“In recognizing the duty to no longer think what I thought,
I aver in the same moment the power to think otherwise than I should.”
—Georges Canguilhem, “L’Erreur”
In his book, Georges Canguilhem, Dominique Lecourt takes issue with Michel Foucault’s account of Canguilhem’s thought and importance for intellectuals of his generation in France. There is no doubt that Foucault’s exuberant praise of his bon maître is extreme in its portrayal of Canguilhem as a central figure in contemporary French philosophy, indeed, in the history of modern philosophy. While it is likely that, as Lecourt suggests, this sort of “hyperbolic praise” would have been disconcerting to Canguilhem, the substantial point of contention concerns the question of error and its epistemological valence. Lecourt disagrees with Foucault’s claim that Canguilhem saw himself as a “philosopher of error,” who produced a profound transformation in the rationalist French tradition to which he belonged through this very focus on error.Footnote 1 Foucault suggests that this transformation impacted the theory of the subject itself, making Canguilhem’s work attractive to many in the 1960s who wanted to question the primacy of the Cartesian cogito. As Foucault has it, the transformation occurs because, for Canguilhem, “rather than knowledge opening itself to the truth of the world, it roots itself in the ‘errors’ of life.”Footnote 2 While Lecourt agrees that knowledge is rooted in the errors of life for Canguilhem, he does not think this excludes openness to the truth of the world. Yet Foucault’s suggestion is not that knowledge does not aim to grasp the truth of the world, but rather that the impetus out of which the will to know emerges involves the errors of life. Lecourt himself brings attention to Canguilhem because he is more than a philosopher of error, and because his work can still help guard against the “stifling moralization” performed in the name of ethics today.Footnote 3 There can be no doubt that, for Canguilhem, philosophy was a moral practice attuned to the dangers of moralizing. Foucault was not, however, incorrect in suggesting that the philosophical problem of error underpins his work from the very beginning, including his most famous text on The Normal and the Pathological, first published in 1943 as an Essay on Some Problems Concerning the Normal and the Pathological [Essay]. Yet Foucault’s argument on behalf of this claim is awkward, though, for most of the allusions in his discussion of error are to Canguilhem’s “La nouvelle connaissance de la vie: Le concept et la vie,” first published in 1966. Guillaume Le Blanc’s book, Canguilhem et la vie humaine gives a similar treatment, closing with the question of error, but primarily through an examination of these later texts.Footnote 4 A more complete discussion of the claim that he was a philosopher of error requires, therefore, an inquiry into the early work and the place of error within it. This is all the more important since the Essay contains no extended discussion of error, indeed the word itself is used only a few times throughout. As it turns out, Canguilhem also gave a course on “Error” in 1943. An examination of this and earlier work from the 1930s not only shows that Canguilhem was animated by the philosophical problem of error from early on, but also illuminates the very meaning of this problem.
1. Error from 1934-1939
1.1. Notes from the Lycée Valenciennes
In his 1934 course at the Lycée de Valenciennes, Canguilhem already poses the questions that will occupy him throughout his work: “Are error and fiction just the negation of truth (discovered by logic and science)? Or, is the irrational solely a negative value? Is there not a positive value to it?”Footnote 5 In this course, he argues that philosophy’s task is to produce a theory of values by investigating the possible positive value of those experiences that are usually ascribed a negative value, such as error, fiction, and irrationality.Footnote 6 Concerning belief and knowledge, he argues that belief, whether credulity or faith, is always a pre-critical affirmation; “thought is confident, believing, before it is disbelieving.”Footnote 7 His notes indicate that:
Faith is always a presumption, it is thought posing in advance the value of its exercise. It is the mode of knowing values. Every judgment supposes a faith, even science (truth). Values are not demonstrated, one justifies an activity through them. It is this that is unshakeable in religious sentiment. Faith is knowledge of liberty.
Whereas credulity is belief that does not judge, but simply accepts, the knowledge of values through faith lays the bases for judgments that justify an activity. If scientific judgment is grounded on a faith in truth, faith in general is knowledge of freely chosen values. And this means that the pursuit of truth is a key mode of human liberty, not an activity that denies this liberty. Yet the pursuit of truth is only realized by the adoption of a stringent discipline for judging and doing. ‘Religious sentiment’ is also understood as the recognition of values that justify a particular way of thinking and a way of living. Thus he defines faith, generally, as the source of a discipline demanding certain actions, certain judgments, certain ways of being. But as belief is pre-critical affirmation, values arise spontaneously in the course of history and become confused in ordinary experience, giving rise to conflicts and obstacles to human freedom, thereby necessitating a critical philosophical practice that judges their authenticity to human freedom.Footnote 8 In this, philosophy requires indocility before and resistance to practices that confuse authentic values. Yet this does not mean that negative values, such as error, fiction, and the irrational more generally, are disqualified as inauthentic.
1.2. On Descartes and Technique
Canguilhem begins to sketch out the positive value of error in 1937 through a close reading of René Descartes’ philosophical writings, which, he finds, aimed to produce knowledge of necessity, not so as to better establish firm limits for human beings, but so as to give us power over nature.Footnote 9 Read in conjunction with his letters, these works show that, despite the lack of any systematic treatment of questions concerning technique, Descartes believed that human technical powers could be greatly expanded, perhaps even leading to the vanquishing of death itself.Footnote 10 Such an expansion, however, could only be undertaken through the guidance of a unifying science, since technique and art are blind without knowledge. Science allows for the improvement of technical skill because it eliminates the confused, qualitative distinctions of ordinary experience by developing a knowledge of the necessary and uniform order of nature.Footnote 11 Such a devaluation of technicians and other artisans who work without such knowledge has, however, its limits, for Descartes recognizes that technical accomplishments must still be sought in the world of ordinary experience. Even if we understand the natural world and grasp it in its quantifiable homogeneity, this does not immediately give us the power to alter this world for our own benefit. Canguilhem writes, “if matter is homogeneous and anonymous for science, the matter that the technician proposes to ‘give over to our usage’ is particular and diverse. This is why Cartesian science avows the necessity of groping about experimentally [tâtonnement expérimental ].”Footnote 12 The application of knowledge, then, can only be pursued in a state that is like that of a blind person finding her way by reaching out and feeling her way through the world. Cartesian science, then, recognizes its own blindness when it comes to applications. Moreover, Canguilhem reveals evidence that Descartes reversed the apparent priority of theoretical knowledge over the blindness of technique. For Descartes’ Dioptrics offers thanks to the lucky successes of the more or less blind human techniques that created early telescopic lenses, leading Canguilhem to write:
Science proceeds from technique not because the true is a codification of the useful, a recording of success, but on the contrary because the technical obstacle, lack of success, and failure, invite the mind to interrogate the nature of the resistances encountered by human art, to conceive the obstacle as an object independent of human desires, and to search for a true understanding.Footnote 13
The will to know and the objective world itself arise from the failure of technique, whereas techniques arise in response to “the exigencies of the living.”Footnote 14
Such an account of the relation between science and technique agrees with the account of error found in the Meditations on First Philosophy.Footnote 15 For there we find that humans, according to Descartes, are precisely like God in one respect, namely, the perfection, the infinite reach of our will, which knows no limits. Unlike God, whose perfect will has priority over his understanding, humans make errors in judgment because we fail to contain our will within the bounds of our imperfect and limited understanding. While Descartes formulated this account so as to understand how error might be overcome once and for all, it reveals that human judgment often extends beyond the known. Thus, Descartes believed that we could avoid falling into error by refraining from judgments about those instances where we lack certain knowledge. But Descartes does think that the failures of technique also hold a positive value insofar as they call forth the search for knowledge. It is well known that, for Descartes, certain knowledge is to be gained through the analysis of ideas; but, as Canguilhem suggests, this depends on the prior syntheses of human technique. He writes, “[s]eeing technique as an action [that is] always synthetic to some degree, thus as such un-analyzable, is not, from the Cartesian point of view itself, it seems to us, to take away all value from it, since this is to see in technique a mode, although inferior, of creation.”Footnote 16 The blind groping about of technical experimentation is creative in its occasional and fortunate successes, but, thanks to its very failures, it also creates the possibility of pursuing theoretical knowledge.
1.3. The Treatise
Canguilhem’s 1939 Treatise on Logic and Morals (Treatise), co-authored with Camille Planet, has been hard to find until recently, but now that it is available, it should help in the assessment of Canguilhem’s own intellectual development. In this paper, I address it in relation to the question of error, since it deploys an account of error that is clearly connected to the earlier work, but that offers further precision about Canguilhem’s thoughts. Yet, perhaps, this text needs to be approached somewhat cautiously not only because it was composed as a textbook, but also because the authorship is not Canguilhem’s alone. Indeed, the first part of the book, which discusses the sciences and contains the most frequent references to, and extended considerations of, error was, claimed Canguilhem, composed primarily by Camille Planet. Indeed, Canguilhem later distanced himself from this part of the text, explaining that he was not the author of the part concerning the philosophy of the sciences, and so did not agree with everything that Planet had written there. But, as Xavier Roth has written, it is clear that Canguilhem’s own notes on error, science, and technique were used by Planet in composing the first part of the treatise. Whatever his later reasons for distancing himself from the considerations of science and error here, Canguilhem was clearly involved in the preparation of the entire text.Footnote 17 Insofar as it is a textbook, one might also worry that particular philosophical topics are not developed in great detail. But Canguilhem and Planet tell the reader that their Treatise aims at systematic exposition in order to give students a taste of the unity characteristic of philosophical work, defined as:
… the search for and choice of particular principles that are supposed to permit the appreciation of information’s [renseignements] value and the judgment of apparently diverse opinions; the choice justifies itself by the coherence of the judgments that it makes possible, and by the unity that it thus assures the mind [l’esprit].Footnote 18
Examining the place of error in this text will be helpful, then, for grasping its status in a systematic account of human activity, one that explicitly aims at establishing coherence between diverse judgments, especially insofar as this diversity entails confusion and ambiguity. Given the tight focus of Canguilhem’s later publications, the systematic treatment given by the two authors here is quite promising for what it might reveal about error. The Treatise is divided into two main parts, the first dealing with the sciences, the second with morals. Most uses of the word ‘error’ occur within the first part and draw on the work of Gaston Bachelard, who already considered the search for error essential for scientific work.Footnote 19 But error is not only a scientific concern, a point to which I will return later.
After a brief overview introducing the history of logic and modern science, the authors of the Treatise begin by defining science as a collection of affirmations about the real, that is, judgments deemed to be true representations of reality.Footnote 20 Such affirmations are not the sole domain of the sciences, since people everywhere depend upon accounts of what is real in order to orient their lives and actions. But science transforms such quotidian affirmations by introducing the importance of method and research, and questions about which methods are appropriate in the search for truth. They find that,“[a] science is, in effect, only a knowledge made methodical for the sake of verification.”Footnote 21 But this raises the question of what could have made the search for a method appear to be a useful way to discover the truth.
Canguilhem and Planet respond to this question by discussing “The postulate of Truth,” which is identified as the a priori conditions of possibility and legitimacy for any theoretical construction that provide rules for the evaluation of the assertions contained within this construction. There are three such conditions contained within the postulate, the first proposing belief in a reality distinct from thought [la pensée] and the third recognizing an obligation for thought to conform to this reality. But the second condition also serves as an essential principle for scientific research, since it introduces the possibility of error.Footnote 22 Without the possibility of error, the authors claim, there would be no reason to search for a true as opposed to a false understanding. Science then is the methodical search for error, which itself makes sense only in relation to truth: an error is that which I believe to be true, but later discover to be false. Although error appears to be a privation of truth, a mere negation from this perspective, even a representative of the theoretical approach as vehement as Descartes understands that error is not nothing, since he recognizes that errors happen when the will judges without knowledge. In agreement with Canguilhem’s earlier thoughts on Descartes, the two authors argue that, error is an a priori condition for theory, even though theory posits truth as the guiding value of its research and regards errors as illusions that must be hunted down and eliminated. Science, like all theoretical enterprises, is austere, severe, and even cruel in its revelation of the individual and collective, “fantasy of passions, desires, hopes, and, above all, the direct urgency of our vital needs that put error in thought.”Footnote 23 In a Cartesian, or even Spinozist, formulation, Canguilhem and Planet suggest that, “the basis of every fiction, collective mythology, or personal fantasy consists in imagining the object on the model of the subject.”Footnote 24 And in a Nietzschean elaboration of this explanation, they assert that, “[e]rror is fiction, and every fiction has its source in the ‘will to power’ of Spirit, first misrecognizing Nature’s autonomous faculty of resistance.”Footnote 25 But if errors are free creations of thought, then the true is “precisely that which one is not free to refuse and, through this, [it is] even legitimately accepted.”Footnote 26 The true, they conclude, is what cannot not be. All theoretical systems, including those of the sciences, thus tend towards determinism since they endeavor to grasp the necessity inherent in nature.
The a priori conditions, or beliefs, that govern and render the scientific work of verification possible are formal not because they are transcendentally a priori, nor because they are ontological facts, but because they govern scientific method in all cases. When error is understood, one grasps that it is not a particular problem that might or might not appear, but that it:
… has the formal value of conditioning the total theoretical problem by giving it its own proper resolution as its aim. It has permitted us to see that this problem, philosophically understood, is above all that of a legitimate transmutation, by which we should be able to substitute a sure value, deduced from the very ‘Object’ of the belief, for the doubtful value of the initial belief.Footnote 27
For a careful scientific mind, such security could only be the final unity at which all its work aims. Thus the reality, being, existence, or materiality of nature are not objects grasped through the work of science; each is instead “only the regulative or normative unity of a theoretical judgment that is constituting itself.”Footnote 28 If science were to achieve knowledge of such things, Canguilhem and Planet write, this would be an achievement and a result, for there is no pre-constituted given [un donné tout fait] for the scientific mind. Quoting Gaston Bachelard, they write: “There are no first truths, there are only first errors.”Footnote 29 But how is it that scientific work can begin if there is no given?
For Canguilhem and Planet, scientific work is essentially analytic, and it is experience that performs, to begin, an initial, but confused synthesis of perception. They write:
Experience is a sort of pseudo-concrete, spontaneous, and uncontrolled judgment in which two abstract systems of judgment are amalgamated without [any] rule. The one [system] is oriented towards the outside and the other towards the inside: since the objective fact expresses the action of Nature on us, and the affective reaction expresses our own needs or tendencies.Footnote 30
Such an uncontrolled, spontaneous mixture of meaning leads to confusion and, potentially, conflict. First, the empirical thought at work in perceptive syntheses tries to make sense of how these two systems of judgment actually agree with each other in experience; “thought, said otherwise, carries out a rash attempt [essai] of practical judgment.”Footnote 31 Such an attempt is risky since empirical thinking supposes it will be able to hastily establish agreement between, “our own desires and the autonomous movement of things.”Footnote 32 And the less conscious one is of the potential conflict, the riskier are these attempts, for such conflicts are eventually realized in “the form of failures, that is to say, precisely by more or less clean ruptures between the two rashly accorded beliefs, from which comes suffering, perhaps even death.” They find that, “[n]either animal instincts, nor the prudent but blind attempts [tâtonnements] of men allow either the one or the other to always escape brutal discordances of this type: if Experience alone was sufficiently concrete and practical by itself, all of the living would be eternal.”Footnote 33 Yet it is precisely because experience is not sufficiently concrete that theoretical consciousness is produced “as a reflection on the conditions of the rupture.”Footnote 34 They think, therefore, that, “human science is effectively nothing other than this reflection, which orients itself towards an abstract and methodical study of things alone, towards pure objective analysis.”Footnote 35 Thus all science depends on a choice, since it is also possible to choose to engage in technical reflection or aesthetic reflection instead. But the very possibility of truth and the scientific spirit [l’esprit] emerges from the failures of technique when “interpreted as an error,” making science an analysis of the objects supposedly existing in Nature as obstacles to our own desires.Footnote 36 Technical successes, by contrast, though implying at least rudimentary foresight:
... must be explained, in the same way as instinctive animal successes, by the so-called method of ‘trials and errors’ [essais et erreurs]. Over the course of a very long development, a technical success can only have been purchased at the cost of multiple failures in a large number of attempts.Footnote 37
The methodical experimentation of the sciences is then a kind of continuation of the trial and error that is blindly attempted by instinct and technique. Instead of proceeding blindly and trusting in techniques that have failed, sciences institute hypotheses about their objects and seek to verify them by looking for possible errors in these.
The authors go on to consider the possibility of moral sciences—to which certain forms of psychology, history, and sociology aspire—that would establish scientific truth about human values and actions. Establishing the truth of history, for example, would mean revealing the unity of the entire past in its very necessity. They worry that such an approach to history would lead one to disregard much of history as unimportant and without value, since it would have the value of a mere obfuscation. The errors and failures of the past would be regarded as so many trivial problems that are not necessary to history itself—for only the truth of history would remain relevant. A universal history, though, would have to be more than scientific, since adventures and accidents, failures and errors are the very source of scientific activity and the value of truth itself. Beyond historical knowledge, the authors worry that if the sciences were successful in all fields, including knowledge of the human being, they would explain the whole of nature and mind. Canguilhem and Planet write that:
… in achieving the elaboration of a true representation of Nature, Science would have directly ruined, devalued the very principle of all error, and, in so doing, abolished every right of Thought to organize experience as it pleases.Footnote 38
Canguilhem remains true to this idea in his later work, and here the authors proceed to establish the difference between science and morals in terms of the unification that each seeks. If unity in science is provided by making judgments of necessity in conformity with logic, in morals it is sought through “judgments of power and value.”Footnote 39 And, as they suggest here, it is necessity as a whole that lacks the power to explain the whole of human activity. Thus, they announce that the sciences need to be thought of not scientifically, but in terms of their powers and value, in terms of their vitality within the range of human activities. The very work of philosophy is to clarify the diverse values of judgment present in experience and seek a coordinated unity for them—and this implies a hierarchy of values.Footnote 40
As a result, the task of philosophy is a moral one. And the work of science, that is, the distinguishing of the true from the false, is only one instance of moral valuation, for there is moral judgment as soon as one admits the opposed values of good and evil, or right and wrong [“le Bien et le Mal”].Footnote 41 Truth is the good of that activity which seeks to grasp the necessity of things. But right and wrong, and good and evil, in their larger scope are values of action, making morals, “the normative organization of conduct in general.”Footnote 42 Canguilhem’s allegiance to the project of a philosophy of values shows through clearly in these pages.Footnote 43 Unlike science, which proceeds by abstractions, moral considerations are concrete and total insofar as they seek to evaluate particular actions, activities, and ways of living. The subordination of truth’s value to that of the good is shown in one of the few discussions of medical practice contained in the Treatise. For learning the truth about why one is in pain might be experienced as anything but good, as in the case that Canguilhem and Planet imagine of a person who seeks medical care only to learn that he is afflicted with a fatal disease.Footnote 44 Morals, therefore, go beyond an interest in the necessary, as their essential concern is human liberty. While truth and liberty are confused in everyday life, the goal of philosophy is to produce a clear “coordination or subordination of the two sets of faculties.”Footnote 45 They write, “[s]cience is certainly not morals, but we see very clearly that a morality [une morale] would only merit its name insofar as it had incorporated science.”Footnote 46 The two make clear their belief that there is no definitive solution to the moral problem, no definitive code or other ordering of human activities and values.
Yet a definition of moral value is still possible, since “the value of morals is creating unity there where there was none.”Footnote 47 This means that there is a different morality for every practice and for each kind of individual, even as all share in the same task of “progressively constructing a responsible personality.”Footnote 48 This means that moral awareness is creative of itself and is, therefore, perpetually in search of that which it is not.Footnote 49 As they state elsewhere, morals consists not in the love of ourselves as we are, “but above all in the will to become what we are not, that is: reasonable, aware, and free [raisonnables, conscients et libres].”Footnote 50 Individual thinking then is to be cultivated with these aims, but given the fact that human life occurs in society, where standards tend to be imposed on individuals from the outside, this cultivation involves a struggle. This means that education should not reproduce these standards, but rather should seek to foster critical consciousness amongst students. Adherence to collective standards, or worse, their imposition, prevents the formation of such ways of thinking. At heart, the risk of error is central to moral development:
To think, this is to risk error, to suffer from it more or less profoundly and recover. There is no thought without an effort [to move] from less to more coherence ... The only proof that individuals are able to give of their thought is that they get things wrong and they are unhappy because of it.Footnote 51
This does not make society a pure obstacle for the individual, but rather, as air is an obstacle and means for the techniques of aviation, so too society is both an obstacle and means for personal creation.Footnote 52 Yet the very process of personal creation implies that we are in a state of becoming and this means that the eventuality of error is engrained within the process of thinking, that is, of thought going beyond itself. If consciousness is, then, bound up with a sense of error, with a worry that one is in error, consciousness is also important for what it does to the sense of life; for, “consciousness of life makes this life immediately appear as a means, thus as a subordinate value.”Footnote 53 Canguilhem and Planet do not hesitate to consider suicide as evidence of this, for this is the starkest and most striking way in which the subordination of life’s value becomes apparent.
It might be tempting to say that suicide is the greatest of moral errors a person can commit. The approach taken by these authors is, however, instructive. For they write that, “it is human, that is to say respectful of value, to refrain from judging suicide.”Footnote 54 They refrain from judgment in part because, “no intelligence is the measure of the power of intelligence,” implying that the very meaning of the act of suicide cannot be established once and for all by any particular intelligence.Footnote 55 Then, in an allusion to the contemporary situation, they write:
Let us suppose, without great rashness, [that] the persecution of racial or political minorities [has] become the rule across the entire earth, the only planet so far accessible to the human species, and we ask ourselves if we would not still pretend to judge the resolution of those who would prefer to die as a result of their courage rather than the ferocity of others. God, if he admits pogroms, why would he not also admit suicides?Footnote 56
Individuality and the meaning of events and actions for the concrete individual are always a primary concern for philosophy. And it is the authentic value of error, of getting things wrong for the individual as such, that allows the two philosophers to defend liberty as the highest value for the concrete human being.
Error is the principle, then, that enables a unity of science and morals in this text, one that Canguilhem saw fit to sign with his own name. It shows that science, the search for truth, cannot simply be extracted from human life without some sort of abstraction being performed, while also suggesting that it is the very sense of truth that allows for the emergence of consciousness and the aspiration for a moral life. What unites both are the trials and errors, the blind groping about of the living that opens the way for failures and the possibility of error in its narrower, theoretical sense, be these scientific or moral in character. Theoretical error and truth must be incorporated into human life, they suggest, without becoming the dominant values deployed for judging all activities, or being treated as relevant only in the search for scientific knowledge. Indeed, it is the value of error before that of truth that appears in both scientific and moral judgments. While Bachelard’s epistemological works had focused on error as a topic in the 1930s, and Sartre had begun a project of freeing philosophy from epistemological considerations by focusing on concrete experience, Canguilhem and Planet insist that error must be understood in relation to concrete individuals, that an awareness of error is constitutive of consciousness itself. And since Canguilhem had already looked to the concrete experiences of technical attempts and failures, [essais et erreurs] as conditions of possibility for our awareness of error, it is no wonder that medical art becomes a central area of investigation.
2. Error in 1943
In 1943, Canguilhem found himself teaching at the University of Strasbourg, relocated to Clermont-Ferrand because of the German annexation of Alsace, as well as publishing his thesis in medicine, the Essay. Among his courses, he taught one on “Error,” in which he continues the project of treating error in its narrower sense, as that which is opposed to truth. His course investigates its concrete conditions of emergence, whereas the Essay partially transforms the problem of error through direct contact with concrete human problems.
2.1. The Course
The course on “Error” endeavours to locate the value of error for truth. I will only consider some of the first pages of Canguilhem’s course notes. To begin, and in accordance with the etymology, Canguilhem finds that the definition of error as, “an act of mind that judges what is false true,” to be the primitive one, since erreur, from the verb errer, is a false direction taken in the course of a journey.Footnote 57 “It is swerving [écart], it is adventure.” Thus error is closer to delirium [délire] than it is to harmful habit [errement].Footnote 58 To go mad [délirer] is to go astray from the track of reality, just as to err [errer] is to quit the path of truth.Footnote 59 Thus, “Error and delirium are species of the genre of extravagances.”Footnote 60 To commit an error is not to wander aimlessly, but to go astray from a certain trajectory. Error, then, strictly speaking, is relative to the pursuit of truth, and has its meaning in relation to theoretical activity, especially that of the sciences.
The course takes positivism as an adversary, which is criticized for tending to treat truth as timeless and failing to see that the truths its methods have uncovered are destined to become errors. Whatever positivism might say, scientific activity is stringently defined by the polarity of error and truth, and its own power of repelling errors, more than a comprehension of eternal truths. Philosophy endeavors, therefore, to criticize the positivist, rousing him from his dogmatic slumber, from the belief in the eternality of truth acquired. In its concern for an authentic awareness of human liberty, philosophical activity becomes a critique of scientific reason and its will to have done with error. Even before positivism, though, Descartes’ desire to eliminate doubt is identified as representative of the problem. Descartes, who does not make out as well here as in 1937, does not allow for a significant recognition of error, since his entire work aims at the elimination of error from human experience and activity.Footnote 61 But, unlike a Descartes or a Comte, Husserlian phenomenology and its various proponents are cast as recognizing the authenticity of error for human experience in the form of anxiety and fear.Footnote 62
Canguilhem begins to explain the value of error by asking, “[w]hat is, therefore, the situation of being in conscious experience such that error is possible in it?” It is not that the object of knowledge is like the object of sight, which I passively perceive as light travels to my eyes; rather it is something that I intend, something that I strive to grasp via a gesture that has its origin in me. Anxiety about our situation, uncertainty as to our place amongst other beings, this incipient fear of error, leads us to the value of truth and to practices that search it out. Thus, epistemology necessarily refers to knowledge’s own past, to the history of judgments that a science has made. He writes: “The possibility of error reveals that knowledge is an operation engaged in a becoming.”Footnote 63
Recognition of a fundamental instability in being calls up every search for truth, showing that knowledge is savoir before connaissance, a search for and way of knowing before anything else. Even knowledge that seems immediate and indubitable can be called into question thanks to the recognition of errors in the history of sciences. The problem is ontological, for, as Canguilhem asks, “[w]hat guarantees that being is being, since being appears to be otherwise than it is?” Because of this fundamental tension in being, which he identifies as the source of anxiety,Footnote 64 any immediate identification of a being comes upon great difficulties when a more general rule for its identification is attempted. While such difficulties are not insurmountable, they do not admit of immediate and spontaneous resolution. Difficulty and obstacle demand mediation, instead, making science an operation, always a becoming, never a being.
Returning to the primitive definition of error, Canguilhem writes:
Error is that which I no longer believe to be true, but what I did believe so. Error is that which I can no longer think, but what I could think before. In recognizing the duty to no longer think what I thought, I aver in the same moment, the power to think otherwise than I should. Consider these two characteristics of judgment brought into evidence by the fact that it is capable of error:
· judgment is an ability [puissance], a power [pouvoir],
· judgment has a history, a becoming.
This ability [puissance] of affirmation that is judgment is, in itself, outside its relation to the duties of affirmation, pure ability of anticipation of what could be being, but also pure mistake. If judgment did not exceed its duty, if it remained pure obedience to the obligations that define adherence to reality, it would be initially and definitively exact. In the intellectual order, “without fear and without reproach,” does not mean anything. It is because one is without fear that one risks reproach. Inversely, knowledge, taken as judgment passed on the real, is the fruit of fear. We have seen that the first experience of error is an attenuated form of anxiety, doubt. There is a kinship between light and knowledge, knowledge dissipates phantoms.Footnote 65
Knowledge is an operation proceeding via the progressive formulation of judgments that strive to affirm what is, and this is possible only because judgment is a pure anticipation of being, that is, a “pure mistake.” A judgment can aim at truth only insofar as it goes beyond the limits of what is immediately apparent. All claims to truth express, therefore, a power that must be understood historically, that is, by reference to preceding ontological judgments. The condition, however, that renders judgment possible in the first place is the perpetual threat or fear of error. Indeed, in order for me to recognize that there are false judgments, I implicitly affirm that I have the power to think “otherwise than I should.” The possibility of error is thus intimately connected to one’s own being, one’s own choice, and thereby is an affirmation of our liberty—even when one makes a judgment that is in error, and especially when one makes a judgment that goes against what is already affirmed as true. Taken further, the very pursuit of truth expresses liberty because it stems from the human power to stray from a reality that is not itself uniform or rational. In this sense, more than a pure mistake, error is creative and indicative of human freedom, the guiding value of philosophy.
As he explains here, the central place of error in all scientific activity requires that Canguilhem’s philosophy of values be pursued through the history of the sciences, even if the sciences tend to deny belief in human liberty. About this study of the sciences, he writes:
The history of sciences is thus a philosophical, rather than scientific discipline. It is epistemology rather than logic. It is anthropology rather than science. It is one of the elements of culture rather than one of the pieces of knowledge [savoir].Footnote 66
The characteristic mark that distinguishes the history and philosophy of sciences from the sciences appears to be set by truth, insofar as this is recognized as immutable and unchanging. Thus the history of science might seem to be the obverse of truth, the history of those errors that have been abandoned on the way to truth. But Canguilhem is not a positivist who relegates errors to the past of contemporary knowledge, for it is in this history of errors, he suggests, that the vitality of contemporary scientific activity is found. Through this history, the epistemologist presents error as a value that cannot be relinquished, since it defines the scientific endeavour from its very beginning, giving it the power and possibility of progression. Error, then, is the generative scientific problem insofar as it is the experience of error that gives rise to a desire to eliminate errors by coming to know the truth. In guarding sciences from their own illusions—naive belief in objective truth—epistemology reveals scientific activity to be one of the ways that humans live, as one element of human culture, and this sets the stage for considering the role that this activity has in this broader context. Investigating the elements of culture, the epistemologist will then be able to reveal the conflicts arising between different activities and the confusion of values and damage that results. And Canguilhem’s work focuses so closely on this one element of culture, rather than political or religious life, for example, because the history of scientific disciplines demonstrates human liberty so vividly, and therefore not in what it has acquired, but in its very becoming.
2.2. The Essay
Does the course on “Error” provide a set of instructions for reading Canguilhem’s thesis in medicine? There are good reasons to think so, though the Essay also continues the earlier work discussed above in its attempt to find positive values implicit within the negative. Part One of the Essay reads very well as an account of the way that the sciences have been progressive without complete awareness. Its genealogy of a scientific dogma shows how the belief in the homogeneity of the normal and pathological save for quantitative variation was appropriated from the life sciences by those working in literature, philosophy, and sociology. And Canguilhem traces the logical confusion—that there is no difference between normal and pathological states because the sciences can explain the pathological as a continuous, measurable variation of a normal function—implicit within various formulations of the dogma back to Xavier Bichat, who claims that “every pathological phenomenon derives from their [the natural properties of the body] increase, decrease, or alteration.”Footnote 67 François-Joseph-Victor Broussais, claims Canguilhem, read this and equated alteration with the notions of quantitative increase or decrease, thinking of it as nothing more than these. Canguilhem finds that such an approach confuses quantitative change, increase or decrease, with qualitative change, alteration, effectively blending the two different perceptive orders of scientific abstraction and concrete experience. Despite such confusion, Broussais’ principle was taken up by the positivist Auguste Comte in his attempt to found sociology, popularizing the dogma, but, just as importantly, it was taken up Claude Bernard, who deployed it as a justification of his experimental studies on physiology and pathology. And despite the confusion within the thesis, and ambiguities present in his thinking, Bernard was able to make important contributions to an understanding of physiology and pathology, contributions that would open therapeutic possibilities for those suffering from diabetes, where no such possibilities had existed before. The Essay, then, despite its criticisms of positivist interpretations of the sciences and medicine, also lauds the work that has been accomplished by individuals such as Claude Bernard. It is telling that Canguilhem does not offer scientific judgments about the historical authors that he is considering. Rather, he shows the extent to which Bichat’s vitalist ideas animate the work of the very positivists who think they have done away with vitalism altogether. Canguilhem does not discuss which scientific accounts were errors and which were true because he is content to show a certain confusion that signals the unabated continuation of vitalism, a doctrine regarded as a failure by the positivists. The closest he comes to a discussion of error as such in Part One is in a quotation from Victor Prus, who was already writing in 1825 against scientific dogmas that would become so famous later in the 19th century. Against the claims that disease merely magnifies what is happening in the normal, or physiological state, and that pathology is, therefore, founded on physiology, Prus states, “pathology, aided by pathological anatomy, has created physiology: every day pathology clears up physiology’s errors and aids its progress.”Footnote 68 So here we have an appeal to “error” in its scientific sense within Canguilhem’s text. And yet it is not Canguilhem who uses the term. He is pleased to report such a scientific critique of the claim, but he, by contrast, assesses scientific claims in terms of their “confusion” and “ambiguity,” that is, in terms of the epistemological approach that he defines in the course on “Error.” Indeed, he explicitly states in the introduction to the Essay that he is treating this scientific dogma from the perspective of its cultural relevance, since no scientific claim to truth can arise outside a cultural context.Footnote 69
Another significant mention of ‘error’ in the context of scientific work concerns Gaussian error curves and is made in Part 2 of the Essay, dedicated to the question of whether sciences of the normal and the pathological exist.Footnote 70 The error curve, also identified as the normal or bell curve, allows for the identification of regular types or variants of a given feature within a population based on randomly taken measurements and establishes the probability that any given member of the population expresses a particular type or variant. Such curves figure, therefore, in Canguilhem’s discussion of norms and averages, their differences, and roles in the life sciences and medical practice. In particular, Canguilhem deals with questions about the extent to which certain traits are normal in the statistical sense, that is, because they are widespread, and the extent to which certain traits are statistically important, that is, widespread, because they are normal in a prescriptive sense, because they are good traits to possess. While the study of probabilities might not seem to conform to scientific determinism, Canguilhem believes that it does, since error curves do actually reduce apparent diversity to basic, unchanging ratios of possibility.Footnote 71 The error curve, in short, is a way to remove error from the study of particular phenomena and thus perpetuates the scientific problem of eliminating error on the assumption that one always begins in error. But Canguilhem develops no lengthy discussion of error in science here, though he does discuss anomalies at length. Anomaly, not to be confused with abnormality, is the rule for living beings, which are defined by their irregularity. Canguilhem can argue, therefore, that regularities within statistical studies do not simply reveal what is normal, but rather what has been normal in a certain situation. That is to say, particular forms are more numerous than others because these forms have been more successful in their milieu than others. But this does not mean that the identification of the average types entails knowledge of which forms are currently normal in the vital, normative sense. In deploying the Gaussian error curve, the life sciences do not grasp error as a normative qualification in the objects that they study, but rather as continuous, quantitative deviation within and from a typical range. Canguilhem does not here pass judgment on the sciences or medical practice that deploys such mathematical functions in the attempt to know and heal, rather he wants to show that averages, whether of the numerical or typical sort, express norms, but are not themselves synonymous with them. So it is clear that Canguilhem thinks that science works by the rectification of errors and that it thus seeks to expel them from its domain in its search for objectivity, but it is unclear how this figures in the larger context of the Essay.
As Canguilhem notes in the introduction, he undertook medical studies as a philosopher for one reason alone: he thought that it would give him an introduction to “concrete human problems.”Footnote 72 If we take this claim seriously, while remembering the scientific desire to eliminate error through abstraction, then we can understand why error does not figure prominently in this text. The problem of error in the sciences alone is not concrete. Instead, it is a generative difficulty for practices of abstraction and remains tied to them. The concrete human problems that Canguilhem discusses in his book are those of human disease [maladie] and suffering, which he treats in relation to the medical art that treats them and in relation to the sciences that support this art.Footnote 73 By refusing to discuss the would-be sciences of physiology and pathology in solely scientific terms, Canguilhem thus puts off any discussion of medical practice in terms of its truth, or lack thereof. So too, the concrete human problems of disease and suffering are protected from treatment in quasi-scientific terms as errors that depart from true health, identifiable by the sciences themselves. Attention to error is replaced, therefore, by focus on the concepts of the normal and the pathological: technical, scientific, and popular terms used in the evaluation of health and illness.Footnote 74 Against the positivist will to eliminate any real distinction between normality and pathology, Canguilhem locates this distinction in the living experiences of impotence before obstacles. In the second part of the Essay, Canguilhem examines the concepts of the norm, the normal, and normativity at length, rooting them in the activity of the living being. Indeed, the primary meaning of normativity is, for him, biological. Canguilhem’s claim is that living things are normative in the most basic sense, since the living strives to institute the norms by which to live.Footnote 75 And since the living is never indifferent to what happens around it and to it, the activities of living beings, from humans to the protista, must be understood in terms of polarity. As Canguilhem writes, “[t]here is a dynamic polarity of life”.Footnote 76 Such a dynamic polarity means that the philosophical sense of normativity, understood as judgment according to a norm, is a development of the biological normativity that leads to the creation of a meaningful environment before the appearance even of any apparent consciousness. While seeking to avoid anthropomorphism, Canguilhem suggests that this means that life is inherently technical.
It is not that all living things make choices that lead to the development of diverse techniques followed by debate about which techniques to deploy and which not. Rather, norms are technical in the sense that they are instituted by living beings as ways of living. In this sense, human techniques and arts, the same as all human activities—including religion, science, and philosophy—emerge out of the normativity of the living. Such an emergence is not tantamount to a clean break, for though it is true that science can be pursued only thanks to the failures of technique, there remains a connection between the sciences and the experience from which they emerge. Though there may be sciences of physiology or pathology, for example, these carry the very experience of the pathological and the normal in their structure as sciences of these different objects. The technical successes and failures of medicine retain, therefore, a primacy in relation to the sciences that came about as a result of technique’s failures. In this, what humans experience as negative in value, as wrong or evil, has taken on a positive value as objects of knowledge through medical practice and the will to know that arises in its failure.Footnote 77 And this is not merely a point about how human experience, art, and science relate for Canguilhem. For he argues that these all arise out of the fundamental way of living shared by all kinds of life. He writes:
[a]ll human technique, including that of life, is set within life, that is, within an activity of information and assimilation of material ... Because life is activity of information and assimilation it is the root of all technical activity.Footnote 78
But this does not mean that knowledge—understood as the methodical pursuit of truth—as such is already at work in living beings. For Canguilhem is careful to claim that:
… human technique extends vital impulses, at whose service it tries to place systematic knowledge which deliver them from much of life’s costly trial and error [essais et erreurs].Footnote 79
Rather than adopt the physiologist’s account of life, Canguilhem looks to the physician, for whom:
… life is not an object but rather a polarized activity, whose spontaneous effort of defense and struggle against all that is of negative value is extended by medicine by bringing to bear the relative but indispensable light of human science.Footnote 80
Science is not human by taking man as its object, but insofar as it is an activity in the service of life’s dangerous “trials and errors.” While Canguilhem uses this phrase only once in the Essay, it clearly draws on his earlier work, and suggests that before error gains its scientific or moral sense, it occurs as a blind straying here and there as the living being attempts to dominate its environment.Footnote 81 Diseases are themselves eventualities for all living beings, insofar as life is a blind risk before a seeing appreciation. There is the suggestion, then, of an emergent rationality within every organism, manifesting itself in the differentiation of health and disease as vital realities. For the conscious being, living with disease takes on the negative value of an ill or evil. And for human beings, living in anxiety before its eventuality, disease itself becomes creative over the course of history, as the experience in which human liberty, human power develops through medical techniques and sciences.
Concluding Remarks
Treating concepts of normality and pathology in concrete experience, Canguilhem’s Essay discusses the sciences of pathology and physiology in relation to the larger context of the medical art, which is not a science itself but at the crossroads of several sciences.Footnote 82 A scientific approach, though generated via a hunt for errors, attempts to get beyond error and eliminate it from its field of study, explaining the desire to suppress the qualitative distinction between the normal and the pathological. But the Treatise suggested error also has a moral sense, not based on a misrecognition of some objective reality, but the failure to establish a kind of unity, a coordinated hierarchy of the various activities and values in an individual’s life. The Essay keeps the question of scientific error at bay, therefore, because it is an attempt to avoid error in the moral sense by introducing the source of normative judgment, the experience of the difference between health and illness that results from the living institution of norms. For the relatively abstract question of error’s role in the sciences is substituted a concrete reflection on the value that disease has in human life in order to risk a way to unify modern values and practices. But an account of moral error is not given here either. Throughout the Essay, instead, one sees the fundamental importance of blind trials and errors, whether in the discussion of life’s anomalies, diseases, and abnormalities, or the very history of the life sciences.
In 1943, then, the problem of error continues to concern the question of the possible positive value of error and thus, ultimately, its place in human life. Canguilhem’s approach is to locate error as the generative problem of the sciences, as an authentically creative value for human beings, while also rooting it in technical and vital activity, which, he suggests, is blind before it sees. Error in its strict sense is conditioned by the failures of living beings to completely dominate their environment in a spontaneous and unimpeded manner. And this means, for Canguilhem, that living proceeds by risky endeavours that, in their very failure and the anxious search thus motivated, provide information about how to live. And here we have the question, the problem that continues to occupy Canguilhem in the coming years. For it seems that either scientific error is reduced to a peculiar form of blindness, becoming merely another kind of failure, and depriving the search for truth thereby of its unique powers and accomplishments. Or, error, and truth with it, really is in living beings, and then the trials and errors of the living are deprived of any positive value, existing only as obstacles to the development of the true form or forms of life. The problem raised by the Essay and earlier work concerns the relation of error to life and the extent to which error is in life and life in error.Footnote 83
In the early 1940s, Canguilhem was, among other things, a philosopher in the process of finishing his education as a medical doctor. The thesis he wrote to complete this education, the Essay on Some Problems Concerning the Normal and the Pathological, demonstrates the transformation that his thinking was undergoing as he deployed his own thoughts about error to concrete human problems of illness, suffering, and health. His work on error led him to downplay the very category of error as such, as a member of the truth-error couplet, in favour of a theoretical approach that sought to clarify the different values at play within physiology, pathology, and medical activity and thereby contribute to a critical and theoretical Essay to coordinate these values. But at the centre of this critical intervention, his earlier philosophical work on the problem of error remains active, especially his account of life’s trials and errors. Truth involves openness to the world because the experience of failure in life’s trials and errors makes objects of scientific investigation emerge as worth knowing. But philosophy must also remain open to the truth of the world in an essay, an endeavour to coordinate the very value of this truth with the other values by which we live.
