performance/competence, cooperation, and the future

[Roeper, Tom]

Info childes    performance/competence and cooperation

Dear  Performance/Competence discussion-

    I'd like to shift the discussion-if it continues-- from that of
errors, which are incontrovertible signs of performance, to questions
of interfaces which is in many ways a new thrust in linguistic theory
partly aimed at incorporating in a single mechanism many phenomena
previously associated with performance.  I think it offers an exciting
future and many avenues for cooperation---let me illustrate.
      I have been silent because I was in Cyprus attending a
fascinating meeting sponsored by the European Union, organized by Uli
Sauerland from ZAS, aimed to developing
instruments to assess children's language in a wide variety of
European
languages.    In the group of 40 were many people associated with Dan
Slobin and Michael Tomasello.  Issues linked to "performance" are
a constant factor in designing experiments---hardly something ignored
in practical terms, because speech pathologists were also a central
part of the group.   How to build those factors into
a language interface that is connected to the subtleties of grammar is
a
real theoretical challenge, but being sensitive to them in day to day
experimental design is just a given.
      Let me offer more  about thiis conference.  After some
discussion we got to work designing experiments.  Using insights
gained from the DELV test,  which Harry Seymour, Jill deVilliers and I
(and many others) devised for English, we looked at  exhausitivity in
"who bought what" (where many children answer  with
one person or a single pair-not all required) and quantifier spreading
in "every dog as a bone", where some children think it means "every
bone".  Now these sentences will be explored in up to 17 different
languages and dialects.   Experimental ideas from many other quarters
were incorporated as well.
      These questions   are askable precisely because they are linked
to linguistic universals that transcend language differences.  They
come straight from linguistic theory---yet everyone who hears about
them  gets interested whether they are followers of generative
linguistics or not.  Experimentation in Polish, Bulgarian, and Romani
has already begun, with interesting results.
      Recent work in linguistic theory has focused on Interfaces,
which includes pragmatics and semantics, and the leader of the
 EU group, Uli Sauerland, is a specialist in precisely these domains.
and a major focus has been how implicatures work.
        The challenge I raised before remains---and it is a tough one-
without obvious answers: how do we build a mechanism---commensurate
with the speed of language---that can build other
compuatational abilities into  language while realizing their
creative and generative power.  How do we represent creativity
in memory to match the creativity of recursion?  How do we
generate unique implicatures in unique situations?  Any model
which avoids the creative algorithms misses the essence of the
mechanism---and I believe ultimately leads to an image of children
that does not honor their dignity.  This is the argument I make
in my book The Prism of Grammar.
       An analogy might help.   While syntax perhaps captures the
skeleton of language---one might holler "how about all the muscles
that enable movement?". Good question, but it requires exactly a
concept of the interface: where exactly do muscles connect to bones so
that they can move them efficiently and how does the physics of
movement
get represented.  A theory of muscles that ignores exactly how
they are connected  with bones won't tell us much.
     Take an example (from Christ Potts) on implicatures and
Point of View.  If  I say "honestly you can climb a mountain"
it is my honesty that is implicated, if I invert the auxiliary
"Honestly, can you climb a mountain" it is yours.   And we
get different readings for: "honestly who can report" and
"Who can report honestly".  The implications are tightly
linked to the syntax.  There is a natural experiment for someone.
to do: when do children control that switch.  (Anna Verbuk has
done interesting relevant work.)
     I think the discussion can be much more productive if we
discuss real examples and project precise theories of how they
can work.  That is the challenge we can explor together, much
more beneficially than by a debate that too heavily linked to
imprecise abstractions.

Tom Roeper

See you at BU if you are coming--


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Info-CHILDES" group.
To post to this group, send email to info-childes at googlegroups.com
To unsubscribe from this group, send email to info-childes-unsubscribe at googlegroups.com
For more options, visit this group at http://groups.google.com/group/info-childes?hl=en
-~----------~----~----~----~------~----~------~--~---