LINGUIST List 7.311

Thu Feb 29 1996

Sum: parameters and minimalism

Editor for this issue: T. Daniel Seely <dseelyemunix.emich.edu>

Directory

1. jaretu.washington.edu, parameters and minimalism

A few days ago I posted a query asking how parameters work in the
minimalist program, and received helpful replies from the following
people: 

Dan Everett
Charles Yang
Marianna Pool
Richard Ingham
Cassian Braconnier
Elizabeth McKeown 
Kiyoshi Ishikara 
Stefano Bertolo

I would like to express my thanks to all of them. I have inserted Charles 
Yang's comments below, which were the most detailed. But first, another 
question. If I understand correctly, the minimalist program is not a 
'modular' theory of syntax, like GB is/was. That is, we don't have 
interacting systems of principles, each comprising a 'module' of UG. It 
would seem odd to call the unparameterized economy principles a 'module'. 
And likewise, I don't see how modules are compatible with morphological 
features being labeled strong vs. weak, overt vs. covert, etc. Is this 
other people's impressions, too? Or is the difference just terminological?

Frank
- ----------------------------------------------------------

Frank, 
 
some remarks on your LINGUIST query. 
in the MP, one way, perhaps the only way, to parameterize lg variation is
the parametric instantiation of features, if we assume syntactic
operations are feature-driven, i.e. to satisfy feature requirements and
PF/LF interface legibility conditions. the basic questions are: what are
these parameters? what values are they set to for different languages? 
how are they represented? how do they enter into syntactic operations? 
how are they acquired? ... it is important to keep in mind that the
principles (e.g. economy, last resort) are *NOT* parameterized -- they are
part of UG, presumably the innate biological endowment.. 

some concrete examples. there are features of the following distinctions: 
strong/weak, overt/covert, on/off.. English has *overt* subject/verb
agreement, therefore subject moves *overtly* to check the agreement
(that's why we have SVO); japanese has SOV because both the subject and
the object check agreement *overtly*.. now this is based on the internal
VP hypothesis *and* something like Kayne's universal SHC hypothesis (BTW,
it appears that Kayne is right in assuming that, but not for reasons he
outlines in his book)... 

anyway, hope this helps a bit.. 
 
charles yang 
MIT AI Lab 
MIT Linguistics