<html>
Colleagues,<br>
<br>
I've been reading and thinking about the apparent problem for Kaqchikel
relative clauses that I mentioned earlier. To recap, I said <br>
<br>
>For example in Kaqchikel, the facts are as follows:<br>
><br>
>(interrogative focus)<br>
><br>
>Who did John
grab?<x-tab> </x-tab><x-tab> </x-tab>grab
is in the neutral voice<br>
>Who grabbed the dog?<x-tab> </x-tab>grab is in the
actor focus voice<br>
><br>
>(contrastive focus)<br>
><br>
>It was the dog that John grabbed (not the
cat).<x-tab> </x-tab>grab
is neutral voice<br>
>It was John who grabbed the dog (not
Mary)<x-tab> </x-tab>grab is actor focus
voice<br>
><br>
>Compare this to topicalization, which doesn't use the actor focus
voice:<br>
><br>
>As for the dog (TOPIC), John grabbed
it<x-tab> </x-tab><x-tab> </x-tab>grab
is neutral voice<br>
>As for John (TOP), he grabbed the
dog<x-tab> </x-tab><x-tab> </x-tab>grab
is neutral voice<br>
><br>
>Now the problem is that relative clauses show the same pattern as
focus constructions, and not the pattern seen in topic
constructions:<br>
><br>
>The dog John grabbed has
mange.<x-tab> </x-tab><x-tab> </x-tab><x-tab> </x-tab>grab
is neutral voice<br>
>The man who grabbed the dog is my
uncle<x-tab> </x-tab><x-tab> </x-tab>grab
is actor focus voice<br>
><br>
>That suggests to me that the head of the relative clause has the
discourse role FOCUS, and not TOPIC.<br>
><br>
<br>
After thinking more carefully about this pattern (and rereading Aissen's
work), I think that the breakdown isn't actually focus constructions vs.
topic constructions but rather "movement" vs.
"non-movement". The AF verb form is restricted to cases
where there is a gap corresponding to a transitive subject.<br>
<br>
Aissen (1992) showed nicely that in Tzotzil there are island effects for
contrastive focus vs. no island effects for topicalization.
She suggests that there is a coreferent null pronoun in a case like
"Xun (TOPIC), [the woman he treated <i>pro</i>] became
well". But there is a trace/gap in the comparable *"Xun
(FOCUS), [the woman he treated <i>t</i>] became well."<br>
<br>
So my initial question suggested that the division between neutral voice
and actor focus voice was TOPIC vs. FOCUS. I now think it's more
likely to be long-distance dependency of the filler-gap type vs.
dependency of the antecedent-anaphor type.<br>
<br>
That does raise the natural question of why interrogatives, relatives,
and focussed elements should be formed with gaps, while topics are
not. But that at least seems like property found in many
languages.<br>
<br>
Aaron Broadwell<br>
<div>-------------------------------------------------------------------------</div>
<div>George Aaron Broadwell, g.broadwell@albany.edu</div>
<div>Dept. of Anthropology, UAlbany, Albany NY 12222</div>
<div>(518)-442-4711 Web page:
<a href="http://www.albany.edu/anthro/fac/broadwell.htm" EUDORA=AUTOURL>http://www.albany.edu/anthro/fac/broadwell.htm</a></div>
<div>--------------------------------------------------------------------------</div>
"I really do not know that anything has ever been more exciting than
diagramming sentences." Gertrude Stein, Lectures in
America, 1935.
</html>