On Tue, 7 Jan 1997 11:09:18 -0800 Johanna Rubba said:
>              I chose things close to the theft in topicality value as
>subject. This resulted in numerous passive sentences in the text.
>I then let the students read the two and choose which one they
>thought worked better. They all chose the second text. Then we examined
>subject choice and found that the topicality of a referent sometimes
>dictated passive as the optimal choice, in spite of the fact that writing
>teachers often preach against the use of the passive voice, and a lot of
>students take this to mean that they must avoid the use of passive at all
>costs. The students really liked this exercise, and reported that it
>taught them a lot about how texts are structured and that, on occasion,
>passive really is the optimal choice for reasons of coherence, not
>formality or agent-hiding.
 
Of course, this is absolutely right.  The choice of voice is directly
related to certain text contraints and the exercise Johanna outlines is
very good.   But that was not the example I was citing.  I used enjoy and
please and was wondering how the configuration of subject and object
and argument structure could be motivated from a strictly semantic
viewpoint.
 
I do not not want to be misunderstood.  Once we begin to get beyond
the sentence there is very little that a formalist approach to knowledge
about language can help us with.  We need to do exactly the kind of
exercises that Johanna has so nicely explained.   I just want to point out
that knowledge about how a passive sentence is related to an active sentence
is different than knowledge about whether the use of the active or passive
sentence is appropriate in a given context.
 
>So you see, 'corpus' work is not inaccessible to students; it _does_
>provide regular results, and it _is_ explanatory, in terms of what
>governs the choice of subject in coherent writing.
 
I am not making a claim that a formalist approach explains everything, but I
am claiming that a strictly functionalist approach can explain everything.
 
I like the performance/competence distinction however difficult it is to
draw the line where one ends and the other begins.
>statement on the Ebonics issue. Watch this space!
 
Bob Yates, Central Missouri State University, [log in to unmask]