Belief Ascription and Model Generative Reasoning: 
joining two paradigms to a robust parser of messages. 
Yorick Wilks and Roger Hartley 
Computing Research Laboratory 
New Mexico State University 
Box 30001, Las Cruces, NM 88003. 
ABSTRACT 
This paper discusses the extension of ViewGen, a program for belief ascription, to the area of inten- 
sional object identification with applications to battle environments, and its combination in a overall sys- 
tem with MGR, a Model-Generative Reasoning system, and PREMO a semantics-based parser for robust 
parsing of noisy message data. 
ViewGen represents the beliefs of agents as explicit, partitioned proposition-sets known as environ- 
ments. Environments are convenient, even essential, for addressing important pragmatic issues of reason- 
ing. The paper concentrates on showing that the transfer of information in intensional object 
identification and belief ascription itself can both be seen as different manifestations of a single 
environment-amalgamation process. The entities we shall be concerned with will be ones, for example, 
the system itself believes to be separate entities while it is computing the beliefs and reasoning of a hos- 
tile agent that believes them to be the same entity (e.g. we believe enemy radar shows two of our ships to 
be the same ship, or vice-versa. The KAL disaster should bring the right kind of scenario to mind). The 
representational issue we address is how to represent that fictional single entity in the belief space of the 
other agent, and what content it should have given that it is an amalgamation of two real entities. 
A major feature of the paper is our work on embedding within the ViewGen belief-and-point-of- 
view system the knowledge representation system of our MGR reasoner, and then bringing together the 
multiple viewpoints offered by ViewGen with the multiple representations of MGR. The fusing of these 
techniques, we believe, offers a very strong system for extracting message gists from texts and reasoning 
about them. 
INTRODUCTION 
The purpose of this paper is to bring together the reasoning techniques used in the MGR and 
ViewGen systems and combine those with the PREMO robust parsing system (see references in separate 
sections below) to yield a combined approach to understanding text, particularly the interpretation or gist- 
ing of text which may be ill-formed, fragmented, stylized or even self-contradictory (when produced 
under stressful conditions, but evaluable against an appropriate knowledge base nonetheless). We also 
aim to build in pointers to the evaluation of the system as we go. Another guiding principle is to combine 
if possible, the benefits of being knowledge and situation based, when needed for interpretation, with the 
benefits that come from large scale automated analysis techniques drawn from the use of machine read- 
able dictionaries and large text samples. 
We shall direct our examples where possible to realistic military scenarios drawn from other 
Army-sponsored work at CRL and constnact scenarios based on ship or troop unit sightings and casualty 
reports. However, our techniques are better, we hope, than the individual scenarios we can construct. In 
other cases we shall use the standard cases with which the systems have been developed. 
219 
The overall aim is to produce a system that takes in noisy messages and data, has access to both 
Blue data-bases and military doctrine (in terms of beliefs and goals) and, in certain circumstances, access 
also to Red Army beliefs and goals. A number of possible scenarios are tentatively put forward at the end 
of the paper, in which the combination of the techniques discussed here could provide advice to, and a 
check upon the decisons of, the G2 officer in a Blue Army Corps, particularly in terms of the consistency 
of the advice offered the commander and consilience of the belief and data structures maintained at other 
lower levels of the system (e.g. Divisions). 
In the first three sections of the paper we summarize the PREMO, MGR and ViewGen techniques 
and then proceed to a discussion of how to link them: in general we envisage an overall system in which 
information moves (outward in the diagram below) from the robust parser PREMO, to the MGR, the gen- 
erator of alternative scenarios/models, to ViewGen the point-of-view shell that considers the possible 
points of view in terms of the environments of other agents (e.g. Blue Generals view of Red General's 
view of a particular tank division). Both MGR and ViewGen also act as filters of possible but inapplica- 
ble models of the situation. 
PREMO: A ROBUST PARSER OF MESSAGES 
PREMO: the PREference Machine Organization is a knowledge-based Preference Semantics 
parser (Wilks 1972, 1975, 1978; Boguraev 1979; Carter 1984, 1987; Fass 1986, 1987, 1988; Huang 1984, 
1988; Slator 1988a, 1988c) due to Brian Slator, with access to the large, text-specific, lexical semantic 
knowledge base created by the lexicon-provider of the CRL project on large scale lexical extraction from 
machine readable dictionaries (Wilks et al. in press). A fuller description of the relationship of PREMO to 
that project appears in (Slator & Wilks 1989). PREMO is the parser we intend to use initially for initial 
processing of military reports (sightings, casualties etc.) that have been generated raapidiy under adverse 
220 
conditions. 
Preference Semantics is a theory of language in which the meaning for a text is represented by a 
complex semantic structure that is built up out of smaller semantic components; this composifionality is a 
fairly typical feature of semantic theories. The principal difference between Preference Semantics and 
other semantic theories is in the explicit and computational accounting of ambiguous, metaphorical, and 
linguistically non-standard language use; which is to say, it is intended to deal with the actual world of 
English texts outside linguistic studies, ranging from newspaper texts (well-formed but full of the non- 
standard phenomena just listed) to rapidly written diagostic reports, like machine-repair, or reports of 
sightings, which have all those features and are grammatically ill-formed in addition.. 
The links between the components of the semantic structures are created on the basis of semantic 
preference and coherence. In text and discourse theory, coherence is generally taken to refer to the mean- 
ingfulness of text. Fass (1987) suggests that in NLP work such as Preference Semantics the notions of 
"satisfaction" and "violation" (of selection restrictions or preferences) and the notion of "semantic dis- 
tance" (across structured type hierarchies) are different ways of characterising the meaningfulness of text; 
they capture different coherence relations. The original systems of Preference Semantics (Wilks 1972, 
1975, 1978), were principally based on the coherence relation of "inclusion" (semantic preferences and 
selection restrictions); the emphasis in PREMO is more on the coherence relation based on semantic dis- 
tance, although the original notions of coherence also survive. 
In Preference Semantics the semantic representation computed for a text is the one having the most 
semantically dense structure among the competing "readings." Semantic density is a property of struc- 
tures that have preferences regarding their own constituents, and satisfied preferences create density. Den- 
sity is compared in terms of the existence of preference-matching features, the lack of preference-breaking 
features, and the length of the inference chains needed to justify each sense selection and constituent 
attachment decision. The job of a Preference Semantics parser, then, is to consider the various competing 
interpretations, of which there may be many, and to choose among them by finding the one that is the 
most semantically dense, and hence preferred. 
PREMO is a robust system for parsing natural language organized along the lines of an operating 
system. The state of every partial parse is captured in a "process control block" structure called a 
language object, and the control structure of the preference machine is a priority queue of these language 
objects. The language object at the front of the queue has the highest score as computed by a preference 
metric that weighs grammatical predictions, semantic type matching, and pragmatic coherence. The 
highest priority language object is the intermediate reading that is currently most preferred (the others are 
still "alive," but not actively pursued); in this way the preference machine avoids combinatorial explo- 
sion by following a "best-first" strategy for parsing. Each "ready" process in the system captures the 
state of a partial parse with priority given to each parse "process" on the basis of a preference semantics 
evaluation. The "time-slice" for each process is whatever is needed to move forward one word in a local 
process sentence buffer (where each process operates on a private copy of the current sentence). After 
every time slice, the preference/priority for the currently "running" parse is re-computed and the 
language object for that process is returned to the priority queue. The first process to emerge from the 
queue with its sentence buffer empty is declared the winner and saved. This strategy is both a run-time 
optimization and an application of the "Least Effort Principle" of intuitively plausible language process- 
ing. The parsing is robust in that some structure is returned for every input, no matter how ill-formed or 
"garden-pathological" it is. 
A principal advance in PREMO over earlier Preference Semantics work is that it overcomes the 
semantic localism of that work that allowed peculiar but telling counter-examples to be created. A better- 
known one is due to Phil Hayes (quoted in Boden 1977): "He licked the gun all over and the stock tasted 
good". A Preference Semantics system might well get "stock" resolved to "soup" rather than "gun stock" 
based on the local preferences of the last clause, because it had no overall GUN context, or indeed any 
way of expressing that. The frame movement of the Seventies was intended to provide such contextual 
theories (e.g. Minsky 1975) but it has not been generally accepted that it did so. The script-based parsers 
from Yale (e.g. Gershman, 1977) were created in the same topic-based spirit tended to fall into the oppo- 
site fault: of not understanding what was said unless it did wholly conform to expectations. 
221 
PREMO is intendeeed to, and does, meet both requirements: the bottom-up coherence of preference 
combined by weightings with the topic-subject codes provided for LDOCE (Longmans Dictionary of 
Contemporary English, Procter et a1.1978) within the lexical acquisition project at CRL mentioned earlier. 
MGR: MODEL GENERATIVE REASONING 
The global objective of this project is to investigate general mechanisms for symbolic problem 
solving in task environments where data are noisy and where the problems addressed require objects to be 
related in ways unanticipated by the designer. This objective is motivated by the frequently reported 
result that problem solvers developed using knowledge-based programming techniques are brittle when 
there is significant noise or novelty in the task environment. (Coombs & Hartley, 1987, 1988a,b, Fields 
et al. 1988a,b). 
Typical AI problem solvers employ fixed semantic relations (e.g. rules) to search for solutions. 
While the structure of these relations plays a significant part in the effectiveness of a problem solver, the 
manipulation of structure is not currently part of AI problem solving methodology. However, in the 
absence of reliable data or complete domain knowledge, i.e. when rules are unreliable, the problem solver 
will have to rely on the structural characteristics of interactions between data and knowledge for control. 
Problem solvers will, therefore, have to be sensitive to structural as well as semantic relations. MGR is a 
problem-solving architecture capable of this sensitivity. 
The basic MGR approach to reasoning is formally related to generalized set covering (GSC), which 
has been used effectively in automated problem solving for tasks where success is highly dependent on 
the quality of hypotheses (e.g., medical diagnosis). MGR extends GSC to its full generality by represent- 
ing problem solving objects in terms of graphs and using the mathematical theory of graphs for the 
specification of problem solving procedures. Graph theory also gives us an ideal vehicle for the explora- 
tion of structural relations. 
Starting with general set covenng as the foundation model, we propose to: i. expand the basic set 
covering operations to their full generality in terms of graph theory to produce a structural formulation of 
covenng operations, ii. relate the graph theoretic descriptions of set covenng operations to MGR opera- 
tions, and iii. prove consistency of the graph theoretic formulation of MGR and investigate the degree of 
its completeness. Following this we will investigate strategies for coping with noise and novelty in terms 
of manipulating relations between the graphs that represent problem data and knowledge. This will 
require the implementation of an explicitly graph theoretic version of the MGR problem solving architec- 
ture, with the capability of computing the structural characteristics of its computational objects and for 
using these as control parameters. These will then be used to address issues of tractability arising from 
problem solving within noisy and novel task environments through a series of experiments on (i) counter- 
ing the effects of noise in data, and (ii) enabling a system to adapt to novel relations between objects. 
The Abductive Mechanism in MGR 
Mechanizing any inference procedure runs the risk of being tied down to syntactic issues, whereas 
the real problems lie with the semantics of the techniques. Our technique is actually far more semantic in 
nature than the logic-based approaches of Charniak, Hobbs and the sort of approach advocated by 
Levesque (Charniak, 86, Hobbs, 89, Levesque, 89). In his paper, Levesque distinguishes the set-covenng 
ideas of Reggia and Allemang et al. (Nau and Reggia, 86, Allemang et al, 87) from the logic-based 
methods. For us the issue rests on the nature of an explanation and the utility of such a structure, were it 
to be available. A logical approach would call anything an explanation as long as the facts being 
explained are entailed by the explanation and some body of assumed knowledge. Schematically this is: 
222 
If there is a set of facts F and a body of domain knowledge D, 
then a formula E explains F iff 
EA D~F 
Mechanizing this involves computing the best possible E. This is rather like computing x in the simple 
equation 3 + x = 5. Calling '2' an explanation of '5' in the context of '3' and the whole of arithmetic is 
what this approach would tend to say. As Levesque points out there are a lot of assumptions built into 
this approach, not the least being that the knowledge in D must be complete (in some non-trivial sense) 
and consistent. For any problem solver that claims generality, these are hard requirements to meet. The 
set covering approach, however, relaxes the stringent requirements of a logic-based system. Instead of 
requiring a consistent knowledge base, we only require that there be coherent susbsets. Coherence in 
MGR rests on the idea of obeying selectional constraints within a type hierarchy. The abductive mechan- 
ism then finds these subsets by insisting that every term in the facts being explained is covered by at least 
one schema in the knowledge base (we could call these rules, but this would mislead, since covenng does 
not impose a preferred form on the components of the knowledge base). Cover can be computed accord- 
ing to parsimony (the best explanation is the simples0 or according to any other suitable criterion. Logi- 
cal entailment also uses parsimony to produce the simplest expressions. The set covenng idea is best 
expressed schematically as: 
Given a set of terms T, a set of facts F and a set of schemata D, 
each of which contain expressions of these terms, an explanation E 
is said to exist if f: 
t(E) = u t(D') where D' c D, and 
t(F) c t(E) 
The computation of E depends on finding the smallest subset D' that contains all of the terms in F. Since 
the relationship between the terms and the compenents of D that contain them is pre-computed in MGR, 
this resolves into an issue of minimizing the Boolean formula that is the conjunction of all the disjunc- 
tions each formed by coveting one term in F with the schemata that contain it. We are also currently 
looking at ways to achieve the same result with less than parsimonious covers to be used when parsimony 
is too strong a principle. The resultant explanations (the models in MGR) are coherent models of some 
aspect of real-world behavior, not merely loose connections of atomic formulae as they would be in a 
logic-based system. They can then be used for prediction (using simulation-like techniques, or deduction 
if you prefer) with real data. Far from being merely syntactic, the models in MGR bear more relationship 
to the models typically produced by a simulation language, rather than the structures employed by, say 
the ATMS of de Kleer (Reiter and de Kleer, 87, Levesque, op.cit.) The successes of set coveting in (Nau 
and Reggia, op. cit.) and (Josephson, 87) are a pointer to the approach work supporting of MGR. 
Formalization of the MGR Symbolic Problem Solver 
MGR employs four operators: Generalize Gn, Merge Mr, Specialize Sp, and Fragment Fr. These 
operators are implemented using the intensional join and project operators as defined by Sowa (op. cit.) 
as analagous with the extensional relational data-base operations join and project. The operators are 
defined as mappings over the sets F ~acts), D, (definitions) and M (models) as follows: 
Gn: PS(M) ---¢ M Mr: PS(M) ----> M 
Sp: M x PS(D) ~ PS(M) Fr: M × PS(F) ~ PS(M) 
The notation PS(X) denotes the power set of the set X. 
The MGR operators are implemented as follows. Gn takes as input a subset M" of M, and returns the 
project of the models in M'. Mr takes as input a subset M" of M, and returns the join of the models in 
223 
M'. Sp takes as input a subset D" of D and a model m ~ M, and returns a set M" of models, each of 
which is a join of m with all of the elements of D" with which it is joinable. Fr takes as input a subset 
F" ofF and a model m ¢ M. Fr then computes the joinfof the elements ofF'. Ifm has a set of discon- 
nected subgraphs gi "" gk such that project(gi ... g/c) =f, then Fr returns {gi ..... gk }" 
Informally, Sp takes a model as input, and generates a set of larger, more specialized models by adding 
definitions. The role of Sp is, therefore, to "glue" knowledge to facts to create models that cover the 
facts. Fr opposes Sp by breaking models into fragments in a way that preserves the information con- 
mined in facts, but may destroy information obtained from definitions. Fr thus generalizes definitional 
information, but not factual information. The role of Fr is to break apart models that do not cohere with 
all of the available the facts in order to generate fragments that can be recombined. Gn and Mr take sub- 
sets of models as input, and generate single models as output which are, respectively, less or more spe- 
cialized than the models from which they were generated. Gn is capable of generalizing both factual and 
definitional information; its role is to maintain coherence with the facts by removing over-specializations. 
Mr merges models whenever possible; its role is to generate models that have the greatest possible cover- 
ing power. All of the operators write over the model population; the set of models available for operating 
on, therefore, changes on every cycle. 
Current Work on MGR 
1. 
. 
. 
The techniques of model generative reasoning are being applied to situation analysis for TRADOC 
through the Army Intelligence School, at Fort Huachuca. MGR is being used to develop hypotheti- 
cal models of enemy intentions from an input of intelligence reports presented in terms of a set of 
'leading indicators'. Stored knowledge consists of schematic definitions of types of maneuver 
(both static and dynamic components) typically carried out by the enemy army. 
A formal analysis of the MGR operators is proceeding, leading towards a better understanding of 
their algorithmic complexity. Eventually this will lead to a parallel version of the MGR architec- 
ture. This work is supported by Harry Diamond Labs. through the ACT program. 
Work has progressed in characterizing the management of alternative hypotheses as generated by 
MGR through dynamic systems control techniques. We have also looked at the genetic algorithm 
as a variant that could be used to control MGR's search for the 'best' model. 
ViewGen: A POINT OF VIEW SHELL FOR REASONING ABOUT BELIEFS 
Introduction 
An AI system that takes part in discourse with other agents must be able to reason about the beliefs, intentions, 
desires, and other propositional attitudes of those agents, and of agents referred to in the discourse. This is especially so in 
those common situations when the agents' beliefs differ from the system's own. Thus, the question of how to represent 
and reason about propositional attitudes is central to the study of discourse. 
Clearly, this question is really about the beliefs, and so forth, that the system ascribes to the agents, on the evidence 
presented by the discourse itself and by context and prior information, since persons have no direct access to each others' 
mental states. We view the ascrpfion problem as being a fundamental one. It has been the focus of our past work on pro- 
positional attitudes and that of our immediate colleagues Afzal Ballim and John Barnden: (Ballim, 1986, 1987, 1988; Bal- 
lira & Wilks, forthcoming; Barnden, 1983, 1986a, 1986b, 1987a,b, 1988a,b, to appear; Wilks & Ballim, 1987, 1988, in 
press (a,b); Wilks & Bien, 1979, 1983). Ascriptional reasoning is profoundly dependent on the communicative context, 
general information that the system has about the world, and special information the system has about the agents at hand. 
Moreover, there are major pragmatic features of discourse, such as speech acts, metaphor, and the determination of the 
intensional entities in play in a discourse, that any system for ascribing beliefs to agents must address. We would go 
further, and assert that even the most apparently superficial aspects of natural language understanding depend on belief 
224 
ascription: such as prepositional phrase attachment. Anyone hearing a sentence with the all-too-familiar structure: 
lie told his mother about the murder in the park. 
will interpret it differently according to whether he believes the speaker believes there was a murder in a park and that the 
speaker believes the hearer believes that too. The function of our basic program Viewgen is to create, or as we shall call it, 
ascribe, environments into which appropriate beliefs can be segregated so that parsing and reasoning can be done in that 
limited environment. 
We have described the basic algorithm in Viewgen in the publications above, and we address the issue of basic pars- 
ing issues seen a belief phenomena elsewhere. Here our purpose is simply to review the basic ascription mechanism and 
then show its extension to the phenomena such as the identification of intensional objects. 
In interpreting an utterance by an agent, the system must ascribe a speech act to that agent; and doing that is a matter 
of ascribing specific intentions, beliefs, desires, expectations and so on to the agent. Thus, speech act ascription is an 
important special case of ascriptional reasoning. That speech-act considerations make reasoning about propositional atti- 
tudes essential for the computational modelling of discourse has been established at least since the work of Perrault and his 
colleagues (e.g. Perrault & Alien, 1980). A major difference between that work and ours is that they took the content of 
belief environments to be already established, whereas our approach is based on the real-time computation of the contents 
of such belief environments. 
The work of Maida (1984, 1986) shares many of the concerns of the current work: his diagrammatic representations 
of nested beliefs are isomorphic to those of Wilks & Bien (1979) and Shadbolt (1983). His concerns are with the problem 
of shared reasoning strategies between believers and how, for example, you could establish that a dialogue partner also 
used modus ponens. We argue, on the contrary, that this phenomenon is best handled by general default assumptions, as are 
the concrete contents of belief. No finite set of dialogue observations ever could establish conclusively that another beli- 
ever was using modus ponens. That being so, concentration on such issues that are not susceptible to proof seems to us 
only to delay the central issue, which is how to infer heuristically the actual beliefs of other believers. Maida (1983) is 
also concerned with the very important, and we believe quite separable issue, of a heuristic rule for identifying intensional 
individuals under different descriptions. Konolige's (1983) work has strong similarities to that just noted; he considers 
what he calls views, for which he writes e.g., v=John,Sue,Kim, to mean John's view of Sue's view of Kim's beliefs. But he 
has no effective construction for the content of such views. Rather, he is concerned with giving an account of limited 
deduction in such views, an important process, but not relevant to issues of constructing individuals' views. Dinsmore 
(1987) has been concerned with what he terms the "algebra of belief spaces" but, although the term is highly general, the 
focus of his attention is always in fact the notions of presupposition and counterfactuals, which are not notions we treat 
explicitly here, and his treatment of them may well be compatible with our own general approach. 
ViewGen: The basic belief engine 
A computational model of belief ascription is described in detail elsewhere (Wilks & Bien, 1979, 1983; Ballim, 
1987; Wilks & Balfim, 1987; Ballim & Wilks, forthcoming) and is embodied in a program called ViewGen, two versions 
of which have been written in prolog by Afzal Ballim. The basic algorithm of this model uses the notion of default reason- 
ing to ascribe beliefs to other agents unless there is evidence to prevent the ascription. Perrault (1987, forthcoming) and 
Cohen & Levesque (1985) have also recently explored a belief and speech act logic based on a single explicit default 
axiom. As our previous work has shown for some years, the default ascription is basically correct, but the phenomena are 
more complex (see below) than are normally captured by an axiomatic approach. 
ViewGen's belief space is divided into a number of topic specific partitions (topic environments). These environ- 
mcnLs can be thought of as a les s permanent version of frames (Minsky, 1975; Charniak, 1978) or more suitably in terms 
of Wilks (1978) as pseudo-texts (henceforth PTs). In effect, a PT is a set of unsorted, unrefined items of knowledge. These 
PTs are general items and are not only stored for individual human beings, but also for groups of humans, objects, and 
abstract ideas. Their hierarchical and inheritance relations are discussed in Wilks (1978) and Ballim & Wilks (forthcom- 
ing). We justify the general notion of explicit environment in the next section. 
Viewgen is a program that generates a type of environment known as a viewpoint. A viewpoint is some person's 
beliefs about a topic. Within ViewGen, all beliefs are ultimately beliefs held by the system (e.g., the system's beliefs about 
France, what the system believes John believes about cars, etc.) and so, trivially, fie within the system's viewpoint. The 
225 
system's view of some topic (say, atoms) is pictorially representeed as: 
~tom \[ Light (atom) 
Small (atom) 
gystem 
This diagram contains two types of environments: First, there is the box labelled with "system" at the bottom. This 
is a "believer environment" or "viewpoint." Viewpoints contain "topic environments," such as the box labelled with 
"atom" at the top of it. A topic environment contains a group of propositions about the "topic." So, for example, the 
above diagram conveys that the system believes that atoms are light and small. Topic boxes are motivated by concerns of 
limited reasoning (see Section 5.1 on relevance, and also Wilks & Bien, 1983). In short, it is envisaged that reasoning takes 
place "within" a topic environment, as if it were the environment of a procedure in a programming language. 
Within ViewGen, environments are dynamically created and altered. ViewGen's "knowledge-base" can be seen as 
one large viewpoint containing a large number of topic environments, with each topic environment containing a group of 
"beliefs" that the system holds about the topic. The reader should note that each proposition in a topic environment has at 
least one symbol identical to the name of the topic. Each such proposition is therefore explicitly about the topic. There 
are, however, implicit ways in which a proposition can be "about" (or "relevant to") a topic. The simplest cases are gen- 
erated by inheritance in the usual way: for example, if John is a man, then any proposition in a "man" topic environment 
is implicitly or indirectly about John. However, we choose not to put such a proposition in the John topic box, and will 
justify that decision in the later section on relevance (5.1 below). Again, the same proposition can occur in more than one 
box, as would the expression asserting that an elephant was larger than an atom, for it is about both atoms and elephants, 
and should appear under both topics.. 
If the "topic" of a topic environment is a person (someone capable of having beliefs themselves), then the topic 
environment may contain, in addition to the beliefs about the person, a viewpoint environment containing particular beliefs 
held by that person about various topics. Normally and for obvious reasons of efficiency, this is only done for those beliefs 
of a given person that are, as some would put it, reportable, where that will often mean beliefs that conflict with those of 
the system itself. For example, suppose the system had beliefs about a person called John who believes that the Earth is 
flat. This would be pictorially represented as follows: 
--John Man (John) 
Height_of (John) is 6-feet 
Earth Flat (Earth) I 
John 
system 
The John viewpoint, shown as the box with "John" on the lower edge, is a nested viewpoint, as it is enclosed 
within the system viewpoint shown (through an intervening topic environment about John, shown as the box with "John" 
on its upper edge). For simplicity, in the diagram of a nested viewpoint we often leave out propositions that are not in the 
innermost topic box: in the above example we would leave out the beliefs that John is a man, and that he is six feet tall. 
Further simplifying this, we often leave out all but the innermost topic box, leaving only it and the viewpoint boxes. 
226 
Hence, the above diagram would be simplified as: 
Earth Flat (Earth) 
-John 
~ystem 
The system stores its own beliefs, and the beliefs of other agents that differ from the system's own beliefs. Others' 
viewpoints are generated on demand, a position we find both computationally and psychologically more plausible than the 
"prestored nesting" view mentioned above (Section 1). The process of generating a viewpoint can be regarded as an amal- 
gamation mechanism that ascribes beliefs from one viewpoint to another (or, "pushing one environment down into 
another"); ascribing certain beliefs, transforming some, and blocking the ascription of others. 
The simplest form of this algorithm, described in Wilks & Bien (1979, 1983), is that a viewpoint should be generated 
using a default rule for ascription of beliefs. The default ascnptional rule is to assume that another person's view is the 
same as one's own except where there is explicit evidence to the contrary. An important special case of such examples is 
when the topic is the same as the agent, and we can illustrate with that. Suppose that at a certain stage in dialogue the sys- 
tem, acdng as a medical diagnostician, has the view of John that he is not healthy, and is six feet tail, while he believes 
himself to be healthy. We shall delay until the section on intensional objects below any question as to whether the 
system's John and John's own could be different entities. This basic situation is represented pictorially as follows: 
--John 
Height_of(John) is 6-feet 
not Healthy(John) 
John HeAthy(John) \[ 
John 
system 
The more complete environment for the system's view of John's view of himself can be generated by trying to ascribe the 
beliefs from the system's topic environment about John to the topic environment about John within John's viewpoint 
(where, as always, the last expression must be glossed as "the system's view of.."). One of the two beliefs survives the 
attempt but the other is blocked, giving the following state: 
227 
--John 
Height of(John) is 6-foot 
not Healthy(John) 
  --John- Height_of(John) is 6-foot Healthy(John) 
John 
~ystem 
This can be pictured in the simplified (or as we shall call it, compressed) manner as 
F John \] 
Height of(John) is 6-foot 
Healthy(John) 
-John system 
We see that in examples of this sort, where the topic is also the agent into whose environment an ascription is being 
attempted, propositions in an outer topic environment E are pushed inwards into a topic environment (for the same topic) 
within a viewpoint nested within E. Such inward pushing is central to our later observations of intensional identification 
and metaphor. 
The above example demonstrates the basic ascription algorithm and a simple case of ascriptions being blocked. How- 
ever, belief ascription is a far more complex phenomenon and the key to our method is the delimitation and treatment of 
cases where the default algorithm is incorrect. But even the default algorithm itself requires for its operation a notion of 
blocking beyond that of explicit contradiction: for example, the proposition Healthy(John) should be able to block 
Sick(John)", ff Sick and Healthy are known to be incompatible predicates. Similarly, we appeal below to blocking that 
arises from incompatible function values, as in the blocking of "'Eye-colour(Frank) = Green" by "Eye-colour(Frank) = 
Blue". The more significant complication is that there is an entire class of beliefs that require the opposite of the default 
ascription rule given above. We call these atypical beliefs and they include technical expertise, self-knowledge (itself a 
form of expertise), and secrets. For example, befiefs that I have about myself, such as how many fillings I have in my teeth, 
are beliefs that I would not normally ascribe to someone else unless I had reason to do so (if, say, the person, to whom l 
was ascribing the belief, was my dentist). A representation based on lambda expressions is used in dealing with atypical 
beliefs, and is described elsewhere (Ballim, 1987; Ballim & Wilks, forthcoming; Wilks & Balfim, 1987), and follows a 
suggestion originally made by McCarthy and Hayes (1969). This combination of a basic default ascription rule, augmented 
by a mechanism for dealing with atypical belief, is an original algorithm and has not, to our knowledge, been described or 
tested elsewhere in the literature. 
The essential feature of this notation is that lambda-expressions, as in the following example 
Cure-for (tuberculosis) is ( (~.x (Cure -for x )) tuberculosis ) 
can only be evaluated by qualified believers (e.g., physicians in this case) in appropriate environments. Yet anyone 
can believe the Fregean triviality that the above sentence expresses unevaluated (and it is vital that they can) but a non- 
228 
trivial interpretation can only be placed on it by those who can evaluate the lambda expression in an environment. In a 
crude sense therefore, the lambda-expressions correspond to intensional representations and their evaluations, when avail- 
able, to extensions, or at least other intensions in those situations where the evaluation of such an expression produces yet 
another lambda expression (see also Maida 1983). 
The above expression, for example, might evaluate to another lambda expression using a predicate Sulfonamide-drug, 
for whose evaluation a particular drug might be an appropriate. Each evaluation would require an environment whose 
"holder" was qualified to perform it. It is really this possibility of successive evaluations of expressions that justifies the 
abstraction capacity of the lambda notation, since it could well result in expressions, such as a conjunction of predicates, 
for which there is no single predicate name. In conclusion, it is this general mechanism in Viewgen that deals with the 
problem of the over-application of the main default rule of ascription, since the ascription of unevaluable expressions, 
about, say, the number of my own teeth to you, does not lead to undesirable results. 
lntensional Objects and Their Identification 
It is natural in a system of partitioned environment notation to treat environment boxes as intensional objects: to treat 
the Jim-object, pushed down into the Frank-object, as not just yielding by computation a set of beliefs that is Frank's- 
view-of-Jim, but also as a sort of intensional object we might call Jim-for-Frank. Let us now consider two simple cases of 
intensional objects and see how the basic default algorithm deals with them: 
CASE 1 (or Two-for-me-one-for-you): The system believes that Frank and Jim's-father are two people, but 
that Mary, whose point of view is being computed, believes them to be the same person. 
CASE 2 (or One-for-me-two-for-you): Mary believes Frank and Jim's-father to be separate people, whereas 
the system believes them to be the same individual. 
Scenarios such as these are common, and arise over such mundane matters as believing or not believing that Johns-house is 
the same as the house-on-the-corner-of-X-and-Y-streets. 
Two- for -me-one-for-you 
Processing of first case will begin with the system having three topic environments: for Frank, Jim's-father and 
Mary. Two questions that arise are: What intensional object(s) should Mary's viewpoint contain? And what should be the 
beliefs about those intensional objects? Let us say that the system has beliefs about Frank and Jim's father as shown below. 
--Frank 
Male (Frank) 
Eye_colour (Frank) = Green 
i Jim's-father Male (Jim's -father ) 
Tall (Jim's -father ) 
Eye_colour (Jim's-father) = Blue 
~ystem 
The first question can be rephrased as "given certain intensional objects in one viewpoint (the system, in this case), what 
are the corresponding intensional objects in the system's version of another viewpoint (Mary's)?" Extending the normal 
default rule for belief ascription to cope with intensional object ascription, we would say, naturally enough, that intensional 
objects in one environment directly correspond to intensional objects in another environment, unless there is counter evi- 
dence to believing this. This notion of correspondence of intensional objects between environments can be expressed as 
beliefs, but these beliefs must be of a type different from those we have previously discussed. 
Although it is true that beliefs about intensional correspondence are beliefs about someone, for instance that the sys- 
tem believes about Mary that her intensional objects correspond to its own in a particular way, they clearly cannot be 
229 
beliefs of Mary herself. Referring to our particular example, we say that the system's belief that Mary has a single inten- 
sional object corresponding to the system's Frank and Jim's-father objects is a belief for Mary. 
These beliefs of intensional correspondence are beliefs of one agent that describe how the agent's intensional objects 
correspond to intensional objects for another agent. Thus, they shall be referred to as beliefs for an agent P, and we shall 
place them in a separate partition as diagrams below will illustrate. A reader might feel here that there is no need for an 
additional box type, but further reflection may correct this. This relationship cannot be captured by co-reference (as 
opposed to what we are calling correspondence) since within these intensional contexts entities are not coreferential, ex 
hypothesi. The For-box (see Fig. below, where it should be noted that the ordered arguments that correspond for-Mary are 
precisely not Konolige's "views", even though the notation is the same. 
In the case at hand (CASE 1), Mary's viewpoint ends up containing a single intensional object O (a topic environ- 
ment) corresponding both to the system's Frank object (topic environment) and to the system's Jim's-father object (topic 
environment). The question now is to decide what should be put inside the environment O. One possibility is to combine 
the information in the system's Frank and Jim's-father objects symmetrically, removing any conflicting information. In the 
present case, this would result in O stating that Frank/Jim's-father is male and tall, but stating neither that he has blue eyes 
nor that he has green eyes. However, we claim that in realistic situations it will often be more appropriate to take an 
asymmetrical view, in which we choose to give precedence either (i) to the information in the system's Frank object over 
the information in the system's Jim's-father object, or (ii) vice versa. Choice (i) reflects the presumption that there is a 
stronger or closer correspondence between Mary's idea of Frank and the system's idea of Frank than there is between her 
idea of Frank and the system's idea of Jim's father. This difference of closeness would be plausible, for instance, if the 
system regarded Mary's view of Frank as being essentially the same as its own except in making the (presumed) mistake of 
taking Frank to have the property of being Jim's father. 
Choice (ii) reflects the converse presumption, which would be most likely to arise from a hypothesis that Mary is 
focussing on the person-description "father of Jim", and that she happens to hold that this description identifies Frank. 
Our claim is that in realistic situations there is more likely to be a reason for making one of these choices than to take the 
symmetrical approach. 
We handle the asymmetrical choices as follows. For choice (i), the system constructs an intensional object O called 
"Frank-as-Jim's-father" inside Mary's viewpoint. This object is so-called because it is, so to speak, "the Jim's-father 
view of Frank" (according to Mary). Notice that we are not here saying that the object is the view of Frank that Jim's 
father holds (according to Mary); rather, the object is a view of Frank that is coloured by the idea that he is Jim's father. 
This way of regarding Mary's intensional object O is directly reflected in the proposed process for constructing O, as we 
shall see in a moment. Mary's Frank-as-Jim's-father object, O, arises in two stages, as follows. 
Stage 1: The system's view of Frank as Jim's father is created. This view is created as a topic environment O' inside the 
system's viewpoint. The creation occurs in three substages: 
(la) Initially, a copy of the system's Frank object (topic environment) is placed inside the Jim's-father object (topic 
environment), as shown in the next figure. Intuitively, the idea so far is that we have not yet tried to identify Frank 
as Jim's father, but have merely established a view of Frank that is, so to speak, in the context of Jim's father. That 
context does not have an effect until substage (lb). 
(lb) We now respect the required identification of Frank as Jim's father. We try to push the beliefs in the system's 
Jim's-father object into the Frank object embedded within it, using the ordinary default rule, with the slight 
modification that Jim's-father is replaced by Frank in a pushed belief. Thus, the beliefs that Jim's father is male and 
is tall are successfully pushed in (although the former happens to duplicate a belief already in the embedded Frank 
object), but the belief that Jim's father has green eyes is blocked by the blue-eye belief already in the embedded 
Frank object. 
(lc) The final substage in constructing the system's Frank-as-Jim's-father object O" is to pull out the Frank object that is 
embedded within the Jim's-father object, making it into an object (topic environment) O" at top level within the 
system's viewpoint. In doing this we replace the "Frank" topic-name by the name "Frank-as-Jim's-father", and 
similarly change the Frank symbols inside the environment to Frank-as-Jim's-father. The following diagram shows 
the result, with the arrow notation indicating the pull-out process. 
230 
-Jim "s-father 
Male (Jim's -father ) 
Tall (Jim's -father ) 
Eye_colour (Jim's-father) = Blue 
Frank Male (Frank) 
Tall (Frank) 
Eye_colour (Frank) = Green 
$ 
--Frank-as-Jim "s-father 
Male (Frank-as -Jim's -father ) 
Tall (Frank -as -Jim's -father ) 
Eye_colour (Frank-as-Jim's-father) = Green 
I Mary I 
system 
Stage 2: We now ascribe the system's beliefs about Frank as Jim's father -- that is, the beliefs inside O" -- to Mary, once 
again using the ordinary default rule. On the assumption that there is no prior information about Mary's view of 
Frank/Jim's-father (e.g. that his eyes are brown), all that will happen is that a copy O of O" will be created inside the 
Mary viewpoint, giving the revised Mary-viewpoint shown in the following figure. 
--Frank-as-Jim "s-father Male (Frank-as -Jim's -father ) 
Tall (Frank -as -Jim's -father ) 
Eye_colour (Frank-as -Jim's -father ) = Green 
Mary 
system 
If we had had prior information from discourse input that Mary believes the person's eyes to be brown, then there would 
already have been a Frank-as-Jim's-father object (topic environment) O inside Mary's viewpoint, and the beliefs in O' 
would all have got pushed into that object except for the green-eye belief. If the system had decided to give precedence to 
the Jim's father information rather than to the Frank information in doing the intensional identification (that is, if it had 
made choice (ii) above) then it would have generated the following state by an analogous process: 
231 
i Jim's-father.as-Frank Male (Jim's -father-as -Frank ) 
Tall (J.im s -father -as -Frank ) 
Eye_colour (Jim s-father-as-Frank) = Blue 
Mary 
system 
It might be thought that a symmetric intensional object, with the feature differences appearing as disjunctions (e.g 
Eyecolor Blue OR Green) would be appropriate as a construct for the Mary environment. We suggest that this is in fact 
psychologically less plausible, and that subjects do construct stronger, and more refutable, hypotheses. A final important 
thing to notice about the process described above is that the crucial pushing of information from the Jim's-father environ- 
ment into the embedded Frank environment (or vice versa) is exactly the type of "inward" pushing used in the basic 
examples with which we illustrated basic belief ascription in section 2. 
In Sections 6 and 8 we shall seek to show that belief ascription (e.g. Jim's-father's-view-of-Frank), intensional 
identification (e.g. Frank-as-Jim's-father), and even metaphor are all different forms of a single fundamental computational 
process. 
Treatment of CASE 2: One-for-me-two-for-you 
In the second case, where the system believes in one individual but Mary two, then the natural computation of. 
Mary's view of either Frank or Jim's-father is simply to push the system's single representation, changing "Frank" to 
"Jim's-father" as necessary. This is shown in the following figure. 
--Frank 
Frank = Jim's-father I Eye_colour (Frank) = Green 
--Mary 
--Frank 
Male (Frank) 
Eye_colour (Frank) = Green 
  -Jim's-father Male (Jim's -father ) Eye_colour (Jim's-father) = Green 
Mary. 
Frank correspond to <Frank, Jim's-father > 
for Mary 
232 
These are not merely aliases, but are best thought of as dual ascriptions, performed by making two identical copies. 
Further information about Mary's beliefs would then presumably cause the contents of the two environments to differ, 
since she presumably has at least some differing beliefs about what she believes to be distinct individuals. 
Discussion 
Neither CASE 1 nor CASE 2 turns out to be particularly problematic, and the situation is no different if the entities 
about whose identity there is dispute are non-believers rather than believers. Those would be like the classic but natural 
cases such as a difference between dialogue participants as to whether Tegucigalpa and Capital-of-Honduras are, or are not, 
the same; or as to whether Rome or Avignon should be identified with City-of-the-Popes. 
More difficult cases, that bring in all the panoply of philosophical distinction and discussion are those conventionally 
discussed under the de re/de dicto distinction. One type is the following: the system reasonably believes Feynman to be a 
famous physicist but encounters Frank who, on the strength of a single appearance on the TV screen, believes him to be a 
famous TV performer. For the sake of this example, it is essential to accept that the two occupations are incompatible. 
Suppose the discussion now forces the system to construct its view of Frank's view of Feynman. Now, there will be no 
point at all in performing that computation unless the system believes Frank's beliefs to be de re. Frank no doubt considers 
his own beliefs de re, as we all do. The crucial thing is that the system believe this, and the test would be some proposition 
in the Frank environment, and ABOUT Frank, equivalent to ("Feynman" names Feynman). If that is not present, the sys- 
tem should infer that Frank has another person in mind: that his beliefs are de dicto FOR THE SYSTEM, and hence any 
pushdown computation would be pointless. 
Consider the relation of this example to the former, simpler, cases, where the system can identify or separate distinct 
environments. This last case would be like that if the system knew which non-Feynman individual Frank was confusing 
Feynman with, perhaps Johnny Carson. In that ease, the system could perform a pushdown, even though it believed Frank's 
beliefs to be de dicto as far as Feynman was concerned, for they would be de re with respect to Johnny Carson. The system 
could then push Carson into Frank, while changing the resulting environments name to "Feynman". To summarize, the 
absence of ("Feynman" names Feynman) in the Frank environment is only a reason for not pushing down Feynman, but 
leaves open the possibility of some other de re push down. 
UNIFYING VIEWGEN AND MGR 
Both ViewGen and MGR are systems producing multiple models or views of a situation, but only in the ViewGen 
system does that multiplicity depend on separate environments representing the views of others explicitly: i.e. the multipli- 
city of ViewGen's models are to be mapped to different agents, while in MGR the multiplicity of models are the alterna- 
tives present in a single agent. Hence the multiplicities of the two theories nest in a natural way, and, we hope to show 
(although our project is still in its early days) combine to filter each other's products, rather than multiply them. 
In its earliest incarnation (Wilks & Bien 1983) ViewGen was also seen as a provider of multiple alternative models 
of text or message content. So, in the following diagram, adapted from that paper, we can consider a message p in a text 
(first frame of diagram) that mentions the agents user, john and frank.The analysis system can then be considered (by 
processes not set out in detail here) as creating the two nested environments shown in the two succeeding frames, each of 
which contains a representation of p, plus whatever beliefs and goals are appropriate to that inner environment, hence, 
given that those two sets differ significantly, the consequences, in those two environments, of p and whatever other beliefs 
are in there, will, taken together, be significantly different, and will be effectively different inteqaretations of the original 
message p. This ambiguity is not syntactic or semantic ambiguity but pragmatic ambiguity against diffenng assumptions. 
233 
P 
agents =john drank ,user 
text 
--user 
F frank P 
user 
~ystem 
--user 
F frank P 
john 
~ystem 
The natural application of that in, say, an Army battle scenario where message-derived intelligence reports and stored 
doctrine provide the data for model construction is one where the Blue has reason to believe the Red commander is cru- 
cially wrong about, say, the position of 22nd Tank Division, a belief that may be accidental or by Blue design and is 
derived from a parsing of Red messages: 
--Tank-corps-C. 
Position -is -(a ,b ) 
Allof RedArmyDoctrineManuaI [ 
-Tank-corps-C Position -is -(A ,B ) 
red-commander 
blue-commander 
234 
The ability of the Blue Commander to compute effectively in this scenario also depends on Blue's access to Red 
Army doctrine manuals, which we have to imagine having been reduced to machine-readable form in some way. An impor- 
tant variant on this theme, for which ViewGen provides the representational mechanism, is a situation where, in Blue's 
view, the new commander has confused and conflated two Blue Tank Divisions, say 1st and 2nd. The problem then, as we 
noted in the ViewGen section above when discussing Jim-as-Frank's-father is how to have a representational entity for an 
object that Blue, in this case, does not believe in the existence of, namely the wrongly conflated 1st and 2nd Tank Divi- 
sions. It will be remembered that it is crucial, in our view that this be an asymmetric representation, not some simple 
union of properties. 
--Ist-Tank-Corps 
l O0-strong ( Ist-Tank---Corps ) 
Position-A -B ( lst-Tank-Corps ) 
--2nd-Tank-Corps 
55-strong (2nd-Tank .--Corps ) 
Position -a-b (2nd-Tank-Corps ) 
$ 
--2nd-Corps-as-lst-Corps 
55-strong (2nd-Corps-as-lst-Corps ) 
Position -a -b (2nd---Corps -as - l st.--Corps ) 
Red-general 
Blue-general 
Our principal architectural aim is to combine these techniques into a system that can play a role within the standard 
Army intellience analysis and battle management infrastructure that can be diagrammatically represented as follows: 
DIAGRAM: DATA AND GUIDANCE 
The G2 is an officer whoses job is to evaluate various types of field data (i.e. weather maps, field observations, sen- 
sor outputs, etc.) and utilizing the 'Red Doctrine' as a set of guidelines, provide explanations or plausible situations to the 
235 
commander. In other words, his job is to integrate the field data with the Red Doctrine and inform the commander of all 
possible explanations of the data. These explanations should be well ordered (i.e. weighted), in order to better inform the 
commander. 
The Red Doctrine is a synthesis of all experiences or facts regarding the operations of the Red Army. Theoretically 
the Red Doctrine is written to anticipate all possible situations, including elements of deception and provide appropriate 
responses for them. We see our project as developing an AI system to act as another computer-based tool for the G2 to 
use. The primary system under development, the Air-Land Battle management (ALBM) system has proven to be strong 
when it comes to handling troop movements in different terrain and in planning a response, but weak in hypothesis genera- 
tion of enemy intentions. 
The goal of this project is for MGR to produce the range of all possible situations (called models in the case of 
MGR) oased on the data and the Red Doctrine. The models will be hypothetical in nature because the input data are noisy, 
thus requiring interrpretation to explain them. MGR looks for coherent subsets of data to produce interpretations. The 
more uncertain the data the greater the number of models, and these models may have components that are incoherent with 
other models. In general multiple models may originate from any of three possible sources: 
The data is incoherent. 
The Doctrine is too general, thus allowing coherent alternative interpretations. 
MGR can generate alternatives from the ambiguity inherent in existentially quantified descriptions. 
The assembled information produced by a G2 and the models produced by MGR should be consistent for the same 
data. Each model or set of alternate models can be used as the basis of prediction of future enemy movements, tactics etc. 
The use of models for both explanations and prediction is an essential element in MGR's use.Let us finally, and tentatively, 
turn to possible ways of integrating these techniques into a more complex system of command and advice. 
We envisage a system in which the PREMO parser takes textual (e.g. Red or Blue Army doctrine) or message input 
and produces an output in Conceptual Graph format. This output is fed to MGR which generates multiple models to 
explain or represent the input. The output of MGR is then represented when required but not in general as different nested 
environments in ViewGen and then evaluated to determine which surviving environment best fits the data (i.e. is most 
coherent) or what to do if things go wrong. 
A suggestion for integrating ViewGen and the PREMO parser into the existing MGR project is shown below. 
g2 ~analysis I I Corps ' I \[ -~- I Commander ". • text I 
i Doctrine ~ ~ .// . 
I D°ctrine I~ f~ "/ IComm- I \[ ~ ~ /i !ander 
-,! oo,oooce i t 
236 
In the diagram, there are several believers at work interpreting data and producing repsonses. Roughly speaking, and 
without any automated help, the sequence is as follows: the Corps G2 assembles all available data and produces a set of 
(weighted) interpretations as to Red's intentions. These are produced from the incoming field data and from his knowledge 
of Red doctrine, which stipulates default tactics and maneuvers in a wide range of different situations. The G2's reports 
are received by the Corps commander and he chooses the basis for a Blue plan based on the situation analysis and his 
knowledge of Red doctrine, and, for planning purposes, Blue doctrine. The commander issues "commander's guidance" 
reports which are then passed on to the division commanders as orders. The division commanders then re-interpret these 
guidances in order to make their plans, in the light of their knowledge of both Red and Blue doctrine. 
There are several belief systems at work, and plenty of room for inconsistency to occur. The clearest area for this to 
arise is not in the different interpretations of Red doctrine, but in intepreting the input. In the case of the G2 this is inter- 
petint the data. For the Corps commnader, this is interpreting the G2's reports, and for the division commanders this is 
interpreting the Commander's guidance. These different interpretations stem from different applications of the same 
knowledge to the input data. The source of the differences could be postulated to be due to various psychological states 
(mistrust, pig-headedness etc.) but even in the "objective" case different weighing of evidence according to personal 
experience is bound to occur. 
MGR is "objective" since it has no possible psychological leanings, but it still only produces one possible set of 
views of the battle. The hope is to make MGR produce more objective alternatives than analysts commonly do, but its 
inclusion in this chain of command is only one factor in the plans actually produced and carried out by the army. How- 
ever, its strength is that it is, at base, algorithmic, and therefore its output is reproducible. It can be therefore be used as a 
benchmark against which other agents' beliefs can be measured. One way to do this is to nest MGR's "beliefs" at the 
core of a nested sequence of belief spaces. Successively these would be G2, Corps Commander, Division Commander. 
These spaces would be maintained by ViewGen by importing the relevant changing data at each level, and recompufing the 
set of beliefs accordingly. The nesting of belief spaces then allows for consistency checking between the agents, using 
MGR's output as a base case in order to evaluate the inconsistencies as they arise. 
As an example, consider a piece of intelligence that places a Red MRD (Motorized Rifle Division) close to a known 
avenue of approach to a potential battle front. MGR would say (objectively) that the division would end up at the ffront on 
that avenue simply based on current activity (direction and speed of movement) and the intervening terrain. It would also 
hypothesise that this division could either play a main or support role in the attack, in conjunction with the other divisions 
known to be in the area. The G2's job is essentially to produce such impartial explanations, albeit weighted by likliehood. 
Now postulate that the Corps commander has detected (perhaps much earlier in the battle) that this particular division has 
been behaving strangely (he may, for instance know the personal preferences of the division commander - something that 
MGR would not know). He may therefore choose to ignore the G2's analysis and assume, for instance that the division 
will play no major part in the battle - i.e. for planning purposes it does not exist. His guidance will the reflect this 
interpretation, providing one level of inconsistency between this guidance and MGR's (or the G2's) output. Now assume 
that one of Blue's division commanders is a little "trigger happy" and is itching for a skirmish. He could detect the Corps 
commander's discounting of the stangely behaving Red division, and choose to bend the orders sufficiently to force an 
encounter with it. His orders will then reflect another level of inconsistency, this time with the Corps commander's gui- 
dance. He may, for instance, believe that the rogue division is an easy target, even though his orders are to ignore it. 
ViewGen can handle this multiply nested set of inconsistencies, and with MGR ouptut as the impartial 'base base' 
can give judgements as to the direction of the inconsistencies, withiout prejudging the fights or wrongs of the case. 
Without MGR the only judgements that can be made are relative, and might be of little use in a real case. Several ques- 
tions come to mind as to how to proceed. Firstly, whereas MGR produces interpretations according to a default set of 
ideas about Red doctrine and how to handle raw data, the belief spaces of individuals is much harder to capture and to 
model. ViewGen cannot create beliefs, only modify statements according to a system of belief already captured and 
represented. Secondly, The text messages that are handed down the chain of command may not turn out to adequately 
reflect the influence of the agent's belief space, and that another round of interpretation may be called for before they can 
be used by ViewGen. Our preliminary examination of sample Commander's guidance messages is inconclusive in this 
regard. Thirdly, although Red doctrine is reasonably easy to represent, beliefs about how it should be used may not be. 
MGR presents a fairly neutral view of this, preferring to err on the side of producing too may models that can be filtered 
out later, rather than missing a potentially important alternative. These differences in inference strategy can be incor- 
porated into ViewGen's evaluation mechanisms, but exactly how is an issue for further research. 
237 
References 
Allemang, D., Tanner, M., Bylander, T., Josephson, J. (1987) On the computational complexity of hypothesis assembly. In 
Proceedings of lJCAl87, Milan, Italy. 
Ballim, A. (1986) Generating point of view. Memoranda in Computer and Cognitive Science, MCCS-86-68. Computing 
Research Laboratory, New Mexico State University, Las Cruces, New Mexico 88003, USA. 
Ballim, A. (1987) Subjective ascription of belief to agents. In Advances in artificial intelligence. Edited by J. Hallam & C. 
Mellish. Chichester, england: John Wiley & Sons. 
Ballim, A. (1988) A language for representing and reasoning with nested belief. First Annual Irish Conference on Artifi- 
cial Intelligence and Cognitive Science. Dublin, Ireland. 
Ballim, A. and Wilks, Y. (forthcoming) Artifcial believers. Hillsdale, N.J.; Lawrence Erlbaum Associates. 
Barnden, J.A. (1983) Intensions as such: an outlined. Proceedings of the 8th International Joint Conference on Artificial 
Intelligence. Los Altos, Calif." Morgan Kaufmann. 
Barnden J.A. (1986a) A view point destinction in the representation of propositional attitudes. Proceedings of the 5th 
Nat. Conf. on Artificial Intelligence (AAAI-86). Philadeophia: Morgan Kaufmann. 
Barnden J.A. (1986b) Imputations and explications: representational problems in treatments of propositional attitudes. 
Cognitive Science, 10, (3), pp.319-364. 
Barnden, J.A. (1987a) Interpreting propositional attitude reports: towards greater freedom and control. In Advances in 
artificial intelligence - II. Edited by B. du Boulay, D. Hogg & L. Steels. Amsterdam: Elsevier (North-Holland). 
Barnden J.A. (1988a) Propositional attitudes, commonsense reasoning, and metaphor. Proceedings of the lOth Annual 
Conference of the Cognitive Science Society. Hillsdale, N.J.: Lawrence Erlbaum Associates. 
Barnden, J.A. (1988b) Propositional attitudes, polysemy and metaphor: initial report. Memoranda in Computer and Cog- 
nitive Science, MCCS-88-139. Computing Research Laboratory, New Mexico State University. 
Boden, M. (1977) Artificial Intelligence and Natural Man. Hassocks, UK: Harvester Press. 
Boguraev, B. K. (1979) Automatic Resolution of Linguistic Ambiguities. Technical Report No.11, University of Cambridge 
Computer Laboratory, Cambridge, England. 
Carter, D. (1984) An Approach to General Machine Translation Based on Preference Semantics and Local Focussing. In 
Proceedings of the 6th European Conference on AI (ECAI-84), pp.231-238, Pisa, Italy. 
Carter, D. (1987) Interpreting Anaphors in Natural Language Texts, Ellis Horwood: Chichester, England. 
Charniak, E. (1978) On the use of knowledge in language comprehension. Artificial Intelligence, 11, pp.225-265. 
Chamiak, E. (1986) Motivation analysis, abductive unification and non-monotonic equality. Artificial Intelligence, 34: 
275-295. 
Cohen, P. and Levesque, H. (1985) Speech acts and rationality. Proceedings of the 23rd Annual Meeting of the Associa- 
tion for Computational Linguistics. University of Chicago. 
Coombs, M. J. and R. T. Hartley (1987) The MGR algorithm and its application to the generation of explanations for novel 
events. International Journal of Man-Machine Studies 27: 679-708. 
Coombs, M. J. and R. T. Hartley (1988a) Explaining novel events in process control through model generative reasoning. 
International Journal of Expert Systems 1: 89-109. 
Coombs, M. J. and R. T. Hartley (1988b). Design of a software environment for tactical situation development. Proceed- 
ings of the US Army Symposium on Artificial Intelligence Research for Exploitation of the Battlefield Environment, 
El Paso, November 1988. 
Dinsmore, J. (1987) Mental Spaces from a functional perspective. Cognitive Science, pp.l-23. 
Fass, D. (1986) Collafive Semantics: An Approach to Coherence. Memorandum in Computer and Cognitive Science, 
MCCS-86-56, Computing Research Laboratory, New Mexico State University, New Mexico. 
Fass, D. (1987) Semantic Relations, Metonymy, and Lexical Ambiguity Resolution: A Coherence-Based Account. In 
Proceedings of the 9th Annual Cognitive Science Society Conference, University of Washington, Seattle, Washington, 
pp.575-586. 
238 
Fass, D. (1988a) Collative Semantics: A Semantics for Natural Language Processing. Memorandum in Computer and Cog- 
nitive Science, MCCS-88-118, Computing Research Laboratory, New Mexico State University, New Mexico. 
Fass, D. (1988b) Metonymy and Metaphor: What's the Difference? Proceedings of the 12th International Conference on 
Computational Linguistics (COLING-88), Budapest, Hungary. 
Fass, D. (1988c) An Account of Coherence, Semantic Relations, Metonymy, and Lexical Ambiguity Resolution. In Steve 
L. Small, Gary W. Cottrell, and Michael K. Tanenhaus (Eds.) Lexical Ambiguity Resolution in the Comprehension of 
Human Language, Morgan Kaufmann: Los Altos, CA, pp.151-178. 
Fauconnier, G. (1985) Mental Spaces: aspects of meaning construction in natural language. Cambridge, Mass.: MIT 
Press. 
Fields, C. A., M. J. Coombs and R. T. Hartley (1988a). MGR: An architecture for problem solving in unstructured task 
environments. Proceedings of the Third International Symposium on Methodologies for Intelligent Systems, Elsevier, 
Amsterdam, 44-49. 
Fields, C. A., M. J. Coombs, E. S. Dietrich, and R. T. Hartley (1988b) Incorporating dynamic control into the Model Gen- 
erative Reasoning system. Proc. ECAI-88, 439-441. 
Gershman, A. (1977) Knowledge Based Parsing. Memorandum from Yale University Computer Science Department. 
Hobbs, J.R., Stickel, M., Martin, P., Edwards, D. (1989) Interpretation as Abduction (Draft). AI Center, SRI International. 
Huang, X. (1984) The Generation of Chinese Sentences from the Semantic Representations of English Sentences. In 
Proceedings of lnternational Conference on Machine Translation. Cranfield, England. 
Huang, X. (1988) XTRA: The Design and Implementation of A Fully Automatic Machine Translation System. Memoran- 
dum in Computer and Cognitive Science, MCCS-88-121, Computing Research Laboratory, New Mexico State 
University, New Mexico. 
Josephson, J.R., B. Chandrasekaran, J. W. Smith, and M. C. Tanner (1987) A mechanism for forming composite explana- 
tory hypotheses. IEEE Trans. on Systems, Man and Cybernetics 17: 445-454. 
Konolige, K. (1983) A deductive model of belief. Proceedings of the 8th International Joint Conference on Artificial Intel- 
ligence. Los Altos, Calif.: Morgan Kaufmann. 
Levesque, H. (1989) A knowledge-level account of abduction. In Proceedings oflJCAl89, Detroit. 
Maida, A.S. (1983) Knowing intensional individuals. Proceedings of the 8th International Joint Conference on Artificial 
Intelligence. Los Altos, Calif.: Morgan Kaufmann. 
Maida, A.S. (1984) Belief spaces: foundations of a computational theory of belief. Tech.Rep. CS-84-22. Dept. of Computer 
Science, The Pennsylvania State University. 
Maida, A.S. (1986) Introspection and reasoning about the beliefs of other agents. Proceedings of the 8th Annual Conf. of 
the Cognitive Science Society. Hillsdale, NJ: Lawrence Erlbaum Associates. 
Maida, A.S, (1988) A syntactic approach to mental correspondence. Proceedings of the Canadian Soc. for Computational 
Studies of Artificial Intelligence. Edmonton, Alberta. 
Minsky, M. (1975) A framework for representing knowlege. In The psychology of computer vision. Edited by P.H. Wins- 
ton. New York: McGraw-Hill. 
Nau, D. and J. Reggia (1986) Relationships between deductive and abductive inference in knowledge-based diagnostic 
problem solving, in L. Kerschberg (Ed.), Expert Database Systems: Proceedings of the First International 
Workshop. New York: Benjamin Cummings. 
Perrault, R. (1987) Invited presentation. Meeting of the Association for Computational Linguistics Stanford, Calif. 
Perrault, R. (forthcoming) An application of default logic to speech act theory. In Plans and intentions in communication 
and discourse. Edited by P. Cohen, J. Morgan & M. Pollack. Cambridge, Mass.: MIT Press. 
Perrault, R. and Allen, J. (1980) A plan-based analysis of indirect speech acts. American Journal of Computational 
Linguistics, 6, pp.167-182. 
Reiter, R., de Kleer, J. (1987) Foundations of assumption-based truth maintenance systems. In proceedings of AAAI87, 
Seattle. 
Shadbolt, N (1983) Processing reference. Journal of Semantics,2, (1), pp. 63-98. 
239 
Slator, B. (1988) Lexical Semantics and a Preference Semantics Parser. Memorandum in Computer and Cognitive Science, 
MCCS-88-116, Computing Research Laboratory, New Mexico State University, New Mexico. 
Slator, B. (1988) Constructing Contextually Organized Lexical Semantic Knowledge-Bases. Proceedings of the Third 
Annual Rocky Mountain Conference on Artificial Intelligence, Denver, CO, (June 13-15), pp.142-148. 
Slator, B. and Y. Wilks (1989) PREMO: parsing by conspicuous lexical consumption. In Proceedings of the International 
Workshop on Parsing Technologies, Carnegie Mellon U., Pittsburgh. Editerd by M. Tomita. 
Wiebe, J. and Rapaport, W. (1986) Representing De Re and De Dicto belief reports in Discourse and Narraative. Proceed- 
ings of lEEE,74, pp. 1405-1413. 
Wilks, Y. (1972) Grammar, Meaning, and the Machine Analysis of Language, Routledge and Kegan Paul: London. 
Wilks, Y. (1975a) A Preferential Pattern-Seeking Semantics for Natural Language Inference. Artificial Intelligence, 6, 
pp.53-74. 
Wilks, Y. (1975b) An Intelligent Analyser and Understander for English. Communications of the ACM, 18, pp.264-274. 
Wilks, Y. (1978) Making Preferences More Active. Artificial Intelligence, 10, pp.75-97. 
Wilks, Y. and Ballim, A. (1987) Multiple agents and the heuristic ascription of belief. Proceedings of the lOth Interna- 
tional Joint Conference on Artificial Intelligence. Los Altos, Calif.: Morgan Kaufmann. 
Wilks, Y. and Ballim, A. (1988) Shifting the belief engine into higher gear. In Proceedings of the international conference 
on AI methodology and systems applications. Edited by T.O'Shea & V. Sgurev. Amsterdam: Elsevier Publishers. 
Wilks, Y. and Ballim, A. (in press, a) Belief systems: ascribing belief. In Proceedings of KIFS-87. Edited by T. Christaller. 
Wilks, Y. and Ballim, A. (in press, b) The heuristic ascription of belief. In Review of cognitive science, 1. Edited by N.E. 
Sharkey. Norwood,NJ: Ablex. 
Wilks, Y. and Bien, J. (1979) Speech acts and multiple environments. 
Proceeding of the 6th International Joint Conference on Artificial Intelligence. Tokyo, Japan. 
Wilks, Y. and Bien, J. (1983) Beliefs, points of view and multiple environments. Cognitive Science,8, pp.120-146. 
Wilks, Y., Fass, D., Guo, C., McDonald, J., Plate, T., and Slator, B., (1988) Machine Tractable Dictionaries as Tools and 
Resources for Natural Language Processing. Edited by J. Pustejovsky, Lexical and Computational Semantics, Cam- 
bridge, MA: MIT Press. 
240 
