At least part of it is that social sciences/humanities is in some sense a 'default', so you see STEM types having hobbies or outside interests which fall into t

At least part of it is that social sciences/humanities is in some sense a 'default', so you see STEM types having hobbies or outside interests which fall into those, but then not vice-versa. Snow's "Two Cultures" thing remarks on this. Thus, STEM types will have a lot more to say about humanities than vice-versa. When you do Simonton-style science-of-science surveys and you survey, say, the National Academy of Science about their leisure-time activities, elite scientists often have deeply-pursued outside interests. Someone like Einstein playing the violin or E. T. Jaynes being so passionate about piano playing he has a whole collection of his own performances on a very expensive piano (https://bayes.wustl.edu/etj/etj.html https://bayes.wustl.edu/etj/etj.html) is not at all remarkable, and is the sort of thing mentioned in passing in obituaries; on the other hand, someone like Goethe or Nabokov who dabbles in optics or in lepidopterology is treated as so remarkable there's probably books written solely about those aspects of their lives. This is true today; there are plenty of STEM nerds who will dabble in odd things like translating Epicurean philosophy in their spare time, but there aren't a whole lot of Latinists whose pride and joy is decapping an old NeoGeo cartridge for the first time. (Now the question you doubtless have, and I had while reading these surveys, is what do the humanities types do in their leisure time? They don't answer that, but the impression I get qualitatively is that it mostly looks like 'either regular hobbies or even more of the same stuff as they do professionally'.) I think this is unfortunate, and the humanities would do much better appreciation of, critique of, and employment of everything in STEM areas if more of them crossed-over. (Sometimes reading philosophy of mind stuff, I think we would be much better off if it were illegal to publish anything in that field without at least an undergrad degree in either AI, neuroscience, or anthropology...) But I have no suggestions as to how to make that happen, since these tendencies seem to come as much from personality differences and preferences as much as any sort of institutional or economic constraint.