CONVERSATIONAL IMPLICATURES IN INDIRECT REPLIES 
Nancy Green 
Sandra Carberry 
Department of Computer and Information Sciences 
University of Delaware 
Newark, Delaware 19716, USA 
email: green@cis.udel.edu, carberry~cis.udel.edu 
Abstract I 
In this paper we present algorithms for the 
interpretation and generation of a kind of particu- 
larized conversational implicature occurring in cer- 
tain indirect replies. Our algorithms make use of 
discourse expectations, discourse plans, and dis- 
course relations. The algorithms calculate implica- 
tures of discourse units of one or more sentences. 
Our approach has several advantages. First, by 
taking discourse relations into account, it can cap- 
ture a variety of implicatures not handled before. 
Second, by treating implicatures of discourse units 
which may consist of more than one sentence, it 
avoids the limitations of a sentence-at-a-time ap- 
proach. Third, by making use of properties of dis- 
course which have been used in models of other dis- 
course phenomena, our approach can be integrated 
with those models. Also, our model permits the 
same information to be used both in interpretation 
and generation. 
1 Introduction 
In this paper we present algorithms for the 
interpretation and generation of a certain kind of 
conversational implicature occurring in the follow- 
ing type of conversational exchange. One partici- 
pant (Q) makes an illocutionary-level request 2 to 
be informed if p; the addressee (A), whose reply 
may consist of more than one sentence, conversa- 
tionally implicates one of these replies: p, "-p, that 
there is support for p, or that there is support for 
"-p. For example, in (1), assuming Q's utterance 
has been interpreted as a request to be informed if 
A went shopping, and given certain mutual beliefs 
e.g., that A's car breaking down would normally 
e sufficient to prevent A from going shopping, and 
i We wish to thank Kathy McCoy for her comments on 
this paper. 
~i.e., using Austin's (Austin, 1962) distinction between 
locutionary and il\]ocutionary force, Q's utterance is intended 
to function as a request (although it need not have the gram- 
matical form of a question) 
that A's reply is coherent and cooperative), A's re- 
ply is intended to convey, in part, a 'no'. 
(1) Q: Did you go shopping? 
A: a. My car~s not running. 
b. The timing belt broke. 
Such indirect replies satisfy the conditions 
proposed by Grice and others (Grice, 1975; 
Hirschberg, 1985; Sadock, 1978) for being classi- 
fied as particularized conversational implicatures. 
First, A's reply does not entail (in virtue of its 
conventional meaning) that A did not go shopping. 
Second, the putative implicature can be cancelled; 
for example, it can be denied without the result 
sounding inconsistent, as can be seen by consider- 
ing the addition of (2) to the end of A's reply in 
(1.) 
(2) A: So I took the bus to the mall. 
Third, it is reinforceable; A's reply in (1) could have 
been preceded by an explicit "no" without destroy- 
ing coherency or sounding redundant. Fourth, the 
putative implicature is nondetachable; the same re- 
ply would have been conveyed by an alternative re- 
alization of (la) and (lb) (assuming that the al- 
ternative did not convey a Manner-based implica- 
ture). Fifth, Q and A must mutually believe that, 
given the assumption that A's reply is cooperative, 
and given certain shared background information, 
Q can and will infer that by A's reply, A meant 'no'. 
This paper presents algorithms for calculating such 
an inference from an indirect response and for gen- 
erating an indirect response intended to carry such 
an inference. 
2 Solution 
2.1 Overview 
Our algorithms are based upon three notions 
from discourse research: discourse expectations, 
discourse plans, and implicit relational propositions 
in discourse. 
B4 
At certain points in a coherent conversa- 
tion, the participants share certain expectations 
(Reichman, 1984; Carberry, 1990) about what kind 
of utterance is appropriate. In the type of exchange 
we are studying, at the point after Q's contribu- 
tion, the participants share the beliefs that Q has 
requested to be informed if p and that the request 
was appropriate; hence, they share the discourse 
expectation that for A to be cooperative, he must 
now say as much as he can truthfully say in regard 
to the truth of p. (For convenience, we shall refer 
to this expectation as Answer-YNQ(p).) 
A discourse plan operator 3 (Lambert & Car- 
berry, 1991) is a representation of a normal or con- 
ventional way of accomplishing certain communica- 
tive goals. Alternatively, a discourse plan operator 
could be considered as a defeasihle rule expressing 
the typical (intended) effect(s) of a sequence of illo- 
cutionary acts in a context in which certain appli- 
cability conditions hold. These discourse plan op- 
erators are mutually known by the conversational 
participants, and can be used by a speaker to con- 
struct a plan for achieving his communicative goals. 
We provide a set of discourse plan operators which 
can be used by A as part of a plan for fulfilling 
Answer-YNQ(p). 
Mann and Thompson (Mann ~z Thompson, 
1983; Mann & Thompson, 1987) have described 
how the structure of a written text can be analyzed 
in terms of certain implicit relational propositions 
that may plausibly be attributed to the writer to 
preserve the assumption of textual coherency. 4 The 
role of discourse relations in our approach is moti- 
vated by the observation that direct replies may 
occur as part of a discourse unit conveying a rela- 
tional proposition. For example, in (3), (b) is pro- 
vided as the (most salient) obstacle to the action 
(going shopping) denied by (a); 
(3) Q: Did you go shopping? 
A:a. No. 
b. my car~s not running. 
in (4), as an elaboration of the action (going shop- 
ping) conveyed by (a); 
(4) Q: Did you go shopping? 
A:a. Yes, 
b. I bought some shoes. 
and in (5), as a concession for failing to do the 
action (washing the dishes) denied by (a). 
(S) Q: Did you wash the dishes? 
A:a. No, 
b. (but) I scraped them. 
3in Pollack's terminology, a recipe-for-action (Pollack, 
1988; Grosz & Sidner, 1988)~ 
4Although they did not study dialogue, they suggested 
that it can be analyzed similarly. Also note that the rela- 
tional predicates which we define are similar but not neces- 
sarily identical to theirs. 
Note that given appropriate context, the (b) replies 
in (3) through (5)would be sufficient to conversa- 
tionally implicate the corresponding direct replies. 
This, we claim, is by virtue of the recognition of 
the relational proposition that would be conveyed 
by use of the direct reply and the (b) sentences. 
Our strategy, then, is to generate/interpret 
A's contribution using a set of discourse plan oper- 
ators having the following properties: (1) if the ap- 
plicability conditions hold, then executing the body 
would generate a sequence of utterances intended to 
implicitly convey a relational proposition R(p, q); 
(2) the applicability conditions include the condi- 
tion that R(p, q) is plausible in the discourse con- 
text; (3) one of the goals is that Q believe that p, 
where p is the content of the direct reply; and (4) 
the step of the body which realizes the direct re- 
ply can be omitted under certain conditions. Thus, 
whenever the direct reply is omitted, it is neverthe- 
less implicated as long as the intended relational 
proposition can be recognized. Note that prop- 
erty (2) requires a judgment that some relational 
proposition is plausible. Such judgments will be de- 
scribed using defeasible inference rules. The next 
section describes our discourse relation inference 
rules and discourse plan operators. 
2.2 Discourse Plan Opera- 
tors and Discourse Relation In- 
ference Rules 
A typical reason for the failure of an agent's 
attempt to achieve a domain goal is that the agent's 
domain plan encountered an obstacle. Thus, we 
give the rule in (6) for inferring a plausible discourse 
relation of Obstacle. s 
(8) 
If (i) coherently-relatedCA,B), and 
(ii) A is a proposition that an agent 
failed to perform an action of 
act type T, and 
(iii) B is a proposition that 
a) a normal applicability condition 
of T did not hold, or 
b) a normal precondition of T 
failed, or 
c) a normal step of T failed, or 
d) the agent did not want to 
achieve a normal goal of T, 
then plausible(Obstacle(B,A)). 
In (6) and in the rules to follow, 'coherently- 
related(A,B)' means that the propositions A and B 
are assumed to be coherently related in the dis- 
course. The terminology in clause (iii) is that of 
the extended STRIPS planning formalism (Fikes 
5For simplicity of exposition, (6) and the discourse rela- 
tion inference rules to follow are stated in terms of the past; 
we plan to extend their coverage of times. 
65 
& Nilsson, 1971; Allen, 1979; Carberry, 1990; Lit- 
man & Allen, 1987). 
Examples of A and B satisfying each of the 
conditions in (6.iii) are given in (7a) - (7d), respec- 
tively. 
(7) \[A\]I didn't go shopping. 
a. \[B\] The stores were closed. 
b. \[B\] My car wasn't run-ing. 
c. \[B\] My car broke doen on the way. 
d. \[B\] I didn't want to buy anything. 
The discourse plan operator given in (8) de- 
scribes a standard way of performing a denial (ex- 
emplified in (3)) that uses the discourse relation of 
Obstacle given in (6). In (8), as in (6), A is a propo- 
sition that an action of type T was not performed. 
(8) Deny (with Obstacle) 
Applicability conditions: 
1) S BMB plausible(Obstacle(B,A)) 
Bo~ (unordered) : 
(optional) S inform H that A 
2) TelI(S,H,B) 
Goals: 
1) H believe that A 
2) H believe that Obstacle(B,A) 
In (8) (and in the discourse plan operators 
to follow) the" formalism described above is used; 
'S' and 'H' denote speaker and hearer, respectively; 
'BMB' is the one-sided mutual belief s operator 
(Clark & Marshall, 1981); 'inform' denotes an il- 
locutionary act of informing; 'believe' is Hintikka's 
(Hintikka, 1962) belief operator; 'TelI(S,H,B)' is a 
subgoal that can be achieved in a number of ways 
(to be discussed shortly), including just by S in- 
forming H that B; and steps of the body are not 
ordered. (Note that to use these operators for gen- 
eration of direct replies, we must provide a method 
to determine a suitable ordering of the steps. Also, 
although it is sufficient for interpretation to spec- 
ify that step 1 is optional, for generation, more in- 
formation is required to decide whether it can or 
should be omitted; e.g., it should not be omitted if 
S believes that H might believe that some relation 
besides Obstacle is plausible in the context. 7 These 
are areas which we are currently investigating; for 
related research, see section 3.) 
Next, consider that a speaker may wish to 
inform the hearer of an aspect of the plan by which 
she accomplished a goal, if she believes that H may 
not be aware of that aspect. Thus, we give the rule 
in (9) for inferring a plausible discourse relation of 
Elaboration. 
e'S BMB p' is to be read as 'S believes that it is mutually 
believed between S and H that p'. 
ZA related question, which has been studied by oth- 
ers (Joshi, Webber ~ Weischedel, 1984a; Joshi, Webber 
& Weischedel, 1984b), is in what situations is a speaker re- 
quired to supply step 2 to avoid misleading the hearer? 
(9) 
If (i) 
(ii) 
coherently-related(A,B), and 
A is a proposition that an agent 
performed some action of act type 
T, and 
(iii) B is a proposition that describes 
information believed to be new to 
H about 
a) the satisfaction of a normal 
applicability condition of T such 
that its satisfaction is not 
believed likely by H, or 
b) the satisfaction of a normal 
precondition of T such that its 
satisfaction is not believed 
likely by H, or 
c) the success of a normal step of T, 
or 
d) the achievement of a normal goal 
of T, 
then plausible(Elaboration(B,A)). 
Examples of A and B satisfying each of the 
conditions in (9.iii) are given in (10a) - (10d), re- 
spectively. 
(I0) \[A\]I went shopping today. 
a. \[B\] I found a store that was open. 
b. \[B\] I got my car fixed yesterday. 
c. \[B\] I went to Macy's. 
d. \[B\] I got running shoes. 
The discourse plan operator given in (11) de- 
scribes a standard way of performing an affirmation 
(exemplified in (4)) that uses the discourse relation 
of Elaboration. 
(11) Affirm (with Elaboration) 
Applicability conditions: 
1) S BMB plausible(Elaboration(B,A)) 
Body (unordered): 
1) (optional) S inform H that A 
2) TelI(S,H,B) 
Goals: 
1) H believe that A 
2) H believe that Elaboration(B,A) 
Finally, note that a speaker may concede a 
failure to achieve a certain goal while seeking credit 
for the partial success of a plan to achieve that goal. 
For example, the \[B\] utterances in (10) can be used 
following (12) (or aIone, in the right context) to 
concede failure. 
(12) \[A\]I didn't go shopping today, but 
Thus, the rule we give in (13)for inferring a 
plausible discourse relation of Concession is similar 
(but not identical) to (9). 
(13) 
If (i) 
(ii) 
coherently-related(A,B), and 
A is a proposition that an agent 
failed to do an action of act 
type T, and 
(iii) B is a proposition that describes 
a) the satisfaction of a normal 
applicability condition of T, or 
b) the satisfaction of a normal 
precondition of T, or 
c) the success of a normal step of T, 
or 
d) the achievement of a normal goal 
of T, and 
(iv) the achievement of the plan's 
component in B may bring credit 
to the agent, 
then plausible(Concession(B,A)). 
A discourse plan operator, Deny (with Con- 
cession), can be given to describe another standard 
way of performing a denial (exemplified in (5)). 
This operator is similar to the one given in (8), 
except with Concession in the place of Obstacle. 
An interesting implication of the discourse 
plan operators for Affirm (with Elaboration) and 
Deny (with Concession) is that, in cases where the 
speaker chooses not to perform the optional step 
(i.e., chooses to omit the direct reply), it requires 
that the intended discourse relation be inferred in 
order to correctly interpret the indirect reply, since 
either an affirmation or denial could be realized 
with the same utterance. (Although (9) and (13) 
contain some features that differentiate Elaboration 
and Concession, other factors, such as intonation, 
will be considered in future research.) 
The next two discourse relations (described 
in (14) and (16)) may be part of plan operators for 
conveying a 'yes' similar to Affirm (with Elabora- 
tion). 
(14) 
If (i) coherently-related(A,B), and 
(ii) A is a proposition that an agent 
performed an action X, and 
(iii) B is a proposition that normally 
implies that the agent has a goal 
G, and 
(iv) X is a type of action occurring 
as a normal part of a plan to 
achieve G, 
then plausible( 
Motivate-Volitional-Action(B,A)). 
15) shows tile use of Motivate-Volitional-Action 
MVA) in an indirect (affirmative) reply. 
(15) Q: Did you close the window? 
A: I was cold. 
(16) 
If (i) coherently-related(A,B), and 
(ii) A is a proposition that an event E 
occurred, and 
(iii) B is a proposition that an event F 
occurred, and 
(iv) it is not believed that F followed 
E, and 
(v) F-type events normally cause 
E-type events, 
then plausible(Cause-Non-Volitional(B,A)). 
/17) shows the use of Cause-Non-Volitional (CNV) 
m an indirect (affirmative) reply. 
(17) Q: Did you wake up very early? 
A: The neighbor's dog was barking. 
The discourse relation described in (18) may 
be part of a plan operator similar to Deny (with 
Obstacle) for conveying a 'no'. 
(18) 
If (i) coherently-related(A,B), and 
(ii) A is a proposition that an event E 
did not occur, and 
(iii) B is a proposition that an action F 
was performed, and 
(iv) F-type actions are normally 
performed as a way of preventing 
E-type events, 
then plausible(Prevent(B,A)). 
(19) showsthe use of Preventin an indirect denial. 
(19) Q: Did you catch the flu? 
A: I got a flu shot. 
The discourse relation described in (20) can 
be part of a plan operator similar to the others de- 
scribed above except that one of the speaker's goals 
is, rather than affirming or denying p, to provide 
support for the belief that p. 
(20) 
If (i) coherently-related(A,B), and 
(ii) B is a proposition that describes 
a typical result of the situation 
described in proposition A, 
then plausible(Evidence(B,A)). 
67 
Assuming an appropriate context, (21) is an 
.example of use of this relation to convey support, 
Le., to convey that it is likely that someone is home. 
(21) Q: Is anyone home? 
A: The upstairs lights are on. 
A similar rule could be defined for a relation used 
to convey support against a belief. 
2.3 Implicatures of Discourse Units 
Consider the similar dialogues in (22) and 
(23). 
(22) Q: Did you go shopping? 
A:a. I had to take the bus. 
b. (because) My car's not running. 
c. (You see,) The timing belt broke. 
(23) Q: Did you go shopping? 
A:a. My car's not running. 
b. The timing belt broke. 
c. (So) I had to take the bus. 
First, note that although the order of the sentences 
realizing A's reply varies in (22) and (23), A's over- 
all discourse purpose in both is to convey a 'yes'. 
Second, note that it is necessary to have a rule so 
that if A's reply consists solely of (22a) (=23c), an 
implicated 'yes' is derived; and if It consists solely 
of (22b) (=23a), an implicated 'no'. 
In existing sentence-at-a-time models of cal- 
culating implicatures (Gazdar, 1979; Hirschberg, 
1985), processing (22a) would result in an impli- 
cated 'yes' being added to the context, which would 
successfully block the addition of an implicated 'no' 
on processing (22b). However, processing (23a) 
would result in a putatively implicated 'no" be- 
in S added to the context (incorrectly attributing 
a fleeting intention of A to convey a 'no'); then, on 
processing (23c) the conflicting but intended 'yes' 
would be blocked by context, giving an incorrect 
result. Thus, a sentence-at-a-time model must pre- 
dict when (23c) should override (23a). Also, in that 
model, processing (23) requires "extra effort", a 
nonmonotonic revision of belief not needed to han- 
dle (22); yet (23) seems more like (22) than a case 
in which a speaker actually changes her mind. 
In our model, since implicatures correspond 
to goals of inferred or constructed hierarchical 
plans, we avoid this problem. (22A) and (23A) both 
correspond to step 2 of Affirm (with Elaboration), 
TelI(S,H,B); several different discourse plan opera- 
tors can be used to construct a plan for this Tell 
action. For example, one operator for Tell(S,H,B) 
is given below in (24); the operator represents that 
in telling H that B, where B describes an agent's 
volitional action, a speaker may provide motivation 
for the agent's action. 
(24) Tell(S,H,p) 
Applicability Conditions: 
1) S BMB plausible( 
Motivate-Volitional-Action(q,p)) 
Body (unordered): 
1) Tell(S,H,q) 
2) S inform H that p 
Goals: 
I) H believe that p 
2) H believe that 
Motivate-Volitional-Action(q,p) 
(We are currently investigating, in generation, 
when to use an operator such as (24). For ex- 
ample, a speaker might want to use (24) in case 
he thinks that the hearer might doubt the truth 
of B unless he knows of the motivation.) Thus, 
(22a)/(23c) corresponds to step 2 of (24); (22b) - 
(22c), as well as (23a) - (23b), correspond to step 
1. Another operator for Tell(S,H,p) could represent 
that in telling H that p, a speaker may provide the 
cause of an event; i.e., the operator would be like 
(24) but with Cause-Non-Volitional as the discourse 
relation. This operator could be used to decom- 
pose (22b)- (22c)/(23a)- (23b). The structure pro- 
posed for (22A)/(23A) is illustrated in Figure 1. s 
Linear precedence in the tree does not necessarily 
represent narrative order; one way of ordering the 
two nodes directly dominated by TeII(MVA) gives 
(22A), another gives (23A). (Narrative order in the 
generation of indirect replies is an area we are cur- 
rently investigating also; for related research, see 
section 3.) 
Note that Deny (with Obstacle) can not be 
used to generate/interpret (22A) or (23A) since 
its body can not be expanded to account for 
b22a)/(23c). Thus, the correct implicatures can 
e derived without attributing spurious intentions 
to A, and without requiring cancellation of spurious 
implicatures. 
8To use the terminology of (Moore & Paris, 1989; Moore 
& Paris, 1988), the labelled arcs represent satellites, and the 
unlabelled arcs nucleii. However, note that in their model, a 
nucleus can not be optional. This differs from our approach, 
in that we have shown that direct replies are optional in 
contexts such as those described by plan operators such as 
Affirm (with Elaboration). 
9Determiuing this requires that the end of the relevant 
discourse unit be marked/recognlzed by cue phrases, into- 
nation, or shift of focus; we plan to investigate this problem. 
68 
Affirm (with Elaboration) 
I 
I went shopping 
(Motivate-Volitional-Action) 
J 
My car's not running 
I 
Tell (CNV) 
I 
(Elaboration) 
Tell (MVA) 
I I 
I bad to take the bus 
(Cause-Non-Volitional) 
The timing belt broke 
Figure 1. A Sample Discourse Structure 
2.4 Algorithms 
Generation and interpretation algorithms 
are given in (25) and (26), respectively. They 
presuppose that the plausible discourse relation is 
available. 1° The generation algorithm assumes as 
given an illocutionary-level representation of A's 
communicative goals.ll 
(25) Generation of indirect reply: 
I. Select discourse plan operator: Select 
from the Ans,er-YHQ(p) plan operators 
all those for ,hich 
a) the applicability conditions hold, 
and 
b) the goals include S's goals. 
2. If more than one operator was selected 
in step I, then choose one. Also, 
determine step ordering and whether it 
is necessary to include optional steps. 
(We are currently investigating how 
these choices are determined.) 
3. Construct a plan from the chosen 
operator and execute it. 
1°We plan to implement an inference mechanism for the 
discourse relation inference rules. 
11 Note that A's goals depend, in part, on the illocutionary- 
level representation of Q's request. We assume that an 
analysis, such as provided in (Perrault & Allen, 1980), is 
available. 
(26) Interpretation of indirect reply: 
I. Infer discourse plan: Select from the 
Ansver-YNQ(p) plan operators all those 
for ,hich 
a) the second step of the body matches 
S's contribution, and 
b) the applicability conditions hold, 
and 
¢) it is mutually believed that the 
goals are consistent with S's goals. 
2. If more than one operator was selected 
in step I, then choose one. (We are 
currently investigatin E what factors 
are involved in this choice. Of course, 
the utterance may be ambiguous.) 
3. Ascribe to S the goal(s) of the chosen 
plan operator. 
3 Comparison to Past Re- 
search 
Most previous work in computational or for- 
mal linguistics on particularized conversational im- 
plicature (Green, 1990; Horacek, 1991; Joshi, 
Webber & Weischedel, 1984a; .\]oshi, Webber 
Weischedel, 1984b; Reiter, 1990; Whiner & Maida, 
1991) has treated other kinds of implicature than 
we consider here. ttirschberg (Hirschberg, 1985) 
provided licensing rules making use of mutual be- 
liefs about salient partial orderings of entities in 
69 
the discourse context to calculate the scalar im- 
plicatures of an utterance. Our model is similar 
to Hirschberg's in that both rely on the represen- 
tation of aspects of context to generate implica- 
tures, and our discourse plan operators are roughly 
analogous in function to her licensing rules. How- 
ever, her model makes no use of discourse relations. 
Therefore, it does not handle several kinds of indi- 
rect replies which we treat. For example, although 
A in (27) could be analyzed as scalar implicating 
a 'no' in some contexts, Hirschberg's model could 
not account for the use of A in other contexts as an 
elaboration (of how A managed to read chapter 1) 
intended to convey a 'yes'. 12 
(27) Q: Did you read the first chapter? 
A: I took it to the beach with me. 
Furthermore, Hirschberg provided no computa- 
tional method for determining the salient partially 
ordered set in a context. Also, in her model, impli- 
catures are calculated one sentence at a time, which 
has the potential problems described above. 
Lascarides, Asher, and Oberlander 
(Lascarides & Asher, 1991; Lascarides & Oberlan- 
der, 1992) described the interpretation and gen- 
eration of temporal implicatures. Although that 
type of implicature (being Manner-based) is some- 
what different from what we are studying, we have 
adopted their technique of providing defeasible in- 
ference rules for inferring discourse relations. 
In philosophy, Thomason (Thomason, 1990) 
suggested that discourse expectations play a role in 
some implicatures. McCafferty (McCafferty, 1987) 
argued that interpreting certain implicated replies 
requires domain plan reconstruction. However, he 
did not provide a computational method for inter- 
preting implicatures. Also, his proposed technique 
can not handle many types of indirect replies. For 
example, it can not account for the implicated nega- 
tive replies in (1) and (5), since their interpretation 
involves reconstructing domain plans that were not 
executed successfully; it can not account for the im- 
plicated affirmative reply in (17), in which no rea- 
soning about domain plans is involved; and it can 
not account for implicated replies conveying sup- 
port for or against a belief, as in (21). Lastly, his 
approach cannot handle implicatures conveyed by 
discourse units containing more than one sentence. 
Finally, note that our approach of including 
rhetorical goals in discourse plans is modelled on 
the work of Hovy (Hovy, 1988) and Moore and 
Paris (Moore & Paris, 1989; Moore & Paris, 1988), 
who used rhetorical plans to generate coherent text. 
12 The two intended interpretations are marked by different 
intonations. 
4 Conclusions 
We have provided algorithms for the inter- 
pretation/generation of a type of reply involving 
a highly context-dependent conversational implica- 
ture. Our algorithms make use of discourse ex- 
pectations, discourse plans, and discourse relations. 
The algorithms calculate implicatures of discourse 
units of one or more sentences. Our approach has 
several advantages. First, by taking discourse rela- 
tions into account, it can capture a variety of im- 
plicatures not handled before. Second, by treating 
implicatures of discourse units which may consist of 
more than one sentence, it avoids the limitations of 
a sentence-at-a-time approach. Third, by making 
use of properties of discourse which have been used 
in models of other discourse phenomena, our ap- 
proach can be integrated with those models. Also, 
our model permits the same information to be used 
both in interpretation and in generation. 
Our current and anticipated research in- 
cludes: refining and implementing our algorithms 
(including developing an inference mechanism for 
the discourse relation rules); extending our model 
to other types of implicatures; and investigating the 
integration of our model into general interpretation 
and generation frameworks. 

References 
Allen, James F. (1979). A Plan-Based Approach to 
Speech Act Recognition. PhD thesis, University 
of Toronto, Toronto, Ontario, Canada. 
Austin, J. L. (1962). How To Do Things With 
Words. Cambridge, Massachusetts: Harvard 
University Press. 
Carberry, Sandra (1990). Plan Recognition in Nat- 
ural Language Dialogue. Cambridge, Mas- 
sachusetts: MIT Press. 
Clark, H. & Marshall, C. (1981). Definite refer- 
ence and mutual knowledge. In A. K. Joshi, 
B. Webber, & I. Sag (Eds.), Elements of dis- 
course understanding. Cambridge: Cambridge 
University Press. 
Fikes, R. E. & Nilsson, N. J. (1971). Strips: A new 
approach to the application of theorem proving 
to problem solving. Artificial Intelligence, 2, 
189-208. 
Gazdar, G. (1979). Pragmatics: lmplicature, Pre- 
supposition, and Logical Form. New York: 
Academic Press. 
Green, Nancy L. (1990). Normal state impli- 
cature. In Proceedings of the 28th Annual 
Meeting, Pittsburgh. Association for Compu- 
tational Linguistics. 
Grice, H. Paul (1975). Logic and conversation. In 
Cole, P. & Morgan, J. L. (Eds.), Syntax and 
Semantics III: Speech Acts, (pp. 41-58)., New 
York. Academic Press. 
Grosz, Barbara & Sidner, Candace (1988). Plans 
for discourse. In P. Cohen, J. Morgan, 
M. Pollack (Eds.), Intentions in Communica- 
tion. MIT Press. 
Hintikka, J. (1962). Knowledge and Belief. Ithaca: 
Cornell University Press. 
Hirschberg, Julia B. (1985). A Theory of Scalar 
Implicature. PhD thesis, University of Penn- sylvania. 
ttoracek, Helmut (1991). Exploiting conversa- 
tional implicature for generating concise expla- 
nations. In Proceedings. European Association 
for Computational Linguistics. 
Hovy, Eduard H. (1988). Planning coherent multi- 
sentential text. In Proceedings of the 26th An- 
nual Meeting, (pp. 163-169). Association for 
Computational Linguistics. 
Joshi, Aravind, Webber, Bonnie, & Weischedel, 
Ralph (1984a). Living up to expectations: 
Computing expert responses. In Proceedings 
of the Fourth National Conference on Artificial 
Intelligence, (pp. 169-175)., Austin, Texas. 
Joshi, Aravind, Webber, Bonnie, & Weischedel, 
Ralph (1984b). Preventing false inferences. 
In Proceedings of Coling84, (pp. 134-138), 
Stanford University, California. Association for Computational Linguistics. 
Lambert, Lynn & Carberry, Sandra (1991). A tri- 
partite plan-based model of dialogue. In Pro- 
ceedings of the 29th Annual Meeting, (pp. 47- 
54). Association for Computational Linguis- tics. 
Lascarides, Alex & Asher, Nicholas (1991). Dis- 
course relations and defensible knowledge. In 
Proceedings of the 2gth Annual Meeting, (pp. 
55-62). Association for Computational Lin- 
guistics. 
Lascarides, Alex & Oberlander, Jon (1992). Tem- 
poral coherence and defensible knowledge. Theoretical Linguistics, 18. 
Litman, Diane & Allen, James (1987). A plan 
recognition model for subdialogues in conver- sation. Cognitive Science, 11, 163-200. 
Mann, William C. & Thompson, Sandra A. (1983). 
Relational propositions in discourse. Technical 
Report ISI/RR-83-115, ISI/USC. 
Mann, William C. & Thompson, Sandra A. (1987). 
Rhetorical structure theory: Toward a func- 
tional theory of text organization. Text, 8(3), 167-182. 
McCafferty, Andrew S. (1987). Reasoning about Im- 
plicature: a Plan-Based Approach. PhD thesis, 
University of Pittsburgh. 
Moore, Johanna D. & Paris, Cecile (1989). Plan- 
ning text for advisory dialogues. In Proceed- 
ings of the 27th Annual Meeting, University of 
British Columbia, Vancouver. Association of 
Computational Linguistics. 
Moore, Johanna D. ~z Paris, Cecile L. (1988). Con- 
structing coherent text using rhetorical rela- 
tions. In Proc. lOth Annual Conference. Cog- 
nitive Science Society. 
Perrault, Raymond & Allen, James (1980). A 
plan-based analysis of indirect speech acts. 
American Journal of Computational Linguis- 
tics, 6(3-4), 167-182. 
Pollack, Martha (1988). Plans as complex men- 
tal attitudes. In P. Cohen, J. Morgan, & 
M. Pollack (Eds.), Intentions in Communica- 
tion. MIT Press. 
Reichman, Rachel (1984). Extended person- 
machine interface. Artificial Intelligence, 22, 157-218. 
Reiter, Ehud (1990). The computational complex- 
ity of avoiding conversational implicatures. In 
Proceedings of the 28th Annual Meeting, (pp. 
97-104)., Pittsburgh. Association for Compu- 
tational Linguistics. 
Sadock, Jerrold M. (1978). On testing for conversa- 
tional implicature. In Cole, P. & Morgan, J. L. 
(Eds.), Syntax and Semantics, (pp. 281-297)., 
N.Y. Academic Press. 
Thomason, Richmond H. (1990). Accommoda- 
tion, meaning, and implicature: Interdisci- 
plinary foundations for pragmatics. In P. Co- 
hen, J. Morgan, &; M. Pollack (Eds.), In- 
tentions in Communication. Cambridge, Mas- 
sachusetts: MIT Press. 
Wainer, Jacques & Maida, Anthony (1991). Good 
and bad news in formalizing generalized impli- 
catures. In Proceedings of the Sixteenth An- 
nual Meeting of the Berkeley Linguistics Soci- 
ety, (pp. 66-71)., Berkeley, California. 
