Maybe the more important problem is: What is the smallest set of signs to most efficiently (or accurately) categorize, understand or communicate? Structure appears to also be important, but how and to what extent seems reliant on signs in the end. Regardless, these three purposes seem both inter-related and entirely independent.
If we assume that thought is ancillary to language, that we think by talking to ourselves, then understanding and communication are fundamentally similar. This requires signs to inherently carry meaning beyond their signifying function, or deeper meaning in language to rely on a long concatenation of metaphors. The latter seems more likely, but then the question of shared metaphors becomes rather pressing. How do these arise? How do we maintain or communicate personal meaning? This seems ridiculous to me. Perhaps repeated interaction forms notions of common signs, their meanings and their poetic uses, but metaphors themselves rely on experience, on understanding which predates language. Then again, it may be that their indexing function is what allows us to work beyond ourselves, so to speak. Placeholders are useful, even if solely personal. This suggests that language is merely a tool for thought.
The next assumption to be made is that categorization is required for understanding (and thus communication). This appears to rest on the assumption that differentiation and holism are logical opposites. If we understand holistically, then categorization is optional. If we understand through differentiation, then categorization is required. Of course, most likely both occur, though language necessarily favors the latter over the former. In fact, the greatest triumphs of the arts appear to be those which create or evince an experience much larger than the words or materials utilized. From this vantage, the question of aesthetic universality and difference becomes intimately related to the social sciences as critical observers of phenomena of the post-.