Dialogical Models of Explanation

This paper takes on the task of providing a formal system of dialogue CE in which the speech acts of requesting and providing an explanation are represented as dialogue moves in the system. CE has opening rules, locution rules, dialogue rules, success rules and closing rules. The system is meant to be simple and basic, to provide a platform for developing more specialized formal dialogue systems of explanation used for specific purposes. The dialogical theory of explanation postulates that an explanation is a dialogue between two parties, one of whom asks a question requesting understanding of something which he or she claims not to understand, while the other offers a response that claims to convey the requested understanding to the party asking the question. In the last half of the twentieth century, the dominant model of explanation was the covering law (deductivenomological) model associated with Hempel (1965), its chief advocate. This model took an explanation to be a deductive inference from a set of facts called initial conditions and a set of general rules to a proposition to be explained. It would have been anathema to the analytical philosophers who accepted this model to suggest that an explanation should be thought of as a dialogue between two parties. Times have changed. Much recent work in AI has been based the dialogue model of explanation. Cawsey’s work (1992) on computational generation of explanatory dialogue used an interactive or dialogue approach, and Moore’s dialogue-based analysis of explanation for advice-giving in expert systems (1995) can also be cited. According to Moore (1995, p. 1) explanation is “an inherently incremental and interactive process” that requires a dialogue between an explanation presenter who is trying to explain something and a questioner who has asked for an explanation. The dialogical model of explanation has also been advocated and developed by Schank and his colleagues in cognitive science (Schank and Abelson, 1977; Schank, 1986; Schank, Kass and Riesbeck, 1994). Although there has been a slow but growing movement in the philosophy of science, away from the covering law model towards consideration of the dialogical model, this process has been very slowly taking place, somewhat like global warming. Copyright c © 2007, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. What is needed is a logical framework in which the dialogical model can be expressed in a precise way so that it can be seen as a worthy competitor to the covering law model. Actually the two models are quite compatible in many ways, but because one is dialogical and the other is emphatically not, they represent different paradigms of explanation. It is the contention of this paper that such a framework can be provided by extending some formal models of dialogue currently being studied in argumentation theory. Any one of a number of formal dialogue systems might be chosen, but in this paper one that is simple and basic is found to be useful because it can easily be extended as a basis for working towards developing more specialized systems of explanation dialogue. This work of extending the model is based on a dialectical (dialogical) analysis of the concept of explanation earlier presented in three places. In a paper (Walton, 2004), and more fully in two book chapters (Walton, 2004a, chapter 2; Walton, 2005, chapter 6), a set of conditions defining what an explanation is as a type of speech act used in a dialogue exchange between two parties was put forward. The analysis portrays both explanation and argument as using reasoning, for different purposes. It sees explanation as a speech act that is a distinctive kind of move made in a dialogue and it evaluates any given explanation on how well or successfully it contributes to the achievement of the collective goal of the dialogue. Each participant is viewed as an agent, an entity that can carry out actions based on its own goals. Each agent can possess something, or fail to possess something, called understanding. The purpose of an explanation is for the one agent to verbally transfer understanding to the other. Thus a basic assumption of the new theory is that agents can communicate with each other in an orderly way by means of what is often called a dialogue, defined as an orderly sequence of moves in which the two participants take turns uttering locutions (Hamblin, 1970, 1971). These locutions, like asking a question or putting an argument to support a claim that has been questioned, are nowadays equated with speech acts (Singh, 1999) of the kind analyzed by Searle (1969) Comparing Argument and Explanation As an example of an explanation, consider some anomalies cited in a book about the flight of Rudolf Hess to Scotland on May 10, 1941, written by a physician, Hugh Thomas, who