Polysemy, the way one symbol can have multiple referents, seems linked to how big language is allowed to get; the smaller your collection of tools the more creative you have to be with them. Gestures in dance or theatre are 'naturally' polysemous because they must stand in relation to multiple concepts.
Some forms take this as a challenge - for instance the rakugo comedian who tells a complicated story without standing up or using more than 1-2 very mild props. Or formal bhara-natyam, where dancers are expected to perform alone for a period of roughly two hours and convey a story from a repertoire of eleven gestures.
But this irks me: from my notes I have what turns out to be a part-quote from Marie-Laure Ryan, that "while polysemy is good for poetry it is bad for science".
Is polysemy good for poetry? I've read a little philosophy of poetry and encountered the idea that what makes a poem 'unparaphrasable', its warp and weft of meanings, is an artifact of the poetic reading itself - much like hyperintensionality, the condition where replacing a term with a logical equivalent no longer works as we would hope, is a byproduct of a context where we care about something (anything) more than logical equivalence. Put more poetically, taking a thing as unique is the recognition of its own self-contained unity, where otherwise from synonymy springs a counterfactual forest.
And while sciences expand language by coining new terms, where we might say that the arts coin new meanings, that's all in service of an ultimate simplification. It benefits us very much to speak polysemically of things we once took to be unrelated, like electricity and magnetism or the motion of planets and the fall of pianos. The Standard Model and the axioms of ZFC both attempt to speak of, ground, all possible phenomena without losing their structure and dissolving into simple reflection ("for all A, A is A"). A model's use is the number of things it can faithfully represent, and useful models proliferate.
I'm being a bit unfair to Ryan, whose next (full) sentence is "It is better to work with a large collection of sharp tools that fulfill precise tasks, rather than a single blunt one, even if everyone cannot share the tools." This is pragmatically and straightforwardly true, and yet the struggle against polysemy that produces these tools, the attempt at doing/saying/pointing-out exactly one thing (and thereby taking that thing as unique), is just as desperately needed in the arts.
Polysemy is the limit natural to language and to communication in general, that there are always fewer words than there are things to be discussed with them. As LM once said, the fundamental issue isn't finding the right adjectives but that adjectives shouldn't have to exist in the first place. We live in a world composed only of uniquenesses, and communication can only serve us insofar as it obscures them.