In THE END OF SCIENCE (first published in 1996, re-edited with a new preface in 2015) John Horgan presents and argues for his thesis of the “end of science”: that we will not be undergoing and revolution in knowledge as substantive as those accompanying Newtonian Mechanics, Einstein’s Relativity, Quantum Theory, and Darwin’s Theory of Evolution. Inshort, that we have essentially “got it right” about the discoverable universe, and the rest is a matter of either accepting undiscoverable or filling in the details.
Horgan’s thesis is rather complicated and finally incoherent, but it is based on a monist model of science as progressive convergence on a single true account of the World. Despite talk about Kuhnian paradigms, it is premised on basically a cumulative vision of discoveries and problem-solving that does not sufficiently take into account conceptual change (the speculative dimension of science). Thus Horgan can only validate his thesis by dismissing bold speculative conjectures (such as multiverse theory) as un-empirical, or untestable in principle, when their testing is merely very difficult to implement.
Cosmologically, Horgan’s thesis of the end of science would be conceivable in a finite cosmos, but whether our cosmos is finite or not is an empirical question, not to be answered by naive inductive argument extrapolating from a very limited sample over a short span of time. In point of fact, Horgan gives us no reason to believe that we are living in a finite cosmos, and seems unaware that he is making an empirical, scientific claim. The discovery that the cosmos has only a finite number of explanatory levels and that we have exhausted them in principle, leaving only details to be discovered, would itself be a major scientific discovery (and a very hotly disputed one).
Further, even in an ontologically finite cosmos, Horgan would have to show that it is both epistemically and hermeneutically constrained. He would have to show not only a vertical limit to explanatory depth, but a horizontal limit to hermeneutic diversity of interpretation. Horgan’s argument is historicist in the worst sense, that of predictive historicism. It consists in extrapolating the future from a narrowly selected and one-sidedly interpreted scientific past.
A counter-example would be the reorganisation of the sciences and humanities around Gaia-in-the-Anthropocene, advocated by Bruno Latour. This would amount to a major revolution in thought, without this revolution conforming to the scientistic model based on modern physics and its pursuit of ever deeper explanatory levels. This reorganisation would lead to a different model of scientific knowledge, less preoccupied by the example of the history of fundamental physics over the last five hundred years and by its Platonic search for transcendence.
Horgan’s model of science as problem-solving envisions it as confronted with a finite list of pre-existent problems that only have to be checked off as they are progressively solved. He ignores or depreciates the speculative playing with concepts and equations that can generate new perspectives and new previously unthought of problems. In a nutshell, the more speculative the physics is the less it is scientific.
Yet Einstein’s Relativity began as just such speculative tinkering, and took many years before receiving empirical confirmation (the Eddington expedition was a bit of a fudge). Galileo was an experimentalist, whereas Einstein was not (he even ironised over the preoccupation with “little effects”, instead of being influenced by the beauty of theory). Galileo’s approach was probably much more speculative than is commonly realised, but he did a lot to hide this speculative component. Horgan’s argument seems to engage much more with a common ideology of science, promulgated also by scientists themselves, than with the practice of science in its full range from experimental to speculative.
Horgan combines an inductivist cumulative vision of progressive discovery of the truths about the world with a Kuhnian overlay. Both of these views are false, and combining them together does not make them any better. The inductive vision ignores the speculative creation of problems and the conceptual reorganisations involved in scientific progress. The Kuhnian view of unitary paradigm followed by revolution followed by new paradigm, etc, ignores the fact that there are multiple rival paradigms competing for favour at any one moment. Both fall foul of the “disunity of science” (Feyerabend, Dupré, Pickering, Latour) by treating scientific progress as far more homogeneous than it actually is.