Deictic Expressions and Discourse Segment in English and Japanese ...

<Topic area: Semantics, Pragmatics, Spoken Discourse>
Deictic Expressions and Discourse Segment
in English and Japanese Task-Oriented Dialogue
Etsuko YOSHIDA
1 Introduction
This paper has three aims: (1) to provide a pragmatic perspective in a contrastive
study of deictic expression in English and Japanese spontaneous dialogue, (2) to suggest
the implications for both GHZ (1993)’s Givenness Hierarchy model and Centreing
model of discourse reference, based on a preliminary study of a task–oriented speech,
and then (3) to predict the way deictic elements can contribute to discourse organisation
in structuring and focusing the specific discourse segment.
After giving some information about data of English and Japanese dialogue
corpus in section 2, I first give the background assumptions on English and Japanese
deictic expressions with some evidence from English and Japanese dialogue in section
3. In Section 4, I present the distribution of deictic expressions and discuss how the
Givenness Hierarchy can contribute to the distribution and its pragmatic interpretation.
Section 5 presents the results of a preliminary analysis on the correlations between
deictic expression and centre transition in Centreing framework. Finally, I present
general hypotheses that suggest ways in which the choice and distribution of
demonstratives of English and Japanese counterparts can be explained in an interactive
discourse model in section 6.
2 English and Japanese Parallel Corpus
2.1 English labelless Map Task Corpus
For the present study, I selected individually collected eight English dialogues
based on maps that did not have written labels to identify the landmarks. This is a small
set of data experimentally collected after completing the original Map Task Corpus
(MTC) compiled by Human Communication Research Centre (HCRC), Edinburgh, UK.
One of the predictions is that the speakers in labelless MTC, especially, the giver, tend
to use deictic pointers more frequently in describing the landmarks and explaining the
routes between landmarks than in the labelled MTC. Therefore, the labelless MTC is
relevant data in the sense that the lack of written labels on the maps encourages the
participants to construct their own expressions to identify entities of landmarks and to
employ a number of deictic expressions than the map of ‘ready-made’ written labels
(Deverell:1994: 22).
2.2 Japanese labelless Map Task Corpus
The Japanese MTC has been conducted as a project at Chiba University
(henceforth Chiba Corpus), Japan, since 1993, based on the Edinburgh MTC with
respect to map and route designs and situational parameters such as familiarity and eye
contact. (see Anderson et al.1991) As in the HCRC study, 128 dialogues have been
collected. Following the specification and transcription policy of Chiba Corpus, the
small data of 8 dialogues based on Japanese labelless MTC have been collected at Mie
University, Japan. using the same labelless maps as MTC and the same experimental
design: familiar and unfamiliar pair, each participant plays a role as giver twice on the
1

---

same map with different followers, and then as follower twice with different givers and
different maps. In Japanese data, I also predict that there are a number of deictic
expressions that may be considered parallel in discourse function to the one I assume in
English labelless MTC.
3. Features of deictic expressions in English and Japanese: Overview
Deixis, in a broad sense, is potentially a context-dependent linguistic expression,
typically anchoring in the perspective of the speaker. Fillmore (1982: 35) defined deixis
as ‘the name given to uses of items and categories of lexicon and grammar that are
controlled by certain details of the interactional situation in which the utterances are
produced’. Deixis includes extended use of linguistic categories such as spatio-temporal
locating adverbials, demonstratives, tense, and social deictic terms. Significantly,
deictic expressions are discussed as crucial discourse-grammatical markers that exhibit
deictic features as one of the most basic functions of human communications. This
view may lead to the hypothesis that the pragmatic properties of demonstratives may be
extended from the distinctive proximal-distal dimension to the speaker-addressee
dimension, that is, ‘interactional’, especially in dialogic discourse rather than narratives
or monologic speech. (see Cheshire 1996) Furthermore, the research based on the cross-
linguistic/contrastive studies in deictic expression in discourse may provide us with
ample evidence from various languages regarding the questions such as ‘which uses of
demonstratives may be universal?’ and ‘which are language specific?’ (Himmelman
1996; Takubo-Kinsui 1990; Cornish 1996, 1999; Valluduvi and Engdahl 1996; and
Diessel 1997, 1999).
For my present purpose three aspects of deictic expression are considered: spatial
deixis, anaphoric demonstratives, and discourse deixis in English and Japanese. The
reason I choose these to focus on is to demonstrate that the actual use of these
categories is interactively related in discourse processing of dialogue. Although each
deictic system of English and Japanese is different in linguistic form and function, there
may be a pragmatically significant parallelism on the correlation between the choice of
deictic expressions and the discourse segment.
4 Distribution of deictic expressions: a preliminary analysis
The Japanese and English labelless MTC contains a number of deictic expressions
that are employed to refer to various discourse entities. I wish to concentrate my
attention on those deictic expressions that are used to refer to the landmarks on the maps
and their notable patterns of distribution in relation with discourse factors.
4.1 Japanese MTC
Table 1 and 2 shows the distribution of deictic expressions in Japanese MTC based
on the same map and the same giver, but in first session as the giver (Table 1) and in the
second session as the giver (Table 2) each time with different followers.
Japanese
Spatial
Anaphoric Demonstratives
Discourse
TOTAL
MTC.ab
Deixis
Deixis
Adnominal
Pronoun
PROX
1
1
MEDIAL
6
17
1
2
26
DISTAL
2

---

TOTAL
6
17
1
3
27
Table 1: The distribution of Deictic expressions in Japanese MTC. ab.
Japanese
Spatial
Anaphoric Demonstratives
Discourse
TOTAL
MTC.ac
Deixis
Deixis
Adnominal
Pronoun
PROX
1
1
MEDIAL
13
9
1
1
24
DISTAL
TOTAL
13
10
1
1
25
Table 2: The distribution of Deictic expressions in Japanese MTC. ac.
Both Table 1 and 2 show that the major type of demonstratives in Japanese dialogue is
the anaphoric use of Medial so-series. There is only one occurrence of Proximal deixis
in each turn of the giver, but none as a Distal use in both turns. Among the Medial use,
in the 1st turn, adnominal demonstratives sono NP are most frequently chosen (17
times), but demonstrative pronoun sore is only used once as anaphoric use and twice as
discourse deixis. In the 2nd turn, on the other hand, the number of sono NP is reduced (9
times), but the occurrence of spatial deixis is twice as frequent (13 times) as in the 1st
turn (6 times). With the limited amount of data, no reliable difference has not been
discussed in the distribution of deictic expressions between the giver’s first and second
sessions. However, this may be a possible factors in the future course of study, because
higher occurrence of sono NP in the 1st session can be related with the Givenness
Hierarchy (henceforth GH; see 4.3)’s implication that adnominal demonstratives require
at least activated. That is, in the giver’s first search of common objects on each map,
sono NP tends to reflect one of the devices called ‘immediate anaphor’ by Lichtenberk
1996 quoted in Himmelmann 1996), which is an immediately subsequent reference after
the first mention of an entity and is realized in the process of establishing an entity in
the discourse. In his or her second turn, on the other hand, the giver’s labor in
describing and identifying a certain object and route across a map become little. Thus,
the finding of the reduced number of adnominal demonstratives with practice may
suggest that practice at a task result in more controlled and centre-oriented treatment of
entities, ‘due to lower conceptual planning demands’ (cf. Branigan, Lickley and
McKelvie:1999) Furthermore, the limited use of the Medial demonstrative pronoun
sore may mean that the speaker avoid putting the hearer to some unjustifiable effort in
retrieving the plausible referent. This is because Medial adnominal demonstrative sono
N is less accessible, but contains more lexical information than pronoun sore, though
both forms would be appropriate only if the referent is currently activated. In section
5.1 we will see the evidence that sono N can be more appropriate when its entity is both
familiar and activated. In addition, I will argue that demonstratives not necessarily refer
to less salient referent. Japanese Medial demonstratives, especially, contribute to
discourse coherence despite their lower ranking of transition. This is also suggested by
the example of ‘focus-shift’ proposed by GHZ (1993:298).
(1)
a. Toori e dete shibaraku hashitteku
street to go out for sometime run
‘He goes out onto the street and runs for some time.’
b. To nanka yatai mitaina omise ga atte.
3

---

and something stall seem shop NOM be
‘There is a shop like a stall.’
c. KARE wa sono omise no toko e itte.
He TOP that shop GEN place to goes
‘He goes to that shop’
This example shows that overt 3rd person pronoun KARE (‘HE’), which is predicted to
behave like demonstratives, with topic marker ha typically implicates that the referent is
not in focus, i.e. it implicates a focus-shift. Also note that sono N can tend to occur as
what Lichtenbark calls ‘immediate anaphor after first mention’, as in that shop
referring to a shop like a stall. Here possible replacement of bare NP omise is rather
ambiguous because the first mentioned NP is not semantically explicit and Medial
pronoun sore/locative soko can be a possible candidate as the subsequent mention.
4.2 English MTC
Table 3 and 4 shows the distribution of deictic expressions in English MTC
based on the same map and the same giver, but in first session as the giver (Table 3) and
in the second session as the giver (Table 4) each time with different followers.
English MTC
Spatial
Anaphoric Demonstratives
Discourse
TOTAL
Lleq4c1
Deixis
Deixis
Adnominal
Pronoun
PROX
1
1
DISTAL
7
6
13
5
31
TOTAL
7
7
13
5
32
Table 3: The distribution of Deictic expressions in English MTC.Lleq4c2.
English MTC
Spatial
Anaphoric Demonstratives
Discourse
TOTAL
Lleq4c8
Deixis
Deixis
Adnominal
Pronoun
PROX
2
2
DISTAL
4
1
11
1
17
TOTAL
6
1
11
1
19
Table 4: The distribution of Deictic expressions in English MTC. Lleq4c8.
Maybe one of the interesting parallelism I find between English and Japanese is that, in
the 1st turn, adnominal demonstratives are used 6 times as Distal and once as Proximal,
but in the second turn, their occurrence are reduced to only once used as Distal
adnominal demonstratives. In English, Distal demonstrative pronouns are much more
frequently used across the task than those in Japanese. Since I have not counted any full
nouns and zero pronouns in Japanese, overall distribution and thorough discussion on
the contrast between the two languages lies in the further research. Only the suggestion
will be that in Japanese, adnominal demonstratives tend to be used either for immediate
anaphora (topic-candidate) or anaphora referring to non-centre referent (topic-shift),
whereas in English, Distal pronominal demonstratives are unmarked choice of referring
an entity as non-centre referent in activated status, and adnominal use is rather marked
because it requires familiar reading. Let us consider several problems on distribution
regarding the matching between the forms of referring expression and their relevant
cognitive statuses in the following section.
4

---

4.3 Givenness Hierarchy
Based on Prince’s (1981) familiarity scale, a series of studies by Gundel,
Hedberg, and Zacharski (1988, 1989, 1990, 1993) successfully define the cognitive
status of referring expressions in the framework called the Givenness Hierarchy (GH).
They claim that different determiners and pronominal forms conventionally signal a
different cognitive status, thereby enabling the addressee to restrict the set of possible
referents (274), which is represented below in English and Japanese:
in
uniquely type
focus > activated > familiar > identifiable > referential > identifiable
that 
English
{it} this  {that N} {the N} {indefinite this N} {a N}
this N
Japanese f kare ‘he’
ano N ‘that N’
kore ‘this’
distal
----------------------fN---------------------------------
sore ‘that’ medial
are ‘that’ distal
kono N ‘this N’
sono N ‘that N’ medial
As the table shows, English weak pronoun (unstressed) correspond to Japanese zero
pronoun, which is roughly acceptable. Since Japanese is a language that has no
definite/indefinite article, entities realized in the form of bare nouns must cover three
range of cognitive status. That is, bare nouns can be compatible with any of type
identifiable, referential, or uniquely identifiable. The first issue is to define how
Japanese bare NP are allocated into any of these three statuses. The reliable
grammatical solution to distinguish them may be that according to morphological
information, bare nouns are basically distinguished into either NP with subject marker
ga as indefinite or NP with topic marker ha as definite. However, this rule, in reality,
does not always work successfully. Then, English that can be compatible with either
Medial so- or Distal a- series in Japanese, but practical division of English that into
Japanese is not easy. Basically, Medial so- series, either adnominal (sono N) or
pronoun (sore), require at least activated, because, as we have noted, they are basically
used as anaphoric. In contrast, Distal a- series in both adnominal (ano N) and pronoun
are basically deictic and requires at least familiar status. This fact is shown in the
following example:
(2)
Kinou Yamada san ni aimashi ta. Ano hito kawatta hito desu ne.
Yesterday f Yamada Mr/Mrs OBJ meet PAST That person different person is FP.
‘Yesterday I met Mr Yamada. That person is different, isn’t he.’
(3)
(The husband ask his wife for something, who has been together with him for many years.)
Oi, are motte kite kure.
INTERJEC that fbring come IMP
‘Hey, bring that to me.’
(Kinsui 1999: 72)
That person in example (2) appears to be anaphoric in the sense that the referent is in
the immediately previous utterance, but ano (that) indicates accessible referent in the
5

---

hearer’s current memory rather than the current discourse. Example (3) is also the
evidence that ‘the entity’ instead of ‘the referent’ is in the hearer’s current memory
rather than in the extra-linguistic situation. Here distal are indicates that the accessible
entity lies in the hearer’s long-term memory that the speaker think should be retrievable
without any linguistic context. This usage is what is called as ‘reminder’ (GHZ) or
‘recognitional’ (Himmelman) and more like ‘interactional’ in Kitano’s (1999) sense.
Therefore:
(3)’
Ano hito/ *Sono hito/ ?Kare kawatta hito desu ne.
‘That person/He is different, isn’t he.’
As example (3)’ shows, purely anaphoric sono is not acceptable, because the entity is
not currently activated. Kare may possibly be an alternative reference, because this
English-translated ‘demonstrative pronoun’ in GHZ’s interpretation is only acceptable if
the speaker knows the person quite well. Otherwise, Kare is rather impolite, because
the speaker only met him once. In English, on the other hand, pure referential pronoun
he will be the most natural translation. Thus, Japanese demonstratives may need more
special account on the correlation between forms and cognitive statues.
Furthermore, Gundel (1998) attempts to integrate this framework with
centreing theory on the ground that GHZ and centreing theory make similar predictions
about the distribution and interpretation of pronouns and full NPs in naturally occurring
discourse (183). GHZ (1993) assumes that pronouns (and zero pronoun in Japanese)
refer to entities at the ‘centre of attention’, whereas demonstratives have less accessible
and less salient referents. In fact, the concept of a ‘focus shift’ (GHZ:1993:297) may be
correlated with the states of SMOOTH- or ROUGH-SHIFT transition in Centreing
terms. Let us now look into the Centreing Theoretic account of reference in discourse
in section 5.
5 Centreing Theory
Centreing is a model of the conversant’s centre of attention in discourse that is
concerned with the relationship of attentional state, inferential complexity, and the form
of referring expressions. (Walker, Joshi, and Prince 1998) What they call ‘centre’ is
distinguished into three different types: Cf(Ui, D), Cb(Ui, D), and Cp. Cf(Ui, D)
indicates FORWARD-LOOKING CENTRES, representing discourse entities evoked by
an utterance Ui in a discourse segment D (Webber 1978; Prince 1981). Cb (Ui, D)
indicates the BACKWARD-LOOKING CENTRES, which is a special member of Cf
representing the discourse entity that the utterance Ui most centrally concerns. The Cb is
the entity that can commonly correspond with what is called ‘topic’ in Givon’s sense as
in current focus of attention in the utterance. Cp is the PREFERRED CENTRE
representing a prediction about the Cb of the following utterance. These three centres
structure in their latest stage three constraints, two rules and the typology of transition
states (CONTINUE, RETAIN, SMOOTH-SHIFT, and ROUGH-SHIFT) as presented
and interpreted by Walker, Joshi, and Prince (1998).
5.1 Centre transition in Japanese MTC
According to Kameyama’s original proposal that zero pronouns in Japanese
correspond to unaccented pronouns in English (Kameyama 1985), Rule 1 is extended
6

---

directly to zero pronouns: ‘If some element of Cf(Ui+1, D) is realized as a pronoun in Ui
, then so is Cb(Ui, D)’ That is, Cb is presumed to be the most salient entity in the current
utterance as ‘a local topic by both the speaker and the addressee’. (Brenan 1995) Cf
ranking for Japanese according to discourse salience is given below (Walker, Iida, and
Cote 1994; Iida 1998):
Cf Ranking for Japanese:
(GRAMMATICAL OR ZERO) TOPIC>EMPATHY>SUBJECT>OBJECT2>OBJECT>OTHERS
Here is a preliminary analysis of transcripts from Japanese task-oriented dialogue:
(4)(a) G: sutaato chiten no minami gawa ni tatemono ga aru no wa wakarimasu
starting point GEN south side at building SUBJ there is TOP see
ka.
Q
‘Do you see there is a building at the south side of the starting point?’
Cb: [?]
Cf: [BUILDING, SOUTH] No Cb
(b) F: hai f ari masu.
Yes SUBJ is
‘Yes, there is (a building).’
Cb: BUILDING
Cf: [BUILDING] CONTINUE
(c) G: mazu sono tatemono o mezashite f susumi masu ga
to begin with DEM(M) building OBJ towards SUBJ go
‘To begin with, (you) go towards that building.’
Cb: BUILDING
Cf: [BUILDING, YOU] RETAIN
(d) G: sono tatemono no nishigawa o tootte f nankashite kudasai
DEM(M) building GEN west side OBJ passing SUBJ go down south please
‘(You) go down south passing the west side of that building, please.’
Cb: BUILDING
Cf: [BUILDING, WEST, YOU] RETAIN
In (4a) a building introduces a centre that is established as the Cb in (4b) by the rule
Zero Topic Assignment (henceforth ZTA) proposed by Walker, Iida, and Cote (1994)
and Iida (1998). Then in (4c) the centre corresponding to a building is realized by
demonstrative determinar (Medial sono) the centre for that building is ranked lower on
the Cf than the centre for you by zero, but that building is still the Cb, so the centreing
transition is a RETAIN. Then the building gains the centre of attention in the
sebwequent utterance:
(e) G: jaa sono tatemono ha ima kita ni mieteimasu ka sutaato chiten
so DEM(M) building TOP/SUBJ now north in seen Q strarting point
kara
from
‘So now is that building seen in the north from the starting point?’
Cb: BUILDING
Cf: [BUILDING, STARTING POINT] SMOOTH SHIFT
(f) F: to ima f nishigawa ni arimasu kedo
now SUBJ west side in is though
7

---

‘(The building) is in the west side now, though..’
Cb: BUILDING
Cf: [BUILDING] CONTINUE
In (4e) that building is re-mentioned on the topic/subject position after intervening
speech that contains the new entity field. The only centre that provides a link to the
prior discourse is the centre for that building, so that the centre is the Cb. However,
since the Cb in the immediately preceding utterance is fields, the Cb must shift in a
SMOOTH-SHIFT transition. Then follower’s utterance (4f) is a CONTINUE transition,
because Cb is the same as in (4e) and because the Cp (4f) is the same as the Cb(4f).
Thus, as we have seen, grammatical or Zero Topic has been the Cb to continue as the
Cp, and that NP occurs only in RETAIN and SMOOTH-SHIFT transition.
5.2 Centre transition in English MTC
Cf ranking for English is:
Subject > Object(s) > Other
As the data shows, the way the speaker of MTC make a discourse entity that is activated
salient is mainly demonstrated by the use of anaphoric demonstratives and the repetition
of the full NP:
(5)
(a) *TA 18
Right, if you go to the left... do you have something just directly
below the cross?
Cb: [?]
Cf: [SOMETHING, CROSS] No CB
*TB 19
Yeah.
(b) *TA 20
< If you go to the left of that, and draw a line down to about... two or
three centimetres above the page /
Cb: SOMETHING(THAT)
Cf: [THAT, LINE, PAGE] RETAIN
(c) *TB 21
Below the level of the object below the cross? >
Cb: SOMETHING(THE OBJECT)
Cf: [OBJECT, CROSS] RETAIN
(d) *TA 22
above the bottom of the page.
Cb: PAGE
Cf: [PAGE] ROUGH SHIFT
(e) *TA 23
Yeah, go to the left of the object, and go down... Do you have another
{a dia} {m erm} object at the bottom.
Cb: OBJECT
Cf: [OBJECT, ] ROUGH SHIFT
Cb: ANOTHER OBJECT
Cf: [ANOTHER OBJECT] ROUGH SHIFT
8

---

*TB 24
No.
(f) *TA 25
Below that one.
Cb: OBJECT
Cf: [THAT ONE] RETAIN
*TB 26
No.
The first-mentioned entity something in (5a) is not pronominalised in the subsequent
utterance. The speaker re-mention the entity by demonstrative pronoun that in (5b), the
full noun the object in (5c) and (5e), and the contrastive use of demonstrative adnominal
that one in (5f) as re-identification. The entity initially introduced in the object position
in (5a), and then subsequently maintained in the object of prepositions, yet never occurs
in the subject position. This can mean that a series of entities cannot gain saliency that
are to become the discourse centre, and ends in the transition of RETAIN and ROUGH-
SHIFT. As the data shows, Cf ranking may properly predict that the deictic forms of
reference are less salient. However, these deictic elements can be pragmatically
coherent in discourse organization despite the location in the lower ranking.
6 Concluding remark and further research
In this paper, I have presented a corpus-based analysis of the distribution and
discourse function of the deictic expressions, and have found out that sono NP and
that/that NP mainly occurred in RETAIN and SMOOTH- or ROUGH-SHIFT centreing
transitions. This result can support the claim that ‘anaphoric demonstratives are often
used to indicate a referent that is somewhat unexpected and not currently in the focus of
attention’(GHZ 1993, Dissel 1999). Especially, demonstrative determiner has some
special effect or implicature, where the referent has been activated but not focus.
(Gundel 1998) That is, both transitions explicitly show that the current centre, that is,
the focus of attention, has been shifted on ‘the discourse node which the pointer is
marked moves from node to node on the tree representing the information of the
discourse’. (Linde:1979: 345) Presumably, contrary to most of the researchers who
considers deixis as organised around ‘deictic centre’ of the speaker, I will argue that the
deictic centre can be shifted from the speaker to the addressee, depending on how the
current discourse is organised: the typical evidence can be the use of the Medial so-
series in Japanese and that/that NP in English, because these are the linguistic forms
that is accessible to the hearer’s domain. I will further investigate what kind of
grammatical and morphological rules can indeed affect the interpretation of
grammatical or Zero Topic and deictic elements, and to what extent other discourse
factors such as connectives and cue phrases can affect the discourse organization in
structuring and focusing the discourse segment. Thus further study needs to explain that
in what condition anaphoric demonstratives are employed, and what types of discourse
factors are correlated with the discourse segment by signalling the boundary, and how
the interaction between grammatical or Zero Topic and deictic elements is interpreted.
REFERENCES
Anderson, A. H., M. Bader, E. G. Bard, E. Boyle, G. Doherty, S. Garrod, S. Isard, Kowtdo, J. McAllister,
J. Miller, C. Sotillo, H. Thompson, and R. Weinert. (1991) ‘The HCRC Map Task Corpus’,
Language and Speech, 23(3): 351-66.
9

---

Branigan, Holly, R. Lickley and David Mckelvie (1999) ‘Non-linguistic influences on rates of disfluency
in spontaneous speech’. In ICPHS
Brennan, S.E. (1995) ‘Centreing attention in discourse’. Language and cognitive processes 10. pp.137-
167.
Cheshire, J. (1996). ‘That jacksprat: An interactional perspective on English that’. Journal of Pragmatics
25, 369-393.
Cornish, F.(1996). ‘’Antecedentless’ anaphors:deixis, anaphora, or what? Some evidence from English
and French’. Journal of Linguistics 32, 19-41.
-------------(1999). Anaphora, Discourse, and Understanding: Evidence from English and French. Oxford
University Press.
Deverell, Jacqui.(1994) ‘Referring expressions in the Map Task Corpus’ unpublished paper
(Centre for Cognitive Science, U of Edinburgh)
Diessel, H.(1997).’ The Diachronic Reanalysis of Demonstratives in Cross-Linguistc Perspective’. CLS
33: The Main Session. 83-97.
----------. (1999) Demonstratives: Form, function, and grammaticalization. John Benjamins.
Ehlich, K. (1982). ‘Anaphora and Deixis: Same, Similar, or Different?’ In R. J. Jarvella and W.
Klein (ed.), Speech, Place, and Action, John Wiley & Sons. 315-338.
Fillmore, C. J. (1982). ‘Towards a Descriptive Framework for Spatial Deixis’. In R. J.
Jarvella and W. Klein (ed.), Speech, Place, and Action, John Wiley & Sons. 31-59
------------. (1997) Lectures on Deixis. Stanford: CSLI Publications.
Fox, B.A. (1987). Discourse Structure and Anaphora: Written and Conversational English.
Cambridge: Cambridge University Press.
Givon, T. (1983). ‘Topic continuity in discourse: An introduction’, in Givon and Ute Language
Program, T. (ed) Topic Continuity in Discouse: A Quantitative Cross-Language Study..
London: John Benjamins. 1-41.
-------------.(1983). ‘Topic continuity in spoken English’, in Givon and Ute Language Program, T.
(ed)
Topic Continuity in Discouse: A Quantitative Cross-Language Study.. London: John
Benjamins. 343-363.
Gundel, J. K., N. Hedberg, and R. Zacharski. (1988). ‘On the generation and interpretation of
demonstrative expressions’, Proceedings of the 12th International Conference on
Computational Linguistics(12), 216-221.
------------. (1989). ‘Givenness, implicature and demonstrative expressions in English discourse’,
Papers from the Regional Meeting of the Chicago Linguistic Society 25(2), 89-103.
------------. (1990). ‘Givenness, implicature and the form of referring expressions in discourse’,
Berkeley Linguistics Society(16), 442-53.
------------. (1993). ‘Cognitive status and the form of referring expressions in discourse’, Language,
69(2), 274-307.
Gundel, J.K. (1998) ‘Centreing theory and the Givenness hierarchy: towards a synthesis’. In M.A.
Walker, A.K. Joshi & E.F. Prince (Eds.), Centreing theory in discourse (pp.183-198). Clarendon
Press
Himmelmann, Nikolaus P. (1996) ‘Demonstratives in Narrative Discourse: A Taxonomy

of Universal Uses.’ In Studies in Anaphora, Barbara Fox (ed). Amsterdam:

John Benjamins. 205-254.
Iida, M. (1998). ‘Discourse coherence and shifting centres in Japanese texts’ In M.A. Walker, A.K. Joshi
& E.F. Prince (Eds.), Centreing theory in discourse (pp.161-180). Clarendon Press
Kameyama, M. (1985). Zero anaphora: The case of Japanese. Unpublished doctoral dissertation, Stanford
University.
Kitano, Hiroaki. (1999) ‘On interaction and grammar: evidence from one use of the Japanese
demonstrative are (‘that’). Pragmatics. Vol.9. No.3.
Kinsui, S (1999) ‘On the relation between the deictic use and the non-deictic use of the Japanese
demonstratives’ In Natural Language Processing (Shizen gengo shori) pp. 67-91.
Lakoff, R. (1974). ‘Remarks of this and that.’ Papers from the Regional Meeting of the
Chicago Linguistics Society(10), 345-56.
Linde, C. (1979). ‘Focus of attention and the choice of pronouns in discourse’, Syntax and
Semantics:Discourse and Syntax, Vol. 12. New York: Academic Press. 337-354
Lyons, J. (1977). Semantics. (Vol. 1 and 2). Cambridge: Cambridge University Press.
Miller, Jim. and Regina Weinert. (1998) Spontaneous Spoken Language. Oxford:
10

---