LINGUIST List 5.488

Wed 27 Apr 1994

Disc: Generative

Editor for this issue: <>

Directory

1. Penny Lee, 'generative'
2. Steven Schaufele, generativity and well-formedness conditions

Re Alex Manaster-Ramer's comment about Dick Hudson's comment about 'generative':

>Re: Dick Hudson's remarks about 'generative', is there anyone
>who still uses it to mean what it originally meant, i.e.,
>'completely well-defined and used to generate all and only
>the well-formed sentences of a given language'? I think
>not, so perhaps it is just as well to allow it to become
>a purely sociopolitical label. Likewise, 'formal'. Instead,
>for the precise mathematical terms, perhaps we need to substitute
>different words, e.g., 'explicit', 'rigorous', 'constructive',,
>or the like.

doesn't Chomsky say quite explicitly, both early and relatively late, that
'generative' only means 'explicit' in his usage? (e.g. Aspects p 4, p 8,
Knowledge of Language p 3).

On p 9 of Aspects he says: The term "generate" is familiar in the sense
intended here in logic, particularly in Post's theory of combinatorial
systems. Furthermore, "generate" seems to be the most appropriate
translation of Humboldt's term *erzeugen*, which he frequently uses, it
seems, in essentially the sense here intended. Since this use of the term
"generate" is well established both in logic and in the tradition of
linguistic theory, I can see no reason for a revision of terminology".

As a naive user of the German dictionary, I find this comment confusing.
'Erzeugen' seems to me to be defined in a way which is compatible with my
ordinary language use of 'generate' in the sense of 'engender' etc. As this
is something which has puzzled me for years, I'd be most appreciative if
someone could enlighten me further on the matter.

Penny Lee.


(Dr P. Lee, School of Education (Soc Sci S), Flinders University, GPO Box
2100, Adelaide SA 5001. Australia.)

In LINGUIST 5-418, Dick Hudson raised the issue of 'the tendency [of
'Chomskyan' linguists] to hijack the term "generative", which is often used
(only by Chomskyans) to refer to Chomskyan linguistics'. Presumably in
response to this, someone last week (i forget who, and seem to have lost the
reference) raise the question of whether in modern theoretical linguistics the
word 'generative' has any value other than as an essentially empty political
label. I thought about it over the weekend, and for what it's worth here's how
 i would like to use the word myself. I think a 'generative grammar of a
language L' should be understood to be a set W of well-formedness conditions
such that one can say 'a string of linguistic elements is language L is well-
formed iff it meets the conditions in W.' Be it noted that these conditions
may be partial; W might be a set of *phonological* conditions applying to
*phonological* strings only, not strings of 'meaning-bearing' elements. More
importantly, note that in this proposed definition i am not making any a priori
 stipulations about the nature or form of the well-formedness condtions.
Although i like working with frameworks that address hierarchical constituent
structure and constituent order explicitly, and am quite willing to work with
transformational frameworks, i am not committed to any particular type of well-
formedness condition. The appropriate conditions may be global or local,
formal or functional, or any combination, in principle. I think the quest for
a fully formal theory of grammar is worth pursuing in at least the sense that
the Neogrammarian Regularity Hypothesis is valid: Even if it should prove to
need amendment, qualification, or restriction, the effort necessary to amend,
qualify, or restrict it will teach us a lot about language. But note that by
my proposed definition even a purely functional framework such as (certain
versions of) Cognitive Grammar would qualify as 'generative'. Driven into me
from the very beginning of my time as a grad student in linguistics over 10
years ago was the belief that what we as linguists are principally looking for
are well-formedness conditions. A theory that further explains why a
particular set W of well-formedness conditions, as opposed to a logically
possible but unattested set W', exists, is of course a highly desirable extra
bonus.

Sincerely,
Steven
--
Dr. Steven Schaufele fcoswsnytud.hu
Room 119
Research Institute for Linguistics (Department of Theoretical Linguistics)
Hungarian Academy of Sciences (Eotvos Lorand University)
P. O. Box 19
1250 Budapest
Hungary

*** O syntagmata linguarum liberemini humanarum! ***
*** Nihil vestris privari nisi obicibus potestis! ***