LINGUIST List 5.558

Tue 17 May 1994

Misc: Binding & coreference, E-mail language

Editor for this issue: <>


  1. Steven Schaufele, Re: 5.533 Misc: Binding & coreference
  2. benji wald, Re: 5.553 E-mail language

Message 1: Re: 5.533 Misc: Binding & coreference

Date: Wed, 11 May 1994 11:46:38 Re: 5.533 Misc: Binding & coreference
From: Steven Schaufele <>
Subject: Re: 5.533 Misc: Binding & coreference

I'm responding to Stephen Straight's recently (LINGUIST 5-533) posted 'Article
Discussion Forum' on Bloom et al.'s paper in the latest issue of Language on
binding & coreference in child language.

First of all, i applaud Straight's initiation of this discussion. I have long
felt that this was one of the many things LINGUIST was particularly good for.
Let's have some more of it!

I haven't had a chance to see Bloom et al.'s paper; my issues of Language
continue to be mailed to my home address in Illinois while i spend a few months
 teaching in Budapest. I look forward to catching up on all this during the
summer. So, most importantly, i haven't had a chance to look at Bloom et al.'s
 bibliography, to see if they address, or are even aware of, something that
struck me as relevant as soon as i started reading Straight's comments. Since
Straight doesn't mention the relevant paper either, i decided i'd better bring
it up.

Jane Grimshaw and Sara Thomas Rosen have a paper in the Spring 1990 issue of
Linguistic Inquiry ('Knowledge and Obedience: the Developmental Status of the
Binding Theory', LI 21:187-222) that argues, if i remember correctly, that the
psycholinguistic experimental data are consistent with the following

Children are aware of Principles A and B of Binding Theory from a very early
age, indeed from the earliest stage at which relevant empirical evidence
becomes available. They also, in Anglophone environments, learn early to
recognize most lexical anaphors (given that English reflexive pronouns are
typically overtly marked by the morpheme 'self'). The fact that in performance
 they often violate the letter of Principle B -- e.g., reading the pronominal
in (1) as coreferential with 'John' -- is due not to a failure to understand
Binding Theory or a misidentification of the pronoun as an anaphor, but to a
conflict between pragmatics and syntax. The operative pragmatic constraint is
that pronouns require antecedents.

(1) John saw him.

We all know that pronouns can derive their antecedents from discourse context,
outside the clause, and the typical adult English speaker confronted with a
clause like (1) will assume that it has been taken from its natural context,
and will allow that the pronoun might be coreferential with an ipso facto
unexpressed NP. Adult speakers are also aware of the possibility of a pronoun
having arbitrary reference. Given these options, the purely syntactic
constraints of Binding Theory are not themselves constrained and can freely
forbid the selection of the only available overt NP as possible antecedent for
the pronoun.

Grimshaw and Rosen argue that the typical young child, insufficiently familiar
with pronouns with arbitrary reference and with grownups' propensity for taking
 clauses out of context and other forms of linguistic abuse, when confronted
with a clause like (1) without an appropriate discourse context, is thereby
confronted with a dilemma. General pragmatic considerations require that a
pronoun have a specific NP as its antecedent; Binding Theory forbids the
(apparently) only available option. The child will often if not always resolve
 this dilemma in favour of pragmatics rather than syntax, thereby forcing the
reading that violates the Binding Theory.

If i remember correctly, what i have just summarized is only one of several
arguments in Grimshaw & Rosen's paper, which would seem to support Bloom et
al.'s claim, as summarized by Straight,

> that Principles A and B of binding theory inhere
> in Universal Grammar and are exempt even from the
> necessity for parameter-setting ... In support of
> this thesis, BBNC present child language-output
> evidence that flaws in performance rather than
> competence account for those few cases in which
> even 2-year-olds violate, either in receptive or
> expressive language processing, the prohibition
> on local coindexing (Principle B) ...

I am not myself familiar with the data discussed by Grimshaw & Rosen, or by
Bloom et al., nor am i committed to Chomskyan Binding Theory in all its details
 as an essential part of Universal Grammar, and Straight's posted critical
questions seem to me to be valid and relevant. But i want to point out that
earlier work has already proposed a principled explanation for apparent Binding
 Theory violations in young children's linguistic behaviour.

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

*** O syntagmata linguarum liberemini humanarum! ***
*** Nihil vestris privari nisi obicibus potestis! ***
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: Re: 5.553 E-mail language

Date: Mon, 16 May 94 18:10 PDT
From: benji wald <IBENAWJMVS.OAC.UCLA.EDU>
Subject: Re: 5.553 E-mail language

With regard to Yates on speech/written differences, particular in the use of
 modal aux's, I have long been interested in this topic and have some
 comments that may be useful. To begin with, Jennifer Coates (1983, I think)
 noted many rather striking differences between British (Southern) spoken
 and written corpora with respect to relative frequency. Underlying such
 differences is the perennial volatility of spoken modal use, compared to
 the relative fixity of written use, at least to the extent of discouraging
 certain widespread spoken uses, e.g., "can" for "may" ("may" is almost
 totally epistemic in spoken languages, not deontic -- epistemic "you
 may go, I don't know" deontic "you may go, I allow it". The persistent
 trend toward favoring historically past-inflected modals is also wide-
 spread in speech, e.g., "might" for "may" in either of the above uses,
 certain contextual uses of "could" for "can" etc. The withering away
 (death in general, I would say) of "shall" in speech is another case,
 its function picked up by "should" (by the pref for past-mark) and "will"
 according to context and/or dialect. Let's not dwell on the conditional
 distinction between "should" and "would" which apparently nobody alive
 understands anymore, though attempts to use it may distinguish speech from
 writing -- and among British dialects there are differences in which uses
 "should" and which uses "would" in relevant contexts -- without making
 the moribund normative distinction.
"must" long ago lost its present "mote" and is another case
where epistemic use is becoming its function at the expense of
its deontic use ("can" takes over the deontic function -- e.g.,"you
shouldn't/can't go" instead of "you mustn't go" -- forget the
possible meaning difference in written English). Coates also noted "hafta"
 on epistemic uses, e.g. "if that's the case, this HASTA be right!",
 although she seemed to think this is more prevalent in American than
 British speech. She also noted that "gonna" (i.e. going to) is much
 less often used in writing than in speech. There are a myriad of other
 features, all of which are quite specific, but which add up to a very
 noticeable difference between speech and writing when summed together (or
 even taken separately when the right contexts are recognised). I would
 be happy to discuss any particular modal developments in greater detail,
 where it might be helpful for analysis. My particular interest is in the
 various uses of WOULD, where I disagree with Coates that it ALWAYS has
 a "hypothetical" implication in contrast to USED TO (so she can count all
 instances by concordance without considering context and painstakingly
 distinguish the various "habitual" uses -- the correct distinction in
 "habitual" contexts is quite different from the "hypothetical" dimension).
 A final interest of mine is in the different relative frequencies between
 "oughta" and "should" (in equivalent contexts for "should", not ALL its
 contexts). The degree of difference varies according to dialect in speech.
 Any differences in different areas of England would be of interest, although
 I am more interested in differences among American dialects, e.g., I think
 Southern US English is most favourable to "oughta" among dialects, but
 there are interesting historical and evolutionary implications to how
 much more (if not less) other dialects are favourable to "should" in
 comparison with "oughta".
 A final point is that is that the conservatism of writing may be the
 primary reason for the difference between speech and writing with respect
 to modals, but it is, of course, not the only or primary reason for
 the difference with respect to many other features. There
 are also considerations of "mode", i.e., differences of favored organising
 principles for the visual and auditory channels, esp important because
 writing lacks the richness of prosodic cues which aids interpretation of
 speech (punctuation is a much less successful adaptation of language to
 writing than the alphabet.) I doubt that genre in itself is responsible
 for linguistic differences, but is only a reflection of the different
 tendencies of organisation favoured by whatever the topic of a particular
 text(/spoken discourse) is about. I concede that there are some genre-
 driven aesthetic considerations, but I don't think they amount to much
 compared to other factors.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue