In all long-standing cultures, the great majority of musical instruments in use today have a history stretching back many centuries, during which their typifying timbres have evolved - if at all - in highly conservative ways. This suggests that their evolution might have been shaped in accord with certain deep-seated timbral preferences.
The origins of these preferences might be culture-specific and individually acquired through early-life exposure within a critical period, akin to infants' learning of phonemes specific to their native language. It's clearly evident that music listeners do tend to be better attuned to detecting and interpreting subtle expressive timbral nuances of instruments of their native culture than they are to foreign instruments, and that, for adults, the foreign-seemingness of non-native timbres doesn't greatly diminish even after years of exposure to them. On the other hand, certain basic instrumental timbres that feature across historically unconnected cultures - flutes and oboes, for example - might be attributable to innate acoustic preferences. Neither of these attibutions, however, account for our ability to fully accommodate the expressive subtleties of instrumental timbres native to our own culture that we first come to experience only as adults (e.g. lute, viols, saxophones). Or for why certain once commonplace instrumental timbres in our own cultural history - e.g., shawms, racketts, crumhorns, rebecs - didn't survive and seem foreign-sounding to our modern-day ears. And what is known regarding common-shared timbral preferences among individuals raised without any exposure to musical instruments, and are these culture-specific or cross-cultural? Clearly, much needs to be disentangled before any innate underpinnings of timbral preferences can be elucidated with any conclusiveness. Investigating the matter might yield insights - questions as well as answers - with implications not only for music-related science but for other fields with the cognitive sciences.
Research in speech phonology (by, for example George Lakoff and colleagues) has accumulated a considerable amount of clear evidence as to the existence of innate human phonetic/phonemic universals. I'd be extremely interested, and most grateful, to learn of any equivalent research investigating musical timbre preferences!
My thanks in anticipation!