You are on page 1of 52

China, 1958.

Hen r i Car tier Bressan/Magn u m P hatas

Machine and Organism
Georges Canguilhem
The relationship between machine and organism has generally been studied
in only one way. Nearly always, the organism has been explained on the basis
of a preconceived idea of the structure and functioning of the machine; but
only rarely have the structure and function of the organism been used to make
the construction of the machine itself more understandable. Even though
mechanistic theory sparked some very impressive technical research, the fact
remained that the very notion of an "organology,;' as well as its basic premises
and methodology, remained undeveloped.
Philosophers and mechanistic biologists approached the machine as a set
of data, or else made it into a problem that they could solve purely through
mental application. To do this, they called on the engineer, who was for them
a scientist in the truest sense. Misled by the ambiguities of their view of
mechanics, they saw machines only as theorems in concrete form. The opera-
tions necessary to construct machines were only secondary considerations
when compared with the all-important idea that the machine revealed their
theories in concreto. To see this, one needed only to acknowledge what science
could accomplish, and from there it was simply a matter of the confident
application of that knowledge. However, I do not believe that it is possible to
treat the biological problem of the "living machine" by separating it from the ..
technological problem it supposedly resolves - namely, the problem of the re-
lationship between technology and science. This problem is normally resolved
by starting with the idea that, logically and chronologically, knowledge pre-
cedes application. What I want to show is that the construction of machines
can indeed be understood by virtue of certain truly biological principles, with-
out having at the same time to examine how technology relates to science.
I shall address the following topics in successive order: what it means to
compare an organism to a machine; the relationship between mechanical
processes, and the results that might be achieved by using them; and the his-
torical reversal of the traditional relationship between the machine and the
organism and the philosophical consequences of this reversal.
Machine and Organism
For those who have carefully studied living beings and the forms they take, it
is rare - and only in the case of the vertebrates - that one notices any truly
mechanical attributes, at least in the sense that the term is commonly under-
stood by scientists. In La Pensee technique, for example, Julien Pacotte notes
that movements of the joints and the eyeball can be paralleled with what math-
ematicians call a "mechanism."2 A machine can be defined as a man-made,
artificial construction, which essentially functions by virtue of mechanical
operations. A mechanism is made of a group of mobile solid parts that work
together in such a way that their movement does not threaten the integrity
of the unit as a whole. A mechanism therefore consists of movable parts that
work together and periodically return to a set relation with respect to each
other. It consists of interlinking parts, each of which has a determinable
degree of freedom of movement: for example, both a pendulum and a cam
valve have one degree of freedom of movement, whereas a threaded screw
has two. The fact that these varying degrees of freedom of movement can be
quantified means that they can serve as tangible guides for. measuring, for
setting limits on the amount of movement that can be expected between any
two interacting solid objects. In every machine, then, movement is a function,
first, of the way the parts interact and, second, of the mechanical operations
of the overall unit. 3
Mechanics is governed by the principle that every movement of a
is geometric and measurable. What is more, every such movement regulates
and transforms the forces and energy imparted to it. Mechanics, though, does
not work in the same way that a motor does: in mechanics, movements are
simply propagated, not created. A rather simple example of how this trans-
formation of movement takes place can be seen in several devices - a wheel
crank or an eccentric crank, for example - that are set into motion by an ini-
tiallateral movement but eventually produce reciprocating, rotary movement.
Of course, mechanical operations can be combined, either by superimposing
them or adding them together. It is even possible to take a basic mechanical
device, modify it and make it capable of performing a variety of other mechan-
ical operations. This is exactly what happens when a bicycle freewheel clutch
is released or stopped.
What constitutes the rule in human industry is the exception in the struc-
ture of organisms and the exception in nature, and I must add here that in
the history of technology and the inventions of man assembled configurations
are not the most primitive. The oldest known tools are made of a single piece.
The construction of axes or of arrows made by assembling a flint and a han-
dle, or the construction of nets or fabrics, are so many signs that the primi-
tive stage has been passed.
This brief overview of some elementary principles of kinematics helps to
~ .
give a fuller sense of the problem without losing sight of a central paradox:
Why was it necessary to turn to the theory of mechanism, as outlined above,
in order to explain the living organism? The answer can be found, it seems,
in the fact that this mechanical model of living organisms does not rely on
kinematics alone. A machine, as defined above, is not totally self-sufficient: it
must receive and then transform energy imparted to it from an outside source.
To be represented in movement it must be associated with an energy source.
For a long time, kinematic mechanisms were powered by humans or ani-
mals. During this stage, it was an obvious tautology to compare the movement
of bodies to the movement of a machine, when the machine itself depended
on humans or animals to run it. Consequently, it has been shown that mecha-
nistic theory has depended, historically, on the assumption that it is possible
to construct an automaton, meaning a mechanism that is miraculous in and
of itself and does not rely on human or animal muscle power.
This is the general idea put forth in the follOwing well-known text:
Examine carefully the physical economy of man: What do you find? The jaws are
armed with teeth, which are no more than pincers. This stomach is nothing but a
retort, or heat chamber; the veins, the arteries and indeed the entire vascular sys-
tem are simply hydraulic tubes; the heart, a pump; the viscera, nothing but filters
and sieves; the lungs, a pair of bellows; and what are muscles if not a system of
cables and ropes. What is the oculomotor nerve, if not a pulley? And so on. Try
as they will, chemists cannot explain nature and set up a separate philosophy sim-
ply by coining a new vocabulary around words like "fusion," "sublimation" and
. "precipitation"; for this does not at all address either the incontrovertible laws of
equilibrium or the laws governing the workings of the wedge, cables, pumps as
elements of mechanical theory.
This text is not where we might think to find it, but in fact comes from the
Praxis medica, written by Baglivi in 1696, an Italian doctor belonging to the
iatromechanical school. This school, founded by Borelli, had apparently been
influenced by Descartes, although for reasons of national prestige, the Italians -
prefer to attribute it to Galileo.
This text is interesting because it treats the
wedge, the rope, the cable and the pump as if they could be seen in the same
terms for formulating explanatory principles. It is clear, however, that from
the mechanistic point of view there is a difference between these devices: a
cable essentially transmits a given movement, whereas a pump transforms a
given movement and is also a motor - admittedly, a motor that returns what-
ever energy it receives; but, at certain intervals, it apparently has a degree of
independence of movement. In Baglivi's text, the heart is the primum movens
- the central pump that serves as the motor for the whole body.
Therefore, a crucial element behind the mechanical explanation of bodily
movement is that, in addition to machines that perform as kinematic devices,
Machine and Organism
there are also machines that act as motors, deriving their energy, at the mo-
ment it is utilized, from a source other than animal muscle. And this is why,
although Baglivi's text seems linked to Descartes, the idea of the body-as-
machine actually goes back to Aristotle. When dealing with the Cartesian
theory of the animal-machine, it is often difficult to decide whether or not
Descartes had any precursors for this idea. Those who look for Descartes's
predecessors here usually cite Gomez Pereira, a Spanish doctor of the second
half of the sixteenth century: Pereira suggested, before Descartes, that he
could demonstrate that animals were wholly machines and that they do not
possess that sensitive soul so frequently attributed to them.
But in other
respects, it is unquestionably Aristotle who saw the congruity between animal
movements and automatic mechanical movements, like those observed in
instruments of war, especially catapults. This idea is treated rather extensively
by Alfred Espinas, who discusses the connection between the problems dealt
with by Aristotle in De Motu animalium and those in his compilation of
Quaestiones mechanicae.
Aristotle draws a clear parallel between the organs
of animal movement and "oTBana ," or parts of war machines, like the arm of
a catapult about to launch a projectile. Thus catapults, typical automatic
machines of the period, seemed to be articulated like a human limb, as they
were pOised and made to release their great stores of pent-up energy. In the
same work, Aristotle carries the analogy even further by comparing the move-
ment of our limbs to mechanisms; and he makes his case in much the same
way that Plato did when, in the Timaeus, he compared the movement of ver-
tebrates to hinges or pivots. " ,
It is true that in Aristotle the theory of movement is somewhat different
from what it would become in Descartes. According to Aristotle, the soul is
the principle of all movement. All movement first presupposes immobility
and then requires a prime mover or some motivating force. Desire moves
the body, and desire is explained by the soul, just as potentiality is explained
by an act. Despite their differing explanations of movement, for Aristotle as
for Descartes later, the comparison of the body with a machine presupposes
that man is composed of automated mechanical parts reliant on aQ energy
source that produces motor effects over time and continue to do so well after
the original (human or animal) energy has dissipated. It is this discrepancy
between the storage of energy to be released by the mechanism and the mo-
ment of release that allows us to forget the relation of dependence between
the effects of the mechanism and the actions of a body. When Descartes looks
to machines to explain how organisms work, he invokes spring-operated and
hydraulic automata. As a result, he owes a great intellectual debt to the ideas
behind the technical creations of his own time, including clocks"and watches,
water mills and church organs of the early seventeenth century. We can say,
then, that as long as the concept of the human and animal body is inextricably
"tied" to the machine, it is not possible to offer an explanation of the body
in terms of the machine. Historically, it was not possible to conceive of such
an explanation until the day that human ingenuity created mechanical devices
that not only imitated organic movements - as in the launching of a projec-
tile or the back-and-forth movement of a saw - but also required no human
intervention except to construct them and set them going.
In two instances, I have asserted that an explanation cannot be formulated
without the existence of certain conditions. Is this tantamount to attributing
a historical necessity to scientific explanation? How do I explain'the abrupt
appearance in Descartes of a lucid mechanistic interpretation of biological
phenomena? This theory is clearly related to modifications that occurred in
the economic and political structure of Western society, but the nature of
this relation remains obscure.
This problem has been treated in depth by P.-M. Schuhl, who has shown
that in ancient philosophy the opposition of science and technique paralleled
the opposition of freedom and servitude and, at a deeper level, of art and na-
Schuhl supports this parallel with Aristotle's assertion that natural and
violent movement are opposed - a violent movement occurs when mecha-
nisms are used against nature, and its characteristics are that it exhausts itself
rapidly and never becomes habitual - which is to say, a permanent tendency
to reproduce itself never obtains.
Here I must turn to the difficult problem of the history of civilization
and the philosophy of history. With Aristotle, the hierarchy of freedom and
servility, of theory and practice, of nature and art, is paralleled by an eco-
nomic and political hierarchy in the cities, namely, the relations of freemen
and slaves. The slave, according to Aristotle in the Politics, is an animated
machine. 10 This is the crux of the problem to which Schuhl only alludes in
passing: Did the Greek conception of the dignity of science lead to their dis-
dain for technique and the resultant paucity of inventions? And did this in
turn lead to the difficulty of applying the results of technical activity to the
explanation of nature? Or, rather, did the Greeks' high regard for purely
speculative science and detached contemplation explain the absence of tech-
nical invention? Did their disregard for work cause slavery, or did the abun-
dance of slaves due to military supremacy explain their low regard for work?
Are we obliged to explain the ideology in terms of the socioeconomic struc-
ture or, rather, the socioeconomic structure in terms of the ideology? Did
the ease of exploiting human beings make it easier to disdain the techniques
that would allow them to exploit nature? Does the arduousness of exploiting
nature justify the explOitation of man by man? Is there a causal relationship
at work here? And if so, in which direction does it go? Or are we dealing with
.1 ... ~ ......
Machine and Organism
a global structure having reciprocal relations and influences?
A similar problem is presented by Father Lucien Laberthonniere, who
contrasts the physics of an artist or an aesthete to that of an engineer and an
artisan.ll Laberthonniere suggests that the determining factor here is ideas,
given that the Cartesian transformation in the philosophy of technique pre-
supposes Christianity. It was necessary to conceive of man as a being who
transcends nature and matter in order to then uphold his right and his duty
to exploit matter ruthlessly. In other words, man had to be valorized so that
nature could be devalorized. Next it was necessary to conceive of men as
being radically and originally equal so that, as the exploitation of humans by
each other was condemned on political grounds, there were increased tech-
nical means to exploit nature and a growing sense of duty to do so. This analy-
sis permits Laberthonniere to speak of a Christian origin for Cartesian physics.
However, he qualifies his own claim: the physics and technique supposedly
made possible by Christianity came, for Descartes, well after Christianity had
been founded as a religion. Moreover, humanist philosophy, which saw man
as master and proprietor of nature, was in direct opposition to Christianity
as humanists saw it: the religion of salvation, of escape into the hereafter, in-
spired by a contempt for the things of this life and unconcerned with whatever
fruits technology might win for mankind in this world below. Laberthonniere
asserts that "time -does not enter into the question," but this is by no means
certain. In any case, several classic texts have demonstrated that certain techni-
cal inventions that transformed the use of animal motor power - for example,
the horseshoe and the shoulder harness - accomplished more for the eman-
cipation of slaves than did the countless preachings of abolitionists.
In Del (jbeI8an8 vomftudalem zum biiI8eIlichen Weltbild, Franz Borkenau
argues that there is a causal relationship between mechanistic philosophy and
the totality of social and economic conditions in which it arises. 12 He claims
that at the start of the seventeenth century the qualitative philosophy of antiq-
uity and the Middle Ages was eclipsed by mechanistic ideas. The success of
these new ideas was, on the level of ideology, an effect of the economic fact
of the new organization and expansion of manufacturing. For Borkenau, the
division of artisanallabor into separate, simplified operations requiring little
skill produced the concept of abstract social labor. Once labor had been de-
composed into simple, identical and easily repeatable movements, price and
wages could be determined simply by comparing the hours worked - and the
result was a process that, preViously qualitative, had become quantifiable. 13
Calculating work in purely quantitative terms that can be treated mathemat-
ically is claimed to be the basis and the starting point for a mechanistic con-
ception of the life world. It is therefore by redUCing all value to economic
value, "to cold hard cash," as Marx puts it in The Communist Manifesto, that
I '
I .
the mechanistic view of the universe is supposed to be fundamentally a
Weltanschauung of the bourgeoisie. Finally, Borkenau claims that the animal-
machine gives rise to the norms of the nascent capitalist economy. Descartes,
Galileo and Hobbes are thus the unwitting heralds of this economic revolution.
Borkenau's theses have been analyzed and criticized more forcefully by
Henryk Grossmann. 14 According to him, Borkenau ignores five hundred years
of economic and ideological history by seeing mechanistic theory as coincid-
ing with the rise of manufacturing at the beginning of the seventeenth cen-
tury: Borkenau writes as if Leonardo da Vinci had never existed. Referring
to Pierre Duhem's Les Ori8ines de la statique (1905), and the publication of
Leonardo's manuscripts (Herzfeld, 1904; Gabriel Seailles, 1906; Peladan,
1907), Grossmann agrees with Seailles that with the publication of Leonardo's
manuscripts it became clear that the origins of modern science could be
pJlshed back by more than a century. The quantification of the notion of work
occurs first within mathematics, well before its economic rationalization.
The norms of the capitalist evaluation of production, moreover, had been
defined by the Italian bankers even in the thirteenth century. Relying on
Marx, Grossmann reminds us that although in general there was no division
of labor in manufacturing properly speaking, manufacturing at its inception
meant the gathering together in the same place of skilled artisans who had
previously worked independently. According to Grossmann, then, it is not
the calculation of cost per hour of work, but the evolution of mechanization
that is the real cause of the mechanical view of the universe. The development
of mechanization begins during the It is, therefore, more accu-
rate to say that Descartes had consciously rationalized a mechanistic technique
than that he had unconsciously expressed the imperatives of a capitalist econ-
omy. For Descartes, mechanics is a theory if machines that presupposes a spon-
taneous invention which science must then consciously promote and develop. -
Which machines did the most to modify the relationship between man
and nature before the time of Descartes, far beyond the wildest imaginations _
of the ancients - and did most to justify and rationalize the hopes men had
vested in machines? Above all there were firearms, which hardly interested
Descartes except in terms of the problem of the projectile. 16 the other
hand, Descartes was very interested in clocks and watches, in lifting machines,
in water-driven machines and other related devices. As a result, one should
say that Descartes made a human phenomenon - the construction of ma-
chines - into an integral part of his philosophy; and one should avoid saying
that he transposed the social phenomena of capitalist production into ideology.
The key question becomes: How does Cartesianism account for an internal
principle of goal-directed activity in mechanisms, as is implied in the compar-
ison of a machine with an organism?
~ .--
Machine and Organism
The theory of the animal-machine is inseparable from "I think therefore I
am." The radical distinction between the soul and the body, between thought
and extension, requires the affirmation that matter, whatever form it adopts,
and thought, whatever function it fulfills, are each an undivided substance. 17
Because the only function of the soul is judgment, it is impossible to admit
the existence of a soul in animals, since we have no proof that animals judge,
incapable as they are of language or invention. 18
For Descartes, though, the refusal to attribute a soul- that is, reason - to
animals, does not necessarily lead to the conclusion that animals are not alive
(since not much more than a warm, beating heart is at issue); nor must ani-
mals be denied sensibility, to the extent that such sensibility is solely a func-
tion of their organs. 19
In the same discussion, a moral foundation for the animal-machine theory
comes to light. Descartes views the animal as Aristotle had viewed the slave,
devalorizing it in order to justify man's using it to serve his own purposes: "My
opinion is no more cruel to animals than it is overly pious toward men, freed
from the superstitions of the Pythagorians, because it absolves them of the hint
of crime whenever they eat or kill animals."2o And it comes as no small surprise
to find the same argument in reverse in a passage of Leibniz: "if we are com-
pelled to view the animal as being more than a machine, we would have to
become Pythagorians and renounce our domination of animals."21 And so we
confront an attitude typical of Western thought. On the theoretical level, the
mechanization oflife only considers animals to the extent that they serve man's
technological ends. Man can only make himself the master and proprietor of
nature if he denies any natural finality or purpose; and he must consider the
whole of nature, including all life forms other than himself, as solely a means
to serve his purposes.
This is how the mechanical model of the living organism, including the hu-
man body, was legitimized; for already in Descartes the human body, if not Animal testing, c. 1970.
man's entire self, is seen as a machine. As I have already noted, Descartes based
his mechanical model on automata, that is, 'on moving machines.
In order to see the full implications of Descartes's theory, I now intend to
look at the beginning of his "Treatise on Man:' which was published for the
first time in Leyden in 1662. He wrote there:
These men will be composed, as we are, of a soul and a body. First I must describe
the body on its own, then the soul, again on its own; and finally I must show how
these two natures would have to be joined and united in order to constitute men
who resemble us.
I suppose the body to be nothing but a statue or machine made of earth, which
God forms with the explicit intention of making it as much as possible like us.
Thus God not only gives it externally the colors and shapes of all the parts of our
bodies, but also places inside it all the parts required to make it walk, eat, breathe,
enabling it to imitate all those functions which seem to proceed from matter and
to depend solely on the interacting movements of our organs.
We see clocks, artificial fountains, water mills and other such machines which,
although only man-made, seem to move of their own accord in various ways;
but I am supposing this machine to be made by the hands of God, and so I think
you may reasonably think it capable of a greater variety of movements than I
could possibly imagine in it, and of exhibiting more artistry than I could possibly
ascribe to it.
Were we to read this text as naively as possible, the theory of the animal-
machine would seem to make sense only if we put forward two important
and often-neglected postulates. The first is the existence of a God who builds
things, and the second that living bodies are given in essence before machines
are constructed. In other words, to understand the machine-animal, it is
necessary to see it as being preceded, logically and chronologically, by God,
who is an efficient cause, and by a preexisting living model after which it is
to be modeled or imitated, which is a formal and final cause. With all this in
mind, I propose to take the animal-machine theory, which is usually seen as
a departure from the Aristotelian concept of causality, and show how all of
Aristotle's types of causality are nonetheless found in it, but not always in the
same place or simultaneously.
If we read the text more closely, we see that in order to construct the liv-
ing machine
it is necessary to imitate a preexisting living model. The con-
struction of a mechanical model presupposes a living original (Descartes is
perhaps closer here to Plato than to Aristotle). The platoniC Demiurge copies
the ideas, and the Idea is the model of which the natural object is a copy. The
Cartesian God, the Art!fox maxim us, works to produce something equivalent
to the living body itself. The model for the living machine is that body itself.
Divine art imitates the Idea - but the Idea is the living body. What is more,
in the same way that a regular polygon is inscribed in a circle, and that one
must pass an infinite distance to deduce one from the other, there is some-
thing of the machine in every aspect of life; but to pass from one to the other
would require crossing over an infinite gap, one that only God can close. This
is the idea brought out at the end of the text: "but I am supposing this machine
to be made by the hands of God, and so I think you may reasonably think it
capable of a greater variety of movements than I could possibly imagine in it,
and of exhibiting more artistry than I could possibly ascribe to it." The the-
ory of the animal-machine would, therefore, have the same relation to life
that a set of axioms has to geometry, that is, nothing more than a rational
reconstruction. Thus, the theory operates by deception: it pretends to ignore
the concrete existence of what it must represent, and it denies that what it
:" 1 \ '
.j- '
I ,


Machine and Organism
actually produces comes only after it has been rationally legitimized.
This aspect of Cartesian theory, moreover, was accurately assessed by a
contemporary anatomist, the noted Nicolaus Steno, in the Dissertation on the
Anatomy if the Brain delivered in Paris in 1665, a year after the "Treatise on
Man" had appeared. While paying homage to Descartes (which was remark-
able, since anatomists had not always been very accepting of Cartesian anat-
omy), he notes that Descartes's man was man reconstituted by Descartes
with God as a foil, but that this was not man as the anatomist understands
him. One can therefore say that by substituting the body for the machine,
Descartes removed teleology from life, but in appearance only, because he
has concentrated it in its entirety at the point at which life begins. A dynamic
structure is replaced by an anatomical one; but since this form is produced by
technique, all possible sense of teleology has been confined to the technique
of production. In fact, it appears that mechanical theory and purposiveness
cannot be placed in opposition, nor can mechanism and anthropomorphism.
If the functioning of a machine can be explained by relations of pure causal-
ity, the construction of a machine cannot be understood without taking two
things into consideration: a specific goal-directed activity and man himself.
A machine is made by man and for man, to achieve specific ends, to produce
a given series of effects. 25
The positive element, then, in Descartes's attempt to explain life mechani-
cally is that he eliminates the need to tie mechanism to finality in its anthropo-
morphic aspect. it seems that in doing this, one anthropomorphism
has been substituted for another. A technological anthropomorphism has been
substituted for a political anthropomorphism.
In "Description of the Human Body and All 0' Its Functions:' a short trea-
tise written in 1648, Descartes addresses the question of voluntary movement
in man: he offers, in terms so lucid that they were to dominate the entire the-
ory of reflex and automatic movements up until the nineteenth century, the
explanation that the body obeys the soul only on condition that the body is
primed mechanically to do so. For the soul to decide to move is not a suffi-
cient condition to induce the body to move. "The soul," writes Descartes
"cannot produce any movement without the appropriate disposition of the
bodily organs which are required for making the movement. On the contrary,
when all the bodily organs are appropriately disposed for some movement,
the body has no need of the soul in order to produce that movement."26
Descartes means that when the soul moves the body it does not act like a king
or a general commanding his subjects or his troops as is popularly conceived.
Rather, by viewing the body as a clock mechanism he envisions each organ
driving the other like interlocking cogwheels. So Descartes substitutes for
the image of the political chain of command - where commands are passed
by signals or spoken orders, through a type of magical causality - the techno-
logical image of "control," in which a desired series of operations is activated
by a controlling device or coordinated by a series of mechanical linkups.
Descartes takes the exact opposite position of Claude Bernard who, in his
critique of vitalism, in Leons sur les phenomenes de la vie communs aux animaux
et aux vesetaux, refuses to admit that a vital force could have a separate exis-
tence because it "cannot do anything" - but he does admit, surprisingly, that
it can "direct phenomena that it does not produce."z7 In other words, Bernard
replaces the notion of a vital-force-as-worker with the idea of vital-force-as-
legislator or guide. This is a way of admitting that one can direct events with-
out taking action - which borders on a kind of magical concept of direction,
implying that the overall operation transcends the execution of individual
operations. On the contrary, according to Descartes, a mechanical operation
replaces the power of direction and command, but God has fixed the direc-
tion once and for all: the constructor includes the guide-controls within the
mechanical process itself.
In short, with the Cartesian explanation, it might appear that we have not
moved beyond the idea of finality or inner purposiveness. The reason for this
is that if we limit ourselves to the workings of the machine, everything can be
explained by the theory of mechanism; but the theory cannot account for the
construction of the machine itself. Machines do not construct other machines,
and it could even be said that, in a sense, explaining organs or organisms
through mechanical models amounts to explaining the organ by means of it-
self. At bottom, then, we are dealing 'with a tautology; for it can be shown -
and I shall indeed try to justify this view - that machines can be considered as
orsans if the human species. 28 A tool or a machine is an organ, and organs are
tools ot machines. And so it is hard to see how mechanism can be distin-
guished from purposiveness. No one doubts that a mechanism is needed to
ensure that a given operation is carried out successfully; and, conversely,
every mechanism must 'follow a precisely determined sequence toward per-
forming some particular task, since a mechanism cannot depend on random-
ness or chance. Therefore, the opposition would be between those mechanisms
whose purpose is manifest and those whose purpose remains latent. In the
case of a lock or a watch, their function is apparent, while the pincers of the '
crab, often considered a marvel of adaptation, have a latent purpose. As a
result, it seems impossible to deny that certain biological mechanisms serve
a set purpose. Let us consider an oft-cited example, which mechanistic biol-
ogists use to argue their case; namely, that of the woman's pelvis, which en- '
larges just before she gives birth. To deny that this enlargement might not in
someway be the fulfillment of a fundamental, purposive activity, we need only
view the question in another way: given that the largest-sized fetus exceeds
Mach i ne and Organism
the maximum size of the pelvis by 1 or 1.5 cm, it would be impossible to give
birth were it not for a loosening of the symphyses and a gradual rocking move-
ment toward the sacrococcygien bone which increases the diameter ever so
slightly beyond its maximum. It is understandable that one would not want to
believe that an act with such a specific biological purpose is allowed to occur
only by virtue of a mechanism with no real biological function. And "allow" is
indeed the word that applies here, since without this mechanism the act sim-
ply could not take place. It is well known that, when dealing with an unknown
mechanism, we have to make certain that it is in fact a mechanism - that is, we
have to know what ultimate purpose or function it is intended to serve. We
can come to no conclusions about how it is to be used, simply on the basis of
its form or its structure, unless we already know how the machine or similar
machines are used. As a result, it is necessary first to see the machine at work
before attempting to deduce the function from the structure.
We are now at the point where we can see the historical reversal of the Cartesian
relationship between the machine and the organism. It is a well-known fact
- and so need not be belabored - that in all organisms we observe the phe-
nomena of autoconstruction, automaintenance, autoregulation and autorepair.
In the case of the. machine, its construction is beyond its power and depends
on the skill of the mechanic. Its maintenance requires the constant attention
and watchfulness of the machinist; for we all know how the complex workings
of a machine can be irremediably damaged due to inattention and carelessness.
As for maintenance and repair, they demand the same periodic intervention
of human action. While there are machines that are self-regulating, these are
in fact machines that man has grafted onto another machine. The construc-
tion of servomechanisms or electronic automata merely displaces the question
of the man-machine relationship without changing it in any fundamental way.
Further, in the case of the machine there is a strict adherence to rational,
economical rules. The whole is rigorously the sum of its parts. The final effect
depends on the ordering of the causes. What is more, a machine functions
within narrowly defined limits, and these limits become all the more rigid with
the practice of standardization. Standardization leads to the Simplification of
basic models and spare parts, and to unified standards of measurement and
quality, which allows for the interchangeability of parts. Any individual part
can be exchanged for any other part meant for the same place - within, of
course, a margin of tolerance determined by manufacturing constraints.
Now that the properties of a machine have been defined in relation to those
of an organism, can one say that there is more or less purposiveness in a ma-
chine than in an organism?
i .
One would surely agree that there is more purposiveness in machines
than in organisms, since a machine seems to move uniformly, unidirectionally
toward completing a particular activity. A machine cannot replace another
machine. The more specific the end-result desired, the more the margin of
tolerance is reduced, and the more the machine's directiveness seems con-
centrated, focused on a particular end. It is well known that functions in the
organism are substitutable, organs are polyvalent. Although this substitutabil-
ity of functions and polyvalence of organs is not absolute, in comparison with
the same qualities in the machine, it is so considerable that any comparison
is quite obviously absurd.
As an example of the substitutability of
I can give a very simple and well-known case, that of aphasia in children. A
hemiplegia on the right side of the child's brain is almost never accompanied
by aphasia, because the other areas of the brain ensure the continuance of the
linguistic functions. In the case of the child who is less than nine months old,
any existing aphasia disappears very quickly. 30 As for the problem of the poly-
valent organs, I need simply note the fact that for a majority of organs, which
we have traditionally believed to serve some definite function, the truth is
that we have no idea what other functions they might indeed fulfill. This is
the reason that the stomach is said to be, in principle at least, an organ of
digestion. However, it is a fact that after a gastrectomy performed to treat an
ulcer, there are fewer problems with digestion than with those we observe
with hematopoiesis. It was finally discovered that the stomach behaves like
an internal secretive gland. And I might also cite yet another example - and
not,at all to be taken as some sort of miracle - which came to light during
a recent experiment performed by the biologist Courrier, at the College de
France. Courrier made an incision in the uterus of a pregnant rabbit, ex-
tracted a placenta from the uterus and placed it in the peritoneal cavity. This
placenta grafted itself onto the intestine and fed itself normally. When the
graft was performed, the rabbit's ovaries were ablated - meaning that the
function fulfilled by the corpus luteum during pregnancy was suppressed. At
that moment, all the placentas present in the uterus were aborted and only
the placenta situated in the peritoneal cavity came to term. Here is an exam-
ple of the intestine behaving like a uterus, and perhaps, one might even say,
more successfully.
In this case, then, it is tempting to reverse one of Aristotle's formulations
in his Politics: "For nature is not stingy, like the smith who fashions the Del-
phian knife for many uses; she makes each thing for a single use, and every
instrument is best made when intended for one and not for many uses."31
On the contrary, it seems that this definition of finality or purposiveness
would be more applicable to a machine than to an organism. One must be
willing to acknowledge, ultimately, that in an organism, a given organ can
r./". "

Machi ne and Organism
accommodate a diversity of functions. Clearly, an organism has a greater range
of activity than a machine. It is less bound by purposiveness and more open
to potentialities.32 Every aspect and every movement of the machine is calcu-
lated; and the working of the machine confirms how each calculation holds
up to certain norms, measures or estimates; whereas the living body functions
according to experience. Life is experience, meaning improvisation, acting as
circumstances permit; life is tentative in every respect. Hence the overwhelm-
ing but often misunderstood fact that life permits monstrosities. There are no
monstrous machines. There is no mechanical pathology, as Xavier Bichat noted
in 1801 in his General Anatomy, Applied to Physiology and Medicine. 33 Whereas
monsters are still living things, there is no way to distinguish between the nor-
mal and the pathological in physics and mechanics. Only among living beings
is there a distinction between the normal and the pathological.
Above all, it is work in experimental embryology that has led to the aban-
doning of such mechanistic representations when interpreting living phenom-
ena, primarily by demonstrating that once the embryo starts to develop, it
does not contain any kind of "specific mechanism" intended to produce auto-
matically one organ or another. There can be no doubt that this was Descartes's
conception as well. In his "Description of the Human Body:' he wrote: "If we
hc..d a good knowledge of what makes up the semen of some species of animal
in particular, for man, then we would be able to deduce from this
alone, using certain and mathematical reasoning, the complete shape and
conformation of each of its members, and likewise, reciprocally, if we knew
many particularities about conformation, it would be possible to deduce
from that what the semen is."34 However, as Paul Guillaume remarks, it seems
that the more we compare living beings to automatic machines, the more we
seem to understand their functions but the less we understand their genesis. 35
If the Cartesian conception were accurate, that is, if the living organism were
both preformed in the embryo and developed mechanistically, any modifica-
tion made in the earliest stages would tend to disrupt the development of the
egg or prevent development altogether.
However, this is hardly the case. According to a study in potential egg,devel-
opment, based on research by Driesch, Horstadius, Speman and Mangold, it
was shown that embryonic development cannot be' reduced to a mechanical
model without running into anomalies. Let us take the example of the exper-
iments conducted by Horstadius on the egg of a sea urchin. He cut an egg A
from a sea urchin at stage sixteen so that each part of the egg maintained a
horizontal symmetry, and then he cut egg B, with each part being vertically
symmetrical. He joined half of A with half of B and the egg developed nor-
mally. Driesch took the sea urchin egg at stage sixteen and pressed the egg
between two thin layers of cells, while modifying the reciprocal position of
the cells at the two poles; still, the egg developed normally. The results of these
two studies allow us to conclude that the same effect is achieved regardless of
how conditions are varied.
There is an even more striking experiment, in which Driesch took blasto-
meres from the sea urchin egg at stage two. By removing the blastomeres, either
mechanically or chemically in sea water lacking calcium salts, the result was
that each of the blastomeres gave birth to a larva which was perfectly normal
down to the smallest detail. Here, then, the result is the same regardless of
how the characteristics of a factor are changed. The quantitative change in
a given factor does not lead to a qualitative change in the result. Conversely,
when two sea urchin eggs are joined they result in a single larva that is larger
than normal. This is yet another confirmation that the result is unaffected by
the quantitative change in one of the factors. Whether the factors are multi-
plied or divided, the experiment yields the same results.
I should add that the development of all eggs cannot be reduced to this
schema. For quite some time there was a problem in knowing whether there
were two different kinds of eggs at issue: regulated eggs, like the eggs of sea
urchins, and mosaic eggs, like those of frogs, whose first blastomeres develop
in exactly the same way, whether they are dissociated or remain together. Most
biologists have recently come around to admitting that what distinguishes the
two phenomena is simply that determination occurs earlier in the so-called
mosaic eggs. On the one hand, the regulated egg starts to act like a mosaic
egg at a certain stage; on the other hand, at stage two the blastomere of the
frog egg yields a complete embryo, as does a regulated egg, if it is reversed. 36
Thus, it is illusory to deny the idea of purposiveness in organisms and to
attribute it to automatic functions, however complex we might imagine these
to be. As long as a machine cannot construct itself, and as long as an organ-
ism is not equal to the sum of its parts, it might seem legitimate to think that -
biological organization is the basis and the necessary condition for the exis-
tence and purpose of a inachine. From the philosophical point of view, it is
less important to explain the operation of a machine than to understand it.
And to understand it means to inscribe it in human history by inscribing
human history in life - not overlooking the fact that with the advent of man
there appeared a culture that was no longer entirely reducible to natural causes.
And so we arrive at the point where the machine is seen as afact of culture,
expressed in mechanisms that are themselves nothing more than an explain-
able fact of nature. In a celebrated text in "Principles of Philosophy," Descartes
writes, "It is certain that all the rules of mechanics belong to phYSics, to the
extent that all artificial thin8s are thereby natural. Since, for example, when a
watch counts the hours, by using the cogs from which it is made, this is no
less natural for it than it is for a tree to produce fruit."37 But, from our point

Machine and Organism
of view, we can and must reverse the relationship of the watch to the tree and
say that the cogs and generally all the components that make up a watch are
designed to produce a desired effect: all the parts of the mechanism are prod-
ucts of imagination, each piece fulfilling some final purpose or design that at
one time was only imagined or dreamed of; they are thus the direct or indirect
products of a technical activity that is as authentically organic as the flower-
ing of trees. And, on a more fundamental level, the process works with great
efficiency even though there is no more conscious observance of the rules
and laws of physics than there might be within vegetal life. Although the con-
struction of a machine might presuppose at some stage the understanding of
the logics of physics, it should not and cannot be forgotten that, as a matter
of chronology and biology, construction of machines took place well before
there was any understanding of physics.
However, another author has asserted, contrary to Descartes, that living
organisms cannot be reduced to a machine and, similarly, art cannot be re-
duced to science. The author in question is Kant, in his Critique ifJudBment.
While it is true that the French have not tended to look to Kant as a philoso'-
pher of technique, it is no less true that German authors greatly interested in
this question, especially after 1870, have done so .
In the "Critique of Teleological Judgment," Kant distinguishes between
the machine and the organism, while drawing on Descartes's favorite exam-
ple of the watch. In a machine, he states, each part exists for the other but
not because of the other: no part produces another part; no one part is pro-
duced by the entire unit; nor does one part produce another part of similar
kind. There is no watch that makes other watches. No part can replace itself.
And no machine can replace one of its own missing parts. And so, while a
machine possesses motor power, it has no transformational energy that might
propagate itself or be transmitted to an object outside the machine itself. Kant
draws a distinction between human skill and technology, which are marked
by intentionality, as opposed to involuntary life processes. But in an impor-
tant passage of the "Critique of Aesthetic Judgment," Kant defines the origi-
nality of human skill as it relates to knowledge:
Art, regarded as human skill, differs from science (as ability differs from knowledge)
in the same way that a practical aptitude differs from a theoretical faculty, as tech-
nique differs from theory. What one is capable of doing, as soon as we merely know
what ought to be done and therefore are sufficiently cognizant of the desired effect,
is not called art. Only that which a man, even ifhe knows it completely, may not
therefore have the skill to accomplish belongs to art. Camper describes very exactly
how the best shoes must be made, but he certainly could not make one.
This text is cited by Paul Krannhals in Der Weltsinn der Technik, and, following
Kant, he acknowledges that all technique is essentially primordial, meaning that
it cannot be reduced to a simple question of rationality. 39 Indeed, we tend to see
the skilled hand that adjusts a machine or the mind that carefully orchestrates
a production process as examples of "ingenuity," having their basis in instinct;
but these are in fact as difficult to explain as the production of mammalian eggs
outside the ovary, even in the event that the physiochemical composition of
protoplasm and of sexual hormones had been made entirely clear to us.
This is why the work of anthropologists (and not engineers) seems to shed
more light, however faint, on the question of the construction of machines. 40
Currently in France, ethnologists have come closest to creating a philosophy
of technique in which the philosophers themselves seem to have lost interest,
their main concern having been chiefly the philosophy of science. On the con-
trary, the ethnographers have generally focused their attention on the rela-
tionship between the production of the earliest tools, the first instruments
that were used to act upon and modify nature, and the ways these tools were
assembled or grouped together. The only philosopher in France I know to
have posed these questions is Alfred Espinas, in his classic text on Les OriBines
de la technoloBie.
This work includes an appendix, the outline for a course
taught at the Faculte des Lettres at Bordeaux around 1890, which dealt with
the will, and in which Espinas addressed, under the guise of will, the ques-
tion of practical human behavior and especially the invention of tools. By
borrowing the theory of organic extension from the German writer Ernst
Kapp, Espinas was able to explain the construction of the first tools. Kapp
first made his theories known in 1877.
According to the theory of exten-
sion, whose philosophical bases go back to Hartmann's The Philosophy if the
Unconscious and further back still to Schopenhauer, the earliest tools were
simply extensions of moving human organs. The flint, the club and the lever
extend and magnify the organic movement of the arm and its ability to strike.
This theory, like all theories, has its limits and runs into certain stumbling
blocks, especially when it is used to explain fundamental inventions, such
as fire and the wheel. In these cases, we would search in vain for the body
movements and the organs that fire and the wheel are supposed to prolong or
extend; but the explanation certainly works for instruments like the hammer
or the lever and all such related tools. In France, then, it was the ethnogra-
phers who sought out and compiled not only the facts but also the hypotheses
from which a biological philosophy of technique could be constituted. The
philosophical path was laid out by the Germans
- for example, the theory
of the development of inventions based on the Darwinian notion of variation
and natural selection, as advanced by Alard Du Bois-Reymond in his EifindunB
und Eifinder (1906), or again, by Oswald Spengler in Der Mensch und die Technik,
which presented the theory that machines are constructed as a "life tactic"44
- and is taken up again, independently it seems, by Andre Leroi-Gourhan
Machi ne and Organi sm
in his book MiJieu et techniques. Leroi-Gourhan attempts to explain the phe-
nomenon of the construction of tools by comparing it to the movement of
the amoeba, which extends substances out beyond its mass so that it might
seize and capture an object it wishes to digest:
If we are drawn to view the act of percussion as the fundamental technical activ-
ity, it is because we witness an act of touch or contact in almost every technologi-
cal process; but even though the amoeba's expansion always leads its prey through
the same digestive process, there is no one way of explaining the working of that
process - whether we view the material being digested or whether we approach
the question from any given view of technology - since our view must change
according to the circumstances, just as the digestive process itself might be like
the various specialized grasping or striking organs. 45
In the last chapters of this work one finds a theory of machine that is alto-
gether different from the traditional theories that, for lack of a better term,
I shall classify as Cartesian - where technical invention amounted to the
application of a given system of knowledge.
Traditionally, the locomotive is presented as a classic example of a "mar-
vel of science." However, the construction of the steam engine is only under-
standable when placed in light of theoretical knowledge that preceded it, as
the culmination of an age-old problem, and a specifically technological one
at that - how to pump water out of mines. And so it would be necessary to
understand the natural history of the development of the pump, and to know
about the fire pump (which at first did not rely at all on vapor but produced a
vacuum via condensation under the pistons, thereby allowing the atmospheric
pressure acting as a motor to lower the piston) in order to see that the essen-
tial "organ" in a locomotive is a cylinder and a piston.
Tracing a similar progression of ideas, Leroi-Gourhan goes even further,
pointing back to the wheel as one of the locomotive's ancestors, in the biolog-
ical sense of the word. "It is machines like the wheel," he states, "that gave rise
to steam engines and modern-day motors. All of the highest technological
achievements of the most inventive minds of our time can be grouped around
the circular movements of the crank, the pedal, the drive belt."47 He then goes
on to add: "The way inventions influenced each other has not been studied
sufficiently and we don't seem to take note of the fact that, without the wheel,
we would not have the locomotive."48 Further on:
At the beginning of the nineteenth century no one had yet recognized how to make
use of the elemental forms that would later give birth to the locomotive, the auto-
mobile and the airplane. The underlying principles of mechanics were spread
throughout twenty applications which had been known for many centuries. It i ~
here we find the principle that explains invention, but the defining characteristic
is that it in someway manifests itself spontaneously. 49
In light of these remarks, we see how science and technique must be consid-
ered as two separate areas; that is, they do not graft onto each other but, rather,
each takes from the other either its solutions or its problems. It is the rational-
izing and ordering imposed by technology that makes us forget that machines
have their origin in the irrational. In this area as in all others, it is necessary to
know how to accommodate the irrational, even when - and especially when
- we want to defend rationalism. 50
It must be added that the reversal of the relationship between the machine
and the organism, brought about by a systematic understanding of technical
inventions as if they were extensions of human behavior or life processes,
is in someway confirmed by the belief that the generalized use of machines
has slowly imposed contemporary industrialized society on man. George
Friedmann has shown very clearly the steps by which "body" gradually
became a first-order term in the human machine-body equation. 51 With
Frederick Taylor and the first technicians to make scientific studies of work-
task movements, the human body was measured as if it functioned like a
machine. If we see their aim as the elimination of all unnecessary movement
and their view of output as being expressed only in terms of a certain num-
ber of mathematically determined factors, then rationalization was, for all
intents and purposes, a mechanization of the body. But the realization that
technologically superfluous movements were biologically necessary move-
ments was the first stumbling block to be encountered by those who insisted
on viewing the problem of human-body-as-machine in exclUSively techno-
logical terms. From here on, the systematic examination of certain physio-
l o g i ~ a l , psychotechnological and even some psychological conditions (since
a consideration of values leads inevitably to questions at the very center of
the origin of human personality) finally culminated in a reversal, called an
inevitable revolution by Friedmann, in which technology would adapt ma-
chines to the human body. As Friedmann saw it, this industrial technology
appeared to take the form of a scientific rediscovery of the same entirely
empirical procedures through which primitive peoples had always sought
to have their tools meet the highest organic norms: that is, their tools had to
carry out a given action effectively while maintaining a biological economy;
. and this occurred at the optimum level, when it most closely approximated
the movement of the body at work, as when the body defends itself sponta-
neously from becoming exclusively subordinate to the mechanical. 52 In this
way, Friedmann could speak, without irony or paradox, of the legitimacy of
considering the industrial development of the West from an ethnographic
point of view. 53
In summary, by considering technology as a universal biological phenom-
and no longer simply as an intellectual operation to be carried out by
Machine and Organism
man, I am led to the following conclusions: on the one hand, the creative
autonomy of the arts and skilled crafts in relation to all forms of knowledge
that are capable of annexing them or expanding on them; and, on the other
hand, to inscribe the mechanical into the organic. It is no longer then, a ques-
tion of determining the extent to which an organism can be thought of as a
machine, whether by virtue of its structure or of its functions. But it is neces-
sary to find the reasons that gave rise to the opposite view, the CartesiaI) one.
I have attempted to shed light on this problem, suggesting that the mechanis-
tic conception of the body was no less anthropomorphic, despite appearances,
than a teleological conception of the physical world. The answer I am tempted
to offer would insist on showing that technology allows man to live in conti-
nuity with life, as opposed to a solution that would see humankind as living
in a state of rupture for which we ourselves are responsible because of sci-
ence. There is no doubt that this answer appears to lend credence to the list
of accusations that all too many writers have offered up nostalgically from
time to time, with no apparent regard to their lack of originality, as they point
out the faults of technology and progress. I have no intention of rushing to
support their cause. It is clear that ifhuman society has embraced the idea of
a technology based on a mechanistic model, the implications are enormous,
and the whole question cannot easily be treated lightly or recalled on demand.
But that model is altogether different from the one just examined.
1. After having been dogmatically accepted by biologists for many years, the mechanis-
tic theory of the organism is now considered narrow and inadequate by those scientists who
call themselves dialectical materialists. But the fact that they still concern themselves with
formulating a philosophical position could easily support the rather widespread idea that
philosophy does not possess its own domain, that it is a poor relation of speculation, and
must clothe itself in the hand-me-downs scientists have used and then discarded. It will be
my aim to show that the problem of machine and organism is much broader in scope and
more philosophically important than is commonly thought; ;md that it is far more than a
theoretical and methodological dispute among biologists.
2. Julien Pacotte, La Pensee technique (Paris: Alcan, 1931).
3. One example of the fundamental principles of a general theory of mechanisms un-
derstood in this way can be found in Franz Reuleaux's Theoretische Kinematik: Grundziiee
einer Theorie des Maschinwesen (Braunschweig: Vieweg, 1875) .
4. For everything concerning machines and mechanisms, see Pacotte, La Pensee tech-
nique, ch. 3.
5. According to Marx, a tool is moved by human power while the machine is moved
by a natural force; see his Capital, trans. Samuel Moore and Edward Aveling (New York:
International Publishers, 1967), vol. 1, pp. 374-79.
6. For more on this, see Charles Victor Daremberg, Histoire des sciences medicales
(Paris: Bailliere, 1870), vol. 2, p. 879.
7. Gomez Pereira, Antoniana Maraarita: Opus physicis, medicis ac theoloais non minus
utile quam necessarium (Medina del Campo, 1555-58) .
8. Alfred Espinas, "L' Organisation ou la machine vivante en Grece au IVe siecle avant
J.-c.," Revue de metaphysique et de morale (1903), pp. 702-15.
9. P.-M. Schuhl, Machinisme et philosophie (Paris: Alcan, 1938).
10. Aristotle's Politics, trans. Hippocrates G. Apostle and Lloyd P. Gerson (Grinnel,
Iowa: Peripatetic Press, 1986), bk. 1, ch. 2, secs. 4-7.
11. Lucien Laberthonniere, Les Etudes sur Descartes (Paris: Vrin, 1935), especially the
appendix to volume 2: "La Physique de Descartes et la physique d'Aristote."
12. Franz Borkenau, Der Uberaana yom jeudalem zum biiraerlichen WeltbiJd (Paris:
Alcan, 1934).
13. Jean de la Fontaine's fable, "The Cobbler and the Businessman" (in La Fontaine:
Selected Fables, trans. Jamie Michie [New York: Viking, 1979], pp. 188-91) is an excellent
illustration of the two different conceptions of work and its remuneration.
14. Henryk Grossmann, "Die gesellschaftlichen Grundlagen der mechanistischen
Philosophie und die Manufaktur," Zeitschriftfor Sozia!forschuna, 4th ser., vol. 2 (1935),
pp. 161-231.
15. "Mechanization" here means the generalized use of machines to replace human
labor. However, it was also used to describe Descartes's theory of animals as machines
before the nineteenth century when the above usage was in force - TRANS.
16. In Descartes's "Principles of Philosophy" (4.187 [AT 8A.314], Descartes: Selected
Philosophical Writinas, trans. John Cottingham, Robert Stoothoff and Dugald Murdoch
[New York: Cambridge University Press, 1988] , pp. 199-200), there are a few passages
that reveal Descartes to be equally interested in gunpowder, but he did not look for an
analogous explanatory principle for the animal organism in the explosion of gunpowder
as a source of energy. It was an English doctor, Thomas Willis, who explicitly formulated
a theory of muscular movement based on the analogy with what occurs when the powder
explodes in a harquebus. Inthe seventeenth century, Willis compared the nerves to pow-
der lines in a manner that remains valid today in some quarters - most notably, W M.
Bayliss comes to mind. Nerves are a sort of Bickford cord. They produce a spark that will
set off, in the muscle, an explosion that, in Willis's view, is the only thing capable of ac-
counting for the phenomena of spasm and prolonged contraction observed by the doctor.
17. "For there is within us but one soul, and this soul has within it no diversity of
parts: it is at once sensitive and rational too, and all its appetites are volitions" ("The
Passions of the Soul" 47, in Selected Philosophical Writinas, p. 236) .
18. "Discourse on Method" 5 (AT 6.56ff.), in ibid., p. 44ff. Letter to the Marquis of
Newcastle, Nov. 23, 1646.
19. Letter to Morus, Feb. 21, 1649, in Descartes, Correspondance, ed. Charles Adam
and Gerard Milhaud (Paris: P.U.F., 1963), vol. 8, pp. 121-39. In order to understand
Machine and Organism
adequately the relationship of sensibility to the arrangement of the organs, we must be
familiar with the Cartesian theory of the degrees of sensej on this subject, see Descartes,
"Author's Replies to the Sixth Objections" 9 (AT 7.436-39), in The Philosophical Writinas
cfDescartes, trans. John Cottingham, Robert Stoothoff and Dugald Murdoch (Cambridge,
Eng.: Cambridge University Press, 1984), vol. 2, pp. 294-96.
20. Descartes, Letter to Morus, Feb. 21, 1649, in Correspondance, vol. 8, p. 138.
21. Letter to Conring, March 19, 1678, in Gotifried Wilhelm Leibniz: Siimtliche Schriften
und Briefe (Darmstadt: Reichl, 1926), 2d ser., vol. 1, pp. 397-401. Leibniz's outline of
criteria in particular, which would allow us to distinguish an animal from an automaton,
should be compared to the analogous arguments adduced by Descartes, and also the pro-
found reflections of Edgar Allan Poe on the same subject in his "Maelzel's Chessplayer."
On the Leibnizian distinction between the machine and the organism, see "A New System
of the Nature and the Communication of Substances" 10, in Leibniz: Philosophical Papers
and Letters, trans. and ed. Leroy Loemker (Chicago: University of Chicago Press, 1956),
vol. 2; and "Monadology" 63-66, in Monadolo8.J and Other Philosophical Essays, trans. Paul
Schrecker and Anne Martin Schrecker (New York: Macmillan, 1985) .
22. It is important to point out that Leibniz was no less interested than Descartes in
the invention and construction of machines, as well as in the problem of automatons. See
especially his correspondence with Duke John of Hanover (1676-1679) in the Siimtliche
Schriften und Briefe (Darmstadt: Reichl, 1927), 1st ser., vol. 2. In a text of 1671, Bedenken
von Atifrichtuna einerAcademie oder Societiit in Deutschland zu Atifnehmen der Kunste und
Wissenschajten, Leibniz exalts the superiority of German art, which has always strived to
produce works that move (watches, clocks, hydraulic machines, and so on), over Italian
art, which has always attached itself exclusively to the fabrication of lifeless objects made
to be contemplated from without (ibid. [Darmstadt: Reichl, 1931], 4th ser., vol. 1, p. 544).
This passage is cited by Jacques Maritain in his Art and Scholasticism and the Frontiers cf
Poetry, trans. Joseph W Evans (New York: Scribners, 1962), p. 156.
23. "Treatise on Man" (AT XI. 119-20), in The Philosophical Writinas cfDescartes, trans.
John Cottingham, Robert Stoothoff and Dugald Murdoch (Cambridge, Eng.: Cambridge
University Press, 1985), vol. 1, p. 99.
24. This phrase is a traditional equivalent of "the human body," especially in the
eighteenth century - TRANS.
25. Moreover, Descartes can only express the meaning of God's construction of ani-
mal-machines in terms of finality: "considering the machine of the human body as having
been formed by God in order to have in itself all the movements usually manifested there"
("Sixth Meditation," in Philosophical Works cfDescartes [1913], trans. Elizabeth S. Haldane
and G. R. T. Ross [New York: Cambridge University Press, 1967], vol. 1, p. 83). [Here the
wording of the older translation is more literal than is the translation of Cottingham et aI.,
Philosophical Writinas cfDescartes, vol. 2, pp. 50-62 - TRANS.]
26. "Description of the Human Body and All of Its Functions" 1 (AT II. 225), in
Philosophical Writinas cfDescartes, vol. 1, p. 315.
27. Claude Bernard, Leons sur les phenomenes de la vie communes aux animaux et aux
veaetaux: 1878-1879 (Paris: Masson, 1936) .
28, For more on this idea, see Raymond Ruyer, Elements de psycho-bioloaie (Paris:
P.ll.F., 1946), pp. 46-47.
29. "Artificial means what is aimed at a definite goal . And is opposed therefore to
livina. Artificial or human or anthropomorphic are distinguished from whatever is only
living or vital . Anything that succeeds in appearing in the form of a clear and finite goal
becomes artificial and this is what tends to happen as consciousness grows. It is also true
of man's work when it is intended to imitate an object or a spontaneous phenomenon as
closely as possible. Thpught that is conscious of itself makes itself into an artificial sys-
tem .. .. If life had a goal, it would no longer be life" (Paul Valery, Cahier B [Paris: Gallimard,
30. See Ed. Pichon, Le Developpement psychique de l'erifant et de l'adolescent (Paris:
Masson, 1936), p. 126; and Paul Cossa, Physiopatholoaie du systeme nerveux (Paris: Masson,
1936), p. 845.
31. Politics, bk. 1, ch. 1 (1252b), in The Basic Works oj Aristotle, ed. Richard McKeon
(New York: Random House, 1941), p. 1128.
32. Max Scheler, in his Mans Place in Nature [1928] (trans. Hans Meyerhoff [Boston:
Beacon, 1961], pp. 75-81), has remarked that it is those living things that are the least
specialized that are the most difficult to explain by the mechanistic idea, pace the mecha-
nists, because in their case all functions are carried out by the whole organism. It is only
with the growing differentiation of functions and the increased complexity of the nervous
system that structures which resemble a machine in some fashion tend to appear.
_ 33. General Anatomy, Applied to PhysioloBY and Medicine, trans. George Hayward
(Boston: Richardson and Lord, 1822).
34. "Description du corps humain" 1 (AT II. 225), in Charles Adam and Paul
Tannery, eds., Ouevres de Descartes (Paris: Vrin, 1974), vol. 11, p. 225. [This pas age is omit-
ted from the English translation of "Description of the Human Body and of All of Its
Functions" - TRANS. ]
35. Paul Guillaume, La' Psycholoaie de la forme (Paris: Flammarion, 1937), p. 131.
36. Pierre Grasse and Max Aron, Precis de bioloaie animale (2d ed., Paris: Flammarion,
1947), p. 647ff.
37. 4.203, in Philosophical Writinas ojDescartes, p. 288. See also my study "Descartes
et la technique," Travaux du Conares International de Philosophie, vol. 2: Etudes cartesiennes
(Paris: Hermann, 1937), p. 77ff.
38. "An organized being is not a mere machine, for that has merely moving power, but
it possesses in itself formative power of a self-propagating kind which it communicates to
its materials though they have it not of themselves; it organizes them, in fact, and this can-
not be explained by the mere mechanical faculty of motion" (Critique oj judament, trans.
J. H. Bernard [New York: Hafner, 1951], p. 22).
39. Krannhals, Der Weltsin der Technik (Munich and Berlin: Oldenbourg, 1932), p. 68.


Machine and Organism
40. The starting point for these works must be sought in Darwin, The Descent if Man -
whose ideas Marx saw clearly as immensely significant.
41. Alfred Espinas, Les OriBines de la techn%Bie (Paris: Alcan, 1897).
42. Ernst Kapp, Grundlinien einer Philosophie der Technik (Braunschweig:
Westermann, 1877). This work, which was a classic in Germany, has remained so misun-
derstood in France that certain psychologists who took up the problem of how animals
utilize tools, and animal intelligence, and who took the research of Kohler and Guillaume
as their starting point, attributed this theory of projection to Espinas himself, without
noting that Espinas states explicitly, at numerous junctures, that he borrowed it from
Kapp. I am alluding here to the excellent little book by Gaston Viaud, Son
evolution et sesJormes (Paris: P.U.F., 1946).
43. See Eberhard Zschimmer's Deutsche Philosophen der Technik (Stuttgart: Enke, 1937).
44. Alard Du BOis-Reymond, EifindunB und Eifinder. (Berlin: Springer, 1906); and
Oswald Spengler, DeI Mensch und die Technik (Munich: Beck, 1931). Alain outlined a
Darwinian interpretation of technical constructions in a fine remark (Les Propos
[Paris: N.R.F., 1920], vol. 1, p. 60), preceded and followed by some others that are most
pertinent to our problem. The same idea is referred to many times in the Systeme des
Beaux-Arts, concerning the making of the violin (4.5), furniture (6.5), houses in the coun-
tryside (6.3, 6.8) .
45. Andre Leroi-Gourhan, Evolution et technique, vol. 2: Milieu et techniques (Paris:
Michel, 1945) .
46. The double-acting engine, in which the steam acted on the upper and lower sides
of the piston alternately, was perfected by Watt in 1784. Sadi Carnot's Rijlexions sur la
puissance motrice du feu dates from 1824, and we know that it was ignored until the middle
of the nineteenth century. On this subject, see Pierre Ducasse, Histoires des techniques
(Paris: P.U.F., 1945), which stresses that technique precedes theory.
On the subject of the empirical succession of the various organs and uses of the steam
engine, consult Arthur Vierendeel's Esquisse d'une histoire de la technique (Brussels and
Paris: Vromant, 1921), which summarizes Thurston's extensive work, History if the Steam
EnBine. For more about the history of Watt's work as an engineer read the chapter enti-
tled "James Watt ou Ariel ingenieur," in Pierre Devaux's Les Aventures de la science (Paris:
Gallimard, 1943).
47. Leroi-Gourhan, Milieu et techniques, p. 100. The same view can be found in an
article by A. Hadricourt on "Les Moteurs animes en agriculture" (Revue de botanique
appliquee et d'aBriculture tropicale 20 [1940], p. 762) : "We must not forget that we owe our
inanimate motors to irrigation: the noria is at the origin of the hydraulic mill, just as the
pump is at the origin of the steam engine." This excellent study sets out the principles for
explaining tools from the perspective of their relationship to organic commodities and
the traditional ways they were used.
48. Leroi-Gourhan, Milieu et techniques, p. 104.
49. Ibid., p. 406.
50. In his The Two Sources of Morality and Reliaion (trans. R. Ashley Andra and
Cloudesley Brereton [New York: Holt, 1949]), Henri Bergson thinks very explicitly that
the spirit of mechanical invention, although it is fed by science, remains distinct from
it and can even, if necessary, be separated from it (pp. 329-30). The fact is that Bergson
is also one of the rare French philosophers, if not the only one, who has considered me-
chanical invention as a biological function, an aspect of the organization of matter by life:
Creative Evolution (trans. Arthur Mitchell [New York: Modern Library, 1944) is, in some
sense, a treatise of general organology.
On the subject of the relationship between explanation and action see also Paul Valery,
"L'Homme et la coquille" and "Discours aux chirurgiens;' in Varifi"te V (Paris: Gallimard,
1945), and his description of boat building in Eupalinos.
And, finally, read the admirable "In Praise of Hands" in Henri Focillon, The Life of
Forms in Art (New York: Zone Books, 1989), pp. 157-84.
51. George Friedmann, Problemes humains du machinisme industrielle (Paris: Gallimard,
1946) .
52. Ibid., p. 96, note.
53. Ibid., p. 369.
54. This attitude is one that has begun to be familiar among biologists. In particular,
see L. Cuenot, Invention etflnalite en bioloaie (Paris: Flammarion, 1941); and Andree Tetry,
Les Outils chez les etres vivants (Paris: Gallimard, 1948) - especially the latter's reflections
on "Adaptation and Invention" (p. 120ff.). It is impossible to mistake the impetus given to
these treatments by the ideas of Teilhard de Chardin.
A new discipline, Bionics, which emerged around ten years ago in the United States,
studies biological structures and systems able to be utilized as models or analogues by
technology, notably by builders of systems for detection, direction and equilibration
meant for equipping planes or missiles. Bionics is the extremely subtle art of information
that has taken a leaf from natural life. The frog, with its eye capable of selecting informa-
tion that is instantly usable, the rattlesnake, with its thermoceptor which traces the blood
of its prey at night, the common fly, balancing itself in flight by means of two vibratile
filaments, have all furnished models for this new breed of engineers. In many American
universities, special training in Bioengineering is available, for which the Massachusetts
Institute of Technology seems to have been the instigator. See the article by J. Dufrenoy,
"Systemes biologiques servant de modeles 11 la technologie," Cahiers des inaenieurs aaronomes
(June-July, 1962), p. 21.
Translated from the French by Mark Cohen and Randall Cherry
The Living and Its Milieu
The notion of milieu is in the process of becoming a universal
and obligatory means of registering the experience and exis-
tence of living things, and one could almost speak of its consti-
tution as a basic category of contemporary thought.
But until
now, the historical stages of the formation of the concept, its
diverse uses, as well as the successive recongurations of the
relationships in which it takes part, whether in geography, biol-
ogy, psychology, technology, or social and economic history, all
make it rather difcult to make out a coherent whole. For this
reason philosophy must, here, initiate a synoptic study of the
meaning and value of the concept. By initiate I do not simply
mean the pretense of an initiative that would consist in taking a
series of scientific investigations for reality and then con-
fronting expectations with results. Rather, it is a question of
using several approaches and engaging them in a critical con-
frontation with each other to locate, if possible, their common
point of departure and to explore its potential richness for a
philosophy of nature that focuses on the problem of individu-
ality. It is therefore appropriate to examine the simultaneous
and successive elements of the notion of milieu each in turn, the
various usages of this notion from 1800 to the present, the many
inversions of the relationship between organism and milieu,
and nally the general philosophical impact of these inversions.
Historically considered, the notion and the term milieu
are imported from mechanics to biology in the second half of
the eighteenth century. The mechanical idea, but not the term,
appears with Newton, and the word milieu is present in
dAlembert and Diderots Encyclopedia with its mechanical
meaning, in the article of the same name. It is introduced to
biology by Lamarck, who was himself inspired by Buffon,
though he never used the term other than in the plural. De
Blainville seals this usage. Etienne Geoffroy Saint-Hilaire in
1831 and Comte in 1838 use the term in the singular, in an
abstract sense. Balzac opens the gates to literature in 1842, in
the preface of the Comdie Humaine, and it is Taine who rst
uses it as one of the three analytical principles used to explain
history, the two others being race and event, as is well known.
It is more due to Taine than Lamarck himself that neo-
Lamarckian biologists in post-1870 France, such as Giard, Le
Dantec, Houssay, Costantin, Gaston Bonnier, and Roule, use
this term. They get the idea, in a sense, from Lamarck, but the
Grey Room 03, Spring 2001, pp. 731. 2001 Grey Room, I nc. and Massachusetts I nstitute of Technology 7
term as an abstract and universal one comes to them from Taine.
French mechanists of the eighteenth century called milieu
what Newton meant when he said uid. The model for this,
if not the sole archetype in Newtons physics, was ether. In
Newtons day, the problem facing mechanics was that of the
action of distinct physical bodies at a distance. This was the
fundamental problem of the physics of central forces. It was a
problem that had not existed for Descartes. For him, there was
only one mode of physical action, impact, in only one possible
physical situation, that of contact. This is why we can say that
in Cartesian physics the notion of milieu has no place. Subtle
matter is not in any way a milieu. But it was difcult to extend
the Cartesian theory of impact and contact to the case of sepa-
rate point particles, since in this case they could not act with-
out being confounded by this action. As a result, we can see
that Newton was led to pose the problem of the means of the
action. Luminous ether was for him the uid that served as the
vehicle of action at a distance. This explains the passage from
the notion of uid as a vehicle to its designation as a medium
[milieu]. The uid is the intermediary between two bodies; it is
their milieu; and to the extent that it penetrates these bodies,
they are situated within it. According to Newton and to the
physics of central forces, it is only because there are centers of
force that we can speak of environment, that we can speak of a
milieu. The notion of milieu is a fundamentally relative notion.
To the extent that we consider separately the body on which the
action, transmitted through the medium, is exercised, we may
forget that the milieu is a between two centers and remember
only its function of centripetal transmission, and one might say
its ambient situation. In this way the milieu tends to lose its rel-
ative meaning and takes on an absolute one. It becomes a real-
ity in itself.
It was perhaps Newton who was responsible for importing
the term from physics to biology. Ether helped him not only
resolve the problem of illumination, but also explain the phys-
iological phenomenon of vision and even explain the physio-
logical effects of luminous sensation, in other words, explain
muscular reactions. In his Optics, Newton considered ether to
be continuous with air, something found in the eye, in the
nerves, even in muscles. It was therefore the action of a milieu
that ensured the dependent connection between the spark of
the perceived luminous source and the movement of muscles
by which man reacts to this sensation. This, it would seem, was
the rst example of an explanation of an organic reaction by the
action of a milieu, that is, of a uid strictly dened by physical
Indeed, the article in the Encyclopedia cited above
confirmed this way of seeing things. All of the examples of
milieus given in the article were drawn from Newtons physics.
8 Grey Room 03
And it is in a purely mechanical sense that one says that water
is a milieu for the sh who move around in it. It is also primar-
ily in this mechanical sense that Lamarck understood it.
Lamarck always speaks of milieus in the plural, and by this
he specifically means fluids like water, air, and light. When
Lamarck wants to designate the whole set of outside actions
that are exercised on a living thing, in other words what we call
today the milieu, he never says milieu but always influ-
ential circumstances. As a result, circumstance is a genus
within which climate, place, and milieu are species. And this
is why Lon Brunschvicg, in Les Etapes de la philosophie
mathmatique, wrote that Lamarck borrowed from Newton the
physicomathematical model of explaining the living through a
system of connections with its environment.
The relationship
between Lamarck and Newton is intellectually direct and his-
torically indirect, as they are linked through Buffon. We can,
for example, recall that Lamarck was Buffons pupil and his
sons tutor.
Buffon, in fact, combines two influences in his conception
of the relationship between organism and milieu. The first is
precisely Newtons cosmology, of which Buffon was a lifelong
The second influence is the tradition of anthropo-
geography, which had been kept alive in France by Montesquieu
before him,
following Bodin, Machiavelli, and Arbuthnot. The
Hippocratic treatise On Airs Waters and Places can be consid-
ered the first work that gave philosophical form to this idea.
These are the components that Buffon brought together in his
principles of animal ethology, to the extent that animal mores
are of a distinct and specic character and that these mores can
be explained by the same method that allows geographers to
explain the diversity of the earths men, races, and peoples.
Therefore, as Lamarcks teacher and precursor in his theory
of milieu, Buffon is positioned at the convergence of the theorys
two components, the mechanical component and the anthro-
pogeographic one. At this point, we are faced with a problem
of epistemology and historical psychology of knowledge that is
far more involved than the specific example that raised it.
Shouldnt the fact that two or more guiding ideas come together
at a given time to form the same theory be interpreted as a
sign that, as different as they may seem when first used in
the analysis, they have a common origin whose meaning and
very existence is forgotten when one considers the different
pieces separately? This is the problem we will come back to
in the end.
The Newtonian origins of the notion of milieu are enough to
account for the initial mechanical meaning of this notion and
the use that was first made of it. The origin determines the
meaning, and the meaning determines the usage. This is so true
Canguilhem | The Living and I ts Milieu 9
that in 1838, in proposing a general biological theory of the
milieu in the fortieth lesson of his Cours de Philosophie posi-
tive, Auguste Comte believed that he was using milieu as a
neologism and claimed the credit for introducing it as a uni-
versal and abstract explanatory concept in biology. And Comte
says that from this point on he would understand the term to
mean not only the uid in which a body is immersed (which
clearly conrms the mechanical origins of the notion), but the
sum total of outside circumstances necessary to the existence
of each organism.
However, with Comte (who has a perfectly
clear idea of the origins of the notion, as well as the new mean-
ing he wishes to give it in biology) we also observe that its use
will remain dominated by the mechanical origins of the notion,
if not of the term.
In fact, it is quite interesting to note that Auguste Comte was
on the verge of creating a dialectical conception of the rela-
tionship between organism and milieu. I am referring to pas-
sages in which he defines the relationship between the
adapted organism and the favorable milieu as a conict of
forces in which action is constituted by function. He posits
that the ambient system cannot modify the organism without
the latter in turn exercising a corresponding influence. But,
except in the case of the human species, Auguste Comte
believes this action of the organism on the milieu to be negligi-
ble. As for the case of the human species, true to his philo-
sophical conception of history, Comte allows that through the
intermediary of collective action humanity modies its milieu.
However, for the living in general, Comte
refuses to consider this action of the organ-
ism on the milieu seriously, reckoning that it
is simply negligible. This is because he is
looking for a very explicit guarantee of a
dialectical connection, of a reciprocal rela-
tionship between milieu and organism, that
would follow the Newtonian principle of
action and reaction. It is in fact clear, from a
mechanical point of view, that the action of
the living on the milieu is practically negli-
gible. And Comte ends up posing the biolog-
ical problem of the relationship between
organism and milieu as a mathematical one:
In a given milieu, given the organ, nd the
function, and vice versa. The connection of
organism and milieu is therefore that of a
function to a set of variables, an equal rela-
tionship that allows us to determine the
function using the variables, and the vari-
ables separately starting with the function,
10 Grey Room 03
all other things being equal.
The analysis of variables for which the milieu turns out to be
the function is conducted by Comte in lesson 43 of the Cours
de Philosophie positive. These variables are weight, air and
water pressure, movement, heat, electricity, and chemical ele-
ments, all factors capable of being studied experimentally and
measured quantitatively. The quality of an organism nds itself
reduced to a set of quantities, despite the skepticism Comte
professes elsewhere toward the practice of treating biological
problems mathematically, a skepticism that, as we know, comes
to him from Bichat.
In short, even a summary history of the importation of the
term milieu to biology in the rst years of the nineteenth cen-
tury brings out the initial, strictly mechanistic use of the term.
If the hint of an authentically biological acceptation and a more
exible usage appears with Comte, it immediately succumbs to
the prestige of mechanics, an exact science that bases predic-
tions on calculations. The theory of milieu appears clearly to
Comte as a variant of the fundamental project that the Cours de
Philosophie positive seeks to fulll: the world rst, then man;
to go from the world to man. If the idea of the subordination of
the mechanical to the vital is assumed, as Le Systme de Politique
positive and La Synthse subjective later suggest, it is never-
theless formally rejected.
But there is still another lesson to get out of the use of the
term milieu that is, beyond any question, denitively conse-
crated by Comte. The equivalent of what this term designates
would be circumstances in the work of Lamarck. Etienne
Geoffroy Saint-Hilaire, in his report to the Academy of Sciences
in 1831, spoke of surroundings. These terms of circum-
stance and surroundings come from a certain intuition of a
centered formation. In the success of the term milieu, the
metaphor of the line or the indenitely extendable plane, being
both continuous and homogeneous, with no denite shape or
privileged position, wins out over the metaphor of the sphere
or circle, shapes that are still defined qualitatively and, we
might even say, attached to a fixed central reference point.
Circumstances and surroundings still retain a symbolic value,
but milieu abandons any evocation other than a position indef-
initely denied by exteriority. The now refers to the future, the
here refers to its beyond, and so forth always ad innitum. The
milieu is really a pure system of relationships without supports.
From this point we may understand the prestige of the
notion of milieu for analytical scientific thought. The milieu
becomes a universal instrument of the dissolution of individu-
alized organic synthesis in the anonymity of elements and uni-
versal movements. When the French neo-Lamarckians borrow
from Lamarckif not the term in the absolute sense and in the
Canguilhem | The Living and I ts Milieu 11
singular, at least the ideathey keep only the formation by out-
side conditioning and, so to speak, the deformation of the mor-
phological character and functions of the living. It is enough to
recall Costantins experiments on the shapes of sagittate leaves
and Houssays experiments on the shape, ns, and metamerism
of sh. In a little book entitled La Vie des Rivires, Louis Roule
was capable of writing, Fish dont lead their lives themselves,
it is the river that makes them lead it, they are persons without
We have here an example of what a strictly mechanistic
use of the notion of milieu must lead to.
We have returned to
the idea of animal-machines. In the end, Descartes was saying
the same thing when he said of animals, It is nature that acts
in them through the medium of their organs.
| | | | |
Starting in 1859, in other words with the publication of Darwins
Origin of Species, the problem of the relationship between
organism and milieu is dominated by the polemical opposition
between Lamarckians and Darwinians. It seems necessary
to recall the originality of these respective starting points to
understand the meaning and importance of the polemic.
Lamarck wrote in his Philosophie zoologique (1809) that if,
by action of circumstance or action of milieus, we understand a
direct action of the external milieu on the living, we are impos-
ing a meaning that is unwarranted.
It is due to a need, a sub-
jective notion that implies a reference to a positive pole of life
values, that the milieu dominates and commands the evolution
of living things. Changes in circumstance bring about changes
in needs; changes in needs bring about changes in actions. As
long as these actions last, the use or nonuse of certain organs
causes them to strengthen or atrophy, and these morphological
losses and gains acquired by individual habit are preserved by
the mechanism of heredity whenever the new morphological
character is common to the two parents.
According to Lamarck, the situation of the living in the
milieu is a situation that we can call both distressful and dis-
tressed. The life and the milieu that is unaware of it are two
asynchronous series of events. The change of circumstances
comes rst, but it is the living itself that, in the end, initiates the
effort to not be let go by its milieu. Adaptation is a repeated
effort on the part of life to continue to stick to an indifferent
milieu. Adaptation as the result of an effort is therefore neither
harmonious nor providential; it is earned and never guaran-
teed. Lamarckism is not mechanism, and it would be inexact to
call it nalism. In reality, it is a naked vitalism. There is an orig-
inality of life that the milieu cannot render, that it does not
know. The milieu is in this case really external in the proper
12 Grey Room 03
sense of the word. It is foreign; it does nothing for life. This is
truly vitalism because it depends on this dichotomy. Life, said
Bichat, is the collection of functions that resist death. In
Lamarcks conception, life resists only by transforming in order
to outlive itself. To my knowledge, no portrait of Lamarck, no
summary of his doctrine, is better than the one given by Sainte-
Beuve in his novel Volupt.
Here we can see how much dis-
tance lies between Lamarckian vitalism and the mechanicism of
the French neo-Lamarckians. Cope, an American neo-Lamarckian,
was truer to the spirit of the doctrine.
Darwin has a totally different explanation of the environ-
ment of the living, as well as the appearance of new forms. In
the introduction to Origin of Species, he writes, Naturalists
are always referring to external conditions like climate and
food as the only possible cause of variations; they are only right
in a very narrow sense.
It seems that Darwin later regretted
having attributed only a minor role to the direct action of phys-
ical forces on the living. This is manifest in his correspon-
dence. On this point, in the introduction he wrote for selected
texts of Darwin, Marcel Prenant published a certain number of
particularly interesting passages.
Darwin was looking for the
appearance of new forms in the interplay of two mechanisms:
a mechanism of production of differences that is variation, and
a mechanism of reduction and criticism of the differences pro-
duced, that is, the struggle for existence [la concurrence vitale]
and natural selection. The fundamental biological relationship,
in Darwins eyes, is a relationship between living things and
other living things. It trumps the relationship between living
and milieu, conceived of as a collection of physical forces. The
primary milieu an organism lives in is the set of living things
around it that are enemies or allies, prey or predators. Among
the living, relationships of use, destruction, and defense are
established. In this test of strength, accidental variations of
morphology play out as advantages or disadvantages. In fact,
variation, that is to say the appearance of slight morphological
differences by which a descendant does not look exactly like
his ancestors, emerges from a complex process: the use or
nonuse of organs (the Lamarckian factor concerns only adults),
correlations or compensations of growth (for the young), or even
the direct action of the milieu (on the germ).
In this sense we can therefore say that according to Darwin,
unlike Lamarck, the initiative of variation sometimes, but only
sometimes, comes from the milieu. According to whether we
emphasize or play down this action, whether we limit our-
selves to his classic works or on the contrary to the whole of his
thought in the way it is revealed by his correspondence, we get a
slightly different idea of Darwins thought. At any rate, for Darwin,
to live is to submit individual difference to the judgment of all
Canguilhem | The Living and I ts Milieu 13
of the living. This judgment has only two outcomes: either
death or ones recruitment in turn, for a time, to the jury. But as
long as we live, we are always judged and judging. We can see,
as a result, that in the body of work Darwin left us, the thread
that ties the formation of living things to the physicochemical
milieu seems quite tenuous. And the day a new explanation of
the evolution of the species, mutationism, was combined with
an explanation that suddenly saw the appearance of specific
variations as hereditary (an explanation that Darwin was aware
of but that he underestimated) was the day that the milieu
was reduced to the role of eliminating the worst without
being involved in the production of new beings, normalized by
their nonpremeditated adaptation to new conditions of exis-
tence, with monstrosity becoming the rule and uniqueness a
eeting banality.
In the polemic that pitted Lamarckians against Darwinians,
it is useful to note that the arguments and objections came under
two categories and had two sets of implications. Finalism was
denounced and mechanicism celebrated, rst on one side, then
on the other. This is a clear sign that the issue was poorly framed.
Darwin, we can say, uses the language rather than the substance
of finalism (he has been sufficiently reproached for using the
term selection). With Lamarck, there is less finalism than
vitalism. Both of these men were true biologists, who take life
as a piece of data that they attempt to characterize without tak-
ing too much time to come to terms with it analytically. In fact,
these two genuine biologists complement one another. Lamarck
thinks of life in terms of duration, and Darwin more according
to interdependence. One life-form implies a plurality of other
forms with which it is in contact. The synoptic vision that
makes up the core of Darwins genius underscores Lamarcks
weaknesses. Darwin is more closely related to geographers,
and we know what he drew from his voyages and explorations.
The milieu in which Darwin imagined the life of the living is a
biogeographical milieu.
| | | | |
At the beginning of the nineteenth century, two names sum up
the emergence of geography as a science newly cognizant of
both its status and its method: Ritter and Humboldt.
In 1817, Carl Ritter published his Geographie gnrale com-
pare ou Science de la Terre dans ses rapports avec la nature
et lhistoire de lhomme. Starting in 1845, over the course of ten
years Alexander von Humboldt published a book whose spirit
is precisely captured in the title Kosmos. These two authors
brought together the traditions of Greek geography, that is, of
Aristotle and Strabos science of the human ecumene, and the
14 Grey Room 03
science of the coordination of human space in relation to celes-
tial congurations and movements, that is, the mathematical
geography whose founders we consider to be Eratosthenes,
Hipparchus, and Ptolemy.
According to Ritter, human history is unintelligible without
understanding the connection of humanity to the land and to the
whole earth. The terrestrial globe, considered as a whole, is the
stable support for the vicissitudes of history. As a result, terres-
trial space and its conguration are the object not only of geo-
metric or geological knowledge, but also sociology and biology.
Humboldt was a naturalist and voyager who traveled several
times over what one could travel of the world of his time and
who applied a whole system of barometric, thermometric, and
other measurements in his investigations. Humboldt was espe-
cially interested in the distribution of plants according to cli-
mate: he was the founder of botanical geography and zoological
geography. Kosmos is a synthesis of learning that focuses on life
on earth and the relations between life and physical milieu. This
synthesis is not an attempt to be encyclopedic, but is rather a
step toward an intuition of the universe. It begins with a his-
tory of Weltanschauung through a history of the cosmos whose
equivalent could not easily be found in a work of philosophy.
It is a critical commentary that is nothing short of remarkable.
It is essential to note that Ritter and Humboldt applied the
category of totality to their object: the relationship between his-
torical man and milieu. Their object is all of humanity over the
whole world. As a result of their work, the idea of a historical
relationship determined by environment
was consolidated in geography, leading rst
to Ratzel and anthropogeography in Germany,
then to geopolitics, and spreading to history
through Michelet. One has only to recall Le
Tableau de la France.
And finally Taine,
as I have already mentioned, contributes to
the spread of the idea in all elds [milieux],
including the literary. One can summarize
the spirit of this theory of the relationship
between man and his geographic milieu by
saying that doing history consists of reading
a map, if we understand by map the congu-
ration of a set of metric, geodesic, geological,
climatological, and descriptive biogeograph-
ical data.
The approach to problems in anthropol-
ogy and human ethology (an approach that
is more and more deterministic, or more pre-
cisely mechanistic, as we get further from
the spirit of the founders) is coupled with a
Canguilhem | The Living and I ts Milieu 15
parallel, if not exactly contemporaneous, methodology in the area
of animal ethology. A mechanistic explanation of the organisms
movement within the milieu succeeds the mechanistic inter-
pretation of the development of organic forms. Let us simply
recall the work of Jacques Loeb and that of Watson. Generalizing
the conclusions of his studies of phototropism in animals, Loeb
considers all movement of the organism in the milieu as a
movement that is forced upon the organism by the milieu. The
reex, considered the elementary response of a part of the body
to an elementary physical stimulus, is the simple mechanism
whose constitution allows us to explain all behavior of the
living. There is no question that such exorbitant Cartesianism
lies, along with Darwinism, at the origin of the postulates of
behaviorist psychology.
Watson assigned the analytical study of the conditions of
adaptation of the living to the milieu as a program for psychol-
ogy by experimenting with the production of relations of exci-
tation and response (the coupling of stimulus-response). The
determinism of the relationship between excitation and
response is physical. The biology of behavior is reduced to neu-
rology, and the latter is reduced to an energetics. Watsons ideas
led him from a conception in which he simply neglected con-
sciousness because he saw it as unuseful, to a conception in
which he rejected it as outright illusory. The milieu nds itself
invested with all powers with respect to individuals; its power
dominates and even does away with the inuence of heredity
and genetic makeup. Once given a milieu, the organism itself
gives nothing that, in reality, it doesnt receive. The situation of
the living, its being in the world, is thus its condition, or more
precisely its conditioning.
Albert Weiss wished to construct biology as a deductive
physics, and he proposed an electronic theory of behavior. It
was left to psychotechnicians, who extended Taylorist tech-
niques of time and motion studies by means of the analysis of
human reactions, to perfect the work of behavioral psychology
and to ingeniously constitute man as a machine reacting to
other machines, an organism determined by the new milieu
In short, as a result of its origins, the notion of milieu rst
developed and spread in a perfectly predictable manner; and
thus we may say, applying to it the methodological approach
that it implies, that its intellectual power was a function of the
intellectual milieu in which it was formed. The theory of
milieu was the positive and apparently verifiable translation
of Condillacs fable of the statue: To us it is a statue smelling a
rose, to itself it is smell of rose.
Similarly, in the physical
milieu, the living simply is light and heat; it is carbon and oxygen,
calcium, and heaviness. It responds by muscular contractions to
16 Grey Room 03
sensory stimuli, from scratching to tickling, from leaking to burst-
ing. But we may, and we must, ask where the living is to be found?
We can clearly see individuals, but they are objects; we see ges-
tures, but they are displacements; centers, but they are environ-
ments; machine operators, but they are machines. The behavioral
milieu coincides with the geographic milieu, the geographic
milieu with the physical milieu.
| | | | |
It was normal, in the strong sense of that word, for this method-
ological norm to be both pushed to its limits and ultimately
overturned in geography. Geography deals with complexes,
complexes of elements whose actions limit each other recipro-
cally and for which the effects of causes in turn become causes,
modifying the causes that brought them into being. For this rea-
son, trade winds provide a prototypical example of a complex.
Trade winds displace surface seawater warmed through con-
tact with the air, deeper cold waters rise to the surface and cool
the atmosphere, the low temperatures lead to low pressures,
which give rise to winds, and the cycle is closed and begins
again. This is a type of complex that we might also observe in
plant geography. Vegetation is spread out in natural groups
within which different species limit each other reciprocally
and in which, as a result, each one contributes to creating an
equilibrium for the others. The whole set of these plant species
ends up constituting its own milieu. In this way exchanges
between plants and the atmosphere end up creating a sort of
screen of water vapor around the plant kingdom that ends up
limiting the effects of radiation, and the cause leads to the effect
that it in turn attenuates, etc.
The same approach should be applied to animals and to
man, although we nd that human response to the stimulus of
the milieu is varied. Man can nd several solutions to the same
problem posed by the milieu. The milieu proposes without
ever imposing a solution. Of course the possibilities are not
endless within a given state of civilization and culture. But the
fact of seeing something as an obstacle at one time that later can
become a tool is clearly tied to the idea, to the representation,
that man (I am speaking of humanity as a whole, of course)
makes of his own possibilities and needs. In short, it relies on
what he sees as desirable, and that is something that cannot be
separated from the whole of his value system.
In this way, we end up inverting the relationship between
milieu and living thing. At this point, to the extent that he
exists in history, man becomes a creator of the geographical
conguration; he is a geographical factor. We may here simply
recall that the work of Vidal-Lablache, Brunhes, Demangeon,
Canguilhem | The Living and I ts Milieu 17
and Lucien Febvre and his school showed that man has no pure
physical milieu. In a human milieu, man is obviously subject
to a determinism, but it is the determinism of artificial con-
structions. The spirit of invention that brought them into exis-
tence has been alienated from him. In the same line of thinking,
the work of Friedmann shows how, in the new milieu that
machines create for man, the same reversal has already occurred.
Pushed to the extreme limits of its ambition, the psychotech-
nique of engineers that grew out of Taylorist philosophy has
succeeded in locating an irreducible center of resistance, the
presence of mans true originality in the form of his sense of
values. Man, even when subordinated to machines, cannot
conceive of himself as a machine. His productive efficiency
increases the more he is aware of his centrality in relation to the
mechanisms that serve him.
Well before this, the same reversal of the relationship
between organism and milieu had occurred in matters of ani-
mal psychology and behavioral studies. Loeb led to Jennings,
and Watson led to Kantor and Tolmann.
At this point, the inuence of pragmatism is clear and well
established. If, in one sense, pragmatism served as an interme-
diary between Darwinism and behaviorism by extending the
idea of adaptation to a general theory of knowledge and, in
another sense, by putting the accent on the role of values
in relation to the interests of action, Dewey was to lead behav-
iorists to regard the connection between organic movements
and the organism itself as essential. The organism was consid-
ered as a being on which not everything could be
imposed, because its existence as an organism con-
sists in presenting itself to things, according to certain
orientations that are specific to it. First explored by
Kantor, Tolmanns teleological behaviorism consists of
researching and recognizing the meaning and inten-
tion of animal movement. It seems essential to the
movement of response to persist in a set of phases that
can be mistakes or unfullled acts, up until the moment
when the reaction puts an end to the stimulus and
reestablishes a state of rest or leads to a new series of
actions that is totally different from the ones that were
closed unto themselves.
Before him Jennings had shown, in his theory of
trial and error, contra Loeb, that the animal does not
react by the sum of molecular reactions to a stimulus
that can be broken down into units of stimulation, but
rather that it reacts as a whole to total objects and that
these reactions regulate the needs that command them.
Naturally, one must recognize here the considerable
contribution of Gestalttheorie, especially the distinction
18 Grey Room 03
between behavioral milieu and geographical milieu that we
owe to Koffka.
Finally, the organism-milieu relationship finds itself
reversed in von Uexklls studies of animal psychology and
Goldsteins studies of human pathology. Each of these illustrate
the reversal with a clarity that comes from a completely philo-
sophical approach to the problem. Uexkll and Goldstein agree
on this fundamental point: that to study a living thing under
experimentally constructed conditions is to create a milieu for
it, to impose a milieu upon it. In fact, it is a fundamental char-
acteristic of the living thing that it makes its own milieu; it
builds one for itself. Of course, even from a materialist point of
view we can speak of the interaction between the living and
the milieu, between the physicochemical system interspersed
within a larger whole and its environment. But talk of interac-
tion is not enough to offset the difference that exists between a
relationship of a physical type and one of a biological type.
From the biological point of view, one must understand that
between organism and environment there is the same relation-
ship that exists between the parts and the whole within the
organism itself. The individuality of the living does not come
to an end at its ectodermal boundaries, no more than it begins
at the level of the cell. The biological relationship between the
being and its milieu is a functional one, and as a result it
changes as the variables successively exchange roles. The cell
is a milieu for intracellular elements; it lives in an interior
milieu that is either on the scale of the organ or the organism,
which organism itself lives in a milieu that is for it, in a sense,
what the organism is for its component parts. We can therefore
move toward using a biological reasoning to evaluate biological
problems. A reading of Uexkll and Goldstein can contribute
a great deal to mapping out this reasoning.
Let us take the terms Umwelt, Umgebung, and Welt. Uexkll
distinguishes between them with great care. Umwelt designates
the behavioral milieu that is proper to a given organism;
Umgebung is the simple geographical environment; and Welt
is the scientic universe. For the living, the specic behavioral
milieu (Umwelt) is a set of stimuli that have the value and sig-
nicance of signals. To act on a living thing, it is not enough
that physical stimuli be produced; they must also be noticed.
As a result, to the extent that a stimulus acts on the living, it
presupposes an orientation of its interest. The stimulus does
not proceed from the object, but from this interest. It is neces-
sary, in other words, for the stimulus to be effective, that it be
anticipated by the subjects attitude. If the living does not go
looking for something, it gets nothing. A living thing is not a
machine that responds by movement to stimuli, it is a machin-
ist who responds to signals by operations. Naturally, this does
Canguilhem | The Living and I ts Milieu 19
not mean that one should call into question the fact that there
are reexes whose mechanism is physicochemical. For the biol-
ogist, the problem is elsewhere. The question is rather to be
found in the fact that out of the exuberance of the physical
milieu, as a producer of stimuli whose number is theoretically
unlimited, the animal retains only a few signals (Merkmale). Its
biorhythm orders the temporality of this Umwelt, just as it
orders its space. Along with Buffon, Lamarck said: time and
favorable circumstances constitute the living little by little.
Uexkll reverses the relationship and says: time and favorable
circumstances exist only in relation to a specic living thing.
Umwelt is therefore a voluntary sample drawn from the
Umgebung, the geographical environment. But the environ-
ment is precisely nothing other than mans Umwelt, that is, the
usual world of his practical perspective and experience. Like
this Umgebung, this geographical environment that is external
to the animal is, in a sense, centered, ordered, and oriented by
a human subject (that is to say a creator of techniques and val-
ues). Similarly, the animals Umwelt is nothing other than a
milieu centered around the subject of life values that makes up
the essential part of what constitutes the living. At the root of
this organization of the Umwelt we must conceive of a subjec-
tivity that is analogous to the one we are bound to think of as
being at the root of the human Umwelt. One of the most com-
pelling examples cited by Uexkll is the Umwelt of the tick.
Ticks grow by imbibing the warm blood of mammals. After
coupling, the adult female climbs to the end of a tree branch
and waits. It can wait eighteen years. At the Institute of Zoology
in Rostock, ticks have stayed alive, closed up, in a state of ina-
nition, for eighteen years. When a mammal passes under the
tree, under the ticks hunting and trapping post, it lets itself fall.
What guides it is the odor of rancid butter that emanates from
the animals cutaneous glands. This is the only stimulus that
can set off the falling motion. This is the rst step. Once the tick
has fallen on the animal, it attaches itself to it. If the odor of ran-
cid butter has been produced articially, on a table, for exam-
ple, the tick does not attach itself, but climbs back up to its
observation post. The only reason it attaches to the animal is its
blood temperature. It attaches to the animal because of its sense
of heat; and guided by its sense of touch, it looks preferably for
areas of the skin that are hairless, it digs in just beyond the
head, and sucks the blood. It is only at the moment when the
mammals blood enters its stomach that the ticks eggs (encap-
sulated since the moment of coupling and able to remain
encapsulated for eighteen years) open up, mature, and grow.
The tick can live eighteen years to complete its reproductive
function in a few hours. It should be noted that, for a consider-
able amount of time, the tick can remain totally indifferent,
20 Grey Room 03
insensitive to all stimuli coming from a milieu like the forest,
and that the only stimulus capable of setting off its movement,
to the exclusion of all others, is the odor of rancid butter.
A confrontation with the work of Goldstein was inevitable,
since his theory is based on a critique of the mechanical theory
of reexes. The reex is not an isolated or gratuitous reaction.
The reaction is always a function of the opening of the senses
to stimuli and its orientation relative to them. This orientation
depends on the meaning of a situation as it is perceived in its
entirety. Separate stimuli may have meaning in the social sci-
ences, but they mean nothing when it comes to the senses of a
living thing. An animal in an experimental setting is in an
abnormal situation that is imposed upon it; it is neither neces-
sary nor of its own choosing. An organism is therefore never
equal to the theoretical sum of its possibilities. We cannot
understand its action without thinking of it in terms of a privi-
leged form of behavior. Privileged does not mean objectively
simpler in this case; it is rather the opposite. The animal nds it
easier to do what it favors: it follows its own norms of living.
The relationship established between the living and the
milieu is like a debate (Auseinandersetzung) in which the liv-
ing brings its own norms of appreciating the situation, where it
is in command of the milieu and accommodates itself to it. This
relationship does not consist primarily, as one might think, of
a struggle or a confrontation. Those are things that characterize
the pathological state. A life that afrms itself in opposition is
already a life threatened. Movements involving strength, as for
example extensive muscular reactions, translate the domina-
tion of the exterior onto the organism.
A healthy life, a life
that is condent in its existence and in its values, is a life that
extends itself yet that is also almost gentle in its exibility. The
situation of the living demanded by the milieu from the outside
is what Goldstein holds up as the prototype of a catastrophic
situation. This is the situation of the living in the laboratory.
The relations between the living and the milieu as they are
studied experimentally and objectively are of all possible rela-
tions those that have the least biological signicance: they are
pathological relations. Goldstein says that the meaning of an
organism is its being; we may say that the being of the organ-
ism is its meaning. Of course, the physicochemical analysis of
the living can and should be undertaken. It has a theoretical
and practical interest. But this constitutes a chapter of physics.
Everything remains to be done in biology. Biology must there-
fore rst consider the living as a meaningful being, and its indi-
viduality not as an object, but as a term within the order of
values. To live is to spread out; it is to organize a milieu starting
from a central reference point that cannot itself be referred to
without losing its original meaning.
Canguilhem | The Living and I ts Milieu 21
While the reversal of the organism-milieu relationship was
being completed in animal ethology and in behavioral studies,
a revolution was occurring in the way that morphological char-
acteristics were being used to explain the autonomy of the liv-
ing relative to the milieu. I am alluding here to the now very
well known work of Bateson, Cunot, Thomas Morgan, H.
Mller, and their collaborators, who took up and extended
Gregor Mendels research on hybridization and heredity. In the
process of creating the science of genetics, these thinkers ended
up claiming that in a given milieu the acquisition of the form,
and therefore the function, of the living depends on its partic-
ular hereditary potential and that the action of the milieu on
phenotype leaves genotype unchanged. The genetic explana-
tion of heredity and evolution (i.e., the theory of mutation) con-
verged with Weissmans theory. The precocious isolation of a
germinating plasma in the course of ontogenesis would nullify
the influence of somatic modifications determined by the
milieu on the evolution of the species. In his book La Vie cra-
trice des Formes, Albert Brachet wrote that the milieu is not,
properly speaking, an agent of formation, but in fact of realiza-
tion, by invoking the multiformity of sea creatures within an
identical milieu in support of his argument.
And Caullery
concluded his study of The Present State of the Problem of
by recognizing that evolution depends much more
on the intrinsic properties of organisms than on the surround-
ing milieu.
Yet we know that the idea of the total autonomy of hereditary
genetic assortment did not go without criticism. At rst critics
emphasized the fact that nucleoplasmatic disharmony tends to
limit the hereditary omnipotence of genes. In sexual reproduc-
tion, if it is true that the two parents each provide half of the
genes, the mother provides cytoplasm for the egg. Given that
the mixed offspring of two different species are not the same,
depending on whether one or the other species is represented
by the father or the mother, we are led to suppose that the power
of genes differs as a function of the cytoplasmic milieu. In addi-
tion to this, H. Mllers experiments (1927) provoking mutations
in Drosophila by the action of a milieu of penetrating radiation
(X rays) seemed to shed some light on the external conditioning
of an organic phenomenon, perhaps too easily underscoring the
distinction between organism and environment.
Finally, Lamarckism has become topical once again thanks
to the ideological, as much as scientific, polemics around the
indignant repudiation of genetic pseudo-science by the
Russian biologists that Lysenko had brought back to the
healthy method of Mitchourine (18551935). Experiments on
the vernalization of cultivated plants like wheat and rye led
Lysenko to claim that hereditary modications can be obtained
22 Grey Room 03
Pennate diatoms.
and consolidated by variations in feeding, upkeep, and climatic
conditions, leading to the dislocation or rupture of the heredi-
tary constitution of the organism that geneticists had falsely
imagined to be stable. To the extent that we can summarize
complex experimental ndings, it should be said that accord-
ing to Lysenko, heredity is dependent on metabolism, and the
latter is dependent on the conditions of existence. Heredity is
to be seen here as the assimilation of outside conditions by the
living over successive generations. Remarks of an ideological
nature concerning these facts and this theory actually help clar-
ify its meaning, regardless of their authors inability to accept,
let alone tolerate, the counterexperiments and criticisms that
are the norm in matters of scientific discussion; all of which
things lie, of course, outside of my realm of competence.
seems that the technical, that is to say agronomic, aspect of the
problem is crucial. In justifying the spontaneous character of
mutations, Mendelian theories of heredity tend to moderate
human, and specically Soviet, ambitions to completely dom-
inate nature and the possibility of intentionally altering living
species. Finally and above all, the recognition of the determin-
ing inuence of the milieu has a political and social impact in
that it authorizes mans unlimited action upon himself through
the medium of the milieu. It justies hope in an experimental
renewal of human nature. In this way, it appears, at rst sight,
to be progressive. Theory and praxis are indissociable, as is
required by Marxist-Leninist dialectics. As a result, we can see
how genetics could be charged with all of the sins of racism and
slavery and how Mendel was presented as
the leading spokesman for a retrograde, cap-
italist, and even idealist biology.
It is clear that the return to legitimacy of
theories of the heredity of acquired charac-
teristics does not in itself authorize us to
unreservedly qualify the recent Soviet bio-
logical theories as Lamarckian. This is
because the essence of Lamarcks ideas, as
we have seen, consists in attributing the
organisms adaptation to the milieu to its
own initiative, needs, and continuous reac-
tions. The milieu provokes the organism
to orient its own development. Biological
response far outweighs physical stimulation.
By rooting adaptive phenomena in necessity,
which means both pain and impatience,
Lamarck was focusing on the point where
life coincides with its own meaning, where
through its sensory experience, the living sit-
uates itself absolutely, for better or worse, in
Canguilhem | The Living and I ts Milieu 23
existence: the indivisible totality of organism and milieu.
With Lamarck, as is the case among the first theorists of
milieu, the notions of circumstances and surroundings
have an altogether different meaning than they have in normal
language. These words genuinely evoke a spherical, centered
disposition. The terms influences and influential circum-
stances, used by Lamarck, take their meaning from astrologi-
cal concepts. When Buffon, in La Dgnration des Animaux,
speaks of the tint of the sky that takes man so long to per-
ceive, he is using, no doubt unconsciously, a term borrowed
from Paracelsus. Even the notion of climate in the eigh-
and early nineteenth centuries is a unied notion com-
mon to geography, astronomy, and astrology. Climate is the
change in appearance of the sky, degree by degree, from equator
to pole; it is also the inuence exercised by the sky on the earth.
I have already indicated that the biological notion of milieu
at rst brought together an anthropogeographic component and
a mechanical one. The anthropogeographic component could
even be considered to make up the whole idea, since it
included in itself the astronomical component, the one Newton
had converted to a theory of celestial mechanics. For in the
beginning geography was for the Greeks the projection of the
sky onto the earth, the coming together of earth and sky, a cor-
respondence that went in two directions at the same time: a
topographical correspondence (geometry and cosmography)
and a hierarchical correspondence (physics and astrology). The
mapping of parts of the earth and the subordination of a
mapped area to the sky were understood in the astrobiological
intuition of the cosmos. Greek geography had its own philoso-
phy, that of the Stoics.
The intellectual relations between
Posidonius on one hand, and Hipparchus, Strabo, and Ptolemy
on the other, are undeniable. It is the theory of universal sym-
pathy, a vitalist intuition of universal determinism, that gives
its meaning to the geographical theory of the milieu. This theory
supposes the assimilation of the totality of things to an organism,
and the representation of this totality in the form of a sphere,
centered on the situation of a privileged living thing: man. This
biocentric conception of the cosmos carried over from the Middle
Ages to blossom in the Renaissance.
We know what happened to the notion of cosmos with the
appearance of Copernicus, Kepler, and Galileo, as well as how
dramatic the conflict was between the organic conception of
the world and the conception of a universe that was decentered
relative to the center privileged in the classical world, the earth
of the living and of man. With Galileo and also Descartes it
became necessary to choose between two theories of milieu,
that is, in the nal analysis, theories of space: a centered space,
defined as being where the mi-lieu is a center; a decentered
24 Grey Room 03
space, defined as being where the mi-lieu is an intermediary
field. Pascals famous text, Disproportion de lHomme, shows
the ambiguity of the term well in a spirit that cannot or will not
choose between its existential security and the demands of sci-
entic knowledge.
Pascal knew that the cosmos had shattered
into pieces, but the eternal silence of innite spaces frightened
him. Man was no longer at the center [au milieu] of the world,
but he is a milieu (a milieu between two infinites, a milieu
between nothing and everything, a milieu between two
extremes); the milieu is the state in which nature placed us; we
are floating on a vast milieu; man is in proportion with the
parts of the world; he relates to all that he knows: He needs a
place to contain him, time in which to endure, movement to
live, elements to make him up, heat and food to nourish him,
air to breath . . . and in the end, everything is his ally. We may
observe that three meanings of the word milieu come into
play here: a median situation, a uid of suspension, a life envi-
ronment. It was in developing this last meaning that Pascal
revealed his organic conception of the world, a return to sto-
icism that went both beyond and against Descartes: All things
being caused and causal, helped and helping, mediated and
immediate, and all intertwined by a natural and insentient con-
nection that links the most distant and different among them, I
hold that it is impossible to know the parts without knowing
the whole, any more than we can know the whole without par-
ticularly knowing the parts. And when he denes the universe
as an infinite sphere in which the center is everywhere, the
circumference nowhere, Pascal is paradoxically using an
image borrowed from the theosophic tradition to try to recon-
cile the new scientic conception that sees the universe as an
innite and undifferentiated milieu and the ancient cosmolog-
ical vision that sees the world as a finite whole connecting to
its center. It has been established that the image used here by
Pascal is a permanent myth of mystic thought of neo-Platonic
origin in which an intuition of the spherical world centered in
and by the living and the already heliocentric cosmology of the
Pythagoreans are reconciled.
Before Newton, the symbolic representation of the potential
ubiquitousness of a spreading action starting from a central
point described in the neo-Platonic cosmology of Jacob Boehme
and Henry More, the Cambridge Platonist, was universally
recognized. Newtonian space and ether, the rst as a means for
the omnipresence of God, the second as a medium and vehicle
of forces, both retain, as we know, an absolute character that
eighteenth- and nineteenth-century scholars failed to remark
upon. Newtonian science, which was to anchor so many
empiricist and relativist professions of faith, is itself founded
on metaphysics. Empiricism masks its theological foundations.
Canguilhem | The Living and I ts Milieu 25
And in this way the natural philosophy or the positivist and
mechanistic conception of milieu has as its source, nds itself
anchored by in fact, the mystical intuition of a sphere of energy
in which the central action is uniformly present and efficient
at every point.
| | | | |
If today it seems normal to anyone trained in the mathematical
and physical disciplines that the objectivity of knowledge
requires a decentering of perspective, the moment also nally
seems to have arrived where, from the perspective of biology,
according to the formulation of J. S. Haldane in The Philosophy
of a Biologist, it is physics that is not an exact science. Indeed,
as Claparde wrote, What distinguishes the animal is the fact
that he is a center relative to surrounding forces that are no
longer, relative to it, anything but signals or stimuli; a center,
in other words an internally regulated system in which reac-
tions are controlled by an internal cause: immediate neces-
In this sense, the milieu on which the organism depends
is structured and organized by the organism itself. What the
milieu gives to the living is a function of its demand. This is
why within what appears to man to be a unique milieu, several
living things draw their own specic and singular milieu. For
that matter, as a living thing, man does not escape the general
law of the living. The milieu that is proper to man is the world
of his perception, that is to say the eld of his practical experi-
ence in which his actions, oriented and
regulated by values that are immanent
to his tendencies, carve out certain
objects, situate them relative to each
other and all of them in relation to him-
self. This occurs in such a way that the
environment he is supposed to be react-
ing to nds itself originally centered in
and by him.
But man the scholar constructs a uni-
verse of phenomena and laws that he
holds up as absolute. The essential
function of science is to devalue the
qualities of objects that make up the
milieu proper, by offering itself as a
general theory of the real, that is to say
nonhuman, milieu. Sensory data are dis-
qualied, quantied, and identied. That
which is imperceptible is first placed
under suspicion, then exposed and
avowed. Measurements are substituted
26 Grey Room 03
for appreciations, laws for habits, causality for hierarchy, and
the objective for the subjective.
In fact, this universe of man the scholar, of which Einsteins
physics is the ultimate representation (a universe in which fun-
damental equations of intelligibility are the same regardless of
the system of reference) because it maintains a direct, if negat-
ing and reductive, relationship with the living mans proper
milieu, endows this milieu with a sort of privilege over the
milieus that are proper to other living things. Living man takes
from his relationship with man the scholar, in whose work
ordinary perceptive experience finds itself contradicted and
corrected, a sort of unconscious fatuousness that leads him to
prefer his own milieu to that of other living things as having not
only a different value, but a higher degree of reality. In fact, as a
proper milieu of behavior and life, the milieu of mans sensory
and technical values does not in itself have more reality than
the milieu proper to a wood louse or a grey mouse. The quali-
cation of real can only be applied rigorously to an absolute
universe, to the universal milieu made up of elements and
movements authenticated by science, in which this recognition
is as such necessarily accompanied by the disqualification of
all subjective understandings of milieu as illusions or errors
of life, including those of man.
The pretension of science to dissolve these centers of orga-
nization, adaptation, and invention that are living things into
the anonymity of the mechanical, physical, and chemical envi-
ronment must be complete, that is to say that it must include
the living human himself. And as is well known, this project
has not seemed too audacious to many thinkers. But we must
then ask ourselves from a philosophical point of view if the ori-
gins of science do not better reveal its meaning than the pre-
tensions of a few scholars. For the birth, development, and
progress of science must be seen as a remarkably audacious
enterprise if we are rightfully to deny the innate genius of
humanity, from the point of view of scientism and even mate-
rialism. If we do not, it would be necessary to admit the absurd
proposition according to which reality contains the science of
reality within itself a priori. And we would then have to ask
what need that has its origins in reality is truly being served by
the ambition to scientically determine that same reality.
But if science is the work of a humanity that is rooted in life
before being enlightened by knowledge, if it is a fact in the
world while also being a vision of the world, then it perpetu-
ates a permanent and necessary relationship with perception.
And therefore mans proper milieu is not situated in the uni-
versal milieu like a thing contained within its container. A cen-
ter does not dissolve into its environment. A living thing does
not reduce itself to an intersection of influences. These ideas
Canguilhem | The Living and I ts Milieu 27
Amoeba proteus.
point to the inadequacy of any biology that would eliminate
any consideration of meaning from its domain out of an utter
submissiveness to the spirit of the physicochemical sciences.
A meaning, from the biological and psychological point of
view, is an appreciation of values in relation to a need. And a
need is, for whoever feels it and lives it, an irreducible system
of reference, and for that reason it is absolute.
28 Grey Room 03
Le Vivant et son milieu was originally presented as a lecture at the Collge
philosophique in Paris in 194647 and was subsequently published in La
Conaissance de la vie in 1952. It is translated and published here with per-
mission from Librairie Philosophique J. Vrin, Paris.
1. I have chosen to translate the authors le vivant as the living, despite
its apparent awkwardness. The French original is similar in this regard, and
other formulations such as living thing, life-form, and organism place
too much emphasis on the denite boundary between these entities and their
surrounding environment, a distinction that the author clearly wishes to
interrogate. I have also kept the term milieu, though its English usage is more
limited than in French. I have used brackets in the text to note those places
where the French use of milieu required a different translation. Along with
the editors, I would like to acknowledge the numerous invaluable suggestions
received from Warwick Anderson, in particular with regard to scientic ter-
minology. Trans.
2. On all these points, see Lon Bloch, Les Origines de la Thorie de lether
et la physique de Newton (1908).
3. Lon Brunschvicg, Les Etapes de la philosophie mathmatique (Paris:
Alcan, 1912), 508.
4. See Georges Canguilhem, La Thorie Cellulaire, in La Connaissance
de la Vie (Paris: J. Vrin, 1992), 54.
5. See Esprit des Lois, books XIVXIX, on the relationship between laws
and climate.
6. The chapter on the degeneration of animals in the Histoire naturelle
des animaux (Paris, 17861791) examines the effects of habitat and food on
the animal organism.
7. I have translated these and other quotations myself, unless otherwise
indicated. The translation of Comte by Harriet Martineau is extremely loose:
see The Positivist Philosophy of Auguste Comte, vol. 2 (New York: D. Appleton,
2 vols., 1853), 364. Trans.
8. Tolmans behavioral psychology also conceives of the relationship
between organism and milieu in the form of the relation of a function to a
variable. Compare Andr Tilquin, Le Behaviorisme (Paris: J. Vrin, 1944), 439.
9. Louis Roule, La Vie des Rivires (Paris: Stock, 1930), 61.
10. A striking summary of this thesis can be found in Houssays Force et
Cause (Paris: Flammarion, 1920), in which the author describes certain types
of units that we call living things, that we set apart as if they had an inde-
pendent and separate existence, when in fact they have no isolated reality and
they cannot be, but for an absolute and permanent linkage with the surround-
ing milieu in which they are but a simple local and momentary concentration.
11. These comments especially concern animals. Lamarck is more reserved
on the subject of plants.
12. Several times each month [dcade] I frequented M. de Lamarcks
Natural History course at the Jardin des Plantes. . . . At that time, M. de
Lamarck was perhaps the last representative of that great school of physicists
and general observers who had reigned from Thales and Democritus to
Buffon. . . . He presented his ideas quite nakedly, with great simplicity and
much sadness. He constructed a world with the fewest possible elements, the
lowest number of crises and the greatest possible duration. A long, blind
patience, that was his idea of the genius of the Universe. . . . In the same way,
in the organic order of things, once he recognized this mysterious power of
life, as small and elementary as it is, Lamarck imagined it developing, building
itself up over time, little by little; deaf necessity, habit alone, gave birth to
Canguilhem | The Living and I ts Milieu 29
organs within a diverse range of milieux, in opposition to the relentless
power of nature that set out to destroy them; for M. de Lamarck distinguished
between nature and life. In his eyes, nature was stone and ash, a granite tomb,
death. Life came into play only as a strange and singularly productive acci-
dent, a prolonged struggle with here or there more or less balance or success,
but always defeated in the end; cold motionlessness reigned afterwards as
before. Sainte-Beuve, Volupt.
13. Here I have translated Canguilhems version of Darwin directly. The
original passage is as follows: Naturalists continually refer to external con-
ditions, such as climate, food, &c., as the only possible cause of variation. In
one limited sense, as we shall hereafter see, this may be true. Charles Darwin,
On the Origin of Species: A Facsimile of the First Edition (Cambridge: Harvard
University Press, 1964), 3. Trans.
14. Marcel Prenant, Darwin (Paris: Editions Sociales Internationales,
1938), 14549.
15. See Lucien Febvres La Terre et lEvolution humaine for a historical
description of the evolution of the idea and a critique of its exaggerations.
16. Andr Tilquin, Le Behaviorisme (Paris: J. Vrin, 1942), 3435. It is of
course from this well-documented work that I have borrowed much of the
information that follows.
17. Condillac, Treatise on the Sensations, trans. Geraldine Carr (London:
Favil Press, 1930), 3.
18. Compare Henri Baulig, La Gographie est-elle une science? Annales
de Gographie 57 (JanuaryMarch 1948); and Causalit et Finalit en
Gomorphologie, Geograska Annaler (winter 1949): 12.
19. A fascinating consideration of this inversion of perspective in human
geography is found in an article by L. Poirier, LEvolution de la Gographie
humaine, which appeared in Critique 89 (JanuaryFebruary 1947).
20. On this point, compare Paul Gillaume, Psychologie de la Forme (Paris:
Flammarion, 1937), and Maurice Merleau-Ponty, Structure du Comportement
(Paris: Presses Universitaires de France, 1942), translated by Alden L. Fisher
as Structure of Behavior (Boston: Beacon Press, 1963).
21. Jakob von Uexkll, Umwelt und Innenwelt der Tiere, 2d ed. (Berlin,
1921); and Theoretische Biologie, 2d ed. (Berlin, 1928); von Uexkll and
Georg Kriszat, Streifz ge durch die Umwelten von Tieren und Menschen
(Berlin, 1934). Goldstein, however, accepts von Uexklls views only with
important reservations. By not differentiating the living from its environ-
ment, any research on their relationship becomes, in a sense, impossible. In
this perspective, determinism disappears in favor of reciprocal penetration,
and taking into consideration the whole effectively stifles knowledge. For
knowledge to remain possible, it is necessary that a nonconventional center
from which a range of relations can emerge appears within this organism-
environment totality. Compare La Structure de lOrganisme, 7576, a critique
of any exclusively environmental theory.
22. According to von Uexkll, the example of the tick is taken up by
Louis Bounoure in his book LAutonomie de lEtre vivant (Paris: Presses
Universitaires de France, 1949), 143.
23. For a discussion of this argument of Goldsteins, compare the conclusion
of Franois Dagognets Philosophie biologique (Paris: Presses Universitaires
de France, 1955).
24. Albert Brachet, La Vie cratrice des Formes (Paris: Alcan, 1927), 171.
25. Maurice Caullery, The Present State of the Problem of Evolut ion
(Washington, 1917).
26. One can nd an anticipation of these ideas in Nietzsches Will to Power.
30 Grey Room 03
To be honest, the criticism that Nietzsche addresses to Darwin would be more
appropriately applied to the Neo-Lamarckians.
27. For an examination of the subject, see Une Discussion scientique en
U.R.S.S., Europe 3334 (1948); and also Claude-Charles Mathon, Quelques
Aspects du Mitchourinisme, etc., in Revue gnrale des Sciences pures et
appliques 34 (1951). On the ideological dimensions of the controversy,
compare Julian Huxley. Jean Rostand has written a good historical and
critical study on the question, LOffensive des Mitchouriniens contre la
Gntique mendelienne, in Les Grands Courants de la Biologie (Paris:
Gallimard, 1951), which is followed by a bibliography. Finally, see the work
of Hovasse, Adaptation et Evolution (Hermann, 1951).
28. See the article on climate in the Encyclopdie.
29. See the excellent abridged history of Greek geography in Theodor
Breiters introduction to volume 2 (commentaries) of the Astronomica by
Manilus (Leipzig, 1908).
30. Pascals Penses, trans. Martin Turnell (New York: Harper and Brothers,
1962), 21520.
31. Dietrich Mahnke, Unendliche Sphre und Allmittelpunkt (Niemeyer:
Halle, 1937); the author devotes several very interesting pages to the usage
and meaning of the expression in Leibniz and Pascal. According to Havet,
Pascal supposedly borrowed the expression from Mademoiselle de Gournay
(see the 1595 preface to Montaignes Essays) or from Rabelais (Tiers livre,
chapter 13).
32. Compare Alexandre Koyr, La Philosophie de Jacob Boehme, 378379,
504; and The Signicance of the Newtonian Synthesis, Archives interna-
tionales dHistoire de Sciences 11 (1950).
33. Preface to F. J. J. Buytendijks Psychologie des Animaux (Paris: Payot,
Canguilhem | The Living and I ts Milieu 31