Dossier Canguillhem and Bachelard

Max-Planck-Institut fr Wissenschaftsgeschichte
Max Planck Institute for the History of Science
2012
Preprint 434
Conference
Epistemology and History

From Bachelard and Canguilhem
to Todays History of Science
TABLE OF CONTENTS
PROSPECTUSES
Epistemology and History
pistmologie et histoire
13
Epistemologie und Geschichte
19
Program
25
INTRODUCTIONS
Erffnung/Ouverture
Henning Schmidgen
27
Historical Epistemology, Old and New

Jean-Franois Braunstein
33
Sur la rception de lpistmologie franaise en Allemagne

Peter Schttler
41
Rheinbergers Take on Historical Epistemology

Henning Schmidgen
47
CONTRIBUTIONS
Lpistmologie historique dans litinraire intellectuel de Georges Canguilhem
Camille Limoges
53
La dissection humaine. De lhistoire des commencements celle des origines

Franois Delaporte
67
Intervals, Possibilities and Encounters. The Trigger of a Ruptured History

in Bachelard
Monika Wulz
77
The Rupture and the Screw. The Structure of History According to Georges
Canguilhem and Helmuth Plessner
Thomas Ebke
91
Kurt Goldsteins Conception of Individuality

Stefanos Geroulanos & Todd Meyers
101
Georges Canguilhem. pistmologie historique et/ou histoire philosophique?

Claude Debru
111
On the Genealogy of Concepts and Experimental Practices. Rethinking Georges

Canguilhems Historical Epistemology
Pierre-Olivier Mthot
117
Le concept dobjectivit chez Canguilhem

Franoise Balibar
145
Scientific Philosophies in the Early 1930s and Gaston Bachelard

on Induction
Sandra Pravica
159
Canguilhem and Blumenberg. Working on the Evolution of Problems

from Both Sides of the Rhine
Cornelius Borck
171
Archologie des concepts et philosophie de la nature

Andrea Cavazzini
181
Histoire(s) de la vie de Canguilhem Foucault

Maria Muhle
187
The Role of Scientific Ideology in Canguilhems Historiography

Cristina Chimisso
197
Ist eine historische Epistemologie der Sozialwissenschaft mglich?

Frieder Otto Wolf
205
Life in Shakespeares Sonnets

Anselm Haverkamp
213
ABSTRACTS OF CONTRIBUTIONS
221
LIST OF AUTHORS AND PARTICIPANTS
229
EPISTEMOLOGY AND HISTORY

From Bachelard and Canguilhem to Todays History of Science

Over the past few years, historical epistemology has had quite a successful
international career. Starting with a week-long conference organized by Ian Hacking in
Toronto in 1993, historical epistemology was and continues to be used as a label for a
wide variety of projects and programs: from Hackings own discussion of styles of
scientific reasoning to Lorraine Dastons historicization of epistemological categories
and values, Arnold Davidsons investigations into the conceptual formation of new
kinds of knowledge and experience and the attempt undertaken by Peter Damerow et al.
to broaden the scope of Jean Piagets genetic epistemology by historical means.
Against this background, a series of recent conferences has been devoted to historical
epistemology in order to explore and discuss in more detail the investigations and
endeavors currently tied to that name. In this context, the key question has tended to be
What is historical epistemology? However, the corresponding answers have turned
out to be diverse. Either they remained general and referred to an interdisciplinary as
well as international project bridging the gap between history of science and philosophy
of science while combining analytic with continental traditions, or responses were very
concrete, referring to a discipline that aimed at introducing historical contingency into
seemingly fixed worlds. Other answers were restricted to programatic statements,
pointing to a general reference frame for all theoretically oriented studies in the history
of science.
Our conference has a different aim. Instead of addressing the what?, it will ask
how? As a consequence, it will also ask who?, when?, where?, and where to?
In other words, this meeting is meant to intervene into the ongoing debate by
dramatizing, as it were, the projects associated with the label historical epistemology.
Its main goal is to historicize and contextualize historical epistemology and, by the
same token, to create what we consider an important prerequisite for concrete and
critical updates.
The central reference point in this connection is the interesting synthesis of history
and epistemology that emerged in the French context during the 1930s and 1940s.
Generally speaking this synthesis resulted from a polemic demarcation of philosophy
with respect to its established versions and the programmatic integration of the natural
sciences. As representatives for this move we refer to the writings of Gaston Bachelard,
the works of Georges Canguilhem, and, at a later point, studies by Louis Althusser,
Michel Foucault and others. In fact, it is precisely in this context that the expression
historical epistemology was initially coined and widely used.
The main purpose of this conference is to start a dialogue between traditional and
recent projects at the interface between history and epistemology. On the one hand, this
shall be achieved by historical and systematic interventions concerning the problematic
field of historical epistemology. On the other hand we invite historical as well as
sociological studies about and with the classical works quoted in this context such
as The Formation of the Scientific Mind (1938), The Formation of the Reflex Concept in
the 17th and 18th Centuries (1955) or Philosophy and the Spontaneous Philosophy of the
Scientists (1967/1974). However, the criticism of historical epistemology and its
transformation into other, seemingly heterogeneous projects, e.g. political aesthetics, are
also of interest here. We believe that a broader and deeper understanding of this
background will enrich future discussions of the historical and epistemological
problems that claim central importance in the history of science today.
Context
I. One of the striking characteristics of the current debate around historical
epistemology is the almost general neglect of historically informed arguments. With
only a few exceptions, the origin of the term historical epistemology has not been of
interest. As a result, the discussion of the project designated by this label, and of its
emergence and evolution, has remained rather restricted. Hacking has related the term
in appropriate ways to Bachelard and the French tradition of science historiography. In
fact, Emile Meyerson had already used epistemology as a key term in his Identity and
Reality (orig. 1908). For Meyerson, it functioned as an equivalent to philosophy of
science. But neither Meyerson nor Bachelard spoke explicitly about historical
epistemology.
It was Dominique Lecourt who first introduced the term in a programmatic way in
1969. In the title of his book Gaston Bachelards Historical Epistemology (Engl. transl.
1975) Lecourt used it as a shorthand for the argument that epistemology, understood
here as the discipline which takes scientific knowledge as its object, essentially is
historical and that its main feature is to proceed historically. This argument was meant
to demarcate epistemology from philosophy, i.e. la philosophie des philosophes: In
opening the field of historical epistemology, he [Bachelard] uncovers [] what
philosophy is eager to cover up: the real historical conditions of the production of
scientific knowledge. With the opening of this field emerged, in the eyes of Lecourt, an
engaging reciprocity between epistemology and the history of science: [I]f
epistemology is historical, the History of the Sciences is necessarily epistemological.
When the term is applied to todays history of science, it seems once again unclear
whether we have to do with the description of the current situation or of a target state.
On the one side, there is a growing number of historians (and philosophers) of science
who argue for a turn to Historical Ontology, i.e. more attention to the material and
constructive aspects of scientific practice. On the other side (and independent thereof)
there is a part of the history of science community that seeks to professionalize itself as
an increasingly autonomous branch of the humanities, deriving its productivity to some
degree, it seems, from having withdrawn from the engaging reciprocity of scientific
and historical knowledge. As a result, both sides pose the question whether or not the
epistemic, i.e. the problem of genuinely scientific forms of cognition, shall remain a
central topic of studies in the history of science. It might also be the case, however, that
this problem constitutes the residuum of a profoundly internalist and insofar idealist
history of science.
II. In a more recent publication (2008), Lecourt has explained that he adopted the term
historical epistemology from Canguilhem. In fact, this expression shows up in
Canguilhems texts as early as 1963, not surprisingly in relation to Bachelard. In his
article The History of Science in the Epistemological Work of Gaston Bachelard,
Canguilhem used it in order to characterize the position of Bachelards epistemology
with respect to what counted, at the time, as traditional history of science: If history
consists in enumerating the variants in the successive editions of a Treatise, then
Bachelard is not a historian of science. If history of science consists in making visible
and intelligible the tedious and contradictory, the constantly renewed and rectified
construction [dification] of knowledge, then Bachelard is doing nothing else than the
history of science.
However, to Canguilhem (and this may come as a surprise) historical epistemology
was not identical with the project of making visible and intelligible the tedious and
contradictory, the constantly renewed and rectified construction of knowledge. In a
sense, the contrary was the case. In order to conceive of the emergence and evolution of
scientific knowledge not as a progressive development [droulement] but as an
adventure, Canguilhem insisted on needing historical history [histoire historique].
This is at least how he continues: Hence the interest that he [Bachelard] has for errors,
horrors, disorders: the margin of historical history that is not covered by historical
epistemology. Here, historical epistemology appears in perhaps rather unexpected
ways as a project akin to a nave historiography of science that hardly ever depicts
the dark spots of thinking. In contrast, histoire historique would be this central
periphery of error and erring that in Bachelards eyes as well as in Canguilhems was
required for understanding the emergence and evolution of scientific knowledge in
appropriate ways.
However, Canguilhem (and this might be the next surprise) did not argue for opening
historical epistemology without reserve to general history. At the same time, he also did
not attempt to pinpoint the close relation between epistemology and history in the sense
of a demarcation with respect to philosophy. He emphasized something different.
Canguilhem stressed the close relation of both sides to science: The historian and the
epistemologist have (or at least should have) one thing in common: the scientific culture
of today.
This was directed not merely at some general scientific culture, but instead referred
precisely to the recurrent character of the entire historico-epistemological project.
Whereas historians, according to Canguilhem, usually start with beginnings to come
from the announcements of the past to the facts of the present, epistemology urges one
to proceed in exactly the opposite way: The epistemologist starts with the current and
goes back to its beginnings so that only part of what yesterday presented itself as
science seems justified to a certain degree by the present.
Compared to a trend in recent history of science that brings back the grand narratives
of cultural history and the history of ideas, this plea for a history of science that would
take issue with contemporary science probably seems rather dated. But perhaps it is
exactly here that we find the key for a convincing continuation of historical
epistemology under the largely altered conditions of the 21st century. Today, the
working place of the historian of science very much resembles the work place of the
laboratory scientists. In both cases, screens and keyboards give access to an
unprecedented amount of data, powerful search engines and ever expanding networks of
communication. In this situation, the border between the archive, the museum and the
laboratory seems to become permeable, if not to vanish. This is a new and unexpected
perspective opening up in the contested field between epistemology and history.
The conference
New editions and translation projects underline the importance of Bachelards and
Canguilhems works but also of the writings by Althusser, Foucault and others in the
current debate. At the same time, a whole series of new studies has been devoted to the
partly historical, partly sociological investigation of the discourses concerning history
and epistemology. These studies have shown that the historical epistemology of the
1930s and 1940s was situated between the Sorbonne and the Rue dUlm but can also be
traced back to the 19th century: to Auguste Comte and Ernst Mach, for example. Against
this background, the aim of this conference is to stimulate an exchange and discussion
between traditional and current projects at the intersection between history and
epistemology.
Systematic and/or historical reference points may be the following:
1) The history of concepts

a) Given the growing importance of visualization and mathematization of scientific
knowledge, does it remain appropriate to conceive of concepts as the central
objects of the history of science, as Canguilhem, following Cavaills,
suggested?
b) If the first use of the expression historical epistemology can be traced back to
the early 1960s, how far back does the history of the concept reach: to Comte,
Hegel or even farther?
10

2) The historicity of objects and spaces of knowledge

a) Bachelard, with his notion of phnomno-technique, seems to have suggested a
central topic for social constructivism and recent science studies. However, the
same constructivism turned emphatically away from the idea of a rupture and/or
break between scientific knowledge and everyday knowledge, between the space
of science and the spaces of literature and the arts. Does this tendency towards
unification remain of critical interest?
b) Historical epistemology seems to have emerged from a rather circumscribed
space of knowledge inside Paris. However, this space has also defined itself by
drawing connections to similar projects, such as the criticism of knowledge in
Nietzsche and the psychoanalysis of scientific knowledge (Freud, Jung, Lacan).
More generally, how can one describe the relation of historical epistemology to
other, seemingly more distant areas of knowledge, such as the logical positivism
(Reichenbach), general biology (Goldstein), and/or metaphorology
(Blumenberg)?
3) The experimental character of scientific practice and the problem of recurrence
a) and b) Hackings notion that experiments have a life of their own was and
still is an important touchstone for much current philosophy and history of
science. Only now do we begin to see, however, that this topic was also of
crucial interest for Bachelard and Canguilhem. Could this be one of the deeper
reasons for the renewed interest in historical epistemology? Or is it exactly this
convergence that is a sign of datedness, given the fact that experiments hardly
exist any more in the techno-practice of todays science?
Literature
Canguilhem, Georges (1963). Lhistoire des Sciences dans loeuvre pistmologique
de Gaston Bachelard, Annales de lUniversit de Paris 1 (1963): 24-39,
reprinted in: Etudes dhistoire et de philosophie des sciences, Paris: Vrin,
1968, pp.173-186.
Daston, Lorraine (1994). Historical Epistemology, In: J. Chandler, A. I. Davidson,
and H. Harootunian, (eds.), Questions of Evidence: Proof, Practice, and
Persuasion across the Disciplines, Chiocago: University of Chicago Press,
pp.282-289.
Davidson, Arnold I. (2001). The Emergence of Sexuality: Historical Epistemology and
the Formation of Concepts, Cambridge, Mass.: Harvard University Press.
Gingras, Yves (2010). Naming without necessity: on the genealogy and uses of the
label historical epistemology, CIRST Note de recherche, 2010-01, pp.
1-17.
Hacking, Ian (2002). Historical Ontology, Cambridge, Mass.: Harvard University Press.
11

Lecourt, Dominique (1969). Lpistmologie historique de Gaston Bachelard, Avantpropos de Georges Canguilhem, Paris: Vrin.
Lecourt, Dominique (1972). Pour une critique de l'pistmologie (Bachelard,
Canguilhem, Foucault), Paris: Maspero.
Lecourt, Dominique (1975). Marxism and Epistemology. Bachelard, Canguilhem and
Foucault, translated by Ben Brewster, London: NLB.
Lecourt, Dominique (2008). Georges Canguilhem, Paris: PUF (Que sais-je?; 3722).
Rheinberger, Hans-Jrg (2010). On Historicizing Epistemology. An Essay, Stanford:
Stanford University Press.
Sturm, Thomas & Feest, Uljana (eds.), What (Good) is Historical Epistemology?,
Berlin: Max Planck Institute for the History of Science, 2009 [Preprints;
386].
Erkenntnis 75/3 (2011), Special Issue.
12

PISTMOLOGIE ET HISTOIRE
De Bachelard et Canguilhem lhistoire des sciences aujourdhui
L pistmologie historique a accompli, depuis vingt ans environ, une carrire

internationale impressionnante. Depuis une rencontre organise en 1993 sous ce titre par
Ian Hacking Toronto, le terme dpistmologie historique est devenu lenseigne dun
grand nombre de projets et de programmes: du traitement des styles de pense
scientifique par Hacking lui-mme lhistorisation des catgories pistmologiques
par Lorraine Daston jusquaux tudes conceptuelles dArnold Davidson sur la formation
des nouvelles formes de savoir et dexprience ou la tentative entreprise par Peter
Damerow et dautres de renouveler et dlargir lpistmologie gntique de Piaget
travers lhistoire des sciences.
Sur cet arrire plan une srie de colloques sest penche sur lpistmologie
historique afin de discuter plus amplement les travaux et les projets associs ce titre. Il
sagissait alors essentiellement de rpondre la question: Quest-ce que
lpistmologie historique? Les rponses furent cependant fort diverses: elles visaient
de manire gnrale une approche aussi bien internationale quinterdisciplinaire, afin de
jeter un pont entre lhistoire des sciences et la philosophie des sciences et de relier les
traditions continentales et analytiques; puis, de manire plus spcifique, une
discipline qui voudrait introduire la contingence historique dans la comprhension
de mondes apparemment stabiliss; et enfin, de manire programmatique, un cadre de
rfrence pour tout travail thorique et rflchi en histoire des sciences.
Le colloque propos vise un objectif diffrent. Il ne pose plus la question quest-ce
que lpistmologie historique? , mais comment est-elle possible? Donc aussi:
qui la pratique? , et quand? , et o? . Ce colloque voudrait donc intervenir en
dramatisant dune certaine manire le projet li lpistmologie historique en
contribuant son historisation et sa contextualisation, afin de crer une des conditions
essentielles pour des actualisations la fois concrtes et critiques.
Au centre nous placerons la relation entre pistmologie et histoire telle quelle
naquit depuis les annes 30 et 40, notamment dans le contexte franais, par une
dmarcation polmique vis--vis de la philosophie traditionnelle et dans le cadre dun
rapprochement avec les sciences de la nature. En tmoignent de manire exemplaire les
crits de Gaston Bachelard, les travaux de Georges Canguilhem, et par la suite
galement les tudes de Louis Althusser, Michel Foucault et dautres encore En fait,
cest dans ce contexte quil fut pour la premire fois littralement question
d pistmologie historique .
Notre colloque veut ouvrir un dialogue concret entre les projets traditionnels et ceux
plus rcents, situs linterface entre pistmologie et histoire. Cela pourrait se faire
13
la fois par des interventions historiques et systmatiques sur la problmatique de

lpistmologie historique et par la prsentation de travaux dhistoire et ventuellement
de sociologie des sciences sur et avec des uvres majeures lies au contexte
esquiss: de La Formation des lesprit scientifique (1938) La Formation du concept
de rflexe (1955) jusqu Philosophie et philosophie spontane des savants
(1967/1974). Mais la critique de lpistmologie historique ou son transfert vers de
nouveaux projets, apparemment trs diffrents, comme lesthtique et la politique,
peuvent galement tre dun grand intrt. En tous cas, cest seulement sur larrirefond dun tel dbat approfondi quil nous semble possible de cerner les problmes aussi
bien historiques qupistmologiques pouvant prtendre aujourdhui encore avoir de
lintrt et un enjeu pour lhistoire des sciences.
Contexte
I. Une des caractristiques du dbat actuel sur lpistmologie historique est labsence
presque totale darguments historiquement informs. Ainsi ne sest-on presque pas
intress, sauf exception, lorigine de la notion d pistmologie historique ellemme. Par consquent le dbat sur le projet du mme nom, son histoire et son
volution, est souvent rest bloqu. Certes, Hacking a mis en relation cette notion avec
Bachelard et la tradition franaise de lhistoire des sciences. Le terme pistmologie
apparat en fait pour la premire fois chez mile Meyerson, dans Identit et ralit
(1908). Mais ni Meyerson ni Bachelard ne parlent dpistmologie historique.
Ce nest quen 1969 que Dominique Lecourt lana pour la premire fois le terme de
manire dmonstrative. En effet, dans le titre de son livre Lpistmologie historique de
Gaston Bachelard il utilisa cette notion pour souligner que lpistmologie entendue
comme la discipline qui prend la connaissance scientifique pour objet est
historique , que son essence est dtre historique . Cette thse voulait tre entendue
comme une dmarcation par rapport la philosophie, ou du moins la philosophie des
philosophes: [E]n ouvrant le champ de lpistmologie historique, il [Bachelard]
dcouvre [] ce que la philosophie sacharne recouvrir: les conditions relles
historiques de la production des connaissances scientifiques. En ouvrant ce champ,
aux yeux de Lecourt, naquit une rciprocit engageante entre pistmologie et
histoire des sciences: [S]i lpistmologie est historique, lhistoire des sciences est
ncessairement pistmologique.
Si cependant on applique cette phrase la situation prsente de lhistoire des
sciences, il reste aussi savoir sil sagit dun tat de fait ou dun tat atteindre. Car
on voit, dune part, un nombre croissant de chercheurs plaidant pour une ontologie
historique et insistant plutt sur la matrialit et le caractre construit de toute pratique
scientifique; et dautre part, une partie de lhistoire des sciences se professionnalise de
plus en plus en tant que science humaine part entire, qui tirerait sa productivit
justement du fait de stre arrache la rciprocit engageante de la connaissance
scientifique et de lhistoire. Ainsi se pose de plusieurs cts la question de savoir dans
quelle mesure l pistmique , cest--dire la problmatique de la connaissance
14
PISTEMOLOGIE ET HISTOIRE
scientifique proprement dite, reste un thme central du travail de lhistoire des sciences.
Ou bien sagit-il seulement du rsidu dune historiographie des sciences profondment
internaliste et mme idaliste?
II. Dans une publication rcente, Dominique Lecourt a concd quil avait repris
lexpression pistmologie historique Georges Canguilhem. Cette notion apparat
en effet chez Canguilhem ds 1963, et cela galement propos de Bachelard. Dans son
article Lhistoire des sciences dans luvre pistmologique de Gaston Bachelard ,
Canguilhem lutilise pour dsigner la position de lpistmologie bachelardienne par
rapport ce qui tait considr lpoque comme une histoire des sciences
traditionnelle: Si lhistoire des sciences consiste recenser des variantes dans les
ditions successives dun Trait, Bachelard nest pas un historien des sciences. Si
lhistoire des sciences consiste rendre sensible et intelligible la fois ldification
difficile, contrarie, reprise et rectifie, du savoir, alors lpistmologie de Bachelard est
une histoire des sciences toujours en acte.
Pourtant, lpistmologie historique ntait pas, pour Canguilhem, identique avec le
projet de rendre sensible et intelligible la fois ldification difficile, contrarie,
reprise et rectifie, du savoir . Dune certaine manire elle en tait mme loppos.
Pour pouvoir concevoir la naissance et lvolution du savoir scientifique non pas
comme un droulement, mais comme une aventure, il fallait ses yeux justement une
histoire historique . Cest en tous cas ainsi quil poursuivit: Do lintrt quil
[Bachelard] porte aux erreurs, aux horreurs, aux dsordres, tout ce qui reprsente la
frange dhistoire historique non recouverte par lpistmologie historique.
Lpistmologie historique apparat ainsi, de manire peut-tre surprenante, aux
cots de cette histoire des sciences trop candide qui ne restitue presque jamais les
obscurits de la pense (Bachelard). Par contraste, lhistoire historique pourrait tre
perue comme cette priphrie centrale de lerreur et de lerrance qui, pour Canguilhem
comme pour Bachelard, formait la condition pralable toute comprhension de la
naissance et du dveloppement du savoir scientifique. Cependant, Canguilhem ne
plaidait pas, et ceci pourrait constituer une autre surprise, pour une ouverture sans
rserve de lpistmologie historique par rapport lhistoire gnrale. De mme quil ne
voulut pas souligner le rapport troit entre pistmologie et histoire dans le sens dune
dmarcation par rapport la philosophie. Il prfra plutt insister sur un autre aspect en
soulignant la proximit des deux par rapport la science: [L]historien et
lpistmologue ont en commun (ou du moins devraient avoir en commun) la culture
scientifique daujourdhui.
Ceci ne devait pas se comprendre dans le sens dune formation scientifique gnrale,
mais portait sur le caractre rcurrent du travail historico-pistmologique proprement
dit. Tandis que lhistorien, selon Canguilhem, a gnralement tendance partir des
origines pour arriver, travers les annonces du pass, jusquau prsent de la science,
lpistmologie tend, au contraire, poursuivre le parcours inverse: Lpistmologue
procde de lactuel vers ses commencements en sorte quune partie seulement de ce qui
se donnait hier pour science, se trouve quelque degr fonde par le prsent.
15
Par rapport la tendance actuelle en histoire des sciences dun retour la grande
narration et au vaste panorama intellectuel et culturel, ce plaidoyer pour une histoire des
sciences oriente vers le prsent de la science, pourrait paratre particulirement dsuet.
Pourtant cest ici peut-tre que lon peut trouver la cl pour une reprise plausible de
lpistmologie historique dans des conditions largement transformes. Le lieu de
travail de lhistorien des sciences ressemble aujourdhui plus que jamais celui du
chercheur dans son laboratoire: lcran et le clavier lui donnent accs un nombre
inimaginable de donnes, de moteurs de recherches puissants et de rseaux de
communication de plus en plus tendus. Dans cette situation, les frontires entre
larchive, le muse et le laboratoire deviennent permables, et cest ainsi que dautres
travaux novateurs pourraient natre la croise de lpistmologie et de lhistoire.
Le colloque
De nouveaux projets ddition et de traduction ont rcemment soulign limportance et
limpact des uvres de Bachelard et de Canguilhem, mais galement dAlthusser et de
Foucault, dans les dbats actuels. En mme temps un certain nombre dtudes rcentes
se sont consacres lanalyse historique ou sociologique des discours dans le champ de
lpistmologie et de lhistoire qui depuis les annes 30 et 40 taient si bien ancrs entre
la Sorbonne et la Rue dUlm et que lon pourrait faire remonter jusquau 19e sicle:
jusqu Auguste Comte, mais galement jusqu Ernst Mach. Sur cet arrire-plan, notre
colloque voudrait lancer un dialogue entre les projets traditionnels et les recherches les
plus rcentes, la croise entre pistmologie et histoire.
Quelques questions systmatiques et historiques nous semblent particulirement
intressantes:
1. Lhistoire des concepts scientifiques

a) Vu limportance croissante de la visualisation et de la mathmatisation du savoir
scientifique, est-il encore adquat de penser les concepts comme constituant
lobjet principal de lhistoire des sciences, comme le faisait Canguilhem,
notamment la suite de Cavaills?
b) Si la premire utilisation de lexpression pistmologie historique remonte aux
annes 1950, jusquo peut-on faire remonter lhistoire du concept: jusqu
Comte ou encore au-del?
2. Lhistoricit des objets et des espaces de savoir

a) En parlant de phnomnotechnique , Bachelard semble avoir lanc un thme
central du constructivisme contemporain. Pourtant le mme constructivisme sest
tourn de faon vhmente contre lide de rupture entre savoir scientifique et
16
PISTEMOLOGIE ET HISTOIRE
sens commun, entre lespace de la science et celui de la littrature et de lart.

Cette tendance lunification est-elle encore pertinente aujourdhui?
b) premire vue, lpistmologie historique semble tre ne dans un espace
parisien assez limit. Mais cet espace na-t-il pas galement t dfini par des
projets comparables comme la critique de la connaissance de Nietzsche ou la
psychanalyse du savoir scientifique (Freud, Jung, Lacan)? Et comment peut-on
dcrire le rapport de lpistmologie historique dautres domaines de savoir,
premire vue loigns, comme, par exemple, le positivisme logique
(Reichenbach), la biologie gnrale (Goldstein) ou la mtaphorologie de Hans
Blumenberg?
3. Le caractre exprimental des pratiques scientifiques et le problme de la rcurrence

a) et b) Depuis les annes 80, la philosophie et lhistoire des sciences ont t
stimuls de faon dcisive par la thse de Hacking selon laquelle non seulement
le travail exprimental, mais aussi les expriences elles-mmes ont une vie
autonome . Mais ce nest quaujourdhui quil savre que ce thme avait dj
jou un rle central chez Bachelard et Canguilhem. Par consquent peut-on y
voir la cause vritable dun renouveau dintrt pour lpistmologie historique?
Ou cette rcurrence est-elle plutt lindice dune obsolescence, puisque, dans
une pratique scientifique devenue de plus en plus technique, lexprimentation
semble presque avoir perdu sa place?
Bibliographie
Balibar, Etienne et Pierre Macherey (1970). pistmologie , in Encyclopaedia
Universalis, tome 6: lastomres Film d'art, Paris: Encyclopaedia
Universalis France, 1970, pp.370-373.
Canguilhem, Georges (1963). Lhistoire des Sciences dans luvre pistmologique de
Gaston Bachelard, Annales de lUniversit de Paris (1963): pp. 24-39,
repris dans: tudes dhistoire et de philosophie des sciences, Paris: Vrin,
1968, pp. 173-186.
Daston, Lorraine (1994). Historical Epistemology, In: J. Chandler, A.I. Davidson, et
H. Harootunian (ds.), Questions of Evidence: Proof, Practice, and
Persuasion across the Disciplines, Chicago: University of Chicago Press,
pp. 282-289.
label historical epistemology, CIRST Note de recherche, 2010-01, pp.
1-17.
17
Lecourt, Dominique (2008). Georges Canguilhem, Paris: PUF (Que sais-je? ; 3722).
Rheinberger, Hans-Jrg (2010). On Historicizing Epistemology. An Essay, Stanford:
Stanford University Press.
Sturm, Thomas & Feest, Uljana (2009) (ds.), What (Good) is Historical
Epistemology?, Berlin: Max Planck Institute for the History of Science
[Preprints ; 386].
18
EPISTEMOLOGIE UND GESCHICHTE

Von Bachelard und Canguilhem zur Wissenschaftsgeschichte von heute
Die Historische Epistemologie hat in den vergangenen zwei Jahrzehnten international

bemerkenswert Karriere gemacht. Seit einer Studientagung, die Ian Hacking 1993 unter
diesem Titel organisierte, ist Historische Epistemologie als Bezeichnung fr eine
Vielzahl von Projekten und Programmatiken verwendet worden: von Hackings eigener
Auseinandersetzung mit den styles of scientific reasoning ber Lorraine Dastons
Historisierung erkenntnistheoretischer Kategorien und Werte bis hin zu Arnold
Davidsons begriffszentrierten Studien zur Herausbildung von neuen Wissens- und
Erfahrungsformen oder etwa dem von Peter Damerow und anderen unternommenen
Versuch, die Genetische Epistemologie Jean Piagets mit den Mitteln der
Wissenschaftsgeschichte zu erweitern und zu erneuern.
Vor diesem Hintergrund hat sich in jngster Zeit eine Reihe von Konferenzen auf die
Historische Epistemologie bezogen, um die gegenwrtig mit diesem Titel assoziierten
Untersuchungen und Unternehmungen genauer zu diskutieren. Dabei stand zumeist die
Frage Was ist Historische Epistemologie? im Vordergrund. Die Antworten, die auf sie
gegeben wurden, fielen allerdings durchaus unterschiedlich aus allgemein, im Sinne
eines ebenso interdisziplinren wie internationalen Ansatzes, der eine Brcke zwischen
Wissenschaftsgeschichte und Wissenschaftsphilosophie schlgt und dabei kontinentale
mit analytischen Traditionen verknpft; deutlich spezifischer, im Sinne einer
Disziplin, die darauf abzielt, historische Kontingenz in das Verstndnis scheinbar
festgefgter Welten einzufhren; und schlielich programmatisch, als Bezugsrahmen
fr jede theoretisch reflektierte Form wissenschaftshistorischer Arbeit.
Die von uns geplante Konferenz zielt in eine andere Richtung. Sie fragt nicht lnger
nach dem Was. Ihr geht es um das Wie, also auch um das Wer, das Wann, das Wo und
das Wohin der Historischen Epistemologie. Diese Konferenz mchte in die laufende
Debatte also dadurch eingreifen, dass sie das Projekt, das mit diesem Titel bezeichnet
ist, in gewisser Weise dramatisiert. Ihre Absicht ist es, zur Historisierung und
Kontextualisierung der Historischen Epistemologie beizutragen, um damit eine wichtige
Voraussetzung fr ebenso konkrete wie kritische Aktualisierungen zu schaffen.
In den Mittelpunkt wird dabei jene Relationierung von Epistemologie und
Geschichte gerckt, die seit den 1930er und 1940er Jahren vor allem im
franzsischsprachigen Kontext aus einer polemischen Abgrenzung zur etablierten
Philosophie und einer programmatischen Annherung an die Naturwissenschaften
entstanden ist. Als exemplarisch dafr werden hier die Schriften von Gaston Bachelard
verstanden, die Arbeiten von Georges Canguilhem sowie, im Anschluss daran, die
Studien von Louis Althusser, Michel Foucault und anderen. Tatschlich ist es dieser
19
Zusammenhang, in dem zuerst buchstblich von Historischer Epistemologie die Rede

gewesen ist.
Ziel der Konferenz ist es, einen konkreten Dialog zwischen traditionellen und
aktuellen Projekten im Schnittbereich von Epistemologie und Geschichte zu erffnen.
Dies soll einerseits durch historische und systematische Interventionen zur Problematik
der Historischen Epistemologie geschehen, und andererseits durch die Vorstellung von
wissenschaftshistorischen und -soziologischen Arbeiten ber und mit den im
skizzierten Zusammenhang einschlgigen Werken von der Bildung des
wissenschaftlichen Geistes (1938) ber die Herausbildung des Reflexbegriffs (1955) bis
hin zu Philosophie und spontane Philosophie der Wissenschaftler (1967/74). Auch die
Kritik der Historischen Epistemologie sowie ihre berfhrung in neuartige, scheinbar
gnzlich anders geartete Projekte etwa die politische sthetik sind von Interesse.
Erst vor dem Hintergrund einer auf dieser Weise vertieften Diskussion scheint es uns
mglich zu sein, jene historischen und zugleich epistemologischen Probleme zu
umreien, die in der Wissenschaftsgeschichte von heute weiterhin Interesse und
Relevanz beanspruchen knnen.
Der Kontext
I. Ein Merkmal der gegenwrtigen Debatte um die Historische Epistemologie ist ihr
weitgehender Verzicht auf historisch informierte Argumente. So hat es, von Ausnahmen
abgesehen, bislang kein ernsthaftes Interesse fr die Herkunft des Ausdrucks
Historische Epistemologie gegeben. Dementsprechend kurzschlssig ist oftmals die
Auseinandersetzung mit dem dadurch bezeichneten Projekt, seiner Entstehung und
Entwicklung, geblieben. Hacking hat den Ausdruck Historische Epistemologie zwar
in zutreffender Weise mit Bachelard und der franzsischen Tradition der
Wissenschaftsgeschichtsschreibung in Verbindung gebracht. In der Tat fungiert
Epistemologie schon bei Emile Meyerson, in Identitt und Wirklichkeit (frz. 1908),
als Synonym fr Wissenschaftsphilosophie. Doch weder bei Meyerson noch bei
Bachelard ist die Rede von Historischer Epistemologie.
Es ist Dominique Lecourt, der den Ausdruck 1969 erstmals an prominenter Stelle
verwendet hat. Im Titel seines Buches Lpistmologie historique de Gaston Bachelard
setzt Lecourt diesen Terminus als Platzhalter fr die These ein, dass die
Epistemologie hier verstanden als die Disziplin, die die wissenschaftliche Erkenntnis
zum Gegenstand nimmt auf genuine Weise historisch ist, dass ihr wesentliches
Charakteristikum also darin besteht, ber eine Geschichte zu verfgen. Diese These
wollte im Sinne einer Abgrenzung gegenber der Philosophie, zumindest der
philosophie des philosophes verstanden sein: Indem Bachelard das Feld der
historischen Epistemologie erffnet, entdeckt er [...], was die Philosophie verdeckt: die
historisch-reellen Bedingungen der Produktion wissenschaftlicher Erkenntnisse. Mit
der Erffnung dieses Feldes entstand in den Augen von Lecourt zugleich eine
mitreiende Wechselseitigkeit von Epistemologie und Geschichte der Wissen-
20
schaften: Wenn die Epistemologie historisch ist, ist die Wissenschaftsgeschichte

notwendigerweise epistemologisch.
Wendet man diesen Satz auf die heutige Situation der Wissenschaftsgeschichte an,
erscheint es allerdings erneut keineswegs als ausgemacht, ob es sich um die
Beschreibung eines Ist- oder eines Soll-Zustands handelt. Denn zum einen mehren sich
unter Wissenschaftsforschern (und -philosophen) die Stimmen fr die Hinwendung zu
einer Historischen Ontologie, die strker als bisher auf das Materielle und das
Gemachte aller Wissenschaftspraxis abstellen wrde. Zum anderen (und dessen
ungeachtet) professionalisiert sich ein Teil der Wissenschaftsgeschichte als zunehmend
eigenstndige Geisteswissenschaft, die ihre Produktivitt auch daraus zu beziehen
scheint, dass sie sich der mitreienden Wechselseitigkeit von naturwissenschaftlicher
Erkenntnis und Geschichte entzieht. Von zwei Seiten steht somit die Frage im Raum, ob
das Epistemische, d.h. die Problematik der genuin wissenschaftlichen Erkenntnis, ein
zentrales Thema der wissenschaftshistorischen Arbeit bleibt. Oder handelt es sich nur
um das Residuum einer zutiefst internalistischen und insofern auch idealistischen
Wissenschaftsgeschichtsschreibung?
II. In einer jngeren Verffentlichung (2008) hat Lecourt eingerumt, den Ausdruck
Historische Epistemologie von Canguilhem bernommen zu haben. Tatschlich
taucht der Ausdruck bei diesem schon 1963 auf, allerdings ebenfalls in Bezug auf
Bachelard. In dem Aufsatz ber Die Wissenschaftsgeschichte im epistemologischen
Werk von Gaston Bachelard setzt Canguilhem ihn ein, um die Position der
Bachelardschen Epistemologie gegenber dem zu bestimmen, was seinerzeit als
traditionelle Wissenschaftsgeschichte galt: Wenn Wissenschaftsgeschichte darin
besteht, die Varianten in den aufeinanderfolgenden Ausgaben eines Traktats
aufzuzhlen,
ist
Bachelard
kein
Wissenschaftshistoriker.
Wenn
die
Wissenschaftsgeschichte darin besteht, den schwierigen und widersprchlichen, stndig
erneuerten und berichtigten Aufbau des Wissens sichtbar und zugleich einsehbar zu
machen dann ist Bachelards Epistemologie nichts anderes als Wissenschaftsgeschichte.
Nun war fr Canguilhem, und das mag berraschen, die Historische Epistemologie
keineswegs mit dem Vorhaben identisch, den schwierigen und widersprchlichen,
stndig erneuerten und berichtigten Aufbau des Wissens sichtbar und zugleich
einsehbar zu machen. In gewissem Sinne war das Gegenteil der Fall. Um die
Entstehung und Entwicklung wissenschaftlichen Wissens nicht als einen Ablauf,
sondern als Abenteuer begreifen zu knnen, bedurfte es Canguilhem zufolge gerade der
historischen Geschichte. So jedenfalls schliet er an: Daher das Interesse, das er
[Bachelard] den Irrtmern, den Greueln, den Unordnungen entgegenbringt: jener
Randzone des historischen Geschehens (histoire historique), die von der historischen
Epistemologie (pistmologie historique) nicht abgedeckt wird.
Damit taucht die Historische Epistemologie in vielleicht unerwarteter Weise an der
Seite jener blauugigen Wissenschaftsgeschichte auf, die die dunklen Stellen des
Denkens kaum je darstellt. Im Kontrast dazu wre die histoire historique als jene
zentrale Peripherie des Irrens und des Irrsals zu begreifen, die fr Canguilhem ebenso
21
wie fr Bachelard unentbehrlich war, um die Entstehung und Entwicklung des

wissenschaftlichen Wissens zu begreifen.
Canguilhem pldierte allerdings, und das ist vielleicht die nchste berraschung,
keineswegs fr eine rckhaltlose ffnung der Historischen Epistemologie gegenber
der Allgemeinen Geschichte. Er versuchte auch nicht, das enge Verhltnis von
Epistemologie und Geschichte im Sinne einer Abgrenzung gegenber der Philosophie
zu pointieren, sondern setzte einen ganz anderen Akzent. Canguilhem hob die
beiderseitige Nhe zur Wissenschaft hervor: Der Historiker und der Epistemologe
haben [...] eines gemeinsam (oder sollten es zumindest gemeinsam haben): die
wissenschaftliche Kultur von heute.
Das war nicht auf eine allgemeine (natur-)wissenschaftliche Bildung gemnzt,
sondern bezog sich auf den rekurrenten Charakter der historisch-epistemologischen
Ttigkeit insgesamt. Whrend der Historiker laut Canguilhem gewhnlicherweise die
Tendenz hat, von den Anfngen auszugehen, um von den Ankndigungen der
Vergangenheit auf die Gegenwart der Wissenschaft zu kommen, drngt die
Epistemologie dazu, genau umgekehrt zu verfahren: Der Epistemologe geht vom
Aktuellen aus und auf seine Anfnge zurck, so dass nur ein Teil dessen, was sich
gestern als Wissenschaft ausgab, bis zu einem gewissen Grad als durch die Gegenwart
begrndet erscheint.
Angesichts einer in der heutigen Wissenschaftsgeschichte neu erwachten Tendenz
zur Groen Erzhlung, zum geistes- und kulturgeschichtlichen Panorama erscheint
dieses Pldoyer fr eine an der gegenwrtigen Wissenschaft orientierte
Wissenschaftsgeschichte vielleicht als besonders antiquiert. Mglicherweise liegt aber
gerade hier ein Schlssel fr die plausible Fortsetzung der Historischen Epistemologie
unter den weitgehend vernderten Bedingungen des frhen 21. Jahrhunderts. Der
Arbeitsplatz des Wissenschaftshistorikers hnelt heute mehr denn je dem des
Laborwissenschaftlers: Bildschirm und Tastatur erffnen einen Zugriff auf ungeahnten
Datenmengen, leistungsfhige Suchmaschinen und sich immer weiter ausbreitende
Kommunikationsnetze. In dieser Situation scheinen die Grenzen zwischen Archiv,
Museum und Labor durchlssig zu werden, und eben darin knnte die Chance fr
weitere innovative Arbeiten im Spannungsfeld von Epistemologie und Geschichte
liegen.
Die Konferenz
Neue Editionsprojekte und bersetzungen haben die aktuelle Bedeutung unterstrichen,
die den Werken von Bachelard und Canguilhem, aber auch von Althusser, Foucault und
anderen in der heutigen Diskussion zukommen. Zugleich hat sich eine Reihe von
rezenten Studien der teils historischen, teils soziologischen Aufarbeitung jener Diskurse
im Spannungsfeld von Epistemologie und Geschichte gewidmet, die seit den 1930er
und 1940er Jahren besonders zwischen Sorbonne und der Rue dUlm verankert waren,
die in ihrer weit verzweigten Entstehung aber bis ins 19. Jahrhundert zurckverfolgt
werden knnen: nicht nur auf Auguste Comte, sondern beispielsweise auch auf Ernst
22
Mach. Vor diesem Hintergrund ist es das Ziel der Konferenz, einen Dialog zwischen
traditionellen und aktuellen Projekten im Schnittbereich von Epistemologie und
Geschichte zu beginnen.
Als systematische und historische Bezugspunkte kommen dabei in Frage:
1.) Die Geschichte wissenschaftlicher Begriffe
a) Ist es angesichts der zunehmenden Bedeutung der Visualisierung und
Mathematisierung wissenschaftlichen Wissens noch angemessen, die Begriffe
als zentralen Gegenstand der Wissenschaftsgeschichte zu betrachten, wie dies
Canguilhem u.a. im Anschluss an Cavaills getan hat?
b) Wenn die erste Verwendung des Ausdrucks Historische Epistemologie sich
bis in die frhen 1960er Jahre zurckverfolgen lsst, wie weit lsst sich dann die
Geschichte des Begriffs zurckverfolgen: bis Comte oder noch weiter zurck?
2.) Die Historizitt von Objekten und Rumen des Wissens

a) Bachelard scheint mit seiner Rede von der Phnomenotechnik dem
Konstruktivismus der heutigen Wissenschaftsforschung ein zentrales Thema
vorgegeben zu haben. Zugleich wandte sich derselbe Konstruktivismus aber
vehement gegen die Idee eines Bruches zwischen wissenschaftlichem und
alltglichem Wissen, zwischen dem Raum der Wissenschaft und dem der
Literatur und der Kunst. Ist diese Tendenz zur Vereinheitlichung weiterhin von
Interesse?
b) Die Historische Epistemologie scheint in einem eng umgrenzten Raum des
Wissens in Paris entstanden zu sein. Doch hat dieser Raum sich nicht auch erst
ber Verbindungen zu hnlich gelagerten Projekten definiert, etwa zur
Erkenntniskritik Nietzsches und zur Psychoanalyse wissenschaftlichen Wissens
(Freud, Jung, Lacan)? Und wie ist das Verhltnis der Historischen
Epistemologie zu anderen, scheinbar weit entfernten Wissensgebieten zu
beschreiben, zum Beispiel zum Logischen Positivismus (Reichenbach), zur
Allgemeinen Biologie (Goldstein) oder zur Metaphorologie von Hans
Blumenberg?
3.) Der quasi-experimentelle Charakter des Wissenschaftsgeschehens und das Problem

der Rekurrenz
a) und b) Die Einsicht, dass nicht nur die Experimentierttigkeit, sondern auch
Experimente ein Eigenleben haben, ist seit den 1980er Jahren eine
entscheidende Anregung fr die neuere Philosophie und Geschichte der
Wissenschaft gewesen. Erst jetzt scheint sich zu verdeutlichen, dass dieses
23
Thema auch fr Bachelard und Canguilhem von zentraler Bedeutung war. Liegt
hier der tiefere Grund fr das neu erwachte Interesse an der Historischen
Epistemologie? Oder ist gerade diese Konvergenz ein Anzeichen der
Antiquiertheit, da das Experiment in der Praxis einer zunehmend technisierten
Praxis der Wissenschaft kaum noch einen Platz hat?
Literatur
Balibar, Etienne und Pierre Macherey (1970). pistmologie , in Encyclopaedia
Universalis, tome 6 : lastomres Film d'art, Paris : Encyclopaedia
Universalis France, 1970, S.370-373.
Canguilhem, Georges (1963). Lhistoire des sciences dans loeuvre pistmologique de
Gaston Bachelard, Annales de lUniversit de Paris 1 (1963): 24-39,
wieder abgedruckt in: ders., Etudes dhistoire et de philosophie des
sciences, Paris: Vrin, 1968, S.173-186.
Daston, Lorraine (1994). Historical Epistemology, In: J. Chandler, A. I. Davidson und
H. Harootunian (Hg.), Questions of Evidence: Proof, Practice, and
Persuasion across the Disciplines, Chicago: University of Chicago Press,
S.282-289.
label historical epistemology, CIRST Note de recherche, 2010-01, S.117.
Lecourt, Dominique (1975). Kritik der Wissenschaftstheorie. Marxismus und
Epistemologie (Bachelard, Canguilhem, Foucault), bers. von Irmela Neu,
Westberlin: Verlag fr das Studium der Arbeiterbewegung, 1975
Lecourt, Dominique (2008). Georges Canguilhem, Paris: PUF.
Rheinberger, Hans-Jrg (2007). Historische Epistemologie zur Einfhrung, Hamburg:
Junius.
Sturm, Thomas & Feest, Uljana (Hg.), What (Good) is Historical Epistemology?,
Berlin: Max Planck Institute for the History of Science, 2009 [Preprints;
386].
Sturm, Thomas & Feest, Uljana (Hg.), What (Good) is Historical Epistemology?,
24
25
Erffnungsvortrag:
Yves Duroux
Histoire(s) dpistmologie(s)
14:30 Thomas Ebke

The Rupture and the Screw.
The Structure of History according to
Georges Canguilhem and Helmuth Plessner
17:00 Begrung und Einfhrung

Henning Schmidgen
Peter Schttler
Schlussdiskussion
Leitung: Hans-Jrg Rheinberger
Anselm Haverkamp
The Bifurcation of Nature and
the Emergence of History
16:30 Frieder Otto Wolf

Une pistmologie historique des sciences
sociales est-elle possible?
16:30 Claude Debru

Georges Canguilhem. pistmologie historique
et/ou histoire philosophique
Pierre-Olivier Methot
From concepts to experimental systems.
Trends in historical epistemology
16:00 Kaffeepause
Cristina Chimisso
The Role of Scientific Ideology
in Canguilhems Historiography
14:30 Maria Muhle

Histoires de la vie de Canguilhem
Foucault
13:00 Mittagessen
Andrea Cavazzini
Archologie des concepts et philosophie
de la nature
11:30 Cornelius Borck

Working on the evolution of problems
from both sides of the Rhine.
Blumenberg and Canguilhem
11:00 Kaffeepause
Sandra Pravica
Bachelard's experimental approach
to rationality and some strands of
scientific philosophy in the 1930s
9:30 Franoise Balibar

Le concept d'objectivit selon Canguilhem
Samstag, 11. Dezember 2010
16:00 Kaffeepause
Todd Meyers
Kurt Goldsteins Revision of Physiology
13:00 Mittagspause
Monika Wulz
Intervals, possibilities, and encounters.
The trigger of a ruptured history in Bachelard
11:30 Franois Delaporte

La dissection humaine. De l'histoire
des commencements celle des origines
11:00 Kaffeepause
Camille Limoges
Lpistmologie historique dans litinraire
intellectuel de Georges Canguilhem
9:30 Moritz Epple

Enchainements without transcendence.
Some remarks on a historical epistemology
of mathematics
Freitag, 10. Dezember 2010
Donnerstag, 9. Dezember 2010
Berlin-Brandenburgische Akademie
der Wissenschaften
(Leibniz-Saal)
Jgerstr. 22/23
10117 Berlin
Veranstaltungsort:
ERFFNUNG/OUVERTURE
Henning Schmidgen
Meine sehr geehrten Damen und Herren, es ist mir eine Freude, Sie zu unserer Tagung
Epistemologie und Geschichte begren zu drfen. Je mappelle Henning
Schmidgen, et je vous souhaite la bienvenue Berlin dans cette belle salle de
lAcadmie des Sciences de Berlin-Brandebourg.
Les organisateurs de ce colloque, cest--dire Peter Schttler, Jean-Franois
Braunstein et moi-mme, sont particulirement heureux de votre prsence malgr le
mauvais temps et la neige. En effet, les aroports de Paris ayant t ferms hier soir et
ce matin, nous attendons encore quelques collgues et amis qui avaient lintention de
venir. Nous esprons dailleurs quils pourront encore nous rejoindre au cours du
colloque. Bien entendu, nous sommes trs heureux que vous ayez accept de venir
participer lexploration de cette histoire si particulire qui mne de Bachelard et
Canguilhem lhistoire des sciences daujourdhui, et nous assisterons trs certainement
des exposs et des dbats passionnants.
Diese Tagung ist eine Veranstaltung von Abteilung III des Max-Planck-Instituts fr
Wissenschaftsgeschichte, die gemeinsam mit dem Pariser Verbundprojekt Exprience
et Cognition (Philosophies contemporaines, EA 3562) ausgerichtet wird. Untersttzt
wird sie durch das hiesige Centre Marc Bloch, dem dafr an dieser Stelle ausdrcklich
gedankt sei.
Der Anla fr diese Tagung ist ein doppelter: premirement, l pistmologie
historique a accompli depuis environ vingt ans une carrire internationale
impressionnante. Depuis une rencontre organise en 1993 sous ce titre par Ian Hacking
Toronto, le terme dpistmologie historique est devenu lenseigne dun grand nombre
de projets et de programmes. Du traitement des styles de pense scientifique par
Hacking lui-mme lhistorisation des catgories pistmologiques par Lorraine Daston
jusquaux tudes conceptuelles dArnold Davidson sur la formation des nouvelles
formes de savoir et dexprience ou la tentative entreprise par Peter Damerow et
dautres de renouveler et dlargir lpistmologie gntique de Jean Piaget travers
lhistoire des sciences.1
Vor diesem Hintergrund hat sich in jngster Zeit eine Reihe von Konferenzen auf die
Historische Epistemologie bezogen, um die gegenwrtig mit diesem Titel assoziierten
Untersuchungen und Unternehmungen genauer zu diskutieren ich erwhne hier
besonders die Tagung mit dem doppelten Titel What is Historical Epistemology?

1
Siehe im einzelnen die Literaturangaben in den Prospectuses zur Tagung (S.7-24).
27

Schmidgen
und What good is Historical Epistemology? , die 2008 am Max-Planck-Institut fr

Wissenschaftsgeschichte stattgefunden hat.2
Unsere Konferenz zielt in eine andere Richtung. Sie fragt nicht oder zumindest
nicht nur Was ist Historische Epistemologie? , ihr geht es um das Wie, also auch
um das Wer, das Wann, das Wo und das Wohin der mit diesem Ausdruck belegten
Programmatik. Diese Konferenz mchte also in die laufende Debatte dadurch
eingreifen, dass sie das Projekt, das mit diesem Titel bezeichnet ist, in gewisser Weise
dramatisiert. Ihre Absicht ist es, zur Historisierung und Kontextualisierung der
Historischen Epistemologie beizutragen, um damit eine wichtige Voraussetzung fr
ebenso konkrete wie kritische Aktualisierungen zu schaffen.
Au centre nous placerons la relation entre pistmologie et histoire telle quelle parut
vers la fin de lentre-deux-guerres, notamment dans le contexte franais, par une
dmarcation polmique vis--vis de la philosophie traditionnelle et dans le cadre dun
rapprochement avec les sciences de la nature. En tmoignent de manire exemplaire les
crits de Gaston Bachelard, les travaux de Georges Canguilhem, et par la suite aussi les
tudes de Louis Althusser, de Michel Foucault et de bien dautres encore. En fait, ce
nest que dans ce contexte que lon commena parler, depuis les annes soixante,
dune pistmologie historique .
Mais nous avons encore une deuxime raison de nous runir. travers ce colloque,
en effet, nous voulons galement honorer le philosophe, biologiste et historien des
sciences Hans-Jrg Rheinberger dont nous fterons bientt le 65e anniversaire et dont
luvre est fortement redevable la tradition franaise de lpistmologie historique.
Depuis 1996, Hans-Jrg Rheinberger est directeur lInstitut Max Planck de lHistoire
des Sciences o il dirige le dpartement III, consacr surtout aux sciences modernes de
la vie. Beaucoup de ceux prsents dans cette salle ont dailleurs, un moment ou un
autre, t membres ou chercheurs invits de ce dpartement. Or, en janvier 2011, HansJrg prendra sa retraite en tant que directeur de cet endroit extraordinaire. Mme sil
continuera lui-mme travailler linstitut en tant que directeur honoraire, son
dpartement en tant que tel va cesser dexister.
Es scheint nicht unangebracht zu sein, kurz in Erinnerung zu rufen, dass diese
Abteilung, die inhaltlich auf die Geschichte der modernen Lebenswissenschaften
ausgerichtet war, entlang von drei Schwerpunkten gearbeitet hat :
1. die Geschichte und Epistemologie des Experiments,
2. die Geschichte der Objekte und Rume des Wissens, und
3. die Herausbildung von Begriffen und der Gebrauch von Theorie.

2
28
Thomas Sturm & Uljana Feest (eds.) (2009), What (Good) is Historical Epistemology?, Berlin : Max
Planck Institute for the History of Science [Preprints ; 386]. Siehe jetzt auch Thomas Sturm & Uljana
Feest (eds.), What (Good) is Historical Epistemology?, Erkenntnis 75/3 (2011), Special Issue.
ERFFNUNG/OUVERTURE
Alle diese Schwerpunkte weisen interessante Resonanzen zu der eben skizzierten

Tradition der Historischen Epistemologie auf. Rheinberger selbst hat in seinen neueren
Arbeiten auf die Bedeutung hingewiesen, die nicht nur Derrida und Husserl, sondern
besonders die Arbeiten von Bachelard und Canguilhem fr seine eigene
Forschungsttigkeit hatten.
Depuis les annes 1980, les travaux de Rheinberger ainsi que la nouvelle histoire et
philosophie des sciences ont t stimuls de faon dcisive par la thse de Hacking
selon laquelle les expriences scientifiques possdent une vie autonome .3 Or,
aujourdhui il savre de plus en plus que ce thme avait dj jou un rle central chez
Bachelard et Canguilhem. Par consquent, peut-on y voir la cause vritable dun nouvel
intrt pour lpistmologie historique? Ou cette rcurrence est-elle plutt lindice
dune obsolescence, puisque, dans une pratique scientifique devenue de plus en plus
technique, lexprimentation semble presque avoir perdu sa place?
Ce sont de telles questions que notre colloque voudrait aborder. Notre but est
douvrir un dialogue concret entre les projets traditionnels et ceux plus rcents, situs
linterface entre pistmologie et histoire.
Dass dieser Dialog, der unserem Verstndnis nach auch und vor allem ein deutschfranzsischer Dialog ist, in den Rumen der Berlin-Brandenburgischen Akademie der
Wissenschaften stattfinden kann, ist nicht nur einem organisatorischen
Erfordernis geschuldet, sondern soll ebenso die besondere Bedeutung unterstreichen,
die dieser Veranstaltung aus unserer Sicht zukommt. Tatschlich erscheint es als
besonders glcklich, da wir uns hier im Leibniz-Saal zusammenfinden, denn Leibniz
steht nicht nur fr die Verbundenheit der franzsischen mit der deutschen Philosophie
und Geschichte, sondern kann, ohne der Sache allzuviel Gewalt anzutun, auch als ein
Historischer Epistemologe betrachtet werden.
Zumindest war dies die Ansicht des Schriftstellers und Philosophen Michel Deguy.
In einer Buchbesprechung aus dem Jahr 1960 beschreibt Deguy das Verhltnis von
Leibniz und Descartes in dem Sinne, da Leibniz der erste Bachelard der
cartesianischen Philosophie gewesen sei, whrend der Autor des fraglichen Buches,
Yvon Belaval, als der Bachelard von Leibniz agiere.
Dailleurs, ce texte de Michel Deguy, publi dans la Nouvelle Revue Franaise,
contient dj lexpression qui nous retient aujourdhui, puisquil y est dit explicitement:
Le livre de Belaval [ savoir son Leibniz, critique de Descartes] se situe dans la ligne
des travaux universitaires franais qui passe par exemple par Brunschvicg et
aujourdhui M. Bachelard, pour qui la philosophie est devenue une pistmologie
historique. 4

3
Ian Hacking, Representing and Intervening. Introductory Topics in the Philosophy of Natural Science,
Cambridge/New York: Cambridge University Press, 1983, p.150.
Michel Deguy, Leibniz, critique de Descartes , La Nouvelle Revue Franaise 8/96 (1960) : 10941101, p.1094 (Hervorh. von mir, H.S.).
29

Schmidgen
Certes cette formule ne commena circuler vraiment qu partir de 1969, lorsque

Dominique Lecourt la plaa dans le titre de son petit livre, tir de son mmoire de
maitrise dirig par Canguilhem, Lpistmologie historique de Gaston Bachelard.5 Mais
Canguilhem, lui aussi, lavait dj utilis auparavant, en 1963, dans son fameux article
Lhistoire des sciences dans luvre pistmologique de Gaston Bachelard .6
Bien entendu, il y a toujours une diffrence entre un concept et un terme, et au cours
de ce colloque nous aurons des communications qui montreront que le concept de cette
branche de lpistmologie est bien plus vieux que le texte cit de Michel Deguy en
1960.7 Cependant, jai voulu attirer votre attention sur cette question, car, comme nous
le savons, il ny a pas de dnominations innocentes.
Dass die Historische Epistemologie seit den 1960er Jahren ein durchaus
beziehungsreicher Titel geworden ist, macht der Titel unseres ersten geplanten
Vortrages deutlich: er spricht nicht von der Historischen Epistemologie, auch nicht von
der epistemologischen Historie, sondern von Geschichten und Epistemologien,
Histoire(s) depistemologie(s) .
Leider werden wir diesen Vortrag heute nicht hren knnen. Yves Duroux hat seine
Teilnahme aus gesundheitlichen Grnden absagen mssen. Duroux wre gerne
gekommen, und angesichts seiner Verwurzelung in der Gruppe um Althusser und den
Cahiers pour lAnalyse, wie auch seiner mageblichen Aktivitten im CNRS, im
Wissenschaftsministerium und am Collge International de Philosophie war er dafr
prdestiniert, diese Tagung zu erffnen. In einem langen Brief verspricht Duroux, einen
Beitrag in schriftlicher Form, sollten wir uns entscheiden, die Tagungsbeitrge in einem
Band zu verffentlichen, und entschuldigt sich nochmals fr seine nicht vorhersehbare
Abwesenheit : Jen suis vraiment dsol, dautant plus que jaurais aim revoir HansJrg Rheinberger auquel je vous demande de transmettre toute mon amiti. Was
hiermit geschehen ist.
Natrlich gibt es keinen Ersatz fr Duroux, es gibt nie einen Ersatz, und in
Absprache mit meinen Ko-Organisatoren wird der eigentliche Beginn der Tagung auf
den morgigen Tag verschoben. Was uns heute Abend bleibt, ist Sie gewissermaen auf
diese Erffnung einzustimmen. Und das mchten wir in drei Schritten von jeweils etwa
zehn Minuten und in zwei Sprachen tun.
Zuerst wird Jean-Franois Braunstein zu Ihnen in englischer Sprache ber die
franzsische Tradition der Historischen Epistemologie sprechen, danach wird Peter
Schttler Ihnen auf Franzsisch einige Stze ber die deutsche Rezeption dieser
Tradition sagen. Abschlieend werde ich Ihnen wiederum auf Englisch einige

5
30
Dominique Lecourt, Lpistmologie historique de Gaston Bachelard, Avant-propos de Georges

Canguilhem, Paris: Vrin, 1968.
Georges Canguilhem, Lhistoire des sciences dans luvre pistmologique de Gaston Bachelard
[1963], in : id., Etudes dhistoire et de philosophie des sciences, Paris : Vrin, 1968, pp.173-186,
citation p.178.
Siehe dazu besonders den Beitrag von Jean-Franois Braunstein (S. 33-40).
ERFFNUNG/OUVERTURE
Einzelheiten zu Hans-Jrg
Epistemologie berichten.
Rheinbergers
Bezugnahmen
auf
die
Historische
Damit schliee ich vorerst. Ich bedanke mich nochmals fr Ihr Kommen und
wnsche uns allen eine anregende und ergebnisreiche Tagung.8

Fr ihre Hilfe bei der Erstellung der vorliegenden Dokumentation sei Sandra Gerhardinger
(Regensburg) herzlich gedankt.
31

HISTORICAL EPISTEMOLOGY, OLD AND NEW
In recent years there have been plenty of conferences dedicated to historical

epistemology. In the last two years alone, three conferences on this topic have met in
Berlin, Columbia and Leuven. The first conference dedicated to historical epistemology
was organized by Ian Hacking in Montreal in 1993. This wealth of conferences has
prompted some to think that such a craze is a mere fad: historical epistemology would
only be a useless label (naming without necessity, according to Gingras)1 or even a
brand that would only serve an advertising purpose: that of justifying the applications
for funding of a particular institution. In other words, all the fuss about historical
epistemology would merely amount to creating a market from nothing. However,
anybody working in the advertising business knows very well that it is impossible to
create a market from nothing: nothing can be sold if it does not answer a certain need,
albeit more or less vaguely felt.
In fact, it seems to me that there exists today, as the title of our conference suggests,
a real necessity to bring epistemology and history closer together. If we have chosen as
a title for our conference History and Epistemology: From Bachelard and Canguilhem
to Todays History of Science, it is indeed because we did not want to stick to the
current meaning of historical epistemology. We hope that one of the original features of
this conference will be that it will also be interested in the long-term history of the
concept of historical epistemology. It indeed seems to me that the least that an
epistemology which boasts to be historical can do is to look into the historical
conditions of its own production. We will not content ourselves with asking what
historical epistemology is or what it is worth; we will also wonder what its history has
been. In that sense we merely pay tribute to Hans-Jrg Rheinberger and carry on the
work of the one who had been able, most notably in his book Historische
Epistemologie, to look into this one hundred-year long history of historicizing
epistemology.
It is also for that reason that our conference has chosen to go back to the French
origins of that historical epistemology. Like Hacking, Rheinberger has always kept in
mind the French setting in which the first invention of historical epistemology took
place. Our conference on historical epistemology is indeed one of the few to welcome
such an important number of French attendees and I would like to take advantage of the
opportunity given to me here to thank my co-organizers for granting something like
political asylum to an historical epistemology which probably is no longer welcome, for
Y. Gingras, Naming without necessity: on the genealogy and uses of the label historical
epistemology, Note de recherche, 2010-01, CIRST.
33
Braunstein
all sorts of reasons that we will certainly have the chance to discuss, in the country
where it was born.
A Marxist oriented invention?

The history of the term historical epistemology is apparently well-known today. If one
sticks to the received view, the locution was invented by Dominique Lecourt in his
work Lpistmologie historique de Gaston Bachelard, first published in 1969. Lecourt
would later credit Canguilhem, who supervised and wrote a preface to this piece of
research, with this invention: since I am told that this point of lexicological history has
been raised, I must say that I owe the locution historical epistemology, which referred
to Gaston Bachelards philosophy of science (in my Masters thesis of 1969 which bore
that title), to Canguilhem.2 However, as it has been quoted somewhere else,
Canguilhem would have preferred the expression epistemological history to that of
historical epistemology.3 By way of this difference in emphasis, Canguilhem
probably indicates where the major difference lies between his own work and
Bachelards.
As some commentators have noted, following Dominique Lecourt the expression
historical epistemology was also used in a Marxist-oriented context by Marx
Wartofsky in 1973: a few years later, I discovered that the same (expression) had been
used in English, completely independently, by Marx Wartofsky, one of the two founders
of the Boston Studies in the Philosophy of Science, to describe his own Marxist-leaning
methodology in the history of science.4 Yves Gingras has also remarked that Polish
logician Jerzy Kmita had published his Problems in historical epistemology in Polish in
1980 (the English translation is from 1988), which was also Marxist in inspiration.
Accordingly, for Gingras the expression historical epistemology would have appeared
in a Marxistoriented context and would have been constructed using the model of the
expression historical materialism.5 I do not find this hypothesis convincing, at least
for the French context. The two expressions are not homothetic since one refers to a
discipline, epistemology, while the other refers to a philosophical doctrine, materialism.
The French writers of the 1970s who, even though they were Marxists, were far from
being true fans of historical materialism, do not seem to have made the connection.
On the other hand, however, I do think that such an historical survey might
nonetheless be relevant for the lessons that can be drawn from it. But it will have us go
back way further. It indeed turns out that the expression historical epistemology did
not emerge in the Marxist-oriented context of the 1970s but almost seventy-years
earlier used by a writer who was no Marxist, even though Lenin read him carefully.
The dating of the appearance of the expression historical epistemology is no mere
2
4
5
34
D. Lecourt, Georges Canguilhem, le philosophe, in: J.-F. Braunstein (ed.), Canguilhem. Histoire
des sciences et politique du vivant, Paris: P.U.F., 2007, p. 30 n.
J. Gayon, Bachelard et lhistoire des sciences, in: J.-J. Wunenburger (ed.), Bachelard et
lpistmologie franaise, Paris : P.U.F., 2003, p. 53.
Lecourt, Georges Canguilhem, le philosophe, p. 30 n.
Gingras, Naming without necessity.
anecdotal curiosity: it also allows a better grasp of the philosophical intentions of those
who created this phrase that is so discussed today.
An old expression for an old style of thinking

I have realized that the expression historical epistemology first appeared in the
writings of an author who was a great coiner of neologisms in philosophy and history of
science and who was also the founder, in 1932, of the Institut dhistoire des sciences et
des techniques of the Universit de Paris, an institution to which many of us have been
affiliated and of which I was still a member until recently. I am talking about Abel Rey
(1873-1940) who, as I have noted, described his own work, many years before Foucault,
as archaeology of scientific ideas.6 This writer is mostly forgotten, but as is often the
case with minor figures, he articulated ideas in a very clear way that would be later be
taken up and developed in a more complex fashion by more original and important
writers in this instance Bachelard and Canguilhem.
It turns out that Abel Rey had used the expression historical epistemology as early
as 1907 to describe his own work in his PhD thesis La thorie de la physique chez les
physiciens contemporains. In his introduction to the book he offered to build, alongside
formal logic, a different kind of logic based on the history of science and more
specifically on what he called a history of the conceptions and institutions of scientists
with respect to methods, where science is advanced enough for its methods to spark off
fruitful investigations.7 Abel Rey notes that philosophy, especially American
philosophy, calls these transcendent investigations on the general principles and
conditions of the sciences epistemology but, he adds so as to mark his difference, that
this word, restricted to a positive meaning, might happily refer to the research of
documents, the set of historical observations required for the establishment of an exact
view of the different sciences, a positive science of the sciences. Historical research
bearing on the general spirit of each science would then constitute one of the great
sections of this historical epistemology.8 The philosophy of the sciences could only be
historical. It is to this kind of investigation that he intended to contribute. What I
primarily intend to do here, he wrote, is precisely to contribute to these historical
investigations bearing on the scientific spirit.9 This historical approach to the
philosophy of the sciences went hand in hand with a determined hostility toward any
theory of knowledge. In the first issue of Thals, the periodical published by the Institut
dhistoire des sciences, Rey introduced the following agenda, which sounds quite
7
8
9
A. Rey, La science dans l'Antiquit. T. I: La science orientale avant les Grecs, Paris: Albin-Michel,
1930, p. 384. Quoted in J.-F. Braunstein, Abel Rey et les dbuts de l'Institut d'histoire des sciences et
des techniques, in M. Bitbol & J. Gayon (eds.), Lpistmologie franaise.1830-1970, Paris: P.U.F.,
2006.
A. Rey, La thorie de la physique chez les physiciens contemporains, Paris: Alcan, 1907, p. 13.
Ibid.
Ibid.
35
Braunstein
Canguilhemian: The theory of knowledge, without a philosophical history of science,

is only a vague ideology or a verbal dialectics.10
Moreover, and that is a second basic point made in the introduction to La thorie de
la physique, if one is to practice the history of the sciences, one should never lose sight
of the methodological goal to which this history contributes:11 the history of the
sciences always remains a history with a philosophical purpose. Elsewhere, still in the
first issue of the periodical Thals, Rey explains that the history of the sciences is no
mere work of erudition. It introduces a crucial component of the history of
civilization, which is not only material but also intellectual, and even spiritual.12
Here, one deals with a philosophical history of the sciences, which Rey also
describes, following Tannery and, above all, Comte, as a general history of the
sciences. One of the volumes of his Science grecque is dedicated to the memory of
Paul Tannery, who reformed the history of the sciences by treating it as a true historian
should, by linking it, as Auguste Comte already intended it, to the history of ideas and
the history of human thought.13 This general history of the sciences, as Comte
properly saw it, [] is either philosophical or it does not exist.14
Thirdly Rey, in La Thorie de la physique, emphasizes that such a philosophical
history of the sciences results in the complete transformation of our conception of truth,
which would no longer conform to the traditional, motionless idea of truth, but which
would not disappear either: truth would be the becoming of an evolution; but it would
exist, since it would constantly realize and complete itself.15 He continues: The
history of science, continuing the psychological history of perception and conception,
would be the history of the acquisition of that truth, the human and necessary outcome
of a human and necessary evolution, within the limits of human nature.16 That history
would enable one to establish a new conception of truth which Rey describes in very
Bachelardian terms as a new rationalism and which he holds to be part and parcel of
the tradition of the philosophical spirit of Comtean positivism.17 Elsewhere Rey
notes that this new rationalism in no way excludes a psychological history of
reason.18
10
11
12
13
14
15
16
17
18
36
A. Rey, Prface, Thals. Recueil annuel des travaux de lInstitut dhistoire des sciences et des
techniques de lUniversit de Paris. Premire anne. 1934, 1935, p. XVIII.
Rey, La thorie de la physique chez les physiciens contemporains, p. 13.
A. Rey, L'Institut d'histoire des sciences et des techniques de l'universit de Paris, Thals. Recueil
annuel des travaux de lInstitut dhistoire des sciences et des techniques de lUniversit de Paris.
Premire anne. 1934, 1935, p. V.
A. Rey, La science dans l'Antiquit. Tome II : La jeunesse de la science grecque. Paris: Albin Michel,
1933, p. 1.
A. Rey, Histoire de la mdecine et histoire gnrale des sciences, Thals Recueil annuel des travaux
de lInstitut dhistoire des sciences et des techniques de lUniversit de Paris. Deuxime anne. 1935,
1936, p. 34.
Ibid.
Ibid., p. 393, 392.
Rey, La philosophie moderne, p. 91.
Eventually, and that is the last important point made in the introduction to La
Thorie de la physique, Abel Rey explains in what lies the crucial present interest of
such a research. The critics of traditional mechanism wanted to take advantage of the
crisis of contemporary physics to proclaim the failure of science, of reason and of
the positive spirit and to revert to a mystical intuition, to a mystical sense of
reality, to mystery.19 According to Abel Rey, it is therefore clear that this work is
not merely theoretical, for it can also answer serious and current concerns20. It might
help in the fight of the Enlightenment against mysticism and Rey would keep on
emphasizing the humanistic dimension of the history of science and the humanism
conveyed by the positive sciences.21 For Rey, as it would also be the case for
Canguilhem later, history of science is always, in a certain sense, a political history of
the sciences.
It would be possible to draw other comparisons between Abel Rey and, on the one
hand, his primary inspiration, Auguste Comte, and on the other hand such followers of
his as Bachelard and Canguilhem, but also in some ways Michel Foucault and
contemporary German-American historical epistemology.
Rey also maintains that science is the result of a choice between different
possibilities and that truth is a value among others: truth is a value, just as the
Beautiful and the Good; the search for truth, which is what both Science and Philosophy
are after, is the pursuit of a value.22 In that respect Rey anticipated Canguilhems most
iconoclastic claims when the latter explained that saying that there is no other
knowledge than scientific knowledge does not amount to saying that there are no other
goals or values for the human mind except truth.23 Curiously enough, both in Rey and
Canguilhem, beyond Max Webers polytheism of values one finds the same reference
to Nietzsche lurking in the background. Canguilhems theory of the axiological bias
for truth refers to Nietzsche, for whom truth is a value that is to be situated among a
plurality of values and certainly not the only value man may devote himself to.24
References to Nietzsche are also significant in Rey who considers Eternal return to be
one of the basic guiding ideas of our science and regards Nietzsche as some sort of a
scientific philosopher pointing out that Nietzsche in 1881, by way of a brilliant
intuition, an intuition originating in one of the greatest philosophical geniuses of our
human history, intended to devote ten years of his life to studying the natural sciences
so as to be able to base his idea of eternal return on atomic theory.25 And when Rey
19
20
21
22
23
24
25

Ibid., p. 20.
Rey, LInstitut d'histoire des sciences et des techniques de l'universit de Paris, p. V.
A. Rey, [Foreword to] Alfred Stern, La philosophie des valeurs. Regard sur ses tendances actuelles
en Allemagne. Paris: Hermann, p. 3.
G. Canguillhem, Philosophie et science (1965), Cahiers philosophiques, Hors srie, juin 1993, p.
22.
G. Canguilhem, De la science et de la contre-science, in: S. Bachelard et al., Hommage Jean
Hyppolite. Paris: PUF, 1971, p. 177.
A. Rey, Le retour ternel et la philosophie de la physique. Paris: Flammarion, 1927, p. 308-309.
37
Braunstein
represents the history of humanity as a sinusoidal curve or as a cyclical fate joining

Apollo with Dionysos, it is also to Nietzsche that he refers.26
Four features of French epistemology

The four characteristic features of what Rey calls historical epistemology seem to be
widely shared by most of the representatives of what I would like to call the French
style in philosophy of science originating with the one who invented it in the middle
of the 19th century, Auguste Comte, and extending at least as far as Michel Foucault.
Accordingly, at least in France, historical epistemology is no recent trend that is bound
to disappear. Its discrete but insistent presence in Abel Reys work testifies to this fact.
Let me summarize these four features of French epistemology.
1. French epistemology always consists in an a posteriori reflection on the sciences.
One may wonder how this epistemology may succeed in being more than an illinformed repetition of science. In fact it is its historical approach that enables French
epistemology not to merely repeat the sciences: the philosophy of the sciences la
franaise is always a history of the sciences. Canguilhem states this relationship very
clearly la manire dAbel Rey: Without a reference to epistemology, the theory of
knowledge would be an empty meditation; without a relation to the history of the
sciences, epistemology would amount to a completely superfluous doublet of the
science it is supposed to discuss.27
The relationship between philosophy and the sciences can be fruitful only if it rests
on the history of the sciences. French epistemology expects to find answers to some
traditional philosophical problems, such as the problem of scientific objectivity or the
question of truth and especially error, within the history of the sciences. Accordingly,
but quite paradoxically, Canguilhem maintains that historicity constitutes a genuine
criterion of demarcation between what is and what is not scientific. The proof of the
scientific nature of a discipline lies in its historical dimension: A science deprived of
history, that is a science within which the rejection of some conditions of objectivity at
a given moment and their replacement by other conditions of objectivity more
objectively defined had not taken place, such a science is not a science.28 Historicity is
what distinguishes true sciences from false sciences, such as astrology, whose
proper feature is to have no history.
2. However, this history of the sciences is not a history in the classical sense of the
word: it claims to be a critical or philosophical history and differs from traditional
history in two main respects. On the one hand this history does judge and value what it
26
27
28
38
A. Rey, La science dans l'Antiquit. Tome III: La maturit de la pense scientifique en Grce. Paris:
Albin Michel, 1939, p. 554, 556.
G. Canguilhem, Lobjet de lhistoire des sciences, in: id., tudes dhistoire et de philosophie des
sciences, 1968), 7me d., Paris : Vrin, 1994, p. 11-12.
G. Canguilhem, Objectivit et historicit, in Collectif, Structuralisme et marxisme, Paris: 10-18,
1970, p. 235.
studies. To the model of the laboratory, claims Canguilhem, one can oppose, in order
to understand the function and meaning of the history of the sciences, the model of the
school or the model of a court of justice, that is of an institution, and a place where
judgments are passed on the knowledge of the past.29 On the other hand it is a history
that judges by way of recurrence; that judges the past in the light of the present.
3. Thirdly, the epistemology developed by these authors leads them to address the
question of the development of reason, which is grasped only through the development
of the sciences. A consequence of this view is that the different forms of reason are said
to be dependent on historical or geographical conditions. Bachelard explains that
since reason must obey science the former must follow the dialectics of the latter:
the traditional doctrine of an absolute and immutable reason, claims Bachelard, is
only a philosophy. It is an outdated philosophy.30 The new rationalism or archrationalism (surrationalisme) advocated by Bachelard, after Abel Rey, is a perpetual
conquest.
Foucault also keeps in mind the Bachelardian and Canguilhemian lesson regarding
the historicity of rationality and acknowledges the fact that he has benefited from
Bachelards idea that reason works on itself at the very moment it constitutes its
objects of analysis.31 What is at stake here is the constitution of a form of rationality
that is presented as prevailing and to which the status of reason is granted so as to make
it appear as one of the possible forms of rationality at work.32 Foucault refuses the
blackmail that has very often been directed at any critique of reason or critical
questioning concerning the history of rationality, and he holds that it is possible to
write a contingent history of rationality just as it is possible to attempt a rational
critique of rationality.33 If Foucault refers several times to Kants text What is the
Enlightenment?, for example in his introduction to the English edition of The normal
and the pathological, it is because he reads it as the first attempt to question the history
and geography of reason, to question its immediate past and the conditions of its
working, to question its moment, its location, and its actuality.34 Rationality,
notwithstanding its claim to universality, adopts historically determined forms.
4. Fourthly, and finally, the history of science is always linked to political goals,
broadly construed, either in Comte, where going through the whole history of the
sciences only aims at the establishment of sociology, the science that would enable the
reorganization of society, or in Canguilhem, with the demonstration that the
deterministic conception of the milieu is not scientifically valid, which he already
judged to be unfair in his early polemical writings against Taine et Barrs, as he would
29
30
31
32
33
34
G. Canguilhem, Etudes d'histoire et de philosophie des sciences, p. 13.

G. Bachelard, La philosophie du non (1940), 9me d. 1983, p. 145.
M. Foucault, Entretien avec Michel Foucault (1980), Dits et crits, t. IV, Paris : Gallimard, 1994,
p. 56.
M. Foucault, Structuralisme et poststructuralisme (1983), Dits et crits, t. IV, p. 440.
Ibid.
M. Foucault, La vie: l'exprience et la science, Revue de mtaphysique et de morale, 90,1, 1985.
39
Braunstein
also argue in his later paper Le vivant et son milieu.35 But it is also true of Foucault.
When he contrasts a philosophy of experience, of meaning, of the subject, that of
Sartre and Merleau-Ponty, with a philosophy of knowledge, rationality and concept,
that of Cavaills, Bachelard and Canguilhem, which he endorses, it should not be
forgotten that the reminder of his text echoes remarks made by Canguilhem on the
occasion of various tributes to the memory of Jean Cavaills: Outwardly, the latter has
remained the most theoretical, the most circumscribed to speculative tasks, the most
remote from current political concerns of the two. Yet during the war it was that
philosophy that took part, very directly, in the fight, just as if the question of the
foundation of rationality could not be separated from the question of its present
conditions of existence. This philosophy also played a crucial role in the sixties when a
crisis emerged that was not only that of the University, but that of the status and role of
knowledge itself. One can then wonder why such a reflection, whilst following its own
logic, ended up being linked so deeply to the present.36 Historical epistemology has
always been a history of the present.
***
If one does not take into account French parochialism which always results in having
the French discovering Anglo-Saxon philosophies with the time-lag of a generation
there is no doubt that the obviousness of the four features I have pointed out explains
why French epistemology did not in recent years manifest the same amount of interest
exhibited by the learned community for the issues raised by the new historical
epistemology. Let us hope that our conference might contribute to the awakening of
such an interest.
35
36
40
Cf. J.-F. Braunstein, Psychologie et milieu: thique et histoire des sciences chez Georges
Canguilhem, in J.-F. Braunstein (ed.), Canguilhem. Histoire des sciences et politique du vivant,
Paris: P.U.F., 2007.
Ibid., p. 4.
SUR LA RCEPTION DE L PISTMOLOGIE FRANAISE

EN ALLEMAGNE
Peter Schttler
Il me revient la tche de dire quelques mots sur la rception allemande de cette

pistmologie historique apparemment si franaise . Bien entendu, il ne peut
sagir que de remarques rapides, partielles et partiales, vu que, dune part, je ne suis ni
philosophe ni pistmologue; et que, dautre part, jai moi-mme t impliqu dans
cette histoire, il y a quarante ans, notamment comme traducteur. Je dirais mme: comme
militant traducteur ou traducteur militant , si on accepte lide que dans certaines
conditions la traduction dun texte thorique peut avoir un enjeu et une fonction
militante, au-del de son enjeu et de sa fonction thorique.
Il serait difficile en effet, et surtout trompeur, de sparer la rception de certains
textes, leur lecture et leur perception, dune conjoncture qui ne se rduit pas une mise
en prsence thorique, voire philosophique. (Mme sil se trouve toujours des
philosophes qui disent: Ah, je ne fais que de la philosophie, la politique ne mintresse
pas Comme si linverse la politique laissait la philosophie tranquille.)
Donc la rception de lpistmologie historique franaise en Allemagne me semble
insparable de lhistoire de laprs-guerre, et plus particulirement de la conjoncture
thorique des annes 1960-70 dabord, et de ses avatars et retournements ensuite. Pour
la caractriser dans le dtail il faudrait videmment un autre colloque. Mais dans notre
cadre il suffit peut-tre de lancer un mot: blocage . En effet, lAllemagne (je ne parle
ici que de la RFA, la RDA serait une autre histoire1), la Rpublique fdrale est une
socit en pleine reconstruction au sens fort et ambigu du terme, cest-dire construction de quelque chose de nouveau et restauration de quelque chose
dancien. Aprs la dfaite et le dmantlement officiel du nazisme, cest une socit la
recherche dune nouvelle identit idologique sur la base de valeurs anciennes: de la
Rpublique de Weimar, du catholicisme rhnan, du protestantisme libral etc., les
1
Signalons cependant que dans une encyclopdie philosophique quasi-officielle nous trouvons la
prsentation suivante (rdige par un auteur franais) : Dans la pense bourgeoise actuelle la
diffrence entre thorie de la connaissance et pistmologie est souligne dans le but de traiter des
questions scientifiques sans recours aux questions soi-disant traditionnelles dune conception du
monde philosophique. Le but principal tant de liquider le matrialisme philosophique. Suit alors un
dveloppement sur Bachelard et Foucault, philosophes idalistes, partir desquels Althusser aurait
en vain tent de mettre en opposition la science et lidologie marxiste. Michel Vade, art.
Epistemologie , in : Manfred Buhr (d.), Enzyklopdie zur brgerlichen Philosophie im 19. und
20. Jahrhundert, Leipzig : Bibliographisches Institut, 1988, pp. 438453 ; cf. aussi lentre
Epistemologie dans le Philosophisches Wrterbuch dirig par Manfred Buhr et Georg Klaus, 11e
d., Leipzig : Bibliographisches Institut, 1974, t. I, pp. 343344. En RDA, la vraie rception
intellectuelle de lpistmologie franaise sera donc oblige de prendre quelques dtours ; cf. le
remarquable recueil dit par Karlheinz Barck et Brigitte Burmeister : Ideologie-Literatur-Kritik.
Franzsische Beitrge zur marxistischen Literaturtheorie, Berlin : Akademie, 1977.
41
Schttler
combinant plus ou moins et cote que cote, tout en se soumettant aux directives des
forces allis doccupation qui reprsentent aussi des forces intellectuelles . Puisque
seuls les allis, comme chacun sait, taient matres bord jusquen 1955 et garderont
encore par la suite un droit de regard et de contrle en dernire instance jusquen
1990, cest--dire jusqu la runification des deux Allemagnes.2
Cette situation de blocage et de reconstruction caractrise tout particulirement le
monde universitaire qui retrouve aprs 1945 lessentiel des structures davant-guerre et
mme, trs largement, le personnel davant-guerre.3 Rares sont en effet les exclusions
dfinitives de professeurs nazis, plus rares encore, et mme rarissimes, les retours de
professeurs ou denseignants exclus au moment de larrive au pouvoir des nazis.4 Do
une trs grande continuit de la culture acadmique, quant aux structures, quant aux
formes denseignement et souvent mme quant son contenu. Ce sont donc les mmes
grands auteurs qui sont au cur de lenseignement philosophique, et les mmes grands
auteurs absents : allez, par exemple, chercher Spinoza dans lenseignement
philosophique de laprs-guerre!
Il y a cependant des variantes selon les Lnder, les lieux et les facults. Ainsi BerlinOuest avec sa situation particulire et son Universit Libre , fonde en 1948 sur le
modle amricain, sera certes un haut lieu de lanticommunisme mais en mme temps
une universit plus librale, plus ouverte vers la pense occidentale et plus critique de la
tradition germanique ce qui sexprime notamment dans une plus forte prsence de
professeurs rmigrs ou, sur un autre plan, dans linterdiction des corporations
dtudiants (Burschenschaften), symboles de la vieille universit teutonique.5 Mais dans
la plupart des autres universits il faudra attendre presque vingt ans pour que soient
vraiment branls les bastions de l idologie allemande traditionnelle. Cest
pourquoi, en philosophie, on pouvait facilement y parler sans risque de surprendre de
mtaphysique, disons pour aller vite: de ltre, du Temps et du Nant, tandis que
dautres thmes, notamment ceux de la philosophie des sciences, de lempirisme
logique, du rationalisme critique, voire du matrialisme dialectique, se voyaient sinon
exclus, du moins fortement marginaliss.
Pour ce qui est de lpistmologie au sens moderne (au-del de la philosophie
grecque), et notamment dans le sens que lui donnrent au dbut du 20e sicle un
2
42
Sur larrire-fond historique et socio-culturel cf. Axel Schildt/Detlef Siegfried, Deutsche

Kulturgeschichte. Die Bundesrepublik von 1945 bis zur Gegenwart, Munich : Hanser, 2009, pp. 122 et
suiv.
Sur ce phnomne de continuit (relative) cf. les travaux de Norbert Frei, notamment : Karrieren im
Zwielicht. Hitlers Eliten nach 1945, Francfort : Campus, 2001. Pour le monde universitaire, cf. Bernd
Weisbrod (d.), Akademische Vergangenheitspolitik. Beitrge zur Wissenschaftskultur der
Nachkriegszeit, Gttingen : Wallstein, 2002.
Cf. le film ralis en 1949 par Josef von Bky, Der Ruf , avec Fritz Kortner dans le rle dun
professeur de philosophie qui ose revenir de Californie lUniversit de Gttingen et se voit confront
des tudiants encore imprgns didologie nazie <http://www.filmportal.de/film/derruf_d9d852bc421441a885c119291963f383>,
Sur les dbuts de lUniversit Libre de Berlin, fonde en 1948, on lira avec profit les mmoires de
lhistorien, politologue et romaniste Gilbert Ziebura, Kritik der Realpolitik . Genese einer
linksliberalen Vision der Weltgesellschaft. Autobiographie, Mnster : LIT, 2009.
SUR LA RCEPTION DE L PISTMOLOGIE FRANAISE EN ALLEMAGNE
Meyerson ou un Abel Rey (dont Jean-Franois Braunstein a rappel le souvenir, Rey

faisant lintermdiaire non seulement entre la philosophie et les sciences, mais aussi
entre la philosophie et lhistoire, puisquil tait un des directeurs adjoints de la Revue de
Synthse dHenri Berr6), pour ce qui est de lpistmologie moderne donc, le terme
mme (et donc le concept ?) en tait totalement absent en Allemagne fdrale jusque
dans les annes 1970.
En voici deux indicateurs: 1) Dans les premires traductions allemandes dAlthusser
et de Foucault, qui datent de 1968/69 (Pour Marx, LHistoire de la Folie et Les Mots et
les Choses, chez Suhrkamp) le terme d pistmologie est systmatiquement vit et
traduit soit par Erkenntnistheorie soit par Wissenschaftstheorie . 2) Dans les
dictionnaires et les encyclopdies le terme est galement absent. Ainsi le monumental
Historisches Wrterbuch der Philosophie (19712007, 13 volumes) que mettent en
chantier cette poque Joachim Ritter, Karlfried Grnder et leur quipe ne possde pas
dentre pistmologie ou pistmologie historique , ni dailleurs dentres
discours , mentalit etc., au point que lon peut se demander, de manire presque
brutale ou nave, quel problme la philosophie acadmique allemande a-t-elle pu avoir
avec la pense franaise , ou du moins avec une partie de cette pense, cette
poque.
La rponse mes yeux, mme sil faudrait approfondir lenqute, risque de se
trouver dans un certain nombre de barrires qui existaient ce moment-l et qui
fonctionnaient comme autant de panneaux avertisseurs afin dempcher les biens
pensants de tomber dans le panneau de ce que lon appellera plus tard la
Franzosentheorie .7
En fait, il sagissait surtout de deux sortes de panneaux, des vieux et des rcents:
Gare au marxisme ! , par exemple. Ce panneau-l tait dj trs vieux et mme
ringard, si bien quil nimpressionnait plus que les milieux conservateurs. Dans les
milieux tudiants par contre, ou plus gnralement dans les milieux de gauche, domins
par les ides de l cole de Francfort , on disait plutt le contraire: gare au nonmarxisme ! , gare la trahison du marxisme sous la forme nouvelle du
structuralisme ! Ainsi, lorsquen 1967 un althussrien Nicos Poulantzas tenta
dexpliquer dans le cadre dun grand colloque lUniversit de Francfort les nouvelles
perspectives dune lecture structurale du Capital le dbat tourna rapidement court :
on ne parlait plus le mme langage.8 Le titre dun article dAlfred Schmid, porte-parole
6
Cf. ric Brian, Cent dix ans de renouvellement incessants. Note sur litinraire de la Revue de
Synthse Historique de 1900 2010 , Revue de Synthse, 131 (2010), pp. 401-438 ; Peter Schttler,
Le Centre International de Synthse et lAutriche. Note pour une enqute , Austriarca, 31 (2006),
63, pp. 99-117.
Formulation lance par Lothar Baier dans son livre Franzsische Zustnde, Francfort : EVA, 1982,
pp. 21 suiv. Notons que le mme auteur a galement pourfendu la tendance inverse, savoir la
fascination aveugle de beaucoup dintellectuels franais pour la pense allemande : Lothar Baier,
Wider die Meisterdenker. Deutsches Philosophieren, Antihumanismus und der Tod des Subjekts
Blicke nach Frankreich , Kommune, 1987, no. 6, pp. 7176.
Nicos Poulantzas, Theorie und Geschichte. Kurze Bemerkungen ber den Gegenstand des
Kapitals , in: Walter Euchner/Alfred Schmidt (ds.), Kritik der politischen konomie heute. 100
Jahre Kapital , Francfort : EVA, 1968, pp. 60 et suiv.
43
Schttler
de lcole de Francfort, devint alors programmatique: Der strukturalistische Angriff auf

die Geschichte.9
Quant ceux qui ne partageaient pas ces prmisses et sorientaient vers un marxisme
plus orthodoxe (en gros : celui propag en RDA), leur refus sera tout aussi net : le
structuralisme , lpistmologie historique, Althusser, Foucault et toute cette
nbuleuse, tombera sous le verdict de lidalisme , donc dune variante de la
pense bourgeoise , donc combattre. Pour cela il suffisait dextraire quelques
citations de Bachelard,10 jusque-l inconnu en Allemagne (aucun livre traduit), afin de
prouver quil ntait pas marxiste. Certains iront mme et aprs-coup cest assez
cocasse (ou ironic au sens anglais du terme) jusqu mettre sur le mme plan le
nouvel althussrisme et lempirisme logique du Cercle de Vienne comme deux
variantes de lobjectivisme positiviste.11
Peut-tre faut-il rappeler ici que la pense franaise tait perue en Allemagne
essentiellement sous deux formes: dune part l existentialisme , qui peut-tre
cause de son succs littraire, thtral, cinmatographique etc. semblait constituer la
philosophie parisienne dominante; or on pouvait lui trouver assez facilement des
alliages, voire mme des racines allemandes. Et puis, dautre part, le positivisme ,
qui semblait reprsenter effectivement la pense occidentale , dmocratique (au
sens encore ambigu du terme, vingt ans aprs le nazisme) et paraissait non seulement
aux yeux des conservateurs, mais aussi des nomarxistes comme le dernier avatar des
Lumires , dans sa variante la plus technocratique et scientiste, menant tout droit de
Diderot et dHolbach Lnine - via Auguste Comte.12
Ceux qui osaient transgresser ces panneaux davertissement et saventurer dans ces
terres dangereuses taient videmment assez rares, et au dbut mme rarissimes.13
9
10
11
12
13
44
Alfred Schmidt, Der strukturalistische Angriff auf die Geschichte , in : id. (d.), Beitrge zur
marxistischen Erkenntnistheorie, Francfort : Suhrkamp, 1969, pp. 194265.
Cf. le petit recueil de textes dit par Dominique Lecourt en 1971 et traduit peu aprs en allemand :
Gaston Bachelard, Epistemologie. Ausgewhlte Texte, Francfort : Ullstein, 1974.
Cf. Wulf D. Hund (d.), Strukturalismus. Ideologie und Dogmengeschichte, Darmstadt : Luchterhand,
1973.
Sur le rapport polmique de lcole de Francfort au positivisme du 19e et au scientisme du 20e
sicle cf. la Dialektik der Aufklrung dAdorno et Horkheimer (1947), qui circulait dans les annes 60
en dition pirate (Raubdruck), ainsi que louvrage de Horkheimer clipse de la raison, Paris : Payot,
1974, dont la premire dition allemande, traduite par Alfred Schmidt, part en 1967 sous le titre
Kritik der instrumentellen Vernunft. Pour toute cette histoire voir le livre dcisif de Hans-Joachim
Dahms, Positivismusstreit. Die Auseinandersetzungen der Frankfurter Schule mit dem logischen
Positivismus, dem amerikanischen Pragmatismus und dem kritischen Rationalismus, Francfort :
Suhrkamp, 1998.
Citons quelques exemples : Helga Gallas (d.), Strukturalismus als interpretatives Verfahren,
Darmstadt : Luchterhand, 1972 ; Rodolphe Gasch, Die hybride Wissenschaft. Zur Mutation des
Wissenschaftsbegriffs bei Emile Durkheim und im Strukturalismus von Claude Lvi-Strauss,
Stuttgart : Metzler, 1973 ; Wolfgang Hagen, Die Schillerverehrung in der Sozialdemokratie. Zur
ideologischen Formation proletarischer Kulturpolitik vor 1914, Stuttgart : Metzler, 1977 ; Karl-Heinz
Ladeur, Rechtssubjekt und Rechtsstruktur. Versuch ber die Funktionsweise der Rechtssubjektivitt,
Giessen : Focus, 1978 ; Klaus-Michael Bogdal, Schaurige Bilder. Der Arbeiter im Blick des Brgers
am Beispiel des Naturalismus, Francfort : Syndikat, 1978 ; Jrgen Link, Die Struktur des Symbols in
SUR LA RCEPTION DE L PISTMOLOGIE FRANAISE EN ALLEMAGNE
Dautant plus quil manquait cruellement de traductions, mme en anglais, et quil

fallait ncessairement passer par les ditions originales.14 Ce nest donc pas un hasard si
la plupart de ceux qui lisaient et citaient les nouveaux travaux franais entretenaient des
liens privilgis avec le monde francophone, parfois mme avec le monde littraire ou
artistique franais, sil est vrai que la rflexion pistmologique, ce moment-l,
transite souvent par des champs para-universitaires. En France, il suffit de penser des
projets et des groupes thoriques comme Tel Quel, les Cahiers du Cinma ou
Cinthique. En Allemagne il faut surtout voquer une revue, aujourdhui un peu
oublie, paraissant Berlin-Ouest sous une couverture rouge brique: Alternative.
Zeitschrift fr Literatur und Diskussion, dirige par Hildegard Brenner. Il sagissait
effectivement dune revue de littrature, mais aussi de cinma et de thorie qui ds les
annes 1960 traduisait et prsentait les dbats franais, alors que les grands diteurs
taient encore sur leur rserve.15 Dans ce contexte signalons aussi une autre revue
faisant fonction de relais, au moins dans son domaine : Filmkritik. Parmi les jeunes
cinphiles et cinastes qui leur manire sy approprirent les dbats doutre-Rhin je
me souviens par exemple dune grande discussion thorique autour de Young Mr.
Lincoln de John Ford lance par les Cahiers du Cinma il y avait notamment :
Hartmut Bitomski, Harun Farocki, Wim Wenders et Hanns Zischler.16
Un deuxime rseau que lon pourrait voquer tait non pas install Berlin-Ouest
mais en province , au fin fond de la Ruhr. Cest en effet luniversit nouvelle de
Bochum17 que se sont retrouvs plusieurs groupes denseignants et dtudiants de
diffrentes disciplines qui ds 1968/70, au lieu de lire tel ou tel classique ou manuel
marxiste, ont commenc tudier , vaille que vaille, les textes dAlthusser, de
Foucault, mais aussi de Bachelard et de Canguilhem, au point sattirer la fois des
ennuis politiques puisque ctaient des auteurs encore non-admis dans la culture
de gauche allemande et mme acadmiques, puisque presque aucun professeur ne
connaissait et ne prenait donc au srieux ces penseurs parisiens .18 Aujourdhui, les
rpercussions de ces dbats acharns (un peu trop) peuvent se lire dans une bonne
dizaine douvrages et de thses de doctorats publis dans les annes suivantes, sans
parler de telle ou telle collection de livres ou dune revue comme KultuRRevolution.
14
15
16
17
18
der Sprache des Journalismus, Munich : Fink, 1978, etc. La prdominance des tudes littraires
reflte bien la prcocit de la rception des thories franaises dans ces disciplines ; cf. Jost
Schneider (d.), Methodengeschichte der Germanistik, Berlin : de Gruyter, 2009, not. pp. 89-107.
Si on regarde la chronologie des publications, on remarquera quune certaine rception et critique
allemande du discours philosophique franais, notamment de la part de Jrgen Habermas, ne
surviendra qu partir du moment o la plupart des textes fondateurs auront t traduits en anglais.
Alternative part de 1964 1982. Sur cette revue et le petit monde berlinois des annes 60 cf. le
rcit de Heinz Dieter Kittsteiner : Unverzichtbare Episode. Berlin 1967 , Zeitschrift fr
Ideengeschichte, 2 (2008), 4, pp. 3144. cot de Alternative citons aussi des revues comme
Das Argument , Sozialistische Politik (SOPO) ou Probleme des Klassenkampfs (PROKLA)
qui dans les annes 70 sengagrent galement dans le dbat.
Filmkritik part de 1957 1984.
Soulignons que dans les annes 1960-70 enseignaient Bochum des philosophes comme Hans
Blumenberg, Hermann Lbbe, Otto Pggeler, Karlfried Grnder, Kurt Flasch ainsi que lhistorienphilosophe Reinhart Koselleck.
Occasion de rendre hommage lhistorien Rudolf Vierhaus (1922-2011) qui en 1970 me permit
Bochum de passer loral de la Zwischenprfung sur lHistoire de la folie de Foucault.
45
Schttler
Zeitschrift fr angewandte Diskurstheorie, fonde en 1982 par Jrgen Link et qui

continue encore travailler dans cette direction.19
Jai parl de groupes au pluriel parce quil faut se rappeler, quand on voque cette
rception prcoce, quelle fut toujours traverse par dpres conflits politiques ,
thoriques et de personnes , ce qui compliquait terriblement les choses.20
Heureusement, ces divergences qui paraissaient si terribles premire vue se sont
estompes aprs quelques annes et notamment aprs la disparition du spectre
maoste21 pour faire place des alliances nouvelles et mme des collectifs de
pense novateurs: face lincomprhension tonnante des reprsentants de
lestablishment en sciences humaines qui ne cesse davoir peur de tel ou tel
tremblement de terre.22
Je marrte l: ces quelques notes navaient dautre but que desquisser rapidement,
et notamment pour les auditeurs franais, la situation initiale par rapport laquelle le
jeune philosophe Hans-Jrg Rheinberger dveloppa partir de la fin des annes
soixante une fascination particulire pour cette pistmologique historique si
franaise .
19
20
21
22
46
Cf. le site de la revue <http://zeitschrift.kulturrevolution.de> ainsi que la bibliographie tablie par Rolf
Parr et Matthias Thiele, Link(s). Eine Bibliographie zu den Konzepten Interdiskurs ,
Kollektivsymbolik und Normalismus sowie einigen weiteren Fluchtlinien, Heidelberg :
Synchron, 2005.
Voir la comptition politique entre les petites maisons ddition de gauche installes BerlinOuest : Wagenbach, Rotbuch, Oberbaum, Merve, VSA, etc.
Cf. Gerd Koenen, Das rote Jahrzehnt. Unsere kleine deutsche Kulturrevolution 1967-1977, Cologne :
Kiepenheuer & Witsch, 2001 ; Andreas Khn, Stalins Enkel, Maos Shne. Die Lebenswelt der
K-Gruppen in der Bundesrepublik der 70er Jahre, Francfort : Campus, 2005.
Cf. Peter Schttler, Nach der Angst. Was knnte bleiben vom Linguistic Turn ? , Internationales
Archiv fr die Sozialgeschichte der deutschen Literatur, 36 (2011), pp. 135151.
RHEINBERGERS TAKE ON HISTORICAL EPISTEMOLOGY
Henning Schmidgen
I would like to say something about the relationship between the work of Hans-Jrg
Rheinberger and the tradition of historical epistemology. That task would seem to be
straightforward since in 2007 Rheinberger published an introductory book on the topic:
Historische Epistemologie zur Einfhrung. The corresponding English version was
published in 2010 by Stanford University Press under the slightly modified title On
Historicizing Epistemology. An Essay.1
In this volumes introduction Rheinberger refers to what is described as the French
practice of using the term epistemology. As he explains, he does not use [this term]
as a synonym for a theory of knowledge (Erkenntnis) that inquires into what it is that
makes knowledge (Wissen) scientific, as was characteristic for the classical tradition,
especially in English-speaking countries. Rather, the concept is used here, following the
French practice, for reflecting on the historical conditions under which, and the means
with which, things are made into objects of knowledge. It focuses thus on the process of
generating scientific knowledge and the ways in which it is initiated and maintained.2
One of the interesting things about the following text is that it is not exclusively
focused on the French practice as exemplified by Bachelard, Canguilhem, Foucault
and/or Althusser. Rather, Rheinberger presents Ludwik Fleck and Edmund Husserl as
having contributed in important ways to historicizing epistemology. According to
Rheinberger the starting point of the entire project is even more remote. The book starts
with historical studies undertaken by scientists such as Ernst Mach and Emil du BoisReymond. Alluding to Althusser, one could say that Rheinberger traces back the
emergence of historical epistemology to the spontaneous historiography of scientists
in the 19th century.
This rather broad understanding of historical epistemology is underscored in
Rheinbergers most recent book entitled An Epistemology of the Concrete. Twentieth
Century Histories of Life. The first section entitled Historical Epistemology is dedicated
to Husserl, Fleck, Bachelard, and Canguilhem.3
In his earlier writings the focus was different. As we all know, Rheinberger entered
the international scene of the history of science as a molecular biologist who in his

1
2
3
Hans-Jrg Rheinberger, Historische Epistemologie zur Einfhrung, Hamburg: Junius, 2007, and id.,
On Historicizing Epistemology. An Essay, transl. David Fernbach, Stanford, CA: Stanford University
Press, 2010.
Rheinberger, On Historicizing Epistemology, p.2 (the emphasis is Rheinbergers).
Hans-Jrg Rheinberger, An Epistemology of the Concrete. Twentieth-Century Histories of Life,
Foreword by Tim Lenoir, Durham & London: Duke University Press, 2010, pp.11-48.
47

Schmidgen
investigations into the history of that field referred to and relied on the grammatological
philosophy of Jacques Derrida from his two papers on Experiment, Difference, and
Writing (1992) to his book on Synthesizing Proteins in the Test Tube (1997) and the
volume Iterationen that was published in the German language as a tribute to Derrida
(2005).4
No doubt that Rheinberger was well prepared for this rather unexpected combination
of fields of interest. On one side he was a trained laboratory biologist who had worked
since the mid-1970s at the Max Planck Institute for Molecular Genetics, and on the
other side he is a philosopher by training who with Hanns Zischler in the early 1970s
translated Derridas Grammatology into German.5
Given this twofold background, it doesnt come as a surprise that Rheinberger
describes his contributions to the history of the life sciences as a translation. This time,
however, it is not a question of translation from one language into another, e.g. from
French to German. Neither are we concerned with the deciphering of biochemical
transmission processes in the laboratory of molecular biology. Rather, Rheinberger
presents his book on synthesizing proteins as translating Derrida deeply into the
contemporary history of the empirical sciences.
Let me quote how Rheinberger himself describes this process in a recent paper: []
after a long decade of experimental work in molecular genetics [] I came back to my
earlier experience of reading and translating Derrida. This time, it resonated with my
attempt at an epistemology and history of experimentation in the empirical sciences, and
now, another translation adventure started: to make sense of Derrida in historical
science studies. As is well known, the natural sciences proper are virtually absent from
the oeuvre of Derrida as a subject of concern. And yet we have to situate Derridas work
with respect to science in a more general sense.6
Implicitly Rheinberger reminds us here of the fact that Derrida was an assistant to
Canguilhem in the early 1960s. In addition, one of Derridas earliest texts is the long
introduction to his translation of Husserls extremely evocative fragment on the
Ursprung der Geometrie.7 And it is precisely in this text by Husserl that the question of
writing surfaces when Husserl says: It is the important function of the written, the
documented linguistic expression, that it allows communication without immediate or

4
48
Hans-Jrg Rheinberger, Experiment, Difference, Writing: I. Tracing Protein Synthesis, and

Experiment, Difference, Writing: II. The Laboratory Production of Transfer RNA, Studies in
History and Philosophy of Science Part A 23 (2):305-331, and (3):389-422. See also Rheinberger,
Toward a History of Epistemic Things. Synthesizing Proteins in the Test Tube, Stanford, CA: Stanford
University Press, 1997, and id., Iterationen, Berlin: Merve, 2005.
Jacques Derrida, Grammatologie, transl. Hans-Jrg Rheinberger and Hanns Zischler, Frankfurt am
Main: Suhrkamp, 1974.
Hans-Jrg Rheinberger, Translating Derrida, CR: The New Centennial Review, 8/3 (2008): 175-187
(Special Issue: Paying Attention to Rodolphe Gasch), quote on p.180.
Edmund Husserl, Lorigine de la gomtrie, Traduction et Introduction par Jacques Derrida, Paris:
Presses Universitaires de France, 1962.
mediated personal allocution, that it is, so to speak, communication that has become
virtual.8
Against this background we might better understand Rheinbergers reference to what
he calls Derridean historical epistemology or his [i.e. Derridas] version of a
historical epistemology.9 It is a kind of historical epistemology that focuses on the
spaces of writing that had to be created for objects to be made into objects of empirical
knowledge under variable conditions.
There is more. In the context of his historical studies of modern biology Rheinberger
returns not only to Derrida. He also returns to Althusser, or so it seems at least to
readers of the German version of Synthesizing Proteins in the Test Tube.
One of the features of experimental practice that is highlighted in this book might be
called the eventality of experiments. In general terms Rheinberger describes
experimental systems as operating with a logic of differential reproduction. In his
understanding such systems are essentially open to the generation of surprise and
novelty.
The key notion in this respect is the unprecedented event, or as listed in the index
to the German version of the book: Ereignis, unvorwegnehmbares. I dont want to
suggest a kind of Rheinberger philology here, but in the case of the event there is an
attractive difference between the German and the English version.
Both editions connect the concept of unprecedented event to the notion of
conjuncture: Junctures or conjunctures, we read in the English version, come along
with unprecedented events and may lead to major rearrangements and re-combinations
between given partial spaces or representation of an experimental system.10 To
illustrate this Hans-Jrg invokes the notion of serendipity a few lines down.
In the German version of the book Rheinberger does pretty much the same but
inserts a further explanation concerning the term conjuncture. Althusser has called
historical conjunctures generally conjunctions, i.e. the aleatory encounter of elements
that, in part, are quite obvious, in part can not be anticipated on.11 Rheinberger then
continues by explaining that he, i.e. Althusser, deemed the notion of conjunction
necessary in order to think the openness of the world towards the event.12
Similar to his use of Derrida, this reference to Althusser can be seen as a return. In
fact, the first philosophical publications by Rheinberger were dealing with Althussers
materialistic and radically anti-dialectical epistemology. In 1975 he published in Das

8
9
10
11
12
Edmund Husserl, Der Ursprung der Geometrie, quoted after Rheinberger, Translating Derrida,
p.183.
Rheinberger, Translating Derrida, pp.182-183.
Rheinberger, Toward a History of Epistemic Things, p. 133
Rheinberger, Experimentalsysteme und epistemische Dinge. Eine Geschichte der Proteinsynthese im
Reagenzglas, Gttingen: Wallstein, 2001, S. 144.
Ibid.
49

Schmidgen
Argument, a Marxist philosophy journal, what seem to be the results of his

philosophical Master Thesis on the epistemological conceptions of Althusser: Die
erkenntnistheoretischen Auffassungen Althussers. One year later another article on
Althusser appeared in the leftist journal Sozialistische Politik, this time on Althussers
view concerning the relation between philosophy and the sciences. Whereas the first
text focuses on Pour Marx and Lire le capital, the second mainly deals with Althussers
course for scientists published in 1974 under the title Philosophy and the Spontaneous
Philosophy of Scientists.13
There is no mention of the event or conjunctures in these articles and the entire tone
of these early texts is rather unusual, at least from todays perspective. There is little
reference to history but much talk of political battles and class struggle. There are few
narrative elements and a lot of assertions, statements, and arguments. There is frequent
questioning of how to distinguish science and ideology but it is not clear if this
distinction in itself was meant to be scientific or ideological.
Despite this strangeness there are elements that seem familiar, or to be more precise
elements that invite us to not resist the temptation to understand them
anachronistically to read them in the light of what was written in the meantime
The first element is the analogy between the process of scientific practice on the one
side and the general process of production on the other. As Rheinberger shows,
Althusser draws this analogy by means of his famous notion of the appareil de pense
as well as his concept of the mode de production dtermin de connaissances.14
Defined as a historically constituted system that is grounded in natural and social
realities, and at the same time described as a system that combines an object on which it
works with the theoretical and technical means of production and historical relations
within which it produces, the appareil de pense seems not without resemblance to
what Rheinberger calls experimental systems.
The second element is a critical one. Rheinberger highlights the fact that Althusser,
in his discussion of Marx, is centering on a theory of reading and writing. However,
according to Rheinberger this theory remains rather immaterial contrary to Althussers
explicit interest in questions of history and materiality. From this perspective it seems
rather obvious why a Derridean philosophy of writing that highlights the exteriority of
writing should be of crucial interest.
The third element concerns the spontaneous philosophy of the scientist. The early
Rheinberger agrees with Althusser that philosophy should never intervene between
science and its own specific practice. However he does not seem fully satisfied with

13
14
50
Hans-Jrg Rheinberger, Die erkenntnistheoretischen Auffassungen Althussers, Das Argument

94/11-12 (1975): 922-951, and id., Philosophie und spontane Philosophie der Wissenschaftler: Zu
Althussers Verhltnisbestimmung von Philosophie und Wissenschaften, Sozialistische Politik 8/3
(1976): 95-102. See also Rheinberger, Erkenntnisproze, Wissenschaft und Philosophie, in Hans
Jrg Sandkhler (ed.), Betr.: Althusser: Kontroversen ber den Klassenkampf in der Theorie, Kln:
Pahl-Rugenstein, 1977, pp. 77-102.
See Rheinberger, Die erkenntnistheoretischen Auffassungen Althussers, p.931.
Althussers suggestion that philosophy should intervene where scientists move into
philosophical terrain. Instead he suggests conceiving of philosophy as a continuation of
science on a different theoretical level.
Against this background it perhaps becomes clear why Rheinberger has always
advocated a kind of history of science that remains open to a dialogue with scientists.
Perhaps one also understands better now why Rheinberger, as noted above, has taken
interest in the spontaneous historiography of scientists.
To sum up, I shall say that (a) Rheinbergers relation to the French tradition of
historical epistemology is deeply rooted in his intellectual and biographical itinerary. In
fact it goes way back to the early 1970s when there was no question whatsoever of
creating a Max Planck-Institute for the History of Science.
And (b) this relation was never simple, never a matter of applying ready made
theories and concepts to existing questions or problems. Rather, Rheinberger has made
his own creative use of historical epistemology. And since it is the use that decides on
the meaning, he has also contributed to transforming the tradition tied to the names of
Bachelard and Canguilhem as well as Derrida and Althusser.
51

LPISTMOLOGIE HISTORIQUE DANS LITINRAIRE INTELLECTUEL

1
DE GEORGES CANGUILHEM
Camille Limoges
Ce titre apparatra saugrenu plusieurs: pourquoi sinterroger sur la place de

lpistmologie historique dans litinraire de Canguilhem, puisque celle-ci aurait servi
de matrice son uvre qui toute entire en porterait la marque. Canguilhem nest-il pas,
dans le sillage de Gaston Bachelard, lincarnation mme de cette pistmologie la
franaise?
Ce fut en effet lopinion gnralement reue, conforte par deux fortes lectures de
luvre de Canguilhem, dcales dune quinzaine dannes, celle de Pierre Macherey
en 19642 et celle de Michel Foucault en 1978.3
Larticle de Pierre Macherey avait en effet prcocement et efficacement camp la
posture dun Canguilhem dont luvre frappe dabord par sa spcialisation, celle
dun pistmologue et historien des sciences.4
Or, on avait l affaire, en 1964, un Canguilhem court, un instantan presque, de
superbe dfinition certes, mais restreignant luvre une tranche de 20 annes
seulement: 1943-1963. Ainsi allait simposer, presque jusqu aujourdhui, lide que la
thse de mdecine de 1943, lEssai sur quelques problmes concernant le normal et le
pathologique tait apparue sui generis, sans antcdent dans luvre de Canguilhem.5
1
Texte rvis en octobre 2012 de la communication prononce Berlin le 10 dcembre 2010 dans le
cadre dun colloque organis sous lgide du Max-Planck-Institut fr Wissenschaftsgeschichte:
Epistemologie und Geschichte Von Bachelard und Canguilhem zur Wissenschaftsgeschichte von
heute. Je tiens ici remercier Madame Nathalie Queyroux responsable des archives et de la
bibliothque du Fonds Canguilhem conserves au CAPHS Paris, de mme que les membres de sa
trs comptente quipe de collaborateurs, pour la gnrosit de leur accueil et lexceptionnelle qualit
des services offerts. Leur apport a t essentiel la conduite de cette recherche.
Pierre Macherey, La philosophie des sciences de Georges Canguilhem. pistmologie et histoire des
sciences, La Pense, no 113, fvrier 1964, p. 50-74 (avec une Prsentation par Louis Althusser
aux p. 50-54).
Dabord parue en anglais en guise dIntroduction dans On the Normal and the Pathological,
traduction par Carolyn R. Fawcett, Dordrecht, Boston et Londres, D. Reidel Publishing Company,
1978, p. ix-xx; le texte original en franais a t publi avec quelques modifications sous le titre La
vie: lexprience et la science, dans la Revue de mtaphysique et de morale, vol. 90, no 1, 1985,
p. 3-14; il a t repris dans Michel Foucault, Dits et crits, d. Daniel Defert et Franois Ewald,
Paris: Gallimard, 1994, vol. IV, p. 763-776.
Les italiques sont miennes. Plusieurs articles avaient t publis sur des travaux de Canguilhem avant
celui de Macherey, sous forme de comptes rendus plus ou moins labors loccasion de la
publication de chacun de ses ouvrages, mais aucun navait port sur une pluralit de ses crits.
Ainsi, Franois Dagognet lui-mme, lecteur autoris de Canguilhem sil en est, pose-t-il lui aussi
lEssai de 1943, comme un commencement: Commencement apparemment limit mais incisif: M.
Canguilhem part dune conception de la mdecine (Une uvre en trois temps, Revue de
Mtaphysique et de Morale, vol. 90, no 1, 1985, p. 29-38, p. 29). Un article de Henning Schmidgen,
53
Limoges
On ignorait encore gnralement que celui-ci avait commenc publier dix-sept ans
plus tt, ds 1926,6 et bien sr on ne pouvait dj savoir en 1964 ce que rserveraient les
trois dcennies suivantes. Il manquait la trajectoire un amont mconnu et un aval non
encore advenu.
Larticle de Michel Foucault, La vie: lexprience et la science, mettait
contribution les travaux poursuivis par Canguilhem jusquen 1977, mais aucun de ceux
qui avaient prcd lEssai de 1943. Mais surtout, de cet article plus complexe que la
plupart de ses chos ne le donneraient penser, ses lecteurs ont gnralement retenu une
portion seulement, les premires pages caractrisant Canguilhem comme lauteur dune
uvre austre, volontairement bien dlimite, et soigneusement voue un domaine
particulier dans une histoire des sciences, qui de toute faon ne passe pas pour une
discipline grand spectacle.7
Il faudra attendre le Canguilhem avant Canguilhem de Jean-Franois Braunstein,
en 2000,8 puis louvrage de Dominique Lecourt en 2008,9 pour que soient restitus
Canguilhem ses dbuts et que se rvle lexceptionnelle richesse philosophique de prs
de deux dcennies de travaux prcdant la thse de mdecine. La publication rcente du
premier tome des uvres compltes de Canguilhem offre dsormais rassembles les
pices lappui de ces nouvelles lectures. 10 Quant au Canguilhem daprs 1964
9
10
54
Georges Canguilhem et les discours allemands est venu rcemment dmontrer la fcondit dune
interrogation sur certains pralables de la pense de Canguilhem au moment de la rdaction de sa
thse de mdecine (dans Anne Fagot-Largeault et al., Philosophie et mdecine. En hommage
Georges Canguilhem, Paris: Vrin, 2008, p. 49-62). Claude Debru aussi, ds 1998, avait dans Georges
Canguilhem et la rationalit du pathologique (Annales dhistoire et de philosophie du vivant, vol. 1,
no 1, 1998, p. 39-58) interrog partir dune analyse interne la gnalogie de lEssai dans ses relations
avec la tradition mdicale strasbourgeoise, relativis lapport de Kurt Goldstein et, juste titre, conclu
que Canguilhem vient dailleurs et va plus loin (p. 51).
Il est cet gard remarquable que les contributions au colloque tenu Paris en lhonneur de
Canguilhem en dcembre 1990 ( deux exceptions prs, celles de Jan Sebestik et dYves Schwartz qui
renvoient tous deux la communication de 1937 sur Descartes et la technique) faisaient totalement
limpasse sur la centaine dcrits publis par Canguilhem avant 1943. Cf. Georges Canguilhem.
Philosophe, historien des sciences. Actes du Colloque (6-7-8 dcembre 1990), Paris: Albin Michel,
1993. Depuis, dans La vie humaine. Anthropologie et biologie chez Georges Canguilhem (Paris:
Presses universitaires de France, 2002), sappuyant sur la communication de 1937 et celle de lanne
suivante, Activit technique et cration , Guillaume Leblanc a montr leur rle pivot dans la pense
de Canguilhem et comment sy jouent et sy articulent dj ses thses fondamentales, et chez lui
prennes, sur les relations de la technique la vie, et de la science la technique (cf. part. p. 146-159).
Les italiques sont miennes. La phrase de cet article de Foucault qui a le plus souvent retenu lattention
suit immdiatement le passage que je viens de reproduire: Mais tez Canguilhem et vous ne
comprendrez plus grand chose toute une srie de discussions qui ont eu lieu chez les marxistes
franais; vous ne saisissez pas non plus ce quil y a de spcifique chez des sociologues comme
Bourdieu, Castel, Passeron, et qui les marque si fortement dans le champ de la sociologie; vous
manquez tout un aspect du travail thorique fait chez les psychanalystes et en particulier chez les
lacaniens, etc. (p.3-4).
Dans le numro spcial de la Revue dhistoire des sciences consacr Georges Canguilhem et son
temps, tome 53, no 1, p. 9-26.
Georges Canguilhem, Paris: Presses universitaires de France, 2008, collection Que sais-je , no 3722.
Georges Canguilhem, uvres compltes. Tome I: crits philosophiques et politiques 1926-1939,
Paris: Vrin, 2011, 1033 pages. Cette dition comportera 6 volumes.
LPISTMOLOGIE HISTORIQUE DANS LITINRAIRE INTELLECTUEL DE GEORGES CANGUILHEM
(Macherey), et mme daprs 1978 (Foucault), les tomes venir des uvres compltes
rservent encore quelques rencontres inattendues.
En accord avec le titre que jai retenu, jentends montrer que les travaux
dpistmologie et dhistoire des sciences ont merg assez tardivement dans litinraire
intellectuel de Canguilhem, et que sils ont ensuite intensment occup sa rflexion
pendant prs de deux dcennies, cest sans suspension de ses proccupations plus
anciennes, surtout philosophiques, qui de nouveau requerront largement son attention au
cours des deux dernires dcennies de sa vie, compter de 1975 environ.
Lpistmologie: une thmatisation tardive chez Canguilhem

En 1991, loccasion dun tmoignage sur Jean Hippolyte, Canguilhem rapportait que
ds ses annes lcole Normale Suprieure11 son ami portait un intrt certain la
philosophie des sciences, une poque o le mot dpistmologie ntait pas encore
banalis dans les amphithtres.12 Cet intrt ntait pas alors partag par Canguilhem.
Ses emprunts la bibliothque de lcole Normale Suprieure, comme ses propres
crits dans les annes 1920, ne traduisent aucune attention privilgie la philosophie
des sciences ou lpistmologie (et encore moins lhistoire des sciences) au-del de
ce quexigeaient la prparation des examens et concours (baccalaurat, licence,
agrgation), cest--dire fort peu de choses, comme plus tard, de 1929 1940, le
programme denseignement et la prparation de ses leons de philosophie dans divers
lyces de province.
De fait, dans la centaine de textes quil publie de 1926 1939 on ne relve que deux
mentions du mot pistmologie (et aucune pour lpithte). Chaque fois, il sagit
seulement dinformer le lecteur sur la signification du terme (dont lusage ntait donc
pas alors effectivement trs courant en France). 13 Il nen ira pas autrement pendant
encore longtemps. Mme dans son premier ouvrage majeur, lEssai de 1943, on trouve
bien lpithte pistmologique, trois fois, mais jamais le substantif. 14 Dans La
Connaissance de la vie, le premier des recueils composites de Canguilhem, publi en
1952 et qui servirait de thse complmentaire lors de la soutenance pour le doctorat s
lettres trois ans plus tard, on repre une seule occurrence du substantif et cinq de
lpithte. Dans aucun cas lusage de ces termes ne fait lobjet dun dveloppement
11
12
13
14
Hyppolite y tait entr en 1925, Canguilhem lanne prcdente.

Tmoignage sur Jean Hyppolite, Bulletin de la Socit des Amis de l'cole normale suprieure,
no 186 (dcembre 1991), p. 20-23. Le terme ntait cependant pas inconnu chez des scientifiques et
philosophes franais de la gnration prcdente, comme la relev par exemple Anastasios Brenner
(Les origines franaises de la philosophie des sciences, Paris : Presses universitaires de France, 2003).
Il le fait dabord dans un compte rendu sur deux ouvrages (Sur la philosophie contemporaine.
H. Srouya, Initiation la philosophie contemporaine (La Renaissance du Livre). J. Benrubi, Les
sources et les courants de la philosophie contemporaine en France (Alcan), Europe, 15 mars 1934,
no 135), puis dans une note de bas de page du Trait de logique et de morale de 1939.
Il sagit chaque fois de mentions en passant, sans plus: des donnes de la psychopathologie au
mme titre de dignit pistmologique que les faits observs chez les normaux ; La polarit
dynamique de la vie et la normativit qui la traduit expliquent un fait pistmologique; une
anecdote de grande porte pistmologique.
55
Limoges
propre. Il en va de mme dans La Formation du concept de rflexe, la thse principale

de 1955, qui use une seule fois du substantif et une autre de lpithte. Dans cet ouvrage
o Canguilhem savre dj trs videmment un subtil pistmologue, lpistmologie
nest pas non plus comme telle thmatise; sa mise en uvre reste silencieuse.
En fait, au cours de ses premires 36 annes de publication, de 1926 1962,
Canguilhem use des termes de cette famille 20 fois seulement, ce qui reprsente moins
dune occurrence annuelle; au cours de la priode terminale de sa production, de 1969
1992, pour 23 annes, on relvera 54 occurrences, donc un peu plus de 2 par anne.
Mais de 1963 1968, son intrt sonder explicitement lpistmologie semble avoir
vritablement explos: pour ces 6 annes seulement, il accumule 56 mentions.15 Cest
dailleurs compter de cette priode que, de faon durable, simposera lidentification
de Canguilhem lpistmologie que lon dira historique.
Alors quau cours des cinq annes prcdentes on trouve seulement 5 mentions du
substantif ou de lpithte, soudainement en 1963, pour cette seule anne, leur nombre
est port 25. Or, ce jaillissement ne tient pas une ncessit interne la conduite par
Canguilhem de ses propres travaux de recherche. Cest la mort de Gaston Bachelard,
survenue lautomne 1962, qui suscite cette floraison concentre de propos sur
lpistmologie, dans la srie dcrits que Canguilhem publie immdiatement en rafale
propos de celui qui fut son directeur de thse au dbut des annes 1950 et son
prdcesseur dans la chaire de la Sorbonne et la direction de lInstitut dhistoire des
sciences de la rue du Four.16
On est donc justifi de dire que si lpistmologie est chez Canguilhem lobjet dune
thmatisation forte, cest nanmoins le fruit dune opration relativement tardive (il a
dj 59 ans en 1963) qui seffectue dabord et surtout dans ses crits sur Bachelard et
non par priorit rflexive sur ses propres travaux.
15
16
56
Ces occurrences sont prises la date de premire publication de chaque crit. On ne retient ici pour la
compilation que des crits publis. Si lon tenait aussi compte dautres prestations, comme par
exemple des confrences radiodiffuses que lon trouvera au tome IV de ses uvres compltes ,
certaines frquences annuelles seraient un peu plus leves, mais les scansions temporelles resteraient
inchanges.
Trois occurrences des vocables qui nous intressent avaient prcd la prolifration de 1963, en 1957;
elles tenaient aussi directement lexamen de luvre de Bachelard, dans la contribution au volume
dHommage Gaston Bachelard (Paris : Presses universitaires de France, 1957) auquel Canguilhem
avait particip: Une pistmologie concordataire. Il sagit du tout premier article consacr
Bachelard par Canguilhem.
Figure 1: volution du nombre des occurrences des vocables drivs du terme

pistmologie dans les publications de Georges Canguilhem
Le soin mticuleux apport par Canguilhem la caractrisation de lpistmologie

bachelardienne, la prsentation si globalement logieuse et positive quil en a donne
de multiples reprises17 ont fait que lexpression de certaines rserves, assez discrte, il
est vrai, reste gnralement ignore. 18 Cest dans ce contexte que ltiquette
pistmologie historique a fait amalgame et en est venue fonctionner comme une
sorte dappellation contrle applicable tout autant luvre de Canguilhem qu celle
de Bachelard.19
17
18
19
Lors dune allocution prononce en avril 1963 devant les membres du Groupe franais dhistoire des
sciences, Canguilhem navait pas craint daffirmer: Bachelard est une grande figure, selon moi la
plus grande figure de lpistmologie franaise au XXe sicle, une haute et noble figure de
philosophe. On pouvait dautant plus tenir cette valuation pour sincre quil avait aussi dclar: Je
ne dirai pas que Gaston Bachelard a t un minent historien des sciences, car ce serait travestir le
sens et la porte de son uvre (Fonds Canguilhem, CAPHS, GC.31.5.7).
Bien que Dominique Lecourt ait rcemment, dans son Georges Canguilhem, consacr un
dveloppement trs clair la conception fort diffrente que se faisaient Bachelard et Canguilhem de la
nature de la philosophie et de ses rapports la science, ce qui nest pas peu. Voir aussi, sur les carts
et les similitudes de leurs pratiques pistmologiques effectives: Jean-Franois Braunstein,
Bachelard, Canguilhem, Foucault. Le style franais en pistmologie, dans Pierre Wagner (dir.),
Les philosophes et la science, Paris: Gallimard, Folio/Essais, 2002, p. 920-963.
On doit noter que la locution pistmologie historique apparat trs rarement dans les publications
de Canguilhem (4 occurrences au total). On la trouve dabord dans lun de ses articles de 1963 sur
Gaston Bachelard, LHistoire des sciences dans luvre pistmologique de Gaston Bachelard. Il
lutilise dailleurs dans ses crits de faon gnrique, lappliquant par exemple aux travaux de Thomas
Kuhn et il nen fait pas la marque dune forme spcifiquement bachelardienne (ou canguilhemienne)
de pratique de lpistmologie. Pierre-Olivier Mthot, dans un article sous presse, a rcemment
explor cette question: On the genealogy of concepts and experimental practices: Rethinking
Georges Canguilhems historical epistemology , Studies in History and Philosophy of Science,
2012. Voir aussi dans le prsent volume la contribution de Mthot (p. 117-144). Cest sans
doute dabord lusage de la locution par Dominique Lecourt dans le titre de son ouvrage sur
Lpistmologie historique de Gaston Bachelard (Paris: Vrin, 1969) qui a donn lappellation une
signification spcifique et qui en a dabord favoris la diffusion que lon sait.
57
Limoges
Or les proccupations originaires de Canguilhem avaient t tout autres que celles

explores dans les ouvrages de Bachelard. Les crits davant-guerre le prouvent
incontestablement. 20 Et lEssai de 1943, avec sa problmatique centrale de la
normativit du vivant, sorte de bute de la philosophie biologique de Canguilhem, ne
devait rien Bachelard, quoi quon ait pu en crire.
Canguilhem a voulu lui-mme le confirmer dans une lettre de 1987 adresse au
chercheur nerlandais Kees Hertogh, qui lui avait communiqu sa thse de doctorat:
Je nai connu personnellement Bachelard quaprs la soutenance de mdecine, en
1943. Ma conception normative de la vie ne doit rien la lecture des uvres de
Bachelard, pourtant connues de moi, la fois par got, et par obligation
professionnelle. [] Je fais seulement des rserves sur la parent, expose par
vous, de ma conception de la vie et de son tude dune part, et de ma
conception pistmologique de lhistoire des sciences dautre part.21
Canguilhem est ici parfaitement clair: llaboration de sa philosophie biologique ne

doit rien Bachelard et nest en aucune faon gnalogiquement relie sa conception
des relations de lpistmologie et de lhistoire des sciences.22
Ces relations, Canguilhem les explorera beaucoup plus tard, vingt ans plus tard
comme je lai dj signal, compter de 1963 surtout.
Mais, dira-t-on, Canguilhem au moment o il sengage dans sa thse de mdecine
nest-il pas dj fru dpistmologie, puisque co-auteur en 1939 dun Trait de logique
et de morale, alors que logique en France tait encore un terme peu prs synonyme
de philosophie des sciences ou dpistmologie? Oui, mais justement, dans la rdaction
de cet ouvrage en collaboration avec son ami Camille Planet, Canguilhem nest pas
lauteur premier l o on lattendrait. En effet, dans une lettre doctobre 1984 au
20
21
22
58
Jean-Franois Braunstein la dailleurs soulign dj: Canguilhem ne rencontre que fort tard dans
son uvre linfluence de Bachelard: ses motifs initiaux sont sans doute plus chercher dans des
engagements thiques de sa jeunesse, tays ensuite sur la mdecine, que dans lhistoire des sciences
(dans son recueil de textes: Lhistoire des sciences. Mthodes, styles et controverses, Paris : Vrin,
2008, p. 15). Dans une thse de doctorat soutenue en dcembre 2010, Xavier Roth a analys avec
prcision les moments de lvolution de la pense de Canguilhem de 1926 1939: Georges
Canguilhem et lcole franaise de lactivit. Juger, Agir (1926-1939); un livre de Roth, rdig sur
la base de cette thse, paratra sous peu Paris aux ditions Vrin.
Brouillon de lettre de rponse de Canguilhem Hertogh. Cette lettre tait insre dans la thse de
Hertogh (cote CAN 274 dans la bibliothque de Canguilhem conserve au CAPHS). Cette thse a
t publie: Cornelius Marinus Petrus Maria Hertogh, Bachelard en Canguilhem. Epistemologische
discontinuteit en het medisch normbegrip, Amsterdam, VU Uitgeverij, 1986.
Comme jai essay de le montrer dans une communication encore indite, sur la gense de lEssai de
1943, prsente en juin 2010 lors dun colloque sur La philosophie en France en 1943. uvres et
enjeux au moment de la guerre, la philosophie biologique de Canguilhem senracine en effet plutt
dans une analyse qui slabore compter de 1937-1938 sur ce quil appelle linitiative de la
technique dans les exigences du vivant et sert de fondement une problmatique travaille surtout
avec les outils conceptuels issus de la philosophie allemande contemporaine des valeurs.
chercheur italien Giuseppe Sertoli qui venait de consacrer un travail son uvre,23
Canguilhem crit:
Votre prsentation de mes recherches et de leur expos est historique,
lgitimement. [] Vous avez pris comme point de dpart de votre historique le
Trait de Logique et de Morale. [] Nous avons24 un jour compar les leons
dactylographies que nous remettions nos lves et dcid de les runir aprs
rvision systmatique. Planet a rdig les leons de philosophie des sciences. Jai
rdig les leons de morale.25
Aprs cinq dcennies de clbrations de Canguilhem grand matre de lpistmologie

historique, cela peut videmment tonner. Cest, encore une fois, que contrairement ce
que nous pensions savoir, pistmologue, en 1939 il ne ltait pas encore, pas plus quil
ne le sera dj en 1943.
Mais en 1955, La formation du concept de rflexe, ouvrage qui ressortit sans conteste
lhistoire des sciences, en dpit de la raret dj mentionne de son recours des
vocables caractre pistmologique, savre de part en part un remarquable exercice
pistmologique: il sy agit bien de la reconstitution de la gnalogie dun concept et de
la critique des interprtations qui en ont t donnes. Canguilhem historien des sciences
apparat ainsi en 1955 un pistmologue en acte. Mais justement, historien des sciences
quand lest-il devenu? Il faut se questionner aussi sur lavnement et sur la courbe
dexercice de lhistoire des sciences dans luvre de Canguilhem.
Lhistoire des sciences, exigence du mtier

Certes, on ne saurait dnier Canguilhem sa qualit dhistorien des sciences.
Rcipiendaire de la prestigieuse Mdaille Sarton de la History of Science Society,
directeur de lInstitut dhistoire des sciences et des techniques pendant plus de 15 ans,
auteur de recueils reconnus canoniques par nombre dhistoriens des sciences,
gnralement associ dans la littrature quelque chose comme une cole franaise de
lpistmologie historique, praticien insigne de lhistoire pistmologique, Canguilhem
a laiss une marque forte sur la discipline. Tout cela est indniable.
Mais on tend ignorer que la pratique de lhistoire des sciences, elle aussi, est tard
venue dans son uvre.
Aucun des articles antrieurs la thse de 1943 nest assimilable un travail
dhistorien des sciences. Cette thse, elle non plus navait pas t conue comme un
23
24
25
Epistemologia e storia delle scienze in Georges Canguilhem, Nuova Corrente, vol. 30, no 90-91
(1983), p. 101-171.
Il sagit bien sr ici de Canguilhem et de son ami Planet alors professeur de philosophie au Lyce
Thiers Marseille, alors que Canguilhem enseignait Toulouse. Tous deux taient natifs de
Castelnaudary.
Les italiques sont miennes. Le document que je cite avait t insr dans le fascicule de la revue
Nuova Corrente cot CAN 1374 dans la Bibliothque de Canguilhem conserve au CAPHS.
59
Limoges
ouvrage dhistoire des sciences.26 Sa vise en tait une de philosophie mdicale ou de

philosophie biologique, selon la locution dont il fera plusieurs fois usage dans La
connaissance de la vie (et dans son article majeur de 1947, Note sur la situation faite
en France la philosophie biologique).27
De fait, avant 1955, lanne de la soutenance des thses pour le doctorat s lettres
puis de la prise de poste la direction de lInstitut dhistoire des sciences et des
techniques de lUniversit de Paris, il ntait venu personne et dabord pas
Canguilhem lui-mme lide de le dsigner comme un historien des sciences, mme
en herbe.28
Les crits indits datant de limmdiat aprs-guerre et conservs dans ses archives
personnelles ne prsentent que de rares traces dintrt pour lhistoire des sciences et
aucun exercice proprement historique. Il existe une note brve, date de mars 1945,
intitule Histoire de la philosophie et histoire des sciences, 29 qui malgr son titre
porte essentiellement sur lhistoire de la philosophie. Il existe aussi quelques feuillets de
notes pour une leon donne dans le cadre de son enseignement Strasbourg (Sur
lhistoire de la Science 28 mars 1946).30 Il sagit dun texte de rflexions critiques sur
la pratique de lhistoire et sur celle de lhistoire des sciences, appuyes notamment sur
une lecture des tudes galilennes dAlexandre Koyr (sans mention de Bachelard).31
Cest dailleurs aussi encore essentiellement sur ces tudes de Koyr que sappuie en
1949 une confrence (publie seulement sous forme de polycopi, circulation limite
et donc peu connue), prononce par Canguilhem dans le cadre de ses fonctions
dInspecteur gnral devant une assemble de professeurs de philosophie et de
professeurs de sciences physiques des lyces, sur le Rle de lhistoire des sciences
dans la philosophie des sciences: Ltablissement des faits fondamentaux de la
dynamique. Cette confrence de 1949 et larticle de 1946 sur la thorie cellulaire,
repris dans La Connaissance de la vie, sont les deux premiers articles publis par
Canguilhem, et les seuls avant 1955, concerner directement lhistoire des sciences.
26
27
28
29
30
31
60
Malgr lintrt des analyses historiques que recle lEssai (sur la pense de Broussais ou celle de
Claude Bernard), comme on en trouve par exemple dans des ouvrages dmile Meyerson ou de Lon
Brunschvicg qui nont jamais t classs comme historiens des sciences. Les lecteurs de lpoque
dailleurs ne sy sont pas tromps, comme en tmoignent les comptes rendus contemporains. Ainsi,
Pierre Brunet, dans une revue toute nouvelle voue lhistoire des sciences commence-t-il son compte
rendu par les mots: Pour ntre pas une tude historique, ce travail dont toute la premire partie est
centre autour des noms de Comte et de Cl. Bernard, est susceptible par l (et en raison aussi de
certaines remarques plus dissmines) dintresser les historiens des sciences (Revue dhistoire des
sciences, vol. 1, no1, 1947, p. 90-91).
Revue de mtaphysique et de morale, vol. 52, nos 3-4, 1947, p. 322-332.
Bien quun auteur ait pu tourdiment prtendre que Canguilhem stait engag dans des tudes de
mdecine lautomne 1936 afin de parfaire ses connaissances en histoire des sciences! Cet t une
bien trange dcision.
Fonds Canguilhem, CAPHS, GC.11.3.10.
Fonds Canguilhem, CAPHS, GC.15.3.3 ; fol. 44-49.
On retrouve dans la premire partie de larticle de 1946 sur La thorie cellulaire des traces des
considrations thoriques esquisses dans ces notes.
Il est notable que, dans La Connaissance de la vie (1952), recueil que Canguilhem a
organis en trois parties, Mthode, Histoire, Philosophie, cet article de 1946
constitue lui seul la deuxime partie. Il est remarquable aussi, bien qu notre
connaissance personne ne lait relev, que cet article paru initialement dans une
publication de la Facult des Lettres de Strasbourg, intitul La thorie cellulaire,
portait aussi alors un sous-titre qui le dsignait comme un texte de rflexion thorique
plutt que dhistoire, sous-titre que Canguilhem fit disparatre en 1952: Du sens et de
la valeur des thories scientifiques. De fait, en 1946, dans la table des matires de la
revue strasbourgeoise, cet article avait t class dans une section intitule Philosophie
des sciences, de sorte que le seul article qui donnerait une partie de La connaissance
de la vie un caractre franchement historique apparat lui-mme dune nature
quivoque.32
Au-del de ces deux seuls crits sur la thorie cellulaire (1946), puis sur la physique
galilenne (1949), il faut attendre 1955, et la thse sur la formation du concept de
rflexe pour que sengage vraiment chez Canguilhem une dmarche continue de
recherches en histoire des sciences.33
En effet, soudainement, compter de lanne suivante, samorce une avalanche de
prestations de cette nature. Quon en juge: dabord une demi-douzaine dmissions
radiophoniques caractre nettement historique, sur Descartes, La Mettrie, Pinel
(1956), sur Lamarck et Darwin, sur Claude Bernard (1957), sur lhistoire de la
physiologie au 19e sicle (1964). De 1957 1965, se multiplient les comptes rendus
douvrages sur lhistoire de la physiologie, sur le premier tome de lHistoire gnrale
des sciences publie sous la direction de Ren Taton et sur lHistoire de la science
dirige par Maurice Daumas, mais aussi sur Darwin, Claude Bernard, Vsale,
Boerhaave, William Harvey, Johannes Mller ou tienne Geoffroy Saint-Hilaire.
Au cours de la dcennie de 1956 1966 paraissent aussi une cadence acclre de
nombreux articles nettement historiques, par exemple, sur Fontenelle, sur La
pathologie et la physiologie de la thyrode au XIXe sicle, sur lhistoire de la
physiologie dans lHistoire gnrale des sciences de Ren Taton et aussi dans le Trait
de physiologie de Charles Kayser, sur Auguste Comte et la pense mdicale de lcole
de Montpellier, sur le concept de rflexe au XIXe sicle, sur Galile, sur Vsale et
Copernic, sur Claude Bernard et Bichat, sur la psychologie humaine et animale chez
Darwin, et avec des membres de son sminaire de lInstitut dhistoire des sciences, la
longue tude historique Du dveloppement lvolution au XIXe sicle.
32
33
Dans la version publie en 1946, Canguilhem avait ddi cet article Marc Klein son collgue
lUniversit de Strasbourg. Ainsi quil le souligne, la prparation de son article lui avait t
grandement facilite par un travail de Klein, Histoire des origines de la thorie cellulaire qui se
prsentait explicitement comme une contribution historique. Louvrage de Klein avait t publi chez
Hermann dans la collection Actualits scientifiques et industrielles en 1936.
Canguilhem avait aussi sign dans la Revue philosophique, cette mme anne 1955, un long essai sur
le Descartes selon lordre des raisons de Martial Gueroult, Organisme et modles mcaniques.
Rflexions sur la biologie cartsienne. Il sagit dune exgse de nature davantage philosophique
quhistorique.
61
Limoges
De cette dcennie trs fertile, plusieurs de ces crits seront intgrs au recueil qui
voit le jour en 1968, les tudes dhistoire et de philosophie des sciences. Cet ouvrage
incorporera aussi plusieurs des articles que Canguilhem avait publis depuis 1963 sur
Gaston Bachelard. Quatre ans aprs larticle de Macherey, la parution de ces tudes
viendrait ainsi en quelque sorte affirmer la figure de Canguilhem la fois historien des
sciences et pistmologue.34
Or jusquen 1955, rien ne pouvait laisser prsager chez Canguilhem une si profuse
production en histoire des sciences. Le ressort de son nouveau rgime de recherches et
de publications tient la haute conception quil se faisait de ce quexigeaient de leur
titulaire ses nouvelles fonctions de professeur dHistoire et de Philosophie des sciences
la Sorbonne et de directeur de lInstitut dhistoire des sciences. Elles lui craient
lobligation et aussi lui multipliaient les occasions dassumer trs intensivement un
rle nouveau.
De ceci, on retrouve comme lcho dans une lettre Michel Deguy souvent cite,
date de dcembre 1990:
Il ne mest pas possible, mon ge, de faire autrement que jai toujours fait, cest-dire de considrer ce quon appelle mon uvre comme autre chose que la trace
de mon mtier.35
Ce mtier qui ses yeux ouvre ou mme impose une voie, et qui exige quon la
jalonne de travaux pertinents faute desquels on dmriterait, cest bien sr son mtier de
professeur. Canguilhem a toujours t parfaitement clair l-dessus, lenseignement est
un mtier, son mtier; qui le choisit encourt des obligations. Non quil ait rpugn de
telles responsabilits, bien sr. Au contraire, aprs les sept annes passes lInspection
gnrale, ces responsabilits il les a souhaites, voulues, et on peut penser que
lorientation donne sa thse visait laccession au type de fonctions qui lui seront
dvolues la succession de Gaston Bachelard compter de lautomne 1955. Il les
assumera intensment durant quelque dix-sept annes, de 1955 lautomne de 1971.
Les deux graphes ci-dessous, qui traduisent nettement la venue tardive de lhistoire
des sciences dans litinraire intellectuel de Canguilhem, vrifient aussi que cest bien
au cours de la priode o il assume ses fonctions professorales la Sorbonne et
lInstitut dhistoire des sciences que le nombre des publications dans ce domaine crot le
plus rapidement.36
34
35
36
62
Dautres facteurs ont contribu la diffusion, en France et ltranger, de cette reprsentation de

luvre et de son auteur: la rfrence forte Canguilhem en 1965 chez les auteurs de la vague
althussrienne (Pour Marx et Lire le Capital) et chez ceux des Cahiers pour lanalyse ( compter de
janvier 1966), enfin, la large diffusion et les traductions des deux premiers livres de Dominique
Lecourt (Lpistmologie historique de Gaston Bachelard, op. cit., qui comporte un avant-propos de
Canguilhem, et Pour une critique de lpistmologie, Paris : Maspro, 1972, dans lequel un chapitre
lui est consacr).
Dans Georges Canguilhem. Philosophe, historien des sciences. Actes du Colloque (6-7-8 dcembre
1990). Paris : Albin Michel, 1993, p. 324.
Ont t ici pris en compte dans la ralisation des graphes les articles, brochures, etc. et les comptes
rendus de plus dune page; la compilation en a t effectue lanne de leur premire publication.
Figure 2: volution annuelle du nombre cumulatif des publications de Georges

Canguilhem selon quatre champs
Figure 3: volution du nombre cumulatif des publications de Georges Canguilhem

selon quatre champs et selon les phases de sa carrire37
37
Une part de subjectivit peut affecter lassignation de certains titres lun ou lautre champ, mais sans
altrer significativement lallure de chacune des quatre courbes.
Les priodes indiques sous la ligne dabscisse correspondent aux phases de la carrire de
Canguilhem: 1926-1936: Canguilhem normalien puis professeur de lyce; 1936-1947: Canguilhem
tudiant en mdecine Toulouse puis Clermont-Ferrand, professeur de khgne au Lyce Fermat de
Toulouse; enseignant la Facult des Lettres de Strasbourg, Clermont-Ferrand puis Strasbourg;
1948-1955: Canguilhem Inspecteur gnral de lenseignement de la philosophie au secondaire et en
rdaction de ses thses du doctorat s lettres; 1955-1972: Canguilhem professeur la Sorbonne et
directeur de lInstitut de la rue du Four; 1973-1995: les annes de la retraite.
63
Limoges
Ces distributions graphiques mettent galement en vidence une certaine similitude

dallure entre la courbe des crits dhistoire des sciences et celle des crits
dpistmologie; la croissance de cette dernire prcde lgrement lautre, de sorte
quau plan intellectuel lpistmologie aurait en quelque sorte servi de moteur ses
intrts nouveaux.
Les titres en histoire des sciences reprsentent un peu plus de 15% de la production
totale de Canguilhem; lpistmologie environ 27%. Les publications philosophiques
sans caractre proprement pistmologique, qui sont les plus nombreuses (un peu plus
de 30% du total), sont aussi celles dont labondance a t la plus prcoce et dont la
croissance du nombre reste la plus uniforme (avec une acclration au cours des annes
de la retraite, alors que les travaux dhistoire des sciences subissent au contraire une
dclration).38
Enfin, autre leon de ces graphes, source dtonnement peut-tre pour les lecteurs qui
sen sont jusquici tenus aux uvres dj consacres, ce sont surtout les articles
caractre politique qui dominent la premire priode de ses travaux publis, pour
ensuite stagner, bien quils ne soient pas totalement absents de la suite de luvre.39
On doit se rendre lvidence: malgr la hauteur de la rputation dpistmologue et
dhistorien des sciences de Canguilhem, le projet qui fut le sien et la porte de ses
accomplissements sont loin de sy limiter; de fait, lanalyse de luvre montre que la
pousse centrale de son itinraire intellectuel vient dailleurs, de son projet proprement
38
39
64
Rappelant que quantit nest pas qualit, on pourrait, au regard de lexcellence reconnue des crits
dhistoire des sciences et dpistmologie de Canguilhem, soulever la question de la signification et
du poids intellectuels des crits rassembls dans les graphes sous la rubrique Philosophie (hors
pistmologie). La question se pose mais ne peut tre vide ici. Toutefois, les lecteurs un peu
familiers avec son uvre auront dj not que ces autres crits ne le cdent en rien aux travaux
dhistoire des sciences et dpistmologie au plan de lintrt, de la novation intellectuelle et de la
qualit. On peut dj le vrifier pour ce qui est des crits antrieurs 1939, maintenant que le tome I
des uvres compltes est disponible. On pourra le faire encore plus compltement la lecture des
tomes IV et V qui incluront plusieurs dizaines de ces crits inconnus ou mconnus. Mais il y a dj
ample matire vrification en se rapportant pour la priode postrieure 1939 un chantillon qui
pourrait comprendre des crits assez aisment accessibles, comme par exemple le substantiel essai sur
le Leibniz et Spinoza de Georges Friedmann (1946), larticle fondamental Milieu et normes de
lhomme au travail (1947), la Note sur la situation faite en France la philosophie biologique
(1947), larticle critique Hegel en France (1949), les Rflexions sur la cration artistique selon
Alain (1952), lessai sur le Descartes de Gueroult (1955), Exprience et aventure (1957), Mort de
lhomme ou puisement du cogito ? (1967), le cours sur laction (1967), les interventions
philosophiques Qualit de la vie, dignit de la mort (1976), Le droit la mort (1977), Quest-ce
quun philosophe en France aujourdhui ? (1991), etc.
Aussi le premier tome des uvres compltes de Georges Canguilhem a-t-il reu pour titre : crits
philosophiques et politiques 1926-1939. On doit distinguer trois plages privilgies dactivit de
publication politique chez Canguilhem: 1926-1936 surtout o la cause pacifiste occupe le devant de
ses proccupations, puis 1944-1946, au moment de la Libration, enfin 1958-1960, au moment de la
Guerre dAlgrie et des troubles marquant la fin de la IVe Rpublique. Il nest pas dans mon propos de
commenter ici ce segment politique des contributions de Canguilhem; jaurai loccasion dy revenir
ailleurs. Il a nanmoins paru intressant de donner aussi la courbe relative ce champ dactivit, de
faon encourager une interprtation mieux pondre de lvolution des autres types de
proccupations intellectuelles de Canguilhem.
philosophique, mu au cours des annes de sa premire maturit en philosophie

biologique, en philosophie du vivant et de la vie, fortement modul vers sa mi-temps et
pour prs de deux dcennies, par les exigences propres aux lieux do il exerait son
mtier de professeur.40
La redcouverte lt 2010 dun entretien de 1972, publi peu aprs son dpart la
retraite, mais oubli ou pass inaperu, probablement parce que paru dans une revue
professionnelle pour mdecins rarement frquente par philosophes et historiens, est
venue conforter cette lecture de la dynamique de la pense canguilhemienne.
Canguilhem y dclare en effet:
Je ne suis pas vritablement un historien des sciences, je suis en fait un professeur
de philosophie qui sintresse un certain nombre de questions, qui sont les
rapports entre la philosophie et la science et en particulier celle de la fabrication,
de la naissance, de limportation et de lexportation dun certain nombre de
concepts interprtatifs de fonctions biologiques.41
Il est clairant qualors questionn sur ses travaux historiques, Canguilhem rponde
en affirmant la priorit de son mtier tel quil lentend, le mtier de professeur de
philosophie. Et dans ce tmoignage pris dans le vif de loralit, ce quil place au foyer
de son activit de philosophe notons-le ce nest pas la philosophie des sciences,
mais les rapports entre la philosophie et la science, ce qui est tout autre chose et
renvoie dans son histoire intellectuelle personnelle la continuit dune problmatique,
ne dans les annes 1930 dun double jeu dinterrogations, dune part sur la valeur de la
vrit au regard des autres valeurs42 et, dautre part, sur la prcarit cratrice du vivant,
perspective engage avec larticle dcisif de 1937 sur Descartes et la technique, et
dont la richesse se dploie ensuite dans lEssai de 1943.
*
**
Dans litinraire intellectuel de Canguilhem, pistmologie historique et histoire des
sciences sont de survenue tardive et leur accentuation fut troitement lie lexercice de
responsabilits professorales, sans jamais devenir lobjet exclusif de son attention.
Cette pistmologie et son doublet historique sont rests pour lui, quant leurs
vises, subordonns des proccupations philosophiques dantriorit et de prsance
40
41
42
Je reprends ici, en la rectifiant quelque peu, linterprtation que jai dabord formule dans une
communication prsente dans le cadre du colloque organis Montral en hommage Yves
Schwartz en avril 2009, sous lgide du Centre interuniversitaire de recherche sur la science et la
technologie: Historien des sciences parce que philosophe: lexercice philosophique selon
Canguilhem.
Entretien avec Franois Proust, La mdecine et son histoire, Tonus, 1972.
Interrogations explicitement nonces notamment dans sa confrence de 1938 Activit technique et
cration et, lanne suivante, dans le Trait de logique et de morale.
65
Limoges
indiscutables, celles de sa philosophie biologique43 dont lessor stait dessin depuis

les annes axiales de dfinition de son uvre, de 1937 1943.
Hans-Jrg Rheinberger, a rcemment relev quaucune tentative pousse de
priodisation de luvre de Canguilhem na encore t tente.44 Les scansions que jai
voulu mettre en vidence ici devraient aider combler cette lacune.
43
44
66
Jean Gayon a soulign dj le caractre essentiel dans toute luvre de Canguilhem des thses
centrales de sa philosophie biologique. Cf. Le concept dindividualit dans la philosophie biologique
de Georges Canguilhem, dans Michel Bitbol et Jean Gayon (dir.), Lpistmologie franaise 18301970, Paris : Presses universitaires de France, 2006, p. 430-462.
Hans-Jrg Rheinberger, An Epistemology of the Concrete. Twentieth-Century Histories of Life.
Durham & London : Duke University Press, 2010, p. 37.
LA DISSECTION HUMAINE
De lhistoire des commencements celle des origines
Franois Delaporte
Dans son dernier livre, Secrets de femmes: Le genre, la gnration et les origines de la
dissection humaine Katharine Park, crit: En tenant compte de lensemble de ces
pratiques au lieu den isoler la dissection acadmique, jai voulu rtablir leur cohrence
culturelle. Cest l un point fondamental: en postulant de manire anachronique que
louverture de corps humains fut en premier lieu une procdure mdicale, les historiens
ont mconnu le phnomne plus large au sein duquel elle a merg -ou relgu ces
autres procdures au rang d arrire plan ou de contexte culturel . Je considre au
contraire louverture de corps humains comme un tout. Ses variantes (la dissection
proprement dite, lembaumement, lautopsie, lexcision de ftus, la reconnaissance
ou inspection des cadavres de saints hommes et femmes) sont comme une srie de
miroirs angulaires qui se rflchissent et sclairent les uns les autres. Aucune [variante]
nest premire, et la dissection (qui fut tous gards la plus sotrique) moins que les
autres. Pour mettre en vidence leurs points communs et leur degr dassociation dans
lesprit des contemporains, jai utilis les mots dissection et plus volontiers
anatomie pour les dsigner toutes, sauf quand lexigence de clart a rclam un
terme plus prcis .1
Quatre remarques serviront dintroduction. Elles indiqueront lordre dexposition des
questions que nous souhaitons poser.
1) Ce texte programmatique prsente une critique de lhistoire des historiens de la
mdecine : ils ont commis le pch danachronisme en postulant que louverture de
corps humains fut en premier lieu une procdure mdicale .
2) Pour Katharine Park, les historiens ont valoris la dissection acadmique et
relgu toutes les autres procdures au rang, dit-elle, de contexte culturel . Elle fait
linverse en rtablissant la prsance du contexte culturel au dtriment de la dissection
anatomique.
3) Cette cohrence culturelle, qui marque la prdominance du contexte culturel,
trouve sa pleine signification dans le fait que Katharine Park considre que
l ouverture du corps humain est un tout . Si louverture du corps humain est un tout,
son histoire doit tre faite comme celle dun tout. En porte directement tmoignage, le
titre de son livre: Secrets de femmes. Le genre, la gnration et les origines de la
dissection humaine.
Katharine Park, Secrets de femmes. Le genre, la gnration et les origines de la dissection humaine,
Dijon : Les presses du rel, 2009, p. 13.
67
Delaporte
4) Pour Katharine Park, on se trouve en prsence dune srie de variantes de

louverture qui doivent tre apprhendes, telles quelles apparaissent sous langle de
leurs points communs et de leur degr dassociation dans lesprit des contemporains .
Pour elle, lhistoire des pratiques devient lhistoire de la manire dont les contemporains
des origines de la dissection humaine se la reprsentent.
1. Anachronisme
Les historiens de la mdecine ont commis le pch danachronisme en postulant que
louverture de corps humains fut en premier lieu une procdure mdicale .
Relativement quoi, Katharine Park oppose lide dune cohrence culturelle des
diffrents gestes douverture du corps humain. Mais pourquoi une histoire qui prend
pour sujet dtude la dissection acadmique serait-elle anachronique?
La thse des historiens de la mdecine est lantithse de celle de Katharine Park. Il
ny a rien danachronique faire lhistoire dune pratique acadmique. Faire une
histoire de cette pratique, ce nest pas nier que louverture du corps humain puisse avoir
bien dautres significations. On pourrait donc retourner contre Katharine Park la critique
quelle adresse aux historiens de la mdecine. Il est anachronique de dire tout ce quon
sait au sujet dune procdure, dans la mesure o tout ce quon sait de cette procdure
relve de la diachronie. Les cas quelle dcrit dans son histoire sinscrivent dans une
succession historique, puisque sa priode stend du XIVe sicle au XVIe sicle. Cette
faon de procder nest pas innocente.
La srie des vnements du bas Moyen ge si bien points par Katharine Park peut
difficilement passer pour quelque chose qui ressemble aux origines de la dissection
humaine . Cest toute lAntiquit grco-romaine et limmense parenthse de lEcole
dAlexandrie qui sont cartes. Il suffit de couper le bas Moyen ge et la Renaissance
de son pass pour proposer une histoire qui occulte deux questions centrales la
Renaissance. Dune part, la question du retour lAntiquit et, par le biais de ce retour,
la description de la transformation des pratiques relatives la connaissance du corps
humain. Dautre part, la question, complmentaire de la prcdente, qui concerne lobjet
dtude: Sagit-il des origines ou des commencements ? Ici, se trouve carte la
question historique de lhistoire des commencements de la dissection anatomique dans
lEurope savante, mais aussi celle de lhistoire de cette pratique avec ses
transformations la Renaissance. Bref, substituer la question des origines celle des
commencements, cest privilgier une priode, du XIVe sicle au XVIe sicle, et un
objet dont il ny a rien dire sous le rapport dune histoire pistmologique, le geste
douverture. Disqualifier cette histoire, cest dprcier la comprhension historique de
la spcificit des diffrentes pratiques qui peuvent tre en rapport (ou non) les unes avec
les autres.
Lide selon laquelle les diffrentes pratiques douverture relvent dune cohrence
culturelle est clairement oriente, voire slective. Elle est le rquisit indispensable pour
tayer la relation entre les gender studies et, les origines de la dissection humaine .
68
Pour promouvoir cette histoire culturelle, Katharine Park procde une dvalorisation
de lhistoire des sciences.
En effet, pour Katharine Park, retenir louverture des corps des fins de
connaissance et denseignement, ce serait choisir une perspective trique. Dun ct, le
chirurgien, -un seul sujet du savoir- ; de lautre, le criminel excut, -un seul sujet
dtude. De plus, dit-elle, on sait que la rgle de la sance de dissection annuelle inscrite
dans les rglements de la fin du XIVe sicle tait plus ignore que suivie. En regard,
Katharine Park montre la richesse de la dissection autour dun ensemble de pratiques
culturelles sans rapport avec lenseignement de la mdecine et la recherche
anatomique : les rituels funraires, lembaumement par viscration, le culte des
reliques, les autopsies pratiques dans le cadre de la justice pnale ou des fin de sant
publique et, pour finir, une pratique obsttrique qui devait prendre le nom dopration
csarienne (sectio in mortua). Autre exemple de la dvalorisation des pratiques qui
relvent de la dissection acadmique. Katharine Park donne une perception toute
ngative de lanatomie des fins denseignement, parce que lindignit sy attache. Par
contraste, aucune des pratiques, comme lembaumement, lautopsie, les reliques
nimpliquait le profond dshonneur li aux leons de dissections publiques o le sujet
nu, anonyme, tait expos aux yeux dun groupe dobservateurs sans lien de parent
avec lui. Son corps, dit-elle, tait en grande partie dmantel, un dispositif qui violait
la fois sa qualit de personne et son identit sociale en le rendant mconnaissable et
impropre recevoir les obsques traditionnelles qui prvoyaient lexposition du corps
vtu sur sa bire. 2
Les leons de dissection publique ne sont pas aussi surdtermines que le dit
Katharine Park. Remarquons au passage que, suivie (ou non) cette rgle de la dissection
annuelle dans les universits prouve au moins que de lide douvrir un corps dans le
cadre de lenseignement commence faire son chemin. Rappelons galement quun
mort est appel un homme par homonymie et que louverture du corps est moralement
indiffrente. Cette ide avait dailleurs contribu depuis bien longtemps lever la
prohibition de louverture des cadavres. Enfin, pour affirmer que louverture porte
atteinte la dignit de la personne, il faut encore oublier quil ny a plus de personne
quand il y a un cadavre. Il ne suffit donc pas de dvaloriser un sujet dtude, en
loccurrence la dissection acadmique, pour valoriser les autres pratiques douverture. Il
ny a rien, dans la description toute relative des aspects ventuellement ngatifs de la
dissection anatomique, qui annule sa spcificit, sa signification et sa porte. Enfin,
porter des jugements de valeur sur les pratiques dont on fait lhistoire nest sans doute
pas le meilleur moyen de les dcrire ou de les rendre intelligibles. Nous ne pensons pas
que la dissection acadmique doive tre rduite au rle de faire valoir dautres pratiques
culturellement et socialement plus intgres comme lviscration des fins
dembaumement.
Ibid., p. 11.
69
Delaporte
2. Contexte
Pour Katharine Park, les historiens de la mdecine ont donn une place de choix la
dissection anatomique. Au dtriment des autres procdures qui seraient ainsi ravales au
rang de contexte culturel . Son histoire sera donc linverse de lhistoire classique de
la mdecine: elle accorde la prsance lhistoire du contexte culturel, qui comprend
toutes les formes de dissection humaine. Mais pourquoi faudrait-il ncessairement
penser lhistoire dune pratique, -que ce soit pour len affranchir ou pour ly inscrire, dans son rapport un contexte culturel?
Il est rducteur de croire que lhistoire dune pratique ne peut chapper lalternative
suivante: soit elle est dtache dun contexte, soit elle sy insre. Pourquoi faudrait-il
donc souponner les historiens de la mdecine davoir en tte le sombre
projet daffranchir lhistoire de la dissection anatomique de son contexte culturel?
Quils soccupent de lAntiquit, ou de la Renaissance, ce nest certainement pas leur
objectif. Du point de vue historique, il faut bien voir que la question nest pas celle du
contexte, mais celle de la coexistence de pratiques diffrentes qui nentretiennent aucun
rapport entre elles ou celle du passage dune pratique lautre. Voyons dabord la
coexistence de pratiques diffrentes qui nont aucune relation entre elles: cest le cas de
lembaumement, dont Katharine Park fait si grand cas, et de la dissection des fins de
connaissance dans la priode alexandrine. Heinrich von Staden a soulign ce point: Le
niveau des connaissances anatomiques requises par les mthodes gyptiennes
dembaumement est plus proche de celui dun habile boucher que de celui dun mdecin
grec [] la momification a trs peu voir avec la mdecine, et Hrophile et Erasistrate
nont probablement tir des embaumeurs gyptiens aucun bnfice pour leur
connaissance anatomique. 3 Voyons ensuite le passage dune pratique une autre: cest
le cas des premires descriptions anatomiques dAristote dans le droit fil des pratiques
sacrificielles en Grce. Jean-Louis Durand a montr que le sacrifice tait au cur de
lexprience quotidienne en Grce, et que le savoir quil suppose sest trouv thoris
par lrudition anatomique au point de constituer le modle partir duquel se construit
la logique aristotlicienne du corps. Cest dire quAristote est aussi observateur des
techniques sacrificielles et que les remarques tires de la hiroscopie lui sont prcieuses.
Assurment, le trajet de la description anatomique emprunte, chez Aristote, les voies
traces par la lame du dcoupeur des btes. Lanatomie des animaux est lespace
projectif o sinscrit secrtement un ordre de ncessits propres la socit des
hommes: une topologique. 4 En quoi le passage du sacrifice lanatomie comporte,
la fois, une filiation et une rupture.
On peut donc adresser Katharine Park, la critique symtrique et inverse de celle
quelle fait aux historiens de la mdecine. Ne soriente-t-on pas vers la plus grande
confusion en rduisant lhistoire des pratiques diffrentes celle dun mme geste
3
70
Heinrich von Staden Herophilus. The Art of Medicine in Early Alexandria, Edition, translation and
essays, Cambridge : Cambridge University Press, 1989, p. 29.
Jean-Louis Durand Btes grecques. Propositions pour une topologique des corps manger , in La
cuisine du sacrifice en pays grec, par Marcel Detienne et Jean-Pierre Vernant, Paris: Gallimard, 1979,
p. 165.
douverture? Sa premire erreur mthodologique, cest daccorder un privilge

pistmologique un geste vide de signification ou, pour le dire autrement, le geste
douverture abstraction faite de ses significations. Sa seconde erreur, consiste faire
passer pour les origines de la dissection humaine , une srie de pratiques diffrentes
et, par consquent, irrductibles les unes aux autres lpoque envisage. Ce que
Katharine Park reconnat bien volontiers: puisque, dit-elle, quand lexigence de clart se
fait sentir, il faut bien parler des variantes de louverture: autopsie, excision du
ftus, inspection des cadavres de saintes femmes, embaument, dissection acadmique.
Laffirmation, a priori, de lexistence dun contexte culturel nest pas fausse. Elle
devient suspecte partir du moment o elle conduit affirmer quaucune variante de la
dissection humaine nest premire. Et la dissection acadmique moins que les autres.
Inscrire cette dernire dans son contexte revient substituer son inscription dans un
environnement culturel sa mise en perspective historique.
Cest au sens journalistique que Katharine Park emploie le terme de contexte. Le mot
se dit dun ensemble de circonstances dans lesquels sinsre un fait: En accordant au
moins autant dattention ces procdures prives peu intrusives quaux dissections
institutionnelles de criminels par des professeurs duniversit, jai voulu restituer la
dissection acadmique son contexte social et religieux5 . A partir de l, quiconque
prtend sen passer commettrait une grave erreur. A vrai dire, tout dpend de quoi on
fait lhistoire. Retenons lessentiel: pour Katharine Park, le contexte a une fonction
explicative de premier plan. Pour elle, le contexte a le sens de milieu dans lequel
sinscrit une pratique. Le contexte est externe. Il est passif: un cadre pralable dans
lequel on insre une action. Restituer la dissection anatomique son contexte religieux
et social, cest donc rappeler qu ct de louverture dans un but de connaissance, il y a
bien dautres raisons douvrir un corps. Mais cette dmarche a ses limites: on est
conduit enregistrer, comme faits dobservation, toutes les diffrentes pratiques. Leur
description, aussi minutieuse et savante soit-elle, ne rend pas compte de leur mergence.
Elles sont toutes l, au dpart ou, plutt, aux origines. Il y a une histoire culturelle, ou
une histoire des mentalits, mais il ny a pas dhistoire des sciences.
Katharine Park admet la notion de contexte sans le moindre examen critique. Il nest
pas inutile de signaler une autre signification du mot. Contexte, contextus:
Assemblage, runion , spcialement: ensemble de relations organises entre les
lments significatifs dun discours . A donner cette notion une plus grande
extension, on dira que ce sont les relations qui rendent possible une pratique, un savoir,
une connaissance. Ce contexte dsigne un jeu de relations. Cest lui quil faut dcrire;
pris en ce sens le contexte est actif, cest--dire constitutif de lobjet dont on fait
lhistoire. On peut donc voir, dans le contexte, lensemble des conditions de possibilit
dune pratique. Autrement dit, le contexte fait partie de linvention des pratiques. Sans
lui, la pratique nexisterait pas. A titre dexemple, comme la montr Heinrich von
Staden, pas de dissection anatomique, voire de vivisection anatomique, sans un faisceau
de conditions: Lexceptionnelle combinaison de la dynastie des Ptolme, des savants
confirms comme Hrophile, une nouvelle cit dans laquelle les valeurs traditionnelles
5
Park, Secrets de femmes, p. 11.
71
Delaporte
ntaient pas considres comme intrinsquement suprieures, et une intelligentsia

cosmopolite ont permis le lever les inhibitions traditionnelles relative louverture de
corps humain6. On retrouve, ici, toute la force du mot contexte: driv de contexere,
ourdir, entrelacer , bref, tisser avec .
3. Procdure
Les historiens de la mdecine ont fait lhistoire de louverture de corps humains dans la
perspective dune histoire de la dissection anatomique. Katharine Park leur oppose une
histoire de louverture de corps humains comme celle dun tout. Do la prpondrance
de la procdure, le geste douverture, au dtriment de ltude des diffrentes pratiques.
Il ne faut pas sy tromper: les origines de la dissection humaine recouvrent une srie de
pratiques diffrentes dont le seul trait commun consiste dans louverture. Mais quest-ce
donc quune histoire qui se donne pour objet dtude la dissection ou, encore,
lanatomie? En loccurrence, louverture du corps humain comme un tout? Jai utilis,
prcise-t-elle, les mots dissection et (plus volontiers) anatomie pour les dsigner
toutes .
Dans la mesure o il ny a pas de procdure sans signification, comment peut-on
donner une histoire du geste douverture comme celle dun tout? Comment peut-on
donner une histoire de la dissection humaine qui ne soit pas celle des diffrentes
pratiques? Katharine Park est bien oblige dapprhender lhistoire dun tout comme
celle dune srie de cas. Autrement dit, elle est bien oblige de revenir aux diffrentes
modalits de louverture de corps humains sous la forme de ce quelle nomme ses
variantes . Pour ce faire, elle subsume diffrentes pratiques, comme lautopsie,
lexcision du ftus, linspection des cadavres, lembaumement, la dissection
acadmique sous la catgorie anatomie humaine . Le reste en dcoule. Pour
sopposer lhistoire de la dissection acadmique, comme discipline autonome,
Katharine Park doit montrer quelle est troitement lie aux autres.
De l, quelques considrations sans grande porte pistmologique. Sous langle du
savoir, le bnfice de lembaumement est mince. On apprend que Berangario Carpi,
loccasion de lviscration de larchevque de Turin, trouve que les personnes grasses
accumulent de la graisse dans la rgion du cur. Plus important, peut-tre, le culte des
nouveaux saints et les pratiques funraires doivent occuper une place de choix: Jai
voulu montrer, dit-elle, que les pratiques sociales et plus spcifiquement religieuses ont
jou un rle bien moins priphrique dans lhistoire prcoce de la dissection que ne
lont laiss entendre la plupart des historiens de lanatomie . Un peu plu loin, on trouve
cette apprciation, sans doute moins priphrique que la prcdente, mais gure plus
centrale: Du fait que ces procdures taient en lien troit avec la pratique de la
dissection au service de la recherche et de lenseignement de la mdecine, et parce
quelles jourent un rle dterminant dans son histoire, les informations rsultant de ces
proccupations faonnrent en mme temps le geste dissecteur comme tel7. Katharine
6
7
72
Von Staden, Herophilus, p. 141.

Park, Secrets de femmes, p. 15-18.
Park reste discrte sur ce point. On peut lgitimement douter de limpact de pratiques
sociales, comme lembaumement, sur le geste dissecteur en particulier et sur la
dissection anatomique en gnral.
Lide selon laquelle doit prvaloir lhistoire de louverture occulte lhistoricit des
diffrentes pratiques douverture. Le corrlatif dun tel rcit, cest une histoire qui est
celle dune collection de cas. De ce point de vue, lentreprise est un succs. Mais elle
comporte la ngation de lhistoricit des pratiques dans leur devenir.
La mthode de Katharine Park est toujours la mme: dvaloriser lhistoire des
sciences et substituer lhistoire des origines celle des commencements. Pour
rhabiliter les saintes femmes et plus gnralement lEglise, elle rappelle la tradition
historiographique de ses adversaires, qui remonte au milieu du XIXe sicle. Cette
tradition oppose la science et les prjugs de la religion comme deux entreprises
culturelles contradictoires. Cette lecture errone, dit-elle, est encore largement
rpandue . Et de citer, comme exemple de choix, un auteur auquel on ne sattendait
pas. Michel Foucault, dans Naissance de la clinique, a crit que lhistoire a prt la
fin de lAncien rgime les couleurs du Moyen ge en ses dernires annes, et confondu
avec les dchirements de la Renaissance les problme et les dbats de lAufklrung. 8 Il
nen faut pas plus pour assurer que cette version donne par Foucault a
considrablement influenc lhistoriographie de la dissection humaine depuis les annes
1980 . De l, la dnonciation de fictions, produisant des mythes fondateurs propres
confirmer les intuitions occidentales bien ancres sur les origines scientifiques de la
modernit. 9
Il nest pas ncessaire dtre un spcialiste de la Renaissance pour savoir quil y bien
eu, lpoque de Vsale, pnurie de cadavre et quil pouvait arriver quon en dterre
quelques-uns. OMalley a rappel ce point. Foucault na donc pas tort de dnoncer,
comme il le fait, la bvue des historiens de lAncien rgime. Il y a plus; chacun sait que
Naissance de la clinique ne soccupe pas du bas Moyen ge, ni de la Renaissance ni de
Vsale. Park a bien raison de partir en guerre contre lide, errone, selon laquelle
lEglise serait fondamentalement hostile la dissection. Mais contre Foucault, sa
critique fait long feu. Enfin, il est clair que Katharine Park ne sintresse pas lune des
tches principales de lhistoire pistmologique: penser lmergence dune pratique,
dcrire les ruptures et les transformations qui prsident ses commencements toujours
relatifs. Si ctait le cas, elle aurait compris, en continuant la lecture du fameux chapitre
Ouvrez quelques cadavres , en quoi la perception de Bichat diffre de celle de
Morgagni et, plus prcisment, en quoi la question de lorigine est un mythe
pistmologique.
Michel Foucault, Naissance de la clinique. Une archologie du regard mdical, Paris : PUF, 1963,
p. 126.
Park, Secrets de femmes, p. 17 et note 20, p. 236
73
Delaporte
4. Reprsentations
Pour Katharine Park, les variantes de louverture doivent tre apprhende dans leurs
ressemblances, telles quelles apparaissent dans lesprit des contemporains . Par l,
elle oppose une histoire des reprsentations une histoire des sciences. Faire lhistoire
de lanatomie, ce nest pas faire lhistoire des reprsentations du corps par les religieux,
les profanes ou les populations, dans un pays donn, un moment donn de lhistoire.
Il doit tre possible de faire une histoire des pratiques qui serait une histoire de ce que
font les hommes, et non celle de ce quils croient faire ou de leurs reprsentations.
A la rigueur, on peut dire que la procdure dembaumement est plus ancienne que
son application la dissection acadmique. Mais, pour le dire, il faut envisager une
priodisation bien diffrente de celle que propose Katharine Park. Personne ne doute
que lviscration des cadavres remonte loin dans lhistoire et que cette pratique
dsigne, dans diffrentes cultures, des rites religieux. Il y a quelques 5 000 ans avant
Jsus-Christ, bien avant les Egyptiens, les cultures Chinchorro dans le nord du Chili
momifiaient artificiellement les corps. Mais au Chili et au nord du Prou, pas plus quen
Egypte, cette pratique na engendr la dissection anatomique. Nietzsche la bien vu: il
faut distinguer la procdure et ce quil y a de fluide en elle, cest--dire le sens, le but,
toutes choses qui se rattachent la mise en uvre de la procdure. Suffit-il vraiment que
des pratiques diffrentes soient dsignes par le mme terme d ouverture
pour quon cesse de les distinguer? Et pour renoncer faire lhistoire de chacune?
Suffit-il que ces pratiques aient des points communs, aux yeux des contemporains, pour
quelles ne soient que des variantes dun mme geste fondateur?
Katharine Park, semble-t-il, donne une solution la dlicate question de la
distinction des pratiques lies au geste douverture en convoquant le concept d airs de
famille (Wittgenstein). Les phnomnes dsigns comme geste douverture ou,
encore, comme dissection humaine sont regroups sur la foi dune ressemblance
quon dcle entre eux. Il faut bien croire lexistence dune structure de parent pour
occulter toutes les diffrences et viser lhistoire dun geste douverture. Pour Katharine
Park, il ny a aucune raison de privilgier certaines pratiques au dtriment des autres, et
dintroduire, parmi elles, des hirarchies. Il ny a aucune raison de privilgier la
dissection acadmique au dtriment de lviscration des fins dembaumement. On
peut toujours renoncer hirarchiser des pratiques diffrentes. Mais ce refus nannule
pas pour autant leur diffrence. On a bien compris lenjeu: Secrets de femme contre
Vsale. Pour en revenir notre question, il se pourrait bien quelle ne soit pas celle de la
hirarchie des pratiques, mais celle de leur relation. En loccurrence, une relation
dexclusion. L ouverture , ou la dissection humaine peut-elle fonctionner comme
un concept unificateur vocation classificatoire ds lors que chacune des formes de la
dissection humaine prsente une histoire spcifique?
Mettre en avant la prgnance des reprsentations du geste douverture par les
contemporains relve du dfi: La spcificit des contextes culturels assortie et des
reprsentations associe mont amen restreindre mon enqute lItalie du Nord,
rebours de la tendance de la plupart des histoires de lanatomie adopter une
74
perspective chronologique universalisante. 10 Cest, un fois de plus, la volont de

sopposer une histoire des sciences dont le dfaut serait de privilgier les sommets:
louverture des fins denseignement et de recherche. Mais, aux yeux des spcialistes,
la priode que scoule entre Mondino (1316) et lpoque de Lonard de Vinci et de
Vsale a connu si peu de progrs quon pourrait penser une sorte de rgression. Pour
la priode choisie, Katharine Park navait peut-tre pas dautre choix que de valoriser
les procdures extra-scientifiques.
Mais pourquoi opposer lhistoire des sciences une histoire rgionaliste? A quoi bon
souligner que les habitants des villes dItalie du Nord comprenaient leur corps avant
tout partir des schmes de la parent et de la religion et que les modles mdicaux
arrivent loin derrire dans lordre des proccupations. 11 On pourrait dire, inversement,
et sans en tirer le moindre bnfice intellectuel, que les mdecins comprenaient le corps
partir des schmes de lanatomie hrits de lAntiquit.
La mthode utilise par Katharine Park est simple. Elle enregistre les traits communs
des variantes de la dissection humaine en soulignant leur degr dassociation dans
lesprit des contemporains. Les mots des contemporains pour dcrire louverture de
corps humains trahissent une grande porosit entre le monde universitaire et ceux de la
parturition, des procs criminels, des maladies chroniques, des funrailles officielles et
du culte chrtien. La littrature mdicale recours au latin anatomia avec se variantes
(nothomia, anathomia) et drivs vernaculaires pour dsigner indiffremment des
pratiques distingues aujourdhui sous les termes de dissection et dautopsie
Lemploi des verbes marque encore moins de distinction entre ces diverses pratiques.
Les auteurs latins ont volontiers recours des termes non techniques inciser, ouvrir,
viscrer- pour dcrire aussi bien une dissection ou une autopsie quun embaumement
ou louverture dune femme morte en couches. Ces quatre pratiques se trouvent
associes avec sparare associ la prparation danimaux des fins culinaires, telle
laction de vider un poisson ou un cochon. 12 La citation un peu longue montre
clairement que lobjectif de Katharine Park est daligner le discours savant sur les
discours populaires.
A partir de l, Katharine Park affirme labsence de coupure entre les diffrentes
pratiques. Cette approche suppose une confusion, savamment entretenue, entre
connaissance vulgaire et connaissance scientifique. Lhistoire culturelle comporte une
histoire qui sinscrit dans la continuit. On est ici loppos dune histoire des ruptures
tel quon la trouve en histoire des sciences. Mais pourquoi donc procder la confusion
des genres? Sous langle dune histoire culturelle, le livre de Katharine Park est
admirable.
10
11
12
Ibid., p. 14
Ibid., p. 17.
Ibid., p. 11-12.
75
INTERVALS, POSSIBILITIES AND ENCOUNTERS

The Trigger of a Ruptured History in Bachelard
Monika Wulz
1. Bachelard
The starting point for my paper was the sharp attention drawn by a sentence that I
happened to read in Bachelards Dialectic of Duration. In the corresponding passage
Bachelard is concerned with the question: what is an action? How is it initiated? How
and from which elements does it emerge? What is its trigger? And which kind of
temporal duration does it generate?
Referring to Eugenio Rignanos Psychology of Reason,1 Bachelard emphasizes that
an action is not just effected by physiological impulses. Instead he points to much more
abstract circumstances that initiate the possibility of activity. Bachelard writes:
Il faut quil y ait permission dagir, adhsion de lesprit ltre. Cette adhsion,
cette prsence de lesprit, nest sentie que dans un repos pralable, en confrontant
nettement le possible et le rel. (DD, p. 73)
There has to be permission to act, and the mind must lend its full support to being.
We only feel this support, we only feel the minds presence, in the repose that
precedes the action, when the possible and the real are clearly compared. (DDe,
p. 86)
So, what happens in this encounter of mind and being? What kind of agency does the
moment of repose initiate? What does it mean to confront the possible and the real in
this instant? Why is the mind only present in this encounter that takes place within the
moment of repose? And why, for Bachelard, is the initiation of an action the origin of a
temporal duration? Why is temporality only an effect of the agency within this
confrontation?
Bachelard ties his metaphysical interest in the agency that produces temporality and
duration to the enterprise of a philosophy of repose (DD, p. V) and a psychology of
annihilation (DD, p. 8). As a consequence his account requires the idea of nothingness
(DDe, p. 29), the possibility of voids. It implies negativism, coercion, inhibition,
hesitation, and destruction. In addition, and in contrast to the idea of a continuous
creation in living processes (as proposed by Bergson), Bachelards philosophy of
temporality is based on ruptures. It allows lacunae: realms of nothingness. It sets as
we will see the possibility of initiating a development within these vacancies. This
instant of nothingness is the point of departure for a temporal sequence.
1
Eugenio Rignano, Psychologie du raisonnement, Paris: Alcan, 1920. All abbreviations used for
referencing quotations in the text are explained in the bibliography at the end of this contribution.
77
Wulz
In his Intuition de linstant Bachelard develops the ontology of being and the
potentiality of development from the moment of isolation; he suggests that the
possibilities of reality are enclosed in isolated instants or isolated points. In other words
there is no direct connection between the scattered possibilities. They are not only
disconnected in space (in the way Leibniz described the monads) Bachelard adds that
these entities are also disconnected in a temporal dimension: there is no connection
between an entity and the past, it is essentially detached and isolated (II, p. 60f.). It is
this spatial and temporal discontinuity to which Bachelard links the potentiality of
emergence. The multiple possibilities appear only in this state of isolation, within the
instant of nothingness (II, p. 67).
It is important to understand that Bachelards concept of nothingness, or the moment
of repose, is a free and undefined sphere where a specific conjunction can take place:
the fusion, the adhesion of the mind and the being, the collision of the possible and the
real. This incidental conjunction of opposites breaks up every continuous development
(DD, p. 71). Hence, there is also no immediate possibility of action. Instead Bachelard
emphasizes that every action is the product of an instantaneous concurrence of
discontinuous and unexpected events. It is initiated by contradictory and counter-acting
elements (DD, p. 71). The collision of the possible and the real, as well as the fusion of
mind and being, take place in the lacunae. As a result they become agencies for an
absolute beginning. They initiate an action that is effective in a temporal dimension.
For Bachelard, the interest in the initiation of an action is important in so far as it
provides the foundation of temporality. Only the concurrence of discontinuous and
counter-acting elements generates duration, temporality, reality, individuality, mind,
memory, and knowledge (DD, p. 73). It continuously initiates a novel temporality and
new temporal objects (DD, p. 69) and thus creates differential durations allowing
multiple and anti-chronological times. It creates a temporal individual with its specific
memory, a temporal object with its specific history and future. An action is not based on
a physiological sequence. Instead, Bachelard emphasizes that the initiation of a
temporal duration is based on a complex system of triggers. A temporal duration can
only emerge from the complex arrangement (le groupement; DD, p. 73) of initiating
elements.
In addition, the mind, according to Bachelard, is not a consistent entity related to an
individual. It is only an interaction of impulses. It only appears in the incidentally
emerging coherent arrangement of impulses that triggers an intelligent action (DD,
p. 69).
It is important to understand what Bachelard suggests with this idea of an
arrangement of initiating triggers at the basis of every action and every temporal
duration. He suggests a de-subjectified understanding of mind, memory, and reason.
With his concept of the instant he dissolves any idea of a pre-existing and permanent
identity or individuality: he emphasizes that there is no persistent identity of the I
outside of the synthesis that is realized within the instant. The I is essentially
discontinuous, differentiating, temporary, and contingent: the individual is only a sum
of coincidences and even this sum itself is a coincidence (II, p. 70). The person can
78
only be found in the dust of events (DD, p. 35). It is only due to the arrangement of
triggers that we are able to desire, to suggest, and to observe (DD, p. 74).
As we move on it becomes rather obvious that Bachelard addresses the isolated
instant as a moment of solitude. However, Bachelard does not talk about an isolation of
the subject from its environment, but about an isolation from ourselves a rupture with
our own past (II, p. 13). Bachelards instant is thus a de-subjectified2 isolation: an
isolation without a subject. The moment of solitude is the pure void devoid of any
concept of individuality. In other words, Bachelards concept of the instant constitutes
the moment in which we depart from ourselves in order to anticipate reality (II, p. 15).
He suggests that we can discover the areas of repose, the realms of nothingness, the
lacunae, or the vacancies exclusively in the impersonal part of the person (DDe, p.
18). Or as he observes, the nothingness is in us (DD, p. 29). The void is within the
instants. They generate an agency of framing that brings forth an individual personality
the mind, memory, a temporal duration of life, and differential layers of knowledge.
Only in the solitary instant does the possibility of time appear. Bachelard thus aims at
developing a differential temporality and individuality based on the idea of nothingness
and a psychology of coincidences (DD, p. 29). By the same token Bachelard calls for
a kaleidoscopic and discontinuous character of material change (DDe, p. 79). The
existence of matter is only a singular event as much as the existence of individuals is:
The coincidences of subject and object will be atomized. They will have no duration.
(DDe, p. 46) There is no continuously existing matter that could be the object of
research. Bachelard emphasizes that subject and object are microphenomena that are
produced in the knot of coincidences (DD, p. 28). The nothingness, the absence, the
lacunae are not only in the individual they are also part of the materiality that is
established in the process of research.3
The moment of isolation constitutes the instant of the dissolved subject and the
dissolved object. This de-subjectified instant is of crucial importance in Bachelards
philosophy: it is the area where an unforeseen event can happen. It opens up the
possibility of coincidences and creates a zone where obstacles, deviations, and
On the concept of de-subjectification and the transgression of the human and the intellectual in
Bachelard see Jean-Michel Le Lannou, Le dpassement de lhumain. Bergson et Bachelard, in:
Bachelard et Bergson. Continuit et discontinuit? Actes du Colloque international de Lyon, 28-30
Septembre 2006, Frdric Worms, Jean-Jacques Wunenburger (eds.), Paris: PUF, 2008, pp. 73-93.
In the Dialectic of Duration Bachelard develops this philosophy of coincidences based on recent
studies in the fields of psychology, neurology, psychiatry, and sociology (referring i.a. to the works of
the psychologist Pierre Janet, the anthropologist and neurologist W.H.R. Rivers, the biophysicist
Pierre Lecomte de Nouy, and the psychiatrist Eugne Minkowski, relying on the psychological works
of Eugenio Rignano and the works on sociology and probability of Eugne Duprel). However, he
does not only use these studies as a critique of Bergsons philosophy. He tries to go even further
extending his philosophy of nothingness and discontinuity to questions of matter, causality, and
microphenomena in the discipline of physics (DD, pp. 52ff.). Moreover, he emphasizes that his
discontinuous account of time and duration is in accordance with Einsteins theory of relativity and
the subsequent critique of the concept of duration as such (II, pp. 29ff.). In his New Scientific Spirit he
also relates his discontinuous philosophy to the concepts of the wave-particle duality and of
indeterminism in quantum theory.
79
Wulz
inhibitions can intervene (DD 86). Bachelard says that Things are but opportunities
[] to be tempted (DDe, p. 41). They are occasions for transformation.
Following the critique of the traditional concept of causality that the Belgian
philosopher Eugne Duprel developed in his philosophy of the interval4, Bachelard
then points to the possibilities that can enter in the interval between what is traditionally
conceived as cause and effect. As he writes in Dialectic of Duration:
En effet, cest dans lintervalle temporel que pourront intervenir les
empchements, les obstacles, les deviations, qui briseront parfois les chanes
causales. Cette possibilit dintervention, il faut la prendre pleinement comme une
possibilit pure et non pas comme une ralit ignore. Ce nest pas parce quon
ignore ce qui interviendra quon manque prvoir lefficacit absolue dune cause
donne; cest parce que, de la cause leffet, il y a une intervention toute
probabilitaire dvnements qui ne sont daucune manire lis la donne
causale. En particulier, on naura jamais le droit de se donner lintervalle. Dans la
science, on peut construire certains phnomnes, on peut protger lintervalle de
certaines perturbations, mais on ne saurait vincer toute intervention de
phnomnes imprvus dans lintervalle de la cause leffet. (DD, p. 85/86)
Indeed, it is in the interval of time that impediments, obstacles, and deviations can
intervene and these will sometimes shatter causal chains. This possibility of
intervention must be wholly regarded as a pure possibility and not as a reality we
do not know. It is not because we do not know what will intervene that we fail to
predict the absolute effectiveness of a given cause; rather, it is because there is
between cause and effect an entirely probable intervention of events which are not
in any way at all connected to the causal datum. In particular, we shall never have
the right to give ourselves the interval. In science, we can construct certain
phenomena, we can protect the interval from certain disturbances, but we cannot
get rid of every intervention of unforeseen phenomena in the interval between
cause and effect. (DDe, p. 97)
In this connection Bachelard emphasizes that there is a free scope for possibilities
of halts or deviations (DDe, p. 98) in every development, every vital potency, and
every continuous movement. The disturbances intervene within these undetermined
zones (DD, p. 87). The obstacles, the incidental occurrences, the disturbances are pure
possibilities for unforeseen deviations and novelties.
For Bachelard, the temporal connection of two events is always contingent. The
concurrence of events is important only because it can become effective as an initiating
instant (DD, p. 53). As a consequence the crucial question is not how a specific
concurrence comes about. Instead the crucial question is whether or not this
concurrence triggers a temporal agency an expectation that develops persistence and
duration. The obstacles, the incidental occurrences, the disturbances operate as triggers
for actions. When intervening within the interval they create attention and expectation
towards further events. In this way they initiate a temporal dimension. An object is
therefore only a temporal phenomenon that is generated due to a frame of expectation
(DD, p. 65). Bachelard points out that even the process of material research is a
4
80
Eugne Duprel, La cause et lintervalle ou ordre et probabilit, Bruxelles: Lamertin, 1933.
temporal activity essentially based on waiting for discontinuous events: You will
have to wait for it [matter] to produce its events (DDe, p. 46).
Quoting Jean-Marie Guyaus La gense de lide de temps,5 Bachelard emphasizes
the role of desire in the emergence of duration: only the ability to desire something
creates time. The future and the past are therefore not dimensions of time as such.
Instead Bachelard points out that the duration of time is produced by the tensions of our
desires (II, p. 51) that effectuate the emergence of habits and intentions. These habits
and intentions create a perspective of actions (II, p. 74) and make us set aims for our
future. Organizing the relationship between the past and the future, they thus generate
the persistence and development of an entity (II, p. 60f.). Nevertheless, Bachelard
emphasizes the discontinuous origin of the desire, the habits, and the intentions. They
are essentially based on the isolated instant and initiated by the intuition emerging from
within (II, p. 51). In this way the intention frames the future it organizes time as a
perspective related to an initiating instant as the centre of projection. As a consequence
it is crucial for Bachelard that the emergence of duration always depends on a viewpoint
(II, p. 34).
In the Dialectic of Duration Bachelard again sets the emergence of duration in
relation to the instant. This time he emphasizes even more the role of the arrangement
and re-arrangement of instants that set a framework of expectation. He points out that
only the arrangement of triggers6 in a framework in an artificial system gives
meaning to them (DD, p. 50). It creates a project of expectation and hence brings forth
temporality: a temporal object and a temporal subject that is acting in a temporal
dimension. It creates both history and future. More generally, Bachelard underscores If
we think time, it means that we place life in a framework (DDe, p. 92). It is this
activity of framing that Bachelard denotes as rational. It enables time, duration, and
history. The activity of framing and re-framing events is, for Bachelard, even the
starting point for the emergence of memory (DD, p. 46). He conceives the arrangement
of triggers as a rationalized memory (DD, p. 74) and emphasizes that only the order
and arrangement of involved instants and elements can be the basis of the rationalized
memory. This kind of memory is not a recollection of duration in itself. On the
contrary, Bachelard points out that the arrangement of initiating triggers is the origin of
duration (DD, p. 73). Hence, memory can only emerge from the present interaction of
elements that generates a framework of attention: the expectation makes temporal
frameworks in order to receive memories (DDe, p. 62). This activity of framing
consolidates memory and history. On the other hand, it is always related to its other: the
unexpected novelty and innovation as the discontinuity of time (II, p. 15). There can
be no [emotional] memory without an initial drama, without surprises by opposites
(DDe, p. 63).
In relation to this notion of a rationalized memory, it is important to understand
that in his Lintuition de linstant Bachelard proposes a concept of rationality that does
not only refer to thinking and reasoning, but to every activity of framing that develops a
5
6
Jean-Marie Guyau, La gense de lide de temps, Paris: Alcan, 1890

Lordre des dclics, schme de dclics, groupement (DD, p. 74).
81
Wulz
temporal duration and persistence based on the incidental coherence of instantaneous

triggers. Bachelard uses the term rational to refer to a de-subjectified system of
triggers that generates permanence: it effectuates a coherent and continuous activity in
time based on the projection of a frame towards the future or towards the past. Since
this is a dynamic process, Bachelard rather uses the term becoming rational (en train
de devenir rationnel, II, p. 91) instead of rationality. The process of becoming-rational
can thus also be attributed to feelings and emotional activities that create permanence:
On ne peut pleurer longtemps quun tre quil est rationnel de pleurer (II, p. 91).7 In
this regard Bachelard emphasizes that le temps est affectif (II, p. 93) time is an
affective formation. It is created from within the isolated instant by the desire to expand,
to develop permanence, a past, and a future. In every present instant this rational
agency aims at preparing the future (II, p. 92), namely its specific kind of future
related to the coordination of the specific possibilities present in the instant. According
to Bachelard this agency of the rational is effective in two ways: by its desire to expand
and to prepare the future, it generates a temporal dimension and at the same time reason
has an anti-hereditary effect it interrupts every kind of memory and heritage (II, p.
97). Bachelard thus proposes a concept of rationality that is not eternally valid and
innate it needs to be shaped temporarily just as the experience does (NES, p. 172).
According to Bachelard, developing a temporal dimension thus means to develop an
arrangement of instants to make use of instants within the constellation of a frame (II,
p. 88). Only in this way can the isolated instants get consolidated in a specific history.
The ongoing adaptation of the framework is the recurrent determination of the present
by the expected future (DD, p. 78). It creates differentiated temporal realities (DD, p.
81) and constitutes specific temporal formations (DDe, p. 91). Bachelard emphasizes
that the future and I would like to add, history is a project, a program of promised
actions. The intuition of the future promises the possibility of action. Anticipating the
future and reverting to the past (that never happened as such) means fixing its frame and
thereby neglecting its initial vacancies (DD, p. 35). Only the framework of expectation
opens duration towards the future as much as towards the past, like a perspective (DD,
p. 40).
Focusing on the relevance of the instant, Bachelard is also interested in the instant
of the emerging knowledge (II, p. 6). He suggests that knowledge (i.e. connaissance)
emerges from the instant that generates a frame of anticipation and expectation (attente)
(II, p. 90). In the same way, Bachelard emphasizes again in an epistemological context
that the contemporary science goes beyond the subject and the object and is instead
based on the project. (NSS, p. 11f.) Science sets programs and thus shapes the
differential temporality of its research, of its objects, and of its subjects. As he writes in
the New Scientific Spirit:
Objective meditation in the laboratory commits us to a path of progressive
objectification that gives reality to both a new form of experience and a new form
of thought. [] The scientist, when he has done with his days objective
meditation, has his program of research for the following day in hand, and at
82
One can only cry for a long time if one has the intelligence to cry. (Translation M.W.)
the end of each working day he repeats the following article of faith: tomorrow I
shall know [].8 (NSS, p. 171)
In other words, we can find the concepts of the repose and the interval also in
Bachelards epistemological writings. Their use is not restricted to his metaphysical
early works. In his writings on epistemological questions these concepts are
transformed into the notion of the rupture and the no. Again, they constitute the
moment when the epistemological obstacle can intervene and transform the whole
epistemic scene. The ontological discontinuity of isolated instants appears again as a
discontinuous epistemology. It constitutes the capability to negate and to make
decisions in view of different possibilities.9 This moment of the rupture where the
obstacle can intervene is the basis of a temporary and provisional epistemology a
historical epistemology. It is concerned with temporal objects and thus with temporary
epistemologies that frame their research in provisional programs at any moment from
the perspective of their actual research.
While Bachelard thinks of life, memory, and reason as discontinuous and
instantaneous activities, he also describes the progression of scientific knowledge as a
discontinuous and unexpected event. He emphasizes that the scientific progress is a
phenomenon of suddenness: the epistemic innovation bursts out. At the same time, it
makes the traditional epistemology burst (clater). (MR, p. 210)
From this perspective Bachelard criticizes a historiography that only builds up a
sum of facts as a chronological line. He emphasizes the uselessness of such a
history (II, p. 82). Bachelards critique thus addresses a history that does not pay
attention to the instant and does not make use of instants. It is the critique of a
chronological kind of history that does not incorporate and appropriate initiating
coincidences for developing its own framework. In contrast, Bachelard suggests a
concept of time and history that is ruptured, framed, serial: history is affected by
intervening obstacles, deviations, and events. Time is a product of the use of instants
(II, p. 88). He thus proposes a dynamic kind of history that operates on the basis of
temporal frames emerging from the synthesis in the instant (II, p. 82). The duration of
time, the characteristic of history, is for Bachelard therefore an ongoing innovation and
invention emerging from the instant the zone of nothingness, the interval, the moment
of the rupture, of the no. As a result we can understand the historical epistemology with
Bachelard as a formation emerging from within the epistemic void: the rupture in our
memory and reason, the break created by the no, the nothingness in ourselves and in the
epistemic objects.
Much could be said about the critical enterprise of the Dialectic of Duration and the
concepts of time and duration based on discontinuity which Bachelard set out against
Bergson and his conception of continuity. Instead, I want to relate Bachelards concept
8
The English translation of this last passage reads Tomorrow I shall know the truth. With regard to
the procedural character of Bachelards epistemology, I suggest following at this point the French
original version more closely: Demain, je saurai (NES, p. 173).
James Williams, How Radical is the New? Deleuze and Bachelard on the Problems of Completeness
and Continuity in Dialectics, Pli. The Warwick Journal of Philosophy, vol. 16 (2005), pp. 149-170.
83
Wulz
of the isolated instant in a different direction: I want to relate it to the philosophy of

Louis Althusser. While the relation between Althusser and Bachelard is widely
acknowledged as far as the concept of the epistemological rupture is concerned, I would
like to relate Bachelards concept of the interval where obstacles and disturbances
intervene and unsettle us to yet another theoretical constellation in Althussers work:
namely to the emphasis on the void and the moment of surprise in view of the incidental
encounter in Althussers conception of an aleatory materialism.
2. Althusser
With his concept of a materialism of the encounter, the late Althusser establishes a
philosophical account that denies any pre-existing purpose or cause of reality. Instead,
he aims at constituting the contingent character of every specific occurrence of reality.
According to him, at the bottom of every reality lies a coincidence: a contingent
encounter of elements in a realm of the vacuum in the void.
Bachelard starts his psychology of coincidences from the interval, the solitude, the
nothingness in us. Althusser places the contingent encounters in the realm of
nothingness, the void, the disorder. While Bachelard understands the instant as a kind
of temporal atom (II, p. 26), Althussers aleatory materialism is based on the Epicurian
atomism. It relies on the idea of a parallel rain of atoms falling in the void. In this rain
of atoms a clinamen, an infinitesimal deviation in the route of the isolated atom, is
possible. It can cause conjunctions of elements (MB, p. 87). This random deviation and
encounter is the origin of reality in a specific formation. Althusser suggests that with
this incidental encounter a world is born (SPh, p. 40). The atoms thus do not have any
reality in themselves they only obtain reality and materiality in the contingent
encounter. As a consequence the aleatory materialism denies any pre-existing purpose,
cause, or ultimate goal preceding and underlying the formation of a specific world;
every reality is based on a mere coincidence (SPh, p. 42). Reason, necessity, and
purpose are constituted only within the formation of a specific reality (SPh, p. 41). It is
an account of materialism without any subject dominating the aleatory material
development (SPh, p. 40).
Althusser uses the term conjuncture for the contingent encounters that underlie
every specific formation of reality.10 With this term he points out the pure coincidence
of a specific encounter that took place and that, however, could also have happened
differently. Every kind of order emerges and exists only within a fundamental disorder.
Every encounter might not have happened although it had taken place. The emergence
of a specific encounter, of a specific reality, can therefore only be understood in a
recurrent action recurring from the actual event that developed specific individuals
and specific formations.
10
84
He develops this concept with regard to his reading of Machiavelli.
Reproduced from: Louis Althusser, Du matrialisme alatoire, Multitudes, vol. 21

(2005), p. 187. See also Louis Althusser, Materialismus der Begegnung: Spte
Schriften, translated and edited by Franziska Schottmann,
Zrich: diaphanes, 2010, p. 87.
This is a sketch of how Althusser pictured his concept of the materialism of the
encounter. There is the void in the center; the circles are the formations of persistent
conjunctions that constitute realities and that develop formations of necessity.
Althussers materialism of the encounter is, however, not only based on the void in
the center. Beyond the formation of a temporarily stabilized reality and its specific
necessities, the material encounters again face the danger of becoming de-stabilized.
They are again confronted with the void in the fringy margins, in the intervals and
neutral spaces that are not yet determined by persistent encounters. These undetermined
interstices give a provisional status to every specific formation of reality.
On this basis Althusser suggests a concept of materiality that can shift and change at
any time (SPh, p. 44). It faces its own instability in the margins of disorder. Since there
is no reason for a conjunction in the elements themselves there is no reason and no
teleology at the bottom of our reality. So, when every encounter is contingent and
temporary, how is a permanent encounter, a permanent kind of reality, possible?
Althusser emphasizes that a permanent encounter (a duration of specific formations of
reality) can only originate from a coerced connection, from external coercions and
necessities that generate persistent conditions of conjunctures. With regard to Bachelard
85
Wulz
this coerced connection might also be called a framework. It creates the reality of our
specific history: the reality of specific individuals, specific societies, languages, the
state, and even of knowledge.11
From his considerations on the changing materiality of every reality, Althusser also
derives a conception of history and historiography. He criticizes that the history of the
classical historians takes its objects as fixed and accomplished in the past (SPh, p. 44).
On the basis of his concept of aleatory materialism he therefore suggests a kind of
history that is based on the singular events causing unforeseen bifurcations: a histoire
au present (SPh, p. 45) that does not deal with objects accomplished in the past but
with objects that are open to a random and uncertain future. The historiography of this
present history emerges from the singular and incidental encounter in the present. It
develops a history from within the current events recurring to the past that has never
happened as such.
3. Rheinberger
In his contributions to the historical epistemology of modern biology, Hans-Jrg
Rheinberger also uses the concept of conjuncture (Konjunktur). He characterizes it as
a kind of structural linking.12 He emphasizes that the conjunctures come along with
unexpected events in the research process. They entail the reorganization,
rearrangement, and recombination of experimental systems. The conjunctures can
effectuate hybrids and bifurcations; they can operate as insemination
(Einverleibung, Ex., p. 148), grafting, or dissemination (Hist., p. 136). The
experimental systems are, for Rheinberger, linked by this material interaction (Ex., p.
150). They have fringy margins (Ex., p. 146), they are not self-contained and they
operate at the edge of their own dissolution and collapse. They linger at the border
between knowledge and non-knowledge. They operate as Rheinberger emphasizes
at the border of their breakdown (Hist., p. 135).
Conjunctures play an important role in the development of an experimental system in
so far as they create an open horizon of the epistemic situation. As unprecedented
events they bring forth solutions to questions that initially were not thought to be solved
11
12
86
With regard to Bachelards emphasis on the instant of nothingness and solitude from which action,
intention, and the reality of time emerges, I moreover want to point out the relevance of Althussers
concept of the void within the philosophical practice. The void is for Althusser also a therapeutic
remedy in the philosophical activity: it can be the basis for a theoretical as well as practical technique,
a kind of philosophical yoga for the self-analysis and self-control of every philosophical position
(MB, p. 89). In this respect the void of the aleatory materialism does not just enable a shifting reality
and materiality. In the same way, the philosophical void effectuates a differentiating philosophical
practice. It does not only enable us to develop a theoretical position; moreover, it allows us to shift,
rearrange, and reframe philosophical positions. As a critique of a philosophy that aims at eternally
valid truths, the philosophical void is thus, for Althusser, the basis for a temporary and provisional
kind of philosophy.
The expression appears only in the German version of Rheinbergers book on experimental systems.
In this passage, he also relates the term conjuncture to Althussers concept of historical conjunctures
(Ex., p.144).
by the means of the experimental system. Rheinberger emphasizes that it is a matter of

intuition as much as of the design of the experimental process itself to follow the
conjunctures or to even provoke them (Hist., p. 134). He understands the experimental
systems as machines for generating these conjunctures, coincidences, and unexpected
events. In so far as they generate an unforeseen future, they do not generate a linear and
chronological temporality.
On this basis, Rheinberger suggests a specific kind of temporality that emerges in the
experimental systems: namely a differential temporality (a concept, which he adapts
from Althusser/Balibars Lire le Capital; see Hist., p. 182; Ex., p. 198 and the
corresponding footnotes). The specific temporality of experimental systems can thus
only be understood recurring from their contingent conjunctures that form their
temporary results. This differential temporality cannot be registered in a chronological
line since it is the effect of the experimental systems and the specific mode of operation
of its differentiating machinery. It involves two aspects of contingency: the actual
unforeseen coincidences and the ongoing re-arrangement of the experimental systems
that created them.
This differential temporality of the experimental system also affects the
historiography of knowledge. It creates an anti-chronological temporality of knowledge.
As a consequence, both the temporality of the experimental system and the temporality
of its history (of the historical epistemology) are based on a retroactive activity that
re-organizes and re-arranges its elements initiated by the conjunctures of unexpected
events. It produces differential shifts and relocations. Only in the re-arrangement of the
system do the current conjunctions appear as the result of a past that has never taken
place as such. With this emphasis on the persistent disruption between the unexpected
events and the ongoing recurrent re-arrangement of the system Rheinberger aims at a
history that makes the irritation of differentiated epistemologies appear.
Conclusion
With his concepts of the instant, the rupture, and the realm of nothingness, Bachelard
creates a sphere for the confrontation of the possible, the unexpected, and the real. It is a
zone for the intervention of unforeseen events that trigger the activity of framing and reframing reality and thus create differential temporal realities.
With his conception of an aleatory materialism, Althusser aims at conceiving how a
specific reality in its necessity opens up to the contingency of incidental events and,
conversely, how the contingent event of an incidental encounter can become a necessary
relation when a persistent connection of its elements is created: when the encounter with
specific disturbances forms the new persisting framework for our reality.
Rheinberger emphasizes the recurrent activity of re-arrangement in the experimental
system that is initiated by the occurrence of unforeseen events. He points to the
differential epistemology that emerges from this fractured temporality.
87
Wulz
In the context of this conference I would like to ask what it could mean for todays
history of science (and the history of knowledge) to treat history as an experimental
system and to understand the materiality of history as contingent a given space that
can be rearranged and that gives way to interaction between its elements, a space for
being surprised about unforeseen events and outcomes. What is the moment of the void,
the coincidence, the encounter, the unexpected event from which we start to arrange and
re-arrange our experimental systems of history? What are the triggers from which we
frame our historical shifts? How do these incidental encounters appear in our recurrent
historical frames of attention? What would it mean to show these events as the starting
point for the differential history of a past that has never happened as such?
The attention to the differential historicity of knowledge, to the historicity of
materiality, to the material aspects of the research processes is thus related both to the
possibility of unforeseen events and to the idea of the void (1) in the process of
research, (2) in our epistemic conditionality, and (3) in the epistemic objects
themselves. In the laboratory as well as in the history of science (or historical
epistemology) the infinitesimal instant of the epistemic void could give us the
opportunity to ask a new, a differentiated, question of research, to re-arrange the
elements of the history in the present with regard to a moment of surprise. The
differential temporality of history relies on the current moment of incidental encounter:
a moment of confusion, of disorientation in view of the unexpected, a moment of shock.
Literature
Althusser, Louis et al. (1965). Lire le Capital, Paris: Maspero.
Althusser, Louis (1994). Une philosophie pour le marxisme. La ligne de Dmocrite,
in: id., Sur la philosophie, Paris: Gallimard [SPh].
Althusser, Louis (2005). Du matrialisme alatoire, Multitudes, vol. 21, pp. 179-194.
Althusser, Louis (2010). Materialismus der Begegnung. Spte Schriften, translated and
edited by Franziska Schottmann, Zrich: diaphanes [MB].
Bachelard, Gaston (1934). Le nouvel esprit scientifique, Paris: Alcan [NES].
Bachelard, Gaston (1938). La formation de lesprit scientifique. Contribution une
psychanalyse de la connaissance objective, Paris: Vrin [FES].
Bachelard, Gaston (1940). La philosophie du non. Essai dune philosophie du nouvel
esprit scientifique, Paris, PUF [PhN].
Bachelard, Gaston (1953). Le matrialisme rationnel, Paris: PUF [MR].
Bachelard, Gaston (1971). LIntuition de linstant [1932], Paris: Gonthier [II].
Bachelard, Gaston (1984). The New Scientific Spirit, translated by Arthur Goldhammer,
Boston: Beacon Press [NSS].
Bachelard, Gaston (2000). Dialectic of Duration [1936], translated by Mary McAllester
Jones, introduction by Cristina Chimisso, Manchester: Clinamen Press
[DDe].
Bachelard, Gaston (2001). La dialectique de la dure [1936], Paris: PUF [DD].
88
Duprel, Eugne (1933). La cause et lintervalle, ou Ordre et probabilit, Bruxelles:

Lamertin.
Guyau, Jean-Marie (1890). La gense de lide de temps, Paris: Alcan.
Le Lannou, Jean-Michel (2008). Le dpassement de lhumain. Bergson et Bachelard,
in: Bachelard et Bergson. Continuit et discontinuit? Actes du Colloque
international de Lyon, 28-30 septembre 2006, Frdric Worms, JeanJacques Wunenburger (eds.), Paris: PUF, pp. 73-93.
Rheinberger, Hans-Jrg (1997). Toward a History of Epistemic Things. Synthesizing
Proteins in the Test Tube, Stanford: Stanford University Press [Hist.].
Rheinberger, Hans-Jrg (2001). Experimentalsysteme und epistemische Dinge. Eine
Geschichte der Proteinsynthese im Reagenzglas, Gttingen: WallsteinVerlag [Ex.].
Rignano, Eugenio (1920). Psychologie du raisonnement, Paris: Alcan.
Williams, James (2005). How Radical is the New? Deleuze and Bachelard on the
Problems of Completeness and Continuity in Dialectics, Pli. The Warwick
Journal of Philosophy, vol. 16, pp. 149-170.
89
THE RUPTURE AND THE SCREW

The Structure of History According to Georges Canguilhem
and Helmuth Plessner
Thomas Ebke
The purport of my considerations today will lead me and I am well aware of this
straight into the lions den, as it were, because they are at least supposed to touch the
conceptual core of what an historical epistemology might be. My point of focus will be
the historicity of, or within, historical epistemology the specific notion of history that
seems to be at work and at stake in the critical tradition of Gaston Bachelard and
Georges Canguilhem. An imaginary reader of recent publications devoted to the
heritage of historical epistemology would not be unlikely to arrive at the impression that
a fair degree of confusion prevails with regard to the question of where exactly the
problem of history enters the arena. Thus, in a preprint published by the Max-PlanckInstitute for the History of Science bearing the title What (Good) is Historical
Epistemology? Theodore Arabatzis, for example, speaks of the Historicity of Scientific
Objects whereas Norton Wise analyses the Historicity of Scientific Explanation,
Michael Heidelberger draws attention to the Historicity of Causes and Philip Kitcher
proclaims that Epistemology without History is blind.1
Even though the subject that I wish to address the peculiar notion of history that we
encounter in historical epistemology seems quite quintessential, I will not come up
with a comprehensive assessment. Rather, I would like to confine myself in this paper
to the particular type of historical epistemology advanced by Georges Canguilhem.
According to a famous remark by Michel Foucault, Canguilhems explicit turn towards
the history of biology and medicine wasnt meant to be the simple application of
Bachelards discipline to a new field. It was meant to transform the discipline
altogether. 2 After all, one might even say that by adding a specific drive to the idea of
an historical epistemology, Canguilhem endowed that project with a conclusive
foundation. And as I am interested in this particular foundation I will concentrate on the
conception of history within Canguilhems approach.
However, I will not tackle my subject head-on. On the contrary, I will employ the
artifice of reading Canguilhem through the eyes of a German author with whom he was
entangled, although neither of the two was aware of this, in a parallactic relationship.
1
Thomas Sturm & Uljana Feest (eds.), What (Good) is Historical Epistemology?, Preprint No. 386 of
the Max-Planck-Institute for the History of Science, Berlin 2009. This Preprint contains the abovementioned articles: Theodore Arabatzis, On the Historicity of Scientific Objects, pp. 63-71;
M. Norton Wise, On the Historicity of Scientific Explanation. Technology and Narrative, pp. 39-53;
Philip Kitcher, Epistemology without History is blind, pp. 191-204. Michael Heidelbergers
contribution (Plurality and/or Historicity of Causes) could not be delivered in time to become part
of the preprint. However, the video of his presentation can be viewed at <http://www.mpiwgberlin.mpg.de/Workshops/en/HistoricalEpistemology/Session2.html>.
Michel Foucault, La vie: lexprience et la science, Revue de mtaphysique et de morale 90, No.1,
janvier-mars 1985, p. 4.
91
Ebke
This author being Helmuth Plessner a major representative of a line of thought called
philosophical anthropology which was a very pertinent movement in German
philosophy in the 1920s. Canguilhem and Plessner never noticed each other. One may
say that they were both absorbed into eminently national sceneries. It is well known
(and Foucault has underscored this point)3 that Canguilhem was a highly influential
figure of reference within the French contexts of debate, although he never really
attracted international attention, at least not in his lifetime.
To cut a long story short, Plessner suffered a similar fate. Leaving aside the fact that
philosophical anthropology, as a whole, had already been eclipsed by the dominance of
Heidegger in the 1920s, Plessner was driven into exile in the Netherlands in 1934; just
as the most crucial stage of his philosophical career was about to begin. When he finally
returned to Germany in 1951, he was appointed to a chair in sociology (not philosophy)
at the University of Gttingen. This institutional coincidence made Plessner a traveller
between the two worlds of philosophy and sociology, and it is a bittersweet irony of
history that the sociologists regarded Plessner as too philosophical, whereas in
philosophical literature he was received as just a sociologist.4 What unites Plessner
and Canguilhem is the fact that internationally, or inter-culturally, they were never
perceived as thinkers in their own right while they were still alive. The name of
Canguilhem used to be invoked alongside those of Bachelard and Cavaills, and he was
notoriously described as the teacher of Foucault. Plessners name, in turn, has
traditionally been subsumed alongside those of Max Scheler and Arnold Gehlen, and all
of these authors were labeled the antagonists of Heidegger.
Many things would have to be said about the promising perspectives of entangling
Canguilhem and Plessner in a systematic manner. However, I will not be able to
develop this relationship in this context. My presentation is merely supposed to
illustrate one piece of this large mosaic. This leads me to a final preliminary remark. If
on the following pages I outline a critique of Canguilhems vision of history in his
historical epistemology, I will do so wearing the mask of Helmuth Plessner and
experimentally assume his point of view in the question that is at stake. Nevertheless, I
should reveal that in my dissertation I have worn both masks alternately, that is I have
also tried to furnish a critique of Plessner seen through the eyes of Canguilhem.5 Within
the particular framework of this text, I will suggest the thought experiment of a
philosophical anthropologist who discovers Canguilhems approach for the first time
and feels obliged to comment on it from his own background. Needless to say, the
game also works the other way around.
As a first step in my deliberations I would like to specify the divergence introduced
by Canguilhem with regard to Bachelards paradigm. We will have to sketch what it
means in broad strokes that Canguilhem did not only apply Bachelards pattern to a new
3
4
92
Ibid.
For a well-informed monograph providing a thorough access to Plessners biography see Carola
Dietze, Nachgeholtes Leben. Helmuth Plessner (1892-1985), Gttingen 2006.
Thomas Ebke, Lebendiges Wissen des Lebens. Zur Verschrnkung von Plessners Philosophischer
Anthropologie und Canguilhems Historischer Epistemologie, Berlin 2012.
sphere, but transformed it altogether when he took the history of the life sciences that
is biology and medicine into account.
1. Transition: From Bachelard to Canguilhem

Many important theorems proposed by Bachelard have undeniably found their way into
Canguilhems notion of an historical epistemology, to be positively accepted by him.
For instance, to understand the specificity of the enterprise of science it is necessary to
register the epistemological rupture with the phenomenological evidence of everyday
experience. Canguilhem also clearly supports Bachelards emphasis on the technical
instruments that help to shape the object of a scientific procedure. And he is definitely
convinced by the dynamic of recurrence that is so crucial to Bachelards approach.
Science literally takes on the mode of verdicts, of condemnations regarding prescientific intuitions and outdated, untenable states of research. The rationality that is
proper to scientific research consists of the incessant correction of errors and this
dynamic does not spare those formations of knowledge which once represented,
overhaul themselves.
It is my impression that Canguilhem ties in with all these Bachelardian motives. His
article of 1966 Lobjet de lhistoire des sciences contains a by now famous passage
which might even have flown from Bachelards very pen:
Lobjet en histoire des sciences na rien de commun avec lobjet de la science.
Lobjet scientifique, constitu par le discours mthodique, est second, bien que
non driv, par rapport lobjet naturel, initial, et quon dirait volontiers, en
jouant sur le sens, pr-texte. Lhistoire des sciences sexerce sur ces objets
seconds, non naturels, culturels, mais nen drive pas plus que ceux-ci ne drivent
des premiers. Lobjet du discours historique est, en effet, lhistoricit du discours
scientifique, en tant que cette historicit reprsente leffectuation dun projet
intrieurement norm, mais traverse daccidents, retarde ou dtourne par des
obstacles, interrompue des crises, cest--dire de moments de jugement et de
vrit.6
If one based ones picture of what an historical epistemology is or should be on this

particular essay by Canguilhem, one would indeed learn a lesson that Bachelard has
insisted on with great vigour: that the problem is not to discuss the ontic status of an
entity which it may have apart from the scientific discourse that is conducted about it
the problem is not related to the pretext, to take Canguilhems term. Quite the
contrary, one has to understand that the object of science is a phenomenon into which
the scientific procedure invests a whole bundle of norms: research expectations,
technical infrastructures, political intentions, and so on. It would not be an exaggeration
to say that the object of science has no history before the moment it is turned into an
eminently cultural object, something endowed with the norms of a scientific practice
Georges Canguilhem, Lobjet de lhistoire des sciences, in: id., Etudes dhistoire et de philosophie
des sciences concernant les vivants et la vie [1968], Paris 1994, p. 17.
93
Ebke
that declares it as its correlate. Two paragraphs before the one I just quoted, Canguilhem
proposes the following formula:
Donc, lhistoire des sciences est lhistoire dun objet qui est une histoire, qui a une
histoire, alors que la science est science dun objet qui nest pas histoire, qui na
pas dhistoire.7
In other words, the question of history is linked to the normative stakes, the enjeux
that are inscribed into a phenomenon, and thereby create the phenomenon in the first
place. Yet scientists themselves do not deal with and do not unfold the normative
investments that constitute their very objects. This is the task of the historian of science.
He operates, as Hans-Jrg Rheinberger has clarified in view of Canguilhems essay, on
a third level of object formation.8 Only in the optics of the historian of science does
the historicity in which a scientific discourse is steeped, become transparent.
If the idea of an historical epistemology la franaise (J.-F. Braunstein) can be
reasonably retrieved from this scheme, we can now also try to explain the specific twist
that Canguilhem has given to historical epistemology against the background of this
scheme. For Bachelard, science finds itself under the constant and heavy pressure of
having to ban all of the fuzzy notions and immature intuitions that threaten to invalidate
the scientific spirit. This is why this spirit has to undergo a relentless psychoanalysis
that purges it from the myths which in one way or another still succeed in penetrating
the territory of scientific knowledge and render it impure.
Interestingly, Canguilhem touched upon the dialectical character of Bachelards
model in his 1963 essay Lhistoire des Sciences dans loeuvre pistmologique de
Gaston Bachelard. Inserting quotations from Bachelards book La Philosophie du Non
(1940), Canguilhem expresses the following comment in this text.
Cest dans la Philosophie du Non, qui se donne pour une philosophie du nouvel
esprit scientifique, que le concept de dialectique apparat, non certes comme une
catgorie, mais comme une norme de la pense pistmologique de Bachelard
() On en revient donc, encore et toujours, la relation interne, intime, de
lpistmologie et de lhistoire. Lhistoire illustre la dialectique de la pense bien
plutt quelle nest elle-mme une dialectique objective. () La philosophie du
non nest pas structure par la dialectique de lhistoire gnrale. Cest elle au
contraire qui confre lhistoire des sciences une structuration dialectique.9
Although Canguilhem makes his point in a discreet and slightly inconspicuous way,
he nevertheless seems to assert that Bachelards dialectic of the scientific spirit boils
down to a negative dialectic. His conception of history may illustrate as Canguilhem
7
8
94
Ibid., p. 16.
Hans-Jrg Rheinberger, Ein erneuter Blick auf die historische Epistemologie von Georges
Canguilhem, in: Cornelius Borck, Volker Hess & Henning Schmidgen (eds.), Ma und Eigensinn.
Studien im Anschluss an Georges Canguilhem, Munich 2004, p. 229 (eine dritte Ebene der
Objektformation).
Georges Canguilhem, Lhistoire des sciences dans luvre pistmologique de Gaston Bachelard,
in: id., Etudes dhistoire et de philosophie des sciences concernant les vivants et la vie [1968], Paris
1994, pp.180.
says the dialectics of thought, but it is not in itself part of a dialectic. The starting
point of Bachelards approach is a negation which is absolute and incapable of any
further dialectical mediation: this negation is the rupture by which the scientific spirit
excludes the intolerable erroneousness of the pre-scientific mind. Maybe one can
paraphrase the situation as follows: Canguilhem honors Bachelards systematic aim to
develop a dialectic of the scientific spirit that does not have a speculative closure, but
constantly needs to be temporalised over and over again and is, in this way, kept open.
But he is critical of the fact that Bachelard, even though he dynamizes the relationship
between history of science and the sciences, does not dynamize the relationship of
scientific knowledge with its objects.
This description finally sets the stage for our interpretation of Canguilhems shift.
We can now see why the history of the life sciences that Canguilhem pursued had to
break the scheme that Canguilhem himself reconstructed from Bachelards works.
Because if it comes to the history of biology and medicine, one can no longer cling to
the formula, evoked by Canguilhem himself, that science involves un objet qui na pas
dhistoire. Living phenomena have a history insofar as they relate to their
environment. They attach a value to the way in which they interact with the
environment; they cannot but assess their vital situation in the light of the polarity
between life and death. It is remarkable that Canguilhem speaks assertively of a history
of the living in this context, as a quotation from his essay on Le normal et la
pathologique reminds us.
Parce que le vivant qualifi vit parmi un monde dobjets qualifis, il vit parmi un
monde daccidents possibles. Rien nest par hasard, mais tout arrive sous forme
dvnements. Voil en quoi le milieu est infidle. Son infidlit cest proprement
son devenir, son histoire.10
As far as the history of the life sciences is concerned, the problem of history does not
commence with the normative operations of the scientist with the construction of a
cultural object. Instead, there is historicity already on the level of the objects of science
themselves, and in fact Canguilhem does his utmost to stress that, after all, medicine
and the same is true of biology in some ways is not a science in the first place, but
rather a technique that is directly touched by the vicissitudes of the living forms that it is
confronted with.
Thus, by proceeding from an historical epistemology of physics, as envisaged by
Bachelard, to an historical epistemology of medicine and biology, Canguilhem alters the
scheme that we have evoked. He goes beyond the scope of Bachelard because he
develops the foundation of what happens, in Bachelards view and also in Canguilhems
own view, on the levels of the scientist and the historian of science. That is, Canguilhem
comes up with a fundamental conception of normativity and he calls to mind that
norms, understood as values, are at stake prior to both the practices of the scientist and
the recurrent judgment of the historian of science. This is why, in the end, Canguilhem
plays with an affinity between the evanescence of scientific truths and the biological
10
Georges Canguilhem, Le normal et le pathologique [1943/1966], Paris 1998, p. 131.
95
Ebke
evolution of living matter. One leaves Canguilhems version of an historical

epistemology with a lesson that differs radically from the one we can draw from
Bachelard. In Canguilhem the structure of history is traced back, albeit in complicated
and often indirect ways, to its source in the activity of the living. Both the position of
the scientist and that of the historian of the sciences can only be comprehended if one
keeps this vital horizon in mind.
2. Helmuth Plessners critique of Canguilhem

I now come to the second part of my paper which is actually also intended to be the core
of my argument. I would like to look upon Canguilhems specific drive of historical
epistemology through the eyes of Helmuth Plessner who, as I have already remarked,
was in fact completely unfamiliar with Canguilhem. My main suggestion has just been
that Canguilhem offers a foundation for the projects of Bachelard (and by the way also
of Foucault) by showing that the problem of normativity, and thus the engine of
historicity, is closely intertwined with the specificity of the living.
Now it wont do any harm to call to mind that Plessner had elaborated a relationship
between the moment of history and the order of living things. In his landmark
publication The Levels of the Organic and Man, of 1928,11 Plessner presents something
that he calls a vertical comparison of the position of man with non-human life
forms.12 What Plessner tries to grasp, following Dilthey rather than Hegel, is the way
in which man arrives at understanding himself as being both the center and the object of
his experience. This situation in which man relates to himself both as the subject and the
object of experience must be taken on, according to Plessner, from two different angles.
From the vertical angle it is necessary to specify in what way humans differ from nonhuman levels of the organic and may be able to represent themselves in a way that is
incommensurable to non-humans. From the horizontal angle the comparison would
have to deal, diachronically and synchronically, with the different cultures in which
human beings interact. In fact for Plessner these two lines of investigation correspond to
the two rhythms that intersect in the human condition, even if neither of the two
rhythms is antecedent. Man stands precisely at the crossroads, or in the rupture between
nature and history. He lives only insofar as he leads his life as he makes himself the
individual that he is.
It is this irreducible ambivalence that Plessner wished to describe in his book when
he preferred to speak of living persons rather than of human beings. What
distinguishes persons from human beings is the fact that the term person first of all
stands for a particular viewpoint from which someone is able to refer to him-or herself,
11
12
96
Original German title: Helmuth Plessner, Die Stufen des Organischen und der Mensch. Einleitung in
die philosophische Anthropologie [1928], Berlin/New York 1975. In order to make the citations from
Plessners texts comprehensible to readers who do not read German fluently, I will take the liberty of
offering my own (if certainly not immaculate) translations of specific passages in this paper. To date,
there is no translation of Plessners Stufen in the English language.
Ibid., p. 32 (my translation, TE).
namely as someone who at the same time has a body, exists in a body and experiences
this distinction from the sphere of a shared world (Mitwelt) that lies outside the
interactions of the organism with its environment.13 Plessners famous expression for
this external point of reference is the eccentric positionality of man.14 Whether a
human being can effectively assume this threefold perspective of a person is a different
question. At least, or so one may presume, a human being can lose that structure as
much as they can gain it. What is more, non-human entities can also be candidates for
personhood: animals, gods, angels, demons etc. Historically speaking, the place of
persons has indeed been occupied by the most perplexing creatures.
We have now stumbled over the concept that allows us to build the bridge which
connects us again with Canguilhem. It is the concept of history. For Plessner, the
horizon of history is inextricably linked to the structure of persons. We can speak of
history only with regard to beings that live in a permanent dissociation between their
intentions and their achievements. In a sense, the condition of man is such that his life
constantly poses a challenge and remains a question for him. Now whatever man does,
whether he makes a gesture, builds a house or writes a book and these examples are
taken from Plessner himself 15 he never actually achieves what he had originally
envisaged. The structure of history, according to Plessner, is reminiscent of a screw
rather than a straight line or a circle.
Der Prozess, in dem er [der Mensch, TE] wesenhaft lebt, ist ein Kontinuum
diskontinuierlich sich absetzender, auskristallisierender Ereignisse. In ihm
geschieht etwas und so ist er Geschichte. Er hlt gewissermaen die Mitte
zwischen den beiden Mglichkeiten eines Prozesses, dessen Sinn im Fortschritt
zur nchsten Etappe besteht, und eines Kreisprozesses, der dem absoluten
Stillstand quivalent ist. Die Vorstellung also, dass der Sinn der Geschichte in
einem ihr vorschwebenden Ziel liegt, welchem sie entgegeneilt, ist ebenso unwahr
wie die entgegengesetzte Vorstellung, dass sie ein groes nunc stans bedeutet. In
der Expressivitt liegt der eigentliche Motor fr die spezifisch historische
Dynamik menschlichen Lebens. Durch seine Taten und Werke, die ihm das von
Natur verwehrte Gleichgewicht geben sollen und auch wirklich geben, wird der
Mensch zugleich wieder herausgeworfen, um es aufs Neue mit Glck und doch
vergeblich zu versuchen. Ihn stt das Gesetz der vermittelten Unmittelbarkeit
ewig aus der Ruhelage, in die er wieder zurckkehren will. Aus dieser
Grundbewegung ergibt sich die Geschichte. Ihr Sinn ist die Wiedererlangung des
Verlorenen mit neuen Mitteln, Herstellung des Gleichgewichts durch
grundstrzende nderung, Bewahrung des Alten durch Wendung nach vorwrts.16
13
14
15
16
For this context, see ibid., p. 293.

The term exzentrische Positionalitt is indeed Plessners key concept. For the development of the
argument see ibid., pp. 288.
Ibid., p. 337.
Ibid., 339 (italics in the original). My translation would be the following: The process in which he
[man, TE] essentially lives is a discontinuous continuum of contrasting, crystallizing events. In this
process something happens and thus it is history. He virtually holds the balance between the two
possibilities of a process whose sense consists in the progress to the next stage, and a circular process
which is equivalent to absolute stagnation. Hence, the notion that the sense of history lies in a goal
which history has in front of itself, and towards which it dashes, is as untrue as the opposite notion
according to which it signifies one great nunc stans. Expressivity is the actual engine of the
specifically historic dynamic of human life. Through his deeds and works, which are supposed to give
97
Ebke
In his book Power and Human Nature (1931), Plessner is more specific about the
three temporal directions that have to be balanced by living persons.
Die Vergangenheit, die uns in allem als verborgene oder bewusste Herkunft
durchdringt und in dem Rahmen der Tradition umfangen hlt, ffnet sich in das
noch zu lebende Leben der Gegenwart hinein. Und die Gegenwart, die uns in
einem anderen Sinne aus dem Woraufhin unserer Lebensfhrung umfangen hlt,
ffnet sich in das hinein, was wir faktisch schon sind, weil wir durch unsere
Vergangenheit so geworden sind.17
The salient point in all of this is that Plessner describes the process of history through
the interplay and the antagonism of two factors. Historicity arises in the difference
between that which man can attribute to himself and that which he can no longer, or not
yet, attribute to himself. Past and future are experienced as temporal orders that are
produced at the same time as they are grown. The present, however, is, as Plessner
writes, a rupture in which a person, a generation or a culture struggles to create a viable
balance of the temporal dimensions.
What happens if one scrutinizes Canguilhems paradigm which as we have seen,
embraces Bachelards paradigm based on Plessners philosophical anthropology? In
an odd way, one could argue, the structure of history, as Canguilhem has it, falls apart
into a dualism of transcendence and immanence. On the one hand, Canguilhem in a
tradition that unites him, among others, with Marx, Nietzsche and Deleuze describes
the process of history as an original productivity that can only appear as a series of
errors to the subjects situated in history. On the other hand, the position of the
epistemologist who goes about reconstructing the past of a science is by no means
indeterminate: in a certain sense, the epistemologist plays off that which he regards as
the immanent values of the objects of a science over against that which he regards as the
values which, he thinks, a science pursued at a certain point in time. In this way, a
second notion of history emerges to crisscross the first one: the future history may be a
productive game that eludes us. But in order to be able to judge the past of a science in
opposition to that science itself, as history, everything depends on the normative reading
that we, as historians of science, practise or dont practise. The history of a science
17
98
him the balance denied by nature and which do in fact give him this balance, man is at the same time
propelled out of this again, only to try it afresh with new luck but still in vain. The law of mediated
immediacy thrusts him out of the position of rest into which he wishes to return again. History results
from this basic movement. Its sense is the recovery of what was lost with new means, creation of the
balance through groundbreaking change, conservation of the old by means of a turn ahead.
Helmuth Plessner, Macht und menschliche Natur. Ein Versuch zur Anthropologie der
geschichtlichen Weltansicht [1931], in: id., Gesammelte Schriften V. Macht und menschliche Natur.
Frankfurt am Main 1981, p. 187. Again, due to the lack of an English translation of this text, I would
like to offer my own attempt at a translation in order to convey at least the intellectual substance of
Plessners argument: The past, which permeates us in everything as a concealed or sensible origin,
and which enfolds us in the framework of tradition, opens up into the life that we still have to live in
the future. And the present, which enfolds us in the here and now of our conducted life, opens into
what we effectively already are because we have become like it through our past.
refers to all the discontinuities, the agonies, the empowerments and the
disempowerments that the genesis of this science contains within itself. Yet we can only
exhibit these things through a well-intended twisting or rather, as Pierre Macherey has
it, a distorsion.
In this way, the philosophical opposition to Canguilhem can be clearly marked from
Plessners perspective. Plessner himself shows, on the one hand, that we can reconstruct
the past and design the future only if we hypothetically proceed from a rupture between
two shares that needed and need to be entangled by every past, present and coming
culture. These shares are the historical power and the natural powerlessness of man. On
the other hand, we ourselves, here and now, face the task of bringing about such an
entanglement which means to eschew the absolutization of our own historic
standpoint.
When criticized in the light of this argument, a chiasmus seems to yawn in
Canguilhems historical epistemology between the immanent perspective of the present,
in which history is entirely appropriated, and the transcendence of a future whose
history can be ascribed to nothing and nobody because, on this reading, history presents
itself as a complex game which evolves an unpredictable productivity. In Canguilhem,
the place of the entanglement by historical power and natural powerlessness is taken by
an unrelated juxtaposition of absolute power and absolute powerlessness: a duality
which also inscribes itself in Canguilhems concept of life and which goes back last, but
not least, to his dependence on Nietzsche.
Literature
Arabatzis, Theodore (2009). On the Historicity of Scientific Objects, in: Thomas
Sturm & Uljana Feest (eds.), What (Good) is Historical Epistemology?
Preprint No. 386 of the Max-Planck-Institute for the History of Science,
Berlin, pp. 63-71.
Canguilhem, Georges (1998). Le normal et le pathologique [1943/1966]. Paris.
Canguilhem, Georges (1994). Lobjet de lhistoire des sciences, in: id., Etudes
dhistoire et de philosophie des sciences concernant les vivants et la vie
[1968], Paris, pp. 9-23.
Canguilhem, Georges (1994). Lhistoire des sciences dans luvre pistmologique de
Gaston Bachelard, in: id., Etudes dhistoire et de philosophie des sciences
concernant les vivants et la vie [1968], Paris, pp. 173-186.
Dietze, Carola (2006). Nachgeholtes Leben. Helmuth Plessner (1892-1985), Gttingen.
Ebke, Thomas (2012). Lebendiges Wissen des Lebens. Zur Verschrnkung von
Plessners Philosophischer Anthropologie und Canguilhems Historischer
Epistemologie, Berlin.
Foucault, Michel (1985). La vie: lexprience et la science, Revue de mtaphysique et
de morale 90, No.1, pp. 3-14.
Kitcher, Philip (2009). Epistemology without History is blind, in: Thomas Sturm &
Uljana Feest (eds.), What (Good) is Historical Epistemology? Preprint No.
99
Ebke
386 of the Max-Planck-Institute for the History of Science, Berlin, pp. 191204.
Plessner, Helmuth (1975). Die Stufen des Organischen und der Mensch. Einleitung in
die philosophische Anthropologie [1928], Berlin/New York.
Plessner, Helmuth (1981). Macht und menschliche Natur. Ein Versuch zur
Anthropologie der geschichtlichen Weltansicht [1931], in: id., Gesammelte
Schriften V. Macht und menschliche Natur. Frankfurt/Main, pp.135-234.
Rheinberger, Hans-Jrg (2004). Ein erneuter Blick auf die historische Epistemologie
von Georges Canguilhem, in: Cornelius Borck, Volker Hess & Henning
Schmidgen (eds.), Ma und Eigensinn. Studien im Anschluss an Georges
Canguilhem. Munich, pp. 223-237.
Wise, M. Norton (2009). On the Historicity of Scientific Explanation. Technology and
Narrative, in: Thomas Sturm & Uljana Feest (eds.), What (Good) is
Historical Epistemology? Preprint No. 386 of the Max-Planck-Institute for
the History of Science, Berlin, pp. 39-53.
100
KURT GOLDSTEINS CONCEPTION OF INDIVIDUALITY

Around the time of his flight from Nazi Germany in 1933-34, Kurt Goldstein wrote a
major theoretical work drawing on some thirty years of his neurological and psychiatric
studies. According to the established myth, Goldstein basically unloaded these years
and studies in a five-week-long dictation of the massive Der Aufbau des Organismus
(The Structure of the Organism) during the short period he spent in Holland before
emigrating definitively to the United States, where with the support of the Emergency
Committee In Aid of Displaced Foreign Scholars, he was first appointed Clinical
Professor of Neurology and Lecturer in Psychopathology at Columbia University in
1935.1 To hear Oliver Sacks tell it, Goldstein dictated The Organism almost nonstop
over a period of five weeksleaving himself (and his typist) in a state of prostration.2
Soon thereafter, Goldstein prepared his book for publication in English, and it appeared
as The Organism in 1939.
This paper aims to demonstrate that the novelty of Goldsteins argument in The
Organism resides in its epistemology of life and its philosophical anthropology; that it
is based on a particular and novel theorization of individuality; and that as a result, the
work formed part of a significant revision of physiology that took place in the 1920s
and 1930s.
1. Goldsteins account of the organism was fundamentally premised on the two
linked objects of his research: first, his extensive writings on aphasia, which dated to
19063 and had already culminated in a book-length publication, ber Aphasie in 1927,4
and second, his WWI-era observation and re-training of brain-wounded soldiers, first as
Professor of Neurology and Psychiatry at the University of Berlin, and then at the newly
established Brain Injury Institute (Institut zur Erforschung der Folgeerscheinungen von
Hirnverletzungen) in Frankfurt. Together with his collaborator, the Gestalt psychologist
Adhmar Gelb, Goldstein challenged a series of central elements of contemporary
With a fellowship underwritten by the Rockefeller Foundation, he would go on to teach at Tufts

College Medical School from 1940 to 1945, and in the winter of 1938-1939, was invited to give the
William James Lectures at Harvard University.
Oliver Sacks, Foreword, in Kurt Goldstein, The Organism, New York: Zone Books, 1995, p. 9.
Sackss anecdote dramatizes Marianne Simmels version in the introduction to The Reach of the Mind,
New York: Springer, 1968, p. 9: Alone, living in a boarding house without family, without any
clinical or teaching responsibilities, a period of enforced, uninterrupted leisure which he used to write
what was to be his magnum opus, The Organism. He dictated the basic draft of this long and complex
book in a matter of five or six weeks, leaving his secretary in a state of utter exhaustion.
Kurt Goldstein, Ein Beitrag zur Lehre von der Aphasie, Journal fr Psychologie und Neuorologie, 7
(1906): 172-188.
Kurt Goldstein, ber Aphasie, Zurich: Orel Fuessli, 1927.
101
Geroulanos & Meyers
accounts, particularly the psychology of reflex,5 and the neurological localization of

mental pathologies.6 Aphasia should not be explained by way of localization, Goldstein
argued, as this method denied the tremendous complexity of pathological
performances exhibited by the organism. Brain injuries in war, moreover, showed that
for the wounded soldier the localization approach offered little by way of therapy, for it
failed to understand the restriction of his abilities resulting from irreparable damage to
areas of brain tissue.
Most interesting to Goldstein and Gelb were cases on the physical end of the
spectrum of Kriegsneurose, cases that exhibited actual neurological damage (not merely
psychic trauma)7 and where the damage was not so devastating as to disallow all
treatment and block all adaptation to a new, restricted kind of normality. These cases
demanded the application of psychologic methods in the clinic, to quote Goldsteins
subtitle to his 1942 Aftereffects of Brain Injuries in War. What this meant, for
Goldstein, was the sustained clinical study of, and care for, individual patients. Pointing
out that only some 4-10% of brain-injured soldiers survived during the period from
1914 to 1917, Goldstein emphasized the breadth of his systematic study of some 2000
patients, but especially the detailed extent of his research, namely working on some 90
of them regularly over the following decade.8 In Aftereffects of Brain Injuries in War,
which he wrote at least in part with a new wartime effort in mind,9 Goldstein repeated
accounts of the type of damage and consequences he was referring to. One 34-year old
patient had been injured by a projectile that had entered the head through the left eye,
passed through the skull, and come out behind the left ear10 and suffered from amnesic
aphasia (today known as anomia) a condition characterized by the inability to name
objects or identify abstract qualities. Another patient, 24 years old, had been injured by
shrapnel in the left side of the skull between the forehead and the ear [such that] it
was assumed that the base of the third frontal convolution was injured.11 Goldstein
noted the absence of neurologic symptoms and the facility of comprehension, which
were matched, however, by an absence of words with concrete meanings; he further
noted that the patient could not identify objects in tests nor understand accented
inflection. He did not classify this patient as suffering from amnesic aphasia as with the
earlier patient but emphasized a visual disturbance that placed him under the category of
other disturbances.12 The same concern with how a patient handles loss, and how he
retrains himself and may be retrained in matters of language and vision, punctuated the
5
6
8
9
10
11
12
Goldstein, The Organism, p.70, p.71, p.78 and following.

See essays in the volume Goldstein edited with A. Gelb, Psychologische Analysen hirnpathologischer
Flle, Leipzig: J. Barth, 1920.
Goldstein remained concerned about possible neurotic origins of certain behaviors and insisted on
repeated occasions that his focus was not on mere neurosis or shell shock. Kurt Goldstein,
Aftereffects of Brain Injuries in War, New York: Grune & Stratton, 1942, p.25.
Goldstein, Aftereffects of Brain Injury in War, p.13.
[This book] will not only be welcomed by his many friends, but is also particularly apposite at the
present time, when [the problems in applied psychology presented by the sequelae of cerebral lesions]
is likely to assume dimensions of national importance. D. Denny-Brown, Foreword, in Goldstein,
Aftereffects of Brain Injuries in War, p.9.
Goldstein, Aftereffects of Brain Injury in War, p.152.
Ibid., p.159.
Ibid., p.160.
102
description of other wounds, especially damage to the frontal lobes.13 The most
important patient, however, for Gelb and Goldstein, was a 34-year-old laborer
colloquially known as Schneider who had been struck in the occipital lobes twice, one
wound clearly penetrating into the brain.14 As Anne Harrington has noted, Schneider
became not merely the subject of several publications by Gelb and Goldstein, but a
veritable Anna O for them because he passed all the traditional tests aiming to point out
reduced capacities (tests ostensibly premised on localizationism and atomistic
symptomatology), but upon closer observation he was incapable of a natural
performance of most of the tasks he had managed. (It is worth noting at this point that
Maurice Merleau-Ponty spends several pages in both his 1938 The Structure of
Behavior and even more in his 1945 Phenomenology of Perception on Schneider.)
Reading, for example, required
a series of minute head- and hand-movementshe wrote with his hand what his
eyes saw. He did not move the entire hand as if across a page, but wrote the
letters one over the other, meanwhile tracing them by head-movements. An
especially interesting aspect of the case was the patients own ignorance of using
this method If prevented from moving his head or body, the patient could read
nothing whatever If required to trace a letter the wrong way, he was quite at a
loss to say what letter it was even circles had to be traced by head-movement
before the patient could say what they were.15
2. Individual cases mattered tremendously to Goldsteinhe went so far as to note

that, quote, important though it may be to seek repeated confirmation of our findings
through new case material, such confirmation adds nothing essential to our
knowledge.16 The specific goal was to [construct] the total picture of the patient,
which allowed him and Gelb to predict with relatively great certainty how he will
behave in any situation.17 This is because, as suggested already, localization was not
the entire object of the critique, nor were restricted, particular responses to tests what
mattered. Central to Goldsteins argument in The Organism was a refusal of atomistic
symptomatology, that is to say of the tendency to identify a single central cause behind
symptoms, to focus on primary symptoms.18 In his theorization of diagnostic and
therapeutic efforts, already before The Organism, Goldstein charged that atomism went
hand-in-hand with a thin comparative analysis that centered on lesions rather than a
deeper understanding of the way the organism attempts to compensate for the loss of
13
14
15
16
17
18
Goldstein, The Organism, p.58.

Goldstein, After-Effects of Brain Injury in War, p.149.
Adhmar Gelb and Kurt Goldstein, Analysis of a Case of Figural Blindness in W. D. Ellis, ed., A
Source Book of Gestalt Psychology, London: Routledge and Kegan, 1938, pp.318-319. Also
(originally) Gelb and Goldstein, I. Abhandlung. Zur Psychologie des optischen Wahrnehmungs- und
Erkennungsvorganges. Psychologische Analyse hirnpathologischer Flle auf Grund von
Untersuchungen Hirnverletzter, Zeitschrift fr die gesamte Neurologie und Psychiatrie 41 (1918),
p.I, pp.18-19, p.21, p.31.
Ibid., p.41.
See ibid., pp.34-35. Also, Goldstein, Selected Writings/Ausgewhlte Schriften. edited by Aron
Gurwitsch, E. M. Goldstein Haudek, and William E. Haudek, The Hague: Martinus Nijhoff, 1971,
pp.155-156.
103
Geroulanos & Meyers
functions.19 This was the heart of the critique of reflex: atomism presumed that
individuals were affected in the same way, and comparative analysis only replicated
itself accordingly. By attending to the individual organism, Goldstein refused to see
facts, reflexes, and particular disturbances merely as factsasking instead what kind
of a fact an observed phenomenon represents.20 In other words, at stake for the patient
and the therapist were not facts or lesions themselves but what Goldstein referred to by
performance as this was affected by the disturbance.21 The organisms original
ordered behavior was disrupted; the organism found itself performing in a disordered
fashion that sought the restoration of order.
In an ordered situation, responses appear to be constant, correct, adequate to the
organism to which they belong, adequate to the species and to the individuality of
the organism, as well as to the respective circumstances. The individual himself
experiences them with a feeling of smooth functioning, unconstraint, well-being,
adjustment to the world, and satisfaction, i.e. the course of behavior has a definite
order, a total pattern in which all involved organismic factorsthe mental and the
somatic down to the physico-chemical processes-participate in a fashion
appropriate to the performance in question. And that, in fact, is the criterion of a
normal condition of the organism.22
It bears repetition that all involved organismic factorsthe mental and the somatic
down to the physico-chemical processes are involved in this performance. Certain
kinds of disorder could result in a catastrophic reaction where the organism found
itself failing through and through to handle any of the environments requirements. In a
passage that intentionally conflates a disordered situation with the presence of
disorder around the patient, Goldstein writes:
The principal demands which disorder makes upon them are: choice of
alternatives, change of attitude, and rapid transition from one behavior to another.
But this is exactly what is difficult or impossible for them to do. If they are
confronted with tasks which make this demand, catastrophic reactions,
catastrophic shocks, and anxiety inevitably ensue. To avoid this anxiety the
patient clings tenaciously to the order which is adequate for him, but which
appears abnormally primitive, rigid and compulsive to normal people. In other
words, the sense of order in the patient is an expression of his defect, an
expression of his impoverishment regarding an essentially human trait: the
capacity for adequate shifting of attitude.23
Order/disorder, compensation, and indeed anxiety are central points in the more
generalized biological and physiological account of The Organism, and in Goldsteins
effort to counter reflex theory (on the basis of the claim that reflexes are themselves
milieu-based and not of equal consequence in different contexts) and atomistic
conceptions of the organism. Moreover, in The Organism, Goldstein did not see disease
19
20
21
22
23
For the critique of comparative analysis, see Goldstein, The Organism, pp.41-42.
Ibid., p.28. Gurwitsch comments on this passage in Studies in Phenomenology and Psychology,
Evanston: Northwestern University Press, 1966, p.80.
Ibid., p.48.
Ibid., p.54.
104
or disturbance as the sole origin of a disruption of behavior. Rather, the organisms

response to such a disturbance within its given milieu was a, if not the, crucial
determinant. Disease and disturbance forced the organism into an adaptive, conservative
situation where its disordered performance tended toward compensation or
readjustment.24 Specifically with regard to the problem of central cortical injury
including the aforementioned cases that resulted in visual and linguistic disturbance,
Goldstein offered a further point, namely that their effect on the organism was that it
destroyed patients capability for abstract thought and left them with only a concrete
behavior. Disease and disturbance caused the organism to lose its abstract capacity. The
patient directly referenced in the following passage is Schneider, but clearly the point is
more general:
whenever the patient must transcend concrete (immediate) experience in order
to actwhenever he must refer to things in an imaginary wayhe fails. On the
other hand, whenever the result can be achieved by manipulation of concrete and
tangible material, he performs successfully. Each problem that forces him beyond
the sphere of immediate reality to that of the possible, or to the sphere of
representation, ensures his failure. This manifests itself in all responses The
patient acts, perceives, thinks, has the right impulses of will, feels like others,
calculates, pays attention, retains, etc., as long as he is provided with the
opportunity to handle objects concretely and directly. He fails when this is
impossible. This is the reason why he does not succeed in intelligence tests. This
is also the reason why he can grasp a little story as long as it concerns a familiar
situation in which he himself has participated. But he will not understand a
storycertainly no more difficult for the average personrequiring him to place
himself, in imagination, in the position of someone else. He does not comprehend
metaphors or puzzles. He can manipulate numbers in a practical manner, but has
no concept of their value. He can talk if there is some concrete subject matter
present but cannot recount material unrelated to him, or report it purely
conceptually. He is incapable of representation of direction and localities in
objective space, nor can he estimate distances; but he can find his way about very
well, and can execute actions that are dependent upon perception of distance and
size. The most general formula to which the change can be reduced is probably:
The patient has lost the capacity to deal with that which is not realwith the
possible. 25
Abstract behavior, an attribute of the human being, he writes later in the book,
involves the ability of voluntary shifting, of reasoning discursively, oriented on selfchosen frames of reference, of free decision for action, of isolating parts from a whole,
of disjoining given wholes, as well as of establishing connections.26 It bears note that
over time, Goldstein would become more emphatic on this superposition-and-collapse
of the abstract attitude on the concrete. In opposition to the abstract, which for him
embraces more than merely the real stimulus in its scope, the concrete attitude is
realistic and does not imply conscious activity in the sense of reasoning, awareness,
or a self-account of ones doing.27
24
25
26
27
Ibid., p.336.
Ibid., pp.43-44.
Ibid., p.301.
Kurt Goldstein and Martin Scheerer, Abstract and Concrete Behavior, Psychological Monographs
53:2 (1941): p.2, p.3.
105
Geroulanos & Meyers
Crucially, the loss of the abstract attitude was central to the elimination of the old
order and to the reduction of the patient to a situation where all he could cope with was
the immediate, concrete situation surrounding him. That allows for a new, reduced order
to emerge, but the loss of words and of the capacity to understand visual wholeness and
found himself amidst a potentially catastrophic situation. What has happened, Goldstein
emphasizes, to the cortically-injured patient is a systemic disintegration rather than
the mere loss of particular performances, as those inevitably affect the organisms
performance wholeness.28 Faced with abstract demands, a patient would begin to sense
intense anxiety at his inability to fulfill them, and would seek emphatically to return to
orderhowever reduced and restricted. These catastrophic situations may involve a
further reduction in the capacity to cope, and the organism, operating without abstract or
discursive thought, performs merely with a restitutio ad integrum in mindhowever
restricted that integrum might be.
3. In order to address the specificity of Goldsteins revision of existing conceptions
of physiology the organisms mental and physiological reality under and resulting
from disease, distress, and disordera reality that is as much actualized as
enacted)it is first necessary to discuss his notion of the individual. It would be
careless to make the claim that Goldsteins concern with individuality is simply borne
out of a methodological perspective. For Goldstein, pathological data is made
meaningful as it arises from the solitary patient, but only when comparing the individual
with himself in similar or different situations, or with himself in a previous state felt by
this patientand it is here that order and disorder hold meaning. In his writings on the
individual, as such, we find a unit of analysis but more importantly also the basis of a
critique of collectivism and classical norms,29 as for Goldstein, the ability to claim
health or illness is a singular human trait that goes hand in hand with the promise of
autonomy,30 of escape from threat, for Well being consists of an individual norm of
ordered functioning.31 As Jean Gayon points out, this fact is not surprising after his
forced leave of Hitlers Germany in 1934. Indeed, when Goldstein writes, The
statistical norm concept cannot do justice to the individual, (our emphasis)32 we are
forced to deal with the implications of what Goldstein means by justice, implications
particular to this moment. Dislodging the identity of the norm with the normal, the
average, the healthy, Goldstein negates the socio-political implications that would
oppose this normality to pathological alternatives, the degenerate or the diseased. What
more powerful theoretical alternative could one imagine to a politicization of biological
life that, in Goldsteins case, had removed him from his social milieu, and would soon
move from marginalization and removal toward extermination.
28
29
30
31
32

See the section Individual and Others, in Goldstein, Human Nature in the Light of Psychopathology,
Cambridge: Harvard University Press, 1940.
Jean Gayon, The Concept of Individuality in Canguilhems Philosophy of Biology, Journal of the
History of Biology, 31/3 (1998), pp.309-310. As well as M. Simmel, Kurt Goldstein: A Life in id.,
The Reach of the Mind, New York: Springer, 1968.
Ibid., p.329.
106
In Goldsteins conception of the individual patient we find the ground for a

philosophy of biology inseparable from a methodology for the study of pathological
phenomena. We find the conditions of experienceof disease and disorder in the
patient, as well as conditions that refuse a generalization out of intervention, treatment,
and healing. We find the motivations and activities of the organism freed from a
misattribution of poorly conceived biological determinism. Regarding this last point,
Goldsteins methodology does not reject discernable patterns of pathological
phenomena (particularly in the case of organic damage as the result of brain injury), but
this pattern would have to be found in the individual himself, as would the conditional
norms delineating the normal and the pathological. To quote Georges Canguilhem:
The borderline between the normal and the pathological is imprecise for several
individuals considered simultaneously but it is perfectly precise for one and the same
individual considered successively.33 Goldstein sees a contest between individual
norms and norms provided by a statistical mean. Despite (or precisely because)
individual norms result from disorder and debilitation, the individual is able to carve out
a space where new norms take hold, however tenuous. Goldstein argues that even the
individuals recognition of a pattern of somatic experience that differs from previous
patterns understood by the patientno matter how imprecisebecomes the basis of
healing. In a long passage from a chapter of The Organism, On Norm, Health, and
Disease, he offers a description along the lines of the clinical encounter:
But let us put aside, for the time being, this problem of definition and see how
patients themselves and the physician go about distinguishing health and disease.
I believe they proceed at first by not focusing on the contents. True, the physician
as well as the patient can become suspicious as to health, when recognizing
deviations for the usual behavior with regard to content, as for example, abnormal
fatigue, palpitations of the heart, nausea, headache, swollen feet, and so on. But
neither for the physician nor for the patient are these manifestations of diseases in
themselves but, at best, signs that disease may exist. The experience of being sick
does not necessarily contain any definite kind of change as to contents.34
Health, illness and recovery cannot rely on a statistical norm, because these states are
founded upon the individuals capacities. As Canguilhem describes, What Goldstein
pointed out in his patients is the establishment of new norms of life by a reduction in the
level of their activity as related to a new but narrowed environment.35 But what, if
anything, does this have to do with a revised understandingan alterationin the
conceptualization of physiology? Here are three suggestions:
a) Pathological data begins with the individual.
In the work of his
contemporaries, the organisms return to a state of equilibrium (homeostasis in
the work of Walter Cannon) is the return to a given order. In Goldsteins
work, the ill or disturbed organism returns to a very definite state, one that is
self-directed but often vastly narrowed vis--vis its milieu.36 His experimental
33
34
35
36
Georges Canguilhem, The Normal and the Pathological, New York: Zone Books, 1989, p.182.
Goldstein, The Organism, pp.327-328.
Canguilhem, The Normal and the Pathological, p.185.
107
Geroulanos & Meyers
methodology centers upon the individual and compares the organism with
itself, either between the same or different situations, or from a previous felt
state with the one at present. Even under experimental conditions, the milieu
is imposed on the organism, but is a milieu nonetheless.37In his William
James lectures, in a chapter beautifully entitled Coming to Terms with the
World, he begins his critique of the concept of reflex with a question: Now
what are the means of determining the capacities of the individual?
Observation under the atomistic method reveals a great variety of
phenomena.38
b) Goldstein actively erases the line separating mental and physical phenomena in
the organism.
c) Finally, and perhaps most importantly for the point regarding revision,
Goldstein demonstrates that the organisms response to disease and disorder
strives towards adaption opposed to reaction. As Canguilhem writes, A life
that affirms itself against the milieu is a life already threatened.39 Goldstein
did not see physiology as series of normal functions held together with
conservative and defensive tendencies, but rather saw only conservative
tendencies in the experience of disease. As a result, a healthy organism is one
characterized by the tendency to face new situations and institute new
norms.40
4. If we look at other histories, Goldstein sits similarly athwart. Goldstein is largely
left aside of histories of physiology proper. And it is largely contingent in histories of
neurologyGoldsteins neurology having been largely explicated already. In his
foreword to the 1995 reissuing of The Organism, Sacks calls Goldstein one of the now
most forgotten figures in the history of neurology and psychiatry.41 One may wonder
about the dramatic element of this descriptionpresumably some others are yet more
forgottenbut the point gives a sense of the awkward place of The Organism,
particularly following the acceptance of Gerald Edelmans neural Darwinism
paradigm that Sacks credits with recasting the neurological debate and rendering much
of Goldsteins thought obsolete. In the history of holism, to take up Anne Harringtons
well-known book Reenchanted Science, Goldstein stands out as someone who did not
quite belong to the reenchantment movement, and did not espouse organicism.42
Rather than the work of a vitalist opposed to mechanization, rather than a work that
belongs within the spectrum of the conservative revolution such as those of Driesch,
37
38
39
40
41
42
Georges Canguilhem, Knowledge of Life, New York: Fordham University Press, 2008, p.111.
Goldstein, Human Nature in the Light of Psychopathology, p. 120; also see Maurice Merleau-Ponty
the classical problem of obtaining conflicting results when trying to reproduce consistent reflex across
various subjects in the laboratory, The Structure of Behavior, New York: Beacon, 1967, pp.44-45.
Canguilhem, Knowledge of Life, p.113.
Canguilhem, The Normal and the Pathological, p.204.
Sacks, Foreword, p.7.
Anne Harrington, Interwar German Psychobiology: Between Nationalism and the Irrational,
Science in Context 4 (1991): 429-447; cited in Mitchell Ash, Gestalt Psychology in German Culture,
1890-1967, New York: Cambridge University Press, 1998, p.282.
108
Klages, or von Uexkll (only the last of whom seems important to Goldstein), The
Organisms forceful refusal of functionalism does not lead to just any holism, to a
generalized psychology, but to an effort to explain the individual case by referring to its
individual, specific way of handling its situation.43 Indeed, rather than holism we find
here a new kind of integral individuality: an individuality that in situations of disease
finds itself broken down, divided, and yearning for a reconstruction, indeed for any
reconstruction that promises to preserve it in some form. What Goldstein does not say,
and what commentators on his work only circle around is this: the physiological
individual strives for wholeness: it is not given whole, and it can be broken up. It is
constantly threatened. But life here is a relation to the production of oneself as
autonomous, as non-threatened, as normative, as individualnot least in the literal
sense. Life, even animal life, maintains itself by instituting its own norms, by seeking to
maintain its individuality. [Animal behavior] points to an individual organization, on
which basis alone it becomes intelligible as the expression of the tendency to actualize
itself according to the circumstances.44 This is the fundamental novelty.
43
44
Goldstein refuses vitalism as well as finalism in The Organism, p.323.

Ibid., p.355.
109
GEORGES CANGUILHEM
pistmologie historique et/ou histoire philosophique?
Claude Debru
Luvre de Georges Canguilhem a t enrle sous la bannire de lpistmologie

historique, une expression quil a lui-mme utilise,1 mais qui ne saurait se rduire
une quelconque vulgate touchant linterprtation philosophique de ce qui se passe
lintrieur des sciences. Je voudrais tenter de le montrer par lexamen du premier texte
publi dans lequel Canguilhem rflchit sur lhistoire des sciences, ses objets et
mthodes, savoir larticle sur la thorie cellulaire publi en 1945 dans des Mlanges
de la Facult de Strasbourg et repris dans La connaissance de la vie. On y lit: Le
bnfice dune histoire des sciences bien entendue nous parat tre de rvler lhistoire
dans la science. Lhistoire, cest--dire selon nous, le sens de la possibilit. Connatre,
cest moins buter contre un rel que valider un possible en le rendant ncessaire. Ds
lors, la gense du possible importe autant que la dmonstration du ncessaire. 2 Il est
assez significatif de voir ici un philosophe qui va consacrer ses rflexions mettre en
regard lune de lautre la vie et la connaissance de la vie, la vie et le concept de la vie,
introduire dentre de jeu la catgorie du possible dans sa discussion de lhistoire des
sciences avant de lappliquer sa prsentation de la thorie cellulaire. Ces deux thmes,
lhistoire des sciences et la thorie cellulaire, formeront le sujet de cet expos, avec une
question sous-jacente: de quoi lhistoire des sciences est-elle rellement lhistoire? Mais
il convient aussi de tenir compte des circonstances, qui ne sont pas trs loin derrire la
rdaction de ce texte publi en 1945, des enjeux, aussi bien intellectuels que politiques,
de ces thmes et donc il convient de proposer des cls de lecture adquates. Deux cls
peuvent tre proposes. La premire concerne lpistmologie et lhistoire des sciences
au sens classique. Linsistance sur la dimension de possibilit, qui fonde lhistoricit de
la science, recle une position antidogmatique concernant la conception mme de la
science, une insistance sur la fragilit des thories scientifiques. Sur ce point,
Canguilhem mentionne certaines rserves sur la validit descriptive gnrale de la
thorie cellulaire.3 Comme toute autre thorie scientifique, la thorie cellulaire a
certaines limites et fragilits, qui invitent ne pas en faire un dogme. Lun des objets de
la dmonstration de Canguilhem est plutt de montrer le travail de limagination
thorique luvre dans la naissance et le dveloppement de la thorie cellulaire. Mais
comme toujours chez Canguilhem il y a plusieurs niveaux de lecture de ses textes, et
laspect proprement pistmologique de mise lpreuve dune thorie scientifique
saccompagne dautres considrations sur lesquelles il convient dinsister. En effet, pour
qui lit (et relit) que lhistoire, cest le sens de la possibilit, il est impossible de ne pas
lire lhistoire avec un grand H, celle qui vient de se drouler, et qui a vu un possible
vaguement entrevu devenir le facteur dun enchanement ncessaire. Il ne sagit plus

1
2
3
Cf. la prsentation de Camille Limoges dans ce volume (p. 53-66).

Georges Canguilhem, La connaissance de la vie, Paris: Vrin, 1965, p. 47.
Ibid., p. 75.
111

Debru
seulement de science, il sagit manifestement dautre chose. Et donc, lpistmologie

historique est aussi prendre dans le sens de la grande histoire. Cest peut-tre le point
le plus important de cette dmonstration.
La thorie cellulaire a donc dautres aspects, elle ouvre dautres considrations que
la pure pistmologie, et dabord elle ouvre linvitable comparaison des organismes
multicellulaires et des socits animales ou humaines, avec leurs divers modes
dassociation. Ici se pose une question assez cruciale, celle de lindividualit, de ses
niveaux ou degrs, et de ses modes dassociation. Cest une question o la logique et la
smantique rejoignent la philosophie politique et la sociologie. Lhistoire du concept
de cellule, remarque Canguilhem, est insparable de lhistoire du concept dindividu.
Cela nous a autoris dj affirmer que des valeurs sociales et affectives planent sur le
dveloppement de la thorie cellulaire. 4 Sagissant de Haeckel, il ajoute: une
philosophie politique domine une thorie biologique .5 Essayons de rsumer quelles
sont donc les valeurs ou les diffrences de philosophie politique qui peuvent
imprgner diffremment la thorie cellulaire. Cet exercice est intressant. Il montre que
ce quon appelle thorie cellulaire, et qui peut tre saisi dans un petit nombre de
propositions sur la composition et la gense des organismes, peut se laisser dominer par
diffrentes philosophies politiques.
Mais dabord, remarquons avec Canguilhem que le terme de cellule est lobjet dune
surdtermination affective considrable, dun surinvestissement par certaines
valeurs attaches certaines mtaphores, valeurs affectives et sociales de coopration
et dassociation .6 La cellule vgtale est comme la cellule de cire de la ruche, a
remarqu Haeckel.7 Ce quvoque cette mtaphore est lide dun travail coopratif.
Cette mtaphore se trouve assez frquemment, y compris chez des auteurs antrieurs
la thorie cellulaire proprement dite, comme Buffon. La thorie des molcules
organiques de Buffon est un mlange datomisme et dassociationnisme qui appelle une
interrogation sur les formes et mcanismes de lassociation et donc sur les proximits ou
les diffrences plus ou moins grandes entre les trois niveaux que forment les socit
humaines, les socits dinsectes, et les organismes. La proximit la plus grande, la
mtaphore la plus oprative, se trouve donc chez Buffon entre les socits dinsectes,
vues comme des assemblages dindividus, et les organismes, vus comme des
assemblages de ces composs individuels que sont les molcules organiques,
assemblages effectus selon le moule intrieur qui assure la conservation de la
forme macroscopique dans lassemblage des parties.8 Une question se pose ici avec une
certaine acuit, celle de la nature et du degr de lindividualit de ces parties
constitutives que sont les molcules organiques, fruits de limagination de Buffon, tres
thoriques, tres de raison, jalons sur la voie de la formulation de la thorie cellulaire,
dont il sagit de montrer quelle dpend moins de faits que de thories prexistantes.
Cest dire quel point la thorie cellulaire est vue fortement comme thorie. Si nous

4
5
6
7
8
Ibid., p. 62.
Ibid., p. 70.
Ibid., p. 49.
Ibid., p. 49.
Ibid., p. 58.
112

GEORGES CANGUILHEM
restons encore un instant avec Buffon, nous navons gure de comparaison possible
entre les socits humaine et les organismes, car les socits humaines et les socits
dinsectes ne fonctionnent pas sur les mmes principes.
Cela a peut-tre pour consquence que ce serait la formulation plus avance de la
thorie cellulaire qui permettrait en quelque sorte la philosophie politique de sinsrer
dans la discussion des rapports entre organismes et socits humaines. Le maillon
suivant dans la gense de la thorie cellulaire est Lorenz Oken, cit par Canguilhem
aprs Marc Klein biologiste mdecin strasbourgeois que Canguilhem aimait
beaucoup, soit dit en passant. Lorganisme, selon Oken, est compos danimalcules. Il
se produit ici, crit Oken, une vritable interpntration, un entrelacement et une
unification de tous les animalcules. Ils ne mnent plus de vie propre partir de ce
moment. Ils sont tous mis au service de lorganisme plus lev, ils travaillent en vue
dune fonction unique et commune, ou bien ils effectuent cette fonction en se ralisant
eux-mmes. Ici aucune individualit nest pargne, elle est ruine tout simplement
les individualits runies forment une autre individualit, celles-l sont dtruites et
celle-ci napparat que par la destruction de celles-l. 9 Commentaire de Canguilhem:
Lorganisme nest pas une somme de ralits biologiques lmentaires. Cest une
ralit suprieure dans laquelle les lments sont nis comme tels. Oken anticipe avec
une prcision exemplaire la thorie des degrs de lindividualit Lorganisme est
conu par Oken limage de la socit mais cette socit ce nest pas lassociation
dindividus telle que la conoit la philosophie politique de lAufklrung, cest la
communaut telle que la conoit la philosophie politique du romantisme. 10 Novalis,
Hegel sont mentionns un peu plus loin, o il est dit que le romantisme a interprt
lexprience politique partir dune certaine conception de la vie, la conception
vitaliste. Mais poursuivons cette enqute franco-allemande: Au moment mme o la
pense politique franaise proposait lesprit europen le contrat social et le suffrage
universel, lcole franaise de mdecine vitaliste lui proposait une image de la vie
transcendante lentendement analytique. () La vie est une forme irrductible toute
composition de parties matrielles. La biologie vitaliste a fourni une philosophie
politique totalitaire le moyen, sinon lobligation, dinspirer certaines thories relatives
lindividualit biologique. 11 Par totalitarisme, il faut entendre la ngation de lindividu
dans le tout.
Venons-en un bien trange pisode, celui du rejet de la thorie cellulaire par
Auguste Comte, qui la considrait comme une rverie allemande, contraire lesprit du
positivisme.12 Mais Canguilhem, soucieux de dfendre Comte, met le doigt sur une
parent profonde entre les ides dOken sur lindividualit prise dans la vie
communautaire et celles de Comte, qui nadmet pas que la vie dun organisme soit une
somme de vies particulires pas plus quil nadmet que la socit soit une
association dindividus. Les monades organiques (les cellules) sont des abstractions.
En quoi pourrait donc consister rellement, demande Comte, soit lorganisation, soit

9
10
11
12
Ibid., p. 61.
Ibid., p. 61.
Ibid., p. 63.
Ibid., p. 65.
113

Debru
la vie dune simple monade ?13 Pour Canguilhem, cette question est pertinente au
regard des conditions contemporaines de la culture des tissus. Aujourdhui, me semblet-il, elle lest moins, du fait des connaissances beaucoup plus prcises sur les
communications cellulaires, les facteurs de croissance etc. La thorie cellulaire a
survcu, elle nest pas quun rve de philosophe en proie la Schwrmerei, ou plus
exactement elle correspond bien une ralit dabord anticipe par limagination.
Canguilhem trouve dailleurs les traces de ce romantisme holiste qui voit dans les
cellules des produits de dissociation de lindividu total, dans les ides des hritiers
positivistes de Comte preuve de la persistance dun rel problme philosophique, celui
de lindividualit, de ses degrs et chelles, de sa position simultane comme partie et
comme totalit selon le point de vue et le niveau auquel on se place. Il y a l
indubitablement une communaut de problmes entre biologie et sociologie.
Claude Bernard, aprs Virchow, est un acteur important de lintroduction de la
thorie cellulaire. Cette thorie constitue en effet un claircissement considrable pour
la physiologie. En outre, elle est une pice ncessaire de la thorie du milieu intrieur.
Quen est-il alors du fameux problme de lindividualit, du tout et de la partie, de leur
gouvernement rciproque? Claude Bernard affirme le principe de lautonomie cellulaire
ce qui aurait pour consquence que la cellule, milieu gal, vit en libert (en culture
cellulaire) comme en socit (dans lorganisme).14 Lorganisme est une cit o chaque
lment cellulaire vit selon les mmes principes que ses congnres tout en exerant une
activit propre. Haeckel crit en 1899 que les cellules sont des citoyens autonomes. La
philosophe politique qui finit par dominer la thorie cellulaire est le rpublicanisme.
Qui pourrait dire si lon est rpublicain parce quon est partisan de la thorie cellulaire
ou si on est partisan de la thorie cellulaire parce quon est rpublicain? commente
Canguilhem.15
A son texte de 1945, Canguilhem a ajout quelques paragraphes pour la publication
de 1952. Ces paragraphes sont une charge contre la biologie sovitique officielle, celle
de Lyssenko, oppose la thorie cellulaire autant qu la gntique, oppose
lexclusion de lhrdit de lacquis par la thse de la continuit du plasma germinatif
soutenue par August Weismann. La charge canguilhemienne, avec des accents
poppriens,16 antidogmatiques, doit se comprendre non seulement par des raisons
historico - scientifiques, mais galement politiques, dans un contexte de plus en plus
tendu de confrontation et de guerre froide. LUnion sovitique nest pas le seul pays de
la vraie science. La vraie science, cest le consentement lerreur.
Pour conclure: on voit dans ce travail assez sminal de Canguilhem sur la thorie
cellulaire se former une pratique de lpistmologie sur laquelle me semble-t-il certains
interprtes de Canguilhem (y compris moi-mme dailleurs) nont pas assez insist.
Lhistoire des thories et concepts, qui formera la vulgate de lpistmologie historique,
laquelle nous apprend que les thories naissent de thories prexistantes, ne peut pas se

13
14
15
16
Ibid., p. 65.
Ibid., p. 69.
Ibid., p. 70.
Ibid., p. 78: une thorie ne vaut rien quand on ne peut pas dmontrer quelle est fausse .
114

GEORGES CANGUILHEM
dissocier de lexercice de limagination thorique. Ce qui se passe avant la fameuse

coupure pistmologique est aussi important que ce qui se passe aprs. Mais le plus
essentiel du texte sur la thorie cellulaire nest pas dans son contenu pistmologique. Il
se trouve dans le constat de la connexion et de la circulation des ides. La science, dans
ces textes, napparat pas comme un processus ferm, une hyperspcialisation de plus en
plus minuscule. Les ides scientifiques, la philosophie des ides scientifiques (plutt
que la philosophie des sciences, expression que je nai jamais bien comprise) sont en
communication avec bien dautres domaines que leurs domaines de pertinence directe.
Les thories de la structure des organismes, avec les discussions sur le caractre holiste
ou associationniste de ces structures, nous ouvrent bien dautres dimensions. On y
dcouvre que lpistmologie historique est recouverte de philosophe politique. On y
dcouvre aussi un Canguilhem, rationaliste antitotalitaire, manifester certaines
inclinations romantiques, et commencer dgager certains aspects du vitalisme, aspects
ambigus, totalitaires et antitotalitaires, quil va bientt approfondir. Lpistmologie
nest donc pas seulement la rigueur et la froideur des concepts, cest aussi le
romantisme de limagination, le sens du possible, et au bout du compte laction
politique antidogmatique dans une histoire qui nest pas seulement celle des ides.
115

ON THE GENEALOGY OF CONCEPTS AND EXPERIMENTAL PRACTICES

Rethinking Georges Canguilhems Historical Epistemology1
Pierre-Olivier Mthot
In a sense, epistemology has always been historical.

G. Canguilhem (1988 [1977])
1. Introduction
Scholars have yet to explore more systematically the thought of philosopher and
historian of science Georges Canguilhem (1904-1995). Indeed, until recently, an
impressive number of his publications were virtually inaccessible because published in
hard-to-access journals, while others were not yet translated from French, or were
simply left unknown. With years, it became a commonality among commentators to
state that Canguilhem wrote very little and on a limited range of topics in the history of
the life sciences. While this situation possibly slowed down the development of a
research industry, if not cult, around Canguilhems person and work, it had the
adverse effect of concealing aspects of the complexity, variety, and scope of his
historical and philosophical interests. crits politiques et philosophiques 1926-1939
(2011), the first volume (out of 6) of Canguilhems Oeuvres compltes, will enable
researchers to gain a better and broader understanding of Canguilhem before
Canguilhem, to use Jean-Francois Braunsteins expression (2000), and to relate his
early writings to later ones in the history and philosophy of biology and medicine for
which he is more well-known. This volume is also timely because Canguilhems views
on health, disease, and the history of science are gaining currency, not only in France
but also in Germany and in the Anglo-American world as well, in particular thanks to
the recent English translations of La connaissance de la vie and crits sur la mdecine,
and the German edition of La formation du concept de rflexe aux XVIIe et XVIIIe
sicles.
In this article I will not focus directly on medicine and biology but rather on
Canguilhems methodology in the history and philosophy of science often labelled
historical epistemology, an expression which has recently stirred much discussion and
controversy in philosophical circles, particularly outside France. Broadly speaking, this
expression captures a certain style or method in philosophy where philosophical
problems are inseparable from their historical milieu and the distinction between
context of discovery and context of justification is at best illusory. As a method,
historical epistemology also refers to the a posteriori reconstitution of forms of
knowledge in a given domain like biology, medicine, psychology, etc., and to the
developments of regional epistemologies.
A version of this paper is forthcoming in Studies in History and Philosophy of Science (Part A).
117
Mthot
Nowadays, the expression historical epistemology designates a French network

(Gutting 1990).2 Although firstly associated to the triad of Bachelard, Canguilhem, and
Foucault (and sometimes the Polish philosopher and immunologist Ludwig Fleck), the
label historical epistemology was marshalled in the mid-1990s to draw attention to new
approaches to the history and philosophy of science.3 In particular, it acted as a guiding
concept for a variety of combined philosophic and historic research projects conducted
at the Max Planck Institute for the History of Science in Berlin (MPI). The diversity of
these projects is too broad to be reviewed here but it must be emphasized that the
concept of historical epistemology became closely associated with the work of Lorraine
Daston, Jurgen Renn, and Hans-Jrg Rheinberger. The latter also used it to characterize
not only individual thinkers but also what he perceives to be a broader movement of
historicization of epistemology and epistemologization of history, running from the
late nineteenth century to the present (Rheinberger 2010b).
A number of publications (Davidson 2002; Chimisso 2003; Hyder 2003; Rheinberger
2005a; Kusch 2011) and international conferences focussing on historical epistemology
followed rapidly in the 1990s and 2000s.4 Those events brought together scholars in the
history and philosophy of science to debate the use (and abuse) of the concept, its future
applications, and indeed its very nature (Sturm and Feest 2011; Gingras 2010). They
covered classical questions like: Is history necessary to epistemology? What is the
relation of history to philosophy of science? But also new ones such as: What are the
specificities of historical epistemology when compared to other approaches in science
studies? And, overall, what good is historical epistemology? A growing number of
philosophers of science whose names are associated with Anglo-American traditions
(especially, Kusch 2011; Hacking 2002; Davidson 2002), but who do not relinquish
exploring perspectives in continental philosophy as well, are critically investigating the
significance of the concept of historical epistemology.5
The programmatic use of the concept of historical epistemology at the MPI, followed
by the organization of several topical conferences and the publication of a special
journal issue, led Yves Gingras (2010) to write a critical essay in which he argued that
this splattering masks a strategy of rebranding, primarily designed to reinforce
institutional identity, to fund new research programs, and to gain in visibility on the
global, international scene (Gingras, 2010, p. 447). The present contribution does not
aim to confirm or contradict this claim although it should be noted that recent work in
historical epistemology has, in my view, provided important conceptual tools, for
instance with the concept of epistemic things and experimental systems
3
4
Although historical epistemology is often characterized as French epistemology, philosophers

working outside France also developed it. See, Schmidgen (2008).
See, Daston (1998). On Fleck and Canguilhem, see Mller-Wille (2011); Braunstein (2008).
In 1993, Hacking and Daston organized a weeklong conference on historical epistemology at the
University of Toronto. For other recent conferences on historical epistemology, see Gingras (2010).
Martin Kusch (2010), for instance, recently pointed out that himself and Hacking are among the very
few interested to work out the ideas, or use the tool boxes, of both Foucault and Wittgenstein
(Kusch, 2010, p. 159; emphasis in original).
118
(Rheinberger 1997), to investigate unexplored epistemological and social dimensions of

contemporary sciences.
In order to clarify the meaning(s) of the expression, Sturm and Feest (2011)
established a typology that maps out the contours of the work of Daston, Rheinberger,
and Renn, respectively. According to them, historical epistemology is often understood
as 1) a study of higher epistemic concept (e.g. objectivity, observation, probability); 2)
the study of historical trajectories of particular scientific objects of research (e.g.
electron, phlogiston, DNA) or epistemic things; and 3) as the long term study of
scientific developments (Sturm and Feest, 2011, p. 285). Sturm and Feest acknowledge
that this typology remains open; they did not seek to be exhaustive but principally to
survey ways of practicing historical epistemology by prominent scholars at the MPI.
While the recently published special issue in Erkentnnis edited by Sturm and Feest
demonstrates that the category of historical epistemology is of interest to more strictly
analytical philosophers, it is also symptomatic of recent attempts to grasp the nature of
historical epistemology, whatever it is, outside reference to the original French context.
Interestingly, the puzzlement of many English-speaking philosophers looking for the
essence of historical epistemology (i.e. what is...?) was hardly ever a source of
concern within French philosophical circles as the concept was rapidly associated to a
particular style of philosophy of science (Gingras, 2010, p. 442). But maybe curiously,
the thinkers we today consider to be part of the French historical epistemology tradition
Bachelard, Canguilhem, and Foucault never proposed this label to characterize their
own practice; this was done a posteriori, partly by some of their students.
A consequence of the international ascent of historical epistemology (understood as a
rebranding strategy or not), where the emphasis is placed on practical aspects of
science, on experimentation, and on experimental systems, as opposed to its conceptual
dimension, is the temptation to draw too sharp a contrast between newer and older
methodologies and projects in historical epistemology.6 Canguilhem for instance is
usually associated with an earlier school of historical epistemology where scientific
concepts unfold according to their own internal logic of change and rectification. In this
article, I re-evaluate the relation between concepts and experimentation in
Canguilhems philosophy of science and I investigate neglected aspects of his work that
are still relevant for todays history and philosophy of science and for the history of
historical epistemology more generally.
Commentators have often left out the experimental and practical dimension of
Canguilhems epistemology, choosing to focus largely (if not exclusively) on the
genealogical aspect of concepts instead. And yet, Canguilhems conceptual history was
from the very beginning practice-oriented, as his first articles on the relations between
science and technology will illustrate. This topic was later explored further in
Canguilhems other major works, including his two dissertations and in an original
essay on Experimentation in animal biology published in 1952 and collected in La
For a recent view that posits that Rheinbergers experimental systems are not composed of material
entities alone but are also inclusive of concepts, see Weber (2012).
119
Mthot
connaissance de la vie.7 My reading of Canguilhem will illustrate how the main

criticisms voiced against his approach, and the perceptible distance separating his work
from recent studies in historical epistemology, dissolve once we move beyond
Canguilhems image one he somehow himself contributed to promote of a
philosopher primarily concerned with disembodied, abstract genealogies of concepts,
disconnected from both real science and wider cultural contexts. Even Canguilhems
perception of his own method can be misguiding, and if we rely on his own account
alone, paraphrasing historian of science Yehuda Elkanas (1987), we will not be able to
break out of the confines of the usual images of Canguilhem as a philosophical historian
of disembodied concepts (Elkana, 1987, p. 136).8 Drawing on his early writings on the
relations between science and technology in the 1930s, on the Essai sur quelques
problmes concernant le normal et le pathologique (1943), and on La formation du
concept de rflexe aux XVIIe et XVIIIe sicles (1955), I argue that the formation,
transformation, and rectification of concepts, in Canguilhems sense, is intrinsically
bound with the experimental, material, technical, and cultural contexts in which
concepts are operationalized. These analyses will show that the nature of concepts in
Canguilhems work has been misunderstood, and especially that the role of practice,
techniques and experimentation in concept formation was largely overlooked. The
interpretation of Canguilhems philosophical legacy I propose differs from most
accounts and suggests firstly, that there is no sharp discontinuity between old and
new historical epistemology perspectives, and indicates, secondly, that Canguilhem
has anticipated a number of claims such as the connections between concept formation
and scientific practices, including measurements and observations, that are now in the
process of becoming the norm among historians and philosophers of science.9 Before
turning to these issues, I introduce the received view about Canguilhem that is still
prevalent nowadays.
2. The received view: a philosophical historian of disembodied concepts

In a tribute article to Canguilhem, Michel Foucault (1991) credited his mentor for
promoting a philosophy of the concept, as opposed to a philosophy of the subject,
placing at the same time epistemologists (Bachelard, Cavaills) and phenomenologists
(Sartre, Merleau-Ponty) in two distinct and rarely overlapping philosophical traditions,
at least in France (Foucault 1991 [1985]). Following Foucault, commentators noted that
Canguilhems approach consists in applying the principle of the genealogy of
Canguilhems medical dissertation was republished as Le normal et le pathologique in 1966. This

edition added three essays in a new section titled Nouvelles rflexions sur le normal et le
pathologique. In the present paper I use both titles (Essai sur quelques problmes concernant le
normal et le pathologique and Le normal et le pathologique) interchangeably.
Canguilhem, for instance, famously suggested that the history of science cannot be satisfied with
collecting biographies of scientists but [] has also to be a history of the formation, deformation,
and rectification of scientific concepts (Canguilhem, 2002 [1963b], p. 235).
Hacking (1983) was among the first philosophers to make this point. But see, Tiles (2011) for an
account of how Bachelard came to a similar conclusion much earlier.
120
concepts (Macherey, 2009, pp. 113-114; Badiou, 2009, p. 7).10 In the early 1970s,
Dominique Lecourt (1975), now director of the Centre Georges Canguilhem in Paris,
also pointed out that Canguilhem concerns himself more with the descent of concepts
than with the concatenation of theories (Lecourt, 1975, p. 171). And more recently
still, according to philosopher Gilles Renard (1996) Canguilhems foremost objective
was to unravel the complex and complicated historical trajectory of concepts. Yet this
feature of Canguilhems methodology the close attention to the trajectory of scientific
concepts was singled out much earlier by Pierre Macherey (1998 [1964]) in an article
published twenty years before Foucaults tribute. For all these thinkers, an historicphilosophical approach la Canguilhem consists primarily in tracking scientific
concepts over space and time, and across disciplinary boundaries, in order to locate
significant shifts regarding meaning, reference, and domains of application.
Following the principle of the genealogy of concept, commentators often argued that
concepts and theories are, for Canguilhem, to a large extent decoupled, in the sense that
concepts can migrate from one theoretical context to the other (Gutting 2001; Chimisso
2003). Gutting claims, for instance, that Canguilhems most important methodological
contribution is precisely the distinction between concepts and theories (Gutting,
2001, p. 229). Drawing on Canguilhems history of the reflex concept (1955), he argues
that Canguilhem demonstrated how a particular concept could operate within very
different theoretical contexts (e.g. mechanism, vitalism), taking on different meanings
and significations. For Gutting (2006), Canguilhems approach posits a neat separation
between concepts and theories that allows for the development of a distinct kind of
history of science, one that is not based on a succession of theoretical frameworks, but
on the genealogy of concepts instead (Gutting, 2006, p. 8). The possibility for concepts
to migrate from one theoretical context to another indicates that they are to some extent,
independent of the theories within which they are invoked.11
Needless to say that Canguilhem was not alone in promoting a conceptual approach
to science in France during the past century; one could think of Cavaills, Lautman, or
Bachelard, to name a few. But in contrast with other philosophers of the concept,
Canguilhem was a singular character in the conceptomania which was largely
indebted to him during the 1950s-1960s (Schwartz, 1993, p. 307). Indeed, his
approach to concepts was, at the time, considered by some to be almost revolutionary.
According to Louis Althusser, Canguilhems scrupulous respect for the reality of real
science permitted the emergence of a new kind of epistemologist who, similar to
ethnologists, go into the field, to overturn the problems of classical epistemology
(Althusser, 1998 [1964], p. 163). Strikingly, Althusser compared Canguilhem to
ethnologists studying scientists in their own milieu, an idea frequently associated with
Bruno Latours laboratory studies that would come only fifteen years later. As Paul
10
11
Note that Macherey (2009) does situate Canguilhems analysis of scientific concepts in a wider
cultural and societal context. He cites, for instance, Canguilhems discussion of the social status of
science in a lecture given at the Sorbonne in 1961-1962 (Macherey, 2009, p. 115).
It would be unwise to think that Canguilhem had no interest in scientific theories and dismissed them
altogether in favor of concepts, however. See his study of cell theory in La connaissance de la vie
(Canguilhem, 2008 [1952]).
121
Mthot
Rabinow (1994) has noted, this image is somewhat misleading, however, because
Latours project aims not only at replacing a positivist view of the science with a
historicized one, but also at dismantling the very idea of science a position as far
from Canguilhems as one could imagine (Rabinow, 1994, p. 13). Nevertheless,
Althussers comments underline Canguilhems authentic concern with concrete aspects
of scientific activity. As the picture of Canguilhem as ethnologist vanished, his conceptoriented epistemology rose to become the dominant interpretation of his approach
today. While the focus on concepts is undeniable, Canguilhems approach also calls for
questioning the relation between concepts and their operationalization within
experimental contexts, however.
In her recent book, Cristina Chimisso (2010) argues that historical epistemology
stems out of a larger project elaborated in Paris between the 1920s and 1940s that
sought to combine the history of science with philosophy. She suggested looking at
Bachelard and Canguilhems work not as the beginning of a new tradition in the
philosophy of science but rather as the culmination of older philosophical debates in
France. Some years before, Chimisso (2003) had assessed Canguilhems historical
epistemology by comparing the concept of a norm, as used in La formation du concept
de rflexe (1955) and in lEssai sur quelques problmes concernant le normal et le
pathologique (1943). Though she accepts much of Canguilhems ideas about health,
disease, the subjectivity of illness and a holistic view of norms, Chimisso (2003) is
critical vis--vis his use of history for philosophical purposes. More precisely, she
charges that it was logic rather than historical considerations proper that drives the
main thesis in La formation du concept de rflexe. In this book, Canguilhem aimed to
evaluate past and present concepts of reflex and see whether they are logically
consistent. Chimisso argues that in discussing almost exclusively the work of
philosophers and scientists, and by leaving out most of the cultural and social
background that led to the first, vitalistic concept of reflex in the work of British
physiologist Thomas Willis, Canguilhem kept history simple (Chimisso, 2003, pp.
322-324). This line of criticism echoes Bruno Latours earlier comments that in France,
science was often depicted as something somehow apart from society and history
(Bowker and Latour, 1987, p. 717). In her paper, Chimisso places Canguilhem closer to
older historians of science of the first half of the twentieth century than to recent work
in historical epistemology: for her, Canguilhem followed a methodology similar to that
of Metzger and Koyr in the sense that his work established close connections between
historical concepts and broader worldviews or systems of thought and scientific ideas.
Chimisso (2003) suggests further that historical epistemology, in Canguilhems sense,
would benefit from opening-up to sociological investigations. That is to say ideas
should be re-embodied (Chimisso, 2003, p. 324). This last point crystallizes what I
take to be the now widespread image of Canguilhem understood as a philosophical
historian of disembodied genealogies of concepts.
While Chimisso blames Canguilhem for his failure to produce a cultural history of
the reflex concept, others have called into question the adequacy of concepts as units
of analysis and narration for the history of science altogether. Historian of biology
Jonathan Hodge (2000) argues that Canguilhems preoccupation with concepts has
122
seriously confined his agenda as an historian of science by neglecting the role played by
interests and institutions and the influences of predecessors and scientists intentions on
scientific progress (Hodge, 2000, p. 72). Hodge is convinced that Canguilhems method
is wholly inadequate to its object of study, and he goes on to claim that, for example,
any historical analysis of the Principles of Geology cannot succeed if it restricts itself
to an analysis of concepts (Ibid.). According to Hodge:
A historiography for science that concentrates our attention on concepts cannot do
justice to the challenges we face as historians. Nor, in any case, is the concept a
very suitable unit of analysis and narration. For a concept is a product, an
achievement not a process or a goal. The concept, then, as a unit of analysis and
narration is not adapted to the historians quest of insight into the process of
enquiry that Lyell or Darwin, or Bernard or Monod, have undertaken, or for
insights into the aims that inspired and directed those activities (Hodge, 2000, pp.
72-73; emphasis added).
For Hodge, conceptual history should be rejected because it neglects the role of the
social. Moreover, concepts are the wrong units of analysis to understand science
historically and dynamically because firstly, concepts are static entities, whereas science
is presumably characterized by dynamic and change; and secondly, because the aims of
scientists are irreducible to the formation of concepts. Thus, Hodge (2000) argues, when
Charles Lyell wrote the Principles of Geology or when Charles Darwin opened up his
Notebooks they both had a number of goals in mind such as reforming the science of
geology, or introducing new methods of classification, none of which are reducible to
an intention to construct one or more concepts (Hodge, 2000, p. 72). On this reading,
Canguilhem provided a wholly internal history of science, one that leaves out the role of
social institutions, intentions, and beliefs. Echoing Latours concerns as well, Hodge
argues that Canguilhems focus on concepts has limited the scope of his historical
analyses by neglecting the role of both the social and political consequences on the
development of the sciences. While this line of argumentation raises important issues it
is, however, based on an oversimplified view of what Canguilhemian concepts are. As
the remainder of this paper will illustrate, a preoccupation with concepts not only
leads to a broad intellectual view of science (Hodge, 2000, p. 71), but also to a more
situated view of concepts and scientific practices.
To put it succinctly, Chimisso and Hodges critical assessments fail to identify an
important aspect of Canguilhems historically informed epistemology: the connection
between the formation of concepts and experimental practices. In contrast, the late
philosopher of science Marjorie Grene (2000) sketched a portrait of Canguilhems
philosophy of science that suggests this point although she did not develop it in full.
Comparing Canguilhems philosophic style to logical positivism, Grene (2000)
concludes that it was Canguilhems emphasis on the historicity of concepts that led him
to stay much closer to the practice of science than his positivist counterparts of the
time (Grene, 2000, p. 60; emphasis added). Grene was familiar with the work of
Canguilhem and had a deep understanding and knowledge of his contributions as early
as the 1960s (Mthot 2009). Along the same lines, historian and philosopher of science
Hans-Jrg Rheinberger, who entertained a long lasting interest in the spirit of French
philosophy of science, reassessed Georges Canguilhems historical epistemology as one
123
Mthot
focussing on the filiation of concepts in relation to the role of experimentation (2005a;

reprinted in 2010a). For him Canguilhems work should be understood as a history of
the displacements of problems which must be reconstructed in their historic context
(Rheinberger, 2010b, p. 66). He concurs with Etienne Balibar (1993) to state that with
Canguilhem, the typical units of knowledge have shifted from scientific theories to
concepts. Rheinberger also discussed, approvingly, the latters conceptualization of the
object of the history of science as one distinct from the object of science. Like Chimisso
(2003), Rheinberger (2010a) considers Canguilhems project of producing a history of
concepts to be an integral part of the reasoned history of ideas represented in the midtwentieth century by people like Alexandre Koyr or Arthur Lovejoy (Rheinberger,
2010a, p. 44). In contrast with Chimisso and Hodge, however, Rheinberger (2005a;
2010a) stressed that for Canguilhem the formation of scientific concepts like Claude
Bernards milieu intrieur is intimately connected to the development of a new space
of experimentation (Rheinberger, 2005, p. 193).12 The experimental aspect of
Canguilhems work contrasts with the work of his predecessors such as Koyr who was
(according to Rheinberger himself) to a certain degree even antiexperimental
(Rheinberger, 2010b, p. 52). Rheinberger aptly seizes Canguilhems views on
experimentation and concept formation but he concentrates exclusively on the latters
post-1960s writings, like Idologie et rationalit dans lhistoire des sciences de la vie
(1977), and tudes dhistoire et de philosophie des sciences sur les vivants et la vie
(1968).
In the present article, I pursue and deepen the research avenue opened-up by Grene
and Rheinberger and I do so by looking at Canguilhems writings pre and post 1960.
Exploring Canguilhems thought from his early writings on technology in the 1930s to
his doctoral dissertation on the reflex concept in 1955, and through his penetrating
analyses of concepts of the normal and the pathological in 1943, this research brings to
the fore a recurrent interrogation Canguilhem attempted to come to terms with, namely
the relation of concepts and experimentation in the life sciences, of theory and practice,
and of thought and action. Before, however, I place Canguilhems views on the relation
between the history and philosophy of science within the larger history of historical
epistemology in France, and I clarify his views regarding this expression.
3. Canguilhem in the history of French historical epistemology

Though the history of twentieth philosophy of science in France can obviously not be
reduced to the works of Bachelard and Canguilhem, these two thinkers exerted a deep
influence on the direction and methods in the history and philosophy of science during
the twentieth century in ways that are not yet fully appreciated.13 In addition to
12
13
Rheinberger quotes Canguilhem on the relation between concepts and techniques in the context of an
essay on Bernards physiology: Let us be clear about this point, it is the concept of milieu intrieur
which is given as a theoretical foundation of the technique of physiological experimentation (in
Rheinberger, 2005a, p. 194; emphases in original).
For a history of philosophy of science in France, see Chimisso (2010); Brenner and Gayon (2009);
and Brenner (2003).
124
promoting a new style of philosophy of science, both had several pupils many of whom
became, in turn, very influential. Taken together, both occupied the key position of
Head of the Institut dhistoire des sciences in Paris for more than three decades.
Between 1964 and 1968 Canguilhem was also president of the aggregation jury, an
important examination in France, and after the Second World War he was appointed
General Inspector in philosophy between 1948 and 1955, until he replaced Bachelard at
the Sorbonne.
The history of science in France has almost always been a philosophical project, and
the idea of an historical epistemology, indeed, goes back (at least) to Auguste Comtes
(1869) saying that we do not know completely a science if we do not know its history
(Comte, 1869, p. 65, quoted in Gingras, 2010, p. 447). Yet while Bachelard probably
never himself used the term historical epistemology, Dominique Lecourt, a former
student of Canguilhem, published a book-length monograph on Bachelards
pistmologie historique (historical epistemology) in 1969, a fortunate choice of words
that may reflect the Marxist orientation of its author (Gingras 2010).14 This book was
shortly followed by Pour une critique de lpistmologie, in 1972 (Lecourt 1972), and
the two books were then translated into English and republished together as Marxism
and philosophy: Bachelard, Canguilhem, Foucault in 1975 (Lecourt 1975). Since then,
the expression historical epistemology has come to designate Bachelards way of
practicing epistemology as well as the works of many of his students and followers,
including Georges Canguilhem and Michel Foucault, but also Francois Dagognet,
Dominique Lecourt, Jean Gayon, Claude Debru, Suzanne Bachelard, Claire SalomonBayet, Mirko Grmek, Francois Delaporte, and Christiane Sinding, to name a few. On
the American scene, Marx Wartofsky, a Marxist philosopher, employed the label
historical epistemology during the 1970s and 1980s, before Ian Hacking took it up (in a
modified form) a decade later.15
In an often-told anecdote Canguilhem is said to have replied to Lecourt who wanted
to use the title historical epistemology for his book on Bachelard, No. Not historical
epistemology but epistemological history (Gayon, 2003, p. 53). Lecourt did not follow
Canguilhems advice although he remained careful to speak of Canguilhems
epistemological history (Lecourt 1975). Some scholars have detected in Lecourts
choice a corroboration of their view that Bachelard and Canguilhems methods differ
significantly (Rheinberger 2005a; Gayon 2003; Gutting 2001). Amidst the recent
debates and controversy regarding the origin and the history of historical epistemology,
Dominique Lecourt admitted he borrowed the notion from Canguilhem (Lecourt 2008)
though Canguilhem did not use it in a programmatic fashion or as a guiding concept for
his philosophical and historical work. As we will see, he used the expression historical
epistemology three times in his publications: it appears twice in the collected essays
(still awaiting translation) tudes dhistoire et de philosophie des sciences concernant
14
15
Gingras suggests that Lecourt chose historical epistemology because it was the natural counterpart
to historical materialism.
Nowadays, Hacking prefers to use the label historical meta-epistemology, a branch of a wider
historical ontology (a concept borrowed from Foucault), to characterize his own project (see
Hacking 2002). On Wartofsky and historical epistemology see, Sturm (2011).
125
Mthot
les vivants et la vie (1968), and once in an article on Bachelard published in 1974 in an
Italian encyclopaedia on science and technology. It is absent from Canguilhems other
major works.16
The expression historical epistemology appears, firstly, in a paper on Gaston
Bachelards conception of the history of science published in 1963 (included in
Canguilhem, 2002).17 Bachelards historical method, Canguilhem tells us, is distinct
from what historians of science usually do. Instead of tracking changes in a scientific
treatise over time, for instance, Bachelard investigates the role of errors and
rectification in science, and in the process of overcoming epistemological obstacles.
In a word, Bachelard examines what is on the fringe of historical history, what
remains in the margins and is not covered by historical epistemology (Canguilhem,
2002 [1963a], p.178). Canguilhem does not theorize the notion further, however. But
we can understand that if, as he says, Bachelards epistemology is a history of science
always en acte (Ibid.), it is precisely because his epistemology is historical.
Chronologically, the expression is used a decade later in another article on Bachelard
and published in Italian. Using similar arguments he developed elsewhere, Canguilhem
argues for the need for philosophers of science to combine the resources of the history
of science and epistemology: On the one hand, he says, epistemology draws on the
history of science to identify science with the dialectic of conceptual rectification (i.e.
epistemology needs the history of science to grasp the idea that science is not a linear
process). On the other, the history of science borrows from epistemology the cognitive
values that are currently in use in the scientific city with the purpose of ordering the
chronological succession of theories following the progression of intellectual growth
(i.e. the history of science needs an epistemological perspective to clarify the nature of
scientific activity and progress according to a particular set of epistemic values).18 What
distinguishes this essay from The objects of the history of science (Canguilhem,
1968) or The role of epistemology in contemporary history of science (Canguilhem,
1988 [1977]) is that the author explicitly brings together historical epistemology and
epistemological history. Indeed, his conclusion from the preceding remarks is that
historical epistemology and epistemological history reciprocally validate their project
and their approach (Canguilhem, 1974, p. 66).
Lastly, Canguilhem uses the expression in a paper on the epistemological status of
medicine published in 1988 (now also collected in tudes). Reflecting on how the
epistemological status of medicine could be clarified, Canguilhem draws on the work of
his colleague, German historian of medicine Karl Rothschuh, who uses the key
concepts of a historical epistemology, that of Thomas Kuhn, to examine this issue
(Canguilhem, 2002 [1988], p. 426). He further reports that, according to Rothschuh, the
16
17
18
In the unpublished manuscript of a talk on Jean Cavaills delivered in 1986 and kept in Paris at the
CAPHS (Centre dArchives de Philosophie, dHistoire et dditions des Sciences), where are located
Canguilhems papers, Canguilhem wrote that Cavaills work remains a classical example of the
historical epistemology of mathematics. I am thankfull to Camille Limoges for this and for the
reference of Canguilhem on Bachelard in the encyclopedia Mondadori.
On Bachelard as a historian of science, see Gayon (2003).
The concept of scientific city (cit savante) is from Bachelard. On this concept, see Gingras (2003).
126
concepts of normal science and paradigm are incapable of capturing the progress
and conceptual changes in clinical medicine because of the variability of its object.
Canguilhem does not dispute these conclusions, which is hardly surprising as for him
Kuhn misconstrues the nature of scientific rationality (Canguilhem, 1988 [1977], p. 13).
What should be noted here is that historical epistemology comes in various kinds, and is
not restricted to French philosophical and historical circles. For Canguilhem, historical
epistemology is not the name for a prestigious tradition in the history and philosophy of
science; it is a descriptive notion. And though he may not have used it as a guiding
concept Canguilhem did use it nevertheless but not to characterize his own methodology
in the history and philosophy of science. However, he fully aligned himself with the
significance of a genuine historical epistemology when he claimed, famously: in some
sense, epistemology has always been historical (Canguilhem, 1988 [1977], p. 10).
Recently, Gingras radicalised the difference between the two approaches arguing that
if we use the language correctly, historical epistemology is kind of epistemology and
not a kind of history and epistemological history is a kind of history and not a kind of
epistemology (Gingras, 2010, p. 444). He argues that historical epistemology places
more emphasis on the philosophical than on the historical aspects of this project (and so
better characterizes thinkers like Bachelard who used history for philosophical
purposes); whereas epistemological history is more adequate for historians like
Canguilhem. Both Gingras and Lecourt agree that epistemological history is overall a
better label for Canguilhems project that is (they gather), of a more historical nature.
While Gingras definition is accurate (i.e. that historical epistemology does encompass
philosophically-oriented histories of science) it is, however, misleading to try to pigeonhole thinkers as being either on the philosophical or on the historical side of the fence:
in my view, the real interest of the concept of historical epistemology is precisely that it
indistinctly links the history and philosophy of science together in a single, coherent
project. Canguilhem was a philosopher, not just an historian of science, though the
history and philosophy of science are always closely connected in his epistemological
writings. As Yves Schwartz noted, Canguilhem has often applied to himself the
expression historian of science because philosopher (Schwartz, 2011, p. 98, n.1). For
Canguilhem, there is reciprocity between historical epistemology and epistemological
history.
Cast in the light of epistemological analysis, the history of science ceases to be
designed to confront or test philosophical claims and allows for a philosophical
understanding of the surprising and growing diversity of scientific practices as seen
through the emergence of methods, objects, theories, styles, and concepts. Taken in this
sense, historical epistemology is presently interesting because it could help us to move
past the debate between the history and philosophy of science in the English-speaking
world where the relation between the history and philosophy of science, since at least
the 1960s, consistently operates on a confrontational mode where philosophical claims
are tested against historical data (Schickore 2011). In the next sections I will illustrate
how Canguilhems historical epistemology weaves together the formation of concepts
and the development of scientific practices and experimentation.
127
Mthot
4. Techniques, science, and the normativity of life

The essay on Machine and Organism (2008 [1952]) underlines Canguilhems
concerns with the question of technology, and with the wider problem of the
understanding relation between life and machines. Nine years before this classic essay,
however, he had already reflected philosophically on the relation between science and
techniques. Prior to Le normal et le pathologique (1943), the young Canguilhem had
elaborated a distinctive and original view on the relation between technique and science
that was significantly different from the more melodramatic discourses on the
technological age pervading the first half of the twentieth century. Indeed,
Canguilhems career as a professional researcher in philosophy begins with a paper on
Descartes and techniques he gave in Paris in 1937, followed by a second one on
technique and creation the year after in Toulouse. There, Canguilhem addressed the
problem of technology, or technique, in explicit metaphysical terms.19 His main
conclusion is that technological developments do not (necessarily) lead to an
instrumental society but rather illustrate a deeper and normative impulse proper to living
systems.20 Whereas for some thinkers like Martin Heidegger machine-based technology
is the hallmark of modernity (1977 [1954]), technological developments broadly
construed express, for Canguilhem, what in Le normal et le pathologique he has called
the normativity of life (1991 [1943]).
In both Descartes et la technique (1937) and Activit technique et cration
(1938), Canguilhem argues that scientific progress is not a direct result of the
application of theoretical knowledge (as positivists like Comte would say) but on the
contrary, that the growth of science necessarily emerges from a series of technical
failures and surmounted obstacles. Most of the time, technique (or technology) is prior
to scientific development, both historically and epistemically, and enables scientific
growth.21 For Canguilhem (1937), the development of technological instruments, as
well as their inevitable shortcomings, not only facilitates the discovery of new
phenomena or the formulation of theories (e.g. Pasteurs germ theory comes from his
work on silk worms; thermodynamics comes from the construction of the steam engine,
etc.), but opens up future research avenues for scientists to engage into more
theoretically-oriented investigations, precisely because of those unavoidable difficulties
provoked by technical obstacles (1937, pp. 83-4).
19
20
21
In a note at the end of Activit technique et cration (1938), Canguilhem writes that his oral
presentation included a section on how biology and sociology can be useful regarding the problem he
was considering. However, he decided to leave out this part in order to confer to the written article an
explicit metaphysical character (Canguilhem, 1938, p. 86).
For instance, human technique extends vital impulses (Canguilhem, 1991, p.130). Those vital
impulses are attributed by Canguilhem to all life forms: therapeutic need is a vital need, which, even
in lower living organisms (with respect to vertebrate structure) arouses reactions of hedonic value or
self-healing or self-restoring behaviours (Canguilhem, 1991, p. 127).
In an essay on Thrapeutique, exprimentation, responsabilit published twenty years later,
Canguilhem expressed the same idea again, writing that it is the spontaneous technique that creates
for knowledge the conditions of its emergence, and thus precedes it (Canguilhem, 2002 [1959b],
p. 387).
128
But what drives technological development over time? For Canguilhem, the answer
is found in the exigencies of life, manifesting and extending those deep vital impulses
further and in various directions (Canguilhem, 1994, p. 225; 1937, p. 84). Canguilhem
(1937) sees techniques (including medicine and agriculture) as reflecting the creative
and original power of man in his quest to master and to become possessor of the
natural world (Canguilhem, 1937, p. 77). As he says in the essay on experimentation in
animal biology: Man first experiences and experiments with biological activity in his
relation of technical adaptation to the milieu (Canguilhem, 2008 [1952], p. 9). Situated
somewhere between life and art, technologies have also for Canguilhem (1938) an
irreducible art-like dimension as they are at once both creative and liberating
(Canguilhem, 1938, p. 85).
These two texts on techniques and science (too) briefly examined here would deserve
more attention as they are key to understand the origin and the originality of Le normal
et le pathologique, written in the early 1940s. In this book, Canguilhem explores how
scientific apparatuses, technological innovations, and laboratory instruments intersect in
producing medical knowledge but also new norms of life within conceptual and
experimental contexts. As we will see next, before developing a more sustained
methodological reflection on scientific concepts in the 1950s, Canguilhem had already
suggested that their elaboration requires setting up a realm of intelligibility.
5. The experimental and the normal22
Medicine seemed to us and still seems to us like a technique or art at the crossroads of
several sciences, rather than, strictly speaking, like one science (Canguilhem, 1991, p.
34). This declaration opens Canguilhems medical dissertation on the concepts of the
normal and the pathological, and is one of the most well known phrases of the book
published during the Second World War. Le normal et le pathologique provides an
original reflection on the relation between values, techniques, norms, and concepts. The
category of experimental is at the core of this critique of medical reason and supports
Canguilhems argument that the determination of the concept of normal, especially in
physiology and clinical medicine, is linked to specific laboratory equipments and sets of
practices; in other words, that there is no absolute concept of the normal state in
medicine or physiology.
Medicine, for Canguilhem, is a technique that extends a spontaneous effort, peculiar
to life, the effort of living beings adapting to their milieu, and adapting their milieu to
them. Located at the intersection of several sciences, medicine exists as the art of life
not because physiology designates certain states as being normal and others as
abnormal, but rather because man, as an individual, has come to call certain states as
being in need of correction in relation to the dynamic polarity of life (Canguilhem,
22
I am grateful to one anonymous reviewer for suggesting adding a section on Le normal et le

pathologique to this essay, and to the editor of Humana.Mente Journal of Philosophical Studies, for
permission to draw on some previously published material (Mthot 2009), and to use it in a modified
form here.
129
Mthot
1991, 126). Medicine, as a human enterprise, reflects this fundamental fact that life is
not indifferent to environmental circumstances but is rather polarity or position of
value, even unconscious (Ibid.) Organisms are normative beings in the sense that,
contrary to inorganic matter, they are inevitably affected by their milieu and will
spontaneously react to external perturbations by making physiological adjustments with
more or less success. Following a change in the environment, healthy organisms will be
those that are able to adapt smoothly to the external modifications. For Canguilhem
(1991), health is thus defined as the possibility of tolerating infractions of the habitual
norm and instituting new norms in new situations (1991, p. 197). If health includes the
possibility of disease, then the latter should be regarded as a new dimension of life, an
innovative experience in the living being, not merely a fact of decrease or increase
(Canguilhem, 1991, p. 184). This last point is a critique directed at Claude Bernards
epistemology on which normal and pathological states are identical in nature and only
differ quantitatively (i.e. for Bernard, pathological phenomena amount to an increase or
a decrease of an otherwise normal function).23 For Canguilhem, an impoverishment in
the normative capacity to react adaptively is a sign of disease as it indicates a reduction
in the margin of tolerance at the level of the whole organism. What distinguishes the
normal from the pathological then is that an organism will be in a pathological situation
if the new norms he has established are inferior in terms of stability, fecundity and
variability of life than the previous ones (Canguilhem, 1991, p. 144). The sick
organism is the one that has lost its normative capacity, the capacity to establish other
norms in other situations (Canguilhem, 1991, p. 183). However, Canguilhems
originality is to argue that a new norm is never a priori normal or pathological; its
normality will come from its normativity (Canguilhem, 1991, p. 144), that is, from the
organisms capacity to organize the milieu according to its own needs. In other words,
norms of life cannot be said normal or pathological a priori, because judgments about
normal and pathological states must take into account the environment into which an
organism lives. Grounded in a Darwinian approach, Canguilhem argued that Taken
separately, the living being and his environment are not normal: it is their relationship
that makes them such (Canguilhem, 1991, p. 143).24
In Le normal et le pathologique Canguilhem argues that physiology cannot secure an
objective foundation for pathology because physiological norms are initially noticed
through a pathological situation in a clinical encounter (1991, p. 209). In that, he agreed
with the French surgeon Ren Leriche, according to whom: at every moment there lie
within us more physiological possibilities that physiology would tell us about. But it
takes disease to reveal it to us (Canguilhem, 1991, p. 100). Although this argument
about the vagaries of the concept of normal in medicine is not a weak one, Canguilhem
complements it with a stronger one that brings together scientific practices and concept
formation. Canguilhems main argument against the possibility to derive an objective
concept of the normal is located in his critique of the science of physiology, and how
laboratory medicine constructs new norms of living that cannot be called normal in an
absolute sense. Although not frequently mentioned by commentators, the relationship
23
24
Bernards classical example is diabetes.

For a critical assessment of Canguilhems thesis concerning the normal and the pathological, see
Keating (2001).
130
of the normal and the experimental, Canguilhem says, is at the heart of our concerns
(Canguilhem, 1991, p.145; emphasis added). At the outset, Le normal et le pathologique
critically analyses how the experimental method expounded by Bernard impacted on the
ways in which physiology and pathology were being realigned with the clinic at the end
of the nineteenth century. For Canguilhem, two important consequences that flew out of
these disciplinary rearrangements were, to begin with, that in discovering the laws of
normal and pathological phenomenon, physiology became epistemically first, and could
pretend to guide and illuminate the clinic. Additionally, in elucidating the statistical
constants representing the normal curve of vital functions of living beings, physiology
went down a reductionist slope and ended up ascribing health and disease not to the
whole organism anymore, but to its most inner constituents (organs, tissues, cells, etc.)
which departed from the statistical mean. Using scientific measurement, health and
disease could then be quantified and assessed objectively. In the second part of the
book, Canguilhem was to turn this view upside down and replace the whole individual,
and the clinic, at the centre of medical epistemology.
Canguilhems concerns then are with the relation between the norms constructed by
the physiologist with the help of laboratory instruments and standardization procedures,
and the living beings functional activity outside the laboratory (Canguilhem, 1991,
p.145). Looking at the distinction between the normal and the pathological from a broad
evolutionary point of view, he frames it in terms of understanding the problem of the
variability of organisms and the significance and scope of this variability
(Canguilhem, 1991, p.141). In the context of physiological experiments, what is the
significance and scope of variability among organisms (and environments) when this
variability is reduced to a minimum by standardization, normalization process, and
control procedures, so that any deviation beyond a particular threshold could be marked
as pathological, at least in principle? Canguilhem argues that the results from
experiments are always context-dependent, as other experimental conditions would have
produced other norms of life, normal or pathological. Thus, while the physiologist can
claim to have established a norm, he does not objectively define which conditions are
normal (Canguilhem, 1991, p.145). If one thinks that it is the relation between
organism and environment that makes them either normal or pathological, it follows
that if one defines the pathological as a deviation from the average, as a statistical
difference, the laboratorys conditions places the organisms in a pathological situation
from which scientists attempt to derive norms understood as the normal state. These
reflections were taken up once again in the essay on experimentation in animal biology
where Canguilhem insisted on the originality of biological method and on the formal
obligation to respect the specificity of its object which, by nature, is not inert but
reactive to external stimuli (Canguilhem, 2008 [1952], p.21).
Although Canguilhem denies the possibility for science to objectively identify what
the normal is, he does not embrace a wholly social normativism either. To the contrary,
his normativism is intended to be universal and biologically grounded: even for an
amoeba, living means preference and exclusion; [...] this point of view is that of vital
normativity (Canguilhem, 1991, p. 136). This normativity does not merely reflect
social preferences at a given space and time but is rather an intrinsic property of living
131
Mthot
beings. The existence, coextensive in space and time with humanity, of medicine as a
more or less scientific technique for healing diseases (Canguilhem, 2008 [1952], p.
132) appears to Canguilhem to be the result of a universal tendency in living organisms
to avoid diseases and prefer health instead.
It is life itself and not medical judgment which makes the biological normal a
concept of value and not a concept of statistical reality. For the physician, life is
not an object but rather a polarized activity, whose spontaneous effort of defense
and struggle against all that is of negative value is extended by medicine by
bringing to bear the relative but indispensable light of human science
(Canguilhem, 1991, 131).
However, in addition to those biological norms, Canguilhem is perfectly aware that

the definition of the normal is partly established by the social context. Technological
innovations having considerably enlarged the possibilities of human beings in terms of
activities, he argues, the fluctuating line between the normal and the pathological ought
to take into account certain activities which have become a need and an ideal for
mankind (Canguilhem, 1991, pp. 200-1). This is why, for Canguilhem, to discern the
normal from the pathological one must also look beyond the body (Ibid.), into the
realm of the social and the political. Citing the hospital reform in France and Austria
during the nineteenth century, Canguilhem notes that reforming medical theory has
rested on the reform of medicine as a practice (Canguilhem, 1991, p. 237). The
process of normalisation in medicine, indeed, reflects a more general demand for
rationalisation which has roots in politics and economy as well (Ibid.). Technology, in
sum, is driven by vital, organic values, but the process of normalisation in society at
large also flows from those technological developments.
6. The formation of the reflex concept

Canguilhems interpretation of the relation between the creative, but also normalizing,
power of techniques and its relation to science feeds back into the formation of concepts
by emphasizing the active role technological apparatuses play. In addition to Le normal
et le pathologique, the mutually engaging, two-way relations between scientific
concepts and laboratory apparatuses appear most clearly in Canguilhems book on the
reflex concept where he contrasts the reflex 1800 with the reflex 1850. The
experimental possibilities generated by the latter are here the main difference between
these concepts. As Foucault (1991) usefully puts it, The concept of reflex was not
formed as a biological concept when Willis applied the image of a reflected light ray to
an automatic movement; but it did happen the day Prochaska could write it down in the
analysis of sensorimotor functions and their centralization in relation to the brain
(Foucault, 1991, p. 21). Following his mentor and friend Gaston Bachelard, Canguilhem
often repeated, a same word is not a same concept (Canguilhem, 2002 [1963a],
p. 177). Scientific discourses sometimes contain terms that are linguistically identical
but are nevertheless epistemically different. To distinguish words from genuine
(scientific) concepts, Canguilhem tells us, one needs to reconstitute the synthesis of a
concept, that is, to reconstruct the specific conceptual context in which the concept is
132
inserted and the goal [intentions directrices] of the experiments or

observations (ibid.). For example, in Thorie et technique de lexprimentation chez
Claude Bernard, Canguilhem (2002 [1959a]) argues that it is by the means of
experimentation in the course of his research on diabetes and with the help of related
concepts (such as internal secretions) and theories (cell theory), that the concept of
milieu intrieur was formed, not through the application of a general scientific
method. Returning to the initial moment of the creation of a concept like milieu
intrieur allows us to witness the construction of a properly biological concept, one
whose elaboration is at once an effect and a cause of experimentation (Canguilhem,
2008 [1952], p.7). In other words, concepts are the result of newly opened spaces of
experimentation and foster at the same time the developments of new forms of
practices. This give-and-take relationship between the construction of a concept and the
experimental set-up in which it is embedded opens up new ways of experimenting and
vice-versa. Forming concepts is no easy task, though, and in biology the core issue is
perhaps not so much using experimental concepts but rather experimentally
constituting authentic biological concepts (ibid., p. 6).
The shift between a word and a concept in the case of the reflex was contiguous to
the moment where the tools necessary to measure it and make it happen, so to say,
became available. As Canguilhem (1977) points out (with an intended play on words),
while the concept of reflex 1800 was a good concept, it was, however, not good for
anything yet (Canguilhem, 1977, p. 161).25 The reflex concept was discussed and,
indeed, was included in physiology textbooks but no one did anything practical with it
yet. Circa 1850, however, things began to change, signposting the crucial passage from
a word to a biological concept proper. In 1850, the reflex was not only written in
books but also in the laboratory, in terms of exploratory and demonstrative
apparatuses, designed for it, and which would not have existed without it (Ibid.). In
that sense the concept of reflex facilitated the creation of new technical support that, in
turn, allowed for a deeper and fuller analysis of the larger phenomenon of reflex action
itself. The reality of the 1850 reflex concept is finally demonstrated as it brings new
objects into existence that it can, in turn, explain (Ibid.). Concepts and experimental
arrangements closely intersect and one could say that for Canguilhem they are even
reciprocally coproduced.
Drawing again on Bachelard, Canguilhem (1977) points out that the reflex concept in
1850 is no longer phenomeno-logical but had become phenomeno-technical
(Canguilhem, 1977, p. 161).26 As early as in La formation de lesprit scientifique
Bachelard (1938) had argued that a concept has become scientific according to the
proportion to which it has become technical, to which it is accompanied by a technique
of realization (Bachelard 1969 [1938], p. 61, quoted in Rheinberger 2005b, pp. 320-1).
Concepts become scientific through their intricate relation with the technical realm. But
if the 1850 reflex concept has become more scientific on the grounds that it has become
more phenomeno-technique, its validity and usage as a scientific concept is not limited
25
26
The book on the reflex was published in 1955 but I quote from the 1977 edition.
On the concept of phenomeno-technique, see Castelo-Lawless (1995), Chimisso (2009) and
Rheinberger (2005b).
133
Mthot
to the frontiers of physiological laboratories understood as spaces of experimental

practices. In effect, one finds traces of the reflex 1850 concept in hospitals, the clinic,
but also in the contemporary culture broadly construed (Canguilhem, 1977, pp. 162-3).
To take a simple but clear example of this, nowadays everyone wants to know whether
they have good reflexes (Canguilhem, 1977, p. 163). Attempting to apply Bachelard's
epistemology to biology, Canguilhem relies on the distinction between phenomenologic and phenomeno-technique, but he assesses the epistemic value of a concept (or a
scientific object) somewhat differently by examining its insertion in a broader scientific
context or culture, and by the changes it brings about at the practical level. For instance,
when medical students use a pocket light to detect pathognomonic signs most of them
ignore who Argyll Robertson is; nevertheless, they construct their diagnosis on the basis
of the presence (or the absence) of a reflex action of the eyes accommodation to light.27
Similarly, the kneecap reflex tested with a small hammer, or the Babinski sign gave the
reflex concept at the end of the nineteenth century the status of a biological fact to the
extent that it becomes unclear whether its existence realizes a concept or if it is its
concept that reflects its existence (Canguilhem, 1977, p. 162). The coming into being
of the concept of reflex as a biological fact is assessed by Canguilhem, finally, not
only by the techniques it generated or the instruments it helped develop but also by its
cultural extension, its public notoriety, that is, by its being rooted in contemporary
culture (Canguilhem, 1977, p. 163).
Canguilhem rarely expressed himself on the nature of concepts, and we can regret he
was not more precise on this issue.28 In an article on the concept of reflex in the late
nineteenth century that follows-up on his dissertation, however, he explains that when
talking about concepts he has in mind a denomination and a definition, that is to say,
a noun endowed with meaning that is able to fulfill a function of discriminating between
a number of observations or experiments (Canguilhem 2002 [1964], p. 295; emphasis
added). Concepts thus perform an operational function as they allow diverse kinds of
judgment (e.g. discrimination) to be performed. For Canguilhem a concept always
contains an operational or judgmental norm (Canguilhem, 2005 [1966], p. 198). In
sum, he argues that what guarantees the theoretical efficacy or the cognitive value
of a concept is its operational function [fonction doprateur] (Canguilhem, 2002
[1966], p. 360). 1850 is an important turning point in the history of physiology because
it is when the reflex concept becomes operational.
As concepts come into being within certain scientific practices (as Canguilhem
claims), if one wants to understand their formation, deformation and rectification, one
also has to consider the experimental and material contexts in which they were
successively elaborated. This process of conceptual correction is indeed not a purely
intellectual or theoretical one. To put it differently, the process of rectification of the
concept is not a matter of logic, but of experimentation (Canguilhem, 2002 [1964],
p. 296). Concepts are thus both formed and rectified in the context of scientific
27
28
Douglas Argyll Robertson (1837-1909) was a Scottish ophthalmologist who described the lack of
motor reflex in the eyes during light exposure.
It is noteworthy for the main thesis of the present article that in a book review Canguilhem writes that
concepts encompass ordered facts and methods of investigations (Canguilhem 1960, p. 158).
134
practices. We could say that the philosophy of Canguilhem is part of a tradition where
concepts are tools29 or operational devices used alongside instruments and other kinds
of scientific apparatuses to intervene in the world, and as such, enlarge both the sphere
of human action and thought. The point concerning the operational character of
concepts was well taken by the French historian of biology and Nobel Prize laureate
Franois Jacob (1973) for whom, indeed, the importance of a concept is given
operationally in terms of its role in directing observation and experience (Jacob, 1973,
p. 11; emphasis added; see Rheinberger 2005a). The epistemic value of a scientific
concept thus refers to the possibility of further theoretical, technical, or experimental
developments it creates. The tightly interwoven relation between concepts and
experimental practices in Canguilhems epistemology suggests a strong link between
concept formation and experimentation but it also gestures towards the wider social and
cultural contexts in which those processes operate. As Rheinberger (2005a) pointed out,
the products of science, including concepts are for Canguilhem cultural-historical
objects par excellence (Rheinberger, 2005a, p. 190). Describing Canguilhems
approach of concepts as one entirely focussing on abstract genealogies of concepts is
thus at best incomplete and, indeed, squarely reductive. Concepts emerge and change
not in the void but in the context of concrete manipulations by human beings who
conduct various operations that produce new meaning or correct past interpretations.
For, to put it differently, tracing the genealogy of a concept presupposes the genealogy
of practices that accompany it and operationalize it in various ways. Scientific concepts
do not emerge out of nothing and do not have a life of their own; their formation,
rectification, and re-organisation across different fields of knowledge are always closely
linked to particular epistemic cultures, while they also relate, in one way or another, to
the cultural context in the broader social and political sense of the word.
7. On science, culture, and society

While the characterization of Canguilhem as being almost obsessed with concepts, truth
and rationality may sound slightly exaggerated it contains a kernel of truth (Rose 1998;
Tiles 2011). For, even if Canguilhem does not isolate concepts from theoretical and
material contexts, it is clear he does grant a certain authority to science, which others
like Bruno Latour would certainly like to undermine. Science, according to
Canguilhem, is one of the many cultural systems that flourish within societies and is,
indeed, a branch of culture (Canguilhem, 2002 [1963b], p. 235). But there is
something specific, or different about it. For Canguilhem, science creates and acts upon
particular norms of rationality that are not found in, say, aesthetics or politics. This is
why Canguilhem would have rejected Hodges earlier comments about his lack of
interest in the role of institutions and political influences on the sciences. For, if one
takes as object of inquiry nothing other than sources, inventions, influences, priorities,
simultaneities, and successions, Canguilhem (1988) tells us, one fails to distinguish
between science and other aspects of culture (Canguilhem, 1988, p. 3). To say that
science and society are wholly coproduced is thus to miss what is specific about the
29
For an analysis of concepts as tools, see Feest (2010).
135
Mthot
former: that it operates according to a set of norms that are not found outside it, and
whose goal is the search for truth (Canguilhem 2005, 204). But as Mary McAllister
Jones has put it, the relation of science to culture for Canguilhem is a complex one
because those domains are different yet somehow interdependent; scientific concepts
can change cultural attitudes while cultural contexts can also facilitate or hinder
scientific progress (Jones, 2000, pp. 124-125). The work of scientists cannot be
detached or concealed from the rest of society and this point exemplifies Canguilhems
position that living beings and that includes humans actively construct their milieu
which in turns reverberates on them.
Canguilhem was no naive realist and he recognized that the truth of today in science
is often tomorrows error. For him, the genesis of scientific facts, to speak with Ludwig
Fleck, does not spring out of unmediated observation of the world but, as we have seen,
is always the result of a configuration of specific technological, conceptual and
experimental practices situated in a given cultural context. Nevertheless, he neither set
himself the task to challenge what science tells us about the world or to mount an attack
on it based on social, economic, or political influences, nor to explain science by
appealing to these very same external factors, which does not mean he was uncritical
towards the authority of science and scientists. However, he did not expose scientific
knowledge as being socially constructed even if his work clearly illustrated that a
number of key concepts such as regulation, disease, healthy, normality, or reflex are
deeply value-laden and have significant impacts on the construction of social,
individual, even biological norms. This respect for science is rooted in the belief that
while science is a product and part of human culture, bound in space and time, scientific
discourses and practices cannot be reduced to aspects of a given cultural context, or
determined by it. Canguilhems respect for the reality of real science (Althusser,
1998, p. 163) may be a little disconcerting from todays perspective. However, it is also
refreshing in some ways as it hands back to scientific discourse the right to being in the
truth, and so it can be read as an attempt to dispose of pervading forms of relativism.
Not only do the evolutions of scientific concepts partake of an experimental culture
but they also constitute a platform onto which broader relations of power are articulated.
The link between power and knowledge, embedded in scientific concepts, was
developed further by Foucault but can already be seen in Canguilhems work. Although
concepts are linguistic entities to which we attach different meanings, they are not just
words, that is, they are not neutral descriptions of scientific objects they also convey
value judgments about people, society, and so on. As Staffan Mller-Wille (2011)
recently emphasized, the meaning of a concept [for Canguilhem] does not exhaust
itself in its discursive relationship to other words and texts only; to the contrary,
concepts articulate dynamic power relationships of authority and resistance by
advancing certain evaluations in order to contest or overcome others (Mller-Wille,
2011, p. 479). For example, the 1850 concept of reflex, spreading across many
disconnected fields, served the purpose of those promoting a mechanistic philosophy of
labour where the actions of the workers are decomposed into smaller reflex actions in
order to fit larger economical needs in the production of goods. Following the analysis
of G. Friedman, Canguilhem (1977) remarked that the reduction of the workers actions
136
to a sum of reflex was only possible once man in general was assimilated to a
machine, as for example in Taylorism, and his actions subordinated to it. As
Canguilhem concluded, however, in so far as the worker refuses practically to be
mechanised, he brings to the fore the theoretical error consisting in decomposing his
own actions into mechanical reflexes (Canguilhem, 1977, p. 166).
The wider social implications of Canguilhems approach become clearer when one
looks at the kind of concepts he focused on health, disease, reflex, regulation
concepts whose impact reaches beyond science deep into the social and political realms.
As a vital rationalist (Rabinow 1994), Canguilhem believed that philosophy must
learn from science because only science can tell us what exists (Canguilhem, 1967, p.
51). This was not a sign of defeat from the part of philosophy but rather the sign of a
deep conviction that philosophy is a normative activity (Ibid). For him, the
fundamental reason why philosophers should gain a minimal understanding of what
scientists take to be real about the world is that it opens up new possibilities as to how
this reality could eventually be changed, or challenged. Indeed, Canguilhems
philosophy of life, embodied in his concept of normativity, is above all a philosophy
that leads one to make decisions and to undertake action. It is thus no coincidence that
the concepts studied by Canguilhem are core concepts of the life sciences; and
Canguilhems project could be summarized as one constantly attempting to come to
terms with advances in biomedicine by demonstrating how the formation of concepts is
one of the multiple forms of living.30 Or, as Foucault (1991) poignantly noted in a
tribute article to his mentor, forming concepts is one way of living, not killing life
(Foucault, 1991, p. 21). The relatedness of concept and experimentation, theory and
practice, thought and action, and indeed, knowledge and life, provide a significant entry
point into the philosophy of Georges Canguilhem. Understanding how these poles come
together and are differently articulated was a task he relentlessly pursued through his
life, not the least when the threat of war was imminent. Then, thought and action
crystallized out of sheer necessity and courage. To quote the last sentence of not wellknown Trait de logique et de morale, a book Canguilhem co-authored with his
philosopher colleague Camille Planet in 1939, on the eve of the Second World War,
Here, like in Shakespeares Hamlet, we must choose (Canguilhem and Planet, 2011
[1939], p. 922).31
8. Conclusion
Building on the interpretation of Grene (2000) and Rheinberger (2005a), this article was
concerned with a neglected but crucial aspect of Canguilhems historically informed
epistemology, namely that for him, the formation of concepts and experimentation in
the life sciences belong together. Reading Georges Canguilhem today from this
perspective makes his work sound much more relevant to current problems addressed
30
31
On life and concepts, see Le concept et la vie (Canguilhem 2002 [1966]). This text is partly
available in English translation in Canguilhem (1994).
On the philosophical significance of this book in Canguilhems writings, see Roth (2011). On the
ethical and political aspects of Canguilhems early writings, see Braunstein (2007).
137
Mthot
by philosophers, historians, and sociologists of science. Despite the fundamental role of

concepts in both science and its history, Canguilhem neither explained what a concept
really is, nor what an understanding of science based on conceptual analyses could be
like; in fact, though he did not say it, he provided countless examples of how such
narratives can be constructed using concepts as units of analysis.
Commentators have tended to treat the notion of a concept as a primitive category,
as if concept was a straightforward, transparent, and unproblematic philosophic
notion in Canguilhems writings. As a consequence, some central aspects of the
formation, rectification and translation of concepts across disciplinary fields of research
have been overlooked, caricatured, or simply ignored. Over time, Canguilhemian
concepts have somehow become understood as free-floating entities divorced from
experimental, theoretical, and wider cultural contexts. They needed to be re-embodied,
as it were. This idealized history of science was further accused of neglecting the role of
institutions, influences, and predecessors in the growth of the sciences, in particular,
biology. The present essay illustrated that, to the contrary, the dynamics of the
formation of concepts is always grounded in a set of experimental practices, itself
embedded at different levels of the social life. It was shown that the kind of historical
epistemology Canguilhem provided us with is one that is deeply grounded in concrete
and real science, where experimentation occupies a privileged place. Once correctly
understood, it becomes clear that a Canguilhemian history of conceptual changes does
not ignore the development of scientific practices in which those concepts are
operationalized, situated, and negotiated.
The analysis provided here naturally gives rise to many follow-up questions,
including ones about how Canguilhems approach might be fruitfully and positively
connected to current debates in the history and philosophy of experimentation such as
Hans-Jrg Rheinberger and Ian Hackings, and how it can lead to a renewed interest in
the operationalization of concepts.32 The fate of Canguilhems historical epistemology
should also be investigated outside mainstream French and German history and
philosophy of science, for instance in the sociology of Pierre Bourdieu, another
theoretician of scientific practice, as well as in the anthropology of reason as developed
by Paul Rabinow. No doubt, when new Canguilhem material emerges in the next few
years, future research will be needed to provide a more complete understanding of his
contributions, and the limits of his conceptual approach, but also to assess his
philosophical legacy for twenty-first century sciences of life.
Acknowledgments
This paper was written at the University of Exeter and at the Universit PanthonSorbonne (Paris 1) and recently completed during a postdoctoral residency at the
Institute for the History of Medicine and Health at Geneva University. I would like to
thank warmly Henning Schmidgen for inviting me to present this paper at the
32
On operationalism as a guide to a practice-oriented history and philosophy of science, see Chang

(2009).
138
conference Epistemology and History: From Bachelard and Canguilhem to Todays

History of Science held in Berlin in December 2010. Another version of the paper was
presented at a workshop on Georges Canguilhem and the Contemporary Life Sciences
in Skagen (Denmark) in May 2012 on the occasion of the 60th anniversary of La
connaissance de la vie. In addition to the pertinent comments of several participants and
organizers of these two events, the paper benefited from stimulating discussions with
Uljana Feest, Jean Gayon, Sabina Leonelli, Michel Morange, Staffan Mller-Wille, and
Hans-Jrg Rheinberger. I am particularly indebted to Camille Limoges for drawing my
attention to a number of unknown articles of Canguilhem, and for his very helpful
remarks on the paper at a later stage of completion. The critical and constructive
suggestions of the two anonymous reviewers strongly improved the paper. Lastly,
financial support from the Social Sciences and Humanities Research Council of Canada
(no. 752-2007-1257), and from the Fonds de Recherche Socit et Culture (Qubec) is
thankfully acknowledged.
References
Althusser, L. (1998). Presentation, in: W. Montag (ed.), In a materialist way. Selected
essays by Pierre Macherey, London, New York: Verso [1964].
Badiou, A. (2009). Pocket Pantheon. Figures of postwar philosophy, Translated by
David Macey, London, New York: Verso.
Bachelard, G. (1938). La formation de lesprit scientifique, Paris: Vrin.
Balibar, E. (1993). Science et vrit dans la philosophie de Georges Canguilhem, in:
Georges Canguilhem. Philosophe, historien des sciences, Paris: Albin
Michel.
Bowker, G. and Latour, B. (1987). A booming discipline short of discipline: (social)
studies of science in France, Social Studies of Science, 17(4): 715-748.
Braunstein, J.-F. (2000). Canguilhem avant Canguilhem, Revue dhistoire des
sciences 53: 9-26.
Braunstein, J.-F. (ed.) (2007). Canguilhem. Histoire des sciences et politiques du vivant,
Paris: Presses Universitaires de France.
Braunstein, J.-F. (2008). Deux philosophes de la mdecine. Canguilhem et Fleck, in:
H.-J. Han et al. (eds.), Philosophie et mdecine. En hommage Georges
Canguilhem, Paris: Vrin.
Brenner, A. (2003). Les origines franaises de la philosophie des sciences, Paris:
Presses Universitaires de France.
Brenner, A. & Gayon, J. (2009). Introduction, in: A. Brenner & J. Gayon (eds.),
French studies in the philosophy of science. Contemporary research in
France, Boston etc.: Springer.
Canguilhem, G. (1937). Descartes et la technique, in: Travaux du IXe Congrs
International de philosophie (Congrs Descartes), tome 2, Paris: Hermann.
Canguilhem, G. (1938). Activit technique et cration, in: Communication et
discussions de la socit toulousaine de philosophie.
Canguilhem, G. (1960). [Compte rendu de Bentley Glass, Owsei Temkin, William L.
Straus Jr. (Eds.), Forerunners of Darwin: 1745-1859, Baltimore, The Johns
139
Mthot
Hopkins University Press, 1959], Archives Internationales dhistoire des

sciences 13 (50-51): 157-159.
Canguilhem, G. (1961). Physiologie animale: II. La physiologie en Allemagne,
III. Jeunes coles de la seconde priode, IV. Techniques et problmes de la
physiologie au XIXe sicle, in: Ren Taton (ed.), Histoire gnrale des
sciences, vol. III: La science contemporaine. Part. 1: Le XIXe sicle, Paris:
Presses universitaires de France, p. 475484.
Canguilhem, G. (1967). Du concept scientifique la rflexion philosophique, Cahiers
de philosophie, no. 1 (janvier): 3951.
Canguilhem, G. (1977). La formation du concept de rflexe aux XVIIe et XVIIIe sicles
[1955], Paris: Presses Universitaires de France.
Canguilhem, G. (1988). Ideology and rationality in the history of the life sciences
[1977], Cambridge, MA: MIT Press.
Canguilhem, G. (1991). The normal and the pathological [1943/1966], Translated by C.
R. Fawcett, New York: Zone Books.
Canguilhem, G. (1994). A Vital Rationalist. Selected Writings from Georges
Canguilhem, with a Critical Bibliography by C. Limoges, ed. by
F. Delaporte, transl. by A. Goldhammer, New York: Zone Books.
Canguilhem, G. (2002). tudes dhistoire et de philosophie des sciences concernant les
vivants et la vie [1968], 7th ed., Paris: Vrin.
Canguilhem, G. (2002). Thorie et technique de lexprimentation chez Claude
Bernard [1959a], in: id., tudes dhistoire et de philosophie des sciences
concernant les vivants et la vie, 7th ed., Paris: Vrin, pp. 143-155.
Canguilhem, G. (2002). Thrapeutique, exprimentation, responsabilit [1959b], in:
id., tudes dhistoire et de philosophie des sciences concernant les vivants
et la vie, 7th ed., Paris: Vrin, pp. 383-391.
Canguilhem, G. (2002). Lhistoire des sciences dans luvre pistmologique de
Gaston Bachelard [1963a], in: id., tudes dhistoire et de philosophie des
sciences concernant les vivants et la vie, 7th ed., Paris: Vrin, pp. 173-186.
Canguilhem, G. (2002). La constitution de la physiologie comme science [1963b], in:
id., tudes dhistoire et de philosophie des sciences concernant les vivants
et la vie, 7th ed., Paris: Vrin, pp. 226-273.
Canguilhem, G. (2002). Le concept de rflexe au XIXe sicle [1964], in: id., tudes
dhistoire et de philosophie des sciences concernant les vivants et la vie, 7th
ed., Paris: Vrin, pp. 295-304.
Canguilhem, G. (2002). Le concept et la vie [1966], in: id., tudes dhistoire et de
philosophie des sciences concernant les vivants et la vie, 7th ed., Paris: Vrin,
pp. 335-364.
Canguilhem, G. (2002). Le statut pistmologique de la mdecine [1988], in: id.,
tudes dhistoire et de philosophie des sciences concernant les vivants et la
vie, 7th ed., Paris: Vrin, pp. 413-428.
Canguilhem, G. (2005). The object of the history of science [1966], in: G. Gutting
(ed.) Continental philosophy of science, London: Blackwell.
Canguilhem, G. (2008). Knowledge of life [1952], Translated by S. Geroulanos & D.
Ginsburg, New York: Fordham University Press.
140
Canguilhem, G. (2011). crits philosophiques et politiques. 1926-1939, Paris: Vrin

(= Oeuvres compltes, tome 1).
Castelo-Lawless, T. (1995). Phenomenotechnique in historical perspectives: Its
origins and implications for philosophy of science, Philosophy of Science
62 (1): 4459.
Chang, H. (2009). Operationalism, in: E. N. Zalta (ed.), Stanford Encyclopedia of
Philosophy.
Chimisso, C. (2003). The tribunal of philosophy and its norms: history and philosophy
in Georges Canguilhems historical epistemology, Studies in History and
Philosophy of the Biological and Biomedical Sciences, 34: 297-327.
Chimisso, C. (2009). From phenomenology to phenomenotechnique: the role of early
twentieth century physics in Gaston Bachelards philosophy, Studies in
History and Philosophy of Science 39 (3): 384-392.
Chimisso, C. (2010). Writing the history of the mind. Philosophy and Science in France,
1900 to 1960s, Aldershot: Ashgate.
Comte, A. (1869). Cours de philosophie positive, 3rd ed., Paris: Baillire et Fils.
Daston, L. (1989). Historical epistemology, in: J. Chandler et al. (eds.) Questions of
evidence. Proof, practice, and persuasion across disciplines, Chicago:
University of Chicago Press.
Davidson, A. (2002). The emergence of sexuality. Historical epistemology and the
formation of concepts, Harvard: Harvard University Press.
Elkana, Y. (1987). Alexander Koyr: between the history of ideas and sociology of
knowledge, History and Technology, 4: 111-144.
Feest, U. (2010). Concepts as tools in the experimental generation of knowledge in
cognitive neuropsychology, Spontaneous Generations, 4 (1): 173-190.
Foucault, M. (1991) Introduction, in: G. Canguilhem, The normal and the
pathological, New York: Zone Books.
Gayon, J. (2003). Bachelard et lhistoire des sciences, in: Wunenburger, J.J. (ed.),
Bachelard et lpistmologie franaise, Paris: Presses Universitaires de
France.
Gayon, J. (2009). Philosophy of biology: an historico-critical characterization, in:
A. Brenner & J. Gayon (eds.), French studies in the philosophy of science.
Contemporary research in France, Boston: Springer.
Gingras, Y. (2003), Mathmatisation et exclusion: socio-analyse de la formation des
cits savants, in: Wunenburger, J.J. (ed.), Bachelard et lpistmologie
franaise, Paris: Presses Universitaires de France.
Gingras, Y. (2010). Naming without necessity. On the genealogy and uses of the label
historical epistemology, Revue de Synthse, 131(3): 439-454.
Grene, M. (2000). The philosophy of science of Georges Canguilhem. A transatlantic
view, Revue dhistoire des sciences 53: 47-64.
Gutting, G. (1990). Continental philosophy and the history of science, in: R. Olby et
al. (eds.) Companion to the history of modern science, London: Routledge.
Gutting, G. (2001). French philosophy in the twentieth century, Cambridge: Cambridge
University Press.
Gutting, G. (2006) (ed.) The Cambridge companion to Foucault, 2nd Edition,
Cambridge: Cambridge University Press.
141
Mthot
Hacking, I. (1983). Representing and intervening. Introductory topics in the philosophy

of the natural sciences, Cambridge: Cambridge University Press.
Hacking, I. (2002). Historical Ontolgy, Harvard: Harvard University Press.
Heidegger, M. (1977). The question concerning technology [1953], in: M. Heidegger,
The question concerning technology and other essays, Harper Colophon
Books.
Hodge, J. (2000). Canguilhem and the history of biology, Revue dhistoire des
sciences, 53(1): 65-81.
Hyder, D. (2003). Foucault, Cavaills, and Husserl on the historical epistemology of
the sciences, Perspectives on Science 11(1): 107-129.
Jacob, F. (1970). La logique du vivant. Une histoire de lhrdit, Paris: Gallimard.
Keating, P. (2001). Georges Canguilhems on the normal and the pathological. A
restatement and a commentary, in: A. Cambrosio & A.-M. Moulin (eds.),
Singular selves: historical issues and contemporary debates in immunology,
Dordrecht: Elsevier.
Kusch, M. (2010). Hackings historical epistemology. A critique of styles of
reasoning, Studies in history and philosophy of science 41: 158-173.
Kusch, M. (2011). Reflexivity, relativism, microhistory. Three desiderata for historical
epistemologies, Erkenntnis 75(3): 483-494.
Lecourt, D. (1969). Lpistmologie historique de Gaston Bachelard, Paris: Vrin.
Lecourt, D. (1972). Pour une critique de l'pistmologie (Bachelard, Canguilhem,
Foucault), Paris: Maspero.
Lecourt, D. (1975). Marxism and epistemology. Bachelard, Canguilhem and Foucault,
London, New York: New Left Books.
Lecourt, D. (2008). Georges Canguilhem, Paris: Presses Universitaires de France.
McAllister Jones, M. (2000). Georges Canguilhem on science and culture. Learning
biologys lessons, French Cultural Studies 11(31): 117-130.
Macherey, P. (1998). Georges Canguilhems philosophy of science. Epistemology and
History of Science [1964], in: W. Montag (ed.), In a materialist way.
Selected essays by Pierre Macherey, London, New York: Verso.
Macherey, P. (2009). Georges Canguilhem. Un style de pense, in: P. Macherey (ed.).
De Canguilhem Foucault, la force des normes, Paris: La Fabrique.
Mthot, P.-O. (2009). French epistemology overseas. Analysing the influence of
Georges Canguilhem in Qubec, Humana.mente. Journal of Philosophical
Studies, 9: 39-59.
Mller-Wille, S. (2011). History of science and medicine, in: M. Jackson (ed.),
Companion to the History of Medicine. Oxford: Oxford University Press.
Rabinow, P. (1994). Introduction. A vital rationalist, in: F. Delaporte (ed.). A vital
rationalist. Selected writings from Georges Canguilhem, Critical
bibliography by Camille Limoges, Transl. by A. Goldhammer, New York:
Zone Books.
Renard, G. (1996). Lpistmologie chez Georges Canguilhem. Nathan.
Rheinberger, H.-J. (1997). Towards a history of epistemic things. Synthesizing proteins
in the test tube, Stanford: Stanford University Press.
142
Rheinberger, H.-J. (2005a). Reassessing the historical epistemology of Georges

Canguilhem, in: G. Gutting (ed.) Continental philosophy of science,
London: Blackwell.
Rheinberger, H.-J. (2005b). Gaston Bachelard and the notion of phenomenotechnique,
Perspectives on Science, 13 (3): 313-328.
Rheinberger, H.-J. (2010a). An epistemology of the concrete. Twentieth-Century
histories of life, Durham: Duke University Press.
Rheinberger, H.-J. (2010b). On historicizing epistemology. An essay. Stanford: Stanford
University Press.
Rose, N. (1998). Life, reason and history. Reading Georges Canguilhem today,
Economy and Society, 27(2-3): 154-170.
Roth, X. (2011). [Prsentation du Trait de logique et de morale], in: Georges
Canguilhem, crits philosophiques et politiques. 1926-1939, Oeuvres
compltes, tome 1, Paris: Vrin.
Schickore, J. (2011). More thoughts on HPS. Another twenty years later, Perspectives
on Science 19(4): 453-481.
Schmidgen, H. (2008). Canguilhem et les discours allemands, in: H.-J. Han et al.
(eds.), Philosophie et mdecine. En hommage Georges Canguilhem, Paris:
Vrin.
Schwartz, Y. (1993). Une remonte en trois temps. Georges Canguilhem, la vie, le
travail, in: Georges Canguilhem. Philosophe, historien des sciences, Paris:
Albin Michel.
Schwartz, Y. (2011). Jeunesse dun philosophe, in: Georges Canguilhem, crits
philosophiques et politiques. 1926-1939, Oeuvres compltes, tome 1, Paris:
Vrin.
Sturm, T. (2011). Historical epistemology or history of epistemology? The case of the
relation between perception and judgment, Erkenntnis 75(3): 303-324.
Sturm, T. & Feest, U. (2011). What (Good) is Historical Epistemology? Editors
introduction, Erkenntnis 75(3): 285-302.
Tiles, M. (2011). Is historical epistemology part of the Modernist Settlement?,
Erkenntnis 75(3): 525-543.
Weber, M. (2012). Experiment in biology, in: E. N. Zalta (Ed.), Stanford
Encyclopedia of Philosophy.
143
LE CONCEPT DOBJECTIVIT CHEZ CANGUILHEM
Franoise Balibar
Lorsque les organisateurs du colloque mont demand de leur proposer un titre pour
mon intervention, jtais en train de lire le livre de Lorraine Daston et Peter Galison
intitul Objectivity.1 Lide men tait venue en lisant Donna Haraway, plus
prcisment le texte intitul Situated Knowledges: The Science Question in Feminism
and the Privilege of Partial Perspective , qui constitue le chapitre 9 de son livre
Simians, Cyborgs and Women: The Reinvention of Nature.2 Jy avais dcouvert, avec la
navet qui caractrise la scientifique que je suis, une dmolition en bonne et due forme
de la notion dobjectivit. Notion dont je mtais persuade au fil des ans quelle est
fondatrice ou plutt le noyau, le coeur mme de la science dont javais fait mon
mtier, la physique. Oserais-je dire que lobjectivit telle que les physiciens lentendent
(jy reviendrai) mtait toujours apparue sinon comme une raison de vivre, du moins
comme une raison subjective de faire ce mtier-l, physicienne, qui plus est
thoricienne: il est si reposant, quand par ailleurs on est en proie au doute, la rage et
lexcs, de se laisser guider par la ncessit l o tout nest que luxe, calme et
volupt et certitude (comme Descartes, jai t nourri(e) aux lettres ds mon
enfance , mais je me plaisais surtout aux mathmatiques, cause de la certitude et de
lvidence de leurs raisons3) oubliant ne serait-ce quun instant le reste, la vie.
Voil quavec Donna Haraway jtais confronte au rcit rcit la premire
personne, acide et joyeux, humoristique et drle, cent lieues de la plainte fminine et
du ressentiment acadmique dont nous rebattent les oreilles certaines universitaires qui
pensent ne pas avoir t reconnues par les hommes leur juste valeur rcit donc, de la
manire dont une scientifique pratiquante , si je puis dire, sest forg une opinion qui
lui soit propre sur la question de lobjectivit . Que lobjectivit soit une question
philosophique, quon puisse sur ce sujet avoir des conceptions philosophiques
diffrentes, je le savais pour avoir lu Bachelard et Canguilhem et de faon plus gnrale
certains des travaux de lpistmologie historique franaise. Jy avais vu laffrontement
entre deux conceptions philosophiques de lpistmologie, lune positiviste, lautre
quavec lloignement que procure lappartenance un autre continent, les Amricains
appellent constructiviste , quon nomme ici, dans ce colloque, pistmologie
historique , appellation invente par Canguilhem, reprise par Dominique Lecourt.
Cette seconde pistmologie me convenait parce quelle paraissait compatible, voire
1
2
L. Daston et P. Galison, Objectivity, New York : Zone Books, 2007.

D. Haraway, Situated Knowledges : The Science Question in Feminism and the Privilege of Partial
Perspective , Feminist Studies, vol. 14 n3, autumn 1988, p. 575-599 ; repris en tant que chapitre 9
dans Siminans, Cyborgs and Women : The Reinvention of Nature, New York : Routledge, 1991 ; les
rfrences de page sont ici celles de la version en ligne de ce chapitre: <http://science.consumercide.
com/haraway_sit-knowl.html>.
R. Descartes, Discours de la mthode, premier discours.
145
Balibar
mme conforme, ce que je connaissais: la physique mathmatique. Quand

lobjectivit, quelle puisse tre une question politique politique parce que pose au
nom du fminisme mavait toujours gne (depuis les annes 60-70); ne serait-ce
quen raison de lpisode dsastreux de la science proltarienne des annes 50,
quand la mcanique quantique tait une science bourgeoise et la vrit de la biologie
une question de classe; et je ntais pas loin de mettre lide dfendue par Haraway
dune objectivit fministe au compte des fameux excs du fminisme mme
si lexpression me mettait hors de moi lorsquelle tait employe par quelquun dautre.
Justement parce que lexpression me dplaisait et me dplat toujours, et parce quil
vaut mieux tre excessive que conformiste, je mtais mise, cet t 2010, lire (enfin!)
Donna Haraway, pour essayer de comprendre.
De Haraway, jtais passe Daston et Galison. Tout cela est bien beau, avais-je
dabord pens, mais ne peut sappliquer la physique: les rflexions de Daston et
Galison portent essentiellement sur les sciences descriptives (du moins dans les
premiers chapitres du livre) et la fabrication datlas na pas grand chose voir avec la
physique mathmatique. Cest dans ce contexte que jai d choisir un titre pour
lintervention daujourdhui. Je me suis dit que Canguilhem, parce quil stait intress
la mdecine plutt qu la physique (contrairement Bachelard) ce qui situait son
objet dtude mi-chemin, en quelque sorte, entre les sciences descriptives et la
physique mathmatique , devait pouvoir maider comprendre les ides dveloppes
par Daston, Galison et Haraway sur la question de lobjectivit.
Jtais soutenue dans cette intuition par le souvenir dun avertissement donn par
Canguilhem:
En raison des spcialits scientifiques physique mathmatique et chimie des
synthses calcules sur le champ desquelles elle a t initialement labore, la
mthode historique de rcurrence pistmologique ne saurait tre tenue pour un
passe-partout.4
Lisant cette phrase, javais t heureuse de trouver chez Canguilhem la confirmation

dun tonnement qui ne ma jamais quitt, ltonnement devant le prestige insens et
draisonnable dont a joui, et jouit encore, la physique, surtout la physique mathmatique
auprs des pistmologues, singulirement ceux de la tradition franaise. La physique
est une science, une parmi dautres; en sorte que dfinir le mot science en gardant
les yeux fixs sur la physique, considre comme le modle laune duquel se
mesurerait la scientificit de toute autre science, na gure de sens. Que des physiciens
se laissent aller une telle pense, passe encore, on peut comprendre; mais que des
pistmologues entonnent le mme couplet, cela me semble incomprhensible.
Parce quil avait pris explicitement ses distances par rapport ce prjug, je me
disais que je trouverais chez Canguilhem lindication dune inflexion, voire mme une
rectification, des thses de lpistmologie historique la plus dure (labore partir de
lexemple de la physique), en vue de son adaptation dautres spcialits
scientifiques . Jimaginais dj Canguilhem me livrant les clefs de portes que mon
4
G. Canguilhem, Idologie et Rationalit dans les Sciences de la Vie, 1977, Paris : Vrin, 2000, p. 24.
146
faux passe-partout, forg mme la serrure de ma propre porte, nouvrait visiblement

pas.
Do le titre de mon intervention: Le concept dobjectivit chez Canguilhem .
***
Je nai pas trouv ce que je cherchais.
Ou plus exactement, je nai pas trouv chez Canguilhem une thorie de lobjectivit.
Jaurais d men douter: Canguilhem est rput pour avoir expressment refus de
produire un discours spar relevant de la catgorie pistmologie , obligeant son
lecteur rechercher dans ses tudes non gnralistes les lments de ce que lon doit
quand mme considrer comme une thorie . Jaurais d savoir que je ne trouverais
dans son uvre le mot objectivit quau dtour de telle ou telle tude portant sur tel
ou tel problme bien prcis.
Pourtant, cest par ce mot, objectivit , quen 1965 Canguilhem est entr, de faon
fracassante, dans le monde de lenseignement de la philosophie au lyce. Interrog par
Alain Badiou pour la tlvision scolaire sur le thme philosophie et science , il avait
affirm: il ny a de vrit que scientifique , lexpression connaissance
scientifique est un plonasme, car expliquait-il:
il ny a pas de connaissance spcifique sans dune part des thories
mathmatiques trs labores, et sans dautre part le maniement dinstruments de
plus en plus complexes.5
Et un peu plus tard, prcisant sa pense, il indiquait quil visait ainsi la philosophie
(et accessoirement le marxisme, peut-tre aussi la psychanalyse): il ny a de vrit que
scientifique, il ny a pas de vrit philosophique , et il ajoutait:
Mais dire quil ny a de vrit que scientifique ou quil ny a dobjectivit que de
la connaissance scientifique (soulign par moi), cela ne veut pas dire que la
philosophie est sans objet.6
Vritscienceobjectivitobjet. Tel est le trajet suivi par Canguilhem en cette

occasion du moins. Ma qute dautre chose que ce que dit la physique sur le sujet de
lobjectivit commenait mal. Il ny a l rien qui remette en cause le statut de la
physique (non nomme, mais reconnaissable la mention des mathmatiques trs
labores et de linstrumentation de plus en plus complexe ) comme modle et
parangon de toute science, plus mme de toute connaissance objective. Il ny a pas
dobjectivit en dehors de ce que la physique entend par l.
5
6
Publi dans la Revue de lEnseignement philosophique, dcembre 1964-janvier 1965, p. 10-17.

Publi dans Structuralisme et marxisme , Paris, U. G. E., 1968.
147
Balibar
***
Il est temps que jindique en quoi consiste ce que la physique entend par
objectivit . Il faut dabord prendre acte de ce que lobjectivit en physique a une
histoire ; la signification que prend ce mot dans la physique actuelle, celle des 20e et 21e
sicles, est le rsultat dune volution historique.
Lobjectivit, telle que la conoit la physique, est au cur mme de lentreprise de
Galile, fondateur, comme chacun sait, de la physique moderne, post-aristotlicienne.
Assurment, et comme la soulign Lorraine Daston,7 le mot objectivit , ne peut pas
figurer et ne figure pas, pas chez Galile puisquil a t introduit bien plus tard dans les
diffrentes langues europennes ; plus mme, ce nest qu la fin du 19e sicle que le
mot a pris le sens dune qualit, propre certains noncs (nonc scientifique ou rcit
dun vnement par un tmoin), dtre non marqus, de ne rien reflter de la subjectivit
de celui ou de celle qui les a produits, dont ils sont donc indpendants. Pourtant lide
que certains noncs concernant le monde puissent tre communs des observateurs
ayant dans ce monde des positionnements (jemploie ce mot pour viter le mot position
car il sagit dun tat et non dune simple localisation) diffrents se trouve bien chez
Galile. Elle est au cur de ce qui par la suite a t appel le principe de relativit .
Principe de relativit qui chez Galile ne se prsente pas comme un principe,
solennellement nonc, mais prend la forme dune fable (voir ce que dit Italo Calvino
ce sujet) munie dune morale (comme il est de rgle pour les fables). Cette morale ne
fait pas, et ne peut faire, rfrence une quelconque relativit puisque le mot na fait
son apparition, lui aussi, qu la fin du 19e sicle, en Angleterre principalement. En
rsum, lobjectivit, sous le nom de relativit, deux mots qui nont pas cours au 17e
sicle, est constitutive de la physique moderne sa naissance. Peut-tre serait il plus
juste, et tout aussi anachronique, dappeler principe dobjectivit ce qui est connu sous
le nom de principe de relativit.
La physique se dveloppe ensuite pendant plus de deux sicles sur la base de, cest-dire en conformit avec, ce principe de relativit, qui nen a pas encore le nom mais
qui en acquiert, au fil des sicles, le statut ( la fois commandement et commencement)
dans la pratique des physiciens. Cette pratique est formalise la fin du 19e sicle,
principalement par Poincar, qui en tant que physicien et mathmaticien donne la
thorie physico-mathmatique (les deux sont indissociablement lis, car les
mathmatiques sont constitutives de la physique) de ce nouveau statut. Fait
remarquable peut-tre unique dans lhistoire Poincar dgage la signification
pistmologique, erkenntnistheoretisch, du principe de relativit en mme temps quil
en fait la thorie physico-mathmatique. Cet exploit est rendu possible par le fait quil
est, avant toute chose, un mathmaticien. En tant que tel, il est mme de reconnatre
dans le principe de relativit lexpression de ce quen termes de mtier on appelle une
invariance, objet dune thorie mathmatique bien prcise, la thorie des groupes. Les
lois de la physique (expression en cours depuis Newton) sont invariantes par certaines
transformations bien prcises (formant un groupe, le groupe dit aujourdhui de
Poincar ), transformations qui concernent le reprage dans lespace et le temps des
7
Daston et Galison, Objectivity, p. 30 et sv.
148
systmes physiques, ce que jai tout lheure appel le positionnement de

l observateur (le terme d observateur est peu satisfaisant, mais il est
canonique), par rapport ce qui est observ (les systmes eux-mmes). Ce qui en
termes pistmologiques revient dire que lnonc des lois de la physique est
indpendant de celui qui les nonce (l observateur est galement nonciateur,
producteur de discours, jy reviendrai), non pas de faon gnrale et universelle, mais
de faon conditionne (les observateurs susceptibles dnoncer de la mme faon les
lois de la physique forment une classe, dfinie par certaines conditions imposes leurs
positionnements relatifs; par exemple, dans le cas de la relativit galilenne, ou
einsteinienne restreinte, ils sont en translation uniforme les uns par rapport aux autres).
En ce sens, les lois de la physique ont, comme dit Poincar dans le langage de son
poque, une valeur objective , mais cette objectivit est conditionne et non
universelle; objectivit sous conditions, objectivit restreinte. Ajoutons, objectivit
discursive.
La conception physico-mathmatico-pistmologique de lobjectivit dgage par
Poincar est encore aujourdhui au fondement de la Weltanschauung des physiciens.
Ces derniers sont coupables de navoir pas cherch faire comprendre en dehors de leur
communaut la subtilit du schma thorique de Poincar, laissant courir
l extrieur deux versions qui toutes deux dforment la nature de lide dobjectivit
en physique. Premire version : lobjectivit de la physique est une ide difficile,
inaccessible au commun des mortels, exigeant pour sa comprhension des
mathmatiques labores et une instrumentation de plus en plus complexe ,
comme dit Canguilhem face Badiou et aux tlspectateurs scolaires. Deuxime
version, inverse de la prcdente: lobjectivit de la physique est prsente, par les
physiciens eux-mmes mais surtout par les vulgarisateurs, comme une ide simple que
tout un chacun peut comprendre : lobjectivit est le nom (en t , signe
dabstraction) que les physiciens donnent au fait quils sont daccord entre eux, que tout
ce qui est nonc a t soumis au contrle, au consensus , de la communaut des
physiciens. De l lide, totalement intriorise par les membres de cette communaut,
que leur activit a un rapport privilgi la vrit.
En ralit, lide dobjectivit en physique nest ni triviale ni difficile ; elle est
complique (au sens positif de ce terme).8 Lexemple fondateur de lnonc du
principe de relativit par Galile permet peut-tre de saisir o se situe la
complication.
Dire, comme le fait Galile, que les choses se passent de la mme faon bord dun
navire quai Venise, attendant de partir, et lorsque ce navire vogue vers Alep en
rgime de croisire en ligne droite vitesse constante, sans heurts cest dire, en
termes plus abstraits, que les rsultats de ltude de la nature (ce quest
tymologiquement la physique) sont les mmes pour deux observateurs, en tat de
mouvement de translation uniforme lun par rapport lautre (et lautre par rapport
lun). Il ne sagit pas dune identit de vue: la trajectoire dune pierre tombant du haut
du mt du navire est, vue par le marin embarqu, un segment de droite (un morceau de
8
Il est compliqu de sortir dun labyrinthe ; il est difficile de gravir le col du Tourmalet.
149
Balibar
verticale) alors que lobservateur quai voyant passer le navire, observe un segment de
courbe (parabole). Ce que dit Galile est plus compliqu : ces deux observateurs ont
sur le monde (la chute de la pierre) des points de vue quivalents: certes, ils ne voient
pas la mme chose, mais ils peuvent se mettre daccord, en rflchissant leurs
positionnements respectifs, cest--dire en tant que sujets, sur le fait quils ont observ
la mme chose: la chute dune pierre. Le point important ici est: il ny a pas identit
mais quivalence; lobjectivit en physique est louverture du sujet autre chose qu
son je , la conscience dappartenir une certaine classe une conscience de classe,
pourrait-on dire.
Une complication supplmentaire surgit lorsquon se pose la question de savoir ce
qui est significatif dans la chute dune pierre. En raison du principe de relativit, la
chute dune pierre acquiert le statut dobjet (au sens de la philosophie kantienne); elle
nest pas simplement ce qui se prsente la vue ; elle est, sous la varit des
apparences, permanence, invariance dun certain type de relations entre la pierre et le
sol. Comme le fait remarquer Poincar bien plus tard, la connaissance objective au
sens de la physique ne porte pas sur la nature de la chute ou de la pierre en soi, mais
bien plutt, et seulement, sur les relations, les rapports, entre les choses (ici les rapports
entre la pierre et le sol). Le principe de relativit nonce que ces rapports sont les
mmes pour les deux observateurs; cest en cela que leurs points de vue sont
quivalents: les rapports entre les choses quils voient sont conservs dun observateur
lautre, mme si la description de ce quils voient diffre de lun lautre (lun voit une
droite, lautre une parabole). Cette proprit stend videmment tous les observateurs
en translation uniforme par rapport lun de ces deux- l. Il existe une classe de points
de vue sur le monde quivalents. Tel est finalement lnonc de tout principe
dinvariance en physique, dont le principe de relativit nest quun exemple. Il se trouve
que, dit trs rapidement et de faon schmatique, en mathmatiques, chaque
invariance correspond un objet mathmatique appel groupe , et chaque groupe une
gomtrie ; la physique, au terme dun long trajet, retrouve donc ses origines: le grand
livre de la Nature est crit en caractres gomtriques.
Si jai dvelopp un peu ce point, au risque dennuyer, cest, je lai dit, pour faire
sentir ce que la notion dobjectivit en physique a de compliqu. Non pas tant parce que
des mathmatiques sophistiques y seraient luvre, ce nest pas le cas; mais bien
parce que la description mme des conditions de lobjectivit, donc des conditions de
possibilit de la connaissance en physique, est assez retorseau point que la vue quen
donnent les yeux de lesprit (comme disait Maxwell) se perd parfois dans le labyrinthe
des changements dobservateur et quil faut avoir recours des schmas, non pas des
reprsentations douteuses de la ralit, mais des crobards dune navet droutante
pour qui les observe de lextrieur, qui nont dautre fonction que dviter au
raisonnement de se perdre dans les couloirs. Les tudiants de physique, apprenant
manipuler les concepts de la relativit (restreinte), peuvent tmoigner de la facilit avec
laquelle on se trompe dans la description dune situation physique vue depuis plusieurs
points de vue quivalents.
Cest pourquoi je ne suis pas satisfaite par certaines formulations de lobjectivit
scientifique telle que je les trouve sous la signature de Lorraine Daston et Peter
150
Galison : ds la premire page de leur livre, ils annoncent: Objectivity is blind sight,
seeing without inference, interpretation, or intelligence ,9 utilisant une mtaphore
connotation footballistique (au football amricain, un joueur peut tre oblig de taper
dans le ballon en arrire, sans pouvoir viser); et plus loin, dreams of a view from
nowherea Gods eye perpective of the universe . Parler de view from nowhere ,10
cest oublier que lquivalence de certains points de vue, qui est au fondement de la
physique post galilenne, nest en rien universelle, quelle est toujours conditionne,
ne concernant explicitement quune classe dfinie par une relation dquivalence. Il ne
sagit nullement de view from nowhere , encore moins de Gods eye perspective;
cest dailleurs en opposition cette ide que sest construite la physique moderne !
En revanche, je suis prte soutenir qu cet gard les noncs de la physique ont
quelque chose voir avec ce que Donna Haraway appelle des situated knowledges ;
les uns et les autres, les noncs de la physique et les noncs situs , sont penss
against the god-trick of seeing everything from nowhere ,11 prcisment; les uns et
les autres sont incomplets tant que le point de vue do ils sont mis nest pas qualifi.
La diffrence tient ce que Donna Haraway accepte tous les points de vue, individuels,
alors que la physique rduit cette diversit (Haraway est en droit de parler cet endroit
de rductionnisme) en dclarant dune part lexistence de points de vue quivalents,
dautre part que seuls les noncs mis depuis ces points de vue sont autoriss , (au
sens du label appellation contrle pour les vins et les fromages). La connaissance,
situated knowledge, que promeut Donna Haraway est une forme de connaissance, brute
de dcoffrage, brutalement diverse, qui ne cherche ni laccord, ni le consensus, au motif
(raliste et justifi !) que celui-ci ne peut tre obtenu quau prix dune rduction de la
diversit. La physique mathmatique, pour sa part, produit une connaissance quelle sait
rduite au regard de la diversit, mais qui prsente lavantage de pouvoir tre
partage, par les porteurs de ces points de vue quivalents, slectionns au sein de
ltendue de la diversit, et capables de discours leur permettant de se mettre
daccord. L encore, Poincar est irremplaable:
Que devons-nous entendre par objectivit? Ce qui nous garantit lobjectivit du
monde dans lequel nous vivons, cest que le monde nous est commun avec
dautres tres pensants Telle est la premire condition de lobjectivit: ce qui
est objectif doit tre commun plusieurs esprits, et par consquent doit pouvoir
tre transmis de lun lautre et comme cette transmission ne peut se faire que par
le discours, nous sommes bien forcs de conclure : pas de discours, pas
dobjectivit.12
Quon ne sy trompe pas le discours dont il est question ici nest pas le langage
formel des mathmatiques, mais bien le laus en termes vernaculaires qui fait dune
formule un nonc, et garantit la discursivit de lobjectivit. Finalement, il apparat que
9
10
11
12
Voir aussi Daston et Galison, Objectivity, p. 161 : a new way of looking at an individual plant, as if
liberated from the second sight or prior knowledge, desire, and aesthetics. In this blind sight lay an
epochal novelty .
Daston et Galison, Objectivity, p. 17, 51, 52.
D. Haraway, op. cit. p. 6
H. Poincar, La valeur de la science 1905, Paris: Flammarion, coll. Champs, p. 178-179.
151
Balibar
cest sur la question du langage que Haraway13 scarte de la conception de lobjectivit

scientifique, telle que je lai prsente ici, sur les pas de Poincar. Science has been
about a search for translation, convertibility, mobility of meanings, and
universality reconnat-elle, et Poincar ne laurait pas dsavoue. Mais elle ajoute:
which I call reductionism, when (soulign par moi) one language (guess whose)
must be enforced as the standard for all the translations and conversions.14 Guess
whose? celui des hommes bien videmment; one language? les mathmatiques. Sur ce
point, je ne suis plus compltement daccord: certes, les mathmatiques distinguent les
deux types de connaissance, situe dun ct, objective au sens de la physique de
lautre ; mais la distinction nest pas aussi clear cut: les mathmatiques, sans le
discours qui doit ncessairement les accompagner en physique, ne fournissent pas
dnoncs physiques. Justement, ce qui mintresserait, cest de savoir ce qui, dans ce
discours, perce de situation (au sens de Haraway) de celui qui lnonce.
Avec le sens de la provocation qui la caractrise, Donna Haraway affirme que la
connaissance situe correspond une forme dobjectivit, lobjectivit fministe. Ce
nest pas le lieu ici de discuter la pertinence de ladjectif fministe . Mais, cest
juste titre me semble-t-il quelle revendique le mot objectivit en lassortissant dun
qualificatif. Car, ce faisant, elle prend revers lusage courant de la langue qui, avide de
consolation cumnique, ne retient de lobjectivit (telle que je viens dessayer den
prciser le sens en physique) que le consensus un consensus mou, aurait dit Gilles
Chtelet, que toute spcification (scientifique, fministe, etc.), en le situant ,
risquerait de remettre en question. Cet usage du mot objectivit comme synonyme
de consensus partag15 va de pair avec lide que lopinion est ce qui ne fait pas lobjet
dun partage consensuel (ce quexprime parfaitement, lexpression populaire franaise:
cest mon opinion et je la partage ). Rien ne prouve que la connaissance situe de
Haraway doive tre range dans la catgorie de lopinion objet de mpris traditionnel
chez les scientifiques. Prcisment parce quelle est situe , cest--dire assortie de la
description du point de vue do elle est produite, alors que lopinion mane dun je
totalement impersonnel, au point de devoir redoubler laffirmation de son identit par
cet irritant moi je quaffecte le langage courant et dont il est si difficile de se garder.
***
Mais revenons Canguilhem. Pour constater quil a raison: proprement parler, il
ny a dobjectivit que de la connaissance mathmatise de la physique. Le mot, dans le
13
14
15
The further I get with the description of the radical social constructionist programme and a particular
version of postmodernism, coupled to acid tools of critical discourse in the human sciences, the more
nervous I get.
Haraway, Situated Knowledges , p. 5.
Voir R. Rorty : For pragmatists, the desire for objectivity is not the desire to escape the limitations
of ones community, but simply the desire for as much intersubjective agreement as possible, the
desire to extend the reference of us as far as we can. Insofar as pragmatists make a distinction
between knowledge and opinion, it is simply the distinction between topics on which such agreement
is relatively easy to get and topics on which agreement is relatively hard to get , Objectivity,
Relativism and Truth, in R. Rorty, Philosophical Papers, Volune I. Cambridge/New York :
Cambridge University Press, 1991.
152
domaine scientifique du moins, et donc en pistmologie, renvoie ncessairement aux

mathmatiques.16
Canguilhem, en somme, adopte le point de vue de la physique. Faut-il mettre cet
accord au compte de cette fascination pour la physique, science idale et idal des
sciences que lui reproche Alain Prochiantz, dans un contexte diffrent?17 Je ne pense
pas; lexplication est trop courte.
Relisant Poincar, je me suis avise de ce quil ne se pose jamais la question de
savoir de quelle nature est lobjectivit. Quelle est la valeur objective de la science? Et
dabord que devons-nous entendre par objectivit?, crit-il.18 Que devons-nous
entendre, certes; et Poincar donne sa rponse la question. Mais quelle catgorie de
la pense appartient le concept, la notion dobjectivit? Sagit-il dune loi, dun
principe, dune qualit, dune relation, dune fonction, dune transformation, dun style,
dun acte de pense? Disons tout de suite, pour viter toute ambigut, que chez
Poincar, lobjectivit nest pas un concept mathmatique: ce nest pas parce que ce
que nous devons entendre par objectivit peut tre rapport la thorie des groupes
que l objectivit est un concept mathmatique.
Pour Lorraine Daston et Peter Galison, qui ne posent pas la question dans ces termes,
il ne fait aucun doute que lobjectivit est une vertu, an epistemic virtue... Une parmi
dautres, car ils en distinguent plusieurs qui ont nom: truth to nature, mechanical
objectivity, trained judgment, chacune caractrisant un certain type de rapport au
savoir une certaine poque un certain type de subjectivit du savant galement
( virtues earn their right to be called virtues by molding the self ),19 chacune poussant
lautre vers la sortie dans une succession historique, qui nest pas celle des rois montant
lun aprs lautre sur le trne, mais doit plutt tre dcrite comme lmergence de
tendances contradictoires qui portent ce que la vertu prcdente soit dlaisse, car elle
ne remplit plus son rle de guide.20 Je dois avouer que cette conception me parat trop
empirique, et surtout trop vertueuse (mme pistmique, une vertu reste une vertu).
Quest-ce que lobjectivit, de quelle nature est lide dobjectivit?
Jen tais l de mes cogitations, lorsque je suis tombe, dans les Archives
Canguilhem, sur les notes de cours, crites alors quil prparait un cours, donn la
Sorbonne durant lanne universitaire 1958-59, intitul Comment concevoir le rapport
entre objet et objectivit? La lecture de ces notes ma finalement conduite la thse
suivante: si, sur la question de lobjectivit, Canguilhem semble accorder tant
dimportance la physique, cest parce que seule la physique donne de lobjectivit une
16
17
18
19
20
En ce sens, qui est celui que donne Poincar au mot, objectivit a indubitablement une connotation
litiste. Ce que signale bien le texte de Thomas de Quincey (1856), cit par Daston and Galison
(Objectivity, p. 31), o son usage est qualifi de pedantic . Dans le Littr (1866-1873)
objectivit est prsent comme terme de philosophie moderne .
A. Prochiantz, Le matrialisme de Georges Canguilhem , in Georges Canguilhem. Philosophe,
historien des sciences, Actes du Colloque (6-7-8 dcembre 1990) organis par le Collge International
de philosophie, Paris : Albin Michel, 1993, p. 276.
Poincar, La valeur de la science, p. 178.
Daston et Galison, Objectivity, p. 41
Ibid., p. 41, 11-113.
153
Balibar
dfinition qui lui soit intrieure, une dfinition qui fait de lobjectivit non pas une vertu
pistmique, une qualit surajoute, un supplment dme, un style, mais le moteur de
la construction de son objet, autrement dit une norme.
Les notes en question tiennent sur deux pages recto-verso, probablement celles que
Canguilhem a apportes avec lui dans lamphi le jour (ou les jours) o il a prononc ce
cours sur le rapport entre objet et objectivit, pour lui servir de pense-bte. Aprs avoir
rappel que le mot objectif est la fois un terme de philosophie scholastique (voir
lusage quen fait Descartes quand il parle de ralit objective21, intermdiaire entre la
chose et lide, pour aller vite) et un terme de la science moderne (il faut entendre, celle
du 17e sicle), fonde sur lusage des appareils doptique, Canguilhem note: Dans lun
et lautre cas, objectif est un concept emprunt lexprience de la vision . Comme
Lorraine Daston, Canguilhem fait remarquer que le problme de lobjectivit de la
connaissance ne sest pas pos dans les mmes termes avant et aprs la fin du 19e sicle.
Vient ensuite la phrase: On peut prsenter les rapports de lobjectivit et de lobjet de
deux faons selon que lon va de lobjet lobjectivit ou inversement .
Truc de prof, dira-t-on. Certes, on sent le pdagogue qui tient lide autour de
laquelle il va construire son cours. Pourtant, et bien que ce ne soit pas explicitement
crit, on comprend que ce balancement est plus quune astuce de vieux routier de
lenseignement ; il met en relief un retournement de situation dans le champ des ides
(et pas seulement des mots), un renversement de perspective rendant ncessaire
lapparition du mot objectivit , prcisment la fin du 19e sicle, un moment o
lide ne pouvait pas ne pas apparatre, compte tenu du dveloppement de la
connaissance scientifique dalors.
1. De lobjet lobjectivit.
La connaissance est une varit de la passivit de lme. Toute la question sera de
distinguer dans cette passivit - la passivit de contemplation. Ide
- la passivit de passion. Idoles.
Thorie chez Bacon des Idoles, c..d. des sources derreurs. Idem chez Descartes.
La thorie de lobjectivit cest une thorie des prjugs (pr-cipitation,
prvention). Etre objectif, cest rsister un entranement.
La rfrence Bacon et Descartes indique assez que la priode considre prcde

le tournant de la fin du 19e sicle. Lobjectivit est la parade quinvente lintellect
soumis aux sollicitations de la passion. Lobjet, par le truchement de la vision, imprime
sa marque dans lme, mlange dide et derreurs; lobjectivit est alors une attitude:
une ascse (en quoi elle a les attributs dune vertu),22 une rpression de la spontanit
21
22
Descartes : Par la ralit objective dune ide, jentends lentit ou ltre de la chose reprsente par
cette ide, en tant que cette entit est dans lide ; car tout ce que nous concevons comme tant dans
les objets des ides, tout cela est objectivement ou par reprsentation dans les ides mmes (Rponse
aux deuximes objections).
Voir certaines affirmations de Bachelard (La formation de lesprit scientifique, Paris : Vrin, 1938,
p. 242-244): Il ny a pas de dmarche objective sans la conscience dune erreur intime et
premire prparons-nous mutuellement par une vritable confession de nos fautes
intellectuelles , par une catharsis pralable cet asctisme intellectuel qui teint toutes les
154
(source derreurs), la recherche sans passion, de lentit qui est dans lide. Pour
Descartes, note Canguilhem dans la section 1. De lobjet lobjectivit , lobjectivit
est garantie par lIde mathmatique, la simplicit de lquation algbrique avec,
ajoute-t-il, les limitations inhrentes la mthode de Descartes.
A cette premire phase de lhistoire de lobjectivit, recherche de ltre de la chose
reprsente par lide, tributaire des limites imposes par la faible extension du domaine
dapplication de la gomtrie analytique sa naissance, Canguilhem fait succder une
deuxime phase (toujours sous la rubrique 1. De lobjet lobjectivit ) inaugure
par Newton qui 1) largit la porte de lanalyse mathmatique en faisant accder
linfiniment petit au statut de ralit objective , donc d tre de la chose et 2)
calcule ce que Kepler (il et lunette) avait cru lire dans le ciel . De lire calculer il
y a la mme diffrence que de passivit activit. Ce que Kant enregistre poursuit
Canguilhem.
Do une nouvelle thorie de lobjet scientifique, partir de lobjectivit entendue
comme une certaine activit constituante de lobjet. Lobjectivit nest plus une
attitude, elle est une action. Elle scarte de plus en plus de la vertu, si je puis
dire. Mais cette thorie de lobjectivit constituante ne peut pas se dbarrasser
totalement de la rfrence lobjet. Kant reste tributaire du rapport de la
connaissance lobjet de la perception. Mme quand il aligne la perception sur la
science, il comprend la science comme une perception : des concepts sans
intuition sont vides. On va donc conserver ici un objet dont on ne peut rien
connatre, bien quon ne puisse sen passer.23
De cette section 1. De lobjet lobjectivit , et plus spcialement de cette

objectivit avec objet (lexpression est de Canguilhem) qui passe par la construction
de lobjet scientifique, partir de limpression produite dans lme par lobjet, Proust
donne un exemple presque parfait. Il sagit du moment o travers les fleurs, dans le
dsordre des outils de jardinage, le narrateur voit Gilberte, la fille de Swann, dont il va
tre amoureux, pour la premire fois:
Ses yeux noirs brillaient et, comme je ne savais pas alors, ni ne lai appris depuis,
rduire en ses lments objectifs une impression forte, comme je navais pas, ainsi
quon dit assez d esprit dobservation pour dgager la notion de leur couleur,
pendant longtemps, chaque fois que je repensai elle, le souvenir de leur clat se
prsentait aussitt moi comme celui dun vif azur, puisquelle tait blonde.24
Tout est l: lobjet (les yeux noirs) produit une impression forte ; le narrateur qui
nest pas un savant (ce qui savre tre une supriorit: serait-il tomb amoureux si,
Gilberte lui tait apparue dans son souvenir avec des yeux noirs?) ne sait pas, par
soustraction ( rduire en lments objectifs ), constituer dans son entendement un
objet de connaissance ( la notion de leur couleur ). Il lui manque pour cela cette
23
24
intuitions . Limpression de dlire en acte se confirme quand on en vient lapplication de ces

principes lducation des jeunes esprits
Cest exactement ce quest lespace absolu de Newton ?
Marcel Proust, Du ct de chez Swann, 1913, p. 141 ; le narrateur voit pour la premire fois Gilberte,
la fille de Swann.
155
Balibar
facult que Kant appelle lintuition (mais que lignorant, ainsi quon dit , appelle
esprit dobservation et que Canguilhem a dsign plus haut dans ses notes comme la
passivit de contemplation ). Manque desprit dobservation (ou de contemplation)
qui laisse le champ libre aux idoles produites par la passivit de passion par exemple,
le prjug selon lequel les blondes ont les yeux bleus. Le procd littraire joue sur le
fait que tout comme le narrateur (qui est amoureux dyeux bleus), le lecteur, attentif au
dtail du texte et donc la couleur noire des yeux de Gilberte, reste tributaire du rapport
de la connaissance lobjet de la perception.
Vient alors la deuxime partie du cours, exact renversement de la premire: 2. De
lobjectivit lobjet.
Aprs avoir distingu lobjectivit avec objet (dj traite dans la premire partie, o
lobjectivit reste dicte par lobjet, o lobjectivit construit son objet, mais en restant
tributaire de lobjet), de lobjectivit sans objet ( objectivit constituant son objet ),
Canguilhem entame la discussion de ce deuxime type dobjectivit. Je copie le texte:
a) lobjectivit est une intuition spcifique. Subjectivit personnelle ngative;
b) lobjectivit est une conception, cest--dire une gnration. Les concepts sont
opratoires. Subjectivit universelle positive;
c) la science moderne est une science des champs, des milieux, lumire,
lectricit, magntisme.
Quelques lignes sont sautes et Canguilhem crit:

Objectivit sans ontologie (Bachelard pre et fille, Oppenheimer25).
Lobjet est un projet, cest lobjectif (technique, militaire jimagine quil faut lire
au sens technique ou militaire de lobjectivit.
Ce qui manifestement est le mot de la fin; le cours en Sorbonne est arriv sa

conclusion. Interprtation soutenue par lexistence dune demi-feuille, demi format
range dans le mme dossier, sur laquelle est crit, dune main pose:
Objet de la connaissance: ce qui est donn, son sujet

: ce quelle a en vue, son projet.

Lobjet cest lobjet de lobjectivit, lobjectif de lobjectivit
Lobjectivit na pas lobjet pour source et origine.26
Avant de commenter cette fin, je cite rapidement quelques phrases crites soit la
suite de ce que je viens de relater, soit sur dautres feuilles demi-format, et qui
apparaissent comme des formulations qui cherchent leur expression
25
26
Sur une autre feuille, moins au net : Oppenheimer 2 juin 1958. Les sciences les plus neuves et les
plus importantes (frquentes a t barr) sont celles dont le public ne sait rien, car il nen a rien appris
en classe.
Que lobjectivit nait pas lobjet pour source et pour origine, on lavait compris, ds la fin de la
premire partie du cours. Quil faille entendre objet au sens de projet , cest ce que Canguilhem
a laiss supposer tout au long de son expos, ne serait-ce quen btissant son expos sur le
retournement objet objectivit, objectivit objet
156
La science nest pas objective en ce sens quelle serait indpendante du savant.27

Lobjectivit cest la proprit de certaines communauts humaines (savants).
Lobjectivit cest le syndicalisme de la connaissance.
Savoureuse formule.
***
a) lobjectivit est une intuition spcifique. Subjectivit personnelle ngative;
b) lobjectivit est une conception, cest--dire une gnration. Les concepts sont
opratoires. Subjectivit universelle positive.
Je cherchais ce quest lobjectivit. Me voil servie, sinon vraiment renseigne. Au

terme des deux alinas a) et b) cits la page prcdente, il apparat que lobjectivit est
la fois une intuition spcifique et une conception, cest--dire une gnration.
Formulation assez peu kantienne, qui justifie que les alinas a) et b) soient suivis de
lalina c), expression un peu dmode, mme en 1958, de lide (Bohr? de Broglie?)
que le procd kantien de construction de lobjet par application des concepts purs de
lentendement des phnomnes apprhends dans lespace et le temps, ne marche plus
sagissant des nouveaux objets de la physique quantique.
Il semble bien que ce soit du ct de Cavaills quil faille chercher ce que
Canguilhem veut dire en donnant une dfinition de lobjectivit deux entres chacune
introduite par lobjectivit, cest . Cavaills, parlant des mathmatiques et
seulement des mathmatiques, a en effet cherch modifier le schma kantien (auquel il
na cess de se rfrer pour le critiquer): Lide est bien kantienne: la mathmatique
est plus que la logiqueen tant quelle suppose lapplication de la pense abstraite
une intuition . Mais. La diffrence avec Kant est quil ny a pas de pense logique
pure, la logique nest quun constituant, non isolable, de toute pense fonctionnant
vritablement.28 Le problme auquel na cess de revenir Cavaills est celui de la
sparation quil juge inexacte, entre intuition et pense abstraite: Il est plus difficile
daller plus avant que Kant dans lanalyse du rle de lintuition; non pas
contemplation dun tout fait, mais apprhension dans lpreuve de lacte des conditions
mmes qui le rendent possible , crit Cavaills29 qui immdiatement va plus avant,
critiquant le fait que chez Kant le concept et lintuition sont impntrables , alors
quil est, lui, la recherche de lalliance interne de lintuition et de labstraction (dont il
pense trouver lexplication dans la thorie du signe de Hilbert): construire un concept
cest prsenter lintuition qui lui correspond, les deux oprations ne sauraient tre
dissocies.
27
28
29
Elle lest en ceci que lobjet est lobjet (au sens dobjectif, de projet) de lobjectivit.
J. Cavaills, Mthode axiomatique et formalisme. Essai sur le problme du fondement des
mathmatiques, Paris: Hermann [Actualits scientifiques et industrielles, n 608, 609 et 610], p. 91-92
(cit par Hourya Sinaceur, Jean Cavaills. Philosophie mathmatique, Paris: PUF, 1994, p. 60).
Cavaills, Mthode axiomatique et formalisme, p. 27.
157
Balibar
Pour complter cette analyse, il conviendrait de rechercher sil y a chez Cavaills

lide que lobjet est lobjet de lobjectivit, lobjectif de lobjectivit. Je nai pas eu le
temps de le faire.
Pour conclure, je voudrais revenir mon projet, suscit par la lecture troublante pour
moi des crits de Daston, Galison et Haraway, qui tait de chercher chez Canguilhem un
sens du mot objectivit qui ne rduise pas cette notion ce quen disent les
physiciens. Je suis loin dtre arrive un rsultat tangible. Il conviendrait maintenant
de dvelopper de faon plus approfondie et plus instruite lide que lobjectivit est
une intuition , intuition qui est en mme temps conception, cest--dire engendrement,
ce que la lecture de Cavaills permet de rapporter des proccupations que les deux
hommes, Cavaills et Canguilhem, semblent avoir partages. Peut-tre y a-t-il l, en
germe, la possibilit de replacer la distinction objectif/subjectif, longuement critique
dans leur livre par Galison et Daston, dans un contexte un peu diffrent, celui de
lopposition traditionnelle entre intuition et concept. A ce propos, une formule de
Canguilhem, souvent cite, mais rarement explicite, me trotte dans la tte, celle de
subjectivit sans intriorit
158
SCIENTIFIC PHILOSOPHIES IN THE EARLY 1930S

AND GASTON BACHELARD ON INDUCTION*
Sandra Pravica
I.
In autumn of the year 1934 a group of philosophers gathered for a little conference in
Prague. They had come from Austria, Germany, France, Poland and the US, or lived
and worked in Prague itself. Most of them held academic positions in philosophy or in
physics. Many had an additional background in mathematics, engineering, or another
scientific discipline. Among them were Moritz Schlick, Otto Neurath, Philipp Frank,
Charles Morris and Hans Reichenbach names which are all in one way or another now
associated with the philosophical movement of logical empiricism or the Vienna
School, as it was mainly referred to at that time, because the epitome logical
empiricism had yet to be coined.
Although the meeting was merely a pre-conference, meant to organize what would
become the First International Congress for the Unity of Science, it is remarkable
because it was the first official context in which members of the Vienna School used the
term philosophie scientifique. Surprisingly it shows up in the French language, although
communications at the conference were predominantly held in German with a few in
English and only one Louis Rougiers in French.1 This term had not been present
when the group introduced its ideas publicly on earlier occasions. For example, five
years before when a programmatic text of the Vienna School had been presented at a
symposium,2 which was carried out with the Berliner Gesellschaft fr empirische
Philosophie, the key concept had been Wissenschaftliche Weltauffassung which can be
translated as scientific world conception or scientific world perspective. Scientific
philosophy, or philosophie scientifique, in contrast had played no discernable role.3 In
1934 however, the conference organizers decided that philosophie scientifique should
be the subject for the following years event. Thus, the First Congress for the Unity of
Science, in 1935, was officially called Congrs International de Philosophie
Scientifique.4
Today the term scientific philosophy is quite restricted. It cant be used as
comprehensively as the term philosophy of science for example. To speak of
scientific philosophy involves, in one-way or another, the context of logical empiricism.

* This conference manuscript contains research that will be in more elaborate form part of the
authors dissertation.
1
See Einheit der Wissenschaft. Prager Vorkonferenz der internationalen Kongresse fr Einheit der
Wissenschaft 1934, Leipzig 1934.
2
Tagung fr Erkenntnislehre der exakten Wissenschaften, Prague, September 15th/16th, 1929.
3
See Wiener Kreis. Texte zur wissenschaftlichen Weltauffassung von Rudolf Carnap, Otto Neurath
Moritz Schlick, Phillipp Frank, Hans Hahn, Karl Menger, Edgar Zilsel und Gustav Bergmann, ed. by
Michael Stltzner and Thomas E. Uebel, Hamburg [etc.] 2006, pp. 5-26.
4
See Einheit der Wissenschaft. Prager Vorkonferenz, pp. 2-204.
159
Pravica
Either it relates in an historic manner to the concrete projects of Otto Neurath, Phillip
Frank, Hans Reichenbach or others,5 or it relates to more recent philosophical projects
that affiliate themselves with the tradition of logical empiricism.6 Against this
background it is particularly interesting that the term appears frequently in Gaston
Bachelards epistemological work, which is often positioned in opposition to the
tradition of the Vienna School, or to later branches of the so-called analytical
philosophy.
With philosophie scientifique Bachelard promoted his own new philosophy, and it
shows up throughout his entire epistemological writings. Most interesting, however, for
the case at hand, is that Bachelard used the term in September 1934 in Prague where he
gave a talk at the VIII International Congress of Philosophy.7 This traditional large
meeting took place three days after the pre-conference that I introduced earlier.
Bachelards lecture was part of the first plenary session. Amongst the audience were
members of the pre-conference such as Hans Reichenbach and Rudolf Carnap.8 The
chair of the session was Philipp Frank, who was at that time director of the Institute for
Theoretical Physics at Pragues German University, and constituent of the Vienna
School since his early academic years in Vienna. Both the encounter between Bachelard
and some representatives of logical empiricism, and Bachelards participation in one of
the large international congresses for philosophy were events that did not recur. This
provides an occasion to consider anew the relationship between the two scientific
philosophies involved. Above all, it allows us to reassess Bachelards role as a
philosopher of science. Bachelards philosophy is frequently seen as a kind of
counter-position to logical empiricism or, more generally, as being somehow outside of
or at odds with the discourse that is usually referred to as philosophy of science be it
with regard to the past or the present. This opinion is generally based on the assumption
of a certain non-communication of philosophical traditions. Mary Tiles, for example,
claims that Philosophers of science in the 1930s were talking about foundations,
simplicity, and observation protocol statements,9 according to her, Bachelard is no
8
9
This can be seen for example in Stadlers equation of scientific philosophy and Logical Empiricism.
See Friedrich Stadler, Wien Berlin Prag. Zum Aufstieg der wissenschaftlichen Philosophie, in:
Wien-Berlin-Prag. Der Aufstieg der wissenschaftlichen Philosophie. Zentenarien Rudolf Carnap
Hans Reichenbach Edgar Zilsel, ed. by Rudolf Haller and Friedrich Stadler, Wien 1993, pp. 11-37.
The title is a reminiscence of Hans Reichenbachs book from the early 1950s. See Hans Reichenbach,
The rise of scientific philosophy, Berkeley [etc.] 1951.
So a workshop that took place under the name Scientific Philosophy. Past and Future, in April 2010
at Tilburg University, alluded to the doctrines of logical empiricism when it defined Scientific
Philosophy as a system in which philosophical theses and arguments should be as clear and precise
as scientific ones <http://www.tilburguniversity.edu/research/institutes-and-research-groups/tilps/
sppf2010/>.
Gaston Bachelard, Critique prliminaire du concept de frontire pistmologique, in: Actes du
huitime Congrs International de Philosophie Prague 2-7 Septembre 1934, Prague 1936, pp. 3-9.
See ibid., pp. 31-36.
Mary Tiles, Technology, science, and inexact knowledge: Bachelards non-Cartesian epistemology,
in: Continental philosophy of science, ed. by Gary Gutting, Malden, MA 2005, pp. 157-175,
on p. 157-158.
160

SCIENTIFIC PHILOSOPHIES IN THE EARLY 1930S AND GASTON BACHELARD ON INDUCTION
ordinary epistemologist; []. He simply isnt looking in the same place that most
philosophers of science look, [].10
Gary Gutting also contributes to this view when in a similar vein he subsumes
Bachelards philosophy under continental philosophy of science in contrast to the
analytical model of philosophy, and asserts that The continental-analytic divide is
grounded in the undeniable fact that, sometime around the end of the 1920s,
philosophers split in two camps that, in short order, had nothing to say to one
another.11 I propose, that in a closer examination the situation doesnt turn out as clearcut as these authors claim. At least with regard to Bachelard, these general
classifications need to be differentiated and even revised, as we shall see.
In the following pages I will first roughly sketch the use of the term philosophie
scientifique as used by members of the Vienna School and by Bachelard. Secondly, I
elaborate on some intersections in their respective research agendas with an emphasis
on the concept of induction. With regards to logical empiricism, I focus narrowly on the
early thirties.12 With respect to Bachelard, I do not restrict myself to a particular period
of his work. Although my arguments may seem to be historical in a sense, my concerns
are philosophical. Im primarily interested in how Bachelards epistemology affects
philosophy today.13
II.
In the early 1930s, logical empiricism as a philosophical movement was at a crossroads.
Its affiliates promoted their doctrines actively on an international scale acting as a
constituted and coherent group that would transform and reinvigorate philosophy. But at
the same time they were increasingly confronted with political pressure in Germany
which forced some of them to emigrate. For example when Otto Neurath visited Prague
in 1934, he was on the way to his exile in the Netherlands.14
For both Bachelard and the representatives of logical empiricism, it was a prior
concern to clarify and redefine the mutual relationship between philosophy and science.
Their respective use of the concept of scientific philosophy indicates a decided attitude
regarding this question. At the large philosophers meeting, attended by Bachelard,
philosophys role vis--vis science was an overarching concern for the ensemble of the

10
11
12
13
14
Ibid.
Gary Gutting, Introduction: What Is Continental Philosophy of Science?, in: Continental philosophy
of science, ed. by Gary Gutting, Malden, MA 2005, pp. 1-16, on p. 2.
For the serious transformations logical empiricism underwent from the thirties on and particularly in
the postwar period, see George A. Reisch, How the Cold War transformed philosophy of science. To
the icy slopes of logic, New York [etc.] 2005, and John McCumber, Time in the ditch. American
philosophy and the McCarthy era, Evanston, IL 2001.
I elaborate on this question extensively in my dissertation Tentative transgressions. Gaston
Bachelards experimental epistemology. See <http://vlp.mpiwg-berlin.mpg.de/exp/pravica/
project.html> (summary).
Stadler, Wien Berlin Prag, p. 31.
161

Pravica
congress sections. So in his inaugural address the president of the organisational

committee, Emanuel Rdl, asks:
How do the diverse domains of philosophy relate to each other? Is it appropriate
to assume that all fields of scholarship eventually depend on the natural sciences,
which have become in a way dictatorial in philosophy? Or shall we not admit
for every single academic field the principle of self-determination, so that a jurist
or a philologist would have the same original right to impose their methods and
results to philosophy?15
Edmund Husserl, by the way, who could not attend the meeting, also sent along a
missive with a negative pronouncement on the authority of science vis--vis
philosophy.16
On the side of the logical empiricists the term scientific philosophy, in the
conference context, appears in texts with a rather programmatic character; exhibiting a
negative as well as a positive function within the rhetoric. Negatively, it deployed
delimitation against approaches that critique was centred on. In the foreword of the preconference papers we can read that philosophie scientifique, in the organisers terms,
meant discussion of a doctrine of science, free of metaphysics (metaphysikfreie
Wissenschaftslehre).17 The term metaphysics pejoratively summarized contemporary
philosophical approaches like the Brentano School or the vitalism of Hans Driesch, all
of which were criticized by the Vienna school because of ambiguous argumentation.
Another word with a comparable delimiting function towards other philosophies,
frequently employed in the congress communications, is a priori or the more
pejorative aprioristic philosophy. The rejection of aprioristic knowledge stands for
a professed consent against philosophies that, in the succession of Kant, still adhered to
the concepts of causality, space, time and substance as unmodifiable foundational terms
assumed in scientific practice, although modern physics refuted them as being
foundational.18 Used in a positive sense, scientific philosophy served as a means of
affiliation with the theoretical positions of, for instance, Henri Poincar, Ernst Mach,

15
16
17
18
Actes du huitime Congrs International de Philosophie a Prague, pp. XXV-XXVI.

See ibid., pp. XLIII-XLIV.
See Einheit der Wissenschaft. Prager Vorkonferenz, p.2. In his definition of metaphysics Philipp
Frank refers to Max Planck: Max Planck has called the opposing views of which we speak the
positivist and the metaphysical conceptions of science. He characterises them perfectly correctly in
this way: according to the metaphysical conception, the aim of science is the discovery of an existing
true world, while according to the positivist, however, it is the construction of a system of
statements with the help of which we can find our way in the world of our experiences. Philipp
Frank, Das Kausalgesetz und seine Grenzen, Wien 1932, and id., The Law of Causality and its Limits,
ed. by Robert S. Cohen, Dordrecht/Boston/London 1998.
See, e. g., Hans Reichenbach, Atom und Kosmos, Berlin 1930, pp. 307-313, and id., Die
philosophische Bedeutung der modernen Physik, Erkenntnis 1 (1930/31): 49-71, on p. 57-58 and
p. 70-71. Beyond the realm of short and pointed congress communications, the role of the a priori
for logical empiricism must nevertheless be differentiated. See Michael Friedman, Coordination,
Constitution, and Convention. The Evolution of the A Priori in Logical Empiricism, in: The
Cambridge Companion to Logical Empiricism, ed. by Alan Richardson and Thomas Uebel,
Cambridge [etc.] 2007, pp. 91-116.
162

Pierre Duhem or Gottlob Frege. In suggesting logical analysis for scientific as well as
for philosophical statements, the logical empiricists furthermore explicitly referred to
the work of Bertrand Russel.
The extensive application of logical analysis was the outspoken new feature that
was integrated with elements of positivistic and empiricist philosophical traditions. The
scientification of philosophy was an aim, extensively professed in the respective texts.
Used in this way, scientific philosophy, secondly, meant orienting philosophy according
to scientific representational modes such as logical stringency and clarity of language
and argumentation. Thirdly, scientific philosophy stood for a close relation between the
philosophical work and current scientific research, be it with regard to theory or in
actual cooperation with scientists and scholars. Not least, it should be mentioned that in
1931 the Society for empirical philosophy, centred around Hans Reichenbach in
Berlin, changed its name to Society for scientific philosophy. This change was due,
according to Friedrich Stadler, to the deliberate integration of logics and mathematics.19
Bachelards talk at the plenary session in Prague was entitled Critique prliminaire
du concept de frontire pistmologique. He argued that the question of scientific
limits, as commonly dealt with in philosophy, doesnt make any sense at all, and
employed the vocabulary common to the logical empiricists polemics against
metaphysics and aprioristic knowledge.20 However, the fact that Bachelard, on this
occasion, spoke in a pejorative way about the metaphysician and the metaphysical
limit did not prevent him from applying the term metaphysics positively on different
occasions and to propose for example, metaphysics that are positive insofar as they
undergo an experiment themselves.21 This demonstrates a certain, lets say, flexibility
in Bachelards use of common philosophical terms. Beyond the context of the
conference, Bachelard generally uses philosophie scientifique as an equivalent to the
respective philosophy he was promoting, be it Le nouvel esprit scientifique, La
philosophie du non, or Le rationalisme appliqu. As Bachelard in contrast to the
philosophers committed to logical empiricism did without a justification or definition
of the term philosophy scientifique, I will move on to a concrete item in his scientific
philosophy.
III.
In contrast to today the concept, or the problem, of induction still produced a lively
discussion during the 1920s and 30s. Its role in epistemology was a controversial issue,
not least among the philosophers of logical empiricism. Bachelard, in turn, was

19
20
21
Stadler, Wien Berlin Prag, pp. 27-28.

See Bachelard, Critique prliminaire du concept de frontire pistmologique, p. 3, and pp. 8-9: Le
devoir de la philosophie scientifique semble alors trs net. [...] La philosophie scientifique doit en
quelque manire dtruire systmatiquement les bornes que la philosophie traditionnelle avait
imposes la science. [] Nous demandons par consquent que la philosophie scientifique renonce
au rel immdiat et quelle aide la science dans sa lutte contre les intuitions premires.
Gaston Bachelard, Noumne et microphysique" [1931-32], in: id., tudes. Paris: Vrin, 1970, pp. 1124, on p. 24 (my translation, S.P.).
163

Pravica
concerned with induction from the early years of his doctoral theses until his later
writings. He gradually assimilated induction with one of his epistemological core
principals the concept of recurrent progress (my words, S.P.). With this, the notion
undergoes a dramatic redefinition, which reopens the problem of induction on new
epistemological grounds. I shall briefly sketch Bachelards induction in its ultimate and
most elaborate stage. In this case, as well as in others when he presents philosophical
insight, Bachelard mainly consults specific episodes or instances from the more or less
recent history of science to show, or demonstrate, rather than explain what is at stake
epistemologically. The examples which outline the framework of Bachelards
involvement with induction all stem from mathematics and focus on what he calls
mathematical productivity. He insists on the latter throughout almost all his
epistemological writings.
In his 1931 article The inferential richness of mathematical physics,22 which resumes
an earlier argument started in one of his dissertations, Bachelard presents the
contribution of Gabriel Lam to the process of the mathematical formulation of heat
conduction as being the best guide to comprehend and invent.23 He claims to
demonstrate mathematics applied in physics as being the actual source of scientific
diversification and discovery.24 Lam discovered the anisotropy of crystals regarding
heat conduction, i.e. the dependence of the conductivity on the direction.25 Striving for
the highest level of generality regarding the mathematical account of heat conduction in
crystals, Lam discloses the postulate of commutativity in the equation as restrictive to
generalization. He emphasizes that equal conductivity in either direction is but a special
case of crystalline systems which revealed to be, in general, anisotropic in several
aspects.
Bachelard stresses that by mathematical necessity Lams modifications, that
simplify and generalize the outcome of the calculus, result in a more complicated
equation as well as in a material differentiation and diversification.26 Bachelard's
example is instructive for the mathematical productivity he insists on frequently. It
shows, for instance, how mathematical reasoning brings about functional differences of
a material that didn't previously exhibit distinguishable structures.27
However, it is most important to note that Bachelard emphasizes the productive
impact of mathematics in physics as being a reciprocal one. He speaks of an
impulsion that mathematics receive, in return, from the respective material worked on
and states that with mathematical physics emerge, at the same time, physical

22
23
24
25
26
27
Gaston Bachelard, La richesse dinfrence de la physique mathmatique, in: id., Lengagement

rationaliste, Paris 1972, pp. 109-119.
Ibid., p. 117: On ne peut avoir de meilleur guide pour comprendre et inventer.
Ibid., p. 110.
See ibid., pp. 116-118.
See ibid., pp. 117-118.
See Maurice Loi, Bachelard et les mathmatiques, in: Guy Lafrance (ed.), Gaston Bachelard.
Profils pistmologiques, Ottawa 1987, pp. 46-56, on p. 50.
164

mathematics.28 A similar lesson can be drawn from a more recent instance in the
history of physics, the so-called wave-particle-dualism in quantum physics and the
correspondent formalism an example Bachelard refers to extensively when
introducing his new scientific spirit.29 The non-commutativity of some operators
inserts, via both the conceptual and instrumental level, a moment of indecision or of
possibility into scientific reality and hence a kind of two-sided material reality results.
Accordingly, what by means of experiment is represented as a difficulty of
localization is theoretically available as the non-commutativity of specific operators in
the quantum physical formalism.30
Furthermore, a chapter in Le rationalisme appliqu, which focuses on induction,
provides an example of mathematical productivity in geometry which is a rare
exception because Bachelard normally is interested in instances of applied rationalism
in physics or chemistry. Bachelard substantiates the inductive allure of mathematics
by presenting the concept of orthogonality within the theorem of Pythagoras in its
succeeding theoretical reformulations. He demonstrates a generalisation of
orthogonality in the Pythagorean theorem with algebraic spaces, its application in set
theory and its role as a fundamental term in the functions of wave mechanics, and
focuses particularly on the principle of identity (A = A) that he also wants to prove as
being productive.31
In the successive diverse reformulations of the orthogonality, identity undergoes a
corresponding reformulation as an operational or functional identity.32 Bachelard
stresses that it is actually not the theorem itself that has been generalized, but the feature
that he labels Pythagoricity.33 The generalisation presented here is not a
generalisation on the basis of particular cases; rather it is a top-down generalisation by
which the initial insight reveals its limit a posteriori.34 This introduces the basic
structure of Bachelards notion of induction. The extension articulated with it is not
conceived as a simple inflation, but is basically an enclosure that establishes the
particularity and narrowness of the initial insight afterwards in a recurrent manner by
encompassing it.35 It becomes apparent that Bachelards induction introduces a
particular time axis, detached from chronological time. This recurrent comprehension
provides an intrinsic rational dynamic appropriate to the respective epistemological or
scientific instance which is studied.36

28
29
30
31
32
33
34
35
36
See Gaston Bachelard, Essai sur la connaissance approche, Paris 1927, p. 190, and id., La richesse
dinfrence de la physique mathmatique, p. 118.
See Gaston Bachelard, Le nouvel esprit scientifique, Paris 1991.
See
Johann
von
Neumann,
Mathematische
Grundlagen
der
Quantenmechanik,
Berlin/Heidelberg/New York 1968, pp. 222-237, Michel Serres & Nayla Farouki (eds.), Thesaurus
der exakten Wissenschaften, Frankfurt a. M. 2001, p. 308.
Gaston Bachelard, Le rationalisme appliqu, Paris 1986, p. 83.
See ibid.
See Jean Gayon, Bachelard. Le rationalisme appliqu, Unpublished Manuscript, Vanves 1995, p. 113.
See ibid., p. 112.
See ibid., p. 113.
See Bachelard, Le rationalisme appliqu, p. 96.
165

Pravica
But how, then, does Bachelards particular notion of induction eventually relate to
induction as a controversial issue in the conference context in 1934? Induction has been
largely discussed for centuries in philosophical methodologies of empirical and
experimental science, not least because it is strongly associated with the issues of
prediction and discovery. The problem with induction, in radical abbreviation of course,
is that induction as a syllogism a mode of logical inference is genuinely meant to
provide a consistent relation of premises and conclusion and therefore to guarantee the
validity of statements about the world. But unfortunately it displays a kind of logical
gap, a leap of uncertainty, when scrutinized in detail because its conclusion, so to speak,
transgresses its premises.
Hans Reichenbach, in the conference context of 1934, clearly is an advocate of the
epistemological use of the term induction. He strongly defends the indispensability of
the principle of induction in science and epistemology.37 He reformulates inductive
inference in terms of the logic of probability that he elaborates apart from bivalent
logic38 and claims having given a final solution to the so-called problem of induction
as well as having provided the logics of scientific knowledge per se.39 With his
statement, Reichenbach triggered a dispute that in 1934 was significantly marked by the
positions of Otto Neurath and Karl Popper. Neurath, in his immediate response to
Reichenbachs intervention, strongly disapproved of the suggestions of the latter. He
pejoratively called Reichenbachs formal logics of probability a machine of induction,
not at all desirable for science because of its lack of scientific value. Neurath
emphasized the unity of science as being primarily historical and social. According to
him, there is no way to establish the unity of science by logical inference.40 Karl
Poppers reply to Reichenbach differed considerably from that of Neurath. Popper,
generally favouring the principle of deduction over induction as being the exclusive
basis for what he called scientific logic, reproached induction with the traditional
accusation: it would lead to an infinite regress and to make matters worse it was an
aprioristic principle.41
A closer examination of Reichenbachs comments in the context of the conference
papers from 1934 reveals that his positive adherence to induction in epistemology is
closely connected to the question of which domains of scientific endeavours are at all
able to be rationalized that means represented by rational argumentation on the
basis of formal-logical means.42 In consequence, Reichenbach proposed to discern the

37
38
39
40
41
42
Hans Reichenbach, Wahrscheinlichkeitslogik, in: Einheit der Wissenschaft. Prager Vorkonferenz,

pp. 37-43, id., Zur Induktionsmaschine, ibid., pp. 172-173, id., Wahrscheinlichkeitslogik und
Alternativlogik, ibid., pp. 177-178m and id., Die Bedeutung des Wahrscheinlichkeitsbegriffs fr die
Erkenntnis, in: Actes du huitime Congrs International de Philosophie a Prague, pp. 163-169.
Reichenbach published in 1935 a comprehensive theoretical account of his new approach in Hans
Reichenbach, Wahrscheinlichkeitslehre. Eine Untersuchung ber die Logischen und Mathematischen
Grundlagen der Wahrscheinlichkeitsrechnung, Leyden 1935.
Reichenbach, Wahrscheinlichkeitslogik, p. 43.
See Otto Neurath, Zur Induktionsfrage, in: Einheit der Wissenschaft. Prager Vorkonferenz, pp. 173174, on p. 174.
See Karl Popper, Induktionslogik und Hypothesenwahrscheinlichkeit, in: Einheit der
Wissenschaft. Prager Vorkonferenz, pp. 170-172.
See ibid., pp. 172-73.
166

procedures by means of which a scholar or scientist discovers a theory

(Auffindungsverfahren) from the operations by which they present and defend their
theories publicly (Rechtfertigungsverfahren). Whereas the first kind of procedures, the
procedures of discovery, are according to Reichenbach, as less able to be rationalized
as a guessing game, the second kind of procedures, the operations of justification, have
the privilege of being elaborated and presented systematically by rational means in
philosophy.43 Hence apparently, Reichenbachs advocacy of the role of induction for
epistemology entails the basic structure of the distinction between a context of
discovery and a context of justification, which later would have an enormous impact on
the prevailing conception of the philosophy of science in the 20th century its subjects,
tasks, and its approved methods.44
Bachelard in turn, with his conception of induction, radically relies and elaborates on
what Reichenbach successively, but decidedly, excludes from its relevance for
epistemology: scientific discoveries. Furthermore, in stressing particularly the
exceeding dynamics of induction, Bachelard turns in an extreme way to the issue that
constitutes the problematic side of induction in the traditional conception. He takes on
the logical gaps that prevail in the common philosophical discussion as a genuine
difficulty, but in his approach they turn out as positive.45
Andr Lalande, in his compendium of theories of induction and experimentation
from 1929, gives an historical overview of positions that range from antiquity to
Charles Sanders Peirce.46 He attaches an appendix on important recent publications on
the subject where he presents Bachelards approach in a chapter of the Essai sur la
connaissance approche from 1927,47 one of Bachelards doctoral theses, together with
works of Lon Brunschwicg, Jean Nicod and Maurice Dorelle.48 Bachelard, in Essai,
announced decidedly that he would not engage in the debate on the legitimacy of
inductive reasoning.49 In doing so he clearly dismissed the justificational questions
Reichenbach was focusing on. Bachelard declared that, on the contrary, he was
interested in the application of induction.50 Notably, Bachelards rather theoretical
elaboration on induction in the Essai doesnt yet exhibit features of the principle that is
exceeding and encompassing at the same time, which I introduced earlier.
It is remarkable that Bachelards use and discussion of the term, in 1927, was still
rooted in the common philosophical debate on the issue then although the approach of

43
44
45
46
47
48
49
50
See ibid.
See Gregor Schiemann, Inductive Justification and Discovery. On Hans Reichenbachs Foundation
of the Autonomy of the Philosophy of Science, in: Revisiting Discovery and Justification. Historical
and philosophical perspectives on the context distinction, ed. by Jutta Schickore and Friedrich Steinle,
Berlin [etc.] 2006, pp. 23-39, on p. 23.
Within the framework of the paper in hand this difference with far-reaching consequences cannot be
more than indicated.
Andr Lalande, Les Thories de lInduction et de lExprimentation, Paris 1929.
See Bachelard, Essai sur la connaissance approche, pp. 127-154 (Linduction, la corrlation et la
probabilit dans leur rapport avec la connaissance approche).
See Lalande, Les Thories de lInduction, pp. 265-283.
See Bachelard, Essai sur la connaissance approche pp. 127ff.
See ibid., pp. 127-129.
167

Pravica
applied induction he introduced was obviously already heading for new

epistemological grounds. Only one year later, in La valeur inductive de la Relativit
Bachelards book on the dynamics of the theoretical renewal, that the theory of
relativity provides his new concept of induction, associated with the epistemic
productiveness of the reciprocal information of mathematics in physics and experience,
is clearly recognizable. Hence, compared to the application of induction in Essai, there
is a clear shift in use in the book on relativity. Induction serves as Bachelards main
concept to elaborate on the particular novelty of relativity.51 It helps to present the
theory of relativity as a method of progressive discovery.52
However, Bachelards sudden shift in the use of induction with his book on relativity
goes at the expense of still being conceived as a contribution to the broader
philosophical discourse on induction. So the reviewer Albert Spaier complains that
Bachelard doesnt even warn the reader that he is not taking into account the habitual
meaning of induction throughout the book.53 Five years later in Le nouvel esprit
scientifique, where he makes extensive use of his new principle of induction for
example when introducing a non-newtonian thinking Bachelard additionally
experimented with the term induction itself. In the course of strengthening the new
epistemological grounds he was heading for with induction, Bachelard introduced the
concepts of contraction, envelopement, and transcending induction as
equivalents.
To sum up, even if Bachelards concept of induction at first sight doesnt seem to
match very well with what others at the same time discussed under this term, his
starting point clearly is the problem of induction as generally referred to in the 1920s
and 30s. In the course of his elaboration on induction, the traditional key problems of
the concept namely scientific novelty and discovery remain integrated into
epistemology in contrast to the solution which is offered by Hans Reichenbach who, in
restricting the consideration of induction to a certain context of justification,
completely dispelled the subject of discovery from the philosophy of science. The topic
was banished from mainstream philosophy until the late 1980s.
IV.
There is, of course, more to be said with regard to the yet unnoticed relations between
Gaston Bachelards research agenda and the philosophies of the logical empiricists. I
talked about this subject on other occasions, and Im investigating it in my dissertation.
Given the juxtaposition of some elements in their respective approach covered in this
paper, I would like to emphasize the mutual relationship between Gaston Bachelard and
logical empiricism in the early 1930s in general as being undecided and ambiguous.
Statements like that of Mary Tiles: Bachelard is no ordinary epistemologist; []. He

51
52
53
See Gaston Bachelard, La Valeur inductive de la Relativit, Paris 1929, pp. 5-6 and pp.10-11.
Ibid. p. 6.
Albert Spaier, [Review of: Gaston Bachelard, La valeur inductive de la Relativit, Paris, Vrin
1929], Recherches philosophiques I (1931-32): 368-373, on p. 369.
168

simply isnt looking in the same place that most philosophers of science look [...],54 in
my opinion are hard to maintain.
There are, nevertheless, differences with far reaching consequences in the
approaches presented here. They cannot be more than indicated in the framework of this
paper because Bachelard also puts forward serious reservation against some authors of
the Vienna school, particularly concerning the question of the role of mathematics for
scientific endeavours. As we have seen, he attaches high epistemological importance on
epistemic novelties produced by means of mathematics. In the case of a reduction of
mathematics to logic, widely assumed in the context of logical empiricism at that time,
this productivity goes unnoticed.
The philosophers espousing logical empiricism felt the same indecision and
ambiguity toward Gaston Bachelard as a philosopher. Elisabeth Nemeth recently
alluded to conflicts within the group that prepared the two congresses for the Unity of
science that were both held in Paris in 1935 and 1937. For example, only two months
after the meeting at the congress in Prague, in a letter to Otto Neurath, Philipp Frank
suggested inviting Gaston Bachelard for a talk. Later when arrangements for the
program had to be fixed, Neurath complained to Frank that too many contributions on
logic but far too few on real science (Realwissenschaften) were scheduled.55 So Frank
repeatedly suggested Bachelard besides the physicist Paul Langevin for an
introductory speech. Hans Reichenbach, who himself intended to give the introductory
lecture, objected to the proposal denouncing Gaston Bachelard as being only of
secondary position.56 Asked by Neurath to comment on Reichenbachs intervention,
Philipp Frank responded that Gaston Bachelard had an excellent reputation in France
and that [] among the French professors he is certainly the one most proximate to
us.57

54
55
56
57
Tiles, Technology, science, and inexact knowledge, pp. 157-158.

Elisabeth Nemeth, Philipp Frank und die Verbindung Wien-Paris, Revue de synthse, 6e srie, n 12 (2007): 209-217, on p. 211.
Ibid., pp. 211-212.
Ibid., p. 212.
169

CANGUILHEM AND BLUMENBERG

Working on the Evolution of Problems from both Sides of the Rhine
Cornelius Borck
A dialogue did not happen. How could it be initiated posthumously and why should it?
What could form a productive link between two authors who never discussed each
others writings or even mentioned each other? At first glance the differences
overshadow any possible commonalities. Whereas one published just two monographs
which were both his theses and specialized in essays devoted to a fine grained
analysis of specific conceptual problems or historiographical questions the other was a
prolific writer penning more than a dozen books. And many of them were substantial
tomes navigating vast amounts of material and pushing fundamental philosophical
questions beyond the limitations of conceptual traditions into deep history.
In light of these well-known differences, any comparison may seem a fairly useless
intellectual exercise, even though both shared an interest in the history of science from a
background in philosophy. In fact, the picture of two philosophers specializing in the
history of science would be a highly misleading one as it would invoke a pattern of an
academic distribution of labor, which both of them opposed, and here, indeed, some
shared ground emerges. Both positioned their rigorous analyses, vis--vis the academic
and philosophical traditions, precisely with the aim of questioning the unfolding and
differentiation of philosophical paradigms and disciplinary regimes. The tradition had to
be searched against the grain in order to reveal a hidden logic, relating blind spots and
guiding assumptions across and through major theoretical transformations.
From this observation a first hypothesis in a more systematic direction becomes
possible. Like Canguilhem, Blumenberg sounds out trajectories of disturbances or
interferences, and they both do not aim to solve conceptual problems or to reconcile
divergent positions, but to trace conceptual genealogies precisely to explore the
dynamics of such turbulences themselves. In his The Normal and the Pathological,
Canguilhem stipulates as his task the re-raising and repositioning of already known
philosophical problems instead of putting them aside, thereby following Leon
Brunschvigs definition of philosophy as the science of solved problems. Calling
Brunschvig as his witness, Canguilhem added: We are making this simple and
profound definition our own.1
Blumenberg shared a somewhat similar understanding of philosophy, and
Canguilhems motto could also be said to be his own who not only revised his books
thoroughly before and after publication, but also conceived of his project at large as a
re-writing of the history of philosophy. In Blumenbergs view, such a revision was not
the consequence of a revolutionary program or fundamentally new insights, but a
1
Georges Canguilhem, The Normal and the Pathological, New York: Zone Books 1991, p. 35.
171
Borck
necessary task in order to keep philosophy alive and up to date. As eternally

questioning, philosophy constantly required renewal and re-appropriation:
Thoughtfulness means: not everything remains as self-evident as it was.2
In addition, Canguilhem and Blumenberg both identify the concept as the smallest
unit of analysis. Speaking of Gaston Bachelard, Canguilhem described the genealogy of
concepts as the center of historical epistemology thereby opposing a straight forward
history of ideas as well as the history of scientific theories.3 In a similar vein,
Blumenberg pushed the history of ideas towards and beyond its limitations in his
methodological program of metaphorology.4
Metaphorology and conceptual history are the programs they followed to trace and
explore the unfolding of problems through their historical transformation. This project
gains its weight not just as a delineation of the shifting meanings, understandings, or
definitions and as a reconstruction of the different theories as they materialize in
historically specific conceptualizations, but also, and furthermore, as the charting of an
underlying conceptual field which transpires only in the process of articulation,
transformation and revision. The metamorphing process itself, and in its endless
entirety, covers or captures something which not only transgresses each individual step
but which lies outside of the epistemological progress. The conceptual steps may follow
a logic of discovery, theoretical revision or paradigmatic revolution, but the conceptual
field, as it is demarcated by the very process of transformation and revision, renders the
space of an underlying problem that escapes the grasp of the theories.
Here, a few remarks about the historical situation of both projects may be in place
for discussing the significance of their historical epistemologies in the political context
of the twentieth century. Blumenberg, sixteen years younger than Canguilhem, had been
denied the honor of delivering the valediction speech when he graduated in 1938 as the
schools and the provinces best pupil, because of the Jewish descent of his mother.
Soon after, he had to stop his academic training in Nazi Germany at a small catholic
university because of anti-Semitism. He was later first forced into a labor camp and
after escaping this into living a life in hiding. Blumenberg always thought of these
years in jeopardy as lost and stolen time for which he had to make good by an extreme
concentration on reading and writing. Later in his life, Blumenberg freely choose a
secluded life almost as withdrawn from the outside world as during his survival of the
Nazi period.
Beyond the personality pattern which may become discernible here, it appears that
the biographical background is important for contextualizing philosophically his
engagement in the debate about secularization, where he argued ardently for the
2
Hans Blumenberg, Nachdenklichkeit, Deutsche Akademie fr Sprache und Dichtung 1980, pp. 5761, on p. 61.
Georges Canguilhem, Lhistoire des sciences dans loeuvre pistmologique de Gaston Bachelard,
in: id., Etudes dhistoire et de philosophie des sciences, Paris: Vrin 1994, pp. 173-186, on p. 184
(une histoire des filiations conceptuelles, mais cette filiation a un statut de discontinuit).
Hans Blumenberg, Paradigmen zu einer Metaphorologie, kommentiert von Anselm Haverkamp,
Frankfurt am Main: Suhrkamp, 2007.
172
legitimacy of the contemporary.5 Where others regarded the catastrophe of Nazism in

particular and the massive problems of modernism in general as a clear and lamentable
sign of the loss of a religious orientation by the society, Blumenberg defended
secularization stoutly. According to him, the contemporary could not, and should not,
be conceived of as a derivative stage in relation to a previous more harmonious or even
golden age of religious order. For him, the contemporary was, and continued to be, a
proper period in its own right, and named precisely because of secular modernity. He
would never call this general assumption into question, although he later gradually
developed a more skeptical and perhaps even melancholic outlook on the human
condition. Such a fundamental philosophical position certainly maps onto Blumenbergs
biographical path but it has also to be put into context within the German post-WWII
debates and the revival of more conservative agendas where Blumenberg took a
decisively modernist position, especially in his extended discussion with Karl Lwith.
The larger political circumstances and developments also crossed through
Canguilhems biographical and intellectual path. As is well known, Canguilhem
ventured into medicine after his philosophical studies and while working as a school
teacher in Toulouse. And he submitted his Essai as a medical dissertation to the
University of Strasbourg when he was already a professor of philosophy there, and
while the university was transferred to Clermont-Ferrand because of the war. Studying
medicine, he declared in the Essai, did not result from a wish to gain knowledge in
another specialty, or with the aim to engage in the history of medicine, or to offer
medical philosophical advice, but for genuinely philosophical reasons to confront
philosophy with concrete problems and to study a domain of technology.6
Before he could follow this agenda further, however, Canguilhems life was also
shaped by the larger political circumstances. Even Clermont-Ferrand was not a safe
place during the German occupation and Canguilhem narrowly escaped the Gestapo
who soon after captured and killed his friend Jean Cavaills. Confronted with this
escalation, Canguilhem became an active member of the resistance and served as a
physician, returning to his academic position only after the defeat of the Germans and
the return of the university to Strasbourg.
Already by its title, the Essai sur quelques problmes concernant le normal et le
pathological indicates how Canguilhem transformed philosophy into a genealogy of
concepts. The book is not so much a historical analysis of the different theories of the
pathological, but an exploration of the relation between the normal and the pathological.
At the core of medicine lies the dynamic polarity between the normal and the
pathological, because the pathological cannot be characterized by finite parameters of
deviation but only by the activity of the normal organism to oppose it. According to
Canguilhem, medicine is so valuable for philosophy not so much because of its
epistemic content but because it is the knowledge domain where the intrinsic
normativity of life itself becomes transparent and epistemically accessible. This
5
Hans Blumenberg, Die Legitimitt der Neuzeit, Frankfurt am Main: Suhrkamp, 1966 (translated by
Robert M. Wallace as The legitimacy of modern age, Cambridge, MA: The MIT Press, 1983).
See Canguilhem, Normal and Pathological, p. 33f.
173
Borck
rehearsal of Canguilhems argument hardly seems relevant or necessary in the context

of our conference, but it is here, I want to argue, where we have to situate Canguilhems
epistemological vitalism and the specific (or peculiar) conceptual continuity between
the primary activity of living systems managing their homoeostasis by interacting
with and structuring their environment and the normativity of abstract structures such
as concepts and theories. From this perspective, concepts may get defined as solutions
to scientific problems or questions but they operate as strategic moves in dynamic
power relations and hence have a history of their own.
Historical epistemology explores the trajectories of these relations and polarities as
they meander through history, transform the perspectives on the matter they relate to
and thus structure underlying problems. Later, and in relation to Bachelards
epistemology of ruptures and transformations, Canguilhem describes this task as the
reconstruction of the history of thinking and reflection in the true sense.7 It is
important to note in this context that such a history does recognize radical paradigm
shifts and revolutions in history, but does not stop there. In contrast, historical
epistemology de-centers the history of science from the history of theories (they may
change radically, causing all sorts of problems of incommensurability as Thomas Kuhn
has alerted the history of science about). But genealogy of concepts reorients the history
of science towards more fundamental constellations of problems which get articulated
in new ways by scientific revolutions yet remain recognizable. Indeed, such problems
become very much recognizable and traceable in their dimensions and implications in
the sequence of such re-articulations and transformations.
Blumenberg shared this critical distance towards the theory of paradigm shifts and he
did so for similar reasons. In a slightly ironic, brief critique, he ridiculed Kuhns choice
of words as confusing scientific theories with linguistic schemes.8 The conceptual
historian Blumenberg knew too well and could hence not ignore that a paradigm was
the example to establish a declension schema (had he known the inflationary usage of
paradigm that was to follow Kuhn!).
The reason behind this terminological nitpicking was less a disagreement with the
concept of scientific revolutions. As a philosopher, Blumenberg was not generally
interested in the history of various sciences, but especially in the major revolutions and
how they changed the dynamic polarity between philosophy and the sciences; indeed,
his Genesis of the Copernican World was a reconstruction of the scientific revolution
and an analysis of the epistemic transformations enabling it.9 But in this reconstruction,
as in his philosophy more generally, Blumenberg aimed at an exploration of those kind
of fundamental questions to which scientific theories always provide only partial
answers. The detection and identification of an incommensurability always already
7
Canguilhem, Lhistoire des sciences de Gaston Bachelard, p. 177 (lhistoire des sciences dtre
authentiquement une histoire de la pense).
Hans Blumenberg, Paradigma grammatisch, in: id., Wirklichkeiten in denen wir leben, Stuttgart:
Reclam, 1999, pp. 157-162.
Hans Blumenberg: Die Genesis der kopernikanischen Welt, Frankfurt am Main: Suhrkamp, 1975
(translated by Robert M. Wallace as The Genesis of the Copernican World, Cambridge, MA: MIT
Press, 1987).
174
required a metatheoretical position for a comparative perspective. He agreed with Kuhn

that such a position could not be reconstructed from within each of the
incommensurable theories, but he disagreed with Kuhn that its recognition was the
philosophers last word. He demonstrated this conviction in his later massive tomes like
his Work on Myth or the Hhlenausgnge [Cave Exits] by extending the history of ideas
to a reconstruction of the human predicament through all its articulations or
Umbesetzungen, as he coined an untranslatable term somewhere between repositioning,
re-alignment and re-valuation.10
This was Blumenbergs version of a history of thinking and reflection in the true
sense, for which he widened and transformed the history of ideas and conceptual
history into a rendering of the conceptual field or space as it escapes the grasp by proper
concepts and theories. And here Blumenbergs big books reconnected with his earlier
interest in metaphors, which he regarded not as premature, preliminary or otherwise
inadequate and derivative concepts, but as notions providing a glimpse upon a
problems dimensions. However, I do not regard Blumenbergs metaphorology as anticonceptual but rather as an opening and extension of conceptual history into its nonconceptual foundations somewhat similar to Canguilhems interest in practices or
institutions and his trans-disciplinary reconstruction of meanings.
In contrast to a history of concepts and philosophy of science as the reconstruction of
a reached discursive consensus or epistemic progress, Blumenberg like Canguilhem
reaches out for exactly the opposite i.e. the rendering of the historical articulation of
problem zones as they manifest along the metamorphosis of truth claims, the failures of
established solution, the dead-ends of conceptual labor. Blumenberg and Canguilhem
share an interest in forgotten detours from the established trajectory of the canonical
tradition, not for the sake of erudite scholarship, but as heightened attention for failures
and breakdowns. Foucault famously declared Canguilhem the philosopher of error, and
Blumenberg explained Shipwreck with Spectator to be the Paradigm of a Metaphor for
Existence.11 Failure is the access to the non-conceptual.
Failure is typically the domain of practical application and of technology. Hence it
seems as if their shared interest in technology is also just a variation on the same theme.
However, technology establishes itself by its successes, and both failure as well as
success point to the intrinsic originality of technology as it lies outside of the realm of
theoretical determination. Since his early study on Descartes and technology,12
Canguilhem understood technology as a dynamic productivity pushing theoretical
science forward by its failures and successes. Surprisingly similar, Blumenberg situates
a radical departure from the ancient conceptualization of ars with the early modern
10
11
12
Hans Blumenberg, Arbeit am Mythos, Frankfurt am Main: Suhrkamp: 1979, and id., Hhlenausgnge,
Frankfurt am Main: Suhrkamp, 1989.
Hans Blumenberg, Schiffbruch mit Zuschauer. Paradigma einer Daseinsmethapher, Frankfurt am
Main: Suhrkamp, 1979.
Georges Canguilhem, Descartes et la technique, Travaux de IXe Congrs international de
philosophie (Congrs Descartes), vol. 2, Paris: Hermann, 1937, pp.77-85.
175
Borck
notion of technology.13 Whereas knowledge, science and technique remained in a state

of mutual correspondence throughout antiquity, the modern age understood their
relationship as being in a state of constant tension since the primacy of technological
innovation went beyond the horizon of theoretical knowledge.
Technology always already offers a solution; even in case of the inexistence of the
appropriate recognition of a scientific problem. For this reason technology cannot be
conceived of as a mere application of theoretical knowledge. This powerful illusion,
lingering on from antiquity, had to be turned upside down as the modern age started to
recognize when the technological progress set in. Science and theoretical knowledge are
secondary to practical curiosity and technical inventiveness. And hence, any
demonization of technology as natures opposite, so fashionable at the dawn of the
atomic age, was met with critical distance by both of them, for once, because this stance
was obviously totally ignorant of the involvement of technology in scientific research
and thus the very nature of the contemporary understanding of nature.
For Blumenberg, however, the recognition of the true nature of technology did not
only speed up the process of technological evolution and industrialization, but marked a
moment of spontaneity and radical contingency in history where human ingenuity and
human history departed from natures path. The Legitimacy of the Modern Age, so one
could follow Blumenbergs argument, very much resided in the recognition of this
radical contingency. And in this respect, the very departure from nature brought human
beings closer to an understanding of their eccentric positionality (Helmuth Plessner).
The invention of technology per se with the modern age thus signified a major, and
for Blumenberg most significant, epistemic shift, because it liberated intellectual life to
independence The invention of technology was also for Canguilhem a diversely
important event in human history, but he enlisted its epistemic surplus in almost
opposite terms in comparison to Blumenberg. Technology is not so much the result of
human inventiveness or a consequence of human needs, but the practical as well as
theoretical appropriation of a fundamental activity of life itself because technology is an
activity rooted in the living beings spontaneous effort to dominate the environment
and to organize it according to his values as a living being.14 Where Blumenberg saw a
continuous process of theoretical curiosity, which is released to its full potential in the
invention of technology, Canguilhem recognized quasi-technological interactions and
appropriations across the entire plane of living phenomena where intellectual activity
simply followed on. As a radical consequence, he turned biology ultimately into a
13
14
Hans Blumenberg, Lebenswelt und Technisierung unter Aspekten der Phnomenologie, in: id.,
Wirklichkeiten in denen wir leben, Stuttgart: Reclam, 1981, pp. 7-54. Here, I follow Mllers analysis
of Blumenbergs philosophy of technology: Oliver Mller, Natur und Technik als falsche Antithese.
Die Technikphilosophie Hans Blumenbergs und die Struktur der Technisierung, Philosophisches
Jahrbuch 115 (2008): 99-124.
Canguilhem, The Normal and the Pathological, p. 228f.
176
discipline of philosophical reflection because it clarified the philosophical position of

the problem of the means of knowledge.15
One should be careful not to mistake this position for a biological theory of cognition
or an evolutionary theory of knowledge. Canguilhem is neither arguing that existent
knowledge claims have survived because of their epistemic superiority over
extinguished ones, nor that neurophysiological processes can explain thinking and
reflection, but he insists that life itself is a structure-setting activity for imposing
supportive conditions upon the environment. In a certain respect Canguilhem and
Blumenberg both research into accidents which they regard as uniquely productive
sources for epistemological and fundamental philosophical inquiries because of the
vivid originality of technological innovation, but they search in opposite directions.
This almost antithetical symmetry results in important divergences between the two.
For Blumenberg, technology as essentially a non-theoretical endeavor is driven by
merely technical imperatives such as rationalization and efficiency and thus offers
human society ever new means of relief which eventually threaten to replace the human
lifeworld by tools and artificial acceleration.16 From this perspective philosophy of
technology gradually turns into risk assessment which is not meant as a critique but to
stress the difference to Canguilhem. In his view, philosophy and historical epistemology
reflect upon how science and technology function and structurally they thereby
resemble, mimic and extend the processes of life. Where Blumenberg diagnosed the risk
of excessive demands by technology, Canguilhem saw productive alienation from the
seemingly natural by technology. The creation of the artificial, which is so decisive for
the modern condition in science and technology, arrives at self-reference and reflexivity
in the modern life sciences with their myriads of hybrid, naturally artificial and
artificially natural objects:
We have to acknowledge the cooperation with the simulation machines for the
functions of life, we will have to accept their integration in our lifeworld if we want to
hold on to our desire to know better, what life is and of what it consists. Only now
transpires how hard human beings must try to defamiliarize the nave objects of their
vital questions in order to gain knowledge and continue science.17
What may be read as a standard lament about the departure of the natural sciences
from common sense into the vaults of complexity and artificiality entails, in fact, the
renewal of the core idea of historical epistemology vis--vis contemporary science.
15
16
17
Georges Canguilhem, LExprimentation en biologie animale, in: id., La Connaissance de la vie,

Paris: Vrin 1965, pp.17-39, on p. 39 (On peut penser que la biologie est aujourdhui une science de
caractre dcisif pour la position philosophique du problme des moyens de la connaissance.).
Hans Blumenberg, Lebenswelt und Technisierung unter den Aspekten der Phnomenologie.
Georges Canguilhem, Sur lhistorie des sciences de la vie depuis Darwin, in: id., Idologie et
rationalit dans lhistoire des sciences, Paris: Vrin, 1977, pp. 101-119, on p. 119 ([La connaissance
de la vie dpend dsormais des nouveaux automates. Ils sont ses modles, ses instruments, ses
dlgus.] Il faut accepter la collaboration de ces simulateurs des fonctions du vivant humain, il faut
dsormais accepter de vivre en leur compagnie, si lont tient mieux savoir en quoi consiste vivre.
Jamais il na t ce point manifeste combien lhomme doit travailler se rendre trangers les objets
nafs de ses questions vitales pour mriter la science.).
177
Borck
Where the critical history of science defamiliarizes the traditional trajectories of

development and evolution by the simple but strategic methodological move to
recognize as its object not the objects of a particular science but the very historicity of
this scientific discourse, the relation to contemporary science must equally be one of
critical distance, a form of dis-understanding that takes scientific discourse seriously in
its epistemic content, artificiality and effectivity in order to detect its normative forces
together with the ruptures, the moments of failure and breakdown where crises and truth
events interrupt or protract: The historian of science is by necessity a historiographer
of truth, tracing moments of cogitation which reflect backwards on the history of
thinking and experience.18
The conscious presentism of historical epistemology does not measure the past
against current understanding but inverts the progress of scientific theorizing to a
tableau of theoretical as well as experimental explorations bearing the traces of truth
events. Its progress coalesces around the productivity of failures, interruptions and
disturbances in form of an abstracted and materialized sensorium.
Blumenberg took the side of rewriting the history of science; criticizing the prevalent
alliance with theoretical progress and recommending instead the methodological tools
of distancing as developed in metaphorology and ethnography.19 His guiding
perspective for such a fundamental philosophy of science and history of reflection was a
radical skepticism against any form of teleology. Philosophy and history had to abstain
from any calculations about a dividend in a brighter future; lessons from history cannot
pave the way for the future, but their primary function is to pay tribute to the past by
keeping its memory alive the endless task to work against its disappearance.20
He cultivated this agenda by developing and permanently refining a sophisticated
system of record keeping, note making, indexing and re-constellation. The perfection of
his machine-like system is demonstrated by the sequence of posthumous publications
from the vaults of his archives. His constant stream of messages will soon be met by
posthumous eloquence from the other side of the Rhine with the new edition of
Canguilhems writings. There was no dialogue across the river, even though their ways
of thinking unfolded along vital resonances revealing kindred forms of inquiry. Among
their shared concerns was that Canguilhems historical epistemology and Blumenbergs
philosophical history both call for scrutiny a scrutiny that starts in this case with the
18
19
20
Georges Canguilhem, Rle de lpistmologie dans lhistoriographie scientifique contemporaine,

in: id., Idologie et rationalit dans lhistorie des sciences, pp. 11-29.
Hans Blumenberg, Ernst Cassirer gedenkend, in: id., Wirklichkeiten in denen wir leben, pp. 163172, on p. 169f (In der Ethnographie ist seit langem nicht mehr mglich, was in der
Wissenschaftsgeschichte noch gang und gbe ist: den raum-zeitlichen Standpunkt des Betrachters
zum Bezugspunkt seiner Faktenwahl und seiner Urteile zu machen. ... Die Wissenschaftsgeschichte
ist fast immer dem Spott derer, die recht behalten haben, gefolgt. ... Die Wissenschaftsgeschichte, mit
ihrem Gegenstand dem Erfolg verbunden, hat erkennbar andere Mhe, den obskur Gewordenen
Respekt zu erweisen.).
Ibid., p.170 ([Lernen aus der Geschichte ist] sekundr gegenber der elementaren Obligation,
Menschliches nicht verloren zu geben, ... denn, dass es kein Ziel der Geschichte gibt, bewahrt uns
davor, auf ein solches Ziel hin vorlufig zu bleiben und aufgefordert zu werden, ihm als Mittel
dienstbar zu sein.).
178
respect for the very absence of a dialogue. It is the lack of communication which may
serve us as truth event for a historical epistemology of the contemporary.
179
ARCHOLOGIE DES CONCEPTS ET PHILOSOPHIE DE LA NATURE
Andrea Cavazzini
Lide mme dune pistmologie historique implique la possibilit dindiquer des
niveaux dtermins dhistoricit lintrieur de ce qui se prsente comme le domaine,
ou les domaines, de la connaissance scientifique une poque donne. On peut
supposer que lhistoricit propre quil sagit de saisir dans une telle entreprise historicopistmologique soit lhistoricit des concepts.
Par l, une dmarcation est pose par rapport dautres dmarches philosophiques
visant analyser la connaissance scientifique. Par exemple, la position orthodoxe du
positivisme logique consiste attribuer un rle crucial la notion de thorie en
entendant par thorie un systme formel axiomatisable lintrieur duquel au moins
deux hirarchies sont tablies : dune part, celle entre les noncs poss comme axiomes
et les noncs quon peut tirer des premiers suivant des rgles dinfrence ; de lautre
celle entre un langage observationnel et un langage thorique, les deux tant mis en
correspondance par des rgles explicites. Selon le rsum par B. Van Fraassen du point
de vue orthodoxe de lpistmologie no-positiviste, une thorie tait conue comme
une thorie axiomatique, cest--dire comme un ensemble de phrases () les axiomes
de la thorie do un autre ensemble de phrases plus vaste les thormes peut tre
tir dune manire purement formelle ; ensuite, une distinction est introduite entre les
thories scientifiques et les interprtations qui leur sont associes et qui relient les
termes de la thorie avec le domaine de rfrence vis .1
Cette vision de la thorie rduit la connaissance une pure dimension syntaxique,
qui peut tre la limite vide de tout contenu, et quil sagit de combiner aprs-coup avec
un contenu rel rduit, lui, des sense-data atomistiques (ou plus prcisment : aux
transcriptions nonciatives de ceux-ci). Du coup, toute paisseur historique est limine
dans lun comme dans lautre de ces deux cts de la connaissance scientifique, lun
tant une structure sans contenu thmatique, lautre un contenu brut dpourvu de
structure, donc impossible articuler selon une intelligibilit propre.
Lapproche que Van Fraassen propose, repose sur la notion de modle, qui peut tre
utilement rapproche de ce que nous entendons par concepts : selon cette approche, une
thorie consiste construire des entits abstraites, pourvues dun contenu dtermin,
dans lesquelles il sagira de plonger des rgularits empiriques, elles aussi dailleurs
dfinies par une mise en forme modlisante.2 Toute dichotomie entre le domaine
thorique et celui des observations est rcuse ds lors, les modles abstraits et les
modles empiriques se posant rciproquement des contraintes, et demandant tant les uns
que les autres la fois la spcification dun contenu et la mise en forme des relations
1
2
Bas C. Van Fraassen, Lois et Symtrie, Paris : Vrin, 1994, p. 339.

Ibid., pp. 337-338.
181
Cavazzini
thoriques que ces deux sries dentits incorporent au mme titre. Leffet de
connaissance de la thorie consiste donc dans lincorporation, et lajustement, entre les
rgularits empiriques conceptualises et les entits thoriques que sont les modles
abstraits : cest par lhomologie entre ces deux sries doprations et de construits que
la connaissance merge.3
Le contenu conceptuel du savoir scientifique, avec ses dterminits thmatiques, est
donc impossible contourner pour lanalyse de ce jeu dadquations et homologies
rciproques. Du coup, lhistoire redevient pensable au niveau de ces dterminits :
contrairement aux syntaxes et aux donnes observationnelles soi-disant pures , la
construction de modles thoriques ou empiriques (cette distinction nest valable
dailleurs que selon une logique du plus-ou-moins-de-thorie-et-d-empirie, les deux
sries tant toujours des mixtes de contraintes relles et didalisation) implique une
stratification historique profonde des problmes, des rapports entre champs scientifiques
diffrents, des interfrences avec les diffrentes formes de la culture, tout cela formant
les prsupposs de la conceptualisation qui reprsente lassise de la construction des
modles et de la modlisation de lempirique.
Si nous nous sommes permis cette digression sur Van Fraassen, qui travaille partir
dune rflexion sur la physique, ce nest que pour mieux marquer la spcificit de la
dmarche canguilhemienne lintrieur dun champ pistmologique oppos
lorthodoxie no-positiviste. Georges Canguilhem a insist avec beaucoup de force sur
le caractre crucial de la dimension propre au concept : cest celui-ci quil reviendrait
de jouer le rle du pivot de ltude du devenir de la connaissance et de la pense
scientifique.
De toute vidence, le champ scientifique que Canguilhem a creus, celui des sciences
de la vie, contribue lgitimer et tester cette position. Le vivant ne se laisse interroger
que par la construction de modles conceptuels, ncessairement qualitatifs, dont la
fonction consiste restituer une ide gnrale des proprits spcifiques de la ralit
vivante. Par exemple, la notion mme de loi a toujours t hautement problmatique en
biologie, et les structures mathmatiques de symtrie et dinvariance, qui jouent un rle
dcisif dans la constitution de lobjet physico-mathmatique, ne peuvent gure puiser
lindividualit irrductible des organismes, dont la forme et les proprits ne peuvent
faire lobjet que dune saisie globale. Si nous considrons les objets physicomathmatiques comme des entits gnriques, dpourvues dindividualit concrte mais
au contraire dfinies entirement par les structures thoriques sous-jacentes, on peut au
contraire considrer les objets biologiques comme des tres rels et pr-donns, dont
lautonomie et lauto-suffisance posent des problmes assez diffrents la construction
thorique. Le vivant se donne comme une ralit archaque qui, sans pouvoir pour
autant tre interroge sans le recours des constructions conceptuels, chappe cependant
toute modlisation dtermine et se prsente comme lhorizon dun inpuisable travail
3
Une vision analogue de la connaissance scientifique est celle expose par Lon Brunschvicg dans
Lexprience humaine et la causalit physique, Paris : Alcan, 1922 ; cfr. aussi Andrea Cavazzini,
Archeologia della fisica I & II , dans id., Il vivente, lanalogia, le scienze. Studi di archeologia dei
saperi, Milan : Mimesis, 2007.
182
dapproximation. Si cela est vrai dans une certaine mesure pour toute conceptualisation
scientifique limage de la ralit archaque est utilise par Alain Connes se
rfrant aux mathmatiques ! la spcificit biologique de cette ralit impose une
dialectique singulire aux relations entre concept et ralit.
Dabord, les modles conceptuels biologiques qui oprent cette saisie dun rel
archaque gardent un lien structurel des mtaphores, des images, des discours et des
pratiques extra-scientifiques. Les sciences de la vie sont beaucoup moins capables que
la physique mathmatique ou les mathmatiques disoler leurs dispositifs thoriques
vis--vis de lensemble des formes de la vie sociale, des discours qui circulent en elles
et aussi des fantasmes idologiques qui les traversent. Cest pourquoi une archologie
des concepts est possible, qui rattacherait les concepts et les procdures les mieux
purs des systmes souterrains de cohrences caches qui tmoignent de lexcs
ineffaable de la significativit dun modle par rapport sa valeur dobjectivation de
lexprience et de vrification des connaissances la prgnance dun concept excde
toujours ses usages vrifis et rgls au sein dun domaine disciplinaire (encore une
fois, cela est vrai de toute dmarche scientifique, des diffrences de degr prs). Cet
excs reprsente par consquent la surdtermination et limprvisibilit de lhistoricit
propre aux modles conceptuels, dont la biologie du fait que son objet ne parvient pas
se constituer en schma mathmatique auto-suffisant fournit les cas de figure les
plus saisissants.
Autrement dit, ds que lhistoricit est rintroduite dans ltude de la connaissance
scientifique ce qui est bien entendu possible par le biais de lapproche par modles il
devient impossible dviter une certaine dispersion des diffrents axes de
lhistoricisation, des diffrents niveaux de pertinence qui dfinissent lobjet de ltude et
les points de vue par rapport auxquels il est accessible. Lhistoricit des sciences est
toujours plus, ou autre chose, que lhistoricit des sciences comme Auguste Comte
lavait dit (et Althusser lavait raffirm), la formation des concepts qui orientent la
rflexion thorique et rendent comprhensible lempirique, implique lhistoire entire de
la civilisation, des idologies, des techniques, de toute lencyclopdie des savoirs, sans
oublier les savoirs prsums, conjecturaux, faux ou tout simplement ces discours
dont la prgnance dpasse la positivit de la vrification.
Cette position, pour laquelle lhistoire des sciences nest pas que lhistoire des seules
sciences, suppose videmment lide que les sciences elles-mmes soient toujours plus,
ou autre chose, que des sciences, cest--dire, des dispositifs de production, organisation
et transmission de savoirs vrifis et de procdures de vrification.
Cependant, ce serait une erreur de rduire ces positions communes des auteurs
tels que Canguilhem, Comte, Foucault et Althusser, entre autres une simple
rsorption de lanalyse de la connaissance scientifique par la sociologie ou ltude de
schmes culturels contingents. Si les sciences sont toujours autre chose quelles-mmes,
si leur structure excde toujours la simple connaissance vrifie, et si cet excs dfinit
prcisment leur historicit, il ne faudra pas oublier que laltrit qui travaille et traverse
lesprit scientifique nest pas uniquement celle que reprsente la contingence de ses
commencements, mais et cela depuis lpoque moderne et sa Rvolution
183
Cavazzini
scientifique galement celle quindique le mot nature : savoir, lautonomie et le

dynamisme dun systme, ou dun systme de systmes, rel, pr-thorique, mais qui
doit nanmoins possder une certaine intelligibilit propre afin que les modles
conceptuels puissent, non seulement en restituer quelque chose, mais galement tre
valus laune de leur capacit restituer des proprits desdits systmes. Ce qui pose
le problme des rapports entre le devenir historique des concepts en tant que porteurs de
lintelligibilit du rel et le rel lui-mme en tant que critre de la capacit des concepts
fournir lintelligence de la chose elle-mme.
Larticle de Canguilhem intitul Le concept et la vie reprsente un cas
paradigmatique de cette dialectique entre lhistoire et la nature dans la construction des
modles conceptuels en lesquels consiste la connaissance. Dans lhistoire esquisse par
ce texte, allant dAristote la biologie molculaire, travers Kant, Hegel, Leibniz,
Bergson, lhistoricit des concepts visant la saisie du vivant ne relve pas, pour
Canguilhem, que de la dispersion des cohrences archologiques : la succession des
modles de la connaissance biologique est traverse galement par un telos dtermin
consistant, justement, dans la saisie de la singularit du vivant. Finalement, ce quoi se
rfrent, et tendent, les diffrents concepts que lhistoire propose comme oprateurs de
la connaissance de la vie, cest bien la spcificit du vivant dans ce quil a
dirrductible. Mais cette irrductibilit relve justement du fait que le vivant est son
propre critre dintelligibilit : le modle adquat du vivant nest finalement que le
vivant lui-mme, dont lintelligibilit concide avec le dploiement de sa propre
existence. Autrement dit, si la vie consiste dans un pouvoir de se donner des normes, et
si les normes immanentes la vie font galement partie de lintelligibilit de la vie,
force est de conclure que les modles conceptuels du vivant pourront et devront tre
jugs laune de cette intelligibilit immanente.
Lhistoricit des concepts donc ne relve pas que de la contingence historique et de
la surdtermination par les pratiques et les discours qui traversent un contexte social
donn : il existe aussi une historicit tlologique, oriente, qui vise ladquation des
concepts, des mthodes, des thories, lintelligibilit intrinsque de la vie en tant que
porteuse des conditions de possibilit de sa propre connaissance. Cette tension interne
lpistmologie historique de Canguilhem, qui caractrise notamment de textes tels que
Le Concept et la vie , se fonde sur la circularit entre les deux modes de lhistoricit :
la validit dun modle conceptuel dpend de sa capacit reprsenter une
approximation convenable des modes dtre du vivant ; mais ces modes dtre du vivant
ne peuvent dployer leur signification que par la saisie quoprent des modles
historiquement dtermins. Le concept suppose la vie comme le prsuppos de
lintelligibilit quil effectue ; et la vie suppose le concept comme la condition de
lintelligibilit dont elle-mme contient les critres.
La dialectique entre le concept et la vie est formule par Canguilhem dans le texte
cit en renvoyant Hegel, et la thorie de la vie comme tre-l immdiat et implicite
du concept : la vie, dit Hegel, est lunit immdiate du concept sa ralit, sans que ce
184
concept sy distingue .4 Pour Hegel, la vie est un auto-mouvement de ralisation

selon un triple processus, et ici Hegel ne fait en somme que reprendre les analyses de
Kant dans la Critique du jugement tlologique. Ce triple processus est : la structuration
de lindividu lui-mme ; son auto-conservation lgard de la nature inorganique ; la
conservation de lespce .5 Or, ce en quoi Hegel se spare de Kant porte justement sur
le statut de la construction des concepts par rapport la ralit pr-thorique et printellectuelle de la vie. Ici, on ne peut viter une longue citation : Mais si nous avons
gagn [avec Kant] la lgitimation dune possibilit, celle de la connaissance par
concepts, naurions-nous pas perdu la certitude que, parmi les objets de la connaissance,
il sen trouve dont lexistence est la ncessaire manifestation de la ralit de concepts
concrtement actifs ? Autrement dit, naurions-nous pas perdu la certitude que, parmi
les objets de la connaissance, se trouvent en fait des tres vivants ? () La logique
transcendantale ne parvient pas, dans sa constitution, a priori de la nature comme
systme de lois physiques, constituer en fait la nature comme le thtre des
organismes vivants. Nous comprenons mieux les recherches du naturaliste, mais nous
ne comprenons pas les dmarches de la nature .6
Difficile dexprimer plus efficacement linsuffisance dune pense de lautonomie
des modles conceptuels pour saisir ce en quoi ces modles trouvent finalement une
adquation des dynamismes rels. Le vivant est lui-mme le premier et dernier modle
de soi-mme, tous les concepts biologiques historiquement proposs et ventuellement
valids ne trouvant de site deffectuation qu lintrieur de ce premier et dernier terme :
Un tre organis est un tre qui est la fois cause et effet de lui-mme, qui sorganise
et qui reproduit son organisation, qui se forme et qui se donne la rplique,
conformment un type, et dont la structure tlologique, o les parties sont en rapport
entre elles sous le contrle du tout, tmoigne de la causalit non-mcanique du
concept .7 Autrement dit, les diffrents modles du vivant que lhistoire de la pense
prsente et dont lhistoire archologique peut mettre en vidence les solidarits
caches, telle celle entre Aristote, Hegel et la biologie molculaire doivent se
comprendre laune de cette dialectique entre la comprhension du vivant que chaque
modle permet et cette autre comprhension que le vivant engendre du simple fait de cet
oprer en quoi consiste son tre.
Dvelopper cette dialectique pourrait permettre dcrire une autre histoire de
lpistmologie historique la franaise , den apprcier diffremment les liens
linspiration de la Naturphilosophie idaliste et romantique (ventuellement filtre par
la tradition des Ravaisson, Lachelier, Boutroux et Bergson), et peut-tre dimpulser une
nouvelle approche des sciences contemporaines, en particulier de la biologie, o la
question de lintelligibilit des objets pistmiques croise la problmatique du
dynamisme immanent la nature . Mais, peut-tre, cela reprsenterait un
dpassement de lpistmologie historique et de la philosophie des sciences en tant que
4
5
6
7
Georges Canguilhem, Le Concept et la vie , dans id., Etudes dhistoire et des philosophie des
sciences, Paris : Vrin, 1970, p. 345.
Ibid.
Ibid., p. 344.
Ibid., p. 344-345.
185
Cavazzini
telles, en direction de quelque chose qui na pas de nom, si ce nest celui de philosophie
tout court.
186
HISTOIRE(S) DE LA VIE DE CANGUILHEM FOUCAULT
Maria Muhle
Dans ce texte, je voudrais interroger les rapports respectifs tablis par Georges
Canguilhem et Michel Foucault entre la vie et le social, voire entre la vie et le pouvoir.
Pour commencer je voudrais rappeler que Foucault ne propose pas de dfinition de la
vie au contraire, il se maintient dans une position purement analytique ou
archologique face la vie qui est, pour lui, uniquement le corrlat des techniques de
savoir et de pouvoir. Canguilhem, au contraire, combine ses tudes des sciences de la
vie avec lintroduction dune dfinition de la vie en termes normatifs, en rupture la
fois avec le paradigme mcaniste incapable, selon lui, de saisir pleinement ce quest la
vie et avec lexplication vitaliste traditionnelle du pouvoir de vivre par un principe
vital. Il aborde la question du rapport entre le social et vital dans la deuxime partie de
la rdition du Normal et le Pathologique, sous le titre des Nouvelles Rflexions
concernant le normal et le pathologique crites 20 ans aprs, donc de 1963 1966, et
rdiges, entre autre, sous limpression de lide dune norme construite par lordre
social et porteuse de normalisation avance par Foucault en opposition la
conception canguilhemienne dune norme produite par la vie .1
Cest ainsi quElisabeth Roudinesco rsume lopposition intellectuelle entre
Canguilhem et Foucault en insistant sur le renversement dans le rapport de matre
disciple qui sopre lors de la publication de la Naissance de la clinique. Ainsi, dans son
fameux texte Un nouveau concept en pathologie: lerreur, Canguilhem attribue
Foucault le mrite davoir montr, en des pages admirables, mouvantes, de la
Naissance de la clinique [] comment Bichat a fait pivoter le regard mdical sur luimme pour demander la mort compte de la vie 2 un regard mdical donc qui
reprsente linstance de normalisation sociale laquelle sont soumis les phnomnes
vitaux, et notamment les phnomnes vitaux anormaux quil sagit, autour de 1800, la
fois de matriser et de fonctionnaliser pour une meilleure connaissance et un meilleur
contrle de la vie. Cest donc partir de ce constat de limportance de la mort pour la
vie que Canguilhem revient sur ses thses du premier essai de 1943 pour les relire la
lumire des thories foucaldiennes, donc la lumire dune interrogation sur les normes
sociales. Et il esquisse une nouvelle science du normal fonde sur un dplacement
qui cimente la priorit de la pathologie sur la normalit. Roudinesco situe dans ce
dplacement, cest--dire dans la reconnaissance de la productivit de la pathologie ou
de lerreur des valeurs vitales ngatives , la rupture de Canguilhem avec lidal
vitaliste de sa jeunesse , un idal vitaliste qui ne tolrerait aucune pathologie.

1
Elisabeth Roudinesco, Georges Canguilhem, de la mdecine la rsistance : Destin du concept de

normalit , in : Franois Bing, Jean-Franois Braunstein et Elisabeth Roudinesco (ds.), Actualit de
Georges Canguilhem, Paris 1998, p. 39.
Georges Canguilhem, Le normal et le pathologique, Paris 92003, p. 215.
187

Muhle
Je voudrais dans ce qui suit revenir sur cette thse de la rupture: Il est bien connu que
Canguilhem rompt avec une conception classique du vitalisme, chose qui dailleurs
sannonce ds son texte sur les Aspects du vitalisme publi dans La connaissance de
la vie bien avant les Nouvelles Rflexions.3 Ici, Canguilhem constate que lopposition
Mcanisme et Vitalisme, Prformation et pigense est transcende par la vie ellemme se prolongeant en thorie de la vie .4 Cest cette thorie de la vie que
Canguilhem va dsigner comme un vitalisme qui est lexpression de la confiance du
vivant dans la vie ,5 au-del de toute supposition de principe vital ou entlchie, qui, en
dernire instance, reprsente le retour du paradigme mcaniste dans les sciences de la
vie et conoivent le vivant comme exception aux lois du milieu physique. Le vitalisme,
selon Canguilhem, est suscit par la vie, est plus spcifiquement par sa spontanit
propre , par ce que Claude Bernard exprimait en disant: la vie cest la cration. 6
Au lieu dune rupture avec le vitalisme, Canguilhem procde donc une
reformulation du vitalisme une reformulation qui prendra tout son sens sur le fond de
la reconnaissance de la productivit normative de lerreur telle quelle a t voque
plus haut. Il sagira donc de monter comment limportance de lerreur, fondamentale
pour la conception de la science de Georges Canguilhem, articule sous le nom
d pistmologie historique , sera galement la pierre de touche pour la
comprhension de la dynamique spcifique de la vie en termes de polarit de la vie que
Canguilhem appellera la normativit vitale , et, comme nous voudrions montrer, tout
aussi fondamentale pour comprendre le lien entre cette normativit vitale et les normes
sociales, ou, pour emprunter le vocabulaire de Michel Foucault, qui sera essentiel dans
ce contexte, le rapport du bio-pouvoir la vie proprement normative.
Dans un premier temps, le texte propose donc une lecture compare de
lpistmologie historique de Canguilhem et de larchologie foucaldienne du savoir,
afin de montrer leurs convergences autour de la pense de lerreur, de lcart ou de la
dispersion comme concepts centraux dune certaine conception de lhistoire. La
deuxime partie est consacre la reformulation du vitalisme partir du constat de la
double dynamique de la vie; la troisime partie interroge limportance de cette
reformulation pour la notion de norme sociale de Canguilhem et pour la thorie du
pouvoir ou du bio-pouvoir de Foucault.
1. Epistmologie historique
Dans son Introduction lpistmologie historique, Hans-Jrg Rheinberger introduit
une notion dpistmologie la franaise entendue comme rflexion des conditions
historiques ainsi que les moyens dans et par lesquels quelque chose devient objet de

3
4
5
6
Georges Canguilhem, Aspects du vitalisme , in : id., La Connaissance de la vie, Paris 1998, p. 83100. Le texte Aspects du vitalisme comme les textes Machine et organisme et Le vivant et
son milieu sont issus de confrences donnes en 1946-47 au Collge philosophique. Cf.
Avertissement la premire dition de La Connaissance de la vie.
Canguilhem, Aspects du vitalisme , p. 85.
Ibid., p. 86.
Ibid., p. 99.
188

savoir, et dans et par lesquels le processus dobtention de connaissance scientifique est

dclench et maintenu en actif .7
Il y aurait donc, autour de 1900, un revirement de la thorie de la connaissance de la
tradition philosophique classique vers lpistmologie historique qui renverse le rapport
entre concept et objet partant du sujet connaissant en un rapport de lobjet au concept
dont le point dappui est lobjet mme. Ce dplacement, crit Rheinberger, est la
fois le noyau de lpistmologie et le point de dpart de son historisation. 8 Ce qui est
en cause nest donc plus la capacit du sujet de saisir les objets scientifiques, mais au
contraire, les conditions ncessaires pour que des objets deviennent des objets de
connaissance empirique. Par ce fait, la thorie de la connaissance fonctionne dsormais
sur un rgime empirico-historique et son objet devient objet historique, cest--dire un
objet soumis des variations historiques et non plus une condition transcendantale ou
une norme invariable.9 Cette pistmologie historique est attribue par Canguilhem
Gaston Bachelard, mais caractrise aussi, en tenant compte de certaines diffrences
galement voques hier, lhistoire de la science labore par Canguilhem lui-mme.10
Au lieu de revenir sur ce rapport, je voudrais plutt rappeler trs brivement
quelques points fondamentaux concernant cette comprhension de lpistmologie
historique avec la notion dhistoire prsente par Foucault dans lintroduction
LArchologie du savoir, publie en 1969. Foucault prsente le dcrochage dune
nouvelle histoire ou histoire gnrale dun modle dhistoire globale cherchant, lui,
restituer la forme densemble dune civilisation, le principe matriel ou spirituel
dune socit [] ce quon appelle mtaphoriquement le visage dune poque .11
Ce dcrochage, nous dit Foucault, sopre autour de la mise en question du
document qui, en surface, provoque deux consquences apparemment opposes: Dun
ct, lhistoire proprement dite efface lirruption des vnements au profit des structures
sans labilit (ou de la longue dure ), de lautre ct, lhistoire de la pense, de la
connaissance, de la littrature etc. multiplient les ruptures et cherchent les
hrissements de la discontinuit .12 Foucault termine son introduction en constatant
que cette mutation pistmologique de lhistoire est toujours en cours et que
lhistoire des ides, ou de la connaissance, semblent y tre particulirement rticentes:

7
8
9
10
11
12
Hans-Jrg Rheinberger, Historische Epistemologie. Zur Einfhrung, Hambourg: Junius 2007, p. 11

(Toutes les traductions sont miennes, M.M.).
Ibid., p. 12.
Rheinberger relve deux conditions fondamentales de cette pistmologie historique, dune part les
interminables rvolutions scientifiques fondes dans le surpassement initial du paradigme de la
physique dans sa forme classique, et dautre part la pluralisation des sciences reniant le projet
moderne dune science unique.
Voir pour une analyse en profondeur de ce lien et de lapparition du terme pistmologie
historique dans luvre de Georges Canguilhem les interventions de Pierre-Olivier Methot, From
concepts to exprimental systems. Trends in historical epistemology et de Camille Limoges,
Lpistmologie historique dans litinraire intellectuel de Georges Canguilhem lors du colloque
Von Bachelard zu Canguilhem zur Wissenschaftsgeschichte von heute , MPIWG, Berlin 9 au 11
dcembre 2010.
Michel Foucault, LArchologie du Savoir, Paris 1969, p. 18.
Ibid., p. 13.
189
Muhle
Comme si [] on [y] prouvait une rpugnance singulire penser la diffrence,

dcrire des carts et des dispersions, dissocier la forme rassurante de lidentique. 13
La mise en question du document de larchologie foucaldienne va donc lencontre
dun modle dhistoire qui vise le dpistage et la reconstitution dun pass inscrit dans
les documents et depuis vanoui. Le document nest plus pour lhistoire la trace des
choses passes, la matire inerte de lhistoire , au contraire, cest dans le tissu
documentaire mme que lhistoire cherche dfinir des units, des ensembles, des
sries, des rapports. Le document, dans sa matrialit, est lobjet de lhistoire sans
pour autant renvoyer lexistence de choses, dobjets naturels ou dtats, sachant que,
dans les mots de Paul Veyne, ces entits ne sont que les corrlats de pratiques
discursives. Larchologie foucaldienne rejoint donc la conception canguilhemienne
selon laquelle lobjet de la science est non naturel: La nature nest pas delle-mme
dcoupe et rpartie en objets et en phnomnes scientifiques. Cest la science qui
constitue son objet []. 14 Un objet qui est par consquent second par rapport lobjet
naturel quil dsigne ironiquement comme un pr-texte , une entit hors texte.
Lhistoire des sciences, comme larchologie, sexerce sur ces objets seconds, culturels
ou discursif. Et Canguilhem continue en dfinissant lobjet du discours historique
comme lhistoricit du discours scientifique dune historicit qui, loin dtre
linaire, cest--dire globale dans les termes foucaldiens, reprsente leffectuation
dun projet intrieurement norm, mais qui est traverse daccidents, retarde ou
dtourne pas des obstacles, interrompue de crises, cest--dire de moments de
jugement et de vrit .15
Sans marrter davantage sur la discussion dpistmologie historique, je voudrais
insister sur lanalogie entre cette historicit entendue comme une normativit intrieure
confronte en permanence des carts, des erreurs et des crises, et la double dynamique
de la vie telle que Canguilhem la introduite dans le Normal et le Pathologique.
2. Vitalismes
Bien avant ses rflexions sur lpistmologie bachelardienne, Canguilhem avait dj
relev limportance fondamentale de lerreur, de la dviation et de lcart par rapport
la normalit pour la connaissance ou la science de la vie: La vie peut uniquement tre
connue travers ses erreurs qui renvoient chaque tre vivant son imperfection

13
14
15
Ibid., p.21.
Georges Canguilhem, Objet de lhistoire des sciences , in : Etudes dhistoire et de philosophie des
sciences concernant les vivants et la vie, Paris 72002, p. 16.
Ibid., p. 17 . Cette historicit se situe au-del dune conception progressiste de la science et scinde la
notion mme de vrit scientifique en un dire vrai et un tre dans le vrai , une distinction que
Canguilhem tablit autour de Galile et que Foucault reprend dans LOrdre du discours propos de
Mendel: Mendel disait vrai, mais il ntait pas dans le vrai du discours biologique de son
poque: ce ntait point selon de pareilles rgles quon formait des objets et des concepts
biologiques. [] Il se peut toujours quon dise le vrai dans lespace de lextriorit sauvage; mais on
nest dans le vrai quen obissant une police discursive quon doit ractiver en chacun de ses
discours. Michel Foucault, LOrdre du Discours, Paris 1971, p. 37.
190

constitutive. Cest ce que Xavier Bichat avait dj expliqu en opposant les sciences
physiques obissant des lois invariables et constamment les mmes, aux sciences
physiologiques, elles seules capables de percevoir un objet dtudes soumis des
variations imprvisibles, ou dans les mots de Bichat, linstabilit et lirrgularit
essentielle aux phnomnes vitaux .
Nanmoins, ce constat de la fondamentale irrgularit de la vie, cest--dire de sa
capacit errer et dvier, de sa rceptivit pour les valeurs ngatives, dploiera toute
sa profondeur thorique dans une perspective non de connaissance mais de normativit
de la vie que plus tard dans luvre de Canguilhem et en rapport avec le constat de la
priorit de linfraction sur la rgle dans les Nouvelles Rflexions. Un constat que
Canguilhem raffirme dans son article Vie publi dans lEncyclopaedia Universalis
en 1973, o il crit que la valeur de la vie, la vie comme valeur senracinent dans
la connaissance de son essentielle prcarit .16
Dans son texte Lhistoire pistmologique de Georges Canguilhem , Rheinberger
constate une certaine parent entre le regard historique que Canguilhem porte sur les
phnomnes tels que la sant, la maladie et la vie et sa propre notion ou comprhension
de ce quest la science car Canguilhem considre la vie autant que les sciences
premirement comme des processus vritablement autocorrectifs et deuximement
comme des processus principalement prcaires .17 Cest aussi dans ce sens quon doit
lire laffirmation canguilhemienne selon laquelle la thorie de la vie doit pouser la
structure de son objet dtude, cest--dire lessence dialectique de la vie.18 Lerreur,
la prcarit, la morbidit donnent lieu la normativit vitale une normativit vitale qui
se ddouble dans cette priorit de valeurs ngatives sur les valeurs positives: Car la vie
tend bien vers son autorgulation ou autocorrection cest en ceci quelle est
organique , mais elle suppose galement le dpassement ou la mise en question
constante de la normalit tablie par lautorgulation. La prcarit des processus
vitaux est donc fondamentale dans ce sens prcis, parce quelle est constamment
actualise par une vie proprement vitale autrement cette vie serait pathologique.
La vie est bien organique par sa tendance lhomostasie, cest--dire au maintien
dun quilibre organique intrieur (du milieu intrieur comme la montr Claude
Bernard). Mais cet quilibre organique et homostatique est remis en question en
permanence par une dynamique vitale dauto-transgression qui nest pas assimilable par
les tendances autorgulatrices de lorganisme mais reprsente un excs par rapport
cette tendance, une rsistance interne limmobilisation dans ltat dquilibre. Cest
cette vitalit qui permet au vivant de scarter de ses propres normes: La vie normale est
donc la vie normative, cest--dire une vie qui ne sadapte pas un milieu et une norme
donnes, mais qui les remet constamment en question et cre ses propres normes: Sa
normativit consiste faire craquer les normes .19

16
17
18
19
Georges Canguilhem La Vie , in : Encyclopaedia Universalis, Paris 2002, p.532.

Hans-Jrg Rheinberger, Die epistemologische Historie von Georges Canguilhem , in: Hans-Jrg
Rheinberger, Epistemologie des Konkreten, Frankfurt/Main 2006, p. 56. (La traduction est mienne,
M.M.)
Canguilhem, Aspects du vitalisme , p. 85.
Canguilhem, Le normal et le pathologique, p. 106.
191
Muhle
Lenjeu de la notion canguilhemienne de vie consiste alors dans le fait que la

normalit organique se voit en permanence expose aux carts normatifs: la vie ne
sarrte pas sur un tat dquilibre (tel quil est cr de manire artificielle au
laboratoire) mais met cet quilibre lpreuve et le dpasse. Autrement, cest--dire si
la vie ntait que organisme, elle se trouverait rgie par des constantes
pathologiques valeur rpulsive et strictement conservatrice .20 Mais au lieu de
cela la vie est polarit et donc organique et cratrice en mme temps, cest--dire
vivante.
Canguilhem reformule par consquent un vitalisme au-del de son opposition
ternelle avec le mcanisme, mais galement au-del des notions classiques de principe
vital ou dentlchie. Le vitalisme est la pense du vivant partir du vivant, dun vivant
lui-mme habit par une double logique qui peut tre formule en termes de prcarit et
dautocorrection si et seulement si la prcarit excde la tendance autocorrective, ou
pour le dire autrement, si lcart excde et prcde la normalit. A la toute fin de son
texte sur lerreur, Canguilhem articule cette conclusion de la manire suivante: La
connaissance de la vie, comme celle de la socit, suppose la priorit de linfraction sur
la rgularit. 21
3. Du vital au social
Canguilhem adresse donc directement la question du social dans ce texte et dans les
autres textes des Nouvelles Rflexions mais uniquement, comme il prcise, en vue dune
meilleure comprhension du vital: Cest en vue de lorganisme que je me permets
quelques incursions dans la socit. 22
Dans son texte sur les Normes vitales et normes sociales, Pierre Macherey a toutefois
montr pourquoi la normativit vitale doit tre comprise ncessairement sur fond de la
confluence des normes vitales et sociales afin dviter laporie du vitalisme classique
tout en maintenant la dynamique interne de la norme. Il faut, crit Macherey,
dynamiser la norme de lintrieur ce qui est prcisment lenjeu du passage dune
doctrine du normal une doctrine du normatif 23 donc lenjeu canguilhemien. La vie
cesse dtre comprise en termes de nature substantielle et devient un projet au sens
dun lan qui la dsquilibre en la projetant sans cesse en avant et lexpose au risque
permanent de trbucher.
Devant cette comprhension dune vie normative dynamise de lintrieur, il reste
pourtant clarifier la question de lorientation que peut prendre cette dynamique, une
orientation qui donne la ralisation du projet vital son allure densemble , grce
une certaine dose de finalisme, une perspective intentionnelle essentiellement

20
21
22
23
Ibid., p. 156.
Ibid., p. 216.
Ibid., p. 173.
Pierre Macherey, Normes vitales et normes sociales dans lEssai sur quelques problmes du normal
et du pathologie , in: Actualit de Georges Canguilhem, p. 74.
192

subjective .24 Car lexprience du vivant, qui selon Canguilhem est une exprience
concrtement vcue, est ncessairement singulire, il ny a pas dexprience du vivant
en gnral: elle est et ne peut tre quune exprience individue .25
Afin dchapper au double dilemme dune normativit purement subjective et par
consquent sans allure commune, et dune normativit dont la dynamique serait
impose de lextrieur, donc explique par une causalit de type mcanique, Macherey
rappelle la diffrence entre lexprience individue et lexprience individuelle:
lexprience individue produit des individus partir de conditions qui ne sont pas
strictement individuelles, cest--dire qui supposent lintervention du milieu humain o
prvalent des formes dexistence qui ne sont pas individuelles mais collectives: La vie
humaine se trouve la confluence de dterminations biologiques et sociales.26 La
normativit interne de la vie ne peut donc tre pense que sur fond de ce recoupement
du social et du vital: Le pouvoir de vivre [] seffectue dans des formes qui, bien loin
dtre librement inventes par des individus qui conditionneraient leurs seuls caractres
biologiques, [] rpondent des conditions qui sont celles dfinissant la constitution
dun milieu humain travers son histoire. 27 Les normes vitales et sociales obtiennent
leur ralit dans linfraction, cest--dire dans laction concrte travers laquelle elles
seffectuent en affirmant, contre les obstacles qui sopposent cette action, leur valeur
normative .28
Face la hirarchisation entre social et vital, Macherey oppose donc un fait
essentiellement ngatif partags par les normes sociales et vitales: elles ne peuvent
offrir des modles dexistence prfabriqus portant en eux la puissance de simposer
elles sont des paris ou des provocations qui nont dimpact qu travers
lapprhension de lanomalie ou de lirrgularit sans lesquelles elles nauraient tout
simplement pas lieu dtre .29
Canguilhem se maintient plus rserv face cette question et limite lexplication du
social par le vital la dimension organique du vital. Ainsi il explique quil faut voir
dans les tentatives des planifications (techniques) des essais de constitutions dorganes
par lesquels une socit pourrait prsumer, prvoir et assumer des besoins, au lieu den
tre rduite les enregistrer et les constater par des comptes et des bilans .30 Lordre
social tend imiter lorganisation vitale en tant que systme auto-organis, capable de
maintenir un quilibre intrieur, une homostasie.
Dans son analyse du rapport entre vital et social, Canguilhem insiste par la suite sur
lextriorit de la norme sociale au sujet norm, la population: dans une organisation
sociale, les rgles dajustement des partis en une collectivit la normalisation
constitutive dun corps de socit, dune communaut sont extrieures au multiple

24
25
26
27
28
29
30
Ibid., p. 74s.
Ibid., p.75.
Cf. ibid., p. 76.
Ibid., p. 77.
Ibid., p. 83.
Ibid., p. 84.
Canguilhem, Le normal et le pathologique, p. 184.
193
Muhle
ajust tandis que, dans un organisme vivant, ces mme rgles sont immanentes,
prsentes sans tre reprsentes, agissantes sans dlibration ni calcul .31 Lorganisme
social comporte donc, selon Canguilhem, quelques caractres du mcanisme: une
socit est la fois machine et organisme. 32 Et plus loin, il constate que les
phnomnes dorganisation sociale sont comme une mimique de lorganisation vitale,
au sens o Aristote dit de lart quil imite la nature : il tend retrouver le sens dune
production. Lorganisation sociale est par consquent invention dorganes .33
Canguilhem propose donc une notion de norme sociale qui dans son mimtisme
renvoie la dimension organique de la vie, et qui, par consquence, ninteragit pas avec
lautre ple de la vie, sa dimension proprement vitale, cest--dire transgressive. De
cette manire, et bien que Canguilhem ait fourni tous les lments ncessaires pour
penser le rapport entre norme sociale et norme vitale dans une optique explicitement
normative, nietzschenne presque, il se rsiste articuler ce dplacement et il faudra se
tourner, nouveau, vers Foucault pour trouver des lments qui permettent de penser
une thorie du pouvoir qui tient compte de la vie comme polarit.
Dans les cours Scurit, territoire, population des annes 1977-1978 Foucault va
procder une reformulation de la dfinition de la biopolitique dabord introduite dans
la Volont de Savoir comme pouvoir investissant la vie de part en part quil loigne
maintenant du rgime disciplinaire en le rebaptisant avec le nom de gouvernementalit
ou de dispositif de scurit. Ce ragencement se fait autour de la question de la norme:
Foucault reformule ici lopration de la norme disciplinaire comme une normation qui
soppose la normalisation des normes de scurit. Tandis que la normation
disciplinaire part dune norme tablie et organise le dressage autour de cette norme qui
sert alors comme pierre de touche pour la diffrence normal-anormal, la normalisation
cre sa propre norme en faisant jouer diffrentes distributions de normalit et en
faisant en sorte que les plus dfavorables soient ramens aux plus favorables .34 La
norme est alors un jeu lintrieur des normalits diffrentielles et lopration de
normalisation consiste jouer et faire jouer les unes par rapport aux autres ces
diffrentes distributions de normalit .35
Les normes biopolitiques ont donc un double rapport la vie qui est dabord leur
objet mais dont ils adoptent ensuite la dynamique comme modle de fonctionnement.
Les dispositifs de scurit laissent libre cours aux phnomnes vitaux mais se greffent
sur leur dynamique: Elles crent les conditions, le milieu, dans lesquelles la vie peut
sautorguler librement, tel que Foucault la montr propos du traitement de la disette
et des techniques dinoculations contre la variole. Car bien que les normes dune socit
de normalisation soient artificielles, elles imitent limpulsion crer des normes qui est
naturelle ou vitale. De mme lagir du pouvoir sur le milieu est un rapport extrieur ou

31
32
33
34
35
Ibid., p. 186.
Ibid., p. 187.
Ibid., p. 188.
Michel Foucault, Scurit, territoire, population. Cours au collge de France, 1977-1978, Paris 2004,
p. 65.
Ibid.
194

artificiel mais la cration ou la modification du milieu est une impulsion vitale dont la
logique est reprise par la biopolitique.
Les techniques de scurit se projettent donc avec la vie, elles ne supposent aucun
modle prfabriqu, mais sont elles aussi des paris ou des provocations pour reprendre
les termes de Macherey. Cest ainsi quelles sont capables, contrairement aux
techniques disciplinaires ou souveraines, de faire face la normativit interne de la vie,
non seulement dans son versant organique mais aussi dans son versant vital, cest--dire
dans sa dimension alatoire, fondamentalement et premirement prcaire. Cest
pourquoi, non seulement la vie et les sciences de la vie, mais aussi les stratgies dun
pouvoir sur la vie sont, pour reprendre les mots de Rheinberger, des processus la fois
autocorrectifs et prcaires, normalisants et normalisateurs.
Cest dans ce sens quil faut comprendre la dfinition foucaldienne de la biopolitique
comme pouvoir positif sur la vie: Cest un pouvoir qui sappuie sur la productivit
interne de la vie et engage sa normativit afin de dominer plus efficacement son objet.
Ainsi devant la menace ventuelle de lquilibre social par un excs normatif de la vie,
le dispositif scuritaire produit, son tour, une nouvelle norme et rduit la normativit
la normalit, cest--dire quil pathologise la vie afin de la rendre gouvernable. La
supposition quune politique de la vie puisse exister au-del des relations de pouvoir,
comme le suggrent quelques interprtations contemporaines de la biopolitique, est
donc contredite avant la lettre par Canguilhem quand il insiste encore et toujours sur
limplication indissociable de la dynamique normale (donc normalisante ou disciplinaire
dans les mots de Foucault) avec la dynamique normative (ou normalisatrice) il ny a
pas de pouvoir de la vie qui puisse exister indpendamment dun pouvoir sur la vie, ni
vice-versa, car la vie est fondamentalement polarit. Ou pour le dire dans les mots de
Foucault: [] il ny a pas de pouvoir sans rsistances; [] celles-ci sont dautant plus
relles et plus efficaces quelles se forment l mme o sexercent les relations de
pouvoir; la rsistance au pouvoir na pas venir dailleurs pour tre relle, mais elle
nest pas pige parce quelle est la compatriote du pouvoir. [] elle est donc comme
lui multiple et intgrable des stratgies globales. 36

36
Michel Foucault, Pouvoirs et Stratgies , Entretien avec Jacques Rancire (1977), in: id., Dits et
crits II, 1976-1988, Paris 2001, p. 425.
195
THE ROLE OF SCIENTIFIC IDEOLOGY IN

CANGUILHEMS HISTORIOGRAPHY
Cristina Chimisso
Introduction
In 1969, Canguilhem gave a paper entitled Quest-ce quune idologie scientifique?,
published the following year in a journal and eventually in the volume Idologie et
rationalit.1 The title may be surprising for two reasons. One is the central term:
ideology. This term had not previously been part of Canguilhems reflection on
science and indeed he did not elaborate on it in later works. The other reason is the
adjective that accompanies it: scientific. To many the expression scientific ideology
may sound as an oxymoron. Why did Canguilhem introduce the concept of ideology at
this stage, after publishing his major books? A general answer could be that it was the
Sixties, and both Marx and the concept of ideology were on the philosophical agenda.
More to the point, these were the years in which Louis Althusser was publishing his
major works. Indeed, when asked by Franois Bing and Jean-Franois Braunstein
whether his use of the concept of ideology was inspired by Althusser, Canguilhem
answered that it was. Unfortunately, he did not elaborate.2
Bing and Braunstein were in fact asking Canguilhem to confirm what he had already
stated in the Preface of Idologie et rationalit, written in 1977. There, Canguilhem told
his readers that he had introduced the concept of scientific ideology in his lectures under
the influence of Louis Althusser and Michel Foucault. His statement seems to suggest
that Canguilhems thought was undergoing an evolution due to the two new stars of
French academia and culture. However, this would be at least partly misleading. Indeed,
in this same preface Canguilhem also wrote that a reader of Idologie et rationalit
would not find signs of change or evolution in his thinking. Despite introducing a new
concept in his philosophy, Canguilhem appeared to look back rather than forward. He
explained that the introduction of the concept of scientific ideology was a way of
refurbishing without rejecting the lessons of Gaston Bachelard.
I intend to focus precisely on the use that Canguilhem made of the concept of
scientific ideology, which he employed in order to re-think and re-adjust the
epistemological ideas that he had borrowed from Bachelard. I will not follow the lead of
Althussers and Foucaults influence on Canguilhem. Although this can certainly be
done, in fact as we shall see the concept of ideology as Canguilhem uses it is much
more securely inscribed in the existing tradition of historical epistemology and in the
1
Published in Georges Canguilhem, Idologie et rationalit dans l'histoire des sciences de la vie, Paris:
Vrin, 1993 [1977].
Franois Bing and Jean-Franois Braunstein, "Entretien avec Georges Canguilhem," in Actualit de
Georges Canguilhem. Le normale et le pathologique, ed. F. Bing, J-F. Braunstein, and Elisabeth
Roudinesco (Paris: Synthlabo, 1998).
197
Chimisso
development of his own thought. Not only did Canguilhem neglect to discuss the
complexities and novelties of Althussers concept of ideology, but he also avoided any
mention of either Althusser or Foucault in his paper on scientific ideology. His only
reference to a previous concept of ideology is to Marxs. In the present paper, I shall
follow Canguilhem and interpret his use of the concept of ideology as a solution, or an
attempted solution, to a long-standing historiographical problem, which has its root in
the classic tradition of historical epistemology.
The problem
The problem to which Canguilhem aimed to give a new solution in his paper is that of
the object of the history of science. What is history of science history of? Science, of
course. However, it goes without saying that this only invites a further question: what is
science? It is no surprise that Canguilhem started his paper with the problem of
demarcation between science and non-science. This may sound like a boring starting
point of some foundation course in philosophy of science, but it is certainly not it. The
epistemological rupture between science and non-science was at the core of the tradition
of historical epistemology. Canguilhem accepted Bachelards epistemological break
between scientific knowledge and common knowledge, and the normative approach that
is its foundation.
In fact, Canguilhem insisted on the importance of sorting out what is science and
what is not. His main aim in doing so was to have a strong epistemological guide to the
writing of the history of science. This is already rather different from Bachelard; in the
latters work the study of history is at the service of epistemology. Canguilhems main
concern was how to construct a narrative out of seemingly disparate facts, and diverse
activities, guided by heterogeneous aims. What is the past of a current scientific
concept? What is its history? These are the questions that preoccupied Canguilhem. For
him the idea of reconstructing the past as it really was is just impossible. He believed
that the historian can only start from the present and try to reconstruct the past of our
present concepts, theories and ideas. In order to trace back the history of a current
concept, for him it is necessary to understand which concepts, theories and practices are
really part of the past of a given concept, and which have no significant links to it. One
could visualize Canguilhems (and any historians) situation as being presented with a
mass of dots representing concept, theories, events, experiments and so on. Which dots
can realistically be linked with each other? Which can be included in the past of the
concept (or theory) under study? Canguilhem believed that, in order to establish the real
links between the dots, and indeed to choose the right dots, a normative approach is
necessary.
The mention of Canguilhems normative approach is sometimes met with a certain
amount of suspicion and even hostility. This is often because some people take it to be
close to the ways of whiggish history. In fact, Canguilhems normative approach is
aimed precisely at uncovering the false trajectories that whiggish history may create
when failing to see the differences, indeed the epistemological breaks, that separate our
198
THE ROLE OF SCIENTIFIC IDEOLOGY IN CANGUILHEMS HISTORIOGRAPHY
current concepts from past concepts. In order to understand whether past concepts are
linked to our current ones, for Canguilhem it is necessary to take the present ones as
norms, and then evaluate past concepts according to those norms. As for Bachelard, for
Canguilhem current science is the norm against which the historian and the philosopher
should judge whether a theory or a concept is scientific, although, as we shall see,
Canguilhem held a more complex view of the relationship between scientific and nonscientific concepts and theories.
Canguilhem provided the clearest example of his application of normativity to
history writing in La formation du concept de rflexe aux XVIIe et XVIIIe sicles.3 In this
book, Canguilhem started with the current concept of reflex movement and used it as
the norm to establish whether that concept existed in the past. A certain whiggish
history had created continuity between the current concept of reflex and Descartes
concept of reflex. For Canguilhem, his own normative method reveals this continuity to
be false, as it is only a projection of our present concept onto a past concept that in fact
had no relation with it. An analysis of the two concepts the current one and
Descartes for him shows that the two concepts are heterogeneous. Descartes concept
of reflex movement does not belong to current science, indeed there is a discontinuity
between the two. There is no need here to go into details; it will suffice to say that for
Descartes the flux of spirits in involuntary bodily movements is always from the brain
towards the periphery, and never in the opposite direction. This goes directly against the
modern concept of reflex movement which originates in the periphery of the nervous
system.
So far, this may not be different from Bachelard. However, Canguilhem created
another connection: that between the present concept of reflex and the concept of reflex
elaborated by the seventeenth-century natural historian and medic Thomas Willis.
Canguilhem explained that Williss concept of reflex, unlike Descartes, includes
movement from the periphery to the centre. In this, it is not unlike the current one. Here,
Canguilhem appears to stride rather far from Bachelard. How can a seventeenth-century
natural philosopher, who had a rather imaginative interpretation of life as light, be part
of the history of science? And yet Canguilhem pointed out that precisely Williss
interpretation of life as light brought about his use of the optical laws of reflection in the
interpretation of biological phenomena. In this unexpected context, Willis conceived of
the reflex movement as originating in the periphery and going towards the brain. For
this reason, Williss concept of reflex is a stepping stone in the history of the current
concept.
These few reminders of Canguilhems history of the concept of reflex already
suggest that it does not sit unproblematically with the Bachelardian view of history of
science. Bachelard envisaged science as having a discontinuous history. In La formation
de lesprit scientifique he proposed three broad periods of the history of thought: the
Georges Canguilhem, La formation du concept de rflexe aux XVIIe et XVIIIe sicles, Paris: Presses
Universitaires de France, 1955.
199
Chimisso
period of the pre-scientific mind, of the scientific mind and of the new scientific mind.4
These periods, for Bachelard, are separated by epistemological breaks. La formation de
lesprit scientifique is dedicated to the obstacles that the mind has to overcome in order
to become scientific. These obstacles are the expression of human desire and
imagination, and they produce eminently non-scientific works. Bachelard drew
examples from different times, including recent periods, as desire and imagination of
course still persist, and indeed for him they should always express themselves, although
not in scientific work. However, before the emergence of the scientific mind, for him all
works that the mind produced were the expression of imagination and desire, and as a
consequence non-scientific. The date that Bachelard assigned to the emergence of the
scientific mind, and therefore of science, is between the end of the eighteenth and the
beginning of the nineteenth century. In his scheme, Williss work falls on the wrong
side of the epistemological break between pre-science and science, indeed it appears to
be an epistemological obstacle rather than a stepping block towards modern science. For
Bachelard, history of science has been rather short: at the time of his writing, only a
couple of centuries. Indeed, he believed that a new epistemological break had occurred
with Einsteins theory of relativity at the beginning of the twentieth century.
In his work as a historian of science, Canguilhem established continuities that went
back to times when for Bachelard science did not exist. Canguilhem did not think that
Williss theories were scientific, but he did think that Williss concept of reflex was part
of the history of the modern concept of reflex. He believed that concepts can have a
continuous history that strides the science and non-science divide. That scientific
concepts can emerge within non-scientific theories is clearly a controversial point that I
am not going to discuss here. What interests me here is that he established continuities
that go back centuries and that he found the roots of current scientific concepts in
concepts that a strict application of historical epistemologys normativity would have
excluded from science. So, how do we classify Williss theories? These are not
scientific, and yet for Canguilhem it would be a mistake to expel them from the history
of science. Indeed, for him if we do not include Williss imaginative analogy between
life and light, we will not be able to understand how the reflex movement came to be
conceptualized within science.
Canguilhem needed to find a theoretical framework that would enable him to reintegrate into history of science those activities, theories and concepts that a rigid
application of Bachelards epistemological break would have excluded. As he explicitly
wrote, far from aiming at rejecting Bachelards lesson, he intended to refurbish it. It is
here that the concept of scientific ideology can provide a solution to Canguilhems
historiographical dilemma of the object of history of science.
Gaston Bachelard, La formation de l'esprit scientifique: contribution une psychanalyse de la

connaissance objective, Paris: Vrin, 1993 [1938].
200
Scientific ideology and historiography

Canguilhem proposed two types of scientific ideology: one that precedes science and
one that it is parasitic to it. The first type is quite close to what Bachelard saw as prescience and indeed as an obstacle that the mind has to overcome in order to become
scientific. This type of ideology comprises the systems of ideas that occupy the same
realm as science will later do, but that do not possess the methodological requirements
and operational possibilities of science. For Canguilhem these systems of ideas cannot
just be categorized as non-science, or anti-science, as some other philosophers put it.5
Canguilhem argued that these systems were considered to be scientific, and were
meaningfully distinct from religion and superstition. Examples of these systems of ideas
that Canguilhem cited are the doctrines elaborated by Democritus and Lucretius. He
argued that, although Democrituss and Lucretiuss theories were not scientific, they
were presented as such; they were aimed at being scientific, in the sense of being
separated from, and opposed to, superstition. For this reason, Canguilhem argued that
they are part of the history of science. They are examples of scientific ideologies.
Another type of scientific ideology that Canguilhem described is parasitic to science,
as it extends scientific theories to fields that are not of their competence. An example of
this type of ideology is Herbert Spencers theory of evolution. Spencer extended the
field of application of laws that belonged to embryology and biology to the totality of
human experience. His theory for Canguilhem is not scientific, although for its
relationship to science, it is scientific ideology. This second type of scientific ideology
in fact fits better with Marxs concept of ideology, as Canguilhem himself remarked. He
argued that this type of ideology serves particular interests. In the case of Spencers
theory of evolution, it provided a justification for industrial society, against traditional
society, and played a role against both religion and socialism. However, the first type of
scientific ideology, that which precedes science, is of particular interest here, as it
provides the best examples of those theories and practices that for Canguilhem are part
of the history of science, and that for Bachelard are not.
For Canguilhem, scientific ideologies occupy an intermediate position between
science and non-science. Scientific ideologies are not science, but nevertheless they
belong to its history. Many of the theories and practices that Canguilhem regarded as
scientific ideology, and therefore as part of the history of science, would have been
categorized by Bachelard as pre-science, or as obstacles to science. For Bachelard,
although they are necessary to science as they provide the dialectical mechanism for its
development, they are external to science, and therefore not part of its history in a strict
sense. By using the category of scientific ideology, Canguilhem not only re-integrated
into the discipline of history of science systems of ideas that Bachelard had at least in
principle excluded from it, but also give them a history. He shared with Bachelard the
5
In Quest ce quune idologie scientifique? Canguilhem discussed in particular the concept of antiscience that Bogdan Suchodolski had presented in his paper at the 12th International Congress of
History of Science in 1968. See Bogdan Suchodolski, Les facteurs du dveloppement de l'histoire
des sciences, in XIIe Congrs International d'Histoire des sciences: Colloques, textes des rapports,
Paris: Blanchard, 1970.
201
Chimisso
view that error, anti-science, or epistemological obstacles, do not have a history. The
theories and practices that Canguilhem regarded as ideologies would have been ahistorical epistemological obstacles for Bachelard. For Canguilhem, however,
ideologies are not simply error, and, unlike epistemological obstacles, develop and
change. Indeed, they may show some elements of continuity with present science, as for
instance in the case of Thomas Williss work. The theories in which he used the concept
of reflex might have been far from scientific, and we can safely guess that Bachelard
would have regarded them as an example of the power of imagination over rationality.
However, not only did Canguilhem think that Williss work deserves a place in the
history of science, but he also regarded Willis as the father of the modern concept of
reflex movement.
The concept of scientific ideology changes the content of the history of science, by
broadening its object. This of course has consequences for history writing. By using it,
Canguilhem partly avoided the type of internalism6 that in fact could be a
consequence of Bachelards historical epistemology. If ideologies are part of history of
science, it means that history of science is not only the history of truth, but also the
history of the attempts at the truth. Indeed, in an article published a couple of years
before the paper on scientific ideology, Canguilhem urged to recognize that superseded
notions, attitudes and methods had in their time superseded previous notions, attitudes
and methods.7 A critical attitude for Canguilhem is a scientific one, and therefore past
notions and methods, even if they no longer obey current norms of scientificity, for him
should still be the object of the history of science. While defending normativity in the
history of science, he claimed that the judgments of epistemology do not translate into
purges or executions of theories that do not live up to the norm of current science. For
him, there is room in the history of science for those theories and practices that are not
science by current standards, but still exhibit the critical attitude of science.
Conclusion
Canguilhem conceived of normativity in a more liberal way than Bachelard did. This is
perhaps inevitable for two reasons. One is that Canguilhem focused on history writing,
as he did in La formation du concept du rflexe. He reconstructed the history of the
concept of reflex by using the current concept of reflex movement as its norm.
However, the resulting narrative cannot exist in a pure form in which truth and error,
science and non-science are separated in a clear-cut way. Canguilhem resolved or at
least attempted to resolve this dilemma by acknowledging a grey area and calling it
ideology. The other reason why it seems inevitable that Canguilhem should have a
less clear-cut notion of the distinction between truth and error is his focus on the life
6
I use the term internalism loosely, only to mean a history of science whose only object is science
itself. Canguilhem criticised both externalism and internalism. About the latter, he believed that
internalists confused the object of history of science with the object of science. See Georges
Canguilhem, Lobjet de lhistoire des sciences [1966] in id., Etudes d'histoire et de philosophie des
sciences concernant les vivants et la vie, Paris: Vrin, 1994 [1968].
Ibid.
202
sciences, and on medicine and psychiatry. In these disciplines, which are concerned
with life and with events that are never precisely repeated, it is much more difficult to
draw clear lines between truth and error, and to have norms that are precise and
universal. When discussing normativity in medicine and psychiatry that is what
should be considered normal in these fields Canguilhem gave subjectivity an
important role to play, and he recognized that different situations in life call for different
sets of norms. The life of a diabetic is normal as the life of a non-diabetic is, but these
are two different types of normality, as the lives of these two individuals obey different
norms. Indeed, each individual lives at the crossroad of social, biological, economic and
individual norms.8 The complexity of the concepts of norm and normativity in the
sciences he studied had a profound, though indirect, effect on his concept of normativity
in history writing.
On the other hand, Canguilhem did not accept to put science and its history on the
same level of the history of other cultural phenomena. Although he conceded that
science is not historically independent from the state of the industry and economy in
which it emerges, he nevertheless believed that science is autonomous from the
ideology of the dominant class, as they are its method and problematic. Indeed, he cited
Marx writing that there is a difficulty with art, because, although it is the product of
specific social conditions, it nevertheless acquires a permanent value that transcends the
conditions of its emergence. Canguilhem argued that Marxism should not refuse to
science what affords art. Indeed, normativity is still at the centre of Canguilhems view
of science and its history; he would not have been prepared to suspend judgement on the
content of truth, of scientificity, of a theory as a current sociologist of scientific
knowledge would do. Moreover, history of science for him has still an internal logic of
development. As he puts it, in matters of the history of the sciences the rights of logic
should not be replaced by the rights of the logic of history.9 For him the analysis of
doctrines reveals continuity and discontinuities in the history of science. Indeed, he
analysed doctrines in order to construct his narrative of the history of the concept of
reflex: for him Descartes theory could not be the past of the current concept of reflex
because the two are logically incompatible. He was not prepared to treat history of
science as a cultural phenomenon among others as other French historians of science
such as Hlne Metzger, whom he cited, did. Unlike Canguilhem, she regarded
scientific doctrines as expressions of particular ways of thinking, which are common to
all other human expressions of a particular time. Similarly, he would not have accepted
the interconnections between religion and science that Koyrs historiography proposed.
The concept of scientific ideology, however, could serve as a corrective for what for
him were the too severe historiographical consequences of the normativity at the centre
of historical epistemology. Without this corrective, it would have been hard to justify
the narratives he constructed, like that of the concept of reflex movement. This is
because these narratives exhibit some continuity across centuries, and include concepts
born out of theories that had little to do with the norm of modern science. Canguilhem
8
Canguilhem discussed these themes in his most famous book, Georges Canguilhem, Le normal et le
pathologique, Paris: Presses Universitaires de France, 1999 [1966].
Canguilhem, La formation du concept de rflexe aux XVIIe et XVIIIe sicles.
203
Chimisso
needed a new category for a class of theories that are not expelled from history of
science, but at the same time are not granted the status of science. In this, the concept of
scientific ideology was a possible solution. Canguilhem used the category of scientific
ideology in order to develop historical epistemology in a direction that in fact fits with
his own historical work.
204
IST EINE HISTORISCHE EPISTEMOLOGIE DER

SOZIALWISSENSCHAFT MGLICH?
Frieder Otto Wolf
Schon Francis Bacon und Thomas Hobbes haben es gewusst: Es gibt einen wichtigen
Unterschied zwischen den Wissenschaften von der Natur (bzw. der Mathematik) und
denjenigen Wissenschaften, die sich mit dem beschftigen, was Aristoteles noch als das
Gemensche (ta anthropina, als Kollektivsingular) bezeichnet hatte. Diesen
Unterschied sahen sie allerdings nicht mehr in einer andersartigen Wissenschaftlichkeit,
sondern in ihrer Umkmpftheit (und in ihrer damit verbundenen andersartigen
Belastung durch Vorurteile). Auch Karl Marx (und Sigmund Freud) haben bekanntlich
noch beansprucht, auf ihren Gebieten endlich den Durchbruch zu einer wirklichen
Wissenschaft geschafft zu haben (und haben dieser Auffassung entsprechend auch
methodologisch durchaus mathematisch-naturwissenschaftliche Modelle in ihre eigene
Forschungspraxis bersetzt).1
Darin wird in der deutschsprachigen Debatte heute wohl berwiegend eine
positivistische (oder zumindest szientistische) Verirrung gesehen. Ich denke, diese
Doxa beruht ihrerseits auf einem epistemologischen Unverstndnis, einer unzulssigen
Verengung, die eine historische Epistemologie zu berwinden geeignet ist, die auch
einer physikalistischen (bzw. auf andere Weise reduktionistischen) oder empiristischen
Verengung des Wissenschaftsverstndnisses kritisch mit guten Argumenten begegnen
kann.
Ich mchte mich aber hier und heute darauf beschrnken, auseinanderzulegen, wie
Georges Canguilhems Konzept des epistemologischen Einschnitts einen anderen
Zugang zu der Frage der Wissenschaftlichkeit in diesen Wissenschaften von
Geschichte, Gesellschaft und Staat erffnet, als den in der deutschen Debatte seit den
1870er Jahren blich gewordenen, in der sie im Kern als Geistes- und
Staatswissenschaften konzipiert worden sind aber auch einen anderen Zugang als den
in den Traditionslinien der social sciences in den USA, wie sie seit 1945 weltweit
hegemonial geworden sind, und auch der sciences humaines in Frankreich, die im
franzsischen Moment der Philosophie (Alain Badiou) das wichtigste alternative
Paradigma2 gebildet haben.3
1
Ich kann hier offen lassen, ob auch auf anderen Gebieten des hier angesprochenen Feldes
vergleichbare epistemologische Einschnitte vollzogen worden sind.
Nicht einmal in erster Linie wg. des in ihrem etablierten Rahmen Erreichten, sondern wg. der von
ihnen ausgelsten Versuche, aus ihren disziplinren Modellen auszubrechen und unter Umgehung der
hier diskutierten Problematik neue Paradigmen der Forschung zu ermglichen, wie dies exemplarisch
etwa von Henri Lefebvre, Cornelius Castoriadis und vor allem von Michel Foucault vorgefhrt
worden ist.
Zumeist wird das konservative englische Modell bersehen, wie es etwa im sog. Oxford Tripos
zum Ausdruck kam, in dem noch in den 1960er Jahren Philosophie, Politik und konomie zur
Ausbildung von Fhrungskrften angeboten wurden: Hier wurde die Herausbildung neuer
205
Wolf
In dieser Beschreibung liegt bereits eine wissenschaftsgeschichtliche Beobachtung,

die zu explizieren und zu erklren ist.
Wir knnen an der Tatsache ansetzen, dass Geistes- und Staatswissenschaften, social
sciences und sciences humaines unterschiedliche Wissenschaftsmodelle darstellen, die
sich nach Gegenstandsbereich und gegenstndlicher Gliederung, nach
Methodenverstndnis und nach ihrem expliziten Praxisbezug signifikant unterscheiden.
Diese Modelle sind als solche im bergang vom 19. zum 20. Jahrhundert
unterschiedlich nach einer gemeinsamen europisch-atlantischen Vorgeschichte bis
zur Mitte des 18. Jahrhunderts und bereits getrennt verlaufenden nationalen
Entwicklungen4 bis zum Ende des 19. Jahrhunderts konstituiert und staatlich
institutionalisiert worden. Dass im Rckblick von der heutigen Situation einer erst
langsam zerbrselnden US-Hegemonie deren Unterschiede de-thematisiert werden,
sollte einem kritischen, wissenschaftsgeschichtlich aufgeklrten Blick nicht lange
standhalten knnen.
Allein schon diese Unterschiedlichkeit dieser imperialen Paradigmen5 der
Wissenschaftlichkeit fr den Bereich von Geschichte, Gesellschaft, Kultur und Staat
stellt aber einen Grund zur Nachdenklichkeit dar. Wenn es denn zutrifft, dass im
bergang von den Vorurteilen des Common Sense zu einer wirklich wissenschaftlichen
Auffassung immer ein epistemologischer Einschnitt erforderlich ist, dann bleiben nur
drei Mglichkeiten, um diese Pluralitt der Paradigmen zu erklren:
Erstens knnte diese Pluralitt gegenstndlich bedingt sein Gesellschaft und
Geschichte in Deutschland, in den USA und in Frankreich wren einfach derart
unterschiedlich,
das
zu
ihrer
Erkenntnis
ganz
unterschiedliche
Wissenschaftskonfigurationen und -paradigmen erforderlich sind, so dass ein
epistemologischer Einschnitt geradezu in diese Pluralitt einmnden msse;
Zweitens knnten wir vermuten, dass in diesem Bereich noch gar kein
epistemologischer Durchbruch stattgefunden htte, so dass die unterschiedlichen
Konfigurationen und Paradigmen als bergangsgestalten auf der Suche nach einem
knftigen epistemologischen Einschnitt aufzufassen wren;
Drittens knnten wir mit den meisten AutorInnen bis 1945 die Auffassung
vertreten, nur eine dieser Konfigurationen und Paradigmen gehe auf einen
Einzelwissenschaften (Psychologie, Soziologie) stark gehemmt, whrend Philosophie, konomie und

Politik im Sinne der neuzeitlichen Tradition generalistisch geprgt blieben (wenn auch mit
vorherrschend analytischen Mitteln).
Die jedenfalls zunchst nicht an Nationalstaaten gebunden waren, wie die Beispiele Schottlands im
18. Jahrhundert und Deutschlands im 19. Jahrhundert deutlich belegen.
Nicht zufllig war jedes dieser Paradigmen an jeweils eine der Mchte gebunden, die im 19. und im
20. Jahrhundert miteinander um die hegemoniale Rolle in der weltweiten Staatenkonstellation
gerungen haben.
206
IST EINE HISTORISCHE EPISTEMOLOGIE DER SOZIALWISSENSCHAFT MGLICH?
epistemologischen Einschnitt zurck und die beiden mit ihr konkurrierenden seien ganz
einfach dadurch epistemologisch berholt.
Alle drei Optionen sind jedoch recht problematisch: die erste treibt den Historismus
und Relativismus auf die Spitze und eliminiert jede Art von Fragestellung nach
gemeinsamen oder bergreifenden Strukturen aus der Wissenschaft; die zweite so
plausibel sie auch empirisch ist: denn deskriptiv ist sie sicherlich schwer zu bestreiten
verweist auf einen knftigen epistemologischen Einschnitt, von dem sie nicht einmal
die Voraussetzungen und Bedingungen spezifizieren kann; die dritte lsst sofort einen
Dogmatismusverdacht entstehen und hat sich in der Zeit von etwa 1890 bis 1945 auch
ziemlich prgnant historisch blamiert.
Wenn wir uns ernsthaft auf das Konzept des epistemologischen Einschnitts beziehen,
dann kann es offensichtlich, sobald dieser einmal erfolgt ist, nicht mehrere richtige
Weiterentwicklungen geben, sondern letztlich, nachdem irrtmliche Anstze eliminiert
worden sind, immer nur eine richtige Linie des wissenschaftlichen Fortschritts selbst
wenn das den Forschenden selber nicht klar ist (und vielleicht auch immer nur
retrospektiv geklrt werden kann).
Nehmen wir einmal an, Marx htte wirklich mit seiner Entdeckung, wie
kapitalistische
Mehrwertproduktion
ohne
Verletzung
der
Gesetze
des
quivalententauschs mglich ist, den epistemologischen Einschnitt erzielt, durch den
auf dem Felde der neuzeitlichen politischen konomie eine wirkliche Wissenschaft
ermglicht wurde, und Freud htte mit seiner Entdeckung des Unbewussten als einer
Erklrungsinstanz menschlichen Handelns und Denkens, die sich dem Bewusstsein
entzieht, fr seinen Bereich Entsprechendes geleistet. Dann wre das
wissenschaftsgeschichtliche Problem, das wir zu klren htten, ein etwas anders
gelagertes: Wir mssten untersuchen, wie es dazu gekommen ist, dass diese
epistemologischen Einschnitte in der nachfolgenden und sich nicht etwa an sie
anschlieenden wissenschaftlichen Entwicklung geleugnet worden sind, und geradezu
Anstze der wissenschaftlichen Entwicklung verfolgt worden sind, die ihre
Weiterentwicklung ausschlieen: konomische Neoklassik und Soziologie auf dem
Felde der politischen konomie und Psychiatrie und Psychologie auf dem Felde der
Psychoanalyse. Es ist nun aber wenig plausibel, in der Entwicklung dieser Disziplinen
in Forschung und Lehre einfach nur einen Machteffekt entsprechender Gruppen von
Herrschenden zu sehen um als wissenschaftliche Disziplinen zu funktionieren,
bedurften sie einer tragfhigen epistemologischen Grundlage. Hier ist sicherlich erst
noch wissenschaftsgeschichtlich zu untersuchen, ob die marxistischen Theorien der
Ideologie und die psychoanalytische Theorie der Rationalisierung dazu in der Lage sind,
diese Entwicklungen zu rekonstruieren und zu erklren. Es ist jedoch m.E. jetzt bereits
absehbar, dass dies nicht mglich sein wird, ohne die wissenschaftsfrmige Spezifik
dieser Entwicklungen deutlich zu artikulieren: Sie sind in sehr unterschiedlicher
Weise jeweils mehr als bloe Systematisierungen des Common Sense oder bloe
Beschreibungen der Oberflche von Freiheit, Gleichheit, Eigentum und Bentham!
(Marx) bzw. des selbstvergessenen Agierens der Subjekte in ihrer Psychopathologie
des Alltagslebens (Freud): Sie erlauben durchaus, etwas zu sehen, wie etwa im
Konzept der Rolle (Mead, Linton) oder auch im Konzept der kognitiven Dissonanz
207
Wolf
(Lewin), was sich in den Mystifikationen der Oberflche und in den Rationalisierungen
des Alltagslebens nicht erschliet und was in den gang und gben Vorstellungen des
Common Sense vllig verborgen bleibt. Louis Althusser (1955; 1963) etwa hat
versucht, diese Art von Erkenntnissen als Momente von Technologien (also eines
bloen Know-how ohne wissenschaftliche Einsicht in die technisch behandelten
Gegenstnde) zu begreifen, zu deren Institutionalisierung er diese Disziplinen erklrte
und sie damit von den zentralen epistemologischen Einschnitten abgesetzt, wie sie fr
Wissenschaften konstitutiv waren. Auch wenn wir heute nicht mehr von einem
historischen Materialismus als konstituiertes wissenschaftliches Gegenstck zu diesen
bloen Technologien ausgehen knnen (vgl. Balibar 1973), knnen wir vielleicht von
oberflchlichen kognitiven Einschnitten reden, um uns auf die zumindest
wissenschaftsfrmige Dynamik dieser Disziplinenentwicklung in Forschung und Lehre
zu beziehen. Dies wrde uns an die Aufgabe erinnern, in der
wissenschaftsgeschichtlichen Aufarbeitung dieser Prozesse niemals die relativen
Erkenntnisgewinne aus dem Auge zu verlieren, wie sie auch unter dem Vorzeichen der
Leugnung und Verdrngung der zentralen epistemologischen Einschnitte erzielt worden
sind.
Diese Entwicklungen deren Spektrum von der Ausdruckspsychologie bis zur
Reflexologie und von der Gruppendynamik bis zur Sozialisationstheorie reicht
haben nicht nur wirkliche Errungenschaften aufzuweisen, in denen ihr (partielle und
sehr ungleich ausgeprgte) Erkenntnischarakter greifbar wird. Sie sind auch deswegen
noch konkreter historisch zu untersuchen, weil sich diese Errungenschaften in aufgrund
staatlicher und kultureller Vorgaben sehr unterschiedlichen Ausgestaltungen
verwirklicht haben. Denn dass diese Negation und diese Exklusion von Marxismus und
psychoanalytischer Theorie von Seiten der institutionalisierten Wissenschaften in den
genannten nationalen Paradigmen unterschiedliche Formen angenommen hat, ist bei
nherem Hinsehen unbersehbar nicht nur wenn wir uns die unterschiedlichen
Gestalten und Verlufe der neoklassischen Revolution bei Walras und Pareto, bei
Jevons und bei Menger (d.h. vor der neoklassischen Synthese) vergegenwrtigen,
sondern auch, wenn wir uns vor Augen fhren, was in Deutschland, Frankreich und den
USA seit 1870 unter Psychologie verstanden wurde6 oder was bis zu Talcott Parsons
hegemonialer Synthese der deutschen und der US-Tradition in Deutschland, in den
USA und in Frankreich unter dem Titel einer Soziologie durchging7
Erst in der neuen Weltordnung nach dem II. Weltkrieg hat sich eine scheinbar
kohrente Vorstellung von Sozialwissenschaften weltweit konsolidiert, so weit sogar,
dass auch noch auf der anderen Seite im Kalten-Kriegs-System jeweils marxistische
Versionen der konomie, der Soziologie oder auch der Psychologie institutionalisiert
worden sind. Auch psychoanalytische Konzeptionen werden in diese Konstellation der
Sozialwissenschaften integriert. Zugleich kommt es allerdings auch nicht nur in den
6
Die schwankende Terminologie (Psychotechnik, verstehende vs. erklrende Psychologie bzw.

allgemeine und angewandte Psychologie) ist nur ein schwacher Indikator der hier auftretenden
Widersprche.
Ich habe selbst noch erlebt, wie eine notdrftig bertnchte Volkskunde in der Tradition von
Wilhelm Heinrich Riehl als Soziologie auftrat.
208
USA in Gestalt der area studies zu einer zunehmenden Inanspruchnahme

interdisziplinrer Wissenschaftsmodelle, durch welche die aufflligen pragmatischen
Defizite der voneinander scharf abgegrenzten Einzelwissenschaften berwunden
werden sollen, was dann allerdings die ungeklrten Grundlagenprobleme keineswegs
einer Lsung nher bringt und allenfalls durch den Rckgriff auf Ideologien des
Common Sense die Wissenschaftlichkeit beschdigt.
Seit der Wissenschaftskritik der 1960er Jahre (vgl. Klver und Wolf, 1972) die sich
(anders als die radical science in der angelschsischen Tradition) zentral auf diesen
Wissenschaftsbereich bezogen hat ist immer wieder deutlich geworden, dass diese
Konstellation problematisch und brchig ist seit den 1990er Jahren mit der
programmatischen Konsequenz, neben den derart disziplinr verfassten
Einzelwissenschaften in der Politikberatung noch eine anders strukturierte mode-2science zu etablieren, die eine problemnhere Arbeit ermglichen soll
(Funtowicz/Ravetz 1993) oder sogar die etablierten Sozialwissenschaften selbst
kaputtdenken zu wollen (Wallerstein 1995).
Das wird jedoch, wie wir wiederum im Rckgriff auf das Konzept des
epistemologischen Einschnitts erkennen knnen, nicht dauerhaft mglich sein, ohne
dass aufgeklrt wre, wie es denn dazu gekommen ist, dass die beidem auf diesem Feld
erzielten epistemologischen Einschnitte von Marx und Freud derart haben negiert und
marginalisiert werden konnten. Das wird sich jedoch nicht ohne eine tiefer greifende
historisch-epistemologische Untersuchung klren lassen: Dazu gehrt die historischkritische Erforschung der Widersprche, in denen sich die unvollendeten
Theorieentwrfe der Pioniere trotz des vollzogenen epistemologischen Einschnitts
immer noch haben bewegen mssen und eine kritische Rekonstruktion der Leistungen
und der Schranken der von Marxismus und Psychoanalyse in sehr unterschiedlichen
Formen verfolgten Strategien der Oppositionswissenschaften. Erst dann kann es
gelingen, einen Prozess des Kaputtdenkens der Sozialwissenschaften und zwar
wiederum gesttzt auf konkrete wissenschaftsgeschichtliche Untersuchungen ihrer
Konstituierungsprozesse einzuleiten, der kein Ruinenfeld zurcklsst, sondern einen
fruchtbaren Boden dafr schafft, dass die epistemologischen Einschnitte, wie sie
Marx und Freud vollzogen hatten, wirklich zum Ausgangspunkt einer
wissenschaftlichen
Behandlung
des
gesamten
Feldes
der
historischen
Gesellschaftswissenschaften werden.
Eine derartige Aktualisierung der epistemologischen Einschnitte, wie sie Marx und
Freud vollzogen haben, bedarf selber inzwischen einer wissenschaftsgeschichtlichen
Untersuchung ihrer Voraussetzungen. Denn auf der Seite der historischen Entwicklung
von Marxismus und Psychoanalyse ist wiederum zu konstatieren, dass sich hier eine
Pluralitt der Entwicklungslinien herausgebildet hat, wie sie dem Gedanken des
epistemologischen Einschnitts diametral widerstreitet. Und auch hier bieten sich
logischerweise dieselben Verarbeitungsmuster an: objektiver Historismus, subjektive
Relativierung und Dogmatisierung mit eben den gleichen offensichtlichen Nachteilen.
Nur mit dem wichtigen Unterschied, dass die groen epistemologischen Einschnitte
innerhalb dieser Entwicklungen noch zu finden sind, so sehr sie auch inzwischen
verschttet sind. Aber zumindest dieses sollten wir aus wissenschaftsgeschichtlichen
209
Wolf
Studien von epistemologischen Renaissancen gelernt haben: Es ist gar nicht mglich,
einfach zurckzugehen und heute gleichsam die Abzweigung zu nehmen und den Pfad
zu betreten, die damals nicht genommen und nicht begangen worden sind also in den
1890er Jahren die Nichtrezeption von Marx durch die Kathedersozialisten und im
ersten Jahrzehnt des 20. Jahrhunderts die Nichtrezeption von Freud durch die etablierte
Psychiatrie. Also auch hier bedarf es der wissenschaftsgeschichtlichen Dekonstruktion
der Formen, in denen sich in einer inzwischen mehr als einhundertjhrigen Entwicklung
in der Gestalt von alternativen Wissenschaften Marxismus und Psychoanalyse die
Weiterarbeit an dem durch die groen epistemologischen Einschnitte erffneten
Kontinent mit ihrer Verkennung und Verdeckung verbunden hat. Daraus ergibt sich
m.E. weder eine unverbindliche Pluralisierung, in der der Anspruch des bereits
vollzogenen groen epistemologischen Einschnitts aufgegeben wrde, noch eine
relativistische Bescheidung, die auch noch auf einen knftigen epistemologischen
Einschnitt verzichten wrde (im Sinne der missverstandenen Devise Das Reale ist das
Unmgliche!, die zwar fr die Therapie, nicht aber fr die Wissenschaft gilt), noch ein
neuer Dogmatismus, der die eigene, unvermittelte Position zur absoluten Wahrheit
erklrt.
Sptestens an dieser Stelle muss ich einen Joker ins Spiel bringen, den ich bis jetzt
mit Flei und ganz bewusst zurckgehalten habe: die Philosophie. Denn was die
Aufgabe und die Ttigkeit von PhilosophInnen ist, fllt nicht nur in den
unterschiedlichen Paradigmen, auf die ich hingewiesen habe, ganz charakteristisch
unterschiedlich aus und es ist auch in den differenten Traditionslinien von Marxismus
und Psychoanalyse vllig unterschiedlich bestimmt worden. Vor allem aber lsst sich
die Philosophie in allen diesen unterschiedlichen Gestalten unter dem Gesichtspunkt des
epistemologischen Einschnitts bereinstimmend negativ kennzeichnen: Fr die
Philosophie ist es offenbar nicht bestimmbar, was in ihr als ein epistemologischer
Einschnitt zu begreifen wre die unterschiedlichen Versuche (zeitlich in etwa parallel
zu den anderen epistemologischen Entwicklungen, auf die wir uns hier bezogen haben),
endlich eine Philosophie als strenge Wissenschaft zu etablieren, sind allesamt
gescheitert. Und die ltere, etwa von den Ideologues,8 den Junghegelianern,9 Marx
und Engels und dem klassischen Positivismus vertretene Vorstellung, deswegen endlich
mit der Philosophie Schluss machen zu knnen, hat erst recht nicht funktioniert
sondern allenfalls dazu beigetragen, immer wieder neue Gestalten philosophischer
Ttigkeit in Gang zu bringen.
Wenn es aber zutrifft, dass sich das Philosophieren weder beenden, noch durch einen
epistemologischen Einschnitt verwissenschaftlichen lsst, dann folgt daraus, dass wir
es in unserer kognitiven Orientierung immer auch mit etwas zu tun haben, was sich
nicht restlos aufklren lsst. Darin liegt dann wiederum eine unbestreitbare Quelle
8
Deren Pionierarbeit auf dem Gebiet einer epistemologischen Strukturierung des Feldes der sciences
morales et politiques deswegen weitestgehend in Vergessenheit geraten ist, weil sie sie vor den
groen epistemologischen Einschnitten, etwa im Kontext dessen, was Marx als Vulgrkonomie
denunziert hat, angesetzt haben.
Die zumeist deswegen berschtzt werden, weil Marx sich in seinem epistemologischen Einschnitt
sie berwindend auf sie bezogen hat.
210
legitimer Pluralitt, die wir nicht mit dem Hinweis auf bereits erfolgte oder knftige
epistemologische Einschnitte abtun knnen. Mir ist es hier aber vor allem darum
gegangen, nachvollziehbar zu machen, dass aus der Einsicht in die Irreduzibilitt
philosophischer Pluralitt auch auf dem Felde der Wissenschaften, die sich mit
Geschichte, Staat, Kultur und Gesellschaft befassen, nicht der Verzicht auf tiefe und
grundlegende epistemologische Einschnitte begrnden lsst und dass es von hier aus
denkbar und begrndbar ist, durch die Untersuchungen einer historischen Epistemologie
die Voraussetzungen dafr zu schaffen, dass die Vorhaben eines Kaputtdenkens der
Sozialwissenschaften und einer wissenschaftlichen Erneuerung von Marxismus und
Psychoanalyse keine leeren Gesten bleiben, sondern in der doppelten Bewegung einer
wissenschaftsgeschichtlichen Untersuchung der pluralen Entwicklungen in den
etablierten ebenso wie in den alternativen Wissenschaften auf diesen Feldern
konkret ausgearbeitet werden.
Zum Schluss noch eine strende Frage: Fhrt mein Insistieren auf dem
epistemologischen Einschnitt nicht zu einer heillosen Akademisierung der Praxis von
Marxismus und Psychoanalyse? Darauf kann ich nur antworten, dass ohne den
Rckgriff auf einen solchen epistemologischen Einschnitt diese Praxis letztlich in den
Illusionen des Common Sense befangen bleibt und keine bewusst befreiende Praxis
werden kann. Und es geht eben nicht um akademische, sondern um streng
epistemologische Mastbe denn ohne wirkliche Wissenschaft gibt es auch keine
wirklich radikale Praxis.
Literatur
Althusser, Louis (1955). Sur lojectivit de lhistoire. Lettre Paul Ricur, Revue de
lenseignement philosophique 5/4: 3-15 [wieder in: ders. (1998), S. 17-31].
Althusser, Louis (1963). Philosophie et sciences humaines, Revue de lenseignement
philosophique 13/5: 1-12 [wieder in: ders., (1998), S. 43-58).
Althusser, Louis (1998). Solitude de Machiavel, hrsg. von Yves Sintomer, Paris.
Badiou, Alain (2012). The Adventure of French Philosophy, London.
Balibar, Etienne (1963). Cinq tudes du matrialisme historique, Paris.
Funtowicz, Silvio O., und Ravetz, Jeremy R. (1993). The Emergence of Post-Normal
Science, in: R. von Schomberg (Hg.), Science, Politics, and Morality,
Dordrecht.
Klver, Jrgen, und Wolf, Frieder Otto (1972). Wissenschaftskritik und sozialistische
Praxis, Stuttgart/Bad Cannstatt.
Wallerstein, Immanuel (1995). Die Sozialwissenschaft kaputtdenken, Neuwied.
211
LIFE IN SHAKESPEARES SONNETS
Anselm Haverkamp
Whose Life? Georges Canguilhem has put the crux most succinctly: Do we proceed from knowledge to
the concept of life, or is it not, rather, the other way round, and we come from the non-conceived unconceptuality of life and living to the conception of things and worlds? If bio-logical life has, and follows,
a logic of its own, whose life is it that we are about to endure and, at best, are able to watch and reflect
upon while living along? The threshold crossed by Darwin was, among other things, history in its old
sense of man made progress; instead, in Darwin an abyssal deepening and radicalization took place, and
a truly new historicism came about which was to transcend in an un-preconceived manner the
transcendental world and the salvation histories that came with it. Life, in other words (and very crudely
put, at that), had no future any more but was to be grasped, if ever (and only metaphorically so), in the
grammar of a second future, in a futurum exactum always already accomplished. Is it ours to fall, more
or less exactly, in a variety open to development, into place? Like the question of capital Being, that fell
under erasure for a short while, Life became a categorically forbidden (and forgotten or repressed)
territory whose infra-structure, hard to trace, but halfway decoded by now along the lines of so called
informational processes, seems to offer nothing more or less than a blind mirror for what the manifold
ways of former forms of life and life expectancy were about. Almost self-evidently, life has turned out to
be some in-explicable type of Being in time, but without anything that would deserve the name of an
evidence in the older sense of a shared experience an unshared non-experience, and the most graphic
instance of what Stanley Cavells reading of Shakespeare has named disowning knowledge. Since
Life in itself is not self-evident and does not lend itself to evidencing, it asks for a different type of
questions. Do we have any other lives outside of the quotation marks that cite the play of differences on
the surface of what we live through? Science maybe nothing but and that, it seems, means already
everything the most indulging mirror stage of human understanding. That wouldnt be so bad, a little
vanity for a kill time, if it were not for the death drive and those politically attracted to it. Thus we come
to face landscapes of life ever since the word landscape came up in modernity as a cultivated field of
perception and conception, the playground of scientific mirroring or, as it came to be called, representation on a stage upon which Benjamins melancholy observer meditates the empty shells drained
from lifes transcendence. Whiteheads bifurcation of nature reaches its full scale in the unbridgeable
bifurcation of life between good old, by now lost, deluded bio-graphical life and the bare life of
Benjamin and Agamben, subjected to, and ready for, new types of violence not only, but to a bio-logic
free from a meta-physical and, that is, epistemological question, for once, and no ethical one, whose
place pragmatic epistemology, no ethics is the unknown as in a cloud of unknowing unknown and
avoided topos of Life in Literature.
These preliminary remarks refer to a much larger project, whose prehistory I attempt to sketch out
here in a provisional, though exemplary fashion. The project as a whole has been outlined in a
companion piece under the heading A Narrow Thing Within One Word (a phrase from Empsons
Structure of Complex Words) and put in the historical perspective of a Foreclosure of Nature in
(precisely) Post-Shakespearian Worlds and Times. For this larger project of an epistemological
history of literature, the following may serve as a prolegomenon. In short, Shakespeare marks a
historical threshold in the emerging epistemology of (a new meaning of) life, whose early format is
here to be viewed; the emphasis is on the emergence of the new conception of life as opposed to an
explicit expression in, or occurrence of a life-related semantics. Nothing seems (or needs) to be
expressed, but a lot has become pressing and emerges in the medium call it the form of the
Literary.
213
Haverkamp
It is impossible to avoid Shakespeare in these matters.1
One way to read my title is to take it like one of the chapter entries of William
Empsons Structure of Complex Words Sense in Measure for Measure, for
example.2 The next step would be to investigate the relationship of Sense and Life
in the light of Evelyn Fox Kellers title Making Sense of Life a title much more
difficult than meets the eye at first glance, since the normal senses of making sense
have come to an end with the new sense of life in the life sciences sciences no
longer content to be natural sciences in the older sense of still reading, after all, the
book of nature.3 For a while, we enjoyed the intricate delusion of witnessing the
transition and, as it were, the withdrawal, of functions of living from the visible world
to an anatomically invisible realm, as described in Michel Foucaults discussion of
Bichats treatise on membranes of 1799.4 At the same time, when Foucault began to
historicize to an unheard of degree the disappearance of life under the hands of biology,
Franois Jacob had proposed a logic of the living (no longer of life no concept of
Life as such left) a logic utterly beyond the old order of things (Foucaults in 1966,
Jacobs in 1970).5 The enormous changes registered and put into new order by Foucault
and Jacob is a manifestation of what I find most convincingly stated in Alfred North
Whiteheads bifurcation of nature in 1920. Whitehead proposed nothing less than an
unbridgeable, though seemingly unfelt unfelt but abyssal gulf between thought
about nature and the sense-perception of nature as a perception, in fact, which
resides in nature itself only insofar as the (mere) fact of sense-perception has (and
must have) an ingredient or factor which is (as a matter of fact) not thought. And
Whitehead enforces the emphasis on what there is not by (explicitly) saying that nature
is closed to the mind.6
This radical closure of nature to thought of nature as the principally un-thought
implies a new type of historicality for epistemology, which escapes the usual histories
and asks for a new discipline of what had never before been conceived of as historical,
William Empson, Seven Types of Ambiguity, London: Chatto and Windus, 1930, revised ed. New
York, NY: New Directions, 1947, p.49 (chapter II, on the Sonnets, pp.50-57).
William Empson, The Structure of Complex Words, London: Chatto and Windus, 1951, ed. Jonathan
Culler, Cambridge, MA: Harvard University Press, 1989, p.270 ff.
Evelyn Fox Keller, Making Sense of Life: Explaining Biological Development with Models,
Metaphors, and Machines, Cambridge, MA: Harvard University Press, 2002, p.130 ff.
Michel Foucault, Les mots et les choses , Paris: Gallimard, 1966; id., The Order of Things: An
Archeology of the Human Sciences, New York, NY: Pantheon, 1971, chapters 5 and 8.
Franois Jacob, La logique du vivant: Une histoire de l'hrdit, Paris: Gallimard, 1970; id., The
Logic of Life [sic!], Princeton, NJ: Princeton University Press, 1993.
Alfred North Whitehead, Concept of Nature, Cambridge: Cambridge University Press, 1920, p.4, p.5,
p.13 (my emphasis and parentheses). The historical dates of the epistemological development are by
no means unimportant, although I cannot do justice on this occasion to the intricate interlacing of
epistemological and philological concerns (Whiteheads and Empsons Cambridge of the 20s), an
interplay implicitly postulated in Foucaults design (Canguilhems Paris of the 60s), but far from
developed in his own account.
214
namely, an historical epistemology.7 What is to be faced with Whiteheads concept of

a closure is to an unexpected degree the question of what it actually is that is known
to us as the direct deliverance of sense-awareness and (let me add on the occasion)
that includes records registering or expressing such deliverance. Although Whitehead is
not interested in aesthetics, he implies the mediating role of a sense-perception, which
masks on the blind flipside of its facticity a certain, if essentially inarticulate latency,
whose impact is to remain as problematic as it remains incommunicable. It seems
there lies an irony buried in the sciences empirical nature, which amounts to nothing
less than an aesthetic self-misunderstanding of the sciences aesthetic in Paul de
Mans sense of an aesthetic ideology and even the dialectics of enlightenment (de
Mans model case Adorno) may turn out to be nothing but the side-effect of the same
deep delusion of the scientific ages vis--vis the closure of nature.8 Whiteheads
Concept of Nature offered a first, epistemologically sober, mathematically informed
and, in this new sense, meta-physical refutation of the (in Bacons early sense of a
critique of ideology) idola fori of a self-deluded scientific self-esteem that left many, all
too many traces on the surface of the literary public sphere ever since, the concept of
life the most enigmatic among them. As a crux in its own secret right, Foucault could
not help naming it, if only in passing on his way to the new strands of bio-power, life
in general while confirming thereby the entry of life into history, that is, the entry of
phenomena peculiar to the life of the human species into the order of knowledge 9
Shakespeares Sonnets offer some rather unexpected insight in the emergence of the
new meaning of life, whose entry into history is registered by Foucault only as part of
the accomplished bio-political agenda. Like the foremost paradigm of the preceding
process of an emergence that accompanies, even mirrors, to a certain extent, the
emergence of life and brings it into the new political state of governmentality, statistics,
the phenomena named by Foucault as peculiar to the life of the human species had
to be recognized somehow, if not as such.10 While already probability itself ran contrary
to all expectations of a straight truth, life seemed to remain uncannily the same. If there
was an exemplary instance in this new grammar of an emerging order, it could not be
more adequately conceived of than as the probable. As such probability is a first result
of the bifurcation within the Concept of Nature, on whose tabula rasa a new mode of
history emerged emergence is this new mode of history, and lifes entry into history
becomes the founding figura of a new history without implementum. While it appears
easy to trace the new historicality of emergence in the technicality of probability, the
new sense of life remains in the unthought realm of the bifurcated nature.
7
10
The term according to Hans-Jrg Rheinbergers Historische Epistemologie zur Einfhrung, Hamburg:
Junius, 2007; see his seminal study Towards the History of Epistemic Things, Stanford, CA: Stanford
University Press, 1997.
Paul de Man, The Resistance to Theory, Minneapolis, MN: University of Minnesota Press, 1986, p.7
ff.
Michel Foucault, La volont de savoir, Paris: Gallimard, 1976; id., The History of Sexuality, vol. I
[skipping the decisive French title!], New York: Random House 1978, pp.141-142. The sentence
continues with what is from now on uppermost on the authors mind: of knowledge and power, into
the sphere of political techniques.
Ian Hacking, The Emergence of Probability: A Philosophical Study of Early Ideas about Probability,
Induction and Statistical Inference, Cambridge: Cambridge University Press, 1975, chapter 4.
215
Haverkamp
There had been always a certain split within the concept of life, which had only
recently deepened to the late medieval divide of individual and collective (in that sense
general) eschatology of man.11 From the start, both of these eschatological tendencies
of life had strengthened in their early modern differentiation the cultural format of life
as opposed to, and ultimately overcoming, the natural state of life that was to re-emerge
and surface from the un-thought bio-logic of the living ground of both individual and
collective ways of living, in the sciences of life. Thus one first and immediate
misunderstanding has to be avoided: we cannot be interested here in Shakespeares
taking part in the collective eschatology of his life-time, as little as we can be interested
in the more or less individual vicissitudes represented, expressed or articulated in the
Sonnets. On the contrary, not what is manifest, but what is latently pressing within the
un-thought part of living nature, as far as it surfaces in the intricate semantic setup of
Shakespeares texts, must be on our mind. Thus, to put it in the very briefest possible
terms, I am inclined to read and identify in Empsons unreasonable simplicity (i.e.
the whole charm of the poem, no lesser phenomenon) Blumenbergs un-conceptual
coming to grips with Whiteheads un-thought nature.12 Life, it shall turn out in various
attempts of 20th century philosophies of life, is to become the problematic center or
even agency of natures foreclosed being. The state of affairs in Blumenberg the
unconceptuality of the unthought nature of life has been confirmed and radicalized to
an essential ambi-valence, which in the account of Fox Keller characterizes the
riddle of life for biology, and it is this ambi-valence, in which the neologism of the
Gene not quite a metaphor, Keller adds operates as a riddle in and of itself.
Explicitly quoting the riddle according Aristotles theory of metaphor, Poetics 1458a
26-27 (and in Derridas reading, at that), Keller points to a deeper metaphorological
dimension of the unconceptual emergence of life, in which Empsons sense of the
surface ambiguities in Shakespeare seems to converge with the unthought nature of
Whitehead and, that is, with the fact that sense-perception has an ingredient or factor
which is not thought (Whitehead as cited above).
In the Sonnets, there is more of that un-conceptuality than we would expect and find
documented in the relevant secondary criticism. Coming to think of it, there it seems to
be almost avoided: the whole predicament of finding a biographical setting in order to
be able to talk at all about what happens in these texts attests to it. They are like
desperately read for an individual eschatology of inclinations and vicissitudes, to be
matched by some collectively shared social horizon or world picture. Among the
intimations, the so-called procreation topic of the first 17 sonnets (left behind in the
18th) is a telling symptom of the perplexed state of affairs, if only for the reason that the
term procreation itself is not to be found in Shakespeares lexicon (Onions lists one
use of the verb), and thus is an entirely modernist, biologically minded imputation. As
11
12
For this distinction see Hans Blumenberg, Die Legitimitt der Neuzeit, Frankfurt/Main: Suhrkamp,
1966, p.39 ff.; id., The Legitimacy of the Modern Age, Cambridge MA: MIT Press, 1975, part I.
Unconceptuality according to Hans Blumenbergs Ausblick auf eine Theorie der
Unbegrifflichkeit (1979), in: id., sthetische und metaphorologische Schriften, ed. Anselm
Haverkamp, Frankfurt/Main: Suhrkamp, 2001, pp.193-209; the existing American translations of this
essay, say, in the appendix of his Shipwreck with Spectator, Cambridge MA: MIT Press, 1996, are
unfortunately not adequate.
216
an insinuation of sorts it lends itself to the most trivial of aims, namely, of persuading a
young gentleman to marriage (Erasmus title), and thereby brings about, as
unnecessarily as it could be, an awkward conjunction of 19th century petit-bourgeois
intentions with that times most grotesque ambition of literary genius. The pragmatics
of Erasmus manual notwithstanding (entirely possible as pretext or satirical backdrop),
the discrepancy vanishes, if we take the mis-named procreation issue as part of a
different, by no means unknown and by no means non-political agenda, whose
predecessor is the cosmological economy of power and glory, recognized only
recently by Giorgio Agamben as instrumental for the advent of the governmentality to
come.13
Like the Virgin Queens glory, the proposed glory of the procreation sonnets is
modelled on a cosmological layer, which had come to serve, since Augustine, as an
embellishment and ontological support of the onto-theological world picture.
Interestingly enough, the Greek cosmos had never vanished completely but remained
present under various covers. Not only did the old cosmos come to a new political role,
this roles post-theological function thrived in the advent of the emerging regimes of
government. No doubt, it also influenced some new, though unclear notions of a procreation avant la lettre, which we find in Shakespeare and elsewhere (and in him not
different from elsewhere).14 The Sonnets point, however, is different. It does agree with
the modes of Elizabethan representation and seems even particularly fitting, if we
include in this picture a likely noble addressee, but this seems, at the same time, entirely
beside the poetical point. Sex, after all (as was not unknown), is a diversity generating
device a genetic lottery, so to speak, beauty not included.15 The lifes gloria, to
which the celebrated addressee is to live up, is not his own (nor his Queens either), but
the creator gods way of being a cosmo- rather than ontological given, which is
already present in Canterburys cosmo-theological proof of Gods existence even
beyond the mythic drama of creation itself and entirely bound to His glorious
completion alone.16 Sonnet 18 offers an unbelievably beyond belief wonderful
transposition and illustration. It is this sense of the fullest completion, which startles and
necessarily defies the poets ambition in sonnet 16. The full glory, however, which
defines this post-creational Gods onto-theological being in time, is given in and as
signature (as Agamben explains in his next book) as a mark of the gloriously untensed way of being, to whose participation the young beauty is to be brought or so it
seems at first, in the necessary delusion of the generational vanity. The literal
transcription of the cosmological signature into the timely process of the noble
generation of the young mans beauty in his offspring is not just another case of
reformed literalization; the book of nature is no longer the twin of the bible. The God of
the two Testaments, whose mythic identity had been a permanent allegorical
13
14
15
16
Giorgio Agamben, Il Regno e la Gloria: Per una genealogia teologica delleconomia e del governo,
Vicenza: Neri Pozza 2007, chapter 8. Agamben discusses for his purpose Leonard Lessius (1620).
Horst Breuer, Theories of Generation in Shakespeare, Journal of European Studies 20 (1990): 325342.
Franois Jacob, The Possible and the Actual, Seattle WA: University of Washington Press, 1982, p.8.
See Hans Blumenberg, Arbeit am Mythos, Frankfurt/Main: Suhrkamp, 1979; id., Work on Myth
Cambridge, MA: MIT Press, 1988.
217
Haverkamp
challenge the Gordian knot cut down to the literal in Reformation had turned into a
God remote even from Genesis. Nature is his thought, but glory the only manifestation
of this thought and, that is, close enough (for our purpose here) to the closure of a nature
unthought by the species unwittingly involved in natures course.
The older bifurcation of life, abysmally deepened in the bifurcation of nature to be
faced with Whitehead, had a well-known prehistory in the two Greek words for life,
bios and zoe. The surfacing of this divide in sonnet 16 is an illustration illustratio in
the strictest of Renaissance senses of Life as a riddle in and of itself (in Kellers apt
quote from Aristotle). Stephen Booth, most Empsonian of commentators, speaks of this
passage, the 3rd quatrain of sonnet 16, as a devils puzzle for an editor, without
naming the crucial notion, the life-word in question.17 Neither he, nor Empson in Seven
Types, more than sense the fundamental shift in the line: So should the lines of life that
life repair. Empson does register the grammatical ambiguity, which defines the second
type and makes Shakespeares Sonnets a superb case study. He does not recognize,
however, the bifurcation that surfaces in the life-words duplicity (10 years after
Whitehead, same Cambridge), although he identifies the essential ambi-valence,
which in Fox Kellers analysis informs the riddle of life in biology and in her pointed
hyphenation of the ambi-valence surpasses Empsons surface ambiguity. In both
Empson and Booth, the analysis points to an ambivalence that might as well be
described as metaphorical in the conventional Aristotelian sense: The way of life
denoting that life in Empsons close reading had been of old (in Aristotles Greek)
transferable to lines of life, which remained the bio-graphical traces of bios (Aristotle
offers even a category of metaphor for this case). Biological zoe remained in reach of
bios, as much as it might reach painfully into bios.
Booths extremely perceptive construction of the So that opens the line understands
the resuming function of this entrance but bemoans, sensitive as he has become as an
editor, the loss of cognitive content that inevitably comes with the result: that a reader
will see the point without understanding (or knowing that he has not understood and
cannot in any usual sense understand) the sentence that makes the point. It is here,
where Kellers analysis of the Genes neo-logical capacity to conceive of life beyond
Aristotles meta-phorical categorization hits the crucial point: the active tension of
two metaphoro-logical orders amalgamated in one new name. Shakespeares riddled
line takes the amalgamation apart while preserving the opaquely split old name. As the
post-theological glory of the higher lines of life makes palpable a glory, whose cosmic
quality is to be implemented in sonnet 18s eternal lines the poets genre and
traditional office, singer of sonnets, takes up and investigates a principle of structure, in
which the hidden double force of life as lived living appears and is offered as the
structure of a para-doxical appearance (far from orthodox), which is hard to distinguish
from the opacity of nature diagnosed by the mathematician of the 20th century.
Shall I compare thee plays skillfully with the metaphorological mechanism that was
the rhetorical ornatus as a mirror of the kosmos in its relation comparison to the
17
Stephen Booth, Shakespeares Sonnets, ed. with analytic commentary, New Haven, CT: Yale
University Press, 1977, Preface, p. xii ff.
218
logos.18 Sonnet 18 includes the heavenly riddle in and of itself, whose old description
had been the natura naturans, but whose secret force is life. The ancient proportion of
kosmos and logos is refigured in Shakespeare as in a lot of Renaissance, but that is not
it. What needs to be perceived is the mask, under whose guise the disappearance of
nature rather than of God would enforce, with forcd fingers rude (no longer
Shakespeare), a new conception of time: in eternal lines to time thou growst is the
prediction. The time, whose eternal lines biological life follows, is de-tensed time, as
recent analytical lore calls it.19 In retrospect, taking Darwin for granted (Darwin without
the disasters of Darwinist misreadings), a sonnet design that respects the un-known
within nature clears up. Analytically speaking, Darwin had diachrony collapse, like
Saussure a little later, into the synchrony of variation and differences. At about the same
time, mad old Hlderlin took up, in a last of his superb poems of old age, the lines of
life motif and brought it to the new state of affairs: The lines of life are variation (Die
Linien des Lebens sind verschieden). The conception of life that announces itself in the
Sonnets and articulates early on the un-thought of nature is a poetics of variation not
evolution, not generation, both of them fallacies of a bygone mythic timing. Something
like this may have been on Georges Canguilhems mind, when he asked, in his lecture
of 1966 (his student Foucault was on his way): Do we come from knowledge to the
concept of life, or is it not, rather, that we come from the un-conceptuality of living to
the conception of un-thought life in nature?20 (Canguilhem was a reader of Whitehead;
the paraphrase is mine.)
18
19
20
Hans Blumenberg, Paradigmen zu einer Metaphorologie, Bonn: Bouvier 1960, ed. Anselm
Haverkamp, Berlin: Suhrkamp 2011, p. 12 (with commentary ad 12 f.).
L. Nathan Oaklander and Quentin Smith, eds., The New Theory of Time, New Haven, CT: Yale
University Press, 1994, Smiths introduction.
Georges Canguilhem, La nouvelle connaissance de la vie [1966], in: id., tudes dhistoire et de
philosophie des sciences, Paris: Vrin 1968, pp. 211-225, on p. 214. See the companion piece The
Concept of Life [1966] in the unfortunate patchwork edition of A Vital Rationalist: Selected Writings
from Georges Canguilhem, ed. Franois Delaporte, New York NY: Zone Books, 1994, pp. 303-319,
on p. 318.
219
Haverkamp
Sonnet 16
But wherefore do not you a mightier way
Make war upon this bloody tyrant time?
And fortify yourself in your decay
With means more blessd than my barren rhyme?
Now stand you on the top of happy hours,
And many maiden gardens yet unset,
With virtuous wish would bear your living flowers,
Much liker than your painted counterfeit.
So should the lines of life that life repair
Which this times pencil or my pupil pen
Neither in inward worth nor outward fair
Can make you live yourself in eyes of men.
To give away yourself keeps yourself still,
And you must live, drawn by your own sweet skill.
Sonnet 18
Shall I compare thee to a summers day?
Thou art more lovely and more temperate:
Rough winds do shake the darling buds of May,
And summers lease hath all too short a date;
Sometime too hot the eye of heaven shines,
And often is his gold complexion dimmed;
And every fair from fair sometime declines,
By chance or natures course untrimmed:
But thy eternal summer shall not fade,
Nor lose possession of that fair thou owst,
Nor shall death brag thou wandrest in his shade,
When in eternal lines to time thou growst.
So long as men can breathe or eyes can see,
So long lives this, and this gives life to thee.21
21
Shakespeares Sonnets, ed. Stephen Booth, New Haven, CT/London: Yale University Press, 1977,
p. 16, p. 19.
220
ABSTRACTS OF CONTRIBUTIONS
LPISTMOLOGIE HISTORIQUE DANS LITINRAIRE INTELLECTUEL

DE GEORGES CANGUILHEM
Camille Limoges
Je veux dans cette communication interroger lvolution des proccupations manifestes
dans la production intellectuelle de Georges Canguilhem, de 1926 jusqu lanne de
son dcs en 1995, et plus particulirement jauger la place quy occupa
l pistmologie historique .
Pour ce faire, je montrerai comment, deux gards, son uvre peut tre utilement
priodise.
Ainsi :
1) La production publie de Canguilhem manifeste au cours du temps des
dplacements significatifs de ses domaines dintrt et de leurs poids relatifs;
2) Une scansion assez nette apparat aussi dans les positions thmatises par
Canguilhem, notamment pour ce qui est de ses proccupations pistmologiques, de la
mise en forme des thses de lpistmologie historique, de mme que de son
engagement en histoire des sciences.
Jentends montrer enfin comment, malgr ses transformations, luvre demeure
gouverne par lunit profonde dun projet qui demeure de part en part un projet
philosophique.
De l'histoire des commencements celle des origines
Franois Delaporte
Dans son dernier livre, Secrets de femmes, Le genre, la gnration et les origines de la
dissection humaine (Les presses du rel, 2009) Katharine Park, crit :
En tenant compte de lensemble de ces pratiques au lieu den isoler la dissection
acadmique, jai voulu rtablir leur cohrence culturelle. Cest l un point fondamental:
en postulant de manire anachronique que louverture de corps humains fut en premier
lieu une procdure mdicale, les historiens ont mconnu le phnomne plus large au
sein duquel elle a merg ou relgu ces autres procdures au rang d arrire plan
ou de contexte culturel . Je considre au contraire louverture de corps humain
comme un tout. Ses variantes (la dissection proprement dite, lembaumement,
lautopsie, lexcision de ftus, la reconnaissance ou inspection des cadavres de saints
hommes et femmes) sont comme une srie de miroirs angulaires qui se rflchissent et
221
ABSTRACTS
sclairent les uns les autres. Aucune (variante) nest premire, et la dissection (qui fut
tous gards la plus sotrique) moins que les autres. Pour mettre en vidence leurs
points communs et leur degr dassociation dans lesprit des contemporains, jai utilis
les mots dissection et plus volontiers anatomie pour les dsigner toutes, sauf
quand lexigence de clart a rclam un terme plus prcis (page 13).
Quatre remarques indiqueront lordre dexposition des questions que nous souhaitons
poser.
1. Ce texte programmatique prsente une critique de lhistoire des historiens de la
mdecine: ils ont commis le pch danachronisme en postulant que louverture de
corps humains fut en premier lieu une procdure mdicale . Katharine Park sen tient
lide, centrale, dune cohrence culturelle . Mais pourquoi une histoire qui prend
pour sujet dtude la dissection acadmique serait-elle anachronique?
2. Pour Katharine Park, les historiens ont valoris la dissection acadmique et
relgu toutes les autres procdures au rang, dit-elle, de contexte culturel . Elle fait
linverse en rtablissant la prsance du contexte culturel au dtriment de la dissection
anatomique. Mais pourquoi faudrait-il ncessairement penser lhistoire dune pratique,
que ce soit pour len affranchir ou pour ly inscrire, dans son rapport un
contexte culturel?
3. Cette cohrence culturelle, qui marque la prdominance du contexte culturel,
trouve sa pleine signification dans le fait que, pour Katharine Park, l ouverture du
corps humain est un tout . Cest dire que lhistoire de louverture de corps humain doit
tre faite comme lhistoire dun tout. En tmoigne, le titre si loquent de son livre:
Secrets de femmes. Le genre, la gnration et les origines de la dissection humaine. Il
ne faut pas sy tromper : Les origines de la dissection humaine recouvrent une srie
de pratiques diffrentes dont le seul trait commun consiste dans un geste douverture.
Mais quest-ce donc quune histoire qui se donne pour objet dtude la dissection
humaine? En loccurrence, louverture du corps humain comme un tout?
4. Pour Katharine Park, on se trouve en prsence dune srie de variantes de
louverture qui doivent tre apprhendes, telles quelles apparaissent sous langle de
leurs points communs et de leur degr dassociation dans lesprit des contemporains .
Pour elle, lhistoire des pratiques devient lhistoire de la manire dont les contemporains
des origines de la dissection humaine se la reprsentent. Mais nest-il pas possible de
faire une histoire des pratiques qui serait une histoire de ce que font rellement les
hommes, et non celle de ce quils croient faire?

The Trigger of a Ruptured History in Bachelard
Monika Wulz
This paper focuses on Gaston Bachelards concepts of the instant, the interval, the
break, the rupture. It presents these concepts as non-temporal moments and non-spatial
realms of nothingness where an unexpected intervention of disturbances takes place and
a deviation of processes can happen. This discontinuous and incidental confrontation is
effective as a trigger for generating a framework of attention and thus temporality.
222
ABSTRACTS
In the paper, I compare this constellation of nothingness, unexpected confrontations,

and the emergence of temporality in Bachelards metaphysics and epistemology to
Louis Althussers conception of an aleatory materialism that is based on contingent
encounters of isolated elements in the void and to Hans-Jrg Rheinbergers emphasis on
the unforeseen events in an experimental process. Based on these three accounts, the
paper aims at developing concepts of temporality, history, and historiography that rely
on contingency and unexpectedness.

The Structure of History According to Georges Canguilhem and Helmuth Plessner
Thomas Ebke
My presentation will revolve around the idea of history that inheres the specific type of
historical epistemology advocated by Georges Canguilhem. Drawing on Canguilhems
essay Lobjet de lhistoire des sciences (1966) as a starting point, I will be concerned
with the question of what may unify the distinct areas of object formation specified by
Canguilhem. On closer inspection, it may well be the case or so I argue that
Canguilhems vision of historical epistemology is torn apart by a dualism that leaves the
historian of science in a position that is at the same time immanent and transcendent.
What is more, my critique will not have a frontal character. Rather I will offer a
reading of Canguilhem through the eyes of an author who knew as little of Canguilhem
as Canguilhem knew of him. The author in question is Helmuth Plessner, proponent of a
philosophical anthropology (Philosophische Anthropologie). Thus, by inserting
Canguilhems position into Plessners, my talk will juxtapose more broadly the projects
of an historical epistemology and a philosophical anthropology with respect to the
structure of history.

Around the time of his flight from Nazi Germany in 1933-34, Kurt Goldstein wrote Der
Aufbau des Organismus, a major theoretical work drawing on some thirty years of his
neurological and psychiatric studies. This paper aims to demonstrate that the novelty of
Goldsteins argument in The Organism resides in its epistemology of life and its
philosophical anthropology; that it is based on a particular and novel theorization of
individuality; and that as a result, the work formed part of a significant revision of
physiology that took place in the 1920s and 1930s.
223
ABSTRACTS
GEORGES CANGUILHEM
pistmologie historique et/ou histoire philosophique
Claude Debru
Georges Canguilhem, en tant qu'auteur d'une importante rflexion mthodologique sur
l'histoire des sciences, peut tre rang sous la bannire de l'pistmologie historique,
expression aux origines vraisemblablement multiples. Cette rflexion est en ralit
insparable d'une rflexion sur l'objet scientifique lui-mme en sciences de la vie et en
mdecine, dont toutes les dimensions philosophiques sont envisages par Canguilhem
dans les textes publis dans le volume La Connaissance de la vie, et dans sa thse sur
l'histoire du concept de rflexe, entre 1945 et 1955. Ces annes de rflexion peut-tre
les plus riches de Canguilhem montrent bien cette intrication des problmes
pistmologiques et ontologiques poss par les sciences de la vie. On tentera de dgager
quelques arguments principaux dans cette intrication et de montrer comment le
rationalisme viscral de Canguilhem se met lui-mme l'preuve de ces sciences.

Rethinking Georges Canguilhems Historical Epistemology
The importance given by historian and philosopher of science Georges Canguilhem
(19041995) to the role of practice, techniques, and experimentation in conceptformation was largely overlooked by commentators. After placing Canguilhems
contributions within the larger history of historical epistemology in France, and
clarifying his views regarding this expression, I re-evaluate the relation between
concepts and experimental practices in Canguilhems philosophy of science. Drawing
on his early writings on the relations between science and technology in the 1930s, on
the Essai sur quelques problmes concernant le normal et le pathologique (1943), and
on La formation du concept de rflexe aux XVIIe et XVIIIe sicles (1955), I argue that
the formation and rectification of concepts in Canguilhems sense are intrinsically
bound with the experimental, material, technical, and cultural contexts in which
concepts are operationalized.
LE CONCEPT D'OBJECTIVIT CHEZ CANGUILHEM

Franoise Balibar
Looking for a concept of objectivity in Canguilhem's work happened to be a failure; no
theory or conceptual analysis of objectivity is to be found. Although Canguilhem seems
to hold to a very classical, not to say conventional, view on objectivity (only scientific
knowledge, implying sophisticated mathematics and refined instrumentation, can be
said to be objective), notes taken while preparing a lecture to be delivered in the
224
ABSTRACTS
Sorbonne in the years 1960 (presently in Canguilhem's Nachlass) point to a somewhat

different view on the subject (although compatible with the previous one): as far as
relations between object and objectivity are to be taken into account, the scope is
different wither you go from object to objectivity or from objectivity to object. From
this, I shall try to draw conclusions regarding new trends in epistemology on the
question of objectivity (Haraway, Daston and Galison).
SCIENTIFIC PHILOSOPHIES IN THE EARLY 1930S AND

GASTON BACHELARD ON INDUCTION
Sandra Pravica
Against the background of the history of logical empiricism in the early 1930s, I will
elaborate on some features of Gaston Bachelards approach to the philosophy of
science. In doing so I wish to counter the tendency to locate Bachelards epistemology
outside of, or at odds with a more common epistemological discourse or to affiliate it
non-ambiguously with a continental tradition. I will explore intersections in the
research agendas of Bachelard and representatives of logical empiricism, their
respective uses of the term scientific philosophy and aspects of their mutual
communication relatively unnoticed so far. Indeed I shall stress that Bachelards
concepts provide serious solutions to the problems topical in the philosophy of science
in the 1930s.

Working on the Evolution of Problems from Both Sides of the Rhine
Cornelius Borck
In this presentation, Georges Canguilhems historical epistemology will be compared
and contrasted with Hans Blumenbergs phenomenology, approaching history of
science from the history of concepts and metaphors. Three areas of comparison shall be
looked at, (a) the trajectory towards a history of concepts, (b) the respective role of
technology in their epistemologies, (c) and finally the relation between historiography,
epistemology and philosophy. Instead of a systematic comparison, however, what is
being sketched out here rather follows historical curiosity into the question how two
philosophical programs with a shared interest in the history of ideas and concepts may
relate to each other albeit being developed (as far as we know) independently from each
other though perhaps in some other form of proximity. Describing Blumenbergs
agenda as a form of historical epistemology may be too far fetched, but he certainly
aimed to move history of philosophy towards historical epistemology. Here, then,
emerges shared territory. Time may have come for a fresh look upon histories of
concepts and historical epistemology, after some decades of highly productive
ethnographic studies of science and fruitful looks into material cultures of science;
something new may be gained from such a return.
225
ABSTRACTS

Andrea Cavazzini
Lide mme dune pistmologie historique implique la possibilit dindiquer des
niveaux dtermins dhistoricit lintrieur de ce qui se prsente comme le domaine,
ou les domaines, de la connaissance scientifique une poque donne. On peut
supposer que lhistoricit propre quil sagit de saisir dans une telle entreprise historicopistmologique soit lhistoricit des concepts.
Georges Canguilhem a insist avec beaucoup de force sur le caractre crucial de la
dimension propre au concept: cest celui-ci qui reviendrait de jouer le rle du pivot de
ltude du devenir de la connaissance et de la pense scientifique. De toute vidence, le
champ scientifique que Canguilhem a creus, celui des sciences de la vie, contribue
lgitimer et tester cette position. Le vivant ne se laisse interroger que par la construction
de modles conceptuels, ncessairement qualitatifs, dont la fonction consiste restituer
une ide gnrale des proprits spcifiques de la ralit vivante. La notion de loi a
toujours t hautement problmatique en biologie, et les structures mathmatiques de
symtrie et dinvariance, qui jouent un rle dcisif dans la constitution de lobjet
physico-mathmatique, ne peuvent gure puiser lindividualit irrductible des
organismes, dont la forme et les proprits ne peuvent faire lobjet que dune saisie
globale.
Les modles conceptuels biologiques qui oprent cette saisie gardent un lien
structurel des mtaphores, des images, des discours et des pratiques extrascientifiques. Les sciences de la vie sont beaucoup moins capables que la physique
mathmatique ou les mathmatiques disoler leurs dispositifs thoriques vis--vis de
lensemble des formes de la vie sociale, des discours qui circulent en elles et aussi des
fantasmes idologiques qui les traversent. Cest pourquoi une archologie des concepts
est possible, qui rattacherait les concepts et les procdures les mieux pures des
systmes souterrains de cohrences caches qui tmoignent de lexcs ineffaable de
la significativit dun modle par rapport sa valeur dobjectivation de lexprience et
de vrification des connaissances. Cet excs reprsente par consquent la
surdtermination et limprvisibilit de lhistoricit propre aux modles conceptuels.
Cependant, ce serait une erreur de rduire ces positions communes Canguilhem,
Foucault et Althusser, entre autres une simple rsorption de lanalyse de la
connaissance scientifique par la sociologie ou ltude de schmes culturels contingents.
Lhistoricit des concepts visant la saisie du vivant ne relve pas, pour Canguilhem, que
de la dispersion des cohrences archologiques: la succession des modles de la
connaissance biologique est traverse galement par un telos dtermin consistant,
justement, dans la saisie de la singularit du vivant. Finalement, ce quoi se rfrent, et
tendent, les diffrents concepts que lhistoire propose comme oprateurs de la
connaissance de la vie, cest bien la spcificit du vivant dans ce quil a dirrductible.
Si la vie consiste dans un pouvoir de se donner des normes, et si les normes immanentes
la vie font galement partie de lintelligibilit de la vie, force est de conclure que les
modles conceptuels du vivant pourront et devront tre jugs laune de cette
intelligibilit immanente.
226
ABSTRACTS
Lhistoricit des concepts donc ne relve pas que de la contingence historique et de

la surdtermination par les pratiques et les discours qui traversent un contexte social
donn: il existe aussi une historicit tlologique, oriente, qui vise ladquation des
concepts, des mthodes, des thories, lintelligibilit intrinsque de la vie en tant que
porteuse des conditions de possibilit de sa propre connaissance. Cette tension interne
lpistmologie historique de Canguilhem, qui caractrise notamment de textes tels que
Le Concept et la vie , se fonde sur la circularit entre les deux modes de lhistoricit:
la validit dun modle conceptuel dpend de sa capacit reprsenter une
approximation convenable des modes dtre du vivant; mais ces modes dtre du vivant
ne peuvent dployer leur signification que par la saisie quoprent des modles
historiquement dtermins. Le concept suppose la vie comme le prsuppos de
lintelligibilit quil effectue; et la vie suppose le concept comme la condition de
lintelligibilit dont elle-mme contient les critres.
La dialectique entre le concept et la vie est formule par Canguilhem dans le texte
cit en renvoyant Hegel, et la thorie de la vie comme tre-l immdiat et implicite
du concept. Dvelopper cette dialectique pourrait permettre dcrire une autre histoire
de lpistmologie historique la franaise , den apprcier diffremment les liens
linspiration de la Naturphilosophie idaliste et romantique (ventuellement filtre par
la tradition des Ravaisson, Lachelier, Boutroux et Bergson), et peut-tre dimpulser une
nouvelle approche des sciences contemporaines, en particulier de la biologie et de la
physique, o la question de lintelligibilit des objets pistmiques croise la
problmatique du dynamisme immanent la nature .

Maria Muhle
Dans mon expos, je voudrais interroger la pense de Georges Canguilhem et de Michel
Foucault autour dun phnomne concret, mais non pour autant bien dfini, la vie. Pour
Canguilhem, la vie est marque par son essentielle prcarit et apporte une
dimension supplmentaire aux distinctions classiques entre lhistoire naturelle qui ne
connat que les tres vivants et la biologie qui, vers 1800, dcouvre la vie dans sa
fonctionnalit organique. Cette prcarit de la vie donne lieu une dynamique que
Canguilhem formule en termes de normativit vitale qui ne se rduit pas la notion de
vie biologique elle introduit au contraire un excs par rapport au modle dune vie
purement organique quon peut appeler vitaliste. La vie est par consquent habite
dune double dynamique : elle ne spuise pas dans la simple tendance maintenir ltat
normal mais remet constamment en cause ltablissement dune normalit organique.
Cest alors quelle est normative. Foucault, au contraire, sabstient de donner une
dfinition de la vie: Cest uniquement comme objet de savoir et de pouvoir que la vie
est prsente dans son uvre. Je voudrais alors interroger limportance de la notion de
vie normative de Canguilhem pour le fonctionnement du pouvoir sur la vie, tel que
Foucault larticule avec le bio-pouvoir, qui se rfre la vie de la population globale de
faon positive: qui fait vivre.
227
ABSTRACTS

Cristina Chimisso
In 1969, Canguilhem gave a talk, published the following year, entitled What is
scientific ideology? The topic may surprise, as Canguilhem so far had not used the
concept of ideology, let alone the apparently oxymoronic concept of scientific ideology.
I will not focus on the new context of the 1960s in which the works of Althusser and
Foucault obviously played an important role. Rather, I will examine Canguilhems
introduction of this concept in the context of the tradition of historical epistemology. In
this, I follow Canguilhem who stated that his aim in the paper is that of refurbishing
without rejecting the lessons of Gaston Bachelard. I argue that Canguilhem introduced
the concept of scientific ideology as a solution to a long-standing historiographical
problem that historical epistemology presented for his history writing. A rigid
application of the norms of historical epistemology would have excluded as non-science
much of the content of his narratives. Moreover, it would not have allowed for the
continuities that he constructed, like that between Thomas Williss concept of reflex and
the current one. Scientific ideology an intermediate category between science and
non-science enabled Canguilhem to re-integrate into history of science those
doctrines, practices and concepts that would otherwise have been excluded.

Anselm Haverkamp
For the project of an epistemological history of literature, Shakespeare marks a
historical threshold in the emergence of a new meaning of life, whose early format is
to be traced in his Sonnets. In "Life in the Sonnets" the emphasis is on the emergence of
the new conception of life as opposed to an explicit expression in, or occurrence of, liferelated semantics. Nothing needed to be expressed, but a lot had become pressing and
emerged in the medium call it the form of the Literary.
228
LIST OF AUTHORS AND PARTICIPANTS
Mme. le Prof. Franoise Balibar

39 rue Gazan
75014 Paris
France
Email: [email protected]
Prof. Dr. Cornelius Borck
Institut fr Medizingeschichte und Wissenschaftsforschung
Universitt zu Lbeck
Knigstr. 42
23552 Lbeck
Prof. Dr. Jean-Franois Braunstein
Philosophies contemporaines EXeCO
Universit Paris 1
UFR de Philosophie
17, rue de la Sorbonne
75231 Paris Cedex 05
Frankreich
Andrea Cavazzini
Email : [email protected]
Dr. Cristina Chimisso
Head of the Department of Philosophy
Arts Faculty
The Open University
Milton Keynes MK7 6AA
United Kingdom
M. le Prof. Claude Debru
10 Avenue de Villars
75007 Paris
France
229

AUTHORS AND PARTICIPANTS
M. le Prof. Franois Delaporte

M. le Prof. Yves Duroux
11 Square Sainte-Croix de la Bretonnerie
75004 Paris
Frankreich
Dr. Thomas Ebke
Universitt Potsdam
Institut fr Philosophie
Am Neuen Palais 10, Haus 11
14469 Potsdam
Prof. Dr. Moritz Epple
Goethe-Universitt Frankfurt
Historisches Seminar
Wissenschaftsgeschichte
Grneburgplatz 1
60629 Frankfurt am Main
Prof. Dr. Stefanos Geroulanos
Department of History
New York University
53 Washington Square South
New York, NY 10012-1098
USA
Prof. Dr. Anselm Haverkamp
Dept. of English
Faculty of Arts and Science
New York University
19 University Place
New York, NY 10003
USA
M. le Prof. Camille Limoges
1206 rue Lajoie
Outremont Montral H2V 1P1
Canada
230

Institute for the History of Medicine and Health
Geneva University, Faculty of Medicine
C.P. 1211, Geneva 4
Switzerland
Prof. Todd Meyers
Anthropology Department, 3057 FAB
Wayne State University
College of Liberal Arts & Sciences
4841 Cass Ave.
2155 Old Main
Detroit, MI 48201
USA
Prof. Dr. Maria Muhle
Merz Akademie
Hochschule fr Gestaltung,
Kunst und Medien, Stuttgart
Teckstrae 58
70190 Stuttgart
Sandra Pravica
Boltzmannstr. 22
14195 Berlin
Prof. Dr. Hans-Jrg Rheinberger
Boltzmannstr. 22
14195 Berlin
Prof. Dr. Henning Schmidgen
Institut fr Information und Medien,
Sprache und Kultur (I: IMSK)
Universitt Regensburg
Universittsstr. 31
93053 Regensburg
231

Prof. Dr. Peter Schttler

Boltzmannstr. 22
14195 Berlin
Prof. Dr. Frieder Otto Wolf
Traunsteiner Str. 8
10781 Berlin
Dr. Monika Wulz
Kulturwissenschaftliches Kolleg der Universitt Konstanz
Otto-Adam-Str. 5
78467 Konstanz
232

Max-Planck-Institut
fr
Wissenschaftsgeschichte
Max Planck Institute for the History of Science
Preprints since 2010 (a full list can be found at our website)
391 Horst-Heino v. Borzeszkowski & Renate Wahsner Die Fassung der Welt unter der Form des

Objekts und der philosophische Begriff der Objektivitt
392 Ana Barahona, Edna Suarez-Daz, and Hans-Jrg Rheinberger (eds.) The Hereditary Hourglass.
Genetics and Epigenetics, 18682000
393 Luis Campos and Alexander von Schwerin (eds.) Making Mutations: Objects, Practices,
Contexts
394 Volkmar Schller Some Remarks on Prop. VIII Probl. II of Newtons Opticks Book I Part I
395 Tams Demeter Humes Experimental Method
396 Fynn Ole Engler, Bjrn Henning und Karsten Bger Transformationen der wissenschaftlichen

Philosophie und ihre integrative Kraft Wolfgang Khler, Otto Neurath und Moritz Schlick
397 Frank W. Stahnisch Der Rosenthalsche Versuch oder: ber den Ort produktiver
Forschung Zur Exkursion des physiologischen Experimentallabors von Isidor Rosenthal

(18361915) von der Stadt aufs Land
398 Angela Matyssek berleben und Restaurierung. Barnett Newmans Whos afraid of Red,

Yellow, and Blue III und Cathedra
399 Susanne Lehmann-Brauns, Christian Sichau, Helmuth Trischler (eds.) The Exhibition as Product

and Generator of Scholarship
400 Fynn Ole Engler und Jrgen Renn Wissenschaftliche Philosophie, moderne Wissenschaft und
Historische Epistemologie
401 M. J. Geller Look to the Stars: Babylonian medicine, magic, astrology and melothesia
402 Matthias Schemmel Medieval Representations of Change and Their Early Modern Application

(TOPOI Towards a Historical Epistemology of Space)
403 Frank W. Stahnisch German-Speaking migr Neuroscientists in North America after 1933:
Critical Reflections on Emigration-Induced Scientific Change
404 Francesca Bordogna Asceticism and Truth: The Case of Magic Pragmatism
405 Christoph Hoffmann and Alexandre Mtraux (eds.) Working with Instruments Three Papers

of Ernst Mach and Ludwig Mach (Translated by Daniel Bowles)
406 Karin Krauthausen Paul Valry and Geometry: Instrument, Writing Model, Practice
407 Wolfgang Lefvre Picturing the World of Mining in the Renaissance
The Schwazer Bergbuch (1556)
408 Tobias Breidenmoser, Fynn Ole Engler, Gnther Jirikowski, Michael Pohl and Dieter G. Weiss

Transformation of Scientific Knowledge in Biology: Changes in our Understanding of the
Living Cell through Microscopic Imaging
409 Werner Kogge Schrift und das Rtsel des Lebendigen. Die Entstehung des Begriffssystems

der Molekularbiologie zwischen 1880 und 1950
410 Florentina Badalanova Geller 2 (Slavonic Apocalypse of) Enoch : Text and Context
411 Henning Schmidgen & Urs Schoepflin (eds.) Hans-Jrg Rheinberger : a Bibliography
412 Renate Wahsner & Horst-Heino v. Borzeszkowski Erkenntnis statt Erbauung: Hegel und das

naturwissenschaftliche Denken der Moderne
413 Mirjam Brusius From photographic science to scientific photography: Photographic

experiments at the British Museum around 1850
414 Viktor J. Frenkel Professor Friedrich Houtermans Arbeit, Leben, Schicksal. Biographie

eines Physikers des zwanzigsten Jahrhunderts. Herausgegeben und ergnzt von Dieter

Hoffmann, unter Mitwirkung von Mary Beer
415 Ilana Lwy (ed.) Microscope Slides Reassessing a Neglected Historical Ressource
416 Andr L. Blum, John Michael Krois und Hans-Jrg Rheinberger (Hrsg.) Verkrperungen
417 Pietro Daniel Omodeo Sixteenth Century Professors of Mathematics at the German
University of Helmstedt. A Case Study on Renaissance Scholarly Work and Networks
418 Peter Schttler & Hans-Jrg Rheinberger (ds.) Marc Bloch et les crises du savoir
419 Albert Presas i Puig (ed.) A Comparative Study of European Nuclear Energy Programs
420 Mathias Grote & Max Stadler (eds.) Membranes Surfaces Boundaries

Interstices in the History of Science, Technology and Culture
421 Frank W. Stahnisch The emergence of Nervennahrung : Nerves, mind and metabolism

in the long eighteenth century
422 Pietro Daniel Omodeo, Irina Tupikova Aristotle and Ptolemy on Geocentrism: Diverging
Argumentative Strategies and Epistemologies (TOPOI Towards a Historical Epistemology

of Space)
423 Han F. Vermeulen Linguistik und Vlkerkunde der Beitrag der historisch-vergleichenden
Linguistik von G.W. Leibniz zur Entstehung der Vlkerkunde im 18. Jahrhundert

[Leicht erweiterte Fassung des Working Papers No. 133 aus dem MPI for Social Anthropology]
424 Alfred Gierer Mit Schiller gegen den Egoismus der Vernunft. Zeitbergreifende Gedanken

zur Natur des Menschen
425 Annette Vogt Die Berliner HumboldtUniversitt von 1945/1946 bis 1960/1961
426 Klaus Geus, Martin Thiering (eds.) Common Sense Geography and Mental Modelling
427 Renate Wahsner Kann eine moderne Naturphilosophie auf Hegelsche Prinzipien gegrndet

werden? Spekulatives und naturwissenschaftliches Denken
428 Stefano Bordoni Widening the Scope of Analytical Mechanics Duhems third pathway to

Thermodynamics
429 Pietro Daniel Omodeo Copernicus in the Cultural Debates of the Renaissance: Reception,
Legacy, Transformation [Part I & II]
430 Mark Geller & Klaus Geus (eds.) Productive Errors: Scientific Concepts in Antiquity

(TOPOI Dahlem Seminar for the History of Ancient Sciences)
431 Klaus Gottstein The Amaldi Conferences. Their Past and Their Potential Future
432 Mikul Teich The Scientific Revolution Revisited
433 Lorraine Daston & Jrgen Renn (Hrsg.) Festkolloquium fr Hans-Jrg Rheinberger

Beitrge zum Symposium am 24. 1. 2011 im Max-Planck-Institut fr Wissenschaftsgeschichte

Dossier Canguillhem and Bachelard

Uploaded by

Copyright:

Available Formats

Dossier Canguillhem and Bachelard

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dossier Canguillhem and Bachelard

Uploaded by

Copyright:

Available Formats

Max-Planck-Institut fr Wissenschaftsgeschichte

Max Planck Institute for the History of Science

Epistemology and History

Epistemologie und Geschichte

Historical Epistemology, Old and New

Sur la rception de lpistmologie franaise en Allemagne

Rheinbergers Take on Historical Epistemology

La dissection humaine. De lhistoire des commencements celle des origines

Intervals, Possibilities and Encounters. The Trigger of a Ruptured History

Kurt Goldsteins Conception of Individuality

Georges Canguilhem. pistmologie historique et/ou histoire philosophique?

On the Genealogy of Concepts and Experimental Practices. Rethinking Georges

Le concept dobjectivit chez Canguilhem

Scientific Philosophies in the Early 1930s and Gaston Bachelard

Canguilhem and Blumenberg. Working on the Evolution of Problems

Archologie des concepts et philosophie de la nature

Histoire(s) de la vie de Canguilhem Foucault

The Role of Scientific Ideology in Canguilhems Historiography

Ist eine historische Epistemologie der Sozialwissenschaft mglich?

Life in Shakespeares Sonnets

LIST OF AUTHORS AND PARTICIPANTS

EPISTEMOLOGY AND HISTORY

EPISTEMOLOGY AND HISTORY

EPISTEMOLOGY AND HISTORY

EPISTEMOLOGY AND HISTORY

1) The history of concepts

EPISTEMOLOGY AND HISTORY

2) The historicity of objects and spaces of knowledge

EPISTEMOLOGY AND HISTORY

L pistmologie historique a accompli, depuis vingt ans environ, une carrire

la fois par des interventions historiques et systmatiques sur la problmatique de

1. Lhistoire des concepts scientifiques

2. Lhistoricit des objets et des espaces de savoir

sens commun, entre lespace de la science et celui de la littrature et de lart.

3. Le caractre exprimental des pratiques scientifiques et le problme de la rcurrence

EPISTEMOLOGIE UND GESCHICHTE

Die Historische Epistemologie hat in den vergangenen zwei Jahrzehnten international

EPISTEMOLOGIE UND GESCHICHTE

Zusammenhang, in dem zuerst buchstblich von Historischer Epistemologie die Rede

EPISTEMOLOGIE UND GESCHICHTE

schaften: Wenn die Epistemologie historisch ist, ist die Wissenschaftsgeschichte

EPISTEMOLOGIE UND GESCHICHTE

wie fr Bachelard unentbehrlich war, um die Entstehung und Entwicklung des

EPISTEMOLOGIE UND GESCHICHTE

2.) Die Historizitt von Objekten und Rumen des Wissens

3.) Der quasi-experimentelle Charakter des Wissenschaftsgeschehens und das Problem

EPISTEMOLOGIE UND GESCHICHTE

14:30 Thomas Ebke

17:00 Begrung und Einfhrung

16:30 Frieder Otto Wolf

16:30 Claude Debru

14:30 Maria Muhle

11:30 Cornelius Borck

9:30 Franoise Balibar

Samstag, 11. Dezember 2010

11:30 Franois Delaporte

9:30 Moritz Epple

Freitag, 10. Dezember 2010

Donnerstag, 9. Dezember 2010