Version classiqueVersion mobile

Ideas in Time

John Potts

Chapter 6. The Age of Rupture

Other Models of Discontinuity

Texte intégral

1This chapter continues the theme of the previous two by appraising influential theories of discontinuity within intellectual history. The work of two theorists considered in this chapter – Bachelard and Canguilhem – mostly predates the writings of Foucault, while the other two theorists – Kuhn and Althusser – published their major works concerning discontinuity within a few years of Foucault’s The Order of Things. The major part of this chapter is devoted to Kuhn’s The Structure of Scientific Revolutions, due to its enduring influence on intellectual history. I offer a critical reading of Kuhn’s influential model of paradigm shifts in the history of science. First, however, I consider the equally significant work, lesser known in the Anglophone world, emanating from the French tradition of the history and philosophy of science.

2My study of these various models of discontinuity proceeds under the guidance of three questions. First, do these theories also contain elements of historical continuity which may have been disregarded by commentators, as I have argued is the case with Foucault’s writings? Second, can it be shown that any of these models has repressed continuity in the process of promoting its opposite, thus leaving the theoretical model open to charges of distorting the picture of historical development? Finally, can these theoretical constructs, especially those developed and published in the 1960s, be regarded as particular products of their time, historically determined by their genesis in that turbulent and disruptive decade? Could that explain the pronounced antipathy to continuity and tradition found in these theories of the break, the rupture, the shift, the abrupt transition? If it is possible to situate these discontinuist models firmly within their historical context, perhaps contemporary attempts to theorise intellectual history may be liberated, to some extent, from their forceful influence. Tradition, continuity, ‘cumulative history’ (Kuhn) may have something of their stigma removed, freeing the way for an approach to intellectual history fusing continuity with discontinuity.

The Bachelard-Canguilhem ‘Network’

  • 1 Gutting (1989: 11) takes the term ‘network’ from an essay by Foucault on Canguilhem (Foucault 1985) (...)

3Gaston Bachelard was director of the Institute d’Histoire des Sciences et des Techniques at the University of Paris from 1940 to 1954; his successor was Georges Canguilhem. The Bachelard-Canguilhem ‘network’ was especially formative for French intellectuals such as Foucault because it was ‘the primary French locus of the historical critique of reason’ (Gutting 1989: 11).1 In the French context it was the history of science that raised ‘the philosophical question of what enlightenment is.’ (Foucault 1985: 4, cited by Gutting 1989: 11)

  • 2 Two studies charting this influence are Dominique Lecourt’s Marxism and Epistemology: Bachelard, Ca (...)

4The mingling of history and philosophy developed by this ‘network’ was one of its many legacies. For Gutting, Bachelard was ‘primarily a philosopher who based his conclusions on historical studies’, while Canguilhem was ‘primarily a historian of science, though one extremely sensitive to the philosophical presuppositions and implications of his work.’ (Gutting 1989: 32) Such a merging of philosophy and history is evident in the work of Foucault, as it is in other French theorists such as Michel Serres. Indeed, much of the following sections on Bachelard and Canguilhem will perhaps appear in the retrospective shadow of Foucault, in that he took much inspiration from their work, while becoming much better known in the English-speaking world. Yet Bachelard and Canguilhem remain highly significant figures in their own right, particularly in France. Their influence has extended in many directions – as reflected in the writings of Althusser, Serres and others.2

5It is in one way misleading to speak of a Bachelard-Canguilhem ‘network’, as this term neglects any differences between their two bodies of work. Gutting notes that ‘there are a number of key points at which [Canguilhem] criticizes and modifies Bachelard’s views.’ (1989: 32) Some of those modifications and differences are pertinent to the theme of continuity and discontinuity in intellectual history: Canguilhem gives greater emphasis to continuities of thought than does Bachelard. Nevertheless, they share a philosophical engagement with the history of science, laid down first by Bachelard, that has had far-reaching consequences. Bachelard’s notion of the epistemological break, to give one example, has become a key concept representing discontinuity in intellectual history.

Bachelard’s Epistemological Breaks

  • 3 Gutting (1989: 13) draws attention to this point: Bachelard argues that Descartes’ principle of the (...)

6The epistemological break was but one concept developed by Bachelard in his theory of scientific change. For Dominique Lecourt, the crucial significance of Bachelard’s ‘historical epistemology’ is that in approaching scientific knowledge as its object, it takes ‘into account the historicity of that object.’ (Lecourt 1975: 25) In The New Scientific Spirit – published in France in 1934 and translated into English fifty years later – Bachelard founds his analysis on this principle: ‘to say that the structure of the scientific spirit changes is just another way of saying that knowledge has a history’ (Bachelard 1984: 171). Bachelard’s critical attention to changes within scientific knowledge reveals the absence of any universal categories within science – and accordingly, within reason. Scientific developments often bring with them refutations of earlier a priori philosophical ideals, such as the refutation – explored in The New Scientific Spirit – of Cartesian principles from the perspective of twentieth century physics and chemistry.3 Bachelard’s exposition of rationality confers the highest privilege on scientific reason; in his view, scientific change is the dynamo of rationality. Science is also splintered, divided into specific areas of knowledge. For Bachelard it is therefore impossible to delineate either a singular history of science or a unified concept of reason. Science is instead composed of ‘regions of rationality’ (Gutting 1989: 14).

7In analysing the specificity of scientific discourse, Bachelard uses the concept of the epistemological break in two ways – one synchronic and the other diachronic. The former designates the split between scientific knowledge and ‘ordinary’ or common knowledge. Science breaks from common sense by defining experience in terms not apparent to everyday perception. For example, a chemistry text declares that glass is very similar to zinc sulfide, due to the similarity of their crystalline structures. Such a comparison, made according to categories within chemistry, would not occur to common sense and is therefore not part of common knowledge (Gutting 1989: 15).

8The diachronic dimension of the epistemological break is more pertinent for the purposes of this book. Bachelard’s history of science opposed the positivist vision of science as a steady accumulation of knowledge; for him science develops by breaking with previous scientific theories. He was particularly impressed by the profound break instigated in the twentieth century by the ‘new scientific spirit’ represented by the rise of relativity and quantum mechanics. The new physics typified for Bachelard ‘the philosophy of no’ – that is, the break from accepted configurations of knowledge. In detailing the discontinuity marked by the novel conceptual framework of the new physics – in a series of books commencing with La Valeur Inductive de la Relativité in 1929 – Bachelard was, as Gutting remarks, in advance of Kuhn and other Anglo-American philosophers of science by ‘two or three decades’ (Gutting 1989: 16). For Bachelard, the theory of relativity was ‘truly and essentially new’ (Bachelard1984: 43). There was no transition from the Newtonian to the Einsteinian system, and it was a ‘mistake’ to look for one (44). Newton’s system was ‘complete, finished’; relativistic astronomy was ‘in no sense a child of Newtonian astronomy’ (43-44). Relativity theory was for Bachelard ‘without antecedents’; it constituted a ‘total novelty’ of scientific knowledge (Lecourt 1975: 35-36).

9Bachelard developed a conceptual apparatus with which to describe scientific change. As well as the epistemological break, he proposed three other epistemological categories: the obstacle, the profile and the act. An epistemological obstacle is a concept, method or mode of thinking that prevents a break from occurring. Generally these obstacles are residues from previous thought or practices. Bachelard mentions common sense in this regard, as well as previously successful scientific models that now obstruct the advent of new theories. From the point of view of Newtonian physics, for example, quantum mechanics seemed absurd; hence the new theoretical model was initially resisted by physicists of the classical tradition.

  • 4 This analysis is provided in The Philosophy of No (1969 [1940]: 36-38); it is cited by Gutting (198 (...)

10The epistemological profile is one of Bachelard’s most intriguing ideas. The profile traces an individual’s understanding of a scientific concept; more specifically, it identifies the various components of that understanding. The epistemological profile has a diachronic dimension, in that it reveals ‘the degree to which the understanding involves elements from the various stages in the concept’s historical development.’ (Gutting 1989: 17) These elements may include common sense versions – sometimes misguided – of scientific theories past and present, as well as an individual’s interpretation of philosophical ideas derived from science. To illustrate what he means by a profile, Bachelard helpfully analyses the profile of his own understanding of mass.4 This profile, according to its owner, is largely composed of the eighteenth century ‘classical’ conception of mass, leavened by two minor and wildly dissimilar ingredients: a childlike naïve notion of mass, and a highly abstract definition derived from relativity theory. Bachelard’s idea of the epistemological profile builds temporality into the individual’s grasp of scientific concepts. An individual’s (normally unexamined) understanding will comprise partial components from different historical stages of scientific knowledge. As Bachelard’s own example demonstrates, a profile may well prove to be an odd mélange, an uncertain composite of disparate elements. Bachelard gave greater emphasis in his later writings – such as The Psychoanalysis of Fire and The Poetics of Space – to non-scientific modes, including the role of imagination.

11Bachelard’s last category – the epistemological act – relates directly to the problematic of continuity and discontinuity in the history of science. The act is a ‘leap’ in scientific knowledge (it may be a leap over an epistemological obstacle). An act entails an improvement in the state of knowledge: it is a leap forward for science. Bachelard’s philosophy of science, then, incorporates a belief in progress: there can be no doubting that present science constitutes an advance on the science of the past. It is therefore appropriate to judge previous science by contemporary standards; as a result, the scientific past can be divided into ‘outdated’ or ‘sanctioned’ history, the latter representing science deemed valid by contemporary standards.

12This formulation may seem close to a teleological perspective; Gutting, however, cites a commentary by Canguilhem absolving Bachelard of teleology on the following grounds. First, there is no imposition of current concepts onto past science, which must be understood in its own terms; and secondly, there is ‘no assumption of the immutable adequacy of the present.’ (Gutting 1989: 20) It is likely that contemporary scientific achievements and standards will be overcome or corrected in the future; the present cannot stand as an endpoint to which previous science leads, and which previous science justifies as a privileged state of knowledge.

  • 5 Bachelard, L’activité Rationaliste de la Physique Contemporaine, published in France in 1951, cited (...)

13Bachelard’s conception of science is one of perpetual process. All scientific achievements are open to revision; many will be rejected in the wake of new theories and methods. This process of revision, however, also yields some scientific work that withstands the scrutiny of later science: these are ‘permanent’ achievements in science. Bachelard offers the notion of ‘specific heat’, which emerged from Black’s work on caloric, as one that is ‘forever a scientific notion’.5 As well as being incautious, this claim seems to contradict Bachelard’s own theoretical model. How can a scientific concept survive unchanged, through the ongoing series of epistemological breaks that is science history?

14Bachelard resolves this problem in the following way, as summarised by Gutting:

Bachelard’s response is that an epistemological break is not merely the rejection of past science but also a preservation, via reformulation, of old ideas in a new and broader context of thought. (Gutting 1989: 20)

15Contemporary scientific theory will refute the status of a past theory as ‘unconditionally correct’, but may allow it to be correct under restricted conditions. Bachelard offers as two examples of this process the acceptance of Euclidean geometry and Newtonian astronomy, under certain restricted conditions, within the Einsteinian theoretical model. In the Newtonian case, Bachelard claims that Newton’s concepts are preserved not just because they produce calculations roughly approximate to Einstein’s, but because Newton’s mass and velocity can be shown ‘to be special simple cases of the corresponding Einsteinian concepts.’ (Gutting 1989: 21)

16This latter theoretical manoeuvre by Bachelard is unnecessary, creating more difficulties than it is worth, as Gutting observes. ‘In precisely what sense, ’ Gutting asks, ‘is Newtonian mass a “special case” of relativistic mass?’ (1989: 22) Kuhn vigorously critiques this very proposition, as articulated by other theorists, in his The Structure of Scientific Revolutions. Bachelard’s argument suffers here in its attempt to have the earlier scientific model absorbed into the later one as a ‘special case’ of the latter. This move pushes the notion of the preservation of past science (in some form) one step too far. The claim made by the previous step – that the past is maintained as valid under certain conditions – is more plausible, and more satisfactory as an account of the preservation and transformation of past thought in the present.

Scientific Change: ‘Discontinuous but Progressive’

17Despite this ‘special case’ difficulty, Bachelard’s theory of scientific change offers a sophisticated model in which both continuity and discontinuity have a role. Bachelard’s picture of science is, as Gutting observes, ‘discontinuous but progressive’ (Gutting 1989: 21). It is discontinuous because of its commitment to the perpetual revision built into the method of science. The theoretical framework of one period (what Kuhn would call a paradigm) will be rejected by a later period. The history of science comprises a series of epistemological breaks, rather than a ‘linear accumulation of truths within a single conceptual framework.’ (Gutting 1989: 21) The history of science for Bachelard is not a continuity but a succession of leaps from one theoretical framework to another, each one an epistemological ‘act’ carrying science forward.

18Within this progress by discontinuities there is, nevertheless, some persistence of the past. These are the ‘permanent’ scientific theories or concepts that transcend epistemological breaks. They take their place in successive theoretical frameworks by a process of ‘generalisation’: that is, they can no longer be held to be universally valid, but they can be assigned a specific place within a general framework (such as Euclidean geometry within a post-Euclidean frame of geometry). Bachelard terms this process of generalisation ‘dialectical’ – and one would be tempted to invoke Hegel if Bachelard had not already rejected dialectical synthesis as a description of his model. Bachelard insists that older concepts are not ‘sublated’ into a higher unity; rather they are corrected from the perspective of a later theoretical model, which accepts their usefulness under certain conditions. Bachelard’s dialectic of science is one of ‘conceptual expansion’: each new conceptual framework marks a progress over its predecessors, because it ‘attains a more general perspective from which the range of validity of previous perspectives can be assessed.’ (Gutting 1989: 21)

19The role allocated to the past by Bachelard may seem restricted, even disdainful. After relativity theory, he claims, ‘the scientific spirit is able to sit in judgement on its own past.’ (Bachelard 1984: 55) In many cases it ‘condemns its historic past’; science history is a series of ‘historical errors’ (172). Yet at least Bachelard acknowledges, as Kuhn does not, the endurance, even importance, of some previous scientific knowledge. Bachelard’s theory is not without its inconsistencies and weaknesses. How, for example, are those ‘permanent’ scientific concepts such as specific heat ‘corrected’ by successive theoretical frameworks? If they are permanent, why do they need to be corrected at all? There is also a certain vagueness surrounding the epistemological breaks: do they occur at the level of specific theories, or at the level of broader theoretical frameworks?

20But inconsistencies are probably inevitable in a project as ambitious as Bachelard’s. His philosophy/history of science is a bold attempt to incorporate the persistence of past scientific knowledge into the ongoing momentum of science, characterised as progress by discontinuity. Scientific knowledge leaps, it jumps across obstacles. It is no smooth progress but a chain of epistemological breaks. Nevertheless, it carries something of the past with it. Bachelard described this process in several ways: Mary Tiles emphasises his notion of ‘recurrent history’, in which past achievements are inscribed into a science history continually retold ‘in the light of the present.’ (Tiles 1984: 14) Elsewhere, Bachelard declares – in the context of the new physics’ break from Newtonian science – that while there is no ‘development’ from the old doctrines to the new, it can be said that ‘the new enveloped the old’ (Bachelard 1984: 60). Bachelard’s complex model of scientific development influenced many later theorists, whether they were explicitly aware of that influence or not.

Canguilhem’s History of Concepts

21Georges Canguilhem occupies a curious place in the intellectual history of the twentieth century, one that highlights the division between French and Anglo-American thought. A crucial influence on Foucault, Althusser and others, he is scarcely known in the Anglophone world: mostly untranslated and largely undiscussed. Yet even the crudest precis of his work will suggest the formative impact of his interests and methods, most obviously on the thought of Foucault.

22Canguilhem’s history of concepts, generally conducted in obscure fields of knowledge, was an alternative form of intellectual history to that conventional history of ideas so lacerated by Foucault and members of the Annales School. For Canguilhem, the search for precursors undertaken in the history of ideas was a misguided project, in that while it is possible to find thinkers or scientists of different eras using the same term, such a ‘discovery’ recognises only superficial similarities. To claim that the ideas of Copernicus or Darwin were ‘anticipated’ by much earlier thinkers is to construct a simplistic narrative based on the mysterious foresight of the precursor, defined by Canguilhem as ‘someone of whom we know only after that he came before.’ (Gutting 1989: 39) But this kind of narrative is only possible if one overlooks basic conceptual differences beneath any apparent similarity.

23Scientists from different historical periods may use the same term or seem to be expressing the same idea, but their conceptual framework will most likely be fundamentally different. They may be using similar words, but they are asking different questions. Canguilhem argues that one can only claim an anticipation or prior discovery if it can be shown that the two researchers, separated by time, are

… asking the same questions and have the same research goal, that their guiding concepts have the same signification and draw their meaning from the same system of concepts… (Gutting 1989: 39)

24For Canguilhem, this possibility is extremely remote. His epistemological model, following Bachelard’s, gives priority to change within scientific knowledge; scientists from different eras will, accordingly, be using different concepts and posing different sets of questions.

25In describing the transformations of scientific knowledge, Canguilhem focuses on the history of concepts, which themselves may appear in different theories at different times. For Canguilhem, a concept identifies a phenomenon about which science will ask questions; a theory is an attempt at answering those questions. Different theories will compete to explain the concept, so that concepts are ‘theoretically polyvalent’ (Gutting 1989: 34). The distinctive characteristic of Canguilhem’s scientific histories is that they trace not the development of theories, but the ‘formation and transformation of concepts.’ (Gutting 1989: 34)

26Canguilhem’s most extensive work in this respect was on the history of the concept of the reflex. His major deviation from conventional science history is to deny the foundation of this concept in Descartes, based on Canguilhem’s observation that although Descartes described ‘reflex movements’, he nowhere explained this phenomenon as disconnected from the ‘central seat’ of the brain. For Canguilhem, the concept of the reflex – as a movement occurring independent of the brain – originated elsewhere, in the seventeenth-century physiological theory of Thomas Willis; Descartes’ theory was in fact an obstacle to the emergence of this concept.

  • 6 The reference is from the second edition of Canguilhem’s work on the reflex: La Formation du Concep (...)

27Canguilhem uses the widespread mistaken belief that Descartes satisfactorily explained the reflex to identify the frequent confusion between phenomenon, concept and theory. The concept of the reflex became an important component in mechanistic theories of the organism; it was mistakenly assumed that Descartes’ early mechanistic model included the concept of the reflex, when in fact it only included descriptions of the phenomenon. To illustrate the independence of concept from theory, Canguilhem points to the concept’s introduction in the non-mechanistic theories of Willis, who explained physiology in vitalistic terms, yet properly conceived of the reflex as a back-and-forth movement independent of the brain. Even though he described this process as a motion of ‘animal spirits’, Willis articulated ‘the thing, the word, the notion’ of the reflex (Gutting: 37).6 Canguilhem’s history of this concept traces its continued use in non-mechanistic models of physiology, until its general acceptance in the nineteenth century as ‘an elementary and rigid mechanism’ (Gutting: 37); later theories have rejected extreme mechanistic interpretations. The concept endures, then, taking its place in succeeding theoretical frameworks. Even if the theory – such as vitalism – is abandoned, the concept survives.

28From this brief account, Canguilhem’s history of science can be seen to share Bachelardian features, while departing from Bachelard in some respects. Canguilhem adopts the notion of science history as a series of epistemological breaks, but adds some significant qualifications. His emphasis on the concept allows him to refine the process, stipulated by Bachelard, in which past science is evaluated, rejected or assimilated by current science. A concept may be retained, while the theoretical framework that attempted to explain the concept is rejected. Gutting discerns another ‘correction’ of Bachelard’s theory made by Canguilhem, regarding the epistemological obstacle. Whereas Bachelard construed the obstacle in negative terms, as a block to scientific progress, Canguilhem allowed that an obstacle may have unexpected value for scientific development. Thus eighteenth century vitalism, while ‘entirely wrong’ in its basic principles, nevertheless served as a reminder that organisms cannot be fully reduced to a model derived from physics and chemistry. In addition, as we have seen, the theory of vitalism was responsible for generating the concept of the reflex.

29Perhaps because of his thorough historical research into concepts, Canguilhem evinces a more sympathetic perspective on past scientific knowledge than does Bachelard. Canguilhem details, in his history of the reflex, the intricate ways in which scientists influence their successors (Gutting 1989: 40). In addition, he freely acknowledges continuities of knowledge across major epistemological breaks. While such breaks are often depicted as ‘global’, Canguilhem insists that:

  • 7 This reference is from the second edition of Canguilhem’s Études d’histoire et de Philosophie des S (...)

We need to know how to uncover, even in the work of a single historical figure, successive breaks and partial breaks. In a theoretical fabric, certain threads can be entirely new, while others are taken from earlier weavings. (Gutting 1989: 40)7

30Here we have a description of the mingling of past and present, the continuous and the discontinuous. (Foucault’s figure of the ‘tangle’ of temporalities, or multiplicity of time-spans, is similar to Canguilhem’s formulation.) Canguilhem is not at all reticent in speaking of the preservation of knowledge, even in the wake of epistemological breaks or ‘revolutions’ in thought: ‘The Copernican or Galilean revolutions also involved the preservation of a heritage.’ (40)

31Gutting remarks that the influence of the Bachelard-Canguilhem network on Foucault was most directly transmitted through the person of Canguilhem, who was the director of Foucault’s doctoral thesis. Whereas Bachelard’s notion of the epistemological break exerted a primary shaping force, Canguilhem’s modifications of Bachelard’s theory found their way into Foucault’s work. One of these modifications was the lessening of Bachelard’s severe distinction between science and non-science. Another, most relevant to my project, concerns the endurance of knowledge from the past. Gutting notes that in making a provision along these lines in The Archaeology of Knowledge, Foucault… follows Canguilhem in emphasizing the various ways that continuities can persist across epistemological breaks. (Gutting 1989: 53)

32Canguilhem’s history of science, then, a model absorbed by Foucault into his own historical method, is an intellectual history stressing the importance of both continuity and discontinuity.

Althusser’s Radical Breaks

33The discussion of Louis Althusser here is very brief and very specific: my sole interest is in his writings that pertain to discontinuity in intellectual history. Althusser, like Foucault a pupil of Canguilhem, may be expected to exhibit significant traces of the Bachelard-Canguilhem network in his analyses; such an expectation is realised in the following passage from Althusser’s Reading Capital, published in France in 1965.

34‘We are beginning to suspect, ’ Althusser states of the history of reason, that it is not ‘a linear history of continuous development’, nor a narrative that fulfils a ‘teleological hope for the religious triumph of reason.’ Rather, he asserts, we are beginning to conceive:

a history punctuated by radical discontinuities (e.g., when a new science detaches itself from the background of earlier ideological formations), profound re-organizations which… inaugurate with their rupture the reign of a new logic, which, far from being a mere development, the ‘truth’ or the ‘inversion’ of the old one, literally takes its place. (Althusser 1970: 44, cited in Smith 1984: 86)

35The history of reason conceptualised as a succession of discontinuities; the vocabulary of ‘rupture’; the pre-eminent role of science: these are familiar ‘network’ categories. Steven B. Smith remarks that Althusser borrowed the notion of the epistemological break from ‘his mentor’Bachelard (Smith 1984: 87). Smith situates Althusser within the ‘new rationalism’ school founded by Bachelard and Alexander Koyré, whose contextualist histories of science concentrated on the seventeenth-century revolution in science (Althusser is listed as one of their ‘followers’, along with Canguilhem and Foucault). For Smith, the significance of this new rationalism for Althusser was twofold: science is seen as ‘a series of continual historical caesuras’; science has a creative or ‘constructive’ aspect – it does not merely reflect and measure the natural world, but is a ‘manufactured object’, an interpretive framework that is always liable to change (Smith 1984 88-89).

36Althusser deployed these concepts within the specific domain of Marxist theory, but his approach, and his perceived target and opposition, were similar to Foucault’s. Smith notes that Althusser’s theory of the problematic – the underlying ‘unitary structure’ beneath thought or ideology – was devised as a counter to humanist and teleological accounts of writers or ideas. The conventional history of ideas is again rebuked for its simplistic and teleological tracing of abstract ideas. The problematic, by contrast, ‘brings out within the thought the objective internal reference system of its particular themes’, and (in a phrase reminiscent of Canguilhem), it reveals ‘the system of questions commanding the answers given by the ideology’ (Althusser 1969: 67, cited in Smith 1984: 84).

37Althusser employed the concept of the epistemological break most notably in his conflict with Marxist humanists over the status of Marx’s writings. This debate was centred on the question of continuity or discontinuity within Marx’s thought. Whereas the humanist theorists argued for a continuity between Marx’s early writings and his later more structural analysis of capital, Althusser refutes this continuity, identifying instead an epistemological break in Marx’s work (specifically in the year 1845 and the announcement of the ‘new’ Marx in the German Ideology and the Theses on Feuerbach). For Althusser, this epistemological break, marking the birth of historical materialism, separates Marx from both his predecessors and his earlier self. There are for Althusser two Marxes: the early ideological humanist Marx and the post-1845 ‘scientific structuralist’ Marx (Smith: 92).

38The opposing humanist interpretation claims a continuity by pointing to later terms such as ‘fetishism of commodities’, which refer back to early concepts like ‘alienation’ (Smith: 84); Althusser dismisses this interpretation with a Canguilhem-style argument. Any supposed continuity in this case is due, he claims, to a superficial and ‘unscientific’ reading of certain terms, without considering the differences in the underlying problematics (Althusser 1969: 55-57, cited in Smith: 84). A deeper analysis will reach to the problematic that endows these concepts with their meaning. Peter Dews comments that the epistemological break functions in this context as an antidote to a Hegelian reading, which would see the later Marx as the ‘truth’ of the young Marx; the teleological interpretation is countered by the breaking of Marx’s thought into two different problematics (Dews 1994: 119).

39However, Althusser’s imposition of a discontinuity within Marx’s writings generates some very serious problems for his analysis. The alleged break is so severe, so abrupt (located in a single year) and so definitive, that it will be weakened by any evidence of continuity across that break. And as Smith remarks, such evidence is not difficult to find. Many of the earlier concepts used by Marx survive into his later writings; the Grundrisse contains a humanist perspective on productive relations that, according to Althusser, should have disappeared (Smith: 92). Smith’s account of Althusserian theory is sympathetic, yet he struggles to justify Althusser’s commitment to this break in Marx, which appears ‘merely arbitrary or imposed’ (86). Perhaps, Smith offers, Marx continued to use an older, inappropriate terminology because he was ‘unconscious’ of the deeper workings of the problematic. This attempted rationale is at least consistent with Althusserian theory, with its elimination of the ‘constitutive subject’ within the ‘mechanism’ of a theoretical structure undergoing ‘mutation’ (Smith 1984: 87). But as an explanation it is far from convincing; as a rescue mission for Althusser’s reading of Marx it arrives too late and with too few resources.

40Althusser has doomed himself through his over-zealous commitment to the epistemological break, as Smith recognises. It is doubtful, he acknowledges, that Marx inaugurated one of the ‘great paradigmatic breakthroughs’, of the type identified by Bachelard and Kuhn in the history of science. Althusser’s Marxist belief in historical materialism as a ‘science’ has encouraged him to make such a claim – but it cannot be sustained even in an analysis of Marx’s corpus. Althusser wants to find a discontinuity marking off this new ‘science’ from previous humanist philosophy, including that of the young Marx. But this proposition can only be advanced by repressing great continuities of thought. As Smith remarks, Marx

no doubt recast Hegelian modes of thought or classical economics in new and unforeseen ways, but can still be seen as carrying on a distinct tradition that he inherited from the past. (91)

41Althusser’s positing of such a severe discontinuity is of course politically motivated: it was his weapon against the Marxist humanists. But he gives so much of the ground to radical discontinuity that historical continuity disappears from view. The contrast with the non-Marxist perspective of Foucault is striking:

At the deepest level of Western knowledge, Marxism introduced no real discontinuity… Marxism exists in the nineteenth century like a fish in water, that is, it is unable to breathe anywhere else. (Foucault 1973: 261-262, cited by Smith 1984: 91)

42In his privileging of the break and the rupture, Althusser exhibits none of the moderation or caution found in Canguilhem, for whom any theoretical fabric will contain some ‘extremely new threads’, while others ‘are taken from earlier weavings.’ (Gutting 1989: 40) Instead, Althusser’s reductive analysis clothes the historical Marx in two entirely different sets of threads, one cast off and replaced by the other in 1845, the two somehow having no material in common.

The Fate of Kuhn’s Theory of Scientific Revolutions

  • 8 These remarks, made by reviewers from, respectively, Science and The Times Literary Supplement, are (...)

43Althusser’s severely discontinuist rendering of intellectual history operated within a particular discursive field, that of Marxist theory. Thomas Kuhn’s similarly discontinuist approach functioned within another specific discourse, that of the philosophy of science; yet Kuhn’s work has exerted a far greater influence, across a broader range of interests, than Althusser’s. Kuhn’s The Structure of Scientific Revolutions, first published in 1962, has had an extraordinary career. Written as an essay on science’s historical development, this book has set off reverberations far beyond its intended domain. Its notion of paradigm shifts and ‘progress through revolutions’ has inspired developments in cultural studies, intellectual history, and other disciplines. It has been critically acclaimed as ‘a landmark in intellectual history’, even one of ‘The Hundred Most Influential Books Since the Second World War’.8

  • 9 In an interview with John Horgan, for the latter’s book The End of Science, Kuhn described his disd (...)

44Kuhn himself was ‘puzzled’ by his book’s widespread influence (1996: 208); he even professed dismay at the way his ideas had been taken out of their field.9 From a cultural studies perspective, this ‘puzzlement’ on the part of the author at the wayward path taken by his inventions poses no problem; indeed it accords with a conception of culture (in large part shaped by Foucault) in which the flow of texts and ideas is elevated, while the intention of mere authors is devalued. The Structure of Scientific Revolutions is in fact a classic case, if not of the ‘death of the author’, then at least of the impotence of the author in delimiting the significance of his work. Its impact has been so dispersed that it has become, against its author’s will, the classic study of discontinuity in intellectual history, rather than the specific essay on the history of science intended by Kuhn.

45It could be argued that Kuhn’s model of paradigm change within the history of science has exerted more influence – when generalised as a description of intellectual history – than the work even of Foucault, as well as of the other theorists discussed in this chapter – at least within the Anglophone world. Partly this is due to publication history: Kuhn’s work appeared a decade before The Order of Things was translated into English. Another factor, noted in Chapter One, is the general ignorance within the English-speaking world of the epistemology of science tradition in France, a tradition that includes the works of Bachelard and Canguilhem. This meant that Foucault’s works of intellectual history, featuring epistemes, ruptures and breaks, were received by Anglophone readers in the context of a prior familiarity only with Kuhn’s theory of paradigm shifts.

  • 10 The only apparent commonality as an intellectual influence on the two theorists is Aexander Koyré’s (...)

46One consequence of this was the inevitable comparison of these two theoretical models. Both are theories of historical discontinuity published in the 1960s; they both insist on underlying deep structures governing intellectual activity; they both conceptualise intellectual history as punctuated by abrupt changes, after which the deep structure is fundamentally altered.10 Dreyfus and Rabinow emphasise at least one aspect of this commonality:

Allowing for differences in their interests and fields of investigation, Foucault would presumably agree with Kuhn’s remark that ‘Rules… derive from paradigms, but paradigms can guide research even in the absence of rules.’ (Dreyfus and Rabinow 1983: 199)

47It is their other common aspect – the conceptualisation of intellectual history as a series of leaps from one epistemological configuration to another – that is my chief concern here. Kuhn and Foucault have been grouped together along these lines as key influences on contemporary critical theory. As one example of this tendency, an extract from The Structure of Scientific Revolutions is included in Thomas Docherty’s Postmodernism: A Reader as a ‘founding proposition’ of postmodern thought (Foucault is also, of course, represented in this Reader). After summarising the theory of the paradigm shift and its ‘enormous influence across all fields of knowledge’ Docherty declares of Kuhn’s book:

It is itself a symptom precisely of a paradigm shift in the field of knowledge and philosophy, away from a model which proclaimed the availability of ‘truth’ towards one which proclaims instead the much more modest ‘pragmatic usefulness’. (Docherty 1993: 36)

48Here Kuhn’s work is celebrated as more than an influential text: it is seen to represent in itself a fundamental shift of knowledge. As we have seen, Foucault’s works have themselves been regarded in this way, as ‘path-breaking’ work, responsible for ‘transforming our understanding of vast swathes of Western intellectuality.’ (Jenkins 1998: 117) Docherty continues in his recruitment of Kuhn to the critical theory cause:

Cultural criticism at this moment has begun to go ‘relativist’, so to speak… after 1968, all such knowledges begin to be deemed ‘local’ and specific to the pragmatic necessities of the specific culture from which the knowledges emanate and whose interests they serve. (Docherty: 36)

49This is an instance of the appropriation of his thought of the kind that so dismayed Kuhn. Docherty has inscribed Kuhn’s book into a mini-narrative of postmodern epistemology, complete with the radical turning-point of 1968. Kuhn’s ideas are seen to feed into a relativising of knowledge, described here in distinctly Foucauldian terms. In instances such as this, one can feel some sympathy for Kuhn as the author who has lost control of his ideas. What has been disregarded, in this enlistment of Kuhn into critical theory ranks, is the specificity of his work – that is, as a study of the discipline of science – on which he is insistent throughout The Structure of Scientific Revolutions. Indeed, Kuhn explicitly denies, in the 1969 Postscript, that his position is relativistic, on the grounds that although he does not believe that science is progressing towards ‘truth’, he nevertheless is ‘a convinced believer in scientific progress.’ (1996: 206) One cannot find all knowledges relatively valid if one believes, as does Kuhn, that ‘ [l] ater scientific theories are better than earlier ones for solving puzzles… ’ (206).

50Kuhn’s pleas have, for the most part, gone unheard. The cultural forces sweeping his theory into other whirlpools of thought have been too strong. Something of their momentum can be gleaned from Hayden White’s invocation, published in 1978:

We require a history that will educate us to discontinuity more than ever before; for discontinuity, disruption, and chaos are our lot. (White 1978: 50, cited in Himmelfarb 1999: 85)

51This will to discontinuity, whether in the name of postmodern history, cultural theory, or other endeavours, has enshrined Kuhn’s book as one of its ‘founding propositions’. For my purposes, a closer critical reading of Kuhn is necessary, to determine the tenor of his discontinuous account of science history.

Paradigms and Revolutions

52In the Postscript to The Structure of Scientific Revolutions, Kuhn neatly summarises his thesis: ‘the book portrays scientific development as a succession of tradition-bound periods punctuated by non-cumulative breaks… ’ (1996: 208). These ‘non-cumulative breaks’ are the revolutions in scientific knowledge through which, Kuhn argues, science progresses. For Kuhn, the history of science is characterised by leaps from one paradigm of knowledge to another. In positing this theory, Kuhn argues against the dominant model of scientific development, which he calls ‘cumulative’ – that is, the gradual ‘accumulation of individual discoveries and inventions.’ (2)

53This ‘cumulative’ conception of intellectual history is supported, in the discipline of science, by a teleological perspective. Scientists are trained to see the ‘discipline’s past developing linearly towards its present vantage.’ (138) Kuhn finds this teleological imperative present in other disciplines besides science: ‘The temptation to write history backward is both omni-present and perennial.’ (138) Yet science is especially affected by this tendency because of the manner in which scientific knowledge is taught. Science textbooks distort the history of science, selecting the work of earlier scientists which can be shown to lead up to the present state of knowledge. The past of science is constructed according to the ‘set of fixed canons that the most recent revolution in scientific theory and method has made seem scientific.’ (138) This means that ‘textbooks and the historical tradition they imply have to be rewritten after each scientific revolution.’ (138) It also means that the revolutions themselves remain largely ‘invisible’, since they come to form the status quo towards which the history of science is thought to proceed.

  • 11 Kuhn shares this conceptual imprecision with Foucault, who likewise has been criticised for the loo (...)

54Kuhn defines ‘paradigm’ at numerous points throughout the book. He first declares paradigms to be ‘universally recognized scientific achievements that for a time provide model problems and solutions to a community of practitioners’ (1996: x). A little later he refers to paradigms as accepted scientific practices that form models, ‘from which spring particular coherent traditions of scientific research.’ (10) The bifurcation here between the actual practices and the ‘models’ they come to form remains a source of minor confusion throughout the book; it is one instance of the work’s theoretical fuzziness, which has drawn much critical fire.11

55In his Postscript, written largely in response to the book’s many critics, Kuhn acknowledges the vagueness with which he uses the term ‘paradigm’; he mentions one sympathetic reader who compiled ‘at least twenty-two different ways’ in which the term is employed (181). By way of clarification, he proposes two different meanings of ‘paradigm’. The first is ‘the entire constellation of beliefs, values, techniques and so on shared by the members of a given community.’ (175) The second is ‘one sort of element’ in that constellation, namely ‘the concrete puzzle-solutions’ of scientific practice that can replace existing rules for scientific puzzle-solving (175). This second, more restricted meaning relates to those ‘exemplary past achievements’ of science which become accepted as models for scientific activity. The first, more expansive or ‘sociological’ (175) meaning, is the definition of paradigm usually implied when Kuhn’s theory is generalised to the broader domain of intellectual history.

56Kuhn argues that paradigms (it is not always clear which of the two meanings applies) support the activities of ‘normal science’. For Kuhn ‘normal science’ is the everyday research activity of scientists; it includes ‘puzzle-solving’ and ‘mopping-up operations’ (24). Much of this activity concerns technical refinement or the solving of minor problems thrown up by accepted scientific theories. Normal science ‘does not aim at novelties of fact or theory’ (52); rather it reconciles newly discovered phenomena to the accepted theoretical framework. Kuhn proposes that normal science proceeds in this way under the rubric of its paradigm – until that paradigm begins to break down.

57A crisis will appear, when new research or technical innovation generates problems that cannot be resolved within the existing paradigm. Kuhn offers as an example the late nineteenth century crisis in physics, when experimental research into light and its transmission failed to produce the findings expected within the dominant Newtonian paradigm (72-74). Something was ‘going wrong with normal research’ in physics (93). The stage was set for a scientific revolution, because the old paradigm was no longer adequate. For a growing number of physicists, the time was out of joint; the physics community was, like an unstable state, ready for a revolution. Kuhn is explicit in paralleling scientific with political revolutions: In both political and scientific development the sense of malfunction that can lead to crisis is prerequisite to revolution. (92)

58Much of the impact of Kuhn’s theory stems from his emphatic use of the word ‘revolution’, with its connotation of violent upheaval. Scientific revolutions, like political ones, are destructive of the old order, as well as productive of a new regime (97). Science for Kuhn proceeds by revolution, not evolution. Einstein’s Special Theory of Relativity of 1905 signalled one such revolution, one of the most dramatic in the history of science. Its historical setting was the serious crisis in physics, to which it proposed a radical solution. Relativity offered itself as a new paradigm. For Kuhn, it became the dominant paradigm of physics in the same way that a new political force displaces the old order – by winning the support of the relevant community: As in political revolutions, so in paradigm choice – there is no standard higher than the assent of the relevant community. (94)

59Kuhn’s analogy is surelyfaulty here: political revolutions aregenerally imposed by force, with or without the assent of the community. A scientific paradigm, by contrast, is accepted because of its explanatory power, as Kuhn suggests. If it is found to be proven by stringent testing, if it provides a satisfactory explanation of observed phenomena, then it will win ‘converts’ from the old paradigm, with which it is incompatible (19). The new paradigm then radically changes the field. It produces a shift ‘in the problems available for scientific scrutiny’; it redefines the standards of ‘what should count as an admissible problem or as a legitimate problem-solution.’ (6)

60Scientific revolutions, then, occur when one paradigm displaces another. This is achieved (unlike most political revolutions) by the intellectual persuasion of a requisite number of professionals. The new paradigm is set, Kuhn observes, when it is able ‘to attract most of the next generation’s practitioners’ (18). Of course, there will be dissenters who reject the new paradigm and cling to the old, but for Kuhn these individuals ‘are simply read out of the profession, which thereafter ignores their work.’ (19) Such a brutal rejection of the old regime and its followers does suggest a parallel with political revolutions, as does the controversy that accompanies a paradigm shift. The final valid parallel suggested by Kuhn is the way a scientific revolution, like a political one, reconfigures the imagination (whether scientific or political), so that there is a ‘transformation of the world within which scientific work [is] done.’ (6) Kuhn goes further, proposing that a paradigm change engenders a new world-view: ‘after a revolution scientists are responding to a different world.’ (111) Kuhn is particularly struck by the rapid developments in astronomy following the proposal of Copernicus’ paradigm:

The very ease and rapidity with which astronomers saw new things when looking at old objects with old instruments may make us wish to say that, after Copernicus, astronomers lived in a different world. (117)

61Kuhn’s argument is at its most persuasive when applied to ‘major turning-points in scientific development’ (6), namely those revolutions of thought associated with Copernicus, Newton, Lavoisier (the eighteenth century ‘chemical revolution’) and Einstein. Whereas normal science is ‘tradition-bound’, these revolutions are ‘tradition-shattering’. They both destroy the old paradigm and, by forging a ‘shift of professional commitments’, they become the new paradigm (6). They are the ‘extraordinary episodes’ by which science progresses.

Problems with Kuhn’s Model

62If, as suggested by Docherty, Kuhn’s book represented its own paradigm shift, then it fulfilled at least one of Kuhn’s criteria for a revolution in thought: controversy. It was heavily criticised within the scientific community on a number of grounds. Karl Popper, to cite one eminent example, accused Kuhn of ‘glorifying subjectivity and even irrationality’ (Kuhn 1996: 186). Such criticism did not of course prevent the book exerting enormous influence outside the scientific world. Interestingly, Kuhn suggests in his Postscript that if his book’s theses have found a ‘wide applicability’, that should provoke no surprise, in that the theses themselves are ‘borrowed from other fields.’ (208) He names the histories of literature, music, the arts and politics as featuring a ‘ [p] eriodization in terms of revolutionary breaks in style, taste, and institutional structure’; his only originality, he proposes, is in applying such a model to the sciences, ‘which had been widely thought to develop in a different way.’ (208) It is perhaps testament to the peculiar and unpredictable dynamic of culture that Kuhn’s thesis on the history of science, inspired by histories of culture and politics, should then be appropriated as a general thesis on intellectual history, thereafter influencing thought within the humanities.

  • 12 Kuhn lists as characteristics specific to science: the ‘relative scarcity of competing schools in t (...)

63Kuhn is adamant to the very end of the Postscript that his book relates to the history of science only; indeed it demonstrates, he insists, that science is ‘strikingly different’ to other disciplines.12 Yet as noted earlier, cultural progeny may disobey their creators; they may be adopted by other guardians; they may be stolen; they may be set to work in fields far from home. This has been the fate of Kuhn’s theses: they have been accepted, judged, lauded and criticised as pertaining to intellectual history in general.

64One of the major criticisms made of Kuhn’s discontinuist theory of science history has been articulated by Christopher Norris. If one accepts Kuhn’s theory, Norris argues, it becomes ‘difficult – if not impossible – to explain how we could ever gain insight into scientific views other than our own’ (Norris 1997: 486). This criticism has also been levelled at other discontinuist models of intellectual history (indeed Norris makes a very similar criticism of Foucault.) In this section, however, I wish to concentrate specifically on Kuhn’s text. By examining any inconsistencies or weaknesses in his argument, I aim to consider the validity of Kuhn’s case for a discontinuist history of scientific knowledge.

65One potential weakness in Kuhn’s account stems from its point of greatest strength: the use of famous breakthroughs in science to exemplify scientific revolutions. Copernicus and Einstein personify major conceptual upheavals – but are they representative of all revolutions in science? Could they in fact be rare exceptions to the norm in the development of science? If they are indeed representative, what then of smaller, less famous breakthroughs? How big must a conceptual change be to constitute a revolution?

66Kuhn is aware of this difficulty, addressing it directly both in the original text and in the Postscript. In the latter, he remarks that some readers’ conclusion has been that he was concerned only with major revolutions in science, on the scale of Newton or Einstein. Kuhn attributes this misunderstanding to the choice of his examples and to his ‘vagueness about the nature and size of the relevant communities’ (180). He proceeds to define revolution as ‘a special sort of change involving a certain sort of reconstruction of group commitments.’ (181) This change, however, need not be large in any way, nor need it ‘seem revolutionary to those outside a single community, consisting perhaps of fewer than twenty-five people.’ (181) A scientific revolution, then, may be small indeed, even largely unknown outside a specific group of scientists.

67How often do such small-scale revolutions occur? Frequently, according to Kuhn, even ‘regularly’:

It is just because this type of change, little recognized or discussed in t he literature of the philosophy of science, occurs so regularly on this smaller scale that revolutionary, as against cumulative, change so badly needs to be understood. (181)

68But this creates another difficulty. If these little revolutions are so regular, can they not be regarded as increments, little steps on the pathway of science? Perhaps one’s perspective on the history of science amounts to a question of emphasis: Kuhn is simply naming as ‘revolutions’ what other observers might call steps or incremental gains.

69Kuhn refutes the incremental interpretation early in the book. For him a new theory brings a fundamental challenge to accepted belief, by implying ‘a change in the rules governing the prior practice of normal science.’ (7) Because it reflects upon much completed scientific work, the new theory will be resisted by some, and its acceptance will necessitate a major re-ordering. Therefore, Kuhn asserts, a new theory, however special its range of application, is seldom or never just an increment to what is already known. (7)

70The assimilation of a new theory, according to Kuhn, requires too much ‘reconstruction’ of the field for this process to represent a mere increment within a cumulative historical development. Assimilation is an ‘intrinsically revolutionary process’; it is ‘an extended process’ – rather than an incremental step – in which ‘prior fact’ is re-evaluated (7).

71Mention of an ‘extended process’, however, introduces another problem. Kuhn acknowledges a certain time-span in which the new theory is assimilated, although he fails to elaborate on this period in the book. Does this assimilation of new theories during an ‘extended process’ undermine the notion of revolutions in thought? Perhaps again it is a matter of emphasis in interpretation. This ‘assimilation’ could be seen not as the aftermath of revolution, but as the absorption of a new theory into the scientific field. If significant changes are as regular as Kuhn himself professes, perhaps scientific development should be regarded as a continual process of growth through absorption. The metaphor would not then be of an unstable state racked by upheaval, but of an organism expanding through the assimilation of new elements. This of course would be to replace Kuhn’s discontinuist model with a continuist one, no doubt to his displeasure. Yet the vagueness and inconsistencies within Kuhn’s own argument undermine his thesis.

72Kuhn is vague as well concerning the advent of scientific revolutions. He suggests that individual scientists, struggling with the crisis gripping their field, somehow give birth to a new paradigm. This may occur ‘all at once, sometimes in the middle of night, in the mind of a man deeply immersed in crisis.’ (90) The exact process, however, remains ‘inscrutable’, possibly ‘permanently so’ for Kuhn (90). This inscrutability at the core of abrupt conceptual change parallels the ‘enigmatic’ nature of epistemological breaks described by Foucault in The Order of Things. Both theorists suffer a failure of articulation at the point of describing the very discontinuities on which their theories are founded. Kuhn can only offer one observation by way of explanation:

Almost always the men who achieve these fundamental inventions of a new paradigm have been either very young or very new to the field whose paradigm they change. (90)

73Elsewhere, Kuhn depends on the abilities of individuals, especially individual ‘geniuses’, to explain the sudden leap from one paradigm to another. But even if one allows a role for the inspiration of extremely talented individuals – such as Galileo or Einstein – Kuhn’s recourse to the concept of genius is unconvincing. In discussing the breakthrough made by Galileo, who ‘saw things differently from the way they had been seen before’ (119), Kuhn asks: ‘Why did that shift of vision occur? Through Galileo’s individual genius, of course.’ (119) Yet in clarifying the operation of this genius, Kuhn conjectures: what seems to have been involved was the exploitation by genius of perceptual possibilities made available by a medieval paradigm shift. (119)

74This notion of genius is so vague that it contributes very little to Kuhn’s argument, while provoking new questions. Does genius of some sort invoke all paradigm shifts – or just the famous ones? If talented individuals are responsible for revolutions, how does a particular individual come to be ‘the one’? The focus on individual agency works to the detriment of Kuhn’s thesis here.

75The most serious difficulty with Kuhn’s argument, however, concerns his insistence on the destructive nature of scientific revolutions. Kuhn argues that because the new paradigm will be incompatible with the old, the acceptance of the new invalidates its predecessor. In pressing this point, Kuhn allows no possibility of overlap, or of the co-existence of different paradigms. In discussing the relation between Einsteinian and Newtonian dynamics, Kuhn avers:

From the point of view of this essay these two theories are fundamentally incompatible in the sense illustrated by the relation of Copernican to Ptolemaic astronomy: Einstein’s theory can be accepted only with the recognition that Newton’s was wrong. (98)

76Here we have the extreme position to which Kuhn’s discontinuist history leads him. Kuhn is adamant that the old paradigm is simply erased – or ‘destroyed’ – as a functional scientific model by the new. There is no persistence of the past in the present; the old may not continue to function, even partially, alongside its usurper. Kuhn admits that his is a minority view, especially in the case of Newton and Einstein (98); he examines the objections raised against it.

77The first objection, and surely the most pertinent, is summarised by Kuhn:

Relativistic dynamics cannot have shown Newtonian dynamics to be wrong, for Newtonian dynamics is still used with great success by most engineers and, in selected applications, by many physicists. (99)

78This objection is so compelling that Kuhn needs to address it directly; unfortunately, he does not. Instead, he adds a ‘furthermore’ to the objection, thereby situating it within an argument from ‘early logical positivism’, which he proceeds to refute. The logical positivists, Kuhn argues, sought to

restrict the range and meaning of an accepted theory so that it could not possibly conflict with any later theory that made predictions about some of the same natural phenomena. (98)

79In relation to the Newton – Einstein issue, the logical positivist ‘furthermore’ is, according to Kuhn, that even in Einsteinian terms, Newton’s equations ‘will be as good as our measuring instruments’ in some conditions (99), so that Newtonian theory may be derivable from Einsteinian theory.

80Kuhn spends several pages dismantling the claim that Newton’s dynamics can be ‘derived’ from relativistic dynamics – a proposition which he regards as a theoretical sleight-of-hand. But he neglects the more general, and more relevant, objection – that Newtonian dynamics is still used with great success in both engineering and physics. This seems a valid claim, one which allows the old paradigm to co-exist, in certain conditions, with the new one. It is a fundamental weakness of Kuhn’s thesis that he does not satisfactorily engage with this objection. Instead, after tussling with the logical positivists for a few pages, he somehow concludes: Let us, therefore, now take it for granted that the differences between successive paradigms are both necessary and irreconcilable. (103)

81A disgruntled reader may at this point concur that this must be ‘taken for granted’, because nothing conclusive has been proven. It seems that Kuhn is so determined to view science history as ‘progress through revolutions’ that he cannot admit the possibility of the old enduring alongside the new. The scientific revolution is, he insists, ‘the displacement of the conceptual network through which scientists view the world.’ (102) An alternative view would be that this ‘network’ contains older paradigms that continue to function under certain conditions: they are not simply, as Kuhn argues, deleted from the network in favour of new paradigms.

82Kuhn represses alternative interpretations of science history elsewhere in his book. In discussing the late nineteenth century crisis in physics, he acknowledges that

[one] root of that crisis can be traced to the late seventeenth century when a number of natural philosophers, most notably Leibniz, criticized Newton’s retention of an updated version of the classic conception of absolute space. (72)

83By labelling Leibniz’ critique a ‘root’ of the much later crisis, Kuhn gestures at a continuity of thought spanning at least two centuries. He allows that Leibniz and others almost destabilised concepts of the absolute in Newton, and that they ‘hinted’ at the appeal of a relativistic model (72). But Kuhn quickly cancels any prospect of continuity. He insists that because their critique was ‘purely logical’, without reference to the application of Newton’s theory, ‘their views died with them’ in the early eighteenth century, to be ‘resurrected’ only in the late nineteenth (72-73).

84Kuhn’s treatment of this ‘anticipation’ is consistent with his thesis: it took a paradigm crisis before the earlier critique of Newton could find a sympathetic place. But his introduction of the ‘root’ to the crisis suggests that another interpretation is possible. Kuhn’s own account of the background to the crisis in physics includes the matter of technical problems arising from the wave theory of light from around 1815; it also includes the paradoxical contribution of Maxwell, whose electromagnetic theory, intended to support Newtonian theory, instead undermined it (73). Given these and other contributing factors – both theoretical and technical – stretching over decades, the period from Leibniz to Einstein may conceivably be regarded as a continuum: a long, gradual process culminating in the acceptance of a non-Newtonian theory of physics.

85Kuhn, of course, refuses continuity; he sees instead paradigm/ paradigm in crisis/ new paradigm. But by his own account of this example, there was a build-up of a number of contributing factors from 1815, with a philosophical background dating to the late seventeenth century. Kuhn’s history of science, in highlighting paradigm change, finds a role for such contributing factors only at the moment of paradigm crisis, when suddenly they seem to converge. The gradual development of thought, the influence felt across diverse fields and disciplines, are repressed by Kuhn in favour of a model in which ‘inscrutable’ scientific revolutions leap up and destroy the older, mortally wounded, paradigm. One is entitled to suspect that Kuhn has overlooked great tracts of scientific development, full of nuanced and interlocking connections, in order to construct a dramatic theory of crises and revolutions.

86Kuhn’s theoretical model is uncompromising in its discontinuity. He seems to grant no place whatsoever to the persistence of the past in the present. John Horgan succinctly summarises Kuhn’s theory of scientific revolution: ‘The proposer of a new paradigm stands on the shoulders of giants (to borrow Newton’s phrase) and then bashes them over the head.’ (Horgan 1997: 43) Kuhn’s commitment to the destructive aspect of scientific revolutions forbids even the absorption of previous knowledge or techniques. The scientific knowledge of the past is simply swept away once it has become outmoded. Kuhn’s model has none of the subtlety of the French histories of science, of Bachelard and Canguilhem. Where those theoretical models allow for continuities and gradual transformation alongside epistemological breaks, Kuhn permits nothing of the past to endure. For this reason he is more difficult to defend and easier to criticise. No one doubts that relativity theory inaugurated a new world for physics (Bachelard announced as much in the 1920s), but few will agree with Kuhn that Newtonian dynamics were dropped into the rubbish bin of physics at that moment.

  • 13 Steven B. Smith: ‘While there may be profound differences between, say, Newton’s and Einstein’s sys (...)

87By insisting on a complete and radical break, Kuhn also overlooks more basic continuities within scientific thought. Steven B. Smith and Anthony Wilden both point to the fundamental continuity linking the thought of scientists such as Newton and Einstein. It is the continuity provided by the ‘scientific enterprise itself’.13 Many scientists conceptualise their discipline’s history as a continuity, or as a dialogue with past scientists, even across paradigm shifts. Stephen Hawking, for example, found it significant that he was born on the anniversary of Galileo’s death and that he held the Cambridge chair once occupied by Newton. Hawking’s version of scientific change does not accord with Kuhn’s, even with regard to the Einsteinian ‘revolution’:

In practice, what often happens is that a new theory is devised that is really an extension of the previous theory… Einstein’s general theory of relativity predicted a slightly different motion from Newton’s theory… However, we still use Newton’s theory for all practical purposes because the difference between its predictions and those of general relativity is very small in the situations that we normally deal with. (Hawking 1988: 10)

88It seems that Hawking is unwilling to bash the giant Newton over the head, as Kuhn’s theory would have him do.

89Kuhn’s annihilation of the past in his model is puzzling in that he does not, as noted earlier, also eliminate the notion of progress from the history of science. While ‘relinquishing’ the idea that ‘changes of paradigm carry scientists and those who learn from them closer and closer to the truth’ (170), he nevertheless maintains that science represents a ‘process of evolution from primitive beginnings’ (170), and that later scientific theories continually improve on their predecessors (206). The history of science is for Kuhn a non-teleological evolution – that is, a progress ‘without benefit of a set goal, a permanent fixed scientific truth’ (173).

90Kuhn’s curious model thus not only eradicates the end-point of truth; it dispenses with the past as well. Science history as a ‘progress through revolutions’ is a process of jumps from one paradigm to the next, each one re-writing the field, each one erasing its predecessor. It cannot be said that Kuhn presents a convincing case to support his model as a general description of science history. Even those major scientific revolutions – those of Copernicus and Einstein – that most forcibly support his argument are, as he admits, ‘famous episodes in scientific development that have often been labelled revolutions before.’ (6) The smaller ‘revolutions’ and paradigm jumps which, he claims, constitute much of the history of science may, more plausibly, be conceptualised as the incremental measures he wants to renounce. Perhaps Kuhn’s model may have been stronger if, like Bachelard, Canguilhem and Foucault, he had assigned some role to continuity or gradual change. Instead, his ruthless repression of continuity, in favour of a series of discontinuous breaks or revolutions, leaves his model seriously unbalanced.

The Picture of Science

91Most of this chapter has dealt with some notable issues in the history and philosophy of science. It may be argued that the propositions and debates within this context are specific to science, and cannot be extended to intellectual history in general. But there is ample evidence to the contrary. As already noted, Gutting’s exposition of Foucault’s archaeology of knowledge places that project firmly within the French tradition of the history of the science. As also noted, Kuhn’s study of scientific revolutions escaped its intended confines, to become one of the most influential works within the study of intellectual and cultural history.

92Peter Watson’s 2001 book A Terrible Beauty: A History of the People and Ideas That Shaped the Modern World, an intellectual history of the twentieth century, defines its mammoth subject primarily in terms of science. For Watson, twentieth-century thought was marked by a profound discontinuity: it was different from all that came before, largely because of its radically new science. To support his thesis, Watson quotes Claude Levi-Strauss who wrote, in 1988, that philosophy has a role in the contemporary world, but ‘only if it is based on the current state of scientific knowledge and achievement’ (Levi-Strauss 1988: 119, cited in Watson 2001: 2). Levi-Strauss here connects with the epistemological theory of Bachelard, enunciated much earlier in the twentieth century. Levi-Strauss adds that science has ‘revolutionised the rules by which the intellect operates.’ (2) This declaration again echoes Bachelard, but also Kuhn, with his refrain of revolution and change.

93If a picture of scientific change has been imprinted on the study of intellectual history, the question remains: what kind of picture is it? Is it purely discontinuous, a series of epistemological breaks by which science invigorates itself? Is it, as Kuhn insists, a picture of progress through revolutions, by which old knowledge is overthrown and replaced by its robust successor? For Watson, Kuhn ‘changed for all time how science was perceived’ (Watson 2001: 488), and he is not alone in modelling intellectual history on Kuhn’s version of science history. But is there another picture of science, one with a place for tradition and continuity of thought?

94I commenced this chapter with three questions, to which I now return. First, do these theories of science history contain a significant role for continuity, one that has perhaps been undervalued or overlooked? In the case of Bachelard and Canguilhem, the answer is affirmative. Although Bachelard’s main emphasis is on the epistemological break – the ‘philosophy of no’ with which a new science announces itself – he nevertheless builds into his model a mechanism by which past knowledge is preserved. Previous scientific knowledge, deemed valid and useful by current practice, is ‘sanctioned’ and maintained in a dialectical process of correction and absorption. Bachelard’s theory of the epistemological break is certainly much better known – and has exerted far greater influence – than his theory of the preservation of the past, but the latter is a significant component of his model. Canguilhem allocates a greater role than does Bachelard to past scientific knowledge, even to ‘the preservation of a heritage.’ His model is in several ways subtler than Bachelard’s, most apparently in his conception of scientific knowledge as a mingling of the contemporary and the past. As has been noted, this conception was adopted by Foucault, although it has generally been overlooked in the latter’s work.

95Secondly, do any of these models distort the picture of historical development, by repressing the role of continuity? I propose that Althusser and Kuhn belong to this category. Both are proselytisers for the cause of radical discontinuity, both banish continuity of thought from their models. Althusser takes the epistemological break from Bachelard’s history of science and applies it zealously within the context of Marxist theory. His conviction that Marxism is a science is untenable to non-Marxists, but even fellow Marxists and sympathisers have been unconvinced by his imposition of a radical break in Marx’s thought. Kuhn’s highly influential model was not meant to serve a political cause, but his theme of ‘revolutions’ in scientific knowledge is ruthless in its treatment of the past. Discontinuity rules in Kuhn; a scientific revolution is destructive of past knowledge, which is obliterated and replaced by a new regime, or paradigm.

96The theories of Althusser and Kuhn were published and promoted in the 1960s, as were those works of Foucault most heavily associated with discontinuity in intellectual history. Hence the third question: can these theoretical models be regarded as products of that disruptive and ‘revolutionary’ decade? This is obviously a matter of interpretation, but such a viewpoint has much to recommend it. It is a view held by Vladimir Biti in his appraisal of Foucault’s 1960s publications:

Foucault neglected the involvement of the present with the past probably for the reason that this concept of periodization based on breaks and upheavals was developed during the revolutionary sixties. (Biti 1999: 178)

97For Biti, a theory of historical discontinuity has significant political ramifications, in that it posits the contingency of history. The implication of such a theory is that:

later periods do not necessarily follow from former ones… [t]herefore the present is by no means indebted to its past, it is announced free to choose and shape the past of a kind it likes or needs. (178)

98The political aspect of discontinuity, as detailed by Poster (1997) and discussed in Chapter Four, finds a sympathetic environment in the 1960s. Intellectual history as conceived of upheavals and radical breaks has a political dimension: it promotes the idea of freedom to make one’s future, liberated from the burden of the past. It was entirely appropriate for theories of discontinuity to find favour in the 1960s, a decade of counter-culture, student protests, the generation gap. A new generation exuberantly proclaimed their disrespect for the past, for the comforts of tradition, for the wisdom of elders. The word ‘revolution’ was widely used, in both political and cultural contexts; the stated aim was to redefine politics, culture, whole societies, ‘to change the world’. In such a cultural environment, continuity of thought or custom is assigned a purely conservative identity. It is seen to serve the very authority that is being challenged. The status quo is sustained by tradition; the present social order justifies itself with reference to the assurances of the past: if that status quo is the enemy, then so is tradition. Kuhn’s model of scientific ‘revolution’, and the theory of breaks and ruptures in knowledge derived from Bachelard, both found ready acceptance and wide application in such an atmosphere.

99But is this picture of breaks, leaps and revolutions a fair account of scientific development? One of the most detailed historical studies of scientific revolutions has been conducted by I. Bernard Cohen, in his book Revolution in Science (1985). Cohen remarks that occurrences of the word ‘revolution’ in the context of science have always reflected current theories concerning political and social revolution (Cohen 1985: x).

100The concept of revolution in science is derived from political revolution; for Cohen the initiating political event was the 1688 Glorious Revolution in England. Prior to that event, a scientific development such as Copernicus’ theory was not conceived as a scientific ‘revolution’ (such an evaluation was made by historians two centuries later). Cohen argues that the Glorious Revolution of1688 launched the idea of ‘revolution’ into various types of discourse; the word embodied two meanings – one of radical change, the other of ‘return’ to an earlier state of things. The sense of radical revolution was adopted into the discourse of science; it was applied to Newtonian thought in the mid-eighteenth century.

101The 1960s references to ‘revolution’ in science or knowledge activated only the radical meaning of the word; this is most apparent in Kuhn’s definition of revolution, with its annihilation of past knowledge. In an article published in 1987, Cohen contends that scientific revolution is more accurately conveyed by using both senses of the word ‘revolution’. Noting the use in French theory of terms like ‘coupure’ or rupture, he argues that such terms ‘overly stress the discontinuities in revolutions in science.’ He asserts that: In my research I have found again and again that the most revolutionary scientific ideas tend to have in them elements of older ideas… (1987b: 200).

102In his book The Newtonian Revolution: Illustrations of the Transformation of Scientific Ideas (1980), Cohen re-names the ‘revolutionary’ creative process as ‘transformation’. By using this term he hopes to encompass the twin senses of revolution – as radical innovation and return to older ideas. A revolution in science produces ‘a new set of beliefs, concepts, theories, methods which are to some degree transformations of older or existing ideas.’ (1987b: 200)

103Cohen is much closer to Canguilhem or even Bachelard, with their ‘correction’ or ‘preservation’ of past knowledge, than he is to Kuhn. For Cohen, revolution in science contains a ‘rolling back’ or return, as much as it constitutes an advance. After considering the claims of the rival ‘evolutionary’ (or incremental) interpretation of scientific development, he is less happy to accept that science progresses solely by radical breaks. Even Einstein, after all, ‘insisted that the introduction and development of relativity theory was an “evolution” and not a “revolution”.’ (203) Cohen also makes the point that the net change in knowledge may be ‘just as profound and far-reaching… when produced by evolution as when caused by revolution.’ (203) Furthermore, the gradual process of scientific development may well incorporate, he conjectures, the many ‘little revolutions’ espoused by Kuhn: I cannot help but observe that a long-term evolution in science can consist – at least in part – of sequences of such lesser revolutions. (204-205)

104We are returned, again, to a matter of interpretation. Where one observer sees the evolution of scientific knowledge, another sees revolutions. My analysis of the various models considered in this chapter leaves me with the following impressions. One is more likely to discern a ‘revolutionary’ pattern of breaks and ruptures in science (or other forms of knowledge) when one is living in a revolutionary (or at least turbulent) time. And change in scientific knowledge – as variously theorised by Bachelard, Canguilhem, Foucault and Cohen – is a matter not just of abrupt but of gradual transition. It contains slow transformation, even the preservation of old concepts and practices. New science maintains some form of dialogue with its predecessors. To deny this element of continuity is to postulate a knowledge that has no contact with its own past. In the next chapters, I focus on that element of continuity, as it has been expounded within the history of science, and within other areas of intellectual history.


1 Gutting (1989: 11) takes the term ‘network’ from an essay by Foucault on Canguilhem (Foucault 1985). My remarks on Bachelard and Canguilhem are indebted to Gutting’s book Michel Foucault’s Archaeology of Scientific Reason (1989). Gutting is largely responsible for alerting the Anglophone world to the major significance of Bachelard and Canguilhem in the context of Foucault’s histories of thought.

2 Two studies charting this influence are Dominique Lecourt’s Marxism and Epistemology: Bachelard, Canguilhem, Foucault (1975) and Mary Tiles’ Bachelard: Science and Objectivity (1984).

3 Gutting (1989: 13) draws attention to this point: Bachelard argues that Descartes’ principle of the intuition of properties is invalidated by twentieth century science. As a result of this process, previous philosophical ideals are revealed, in Gutting’s words, as ‘contingent conditions derived from philosophers’ inability to think beyond the framework of present science.’ (1989: 13)

4 This analysis is provided in The Philosophy of No (1969 [1940]: 36-38); it is cited by Gutting (1989: 18).

5 Bachelard, L’activité Rationaliste de la Physique Contemporaine, published in France in 1951, cited by Gutting (1989: 20). Gutting quotes Bachelard’s reflections on his own assertion: ‘One may smile at the dogmatism of a rationalist philosopher who writes “forever” regarding a scholastic truth. But there are concepts so indispensable in a scientific culture that we cannot conceive being led to abandon them.’ (Gutting’s translation, 1989: 20)

6 The reference is from the second edition of Canguilhem’s work on the reflex: La Formation du Concept de Reflex aux xvii et xviii siècles Paris: Vrin (1977: 68).

7 This reference is from the second edition of Canguilhem’s Études d’histoire et de Philosophie des Sciences Paris: Vrin (1970: 25) (Gutting’s translation).

8 These remarks, made by reviewers from, respectively, Science and The Times Literary Supplement, are featured on the jacket of the Third Edition, University of Chicago Press, 1996. I refer to this edition, which includes Kuhn’s 1969 Postscript, throughout.

9 In an interview with John Horgan, for the latter’s book The End of Science, Kuhn described his disdain for some of the interpretations of his work: ‘I’ve often said I’m much fonder of my critics than my fans.’ On the idea that science is a reflection of the prevailing power structure, Kuhn remarks: ‘In retrospect, I begin to see why this book fed into that, but boy, was it not meant to, and boy, does it not mean to.’ (Horgan 1997: 44-45)

10 The only apparent commonality as an intellectual influence on the two theorists is Aexander Koyré’s work in the history of science. Kuhn (1996: viii) and Foucault (1980a: 54) both credit Koyré as a formative influence on their work. Koyré, who taught in both France and the United States, was the subject of a special edition of the journal History and Technology (Vol 4, Nos 1-4, 1987), following a conference devoted to his work at the College de France. Koyré is celebrated by Canguilhem as ‘une figure, une œuvre, une leçon’ (Canguilhem 1987: 7), while I. B. Cohen quotes Kuhn’s assertion that his own ‘approach to the history of science owes more to Koyré than to any other living scholar.’ (Cohen 1987a: 63). Curiously, Kuhn also mentions as an influence on his work A. O. Lovejoy’s Great Chain of Being (Kuhn 1996: viii). Kuhn manages to build his discontinuous edifice on the platform of Lovejoy’s unchanging ‘unit-ideas’.

11 Kuhn shares this conceptual imprecision with Foucault, who likewise has been criticised for the looseness with which his own terms are applied. Clare O’Farrell (1989: 54-55) observes of Foucault’s work on intellectual history that ‘episteme’, depending on when and how it is used, may mean the configuration of knowledge of a given historical period, or of a civilisation – ‘the West’.

12 Kuhn lists as characteristics specific to science: the ‘relative scarcity of competing schools in the developed sciences’; the crucial role of the science community as providing the ‘only audience and the only judges of that community’s work’; the ‘special nature of science education’; the goal of puzzle-solving; and the value system specific to scientific groups. These and other features, Kuhn argues, are ‘none necessarily unique to science’, but in combination they set science apart (1996: 209).

13 Steven B. Smith: ‘While there may be profound differences between, say, Newton’s and Einstein’s systems of science, there are nevertheless direct continuities between them at the same time, that is, they are both systems of science… and therefore must be judged in terms of how well they satisfy criteria of scientific scope, accuracy, and so on.’ (Smith 1984: 91) Anthony Wilden: ‘The overt discontinuities of the Kuhnian “paradigms” all depend for their existence on an essentially continuous epistemological and ideological agreement about the nature and the goals of the scientific enterprise.’ (Wilden 1984: xxii)

© Presses universitaires de Provence, 2019

Licence OpenEdition Books

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search