form without meaning

[Edith A Moravcsik]

===> To LIZ BATES:

     Liz, you suggested (Saturday, January 11) that the claim that
     syntactic form could be described independently of meaning assumed that
     syntactic classes were of the classical type, with strict category
     membership. You further pointed out that syntactic categories were
     not in fact of this sort and that their fuzzy nature was difficult
     to explain without reference to their meanings.

     I agree that, for EXPLAINING the existence of natural syntactic
     categories, we may have to resort to studying their meanings.
     This does not mean, however, that natural classes cannot be discovered
     and described on strictly formal grounds; in this respect,
     I agree with Dan Everett's (same-day) response to you.

     I found your later message on the four problems in obtaining
     grammaticality judgments very interesting and instructive!


===> to JON ASKE:

     Jon, in your response to my response to your original posting
     (Sunday, January 12), you wrote this:

     "I just don't see why we would want to restrict our linguistic
     analysis to the more formal aspects of constructions and ignore
     their function/meaning pole, their history, and so on. .... To me,
     studying the formal aspects of such constructions without looking at
     what they are made for, how they are made, etc. *as a matter of
     principle*, just does not make sense. I came to the early
     realization that these constructions should not be studied as
     merely formal operations. These constructions exist for a purpose,
     and their form reflects the function that they arose for in the
     first place, even if they have picked up additional bagage along
     the way. And to me that is the most interesting part of analysing
     language/grammar."


     I agree with almost all of this. In particular, I agree that
         a/ linguistic analysis should not be restricted to form,
            with function ignored (in my contribution, I did not
            mean to suggest the opposite)
         b/ constructions exist for a purpose and figuring out
            the extent and the ways they reflect function is the most
            interesting part of analysing grammar.


     Where I may not agree is that studying the form of constructions
     without looking at their function makes no sense. It really depends
     on what you mean by "studying". If you mean "giving a complete
     account", then I fully agree: describing the functions of linguistic
     form and how they correlate with form is part of a complete account.
     But if by "studying" you mean "restricting momentary attention"
     (where "momentary" may be taken on a grand sale, possibly extending
     to the lifetime of a linguist), then I cannot agree. I see the
     study of linguistic form as a logically necessary step in arriving
     at a complete account of linguistic constructions since, as Talmy
     Givo'n pointed out (Saturday, January 11), if a complete account
     involves specifying a (cor)relation between form and function,
     this presupposes that we have an independent characterization of
     both form and function.

     This point does not have to do with research schedule: I am
     not proposing that all of form needs to have been discovered
     before we can begin to look at function. The two lines of
     research usually go in tandem I believe. Rather, the point has to do
     with the logical priority of a description of form and a
     description of meaning over an account of the relationship between
     the two.

     At the very real risk of battling a straw man or beating a dead
     horse (and without attributing this extreme view to Jon Aske),
     let me note that what the idea - when taken in a literal sense -
     that the form of functional objects cannot be described unless
     one knows the associated functions would amount to is that the
     usual descriptive tools we use for characterizing the form of a
     non-functional object would simply fail us in the case of
     functional objects: we would have to hold off on their  formal
     description until we found out about their functions. This would mean,
     for example, the following:

        - One could describe the formal structure of a string of beads
          a child would create with no purpose in mind but not the
          shape of a rosary - unless one knew that the beads stood
          for various prayers.

        - One could describe the chemical composition of naturally
          occurring materials but not that of synthetic drugs, unless
          we knew what each component was supposed to contribute
          to the intended healing effect.

        - One could describe the form of a musical composition free of
          containing designated motifs with explicit meanings but not
          the form of a Wager opera - unless one knew what each motif
          stood for.

        - One could describe the random hand-flailings of an infant,
          but not the hand gestures of body language or sign language,
          unless one knew the meanings of the gestures.

        - One could describe the form of a piece of rock naturally
          shaped as a hammer but not the shape of a real hammer, unless
          one recognized it for an instrument for pounding in nails.

        - One could describe the form of the Easter Island statues
          just in case they were meant to be non-functional; if they
          were meant to be functional, no description would be possible
          unless one learnt what the functions were.

        - One could describe the form of non-symbolic carvings on a
          rock surface but not if those carvings happened to be samples
          of writing; in which case we would have to discover what
          each symbol stood for before being able to characterize the
          forms of the symbols.

       Such descriptive impoasses caused by lack of knowledge about
       the function of the object to be described clearly do not arise.
       Where knowlefge about function comes in for the analysis of
       form is on the explanatory, rather than descriptive, level:
       in helping to explain why the form of a functional object is the
       way it is.

       Best - Edith
   ************************************************************************
                                 Edith A. Moravcsik
                                 Department of Linguistics
                                 University of Wisconsin-Milwaukee
                                 Milwaukee, WI 53201-0413
                                 USA

                                 E-mail: edith at csd.uwm.edu
                                 Telephone: (414) 229-6794 /office/
                                            (414) 332-0141 /home/
                                 Fax: (414) 229-6258