I’m not strong on cognitive linguistics, but it’s an intriguing A2A.
What does it take for a word to become a conceptual metaphor? The meaning it expresses needs to be transferred to an analogous conceptual domain from its normal meaning; as a result of this, some of its meaning is preserved (the meaning that survives the transfer), some does not.
Now, function words clearly have meaning, though they do not always have denotation (they don’t always point to things in the world). But can their meaning always be transferred into an analogous domain?
That’s routinely true for prepositions. Prepositions historically tend to start off as spatial relations, and those relations are used as metaphors for other things all the time. If I say you’re on drugs, I’m not saying that you’re lying on top of some ecstasy pills. I’m saying that the dependency between you and drugs is analogous to the spatial proximity of something on top of something else—drawing on metaphors of closeness, coverage, foundations, and so on.
Conjunctions? They tend not to be spatial but conceptual already; but they’re no less subject to analogy. Transferring causation from the locutionary domain (causes for things happening in the world and described through speech acts) to the illocutionary domain (causes for speech acts themselves), for example: I’m late because there was a traffic jam (this happened because…) vs I’ll be there, because I’m a man of my word (I say this because…)
Pronouns and articles are trickier, but still doable. 1st and 2nd Personal pronouns presuppose personhood, but they can be metaphorically used of anything that can be individuated or in a collective. If I say “we spoons are dumped at the bottom of the pantry”, there’s a metaphor of personhood being imbued to spoons, alright; but the only place that metaphor resides linguistically is in the pronoun. Every other word is literally true.
https://www.quora.com/profile/Hu… issued a challenge on indefinite articles. Indefinite articles have a meaning: they indicate that the referent of noun phrase is not previously defined in the discourse, or is generic.
What’s going on when some boxer, say, says I’m not just *AN* athlete: I’m the best athlete there has ever been?
Well, the sentential stress on an should be telling you immediately that something unusual is going on: the whole point of the indefinite article is that it’s not something you emphasise, even contrastively. And the phrase cannot be literally true: a boxer is an athlete. An here is being used metaphorically: it’s being transferred from the conceptual domain of “generic” (which “an athlete” is: it’s a type) to the metaphorical use of “generic”: “uninteresting, ordinary” (which is not intrinsic to an: Usain Bolt is an athlete is true, and that does not imply he is uninteresting or ordinary).
The claim at first glance looks overwrought, but all words have some meaning, all meaning has at least a default semantic domain, and all meaning can be transferred to an analogous semantic domain, shedding some of its sense and retaining some of its sense. The only way that would be impossible is if a meaning were both atomic (and I’ll concede there are some semantic atoms), and not in a semantic network with other meanings and presuppositions (which is clearly not true, as we saw with we).
So… yeah. All words (including function words) can indeed be used in a metaphorical manner.