[Corpora-List] My semantic prosody questionnaire

[Martin Wynne]

Irena,

Thank you for your interesting email.

If I understand you correctly, your hypothesis seems to be that native
speakers will perceive usual and unusual semantic prosodies in your
example sentences, and non-native speakers will not.

Louw (1993), widely cited as the classic text on semantic prosody,
suggests that intuition is not a guide to semantic prosodies:

"They [semantic prosodies] are essentially a phenomenon that has only
been revealed computationally, and whose extent and development can only
be properly traced by computational methods."

"It may well turn out to be the case that semantic prosodies are less
accessible through human intuitition than most other phenomena to do
with language."

Also, he suggests that semantic prosodies may exist only in texts, not
de-contextualised sentences:

"First, it is clear that in many cases semantic prosodies 'hunt in
packs' and potentiate and bolster one another in rather the same way
that images are forced to cluster in poetry in order to prevent full
'intuitive' meaning from ebbing away into delexical meaning."

It would also be interesting to look at the recent work of Michael Hoey
on lexical priming, as this theory looks into the individual and
psychological aspects of the effects of repeated collocations. In
particular, he states:

"Collocation priming is not a permanent feature of the word. Each use we
make of the word, and each new encounter, either reinforces the priming
or loosens it, if we use it in defiance of the priming."

and:

"Collocational priming is sensitive to the domain in which the lexical
item is encountered."

In fact, it sems to me that a more reasonable hypothesis based on the
relevant literature would have been that native speakers are no better
at perceiving semantic prosody in isolated sentences than non-native
speakers. And your results may have supported this hypothesis (though I
am dubious about the usefulness of such questionnaires).  I agree that
looking in a corpus of non-native speaker writing would now be a good
way to proceed.

References:

Louw, W. (1993). 'Irony in the text or insincerity in the writer? The
diagnostic potential of semantic prosodies' in Baker, M., Francis, G. &
Tognini-Bonelli, E. (eds.). /Text and technology/. Amsterdam: John
Benjamins. 157-176. [Out of print, but reprinted in Sampson, G. and
McCarthy, D. (eds.) (2004). /Corpus linguistics: readings in a widening
discipline/. London: Continuum. 229-241.]

Hoey, M, 'Lexical Priming and the Properties of Text',
http://www.monabaker.com/tsresources/LexicalPrimingandthePropertiesofText.htm

Best wishes,
Martin

--
Martin Wynne
Head of the Oxford Text Archive and
AHDS Literature, Languages and Linguistics

Oxford University Computing Services
13 Banbury Road
Oxford
UK - OX2 6NN
Tel: +44 1865 283299
Fax: +44 1865 273275
martin.wynne at oucs.ox.ac.uk



Irena Shuke wrote:

> Dear all,
>
> I just received a questionnaire I had sent to England to be filled out
> by native speakers to see their awareness of semantic prosody. I knew
> (AND had been told) that it was best to just observe it in a corpus,
> but my aim was to see (compare) the difference between native and
> non-native speakers with respect to this. To test the intuitions of
> the both sides, so to say. (Well, I guess I did always want to see
> with my own eyes how native speakers would circle the "correct"
> sentence, such as 'She caused me a lot of trouble', and leave one like
> 'She caused me a lot of joy' alone.) Anyway, I thought a questionnaire
> given to both native and non-native speakers would show clearly the
> difference. First, I got them filled out by some of my fellow MA
> students (Linguistics, English Literature & ELT Methodology) who were
> all non-native speakers of English. Only one person scored enough for
> me to see she was showing consistent awareness of this phenomenon, but
> then, she and I had been discussing this, so I guess that doesn't
> count, although she had only been given one particular example
> ("cause") earlier in our discussions. (By "scored enough" I do not
> mean any particular percentage since I still have to decide how to
> best display my clever "findings".) A LOT of other respondents ticked
> off "Engineering of these goods will favourably affect the market" and
> "She caused me a lot of joy" as correct (acceptable, not "correct",
> but for some reason I tend to use the words "right/correct" and
> "wrong" when I think of it), or both the sentence that contained
> "caused me joy" and the one with "caused me trouble" in the same set,
> or ticked off "positively affect" and "favourably affect" but left out
> "seriously affect" and "badly affect" as unlikely. I decided to wait
> for the same questionnaire to come back from England where it would be
> administered to native speakers. I have to say that most of the
> sentences were either taken from a corpus AND modified or plain
> invented by me (esp. the provocative ones with the "unacceptable"
> collocations), because I didn't really want to underline the
> collocations that were to be thought about - so I decided to try to
> make all the sentences in one set look similar in form, so that only
> THE collocation would stand out in each of them by being the only
> different element. I did suspect that the respondents, esp. native
> speakers, might be put off by just any wrong collocation in a sentence
> before they even got to rejecting the sentence due to the clash of
> semantic prosodies. I also stated in the instructions that all of the
> sentences are correct grammatically. And before administering it to
> anyone at all, I had also shown the questionnaire to a tutor of mine
> just to make sure in general (she is a non-native speaker too).
> So I waited.
>
> Well, today I got 15 precious questionnaires from England. (Obviously
> it was not possible to get more in, but even these do illustrate
> something.)
>
> I looked at some of them and then double-checked the last "question"
> on each of them, i.e. 'Is English your mother tongue?' All said yes.
> Anyway, here are the results. 'She caused me a lot of trouble', 'She
> caused me a lot of misery', 'She caused me a lot of happiness', 'She
> caused me a lot of joy' are all ticked off as "likely to be used" in
> quite a few(!) people's responses. 'Engineering of these goods will
> favourably affect the market' - just fine in someone else's work.
> 'Caused me a lot of joy' - OK, but 'caused me a lot of happiness' -
> left out as "unlikely". I really don't know what to think. In some
> cases it might have been even ambiguity that I myself hadn't noticed,
> as in "Engineering of these goods will positively affect the market',
> which possibly might have been interpreted as "definitely". Yet, that
> is still not the only case. I myself really liked "cause" because it
> just seems very diagnostic to me, so I was hoping to see some
> consistency there. Alas.
>
> I am thinking now that perhaps it would make sense to put together a
> corpus of non-native students' works (for example) and see how they do
> or do not distinguish any semantic prosody patterns there, and of
> course there are a lot of corpora for researching native speakers'
> language.
>
> I really apologize if this seems outrageous, I do not mean to upset or
> disgust anyone. I guess I just had slightly different expectations (or
> hopes) of this experiment.
> I would be thankful for any opinions and input.
>
> Irena
> (MA Student of Linguistics, University of Latvia)
>
> _________________________________________________________________
> FREE pop-up blocking with the new MSN Toolbar - get it now!
> http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/
>
>
>


--
Martin Wynne
Head of the Oxford Text Archive and
AHDS Literature, Languages and Linguistics

Oxford University Computing Services
13 Banbury Road
Oxford
UK - OX2 6NN
Tel: +44 1865 283299
Fax: +44 1865 273275
martin.wynne at oucs.ox.ac.uk