flip-flop predicates

[Joyce Tang Boyland]

Along lines similar to Greg Thomson's:

My informal thought on how meanings can get and stay flipped
is that there's a lot more uncertainty+indeterminacy going on when
people use language than linguists often acknowledge.
In on-line sentence production, people are having to choose
something quickly that's in the right semantic neighborhood
(syntactic too, for that matter), even if they don't quite
"know" the "correct" form at that moment.  I made that argument
in my dissertation for why people sometimes say things like
"Why would've you done that?".  When you ask them in a survey
whether they'd say "Why would you have?" or "Why would've you?"
they are often puzzled, and say, Hm, I don't know.  When people
speak, though, they have to make these split-second decisions,
and all sorts of cognitive (and social) factors start
playing a part in what form gets selected.
Anyway, then, once the heterogeneity starts, it propagates,
(by frequency effects etc.), particularly when there is some
uncertainty in the first place.

If you think about the cognitive aspects of hypercorrection,
you have some of the same features -- reversal of the originally
sanctioned usage, arising (at least partly) from uncertainty and
surviving (at least partly) because of continued flip-flopped input.
I've recently finished a paper on this, and in the works is
another paper giving empirical + computational evidence for
cognitive factors in the propagation of arbitrary forms in general.

___________________________________________________________________
Joyce Tang Boyland
Alverno College (Dept. of Psychology)
& UW-Milwaukee (Dept. of Foreign Languages + Linguistics)