This project continues our interdisciplinary research into computational and cognitive aspects of narrative comprehension. Our ultimate goal is the development of a computational theory of how humans understand narrative texts. The theory will be informed by joint research from the viewpoints of linguistics, cognitive psychology, the study of language acquisition, literary theory, geography, philosophy, and artificial intelligence. The linguists, literary theorists, and geographers in our group are developing theories of narrative language and spatial understanding that are being tested by the (...) cognitive psychologists and language researchers in our group, and a computational model of a reader of narrative text is being developed by the AI researchers, based in part on these theories and results and in part on research on knowledge representation and reasoning. This proposal describes the knowledge-representation and natural-language-processing issues involved in the computational implementation of the theory; discusses a contrast between communicative and narrative uses of language and of the relation of the narrative text to the story world it describes; investigates linguistic, literary, and hermeneutic dimensions of our research; presents a computational investigation of subjective sentences and reference in narrative; studies children’s acquisition of the ability to take third-person perspective in their own storytelling; describes the psychological validation of various linguistic devices; and examines how readers develop an understanding of the geographical space of a story. This report is a longer version of a project description submitted to NSF. This document, produced in May 2007, is a L ATEX version of Technical Report 89-07 (Buffalo: SUNY Buffalo Department of Computer Science, August 1989), with slightly.. (shrink)

A computer can come to understand natural language the same way Helen Keller did: by using “syntactic semantics”—a theory of how syntax can suffice for semantics, i.e., how semantics for natural language can be provided by means of computational symbol manipulation. This essay considers real-life approximations of Chinese Rooms, focusing on Helen Keller’s experiences growing up deaf and blind, locked in a sort of Chinese Room yet learning how to communicate with the outside world. Using the SNePS computational knowledge-representation system, (...) the essay analyzes Keller’s belief that learning that “everything has a name” was the key to her success, enabling her to “partition” her mental concepts into mental representations of: words, objects, and the naming relations between them. It next looks at Herbert Terrace’s theory of naming, which is akin to Keller’s, and which only humans are supposed to be capable of. The essay suggests that computers at least, and perhaps non-human primates, are also capable of this kind of naming. (shrink)

We present a logic-based formalism for modeling ofdialogues between intelligent and autonomous software agents,building on a theory of abstract dialogue games which we present.The formalism enables representation of complex dialogues assequences of moves in a combination of dialogue games, and allowsdialogues to be embedded inside one another. The formalism iscomputational and its modular nature enables different types ofdialogues to be represented.

Utterances of natural language sentences can be used to communicate not just contents, but also forces. This paper examines this topic from a cross-linguistic perspective on sentential mood. Recent work in this area focuses on conversational dynamics: the three sentence types can be associated with distinctive kinds of conversational effects called sentential forces, modeled as three kinds of updates to the discourse context. This paper has two main goals. First, it provides two arguments, on empirical and methodological grounds, for treating (...) sentential force as part of a compositional dynamic semantics, rather than a dynamic pragmatics. Second, it formulates a minimal dynamic semantic analysis that covers the data at the heart of these arguments, incorporating existing analyses of the three major moods, evidentials and conjunction. A further aim of the paper is to sharpen the distinction between sentential force and utterance force, and discuss its implications. (shrink)

To perceive an affordance is to perceive an object or situation as presenting an opportunity for action. The concept of affordances has been taken up across wide range of disciplines, including AI. I explore an interesting extension of the concept of affordances in robotics. Among the affordances that artificial systems have been engineered to detect are affordances to deliberate. In psychology, affordances are typically limited to bodily action, so the it is noteworthy that AI researchers have found it helpful to (...) extend the concept to encompass mental actions. I propose that psychologists can learn from this extension, and argue that human subjects can perceive mental affordances, such as affordances to attend, affordances to imagine and affordances to count. (shrink)

Previous plan‐based models of dialogue understanding have been unable to account for many types of subdialogues present in naturally occurring conversations. One reason for this is that the models have not clearly differentiated between the varoius ways that an utterance can relate to a plan structure representing a topic. In this paper we present a plan‐based theory that allows a wide variety of utterance‐plan relationships. We introduce a set of discourse plans, each one corresponding to a particular way that an (...) utterance can relate to a discourse topic, and distinguish such plans from the set of plans that are actually used to model the topics. By incorporating knowledge about discourse into a plan‐based framework, we can account for a wide variety of subdialogues while maintaining the computational advantages of the plan‐based approach. (shrink)

This essay presents a philosophical and computational theory of the representation of de re, de dicto, nested, and quasi-indexical belief reports expressed in natural language. The propositional Semantic Network Processing System (SNePS) is used for representing and reasoning about these reports. In particular, quasi-indicators (indexical expressions occurring in intentional contexts and representing uses of indicators by another speaker) pose problems for natural-language representation and reasoning systems, because--unlike pure indicators--they cannot be replaced by coreferential NPs without changing the meaning of the (...) embedding sentence. Therefore, the referent of the quasi-indicator must be represented in such a way that no invalid coreferential claims are entailed. The importance of quasi-indicators is discussed, and it is shown that all four of the above categories of belief reports can be handled by a single representational technique using belief spaces containing intensional entities. Inference rules and belief-revision techniques for the system are also examined. (shrink)

“Philosophy of action” is a recognized specialty in contemporary philosophy, and the literature on action is fairly extensive: see, for instance, (Care & Landesman 1968; Goldman 1970; Hornsby 1980). The relation of actions to their effects is formulated most clearly in the more specialized literature on the logic of action; see (Belnap & Perloff 1988; Chellas 1992; Czelakowski 1996; Segerberg 1982).

How is it that many schizophrenics identify certain instances of verbal imagery as hallucinatory? Most investigators have assumed that alterations in sensory features of imagery explain this. This approach, however, has not yielded a definitive picture of the nature of verbal hallucinations. An alternative perspective suggests itself if one allows the possibility that the nonself quality of hallucinations is inferred on the basis of the experience of unintendedness that accompanies imagery production. Information-processing models of “intentional” cognitive processes call for abstract (...) planning representations that are linked to goals and beliefs. Unintended actions - and imagery - can reflect planning disruptions whereby cognitive products do not cohere with concurrent goals. A model of schizophrenic speech disorganization is presented that postulates a disturbance of discourse planning. Insofar as verbal imagery can be viewed as inwardly directed speech, a consequence of such planning disturbances could be the production of unintended imagery. This link between the outward disorganization of schizophrenic speech and unintended verbal imagery is statistically supported by comparing the speech behavior of hallucinating and nonhallucinating schizophrenics. Studies of “borderline” hallucinations during normal, “goal-less” relaxation and drowsiness suggest that experiential unintendedness leads to a nonpathological variant of hallucinatory otherness that is correctable upon emerging from such passive cognitive states. This contrasts with the schizophrenic case, where nonconcordance with cognitive goals reinforces the unintendedness of verbal images and sustains the conviction of an external source. This model compares favorably with earlier models of verbal hallucinations and provides further evidence for a language production disorder in many schizophrenics.Short Abstract: How is it that many schizophrenics identify certain instances of verbal imagery as hallucinatory? This paper proposes that the critical feature identifying hallucinations is the experience of unintendedness. This experience is nonpathological during passive conscious states but pathological if occurring during goal-directed cognitive processing. A model of schizophrenic speech disorganization is presented that postulates a disturbance of discourse planning that specifies communicative intentions. These alterations could generate unintended verbal imagery as well. Statistical data are offered to support the model, and relevant empirical studies are reviewed. (shrink)

This paper explores principles governing the rational balance among an agent's beliefs, goals, actions, and intentions. Such principles provide specifications for artificial agents, and approximate a theory of human action (as philosophers use the term). By making explicit the conditions under which an agent can drop his goals, i.e., by specifying how the agent is committed to his goals, the formalism captures a number of important properties of intention. Specifically, the formalism provides analyses for Bratman's three characteristic functional roles played (...) by intentions [7, 9], and shows how agents can avoid intending all the foreseen side-effects of what they actually intend. Finally, the analysis shows how intentions can be adopted relative to a background of relevant beliefs and other intentions or goals. By relativizing one agent's intentions in terms of beliefs about another agent's intentions (or beliefs'), we derive a preliminary account of interpersonal commitments. (shrink)

This volume is about abstract objects and the ways we refer to them in natural language. Asher develops a semantical and metaphysical analysis of these entities in two stages. The first reflects the rich ontology of abstract objects necessitated by the forms of language in which we think and speak. A second level of analysis maps the ontology of natural language metaphysics onto a sparser domain--a more systematic realm of abstract objects that are fully analyzed. This second level reflects the (...) commitments of real metaphysics. The models for these commitments assign truth conditions to natural language discourse. Annotation copyright by Book News, Inc., Portland, OR. (shrink)

Natural dialogue does not proceed haphazardly; it has an easily recognized “episodic” structure and coherence which conform to a well developed set of conventions. This paper represents these conventions formally in terms related to speech act theory and to a theory of action.The major formal unit, the dialogue game, specifies aspects of the communication of both participants in a dialogue. We define the formal notion of dialogue games, and describe some of the important games of English. Dialogue games are conventions (...) of interactive goal pursuit. Using them, each participant pursues his own goals in a way which sometimes serves the goals of the other. The idea of dialogue games can thus be seen as a part of a broader theoretical perspective which characterizes virtually all communication as goal pursuit activity.We also define and exemplify the property of Motivational Coherence of dialogues. Motivational Coherence can be used as an interpretive principle in explaining language comprehension.Actual dialogue games have a kind of causal connectedness which is not a consequence of their formal properties. This is explained in terms of a theory of action, which is also seen to explain a similar attribute of speech acts. (shrink)

This research is concerned with the innate predispositions underlying human intentional communication. Human communication is currently defined as a circular and overt attempt to modify a partner's mental states. This requires each party involved to posse ss the ability to represent and understand the other's mental states, a capability which is commonly referred to as mindreading, or theory of mind (ToM). The relevant experimental literature agrees that no such capability is to be found in the human speci es at least (...) during the first year of life, and possibly later. This paper aims at advancing a solution to this theoretical problem. We propose to consider sharedness as the basis for intentional communication in the infant and to view it as a primitive, i nnate component of her cognitive architecture. Communication can then build upon the mental grounds that the infant takes as shared with her caregivers. We view this capability as a theory of mind in a weak sense.›. (shrink)

I show how a conversational process that takes simple, intuitively meaningful steps may be understood as a sophisticated computation that derives the richly detailed, complex representations implicit in our knowledge of language. To develop the account, I argue that natural language is structured in a way that lets us formalize grammatical knowledge precisely in terms of rich primitives of interpretation. Primitives of interpretation can be correctly viewed intentionally, as explanations of our choices of linguistic actions; the model therefore fits our (...) intuitions about meaning in conversation. Nevertheless, interpretations for complex utterances can be built from these primitives by simple operations of grammatical derivation. In bridging analyses of meaning at semantic and symbol‐processing levels, this account underscores the fundamental place for computation in the cognitive science of language use. (shrink)

Successful language use requires accurate intention recognition. However, sometimes this can be undermined because communication occurs within an interpersonal context. In this research, I used a relatively large set of speech acts (n = 32) and explored how variability in their inherent face‐threat influences the extent to which they are successfully recognized by a recipient, as well as the confidence of senders and receivers in their communicative success. Participants in two experiments either created text messages (senders) designed to perform a (...) specific speech act (e.g., agree) or interpreted those text messages (receivers) in terms of the specific speech act being performed. The speech acts were scaled in terms of their degree of face threat. In both experiments, speech acts that were more threatening were less likely to be correctly recognized than those that were less threatening. Additionally, the messages of the more threatening speech acts were longer and lower in clout than the less threatening speech acts. Senders displayed greater confidence in communicative success than receivers, but judgments of communicative success (for both senders and receivers) were unrelated to actual communicative success. The implications of these results for our understanding of actual communicative episodes are discussed. (shrink)

Distributed Artificial Intelligence (DAI) deals with computational systems where several intelligent components interact in a common environment. This paper is aimed at pointing out and fostering the exchange between DAI and cognitive and social science in order to deal with the issues of interaction, and in particular with the reasons and possible strategies for social behaviour in multi-agent interaction is also described which is motivated by requirements of cognitive plausibility and grounded the notions of power, dependence and help. Connections with (...) human-computer interaction are also suggested. (shrink)

In order to be capable of exhibiting a wide range of cooperative behavior, a computer-based dialog system must have available assumptions about the current user's goals, plans, background knowledge and (false) beliefs, i.e., maintain a so-called “user model”. Apart from cooperativity aspects, such a model is also necessary for intelligent coherent dialog behavior in general. This article surveys recent research on the problem of how such a model can be constructed, represented and used by a system during its interaction with (...) the user. Possible applications, as well as potential problems concerning the advisability of application, are then discussed. Finally, a number of guidelines are presented which should be observed in future research to reduce the risk of a potential misuse of user modeling technology. (shrink)

This chapter investigates the computational consequences of a broadly Gricean view of language use as intentional activity. In this view, dialogue rests on coordinated reasoning about communicative intentions. The speaker produces each utterance by formulating a suitable communicative intention. The hearer understands it by recognizing the communicative intention behind it. When this coordination is successful, interlocutors succeed in considering the same intentions— that is, the same representations of utterance meaning—as the dialogue proceeds. In this paper, I emphasize that these intentions (...) can be formalized; we can provide abstract but systematic representations that spell out what a speaker is trying to do with an utterance. Such representations describe utterances simultaneously as the product of our knowledge of grammar and as actions chosen for a reason. In particular, they must characterize the speaker’s utterance in grammatical terms, provide the links to the context that the grammar requires, and so arrive at a contribution that the speaker aims to achieve. Because I have implemented this formalism, we can regard it as a possible analysis of conversational processes at the level of computational theory. Nevertheless, this analysis leaves open what the nature of the biological computation involved in inference to intentions is, and what regularities in language use support this computation. (shrink)

We outline a theory of human agency and communication and discuss the role that the capability to share (that is, intersubjectivity) plays in it. All the notions discussed are cast in a mentalistic and radically constructivist framework. We also introduce and discuss the relevant literature.

We consider two issues relating to WH-questions:(i) when you ask aWH-question you already have a description of the entity you are interested in,namely the description embodied in the question itself. You may evenhave very direct access to the entity – see (1) below.In general, what you want is an alternative description of some item thatyou already know a certain amount about.

I show how a conversational process that takes simple, intuitively meaningful steps may be understood as a sophisticated computation that derives the richly detailed, complex representations implicit in our knowledge of language. To develop the account, I argue that natural language is structured in a way that lets us formalize grammatical knowledge precisely in terms of rich primitives of interpretation. Primitives of interpretation can be correctly viewed intentionally, as explanations of our choices of linguistic actions; the model therefore fits our (...) intuitions about meaning in conversation. Nevertheless, interpretations for complex utterances can be built from these primitives by simple operations of grammatical derivation. In bridging analyses of meaning at semantic and symbol‐processing levels, this account underscores the fundamental place for computation in the cognitive science of language use. (shrink)

In this paper we consider the potential role of metarepresentation—the representation of another representation, or as commonly considered within cognitive science, the mental representation of another individual's knowledge and beliefs—in mediating definite reference and common ground in conversation. Using dialogues from a referential communication study in which speakers conversed in succession with two different addressees, we highlight ways in which interlocutors work together to successfully refer to objects, and achieve shared conceptualizations. We briefly review accounts of how such shared conceptualizations (...) could be represented in memory, from simple associations between label and referent, to “triple co-presence” representations that track interlocutors in an episode of referring, to more elaborate metarepresentations that invoke theory of mind, mutual knowledge, or a model of a conversational partner. We consider how some forms of metarepresentation, once created and activated, could account for definite reference in conversation by appealing to ordinary processes in memory. We conclude that any representations that capture information about others' perspectives are likely to be relatively simple and subject to the same kinds of constraints on attention and memory that influence other kinds of cognitive representations. (shrink)

This paper pursues a formal analogy between natural language dialogue and collaborative real-world action in general. The analogy depends on an analysis of two aspects of collaboration that figure crucially in language use. First, agents must be able to coordinate abstractly about future decisions which cannot be made on present information. Second, when agents finally take such decisions, they must again coordinate in order to interpret one anothers’ actions as collaborative. The contribution of this paper is a general representation of (...) collaborative plans and intentions, inspired by representations of deductions in logics of knowledge, action and time, which supports these two kinds of coordination. Such representations.. (shrink)

We address the issue of manipulating games through communication. In the specific setting we consider (a variation of Boolean games), we assume there is some set of environment variables, the values of which are not directly accessible to players; the players have their own beliefs about these variables, and make decisions about what actions to perform based on these beliefs. The communication we consider takes the form of (truthful) announcements about the values of some environment variables; the effect of an (...) announcement is the modification of the beliefs of the players who hear the announcement so that they accurately reflect the values of the announced variables. By choosing announcements appropriately, it is possible to perturb the game away from certain outcomes and towards others. We specifically focus on the issue of stabilisation: making announcements that transform a game from having no stable states to one that has stable configurations. (shrink)

In this article we present the bases for a computational theory of the cognitive processes underlying human communication. The core of the article is devoted to the analysis of the phases in which the process of comprehension of a communicative act can be logically divided: (1) literal meaning, where the reconstruction of the mental states literally expressed by the actor takes place: (2) speaker's meaning, where the partner reconstructs the communicative intentions of the actor; (3) communicative effect, where the partner (...) possibly modifies his own beliefs and intentions; (4) reaction, where the intentions for the generation of the response are produced; and (5) response, where an overt response is constructed. The model appears to be compatible with relevant facts about human behavior. Our hypothesis is that, through communication, an actor tries to exploit the motivational structures of a partner so that the desired goal is generated. A second point is that social behavior requires that cooperation be maintained at some level. In the case of communication, cooperation is, in general, pursued even when the partner does not adhere to the actor's goals, and therefore no cooperation occurs at the behavioral level. This important distinction is reflected in the two kinds of game we introduce to account for communication. The main concept implied in communication is that two agents overtly reach a situation of shared mental states. Our model deals with sharedness through two primitives: shared beliefs and communicative intentions. (shrink)

An integrated statement is made concerning the semantic status of nodes in a propositional semantic network, claiming that such nodes represent only intensions. Within the network, the only reference to extensionality is via a mechanism to assert that two intensions have the same extension in same world. This framework is employed in three application problems to illustrate the nature of its solutions.The formalism used here utilizes only assertional information and no structural, or definitional, information. This restriction corresponds to many of (...) the psychologically motivated network models. Some of the psychological implications of network processes called node merging and node splitting are discussed. Additionally, it is pointed out that both our networks and the psychologically based networks are prone to memory confusions about knowing unless augmented by domain‐specific inference processes, or by structural information. (shrink)

In spite of alleged differences in purpose, descriptive and computational linguistics share many problems, due to the fact that any precise study on language needs some form of knowledge representation. This constraint is mostly apparent when interpretation of sentences takes into account elements of the so-called “context”. The parametrization of context, i.e. the explicit listing of features relevant to some intepretation task, is difficult because it requires flexible formal structures for understanding or simulating inferential behaviour, as well as a large (...) amount of information about conventional structures in the given language. This paper aims at illustrating major difficulties in these two fields, in relation with the necessity of a contextual approach. It offers a (clearly partial) enumeration of open problems inthe reperesentation of commonsense knowledge and languages-dependent structures, with some attempt to delineate future solutions. (shrink)

As a graduate student in Linguistics at UMass/Amherst in the 1980s, I was fortunate to be exposed to a number of new developments bearing on the relationship between formal semantics and pragmatics. In the 1970s under the influence of Cresswell, Lewis, Montague, and Partee, enormous progress in semantics was made possible by narrowing the focus of the field mainly to the consideration of the conventional, truth conditional content of an indicative utterance, calculated compositionally as a function of the semantic contributions (...) of its parts and its syntactic structure. Context was typically relegated to the background, in the form of indices of evaluation, though occasionally popping out for more serious consideration, as in the work of Kaplan, Karttunen, and Stalnaker. But eventually the nature of the rigorous formal enterprise itself, confronted with phenomena like Geach’s donkey sentences, presupposition projection, and the context-dependence of tense and aspect, forced the field into a more careful, thorough reconsideration of the relationship between context and content. Hence were born the dynamic theories of interpretation, beginning with Hans Kamp (1981) and Irene Heim (1982), who were both at UMass during my stay there, Heim completing her dissertation and Kamp on the faculty in Philosophy. Such frameworks posit a dynamic interchange between content and context, each dependent on the other, even in the course of interpreting a single utterance. Mats Rooth was also a graduate student at UMass at that time, and wrote his influential dissertation (1985) proposing an alternative semantics for the interpretation of prosodic focus, inter alia shedding light on how focus contributes to the contextual domain restriction of various operators. (shrink)

Agent communication languages (ACLs) invoke speech act theory and define individual message types by reference to particular combinations of beliefs and desires of the speaker (feasibility preconditions). Even when the mental states are restricted to a small set of nested beliefs, it seems that there might be a very large number of different possible preconditions, and therefore a very large number of different message types. With some constraints on the mental attitude of the speaker, we enumerate the possible belief states (...) that could serve as preconditions for individual message types, and we identify how these states correspond to different possible message types. We then compare these with FIPA’s primitive message types. Our approach clarifies the nature of core message types in an ACL, and perhaps settles issues concerning just how many, and what types of, speech acts should be seen as primitive in such languages. (shrink)

From a cognitive perspective, intentional communication may be viewed as an agent's activity overtly aimed at modifying a partner's mental states. According to standard Gricean definitions, this requires each party to be able to ascribe mental states to the other, i.e., to entertain a so-called theory of mind. According to the relevant experimental literature, however, such capability does not appear before the third or fourth birthday; it would follow that children under that age should not be viewed as communicating agents. (...) In order to solve the resulting dilemma, we propose that certain specific components of an agent's cognitive architecture (namely, a peculiar version of sharedness and communicative intention), are necessary and sufficient to explain infant communication in a mentalist framework. We also argue that these components are innate in the human species. (shrink)

Most areas of the cognitive and social sciences assume that knowledge representations are constructed and used during communication and that much of its content is social. Those of us who build computer models of comprehension and conversation are forced to be explicit about the nature of these knowledge representations and affiliated processes. There are some conditions when knowledge is not sufficiently social, and other conditions when knowledge is overly grounded in social mechanisms. The argument is advanced that constraints, coherence, and (...) precision are very much at the heart of an explanation of the nature and amount of social knowledge. Asocial content reigns supreme when the referential world knowledge is highly constrained, coherently structured, and precisely specified. The social context of knowledge construction becomes progressively more influential to the extent to that world knowledge is vague, open-ended, imprecise, underspecified, and fragmentary. (shrink)