Stud. Hist. Phil. Biol. & Biomed. Sci.

34 (2003) 297327
www.elsevier.com/locate/shpsc

The tribunal of philosophy and its norms:

history and philosophy in Georges
Canguilhems historical epistemology
Cristina Chimisso
Department of Philosophy, Arts Faculty, The Open University, Walton Hall, Milton Keynes MK7 6AA,
UK

Received 15 June 2002; received in revised form 23 December 2002

Abstract

In this article I assess Georges Canguilhems historical epistemology with both theoretical
and historical questions in mind. From a theoretical point of view, I am concerned with the
relation between history and philosophy, and in particular with the philosophical assumptions
and external norms that are involved in history writing. Moreover, I am concerned with the
role that history can play in the understanding and evaluation of philosophical concepts. From
a historical point of view, I regard historical epistemology, as developed by Gaston Bachelard
and Georges Canguilhem, as a conception and practice which came out of the project, elabor-
ated in France from the 1920s to the 1940s, of combining history of science and philosophy.
I analyse in particular Canguilhems epistemology in his theory and practice of history of
science. What he called normative history is the focus of my analysis. I evaluate the question
of the nature and provenience of the norm employed in normative history, and I compare it
with the norm as discussed by Canguilhem in Le normal et le pathologique. While I am critical
of Canguilhems treatment of history, I conclude that his philosophical suggestion to analyse
the formation of scientific concepts from below represents a useful model for history and
philosophy of science, and that it can be very profitably extended to philosophical concepts.
2003 Elsevier Science Ltd. All rights reserved.

Keywords: Canguilhem; Bachelard; Koyre; Historiography; Normativity; Historical epistemology

E-mail address: C.Chimisso@open.ac.uk (C. Chimisso).

1369-8486/03/$ - see front matter 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S1369-8486(03)00027-X
298 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

1. Introduction

The relation of philosophy with history has often been a difficult one. The great
majority of philosophers who are included in the canons and in most compendia
of the history of philosophy appear not to have been keen on historical research.
Before the eighteenth century, philosophical reflection hardly ever engaged with his-
tory. Enlightenment philosophers did take history as the core of their investigation
but, often, as in the case of Condorcet, Turgot and Rousseau, their histories were
highly philosophical, and were little concerned with the construction of small-scale
narratives through original sources, or the establishment of particular events. Many
philosophers have engaged with the history of their own discipline; but they have
largely seen it as a succession of ideas and thinkers sitting on their own time arrow,
undisturbed by external events. Although this picture may be readily accepted by
those philosophers (and historians) who do not consider the relationship between
philosophy and history a fruitful one, it is, if not false, at least incomplete: in fact
many philosophers have engaged with history, including several whom I am going
to discuss in this paper. This picture may also be unduly simplified, as the relation
between history and philosophy has assumed different forms and has played different
roles in different institutional settings and national traditions. In many quarters this
relation still seems to remain unresolved, or unwelcome. In some philosophical tra-
ditions still alive, notably the analytical tradition, history has been seen as irrelevant,
indeed as the opposite of what philosophy is about: history is about contingency and
change, while philosophy is about necessity and timeless truths.
In this context, the notion of historical epistemology may be seen as immediately
striking by some, for history and philosophy appear to be securely interlocked. To
make things more intriguing, the history in question is history of science rather than
philosophy. Recently, Lorraine Daston and Ju rgen Renn have described their research
programmes as historical epistemology (Daston, 1994, 1997; Renn, 1995). I shall
not be concerned here with their programmes, which share the name but not the
methods and aims with the doctrines that I am going to discuss. I shall concentrate
on the emergence of historical epistemology as a new approach to the study of
scientific knowledge at the hands of Gaston Bachelard and Georges Canguilhem.
My focus will be in particular Georges Canguilhems historical epistemology, and
the type of history that he practised and defended in his work. He called his style
of history normative history, that is, a narrative constructed by assuming a norm
that allows one to evaluate and judge past doctrines. Normative history and historical
epistemology are indeed two sides of the same type of project. Gaston Bachelard
was Canguilhems recognised point of reference in this type of enterprise; for this
reason comparisons between their projects will run throughout this article.
The question on which I shall focus in particular is the origin and nature of the
norm employed in order to judge concepts and theories. Canguilhems work is
extremely interesting in this regard, because not only did he write normative histories
and defend this practice in several articles, but he also wrote a history of the concept
of the norm in his Le normal et la pathologique. Is the norm as conceived in Le
normal et la pathologique the same as the norm employed in the construction of
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 299

normative histories? Does it have the same origin? These will be the main questions
I shall address with regard to the norm in Canguilhems work.
In sympathy with the topic of this articleand the convictions of its authormy
treatment will be not only theoretical but also historical. I shall discuss Canguilhems
conception and practice of history and philosophy in their interaction, and situate
them in the debates which took place among French scholars in the period before
and during Canguilhems career. My view of historical epistemology is as a concep-
tion and practice which came out of the project of combining history of science and
philosophy which had been elaborated from the 1920s to the 1940s in Paris. This
approach favoured both a philological attention to texts and a consideration of
science within its cultural and social context, and at the same time addressed philo-
sophical questions about the mind and knowledge. The relation between historical
epistemology and the original project of combining history of science and philosophy
is not one of simple derivation, for historical epistemology partly abandoned one of
the main aspirations of that project, namely a detailed and comprehensive historical
account of past doctrines. My version of the historical place of Bachelards and
Canguilhems projects, and their relationship with the contemporary historiography,
is rather different from most of the current ones, notably from that of Michel Fou-
cault.

2. Histories of the history of the sciences

In the Archaeology of knowledge, Michel Foucault provides a synthetic account

of recent historiography of the sciences. He distinguishes three types of history of
the sciences: that which employs a recurrential analysis,1 epistemological history
and archaeological history. The first type for Foucault belongs to mathematics, and
consists in reinterpreting past theories so as to make them particular cases of current
theories. Although Foucault only cites Michel Serres, the concept of recurrent history
was elaborated by Bachelard in essentially the same terms as Foucaults. Unlike the
other two types of history, this history is an integral part of science itself. The second
type of history of science is for Foucault epistemological, and it is exemplified by
the works of Gaston Bachelard and Georges Canguilhem. Foucault describes it as:

A type of historical analysis . . . [that] takes as its norm the fully constituted
science; the history that it recounts is necessarily concerned with opposition of
truth and error, the rational and the irrational, the obstacle and fecundity, purity
and impurity, the scientific and non-scientific. It is an epistemological history of
science. (Foucault, 1972, p. 190)

1
Recurrential analysis is the expression used in the English translation of The archaeology of knowl-
edge (Foucault, 1972, p. 190). It translates the French analyse re currentielle (Foucault, 1996 [1969],
p. 248).
300 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

The third type is Foucaults own archaeology. Archaeology abandons the normative
point of view and, as a consequence, the concern with the distinction between scien-
tific and non-scientific forms of knowledge. In fact, Foucaults archaeology, in his
own description, is aimed to uncover descriptive practices in so far as they give
rise to a corpus of knowledge, in so far as they assume the status and role of a
science (Foucault, 1972, p. 190).
Foucaults classification of the historiographies of science is useful as an analytical
tool and as a starting point of a reflection on the aims of different types of history
of science. One cant hide from the fact, though, that Foucaults is a rather recurrent
reconstruction of the history of historiography. Just as the sciences, in Bachelard
and Foucaults view, reconstruct their history in a rational manner from the point
of view of the present, and condemn to oblivion those theories and practices that
cannot be integrated in this history, so Foucaults account of the historiography of
science is simplified and progressive. It is progressive in that his three types of
historiography are in order of increasing depth of analysis and reflexivity. Recurrent
history is unreflexively normative: it does have a norm of scientificity, which
coincides with the norm of current science, but this norm is unreflexively assumed.
As a consequence its historicity is hidden, and scientificity can be anachronistically
projected back to past theories and practices, which are reinterpreted and recon-
structed so that they conform to the norm.
Epistemological history is reflexively normative, for it consciously employs cur-
rent science as a norm to judge knowledge, and, unlike recurrent history, it is able to
analyse and determine when the leap between non-scientific and scientific discourses
occurs. Epistemological history understands the mechanism which produces recurrent
and lapsed history and understands how both non-science and science are produced.
This type of history does not de-historicize current science; indeed, although it does
take current science as its norm, it nevertheless regards its norm as the result of an
open-ended historical process.
Finally, Foucault presents his archaeology as a step forward, posing itself at a
deeper and more general level than epistemological history. Archaeology is reflexive
without normatively assuming the point of view of current science. Its analyses of
discursive practices apply to alchemy as well chemistry, to sciences as well as
humanities and crucially to history and philosophy, potentially including the first
two types of history of science. Moreover, the current norm undergoes the same
analysis as any other discursive practice.
Foucaults account of the history of history of science outlined above is recurrent
also in what it leaves out. Bachelard and Canguilhems normative approach to history
of science was by no means the only alternative to an unreflexive, recurrent history
of science. Needless to say, the historiography of the sciences had a much more
troubled history than a brief outline could show. In particular, Foucault leaves out
the historiographical trends that took the historical craft very seriously, and regarded
an exhaustive historical reconstruction as necessary for providing answers to philo-
sophical questions about the mind and knowledge. It is not just that there have been
many approaches to the study of the history of science, of which Bachelards and
Canguilhems represent two related types, united by their normative character. His-
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 301

torical epistemology also stands in a close historical relationship with other

approaches, in particular that of studying history of science as part of epistemology,
which was promoted by philosophers from the 1920s onwards.
An important centre for the elaboration of the project of the combination of history
and philosophy of science was the Institut dHistoire des Sciences et Techniques,
founded in 1932 by Abel Rey, professor of history and philosophy of the sciences
at the Sorbonne, who was succeeded on the directorship and chair by Bachelard in
1940, and then by Canguilhem in 1955. As conceived by Abel Rey, the ideal of the
Institut dHistoire des Sciences et Techniques was the fusion of history and philo-
sophy: for him, as for other contemporary philosophers, the philosophical study of
the mind, and epistemology, required a comprehensive historical study and a careful
analysis of sources. The history of the sciences as taught in the 1920s in science
faculties, though, was of no use to philosophers. A new history of the sciences was
needed. This new history was to be general, as Abel Rey argued in one of his talks
on the methods and aims of the discipline (Rey, 1935, pp. 3449). The syllabuses
of the courses offered at the Institut dHistoire des Sciences et Techniques reflected
this broad approach to history of the sciences: they spanned from history of physics
to history of sociology, and from antiquity up to the present day.2 This approach
was not just the expression of the French educational ideal of culture ge ne rale.
Philosophers such as the director Abel Rey sought to investigate the mentality behind
the cultural production of a certain period, expressed by the various sciences.3 At
the same time, this study required punctilious examination of sources and an exhaus-
tive account of the wider culture and society in which the sciences under study had
developed. The historian and philosopher of science He le`ne Metzger, associated with
the Institut dHistoire des Sciences et Techniques, thus illustrated this point in the
field of history of chemistry:

. . . no point of human history is irrelevant for the history of the sciences; social
and political history, history of philosophy, history of literature including history
of theatre . . ., history of industry and trade, to which could be also added history
of art and history of occultism, are [all] useful to the researcher who proposes to
study the chemists writings. (Metzger, 1935a, pp. 163164)

What Metzger was expressing here is the ideal of total history, which was promoted
by historians and philosophers of science, some of whom, like herself and Abel Rey,
belonged at the same time to the Institut dHistoire des Sciences et Techniques and

2
See Thale`s, 1 (1934), pp. viviii; 2 (1935), pp. iiii; and following years.
3
The word mentality had been introduced by Lucien Le vy-Bruhl, historian of philosophy, founder
of the Institute of Ethnology at the Sorbonne, and author of La mentalite primitive (Le vy-Bruhl, 1922).
The term quickly spread to other disciplines, including philosophy, history, psychology, pedagogy and
sociology.
302 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

the Centre de synthe`se historiquedirected by Henri Berrthe aim of which was

precisely the development of total history.4
Far from being a simple opposition to recurrent history, as Foucaults account
suggests, the work of Georges Canguilhem represents a very important splinter
resulting from the disintegration of the ideal of combining total history, philological
analysis of texts and philosophical investigation. This ideal was perhaps just too
ambitious, and the result was that students of history of science eventually renounced
either one or another part of the programme. Michel Fichant attributes to Gaston
Bachelard and Jean Cavaille`s (19031944),5 the development of historical epistem-
ology, to Alexandre Koyre the practice of epistemological history, and to Georges
Canguilhem that of the epistemology of the biological sciences, based on a Bachel-
ardian history of science (Fichant, 1973, pp. 168171). We can see here two of
the productsalthough related onesof the disintegration of the ideal: historical
epistemology on one side (Bachelard, Cavaille`s, Canguilhem) and epistemological
history on the other (Koyre ). In both cases, an experiment in the fusion of history
and philosophy takes place, but in the former case it is epistemology to set the
questions, while in the second it is history.
I have analysed elsewhere the development and ultimately the failure of this fasci-
nating and ambitious ideal within the institutional spaces of 1920s, 1930s and 1940s
Paris, and I have shown how the work of Gaston Bachelard represented the renunci-
ation of total history and of the philological analysis of sources in favour of a more
philosophical approach, mainly concerned with epistemological questions (Chimisso,
2001a, Ch. 5; 2001b). I have also shown how the disintegration of that theoretical
ideal went hand in hand with a weakening of the links among departments and
groups dedicated to the study of science. The Institut dHistoire des Sciences et
Techniques in the 1930s represented a point of contact between different approaches
to the study of science. Abel Rey was also a personal friend of Henri Berr and
Lucien Febvre. Together with the ideal of merging the different approaches to history
with philosophy, apparently the role of the Institut as point of contact between
research programmes and scholars failed as well.
Here I shall focus on texts rather than the institutional and intellectual networks,
but my reading will be guided by the historiographical questions posed by the project
of a synthesis between history and philosophy developed in the Parisian academia
in the first half of the twentieth century. I shall assess Canguilhems work as one
of the solutions that the tension between history and philosophy produced. My spe-
cific aim, however, is that of assessing the relation of history and philosophy in the
work of Canguilhem, and in particular his theory and practice of normative history.
But can we still use the heading history of science for Bachelards and Cangu-

4
At the Centre de Synthe`se, Rey directed the Unit of natural sciences and later the Unit of general syn-
thesis.
5
Cavaille`s, mathematician and philosopher, founded the journal Dialectica together with Bachelard.
During the war, he played a prominent role in the resistance; he was captured and killed by the Nazis
in 1944. Canguilhems commemorative speeches in honour of Jean Cavaille`s are published in a small
volume (Canguilhem, 1976), republished by Allia in 1996.
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 303

ilhems work? Indeed, not only the objects of their works, but also their scope, may
raise doubts about their disciplinary classification. Are they really histories? Foucault
himself regarded the work of Bachelard and Canguilhem, together with that of
Cavaille`s, as at the basis of the philosophie du savoir, opposed to the other great
philosophical current of post-war France, that of philosophy of existence (Foucault,
1985). In Bachelards and Canguilhems epistemological histories, what is more
important, epistemology or history? And can the two be disentangled?

3. The laboratory and the tribunal: some background to the normative turn
in history of science

In La formation de lesprit scientifique (Bachelard, 1993 [1938]), Gaston Bachel-

ard investigated the general conditions of the development of science. He was cer-
tainly after a certain kind of totality, in that he believed that the mentality that he
found behind a scientific tradition in fact informed the whole cultural production of
a certain epoch. However, he did not pursue the totality that the members of the
Centre de synthe`se were after, or that Metzger described in the quotation above. For
Metzger, everything was relevant, because the mentality of a certain epoch is to
be studied for itself. In other words, for Metzger the challenge was to understand
the way of thinking behind past texts she studied, and the analysis of all that remains
from that period is useful to meet it.6
Bachelards history, by contrast, was a judged history; in his words:

The history of the sciences is essentially a judged history . . . The historian of the
sciences, in order to judge the past properly, needs to know the present.
(Bachelard, 1972 [1951], pp. 141, 142)

Bachelard intended to sort the scientific from the non-scientific and to explain the
mechanisms of both. He dedicated keen attention to the imagination and emotions.
These, for Bachelard, are only to be expressed in reveries, poetry and art: employed
as organs of the inquiry into nature, they prevent objectivity, and produce obstacles
to scientific knowledge. Some of the main obstacles that Bachelard analysed in La
formation de lesprit scientifique are: a sexualised view of nature, an attraction for
small and precious objects, the instinct to possess these objects and the consequent
desire for them to be real, and a disposition for generalisation and for attributing
a soul, or at least life, to any object or substance. Bachelard thought that the non-
rational part of the human mind did not change, and therefore the obstacles to scien-
tific knowledge tended to turn up in essentially the same form. It was then not neces-
sary for him to reconstruct a historical milieu in order to understand a past theory
or text. He could find the same attitudes in an eighteenth-century alchemy text as

6
Metzger theorized total history without really practising it; for most of time she focused on scien-
tific texts.
304 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

in a nineteenth-century poem, as for him these were constant features of the human
mind (Bachelard, 1993 [1938]).
For Bachelard, the history of science thus began when knowledge had been disen-
tangled from imagination, feelings and personal ends. Unlike the imaginative
approach to nature, rational and objective knowledge was in his view able to evolve,
and had a history. Bachelards kind of history does not require a philological attention
to documents, or a careful reconstruction of milieus or institutions. Actually, he did
not even aim to write a history, for his objective was rather that of examining the
nature of scientific inquiry. By using modern science as a norm, he was able to
isolate the profound differences, or ruptures, which separated science from previous
inquiry into nature, or as he called it, pre-science. Bachelards aim was then epis-
temological rather than historical.
Georges Canguilhem followed Bachelard in this normative view of history of
science and in the epistemological end of historical research. Canguilhem certainly
inherited Bachelards way of conceiving of the relationship between history and
philosophy. He also inherited his disregard for historical research as conceived by
historians who did not intend to judge history of science. This may be the reason
why some of them did not welcome Canguilhems appointment to the Instituts direc-
torship in 1955. One was Rene Taton, at the time research director at the CNRS. In
a letter to Georges Sarton, he commented that the new appointment hardly left any
hope that the Institut dHistoire des Sciences et Techniques would awaken and play
an active role (Taton, 1956). At that time, the Instituts journal Thale`s, which had
been an important organ of innovation in history of science, was not even being
published. Its publications were resumed in 1958, after a five-year gap. The activities
of the Institut and the much reduced breadth of its output show that the directors
Bachelard and Canguilhem did not pursue the goal of broad participation in a wide
network which had characterized the early years of the Institut. The many insti-
tutional links of the Institut under Abel Reys directorship (19321940) reflected the
inclusiveness of the approaches Rey and his associates took to the history of science.
The interest of French philosophers in intellectual history dated back to the mid
nineteenth century. As Martial Gueroult has pointed out, Antoine Cournot (1801
1877) already conceived of philosophy as knowledge of the mind achieved through
the study of the efforts made, in the course of history, by the intelligence in order
to constitute a science (Gueroult, 1988, p. 877). Cournots conception of philosophy
would be developed and consolidated after him. Le on Brunchvicg, Bachelards
supervisor on one of his doctoral theses and his recognised mentor, reconstructed
his own practice of philosophy as a reflection on the limits and capabilities of the
mind through the study of the history of science (Gueroult, 1988, p. 877). This role
of history also explains the prominence of history of philosophy in the philosophy
curriculum at the Sorbonne and in general its prestige as a philosophical subject.7

7
One of the indicators of the importance of history of philosophy is the number of chairs in history
of philosophy at the Sorbonne. Professors, when they obtained a chair, could choose its name; therefore
the number of chairs in a certain discipline reflects the interests of the most distinguished academics.
Between its foundation (1808) and 1935, at the Faculty of Lettres of the Sorbonne 4 professors chose
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 305

An example of the philosophical role of history of philosophy is exemplified by the

unsigned editorial of the first issue of Revue dhistoire de la philosophie, attributable
to the director Emile Bre hier (Anonymous, 1927, p. 4). The author argued that history
was the substitute of the experimental method in philosophy: in order to study the
mind, we need to see it at work, and history represents the laboratory in which to
carry out observations. This metaphor of intellectual history as the experimental
method, or the laboratory, of philosophy was rather successful. Canguilhem himself
employed it, attributing it to Edward J. Dijksterhuis (Canguilhem, 1994 [1968], p.
12; Canguilhem, 1993 [1977], pp. 1213). Canguilhem, however, argued that the
tribunal or a school8 would be a better metaphor than the laboratory. This tribunal
is an institution and a place in which judgements on the past of knowledge, or on
the knowledge of the past, are passed (Canguilhem, 1994 [1968], p. 13). A tribunal,
Canguilhem goes on, needs a judge: in the history of science tribunal, the judge
should be epistemology (Canguilhem, 1994 [1968], p.13).
The metaphor of the tribunal provides us with a good example of the difference
between Canguilhems conception of the relation between history and philosophy
and that of previous historiography of science, at the height of the drive to combine
the two without sacrificing either. In 1935 He le`ne Metzger gave a talk at the Centre
de synthe`se historique titled Tribunal de lhistoire et the orie de la connaissance
scientifique (Metzger, 1935b, pp. 114). Metzgers tribunal is rather different from
Canguilhems. First of all, the terms of the problem are reversed: Metzger addresses
the question of whether history can pass definitive judgements on the theory of
knowledge. If history is to provide definitive answers to epistemological questions
by showing the development of the human mind, Metzger argues, then the theory
of scientific knowledge would no longer belong to philosophy. She declared this
hypothesis untenable. For her, only historians who are a little nave can think that
the mere accumulation of scientific texts can reveal the real development of our
intelligence without any interpretation and critique (Metzger, 1935b, p. 8). Her pos-
ition here could be taken as an attack on some people in the audience. The Centre
de synthe`se, where she was giving her talk, had a full programme of collection of
documents and creation of comprehensive dictionaries aimed at the realisation of
total history or historical synthesis, in the expression of its founder and director
Henri Berr (Berr, 1911, 1921, 1925). Metzger was well aware of history as recon-
struction; in this very talk she declared that history is nothing outside the intelligence
of the historian, and physics is nothing outside the physicists intelligence (Metzger,

the chair in history of philosophy, 6 in history of modern philosophy, 6 in history of ancient philosophy,
2 in philosophy and history of philosophy, and 1 history of philosophy in its relationship with the sciences,
for a total of 19 chairs. In the same period, the other philosophy chairs were: 8 chairs of philosophy, 1
of philosophy and opinions of philosophers, 2 of aesthetics, and 2 of psychology and philosophy, for a
total of 13 chairs, to which one could add 3 of science of education, one of which was renamed by
Durkheim science of education and sociology. The doctoral theses in history of philosophy constituted
the single largest group of philosophy dissertations in the 1930s: 161 in 1937, followed by psychology
(71). (Source: Guigue, 1935, pp. 1429.)
8
The metaphor of the school had often been employed by Bachelard; see for instance Bachelard (1993)
[1938], p. 252.
306 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

1935b, p.10). The reactions to Metzgers talk were rather varied: some, like Henri
Berr and Alexandre Koyre , thought that she did not give history enough importance,
others thought the she gave it too much.9
Koyre s interventions, though brief, are of interest here. First of all, he declared
that the history of science could provide philosophy with important lessons, more
than science could. He defended his claim by arguing that by studying different
mentalities, we can discover what is permanent in our mind. Current science for him
could not provide this because, he implied without saying, it is only one possible
version of our knowledge of the world, and we cannot know whether it is the defini-
tive truth. Koyre added that Duhem and Mach were right to try to prove their positiv-
ism through history.
Metzger was equally critical of the philosophers who thought that they could dis-
pense with history. She included the Cartesians in this category, and Auguste Comte.
Comte did refer to history of science, had a theory of history in his law of three
stages, and, as Metzger reminded her audience, was the first to promote a chair of
history of science in France.10 However, Metzger denied that history played any
serious role in Comtes philosophy. Comte, who, she argued, was not inclined to
patient and meticulous historical study, did employ historical examples, but these
only proved what they had been set to prove, that is, the truth of pre-imposed models,
such as the law of the three stages (Metzger, 1935b, p. 4). She concluded her talk
by stating the lessons that history of science could provide: first of all, to cure
philosophy of the strange mania to want to pose a priori or a posteriori definitive
concepts, which for her only satisfied a psychological need of certainty (Metzger,
1935b, p. 13). Moreover, she argued that the awareness of the possible responses to
similar experiences, and of the different consequences drawn from similar hypoth-
eses, should cure one from dogmatism as well as from pointless scepticism.
Although Metzger declared that the observation of the plasticity of the mind was
very important for scientific research itself, it is clear that for her the lessons to be
drawn from history are philosophical ones, about the variety of ways in which our
knowledge of the world can be organized and expressed.
Notwithstanding her firm conviction that history of science could teach us a few
lessons about the mind, Metzger refused to judge the truth content of past theories
by using an external norm. Indeed, her efforts went towards understanding the men-
talities behind doctrines which may at first seem absurd. For instance, it may seem
peculiar that the philosophers of metals never doubted the possibility of turning lead
into gold, despite their frustrating lack of success. However, Metzger argued that
when one considered their world-view, they appeared to be much less irrational than
at first sight. They saw all things in a relationship of analogy. For them lead was to
gold what an infant was to a man: just a less developed, or less perfect version of

9
The minutes of discussion which followed Metzgers paper are published in Archeion. See Fondation
Pour la science, Centre International de Synthe`se, Section dhistoire des sciences (1935), pp. 8184.
For a comment on this session see Redondi (1997)
10
For a history of the chair of general history of the sciences see Paul (1976); and for an account of the
fortunes of history of science in France in the first half of the twentieth century, see Chimisso (2001a), Ch. 5.
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 307

the same thing. The actual transformation of one into the other was something simply
not yet achieved, but perfectly consistent with the order of things (Metzger, 1969
[1923], pp. 108 ff.). Thus, Metzger aimed at finding the internal norm that governed
a certain period of intellectual history. She called this norm the a priori of thought.11
Koyre s position was different from Metzgers, although not completely opposed
to it, especially as we see it in his work on history of science, the publication of
which commenced a few years after that seminar at the Centre de synthe`se. Koyre
too aimed at understanding the ways of thinking of natural philosophers. What he
called conceptual structures is similar to Metzgers a priori, and both these notions
have strong connections with Le vy-Bruhls mentality. Koyre believed that conceptual
structures and world-views changed in intellectual history, and quite significantly
too. The name of Koyre is indissolubly connected with that of the scientific revol-
ution, for he concentrated on the rupture which divided the Greek-medieval closed
world from the open universe of modern cosmology (Koyre , 1958), and the quantitat-
ive approach of Galileo and Newton that broke with the qualitative approach of
previous Aristotelian science (Koyre , 1939, 1965). Nicholas Jardine has argued that
Koyre held that the ontologies resulting from a periods conceptual structures were
real entities, rather than mere useful historical categories (Jardine, 2001, p. 16). To
rephrase Jardines point into Canguilhems language (and theory), Koyre assumes a
series of norms, e.g. Aristotelianism, Platonism and the Scientific revolution, that he
employs to judge texts and ideas, and that do not depend on other norms. Koyre ,
however, sometimes slipped in an unclarified metanorm which allowed him to
judge those very ontologies. For instance, when Koyre judges Giordano Bruno a
very poor scientist (Koyre , 1958, p. 54), or when he declared that Aristotles phys-
ics is false, it is irremediably superseded (Koyre , 1978 [1939], p. 4), he was not
evaluating these philosophies by the standards of their times, but by an external
norm. In these cases, Koyre did judge past theories, as much as the historian Metzger
avoided giving good marks to some scientists and bad marks to others (Metzger,
1935b, p.14).
In the eyes of many historians, Koyre remained a philosopher, and only a philos-
opher. Jacques Le Goff has remarked, while recognizing his attentive reading of
texts, that the great Koyre could not push his investigation further than the analy-
sis of concepts and systems. He concluded that:

. . . he always remained a philosopher, a type of scholar with whom, at least in

France, historians have always had difficulty in establishing a dialogue, due to
their distrust for the philosophy of history. (Le Goff, 1983, p. 412)
In Le Goffs view, Koyre s attentive reading of texts did not automatically make
Koyre a historian in the historians eyes. Although Le Goff did not go into details,
it is not too arduous to guess some of the reasons for his criticism. For a historian
of mentalities like himself, the challenge was to examine the links between ideas
and social realities, while Koyre dispensed with the latter. To Le Goff, Koyre s

11
See, for instance, Metzger (1936).
308 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

analysis of concepts and systems must have appeared the ultimate objective of his
research, while the historians objective should have been that of analysing the his-
torical formation of those concepts and systems, and their possibility of existence and
development in a given time and society. Moreover, although Koyre s conceptual
structure has close links with the concept of mentality and Febvres outillage
mental, in fact it applies particularly to thought in its highest forms (Koyre , 1966
[1951], p. 1). In reality, Koyre only analysed high culture, and the links that he
established were between philosophy, science and religion, the latter conceived as
theology rather than popular expressions of beliefs. By contrast, Febvres outillage
mental was aimed precisely at the overcoming of the idea that high culture had a
life independent of the broader culture, in the ethnological sense of the word.
The practice of careful reading is by no means foreign to some philosophical traditions,
and Koyres education in Germany makes his interest in it quite unsurprising. In the
context of French history and philosophy of science, however, his philological attention
to texts, while in line with an existing historiographical tradition, marks an important
difference between himself and Bachelard. In the split which I argued occurred in the
programme of combining history and philosophy of science, one of the activities that
philosophers such as Bachelard and Canguilhem dropped was precisely the philological
aspect of their research. However philosophical Koyre s interests were, his main aim was
not that of explaining the working of the human mind, but more modestly the intellectual
changes in the scientific revolution, or Galileos metaphysical assumptions.
However, general historians did not see much of a difference between Bachelard
and Koyre ; if anything, Bachelard was probably seen as more of a philosopher and
less of a historian. As evidence of this lack of communication stands the only review
of Bachelards books to have appeared in the Annales. The mouthpiece of the his-
torians of mentalities did not as a rule host the work of the historians and philos-
ophers of science. Roger Chartier has argued that this lack of interest bore negative
consequences for French historiography. For him French historians confidence in
quantitative methods and their faithfulness to their own definition of history of men-
talities prevented them from recognizing the importance of the epistemological
model offered by the work of Gaston Bachelard, Alexandre Koyre and Georges
Canguilhem to the procedures of intellectual history (Chartier, 1988, p. 35).
This lack of communication may come as a surprise especially because the con-
tacts between Febvre and the historians of science knew no practical obstacles;
indeed Febvre was on the management committee of the Institut dHistoire des
Sciences et Techniques (Thale`s, 1 (1934), p. xi) and for two years (19431945) he
was in the same department (Religious Studies) of the Ecole Pratique des Hautes
Etudes as Koyre . He was also an important member of the Centre de synthe`se, but
rarely ventured to the Unit for history of science.12 The weakness of the relations

12
Febvre did attend Metzgers talk on Les diffe rents aspects de la me me e poque dune civilisation
(lettres, sciences, arts) peuvent-ils e tre conside re s comme autant de projections varie s dun me me e tat
desprit? Ou au contraire leurs modifications diverses ont-elles agi individuellement sur le volution de
cette civilisation en ge ne ral? (Fondation Pour la science, Centre International de synthe`se, Section
dhistoire des sciences, 1930).
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 309

between Febvre and the historians of science should, however, be qualified, not
least chronologically. Febvre did collaborate with Abel Rey. They edited the Revue
de synthe`se historique together, and Febvre, as general editor of the Encyclope die
franc aise, entrusted the first volume, on the outillage mental, to Abel Rey (together
with Antoine Meillet and Paul Monteil). In Le proble`me de lincroyance au XVIe
sie`cle (Febvre, 1942), Febvre quoted Rey on various matters, including Greek science
(pp. 393, 403, 432, 437), Duhem on physics (p. 384), Lynn Thorndike (pp. 382,
396), and other historians and philosophers of science.
In a review of works in history of philosophy, Febvre lauded Abel Reys collabor-
ators for their work of editing and translating philosophical texts, for they paid
attention not only to pure ideas, or their logical filiation, but also to their historical
emergence and their relationship with the wider context (Febvre, 1992 [1938], p.
283). However, it is quite clear that the series he was reviewing was for him the
exception rather than the rule as far as the methodology of historians of philosophy
was concerned. He thus excluded, I suspect, most historians of philosophy and of
science from the community of historians:

Of all workers who cling to the generic title of historian, with or without a qualify-
ing adjective, there are none who fail to justify some part of it in our eyes save,
all too often, those who, applying themselves to rethinking for their own purposes
systems that are sometimes several centuries old, with not the slightest effort to
show their connections with other manifestations of the epoch which saw their
rise, end up doing precisely the opposite of what historical methods demands.
Then, faced with these concepts engendered by disembodied mindswhich take
on a life of their own outside time and spacethey forge strange chains the links
of which are both unreal and closed.13

Febvres criticism seems rather unfair if applied to all of his contemporary historians
of philosophy, and to those philosophers who had turned their attention to the history
of science. It is undeniable, though, that Febvre had little patience for the historians
of sciences theoretical discussions, as showed by his interventions at Metzgers talk.
A fortiori it is difficult to imagine a collaboration between him and Bachelard or
Canguilhem, that is to say, after the normative turn that transformed history of
science into historical epistemology. In the range of orientations of the general his-
tory of science, Febvre would have been interested only in the more historical and
philological, whereas the more philosophical and epistemological ones were going
to succeed. In Le Goffs judgement, Bachelards work has had the same reception,
or lack of it, that Koyre s had; he points out that Bachelard, despite his desire to
write history, has appeared, rightly or wrongly, to historians as a philosopher rather
than a historian (Le Goff,1983, p. 412).
From the 1940s onwards, normative history of science opened up a new style of
research, significantly contributing, in the opinion of Foucault and many others, to

13
Febvre (1992) [1938], p. 278, quoted in modified English translation from Chartier (1988), pp. 2223.
310 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

the most important philosophical debates of post-war France. I shall now investigate
the differences between Bachelards and Canguilhems approaches in order to shed
more light on Canguilhems conception of norm and on how he applied it to history.

4. Bachelards and Canguilhems normative history

Canguilhem advocated the normative approach to history of science in several

articles and talks. When he wrote his theoretical remarks on history of science and
epistemology (Canguilhem, 1993 [1977], 1963, 1994 [1968]), the normative turn in
my view had been fully realized. The epistemological approach had disentangled
itself from the ideal of total history and had chosen normativity over a relativistic
and philological approach. The goal of combining the historical craft with philosophi-
cal aims had been abandoned, and the tension between these two poles of history
of science had been resolved. It had been resolved thanks to the success of the
epistemological approach in the hands of Gaston Bachelard, closely followed by
Canguilhem. In his theoretical essays, Canguilhem indeed relied on the ideas and
authority of his predecessor in the Sorbonne chair of history and philosophy of the
sciences. Indeed, Canguilhem made important reflections on this type of history in
articles on Bachelard. Bachelards and his own method and authority were too well
established to call for a polemical advocacy. Rather, at that point, that is, in the late
1960s and 1970s, Canguilhem was able to show a coherent approach which had
already evolved in the hands of successive scholars, first of all Michel Foucault.
In the Introduction to Ide ologie et rationalite , titled Ro le de le piste mologie dans
lhistoriographie scientifique contemporaine, Canguilhem argues that, without a nor-
mative point of view, the historian of science cannot construct any history. Without
a norm of her own, the historian would consider botanical what was considered to
be botanical in the past. Yet, eighteenth-century botany, he argues, does not necessar-
ily constitute the past of modern botany. Indeed, Canguilhem claims that taken in
its absolute sense, the past of a science is a vulgar concept (Canguilhem, 1993
[1977], p. 13).
In his view only epistemology can solve this problem:

the history of science is entitled to expect from epistemology a set of ethical

criteria, by which I mean a set of criteria for judging which moves within the
vast expanse of the past are legitimate and which are not.14

What does Canguilhem mean by legitimate moves? In order to understand this, I

think it is worth looking at La formation du concept de re flexe aux XVIIe et XVIIIe

14
Canguilhem (1993) [1977], p. 14; in English translation, Canguilhem (1988b), p. 4. I quote here
from the English translation of Ide ologie et rationalite for its clarity. However, this translation is not
close to the French original: On peut penser que ce que lhistoire des sciences est en droit dattendre
de le piste mologie, cest une de ontologie de la liberte de de placement re gressif sur le plan imaginaire du
passe inte gral.
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 311

sie`cles (Canguilhem, 1955). In this book, Canguilhem aimed to disprove what he

called two prejudices. The first is that a concept can only emerge in the context of
a theory, or at least within a heuristic inspiration which is homogeneous with the
theory that will later explain the observed facts (Canguilhem, 1955, p. 3). In other
words, this prejudice denies that a concept employed in a current scientific theory
could have emerged within a theory which is now deemed non-scientific, or which
is incompatible with the current theory. The second, more specific, prejudice is that in
biology only mechanistic theories have led to positive results and fruitful applications
(Canguilhem, 1955, p. 3). For Canguilhem, these two prejudices have distorted the
history of the concept of the reflex. He therefore set out not only to reconstruct a
more truthful history of this concept, but also to use his version of facts as evidence
against those two assumptions. The first prejudice Canguilhem eliminated was the
claim that Descartes provided the first coherent formulation of the concept of the
reflex. He argued that this could not be the case because, in Descartes physiology,
the flux of spirits in involuntary bodily movements is always from the brain towards
the periphery, and never in the opposite direction. Rather, for Canguilhem, the first
to propose the concept of the reflex was in fact Thomas Willis (16211675), pro-
fessor of natural history at Oxford, and of medicine in London. Willis used this
concept within a rather imaginative theory in which life is interpreted as light. The
metaphor of light brought about his use of the optical laws of reflection in the
interpretation of biological phenomena. The concept of reflex then came as a natural
consequence of Willis general theory. Canguilhem argued that Willis had been for-
gotten by historians because his theory had been doomed to be judged untenable.
Willis reputation was the victim of the prejudice according to which the truthfulness
of concepts depends on the truthfulness of the theories of which they are part. Cangu-
ilhem used this historical example to prove his theory that, in fact, concepts are
independent, in their validity, of the theory in which they are embedded.
In La formation du concept de re flexe, Canguilhem blocked the move between
Descartes notion of reflex movement and the modern concept of reflex, while
allowing the one connecting the latter with Willis concept. Canguilhems assumed
starting point was the current concept of reflex: he employed this to assess whether
past concepts are related to it. His answer was negative in the case of Descartes,
and positive in the case of Willis. In a Bachelardian manner, Canguilhem isolated
the epistemological obstacle that prevented Descartes from having a concept of
reflex, that is, Descartes general physiology. Canguilhems argument against the
continuity of the modern concept of reflex with Descartes was not based on histori-
cal evidence, but on the analysis of the internal consistency of the various doctrines.
He defended his approach when he wrote that in matters of the history of the
sciences the rights of logic should not be replaced by the rights of the logic of
history (Canguilhem, 1955). The logic of history may suggest that the modern
concept of reflex had its origins in a mechanistic theory. However, Canguilhem
aimed to show that the modern concept of reflex would have been inconsistent with
Descartes mechanistic physiology: the internal logic of Descartes doctrine dis-
proves, in his view, what the logic of history suggests. What Canguilhem was
interested in is the evaluation of past concepts on the basis of present ones, rather
312 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

than a study of the past in its own terms. His method should not be confused with
a recurrent and anachronistic historiography, for it is indeed the opposite. Although
he did read past texts with modern science in mind, he did not anachronistically
force them to fit into it. He thought that his analysis of past concepts would serve
to indicate which links could be established. His method was most effective in its
negative part, when it allowed him to establish that certain continuities were
delusions created by recurrent reading of past texts.
One of the differences between Bachelard and Canguilhem that immediately
strikes the reader is that the former emphasises discontinuities, while the latter estab-
lishes both continuities and discontinuities, as for instance in La formation du concept
de re flexe. Franc oise Dagognet has argued that Canguilhem regarded discontinuities
as partial and complex, as exemplified in his criticism of Alexandre Koyre s
interpretation of the scientific revolution (Dagognet, 1997, p. 162). Canguilhem took
exception to Koyre s claim that Galileo rejected all forms of Aristotelianism in fav-
our of an Archimedean and Platonic world-view. Canguilhem agreed with Ludovico
Geymonat that Galileo kept important aspects of the Aristotelian tradition
(Canguilhem, 1993 [1977], p. 25).
It is interesting that Canguilhem criticised Koyre rather than Bachelard for exag-
gerating epistemological ruptures. Bachelard was not only Canguilhems mentor and
predecessor in the Sorbonne chair; he was the scholar who had given epistemological
history its crucial place in the French intellectual landscape. In the division of camps
following the split (which had been complex, slow and contradictorynot a sudden
epistemological rupture) of French history of science, Koyre occupied a slightly
ambiguous position. His links with the Sorbonne, which had been significant before
the second world war, weakened considerably after the war. From a methodological
point of view, Koyre did not give up a thorough attention to texts and his aim was
to understand an author or period rather than reaching general epistemological or
historiographical conclusions. On the other hand, Koyre s work does have a strong
and explicit philosophical framework, and proposes a grand narrative that emphasises
epistemological ruptures. For the latter reason, Canguilhem sometimes mentioned
Koyre as close to Bachelard in his conception of history of science, even though
he then stressed their differences (see, for instance, Canguilhem, 1994 [1968], pp.
1314).
The link between Bachelard and Canguilhem has been seen as so strong that this
obvious difference in their views of historical discontinuities has demanded an expla-
nation. Dagognet has argued that their divergence originates in the sciences that
they respectively studied. Bachelard analysed mature sciences, especially physics
and chemistry, while Canguilhem devoted himself to the history of the life sciences.
The latter, Dagognets argument goes, were still in their infancythat is, in a
moment of slow evolution, when continuities are stronger (Dagognet, 1997, p. 162).
In this observation, Dagognet follows Canguilhem himself who in turn reported
Bachelards view (Canguilhem, 1994 [1968], p. 14).
Although Dagognet points to an important feature of Canguilhems work, this
does not constitute, in my view, a full explanation of his attention to continuities. I
believe that the implications of his difference from Bachelard are much more
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 313

important than this simple distinction would suggest. First of all, it is worth noting
that it is not just the sciences under study that differ, the focus of their works does
too. This is quite evident in the two Formations: Canguilhems La formation du
concept de re flexe and Bachelards La formation de lesprit scientifique. The breadth
of scope could not be more at variance: Canguilhem analyses a specific concept
within physiology, Bachelard nothing less than the scientific mind. Moreover, Cangu-
ilhem affirms the independence of concepts from the theory in which they are
inscribed, so that Willis concept of reflex can be seen as the seed of the modern
concept, even if his physiological theory is completely at odds with modern science.
In other words, Canguilhem establishes continuities across incompatible theories,
indeed across different world-views.
By contrast, in his Formation, Bachelard aims to show that there is a break
between pre-science and science. Most of the examples employed in La formation
de lesprit scientifique are drawn from the eighteenth century, which Bachelard
characterises as part of the pre-scientific period, which for him stretches from
antiquity to the end the eighteenth century. He argues that no science was possible
in that period. For Bachelard scientific knowledge emerges by contradicting previous
conceptions. For instance, while the study of metals was dominated by sexual ima-
gery, no chemistry could develop; alchemy flourished instead, with its interpretation
of metals as gendered and of mixtures of substances as copulation. Bachelard denies
that a scientific concept, or the root of a scientific concept, could emerge not only
within a non-scientific theory, but in general within a non-scientific mentality. In
what he calls the pre-scientific period the whole approach to nature could not be
conducive of objectivity, as it was dominated by an imaginative world-view, guided
by subjective needs and desires. Moreover, scientific discourse for Bachelard is only
produced socially, through discussion, rather than by isolated individuals. The social
production of scientific discourse requires rational and objective relationships
between individuals that for Bachelard did not take place in the pre-scientific period.
Private relationships, based on personal authority, cannot generate objective knowl-
edge, and, before the nineteenth century, he can observe neither the psychological
nor the social conditions for science. He warns that, though pre-scientific concepts
may seem similar to scientific ones, actually they are at odds with anything scientific,
since they are the results of subjective, imaginative and a-rational approaches. In his
view, the modern sciences did not inherit any positive contents from natural philo-
sophy of any kind. Rather, they developed by opposition to previous knowledge.
For Bachelard, without the dialectical overcoming of epistemological obstacles, no
science is possible, and therefore no scientific theory and no scientific concept. It is
safe to conclude that the continuity between Willis concept of reflex and the modern
one established by Canguilhem is in contradiction with Bachelards theory of the
development of science.
Canguilhems view of discontinuity is certainly more complex, more partial and
fragmented than Bachelards, and does not necessarily take the shape of an overall
change of mentality and social relationship as does Bachelards. Is the difference in
their views a consequence of the fact that the former studied mature sciences while
the latter studied young sciences? Canguilhem establishes continuities, such as that
314 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

of the concept of reflex, that go back four centuriesmuch longer than any continuity
accorded by Bachelard to physics and chemistry. For Bachelard no eighteenth-cen-
tury observation has originated a nineteenth-century technique (Bachelard, 1993
[1938], p. 71). Paradoxically, Canguilhems physiology, or at least some of its con-
cepts, is much older than Bachelards physics. Although this difference in their object
of study is very important, I believe that it is insufficient to resolve their divergence
on the issue of the epistemological rupture.

5. Historical epistemology and the living being

In La formation du concept de re flexe, the norm is provided by current science.

There is no discussion, though, of how science produces normative concepts, and
whether their origins lie inside or outside science. The question of how a norm
emerges remains unanswered. However, when La formation was published, Cangu-
ilhem had already written a history of the concept of norm in his Le normale et le
pathologique. This book well exemplifies the tenets of historical epistemology: its
main goal is to discuss and clarify the concepts under study, their origin and their
applications; and the way of investigating them is historical. The book is divided
into two parts, aimed to answer the questions posed in their titles: Is the pathological
state merely a quantitative modification of the normal state? and Do sciences of
the normal and pathological exist?. Canguilhem describes the first part as research
of historical sources, and the second as the critical presentation of his own doctrine.
Indeed, the chapters of the first part focus more specifically on past doctrines, from
references to the Egyptian and the Hippocratic tradition to in-depth discussions of
Auguste Comtes, Claude Bernards and Rene Leriches ideas. The material of the
second part is arranged according to issues: first an examination of the concepts of
normal, anomaly and disease, and of normal and experimental, then discussions of
the relation between the norm and average, between disease, cure and health, and
between physiology and pathology.
However, the method that Canguilhem follows in the two parts is not radically
different: in both cases it is a critical dialogue with other doctrines. The first part is
not historical in the sense in which most historians would understand history: there
is no attempt to build a historical narrative, or to reconstruct a historical milieu.
What Canguilhems approach does have in common with that of Metzger, Koyre
and the historians of mentalities is that it links conceptions of illness and health to
more general metaphysical assumptions. A good example of this method is Cangu-
ilhems treatment of the change from a qualitative to a quantitative conception of dis-
ease.
Canguilhem explains that the quantitative conception of disease, the champion of
which is Claude Bernard (18131878), is that according to which the pathological
state is a quantitative variation of the normal state. For instance, the presence of
sugar in the bloodstream is a normal physiological phenomenon. In the diabetic, the
quantity of sugar is higher: the pathology consists in a quantifiable variation of a
normal state, rather than in a qualitative difference. This conception differs from the
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 315

Hippocratic view of disease, whereby illness is seen as a state of disequilibrium

qualitatively distinct from the harmonic state of health. It is also at odds with an
ontological view of the disease, exemplified in modern times by the microbiological
theories of contagious diseases (Canguilhem, 1999 [1966], p. 12).
The emergence of the nineteenth-century quantitative conception of disease for
Canguilhem rests on a series of philosophical premises which emerged from the
Renaissance onwards. The more general one concerns the relationship between man
and nature. In the Hippocratic view, human beings were part of nature, and their
internal equilibrium was the same as that outside them (Canguilhem, 1999 [1966],
p. 12). With Francis Bacon, the unity of human beings and nature was broken; since
then, the dominant view was that human beings could and must intervene in nature
in order to achieve their goals (Canguilhem, 1999 [1966], p. 13). Intervention in the
course of nature was aimed at restoring a desirable state of affairs. In the case of
medicine, this was health, hence the need to have a good knowledge of the healthy
state. The goal of intervention made physiology become the basis of pathology,
which eventually became a branch of the former. The identification of physiology
and pathology made the qualitative view of disease untenable, and lead, for Cangu-
ilhem, to the quantitative view (Canguilhem, 1999 [1966], pp. 1314).
In Canguilhems view, the quantitative conception of disease not only needed a
Baconian view of the relationship between man and nature, but also the dissociation
of science and religion, and the emergence of nineteenth-century rationalistic opti-
mism. This optimism, Canguilhem explains, denied the reality of evil and the asso-
ciated Manichean view that Health and Disease fought over man the way Good
and Evil fought over the World (Canguilhem, 1999 [1966], p. 61). Claude Bernards
quantitative and optimistic view also rested on the positivism of its time, exemplified
in particular by Comte, according to which a technology is the application of a
science, as summarized in the motto: to know in order to act.15 Following this
general assumption, physiology explained pathology which in turn provided the basis
of a therapeutics. This implied the monistic idea of the qualitative indifferetiation
of illness and health.
Following a methodology similar to that of Metzger and Koyre , Canguilhem links
historical concepts of health, disease and normality to broader world-views and philo-
sophical systems, but, unlike them, his scope is pre-eminently epistemological and
normative. Canguilhem analyses past doctrines not for their own sake, but in order
to reach conclusions about the concepts of the normal and the pathological. In his
own words, the historical part of his book served to establish the narrowness and
inadequacy of the principle of pathology . . . according to which the morbid state in
the living being is only a simple quantitative variation of the physiological phenom-
ena which define the normal state of the corresponding function.16
In order to understand Canguilhems own concept of norm, his distinction between
norm and average is particularly illuminating. He discards the theories which equate

15
Canguilhem (1999) [1966], p. 58; in English translation, Canguilhem (1989), p. 99.
16
Canguilhem (1999) [1966], p. 155; in English translation, Canguilhem (1989), p. 277.
316 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

norm and average. In his view, the norm cannot be calculated from a collection of
data, because it represents a desired state of affairs rather than an average. Cangu-
ilhem insists that the norm refers the real to values, and therefore it cannot be reduced
to an objective concept that we can determine scientifically, either in physiology or
elsewhere. Human beings are normative, which means that they establish norms, and
discard old norms in favour of new ones (Canguilhem, 1999 [1966], p. 106). In other
words, they are not merely the expression of norms, or the embodiement of given
biological realities (as in the normal man), but rather they create their own norms.
Indeed, if Canguilhem has a concept of human nature, it seems to lie precisely in
the human ability to adapt to new situations by establishing new norms.
To illustrate what Canguilhem means, one can say that an individual has a concept
of good health and fitness and she judges her own state by this normative concept.
This concept of good health does not correspond to a normal state of affairs, for
most people are not continuously, if ever, in this state of good health and fitness.
However, as a rule human beings do not aim at being average, for example trying
to catch an average number of colds and flus, but generally, although not necessarily,
they try to get as close as possible to the normative concept of good health. This
normative concept of good health varies depending on a variety of factors. An obvi-
ous one is age: a seventy-year-old would not have the same norm as a twenty-year-
old. Canguilhem especially emphasises that what counts as normal depends on the
creative way in which groups of human beings have adapted to their environment
and life circumstances. For instance, he believes that ethics and religion play an
important role in human beings determination of normality. He cites the example
of yogis, who are able to alter their vegetative functions in a way that would be
dangerously pathological for other people. In the yogis case, there is nothing patho-
logical: they have simply instituted different norms (Canguilhem, 1999 [1966], pp.
106107).
Canguilhem even claims that illness can be seen as a type of normality. This
apparently paradoxical claim is explained by bearing in mind that for him illness
also refers to norms, rather than being just a neutral disorder. A diabetics life is
regulated by norms created by her new relationship with her environment. What
really changes in the pathological state is that the patient is dominated by these
norms and looses the ability to adapt to new situations (Canguilhem, 1999 [1966],
p. 120). The process of healing consists for Canguilhem in imparting new norms on
life, norms which are clearly superior to the old ones (Canguilhem, 1999 [1966], p.
156), for patients, once recovered, are able to respond more positevely and creatively
to their environment.
This creative response to the environment is particularly important because the
human environment is unstable and requires different responses at the different times.
In his words: health is a margin of tolerance for the inconstancies of the environ-
ment. Canguilhems defence of his own view that the environment is inconstant is
very interesting because it is centred on the obvious, but often neglected, fact that
the scientific image of the world (the world of laws and regularities) is not the world
in which human beings live. His observation deserves a long quotation:
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 317

Health is a margin of tolerance for the inconstancies of the environment. But isnt
it absurd to speak of the inconstancy of the environment? This is true enough of
the human social environment where institutions are fundamentally precarious,
conventions revocable, and fashions as fleeting as lightning. But isnt the cosmic
environment, the animal environment in general a system of mechanical, physical
and chemical constants, made of invariants? Certainly this environment, which
science defines, is made of laws but these laws are theoretical abstractions. The
living creature does not live among laws but among creatures and events which
vary these laws. What holds up the bird is the branch and not the laws of elasticity.
If we reduce the branch to the laws of elasticity, we must no longer speak of a
bird, but of colloidal solutions. At such a level of analytical abstraction, it is no
longer a question of environment for a living being, nor of health nor of disease.
Similarly, what the fox eats is the hens egg and not the chemistry of albuminoids
or the laws of embryology. Because the qualified living being lives in a world of
qualified objects, he lives in a world of possible accidents. Nothing happens by
chance, everything happens in the form of events. Here is how the environment
is inconstant. Its inconstancy is simply its becoming, its history.17

For Canguilhem, normality is a qualitative concept rather than a quantitative one,

and so is abnormality. Just like the normal, what counts as abnormal, or pathological,
for Canguilhem refers to values rather than facts. Pathology involves pathos, suffer-
ing and disempowerment, whereas anomaly does not. The difference between abnor-
mality and anomaly then lies on the subjective experience of the sick person, and
on the values she attaches to her state. Similarly, Canguilhem thinks that the human
bodys normal state that medicine aims to restore is normal by virtue of the fact that
the sick person deems it to be normal (Canguilhem, 1999 [1966], p. 77). Subjective
here does not merely mean individual. Indeed, Canguilhem rightly thinks that values
are generated socially. Notwithstanding his discussion of the difference between bio-
logical and social norms,18 he believes that biology and sociology are inextricable.
This is clear in his evaluation of biological facts in social and cultural terms, which
is strictly connected with his conviction that a human trait is normal not because is
statistically frequent, but is frequent because it is normal. In other words, if, in a
certain form of life, something is considered to be the norm, it will occur more
frequently. For instance, life expectancy depends on social norms of hygiene, work
conditions, attention to fatigue and disease. Canguilhem concludes that:

the techniques of collective hygiene which tend to prolong human life, or the
habits of negligence which result in shortening it, depending on the value attached

17
Canguilhem (1999) [1966], pp. 130131; in English translation, Canguilhem (1989), pp. 197198.
18
See in particular the first of the three Nouvelle re flections appended to Le normal et la pathologique,
written twenty years after the first edition of the book: Du social au vital (Canguilhem, 1999, pp.
176191).
318 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

to life in a given society, are in the end a value judgement expressed in the abstract
number which is the average human life span.19

Canguilhem also maintains that social norms may have their origin in biological
norms. For instance, he argues that there would be no medicine if there were no
biological norm that makes organisms preserve themselves. He does not attempt an
impossible and ultimately pointless distinction of social norms and biological ones.20
Canguilhem believes that life poses values on its own environment and also on the
organism itself. It is then life that generates the norm. What Canguilhem means
by life is an undivided whole including biology as well as sociality, rationality as
well as emotions. The normal is a value and not a fact, and this value is established
by the organism, or the human being in his or her entirety: as a biological, social,
rational and emotional being.
Le normale et le pathologique is a normative book about the norm and normativity.
Is the normativity Canguilhem discusses in Le normal et la pathologique the same
normativity that he employs in his practice of normative history? Is his historical
epistemology based on the concept of norm exposed and espoused in his writing on
medicine? Once again, a comparison with Bachelard will shed light on these issues.

6. Medical norms and historical norms

Canguilhem defines the concept of normative in the following manner:

Normative, in philosophy, means every judgement which evaluates or qualifies a

fact in relation to a norm, but this mode of judgement is essentially subordinate
to that which establishes norms. Normative, in the fullest sense of the word, is
that which establishes norms.21

This definition naturally provokes the question of what establishes the norms. In
an article on Canguilhem, Michel Fichant asks a similar question: where does the
norm come from? (Fichant, 1993, p. 38). Fichants answer to his own question is,
in brief, that for Canguilhem the norm comes from the living beings values. The
living being creates values which establish norms. Fichant contrasts Canguilhems
position with Bachelards, according to which the norm is given by mathematization.

19
Canguilhem (1999) [1966], p. 103; in English translation Canguilhem (1989), p. 161.
20
Malcom Nicolson has argued that Canguilhems thesis of the continuity of biological and social
norms would provide a solid basis for current sociology of scientific knowledge. He maintains that not
only can [Le normal et la pathologique] help us towards a more sophisticated account of the materialist
basis of scientific knowledge; it also provides us with grounds from which to argue that the acknowl-
edgement of such a materialist basis need not entail any weakening of the strength or scope of the
sociological approach to scientific knowledge in general or of relativism or constructionism in particular
(Nicolson, 1991, p. 361).
21
Canguilhem (1999) [1966], p.77; in English translation, Canguilhem (1989), pp. 126127.
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 319

Fichants comparison take us back to the question of the objects of normative

histories. For Bachelard current science sets the norm. He focuses on physical
sciences, and therefore current physical sciences, which are quantitative, provides
his norm of scientificity, and consequently of his historical epistemology. Bachelards
view of physical sciences is that they are not only obviously quantitative, but also
rational, objective and based on a dialogue between practitioners. He argues that
their rationality is the result of a process of purification from irrational and emotional
perspectives that practitioners would instinctively bestow on them. For him their
objectivity stems from the same process of rationalisation: he identifies subjectivity
with imagination, desires and private goals; only by removing the latter does he
think that knowledge can achieve objectivity. In his view the quantitative and social
character of science are indissolubly linked; in his words: The isolated individuals
science is qualitative. Socialized science is quantitative (Bachelard, 1993 [1938],
p. 242).
Like Bachelard, Canguilhem thinks that philosophy should follow science: at the
beginning of Le normal et la pathologique, Canguilhem claims not only to take
current science as a benchmark, but also to aim at an integration of scientific results
into philosophy.

The present work is . . . an effort to integrate some of the methods and attainments
of medicine into philosophical speculation . . . we want to contribute to the
renewal of certain methodological concepts by adjusting their comprehension
through contact with medical information.22

Canguilhems view of the relationship between science and philosophy is here less
univocal and binding than Bachelards. Moreover, Canguilhem aims to integrate into
philosophy methods and attainments of medicine and the life sciences, rather than
physics and chemistry. However contingent Bachelards and Canguilhems respect-
ive choices of sciences may have beendepending, inter alia, on their biography
they engendered a philosophical difference of crucial importance. First of all, in the
sciences that Canguilhem examines, the object is in a position to establish norms,
that is, to participate in the creation of science itself. This is most evident in medicine
and psychiatry. Indeed, the very existence of medicine depends on the patients valu-
es:

we think that medicine exists as the art of life because the living human being
himself calls certain dreaded states or behaviours pathological (hence requiring
avoidance or correction) relative to the dynamic polarity of life, in the form of
negative value.23

22
Canguilhem (1999) [1966], p. 8; in English translation, Canguilhem (1989), p. 34.
23
Canguilhem (1999) [1966], pp. 7677; in English translation, Canguilhem (1989) p. 126.
320 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

The objects of these sciences are also subjects, and their values guide scientific
development:

The sick person is a Subject, capable of expression, who recognizes himself as

a subject in all that which he does not know how to designate other than with
possessives: his pain and the representation that he makes of it, his angst, his
hope and his dreams. (Canguilhem, 1994 [1968], p. 409)

Angst, hope and dreams are not then epistemological obstacles to be overcome, as
they are for Bachelard. In medicine and psychiatry as conceived by Canguilhem,
feelings and emotions contribute to creating the field of intervention. A state of
affairs is pathological, and calls for medical intervention, because the patient experi-
ences suffering. For Canguilhem, the patients experience is irreducible to any scien-
tific objectivity. However, in his view, the patients subjectivity and conception of
illness are not merely individual. Rather, they are consequences of culture and his-
tory (Canguilhem, 1994 [1968], p. 409). He points out that modern western individ-
uals, who live in industrial and democratic societies do not share Pascals view
that illness is human beings natural state, while health is a danger for their souls
(Canguilhem, 1994 [1968], p. 409). Obviously, these different cultural views produce
different conceptions of medicine and of the proper extent of its intervention.
Canguilhem agrees with Bachelard that norms are produced socially. However his
meaning of society does not overlap with Bachelards. When Canguilhem spoke
of socially produced norms, he referred to any society, as community of people,
while Bachelard intended a scientific community. In Bachelards scientific city the
culture is one of rationality and objectivity (Bachelard, 1986 [1949], Ch. 7). In
society at large, the production of norms is far more comprehensive and complex.
Canguilhem recognises that feelings, emotions and desires are also dependent on
culture and history. Fear of illness, or even endurance of pain, can be quite different
in different cultures, in which different values are attached to the presence or absence
of illnesses and pain. By contrast, for Bachelard scientific norms are created only
by scientific activity, by denying and contradicting the norms created by everyday
life and experience. A possible consequence of Canguilhems arguments is that the
establishment of norms is not confined, as for Bachelard, to realm of the experts,
and is potentially more broad-based and democratic. However, this is not quite the
consequence that Canguilhem drew. He called for a revision and enlargement of
scientific rationality so as to consider the irreducible subjectivity of patients, as well
as the limits of scientific rationality, but he never thought that non-experts should
take their health into their own hands. His comments on the diffusion of pre-rational
medicine and the growing refusal of official medicine that he saw in the 1970s are
revealing of his position. He thought that they were evidence of the need for a
reflection on medical rationality and on its limits, but he called them an amalgam
of banalities (Canguilhem, 1994 [1968], p. 401).
Canguilhems conclusions are not, in my view, simply a consequence of his study-
ing life sciences, but rather of his conception of what life sciences should be. He
advocated a type of medicine and psychiatry in which patients feelings and emotions
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 321

and their subjectivities are recognised in their fundamental importance.24 This is the
reason why Canguilhem normatively rejected quantitative conceptions of pathology.
Quantitative versions of pathology aims to fully objectify illness, that is, to reduce
it to quantitative variations which can be dominated and manipulated by the prac-
titioner. In these views, the patient becomes a mere object and does not play a role
in the shaping and progress of the discipline.

7. Conclusion

Are the norms in the life sciences as presented in Le normale et le pathologique

and the norms of his history of science the same? Normative history as Canguilhem
practised it in La formation du concept de re flexe does not suggest that the norms
which Canguilhem employed to judge past science came from life, as did those
described in Le normale et le pathologique. Canguilhem based his history of the
concept of reflex on the current scientific concept and employed logic rather than
strictly historical considerations in order to carry out his evaluation of the links
between past and present concepts of reflex. In La formation du concept de re flexe
the norm comes from current science, as its aim is to judge which past concept is
consistent with the modern one. However, in Le normale et le pathologique, Cangu-
ilhem aimed to show that norms on which science is based are instituted by life.
The norms employed by Canguilhem in his normative history are therefore directly
those of science, but indirectly those that social and biological life creates.
Nonetheless, the concept of norm that Canguilhem defended in his book does not
belong to science in the same way in which the concept of reflex does. Whereas the
concept of reflex can be scientifically defined, the concepts of norm, normality and
pathology are, so to speak, regulative ideals that guide the life sciences. Canguilhem
was of course well aware of this, and indeed concluded Le normale et le pathologique
by saying that the concept of norm cannot be scientifically established. Similarly,

24
Canguilhems view of the concept of normal in medicine and psychology explains his participation
in a conference that, in the words of its organizer, was prompted by the need to discuss a reform of the
national health service in Italy (Manuali, 1988, p. 7); Canguilhems paper was on Le statut e piste molo-
gique de la me dicine (Canguilhem, 1988a). This conference, titled Medicine and Epistemology. Health,
Disease and Transformation of Knowledge was held in Perugia (Italy) in 1985 and chaired by the psy-
chiatrist Carlo Manuali. In the mid-Sixties, Perugias psychiatric institutions participated in one of the
pioneering applications of ideas coming from the new psychiatric movement, aimed at the de-insti-
tutionalization of the mentally ill and at a re-definition of mental illness. The director of this experiment
which was going to find a more widespread application in the Italian reform of the laws governing
psychiatric care in 1978was Manuali himself. In his Introduction to the Perugia conference, Manuali
declared that the new legislation envisaged an improved service that was more equitable, more accessible
and more controllable by the community. Moreover, its aim was to promote a collective health culture
that would counteract the tendency in sanitary policy to medicalize increasingly large parts of individual
behaviour and experience . . . The greatest obstacles to the reform, or at least the obstacles that I shall
discuss here, lie in medicine itself. They are inherent to its epistemological structure and related to the
ways in which medical knowledge defines its subject, and organizes and transmits its knowledge
(Manuali, 1988, p. 7).
322 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

he argued that there could be no objective method to determine what counted as

disease, as this would always depend on the subjective experience of the sick person.
If this is the case, and indeed I think that Canguilhem is correct, he is implicitly
vindicating a role for philosophy that is quite independent from science. Philosophy
appears as a discipline capable of studying the space and relation between life and
science. Canguilhem aimed to show that the concepts, and indeed the raison de tre,
of science are produced by life as a whole, and not only by rationality, as Bachelard
wanted.25 The relation between life and science can only be analysed by philosophy,
which can apply its critique to both. At a general level, Canguilhem retained the
traditional role of philosophy (at least since Kant) as critique, but his is an open-
ended critique. This for two reasons: because of the historical character of his epis-
temology, and because of his type of rationalism, which is very similar to Bachel-
ards. Their rationalism does not aim to close discussions and to solve problems once
and for all. Rather, Canguilhem often problematized what may have been assumed as
unproblematic, for instance, what disease or health are. He claimed that the aim of
philosophy was that of making life more complicated rather than simpler:

The author maintains that the proper function of philosophy is that of making
human existence more complicated, including the existence of the historian of the
sciences. (Canguilhem, 1993 [1977], p. 139)

But can history of science complicate the philosophers life? My view is that, in
order to develop his philosophical theses, Canguilhem kept history simple. In his
work, history is at the service of philosophy: he judges a selection of past concepts,
without presenting them in a historical contest. Canguilhem evaluates Willis concept
of reflex and Descartes concept of involuntary movement in relation to their general
physiological theories, but he does not explain which cultural values support them,
on how these values came about. Canguilhem approvingly cites Henry Ernest Sigerist
who said that all changes in the medical culture have been conditioned by the changes
in contemporary ideas. (Canguilhem, 1999 [1966], p. 61). However, Canguilhem
does not offer a historical treatment of the society and culture in which the ideas that
he examined developed, and he does not link these ideas to any place in particular. In
other words, he does explain the origins of current norms, but does not give an
account of the formation of past norms. His interest lies in the present, not in the past.
Canguilhem argues that rationality does not come all from above; indeed he
argues that the rationality of expertsof scientists and techniciansdepends on
rationality from below, on the rationality of those who are directly affected by the
advances in science. He cites the example of organ transplantation: this practice

25
Canguilhem only referred to medicine and life sciences rather than all sciences. However, I think
that his argument could be extended to the so-called exact sciences. Certainly their existence depends on
social values attached to scientific research, and indeed its technological uses.
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 323

presupposes a general indifference to the problem of the identity of individuals

with the whole of their organism. (Canguilhem, 1994 [1968], p. 401). This thesis of
the rationality from below is affirmed, implicitly or explicitly, in large part of his
work, and especially in Le normal et la pathologique. He argues that the creation
of norms and values is not the reserve of the experts, but it is rather a complex
process in which every aspect of lifebiological, social, cultural, political and insti-
tutionalcontribute.
Canguilhems view of the formation of scientific concepts is, however, not simply
descriptive but normative. The most important examples of this is his concept of
illness. He thought that what constitutes an illness should be founded on the subjec-
tive experiences of patients, which in turn depend on biological, social and psycho-
logical factors. In his convincing account, this, however, has not always be the case.
In a talk in which his appreciation of Foucaults study on the birth of the clinic is
apparent, Canguilhem recognized that there has been a progressive elimination of
the patients life circumstances from the knowledge of illnesses. He indicates two
roots of this trend. The more apparent, but maybe not the more important, is for
him the colonization of medicine by the fundamental sciences. The second is the
characteristic link between politics and medicine that has existed in industrial
societies from the beginning of the nineteenth century onwards. In these societies,
he argues, a keen attention is given to the health of working people as productive
forces, rather, one could add, than as complete human beings. Political authorities,
Canguilhem continues, establish rules and procedures in order to improve and control
peoples life conditions. A major expression of this aim to control and discipline
peoples health and lives is the hospital. Another expression, in Canguilhems view,
is the rise in prestige of practitioners of specialized branches of medicine and the
decline of that of general practitioners. The former do not focus on individuals as
wholes but rather on specific parts or aspects of their bodies (Canguilhem, 2002
[1989]).
Canguilhems holistic view of how norms come to be established may be the
foundation for a revision of the epistemological break between science and opinion.
In Le normal et la pathologique, scientific norms are not constituted against opinion,
or against every-day desires and impulses, as Bachelard maintained. However, Cang-
uilhem did not aim to eliminate the epistemological break: for him there is science
and there is non-science. Unlike Bachelard, though, the difference between the two
is not determined by the presence or absence of subjectivity and extra-scientific
values and aims. For Canguilhem, science, or at least medicine and life sciences,
are always in a continuum with values and aims that are not produced by science.
However, some of these are legitimate, while others are not. The normativity of
Canguilhems project is expressed in the judgement of values and of their proveni-
ence. For instance, the individuals subjective perception of his or her health should
be the basis upon which medicine operates. This perception, which varies according
to specific cultures, societies and circumstances, becomes part of medicine itself. For
Canguilhem, other values and aims should stay extra-scientific. The aim to control
and discipline peoples lives is not for him an acceptable value, and must be expelled
from the organization of medicine.
324 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

In his article Quest-ce quune ide ologie scientifique? (Canguilhem, 1993 [1977],
pp. 3345), Canguilhem implicitly proposes two epistemological breaks. The first is
that between pre-science and science, along Bachelards lines, that is, where pre-
science is a false science. The other is between science and ideology, which is a sort
of post-science (Canguilhem does not employ this term). Canguilhems scientific
ideology is post-science because is parasitic on an existing science. He cites the
example of Herbert Spencers evolutionism. He observes that Spencer claimed to
have based his theory on Karl-Ernst von Baers embryological research and to have
had it confirmed in Charles Darwins Origin of Species. Canguilhem argues that van
Baers biology and Darwins theory lent scientific credibility to a project of legi-
timation of free enterprise, [and] of the correspondent political individualism and
competition. Here Canguilhems concept of scientific ideology is close to Karl
Marxs classic concept of ideology in that for Canguilhem evolutionistic ideology
works as a self-justification of the interests of a type of society, although he does
not bring in class conflict (Canguilhem, 1993 [1977], p. 43). It is clear that for
Canguilhem the justification of the interests of the industrial society cannot and must
not become part of science.
In Canguilhems work, however, we do not have any historical, or sociological,
study of the formation of rationality from below, or of how scientific concepts are
distorted from above. La formation du concept de re flexe is limited to the examin-
ation of the work of philosophers and scientists. Le normal et la pathologique is a
defence of formation from below of the norms of medicine and psychiatry. Cangu-
ilhem aims to bring the patient back not simply into the process of decision-making
as far as her own health is concerned, but also into the broader formation of the
values and norms that form the very basis of medical knowledge. He chooses, though,
to conduct this defence in philosophical terms, and within ideal discussions with
other philosophers and physicians. Only the experts are embodied in Canguilhems
account; the producers of the rationality from below have neither history, nor social
and cultural identities, not to mention personal ones. They are patients, organisms
or even an expression of life.
The fact that Canguilhem did not substantiate his theses historically does not
detract, however, from their philosophical value. Indeed, some of his ideas and
intuitions can still serve history and philosophy rather well. Precisely the study of
the dynamics from below of the formation of norms, values and concepts that
become part of disciplines, seems to me of particular interest. However, this study,
in order to have real strength, would need to recuperate what historical epistemology
dropped: if not an actual total history, a commitment to an exhaustive historical and
sociological substantiation of any theory of formation of norms and concepts. In
other words, ideas should be re-embodied. Careful analysis of historical settings and
events would also prevent implausible generalizations and the risk of attributing ideas
to entire societies or historical periods without investigating their actual emergence
and roles. In science studies the reconnection of history and society with knowledge
has been pursued in many different manners and from many different angles. How-
ever, philosophy seems to be preserved from the attacks of contingency, of historical
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 325

events and persons.26 The reflexivity that sociologists like Pierre Bourdieu have
brought to their own discipline would inject new life into philosophy as well. As a
result, we could also treat philosophical concepts as objects of study, and analyse
their formation from below as Canguilhem has taught us.

References

Anonymous. (1927). Notre programme. Revue dHistoire de la Philosophie, 1, 14.

Bachelard, G. (1972). Lengagement rationaliste. Paris: Presses Universitaires de France.
Bachelard, G. (1986). Le rationalisme applique . Paris: Presses Universitaires de France. First published
1949.
Bachelard, G. (1993). La formation de lesprit scientifique: Contribution a` une psychanalyse de la con-
naissance objective. Paris: Vrin First published 1938.
Berr, H. (1911). La synthe`se en histoire; essai critique et the orique. Paris: Alcan.
Berr, H. (1921). Lhistoire traditionelle et la synthe`se historique. Paris: Alcan.
Berr, H. (1925). Pour la science. Revue de synthe`se, 18, 516.
Canguilhem, G. (1955). La formation du concept de re flexe aux XVIIe et XVIIIe sie`cles. Paris: Presses
Universitaires de France.
Canguilhem, G. (1963). LHistoire des Sciences dans luvre e piste mologique de Gaston Bachelard.
Annales de lUniversite de Paris, 33, 2439.
Canguilhem, G. (1976). La vie et mort de Jean Cavaille`s. Ambialet: Pierre Laleure.
Canguilhem, G. (1988a). Le statut e piste mologique de la me dicine. History and Philosophy of the Life
Sciences, 10(Suppl), 1529.
Canguilhem, G. (1988b). Ideology and rationality in the history of the life sciences (A. Goldhammer,
Trans.). Cambridge, MA: The MIT Press.
Canguilhem, G. (1989). The normal and the pathological (G.R. Fawcett, Trans.; R.S. Cohen, Ed.; with
an introduction by M. Foucault). New York: Zone Books.
Canguilhem, G. (1993). Ide ologie et rationalite dans lhistoire des sciences de la vie. Paris: Vrin. First
published 1977.
Canguilhem, G. (1994). Etudes dhistoire et de philosophie des sciences. Paris: Vrin. First published 1968.
Canguilhem, G. (1999). Le normal et le pathologique. Paris: Presses Universitaires de France. First pub-
lished 1966.
Canguilhem, G. (2002). Les maladies. In G. Canguilhem (Ed.), E crits sur la me dicine (pp. 3348). Paris:
Seuil. First published 1989.
Chartier, R. (1988). Cultural history. Between practices and representations. Ithaca, NY: Cornell Univer-
sity Press.
Chimisso, C. (2001a). Gaston Bachelard: Critic of science and the imagination. London & New York:
Routledge.
Chimisso, C. (2001b). He le`ne Metzger: The history of science between the study of mentalities and total
history. Studies in History and Philosophy of Science, 32, 203241.
Collins, R. (2002). The sociology of philosophies: A global theory of intellectual change. Cambridge,
MA & London: The Belknap Press of Harvard University Press.
Dagognet, F. (1997). Georges Canguilhem: Philosophe de la vie. Le Plessis-Robinson (Essone): Insti-
tut Synthe labo.
Daston, L. (1994). Historical epistemology. In J. Chandler, A. I. Davidson, & H. D. Harrotunian (Eds.),
Questions of evidence: Proof, practice and persuasion across disciplines (pp. 282289). Chicago:
University of Chicago Press.

26
Recently, some science studies scholars have engaged in this type of enterprise. See Kusch (1995,
2000) and Collins (2002).
326 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327

Daston, L. (1997). Recherche en le piste mologie historique des sciences: Empirisme et objective . Berlin:
Max-Planck-Institut fu r Wissenschaftsgeschichte, Preprint 64.
Febvre, L. (1942). Le proble`me de lincroyance au XVIe sie`cle: la religion de Rabelais. Paris: A. Michel.
Febvre, L. (1992). Combats pour lhistoire. Paris: Colin. First published 1938.
Fichant, M. (1973). Le piste mologie en France. In F. Chatelet (Ed.), Histoire de la philosophie, vol. 8:
Le 20e sie`cle (pp. 135178). Paris: Hachette.
Fichant, M. (1990). Georges Canguilhem et lide e de la philosophie. In E. Balibar (Ed.), Georges Cangu-
ilhem: Philosophe, historien des sciences. Actes du colloque (68 December 1990). Paris: Albin Mich-
el.
Fondation Pour la science, Centre International de Synthe`se, Section dhistoire des sciences. (1930).
Se ance du 19 mars 1930. Archeion, 12, 375381.
Fondation Pour la science, Centre International de Synthe`se, Section dhistoire des sciences. (1935).
Se ance du 23 Janvier 1935. Archeion, 17, 8184.
Foucault, M. (1972). The archaeology of knowledge (A.M. Sheridan Smith, Trans.). London: Tavistock.
Foucault, M. (1985). La vie: lexpe rience et la science. Revue de Me taphysique et de Morale, 90, 314.
Reprinted in English translation in Canguilhem (1989), Introduction.
Foucault, M. (1996). Larche`ologie du savoir. Paris: Gallimard. First published 1969.
Gueroult, M. (1988). Lhistoire de lhistoire de la philosophie, 3: En France de Condorcet a` nos jours.
Aubier: Paris.
Jardine, N. (2001). Koyre s intellectual revolution. La lettre de la maison franc aise dOxford, 13, 1125.
Guigue, A. (1935). La Faculte de lettres de lUniversite de Paris depuis sa fondation (17 Mars 1808)
jusquau 1er Janvier 1935. Paris: Alcan.
Koyre , A. (1939). Etudes Galile ennes. Paris: Hermann.
Koyre , A. (1958). From the closed world to the infinite universe. New York: Harpers.
Koyre , A. (1965). Newtonian studies (I.B. Cohen, Ed.). Cambridge, MA: Harvard University Press.
Koyre , A. (1966). E tudes dhistoire de la pense e scientifique. Paris: Presse Universitaires de France. First
published 1951.
Koyre , A. (1978). Galileo studies (J. Mepham, Ed.). Hassocks, Sussex: Harvest Press.
Kusch, M. (1995). Psychologism. A case study in the sociology of philosophical knowledge. London &
New York: Routledge.
Kusch, M. (2000). The sociology of philosophical knowledge: A case study and a defense. In M. Kusch
(Ed.), The sociology of philosophical knowledge. Dordrecht: Kluwer.
Le Goff, J. (1983). Histoire des sciences et histoire des mentalite s. Revue de Synthe`se, 104, 405415.
Le vy-Bruhl, L. (1922). La mentalite primitive. Paris: Alcan.
Manuali, C. (1988). Introduction: The cultural reasons underlying this meeting. History and Philosophy
of the Life Sciences, 10 (supplement), 713.
Metzger, H. (1935a). La philosophie de la matie`re chez Lavoisier. Paris: Hermann.
Metzger, H. (1935b). Tribunal de lhistoire et the orie de la connaissance scientifique. Archeion, 17, 1
14. Reprinted in Metzger, 1987.
Metzger, H. (1936). La priori dans la doctrine scientifique et lhistoire des sciences. Archeion, 18, 2942.
Metzger, H. (1969). Les doctrines chimiques en France du de but du XVIIe a` la fin du XVIIIe sie`cle. Paris:
Presses Universitaires de France.
Metzger, H. (1987). La me thode philosophique en histoire des sciences, textes 19141939 (G. Freudenthal,
Ed.). Paris: Fayard.
Nicolson, M. (1991). The social and the cognitive: Resources for the sociology of scientific knowledge.
Studies in History and Philosophy of Science, 22, 347369.
Paul, H. W. (1976). Scholarship and ideology: The chair of the general history of science at the Colle`ge
de France, 19821913. Isis, 67, 376397.
Redondi, P. (1994). He le`ne Metzger et Alexandre Koyre : La religion dHenri Berr. In A. Biard, D.
Bourel, & E. Brian (Eds.), Henri Berr et la culture du XXe sie`cle. Actes du colloque international
2426 octobre 1994, Paris (pp. 2426). Paris: Albin Michel.
Renn, J. (1995). Historical epistemology and interdisciplinarity. In K. Gavroglu, J. Stachel, & M. W.
Wartofsky (Eds.), Physics, philosophy and scientific community (pp. 241251). Dordrecht: Kluwer.
 C. Chimisso / Stud. Hist. Phil. Biol. & Biomed. Sci. 34 (2003) 297327 327

Rey, A. (1935). Indications concernant une direction de tude relative a` lhistoire des sciences, du point
de vue de leurs me thodes (dapre`s une confe rence de M. Abel Rey). Thale`s, 2, 3449.
Taton, R. (1956). Manuscript letter to George Sarton, 28/1/1956. Held at the Houghton Library, Cambridge
MA, bMS Am 1803 (1655).
Thale`s, recueil annuel des travux de lInstitut dHistoire des Sciences et Techniques de lUniversite de
Paris. Paris: Alcan.