[Prev][Next][Index]

CSLI Calendar, Oct. 8, 3:2


       C S L I   C A L E N D A R   O F   P U B L I C   E V E N T S
_____________________________________________________________________________
8 October 1987                    Stanford                      Vol. 3, No. 2
_____________________________________________________________________________

     A weekly publication of The Center for the Study of Language and
     Information, Ventura Hall, Stanford University, Stanford, CA 94305
                              ____________
          CSLI ACTIVITIES FOR THIS THURSDAY, 8 October 1987

   3:30 p.m.		Tea
     Ventura Hall		

   4:15 p.m.		CSLI/Philosophy Colloquium
     Philosophy		Providing a Rational Basis for Morality
     Bldg. 90:91A	Holly Smith, Dept. of Philosophy, 
			University of Arizona

                             --------------
           CSLI ACTIVITIES FOR NEXT THURSDAY, 15 October 1987

   2:15 p.m.		CSLI Seminar
     Classroom          The Acquisition of Morphology
     Ventura Trailers  	Discussion of the debate between
			Rumelhart/McClellan and Pinker/Prince  
			Dave Rumelhart, (der@psych.stanford.edu)
			Abstract in next week's Calendar
			
   3:30 p.m.		Tea
     Ventura Hall		

   4:15 p.m.		CSLI Colloquium
     Classroom		A Logic for Practical Reasoning
     Ventura Trailers	Tim Flannagan, Logica Cambridge
			Abstract below

                             --------------
                          NEXT WEEK'S COLLOQUIUM
                     A Logic for Practical Reasoning
                     Tim Flannagan, Logica Cambridge
                               October 15

   We construct a formal logic for practical reasoning (PR) from an
   analysis of Kenny's informal `logic of satisfactoriness' [K], which is
   intended to preserve `satisfactoriness' in passing from premises to
   conclusions just as deductive logic peserves logical truth.
       Whereas Kenny's logic argues to sufficient means to ends, PR argues
   both to necessary means and to sufficient means.  It models intuitive
   sufficiency as material implication and satisfactoriness as relative
   consistency. Unlike Kenny's logic it has a precise definition of
   defeasibility.
       PR extends first-order logic with the following rule of inference
   which we call comodus ponens. From G (understood as a goal) and C
   implies G, infer C, provided that C and G are jointly consistent
   relative to X (the nonlogical axioms).  PR has the following
   properties:

   1. If A is a logical consequence of X, then it is derivable in PR from
      X but the converse is false.

   2. If A is derivable from X, then it is consistent with X.

   3. The separate derivability of two sentences from a set X does not
      imply that their conjunction is derivable.  It is thus possible for
      A and `not A' to be derivable without this resulting in a
      contradiction.

   4. PR is a defeasible logic and hence strongly nonmonotonic.

                             --------------
                    SITUATION SEMANTICS WORKING GROUP
                   Tuesdays 10:00-11:45, Ventura Hall
                  (First meeting: Tuesday, 13 October)
                                Contacts
               Craige Roberts (Croberts@csli.stanford.edu)
                Stanley Peters (Peters@csli.stanford.edu)

   Members of this group will explore Situation Semantics from a
   linguistic point of view.  To this end we hope to first consolidate
   our understanding of the theory in its current state of development,
   and then use this perspective to explore some current issues in
   linguistic semantics.
      In keeping with the trend of current research in semantics, we
   assume that three general criteria for an adequate semantic theory
   are:

     (i) partiality in modeling,

    (ii) adequate tools for talking about context-dependence, and

   (iii) ability to assign sufficiently fine-grained meanings.

      We plan to entertain more specific questions and criteria as
   refinements of (i), (ii), and (iii), and explore how Situation
   Semantics addresses these questions and fulfills these criteria, and
   how it compares in these respects with other current theories.
      In order to establish a basis for discussion, at our first session
   Stanley Peters will give an overview of the version of Situation
   Semantics developed in Barwise and Etchemendy's THE LIAR.  It might be
   useful for those planning to attend to have a look at that work
   beforehand.  (Note: For those interested in Situation Semantics,
   attendance at John Etchemendy's seminar on THE LIAR would be helpful,
   but is not necessary for participation in this group.)
      After some further preliminaries, there are a number of particular
   issues which we hope to explore.  To begin, we will focus on
   conditionals and modality, linguistic phenomena which require a
   semantic theory based on situations (and/or partial worlds).  These
   phenomena also introduce questions of context-dependence and
   fine-grainedness.  We will compare work on these issues within
   Situation Semantics to other contemporary theories such as that of
   Kratzer (in her paper on the ``Lumps of Thought''), Data Semantics
   (Landman and Veltman), and to Muyskens' work on partiality in
   possible-worlds semantics.
      At the first meetings, other possible topics will be suggested.  In
   addition, participants are urged to propose topics and lead sessions
   themselves in accordance with their own research interests.
      Although we will focus on linguistic issues, all members of the
   CSLI community, students, and visitors are invited to attend and
   participate fully.
-------