I'm back at work for a day or two which means, ironically, that I have time to respond to interesting posts again! Hopefully the momentum isn't entirely dead...
It would seem then the distinction between grammatical form and semantic content is not clear-cut.
But I agree with this! In fact this was part of the point of the OP: that grammatical categories
just are semantic categories. Grammar is not just a formal scaffolding of lingustic organization but reflective of - to use the Wittgenstinian lingo - a form-of-life. The whole point is that this informs Witty's statement that 'if a lion could speak, we would not understand him'. His form-of-life, reflected in his grammar, would be radically different from ours (not to be confused with 'in/commensurte' with ours). The focus on grammar here is to specify a
mechnaism which would explain how this difference would come about/operate.
That all said, and thinking a little bit more carefully about your comments, perhaps I was too quick to assimilate art and perception together as two categories to set 'against' language-qua-symbolism. I think on reflection that the category to set apart is
perception rather than art insofar as part of my motivation with the focus on grammar was to recognize the way in which it (grammar) allows for the creation of context 'out of thin air', as it were (also, reading back, our conversation began with a discussion of perception rather than art, and I think I allowed myself to get confused in the flow of it).
Anyway, the idea is that with grammar, I am no longer tied to a particular here and now, words can be used not simply as indexes or icons but as full blown symbols (to employ the tripartite semiotic distinction). This is something Dor gets at when he explains the specificity fo language:
"The claim is that the uniqueness of language lies in this very specific functional strategy. All the other systems of intentional communication, used by humans and the other species that we think we understand, work with different variations of the functional strategy that I call experiential: all these systems allow for (different variations of) the communicative act of presenting: “
this is my experience”. This very general characterization captures the foundational fact that experiential communication is inherently confined to the here-and-now of the communication event, where experiences can be presented.
Language is the only system that allows communicators to communicate directly with their interlocutors’ imaginations, and thus break away from the here-and-now of co-experiencing: instead of presenting the experience to their interlocutors for perception, communicators translate their experiential intents into a structured code, which is then transmitted to their interlocutors and instructs them in the process of imagining the experience –
instead of experiencing it. What the interlocutors do is use the code to bring back from their own memory experiences connected to the components of the code, rearrange them according to the structural configuration of the code, and construct a new, imagined experience."
link
So yeah, okay, language is indeed closer to art than I was willing to give credence to, but further away from perception insofar as perception is indeed tied to the affordances of the environment which acts as the 'external' grammar of our perception (it constrains what we perceive, even while our perceptions are co-informed by our 'sensory-motor schemas').