When folk linguify
A recent entry into the growing field of linguification led me to do some thinking. There seems to be a subtype of linguification that has do to with, in a broad sense, collocations. That is, what sorts of words or parts of words can appear next to each other, in the same sentence, in the same paragraph, or document. Phrases like “word1 and word2 don’t belong in the same phrase/sentence/story” or “word1 rarely/often/always appears near word2″ are examples of these. This LL entry has a large summary of such stock phrases (some of which, however, are not about collocations, such as the “can’t spell x” variety).
There are some extensions: “You can’t mention word1 without also mentioning word2″ is also a slight variation of collocation pattern. Another, slightly less related sort of phrase is exemplified by something I heard on CNN the other day: “You can’t go one sentence without mentioning Barack Obama” [it proves itself!].
What it mean that people like to frame non-linguistic facts in terms of collocational “facts”? An initial hypothesis is this: people have a very naive folk theory about discourse. Namely, any given bit of discourse (from word up to document/conversation) is supposed to be on a single topic and is supposed to advance a single argument. Nothing that would potentially counter that argument (including entities that prototypically exhibit properties contrary/opposite to the topic*) can even be mentioned, let alone used to argue one’s point.
A potential sub-part of this theory (or perhaps just a corollary) is that discourses are saturated with positive epistemic stances. For instance, you do not argue X by saying (not A, not B, not C, not D, etc.). Similarly, a reason for X cannot be “not-A” (e.g., saying that you don’t like profession X because they do not have attribute A is not allowed, because you “rarely utter X and A in the same breath”). This may end up being a property of the linguification family of snowclones/constructions, and about the pragmatics of mentioning words (as opposed to using them), but it seems like a good working hypothesis.
(*From such a belief comes assertions like “‘linguistics’ and ‘practical’ don’t belong in the same/adjacent sentence/phrase/whatever”)