Lakatosian Rational Reconstruction Updated

ABSTRACT

I argue in this article that an aspect of Imre Lakatos’s philosophy has been largely ignored in previous literature. The key feature of Lakatos’s philosophy of the historiography of science is its non-representationalism, which enables comparisons of alternative ‘historiographic research programmes’ without implying that the interpretations of history re-present or mirror the past. I discuss some problems of this interpretation and show specifically that Lakatos’s philosophy does not distort the history of science despite its normative ambitions. The last section is devoted to updating Lakatos’s programme to answer the needs of contemporary history and philosophy of science. The standard of rationality used in comparative assessments should be understood as a tool for measuring the coherence of an account of history with regard to the ‘actual history’. This standard takes two forms: framework-dependent and framework-independent rationality. The latter is decisive in comparative assessments.

Acknowledgements

I thank three anonymous referees of this journal for their detailed feedback and useful suggestions on an earlier draft of this paper.

Notes

1 For example, Lakatos (1978a, 104, 107, 129, 134, 138, 192); see also Proofs and Refutations, in which the contrast between a rational reconstruction and ‘actual history’ features prominently; on the term ‘actual history’ in Proofs and Refutations, see Lakatos (1976, 48n1, 84n1).

2 It is worth stressing that Lakatos’s ‘philosophies of science’ does not refer to the philosophy of science as a disciplinary field, in contrast, say, to the history of science. Rather, ‘philosophies of science’ are akin to interpretative frameworks.

3 For a detailed analysis of the narrativist philosophy of historiography, see Kuukkanen (2015, chapters 2–4).

4 Rankean philosophy of historiography provides the most famous expression of this aspiration through its demand to write history wie es eigentlich gewesen. See Kuukkanen (2012a) for an analysis of historical realism in contemporary historiography of science.

5 Two examples could be Christopher Clark’s (2012) thesis that European powers went to war like ‘sleepwalkers’, or Eric Hobsbawm’s (1995) claim that the twentieth century was ‘short’. For more, see Kuukkanen (2015).

6 Latour asks the historian to empty her mind and feign ignorance with respect to science and knowledge. Latour calls on the researcher to ‘just describe the state of affairs at hand’ (Latour 2005, 144), reject all theoretical frameworks and ‘add more details’ (Latour 2005, 137). The historian must merely follow ‘scientists around’ (Latour 1987), ‘the social fluid wherever it leads’ (Latour 2005, 29), and ‘the veins and arteries’ of science (Latour 2005, 77). ‘The name of the game is [thus] to go back to empiricism’ (Latour 2005, 144) and to spell out ‘what scientists really did in the past’. According to Latour, science studies has added ‘realism to science’ and it believes in ‘the objectivity of science more than anyone else’ (Latour 1999, 2, 3). Golinski has called this aspiration a ‘pretence of just telling like it was’ (Golinski 2005, 204).

7 Fittingly, if unorthodoxly, Lakatos defines historiographic positivism as a position, according to which ‘history can be written as a completely external history. For historiographical positivist history is a purely empirical discipline’ (Lakatos 1978a, 135n4; emphasis added). Lakatos’s view of positivism chimes surprisingly well with Latour’s description of his kind of empiricist social studies of science. This harmony between Lakatos’s criticism and Latour’s characterizations is also evident on the level of terminology. The actor-network theory (ANT) professor tells a student in an interlude in Reassembling the Social not to look for theories, but to produce more descriptions and collect more details (Latour 2005, 137). The ANT professor identifies himself as ‘a naïve realist, positivist’ (Latour 2005, 156).

8 Larvor and Motterlini challenge this reading and suggest that there is room for a ‘very un-Hegelian conception of truth’ (Larvor 1998, 64; Motterlini 2002). It is true that Lakatos refers to the representational ‘Blueprint of the Universe’ (e.g. Lakatos 1978a, 101, 154). There is no space to discuss this interpretation at length. Nevertheless, a number of relevant issues can be mentioned briefly. The first thing is that often his discussion about the ‘blueprint truth’ occurs in the context of Popper’s criticism, suggesting that Popper succumbs to scepticism without such grounding (e.g. Lakatos 1978a, 164–166; 1978b, 122). The second issue is Lakatos’s deep historicism, which does not leave room for absolute standards and evaluations. For example, a ‘fallible’ judgement of reliability or verisimilitude of Newton’s theory can be provided from the point of Einstein’s theory, but this kind of judgment cannot be given with most recent theories: ‘Thus we cannot grade our best available theories for reliability even tentatively, for they are our ultimate standards of the moment’ (Lakatos 1978b, 185). ‘Only God could give us a correct, detailed estimate of the absolute reliability of all by checking them against his blueprint of the universe’ (Lakatos 1978b, 185). Hacking also recognizes Lakatos’s aspiration with regard to the ‘Blue Print conception of Truth’, but concludes that ‘although he takes up the theme of an inductive principle from time to time he never does posit such a principle’, and that this conception is indeed almost always written with ironic capital letters (Hacking 1979, 386).

9 The dichotomy of ‘rational internal’ and ‘irrational’ or ‘non-rational external’ has been the subject of wide-ranging criticism by historians and sociologists of science. For useful overviews, see for example, Shapin (1992) or Golinski (2005).

10 Cf. also the following statement: ‘Unfortunately there is only one single word in most languages to denote history₁ (the set of historical events), and history₂ (a set of historical propositions). Any history₂ is a theory- and value-laden reconstruction of history₁’ (Lakatos 1978a, 121n1). In this article, history₂ is replaced by ‘historiography’.

11 It was shown above how recent historiography of science (and above all Latour) has recommended that historians should merely produce more (detailed) descriptions. More broadly, the historiography of science turned away from normativity towards non-judgmental descriptions of past science. Barry Barnes remarked that in practice ‘rationality’ acts as an evaluative, not an explanatory, term in philosophical schemes (Barnes 1972, 283; see also Zammito 2004, 133). According to Shapin, few historians see the consideration of what ‘the ideal type of the modern scientist should take into account’ as an ‘essential and proper part of their activity’ (Shapin 1982, 198). Another good expression of what has happened in the historiography of science in recent decades can be found in Shapin’s phrase ‘lowering the tone’: the task of a historian is ‘not to celebrate its [science’s] contribution to the future but to [just] describe and interpret its historical situatedness’ (Shapin 2010, 6).

12 It is true that Lakatos was concerned here with explaining progress and objective knowledge in the history of science, but there is no reason to assume that the same presupposition of rational reconstruction does not apply to any framework, whether or not it characterizes history of science in these terms. His claim that all historical interpretations always contain some ‘bias’ and some rationality principles essentially makes the same point.

13 Kuhn continues: ‘if history or any other discipline leads us to believe that the development of science depends essentially on behavior that we have previously thought to be irrational, then we should conclude not that science is irrational but that our notion or rationality needs adjustment here and there’ (Kuhn 1971, 144). According to Lakatos, if in an ‘actual history’ scientists did not abandon Newton’s gravitational theory after the discovery of Mercury’s anomalous perihelion (a ‘basic value judgement’ of scientists; Lakatos 1978a, 152), then methodology used to study this and other episodes of the history of science should learn from this. The methodology should give a rational account of this episode, and not interpret this behaviour of Newton as irrational. To put this in contemporary terms, this is surprisingly ‘naturalistic’.

14 For example, many historians suggest that science should be situated in a ‘context’ and its ‘decontextualisation’ or ‘delocalisation’ then explained (e.g. Schaffer 1991, 23; Galison 1997a, 781–844; 1997b; see also Kuukkanen 2011). Further, Latourian actor-network theory gives an account of science and its development in terms of networks (e.g. Latour 1987). See also Shapin (1992).

15 According to Larvor, ‘reason’ is the fundamental category in historical explanation for Lakatos (Larvor 1998, 76): ‘Lakatos believed in theory neutral universal rationality’ (Larvor 1998, 90) and that the rational method guides individuals (in science) in their reasoning and decision making (Larvor 1998, 79). Lakatos’s celebrated dissertation, later published as Proofs and Refutations (Lakatos 1976), exemplifies his Hegelianism at its best. It may be said that in Proofs and Refutations Lakatos wished to apply the idea of Hegel’s dialectical logic to the history of mathematics and to the debate on Euler’s proof; an attempt to represent the emergence of new concepts as a rational process (see Larvor 1998, 8; Lakatos 1978b, 5; see also Kadvany 2001, section I). It could therefore be said that Lakatos wished to write this kind of abstracted (dialogical) rational history of science, because he wanted to focus on what he saw as crucial in the dynamics of scientific thinking. Everything else appears inessential and the success of a ‘methodology of historiographical research programme’ (Lakatos 1978a, 131–132) can be measured according to how well it captures this operative rationality in science.

16 It is true that there have been some programmes for comparative historiography of science. However, they typically compare historiographical notions, rather than entire interpretative models, against history (Cohen 1994). Moreover, no explicit criteriology—save the presumption of neutral data—has been developed for comparisons (e.g. Donovan, Laudan, and Laudan 1988).

17 Some contend that historical data is theory-laden in the sense that it is interpreted through a theory or philosophical model and that there is no neutral arbiter to decide between them. Hull argued that if the historian approaches his data, say, from the perspective of a Darwinian theory, one ‘should not be surprised when  … [one’s] observations support  … [one’s] theory’ (Hull 1993, 471; for similar views of theory-ladenness e.g. Pitt 2001; Schickore 2011; Kinzel 2016).

18 Compare how Lakatos characterizes progressive shifts in the historiography of science. He identified progressive shifts from inductivism to falsificationism, and then from falsificationism to his methodology of research programmes: The latter ‘rehabilitated the scientific status of falsified theories like phlogiston theory’ (the ‘justificationist’ theories had pushed it and similar cases to the history of ‘prescientific preludes’). Furthermore, the methodology of research programmes provides a

coherent account of more old, isolated basic value judgments  … [while] for instance, according to Popper’s theory, it was irrational to retain and further elaborate Newton’s gravitational theory after the discovery of Mercury’s anomalous perihelion; or again, it was irrational to develop Bohr’s old quantum theory based on an inconsistent foundation. From my point of view these were perfectly rational developments. (Lakatos 1978a, 31)

In other words, the notions of coherence and consistency feature prominently.

19 A long list of successful non-referring theories can be found in Laudan (1981), 33.

20 Although there may not be so many strictly instrumentalist explanations in the historiography of science, there are many anti-realist accounts, which accept a similar explanation about success and failure in science. Kuhn and his idea of paradigms and fundamental scientific transitions is arguably the best and most influential example.

21 Cf. Psillos’s candid and apt remark: ‘I do not deny that my use of historical evidence is not neutral—what is?—but rather seen in a realistic perspective’ (Psillos 1994, 183).

22 For further discussion and developments of Psillos’s scientific realism, see e.g. Harker (2013); Vickers (2013, 2016).

23 A lot of older historiography of science and current philosophy of science relies on some kind of universalism in relation to science and scientific research about nature. See e.g. Kuukkanen (2012b).

24 Some prominent expressions of localism include: Latour (1987); Ophir and Shapin (1991); Livingstone (2003); and Secord (2004). For more discussion and further reference to localist scholars, see Kuukkanen (2011, 2012b).