Dear Irene,
The judgments of your informants seem to reflect what is *possible* in the
language, not what is typical (which is what corpus-derived semantic prosody
is about). In other words, your informants are *imagining any possible
contexts* where, for example, "caused me a lot of joy" is at all possible.
(For me, I can think of a situation where a creative writer is using
parallelism: "She caused me a lot of heartache, but also a lot of joy" --
this certainly sounds ok to me, because balance and aesthetics (should?)
outweigh conventional prosodies.). Under such a "possible worlds" framework,
almost all your sentences would be accepted by many(most?) native speakers.
To get at judgements of semantic prosody, your *instructions* could perhaps
be changed to read: "For each set, please tick a maximum of TWO sentences
that in your opinion sound the most natural or typical in English." Forced
to choose, native speakers *may* (who knows?) go for the ones that corpus
studies reveal as most typical.
Regards,
David Lee
This archive was generated by hypermail 2b29 : Tue Jan 25 2005 - 10:14:08 MET