This paper examines a formal model of how specifications can be constructed from multiple viewpoints and presents some tools to support this approach. The development of specifications is presented as a dialogue in which the viewpoints negotiate, establish responsibilities and cooperatively construct a specification. The model is illustrated by means of some small examples. Keywords: formal specification, distributed artificial intelligence, dialogue, logic, tool support 1 Introduction "Specification-in-the-large", that is the development of requirements specifications for systems of substantial complexity and scale, mirrors "programming-in-thelarge " in raising a variety of difficulties that lie beyond the clerical problems of handling large amounts of information (Cunningham, Finkelstein et al 1985, Finkelstein & Potts 1987). One such difficulty is that of specification from multiple viewpoints (Niskier 1987). Specification-in-the-large is an activity in which there are many particip...

This paper is a development and clarification of certain ideas put forth in Allwood & Haglund 1991 and its aim is to explore some ways in which deontic notions like obligation and option can be used to analyze some of the dependencies and regularities which exist in dialogue. 2 . Obligations and options of a dialogue contribution

The thesis specifies requirements for robust, cooperative and coherent dialogue systems and advocates a new approach, Constructive Dialogue Management. A prototype of a system capable of dealing with the planning of system reponses in factual information seeking dialogues is presented. Constructive Dialogue Management is based

Abstract not found

This paper presents a semantic/conceptual characterization of understanding based on a combination of 3 types of semantic analysis. The analyses lead to a characterization of understanding as the ability to relevantly engage in a partly controllable process of establishing integrated modally qualified meaningful connections between linguistic or nonlinguistic input information and background information, resulting In a relevant factive stative relation between an agent and some object of understanding.

This paper characterizes and exemplifies management in dialogue. Own communication management (choice and change directed) is distinguished from interactive communication management (sequences, turn management and feedback). An attempt is then made to motivate and explain the existence of various types of interactive management in dialogue. The suggested explanations involve a combination of general rational and ethical factors with more specific factors related to particular types of management. 1. Purpose A point of departure for this paper is that a number of different phenomena in spoken dialogue (like self correction, hesitation, feedback, and turntaking) exist primarily in order to enable management of dialogue. The term management has been chosen instead of the related terms regulation and control because it is less authoritarian and machine-like and allows, but does not require, intentional control. The purpose of the paper is to briefly

ION Acta Philosophica Gothoburgensia No 1 Gteborg 1986 2 2. AN INITIAL CHARACTERIZATION OF UNDERSTANDING Let me start by the following abstract characterization of understanding. Understanding is a relation between an agent and some particular type of information. In the discussion below I will try to become more specific about the nature of the relationship and about the nature of the relational objects - the agent that understands and the information that is understood. Some initial guidance can perhaps be found by studying the lexical item understand and its derived forms understanding (noun or adjective), understandable (adjective) and ability (noun). If we look at the meaning of these forms, the stemform understand is by itself most naturally taken as a verb denoting a relational process or state. The noun understanding denotes the reified relational process or state of understanding. The adjective understanding denotes a quality which an agent who understands has. The adjec...

: The paper presents a theoretical discussion of the relation between social structure, language and communication. Ten axiomatic theses on this relationship are stated and elaborated through a discussion of social institutions and social activities in the light of activity based communication analysis. Finally, an overview is given of activity based communication analysis. 1. Background The origins of this paper lie in puzzlement over a number of issues connected with the relation between language and social structure. They also lie in discontent with a dominant trend in 20th century linguistics toward what Hjelmslev (1943) called immanent linguistics, i.e., the study of language as a sui generis structure. Among the puzzling issues I would first like to mention the issue of holism versus atomism and reductionism in social theory. Do social phenomena in some sense (ontological, epistemological or heuristic, cf. Allwood 1973) have an independent status or are they reducible to more ...

video recorded interactions. Quantitative measures, such as the CHILDES (MacWhinney, 1991), and SALT (Miller and Chapman, 1986) transcription and calculation tools on conversation, which have been used e.g. by Holland, can be applied to transcriptions of video recorded interactions. Conversation Analysis and other microanalysis of single pragmatic phenomena, e.g. by the use of coding schemas, is increasingly used in pragmatic analysis of aphasia (e.g., Crockford and Lesser 1994, Ahlsén 1991, Linell and Gustafsson 1987, Leiwo, 1994). This type of analysis is also based on video recorded interactions. Environmental needs assessment, i.e. the analysis of communicative needs in different situations that the patient is likely to face (e.g. Hartley and Jensen 1991, 1992, Ahlsén 1995) supplements direct observation of naturalistic interactions with a focus on important factors of the situation. 3 Measures of communicative effective

this paper can be seen as influenced by at least the following parameters: (i) Collective - global (ii) Collective - local (iii) Individual - global (iv) Individual - local 2.1.1 Influencing collective global parameters Type of parameter Value in the investigated seminar (Natural) surrounding: Indoors in a recording studio, evening