Proceedings of the 7th SIGdial Workshop on Discourse and Dialogue, pages 68–75,
Sydney, July 2006. c©2006 Association for Computational Linguistics
Semantic and Pragmatic Presupposition in Discourse Representation
Theory
Yafa Al-Raheb
National Centre for Language Technology
School of Computing
Dublin City University, Ireland
yafa.alraheb@gmail.com
Abstract
This paper investigates semantic and prag-
matic presupposition in Discourse Repre-
sentation Theory (DRT) and enhances the
pragmatic perspective of presupposition in
DRT. In doing so, it draws attention to the
need to account for agent presupposition
(i.e. both speaker and hearer presuppo-
sition) when dealing with pragmatic pre-
supposition. Furthermore, this paper links
this pragmatic conception of presupposi-
tion with the semantic one (sentence pre-
supposition) through using ‘information
checks’ which agents are hypothesized to
employ when making and receiving utter-
ances.1
1 Introduction
DRT, with its detailed apparatus for the representa-
tion of context, offers the most obvious framework
for investigating presupposition in depth (Kamp
1984, 1988, 1990, 1995, 2001a, 2001b; Kamp and
Reyle 1993; Kamp et al. 2005). However, de-
spite the suitability of DRT for pursuing a detailed
account of presupposition, it is argued that in or-
der to enrich our understanding of presupposition
within the DRT framework, this framework itself
needs to be modified (cf. Al-Raheb 2005). The ap-
proach presented in this paper understands presup-
position within the parameters of dynamic seman-
tics (van der Sandt 1992), part of which is DRT,
but attempts to go beyond that in order to make the
understanding of presupposition within DRT more
pragmatic. The dynamic semantics view of pre-
supposition is incomplete from a pragmatic stand-
point because it neglects the connection between
1I gratefully acknowledge support from Science Founda-
tion Ireland grant 04/IN/I527.
beliefs and presupposition, hence neglecting the
connection between pragmatic presupposition and
semantic presupposition in DRT.
To account for pragmatic presupposition as well
as making presupposition within DRT more prag-
matic, presupposition is understood to be a prop-
erty of the agent. In essence, the effect of pre-
supposition is to give insights about the speaker’s
beliefs as well as the speaker’s beliefs about the
hearer’s beliefs. Speaker belief leads to presuppo-
sition, which indicates the beliefs of the speaker
to the hearer. Presupposition is a reflection of the
speaker’s state of mind. This is stronger than what
is generally conceded in the literature. Geurts
(1996, 1998, 1999) maintains that a presupposi-
tion should not necessarily reflect the beliefs of
the speaker, but rather the speaker’s commitment
to the truth of the presupposition. If, for example,
we were to use Stalnaker’s (2002) example,
(1) I have to pick my sister up from the air-
port.
Geurts argues that the speaker does not have to be-
lieve she has a sister, but just needs to be ‘commit-
ted to’ the truth of the presupposition that she has a
sister. In other words, the speaker need only com-
mit to the presupposition (P) being true. The ap-
proach presented here takes a somewhat stronger
position than Geurts’ position (Geurts 1999) be-
cause it assumes that Grice’s Cooperative Princi-
ple is in place (Grice 1975, 1989). If we make the
simplifying assumption that the agents in the dia-
logue are being cooperative, are not lying, are be-
ing relevant, etc., we can take the stronger position
that the information introduced by the presupposi-
tion, here ‘having a sister’, is indeed a belief held
by the speaker.
68
	





	












Figure 1: Speaker, Sentence and Hearer Presupposition
Viewing presupposition from the more prag-
matic perspective of the agents’ point of view
in dialogue leads us to viewing presupposition
from both the speaker’s and the hearer’s points
of view. Distinguishing speaker presupposition
from hearer presupposition helps make the ap-
proach to presupposition within DRT more prag-
matic, since we are speaking not of truth condi-
tions but of states of mind in communicative in-
teraction. Therefore, this paper deals with two
types of agent presupposition, speaker presuppo-
sition and hearer presupposition (cf. section 3).
This is different from semantic presupposition, i.e.
sentence presupposition (cf. section 2). Agent
presupposition differs from sentence presupposi-
tion in that the latter stems from sentence mean-
ing, whereas the former attaches itself to the be-
liefs of the speaker and her intentions. It is argued
here that the semantic and pragmatic notions of
presupposition in DRT can be linked through link-
ing agents’ beliefs to the utterance being commu-
nicated.
2 Semantic and Pragmatic
Presupposition
The literature has mostly considered the hearer’s
side of receiving the presupposition when deal-
ing with agent presupposition (pragmatics) as op-
posed to sentence presupposition (semantics). For
instance, van der Sandt (1992) deals with accom-
modation from the hearer’s perspective, not dis-
tinguishing between speaker and hearer presup-
position. However, the relationship between sen-
tence presupposition and agent presupposition can
be explained by dividing agent presupposition into
speaker presupposition and hearer presupposition.
From the speaker’s point of view, speakers
make utterances to communicate new informa-
tion. Generally speaking, to generate a commu-
nicatively meaningful utterance, there would be
some discrepancy between the speaker’s beliefs
and the speaker’s beliefs about the hearer’s be-
liefs. The discrepancy leads to an assertion, A,
which may need presupposed arguments to be un-
derstood. First, the speaker decides on the asser-
tion after checking belief discrepancies. Then, the
speaker finds the right presuppositions to be able
to communicate the assertion.
Hearer presupposition differs in that utterances
are split into presupposition and assertion, where
possible, and presuppositions are first needed to
establish links to objects in order for the new in-
formation to be understood by the hearer. For a
hearer, assertions build on presupposition and the
procedure is bottom-up (assertion is supported by
presupposition).2
Therefore, in line with linking the speaker’s
beliefs with the linguistic utterance and the lin-
guistic utterance with the hearer’s beliefs, the
speaker’s presupposition is conveyed through the
speaker’s utterance (sentence presupposition), and
the speaker’s utterance leads to the hearer’s pre-
supposition. This interaction between the seman-
tic and pragmatic notions of presupposition is a
more balanced conception of presupposition (cf.
Figure 1).
With regard to the A part of an utterance re-
ceived by the hearer, the hearer can first ‘accept’,
or ‘weakly believe’, the new information and later
on turn that weak belief into a belief, by adding it
to her belief set (Al-Raheb 2005). However, it is
worth mentioning that when making an utterance,
both the speaker and the hearer focus their atten-
tion on A, which can get accepted by the hearer.
In such a case, the hearer may later adopt A as
a belief and indicate so to the speaker, making A
a mutual belief, which may or may not serve as
a presupposition afterwards in the dialogue. It is
possible for P to be a mutual belief that both agents
in the conversation mutually know they hold, or a
2It remains for future work to test the psychological real-
ity of these hypotheses.
69
new piece of information packaged as P. From this
discussion, it can be seen that beliefs impose some
constraints on making an utterance. The follow-
ing section distinguishes between agent presuppo-
sition (speaker presupposition and hearer presup-
position) and links agents presupposition to sen-
tence presupposition.
3 Agent Presupposition
Speaker presupposition differs from hearer pre-
supposition in terms of three ‘information checks’
agents are hypothesized to perform when intro-
ducing or dealing with presupposition. The checks
are (1) clarification check, (2) informativity check,
and (3) consistency check. The checks are sim-
ilar in principle to Purver (2004) and van der
Sandt (1992). However, they are developed here
as a process which distinguishes speaker genera-
tion from hearer recognition, allowing us to dif-
ferentiate speaker presupposition from hearer pre-
supposition, hence establishing the link between
speaker presupposition and sentence presupposi-
tion, and between sentence presupposition and
hearer presupposition. The three checks apply to
both speaker and hearer.
The clarification check may be used at the be-
ginning of the process of checking. It corresponds
to Grice’s maxim of manner on the part of the
speaker (1989). Nonetheless, as there are different
kinds of clarification requests (Purver et al. 2003),
clarification can also be initiated at various stages
of the check process, indicating a different kind of
clarification.
The purpose of the informativity check is to
check whether the presupposition is new or old
information to the speaker and the hearer. This
check is a modification of Grice’s (1989) qual-
ity maxim, which has been reworked to include
two degrees of beliefs, acceptance and belief (Al-
Raheb 2005). In addition, it checks whether the
information is new or old to the other agent, based
on the beliefs of one agent about the other. The
process of checks for the speaker mirrors that of
the hearer. However, as the process of recogni-
tion is different from the process of generation, the
‘information checks’ are described for the speaker
and hearer individually.
Similarly, the consistency check determines
whether the presupposition is consistent with
the agents’ beliefs – in accordance with Grice’s
maxim of relevance (1989). For presupposition,
as part of the consistency check, another check is
performed, more specifically for the hearer’s ben-
efit, which checks whether the presupposition is
remarkable or unremarkable. Generally, informa-
tion can be accommodated, so long as it is ‘unre-
markable’ (Geurts 1999: 36). For example,
(2) The car across the street from my house
belongs to my neighbour.
is less likely to cause problems than
(3) The small jet across the street from my
house belongs to my neighbour,
when the hearer knows that the speaker lives in the
city centre.3
The process of ‘information checks’ influences
how speakers make their utterances and how hear-
ers recognize those utterances. Section 3.1 follows
the information check process for presuppositions
for the speaker, whereas section 3.2 demonstrates
that process for the hearer.
3.1 Speaker Presupposition
Speaker presupposition differs from hearer pre-
supposition in terms of checks. When a speaker
generates a sentence presupposition (via the com-
municated utterance), we are assuming that the
speaker is bound by Grice’s Cooperative Princi-
ple (1975, 1989). To utter a sentence trigger-
ing a presupposition, the speaker needs to have
reason to believe that her presupposition is go-
ing to be ‘clear’ and ‘consistent’. The speaker
may have previous context in memory that shows
her presupposition to be consistent with her be-
liefs about the hearer’s beliefs. However, when
such evidence is lacking, the speaker may still
make presupposition–triggering utterances (sen-
tence presupposition) and then make the judge-
ment that the presupposition is consistent if there
is no negative feedback; alternatively, the speaker
might receive evidence that shows the presupposi-
tion to be contradictory with her beliefs about the
hearer’s beliefs.
The informativity check comes into play when
the speaker elaborates on given or known infor-
mation by packaging it as a presupposition and fo-
cusing attention on the assertion part of her utter-
ance, i.e. on the new information. In this case, the
3Of course, anything can be ‘out of the ordinary’ or its
reverse for a specific set of circumstances. The speaker is
making assumptions about shared conceptions of the ‘ordi-
nary’.
70

	
	


	


	
		

	
 	
 
 
Figure 2: Speaker Presupposition
speaker needs to have reason to believe that the
hearer is already aware of this presupposed infor-
mation, therefore, that it is known. For example,
(4) My grandchild loves horses.
To be consistent, the speaker checks her mem-
ory to see if the speaker has record that she, the
speaker, has reason to believe that the hearer be-
lieves that the speaker has a grandchild. The
speaker, being in a retirement home, discussing
her grandchildren with the carer, and having had
previous conversations with the same carer about
her family, has reason to believe that the hearer al-
ready knows she has a grandchild. She, therefore,
presupposes ‘I have a grandchild’.
Another example of elaborating on given in-
formation is when the speaker believes the given
information has been established, i.e. both the
speaker and the hearer believe that the information
is part of their mutual or common beliefs. This
constitutes a case of strong speaker belief. Con-
sider example (5):
(5) Sylvia’s will means we have to move
out.
In this case, the speaker and the hearer have been
talking about Sylvia’s will in their dialogue and
both have reason to believe that Sylvia has a will
and that they both know the other person has rea-
son to believe Sylvia has a will.
Generally, if the speaker assumes the informa-
tion presented in the presupposition to be known
to the hearer, the speaker would expect the hearer
to accept the information provided by the sentence
presupposition by default, or even believe it. This
process is generally referred to as binding in dy-
namic semantics.4 Of course, the hearer may ex-
perience some difficulty in understanding and ask
for a ‘clarification’, check 1.
However, the information presented as a sen-
tence presupposition may be new. The speaker
may wish to introduce a topic into the dialogue,
knowing that the hearer has no previous knowl-
edge of the topic. The new information (speaker
presupposition) is then checked by the speaker for
consistency, where it may be remarkable or unre-
markable. Here, we follow Geurts’s (1999) classi-
fication of remarkable and unremarkable presup-
position.5 Thus, examples (6) and (7), given a cer-
tain situation and agents, are more unusual to ac-
commodate without questioning than example (1),
where many people may have sisters.
(6) I have to pick my personal trainer up
from the airport.
(7) I have to get the keys for my private jet.
Unremarkable information is information that
people may accept without too much questioning,
4The lack of feedback about this information is consid-
ered ‘weak positive feedback’ that the hearer has accepted
the information (Al-Raheb 2005).
5This further subclassification of presupposition is not in-
dicated in the classification of checks for reasons of clarity,
but it is incorporated in the DRT model presented in this pa-
per.
71

Figure 3: Hearer Presupposition: Check 1 (Clarification Check)
such as having a brother or a sister. An example of
remarkable information might be:
(8) My private jet arrives this afternoon.
In social contexts in which it is not expected that
everyone owns a private jet, such information will
at least raise an eyebrow. Being cooperative, the
speaker will assume, unless the hearer indicates
otherwise, that the information she provides in
the presupposition is unremarkable for the partic-
ular hearer in the particular context, and that the
hearer will accommodate the information by either
accepting it or believing it. Whether something
is remarkable depends on the specific participant
and type of communicative situation. For exam-
ple, two film stars talking together would presum-
ably not find example (8) ‘remarkable’, nor might
a journalist interviewing a celebrity.
The speaker has to be prepared for cases when,
despite being cooperative, the hearer might per-
ceive sentence presupposition as unclear and/or
contradictory. What this means for the present
treatment of presupposition is that generally the
speaker believes that the new information pre-
sented in the presupposition is unremarkable;
therefore, the speaker will expect the hearer to
accommodate the new information. However, in
case the hearer should find the new information
unusual or remarkable, the speaker will, we as-
sume, expect the hearer to check whether the pre-
supposition is consistent with her beliefs or not.
The speaker may also expect the hearer to ask for
clarification if the sentence presupposition is not
clear.
If clear, the speaker may expect the hearer to ac-
commodate the sentence presupposition and may
safely assume that the information has been ac-
cepted, unless it is indicated through ‘strong pos-
itive feedback’ that the information is actually
strongly believed (Al-Raheb 2005). However, if
the presupposition is not clear, the speaker may
expect the hearer to ask for clarification and a clar-
ification process takes place, in which the hearer
might ask for more clarification if the information
is still not clear. When the information is finally
clear, the hearer may provide feedback.
Despite the speaker’s best efforts to be coop-
erative, there are cases where the presupposition
contradicts the hearer’s previous beliefs. Speakers
usually do not expect this to happen, but are gener-
ally prepared to produce a clarification or attempt
to fix the dialogue when such a problem occurs.
Figure 2 is a flowchart displaying the speaker’s
expectations in terms of presupposition according
to her beliefs and on the assumption that she is
being cooperative. According to this treatment of
presupposition, whether the speaker believes the
information in a presupposition is new or old, the
result is the same in terms of how the speaker ex-
pects the hearer to act. The only difference is
that new information gets accommodated by the
hearer, while known information is ‘bound’ and
either already accepted or believed (Asher and
Lascarides 1998; van der Sandt and Geurts 1991).6
It has to be said that this is of course an ideal sit-
uation. The speaker does not always have beliefs
concerning whether the hearer already believes the
presupposed information or not.
To sum up, when initiating the topic of a
presupposition, we can conclude the following:
bel(S, P), bel(S, clear(P)), bel(S, consistent(P)),
and bel(S, accept(H,P)).7 The speaker may have
either the belief bel(S,¬ bel(H,P)) or the belief
bel(S,bel(H,P)). However, in our implementation
of the pragmatic and semantic notions of presup-
position in DRT, the only beliefs represented after
S’s utterance are: bel(S, P), or accept(S,P) and if
6Discourse referents of known presuppositions attach
themselves to previous discourse markers referring to the
same object or person.
7S stands for the speaker, H stands for the hearer and bel
stands for believes.
72

		

 




 

 	 

		



Figure 4: Hearer Presupposition: Check 2 (Informativity Check: Old)
sufficient previous information is available, either
bel(S,¬ bel(H,P)) or bel(S,bel(H,P)).
3.2 Hearer Presupposition
As a result of the speaker’s initiating the topic of
P indicated through sentence presupposition, the
hearer acquires the belief that the speaker believes
the presupposition. The first information check
to apply to hearer presupposition is the clear/
not clear check. That is to say, upon hearing
P, the hearer first checks whether the presuppo-
sition is clear (e.g. hearer has no problems with
perception). As mentioned previously, there are
other types of clarification requested when incon-
sistency arises. However, what we are concerned
with here is whether the hearer has been able to
receive the message or not. Other clarification
checks may take place after the hearer performs
the new/old (informativity) check and the consis-
tency check.
If the presupposition is not clear in the above
sense, the hearer may ask the speaker to clarify
her statement. As a simple example, consider an
imaginary dialogue between a customer and cus-
tomer service assistant about a gas heater:
(9) Customer: How long does it take to fix
my gas heater?
Customer Service Assistant: Your
what?
Customer: My gas heater needs fixing.
After checking whether the information, sentence
presupposition, is clear, the customer service as-
sistant asks the customer to clarify. In this partic-
ular case, the lack of clarity may be attributed to,
e.g. not hearing very well. The hearer expects the
speaker to provide an explanation or clarification.
The speaker is then obliged to provide a further
explanation. If the information is still not clear,
the hearer may ask for more clarification and the
hearer needs to provide an explanation. Figure 3
incorporates this potentially iterative loop. This
is consistent with conversation analysis research,
which assumes that information may be cleared up
after an explanation is provided, but also allows
for further clarification if needed (Schegloff et al.
1977). If the sentence presupposition is cleared
up, then the hearer may provide feedback that the
information is clear. However, lack of feedback
is also considered a case of ‘weak positive feed-
back’ (Al-Raheb 2005). Generally, after provid-
ing an explanation, the speaker’s assumptions are
likely to be that the hearer now has no problems
with the sentence presupposition.
Having made sure that the sentence presupposi-
tion is clear, the hearer may now move on to per-
form the informativity check, check 2. If the in-
formation the speaker presents as a sentence pre-
supposition is known to the hearer, in the sense of
being in his acceptance space, i.e. already a hearer
presupposition, the hearer may strengthen that ac-
ceptance by now believing the presupposition (cf.
Al-Raheb 2005). The hearer may previously hold
a strong belief about the presupposition, i.e. the
hearer may already believe P. In this case, it is
not necessary to add a new belief that P to hearer
presupposition. We are assuming here that the
hearer’s knowledge of a sentence presupposition
means that this presupposition does not contradict
previous beliefs held by the hearer. Figure 4 shows
the hearer’s options if the sentence presupposition
is already a hearer presupposition, or known to
him. The speaker ideally expects the hearer will
accept P (i.e. P will have become hearer presup-
position), unless negative feedback is provided by
the hearer. If strong positive feedback is provided
by the hearer, the speaker may thereby form the
meta-belief that bel(H,bel(S,bel(H,P))).
If the sentence presupposition provides new in-
formation, the hearer then performs the consis-
73

		

 


 




















 
		

!


 "
	

Figure 5: Hearer Presupposition: Check 3 (Consistency and Informativity Check: New)
tency check. If the information provided contra-
dicts previous beliefs, the hearer may reject the
sentence presupposition and an attempt would be
made to remedy or fix the dialogue. For example,
consider:
(10) Speaker: Julia’s husband is coming for
dinner.
Hearer: This can’t be! Julia is widowed!
When the information presented by the sentence
presupposition is consistent with the hearer’s be-
lief space or acceptance space, the hearer makes
a judgement about whether the information is re-
markable (odd or unusual) or unremarkable (cf.
Geurts 1999). If the information is unremarkable,
the hearer accommodates the new information by
either accepting it or believing it. In other words,
it becomes hearer presupposition. Figure 5 shows
presupposition processing from the hearer’s per-
spective when the sentence presupposition con-
tains information new to the hearer.
If the presupposition is remarkable, the hearer
may check for clarity. This is a different type
of clarity check from the one performed initially.
Clarification checks can arise from different rea-
sons and not just because of difficulty in hearing.
This time the hearer requires an explanation for the
oddness of the information used as a presupposi-
tion. This is when the clarification process starts
again. For example,
(11) Speaker: My pet lion requires a lot of
attention.
Hearer: Your pet what?
Speaker: Oh sorry, I mean one of those
virtual pets you take care of.
Here, we may assume, the hearer has not found
the appropriate discourse referent for ‘pet lion’
and thus goes through ‘remarkable’ check after the
consistency check.8
Again, the hearer may provide feedback con-
cerning whether the explanation has been ac-
cepted or not. Unless negative feedback is pro-
vided, the hearer is expected to at least accept
the presupposition bel(S,accept(H,P)). In addition,
the other agent (speaker) may assume that the
hearer now has no problem with the presupposi-
tion, bel(S,clear(H,P)). If the hearer is not con-
vinced by the speaker’s explanation, an attempt at
repairing the dialogue is needed.9
To sum up, at the stage of the hearer’s receiv-
ing the speaker’s utterance, the hearer may make
the judgement that the speaker believes the pre-
supposition (speaker presupposition). In addition,
unless the hearer gives the speaker reason to think
that the hearer disagrees with the presupposition,
the speaker assumes that the hearer has no prob-
lem understanding the sentence presupposition,
and further, that the hearer has now come to ac-
cept the presupposition (hearer presupposition). It
must be pointed out that generally speaking, un-
less the speaker has introduced as her presupposi-
tion a topic perceived to be new and very unusual,
the hearer does not need to go through the clarifi-
cation process for each presupposition, since gen-
erally the presuppositions are not the focus of the
speaker’s utterance (Levinson 1983).
8This example raises a lot of interesting cognitive and
pragmatic issues, which will be ignored here so as not to dis-
tract from the main focus of this argument.
9Fixing a dialogue process is not addressed here. It is
assumed that if fixing the dialogue is successful, the agent
will then continue with the consistency check in order to carry
on with the dialogue, unless one of the agents simply gives
up.
74
4 Conclusion
This paper has enhanced the pragmatic conception
of presupposition in DRT by considering presup-
position from both the speaker’s and the hearer’s
point of view. It has also linked semantic presup-
position with pragmatic presupposition through
linking the speaker’s presupposition with the pre-
supposition communicated by her utterance (sen-
tence presupposition) on the one hand, and sen-
tence presupposition with hearer presupposition
on the other hand. This it was argued is helped
by the information checks which both the speaker
and the hearer perform.

References
Al-Raheb, Y. 2005. Speaker/Hearer Representation in
a Discourse Representation Theory Model of Pre-
supposition: A Computational-Linguistic Approach.
Phd. University of East Anglia.
Asher, N. and Lascarides, A. 1998b. ‘The Semantics
and Pragmatics of Presupposition’. Journal of Se-
mantics 15, pp. 239–299.
Geurts, B. 1996. ‘Local Satisfaction Guaranteed: A
Presupposition Theory and Its Problems’. Linguis-
tics and Philosophy 19, pp. 259–294.
Geurts, B. 1998. ‘Presupposition and Anaphors in At-
titude Contexts’. Linguistics and Philosophy 21, pp.
545–601.
Geurts, B. 1999. Presuppositions and Pronouns: Cur-
rent Research in the Semantics/ Pragmatics Inter-
face. Oxford: Elsevier.
Grice, P. 1975. ‘Logic and Conversation (from
the William James Lectures, Harvard University,
1967)’. In: P. Cole and J. Morgan (Eds.). Syntax
and Semantics 3: Speech Acts. pp. 41-58. New York:
Academic Press.
Grice, P. 1989. Studies in the Way of Words. Cam-
bridge, MA: Harvard University Press.
Kamp, H. 1984. ‘A Theory of Truth and Semantic Rep-
resentation’. In: J. Groenendijk, T. Janseen, and
M. Stokhof (Eds.). Truth, Interpretation and Infor-
mation: Selected Papers from the Third Amsterdam
Colloquium. Amsterdam: Foris Publications.
Kamp, H. 1988. ‘Comments on Stalnaker’. In: R.
Grimm and D. Merrill (Eds.). Contents of Thought.
Proceedings of the 1985 Oberlin Colloquium in Phi-
losophy. pp. 156–181. Tucson State: The University
of Arizona Press.
Kamp, H. 1990. ‘Prolegomena to a Structural Account
of Belief and Other Attitudes’. In: C. Anderson and
J. Owens (Eds.). Propositional Attitudes: The Role
of Content in Logic, Language, and Mind. Stanford,
CA: CSLI Publications.
Kamp, H. 1995. ‘Discourse Representation Theory’.
In: J. Verschueren, J. Ostman, and J. Bloomaert
(Eds.). Handbook of Pragmatics Manual. Amster-
dam: John Benjamin.
Kamp, H. 2001a. ‘Computation and Justification of
Presuppositions: One Aspect of the Interpretation
of Multi-Sentence Discourse’. In: M. Bras and L.
Vieu (Eds.). Semantics and Pragmatics of Discourse
and Dialogue: Experimenting with Current Theo-
ries. pp. 57–84. Oxford: Elsevier Science.
Kamp, H. 2001b. ‘The Importance of Presupposition’.
In: C. Rohrer and A. Rossdeutscher (Eds.). Linguis-
tic Form and its Computation: Selected Papers from
the SFB 340. pp. 207-254. Stanford: CSLI.
Kamp, H. and Reyle, U. 1993. From Discourse to
Logic: Introduction to Model Theoretic Semantics
of Natural Language, Formal Logic and Discourse
Representation Theory. Boston, Dordrecht: Kluwer.
Kamp, H., van Genabith, J., and Reyle, U. 2005.
The Handbook of Logic. Unpublished Manuscript.
http://www.ims.uni-stuttgart.de/˜hans/.
Lambrecht, K. 1994. Information Structure and Sen-
tence Form: Topic, Focus and the Mental Represen-
tations of Discourse Referents. Cambridge: Cam-
bridge University Press.
Levinson, S. 1983. Pragmatics. Cambridge: Cam-
bridge University Press.
Purver, M., Healey, P., King, J., Ginzburg, J., and
Mills, G. 2003. ‘Answering Clarification Ques-
tions’. In: Proceedings of the 4th SIGdial Workshop
on Discourse and Dialogue, Association for Compu-
tational Linguistics. pp. 23–33. Sapporo.
Purver, M. 2004. ‘Claire: the Clarification Engine’.
In: J. Ginzburg and E. Vallduvi (Eds.). Proceedings
of the Eighth Workshop on the Semantics and Prag-
matics of Dialogue. Catalog ‘04. pp. 77-84. Univer-
sitat Pompeu Fabra: Barcelona.
Schegloff, E., Jefferson, G., and Sacks, H. 1977. ‘The
Preference for Self-Correction in the Organization
of Repair in Conversation’. Language 53, pp. 361–
382.
Stalnaker, R. 2002. ‘Common ground’. Linguistics
and Philosophy 25(5-6), pp. 701–721.
van der Sandt, R. and Geurts, B. 1991. ‘Presupposi-
tion, Anaphora, and Lexical Content’. In: O. Her-
zog and C.-R. Rollinger (Eds.). Text Understand-
ing in LILOG. pp. 259–296. Berlin, Heidelberg:
Springer Verlag.
van der Sandt, R. 1992. ‘Presupposition Projection as
Anaphora Resolution’. Journal of Semantics 9, pp.
333–377.
