gayon

Journal of the History of Biology 31: 305325, 1998. 1998 Kluwer Academic Publishers. Printed in the Netherlands. 305

The Concept of Individuality in Canguilhems Philosophy ofBiology

JEAN GAYONUniversit Paris 7-Denis Diderot2 place Jussieu75251 Paris, France

Although he did not write many books and although he wrote on a limitedrange of subjects, Georges Canguilhem (19041996) is without doubt oneof the most influential figures in French intellectual history in the secondhalf of the twentieth century. His influence has been particularly importantamong philosophers, historians of science, and physicians. For the Frenchlearned public, it is almost impossible to dissociate the three main aspects ofCanguilhems intellectual activity: medical philosophy, the history of biology,and the philosophy of science. Abroad, especially in America, the two firstaspects have become relatively well known. Canguilhems book The Normaland the Pathological (first French edition: 1943) was first translated in 1978,and was published again in 1989; in the 1970s, Canguilhem was awardedthe Georges Sarton Medal by the History of Science Society for his workon the history of biology. However, in terms of the philosophy of science,and especially the philosophy of the life sciences, until recently Canguilhemwas almost completely unknown among American scholars. Hardly a singlereference to him could be found in the current literature of the philosophy ofbiology.1

This paper does not intend to provide an exhaustive account ofCanguilhems thinking. It will focus on his philosophical approach to thebiological sciences. Strictly speaking, I should avoid using the expressionphilosophy of biology because, in spite of its apparent neutrality, it refers toa specifically American tradition of the past thirty years. Canguilhem, like allsimilar European thinkers of the same vein, sometimes used the expressions

1 In his Philosophy of Biology Today (Albany: State University of New York Press, 1988),Michael Ruse devotes one chapter to other lands, that is to say, other than Canada, theUnited States, and England. The only sentence on France says: France has little interest onthe subject (p. 86), which is frankly astounding for any French philosopher. The fifty-twopages of bibliography do not mention Canguilhems name.

306 JEAN GAYON

biological philosophy or philosophy of the life sciences, but he nevermet actually he probably ignored the expression philosophy of biology,a new phrase that appeared in the United States in the late 1960.2

Canguilhems biological philosophy will not be found in a single book,or even in a single paper. It is scattered throughout his publications, frombeginning to end. It will be useful here to briefly review the three majorphases of Canguilhems thought.3 The first phase, the philosophy of medi-cine, consists of a medical dissertation, The Normal and the Pathological, abook he wrote and published during the war (this was hardly anodyne, giventhat Canguilhem played an important role in the Resistance) when he wasthirty-nine (1943). Note that Canguilhem was first trained in philosophy, notmedicine; he began studying medicine only in the mid-1930s, while he wasteaching philosophy in a lyce.4 After the war, Canguilhem moved toward thehistory of biology. His main achievements during this second phase were TheFormation of the Concept of Reflex in the 17th and 18th Centuries (1955)5 in which he demonstrated the fertility of vitalistic traditions and a seriesof essays published as La Connaissance de la Vie (1952).6 These essaysdealt with biological experimentations, cell theory, vitalism, mechanicism,and the notion of milieu. The last and longest period begins after 1955, theyear when Canguilhem succeeded Gaston Bachelard as professor of philos-ophy and history of science at the Sorbonne, and as director of the Institutdhistoire des sciences. In this phase, Canguilhem wrote a variety of mono-graphs that intimately associated epistemology, the history of science, andthe history of philosophy, and that illustrate what has sometimes been called

2 David Hull, What Philosophy of Biology Is Not, J. His. Biol., 2 (1969), 241268. Onphilosophy of biology versus biological philosophy, see Jean Gayon, La philosophie etal biologie, in J. F. Mattei, ed., Encyclopedie Philosophique Universelle, vol. IV (in press).

3 I rely here on Franois Dagognets periodization in Une Oeuvre en Trois Temps, Revuede Metaphysique et de Morale, 90 (1985), 2938.

4 Georges Canguilhem, Essai sur Quelques Proble`mes Concernant le Normal et lePathologique, Publications de la Faculte des Lettres de lUniversite de Strasbourg (Clermont-Ferrand: Imprimerie La Montagne, 1943). A second edition appeared in 1950 under thesame title, with no alteration except a new preface (Paris: Les Belles Letres, 1950). In 1966,the 1950 edition became the first section of a book entitled Le Normal et le Pathologique(Paris: Presses Universitaires de France, 1966). The second section of this book included threenew short essays written 19631966. A second revised edition of the 1966 book was publishedin 1972 (same publisher); it was translated into English by Carolyn Fawcett under the title TheNormal and the Pathological (Dordrecht: Reidel, 1978). This translation was itself reprintedby Zone Books in 1989. This is the edition quoted in the present paper.

5 Georges Canguilhem, La Formation du Concept de Reflexe aux XVIIe et XVIIIe sie`cles(Paris: Presses Universitaires de France, 1955; reprinted, 1977). Not yet translated intoEnglish.

6 Georges Canguilhem, La Connaissance de la Vie (Paris: Hachette, 1952; 2nd rev. edn.,Paris: Vrin, 1965). There have been many reprints. Not yet translated into English.

CANGUILHEMS PHILOSOPHY OF BIOLOGY 307

Canguilhems epistemological history7 or historical epistemology.8 Themain essays of this period have been gathered into two volumes: Studies inthe History and Philosophy of Science,9 and Ideology and Rationality in theHistory of Life Sciences.10

This thematic account of the development of Canguilhems work iscorrect, but it does not reveal much about the possible evolution of his philo-sophical theses. The important point is that all periods and thematic (if notdisciplinary) aspects of Canguilhems work are equally relevant for the under-standing of his biological philosophy. Indeed, each of the three thematic fieldsthat have just been enumerated (philosophy of medicine, history of biology,philosophy of science) was marked by the development of a characteristickind of reflection about the ultimate philosophical significance of the exis-tence of living beings. In a recent book, Franois Dagognet has argued that theentire work of Canguilhem can be understood as a deepening of his 1943idea of the normativity of life.11 Although I am convinced by Dagognetsargument, I have chosen a different approach. Instead of focusing on a singlethesis, I deal with a question that led Canguilhem to make significant changesin his approach to biological philosophy. My Ariadnes thread is the conceptof individuality. The successive treatments of this concept indicate a signifi-cant evolution in Canguilhems biological philosophy. At three successivepoints in the development of his thought, the concept of individuality playeda crucial role. First, individuality is a key word in Canguilhems originalstudy, The Normal and the Pathological (and therefore in his philosophyof medicine). Second, the problem of the nature of biological individuals is

7 Dominique Lecourt, Pour une Critique de l Epistemologie (Paris: Maspero, 1972), p. 64.8 Franois Delaporte, ed., A Vital Rationalist: Selected Writings from Georges Canguilhem

(New York: Zone Books, 1994), p. 43. According to Lecourt, who coined both epistemolog-ical history and historical epistemology, the latter applies to Gaston Bachelard; the formerto Georges Canguilhem. The origin of these expressions is quite interesting. When he waswriting his masters thesis on Bachelard under the direction of Canguilhem, Lecourt toldCanguilhem that he would describe Bachelards philosophy of science as historical episte-mology. Canguilhem answered: epistemological history (D. Lecourt, personal communica-tion). See also L Epistemologie Historique de Gaston Bachelard (Paris: Vrin, 1969). Finally,Lecourt retained the first expression on the title of his thesis and of the corresponding book. Itis obvious that epistemological history corresponds better to what Canguilhem himself did.

9 Georges Canguilhem, Etudes dHistoire et de Philosophie des Sciences (Paris: Vrin,1968; 2nd edn., 1970; 3rd edn., 1975). Here I quote the third edition. This book has not beentranslated into English.

10 Georges Canguilhem, Ideologie et Rationalite dans lHistoire des Sciences de la Vie.Nouvelles Etudes dHistoire et la Philosophie des Sciences (Paris: Vrin, 1981). Translatedinto English by Arthur Goldhammer as Ideology and Rationality in the History of Life Sciences(Cambridge: MIT Press, 1988).

11 Franois Dagognet, Georges Canguilhem, Philosophie de la Vie (Paris: Les Empcheursde Penser en Rond, 1997).

308 JEAN GAYON

predominant in a series of articles published between 1945 and 1960, whichdeal with various biological theories. In the end, individuality is the enigmathat lies behind Canguilhems reflection in the 1960s on the relationshipbetween life and knowledge. In these three situations, individuality standsas a problematic notion that is dealt with counterintuitively. In the context ofmedical philosophy (questioning the nature of illness), Canguilhem viewedindividuality as an axiological rather than as an ontological notion. In thecontext of his examination of biological theories such as cell theory or reflextheory, Canguilhem addressed the ontological issue, the issue of what indi-viduals consist of. His answer was that individuals should not be conceivedof as beings but as relations. Finally, in his reflection upon the relationbetween knowledge and life, Canguilhem finally argued that modern molecu-lar genetics, insofar as it provides an interpretation of biological individualityas a communication of information, enables us to provide a precise mean-ing to Hegels obscure but interesting identification of life and concept (inCanguilhems terms: there is a logos inscribed, preserved and transmitted inliving beings, hence life is concept).12

I will return to these issues below. For the moment, it will suffice tounderstand that individuality played the role of a problematic concept, whichhappened to drive the three successive major dimensions of Canguilhemsphilosophy of biology: the axiological, the ontological, and the gnoseological(theory of knowledge). In other words: three times running, Canguilhem feltit necessary to reassess the concept of individuality in order to develop hisconception of life as value, as being, and as knowledge. These assertionswill be explained in the rest of this paper. In my conclusion, I will makesome observations on the relations between the philosophy of medicine,the philosophy of biology, and the history of biology on both sides of theAtlantic. In contrast to a dominant and deliberate stream in American sciencestudies, Canguilhem work offers a strong integrated approach to these fieldsof research.

As far as possible, Canguilhem will be quoted in English translation.When no translation is available, or when translation raises problems, Iwill refer to the most available French editions. Thus far, only two bookshave been completely translated, The Normal and the Pathological,13 andIdeology and Rationality in the History of the Life Sciences,14 that is to say,the first and the last book. A number of significant excerpts from all aspects

12 Georges Canguilhem, Le Concept et la Vie (1966), Etudes dHistoire et de Philosophiedes Sciences, 3rd edn. (Paris: Vrin, 1975), pp. 362, 364. (English translation (partial): A VitalRationalist (above, n. 8), pp. 317318.

13 See above, n. 4.14 See above, n. 10.

CANGUILHEMS PHILOSOPHY OF BIOLOGY 309

of Canguilhems production, including some unpublished manuscripts, havebeen translated in a remarkable volume edited by Franois Delaporte underthe title A Vital Rationalist: Selected Writings from Georges Canguilhem.15This volume includes the most complete bibliography of and on Canguilhempublished in any language.

The first and major aspect of Canguilhems reflection upon biological indi-viduality can be found in his 1943 Essay on Some Problems Concerning theNormal and the Pathological.16 This essay in the philosophy of medicine waspublished during the war as a doctoral dissertation in medicine. Many peopleconsider it to be Canguilhems masterpiece, probably because it is the mostsystematic text he published in philosophy proper. In fact, if we leave aside alow-circulation textbook of morals published just before the war,17 the 1943essay is Canguilhems only synthetic book on philosophy. Since 1966, thisbook has been issued several times as The Normal and the Pathological. The1966 edition includes three additions, written between 1963 and 1966, andentitled New Reflections on the Normal and the Pathological. Except forvery minor corrections, the 1943 essay is unaltered.

The theses exposed in the book were largely inspired by Kurt GoldsteinsStructure of the Organism (1934).18 There is no doubt that Goldstein wasCanguilhems real master, that is, his chosen master. Canguilhem took fromGoldstein both his major subjects of investigation (health and illness, reflex),and a number of fundamental theses. In The Structure of the Organism, thereis a chapter entitled Norm, Health and Illness. Anomaly. Heredity and Selec-tion (chap. 8). In themselves, the terms of this title shed light on the structureof the 1943 essay (especially the second or philosophical part). However, thechapter also contains several theses that are central to Canguilhems argumentabout the notions of the normal and the pathological. These theses can beeasily set out. Those who are familiar with Canguilhems 1943 book willimmediately recognize them. Goldsteins basic idea was that the philosoph-ical determination of the concepts of normality and illness initially requirethe concept of the individual being.19 Health cannot be properly under-stood as a statistical norm; illness is not merely a deviation from such anorm.20 For Goldstein, health, illness, and recovery must be understood on

15 See above, n. 12.16 See above, n. 4.17 Georges Canguilhem and Camille Planet, Traite de Logique et de Morale (Marseille:

Imprimerie F. Robert & fils, 1939).18 Kurt Goldstein, Der Aufbau des Organismus (La Haye: Nijhoff, 1934).19 Ibid., p. 269.20 Ibid., pp. 266267.

310 JEAN GAYON

the basis of the notion of an individual norm: there is only one relevantnorm; that which includes the total concrete individuality; that which takesthe individual as its measure; it is therefore a personal individual norm.21Consequently, health should be conceived of as an adaptation to a personalmilieu;22 illness should be seen as a qualitative modification leading to anarrowing of the personal milieu23 and recovery will not be conceivedof as a return to a previous state but as the accession to a new individualnorm.24 Furthermore, for Goldstein, an important corollary of these concep-tions was the distinction between an anomaly and an illness. An anomalyis a deviation from a supraindividual norm, whereas an illness must refer topersonal individuality.25 Goldstein thought one should be cautious in usingthe concept of anomaly in medicine because the concept paves the way tosocial discrimination and possibly to racism.26 Finally, Goldstein concludedthat illness was a specifically and essentially human trait, one that goeshand in hand with human liberty.27 It is perhaps significant that Goldsteinpublished his book in 1934, when he decided to leave Germany after Hitlersaccession to power and take refuge in Holland.

Canguilhems important debt to Goldstein, which has oddly goneunnoticed among his commentators, need not be dealt with in any moredetail. All the theses quoted above are explicit and central in Canguilhemsessay. Furthermore, it was certainly not a neutral act to set out the views of adissident German physician (and a Jew at that) on the individuality of illnessand recovery in the historical context of the war and the Nazi occupation ofFrance.

This being said, Canguilhems essay was not merely a repetition ofGoldsteins theses.28 The originality of the book does not consist so much inthe theses as in the argument constructed by Canguilhem in support of Gold-steins theses on illness. I will therefore consider the argumentative structureof The Normal and the Pathological and show how this book constitutes ameditation on biological individuality.

21 Ibid., p. 269.22 Ibid., p. 271.23 Ibid., pp. 276280.24 Ibid., p. 272.25 Ibid., p. 283.26 Ibid., pp. 287291.27 Ibid., p. 292.28 To the end of his life, Canguilhem went on quoting Goldstein, always with great esteem

and approval. See, for instance, La Connaissance de la Vie (above, n. 6), pp. 1113, 24, 146;Etudes dHistoire et de Philosophie des Sciences (above, n. 9), p. 347; Ideologie et Rationalitedans lHistoire des Sciences de la Vie (above, n. 10), p. 138.

CANGUILHEMS PHILOSOPHY OF BIOLOGY 311

The 1943 essay addresses two questions: (1) Is the pathological statemerely a quantitative modification of the normal state? (2) Do sciences ofthe normal and pathological exist?29 The first question refers to a commonconception of illness in modern medicine, according to which pathologicalphenomena.30 Canguilhem provides a fascinating comment on the historyof this idea, often referred to as the positivist conception of illness.I will merely point out his criticism of this quantitative conception. ForCanguilhem, an illness cannot be reduced to an increase or decrease of agiven physiological parameter. Such a quantitative modification may be a signof illness, but it will be pathological only insofar as it reflects an alterationin the organism as a whole.31 For instance, the same quantity of glucose inthe blood will be pathological for one individual but not for another, as afunction of other parameters that interact with glycemia. Hence the formula:what makes a symptom or a functional mechanism pathological is its innerrelation in the indivisible totality of an individual behavior.32 This could bedescribed as the ontological side of Canguilhems criticism of the quantitativeconception of illness. Illnesses are not beings, but they affect certain kinds ofbeings, at a certain level of organization: strictly speaking, it is not moleculesor tissues or even organs that are normal or pathological; it is the individualorganism as a whole, in a given environment.33

The other aspect of the criticism can be described as the axiological side.For Canguilhem, the concepts of the normal and the pathological have tobe interpreted in terms of vital values. Illness cannot be properly under-stood without postulating the existence of negative values, even amongvital values.34 Illness cannot be reduced to a deviation from the normal orphysiological state, it must be philosophically represented as intrinsically ill(this is why, in general, I translate Canguilhems maladie as illness ratherthan disease, as translators commonly do). Once again individuality servesas the key argument: it is because actual human individuals feel sick that theart of medicine exists, and not vice versa.35 In this context of argumentation,the concepts of vital values, norms, and individuality and their rela-tions are taken as referring specifically to human subjective experience, withits existential and psychological aspects. Although this does not make theargument a weak one, Canguilhem could not be satisfied with it. He wanted

29 These two questions are the titles of the two parts of the book.30 Canguilhem, The Normal and the Pathological (above, n. 4), p. 42.31 Ibid., I, III.32 Ibid., p. 88.33 Ibid., p. 223.34 Ibid., p. 104.35 Ibid., pp. 9293.

312 JEAN GAYON

also to provide a more objective formulation of his axiological interpretationof illness. Putting together objective and axiological may seem strangehere because the current philosophy of medicine opposes the subjectiveand the objective of illness. In fact, Canguilhem wanted to generalize theaxiological concept of illness (in terms of norms, values, and individual expe-rience) in such a way that its formulation would be freed from properly humanpsychological testimony. This did not mean denying the importance of thesubjective and psychological aspect of illness in man; the issue was rather tosituate the strange experience of illness in the wider context of life. In otherwords, the issue was to understand better the relation between medicine andbiology, each one enlightening the other.

As I see it, this is the reason that motives the second part of The Normaland the Pathological. It is at this point that we can find the most original partof Canguilhems interpretation of illness, an interpretation that also led him toa somewhat unusual view of biological individuality. The title of the secondpart addresses the following question. Do sciences of the normal and thepathological exist? At first sight, this rather strange question is equivalentto the traditional one: Is medicine a science? In itself this question is notoriginal, nor is Canguilhems conviction that medicine is not a science butan art. What is original is the argument. After the first part of the disserta-tion, we have indeed two reasons to suspect that medicine might be a veryuncommon science. First, if illness has to be conceived as a global behavioralalteration in an individual organism, the science of illness will hardly providean example of the classical idea of science as a knowledge of laws. Second,if the normal and the pathological are irreducible axiological categories, thenmedicine will not consist only of descriptive statements; it will indeed criti-cally involve value judgements.36 Science can hardly harbor such judgments.In evoking Goldstein, I have already mentioned Canguilhems final positionon these difficulties, which consisted of assuming and exploring the conceptof individual norm, a concept that implies, as for Goldstein, refusing toidentify norm and mean, anomaly and illness. What distinguishesCanguilhem is his attempt to clarify the concept of individual norm, andmore generally the significance of the concepts of positive and negativevital values. Canguilhem explores two lines of argument. One of them bearson medicine proper. The other bears upon the inevitability of axiology inbiology in general.

The clinical argument, again, is borrowed from Goldstein.37 Goldstein,who reasoned in the context of neurology, refused to interpret illness as ananomaly (a deviation from the statistical mean) because he considered that

36 Ibid., II, I.37 Canguilhem, The Normal and the Pathological (above, n. 4), II, IV.

CANGUILHEMS PHILOSOPHY OF BIOLOGY 313

the frontier between the normal and the pathological was uncertain for manyindividuals taken simultaneously. Correlatively, he asserted that this limit wasperfectly clear for the same individual considered at different times. Conse-quently, illness is not a deviation from the norm but the emergence of a newindividual norm, the notion of norm being interpreted as a certain struc-turing of the relation between the individual organism and its environment.Illness, Goldstein admitted, is an inferior norm, but only in the sense that thesick person is an individual whose environment is narrowed in comparisonto that of a normal individual. Similarly for Goldstein, healing did not meana return to a previous or normal state: To become well again, in spite ofdefects, always involves a certain loss in the essential nature of the organism.This coincides with the reappearance of order. A new individual norm corre-sponds to this rehabilitation.38 On these matters, Canguilhem acknowledgeshis debt to Goldstein through extensive quotations.39 However, there are threedistinctive aspects of his treatment of Goldsteins idea that both illness andcure imply the institution of a new individual norm. First, Canguilhemextends Goldsteins neuropsychiatric theses to any kind of illness. Second,he insists on the objectivity of these theses: One would gladly emphasizehere as opposed to one way of citing Goldstein which gives the appearanceof initiation into a hermetic or paradoxical physiology the objectivity andeven banality of his leading ideas. It is not only the observations of clinicians. . . but also experimental verifications which go along the lines of his ownresearch.40 Third and above all, Canguilhem proposes a new concept thatof normativity in order to analyze better the role of norms in the notionsof illness, cure, and health. Whereas normality is a statistical concept thatrefers to the commonest adaptation to ordinary conditions of life, norma-tivity is more than that: it means an organisms ability to adopt new normsof life. Illness does not mean primarily that there is a deviation from somestatistical norm but that the organisms ability to tolerate an environmentalchange has been reduced.41 Correlatively, Health is a margin of tolerancefor the inconstancies of the environment . . . Man feels in good health whichis health itself only when he feels more than normal that is, adapted tothe environment and its demands but normative, capable of following newnorms of life.42 And finally, cure does not mean a return to some normal

38 Kurt Goldstein, quoted in Canguilhem, The Normal and the Pathological (above, n. 4),p. 194.

39 See, for instance the quotation, ibid., pp. 194195.40 Canguilhem, The Normal and the Pathological (above, n. 4), p. 195.41

Disease is characterized by the fact that it is a reduction in the margin of tolerance forthe environments inconstancies. Ibid., p. 199.

42 Ibid., pp. 197, 200.

314 JEAN GAYON

or typical state but always a genuine physiological innovation:43 in anycase no cure is a return to biological innocence. To be cured is to be given newnorms of life, sometimes superior to the old ones. There is an irreversibilityof biological normativity.44 This is indeed the most famous sentence of thewhole book. In Canguilhems mind, the philosophical redefinition of illness,health, and cure in terms of normativity did not imply the total irrelevanceof the current notions of the normal and the abnormal. A disease or apathological state cannot be called abnormal in an absolute sense, but it canbe said to be abnormal in a well-defined situation.45 A disease is thus akind of biological norm, whereas being healthy is not strictly synonymouswith being normal: being healthy means being not only normal in agiven situation but also normative in this and other eventual situations.What characterizes health is the possibility of transcending the norm, whichdefines the momentary normal, the possibility of tolerating infractions of thehabitual normal and instituting new norms in new situations.46 We beginto understand the unusual lesson conveyed by the apparently neutral title ofthe 1943 essay: the pathological is not opposed to the normal. Health,illness, cure, all these fundamental medical notions indicate the openness andirreversibility of life. Canguilhems medical meditation on normativity is ameditation on individuality.

Besides these clinical arguments, Canguilhem proposes another argumentin favor of the thesis that biological norms should always be interpreted onan individual basis. This argument consists of questioning the significanceof the concept of normality in biology in general. The schema is utterlyand explicitly Darwinian. For any living being, normality does not consist inconforming to a type, even a statistical type. Normality and abnormality haveno biological signification whatsoever if one does not take into account theenvironment. Canguilhems move from medical philosophy to philosophicalbiology is worth locating precisely. Here is the crucial passage in The Normaland the Pathological:

The problem of distinguishing between an anomaly whether morpho-logical like the cervical rib or sacralization of the fifth lumbar, or func-tional like hemophilia, hemeralopia or pentosuria and the pathologicalstate is not at all a clear one; but it is nevertheless quite important from thebiological point of view because in the end it leads us to nothing less thanthe general problem of the variability of organisms and the significanceand scope of this variability. To the extent that living beings diverge from

43 Ibid., p. 196.44 Ibid., p. 228.45 Ibid., p. 196.46 Ibid., pp. 196197.

CANGUILHEMS PHILOSOPHY OF BIOLOGY 315

the specific type, are they abnormal in that they endanger the specific formor are they inventors on the road of new forms? . . . When a Drosophilawith wings gives birth, through mutation, to a Drosophila without wingsor vestigial wings, we are being confronted with a pathological fact ornot?

The Canguilhem refers to Georges Teissier and Philippe LHritiers exper-imental work on natural selection in populations of Drosophila in the1930s:

Teissier and Philippe LHritier have demonstrated experimentally thatcertain mutations, which can seem disadvantageous in a species usuallyappropriate environment, can become advantageous should certain condi-tions of existence vary. In a sheltered and closed environment Drosophilawith vestigial wings are wiped out by Drosophila with normal wings.But in an open environment the vestigial Drosophila do not fly, feedconstantly, and in three generations we see sixty percent vestigialDrosophila in a mixed population. This never happens in a closedenvironment.47

It is hardly necessary to underline Canguilhems interest in Teissier andLHritiers work on natural selection. The two young biologists were theonly ones to be interested both in population genetics and Darwinism inFrance at that time; and Canguilhem is the only philosopher who noticedthis work. The important point, however, is the conclusion that Canguilhemdraws from his comparison between the Darwinian approach to variabilityand the problem of the normal and the pathological: An environment isnormal because a living being lives out its life better there, maintains its ownnorm better there. An environment can be called normal with reference tothe living species using it to its advantage. It is normal only in terms of amorphological and functional norm.48

Thus the medical problem of what is normal and what is pathological isonly a particular case of the general biological problem of the nature vitalvalues. For any living being, the norm is nothing other than the kind ofdeviation that natural selection maintains.49 In other writings, Canguilhemused even more vigorous expressions. In 1951, for instance, in a new articleon the normal and the pathological, he came to stress the inseparability of the

47 Ibid., p. 142.48 Ibid.49 Georges Canguilhem, Nouvelles Reflexions Concernant le Normal et le Pathologique

(19631966), in Le Normal et le Pathologique (Paris: Presses Universitaires de France, coll.Quadrige, 1972), p. 197 (my translation). Corresponding passage in English: Canguilhem,The Normal and the Pathological (above, n. 4), p. 263.

316 JEAN GAYON

fact of individual variation within species and the problem of biological value:[I]rregularity and anomaly are not mere accidents affecting the individual,they constitute its very existence . . . Individual singularity can be interpretedas a failure or an attempt, as a fault or as an adventure. . . . There is no a prioridifference between a successful form and a wasted form. . . . All successesare threatened because individuals, as well as species, die. It is the futureof forms that determine their values.50 In one of his last writings on thequestion of normality, Canguilhem was as clear as he could be about theDarwinian schema that provide his answer to this problem: Although Darwinintroduced into biology a criterion of normality founded on the relation ofthe living being with life and death, he in no way eliminated the notion ofnormality in the determination of the biological object.51 In fact, as we havealready seen, Canguilhems reference to Darwinism was no less clear in the1943 essay. This reference was indeed extremely important: Darwins naturalselection already provided the crucial argument that allowed the concept ofnormativity to be extended from the medical sphere to biology as a whole:

There are some thinkers whose horror of finalism leads them to rejecteven the Darwinian idea of selection by the environment and strugglefor existence because of both the term of selection, obviously of humanand technological import, and the idea of advantage which comes intothe explanation of the mechanism of natural selection. They point outthat most living beings are killed by the environment long before theinequalities which they can produce even have a chance to be of use tothem because it kills above all sprouts, embryos or the young. But asGeorges Teissier observed, the fact that many organisms die before theirinequalities serve them does not mean that the presentation of inequal-ities is biologically indifferent. This is precisely the one fact we ask tobe granted. There is no biological indifference, and consequently we canspeak of biological normativity.52

Let me now summarize the kind of meditation on biological individualitydeveloped by Canguilhem in his writings on the normal and the pathological.As a starting point, we have a thesis of the philosophy of medicine, borrowedfrom Goldstein: health, illness, and recovery should always be conceived ofas individual norms. But Canguilhem finally reinterpreted this medical idea

50 Canguilhem, Le Normal et le Pathologique (1951), reprinted in La Connaissance de laVie (above, n. 6), pp. 159160 (my translation).

51La question de la normalite dans lhistoire de la pensee biologique, in Canguilhem,

Ideologie et Rationalite dans les Sciences de la Vie (above, n. 6), p. 132 (my translation).English translation: Ideology and Rationality in the History of Life Sciences, trans. ArthurGoldhammer (above, n. 10).

52 Canguilhem, The Normal and the Pathological (above, n. 4), p. 129.

CANGUILHEMS PHILOSOPHY OF BIOLOGY 317

in the framework of a Darwinian view of life. For Canguilhem, Darwinismwas not an end in itself. In fact, the Darwinian detour led to a major thesisin the philosophy of biology: the thesis of the intimate relation betweenindividuality and vital value. A living being is always an absolute andirreducible system of reference.53 This thesis boils down to saying thataxiology is rooted in the conceptual foundations of the biological sciences.Although not stated in the 1943 essay, this thesis later came to be perfectlyexplicit: [B]iology must primarily hold the living being as a significantbeing, and individuality, not as an object, but as a character in the kingdomof values. Life . . . consists of organizing the milieu from a center of referencewhich cannot itself be referred to anything without losing its originalsignificance.54

I now come to Canguilhems ontological discussion of individuality. Thisrefers to a questioning of the nature of entities that can rightly be calledindividuals. There is no doubt that this problem was a major one forCanguilhems philosophy of biology, but I am not sure that he ever cameto a definitive solution. His strategy seems instead to have been to definestrong limits to the extension of the notion of individuality in natural philos-ophy. Although the problem pervades almost all of Canguilhems writingson the philosophy of biology or medicine, two major articles deserve partic-ular attention because of their systematic nature: his article on cell theory(1945), and his study on wholes and parts in biological thinking publishedtwo decades later (1966).55

First, what levels of organization did Canguilhem consider to be potentialcandidates in the debate over biological individuality? Canguilhem wouldcertainly be surprised to learn that many contemporary American philos-ophers of biology think that genes, proteins, cells, organisms, populations,species, and ecosystems are all individuals. Indeed, he would probably nothave accepted such terms, based on rather lax definition of the individualas a spatio-temporally bounded entity.56 In Canguilhems philosophy ofbiology, only three candidates for the status of individual are seriouslyconsidered: cells, organisms, and societies. However, Canguilhems final

53 Georges Canguilhem, Le vivant et son milieu (1947), reprinted in La Connaissance dela Vie (above, n. 6), p. 154 (my translation).

54 Ibid., p. 147 (my translation).55 Georges Canguilhem, La Theorie Cellulaire, reprinted in La Connaissance de la Vie

(above, n. 6), 4380; partial translation in Delaporte, A Vital Rationalist (above, n. 8), pp. 161177. Georges Canguilhem, Le tout et la partie dans la pensee biologique (1966), reprintedin Etudes dHistoire et de Philosophie des Sciences (above, n. 9), pp. 319333.

56 For a general discussion of this problem, see Jean Gayon, Critics and Criticisms of theModern Synthesis: The Viewpoint of a Philosopher, Evol. Biol., 24 (1990), 149.

318 JEAN GAYON

position was that societies are not individuals because they are not genuinewholes, whereas cells are individuals in spite of their being parts. I amconvinced that Canguilhems ontology of life was conceived in order tojustify these two assertions. Let us clarify this point.

Canguilhems final ontological reflection on living beings consists of twodoctrines. The first deals with the relation between wholes and parts in biolog-ical objects. Following an Aristotelian definition, Canguilhem argues that fora genuine whole its unity implies more than the sum of its parts.57 Further-more, he argues that in the history of biology the part-whole relation hadbeen represented by means of two metaphorical models. Up until the nine-teenth century, biologists had conceived of the relation between wholes andparts according to a technological model: parts (i.e., the organs) were similarto a series of differentiated tools that converged on the maintenance of thewhole (i.e., the organism).58 From this point of view, parts are not individuals,whereas the organism is an individual. The second model of the part-wholerelation in biological objects is a political one. It emerged in the nineteenthcentury. In a political model of the part-whole relation, parts are individuals.This is precisely the position of cell theory as it was formulated in the secondhalf of the nineteenth century. As Ernst Haeckel or Claude Bernard liked tosay, cells are virtually autonomous elements, like citizens in the Republic.59In such a metaphor, it can be said that cells are not instruments for the wholeorganism, but rather it is the organism that aims at the survival of the cells.60However, this analysis relates only to the role of the political metaphor inbiology. When Canguilhem dealt with genuine human societies, in the literalsense of the term, he refused to see them as genuine wholes,61 and hedid not accept the comparison of human societies and organisms. Whereasan organisms norm of life is furnished by the organism itself,62 socialnorms are to be invented and not observed.63 This doctrine can be relatedto Canguilhems early critical interest in Auguste Comtes sociological andpolitical thinking.64

Thus Canguilhems ontology of biological individuality is not an intuitivedoctrine. However, there is a second doctrine, which perhaps suggests a

57 Canguilhem, Le tout et la partie dans la pensee biologique (above, n. 55), p. 320.58 Ibid., pp. 320323.59 Ibid., p. 330.60 Ibid., p. 331.61 Georges Canguilhem, From the Social to the Vital, in The Normal and the Pathological

(above, n. 4), p. 256.62

On Organic Norms in Man, in The Normal and the Pathological (above, n. 4), p. 258.63 Ibid., p. 259.64 Georges Canguilhem, La Theorie de lOrdre et du Progre`s (Diplme detudes

superieures, the rough equivalent of a masters thesis, Universite de la Sorbonne, 1926).

CANGUILHEMS PHILOSOPHY OF BIOLOGY 319

solution to the problem. In his 1945 article on cell theory, Canguilhemsmain conclusion is that individuality does not describe a being but a relation.In a famous passage he argues that an individual is that which cannot bedivided without losing its own characteristics. It is a minimum of being. Butno being is a minimum. The individual implies its own relation to a widerbeing.65 But what is this wider being? This is the decisive step which isall the more difficult to understand because Canguilhem never systematicallydeveloped this idea. I think, however, that we can reconstruct his thinking.For a cell, the wider being is the organism, therefore a genuine whole. Foran organism, the wider being is not a strongly integrated whole; it is theexternal milieu, a complex web or organic and inorganic beings endowedwith a certain significance for the organism. Individuals are beings thathave needs, and therefore constitute an absolute and irreducible referencesystem in a given environment.66 Canguilhems ontology of individualitythus boils down to reasserting the primacy of axiology in the philosophy ofbiology. Or in other words, the ontology of life must be subordinated to theaxiology of life.

The third dimension of Canguilhems discussions of biological individu-ality is gnoseological. As with a number of twentieth-century philosophersinterested in biological sciences, Henri Bergsons reflections on the relationbetween life and knowledge were from the beginning an important sourceof stimulation for Canguilhem. Jean Piaget would be another example. Bothauthors were fascinated by Bergsons complex and ambivalent views on therelation between life and concept. On the one hand, Bergson emphasizedthe opposition between life and concept: concepts, just as tools, have arisenin the course of evolution as artificial constructs that constitute a mediationbetween the human organism and its environment.67 As such, conceptualknowledge is useful, but it is made of fictions (spatial schemata) that areunable to reveal the genuine nature of life (the understanding of whichrequires the intuitive knowledge of duration). In this perspective, conceptopposes life, and the philosophy of concept is of no use for a philosophyof life. On the other hand, in La Pense et le Mouvant, Bergson also empha-sized that because all organisms share the property of assimilating externalmatter for their nutrition and survival, they can be said to generalize.Canguilhem himself reports this idea: Bergson explains that it is not thecompleted animal, the macroscopic animal, which generalizes. Everything

65 Canguilhem, La Theorie Cellulaire (above, n. 55), p. 71 (my translation).66 Georges Canguilhem, Le Vivant et son Milieu (19461947), reprinted in La Connais-

sance de la Vie (above, n. 6), p. 154.67 See Henri Bergson, L Evolution Creatrice (Paris: Alcan, 1907), chap. 2.

320 JEAN GAYON

that is living, the cell, the tissue, generalizes. At any scale whatever, to livemeans to choose and to neglect.68 In this perspective, concept and life are notopposed: human conceptual knowledge is only a particular and spectacularextension of a characteristic belonging to all living beings.

This evocation of Bergson is prerequisite for understanding Canguilhemsstrange reflection upon knowledge and life, knowledge of life, andconcept of life, three expressions that constantly interfere in his discourse,sometimes serious ambiguities. Two very different stages must be distin-guished in the evolution of Canguilhems view on this subject. A first stageof the reflection can be found in the book The Knowledge of Life, a collectionof essays first published in 1952. The second stage is illustrated by two textswritten in the midsixties: A New Concept in Pathology: Error (publishedas an addition to the 1966 edition of The Normal and the Pathological), andConcept and Life, also first published in 1966. Michel Foucault consideredthis last paper as the most inventive text of Canguilhems.69 In The Knowledgeof Life, the question is: How can life be an object of (scientific) knowledge?Canguilhems answer is that there exists a major dissymmetry between lifesciences and other natural sciences. In the two papers published in 1966,the issue is quite different. These two papers were written immediately afterJacob, Lwoff, and Monod received a Nobel Prize in 1965; both papers empha-size that molecular biology has introduced a new general conception of life(as information). In this new context, Canguilhem argued that this new visionof life was a rehabilitation of some sort of Aristotelianism because molecularbiology admits there is a logos (or concept) inscribed in all individual livingthings. Hence the bold formula: Life is concept.70 In the two successivereflections on the relation between knowledge and life, the notion of indi-viduality plays a central role, but not the same one. I will now develop thispoint.

The book Knowledge of Life is made up of six essays (seven in the secondedition), which are probably the best known of Canguilhems writings forthe French public. The pieces were written between 1945 and 1951. Five areconstructed as historical narratives (on cell theory, the mechanistic conceptionof life, vitalism, experimentation in biology). But all illustrate the commonphilosophical thesis of the autonomy of biological sciences. Intelligencecannot apply to life without recognizing the originality of life. The knowl-edge of life must draw the idea of the living thing from the living thing

68 Georges Canguilhem, Le Concept et la Vie (1966), reprinted in Etudes dHistoire et dePhilosophie des Sciences (above, n. 9), p. 350.

69 Michel Foucault, La Vie: lExperience et la Science, Rev. Metaphys. et de Morale,special issue on Canguilhem, 90 (1985), 314. See also Paul Rabinow, Introduction: A VitalRationalist, A Vital Rationalist (above, n. 8), pp. 1122.

70Le Concept et la Vie (above, n. 68), p. 364 (A Vital Rationalist [above, n. 8], p. 318).

CANGUILHEMS PHILOSOPHY OF BIOLOGY 321

itself.71 In the introduction to the book, Canguilhem draws heavily on KurtGoldstein, who provides three quotations out of five. One of these quotationsprovides a precise formulation of the epistemological conception of biologythat dominates The Knowledge of Life: Biology has to deal with individualswho exist and tend to exist, i.e. who tend to fulfill their capacities in a givenenvironment.72 By saying this, Goldstein meant that knowledge about some-thing happening in an organism (e.g., a reflex) should be called biologicalonly if it is able to indicate the relation that process to the organism as awhole because it is only at that level that we know whether the establishedfacts (morphological, physiological, biochemical) are or are not significant.Canguilhem totally adhered to this kind of holism, and on its basis developeda brilliant plea for the autonomy of the life sciences. One of the essays inThe Knowledge of Life, the one devoted to the notion of milieu, concludeswith these words: A living thing cannot be reduced to a sum [carrefour]of influences. Hence the insufficiency of any biology which, by completesubmission to the spirit of the physicochemical sciences, completely refusesto take into account meaning.73 Such declarations brings us back, once again,to Canguilhems original axiological view of biological individuality.

In 1966, the knowledge and life question took a very different turn. Ina long paper entitled Concept and Life, Canguilhem traces the relationbetween theory of knowledge and the concept of life from Aristotle to thepresent. This is a rather complicated paper that embraces more or less theentire history of philosophy. Obviously, the author tried at least once toindicate the main links between his biological philosophy and a number oflandmarks in the history of philosophy in general. I will not enter into thedetail of this difficult but fascinating analysis. Actually, the outcome of thepaper is simple. Canguilhem argues that modern molecular genetics can bephilosophically interpreted as a return of Aristotelianism. For Aristotle, soulwas not only the nature but also the form of the living thing. Soul was atonce lifes reality [ousia] and definition [logos]. Thus, the concept of theliving thing was, in the end, the living thing itself.74 Of course, huge prob-lems in Aristotles thinking resulted from the two meanings he attributedto the soul: the soul was both a formal and eternal principle (containingthe common characters shared by all members of a species) and an activeprinciple explaining the generation and functioning of a concrete individual.

71 Kurt Goldstein, Remarques sur le Proble`me Epistemologique de la Biologie, (Congre`sInternational de Philosophie des Sciences (Paris: Hermann, 1951), quoted in GeorgesCanguilhem, introduction to La Connaissance de la Vie (above, n. 6), p. 13.

72 Ibid., p. 11.73 Canguilhem, Le Vivant et son Milieu (above, n. 66), p. 154.74 Canguilhem, Le Concept de la Vie (above, n. 68), p. 336 (A Vital Rationalist [above, n.

8], p. 303).

322 JEAN GAYON

Nevertheless, precisely because of this difficulty, Canguilhem argues thatAristotles conception of soul and life can be compared with the moderninformational conception of life in molecular biology. I would like here toquote a significant passage of the 1966 paper:

In changing the scale on which the characteristic phenomena of life which is to say, the structuration of matter and the regulation of func-tions, including the structuration function are studied, contemporarybiology has also adopted a new language. It has dropped the vocabularyof classical mechanics, physics and chemistry, all more or less directlybased on geometrical models, in favor of the vocabulary of linguistics andcommunications theory. Messages, information, programs, code, instruc-tions, decoding: these are the new concepts of life sciences. . . . When wesay that biological heredity is the communication of a certain kind ofinformation, we hark back in a way to the Aristotelian philosophy withwhich we began. . . . To say that heredity is the communication of informa-tion is, in a sense, to acknowledge that there is a logos inscribed, preservedand transmitted in living things. Life has always done without writing,long before writing even existed what humans have sought to do withengraving, writing and printing, namely, to transmit messages.75

Today, of course, we are very familiar with such declarations. Rememberthat this was written more than thirty years ago. What is interesting, however,is Canguilhems attitude toward the philosophical meaning that must be givento this kind of analogical statement. For Canguilhem, the analogy had to betaken in a strong and realistic way, not only as a metaphor. He made this pointexplicitly on the occasion of a reflection upon inborn errors of metabolism:

At the outset, the concept of hereditary biochemical error rested onthe ingenuity of a metaphor; today it is based on the solidity of ananalogy. . . . It would be very tempting to denounce an identification ofthought and nature, that error is characteristic of judgment, that nature canbe a witness but never a judge, and so on. Apparently everything happens,in effect, as if the biochemist and geneticist attributed their knowledge aschemist and geneticist to the elements of the hereditary patrimony, as ifenzymes were supposed to know or must know the reactions accordingto which chemistry analyzes their action and could, in certain instancesor at certain times, ignore one of them or misread the terms. But it mustnot be forgotten that information theory cannot be broken down, and thatit concerns knowledge itself as well as its objects, matter or life. In thissense to know is to be informed, to learn to decipher or decode. There is

75 Ibid., pp. 360, 362 (A Vital Rationalist [above, n. 8], pp. 316317).

CANGUILHEMS PHILOSOPHY OF BIOLOGY 323

no difference between the error of life and the error of thought, betweenthe errors of informing and informed information. The first furnishes thekey of the second. From the philosophical point of view it would be aquestion of a new kind of Aristotelianism, on the condition, of course, thatAristotelian psychobiology and the modern technology of transmissionnot be confused.76

In fact, Canguilhem was ambivalent toward this modern view of life. Onthe one hand, he was extremely suspicious of the possible medical and socialapplications of genetic conceptions. If muted genes are errors, we shouldfear the development of a genetic inquisition: At the beginning of thisdream we have the generous ambition to spare innocent and impotent livingbeings the atrocious burden of producing errors of life. At the end there are thegene police, clad in the geneticists science. . . . To dream of absolute remediesis often to dream of remedies which are worse than the ill.77 On the otherhand, Canguilhem thought that the analogy between biological and cognitiveinformation should be taken seriously by the philosopher of biology. I havealready quoted the bold formula Life is concept (or Life is meaning andconcept, or again The concept is in life).78 Such formulas must understoodas meaning that there is a deep analogy (an identity from a certain point ofview) between what humans call a concept in the domain of mental life, andgenetic information. Canguilhem is certainly not as explicit as one would likehim to be on this subject, but he provides some clues about the nature of theidentity between life and concept, a delicate expression that he sometimesborrows fro Hegel (but without borrowing Hegels philosophy).79 An encodedgenetic information can be called a logos (or in modern terms, a concept)insofar as it is a material, not abstract, principle of definition (determination ofthe essence of something): To define life as a meaning inscribed in matter isto acknowledge the existence of an objective a priori that is inherently mate-rial and not only formal.80 Such a conception of meaning or conceptinscribed in matter is of course counterintuitive. For philosophical commonsense, a concept will always be an abstract. This is not the case here. What is

76 Georges Canguilhem, A New Concept in Pathology: Error (19631966), in The Normaland the Pathological (above, n. 4), pp. 276278.

77 Ibid., pp. 280281.78 Canguilhem, Le Concept et la Vie (above, n. 68), p. 364 (A Vital Rationalist [above, n.

8], pp. 363364).79 Canguilhem, Le Concept et la Vie (above, n. 68), p. 347. In fact, all quotations from

Hegel in this paper seem to be inaccurate; most probably, Canguilhem took them from acommentator. I thank Andre Doz for his expertise on this point.

80 Canguilhem, Le Concept et la Vie (above, n. 68), p. 362 (A Vital Rationalist [above,n. 8], p. 317). The available English translation (a priori objective) is incorrect. The Frenchoriginal says: a priori objectif, hence in English: objective a priori.

324 JEAN GAYON

at stake is an informing (although material) principle that determines an indi-vidual as such. Some sentences express this idea in a vivid form, for example:Heredity is the modern name of substance.81 Or else, when Canguilhemcompares the informational capacity of DNA with Leibnizs definition of theindividual substance: The formula is quite close to Leibnizs definition ofindividual substance: Lex seriae suarum operationum, the law of the series inthe mathematical sense of the term, a series of operations. This almost formal(logical) definition of (biological) heredity can now be interpreted in the lightof the fundamental discovery of molecular biology, the structure of DNA, thekey constituent of chromosomes.82

We see then that Canguilhems reflection on concept and life broughthim close to gnoseology. Of course, there is nothing like a systematicallydeveloped Canguilhemian theory of knowledge (epistemology in thetechnical sense of English philosophy). But there is a Canguilhemian theoryof life as knowledge. And, as I have just tried to show, this theory is aphilosophical theory of biological individuality.

To sum up: I do not pretend to have provided an exhaustive interpretationof Canguilhems work. His contributions to the history of science shouldbe considered in their own right (especially his outstanding history of theconcept of reflex), as should his general conceptions concerning the relationbetween the history and philosophy of science (although they may have beentoo much commented by philosophers, to the detriment of the other aspects ofCanguilhems work). Nevertheless, I have tried to identify a theme that deeplymotivated both his philosophy of medicine and his philosophy of biology.The theme of individuality led Canguilhem to put forward three major ideas:axiology is inescapable in the life sciences; individuality is better understoodas a relation than as a being; the modern conception of heredity suggests anunprecedented proximity of life and concept. All three theses have an obviousrelation with what is known as Canguilhems vitalism. Most commentatorshave interpreted this vitalism as a factual or historical claim: vitalism, not as aspecial biological doctrine but as a fertile philosophy of nature in the historyof biology. I have tried to show that Canguilhem also provided a philosophicalelaboration of the idea of the uniqueness and originality of life. From thebeginning, this idea involved a subtle articulation of medical philosophy andbiological philosophy: Canguilhems original interest in medicine led him toa certain view of life; but conversely, Canguilhems medical philosophy hasits theoretical foundations in a biologically informed view of life. In recent

81 Canguilhem, A New Concept in Pathology (above, n. 4), p. 280.82 Canguilhem, Le Concept et la Vie (above, n. 68), pp. 359360 (A Vital Rationalist

[above, n. 8], pp. 315316).

CANGUILHEMS PHILOSOPHY OF BIOLOGY 325

times, such intimate and rigorous intertwining of the philosophy of medicineand the philosophy of biology have been quite rare.

Acknowledgments

The first version of this paper was given at Boston University, in the context ofa Boston Colloquium for Philosophy of Science, (Topics in French Philos-ophy of Science: a Franco-American Dialogue, May 67, 1996). BostonUniversity and the French Embassy in the United States are thanked fortheir support. Matthew Cobb is thanked for linguistic corrections. RichardM. Burian is warmly thanked for his fruitful observations on a prelimi-nary version of this paper. Andr Doz is thanked for his observations onCanguilhems quotations of Hegel.