LINGUIST List 5.750

Mon 27 Jun 1994

Sum: Backwards anaphora

Editor for this issue: <>


  1. Dan Hardt, backwards anaphora (summary)

Message 1: backwards anaphora (summary)

Date: Mon, 27 Jun 1994 15:00:19 backwards anaphora (summary)
From: Dan Hardt <>
Subject: backwards anaphora (summary)

thanks to all those who responded to my query about
backwards anaphora. Below is a summary of the responses:

Hi. I'd say that Guy Carden has the franchise. He's

 The phenomenon of backward anaphora to which you refer is
termed "cataphora" by Halliday & Hasan in their work _Cohesion in
English_. For a discussion and examples you might refer to that
book. It does not present statistical information, but it is
very informative.

As for natually occuring examples, lots of works of fiction seem to
begin this way -- introducing a character with a pronoun. If you allow
deictics like "this", you get lots more cases.
also, there was an LSA paper by
Kathleen Ferrara, then of Texas A&M (no idea whether she's still
there). The title is "Cataphor in conversation: Looking Forward to

About ten years ago, I have investigated the functional aspects of
the use of backward anaphora in Dutch sentences. You can find the
results in the article 'The pragmatic value of cataphoric relations,
in: J.NUYTS & G. DE SCHUTTER (eds.), Getting one's words into line. On
word order and Functional Grammar, Foris, Dordrecht, 1987, 131-146.

refers to an:
***Analysis to see what kind of contructions that contain anaphora and
*** with what distribution occur in a sample of real text.
using Penn Treebank.

we did some
work a couple of years back on creating an "anaphoric treebank",
using parsed AP newswire material as input. described in:
Garside, Roger: The marking of cohesive relationships:
Tools for the construction of a large bank of anaphoric
data. ICAME Journal, No. 17, pp 5-28, Bergen, Norway, April 1993

Other documents describing the work are referred to in the

According to Garside, in the first 50 blocks of text analysed
(5442 sentences, 110822 words), there were 4151 cases of
a co-referential proform for which an antecedent could be
recovered from the text, of which 3998 were anaphoric,
134 cataphoric (backwards) and 19 of uncertain directionality.

I think (not sure) that there's some text count data in Carden, Guy &
Dieterich, T. (1980) Introspection, observation and experiment: An
example where experiment pays off. Journal of the Philosophy of Science
Assocation, vol. 2. I'm citing from my dissertation, Smyth, Ron (1986)
Cognitive Aspects of Anaphora Judgment and Resolution, distributed by the
Indian University Linguistics Club (it may well be in your library). My
reference to it discusses only their experiment, but I seem to remember
that their paper refers to some count data by Carden.

Try TV plot synopses (in TV guide or newspaper). You get things like "When
he realizes that X, Y does Z" fairly frequently, I think -- with no under-
standing probs.

Fromx: (Caoimhin P. ODonnaile) Kevin Donnelly

I think they are common in Gaelic. Gaelic doesn't like large and
elaborate noun phrases as a constituent of a sentence, so quite often
a pronoun is used, and this may either precede or come after its
antecedent noun phrase.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue