I've read an exam question given in a class on Semantics, that was asking
Why is language necessarily underspecified
I did not find much about this at the time, which is surprising because it intuitively strikes me as an important concept.
The class was using J. I. Saeed, "Semantics", which albeit does not index the term. The notion must have been given separately in class. The student couldn't guide me any further either. What I found is too formal, e.g. there's the Wikipedia article Underspecification focused mainly on phonetic and morphologic underspecification. A random .pdf jumped right inwith syntax trees. Now I found Ambiguity in the Stanford Encyclopedia of Philosophy, which in turn may be too broad (I should read it eventually).
At any rate, I hope the instructor was not asking the students to synthesize an answer. Rather, I suppose they referred the idea about "necessity" to a certain line of thought, a scholar, or seminal work, in the sense that e.g. "one cannot not communicate" is attributed to Paul Watzlawick, or "I know that I know nothing" to Plato's Sokrates, even if this isn't always completely accurate.
So, looking for an authoritative answer, this is basically a reference request. A discussion, or rather a summary of the context is welcome, but I'm affraid it would have to be necessarily underspecified and thus may be kept brief.
PS: A search for "semantics underspecified" yields many results. The reference request still stands, as long as I can't answer it myself, while I postpone flipping through the results (18 months later I am none the wiser; who developed the idea?)