Eric Christopherson wrote:
>
>Thanks! That's a good term to know; also on Wikipedia are "rebracketing"
>and "metanalysis". Their meanings overlap at least partially.
metanalysis and reanalysis are the terms I've heard most commonly.
>
>However, it's still not the whole story behind what I'm talking about.
>With processes like "a napron" -> "an apron", the beginning and end states
>have substantially the *same* function and semantics -- "singular
>indefinite" plus "apron". What I'm talking about is more like if the
>starting point were "plural" + "apron" but it got reanalyzed as
>*"singular"* + "apron".
Engl. (old) pease / modern pea ~peas < Fr. pois (sg., i think) or Engl
(native) "(an) adder" = Germ. nader-- so it's not just restricted to loan
words, as one might suspect.
>
>Again, I know this is a kind of change that does occur, but what I'd like
>to know specifically is: what kinds of factors encourage or discourage it,
>what limitations it has, and what the intermediary steps are in its
>development.