This post reminds me of this piece by Paul Krugman, who likened the history of economic theory-building to cartography.
Krugman suffers from a bit of economics myopia in this piece — his idea of maps is a bit too narrowly focused on econ-like perspective, and he is/was dangerously ignorant of what Hirschman and his disciples did after rejecting the econ-like approach, but this is still an insightful piece.
Krugman’s version of how economic theory is like maps is something like this: in the beginning, maps were incredibly rich and informative, but a lot of information that it contained were fanciful and wrong. There were dragons, griffons, and pelicans, the creatures cannot possibly exist. (The bit about pelicans is a semi-joke: when Carl Linne started his work on taxonomy, he assigned a lot of creatures that had been described in various works as fanciful creatures that cannot possibly exist, and these creatures included, among others, Pelicans.) When the scientific standards of evidence were applied to the information, a lot of wrong information were thrown out, but so were many accurate ones, and maps became increasingly empty because there was little or no good information available about underexplored regions. Only when these regions were explored and evidence meeting scientific standards of evidence collected, could they be filled in. So, in course of developing maps, much information is lost, then regained.
What Krugman is missing is that the loss and regaining of the information is done through the lens of economics. Hirschman and his disciples rejected this approach and joined another tribe — political science, for the most part. After they went “native” in the wrong tribe, their writings became undecipherable to the economists, and it took decades of reprocessing information into language that fit economics perspective for their insights to be understandable.
So the idea of a discipline-building works something like this: on day 0, the “founders” of a discipline somehow decide “this is how we are going to be thinking”; after that, much of the discipline goes about processing the “known” information through the lens of the divine commands of the founders; the information that fit is retained, information that can be made to fit is reframed; the information that doesn’t is thrown out, at least temporarily, until new theories can be developed. Essentially, theory building on a grand scale, with the caveat that axioms are defined a bit restrictively and everything will have to be built up from those axioms one step at a time. Not such a bad thing, but it does take a long time and different disciplines, with different founding fathers, will have different foundational axioms. As t → infinity, everything will eventually converge (hopefully), but infinity is a long time.
What your take on Dewey reminds me of is that if people hop across disciplines, but with a broader/deeper perspective, the process can be sped up. The founding axioms need not be taken as “gospel truths,” but simply founding axioms. Firms that do not maximize profit are not ruled out of existence a priori, because founding axioms of economics say so, for example, but simply a convenient assumption made to simplify theory-building (which is what it really is). A black swan is not proscribed because it does not exist, as per the founding axioms, but speculation about what a black swan might be like, if the founding axioms might be adjusted somewhat, would be permitted. (like the whole history of non-Euclidean geometry and Euclid’s Fifth Axiom (I hope I’m pointing at the right axiom).
PS. I had to wander off so I wound up cutting my train of thought short. So here’s the real wrap up. Hopping across multiple disciplines can, be difficult, in a sense, akin to learning multiple languages, and, perhaps difficult to do well. But it can be assisted by an attempt at building theory of theories, a la Stefan Banach, to frame the effort. The potential gains are considerable, both in the process of learning and applying knowledge to practice. One rather odd example that I came across in the past was how the Soviet and, later, Russian military is uniquely efficient (at least in principle) in coordinating triphibious operations. Given the way the Russian military is organized, all their staff officers, from fairly early on in their careers, are required to undergo a unified training programs, whereas in Western militaries, joint service staff training is something that fairly high ranking officers do late in their careers. This does emerge from an organizational drawback in the Russian military: the army dominates their military and air and naval forces are supporting arms for ground operations, rather than “strategic” arms in their own right, whereas much of interservice bickering in US military, for example comes from all the services trying to be the dominant branch while subordinating the others to back them up. But the consequence, at the middle levels at least, is that Russian staff officers, if properly trained, can play army, navy, and air force alike, while their Western counterparts are limited to their own branch. (I don’t know how true this is, since I have no firsthand experience, but I repeat the story often since it fits my way of thinking well. Kinda self-serving, eh?)