Linguists assume, as non linguists probably do too, that idioms and schemes like “what’s it to you?” are little blocks of language that are impervious to analysis — their parts can’t be separated; they’re stuck as one piece, almost like a single word.
So it’s surprising that among the handful of persistent problems I have is stumbling over idioms and schemes. I’ve been talking fluently on new topics ex tempore for a couple of years now, almost since the stroke, but even now I trip up on idioms. Two days ago I said, hesitantly and frustratingly, “Speak … to yourself,” meaning “Speak for yourself.”
The best in science is revelatory surprise, but a lot of it is unoriginal mimicry of naive folklore, a sort of grasping at straws. Here’s a case where maybe the folklore needs updating with new insight. If idioms and schemes are impervious, it[‘]s not because they are unanalysed in the mind. Maybe they are impervious just because they are habits, and habits become so familiar that they aren’t easy to break.
So, assuming that I process syntax or semantics more slowly and carefully than before, I wonder whether the idiom-habits are too fast for me to process, so
a [I — I often type ‘a’ for ‘I’: maybe because it’s a phonetic spelling] stumble on them.
The distinction between rule and habit is supposed to lie in productivity: a rule can range over many substitutions (I laugh, you laugh, they laugh; I slept, I left, I leapt); a habit is fixed (speak for yourself; quote your professor for yourself??). But what if rules are also habits but with lattitude for substitutions? That’s quite a compromise to Universal Grammar and innateness. And how would anyone tell the difference between a habit and an innate parametrized rule?
I often feel that syntacticians are engaged in an untestable, and so unfalisfiable, domain where the data are ambiguous and don’t decide between competing theories.