top of page

A Kuhnian History of Science versus Scientific Realism

Scientific realism, as philosopher of science Stathis Psillos (1999) puts it, is the view that mature and genuinely successful scientific theories should be accepted as nearly true. This article aims to show that the so-called ‘Kuhn Cycle’—proposed in Thomas Kuhn’s picture of science's evolution—is not only incompatible with this view but directly opposes it.

The article starts by outlining the scientific realism debate, the very catalyst for theories like that of Kuhn’s (1962) version of the history of science. It then goes on to explain the pessimistic meta-induction (‘PMI’) and Kuhn’s radical approach in contrast to such a view. Be that as it may, the main ‘bulk’ of this feature is found in the latter half of the article, considering two different perspectives on Kuhn’s theory and its compatibility (i.e., its coexistence) with scientific realism. The first point of view, ‘Perspective one’, claims that Kuhn’s picture of scientific revolutions directly undermines scientific realism. ‘Perspective two’, on the other hand, claims that a Kuhnian picture can coexist with a particular variation of scientific realism, namely structural realism. In weighing up the two perspectives, however, the article shows and concludes that the first—less lenient—perspective is a more fitting way to think of Kuhn’s approach and scientific realism.

Figure 1. American historian and philosopher of science Thomas Kuhn (2021).

The Scientific Realism Debate

Supposedly, scientific theories describe the world one lives in. There is, however, a notable debate (known as the scientific realism debate) over their use of unobservable entities and mechanisms to describe phenomena that is observable (Psillos, 1999). For example, one must ask why scientific theories should be taken as true. More specifically, it must be questioned why one should believe that all these entities posited by the ‘best’ scientific theories are real (even when they’re unobservable or undetectable). This indeed is the debate over realism about such entities posited (i.e., claimed to exist) by theories, whereby realism—the view that scientific theories may be accepted as (nearly) true—is opposed by varying levels of anti-realism (Rowbottom, 2019).

Stathis Psillos (1999) takes scientific realism to incorporate three stances (as set out below, 1-3). Each realist dimension indeed results in multiple anti-realist alternatives proposed in response. The metaphysical, semantic, and epistemic dimensions of realism—reflecting what realism amounts to in the context of sciences—are, however, as follows:

  1. The metaphysical stance: the world has a mind-independent structure, meaning that it exists independently (i.e., regardless) of one’s thoughts and knowledge of it.

  2. The semantic stance: scientific theories are capable of being true or false. Firstly, theoretical assertions cannot be reduced to claims about the behaviour of observable phenomena. Secondly, theoretical assertions are not merely instrumental devices for establishing connections between observables. Instead, theoretical terms in scientific theories have a putative factual reference. So, if scientific theories are true, unobservable entities that they posit populate the world. Hence, semantically, realism is committed to a literal interpretation of scientific claims about the world. This is to say that realists take theoretical statements (i.e., the statements about unobservable entities such as subatomic particles) at face value, so to speak. For realists, claims about unobservable or observable scientific objects, events, processes, properties, and relations should therefore be construed literally as having truth values, whether true or false. As such, unobservable phenomena are not simply useful instruments for the prediction of observable phenomena, or for systematising observation reports (Chakravartty, 2007).

  3. The epistemic stance: this regards successful scientific theories as well-confirmed and approximately true of the world. Hence, the entities posited by them do inhabit the world.

As mentioned, each scientific realist stance (1, 2, and 3 above) progressively defends itself against opposing views known collectively as anti-realism (Psillos, 2005). Various ‘kinds’ of anti-realism, therefore, deny different parts of scientific realism. Reductive empiricism (Psillos, 1999), to give an example, denies only the claim that scientific theories give literal descriptions of their subject matter (i.e., rejecting the semantic stance). As the article will discuss later (when outlining ‘Perspective Two’), a Kuhnian picture of how science evolves could then coexist with at least some variation of scientific realism.

Figure 2. A particle track in a cloud chamber in the early 1930s was the first evidence of a positron, a positively charged particle (Conover, 2021).

The Kuhnian Picture

Anti (or ‘non’) realists often use the aforementioned PMI to undermine scientific realism (Müller, 2015). Rather persuasively, the PMI makes use of the history of science to do this. The view shines a light on the apparently ‘empirically successful’ scientific theories that are ‘successful’ merely subject to time (Müller, 2015). PMI theorists, therefore, argue that all the currently successful theories will very probably turn out to be false, just like previously accepted ‘successful’ theories such as the caloric theory of heat (see Figure 8) or aether theories, to name but a few examples of theories once accepted but later rejected in light of new emerging evidence of the time (Psillos, 1994). The empirical success of science thus does not warrant the claim that theories are approximately true (like the epistemic stance claims in scientific realism). More specifically, the PMI denies the first and second premise of Hilary Putnam’s (1975) famous No Miracles Argument (‘NMA’) in favour of scientific realism. The NMA is perhaps the most prominent argument in the scientific realism debate, contending that the truth of the best scientific theories is the only hypothesis that does not make the shocking predictive and explanatory success of science a mystery (Hartmann, S., & Sprenger, J. 2012). Put differently, it is no miracle that the very best scientific theories provide true descriptions of the world (or approximately so). Scientific realism, Putnam (1975) therefore argues, is the best explanation for such success. When the PMI denies this, it rejects the idea that the success of scientific theories can be understood (i) via the reference of theoretical (T) terms and (ii) via the truth of T terms (Müller, 2015). Broadly speaking, one simply cannot reconcile the history of science with the claim that successful theories are approximately true, according to the PMI view. The PMI seriously threatens scientific realism as such, since one cannot deny how the history of science has gone.

Most importantly, the PMI shows just how radical Kuhn’s (1962) version of the history of science is. Indeed, to reiterate, the PMI highlights those different parts of history that tend to favour different scientific theories (thus undermining their supposed ‘success’). Kuhn (1962), however, entirely rejects any ‘traditional’ picture of the history of science which suggests a cumulative process of improvement with scientific theories, as the PMI even might allow for. Instead, the history of science is a non-cumulative cycle which has become known as the ‘Kuhn Cycle’, as illustrated below in Figure 3.

Figure 3. The Kuhn Cycle (Henderson, 2021).

Scientific theories do not gradually evolve towards truth in this picture. Science, in Kuhn’s (1962) view, is instead a paradigm. To begin with, science has a paradigm which remains constant. Paradigms remain constant for a while, until a paradigm shift occurs. Here, scientific theories can no longer explain certain phenomena after all. New theories are then proposed, and the eventual result is a scientific revolution. Revolutions come about when (i) the new paradigm better explains the observations, thus offering a model that is closer to the objective; and (ii) the new paradigm is incommensurate with the old (Kuhn, 1962). This means that the new paradigm cannot be proven or disproven by any rules of the previous paradigm (and vice versa). Hence Kuhn’s use of the term revolution, as a new system is established in this pattern of events.

The Kuhnian picture of how science evolves is a picture of discontinuous theory change. Kuhn (1962) in fact notes that theory change would be genuinely cumulative if all theories were exemplified by the historical process through which science has developed. Yet this traditional view of scientific theories is false, claims Kuhn (1962). One does not witness any gradual progress; one experiences scientific revolutions. The history books only show this to be the case, as reflected by the example of the eventful history of continuously proposed but later rejected theories of light (see Figure 4 below).

Figure 4. Two rival (incompatible) theories of the nature of light were proposed in the seventeenth century, the wave theory and the corpuscular theory. Both were later abandoned (Xu, 2015).

Perspective One: Kuhn’s Picture Is at Odds with Scientific Realism

Evidently, the Kuhnian picture is quite radical. Hereafter, the article aims to show—via Perspective One (‘P1’) to begin with—that Kuhn’s drastic paradigm account is incompatible with full-fledged scientific realism. To this end, the article also considers an already mentioned counterargument (or counter-perspective) arguing that Kuhn’s account can be compatible with another form of scientific realism.

P1 holds that the Kuhnian picture of science cannot coexist with scientific realism. In particular, scientific revolutions undermine the realist story of scientific success (Psillos, 1999). Revolutions are in fact what Kuhn (1962) describes as a change of worldview. There is no gradual progress towards ‘truth’ via success like scientific realism (i.e., the traditional picture of the history of science too) suggests. For this reason, Kuhn’s (1962) non-cumulative cycle views the success of science entirely differently from realism. Whilst realism holds that one can trust science as a whole because of its general success and development, Kuhn’s picture only allows for progress within paradigms and prior to shifts. The overall success of science is therefore a myth, as is the argument that such success leads scientists towards the objective of 'ultimate truth’. This is Kuhn’s rejection of the traditional picture of science's history.

Figure 5. American philosopher, mathematician, and computer scientist Hilary Putnam, proponent of scientific realism in his NMA (Ben-Menahem, 2023).

Since Kuhn (1962) rejects the so-called ‘traditional’ picture of the history of science, he proposes a non-cumulative alternative which reflects the incompatibility of his view with scientific realism. One dominant paradigm is overthrown and a new paradigm is subsequently embraced, thus generating a ‘different normal science’ (Doppelt, 2012). Hence, and to clarify:

  1. On the one hand, scientific realism is the view that one should accept scientific theories as true, largely due to their impressive success rate (Psillos, 1999).

  2. On the other hand, the Kuhnian (1962) picture of how science evolves is a radical approach rejecting any traditional historical view of science. With each paradigm shift in the newly proposed historical picture, one sees a whole new set of criteria for success (i.e., a new worldview with each scientific revolution).

For (1) and (2) to coexist, it would have to be the case that both scientific revolutions occurred and that ‘successful’ scientific theories were accepted as true. Scientific realism says that the successes of science contribute to the very acceptance of theories as ‘true’. But how can one accept science as true when stuck in a vicious cycle whereby the success criteria changes with each scientific revolution?

It is difficult to deny that scientific revolutions undermine scientific realism. This is also compounded by the notion of a Kuhn-loss, as coined by Heinz Post (1971), describing a paradigm’s (or theory’s) loss of its predecessor’s ability to offer explanations in a certain puzzle-solving domain (‘puzzle-solving’ since scientists on Kuhn’s view solve puzzles: problems whose solutions reinforce and extend the scope of the paradigm rather than challenging it). The presence of Kuhn-loss certainly doesn’t imply any cumulative growth like in the scientific realist’s picture. The idea that scientific knowledge is cumulative (i.e., the claim that successor theories preserve all of their predecessor’s successes) is entirely different to what Kuhn (1962) proposes. A Kuhnian picture of how science evolves is thus entirely at odds with scientific realism. Somewhat amusingly, but to illustrate the point nicely, Kuhn (1962) asserts that what would have been ducks in a scientist’s world before a revolution are rabbits afterwards. It is thus hard to imagine how changes in worldview post-revolution could be compatible with unidirectional scientific growth in the scientific realist’s picture.

Figure 6. Quantum mechanics differs from classical physics in that quantities of a bound system are restricted to discrete values (van Holten, 2020).

Perspective Two: Kuhn’s Picture Is Compatible with ‘Unqualified’ Scientific Realism

Notably, the purpose of this article is not to pick a side between Kuhn and scientific realism. Alternatively, it is to decide whether a Kuhnian picture is—or could be—compatible with scientific realism. So far, the article has therefore explained why it is difficult to reconcile scientific revolutions with scientific realism. The essay must, however, now also consider another perspective on the matter: that perhaps Kuhn’s (1962) approach can coexist with a less ‘qualified’ form of scientific realism. This is Perspective Two (‘P2’), which situates itself between the two extremes. The idea is that, although Kuhn’s theory may not be compatible with full-blown realism, it could still coexist with a less demanding realist picture of science.

According to P2, one can argue that Kuhn’s (1962) version of the history of science is compatible with another variation of realism such as John Worrall’s (1989) structural realism. This view holds that science has identified real patterns, relationships, and structures in nature (Worrall, 1989). Seemingly, structural realists can maintain the NMA and avoid the PMI (making it arguably quite an appealing stance to take, which Worrall (1989) claims is the best of both worlds). Unlike full-fledged scientific realism, which holds that the nature of things is correctly described by the best (most successful) scientific theories, structural realism holds that one should be a realist about structure only (Worrall, 1989). So, what it is that instantiates a structure will not stay the same across theory change, but the structure itself does. Science on this view, therefore, discovers the structure of something in nature that bears these structures. Hence structural realism can avoid the threat of the PMI since it is not a miracle that science can account for its subject matter, nor is the history of science too dire after all. Further, in virtue of rejecting the realist explanation of empirical success in theories, structural realism may promise a better realist explanation since it is ‘more’ compatible with radical shifts in ontology (Doppelt, 2012).

Figure 7. John Worrall, creator of Structural Realism (LSE Library, 2010).

As discussed, the history of science shows non-cumulative theory change across time according to Kuhn (1962). This undoubtedly clashes with the realist picture; however, it may not be entirely different to structural realism since structure may be preserved across theory change. The two could arguably coexist. P2 is therefore an interesting take on the matter, whereby a less radical kind of scientific realism is compatible with Kuhn’s radical theory. Perhaps it can be the case that structure is maintained in theories post-scientific revolution. Put differently, scientific revolutions involve complete (and still radical) changes in the theory that instantiates a structure, but the structure itself remains the same (Worrall, 1989). Importantly, although the two views may not completely agree on the nature of scientific theories or scientific progress, they could still coexist. Their coexistence is simply easier to reconcile than that of Kuhn’s (1962) theory with developed, more full-blown, scientific realism.

Weighing Up

P2 is a desirable standpoint maintaining both a form of scientific realism and a Kuhnian picture of science. However, does this mean that a Kuhnian picture of how science evolves is compatible with scientific realism? For Kuhn’s (1962) account to coexist with scientific realism, P2 shows how one must alter scientific realism itself. This indeed involves another variation of scientific realism (i.e., structural realism) since the Kuhnian picture is utterly at odds with the full-fledged realist view. As discussed in P1, scientific revolutions in fact undermine the scientific realist picture completely. The two are by no means compatible. P1 consequently carries more weight. P2, on the other hand, is not necessarily far-fetched but highlights how the Kuhnian version of the history of science could only ever potentially co-occur with a certain form of realism.

Figure 8. Laplace's Calorimeter. The caloric theory of heat was later abandoned and replaced (Sella, 2016).


Kuhn’s (1962) history of paradigm shifts and scientific revolutions is difficult to imagine in the scientific realist picture, this article has argued. The feature indeed started by outlining the scientific realism debate, with the aim to emphasise just how radical an approach Kuhn’s theory is when it opposes the realist argument. Kuhn even goes further than the PMI and rejects the widely accepted traditional picture of the history of science (of cumulative growth towards truth). The so-called Kuhn Cycle is the result, which completely opposes scientific realism since it suggests a new world science with each revolution that occurs (Kuhn, 1962). Whereas the scientific realism view pictures the development of scientific theories as cumulative, unidirectional, all-encompassing, and triumphal success culminating in current science as the pinnacle of truth and knowledge (Doppelt, 2012), Kuhn’s picture of science does just the opposite. Cumulative and non-cumulative accounts are simply not compatible, as discussed at length via P1. P2 was then a counter-example of how one might unite Kuhn’s history of science with scientific realism. Broadly, however, this involves another variation of scientific realism. In weighing up the two perspectives, the article therefore concluded that a Kuhnian picture of how science evolves is incompatible with scientific realism.

Bibliographical References

Chakravartty, A. (2007). A metaphysics for scientific realism: Knowing the unobservable. Cambridge University Press.

Doppelt, G. (2012). Explaining the Success of Science: Kuhn and Scientific Realists. Topoi, 32(1), 43-51. doi: 10.1007/s11245-012-9135-x

Hartmann, S., & Sprenger, J. (2012). The future of philosophy of science. European Journal For Philosophy Of Science, 2(2), 157-159. doi: 10.1007/s13194-012-0052-z

Kuhn, T. (1962). The Structure of Scientific Revolutions, Chicago: University of Chicago Press

Müller, F. (2015). The pessimistic meta-induction: obsolete through scientific progress?. International Studies in the Philosophy of Science, 29(4), 393-412.

Post, H. R. (1971). Correspondence, invariance and heuristics: In praise of conservative induction. Studies in History and Philosophy of Science Part A, 2(3), 213-255.

Psillos, S. (1994). A philosophical study of the transition from the caloric theory of heat to thermodynamics: Resisting the pessimistic meta-induction. Studies in History and Philosophy of Science Part A, 25(2), 159-190.

Psillos, S. (1999). Scientific Realism: How Science Tracks Truth. Routledge.

Psillos, S. (2005). Scientific realism and metaphysics. Ratio, 18(4), 385-404.

Putnam, H. (1975). What Is"Realism"?. In Proceedings of the Aristotelian Society, 76, 177-194. Aristotelian Society, Wiley.

Rowbottom, D. P. (2019). Scientific realism: what it is, the contemporary debate, and new directions. Synthese, 196, 451-484.

Worrall, J. (1989). Structural Realism: The Best of Both Worlds?. Dialectica, 43(1-2), 99- 124. doi: 10.1111/j.1746-8361.1989.tb00933.x

Visual Sources


Author Photo

Rebecca Ivory

Arcadia _ Logo.png


Arcadia, has many categories starting from Literature to Science. If you liked this article and would like to read more, you can subscribe from below or click the bar and discover unique more experiences in our articles in many categories

Let the posts
come to you.

Thanks for submitting!

  • Instagram
  • Twitter
  • LinkedIn
bottom of page