the basis of byproduct theory in the scientific literature, before it gets applied to higher level phenomena like religion, art, etc.

It seems to me that everyone uses that word quite liberally. But I think the general trend is to define cognitive byproducts as all mind-software that is compiled from more generic modules or runs on a lower level “circuitry”.

In one of the articles I linked to, you will find the ground for this reasoning right in the abstract — “despite differences in religious background, individuals show no difference in the pattern of their moral judgments for unfamiliar moral scenarios.” That’s why I brought atheists to the table, and why I’m fairly convinced that ancient animists were not religious (otherwise we should consider atheists as religious too, which would be an oxymoron).

So, while I agree that Gould and Vrba twist themselves into pretzels with spandrels, exaptations, and all that stuff, I’m sort of in the (very broad) byproduct camp too. But since spandrels would define the higher level cognitive phenomena as useful unintended consequences, and exaptations would imply that there is an underlying hardwired functionality that got hijacked and is now used for another essential purpose, I would say that both of these approaches miss the mark — they preclude that e.g. religiosity is a given and indispensable part of our cognitive and/or genetic makeup, which is (thankfully) not the case. And this also highlights another problem with Gould’s definition as he lumps higher level, but still primary, cognitive functions with forms they can take (that is, e.g. religiosity is only one form of cognitive map of universe).

Anyway, I’m looking forward to your article.