Shedding Disconfirmed Theories and False Prophets

The View from Oregon – 288: Friday 10 May 2024

In last week’s newsletter I repeated a claim that I have made previously, specifically, that there is no science of science. As I have been thinking about this claim over the past few years, the consequences of individual sciences (which I often call “the special sciences” in order to distinguish them from science simpliciter) being reasonably well-founded while the enterprise of science itself is, on the whole, not well-founded, have slowly unfolded for me, and every so often I see a new angle to it. I also discussed this in my Wilhelm Windelband and the Place of History among the Sciences, and here the new angle was my realization that idiosyncratic efficacy in science—some people are good at science and others are not—means that individuals and their personalities loom large. This partly explains the heroic narrative of the history of science, treated as an intellectual parallel to Carlyle’s “great man” theory of history.

I have also argued that, contrary to appearances, ours is not a scientific civilization. Many people who have not thought deeply on civilization have asserted that contemporary civilization is scientific. I would say that science plays a crucial role in the development of technologies, but when you step back from all the claims made about science today, it is easy to see that it is not science calling the shots. During the Enlightenment, there was an integration of Enlightenment political thought and science, and it seemed that Enlightenment ideologies translated into state structures (the American Revolution and then the French Revolution) would place science at the center of political deliberations. However, as Enlightenment ideology has developed over the past quarter millennium, it has increasingly diverged from science. With Enlightenment institutions effectively in charge of scientific institutions (higher education, scientific publishing, state funding of science, etc.), Enlightenment ideology has been gradually shaping science to suit its purposes.

If ours were a scientific civilization, it would be science calling the shots, and Enlightenment institutions would be bent to the will of science, and not vice versa. But we can easily see the problem here. Since science cannot make itself fully and completely scientific, science answers to the dominant personalities within scientific institutions, and these personalities can be as benighted and corrupt as any other human beings on the planet. One of things that I have learning by attending conferences over more than a dozen years is that individuals focused on their field can be rational, creative, and interesting in their work, but as soon as they leave their specialization, they are babes in the woods. James Burnham in his The Machiavellians: Defenders of Freedom, made a passing remark about Einstein’s views on economics, and this is very much to the point: an individual can be so gifted that they see something that everyone else has missed, but outside their field they endorse ideas that are insipid.  

So if Enlightenment era institutions were bent to the will of science, that would mean little more than that these institutions would be bent to the will of prominent scientists, and prominent sciences are men of their time: they would (and they do) simply repeat the platitudes of the society of which they are a part. This is how Enlightenment ideology triumphs over rationality, no matter how fulsome representatives of the Enlightenment are in the praise of science and rationality. It’s all hollow. It is a political program that is in the driver’s seat. And, invoking this metaphor, we can compare our Enlightenment civilization to Mark Manson’s consciousness car metaphor: reason has the map, but it’s sitting in the passenger seat, while emotion is at the wheel. So it is with our civilization: science has the map, but it is in the passenger seat, while Enlightenment ideology is in the driver’s seat, and it is free to ignore the advice of science (with the map) at any time.

Science in Enlightenment societies is not, however, a perfectly rational actor, in possession of a map that the rest of society lacks. Since institutional science, also known as “big” science, involves large institutions like universities and the government funding that universities receive for their research, these institutions have now shaped generations of scientists after their own image. So, to recur to the consciousness car metaphor, we have to imagine that science is not only sitting in the passenger seat, but that the only thing that science can see is the map it holds, and it must rely on reports from the driver about what the driver sees out the window: landmarks, signs, other traffic, people in the road, and so on, are all communicated to science by the driver, i.e., by Enlightenment ideology, and then science must attempt to figure out where the car is at, and how to get to the next destination. Needless to say, it is the driver making the choice as to what the next destination is to be. Hume, among the greatest of Enlightenment thinkers, said that “Reason is and ought only to be the slave of the passions.” I don’t think that Hume intended this as a prediction, but that is exactly where we find ourselves today.

One could plausibly argue that no institution at the center of a civilization has ever been a pure exemplar of itself and nothing else, a self-contained and autonomous monad, which is what I seem to be implying with the possibility of science practiced according to a science of science, and not according to the non-scientific whims of its practitioners. The institutional religions that served as the focal institutions of pre-modern agricultural civilizations had long histories and were embedded in a social and agricultural context that gave them meaning. We could point to any number of developments in these traditions where it was climate, geography, or available cultivars that were really driving the development of institutional religion, and not the other way round.

Not directly relevant here, but also not entirely irrelevant, I recently realized a structural similarity between Chinese and Western civilization, and this is that the axialization of these traditions did not emerge from a traditional mythology. The Greeks that their traditional mythology, but it was the expansion of philosophy, and especially the turn to moral and political philosophy after the Peloponnesian War, that served the function that appeared in other societies of the Axial Age through the medium of moralizing religions. China, too, has a traditional mythology, which is as marginal and as entertaining as Greek mythology today, but the axialization of Chinese civilization appeared through Confucianism. Of course, it has been traditional to count Confucianism as a religion, sometimes with certain qualifications, but Confucianism is no more of a religion than the tradition of philosophical ethics in Western civilization. I think it would more plausible to argue that Greek philosophical ethics made possible the appearance of the religious ethic of Christianity as it eventually claimed its central status within Western civilization than to argue that Confucianism is a religion.

The two other contenders here for status as an institutional religion determinative of a civilization are Hinduism, and Buddhism, which latter grew out of Hinduism. Zoroastrianism may have been a contender in its day, as Egyptian civilization was once a contender (not as a religion, but as a civilization), but both ceased to be influential before the advent of modernity. We could also count Judaism, but Judaism has never been a proselytizing religion, so it didn’t have the same influence on history of religious traditions that were actively expansionist. Christianity and Islam both appear well after the Axial Age, and, I would argue, as consequence of the Axial Age, continuing to play itself out in history over a civilizational scale of time.   

In any case, institutional religions as the central institutions of agricultural civilizations were in no sense pure in their provenance. Why should science need to be pure in its provenance—which, in this context, means being in possession of a science of science that completes the extant special sciences and provides a map for the indefinite elaboration of science? Arguably, science derives its legitimacy from its rationality, and this requirement of rationality can be turned against itself. Analogously, institutionalized religions derive their legitimacy from a metaphysical claim about the supernatural world, however, in the case of institutionalized religion, this claim doesn’t have traction when invoked reflexively.

The supernatural is the gift that keeps on giving: anything unprecedented can be credited to a new manifestation of the supernatural within the mundane world. It is subject to change without notice, so it can literally accommodate anything that happens, or anything that fails to happen. This is not the case with science. Rationality, too, is a gift that keeps giving, but in a rather different sense than the appeal to the supernatural. Disconfirmed theories can be shed as easily as false prophets, but the evidence upon which the disconfirmed theory is built is stubborn fact that cannot be wished away. Not only can stubborn fact not be wished away, but the new scientific theory that takes the place of the disconfirmed theory must not only explain everything that the disconfirmed theory explained, but it must also explain the result of the crucial experiment that became the pretext for the disconfirmation of the old theory. With false prophets it is rather different: any miracles attributed to them can be consistently denied with no knock-on consequences for the institutionalized tradition that managed to rid itself of the unwelcome prophet.

On stubborn fact—which I will assume is one and the same as brute fact—and returning to Windelband, Windelband provided a gloss (also quoted in my Wilhelm Windelband and the Place of History among the Sciences) on this nomothetic/idiographic distinction that is relevant here:

“A description of the present state of the universe follows from the general laws of nature only if the immediately preceding state of the universe is presupposed. But this state presupposes the state that immediately precedes it, and so on. Such a description of a particular, determinate state of the arrangement of atoms, however, can never be derived from the general laws of motion alone. The definitive characteristics of a single point in time can never be immediately derived from any ‘cosmic formula.’ The derivation of the description of a single temporal point always requires the additional description of the previously existing state which is subordinated to the law. General laws do not establish an ultimate state from which the specific conditions of the causal chain could ultimately be derived. It follows that all subsumption under general laws is useless in the analysis of the ultimate causes or grounds of the single, temporally given phenomenon. Therefore, in all the data of historical and individual experience a residuum of incomprehensible, brute fact remains, an inexpressible and indefinable phenomenon.”

The juxtaposition within empirical science that Windelband here describes between general laws and brute fact is the empirical parallel to the problem that the formal sciences face in regard to their foundations. A formal system must begin with axioms (or, today, formation rules and transformation rules, but it’s the same thing), and if we don’t accept the axioms, at least hypothetically, then we can’t go any further. Well, sort of. The knowledge of intuitive mathematics that is formalized in an axiom system could be pursued even if it is never formalized, or never fully formalized. Accepting partial formalizations is like accepting the results of the special sciences without a science of science that can be used as a point of reference that all is well in the special science. This is, indeed, the state of formal knowledge, as full formalization is rare, but it exists as an ideal, and the reasoning that lies behind this ideal has been developed relentlessly over more than two thousand years. The current compromise is to accept axioms in the spirit of hypothetico-deductivism: we don’t claim that they are true, certain, necessary, or anything of the properties traditionally ascribed to axioms; we only claim that we can’t go further in our deductions without accepting an axiom hypothetically.

Empirical science, as Windelband has shown, faces a similar dilemma, but the underlying reasoning has not been as relentlessly worked out. We have to accept brute fact in the same way that the formal sciences have to accept axioms, but in addition to the brute fact we also require the theoretical framework within which the brute fact can be rationalized, and this theoretical framework includes the mathematics that has, as we have seen, its own theoretical compromises. Thus empirical science has a double compromise, with the source of brute fact and with the source of the principles to which it must appeal if it is to rationalize brute fact. With this degree of theoretical complexity, it is no wonder that there is no science of science.  

---