Analysing =?windows-1252?Q?=93starting_points=94_?=in complex sentences

Tue Oct 8 19:29:57 UTC 2013

Dear Lars (and Tom),

Good question.  Unfortunately, my answer will be rather long-winded.

My 1977 version of the Perspective Hypothesis, as well as the experimental evidence from Gernsbacher (1990), focuses mainly on the forces determining selection of the subject of the main clause.  However, it was already clear then that only nominals constitute real starting points.  So, in the Adv-V-S order of German as in "Hier kommt der Mann" there is no real sense in which the initial element is a perspective.  It is an attentional focus but not the fundamental structural building block of Gernsbacher 1990.

A fuller answer to your question involves a discussion of perspective shifting or switching.  In 1977, I saw that the data from Herb Clark's (1969) sentence-picture verification task could be addressed in terms of something close to perspective-shifting operations.  Then, in MacWhinney and Pleh (Cognition, 1988), we found that it was necessary to extend the item of perspective-shifting still further to get a fuller understanding of the dynamics of processing for relative clauses in Hungarian (as well as related data for relative clauses in other languages).  

Also, in MacWhinney (1975), which eventually surfaced as MacWhinney and Bates (1978) we provided evidence for online perspective-shifting from marked verb forms like "get" to unmarked forms like "get" along with retracing and pausing in picture descriptions by both children and adults.

Returning to these issues much later in these two papers, I tried to explain in greater detail how perspective shifts between clauses:
MacWhinney, Brian. (1999). The emergence of language from embodiment. In B. MacWhinney (Ed.), The emergence of language (pp. 213-256). Mahwah, NJ: Lawrence Erlbaum.
MacWhinney, Brian. (2008). How mental models encode embodied linguistic perspectives. In R. Klatzky, B. MacWhinney & M. Behrmann (Eds.), Embodiment, Ego-Space, and Action (pp. 369-410). Mahwah: Lawrence Erlbaum.
Your best reference would be the 2008 paper.  For English, my claim is that the subject (which is nearly always preverbal) is the initial perspective and must be referential.  However, if there is a preceding PP or subordinate clause, then specific cues within that phrase can set up cataphoric expectations for a following referential perspective.  More generally, I have become increasingly convinced that the primary function of grammatical markings is to signal perspective shift and maintenance.  There are far too many constructions and cues involved here to capture in a brief email message and even the 2008 paper is just a sketch of this territory.

In regard to freer word order languages like German, Russian, or Hungarian, the placement of an NP with accusative marking before the verb does not establish a true perspective.  It does serve the role of "placing an actor on stage" as Chafe (1974) argued, but if the accusative marking is clear then the NP is not processed as a perspective, and the processor holds that role open, but as in the case of preposed cataphoric subordinates.  This is perhaps clearest in Hungarian and Japanese where case marking is far less ambiguous than in German.  Sanako Mitsugi and I have various eye movement and self-paced reading studies in Japanese demonstrating this.  Ina Bornkessel-Schlesewsky and Matthias Schlesewsky have also done wonderful crosslinguistic work demonstrating similar principles using ERP methodology.  The basic idea is that the whole system of starting points exists as a general default overlay on the more item-based system of grammatical role marking.

Apart from cues found in preposed clauses, there are a myriad of interesting perspective shift effects in deictics (here, now) and the anaphoric systems studied by government-and-binding theory.  There are also crucial perspective-shifting effects within the clause as attention shifts from the starting point to the direct and indirect objects.  Basically, the Perspective Hypothesis offers a pragmatic explanation for the various principles alternatively formulated through c-command.

Some of the effects regarding GPOV that Tom mentions interact with clausal syntax, as in the processing of control verbs such as "easy to see" vs. "eager to see".  Peggy Speas, Carol Tenny, and others have studied a variety of perspective-shifting emotion words and expressions such as "happily" or "that darned ..."   that do not touch clausal syntax that much.  I would like to think of perspective-shifting as a high-level frontal lobe function based on role assignment and scene construction that then connects with syntax and lexicon as interactive inputs. Maybe this is what Jackendoff means about interfaces?

In terms of practical application for your work, the default assumption is one of perspective continuation across clauses.  However, to tell a story effectively, perspectives must be frequently shifted.  As MacWhinney and Bates (1978) showed, young children do not seem to have full control over the use of grammatical marking to mark these shifts.  So this is what often makes it hard to follow their stories.  

--Brian MacWhinney

On Oct 8, 2013, at 3:46 AM, Gerlind Gro�e <Gerlind.Grosse at eva.mpg.de> wrote:

> Dear colleagues,
> 
> We would like to analyze a corpus of children’s stories (5 year olds) regarding the type of starting points. We depart from the Perspective Hypothesis (MacWinney, 1977, 2008, Gernsbacher 1990) which holds that “…given a choice between two starting points, speakers and listeners prefer the starting point closest to the one they assume or which to assume in their own interactions with the world.” Now, our current question is about how to determine starting points in complex sentences or sentence fragments. Does each sub-clause indicate a new starting point? Is the starting point always the very first element or the first main functional element, i.e. (SVO)? We are dealing with German which has a flexible word order.
> Any ideas, references or advices are welcome.
> 
> 
> Many thanks!
> 
> Lars White, and Gerlind Grosse
> (Leipzig)
> 
> 
> -- 
> You received this message because you are subscribed to the Google Groups "Info-CHILDES" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to info-childes+unsubscribe at googlegroups.com.
> To post to this group, send email to info-childes at googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/info-childes/5253B83A.2050904%40eva.mpg.de.
> For more options, visit https://groups.google.com/groups/opt_out.

-- 
You received this message because you are subscribed to the Google Groups "Info-CHILDES" group.
To unsubscribe from this group and stop receiving emails from it, send an email to info-childes+unsubscribe at googlegroups.com.
To post to this group, send email to info-childes at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/info-childes/B79EDB7D-31D5-436A-8EC8-A8C8D83A5C98%40cmu.edu.
For more options, visit https://groups.google.com/groups/opt_out.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/info-childes/attachments/20131008/a875154e/attachment.htm>