I. Introduction: The structure of a scientific involution

A discipline as anxious to shore up its scientific credentials and as inclined to think in terms of ‘demarcation lines’ as International Relations (IR) is bound to take very seriously any discussion of what sociologists of science call ‘boundary work’, that is, the policing of what counts as ‘science’ and what does not.Footnote ¹ It should not come as a surprise that IR has paid much attention to debates in the philosophy of science, since from its very beginning the discipline has entertained a fraught relationship with scientific rationalism. But there is something utterly perplexing about the infatuation of the discipline with the work of Thomas Kuhn starting in the mid-1960s, not least because it was followed by the rejection of the various shibboleths previously gleaned in the pages of The Structure of Scientific Revolutions, from ‘paradigms’ to ‘incommensurability’. There is indeed something of a Kuhnian Gestalt switch whereby the very things that IR theorists initially found so exciting in Kuhn were suddenly jettisoned, leading an entire cohort of scholars to look for new idols among the ranks of the Popperian Kuhn-bashers gathered at the International Colloquium in the Philosophy of Science or in the ‘critical realism’ movement associated with Rom Harré or Roy Bhaskar.Footnote ² Rarely has an object of so much adulation become reviled so rapidly. In retrospect, these ambivalent feelings raise the question of what, exactly, Kuhn stood for in IR. The question is compounded by the fact that the reasons for which the discipline embraced Kuhn in the first place were often self-contradictory.

Consider indeed the causes of Kuhn’s initial popularity among IR theorists: for some, he provided a model for the history of science that could be applied with minor tweaks to political science and suggested that structural realism was indeed a scientific-grade theory of international politics.Footnote ³ For others, his sociological vision of science was a formidable weapon against ‘positivism’, which could be putatively deployed against realism.Footnote ⁴ Philip Mirowski points out that for many in the social sciences, Kuhn offered a fresh departure from the algorithmic notions of rationality that were dominant at the time.Footnote ⁵ In IR, the notions of paradigms and incommensurability meant that there was no need to feel sorry about the state of a discipline marred by fragmentation, but that there was no need either to relinquish pretensions to scientific rigour.Footnote ⁶ And by enshrining the impossibility of validating or refuting a scientific paradigm on the basis of a set of neutral logical criteria, the incommensurability thesis played a protectionist role in carving out a space for the development of intellectual ‘infant industries’ – such as transnational studies or international political economy – outside the realist mainstream and shielded them from criticism.Footnote ⁷ Some saw Kuhn as a revolutionary thinker, who encouraged ‘radical re-thinking, focused on “deep” problems rather than … the routine “puzzle-solving”’;Footnote ⁸ yet this did not deter others to suggest that he had certainly set IR on a course leading toward ‘normal science’ and its puzzle-solving powers.Footnote ⁹ All these contradictory assessments suggest each had their own Kuhn, who may have been a very different person from the author of Structure.

By the 1990s, however, the Kuhnfest had come to an end and the hangover was tangible. Yet, the reasons for ditching Kuhn were just as self-contradictory as those that had made him so popular in the first place. Many now claimed that Kuhn’s model did not apply to the social sciences after all: the attempts at enlisting Kuhn to justify progress were simply a ‘misuse’ of his work;Footnote ¹⁰ his sociology of science was inadequate to understand the structure of IR as a discipline, although a few admirers felt nostalgia for the lost world of paradigms;Footnote ¹¹ his reconstruction of the history of science was equally deficient.Footnote ¹² Even worse, rather than facilitating the coexistence of various approaches in IR, Kuhn had played an essentially conservative role by consolidating realism into ‘normal science’.Footnote ¹³ Unless what had happened was just the opposite: the ‘paradigm mentality’ had fostered the ‘Balkanisation’ of IR into small, fiercely independent territories.Footnote ¹⁴ As one can expect, the most unforgiving critics often happened to be former fans: after participating in the bacchanal and proposing an über-Kuhnian interpretation of the evolution of the discipline in 1987, Steve Smith later recanted by confessing that despite the original impression that paradigm-thinking was ‘liberating’, it turned out to be ‘just another gate-keeping device for maintaining the status-quo’.Footnote ¹⁵ Others followed suit and denounced the incommensurability thesis for being a ‘legitimating device … for theoretical fragmentation’ that stifled criticism and thus, presumably, could only work to the benefit of a still hegemonic realism.Footnote ¹⁶ Besides, the discipline had moved on – unless, as some started to suspect, it had never really looked like the archipelago of insulated paradigms that IR scholars inebriated with Kuhn would have us believe: the so-called ‘third debate’ between realism, pluralism, and structuralism may have never taken place.Footnote ¹⁷ These theoretical aggregates were far from incommensurable, as a result of which the notion of ‘paradigms’ generated ‘deleterious effects’ in a discipline better understood in conventional, ideal-typical Weberian terms.Footnote ¹⁸ In either case, only confusion had come out of the intemperate consumption of Kuhn.

One is tempted to put this dramatic reversal of fortune on the account of the more general waning of Kuhn’s prestige in science studies, where the author of Structure has been for some time the object of much soul-searching into the reasons of what many now view as an unwarranted intellectual success. Steve Fuller blames him for having produced a duplicitous version of science history meant to protect the patronage of science while generating accounts acceptable to all parties, and for having eased humanists and social scientists, among others, ‘into a postmodern mindset’.Footnote ¹⁹ David Stove has excoriated Kuhn for being the founding father of postmodern irrationalism.Footnote ²⁰ Phil Mirowski has pointed out his connection with operation research and the emergence of a science management regime shielded from public scrutiny and subservient to the needs of the Cold War establishment.Footnote ²¹ Even the more sympathetic accounts collected in a recent celebratory issue of the Journal of Modern History do not fail to emphasise Kuhn’s fundamental ambiguities. By reproducing this pattern, it would not be the first time, after all, that IR passively registers trends happening elsewhere. Yet, a number of things suggest that Kuhn has not been just a passing fad in IR.

It is common knowledge that IR has largely misused and misinterpreted Kuhn, but this diagnosis, along with the shift en masse to Imre Lakatos, has prevented us to probe deeper into his actual impact. One of his legacies was surely to encourage the representation of IR as a congeries of contending ‘paradigms’, despite the fact that such a situation was not considered by Kuhn himself, who envisioned paradigms only in their succession and never really contemplated their simultaneity, and despite the fact that the ‘-isms’ of IR theory can hardly qualify as genuinely Kuhnian paradigms. IR thus started to look very much like the fragmented world of sovereign states it sought to understand, rather than like a succession of epistemological empires. More surprisingly, Kuhn’s capacity to serve agendas in every quarter of IR and in particular to support both the agenda of neorealists and that of their postpositivist critics suggests that the reception of Structure was a paradoxical factor of integration at a time of strong disciplinary divisions. As everybody could pick their own Kuhn, he became the point of intersection between allegedly incommensurable worldviews. It must be said that Kuhn himself was partly responsible for this state of affairs. The fact that there was a myriad of ‘possible Kuhns’ to choose from in the absence of arbitration by Kuhn himself seems to vindicate those who, like Paul Feyerabend, had suggested early on that Kuhn’s ambiguity was ‘intended and that [he] want[ed] to fully exploit its propagandistic potentialities’, or those who, like Terence Ball, claimed that while political scientists had taken Structure’s message to heart, ‘few can say with certainty what that message is’.Footnote ²² Logical undecidability always means that the decision is, ultimately, political – something that, incidentally, was central to Kuhn’s vision of science.

In tracing the ‘Kuhn effect’ in IR, I must clarify upfront that I am not seeking to adjudicate between erroneous and correct interpretations of Kuhn: what I am interested in is the reception of Structure, no matter how creative or abusive some interpretations have been. As long as Kuhn was invoked to make specific moves in the discipline, it does not matter whether the underlying interpretation of his theory was flawed or not as it had real consequences. Considering the various uses of Kuhn, the article makes two interrelated arguments about his influence. First, it suggests that the main reason for Kuhn’s felicitous career in IR was that he offered a way out of the ‘second great debate’ by providing an image of science that could accommodate realism’s traditional suspicion vis-à-vis any form of robust rationalism. What Structure did was to break the framework of classical realism, the exemplary – dare I say paradigmatic? – expression of which was Morgenthau’s Scientific Man versus Power Politics, which had established a radical opposition between science and politics. Structure suggested that science too had its dark side, its irrational areas, and even its power politics. By suggesting that scientific revolutions were essentially a matter of power, Kuhn was applying something akin to realist principles in the field of epistemology. This did not only suggest that ‘science’ was safe for realism after all: it also suggested that political realism was nothing short of a theory of scientific knowledge. To put it differently, if Structure proved so fascinating to a discipline that initially did not want to have anything to do with science, it is because it resonated, at some level, with realism itself.Footnote ²³ Kuhn’s realism has so far remained the proverbial elephant in the room, and it took an outsider – a historian of science – to spell out the political implications of Structure that some realists had intuitively grasped.Footnote ²⁴

This leads us to the second argument developed in this article: if the Kuhn effect proved so important in IR, it is because, as Steve Fuller observed, ‘international relations is perhaps the one field in which Kuhn has been used to model not the field’s research trajectory but its very subject matter’, namely decision-making.Footnote ²⁵ It is indeed striking that while Kuhn was discussed by many in the field of International Relations, those who looked up to him in order to actually explain international politics were primarily concerned with establishing criteria of rationality for political decisions. Until then, the entire field of IR was premised on a ‘decisionist’ vision of politics that emphasised the contingency and situatedness of politics in order to repel the notion that political conduct could take the form of rule following and that politics could be comprehended by scientific rationalism.Footnote ²⁶ Against the prior focus of the field of international law, postwar realism had drawn discretely on the critiques of legal positivism as a ‘scientific’ codification of International Relations: the rules of international law had no import for power relations, which were essentially a matter of decisions. No ‘science’ of such decisions could ever obtain – only a prudence informed by historical knowledge. In this context, Kuhn was read in a very specific way: he was seen as offering an image of scientific rationality that could be used as a benchmark for politics because it also broke with legalism and the representation of science as a corpus of laws growing organically. Science too was now characterised by the uncertainty, relativity, and limitations found in political rationality. This is what made Kuhn so congenial to some IR theorists: he allowed them to be ‘scientific’ without having to relinquish a single assumption of political realism. Graham Allison’s Essence of Decision (1971) or Robert Jervis’s Perception and Misperception in International Politics (1976) thus used Kuhn to guard themselves against an over-rationalisation of the process of decision-making while reaping the benefits of a ‘scientific’ analogy. Kuhn himself was aware of the analogies between politics and his vision of science, and by remaining suggestively elusive as to their implications he left opened the possibility of far-fetched interpretations. Much more than his role in shaping the self-image of the discipline, Kuhn’s deep and lasting impact on IR was to reconcile scientific man with power politics by suggesting that the former was just another rank-and-file realist.

II. Science is like a box of chocolates…

In order to understand the impact that The Structure of Scientific Revolutions had on International Relations theory, it is useful to first set the stage on which Kuhn made his triumphal entrance in the discipline. Although the book was published in 1962, its digestion by IR started after a brief physiological lag, during which it first reached the shores of political science and psychology, in particular in relation to strategic decision-making problems, where it was enlisted to build consensus around abstract behavioural models and secure the paradigmatic status of the notion of ‘system’.Footnote ²⁷ By the time Kuhn was metabolised by IR in the late 1960s, the discipline was in the throes of the internecine guerilla warfare known as the ‘second debate’, which opposed ‘traditionalist’ thinkers who defended the idea that politics was an existential and historical phenomenon escaping the strictures of scientific knowledge, and ‘behavioural’ social scientists entranced by their recent discovery of cybernetics, general systems theory, and games.Footnote ²⁸ The former insisted on the constitutive circularity of knowledge, which, even in its most rationalised manifestations, could not escape its situated historical condition, whereas the latter believed that scientific knowledge was defined precisely by its complete detachment from historical reality and its confinement to the netherworld of models and simulations. While the actual positions were probably much closer than it appeared at first sight and remained largely internal to the ‘realist’ mainstream, the debate was dramatised as a radical opposition between ‘realism’ (or ‘history’) and ‘science’, which explains why it remained largely inconclusive.Footnote ²⁹

While Steve Fuller has proposed to see Kuhn as an alter ego of ‘Chance’, the main character in Jerzy Kozinski’s novel Being There, whose simple pronouncements are taken as the expression of a profound wisdom, Forrest Gump may provide a better interpretive frame for Kuhn’s influence in IR. Just as Gump finds himself caught in the midst of momentous historical events which he has nothing to do with, yet contributes to defining, Kuhn walked by chance into the discipline in the midst of the so-called ‘second debate’ and unwittingly delineated a way out of it. While factions were fighting over the meaning of science and the role of history, he was conscripted by both sides to support their position – and he obliged (‘I just be doin what I’m tole’ says Gump to his football coach, and this may well be the key to Kuhn’s success in all the different precincts of IR).Footnote ³⁰ Kuhn offered indeed a version of science everybody could embrace: he vindicated the behaviouralist claims to represent a truly scientific approach to international politics, not least by offering a perspective on the history of their discipline in which they featured as the vanguard of the scientific revolution, but he also seemed to give epistemological substance to their opponents’ claim that any meaningful statement about international relations ‘must … derive from a scientifically imperfect process of perception and intuition’ rather than ex tempore laws or rules, and thus that even ‘science’ was still expressing a partial perspective on the world.Footnote ³¹ Kuhn’s formidable traction in the discipline was directly connected to the fact that he seemed to reconcile these arguments as the two faces of the same coin. Science was indeed premised on two different ‘truths’ – one for the historians, one for the scientists – and this suddenly cast the second debate in a different light by suggesting that it was entirely moot. Just like Forrest Gump’s biography is a conservative American pastoral leading up to the cultural reconciliation of a divided nation, Kuhn’s two-decades long marathon through IR secured the unity of the discipline during a time of deep divisions and ultimately reconciled the historically minded IR scholar with his rocket-scientist counterpart. Let us examine how the historical and the epistemological argument have been used.

It was in his capacity as a historian of science that Kuhn was conscripted by those who advocated the mainstreaming of the behavioural science model. If the history of their discipline could be plausibly told as a paradigm-anomalies-scientific revolution sequence, then surely it was a science. In the mid-1960s, Kuhn thus starred prominently in the keynote speeches given by the presidents of the American Political Science Association: David Truman in 1965 and then Gabriel Almond in 1966 established a new historical interpretation of the development of the discipline. Beyond minute differences, the general canvass they offered was roughly the same: from the mid-nineteenth century to the 1930s, a first ‘paradigm’ had prevailed, characterised by an unsophisticated empiricism premised on the parochial belief in the naturalness of ‘the separation of powers, checks and balances, and the mixed constitution’. Its paradigmatic force was illustrated by the fact that even the challenges formulated at the end of the century – interest groups theory, administrative complexity, etc. – ‘were still prisoners of the theory’ and were expressed in its language.Footnote ³² It took ‘experiences’ falling ‘outside the implicit general agreement’ as well as the accumulation of ‘anomalous findings’ throughout the 1930s and after the Second World War to undermine it.Footnote ³³ From their vantage point, Truman and Almond both concurred that ‘a new paradigm [was] developing in political science’, and that it was defined by the concept of ‘system’: ‘the introduction of the system concept represents a genuinely important step in the direction of science … Taking place in the era of the scientific revolution it represents a surer thrust into the culture of modern science.’Footnote ³⁴

Almond’s view of the new political science contained two elements that predisposed it to be reproduced in the field of International Relations. The first was that the traditional paradigm of the separation of powers in political science was, mutatis mutandis, easily translatable into a paradigm of the balance of power. The second was that the crisis and the subsequent scientific revolution leading to ‘systematic’ science were premised on anomalies which, combined, amounted to ‘the disproof of the Enlightenment historical predictions’ – an assumption that could only resonate with the ideas of early realists, also premised on a critique of the Enlightenment.Footnote ³⁵

The first adaptation of this historical argument to IR was first made two years later by a student of Harold Sprout, Ronald Rogowski. Rogowski provided a historical reconstruction of the discipline in explicitly Kuhnian terms, which assumed that following the ‘profound anomaly’ of the First World War, the traditional reliance on international law as the master representation of international relations was displaced by an approach that stressed the relations between multiple actors and did not take for granted the naturalness of the balance of power. As investigations moved beyond the old legalistic vision, they gradually evolved all the attributes of an alternative paradigm, concerned with the ‘realities’ of a balance that had to be maintained through the purposive decisions of statesmen. The debate between realists and idealists, Rogowski pointed out, was nothing but an ‘internal squabble’ over the best ways to achieve this – it was just that ‘a victorious generation of “realists” found it useful in mopping up to consign all “idealists”, prescriptive international lawyers and altruistic students of international relations alike, to the same limbo’. The result was the modern, realist paradigm of IR that ‘attained textbook status at Morgenthau’s hands in 1948’. This paradigm, however, came under attack a decade later as a variety of authors – Morton Kaplan, Richard Rosecrance, Charles McClelland, Thomas Schelling, Kenneth Waltz – analysed international politics using approaches relying on the notion of ‘system’.Footnote ³⁶

This reading seemed to be confirmed a few years later by Arendt Lijphart. Walking in the steps of Almond but seemingly unaware of Rogowski’s earlier piece, Lijphart suggested that the history of IR exhibited a ‘clear pattern’ of development.Footnote ³⁷ This pattern consisted in a ‘traditional paradigm’ going back to Thucydides and based on the notion of sovereignty and anarchy. The last remnants of this paradigm were to be found in the writings of ‘traditionalist’ realists such as Hans Morgenthau, Arnold Wolfers, Stanley Hoffman, or Raymond Aron.Footnote ³⁸ Finessing the discrepancies between IR and Kuhnian ‘mature science’, Lijphart suggested that something close to ‘normal science’ obtained in the discipline until the Second World War, when the cumulative effect of ‘discrepancies’ between the traditional paradigm and reality displaced it and allowed for the emergence of new methods reflecting a fundamentally different perspective. What characterised these was that they no longer saw the world as anarchical, since they subsumed sovereign decisions and conflict under the all-encompassing concept of ‘system’. A scientific revolution thus took place in the 1950s and saw the emergence of the ‘behavioral paradigm’. It is precisely because the second debate was a dialogue of the deaf – ‘the two schools are really looking at different worlds’, Lijphart saidFootnote ³⁹ – that it was proof that a paradigmatic shift was taking place and a sure sign of the scientific normalisation of IR. Years later, Waltz’s Theory of International Politics would enshrine the new paradigm

III. … you can pick whichever ones you like.

But if Kuhn provided fodder for those who defended the scientific normalisation of IR, he also brought grist to their opponents’ mill. While the former read Kuhn as an ally in their bid to establish a unique methodological standard built around the notion of ‘system’, the latter saw in The Structure of Scientific Revolutions an attack on the last remnants of logical positivism and on the possibility of a neutral scientific language, and thus something to be deployed against the normalisation of the discipline around a set of system-theoretical concepts. In political science, the Truman-Almond line was rapidly countered by Sheldon Wolin, in a judo move that turned Kuhn against its first promoters and impugned the behaviouralists’ pretense to dictate the terms of political research by pointing out that behaviouralism was only one paradigm of political theory.Footnote ⁴⁰ While some political scientists were afraid of the ‘non-rational and relativistic implications’ of Kuhn’s account of science and sought to counter them by reinstating validation criteria that they found in Lakatos, others simply embraced them.Footnote ⁴¹ The same thing happened in IR, where there was a real fear that any scientific normalisation of the discipline would entrench an updated realism operating with the concept of ‘system’ as the only legitimate ‘paradigm’ at the expense of alternative approaches (a fear fuelled by Kuhn himself, who had surmised that ‘it is precisely the abandonment of critical discourse that marks the transition to a science’.)Footnote ⁴² No matter how implausible this conflation of realism with ‘normal science’ was (not least because of realism’s fundamental tension with the idea of a science of politics), it seemed to validate a contrario the notion that attacking positivist models of science was tantamount to pulling the rug under the realists’ feet. The contribution of behaviouralism, as Smith and many others believed, was indeed ‘to continue dominance by bringing the semblance of scientific rigor to Realism’.Footnote ⁴³ For these critics, realism and positivism were virtually conflated, since ‘realism in international relations inherited … the presumption that the methods of natural science are the key with which to unlock the social world’Footnote ⁴⁴ – an assumption which, as we shall see, turned out to be plain wrong. The odd result of all this was that ‘postpositivism’ was somehow understood to be a bulwark against realism.

Those who did not recognise themselves in the emerging science of political ‘systems’ thus found that Kuhn was a true Janus bifrons, who provided ammunition against those who used him to defend the agenda of a modernised, structural realism. In this new incarnation, Kuhn was also a constructivist thinker,Footnote ⁴⁵ one for whom ‘the progressiveness of science lies not in the nature of the work scientists do’ or its relationship to ‘reality’ (however understood), ‘but in the control they exercise over how they recount their collective history’.Footnote ⁴⁶ The history of scientific progress told by Truman, Almond, Lijphart, and others suddenly appeared to be optical illusion, since for Kuhn each scientific revolution is followed by an ‘Orwellian’ rewriting of history, which casts prior scientists as precursors of present ones, working on the same problems and contributing to the gradual emergence of the present paradigm (thus denying revolutions any function and making them ‘invisible’).Footnote ⁴⁷ Textbooks are rewritten, the work of previous scientists is ‘revised’, and history is ‘truncated’ – but all for the good cause of normalising current science: once this shift in perspective is adopted, history tells us that the current paradigm was, somehow, always in the making. Part of what made Kuhn exhilarating is that he was giving away this dirty little secret – histories of disciplinary progress are always deceptive – even though he was at the same time justifying Whig history as necessary to the proper operation of a scientific community.Footnote ⁴⁸

In any case, the disjunction between the official history of a discipline and its actual history contributed to relativise the behaviouralists’ scientific claims and to develop a more complex sociological argument about the capacity to establish paradigm dominance through non-scientific means. What mattered was that the validity of a scientific paradigm was not established through some universal or neutral logical criteria, but exclusively through institutional means, ranging from the control of educational programmes to the inculcation of a whiggish history to gullible aspiring scientists. If realism was still dominant, it was not because of some demonstrable theoretical superiority, but because, as Steve Smith put it, it was ‘the largest and most well-financed academic community’.Footnote ⁴⁹

The notion that normal science depended on the capacity of specific groups to command consent had important epistemological consequences. Kuhn fascinated also because his model of science was premised on ‘an explicit rejection of the model of algorithmic decision theory’.Footnote ⁵⁰ Incommensurability between paradigms was not only a characteristic of the discontinuous historical development of science, as the advocates of behaviouralism had assumed, but a fundamental epistemic structure in a discipline where different ‘schools’ coexisted contemporaneously. This certainly thwarted the claim by neorealists that their approach was demonstrably superior on rational grounds, but it also eased the discipline into a relativistic mindset: since incommensurability implied that each paradigm defined its own criteria of validity and that, between paradigms tertium non datur, the notion of truth and a strong concept of rationality were among the first casualties. Even though Kuhn’s pronouncements remained ambiguous as he both ‘endorsed and qualified statements of anti-realism’,Footnote ⁵¹ his suggestion that the traditional notion of truth as adaequatio rei et intellectus was not a useful benchmark of scientific ‘progress’ (since all this was in the end ‘in the eye of the beholder’)Footnote ⁵² was bought wholesale by critics of realism and interpreted as a strong form of relativism. Michael Banks, for instance, saw in Kuhn somebody who argued that ‘there is no absolute truth’.Footnote ⁵³ Reality per se was not accessible, because it was only a kaleidoscopic effect of paradigmatic vision. There was no proof, no ‘evidence’, that could be used to require submission to a realist paradigm now parading as ‘normal science’, since evidence, as Steve Smith wrote following Kuhn’s footsteps, ‘is evidence precisely, and only, because it fits within a theoretical framework’.Footnote ⁵⁴ It is not that these critics were against science, on the contrary: they were all too glad to accept Kuhn’s description of science as a social practice built upon non-rational foundations, as this allowed them to suggest that realist principles only applied to the world realists saw.

By offering a multifaceted theory of science, Kuhn allowed advocates of scientific IR and their critics alike to hold on to their different notions of what defined the discipline, since these notions were equally valid. The ‘great debates’ of the discipline had no solution in the sense of a demonstration that one side was right. When Ole Wæver suggested in 1996 that the ‘third’ or ‘inter-paradigm debate’ had not taken place (because there were no agreed upon terms for debating), he was unwittingly picking up a well established argument: Lijphart had written twenty years earlier that scientific revolutions ‘do end … even though rational arguments are not the means to solve them’.Footnote ⁵⁵ And as early as 1968, Rogowski had suggested that the ‘second debate’ too was a non-debate, marred by the fact that both sides were captive of their own inarticulate and incommensurable assumptions about what counted as ‘facts’.Footnote ⁵⁶ The various ‘schools’ of International Relations simply inhabited parallel universes that hardly intersected. The seeds of the ‘postpositivist’ moment that characterised IR in the 1980s and 1990s were thus planted three decades earlier, and Kuhn had much to do with it. In this perspective, any talk of progress was an Orwellian illusion, necessary to the proper operation of scientific collectives and fit for ISA presidential speeches. But such ‘progress’ could not be equated with the progress of Reason projecting its light over ever wider swathes of reality: rationality became a circumscribed phenomenon, dependent upon the prior adoption of a certain perspective on the world that, in turn, could not be justified in rational terms and which could be preached only to the converted. The choice of a paradigm created its own justification in retrospect, on the basis of its efficiency and expediency in solving puzzles and ‘rationalising’ a set of phenomena – just as the exceptional political decision becomes legitimate in retrospect, when it succeeds in establishing a new stable order.

IV. The realist epistemology of Thomas Kuhn

Starting in the 1970s, however, assessments of Kuhn became somewhat mixed. Even though in 1989 Yosef Lapid could still suggest that the dawning postpositivist era in IR was premised upon ‘the possibility that, within limits, diversity of viewpoints might be fully compatible with scientific rationality and objectivity’, this optimism was viewed with suspicion. ‘Paradigmatism’ was considered a pathology that had blunted criticism and impaired scientific progress.Footnote ⁵⁷ Less threatening figures like Lakatos were summoned to shore up the antipositivist position while rescuing scientific rationalism, lest one would be accused of postmodernism and relativism. Last but not least, it was clear that the attack on positivism had failed to destabilise a realist mainstream made immune to criticism.Footnote ⁵⁸

These second thoughts, however, barely scratched the surface of what had really happened. It is true that the reaction against the project of turning IR into a behavioural science had failed to unsettle realism, but the reason was not because realism benefitted from paradigmatic immunity. What the postpositivist critics of realism failed to understand is that the advocates of the ‘scientisation’ of realism in the 1960s and 1970s were not positivists. Flummoxed by the high-tech gimmickry of ‘system’ theorists, with their set theoretic games, their simulations and their algorithmic schemata, their critics jumped to the conclusion that this had to be a resurgence of positivism – a conclusion that, unfortunately, has been since the accepted doxa in the discipline, with debilitating effects on any critical project. Yosef Lapid, for instance, mistakenly associated the scientific side of the ‘second debate’ with ‘the ascendance of positivism in Western social science’, while R. B. J. Walker saw Waltz as a die-hard advocate of the ‘positivist method’.Footnote ⁵⁹ Hypnotised by the notion of ‘postpositivism’, which had the unfortunate effect of implying – wrongly – that the past of the discipline was positivistic, many IR theorists showed little or no awareness that the kind of ‘science’ peddled by Kaplan, Schelling, Riker, or Waltz had deep roots in cybernetics, with its celebration of artificiality and its ontological agnosticism, and not in positivism.Footnote ⁶⁰ They also failed to notice the very explicit disclaimers of their opponents: Morton Kaplan, for instance, had candidly confessed that he was ‘not a positivist’, and that system-models were self-referential constructs that bracketed out reality.Footnote ⁶¹ If anything, their neorealist successors were even more outspoken: Waltz’s Theory of International Politics inherited from cybernetics a fuzzy relationship between theoretical models and reality: ‘theories do construct a reality, but no one can ever say that it is the reality’.Footnote ⁶² Waltz pointed out that this kaleidoscopic nature of ‘reality’, accessed only through incommensurable theories, implied that participants in debates over the empirical merits of different theories ‘are talking about different things while using the same terms for them’.Footnote ⁶³ This was a sufficient reason to distinguish his own theory from scientific positivism: it was fundamentally different from the ‘analytic method’ which was ‘preeminently the method of classical physics and … often thought of as the method of science’.Footnote ⁶⁴ While Waltz did not elaborate upon this distinction, it is clear that the tension between realism and science had not subsided with structural realism and was now taking the form of a rejection of positivism, which only became stronger in Waltz’s later work. Rejecting Popper’s falsification proposal, he attacked ‘positivism’ by suggesting that falsifying a theory was made difficult by ‘the interdependence of theory and fact’. As a result, ‘what is to be taken as evidence for or against a theory is always in question’, and all one can do is going back and forth between theory and ‘an uncertain state of affairs that we take to be … reality’.Footnote ⁶⁵

The idea that realism would be destabilised by an attack on its putatively positivist foundations thus turned out to be a Quixotic fight against epistemological windmills. It not only failed miserably: it backfired. In deploying a Kuhnian arsenal, the postpositivist critics of the realist mainstream were unwittingly doing the groundwork for the enemy. By the time they started having second thoughts about Kuhn, it was too late: the enemy had crushed the gates, and science itself was now premised on realist principles. Kuhn, in fact, viewed science very much like realists view the world.

Arguments supporting Kuhn’s affinities with political realism have been made on the basis of biographical considerations, in particular his association with James Conant at Harvard, closely studied by Steve Fuller. According to Fuller, it is through his mentor that Kuhn inherited a number of realist thought patterns, which he then sublimated into a theory of science.Footnote ⁶⁶ It is indeed tempting to read back some of Kuhn’s core tenets into Conant’s vision of science – in particular a political approach to science, shaped by the political and institutional imperatives of the Cold War, which acknowledged the importance of non-scientific decisions in scientific work, as well as the idea that there is ‘no such thing as the scientific method’, in the sense of an algorithmic logic capable of endlessly generating theories from facts.Footnote ⁶⁷ But Fuller does not specify the exact pathways of this influence, which he takes for granted as part of a general intellectual climate.Footnote ⁶⁸ This being said, one does not need to delve deep into Kuhn’s biography to make the case for the political implications of Structure, since Kuhn himself made them explicit: ‘the genetic aspect of the parallel between political and scientific development’, he wrote, ‘should no longer be open to doubt’.Footnote ⁶⁹

While much has been written about Kuhn’s role in shaping the self-image of the discipline as a congeries of ‘paradigms’, little attention has been paid to these political analogies. One of the core propositions of Structure was that (scientific) authority was vested not in a system of rules, but in a (scientific) collective that decided to recognise certain rules as valid and had the means to enforce that decision. While this may have been an earth-shattering proposition for the history of science, it was hardly big news in IR. Postwar realism too was premised on a break with representations of politics as a legalistic or rule-based activity: instead, realists suggested, understanding international politics meant shifting from the rationalism of legal structures to the messiness of power relationships. The shift from science to politics was not difficult: Kuhn was building upon a long historical tradition in Western thought, in which the concept of law, as ‘natural law’, was for a long time common to jurisprudence and science with little by way of differentiation.Footnote ⁷⁰ The move from a system of laws to power relations that was one of the subtexts of Structure certainly resonated with realists, since Kuhn was doing for scientific positivism what they had done earlier for legal positivism. The rejection of legalism was something that Kuhn shared with the realists.

This, in turn, raises the issue of the analogy between paradigm choice and political decision. On the one hand, Kuhn often describes the adoption of a new paradigm in terms of Gestalt switch, meaning a process that is more of a passively registered change of perspective than an active decision. But Kuhn also cultivates ambiguity in this respect and in many instances he explicitly presents paradigm change as a matter of ‘decision’ or ‘choice’, including between competing alternatives. The analogy between paradigm choice and political decision was not lost on IR theorists, notably on Waltz, who distinguishes between scientific decision and scientific law in Theory of International Politics.Footnote ⁷¹ Behind every paradigm thus lies a constitutive decision, prior to the system of rules it gives rise to, in the same way that a legal system is premised on an unconditioned, sovereign decision which by its very nature is exempt from its strictures. Kuhn himself warranted a judicial analogy by suggesting that a scientific paradigm, ‘like an accepted judicial decision in the common law … is an object for further articulation and specification’.Footnote ⁷² If anything, the oscillation between ‘Gestalt switch’ and ‘decision’ in Structure seems to suggest that the former qualifies the latter as non-rational, non-discursive, and non-deliberative. While for other philosophers of science, like Pierre Duhem, the choice between theories can be made on the basis of conventional or commonsensical deliberation even if there is ‘no logical or formal method’ to arbitrate between them, Kuhn presented a more radical version of this argument by suggesting that there was no rational ground for comparing paradigms.Footnote ⁷³

But the analogy between paradigm choice and political decision cuts deeper, for in advancing an argument in which science rests on theory-choices that cannot be accounted for rationally (Kuhn compares them to something akin to a religious conversion),Footnote ⁷⁴ Kuhn ultimately conjures a vision of science that is disturbingly equivalent with the realist vision of politics. This convergence is better teased out when one examines the implications of ‘incommensurability’, which built upon insights originally developed by Duhem and extended by Willard Quine. By showing that any finite set of observations could be accommodated by an indefinite number of theories, Duhem had established that there was no scientific criteria for establishing the superiority of one theory over another. Quine later extended the ‘underdetermination of theory’ proposition to linguistic sets, and suggested that it was impossible to stipulate unambiguous translation criteria between different languages, even when empirical correlations limit the set of meanings possibly intended: a central, authoritative, translation code could not obtain.Footnote ⁷⁵ While in linguistic communication, misunderstandings can abound and interpretations can be challenged without necessarily having dramatic consequences, this problem took a special urgency in science, where there was a need to secure a degree of stability required by ‘normal science’ in the absence of a logically demonstrable foundation. The notion of incommensurability thus pointed at the problem of achieving a stable and rational order in the absence of an authoritative system of rules capable of adjudicating between paradigms.Footnote ⁷⁶ But wasn’t this precisely the core question that IR theory sought to address – stability in the absence of authoritative and centralised rules, order under conditions of anarchy?

In a very direct way, IR theory too was born as an answer to the problem of interpretive authority, when it emerged in the late 1940s and early 1950s as a critique of international law. Hans Morgenthau had pointed precisely at the impossibility of stipulating a univocal and authoritative translation code lifting the indeterminacy of international law.Footnote ⁷⁷ For legal realists in general, the determination of the meaning and the applicability of legal rules was considered a matter of decision that could not be itself settled by the rules under consideration. A conflict over which a given set of rules applies cannot be solved in a rule-based manner and may be settled by recourse to force. The very same logic underpins Kuhn’s explanation of conflict between paradigms: there are no neutral procedures for resolving such conflicts, since any rational argument is conducted on the basis of a prior adherence to a paradigm: ‘the choice [between competing paradigms] is not and cannot be determined merely by evaluative procedures characteristic of normal science, for these depend in part upon a particular paradigm, and that paradigm is at issue. Each group uses its own paradigm to argue in that paradigm’s defense.’Footnote ⁷⁸ Science and International Relations thus faced the same problem: in the absence of a tertium quid, what counts as ‘facts’ and the choice of the relevant procedures for addressing them cannot be established by rule-governed procedures, but through power relations. As Lakatos put it in his coruscating critique, in Kuhn’s science, ‘truth is power’.Footnote ⁷⁹

Kuhn’s discussion of normal science and paradigm change is thus uncannily reminiscent of the world of politics. Scientific crises are indeed formally equivalent with international ones. Normal politics, like ‘normal science’, takes place when the application of rules goes without saying; the rules are not in question and are tacitly agreed upon. In situations of ‘crisis’, however, the determination of which rules apply and the correlate decision over their relationship to ‘facts’ is precisely what is at stake and what cannot be established in a rule-governed fashion. Any adjudication of competing claims by rival countries (any choice between paradigms) cannot be done through rules. The rules of international law (the logic of scientific activity) do not provide a path to a solution because they must be interpreted, and the ultimate instance of interpretation is always party to the dispute (that is, subscribes to a paradigm). Lest the reader believes these are wild extrapolations, I hasten to add that Kuhn leaves no doubt whatsoever about the equivalence of paradigm choice and political or constitutional decision: ‘like the choice between competing political institutions, that between competing paradigms proves to be a choice between incompatible modes of community life’.Footnote ⁸⁰ Hidden in plain sight in the pages of Structure, the political distinction between friend and foe had thus made its discrete entrance in science and was now becoming an unconditioned premise of scientific rationalism.

V. The elusive essence of (rational) decision

These analogies between paradigm choice and fundamental political decisions were not lost on IR scholars. But as Kuhn became widely read in the discipline, the distinction between theory choice and political decision became blurry. For in a realist world where the opponent’s motives are uncertain and can be interpreted in a variety of ways, the ‘rational’ course of action is entirely dependent on the prior adoption of an interpretive grid that helps making sense of that opponent’s moves. The problem for decision-makers thus becomes the problem of deciding between different sets of hypotheses or ‘models’ that all make the enemy legible, yet in different ways and presumably with different practical consequences. To put it differently: in the uncertain world of realism, in a very fundamental way, the political decision is theory choice, and theory choice is a political problem. As a result, explaining decision-making too becomes a matter of choosing a paradigm.

The widespread interest in Kuhn among IR theorists working on decision-making certainly warrants Steve Fuller’s observation that it is the only discipline in which Kuhn has been used not only to map the discipline itself, but also to conceptualise its very subject matter.Footnote ⁸¹ Again, these were often creative interpretations, but they would have been impossible if Kuhn’s ambiguous formulations had not made them plausible. What IR theorists found in Kuhn was not so much a substantive orientation that influenced them – indeed, the whole discipline was already premised on decisionism – but an account of science that was compatible with it and gave it scientific status: by breaking with legalistic representations of science, Kuhn had effaced the distinction between science and politics that was at the core of postwar realism.

The first and arguably most influential work amalgamating political decision and theory choice is Allison’s Essence of Decision, published in 1971. The book reflected the internal discussions of the ‘bureaucratic politics’ group led by Ernest May at Harvard’s Institute of Politics, which operated as the research arm and intellectual beacon of the John F. Kennedy School of Government. It originated ‘at least in the spring of 1966’ when the May group started to convene, and it represented the ‘still unfinished “Evolving Paper” of [the May] Group’.Footnote ⁸²

An in-depth analysis of the Cuban missile crisis based on first-hand information provided by key players in the decision process, Essence of Decision was an innovative book that did not have a thesis but instead applied three different interpretive grids to the same material and outlined the payoffs as well as the blind spots of these ‘models’. For each of the models, Allison provided a narrative ‘cut’ of what the Cuban crisis looked like when it was interpreted through that particular lens. Yet, like the models, the narratives where not related in a logical fashion but simply juxtaposed as alternative versions of reality. The main purpose of the book was not to provide an ultimate pronouncement or to uncover the ‘truth’ of what had actually happened, but rather to accommodate the flux and uncertainty of the actual decision-making process by shedding light on the taken-for-granted assumptions that inform our understanding of decisions. Essence of Decision thus sought to identify the ‘conceptual lenses’ affecting foreign policy analysis, knowing that each of these lenses ‘magnifies, highlights, and reveals’ but also blurs or neglects, in an effort to ‘demonstrate how alternative conceptual lenses lead one to see, emphasize, and worry about quite different aspects of events like the missile crisis’.Footnote ⁸³

Although Kuhn is not cited, his influence on Allison and on the May group in general, flagged out by Bruce Kuklick in Blind Oracles, surfaces in the minutes and documents of the group meetings.Footnote ⁸⁴ In 1969–70, Allison and his associates read Kuhn and saw in him a crucial ally as they developed a theory of decision that refused to arbitrate between incommensurable models. In particular, the reading of Structure led them to justify the dominance of the rational choice model, even though they sought to develop an alternative to it on the basis of its perceived flaws. Multiple examples of irrational decisions by leaders or nations were not cause for discarding the model: ‘it must be replaced with something better before we can realistically counsel the discontinuation of its use’, Steinbrunner wrote in a 1969 memo with reference to the resilience of normal science in the face of contradictory data. More importantly, the members of the group saw in Kuhn a justification for the equivalence of scientific and political rationality. In both cases, rationality obtained only within the framework of a prior decision: ‘The point to which Kuhn sensitises us,” Steinbrunner went on, ‘is that in the most productive scientific disciplines, all work has been organized by a fundamental paradigm, i.e., a set of assumptions defining a family of models. I believe that the argument can be made that the enterprise of making policy tends similarly to be organized by a fundamental paradigm.’Footnote ⁸⁵ For Harvard’s policy wonks, policymaking was akin to scientific practice in terms of cognitive setup: what Kuhn wrote about science could be transposed to politics without further qualifications. For the post-Kuhn realists, the prerogative decision, taken under time constraints and within a context of normative and practical uncertainty, was suddenly expressing the same rationality as scientific activity. An implication of this popularised understanding of Kuhn was that to in order to be ‘scientific’, policy did not have to reach a higher degree of rationality, since an underdetermined decision remained its unfathomable basis.

This intellectual backstage explains the Kuhnian flavour of Allison’s models of decision analysis. These decisional models are solipsistic entities that do not so much distort the perception of a reality ‘out there’ as they constitute what counts as real and what counts as evidence: ‘what we see and judge to be important and accept as adequate depends not only on the evidence but also on the “conceptual lenses” through which we look at evidence’.Footnote ⁸⁶ Even though Allison gestured towards the possibility of integrating the three models and repeatedly claimed he was subscribing to the standard formula of social scientific positivism à la Hempel, the structure of the book as well as the nature of the ‘conceptual lenses’ under scrutiny undermined such claims.Footnote ⁸⁷ Like Kuhn’s paradigms, Allison’s ‘models’ led analysts to see different worlds and were, in the end, incommensurable: ‘while at one level three models produce different explanations of the same happening, at a second level the models produce different explanations of quite different occurrences’.Footnote ⁸⁸ The result was a string of contradictory statements and positions, which only got more confused in the 1999 edition. In the end, the ‘grand model’ that Allison auspicated and which was expected to integrate the different policy analysis paradigms into a general theory was ‘really a metaphor’.Footnote ⁸⁹ All there was left was the undecidable alternative between different tunnel visions, or rather a fundamental decision that was not dependent on a formal set of rules. By the same token, there were no criteria from which such a decision could be judged. According to Kuklick, much of Essence was intended to absolve the mistakes of the best and the brightest in government through a ‘stylish relativism’, which suggested that any decision was dependent on prior assumptions, between which there was no way of arbitrating rationally, and that it was not possible to isolate the factors that had really generated the decision. The title of Allison’s book and the John F. Kennedy quotation from which it is excerpted give away what may be the only message of Essence – the political decision is ultimately an inscrutable arcanum arcarnorum: ‘the essence of ultimate decision remains impenetrable to the observed – often, indeed, to the decider himself … There will be always the dark and tangled stretches in the decision-making process – mysterious even to those who may be most intimately involved.’Footnote ⁹⁰Essence thus boiled down to a restatement of a traditional metaphysics of sovereignty, now refurbished with a scientific imprimatur. Long confined to the darkest recesses of political theory before being reprocessed by IR in the postwar years, decisionism was now worming its way into the very definition of scientific rationality.

A similar example of an intellectual project that set out with positivist overtones but ended up endorsing Kuhnian perspectivism is Robert Jervis’s Perception and Misperception in International Politics, published five years after Allison’s book. The book was based on the assumption that ‘perceptions of the world and of other actors diverge from reality in patterns that we can detect’ (p. 3) – yet, the idea of measuring and understanding this ‘deviation’ quickly unravelled, as the distinction between ‘perceptions’ and ‘misperceptions’ gradually becomes meaningless, or at least makes sense only in retrospect, when the consequences of decisions tell them apart.

Along with Essence of Decision, Perception is the product of an intellectual context influenced by the impact of nuclear strategy on IR theory and the correlate shift of emphasis from material-military factors to the psychological factors involved in the ‘efforts to influence the decision to attack’.Footnote ⁹¹ Psychology, and in particular Gestalt psychology, provided a perfect meeting ground with Kuhn’s philosophy of science and its fascination with duck-rabbit problems. In this respect, Perception is superior to Essence of Decision, since it makes explicit what was largely implicit and inarticulate in Allison’s book (but spelled out in the minutes and memos of the May group), namely the equivalence between theory choice in science and political decision in international politics. In his attempt to specify ‘rational’ criteria for decision-making, Jervis turned to what appeared to be the most constraining concept of rationality available, that is, scientific rationalism: ‘I see no reason to believe’, he suggested, ‘that political decision-makers are less rational, sophisticated, and motivated to understand their environment than are scientists’.Footnote ⁹² But it’s not just that political decisions should be evaluated according to the standards of rationality that had currency in science: Jervis’s argument went further and was explicitly premised on the equivalence of scientific investigation with the cognitive process that informed political decisions (which, then, become the equivalent of a test of the theory or hypothesis behind it). The Kuhnian analogy between scientific communities and political ones was now given free rein and allowed to deploy all its effects, with due credits to Kuhn’s Structure, which became the organising reference of one of the key chapters of the book, dealing with ‘cognitive consistency and the interaction between theory and data’.

The sustained analogy between scientists and decision-makers that runs through Perception allows Jervis to follow the Kuhnian script very closely and to posit the paradigmatic nature of perceptions in international politics: ‘in the interpretation of other states’ behavior, and in the scientific laboratory, expectations create predispositions that lead actors to notice certain things and to neglect others’.Footnote ⁹³ This selective framing of reality, however, is perfectly rational, since the processes that consist in ‘fit[ing] incoming information into pre-existing beliefs … characterize investigations in science’.Footnote ⁹⁴ This, in turn, means that in the face of unexpected or incoherent information, decision-makers will most of the time stick to their beliefs rather than revise them, just as paradigms are not challenged by discrete anomalies. Even seemingly irrational cognitive distortion (evidence is ignored or twisted to accommodate existing beliefs) is a rational feature of the decision, for at least two reasons: first, it does not produce inaccurate pictures of reality, since by virtue of the theory-dependency of facts, ‘evidence always permits multiple interpretations’;Footnote ⁹⁵ second, just as ‘normal science’ could not operate and produce results if its fundamentals were permanently questioned, no sound decision could be taken if policy expectations were constantly changed and destabilised: ‘Like the scientist, the statesman who wants to create and maintain a view that has even minimal coherence will have to refuse to give full weight to evidence that others would see as discrepant with his beliefs … in politics too, ignoring discrepant information or assimilating it to pre-existing beliefs will perpetuate inaccurate images and maintain unsatisfactory policies, but these processes are necessary if decision-makers are to act at all’.Footnote ⁹⁶

In good realist fashion, any decision results into a mixture of good and bad, and international politics knows only partial, tunnel visions that will never capture the ‘truth’ of the situation. The only important question is to strike a balance between ‘stability’ and ‘flexibility’ of expectations or, in other words, to decide when normal science ought to proceed and when a paradigmatic revolution is in order. But Jervis’s wholesale adoption of the Kuhnian model ends up blurring the distinction between a ‘correct’ perception and a misperception, since the latter is perfectly rational and even necessary if decisions are to be made at all. The distinction, then, only makes sense in retrospect, after a paradigm shift that casts new light on prior ‘evidence’: ‘because established theories give a coherent, interrelated view of reality, contradicting facts cannot be appreciated until the theory is displaced’ and until, seen through new lenses, ‘the world looks different’.Footnote ⁹⁷ No judgment about the validity of specific inferences, therefore, can be dissociated from ‘the larger question of the merits of contending theories and images’. But in the Kuhnified decision process analysed by Jervis, these images are, like paradigms, incommensurable, and their merits cannot be adjudicated on the basis of a single set of rules: ‘the interaction between theories and facts renders difficult a debate on the merits of opposing paradigms, and, to a lesser extent, images’.Footnote ⁹⁸ In the end, like in Allison, no rational discussion of the ‘essence of ultimate decision’ can be conclusive.

Remarkably, both books displayed feats of social scientific erudition and yet ended up reasserting the traditional vision of the sovereign decision as something that exceeds the bounds of rationality and thwarts any attempt at rule-based codification. Not that this limitation was specific to Allison or Jervis: despite the vast resources poured into their development, the postwar decision sciences as a whole never delivered the goods they promised, and in particular a scientific formula for making decisions.Footnote ⁹⁹ This was certainly the case in the IR theory, where most of the work done on decision-making, including the more formal work of Schelling, Kuhn and others ended up emphasising the importance of the non-formalisable elements of the decision. Yet, the real stakes of the engagement with Kuhn laid elsewhere.

What was really new in the adoption of Kuhn’s Structure as the ‘paradigmatic’ account of science by IR theorists was the complete transformation of the terms of debate within the discipline and the new articulation of science and politics thus made possible. In many ways, the postwar social science movement was also an attempt at bringing science and politics together by bringing science into politics and transforming politics into applied social science. This is what triggered the reaction of political realists, who believed like Morgenthau that ‘a science of politics deals with a subject that is existentially alien to it’.Footnote ¹⁰⁰ Until Structure, scientific man and the man of power politics were fundamentally at odds, and this tension found an outlet in the episode known as the ‘second debate’, which opposed in the mid-1960s classical realists and a younger generation of scholars using system concepts.

If IR theorists found so much to like in Kuhn, it is because rather than pressing scientific concepts onto politics, he did exactly the contrary and brought politics into science. After Structure, being a realist and subscribing to an ideal of scientific rationality was no longer impossible, since scientific rationalism now appeared to rest upon non-scientific premises and even a dose of power politics. As the tension between realism and scientific rationalism disappeared, the second debate lost its raison d’être. The tension between history and science also subsided: a Kuhnified vision of history as a repository of analogies – read: paradigm-reinforcing cognitive consistencies – informing perceptions represented nothing short of ‘a useful shortcut to rationality’.Footnote ¹⁰¹ Science and politics apprehended the world in the same manner and with the same limitations. Adhesion to a partial and even flawed image of the world was necessary to make political decisions, just as the limitations of rationality were necessary to generate scientific knowledge. Kuhn was thus crucial in authorising realism to claim the ‘scientific’ credentials that the fledgling discipline had sought in vain since its inception, yet without relinquishing its vision of politics as a sphere of limited rationality. The price to pay, however, was the diffusion of a weakened concept of rationality that could never be opposed to political realism because it was already premised on it. Kuhn’s critique of the ‘textbook’ version of the history of science was indeed nothing else than another critique of idealism.