 210 pp.

17. C. Truesdcll, J. Rat. Mcch. Anal. 1,125 (1952); 2, 593 (1953); C. Trues-dell, R. Toupin, "The Classical FieldTheories" in Handbuch der Physih,(Springer-Verkig, Berlin, I960), vol.I I I / l .

18. P. Duhem, Le Potentiel Thermodyna-mique et ses Applications a la Me'-canique Chimique et a la Theorie desPhenomenes Electriques (Hermann,Paris, 1886), 248 pp.

19. P. Duhem, "Thermochinric, a proposd'un livre recent dc M. Marcelin Ber-ihelot," Rev. Quest. Sci. 42, 361(1897).

20. W. Ostwald, J. Cliem. Soc. (London)85, 506 (1(JO1).

21. P. Humbert, Pierre Duhem (Bloud etGay, Paris, 1932) 147 pp.

22. E. Jordan, Ann. Assoc. And ens El evesEcole Norm. Sup. 158 (1917); Mem.Soc. Sd. Phys. Nat. (7), 1, part 1(1917).

23. E. Picard, La lie et VOeuvre dePierre Duhem, Gauthier-Villars, Paris,1921), 44 pp.

24. Helene Pierre-Duhcm, Un SavantFrancais, Pierre Duhem (Plon, Paris,1936).

25. L. de Broglie, foreword to ref. 3.

PHYSICS TODAY • DECEMBER 1966 • 53