Title: On the limitations of Abstract Argumentation Theory

By: Martin Caminada (University of Luxembourg)

Abstract: 
Since Dung's 1995 AIJ paper, abstract argumentation has become the main
research topic in the field of formal argumentation. Instead of aiming
to model natural concepts like how people use reasons to support claims,
these aspects are simply abstracted away, leaving just a graph
(consisting of "arguments" that "attack" each other) as the base
formalism. In the current presentation we point out the inherent
shortcomings of this approach. Our claim is that pure abstract
argumentation approaches are simply not rich enough for any meaningful
forms of knowledge representation. Worse yet, much of the recent work on
abstract argumentation cannot be enriched in any straightforward way
with the aspects that have been abstracted away. In order to deal with
this, we outline a number of research questions whose answers could help
to provide formalisms that are suitable for real world applications of
argumentation theory.