The Structure of Scientific Revolutions by Thomas Kuhn
Any exposure to the philosophy of science will at some point cover Kuhn’s Structure of Scientific Revolutions, somewhere between Popper’s falsificationism and Feyerabend’s methodological anarchism. In part the book is a response to a view of the history of science as a linear, cumulative enterprise from a dark age of superstition that heads ever closer towards Truth. In its place Kuhn introduces scientific paradigms, and divides Science into periods of “normal science” that are punctuated by “revolutionary science”.
Critics of Kuhn’s work have noted that he seems to have overloaded the word “paradigm”, but I think it essentially means the constellation of concepts and procedures that a scientific community coheres around. An example might be Ptolemaic astronomy, where a geostationary Earth was surrounded by crystal spheres that support the other planets, the Sun, and the stars. During a period of normal science, researchers are engaged in what Kuhn calls “puzzle solving”, where the paradigm is used to craft problems that can directly addressed by the methods and procedures of that paradigm. These periods are characterized by a cumulative increase in the number of solved problems and an increase in precision as scientists focus improving measurement techniques.
However, gradually anomalies will emerge (“epicycles on epicycles”) that remain unresolved despite great effort, and eventually an enterprising scientist will engage in revolutionary science where the paradigm itself is replaced. The example here would be the Copernican system with the Sun at the center of the World with the planets traversing circular orbits. If the new paradigm is successful by some criterion the scientific community will switch over (this can often be slowed down by at first skepticism and later intransigence) and a new period of normal science will resume.
That as such I think isn’t too controversial, as an interesting model of how science develops, though I think there is likely to more of a continuum between normal science and revolutionary science and varying degrees of paradigm shift extremity than the terminological dichotomy suggests. Kuhn enters more treacherous waters that incited backlashes when he talks about “the world itself changing” between paradigm shifts and how different paradigms are “incommensurable”. Kuhn tries to address these issues in an added postscript, which is evidence itself I think for the difficulty and confusion of his ideas.
I’m not entirely sure I understand what Kuhn means when he talks about the world changing when paradigms change. If two scientists from two different paradigms observe the same experiment, we expect them to receive the same stimuli, but Kuhn says their “sensations” will be different as how we perceive the world is model-dependent, such that the two will see different things. One response is that there’s an objective reality Out There and its simply the cognitive machinery of the two scientists that differs in interpretation. This has probably been the dominant view in science and philosophy since at least Descartes, though Kuhn argues that this itself is a paradigm. Again, I’m not sure I quite follow what he’s saying, and suspect he may simply be using the word “world” in a weird way.
In my opinion the main text of the book and the postscript seem to color the discussion of “incommensurability” quite differently, though Kuhn himself says that he’s always been consistent. In one sense its fairly trivially true, as terms will change meaning when paradigms shift; for example, space and time have different meanings in the Newtonian framework than in the Einsteinian framework. My first reaction was that scientists are smart people and can hold and compare two different conceptions, so the wording of “incommensurable” seems a bit strong. In the postscript, though, he seems to say something milder when he compares it to translation: there’s a difference between translating your language into another language and “going native” and thinking directly in terms of that other language. Kuhn argues against typical conceptions in science that newer paradigms simply reduce to older ones in some limit, like how Einstein reduces to Newton in the limit of small velocities, but I guess I don’t have nearly as much problem with such language as Kuhn.
In the postscript he argues for a kind of progress in science which certain readers of his had assumed he rejected. He imagines the history of science like an evolutionary tree of life, with branchings occurring with paradigm shifts. As you move up the tree, paradigms can address a wider array of puzzles-to-be-solved and with higher precision, and this is a type of progress. He seems to be agnostic as to whether there’s some ultimate Truth to Nature that we’re groping towards, and leaves the question open as to what kind of world we could inhabit that didn’t have such a Truth but in which Science would still work.
I can definitely see why this book seems to engender so much confusion and disagreement. Even with the postscript I still wasn’t sure at times how strongly Kuhn meant some statements. I’ve been very interested in “philosophy of” scientific disciplines, like the philosophy of biology or the philosophy of quantum mechanics. However, with regards to, say, Kuhn and Popper, the Grand Books of the canon of philosophy of science writ large, I’ve been more disappointed. I’ll read something like Structure, go “that was kind of interesting”, and then not really change how I think about Science. It didn’t really shake my naive perception that Science is a special discipline distinct from other things humans do by its tools and methods, secured with great difficulty over the centuries, that gradually discovers Nature’s Deep Truths and more and more approximates some objective reality. Even though Kuhn’s notions of paradigms is interesting, I don’t quite buy that, like evolution, it’s just a random-walk over some concept space. This naive view is a resultant of the meta-experiment of science, where we’d expect to find ourselves in a world as outlined above when we also find that science works so well.