Argumentation and Disputation as Issues of Philosophical Investigation
Dialectics denotes the art to carry out a dispute with opposing views. In a
dispute, or persuasion dialogue, arguments in favor of and against some claim
are exchanged between two parties according to certain rules of behavior.
Dialectics, as a scientific discipline, includes the investigation of questions
such as
-
what counts as an argument, what as a counterargument?
-
when is an argument defeated by another one?
-
when is an argument justified?
-
according to which dialogue rules should the dispute be carried out?
In the history of philosophy, the theory of argumentation has been closely
related to the development of formal logic whose theme is, however, much
narrower. While logical inference is based on valid sentences and conclusive
reasons and leads to the derivation of logically implied sentences,
argumentation is based on accepted sentences and prima facie reasons and leads
to the acceptance of further sentences after they have been justified.
Hamblin [1970] has proposed the notion of a "dialectical system" as an
extension of the notion of a logical system by including "features of
dialectical contexts within which arguments are put forward". Thus, in
dialectics, dialogue rules and further elements of the dialogue context, such
as the participants' beliefs and goals, resource limitations, and efficiency
constraints may be taken into consideration.
Formalizing Commonsense Reasoning in Artificial Intelligence Research
Researchers in Artificial Intelligence have soon noticed that classical logic
is not a suitable tool in their attempts to formalize commonsense reasoning.
One of the main reason for this is the recognition that most of the rules used
in commonsense reasoning are not universally valid but rather defeasible. This
insight has led to the development of formalisms supporting various forms of
nonmonotonic reasoning where the set of conclusions does not monotonically
increase with the set of premises.
Several of these nonmonotonic formalisms are based on a notion of acceptable
argument [Pollock 1987, Simari & Loui 1992, Vreeswijk 1993a, Dung 1995,
Bondarenko et al. 1997]. Nonmonotonicity arises from the possibility that an
argument may be defeated by a stronger counterargument. There are both
declarative and procedural notions of acceptable argument. In the decalarative
approach, arguments are defined as acceptable if they belong to some designated
set of arguments that is typically given by means of a non-constructive
fixpoint (or closure) definition. In the procedural approach, arguments are
defined as acceptable if they can be justified according to a certain
justification procedure. This procedure may consist of a well-defined dialogue
game between a proponent and an opponent. An argument in favor of some
proposition, then, is justified if it can be successfully defended against
every possible attack. The most important examples for these two approaches are
[Dung 1994] and [Loui 1993, Vreeswijk 1993b, Brewka 1994].
Computational Dialectics
Computational dialectics is concerned with the formal definition of justified
arguments and well-formed disputes, with their representation in the form of
suitable data structures, and with their operational semantics and its
efficient implementation. Typically, argumentation and disputes take place on
the basis of incomplete and uncertain information requiring advanced knowledge
representation and reasoning techniques.
In order to formalize disputes one needs more than a concept of justified
arguments. A dispute consists of a sequence of communication acts following a
certain protocol that defines the admissible patterns of interaction.
The following issues have to be investigated:
-
Are there alternative definitions of justification procedures? if yes, which of
them are most important?
-
How can disputes and protocols be defined in a formal and generic way?
-
What are the relevant speech acts in a dispute?
-
How can desirable properties of protocols such as fairness and efficiency be
formalized?
-
In which ways can protocols be modified online?
-
Which extensions have to be made for modeling negotiation dialogues?
References
[Bondarenko et al. 1997] A. Bondarenko, P.M. Dung, R.A. Kowalski and F. Toni.
An abstract argumentation-theoretic approach to default reasoning.
Artificial Intelligence 93 (1997), 63-101.
[Brewka 1994] G. Brewka. Reasoning about priorities in default logic.
Proceedings AAAI-94, 247-260.
[Dung 1994] P.M. Dung. Logic programing as dialogue games. Unpublished
manuscript.
[Dung 1995] P.M. Dung. On the acceptability of arguments and its fundamental
role in nonmonotonic reasoning, logic programing, and n-person games.
Artificial Intelligence 77 (1995), 321-357.
[Hamblin 1970] C.L. Hamblin. Fallacies. Methuen, London.
[Loui 1993] R.P. Loui. Process and policiy: resource-bounded non-demonstrative
reasoning. Report WUCS-92-43, Washington University, 1993.
[Pollock 1987] J.L. Pollock. Defeasible reasoning. Cognitive Science 11
(1987), 481-518.
[Simari & Loui 1992] G.R. Simari and R.P. Loui. A mathematical treatment of
defeasible argumentation and its implementation. Artificial Intelligence 53
(1992), 125-157.
[Vreeswijk 1993a] G. Vreeswijk. Studies in defeasible argumentation. Doctoral
dissertation. Free University Amsterdam, 1993.
[Vreeswijk 1993b] G. Vreeswijk. Defeasible dialectics: a controversy-oriented
approach towards defeasible argumentation. Journal of Logic and Computation 3-3
(1993), 317-334.
In Deutsch nicht verfügbar
Top
|
Home
|
Sitemap
|
