Do Kuhnians have to be anti-realists? Towards a realist reconception of Kuhn’s historiography

It is quite unequivocal that Kuhn was committed to (some version of) naturalism; that he defended, especially in his later work, the autonomy of scientific rationality; and that he rejected the correspondence theory of truth, i.e., the traditional realistic conception of the world’s mind-independence. In this paper, I argue that these three philosophical perspectives form an uneasy triangle, for while it is possible to coherently defend each of them separately or two of them combined, holding all three leads to incoherence. Hence, the only option is to reject one of them in a way that is both compatible with Kuhnian historiography and philosophically fruitful. My purpose in this paper is not exegetical. My ambition is to sketch a coherent philosophical framework which does justice to Kuhn’s historiographical model for scientific change and avoids the philosophical costs of relativism and anti-naturalism. In particular, I argue in favor of the following interconnected positions: (1) Given the role of anomalies in Kuhn’s historiographical model, the latter satisfies the two fundamental conditions for acknowledging the mind-independence of the world: the ‘irreducibility condition’ and the ‘objectivity condition’, (2) Kuhn’s rejection of the notion of an ‘Archimedean platform’ does not entail that the truth-value of beliefs doesn’t play a role in the explanation of their change, (3) Adopting scientific realism is the only way to defend the autonomy of scientific rationality within a naturalistic philosophical framework. However, adopting realism leads to a liberal rather than an eliminative version of naturalism.


Introduction
sophical framework which does justice, first and foremost, to Kuhn's historiographical model of scientific change, and also avoids the philosophical costs of relativism and anti-naturalism. My account remains Kuhnian only with regard to the philosophical commitments which are presupposed by the cyclical model (e.g. the theory-ladenness of observation), but it is also revisionist with regard to philosophical views which have been explicitly defended by Kuhn himself on the issue of scientific realism. In fact, I suggest that there is an underdetermination of those philosophical views by the philosophical commitments of the cyclical model. In other words, the philosophical presuppositions of the cyclical model and the cyclical model itself are compatible with a range of philosophical views, not only those explicitly defended by Kuhn. Hence there is logical room for an account to be Kuhnian with respect to the cyclical model and revisionist with respect to the issue of scientific realism.
In what follows, I suggest that there is an alternative path which avoids both naturalistic relativism and Neo-Kantian anti-naturalism. This path adopts scientific realism, preserving both naturalism and the autonomy of scientific rationality, without being incompatible with the Kuhnian historiographical model. I argue in favor of the following interconnected positions: 1) Given the role of anomalies in Kuhn's historiographical model, the latter satisfies the two fundamental conditions for acknowledging the mind-independence of the world: the 'irreducibility condition' and the 'objectivity condition' (Psillos 2017). 2). Kuhn's (2000b) rejection of the notion of an 'Archimedean platform' does not entail, as he seemed to think, that the truth-value of beliefs doesn't play a role in the explanation of their change. It only presupposes that the 'subject-side components' (Hoyningen-Huene, 2021) of scientific knowledge are constantly attuned by the object-side elements. 3) Adopting realism is the only way to defend the autonomy of scientific rationality within a naturalistic philosophical framework. However, adopting realism leads to a liberal rather than an eliminative version of naturalism.
In particular, my argumentation takes the following course. In the next section, I present Kuhn's main philosophical insights. In the third section, I explicate why those insights form an uneasy triangle and how the Strong Programme and Friedman's Neo-Kantian view are options for avoiding the impasse of this uneasy triangle. In the fourth section, I present the parts of Kuhn's thought which prevented his adherence to realism. Finally, in the fifth section, I provide a sketchy realist-Kuhnian alternative to the Strong Programme and to Friedman's account. My conclusion is similar to Howard Sankey's (2018, 82) conclusion that 'it should not be thought that Kuhn's views are comprehensively opposed to scientific realism'. However, it is based on a different line of argumentation, for it takes realism as a general philosophical perspective for addressing the main problem of the philosophy of science, i.e., how to understand scientific belief modifications. Taken as such, scientific realism is symmetrical but contrasting to radical naturalism and neo-Kantianism. It is also based on a very specific version of scientific realism, i.e., 'the explanationist defense of realism, namely, the strategy of showing that the basic realist tenets offer the best explanation of the empirical and predictive successes of scientific theories' (Psillos, 2017b, 20, emphasis in the original).

The philosophical agenda and the philosophical insights
Before I examine Kuhn's main philosophical insights, I would like to refer to another aspect of the Kuhnian perspective which has significant philosophical importance in setting the agenda for the philosophy of science.

The philosophical agenda
According to Kuhn, the main problem for the 'philosopher who adopts the historical perspective' 3 consists in understanding small incremental changes of belief. When questions about rationality, objectivity, or evidence arise in that context, they are addressed not to the beliefs that were current either before or after the change, but simply to the change itself […] From the philosophical point of view, the difference between those two formulations-the rationality of belief versus the rationality of incremental change of belief-is vast (Kuhn, 2000, 112).
All major shifts in the philosophy of science consisted, first and foremost, in posing the crucial problems or questions of the discipline 4 . Logical positivists attempted to 'formulate the criterion which enables us to test whether a sentence expresses a genuine proposition about a matter of fact, and then point out that the sentences under consideration fail to satisfy it […] The criterion which we use to test the genuineness of apparent statements of fact is the criterion of verifiability' (Ayer, 1971, 15). In short, they took it that the main task of philosophy of science is to demarcate meaningful from meaningless statements. 5 Popper, on the contrary, stated that 'the problem which [he] tried to solve by proposing the criterion of falsifiability was neither a problem of meaningfulness or significance, nor a problem of truth or acceptability. It was the problem of drawing a line (as well as this can be done) between the statements, or systems of statements, of the empirical sciences, and all other statements-whether they are of a religious or of a metaphysical character, or simply pseudo-scientific' (Popper, 1962, 39). In short, the main task of philosophy of science is demarcating scientific from pseudo-or non-scientific statements.
Kuhn qua historicist rejected armchair philosophy. In the very first lines of The Structure of Scientific Revolutions (SSR hereafter) he noted: 'History, if viewed as a repository for more than anecdote or chronology, could produce a decisive transformation in the image of science by which we are now possessed' (Kuhn, 1996, 1). Rejecting armchair philosophy implies replacing its agenda. Formulating an abstract criterion is not the main task for the philosopher who takes history seriously. The main task is to understand, or to explain, scientific belief modification.

The philosophical insights
I suggest that Kuhn's main philosophical insights are related to the abovementioned task. Naturalism, Neo-Kantianism, and nonrealism provide -in a positive and a negative way-general guidelines to explain the changes in scientific beliefs. Given the Constitution Thesis (Psillos, 2009, 32) that Kuhn adopts from Hanson, the objects of scientific inquiry are constituted by the conceptual schemes 6 of scientific theories. Thus, radical belief changes amount to changes in the conceptual framework (CF) which is constitutive of scientists' experience. Therefore, the question is how to make the CF changes intelligible, i.e., how we should thoroughly understand the major changes of the conceptual categories of a scientific discipline. 7 Naturalism, Neo-Kantianism, and nonrealism entail different philosophical commitments in the course of answering this question.

Naturalism
As I have already said, Kuhn's work has been associated with naturalism. The essential engagement of historical study in the process of forming a philosophical picture of science is the first reason. 8 The second reason is that Kuhn, especially in the early stage of his career, followed a blatant Quinean path concerning the explanation of scientific change. A few years after the publication of the SSR he claimed: Until we can answer more questions like these [i.e., questions about scientific change], we shall not know quite what scientific progress is and cannot therefore quite hope to explain it. On the other hand, answers to those questions will very nearly provide the explanation sought. The two come almost together. Already it should be clear that the explanation must, in the final analysis, be psychological or sociological. It must, that is, be a description of a value system, an ideology, together with an analysis of the institutions through which that system is transmitted and enforced. Knowing what scientists value, we may hope to understand what problems they will undertake and what choices 6 'What I have been calling a lexical taxonomy might, that is, better be called a conceptual scheme, where the "very notion" of a conceptual scheme is not that of a set of beliefs but of a particular operating mode of a mental module prerequisite to having beliefs, a mode that at once supplies and bounds the set of beliefs it is possible to conceive' (Kuhn, 2000, 94). 7 Note that the philosophical prioritization of the question of CF changes coincides with the Kuhnian historiography. Given a dominant paradigm, the belief modification is dictated by the paradigm. During periods of normal science, the processes of belief modification can be fascinating from a scientific point of view but they are philosophically uninteresting. What is philosophically intriguing is the paradigm shift, that is, the processes which lead to CF changes. 8 As Alexander Bird (2012, 206-207) suggests '[t]here are two strands to Kuhn's naturalism. The first concerns his use of history […] The second […] concerns his willingness to deploy evidence drawn from psychology'. they will make in particular circumstances of conflict. I doubt that there is another sort of answer to be found (Kuhn, 1970, 21).
The aim of this passage is pretty clear. It gives a programmatic response to what Kuhn takes to be the main problem for the philosophy of science, viz. the explanation of scientific belief modification. He stresses that the explanation should, in the final analysis, be psychological or sociological. Thus, understanding scientific change thoroughly entails giving a description of the social and psychological factors which determine the belief modification of the scientists.
The response is programmatic and therefore philosophical, and more specifically, naturalistic. Within the broad framework of naturalism there is room for various disagreements. One may favor sociological research as the privileged way of making scientific change intelligible, as the proponents of the Strong Programme do. Others, like Bird (2005) and Nickles (2003), suggest that it is cognitive science which is appropriate to flesh out the Kuhnian programmatic aspiration. In this sense, Bird (2012, 211) claims that 'insofar as they take a predominantly externalist approach, students of science studies are not Kuhn's true heirs. Kuhn had very good theoretical reasons for rejecting externalism' 9 . However, these disagreements are not the issue here. My point is that in the abovementioned passage, Kuhn embraces a radical naturalist perspective. Irrespective of whether he favors sociology, cognitive psychology, or any other scientific discipline, the point is that making scientific change fully intelligible requires an appeal to the explanatory patterns of the empirical sciences. This is the only necessary presupposition in order to classify him as a radical naturalist.
One way to determine the radical naturalist perspective about scientific knowledge is by reversing Wilfrid Sellars's following quote: In characterizing an episode or a state as that of knowing, we are not giving an empirical description of that episode or state; we are placing it in the logical space of reasons, of justifying and being able to justify what one says (Sellars, 1997, 76). 9 The internalism/externalism distinction as it is employed in the HPS scholarship should not be equated with the externalism/internalism distinction in epistemology or in philosophy of language. In HPS, internalism stands for the view that scientific development is the result of 'internal', that is, intellectual or rational processes. For that reason internalists aim to describe those processes by reconstructing criteria of (scientific) rationality. One of the most characteristic examples of internalist perspective is Impre Lakatos' (1978) Methodology of Scientific Research Programmes. Externalists, on the other hand, seek to explain scientific development by appealing to psychological, sociological, or other contingent factors which are considered by the received view as 'external' to scientific practice. Paul Feyerabend is a paradigmatic externalist. In epistemology, the internalism/externalism debate applies both to theories of knowledge and to theories of justification. Internalism implies that epistemic justifiedness is a process exclusively internal to the cognizer. Externalists-like reliabilists (Papineau, 1993)--reject internalism. Finally, in the philosophy of language externalism refers to causal theories of reference, while internalism to traditional descriptive theories. Despite the differences one may observe an affinity between the aforementioned distinctions. In particular, externalism in all different areas can be associated with a general naturalist view. The determination of those affinities is a complex philosophical issue which rests way beyond the scope of this paper.
Radical naturalists advocate exactly what Sellars rejects, that is, making the episodes of knowledge intelligible by giving an empirical description and furthermore an (empirical) scientific explanation of these episodes. They claim that placing the episodes in the logical space of reasons, i.e., making them intelligible in normative terms, is just a provisional way of understanding them which needs to be reconstructed out of conceptual materials that belong to empirical scientific understanding. Scientific knowledge, just as every other part of human activity, should 'fit with the ordinary stream of explanation' (Turner, 2010, 30) which is no other than the available explanatory patterns of (some of) the empirical sciences. In this sense, radical naturalists about scientific knowledge are eliminativist concerning the normative content of the latter. Kuhn seems to embrace this radical naturalist view in the passage above since he claims that there is no other sort of answer to be found than the answer which takes into consideration the social and psychological factors which determine the value system of the scientists of a particular era. Those factors, of course, are made intelligible by fitting with the ordinary stream of explanation.
In the later stages of his career, Kuhn abandoned this kind of crude naturalism. He explicitly rejected the sociologism of the Strong Programme (Kuhn, 2000, 91, 110-111), which is one of the purest expressions of radical naturalism, while he stressed self-critically: When I first got involved, a generation ago, with the enterprise now often called historical philosophy of science, I and most of my coworkers thought history functioned as a source of empirical evidence. That evidence we found in historical case studies, which forced us to pay close attention to science as it really was. Now I think we overemphasized the empirical aspect of our enterprise (an evolutionary epistemology need not be a naturalized one) (Kuhn, 2000, 95).
I think that we should take into careful consideration the wording here. Kuhn says that he 'overemphasized' the empirical aspects of historical philosophy of science. This means, I take it, that he never altogether rejected the idea that the philosophical image of science should base 'itself upon observations of the historical record' (Kuhn, 2000, 118). Therefore, when he suggests that an evolutionary epistemology (i.e., his own view) need not be a naturalized one, he means radically naturalized. He does not reject a moderate version of naturalism. Naturalism in general entails the use of 'deliverances of various a posteriori sciences and other disciplines (notably psychology and history) in dealing with questions that are at least partially philosophical' (Bird, 2004, 337-338). Radical naturalism implies that those deliverances are the only means for replying to these questions. 10 While Kuhn, in his later work, rejected the radical version of naturalism and self-consciously endorsed a version of neo-Kantianism, he kept characterizing his view as historical philosophy of science (Kuhn, 2000, 105-120). Since he also kept understanding the history of science as an empirical enterprise (Kuhn, 1977a, 5), this shows the inherently naturalistic aspect of his perspective.

Neo-Kantianism
At the same time, Kuhn explicitly embraced 'a sort of post-Darwinian Kantianism' (Kuhn, 2000, 104) or Kantianism 'with moveable categories' (Kuhn, 2000, 264). While he never addressed the multiple ramifications that spring from postulating the moveability of the Kantian categories, 11 we can detect some points where his view coincides with the Kantian epistemology.
First, he explicitly acknowledged the role of a priori synthetic principles in science. Discussing Newton's second law, he says: I do not quite want to call such laws analytic, for experience with nature was essential to their initial formulation. Yet they do have something of the necessity that the label 'analytic' implies. Perhaps 'synthetic a priori' comes closer (Kuhn, 2000, 71).
The synthetic a priori principles rest on the heart of the lexicon (Kuhn's later term for the 'conceptual vocabulary of a scientific theory' (Massimi, 2005, 82) 12 and determine the framework of possible experience for scientists.
Though it is a more articulated source of constitutive categories, my structured lexicon resembles Kant's a priori when the latter is taken in its second, relativized sense. Both are constitutive of possible experience of the world, but neither dictates what that experience must be. Rather they are constitutive of the infinite range of possible experiences that might conceivably occur in the actual world to which they give access. Which of these conceivable experiences occurs in that actual world is something that must be learned, both from everyday experience and from the more systematic and refined experience (Kuhn, 1993, 331-332).
Second, in a way that is not typically Kantian but it is compatible with neo-Kantianism, Kuhn claims that belief modification in the historical course of science is a rational enterprise.
From the historical perspective, however, where change of belief is what's at issue, the rationality of the conclusions requires only that the observations invoked be neutral for, or shared by, the members of the group making the decision, and for them only at the time the decision is being made. By the same token, the observations involved need no longer be independent of all prior beliefs, but only of those that would be modified as a result of the change. The very large body of beliefs unaffected by the change provides a basis on which discussion of the desirability of change can rest. It is simply irrelevant that some or all of those beliefs may be set aside at some future time. To provide a basis for rational discussion they, like the observations the discussion invokes, 11 For a presentation of these ramifications see (Dimitrakos, 2016(Dimitrakos, , 2018. 12 For a thorough examination of the term see (Politi, 2020). need only be shared by the discussants. There is no higher criterion of the rationality of discussion than that (Kuhn, 2000, 113, emphasis in the original).
As the passage indicates, Kuhn acknowledges the rational character of scientific change despite the fact that he rejects the idea of a fixed Archimedean platform which serves as a universal and transhistorical standard for the rational evaluation of beliefs. Each time the basis for the evaluation of the modified beliefs is the very large body of beliefs which are shared by the scientists, and which are unaffected by the modification. 13 Third, as Hoyningen-Huene (2021, 24) points out: […] in a fundamental respect Kuhn's spirit is the same as Kant's: What we correctly take as objective reality is nevertheless somehow shaped by genetically subject-sided components.
In other words, the constitutive function of paradigms or lexicons is the same as the constitutive function of the Kantian categories and this function springs from the subject-sided components of cognition. Despite the fact that for Kant those components are absolute, while for Kuhn historically changeable, both of them consider as the main task of epistemology shedding light on these constitutive components. Finally, Kuhn embraces the Kantian notion of the thing-in-itself (Ding an sich): […] lexical categories, unlike their Kantian forebears, can and do change, both with time and with the passage from one community to another.

Nonrealism
The notion of the thing-in-itself brings us to the issue of realism. The rejection of the correspondence theory of truth is probably Kuhn's most permanent philosophical conviction. 13 Kuhn never thoroughly elaborated his conception of scientific rationality. In his mature work, he even provided some contradictory remarks on the issue. For instance, in he both says that he disagrees with Carnap's view that language change is merely pragmatic (Kuhn, 1993, 313-314) and that the logical status of the lexicon is conventional, and that the 'justification of lexicons or of lexical change can only be pragmatic' (Kuhn, 1993, 330-331). However, it is clear that he rejected relativism, and, in this sense, he defended some version of scientific rationality.
[…] what is fundamentally at stake is rather the correspondence theory of truth, the notion that the goal, when evaluating scientific laws or theories, is to determine whether or not they correspond to an external, mind-independent world. It is that notion, whether in an absolute or probabilistic form, that I'm persuaded must vanish together with foundationalism (Kuhn, 2000, 95).
The reason for the rejection in question is the lack of a fixed transhistorical, conceptual framework which plays the role of a fixed Archimedean platform.
Only a fixed, rigid Archimedean platform could supply a base from which to measure the distance between current belief and true belief. In the absence of that platform, it's hard to imagine what such a measurement would be, what the phrase 'closer to the truth' can mean (Kuhn, 2000, 115).
The argument goes as follows: P1 Converging to the truth requires a transhistorical fixed conceptual framework for assessing the convergence.
P2 Historical philosophy of science, and especially adopting the notion of incommensurability, imply that there can be no such conceptual framework.
Therefore, C Evaluating scientific theories is not to determine whether or not they correspond to an external, mind-independent world.
In short, Kuhn rejects the idea that the goal of science is to depict the external, mindindependent reality. In this sense, his view is incompatible with the correspondence theory of truth (Devlin, 2021). Nonetheless, Kuhn (2000, 115) is willing to retain an alternative conception of truth. There are various interpretations of this notion of truth (see Kuukkanen, 2021;Massimi, 2015;Marcum, 2015;Devlin, 2015). Despite the reasonable disagreements, most of the interpretations conclude that Kuhn held an epistemic or pragmatic theory of truth which is compatible with the defense of scientific rationality 14 . The epistemic notion of truth, though, is exactly what scientific realists reject (Psillos, 2017a). Therefore, it is probably an unequivocal conclusion that Kuhn's actual perspective is a version of nonrealism. Of course, one may object that the term 'nonrealism' is quite vague. In fact, it is a negative determination. It signifies what the Kuhnian perspective is not, that is, a version of scientific realism. However, this negative determination entails specific consequences for what I presented as the main problem for the historical philosophy 14 At certain points, Kuhn (e.g., 1993) appears sympathetic to Paul Horwich's minimalist or deflationary theory of truth, which is not an epistemic theory of truth. However, the focus here is not to provide an exegesis of Kuhn's work.
of science, i.e., the explanation of scientific belief modification. The following passage is indicative: […] I aim to deny all meaning to claims that successive scientific beliefs become more and more probable or better and better approximations to the truth and simultaneously to suggest that the subject of truth claims cannot be a relation between beliefs and a putatively mind-independent or "external" world (Kuhn, 1993, 330).
The first part of the passage rejects the epistemic thesis of scientific realism, the idea that 'mature and predictively successful scientific theories [are] well-confirmed and approximately true of the world' (Psillos, 1999, xvii). The second part implies that the truth claims, and therefore the successive scientific beliefs, cannot be explained on the basis of a relation between these beliefs and a mind-independent world. The same idea is expressed by Hacking (1992, 14) when he says: [t]he truth of a proposition in no way explains our discovery of it, or its acceptance by a [knowing subject], or its staying in place as a standard item of knowledge.
In short, while Kuhn (2000, 110) rejected the Strong Programme because, according to their view, '[n]ature itself, whatever that may be, has seemed to have no part in the development of beliefs about it', his own view about realism has an analogous effect: the truth as corresponding to mind-independent reality plays no role in the explanation of the belief modifications about this reality.

The uneasy triangle and the available options
Given nonrealism, truth as correspondence to mind-independent reality cannot be explanatory of scientific belief modifications. In particular, it cannot be part of the explanation of the radical CF changes. Thus, we face the following dilemma: either CF changes are explained exclusively by the empirical research of the various social or psychological (or other contingent) factors (radical naturalism), or they are made intelligible by appealing to (relatively) autonomous rational processes (Neo-Kantianism).

The strong programme as radical naturalism
The Strong Programme obviously goes for the first horn of the dilemma. It is a version of radical naturalism which favors sociology as the privileged discipline for the study of science. As some proponents of this view suggest, '[f]or the scientist the world is the object of study; for the sociologist it is the scientist-studying-the-world that is the object' (Barnes, Bloor, and Henry 1996). The Strong Programme retains the Constitution Thesis, i.e. the experience of the individual scientists is always con-stituted by a CF. But the CF changes can only be made intelligible through sociological research.
Does not individual experience, as a matter of fact, take place within a framework of assumptions, standards, purposes and meanings which are shared? Society furnishes the mind of the individual with these things and also provides the conditions whereby they can be sustained and reinforced.
[…] As well as the individual's sensory experience of the natural world, there is, then, something that points beyond that experience, that provides a framework for it and gives it a wider significance. It fills out the individual's sense of what that overall Reality is, that his experience is experience of. The knowledge of a society designates not so much the sensory experience of its individual members, or the sum of what may be called their animal knowledge. It is rather, their collective vision or visions of Reality. Thus the knowledge of our culture, as it is represented in our science, is not knowledge of a reality that any individual can experience or learn about for himself. […] Knowledge then, is better equated with Culture than Experience (Bloor, 1991, 15-16).
The rationale of the passage is clear. Science cannot be reduced to the experience of the individuals. This kind of experience is always constituted by a framework of assumptions, standards, purposes, and meanings (i.e., CF), provided by society. Thus, understanding scientific changes ultimately means giving a sociological explanation of the CF changes.
This conception leads to the famous impartiality and symmetry theses which entail that the sociology of scientific knowledge should.
[…] be impartial with respect to truth and falsity, rationality or irrationality, success or failure. Both sides of these dichotomies will require explanation.
[…] be symmetrical in its style of explanation. The same types of cause would explain, say, true and false beliefs (Bloor, 1991, 7). Impartiality and symmetry are totally compatible with nonrealism. In fact, they are necessarily incompatible with any form of scientific realism (Psillos & Shaw, 2020). Bloor (1999) is pretty straightforward on this: The point is that, on a naturalistic approach, 'correspondence' has to be seen as a relation which actors themselves assert or impute or accept, rather than something operating as a real cause. These assertions, imputations and acceptances are effects to be explained, rather than causes which can be cited in explanations. More specifically, 'corresponding', or not 'corresponding' to reality, are not causal relationships that bodies of belief bear to their referent. Beliefs do have causal connections to things in the world, but the words 'correspond' and 'do not correspond' do not capture those connections.
The Strong Programme retains two of Kuhn's philosophical insights: radical naturalism and nonrealism.
The problem is that radical naturalism and nonrealism cannot coincide with the defense of scientific rationality. If the explanation of CF changes is sociological, relativism is inevitable. 'The relativist crux [. .] is that there is no distinction between being taken-to-be-justified (by a community) and being-justified' (Psillos & Shaw, 2020, 408). In short, relativism entails 'relativism about epistemic justification' (Sankey, 2018, 74). However, the distinction between being taken-to-be-justified and being-justified is exactly what the defense of scientific rationality requires. We have to be in a position to distinguish between cases where scientists were merely thinking that a belief modification is rational from cases where the scientists were justified in thinking so. We have to be in a position to distinguish between genuine reasons from impostors 15 . Note that it is not sociology as a favored discipline which makes the distinction impossible, it is radical naturalism itself. Irrespective of whether it is sociology, anthropology, cognitive science, or any other empirical scientific discipline which is considered as privileged for making scientific change intelligible, the indistinguishability between being taken-to-be-justified and being-justified is the same. The indistinguishability stems from the elimination of the normative content advocated by the radical naturalists. Empirical research can only reveal what the actual scientists take to be justified, that is, what the actual scientists take to be rational belief modification. In the framework of radical naturalism there is no logical room for assessing whether or not the belief modification is actually rational.
In order to show that radical naturalism per se is incompatible with the defense of scientific rationality, let's take a look at another instantiation of it: Larry Laudan's (1990) normative naturalism. Laudan suggests that empirical study can reveal the successful methodological rules that promote pre-determined epistemic ends: • normative rules of epistemology are best construed as hypothetical imperatives, linking means and ends; • the soundness of such prudential imperatives depends on certain empirical claims about the connections between means and ends; • accordingly, empirical information about the relative frequencies with which various epistemic means are likely to promote sundry epistemic ends is a crucial desideratum for deciding on the correctness of epistemic rules; […] (Laudan, 1990, 46, emphasis in original).
In this context, empirical study can show the most effective normative rules for achieving specific epistemic goals. Normativity is preserved but also restricted to hypothetical imperatives. 16 But what about the goals? Which are the appropriate epistemic goals for science? In other words, what justifies us in pursuing one or another epistemic goal? The Laudanian normative naturalist faces a dilemma. Either 15 At this point, one may object that a pragmatist view could neither maintain the universalistic distinction between being taken-to-be-justified and being justified, nor lead to relativism. Pragmatism simply entails the distinction between warranted and unwarranted claims. However, it is worth considering that the pragmatist distinction itself could be further challenged. Is every distinction between warranted and unwarranted claims truly justified, or is it merely a seemingly justified distinction? Nevertheless, a detailed discussion of pragmatism is beyond the scope of this paper. 16 As the Humean line of thought dictates.
scientific epistemic goals are determined by empirical study too or they are determined in other ways (e.g., traditional conceptual analysis). Choosing the second horn of the dilemma equals abandoning radical naturalism, since empirical study ceases to be the exclusive conceptual tool for making scientific change intelligible. 17 Choosing the first horn of the dilemma ends necessarily up to relativism. For there is no way to distinguish between what epistemic goals are simply taken-to-be-justified by a scientific community and what epistemic goals one is actually justified to pursue. 18 However, the epistemic goals are constitutive of the notion of scientific enterprise for they distinguish it from other cognitive activities 19 . One may think that my argumentation repeats the well-known internalists' arguments against epistemological externalism (e.g. reliabilism). 20 In fact, there are several affinities between my argument concerning the necessity of distinguishing between being justified and being-taken-to-be-justified in order to preserve the notion of scientific rationality and the internalist argument against epistemological externalism. The main affinity is that both, epistemological internalism and my view, take genuine epistemic normativity --that is, non-reducible to empirical descriptions-as constitutive of the epistemic episodes. Nonetheless, there is an important difference. Internalism requires the cognizer's awareness of the reasons that ground her beliefs. In my view, the distinction between being justified and being-taken-to-be-justified could be preserved even retrospectively. 21 In other words, an account of scientific change which aims to defend scientific rationality needs to show solely that scientific development takes place according to some rational standards. It doesn't have to imply that each time actual scientists are aware of those standards. 22 But if my argument is sound, radical naturalism qua radical cannot retain the distinction, neither in 22 In this sense, my view can be seen as close to epistemological externalism too. The i/e debate in epistemology is marked by methodological individualism. The debate is about individual cognizers' beliefs, reasons, consciousness or, in short, 'subjective warrants' (Papineu 1993, 145). Making scientific change intelligible, on the contrary, focuses on scientific communities. As Kuhn (2000, 103) stresses 'The primary unit through which the sciences develop is thus, […] the group, and groups do not have minds'. This fundamental difference creates asymmetrical philosophical dichotomies. But discussing this kind of asymmetry cannot be the case here. For a further argumentation on the issue, see (Dimitrakos, 2021). a prospective nor in a retrospective sense. Employing exclusively empirical study to make scientific change intelligible leads necessarily to the total abolition of the distinction between being justified and being-taken-to-be-justified, and hence to the abolition of a genuine notion of scientific rationality.
Returning to the Strong Programme, we can see that it maintains the Constitution Thesis, radical naturalism, and nonrealism. However, exactly because it is an instantiation of radical naturalism, this view is fundamentally incompatible with the defense of scientific rationality.

Freidman's Neo-Kantianism
Friedman (2001,2002,2011), on the other hand, aims to complete Kuhn's neo-Kantian turn by detaching Kuhn's historiographical model -'our best current historiography of science' (Friedman, 2001, 43)-from Quinean radical empiricism and by incorporating it into the philosophical framework of critical philosophy. Friedman too adopts the Constitution Thesis, for he takes science to go through 'radical transitions between incommensurable or nonintertranslatable conceptual frameworks' (Friedman, 2001, 56). However, his account attempts to make CF changes intelligible in a way that preserves the autonomy of scientific rationality.
In every scientific theoretical system, there are empirical parts (e.g., the empirical laws) and constitutive parts. Empirical laws face the 'tribunal of experience' immediately, while the constitutive principles provide the conditions of possibility for the articulation and empirical testing of the empirical laws. The constitutive principles determine the context of possible experience of the scientists who work within a theoretical framework (Friedman, 2001, p. 85). These relativized and constitutive a priori principles are taken to be what Kuhn calls paradigms and their replacement amounts to the Kuhnian revolution (Friedman, 2001: 45). Since the two paradigms remain incommensurable in the Kuhnian sense (Friedman, 2001, 60, 98), the question is how we can defend a 'prospective notion of inter-framework rationality accessible from the point of view of the earlier framework' (Friedman, 2002, 186; emphasis in original). Friedman suggests that the prospective notion of rationality is ensured by the role played by scientific philosophy. There are three key points in his account of scientific change: [F]irst, that the new conceptual framework or paradigm should contain the previous constitutive framework as an approximate limiting case, holding in precisely defined special conditions; second, that the new constitutive principles should also evolve continuously out of the old constitutive principles, by a series of natural transformations; and third, that this process of continuous conceptual transformation should be motivated and sustained by an appropriate new philosophical meta-framework, which, in particular, interacts productively with both older philosophical meta-frameworks and new developments taking place in the sciences themselves. This new philosophical meta-framework thereby helps to define what we mean, at this point, by a natural, reasonable, or responsible conceptual transformation. (Friedman, 2001, 66, emphasis added).
Scientific philosophy is supposed to provide the conceptual context within which the replacement of one constitutive framework by another incommensurable framework is taken-also from the standpoint of the old framework-as a rational process 23 .
Friedman's effort to fulfill Kuhn's aspiration for a neo-Kantian turn is remarkable. He addresses the philosophical worries springing from the moveability of the constitutive categories. At the same time, he retains the Kuhnian historiographical model and some of Kuhn's philosophical insights. His view is a version of nonrealism. 24 The successive scientific CFs converge to an 'ideal state of maximally comprehensive communicative rationality in which all participants in the ideal community of inquiry agree on a common set of truly universal, trans-historical constitutive principles' (Friedman, 2001, 67). Thus, science does not gradually approximate an external reality, but aims at a Perceian consensus, that is, an ideal consensus of an ideal scientific community (Friedman, 2001, 68). He also defends the autonomy of scientific rationality. Given the role of scientific philosophy, being-justified and being taken-to-be-justified is in principle distinguishable even in periods of extraordinary science and revolutionary changes. However, precisely because of the role of philosophy, Friedman's perspective is incompatible with every version of naturalism, which would reject the idea that philosophy can ipso facto guarantee the rational character of a scientific theory or a theory change. This kind of external vindication for science recalls the pre-critical (i.e., pre-Kantian) role of first philosophy as the ultimate source of warranting and it is exactly what philosophical naturalism rejects.

Kuhnian misunderstandings: What is NOT a problem for realists
Let's sum up what I have said thus far. According to Kuhn, the main problem for the historical philosophy of science is the explanation of scientific change, and in particular the explanation of the shift between incommensurable CFs. His insights concerning the general response to this problem include nonrealism, naturalism, and a neo-Kantian defense of scientific rationality. However, these three perspectives form an uneasy triangle, for they lead to an explanatory conflict between incompatible views concerning scientific change. The Strong Programme retains nonrealism and radical naturalism and abandons the defense of scientific rationality. Friedman retains nonrealism and the defense of scientific rationality while he abandons naturalism. Given nonrealism, Friedman's view and the Strong Programme, despite their vast 23 Kuhn suggests that there is a small set of epistemic values that are 'widely shared among different communities' (Kuhn, 1996, 184). One may think that those values could secure a prospective notion of inter-framework rationality, but Kuhn himself never arrived at this conclusion. This is probably because, as he stressed, '[v]alues like accuracy, consistency, and scope may prove ambiguous in application, both individually and collectively; they may, that is, be an insufficient basis for a shared algorithm of choice' (Kuhn, 1977b, 331, emphasis in original). 24 '[…] although I have argued for an essential element of convergence in the historical evolution of successive constitutive frameworks, this is explicitly not convergence to an entirely independent "reality" (however conceived) but rather convergence within the evolving sequence of constitutive frameworks itself. Indeed, at the end of Lecture III I suggested that the present conception of scientific rationality does not even require that there be a uniquely correct sequence of convergent successor theories -something that would certainly be required by any version of "scientific realism"' (Friedman, 2001, 118). differences, are characterized by an explanatory symmetry. They seek an explanation of scientific change by taking into account the subject-sided components of scientific knowledge. The Strong Programme advocates that those components should be examined with the conceptual tools of the empirical sciences (sociology in particular). Friedman, on the other hand, rejects naturalism and takes the conceptual analysis provided by philosophy as the privileged way of making radical scientific changes intelligible.
I suggest that these are not the only options for those who adopt Kuhnian historiography. Nothing prevents someone who adopts the Kuhnian historiographical model to embrace a version of scientific realism. Embracing a version of scientific realism is compatible with the Constitution Thesis and the defense of scientific rationality. Furthermore, this strategy of defending scientific rationality does not violate the general framework of naturalism which is dominant in contemporary analytic philosophy. In what follows, I shall argue that (a) Kuhn's view is compatible with the metaphysical doctrine of scientific realism: 'The world has a definite and mind-independent structure' (Psillos, 2009, 4); Kuhn's view, even pace Kuhn himself, does not necessarily entail that the mind-independent world is undiscussable, undescribable, or ineffable. Furthermore, (b) incommensurability is not an insurmountable obstacle for scientific realism (Sankey, 2017(Sankey, , 2018, and (c) scientific realism does not entail a teleological conception of science.

The lack of the Archimedean platform and the mind-independence of the world
Let me first discuss the issue of mind-independence. I follow Psillos (2017, 211) in breaking up 'the realist commitment to mind-independence into two components: irreducible existence and objective existence': Irreducible existence means existence in its own right; that is, not dependent on the existence of something else. To say that a kind of entities X exist irreducibly is to say their existence is not grounded, in whatever way, in the existence of another kind of entities Y; or it does not depend on the existence of another kind of entities Y. Hence, to say that entity K exists reducibly is not to say that K is not real but that its reality depends on (or is reducible to) the reality of some other entity (or entities) (emphasis in the original).
The rejection of the irreducible existence of the world is advocated only by idealists à la Berkley. Kuhn explicitly denies this kind of rejection. As he points out: […] the metaphors of invention, construction, and mind-dependence are in two respects grossly misleading. First, the world is not invented or constructed. And he continues, Can a world that alters with time and from one community to the next correspond to what is generally referred to as "the real world"? I do not see how its right to that title can be denied. It provides the environment, the stage, for all individual and social life (Kuhn, 2000, 101-102). Therefore, the first component of the realist commitment to mind-independence is totally in line with the Kuhnian view.
Things are more complicated with the second component.
Objective existence means existence independently of epistemic or cognitive conditions that require the verification, recognition or knowledge of existence.
To say that a kind of entities X exist objectively is to say that their existence is not in any way constitutively connected with epistemic procedures Φ that allow or enable cognizers to decide, or otherwise certify, their existence (Psillos 2017, emphasis in the original). 25 As Psillos stresses, a characteristic example of denying objective existence is the Dummettian justificationism: On a justificationist view […] what we could have known extends only so far as the effective means we had to find out: the entailment is not from its being true to the possibility of knowing it, but in the opposite direction. It would be wrong to say that we construct the world, since we have no control over what we find it to be like; but the world is, so to speak, formed from our exploration of it (Dummett, 2006, 92, emphasis added).
The requirement of objective existence is incompatible with the epistemic theory of truth according to which truth is identified with the result of some epistemic practices. In this view, there is no possibility of divergence between the result of a successful epistemic practice and truth. As we have already seen, it is widely accepted that Kuhn held something like an epistemic theory of truth. 26 This implies that his view denies the requirement of objective existence. But let's take a closer look at this. It is reasonable to suppose that an epistemic theory of truth fits the periods of normal science. During those periods, what is taken to be true is the result of successful epistemic practices dictated by the dominant paradigm. The paradigm provides the general metaphysical framework of the scientists (Kuhn, 2000, 168), the problems they need to solve, and the criteria of successful solving.
One of the things a scientific community acquires with a paradigm is a criterion for choosing problems that, while the paradigm is taken for granted, can be assumed to have solutions. To a great extent these are the only problems that the community will admit as scientific or encourage its members to undertake (Kuhn, 1996, 37).
In this sense, the world is formed by our exploration of it.
However, the periods of normal science are not the most interesting ones from a philosophical point of view. The crucial point, as we have seen, is the explanation of the transition from one CF to another. In the course of a paradigm shift, the role of anomalies is central. Strictly speaking, the existence of anomalies is a condition of possibility for paradigm change.
Anomaly appears only against the background provided by the paradigm. The more precise and far-reaching that paradigm is, the more sensitive an indicator it provides of anomaly and hence of an occasion for paradigm change. In the normal mode of discovery, even resistance to change has a use that will be explored more fully in the next section. By ensuring that the paradigm will not be too easily surrendered, resistance guarantees that scientists will not be lightly distracted and that the anomalies that lead to paradigm change will penetrate existing knowledge to the core (Kuhn, 1996, 65).
The new incommensurable paradigm provides conceptual tools capable of resolving persistent anomalies. The new paradigm comes along with what we may call empirical relief: a better matching between the theoretical predictions and the empirical data. This matching 'relieves' the scientific community from the pressure of empirical problems.
I suggest that the role of anomalies in paradigm change fulfills the requirement of objective existence, for they ensure that there can be a gap between the world and the epistemic practices dictated by the paradigm. If there were only one constitutive framework, absolute or transhistorical, there would be no possibility of divergence between the constitutive principles of the framework and the world. Let's think about it by using the famous 'cookie cutter metaphor': 'The things independent of all conceptual choices are the dough; our conceptual contribution is the shape of the cookie cutter' (Putnam, 1987, 32). According to an absolutist conception of the conceptual scheme, the fundamental outlines of our empirical images of the world would be determined by this very scheme. The things as empirical objects could never fail to conform to the absolute or transhistorical cookie-cutter. In this case, the requirement of objective existence is not fulfilled, and therefore the criteria of scientific realists cannot be met. Only a Kantian version of empirical realism could be defended in this case. However, 'Kant's empirical realism can hardly honour the spirit of the objectivity condition' (Psillos, 2017, 216). The Kuhnian anomalies, on the contrary, ensure the 'resistance of nature' (Hoyningen-Huene, 1993, 75), i.e., that there can always be a divergence between the worldly objects and our conceptual schemes, and therefore the objective existence of the former is protected in our philosophical picture.
As we have already seen ( § 2.2.3), Kuhn (2000, 115) used to think that scientific realism requires 'a fixed, rigid Archimedean platform [that] could supply a base from which to measure the distance between current belief and true belief'. He used to think that the lack of a platform which rests 'outside of history, outside of time and space' prohibits a robust conception of truth as correspondence with reality, and therefore the adherence to scientific realism. But I think that this view is mistaken in any sense of the term 'Archimedean platform'.
If the term 'Archimedean platform' is taken as an absolute conceptual scheme, à la Kant, then, as we saw, this is exactly the opposite of what a scientific realist requires. A Kantian platform fails to meet the objectivity condition. On the contrary, the mindindependence of the world is secured, and hence, the philosophical presuppositions for scientific realism are provided only within a philosophical picture which lacks such an Archimedean platform. Given a Kantian platform, we could not measure the distance between current belief and true belief if truth is taken in its robust non-epistemic sense. If a platform like this were to exist, we could only assess the approximation to the phenomenal world constituted by this very platform.
If the Archimedean platform means something even stronger, like a neutral, theory-free observational vocabulary, as sometimes Kuhn (1996, 206;2000, 113) seems to imply, then scientific realism is presented simply as a strawman. Scientific realism does not presuppose some kind of neutral, so to speak, connection with reality without the mediation of our concepts. It does not presuppose the god's point of view or the 'view from nowhere' (Nagel, 1986). Kuhn appears to believe that a defense of scientific realism presupposes the possibility to 'get outside our [conceptual] skins', according to Donald Davidson's phrase. 27 But it simply doesn't. Scientific realism just entails that the explanation of CF changes should, in some way, take into consideration the relation between those CF and the mind-independent world. The relation is taken into consideration as far as the truth plays an explanatory role for these changes. 28 The fact that he mistakenly attributes to scientific realism the presupposition of a fixed Archimedean platform leads his philosophical reflection to recoil from this untenable presupposition and inauspiciously to a subjectivist path. From the mindindependence of the world he infers the world-independence of the scientific mind. The successive conceptual frameworks are represented as merely the subject-sided components of cognition, while worldly objects which are the object-sided components of scientific knowledge are depicted as 'mute', i.e. as ineffable, undescribable, and undiscussible like the Kantian Ding an sich. The ineffability of the object-sided components of scientific knowledge means exactly that they play no role in the explanation of the CF changes. The interaction between the subject-sided and the objectsided of scientific knowledge cannot be accommodated in this philosophical picture. CF changes are explanatorily reduced to changes in the enclosed sphere of the subject-sided components. The unwanted dilemma between naturalistic relativism and Neo-Kantian anti-naturalism becomes inevitable. The former makes the changes in question intelligible by placing them in the ordinary stream of empirical-scientific 27 'We have been trying to see it this way: a person has all his beliefs about the world-that is, all his beliefs. How can he tell if they are true, or apt to be true? This is possible, we have been assuming, only by connecting his beliefs to the world, confronting certain of his beliefs with the deliverances of the senses one by one, or perhaps confronting the totality of his beliefs with the tribunal of experience. No such confrontation makes sense, for of course we can't get outside our skins to find out what is causing the internal happening of which we are aware' (Davidson, 2001, 143-144). 28 I will elaborate more on this explanatory role in the last section. explanation and stripping them of normative content; the latter retains the normative content by appealing to the autonomous function of scientific rationality, which is secured by scientific philosophy and ends up violating the naturalistic consensus of contemporary analytic philosophy.
However, as I already said, scientific realism does not imply a neutral, theory-free Archimedean platform. The main point in Kuhn's argument is that there is no direct and unproblematic access to truth, and therefore, the evaluation (of the truthfulness) of beliefs and theories necessitates the employment of some other epistemic values (Kuukkanen, 2007, 557, emphasis in the original). But the realists can easily bite this bullet. Their argumentation begins when the employment of epistemic values becomes subject matter of philosophical examination. Therefore, the lack of an Archimedean platform is far from being a problem for scientific realism.
At this point, the orthodox Kuhnian could object against my revisionist account by pointing out that Kuhn explicitly rejected the mind-independent world of scientific realists, as he explicitly rejected the mind-dependent, constructed or invented world (Kuhn, 2000, 101). I can't argue that. Kuhn indeed sought for a middle way between those two images and he used the biological metaphor in order to achieve this middle way, which entails a 'mutual plasticity' between the world as subject matter of scientific enterprise and the practices of the enterprise in question. The key concept for this metaphor is that of 'nich' 29 : [W]hat replaces the one big mind-independent world about which scientists were once said to discover the truth is the variety of niches within which the practitioners of these various specialties practice their trade. Those niches, which both create and are created by the conceptual and instrumental tools with which their inhabitants practice upon them, are as solid, real, resistant to arbitrary change as the external world was once said to be. But, unlike the socalled external world, they are not independent of mind and culture, and they do not sum to a single coherent whole of which we and the practitioners of all the individual scientific specialties are inhabitants (Kuhn, 2000, 120).
The passage is pretty clear. Kuhn explicitly discards the view of 'one big mind-independent world' and adopts what Hoyningen-Huene (1993, 31) calls 'the Plurality-of-Phenomenal-Worlds Thesis'. Therefore, the orthodox Kuhnian could continue, the appeal to scientific realism is not just revisionist; it is not Kuhnian at all.
Against this conclusion, two points can be made. First, taking a closer look at the metaphor can reveal that it blocks scientific realism only because it mistakenly presupposes that the latter rejects the image of mutual plasticity. More particularly, in Kuhn's image, scientific realism necessarily implies the existence of an Archimedean platform which can measure the ongoing adjustment of scientific theories to the mind-independent world. Plasticity is only on the side of the theories. But, as I attempted to show, scientific realism should not be conflated with the view that any such platform is possible. A realist view can accommodate the idea that scientific theories are also shaped by idiosyncratic (sociological, psychological, etc.) factors. The only thing that precludes is the doctrine that those are the exclusive determining factors. Scientific realists aspire to show that the structure of the world also shapes our scientific image and this is exactly the picture of mutual plasticity. Kuhn aims to avoid radical constructivism which acknowledges only one, let me say, direction of fit: from the theories to the world. He also aims to avoid realism, as he gets it, which acknowledges only the opposite direction of fit: from the world to the theories. But if my argumentation is sound, this is a mistaken picture of scientific realism which confuses the latter with the view that an Archimedean platform is possible.
Second, the metaphor of mutual plasticity doesn't necessarily entail the Pluralityof-Phenomenal-Worlds Thesis. Mutual plasticity follows the rejection of empiricist foundationalism and the consequent theory-ladenness of observation. In this perspective, the scientific image of the world cannot rely on a theory free observational basis. It is constituted by our theories and our practices in general. Those practices, in turn, are shaped by various contingent factors. This view is a philosophical commitment presupposed by the cyclical model. However, why does the mutual plasticity thesis necessarily entail the variety of niches? The latter is of course Kuhn's view, no doubt about that. However, Kuhn needs an additional argument in order to show that the one thesis necessarily implies the other. But he provides none. The mutual plasticity thesis is compatible with Friedman's convergence to a Perceian consensus and with various versions of sociologism (e.g. Pickering, 1995). It is also compatible, I suggest, with scientific realism and the mind-independence of the world as long as this view ceases to be equated with the presumption of an Archimedean Platform. 30

Incommensurability and referential continuity
Given the lack of a fixed Archimedean platform, incommensurability is presented by Kuhn as a problem for scientific realists. In his later work, the term is exclusively expressed in linguistic terms, tied to the conceptual and meaning change. 'Applied to the conceptual vocabulary deployed in and around a scientific theory, the term 'incommensurability' functions metaphorically. The phrase 'no common measure' becomes 'no common language'' (Kuhn, 2000, 35-36). This means that incommensurability blocks the possibility of translating two successive paradigms into a common observational language and of assessing them in light of this common language. This also means that the referential continuity of scientific terms during the transition from one paradigm to a successive one is at stake; hence, the explanatory role of truth in paradigm change is equally at stake. In this sense, incommensurability is an obstacle for scientific realists.
Against this conclusion we should object that incommensurability alone is not a threat for scientific realism. The referential continuity is threatened by incommensurability in combination with a descriptive theory of reference and radical semantic holism. Only if (a) reference is fixed by descriptions and (b) it is fixed by descriptions that pertain to the whole range of the theoretical network, then the invariance of reference is at stake when there is a replacement of a paradigm by an incommensurable one. Actually, 'it is holism that does most of the work in getting radical reference variance' (Psillos, 2012, 217). However, two points can be made against this combination of philosophical views. First, scientific realists are not committed to the descriptive theory of reference. They have both developed causal theories and also hybrid causal-descriptive theories 31 of reference (Sankey, 2017, 212;2018, 81). While the former are hardly compatible with Kuhnian historiography, the latter do not entail such incompatibility. Second, Kuhn's later conception of incommensurability restricts the scope of semantic holism. He uses the term 'local incommensurability' and admits that '[t]he terms that preserve their meanings across a theory change provide a sufficient basis for the discussion of differences and for comparisons relevant to theory choice' (Kuhn, 2000, 36). This basis is also responsible for the possibility of rationally assessing belief modification during theory-change. The locality of incommensurability signifies the transition from a radical to a moderate form of semantic holism. Therefore, given these two points, incommensurability per se is not incompatible with scientific realism.

The aim of science and truth as explanatory concept
Finally, another point which obscures the possible convergence between Kuhnianism and scientific realism is Kuhn's confusing view about the correspondence theory of truth. In his words, the latter is 'the notion that the goal, when evaluating scientific laws or theories, is to determine whether or not they correspond to an external, mind-independent world' (Kuhn, 2000, 95). As Jouni-Matti Kuukkanen (2012, 136), stresses 'The first and arguably most important issue for Kuhn was the rejection of teleology in the history of science, that is, the conviction that science is not goal oriented in any sense of the word'. Thus, by depicting scientific realism as a view which takes science to aim at truth as correspondence with an external reality, Kuhn makes the convergence with the latter untenable almost by definition. The incompatibility is not with a view which takes science to aim at truth. The incompatibility is with a teleological view of science in general 32 .
Let me be more precise here. According to Kuhn (2000, 95), the correspondence theory of truth entails that 'the goal, when evaluating scientific laws or theories, is to determine whether or not they correspond to an external, mind-independent world'. This means, I take it, that the correspondence theory of truth has a direct implication for the main problem of the philosophy of science as we have already described it, i.e. the explanation of scientific belief modification. The implication is that given a scientific belief-change, b1→b2, b2's prevailing is explained on the basis of showing that b2 fits better to an external, mind-independent world. This image of scientific development is teleological in the sense that it depicts science as 'pulled from ahead' (Kuhn, 2000, 96). Kuhn compares his view on the evolution of knowledge 31 Psillos (2012). 32 It should be clear that the notion of a goal is not a part of the notion of correspondence. Truth is the goal of scientific enterprise. Therefore, the notion of a goal is inherent to the notion of science, at least according to traditional realist views.
with Darwin's view on the evolution of biological species. In one sense, if we extend the analogy, he compares all the traditional conceptions of scientific change with the Lamarckian teleological view on biological evolution. The correspondence theory of truth is an instance of this teleological view because it presupposes that scientific change can be made intelligible, first and foremost, by assessing the proximity of a theory with objective reality; therefore, it inevitably depicts scientific evolution as a process directed to a goal. But, according to Kuhn, this kind of assessment is untenable because of the lack of an Archimedean platform. The evaluations of successive theories 'are necessarily comparative: which of two bodies of knowledge-the original or the proposed alternative-is better for doing whatever it is that scientists do' [and hence] 'scientific development must be seen as a process driven from behind, not pulled from ahead-as evolution from, rather than evolution toward' (Kuhn, 2000, 96).
Teleological concepts are traditionally conceptually tied to normative concepts. 33 Kuhn understands the traditional view on scientific rationality 34 as teleological. Scientific beliefs get more and more rational, and this explains their change. They are 'moving' towards a final rational state. He also understands the correspondence theory of truth as an instance of this conception: Scientific beliefs get more and more true, and this explains their change. They are 'moving' towards truth. But this assessment is impossible given the lack of an Archimedean Platform. Kuhn aspires to untie normativity and teleology. He retains the former (i.e. scientific rationality) and rejects the latter. 35 He acknowledges that science advances based on certain rational standards but he rejects the view that this progress is towards some kind of final state. On the contrary, he takes this process to be 'driven from behind', which means that it is not assessed with regard to a final state but with regard to the previous theories. He ascribes the same anti-teleological view both to van Fraasen and to the Strong Programme. 36 Given that Kuhn understands the correspondence theory of truth as an instance of the traditional teleological conception of scientific rationality, he inevitably concludes that it is incompatible with his own anti-teleological (Darwinian) view.
Nonetheless, the teleological view of science is not necessarily the view of scientific realists. The 'explanationist defense of scientific realism' (Henderson, 2017;Psillos, 2017b) is not a teleological view on science at all. This view which is inspired by the 'no miracle argument', 37 suggests that scientific realism is the best explanation for the long-term empirical success of science. In this sense, 'the defense of scientific realism can only be a posteriori and broadly empirical' (Psillos, 2017b, 24). As 33 For a recent exposition of this conceptual tie in metaethics and in Philosophy of Biology, see Moreno and Mossio, 2015;Silverstein, 2016, respectively. The same tie can easily be revealed with regard to the philosophical comprehension of belief modification. 34 Scientific rationality is the central normative concept, according to the traditional view in question. 35 Kuhn's willingness to reject the teleological view on scientific changes is another aspect of his naturalism. 36 'And that is the case whether what scientists do is solve puzzles (my view), improve empirical adequacy (Bas van Frassen's), or increase the dominance of the ruling elite (in parody, the strong program's)' (Kuhn, 2000, 96). 37 'The positive argument for realism is that it is the only philosophy that doesn't make the success of science a miracle' (Putnam, 1975, 73). Chakravartty and van Fraassen (2018, 18) note, the difference between a realist and an anti-realist is 'not in what they believe about the natural world, but what they say about this belief'. In particular, in how they explain this belief. Scientific realists can be as anti-teleological as van Frassen is. They just add another explanatory step in the process of making scientific belief modification intelligible. van Frassen explains scientific change as an episode that enhances empirical adequacy, and scientific realists go further by adding another step which explains the improved empirical adequacy as 'symptom' of the fact that the new theory is approximately more true compared to the old one. Thus, pace Kuhn's obscure way of presenting the correspondence theory of truth, the latter can be a fruitful explanatory concept for scientific change, not an untenable goal of science. The explanatory role of truth can play a grounding role for scientific rationality and therefore it can place the Kuhnian historiography into a philosophical framework which avoids both relativism and anti-naturalism.

The realist alternative
Thus far, I have argued why Kuhn's historiography, pace Kuhn's own philosophical reflections, is not incompatible with scientific realism. Now I have to provide a realist alternative to the Strong Programme and to Friedman's neo-Kantianism. I cannot but be extremely sketchy here.
It should be clear by now that there is a particular kind of symmetry between naturalism, neo-Kantianism, and realism. They are all philosophical perspectives concerning the explanation of scientific belief modification. However, as I tried to show, radical naturalism and Neo-Kantianism end up necessarily focusing on the subjectsided elements of scientific knowledge. On the contrary, a realist context entails that scientific belief modification should be explained by showing how the subject-sided components are constantly attuned by the object-sided components of scientific knowledge. But how can this kind of attunement find a place in our philosophical image? In other words, what does the placement of the Kuhnian historiography into a realist philosophical framework entail?

Against radical naturalism: the ineliminability of normative explanations
The first thing that realist-Kuhnianism entails is, pace radical naturalism, that normative explanations are ineliminable. A normative explanation makes a belief modification intelligible by showing that it conforms to a norm or a set of norms. The employment of normative explanations is based on the assumption that '[r]easons must be capable of explaining [beliefs]' (McDowell, 1998, 108). The ineliminability of the domain of normative explanations implies that scientific change cannot be made intelligible exclusively by the conceptual tools of empirical sciences (sociology, psychology, etc.). Before I proceed, one clarification is needed. In some philosophical contexts, it may be crucial to distinguish between normative explanations which include the self-conscious application of normative standards by the actual scientists and normative explanations which do not. The former applies in cases where actual scientists achieve a belief modification following self-consciously some kind of epistemic rules. For instance, we can make the Copernican revolution intelligible by appealing to Copernicus' explicit preference for simplicity. The latter applies in cases which can be reconstructed as fitting to some rational standards without presupposing that the actual scientists follow those standards self-consciously. For instance, we can reconstruct the Newtonian laws of mechanics as untestable constitutive principles despite the fact that Newton himself took them as empirical generalizations. 38 But this distinction is irrelevant to my argument here. 39 All that is required is the admission that normative explanations, in general, are indispensable for making scientific development intelligible. This is the only requirement in order to defend the autonomy of scientific rationality. 40 Scientific realism offers a philosophical ground for the ineliminability of normative explanations within the framework of the contemporary naturalistic consensus. Following epistemic norms is a relatively stable (but not infallible) way of acquiring cognitive access to both observable and unobservable aspects of the world. Correspondingly, normative explanations of belief modification reveal specific aspects about the relation between the scientists and both the observable and unobservable parts of objective reality. They provide an extra level of understanding in comparison to empirical-scientific explanations. The latter can only cognitively grasp the peculiarities (sociological, psychological, or other) of the knowing subjects. Normative explanations, on the contrary, are meant to grasp the relation between the subject and the object of knowledge. If an epistemic norm is valid and if its application is correct, then the application in question leads-for the most part-to the truth. Accordingly, normative explanations express the relation between the rational subject and the world in a normative way. In this sense, they can also grasp the constant attunement of the subject-sided by the object-sided components of scientific knowledge.
A revisionist realist-Kuhnian perspective rejects first and foremost Hacking's conclusion (see § 2.2.3.) that truth plays no role in the explanation of belief modification. The role of truth is reflected in the normative explanations which should be taken as ineliminable. By grounding the autonomy of normative explanations in a robust notion of truth the following desiderata are satisfied: a) The distinction between being-taken-to-be justified and being-justified is ensured.
The normative explanations create this distinction by definition. Consequently, scientific rationality can be properly defended and relativism can be avoided. 38 Given his famous 'hypotheses non fingo'. The reconstruction in question obviously takes untestable constitutive principles as an indispensable, and therefore a rational aspect, of scientific enterprise, contrary to what Newton himself, as an empiricist, considered as (scientifically) rational. 39 I have made this kind of distinction (in Dimitrakos, forthcoming) and I have argued for the proper use of each kind of normative explanation. 40 The acknowledgment of the ineliminability of the normative domain brings my account closer to epistemological internalism. However, the admission that it is possible to apply normative explanations in cases where the actual scientists do not self-consciously apply the normative standards that correspond to the explanations employed makes the account look like some version of reliabilism. This is why I claimed in § 3.2 that the dichotomy between radical naturalism and the defense of the autonomy of the normative domain, as I present it here, and the dichotomy between internalism and externalism in epistemology, are asymmetrical.
b) The defense of scientific rationality takes place within a naturalistic context. Understanding normativity as an explanatory concept and also understanding the autonomy of normative explanations as being grounded in the framework of scientific realism does not entail an appeal to any kind of spooky entities or other mysterious epistemic faculties. It just takes truth into consideration in the course of explaining scientific belief modification. This view is only compatible with a liberal (McDowell, 2009;de Caro and Macarthur, 2004;2010, 2022 rather than a radical (or scientific or eliminativist) version of naturalism. 41 The former takes normative relations as part of nature and therefore the normative explanations as genuine, that is, not eliminable. The latter proclaims the explanatory reduction of normative relations to the explanatory patterns of empirical science. As long as we do not identify naturalism with its radical version, we can conclude that realist-Kuhnianism does not violate the contemporary naturalistic consensus.
One final remark is required. The ineliminability of the domain of normative explanation which, in my view, secures that truth plays a role in the explanation of scientific belief modification, in no way entails that normative explanations are the only kind of explanations that should be employed in the course of making scientific change intelligible. Since scientific theories are shaped also by idiosyncratic (sociological, psychological, etc.) factors (see § 4.1), empirical-scientific explanations are an indispensable part of understanding the historical course of scientific enterprise. 42

Against anti-naturalist kantianism: the direction of warranting
At this point, one may object that realism adds nothing to the general explanatory strategy concerning scientific belief modification in comparison to a non-naturalistic defense of scientific rationality, like Friedman's neo-Kantian account. After all, I just argued that making scientific change intelligible requires appealing to rational processes and therefore to the kind of explanations that I called normative. The same goes for Friedman's account. The fact that I ground normativity in a robust version of truth can be taken as a superficial dogmatic element which adds nothing essential to the general way that we explain scientific change. The key point, the objection may continue, is that we use normative explanations in order to make scientific change intelligible and involving a robust version of truth can make no substantial philosophical difference. However, I think that realism as a philosophical ground for the genuineness of normative explanations makes a great difference in explaining scientific change. The relation between science and scientific philosophy can only be exactly opposite to the one that Friedman's view entails. Endorsing scientific realism, and consequently naturalism, implies that the direction of warranting goes from science to philosophy and not vice versa. Advancements on the scientific level, which entail the expansion of the empirical space of reasons and, consequently, empirical relief from the extinc-tion of previous anomalies, can provide a rational warrant for the philosophical views that recapitulate the advancements in question. This means that philosophy, even if it can be a source of inspiration for scientists, 43 it cannot provide the ground for a prospective notion of rationality.
If scientific rationality consists in following epistemic norms and if following epistemic norms is a relatively stable (but not infallible) way of acquiring cognitive access to both observable and unobservable aspects of objective reality, then philosophy cannot dictate what these norms are. Philosophy can only, after each scientific revolution, recapitulate and reconstruct the norms followed by the scientists. Therefore, in the view I propose, our philosophical view of rationality is determined by the history of science in both senses of the term history: as historiography and as the historical course of science. The determination of the epistemic norms and, consequently, of the normative explanations that make scientific change intelligible, are subject to the constant revisions that result from both the historical study of past science and the course of present and future science. In this sense, science is depicted as a self-vindicating process. This picture is, I think, more compatible with both Kuhnian historiography and Kuhn's own philosophical perspective.

Conclusions
I argued that Kuhn's philosophical contribution can be found in setting the agenda for philosophy of science and also in three major philosophical insights: naturalism, nonrealism, and a neo-Kantian defense of scientific rationality. In particular, I took those insights as providing general guidelines for making scientific belief modification intelligible. I also argued that these insights form an uneasy triangle and this is why Kuhnians should reject one of them. I examined the choice of abandoning the defense of scientific rationality via the Strong Programme, as well as the choice to abandon naturalism via Friedman, and showed their disadvantages. I suggested the option of embracing the explanationist version of scientific realism as a philosophical ground for scientific rationality by contending that some views which actually blocked Kuhn's adherence to realism are unfounded. Finally, I provided an extremely sketchy realist alternative to both the Strong Programme and Friedman's neo-Kantianism. Against the former, the alternative I proposed claims that normative explanations are ineliminable. Against the latter and in accordance with naturalism, my view entails that philosophy cannot provide rational warrants for science and that science is a self-vindicating process.
My view is programmatic and, in this sense, far away from being complete. Many philosophical problems must be solved or dissolved in order for it to be viable. For instance, what kind of philosophical theory of reference should be employed in order to make the convergence between scientific realism and Kuhnianism a real possibility? What version of scientific realism should be favored for the same purpose? Furthermore, what philosophical theory of perception should we adopt? In general, what philosophical stance should we take vis-à-vis the relation between concepts and experience. These are some of the most crucial questions that should be answered. However, I hope that I showed that (a) the convergence between (the explanationist version of) scientific realism and Kuhnianism is not impossible, and (b) this convergence can lead to fruitful philosophical elaborations concerning the philosophical comprehension of scientific change. These elaborations coincide with Kuhn's fundamental philosophical insights and desiderata: the defense of scientific rationality, naturalism, the idea that science is a self-vindicating process, and the idea that nature itself has part in the development of beliefs about it.