Thinking in the shower this morning — really, why don’t I just spend the whole day there? it’s when I arrive at most of my best ideas — it occurred to me to summarize my area of scholarly interest as “systems of expression and inference.” That touches on the way that articulating and uptake constitute complementary aspects of the same process: we speak/write as “I want David to understand this when I address him, so I’ll say that, which seems most likely to evoke the reaction I want,” and we hear/read as “I’d most likely have chosen those words to evoke that reaction.” The expression and inference are systemically related, and no single “law of meaning” governs all such systems. They interact and deflect one another such that one can never fully isolate a natural sign or a conventional signifier and assert a single determinate meaning for it.
Thus assertions about reading “literally,” whether in favor or against, always operate by excluding pertinent contextual data; there’s no “literally” there. (Fred Clark has been pursuing this topic with his characteristic exquisite patience here and here.) Words never arrive at our attention without some accent or inflection, and if we devised a way to transmit them “neutrally,” that very “neutrality” would communicate some metatextual data, in the way that people frequently infer a great deal from a “robotic” voice. Words in a book signify differently from words spray-painted on a wall; words spoken in a flat, unmodulated tone signify differently from words whispered into one’s ear or shouted enthusiastically. But there’s no acontextual venue for words, so even the OED constitutes a context for meaning that affects interpretation (start, for instance, from its Englishness).
Anyway, the shower ended, so I have to go get grubby and shower again to figure out what comes next. But that phrase, “systems of expression and inference,” I want to save and return to.