Autor:Martin Everaert.


To begin with, the question what 'unsolved problems' there are in the
context of generative grammar is difficult to answered satisfactorily
because 'problems' only exist in the context of an explicit theory, and
theories change. That means that problems disappear or are created when new
theories arise. The minimalist program (Chomsky 1995) is a good case at
hand. Some of the basic assumptions in this model are so radically different
from the 'standard' P&P model that we can almost start afresh, which means
that whatever was supposed to be solved, is either not a problem any more or
has become a problem again.

Let me give an example. Binding Theory as we know it in generative grammar
has been an area in which much work has been done, and some progress has
been made. Still numerous issues remained far from resolved. In the
minimalist program the theoretical goal is that syntactic operations should
be driven only by purely formal and mechanical considerations, like checking
morphological features. Chomsky notes that "indices are basically the
expression of a relationship and they should be replaceable without loss by
a structural account of the relation they annotate". This has far-reaching
consequences on how a theory of anaphora should like. How are we to encode
binding in the computational system if computations are limited to a
morphology-based vocabulary, which excludes indices? (cf. Reuland, NELS 1996)

Below I have listed some issues that are mentioned in the context of the
research program of my own institute. The list below contains areas (both
more general and purely (morpho-)syntactic) in which we have gained some
insight but in which we have either successful but incompatible analyses, or
just no satisfactory analysis (in random order and very limited):

1. To what extent does first language acquisition differs from second
language acquisition, and what is the role of explicit instruction in the
latter.
2. How is knowledge of different languages stored and accessed? Are they
mentally represented as separate systems, or largely identified, only being
different in fairly late options?
3. Is parameter setting an irrevocable act, or can the child resort to
previous or coexisting systems of grammatical knowledge?
4. What is the relation between structure and word-order
5. What is the status of the strict cycle,
6. What are the legitimate boundaries of the feature systems used in
syntactic computations.  (Should, for instance, the syntactic codification
of notions such as referentiality, definiteness, familiarity, be allowed?)
7. Why would natural languages have a contrast between anaphors and
pronominals at all?
8. What are the links between argument structure and properties of aspect
and/or event structure?
9. What is the role of the feature specification for local binding of pronouns?
10. Is inherent reflexivity a unitary phenomenon, and is it correctly
understood as a lexical property?
11. Does 'morphological' computation differ from 'syntactic' computation, and
if so, what are the differences?

************************************************************
Martin Everaert         Utrecht institute of Linguistics OTS
Trans 10, 3512 JK Utrecht, The Netherlands
tel: 31-30-2536528; fax: 31-30-2536000
http://wwwots.let.ruu.nl/user/uilots/everaert/home.html
************************************************************