LINGUIST List 12.1657

Mon Jun 25 2001

Sum: Discourse Referents and Time

Editor for this issue: Marie Klopfenstein <>


  1. Robert Belvin, Discourse Referents and Time

Message 1: Discourse Referents and Time

Date: Fri, 22 Jun 2001 12:12:23 -0700
From: Robert Belvin <>
Subject: Discourse Referents and Time

Original query from Feb. 22 (the long time delay
in answering apparently doesn't affect the accessibility
of the general topic):

Does anyone know of studies which look at the extent to which the
passage of time determines whether a discourse referent is "current"
or not. For example, given a discourse like the following:

A: "Is John here?"
B: "I don't know"
 ...time delay...
B: "Is he in his office?"

How long a time delay might there be before "he" in the third sentence
could no longer felicitously be interpreted as "John", assuming no
other discourse referents had been introduced during the delay. I
realize this example is simplistic, but hopefully it suffices to
illustrate the area of study I'm interested in.

Summary of responses:

The vast majority of the responses converged on one central point,
which is that the passage of time, in and of itself, has little or
nothing to do with the accessibility of a discourse referent. It is
only because the passage of time tends to correlate with the
introduction of other referents that it might appear to influence
accessibility (of course the new referents in the above example must
be appropriate for "he" to refer to). Most of the responses that fall
into this category are reproduced below, though I have omitted some
who did not wish to be cited and/or whose remarks were essentially
identical to those shown. There was one response I wish to highlight:
It gives a somewhat different and very interesting perspective, having
to do with studies of anaphora in German Sign Language. Although the
reader will no doubt recognize that the basic generalization still
holds (time, by itself, is at most only marginally relevant), I think the
results discussed highlight the fact that it is the interaction of a
number of factors (including the passage of time), that determine the
accessibility of discourse referents:

J�rg Keller:

In fact I did some work on discourse referents and time in German
Sign Language (DGS). However, the book is written in German - so I
will give you a very brief summary on the findings. Overall, my study
focusses on the relation of space and reference (discourse-syntax
interface). In DGS like in American Sign Language (ASL), pronominal
forms are interpreted via a locus in signing space. E.g., the
(anaphoric) personal pronoun in SLs looks like the (deictic) pointing
gesture in English/German, i.e. pointing with the index finger to
someone/something present. To properly use a pronoun anaphorically in
a SL-utterance, the signer needs to place a referent first. One way to
achieve this is naming the referent and adding a locus-assigner (this
may be another pointing gesture). By doing so, the place pointed out
by the locus-assigner (the locus-marker) is associated with the
referent (this is morphosyntactic licensing). Next, any pronominal
form (or agreement verbs) may be directed to this place and will be
interpreted as if addressing the priorly place referent (this is
identification). So much about the general properties.

SL do not mark TNS but rather use TIME-adverbials to mark the
setting. As long as locus-markers are not wiped out by a signer/by the
introducting of a new discourse/different time etc. the locus-marker
remains intact and may be accessed. No, per se limitation of time or
number of utterances. Furthermore, some areas of space have
conventionalized meanings. E.g. one of these is/are the
locus-marker(s) for past/present/future. Another one is the
locus-marker for "specific-person-momentarily-not-present". The latter
is interesting insofar, as its meaning may change whereas its location
remains: Imagine several signers talking and one of them leaving for a
short time. While he is gone, pronominal access to the discourse
referent for the person absent may be accomplished via the
locus-marker for "specific-person-momentarily-not-present" (on the
right, over the shoulder of a signer). Suppose that the person does
not show up again. Then the locus-marker will gradually loose its
accessibility, but there is no definite stretch of time how long it
will last (also depends on discourse). Of course, if another person
leaves, the content of that locus-marker is replaced, pointing now to
the discourse referent of the person who left last. Also, the
"specific-person-momentarily-not-present" locus-marker will be no
good, if some new person arrives who does not know its content.

For DGS (and likely for other SLs too), therefore, it is not always a
matter of introducing new discourse referents to wipe out existing
ones, nor a simple matter of time. Rather, it is a matter of discourse
(e.g. accessbility of referents, new discourse substructures, closed
substructures), and/or time, and/or new referents, and/or new
interlocutors and/or processing limitations (e.g.working memory) and
conventions to name a few. For some more details see my abstract in
English in the dissertation abstracts of LinguistList; also an English
summary is given in Keller,J. (1999), Aspekte der Raumnutzung in der
Deutschen Geb�rdensprache. [The use of space in German Sign Language.]
Sign Language & Linguistics 2, S.249-257. Finally, the German book
is: Keller,J. (1998) Aspekte der Raumnutzung in der Deutschen
Geb�rdensprache. Hamburg: Signum Verlag.

Michael Niv:

First an anecdote by Ellen Prince.

A husband is trying to get back in bed in the middle of the night,
smashes his toe, and yells in pain "Damn, my foot!"
His wife wakes up and says with concern "same foot?"
She could be felicitously referring to an injury he
sustained decades before he'd met her, and that hand't
been overtly discussed in years.

Second, a little off topic is an old Stephen Wright joke:

Two babies are born within minutes of each other and are lying
in adjacent cribs in the hospital. 82 years later, they're both
dying of old age, and happen to be in adjacent hospital beds.
The one says to the other "so, what did you think?".

I think the joke is funny because the zero anaphor (think of this
life) is felicitous, in the face of the other implausibilities.

These both illustrate to me that there really isn't a deadline for
felcity of refrent, because a referent can be in short-term memory,
long term memory, or even be brought back into memory by an
inference-requiring referring expression.
(John bought a new car. The leather smelled great.)

Ronnie Sim:

you ask (Linguistlist 12. 520) how long a time delay may elapse before a
discourse referent is no longer considered current.

I suggest the answer is determined by pragmatics, not by any kind of
structural analysis for example.

Once, in the month of April, on visit to another part of the country, I
spoke to a gardener friend about a flowering shrub in his garden, 'That's a
nice bush'.

'I'll get you one' he replied. I was around a further week, but didn't see
him, nor get a plant to take home.

I was visiting the same place again in November of that year. My friend came
up to me and said 'I've got the bush for you.'

He clearly expected me to be able to uncover enough context to make sense of
this, in spite of the six month gap.

James Fidelholtz:

No concrete references or research to offer you, just a comment:
I think that, under given circumstances, any time interval at all may be
'too long' for 'he' to refer to John (eg, with gestures pointing to
another referent). On the other hand, no matter how long the interval,
if eg the referent has been 'saved' by a gestural icon in a certain
place in the speaker's gesture space, then 'he' still could refer to

****** Also, reference to a new book**
Albert Krahn:

I just got a flyer from Lincom Europe academic publishers.

A book they are offering is:
by Yan Zuo

Thanks to everyone who sent a reply:
Michael Niv
R.J. Sim
James L. Fidelholtz
Monica Budde
Albert Krahn
J�rg Keller
Matejka Grgic
Michal Lisecki
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue