Gaston Bachelard and

the Notion of
Phenomenotechnique
Hans-Jrg Rheinberger
Max Planck Institute for the
History of Science

The paper aims at an analysis of the oeuvre of the French historian of science
and epistemologist Gaston Bachelard (18841962). Bachelard was the
founder of a tradition of French thinking about science that extended from
Jean Cavaills over Georges Canguilhem to Michel Foucault. In the past, he
has become best known and criticized for his postulation of an epistemological
rupture between everyday experience and scientic experience. In my analysis,
I emphasize another aspect of the work of Bachelard. It is the way he concep-
tualizes the relation between scientic thinking and technology in modern sci-
ence. Within this framework, the notion of phenomenotechnique is of cru-
cial importance. It is one of the organizing concepts of Bachelards historical
epistemology, and it serves as the organizing center of this paper.

1. Introduction
As David Hyder has suggested in his recent essay Foucault, Cavaills,
and Husserl on the Historical Epistemology of the Sciences (Hyder
2003), we can distinguish two lines of French concern within the history
and philosophy of science in the middle of the twentieth century. Both are
critically engaged with a reception of Edmund Husserls phenomenology.
One of them, the subjectivist side, stands in the tradition of a philosophy
of consciousness and is represented by the phenomenology of Maurice
Merleau-Ponty among others; the other, the conceptualist side, is repre-
sented by a genealogy of epistemologists and historians of knowledge
ranging from Gaston Bachelard and Jean Cavaills to Georges Canguil-
hem to Michel Foucault. It is with the founder of the latter tradition that
this paper is concerned. The paper has the rather restricted goal of looking
I thank an anonymous reviewer for extremely valuable critical comments on an earlier draft
of this paper.

Perspectives on Science 2005, vol. 13, no. 3

2005 by The Massachusetts Institute of Technology

313
314 Gaston Bachelard and the Notion of Phenomenotechnique

in some detail at Gaston Bachelards (18841962) notion of phenomeno-

technique as one of the organizing concepts of his historical epistemol-
ogy.1 So far, this aspect of Bachelards work has received comparatively lit-
tle attention.2 It is particularly intriguing to note that the ideas of the
French philosopher and historian of science appear to have played no role
when proponents of the notion of technoscience such as Bruno Latour
developed their social constructivist and post-constructivist agendas of
looking at science in action.3 They instead polemicized against another
characteristic feature of Bachelards epistemological writings, which had
dominated their reception, particularly in France, in the period between
the 1960s and 1970s. The issue at stake was Bachelards postulate of an
epistemological rupture between pre-scientic thinking entrenched in the
immediacy of unquestioned everyday life, and scientic thought, which
needed the connes of a scientic culture to progress.4 This dichotomy
was understood to go against the grain of a symmetric assessment of
scientic activity which refused the distinction between true and false its
organizing power in structuring narratives on science.
It is obvious that there is an intriguing element of faith and belief in
scientic progress in Bachelard, his writings at times even bubbling
over with his enthusiasm for science, as Mary Tiles once remarked (Tiles
1984, p. 4). On the occasion of the International Congress of Philosophy
of Science in Paris in 1949, he even went so far as to claim that it appears
that the very existence of science is dened as a progress of knowledge
(Bachelard [1951] 1972b, p. 36). However, progress in Bachelards
sense, as we shall see, means permanent goings-on, a continuous move-
ment of differential reproduction rather than a movement toward a pre-
conceived end, or an approximation of an ultimate reality (Bachelard
[1928] 1987).
In a recent book, Cristina Chimisso has argued that many features of
Bachelards work can be understood better if one recognizes that he was
deeply embedded in the French education system of the early twentieth
century with its idea of scientic spirit, and if one recalls his own long
practice as a school teacher of physics and chemistry at the Collge of his
home town Bar-sur-Aube (Chimisso 2001, especially Chapter 3). We must
also not forget, as Jean Gayon reminds us, that Bachelard, before choosing
to teach, had considered becoming an engineer (Gayon 1995, pp. 311).
This may help us to understand that his epistemology has two irreducible
1. The term is Dominique Lecourts. See Lecourt 2002, introduction.
2. Notable exceptions are Gaukroger 1976; Castelao-Lawless 1995; and particularly
Gayon 1995.
3. See particularly Latour 1993.
4. Examples of this trend are the writings of Louis Althusser and Etienne Balibar.
 Perspectives on Science 315

dimensions. From the perspective of the scientist as a person, it takes the

form of a psychology or even a psychoanalysis of the scientic spirit; from
the perspective of the scientic process, it takes the shape of a praxeology of
scientic work. In this paper, I would like to address those aspects of his
conception of the contemporary sciences that appear to me to contain key
elements for understanding their peculiar dynamics. Much of the argu-
ment revolves around Bachelards conception of what it means to be a
scientic object in the sense of a technophenomenon.5

2. Premises
Although the cluster of epistemological books that Bachelard wrote be-
tween 1949 and 1952 may be said to represent a radicalization of his
thoughts about the relation between science and technology (Dagognet
2003), there is also a remarkable thematic continuity to be observed with
the rst series of epistemological works in the late 1920s and the 1930s
(Gayon 2003). At the center of this continuity stands the notion of
phenomenotechnique. Gayon has claimed that Bachelard tried to cir-
cumscribe the precise nature of the technical aspect of science. The rst
part of this enterprise consisted of the elaboration of the concept of
phenomenotechnique. This concept is without a doubt the essential piece
of the philosophy of applied rationalism (Gayon 1995, p. 39). In a similar
vein, Teresa Castelao concludes: Phenomenotechnique is one of the most
potentially rich concepts that Bachelard has to offer to contemporary phi-
losophy of science and to science studies in general (Castelao-Lawless
1995, p. 45). The concept aims at conceiving of technology not as an
eventual byproduct of scientic activity, as a derivative product through
which science manifests itself in society, but as constitutive of the contem-
porary scientic modus operandi itself. Insofar as the technological mode
of action is engaged in the core of the scientic enterprise, the technologi-
cal object itself acquires an epistemic function. In a paper on microphysics
from 1931, Bachelard expresses himself as follows: We could therefore
say that mathematical physics corresponds to a noumenology quite differ-
ent from the phenomenography to which scientic empiricism connes it-
self. This noumenology implies a phenomenotechnique by which new
phenomena are not simply found, but invented, that is, thoroughly con-
structed (Bachelard [193132] 1970, pp. 1819).
For the purposes of the present essay, we can take Bachelards notion of
noumenon to be roughly equivalent to the ordinary notion of concept or
scientic law, provided we respect Gayons caveat that it has to be taken
5. For a valuable collection of texts addressing the major epistemological concerns of
Bachelard see Bachelard (1971) 2001.
316 Gaston Bachelard and the Notion of Phenomenotechnique

neither as an arbitrary hypothesis nor an immutable essence, but rather as

an entity in ux, a sort of progress of thought (Gayon 1995, p. 43). In
invoking the example of the separation of isotopes in mass spectrometers,
Bachelard comes back to this point almost twenty years later, concluding
that the scientic activity of our age consists in the noumenal preparation
of technically constituted phenomena. The trajectories along which the
isotopes in the mass spectrometer are separated do not exist in nature; they
have to be technically produced (Bachelard [1949] 1998, p. 103). What
tends to be perceived as a fact, as something given in the real world, has to
be seen as a result of and even derives its existence from a circuit that is at
the same time material and discursive. Technophenomena are theoretically
invested entities. The philosopher of science must try to analyze this in-
vestment in order to understand the enormous productivity of contempo-
rary science and technology. This material-discursive circuit is dialecti-
cally constituted. It has no clearly denable starting point on either the
noumenal or the phenomenal side. Its thorough philosophical understand-
ing is as far away from empiricism as it is from rationalism. One can only
assess it by immersing oneself in the particulars of the phenomenotechni-
cal work of the sciences. Bachelards technical phenomenology can be
seen as an answer to Husserls phenomenology, which remained rmly en-
trenched in the life-world. But whereas Husserl wanted to save Western
scientic rationalism by reconnecting the sciences, at least potentially, to
everyday experience, Bachelard contended that the sciences themselves
were to be assessed as particular, concrete forms of life.
The decisive philosophical challenge was for Bachelard less a radical
constructivist stance, as might perhaps be guessed from the above quota-
tions. His conclusion pointed in another, rather surprising direction. He
contended that with the ever tighter interplay between ever more specic
forms of knowledge and the phenomenal world, the sciences necessarily
became fragmented into different epistemological regions. He postulated
that their conceptual dynamics nally became inseparable from the phe-
nomena in which and through which they expressed themselves. There-
fore, so it appeared to Bachelard, in order for the sciences to be understood
properly, they had to be studied in exactly these regional manifestations
and in all the details of their diverse technical realizations. In order to do
this, he tried to situate himself beyond the received brands of philosophy
of science: beyond positivism and formalism, empiricism and convention-
alism, realism and idealism, positions that he judged to be the results of
impermissible abstractions that did not do justice to the complexity of the
contemporary sciences. He did not want to side with any of these philo-
sophical traditions. In order to express the dialectical tension in his own
 Perspectives on Science 317

Figure 1. Taken from Gaston Bachelard, Le rationalisme appliqu. Presses

Universitaires de France, Paris 1949, 3rd edition, Quadrige/PUF, Paris 1998,
p. 5.

approach, he called it applied rationalism or technical materialism (see

Figure 1, Bachelard [1949] 1998, p. 5).
Consequently, in his Philosophy of No in 1940 Bachelard called for a
dispersed or distributed philosophy, a philosophical position that thus
did not shy away from engaging with the diverse sciences in the process of
their becoming. He pleaded for a philosophy that would be able to ac-
count for the vastly different types of theory involved in the multiple and
various activities of the sciences, for the wide range of their reications,
and particularly for the multifarious procedures of discovery (Bachelard
1940, p. 10). He even went so far as to postulate that each interesting
problem, each experiment, or even each equation required a philosophical
reection of its own. According to Gayon, Bachelard the epistemologist
has over and over again renewed his adhesion to a rationality of detail that
pluralizes the object at the very same time it recties and realizes it
(Gayon 1995, p. 4). Bachelard was convinced that it was not the task of
epistemology to spell out general, timeless norms for scientic knowledge
acquisition. According to him, it was not the epistemologists, but the sci-
entists alone who had the right to dene the continuously shifting bound-
aries of their trade. Consequently, he summarized his program by rhetori-
cally asking the scientists to make available their daily laboratory
experience, the daily dramas of their daily work, to the philosophers of sci-
ence, so that they could reect properly about the practice of contempo-
318 Gaston Bachelard and the Notion of Phenomenotechnique

rary science: Tell us what you think, not when you quit the laboratory,
but during the hours when you leave ordinary life behind you and enter
scientic life. Instead of leaving us with your empiricism of the evening,
show us your vigorous rationalism of the morning (Bachelard 1940,
p. 11).
If the dynamics of the process of knowledge acquisition depended on
regionalization, a plurality of methods had to be acknowledged as well as
an inherent openness toward the process as such. We could speak of a pro-
cess epistemology in this respect, an epistemology of emergence and of
innovation, a philosophy at work (Bachelard [1949] 1998, p. 9). For
Bachelard, it was essential for philosophers of science to keep in touch
with the development of the sciences. Epistemology had to be understood
as a permanent reection of that development. Following this line, Bach-
elard claims: We must attempt a rationalism that is concrete and in line
with the precision of particular experiments. It is also necessary that our
rationalism is sufciently open to receive new determinations from the ex-
periment (Bachelard [1949] 1998, p. 4).
In short, an epistemology that tries to assess scientic thinking in its
dynamicity must be as plastic, as mobile, as uid, and as risky as scientic
thinking itself. Two consequences follow. The rst is a regional mobility
of epistemology according to the regionalization of knowledge. Modern
science creates what Bachelard claims to be kernels of apodicticity that
can only be assessed from inside, if one is ready to play the game according
to the rules of each of the kernels. In this respect, Bachelard often also
talks of cantons, regions, or domains of knowledge within the city of sci-
ence, such as the relativistic canton in the city of mechanics (Bach-
elard [1949] 1998, pp. 132133). These cantons are islands of scientic
culture; they create their own cultural codes and forms of emergence,
which only an intimate knowledge of the respective region allows one to
judge. This means that epistemology must engage itself with the intrica-
cies of these particular islands.
The second consequence is mobility along the historical axis. Scientic
thinking is essentially a rectication of knowledge, claims Bachelard. It
judges its historical past by discarding it. Its structure is the consciousness
of its historical errors. Scientically one thinks of truth as the historical
rectication of one long error, one thinks of experience as rectication of a
primary and common illusion (Bachelard [1934] 1968, p. 173). One is
tempted here to think of Karl Poppers critical rationalism. But here we
also come to the core of what Bachelard tried to capture with his notion of
recurrence. What is at stake is probably less the epistemological rupture be-
tween a scientic mode of thinking and a primary illusion. Much more
important appears to be the fact that within scientic activity itself, a per-
 Perspectives on Science 319

manent process of rectication and reorientation is going on, a process

that permanently transforms the truth of today into the error of yesterday.
And since the truth of today is constituted through this very movement, it
receives from it that strange double character of being an arbiter of the
past and at the same time being subjected to the very same ongoing
rectication process. Any scientic truth of today may thus itself one day
become an error of the past. This is the core of Bachelards idea of the his-
toricity of science, a historicity that marks it as a peculiar culture of truth.
The spirit has a variable structure from the very moment where knowl-
edge has a history (Bachelard [1934] 1968, p. 173). Knowledge is an evo-
lution of the spirit, it is nothing that is accomplished once and forever, it
is grounded in its very own discardability, and not in the timeless unity of
a thinking ego. Reason is always cunning reason, for methods are strate-
gies of acquisition with the precarious status of risking themselves in a
new acquisition on the basis of being enforced through a former acquisi-
tion (Bachelard [1951] 1972b, p. 39, emphasis added). They modify and
at times even consume themselves in their own application.

3. Major Aspects of Bachelards Epistemology

This is the place to have a closer look at a few major aspects that character-
ize Bachelards epistemology. One of them involves scientic object, scien-
tic spirit, and objectivity; the other has to do with the social character of
science.

Scientic Object, Scientic Spirit

Let us rst see what consequences Bachelards version of process episte-
mology has for the conception of scientic objects and of scientic objec-
tivity. With that, I come to the core of my assessment of Bachelards
phenomenotechnology. Two aspects of the problem can be distinguished.
The rst, already mentioned at the beginning, is that in general, the ob-
jects on which the sciences workthe phenomena they are concerned
withare not immediately given to the senses. For Bachelard, it is built
into the very notion of the scientic spirit that its objects are entities that
cannot be grasped without mediation, because they are always a product of
scientic work. Scientic objectivity is only possible if one has broken
with the immediate object, if one has refused the seduction of the rst
choice, if one has arrested and contradicted the thoughts that spring from
a rst observation (Bachelard 1938, p. 9). On another occasion Bachelard
notes that one must accept a veritable rupture between sensual knowl-
edge and scientic knowledge (Bachelard [1938] 1969, p. 239). This
epistemological rupture, in a kind of originative epistemic gesture, or
epistemological act (Bachelard 1951, p. 25), constitutes and transforms
320 Gaston Bachelard and the Notion of Phenomenotechnique

a perceived phenomenon into a scientic object. What is achieved in this

act is the transformation of something that had been taken as given into a
problem. A scientic object is a phenomenon that has been drawn into a cy-
cle of rectication; it is not constituted once for ever, but it remains a
scientic object only through its being constantly reconstituted and
rectied. The epistemological rupture serves to mark the transition from
everyday knowledge to the act of scientic thinking, while at the same
time it inscribes itself into this very act and thus becomes an intrinsic
hallmark of a continued scientic engagement with the world.
The second aspect of Bachelards conception of the objectivity of science
is that not only are its objects mediated, but that the knowing spirit is not
immediately given either. Bachelard claims that the march toward the
object itself is not objective from the beginning (Bachelard [1938] 1969,
p. 239). The knowing spirit has to externalize itself and become instru-
mental (Bachelard [1938] 1969, p. 218), that is, its formation is itself
technically mediated. It must engage in a cycle as wella cycle of instruc-
tion.
As a result, both scientic spirit and scientic object enter in a mutu-
ally exteriorizing and at the same time mutually interiorizing relation
with each other. This double movement is instantiated and epitomized by
scientic instruments. They are embodiments of acquired knowledge and
at the same time help to produce the object as a technophenomenon. In
contemporary science, Bachelard claims, the instrument is the necessary
intermediary in the study of a denitely instrumented phenomenon that
has been designed as an object of a phenomenotechnique (Bachelard
[1949] 1998, pp. 23). As a materialized noumenon, the instrument sits
in the center of the epistemic ensemble: In modern science, the instru-
ment is veritably a reied theorem (Bachelard 1933, p. 140). It is not a
passive device positioned between a Cartesian mind and the outward
world, destined to enhance the subjects discriminatory capacities. At any
given time, the instrument represents the material existence of a body of
knowledge. The phenomenon is provoked as a problem at the knowledge
horizon and may itself require new concepts in order to be accommodated.
Phenomenon and instrument, object and scientic spirit, concept and
method are all joined in a process of mutual instruction.
Occasionally, Bachelard speaks of construction in this context.
Nothing is given. Everything is constructed (Bachelard [1938] 1969,
p. 14), we read for instance in The Formation of the Scientic Spirit. But in
general, he prefers the term realization: Science realizes its objects with-
out ever nding them readymade. Phenomenotechnology extends phenom-
enology. A concept has become scientic according to the proportion to
which it has become technical, to which it is accompanied by a technique
 Perspectives on Science 321

of realization (Bachelard [1938] 1969, p. 61). Whereas the term con-

struction stresses the part of the material object, the term realization
stresses its conceptual counterpart. Most appropriate, however, is the term
instruction, for it insists on the mutuality of the process in which the
object itself becomes an agent in an intimate knowledge relation, in an
epistemic engagement in the words of Bachelard: The position of the
scientic object, actually the object as an instructor, is much more com-
plex, much more engaged. It claims a solidarity between method and expe-
rience. One must therefore know the method of knowing in order to grasp
the object to be known, that is, in the realm of methodologically valued
knowledge, that object which is capable of transforming the method of
knowing (Bachelard [1949] 1998, p. 56). Thus, we are confronted with a
deeply non-nativist, non-naturalistic procedure on both sides. The
scientic spirit must form itself against nature and against its own inborn
inertia. It can become instructed not by self-education, but only by engag-
ing with the objects of the world, by purify[ing] the natural substances
and thus bring[ing] order into nebulous phenomena (Bachelard [1938]
1969, p. 23). Just as the scientic spirit is not ready-made, the objects of
science are neither immediately given nor immediately ready to be
grasped. They take shape in a long and tedious historical process of
purication and ordering.
In its subjective dimension, Bachelard has described this engagement
in what he calls a psychoanalysis of scientic knowledge. It is one of the
characteristics of his historical epistemology that it does not rely on postu-
lating a certain structure of scientic thought or a particular logic of
scientic thinking. In contrast, Bachelard sees it as a deliberately psycho-
epistemic activity. The act of gaining knowledge is an activity in which the
person of the scientist is involved as a whole. The process of gaining
knowledge is work. In the center of its description, amounting to a kind of
phenomenology of work (Bachelard [193132] 1970, p. 14) stands what
Bachelard calls the epistemological obstacle. This is not, as he explains
right at the beginning of his Formation of the Scientic Spirit, some difculty
that may impose itself from outside, such as the sheer complexity of the
world for instance, or the physical limitations of our senses. It is rather in
the act of knowing itself that slownesses and troubles appear, intimately,
in a kind of functional necessity (Bachelard [1938] 1969, p. 13). These
slownesses and troubles have to be surmounted if the objectivizing process
of knowledge acquisition is to work. In it, there can be neither immediacy
nor belief: The real is never what one might believe, it is always only
that which one should have been thinking. Empirical knowledge is lucid
only after the event, after the apparatus of reasoning has been set in motion
(Bachelard [1938] 1969, p. 13). Knowledge can be sanctioned in an act of
322 Gaston Bachelard and the Notion of Phenomenotechnique

recurrence, but there is no straightforward procedure that would grant its

acquisition. The emergence of knowledge remains in the realm of messy
confusion, of trying out, in the inertia of preconceived opinions; it is again
and again in need of those epistemological acts . . . that bring unexpected
impulses into the scientic development (Bachelard 1951, p. 25). In con-
trast to many of his contemporaries, such as Hans Reichenbach, Bachelard
does not exclude this messy space of discovery from the domain of episte-
mology. He rather declares it as its center.

Science as a Social Process

It is another distinguishing feature of Bachelards conception that he also
attributes a social or communitarian dimension to this process. Modern
knowledge acquisition can no longer be conceived of within the connes
of the traditional relation between a solitary subject and its object as stip-
ulated by theories of knowledge in the Cartesian tradition. It is set in mo-
tion and carried out as a collective enterprise effectively performed by a
scientic community and bound together through the effective
interpsychological work of language and experiment (Bachelard [1949]
1998, p. 55). As mentioned already, it is resumed in regional rationalisms,
cantons or quarters within a scientic city which Bachelard expressly
characterizes as scientic cultures. A culture, for Bachelard, is dened as
an accession to an emergence (Bachelard [1949] 1998, p. 133), that is,
as a milieu that allows for the appearance of novelty, of unprecedented
events. If the artistic act of creating novelty might be more of an individ-
ual nature, in the sciences in contrast, Bachelard contends, these emer-
gences are effectively constituted socially (ibid., p. 133). Even the
epistemological acts of scientic geniuses are, as a rule, rmly embedded
in these cultures.
Moreover, we have to distinguish several different aspects of modern
scientic cultures. The rst one is grounded in the very psychological con-
stitution of an individual versus a group of tightly bound individuals en-
gaged in a common epistemological project. Man hesitates, says
Bachelard. But the schoolin the sciencesdoes not hesitate. The
schoolin the sciencescarries along (Bachelard 1951, p. 6). The
school, one could say, is the social expression of the process of recurrence
that characterizes scientic knowledge acquisition. It provides the neces-
sary bonds for the dialectics of attachment and engagement to play out
their role (ibid., p. 12).
Another aspect of modern scientic culture is specialization. The ratio-
nality of the modern sciences is cantonized. In contrast to many of his con-
temporaries who deplored the overspecialization of the sciences, Bachelard
sees specialization in its positive aspects as an actualization of a general-
 Perspectives on Science 323

ity (ibid., p. 11). It dynamizes the spirit. It works. It works endlessly,

and besides that, it is the most specialized cultures that are most liable to
substitutions (ibid., pp. 14, 12). On the one hand, they set free the
forces of xation that are the counterpart of the disposability of the
modern scientic spirit (ibid., p. 13). On the other hand, they instigate
and maintain the game of substitution of methods, categories, and phe-
nomena. The more narrowly dened an area, the more readily conventions,
measurements, descriptions, and classications can become altered and
subjected to modication, and then eventually also may be carried over
into other areas of research. Specialization creates epistemic exibility. We
see here a glimpse of the idea of patchwork productivity of knowledge ac-
quisition. Finally, there is the cooperation between the theoretical and the
technical society in the scientic city. These two societies are engaged in
mutual comprehension. It is this mutual, intimate, and agitating com-
prehension that constitutes, to Bachelard, the really new philosophical
fact of modern science. The convergence of precision and power in local-
ized technophenomenologies is a veritable social feat. It is not of the order
of a natural necessity. It constitutes the very fabric of phenomena of the
contemporary scientic world (Bachelard 1951, pp. 910).
These few remarks may sufce to demonstrate that a social as well as a
cultural dimension is irreducibly built into Bachelards historical episte-
mology. At this point, we can also see how the dialectics of the technical
and the noumenal within the epistemic core of the research process trans-
lates and redoubles itself in an intimate relation between science and tech-
nology on the societal level. This aspect particularly characterizes the later
epistemological work of Bachelard, beginning with Applied Rationalism of
1949. But it is also clear that there remains an inescapable tension that
Bachelard already recognized in his rst book, Approximate Knowledge, in
1928. Gayon formulates this Bachelardian tension as follows: Science
constantly faces the fundamental irrationality of the given. Industry, in
contrast, materially realizes a clearly recognized and pursued rationale. On
one side, one seeks the rational, on the other, one imposes it. Technology
plainly realizes its object, and for that very reason, radically escapes the spon-
taneous skepticism that is one of the most distinctive cognitive attitudes
of scientic thinking (Gayon 1995, p. 38).6

4. Technoscientic Productivity
With this double aspect of the relation between the technical and the
scientic, I come back to the central theme of this special issue,
technoscientic productivity. As already mentioned, in order to stress the
6. The quotes are from Bachelard 1938a, p. 160.
324 Gaston Bachelard and the Notion of Phenomenotechnique

inextricability of this relation, in his later epistemological work Bachelard

uses the compound notions of applied rationalism and technical mate-
rialism for his own position. Materialism here refers to a reality trans-
formed by rationalism and thus carrying the mark of rationalism
(Bachelard [1949] 1998, p. 8). It is crucial in this context to explicate
Bachelards concept of rational application in more detail. It operates on a
level distinctly different from the well-known and worn distinction be-
tween basic science and applied science. In its traditional form, the di-
chotomy can be formulated as follows. Basic research is carried out in a
space in which one is free to play around, to experiment and to hypothe-
size. Sooner or later, fundamental solutions arrived at in basic science will
nd their technological applications in society at large. The notion im-
plies that basic science operates in a space that is essentially value-free and
becomes socially and ethically laden only in the process of its translation,
dissemination, and scaling up: in short, in its industrial application for
certain purposes, for good or for bad.
Bachelard however sees the epistemic and the technical in no such lin-
ear relation, but in a much more intimate one, in fact in a mutual co-
evolution that carries the whole burden of modern sciences productivity.
Application is not extrinsic to modern knowledge, it is not just added to
some epistemic core; it exerts its action at the very level of concept forma-
tion itself; the technical belongs to the essence of the modern sciences
themselves. In The Formation of the Scientic Spirit, Bachelard formulates
quite expressly: In order to accommodate new experimental proofs, one
must . . . deform the primitive concepts. One must not only study the con-
ditions of application of these concepts, but one must incorporate the con-
ditions of application of a concept into the very meaning of the concept itself
(Bachelard [1938] 1969, p. 61). Consequently, applicability is built right
into the core of modern sciences concept formation. Stephen Gaukroger
has called this the essence of [Bachelards] phenomenotechnics (Gau-
kroger 1976, p. 221). Applied rationalism is thus technically imple-
mented materialism. It is not the idea of a science in search of application,
but of a science that is taken and accepted as science because it moves in
and has always existed in the realm of the applicable, because its very
epistemological constitution has a technical dimension, because applica-
tion is built into the very meaning of concepts and into the rules of con-
cept formation, because the technical is built into the experimental phe-
nomena, and because, just the other way around and in a symmetrical
fashion, the noumena are built into the instruments and take on an instru-
mental form that further serves to develop the whole phenomenotechnical
machinery. Here, too, Bachelard distances himself from describing the sit-
uation with terms like construction or artice: The factitious, he
 Perspectives on Science 325

contends in a discussion of Hans Vaihingers ctionalism, may well pro-

vide a metaphor for the situation, but it cannot deliver a syntax, like
the technical, capable of linking arguments and intuitions (Bachelard
1933, pp. 141142). The same holds for utilitarian and pragmatist inter-
pretations of science whose scattered pluralism Bachelard deliberately
rejects, too, in the name of a coherent pluralism (Bachelard 1932).7
Against any supercial methodological mobility he insists on the indis-
pensable virtues of the claim to truth as the constant power of integration
of modern scientic knowledge (Bachelard [1951] 1972b, p. 40), thus
balancing its productive regionalization, the latter being equally indis-
pensable, for without the multiplication of perspectives, there is no ob-
jectivity (Gayon 1995, p. 4). It is its networked material entrenchment,
mediated by instruments and dispersed throughout society in the form of
standardized industrial products, which conveys the necessary minimum
of coherence to the sciences in all their diversications and despite the fact
that the language of science is in a permanent state of semantic revolu-
tion (Bachelard 1953, p. 209). This minimum of coherence is precisely
due to the deep structure of application of modern knowledge. In the
new rationalism, the chief notion is that of technique. Modern science is
an intrinsically technical science (Gayon 1995, p. 36).

5. Historical Epistemology
These aspects of Bachelards thought, pertinent as they are for an under-
standing of technoscientic productivity, have barely found their place in
recent science and technology studies. The aim of this piece of exegesis has
been to point to this gap rather than to close it. To conclude, I would like
to come back to Bachelards project of a historical epistemology. In fact,
the project has two aspects. The rst is connected to the problem of
technoscientic productivity and the particular slant that Bachelard gives
it with his notion of phenomenotechnique. Scientic objectsor techno-
phenomena for that mattercarry an intrinsic history along with them,
because experimental work has always already been invested in them, once
the immediacy of the common sense grip on whatever was thought to be
there has been broken. The history of science possesses a very peculiar con-
nectivity that follows from the lingering path of its auto-rectication. It is
essentially open-ended, surpassing and at the same time bent back on it-
self. In that respect, the history of the sciences appears as the most irre-
versible of all histories (Bachelard 1951, p. 27). Scientic objects are al-
ways transformations of earlier scientic objects and thus intrinsically
historical entities. So, for instance, the electrical reality of the nine-
7. He uses the latter term as early as 1932 prominently in a book title.
326 Gaston Bachelard and the Notion of Phenomenotechnique

teenth century is quite different from the electrical reality of the eigh-
teenth century (Bachelard [1949] 1998, p. 9), and yet they are perceived
as transformations of the same range of phenomena. Through the tight
coupling of the noumenal and the technical within scientic activity, and
of science and technology on the societal level, applied rationalism be-
comes part and parcel of the material culture of humanity. It is important,
however, to recognize that despite this irreversible historicity, there is, ac-
cording to Bachelard, no historical necessity for the sciences to arise, there
is no historical reason in the strong Hegelian sense of Vernunft at work
here (Bachelard 1951, p. 23). The synthetic achievements of the sciences
are thoroughly emergent phenomena; their emergence is not programmed
in a teleological manner. Bachelard quotes Louis de Broglie as his witness
in this context: Many scientic ideas of today would be different from
what they are if the paths followed by the human spirit to approach them
would have been different (de Broglie 1947, p. 9; quoted in Bachelard
1951, p. 21).
The second aspect of Bachelards project of a historical epistemology is
that despite the impossibility, in principle, of anticipating scientic prog-
ress, despite the non-existence of a historical reason, the actual state of
knowledge cannot but serve as a grid for the evaluation of science past.
Historical judgment is a recurrent action that carries a kind of teleology of
hindsight along with it; history is illuminated by, and always appears in
the light of, a nality of the present (Bachelard 1951, p. 26). In the
light of this nality, history of science divides itself into a sanctioned part
and into one that has been superseded. Since science itself makes this
distinction, since it is in essence a process of self-detachment, of self-
distinction, of self-polemics, and of self-negation, the historical epistemol-
ogist who wants to be up to date and follow the movement of his objects,
will nd his own activity falling under the same rules. However, handling
these recurrences requires of him a veritable tact, for they carry a ruin-
ous element along with them: The philosophical position I assume
here, Bachelard asserts, is certainly not only difcult and dangerous. It
even contains an element of self-ruin: this ruinous element is the ephem-
eral character of the modernity of science. If one follows the ideal of the
modernist tension I propose for the history of the sciences, it becomes nec-
essary that the history of the sciences be frequently redone, be often recon-
sidered (Bachelard [1951] 1972a, pp. 143144). If the sciences them-
selves constantly alter their judgment of the past, epistemology is bound
to follow them in a process of co-transformation. Historical epistemology
itself then becomes a historically changing enterprise. If the culture of the
scientist is a history of permanent reformation (Bachelard 1951, p. 13),
the culture of the epistemologist cannot be different in this respect. While
 Perspectives on Science 327

the historian, in principle, may not be out for judgments, the historical
epistemologist must judge. But he also must be ready to change his judg-
ment in the face of new scientic developments.

References
Bachelard, Gaston. (1928) 1987. Essai sur la connaissance approche. Paris:
Vrin.
. (193132) 1970. Noumne et microphysique. Pp. 1124 in
Etudes. Paris: Vrin.
. 1932. Le pluralisme cohrent de la chimie moderne. Paris: Vrin.
. 1933. Les intuitions atomistiques. Paris: Boivin.
. (1934) 1968. Le nouvel esprit Scientique. Paris: Presses
Universitaires de France, 10th edition.
. 1938. La psychanalyse du feu. Paris: Gallimard.
. (1938) 1969. La formation de lesprit Scientique. Paris: Vrin, 6th
edition.
. 1940. La philosophie du non. Paris: Presses Universitaires de
France.
. (1949) 1998. Le rationalisme appliqu. Paris: Presses Universitaires
de France, 3rd edition, Quadrige/PUF.
. 1951. Lactivit rationaliste de la physique contemporaine. Paris:
Presses Universitaires de France.
. (1951) 1972a. Lactualit de lhistoire des sciences. Pp. 13752
in Lengagement rationaliste. Paris: Presses Universitaires de France.
. (1951) 1972b. Le problme philosophique des mthodes
scientiques. Pp. 3544 in Lengagement rationaliste. Paris: Presses
Universitaires de France.
. 1953. Le matrialisme rationnel. Paris: Presses Universitaires de
France.
. (1971) 2001. Epistmologie. Textes choisis par Dominique
Lecourt. Paris: Presses Universitaires de France, 7th edition.
de Broglie, Louis. 1947. Physique et microphysique. Paris: Albin Michel.
Castelao-Lawless, Teresa. 1995. Phenomenotechnique in historical per-
spective: Its origins and implications for philosophy of science. Philos-
ophy of Science, 62:4459.
Chimisso, Cristina. 2001. Gaston Bachelard: Critic of Science and the Imagi-
nation. London: Routledge.
Dagognet, Franois. 2003. Sur une seconde rupture. Pp. 1317 in
Bachelard et lpistmologie franaise. Edited by Jean-Jacques
Wunenburger. Paris: Presses Universitaires de France.
Gaukroger, Stephen W. 1976. Bachelard and the problem of epistemo-
logical analysis. Studies in History and Philosophy of Science, 7:189244.
328 Gaston Bachelard and the Notion of Phenomenotechnique

Gayon, Jean. 1995. Bachelard: Le rationalisme appliqu. Centre de Vanves,

Ministre de lEducation Nationale, Centre National dEnseigenement
Distance.
. 2003. Bachelard et lhistoire des sciences. Pp. 51113 in
Bachelard et lpistmologie franaise. Edited by Jean-Jacques
Wunenburger. Paris: Presses Universitaires de France.
Hyder, David. 2003. Foucault, Cavaills, and Husserl on the Historical
Epistemology of the Sciences. Perspectives on Sciences, 10(4):107129.
Latour Bruno. 1993. We Have Never Been Modern. Cambridge: Harvard
University Press.
Lecourt, Dominique. 2002. Lpistmologie historique de Gaston Bachelard
[1969]. Paris: Vrin, 11e tirage augment dune postface.
Tiles, Mary. 1984. Bachelard: Science and Objectivity. Cambridge: Cam-
bridge University Press.