The Local Organization of Text 
Penelope Sibun 
Department of Computer and Information Science 
University of Massachusetts 
Amherst, MA 01003 
sibun@cs.umass.edu or penni@umass.bitnet 
Abstract 
In this paper, I present a model of the local organization 
of extended text. I show that texts with weak rhetorical 
structure and strong domain structure, such as descrip- 
tions of houses, digital circuits, and families, are best 
analyzed in terms of local domain structure, and ar- 
gue that global structures that may be inferred from 
a domain are not always appropriate for constructing 
descriptions in the domain. I present a system I am im- 
plementing that uses short-raTtge strategies to organize 
text, and show how part of a description is organized by 
these strategies. I also briefly discuss a model of incre- 
mental text generation that dovetails with the model of 
local organization presented here. 
Motivation for local organization 
The approach to organizing extended text described 
here has both psychological and computational moti- 
vation. It aims both to model how people use language 
and to provide a flexible architecture for a system's lan- 
guage use. In this section, I describe the empirical data 
that form the basis of this research, and characterize 
the local organization of the collected texts. In the 
next two sections, I describe a computational architec- 
ture to implement local text organization and discuss 
its advantages of generality and flexibility, and give an 
example of how this architecture works. 
An extended text has a structure; this structure is a 
description of how the components relate so that sense 
can be made of the whole. Two sources of this organiza- 
tion are rhetoricial structure, which describes the way 
elements of the text fit together, and domaitt structure, 
which describes relations among domain objects. For 
this research I chose three domains with strong domain 
structure, and a task--description--with weak rhetor- 
ical structure. I have tape-recorded 29 people giving 
descriptions of house layouts, electronic circuit layouts, 
and family relationships. Description fragments of a 
house and of a family, and the questions asked to ob- 
tain the descriptions, are given in figure 1. (Because of 
space considerations, the fragments are somewhat ab- 
breviated.) 
Many approaches to text organization 1 are based 
on analyses of text in terms of rhetorical structure. 
However, there are few segments of text with inter- 
esting rhetorical structure in my corpus. For exam- 
ple, an analysis of the texts using Mann and Thomp- 
son's (1987) Rhetorical Structure Theory (RST) would 
result primarily in the relations sequence and joint 
and would contain few of the the relations like evi- 
dence or justify that give RST its descriptive power. 
Similarly, it is unclear what work a system like that of 
Grosz and Sidner (1986) would do in analyzing a de- 
scription. 
Since the structure of descriptions cannot be ana- 
lyzed adequately with rhetorical relations, perhaps it 
can be explained in terms of the domain. Houses, chips, 
and families are strongly structured. A family's rela- 
tionships can be captured in a family tree; one might 
suppose that a description of the family would also be 
organized in this way. A house can be encoded in a 
number of ways; for instance, it has a component hier- 
archy, being composed of rooms composed of furnish- 
ings. Linde (1974) has proposed another comprehensive 
structure for houses: a phrase structure grammar that 
determines how the rooms may be visited in a traversal 
of a house layout. 
Surprisingly, these global, hierarchical domain struc- 
tures are not exploited in the organization of descrip- 
tions in my corpus. While family trees and composition 
hierarchies can be inferred from descriptions of families 
and houses, this does not mean that these structures 
guide the process of organizing them. For instance, 
my family informants did not simply construct their 
descriptions by starting at the root of the appropriate 
tree and doing a depth-first or breadth-first traversal 
of it. Instead, to select a next family member to talk 
about, they would apply one of several criteria. Gener- 
ally, a sibling, spouse, parent, or child would be the next 
choice, and this might incidentally constitute part of a 
tree walk. But that this choice is local is evidenced by 
the next choice, which may not be construable, in any 
XWhat I call text organization is usually referred to as 
text planning. 
120 
on our righthand side would be a door, 
which leads to Penni's room 
and you walk in there... 
and there are two windows, 
in the...opposite corner from the one in 
which you enter 
one' s on the lefthand wall, 
and one's on the wall that you would be 
facing 
then, on the righthand side...of her room, 
is the closet 
Fragment of a house description. 
In response to the question: "Could you please describe for me the layout of your house." 
there's my mother Katharine, my father John, 
my sister Penni, and me 
it's my mother's relatives that we go to see 
Margaret and Bill, 
who are Mommy' s... 
urn...Margaret's my great-aunt 
so it must be Mommy's aunt 
Fragment of a family description. 
In response to the question: "Can you tell me how everyone who comes to Thanksgiving is related to each other?" 
Figure I: Sample descriptions. 
principled way, as part of the overall structure of the 
description that one might have postulated at the pre- 
vious step. Where to begin the family description also 
appears to be a locally conditioned choice. Informants 
begin at various points, such as themselves or long-dead 
progenitrixes, but the majority start their description 
of the family by mentioning the hostess and host of the 
Thanksgiving dinner they are attending; we may sup- 
pose that at a different time of year the descriptions are 
likely to start off differently. 
Further evidence that people do not structure their 
descriptions using obvious global domain structures 
may be adduced from examples in which speakers ex- 
plicitly deny knowledge of such structures, as in the 
following fragment. 
and also...tun... 
Eleanor and Elizabeth come 
who are...cousins of...all of us...um... 
I don't know what generation cousins they 
are 
Here, the speaker shows by her description of two 
family members that she does not know her relationship 
to them, even though it would be clear if a family tree 
were being used to organize the description (the women 
in question are in fact first cousins twice removed of the 
speaker). 
Genealogical trees, phrase structure grammars, and 
component hierarchies are useful for succinctly repre- 
senting information about houses, chips, and families 
But there is no a priori reason to suppose that a de- 
scription of such things are the products of such easily 
articulable schemas or grammars. When we examine 
texts of the sort that we wish to generate, we must 
distinguish the mechanisms that direct the process of 
choosing what to say next from a retrospective descrip- 
tion of its result. 
The texts I have collected can be best analyzed as 
locally organized by a process of deciding what to say 
nezt This decision is based principally on what has 
already been said and what is currently available to 
say. For example, if one has just mentioned a large 
window in the kitchen, one can mention whatever is to 
the left of the window, whatever is to the right of it, 
which way it is facing, what it looks like, or how it is 
similar to a window in another room of the house. If 
one has mentioned an aunt, one can give her name, say 
whether she is married, mention her sister, enumerate 
her children, or talk about how much money she earns. 
The strong domain structure of subjects like houses, 
chips, and families ensures that a description can be 
continued from any point: once a description has been 
started, there is always something, often many things, 
that can be said next. In structured domains, there is 
always a default choice for the next thing to say. In 
spatial domains like houses, spatial proximity provides 
this default. Everything in a house, be it a room, a 
wall, or a kitchen appliance, is next to something else. 
Spatial proximity does not constrain house descriptions, 
121 
but it ensures that a description does not come to a 
premature dead end. 
Descriptions are finite. Though there may always be 
more to say, there are points at which a description 
may stop, when the task may be considered accom- 
plished. Linde (1974) proposed s completeness criterion 
for house descriptions, which is reflected in my data as 
well as hers. It states that a description may stop any 
time after all the rooms have been mentioned, but it is 
not complete if it stops before. A similar criterion holds 
in the family descriptions collected: they were given in 
answer to the question, "Can yoti tell me how everyone 
who comes to Thanksgiving is related to each other?" 
In this case, then, the criterion is mentioning everyone 
who attends. 
Knowing how to continue and knowing when to stop 
together ensure that a description can be generated de- 
pending solely on local organization. The strong do- 
main structure of houses and families makes the work- 
ing of these mechanisms for continuation and termina- 
tion particularly clear, and thus these domains are a 
good site for studying this approach to organizing text. 
However, the local organization of text is also evident 
in many other uses of language. People's conversation 
is often locally organized (Levinson, 1983); some inter- 
active systems are currently being designed with this 
approach (Frohlich & Luff, 1989). 
Because I am interested not only in how a program 
may organize text but also in how people do so, I study 
people speaking rather than people writing. A written 
text may be edited and reorganized, and this process 
often involves explicitly thinking about rhetorical struc- 
ture. A spoken description is likely to require that the 
speaker organize her text locally--she cannot plan it 
out ahead of time. Studying spoken text reveals more 
of the underlying mechanisms of language, because time 
constraints and the inability to edit what has already 
been said make post-processing impossible. 
Computational architecture 
I am implementing a system that employs local orga- 
nization of text as described in this paper. The imple- 
mentation comprises: a semantic net knowledge base; 
an organizer composed of strategies and metastrategies; 
and a generator. Local organization is achieved using 
short-range strategies, each of which is responsible for 
organizing only a short segment of text, between a word 
and a clause in length. 
Until recently, I have employed Mumble-86 (Meteer 
et al., 1987) as the generator for this system. Construc- 
tion is underway, however, on a simpler generator that 
more accurately implements the principles of generation 
implied by the structure of the organizer. The system is 
currently implemented only for descriptions of houses; 
the examples in this section will thus be drawn from 
the house domain. 
The knowledge base is a semantic net that encodes 
the objects, properties, and relations of a domain. The 
organizer keeps a pointer to the current node in the de- 
scription. A strategy describes the current node and 
others related to it via local connections in the net- 
work. Strategies are selected sequentially, based on 
local conditions; if there is no strategy available or if 
more than one strategy is appropriate, metastrategies 
(Davis, 1980) resolve the conflict, using a technique 
similar to universal subgoaling in Soar (Laird, Newell 
Rosenbloom, 1987). Metastrategies are responsible 
for control: they sequence and combine strategies. Like 
strategies, they can cause the production of text. 
This architecture has several advantages. First, it is 
flexible: because there is no fixed priority scheme and 
because strategies are small, the strategies can be com- 
bined in a variety of ways to produce texts that are 
constrained only by the appropriateness of each strat- 
egy as determined by the strategies' interaction with the 
knowledge base. Second, the architecture is extensible: 
new strategies can easily be added to extend the orga- 
nizer to different types of text. Finally, the organizer is 
mainly domain-independent: while some strategies may 
be particular to houses, most strategies are not. 
The strategies are applied to the knowledge base and 
select the items that make up the description. The 
strategies find the appropriate lexical items(s) for each 
knowledge base item that is expressed; these lexical 
items and the knowledge base items themselves are de- 
termined by the domain. While the strategies are sim- 
ple, complex behavior emerges from their interaction 
with the knowledge base; this locally organizes the ex- 
tended text. 
Each strategy falls into one of four classes, with 
varying degrees of domain independence: discourse cue 
strategies; linguistic strategies; parameterizable domain- 
independent strategies; and semi-domain-independent 
strategies. Figure 2 gives examples of each. 
Of the domain-independent strategies, discourse cue 
strategies focus attention, in a way similar to the clue 
words described by Reichman (1985), and linguistic 
strategies mention objects and associated properties. 
mentlon-sallent-object, used to say "there is a win- 
dow," may as easily express "there is a penguin" or 
"there is a policy." describe-object is similarly all- 
purpose, and can produce "the window with two flank- 
ing windows" or "the man with one black shoe." 
The parameterizable domain-independent strategies 
have slightly different textual realizations in different 
domains, but these differences can be captured by pa- 
rameters. The example given in figure 2 is typical: the 
strategy is realized as a prepositional phrase in each 
domain, and only the preposition changes. 
The semi-domain-independent strategies accomplish 
tasks such as a sweep that seem particular to the do- 
main, but are similar to tasks in other domains. A 
sweep begins at an object and names another bearing 
some spatial relationship to it, and then another object 
122 
Discourse Cue Strategies 
introduce/shift-attention "then," "and," "now" 
refer-back/relnforce "again," "once more" 
Linguistic Strategies 
mentlon-sallent-object "there is z" 
"we have z" 
describe-object "the z is ~," 
"the y z" 
"the z (which) has z" 
"the z with z" 
Parameterisable Domain-independent Strategies 
situate "in the kitchen" 
"during the morning" 
"about the election" 
Semi-domain-independent Strategies 
sweep Enumerate objects connected each to the 
next by the same relation. 
follow a path Traverse a natural connection be- 
tween parts of the knowledge to be described. 
Figure 2: Strategies. 
that bears the same relationship to the just-mentioned 
one, until there are none left; for example, "to the left 
of the window is a stove and then a refrigerator" is 
a sweep-left. Similar constructions may be found in 
other domains. A description of one's day may start at 
some event and mention the next event and the event 
after that. In this case, the relationship is temporal, 
rather than physical. 
Metastrategies select strategies based on the contezt, 
which comprises: 
Local Context 
• What has just been said. 
• What is immediately connected to the current item 
in the knowledge base. 
• What strategies are applicable. 
Global Context 
• Discourse and speaker parameters. (For example, a 
speaker's propensity to mention objects to the fight 
before objects to the left.) It is here that anything 
considered a global goal would be encoded. 
• The completeness criterion. 
In future implementations, the context may also in- 
volve some model of the hearer. This would be part of 
the local context: what one knows about one's hearer 
and about what one's hearer knows changes, particu- 
larly under the assumption that hearer and speaker are 
engaged in a two-way interaction. 
A strategy conflict set contains whatever strategies 
are currently applicable. If it contains a single strat- 
egy, that one is selected. If more than one is in the 
set, met"strategies may resolve the conflict by selecting 
one or by combining some or all of them. Finally, if 
no strategy presents itself, the met,strategies apply a 
default strategy. 
The met"strategy find-"interesting"-llnk is trig- 
gered after the introduction of a new topic, which has 
links to many other items in the knowledge base. What 
is "interesting" or salient depends on: 
• The domain: In spatial description, objects that 
are large or have many features are interesting. 
• The structure of the domain: Objects that are 
more connected to other objects are more interesting. 
• The local context: If a window has just been men- 
tioned, there is reason to mention other windows. 
• The global context: There may be an inclination 
to mention furnishings but not structural features of 
the house, or vice versa; there may be differing levels 
of detail required. 
123 
Some metastrategies combine strategies. Such metas- 
trategies apply when several strategies are appropriate 
at some point in the description and there is a felicitous 
way to combine them. clrcular-sweep is an example: 
it combines a number of sweep strategies (sweep-left, 
sweep-right, and sweep-under), and includes addi- 
tional orientation strategies to orient hearers between 
sweeps, kltty-corner is a metastrategy that is used 
to describe s room in which the most salient feature 
is diagonally opposite the current location. The object 
that is "kitty corner" from it is mentioned first, and 
the rest of the room is described in relation to it. 2 The 
first fragment in figure 1 exemplifies the kltty-corner 
metastrategy. 
find-new-topic is the default metastrategy just in 
case there is nothing "interesting" to say. In a spatial 
domain, the default is to select an object to describe 
using spatial proximity. 
Example of description organization 
In this section, I describe the organization of a fragment 
of a description in my corpus. While the system is 
not yet fully implemented to handle all the details, this 
example is sufficiently complex to show the operation of 
the architecture in selecting appropriate strategies and 
metastrategies. Though the strategies used are simple, 
complex choices, varying with context, are made. On 
the following page are the text and a sketch of the area 
being described. 
In the fragment in figure 3, the speaker is describing 
the bedroom he shares with his wife Carol. Each line 
of the fragment, in most cases, is the result of a single 
strategy. 
The sketch in figure 4 of the bedroom is provided as 
an aid to the reader in understanding John's descrip- 
tion. There is no corresponding representation in the 
system. 
The global context used by this speaker includes pa- 
rameters that predispose him to mention rather than 
ignore room furnishings, as well as "stuff"--small arti- 
cles than can be found in, on, or near pieces of furniture. 
The strategy mention-stuff is a particular form of 
descrlbe-object, and as such is concerned with men- 
tioning associated properties of the object rather than 
physical proximity; this is suggested by the speaker's 
typically using the preposition "with," rather than an 
obviously spatial one. 
The global context also of course includes the com- 
pleteness criterion which is unsatisfied throughout this 
stretch of text. The fragment starts when the speaker 
has just finished describing the kitchen and the next 
spatially proximate thing is the door to the bedroom. 
When the node to be described is a physical object, 
2For a fuller treatment of how the system computes de- 
ictic and other spatial terms Uke "kitty corner," "left," and 
"right," see (Sibun & Huettner, 1989). 
the mention-object strategy is always available; be- 
cause this object is a door, mention-room is available 
to talk about the room that the door leads into. The 
metastrategies resolve this conflict in favor of the more 
particular mention-room {1}. 
Because the last strategy used was mentlon-room, 
mention-salient-object becomes available; if there 
is a particularly "salient" object in a room, descrip- 
tions can then be organized around it. As it happens, 
salience in this domain tends to depend primarily on 
size; the kingsize bed fits the bill {2}. There are two 
objects spatially proximate to the bed; one is selected 
and the strategy mentlon-object is used to mention 
the endtable {3}. The endtable is connected to several 
other items in the knowledge base, but because there 
is a context parameter to mention "stuffs" this is what 
happens next {4}. 
Now, there are two unmentioned objects spatially 
proximate to the endtablemthe window and the "wall" 
(which is actually a covered-up chimney). The "wall" 
is mentioned because, like furnishings, it has extent in 
the room, and the context disposes the process toward 
mentioning such objects {5}. The local context keeps 
track of what has just been mentioned; another feature 
it records is the direction in which the spatial proximity 
finks have been followed. The "wall" is next along the 
trajectory from the bed through the endtable. 
The "wall" is spatially linked to three things: the 
window on the endtable side of it, the window that is 
along the same trajectory that has been followed, and 
the chests, which are also along that trajectory. The 
window along the trajectory is the choice selected from 
these three for two reasons: there is a tendency to main- 
tain trajectories; 3 and the last thing mentioned is part 
of the structure of the room, as is the window, but not 
the chests. But this selection of the window presents a 
problem (at least, we can infer that it presents a prob- 
lem to the speaker), because this window is connected 
via a similarity link to the previous window, which has 
not been mentioned. So the speaker performs a repair 
consisting of backing up, mentioning the overlooked 
window, reiterating mention of the wall, and, finally, 
mentioning the selected window {6}. 
While the next window is a promising candidate be- 
cause it is the same sort of thing as that just mentioned, 
the parameter for mentioning furniture overrides this, 
and the cedar chests come next in the description {7}, 
followed by their associated "stuff" {8}. The window 
is again available but so is the bureau, which is the 
preferred choice because it is furniture {9}. The bu- 
reau has spatial proximity links to two windows and 
the "small thing," as well as links to its "stuff," so the 
set of available strategies comprises ones that mention 
each of these things. A metastrategy can resolve this 
3Ullmer-Ehrich (1982) notes a similar tendency in the 
dorm room descriptions that she collected. 
124 
and then there's the bedroom {1} 
and there's a huge kingsize bed {2} 
and there's an endtable next to it {3} 
with a lamp and a clock rad/o {4} 
and some stuff of mine 
like the Boston University cup 
and...then there's a wall-- {5} 
and then there's a window behind that {6} 
and there's a wall, and there's another window 
and there's some...cedar chests of Carol's {7} 
that have blankets and sheets in them {8} 
and...there's her bureau {9} 
in the middle of two windows on either side {10} 
with all of her makeup on top of it and clothes 
and there's...a small thing with all her clothes 
and there's another great big bookshelf 
and all her spare books 
and there's a small end table over on her side 
um...a small digital clock and more kleenex 
and...OK 
Figure 3: John's description. 
,J. 
I window I 
I on0,a ,o J I 
bed 
I end table 
I 
I window I 
\[ bookshelf \[ 
Qmall thing 
R 
Figure 4: Sketch of bedroom. 
125 
conflict by realizing that strategies for saying that there 
is an object of the same sort on two different sides of 
the current object can be combined by saying that the 
bureau is between the two windows {10}. 4 
The description for the rest of the room continues in 
a manner similar to that already discussed. 
Incremental generation 
The organizer described here is composed of strategies 
that are often responsible for sub-clausal units of text; 
furthermore, the strategies have already imposed an or- 
ganization on the text, obviating the need for the gen- 
erator to do more than enforce gramma~icalitT/. The 
model of local text organization I van developing is cou- 
pled with a model of incremental generation, in which 
the increments are often smaller than a sentence. (Gaz- 
rett's investigation of speech errors (1975) constitutes 
early work in this area; De Smedt & Kempen (1987), 
and Kempen & Hoenkamp (1987) discuss a similar, 
more fully-developed incremental generation project.) 
A typical generator produces a sentence at a time. How- 
ever, spoken text is replete with restarts, fragments, 
and ungrammaticalities. This suggests that not only 
do people organize their text incrementally, but gener- 
ate it in increments as well. 
Usually, an incremental generator is successful in gen- 
erating grammatical text. The text 
then / 
in the kitchen / 
there is a large window 
is the result of three sequentially operating strategies: 
introduce/shltt-attention, situate, and mention- 
sallent-object. An incremental generator will be able 
to produce this text in the increments specified by the 
strategies. 
A system that generates in strategy-sized increments 
can result, in principled ways, in ungrammatical text. 
A common error, exemplified by 
and a lot of other people / 
who / 
I wasn't quite sure what they did 
can be explained by the operation of the strategies 
mention-object, add-clausal-modifier, and a set 
of strategies to express some additional information, 
which, because a dependent clause has already been 
started, happens to result in the ungrammatical re- 
sumptive pronoun "they." 
Locally-organized, occasionally ungrammatical text 
may be prototypical, but it is certainly not the only sort 
'Note that the previous window conflict was not resolved 
by saying that the "wall" was between the two windows. 
The difference can be explained by the observation that the 
"wall," despite its extent into the room, is a structural ob- 
ject, wld/e the bureau is furniture. 
of text we wish to generate. To be comprehensive, my 
system will require some capability to post-process text 
after it is organized and before it is generated (Hovy, 
1989, suggests a post-processing model). Output of my 
system might also be appropriate input for a text revi- 
sion system (e.g., Meteer & McDonald, 1986). 
Related research 
There are many other projects whose goal is to organize, 
or plan, extended text. The main difference between 
these and mine is flexibility and level of organization: 
most text planning systems rely on global structures to 
organize paragraph-sized text. These structures, which 
are usually schemas or plans, constrain the text to re- 
flect particular rhetorical and domain relations, but at 
the expense of flexibility. My system builds structure 
locally with no recourse or reference to an overall struc- 
ture. 
Through analysis of written texts, McKeown (1985) 
has developed a small set of schemas, built from rhetor- 
ical predicates, that would provide types of descriptions 
(for example, constituency) for objects in a knowledge 
base. ParAs has extended this work to use a process 
trace which derives its structure in part from the do- 
main (Paris & McKeown, 1987; Paris, 1988). This al- 
ternative strategy builds and traverses a path through 
the knowledge base when it is determined, on the basis 
of a sparse model of user expertise, that a process de- 
scription of an object is more appropriate than a declar- 
ative one; the two strategies may be interleaved. Dale 
(1989) similarly organizes text by means of a domain 
structure, in this case, a recipe plan. 
Rhetorical Structure Theory (Mann & Thompson, 
1987) is also drawn from an analysis of written texts; 
it differs from McKeown's work in that it is composed 
of a large number of rhetorical relations, rather than 
a small number of schemas. RST may thus be more 
flexible, but is is still assumed that the relations will be 
combined into a single global tree covering an extended 
text. 
While most text planning systems have worked on 
producing single texts in response to queries, some re- 
search has been more particularly concerned with inter- 
active text. Much recent work in this area has been for 
explanation systems (e.g., Maybury, 1989), and some 
of this work explicitly addresses allowing a human user 
to ask follow-up questions (Moore & Swartout, 1989). 
However, such systems still build and use global struc- 
tures for extended texts. 
An argument is sometimes made that global struc- 
ture is needed to capture high-level rhetorical goals in 
the output text (see Appelt, 1985); Gricean Maxims 
(Grice, 1975) are often invoked. Hut, as Hovy (1988) 
points out, "be polite" is not a decomposable goal; the 
objective of being polite is achieved through local de- 
cisions. Such local decisions can comfortably be inte- 
grated with a model of local organization of text. 
126 
Acknowledgements 
I thank Bonnie Webber and David Chapman for much 
useful discussion in the preparation of this paper. I also 
thank Bruce Leban for helping to make the Mac draw 
the picture. This work was supported in part by the Air 
Force Systems Command, Rome Air Development Cen- 
ter, Grii~ss AFB, New York, 13441 under contract No. 
F30602-85-C-0008, as part of the Northeast Artificial 
Intelligence Consortium (NAIC). 

References 
Appelt, D., "Planning English Referring Expres- 
sions." Artificial Intelligence 26, pp 1-33, 1985. 
Dale, R., "Generating Recipes: An Overview of Epi- 
cure." In Eztended Abstracts Presented at the Sec- 
ond European Natural Language Generation Wor~hop, 
pp 7-13, 1989. 
Davis, R., "Meta-rules: Reasoning about Control." 
Artificial Intelligence 15, pp 179-222, 1980. 
De Smedt, K. and G. Kempen, "Incremental Sen- 
tence Production, Self-correction, and Coordination." 
In G. Kempen (ed), NaturaILanguage Generation: Nelo 
Results in Artificial Intelligence, Psychology and Lin- 
guistics, Martinus Nijhoff Publishers, pp 36,5-376, 1987. 
Frohlich, D. and P. Luff, "Applying the Technology of 
Conversation to the Technology for Conversation." In 
Proceedings of the Computers and Conversation Sym- 
posium, University of Surrey, England, 1989. 
Garrett, M., "The Analysis of Sentence Produc- 
tion." Psychology of Learning and Motivation, Volume 
9, pp 133-177, 1975. 
Grice, H., "Logic and Conversation," In P. Cole and 
J. Morgan (eds.), Syntaz and Semantics 3: Speech Acts, 
pp 41-58, 1975. 
Gross, B. and C. Sidner, "Attention, Intentions, and 
the Structure of Discourse." Computational Linguis- 
tics, (12)3, 1986. 
Hovy, E., "Two Types of Planning in Language Gen: 
eration." In Proceedings of the ~6th Annual Meeting of 
the Association for Computational Linguistics, pp 179- 
186, 1988. 
Hovy, E., "Unresolved Issues In Paragraph Plan- 
ning." To appear in Proceedings of the Second European 
Natural Language Generation Wor~hop, 1989. 
Kempen, G. and E. Hoenkamp, "An Incremental 
Procedural Grammar for Sentence Formulation." Cog- 
nitive Science 11, pp 201-258, 1987. 
Laird, J., A. Newell, and P. Rosenbloom, "Soar: An 
Architecture for General Intelligence." Artificial Intel- 
ligence, 33, pp 1-64, 1987. 
Levinson, S., Pragmatics. Cambridge University 
Press, 1983. 
Linde, C., The Linguistic Encoding of Spatial Infor- 
mation. Doctoral Dissertation, Columbia University, 
1974. 
Mann, W. and S. Thompson, Rhetorical Structure 
Theory: A Theory of Tezt Organization. Report 
ISI/RS-87-190, Information Sciences Institute, 1987. 
Maybury, M., "Enhancing Explanation Coherence 
with Rhetorical Strategies." In Proceedings of the Euro- 
pean Association of Computational Linguistics,, pp 168- 
173, 1989. 
McKeown, K., Test Generation. Cambridge Univer- 
sity Press, 1985. 
Meteer (Vanghan), M. and D. McDonald, "A Model 
of Revision in Natural Language Generation. n In Pro- 
ceedings of the ~th Annual Meeting of the Association 
for Computational Linguistics, pp 90-96, 1986. 
Meteer, M., D. McDonald, S. Anderson, D. Forster, 
L. Gay, A. Huettner, and P. Sibun, MUMBLB-86: De- 
sign and Implementation. COINS Technical Report 87- 
87, University of Massachusetts, 1987. 
Moore, J. and W. Swartout, "A Reactive Approach 
to Explanation." In Proceedings of the International 
Joint Conference on Tezt Generation IJCAI, 1989. 
Paris, C., "Tailoring Object Descriptions to a User's 
Level of Expertise." Computational Linguistics 14(3), 
pp 64-78, 1988. 
Paris, C. and K. McKeown, "Discourse Strategies for 
Describing Complex Physical Objects." In G. Kempen 
(ed), Natural Language Generation: New Results in Ar- 
tificial Intelligence, Psychology and Linguistics, Marti- 
,us Nijhoff Publishers, pp 97-115, 1987. 
Reichman, R., Getting Computers to Talk Like You 
and Me. MIT Press, 1985. 
Sibun, P. and A. Huettner, "Spatial Deixis in Gen- 
erating Descriptions." COINS Technical Report 89-34, 
Department of Computer and Information Science, Uni- 
versity of Massachusetts, 1989. 
Ullmer-Ehrich, V., "The Structure of Living Space 
Descriptions." In R. Jar,ella and W. Klein, eds., 
Speech, Place, and Action, John Wiley & Sons, Ltd., 
1982. 
