Analysis, Reconstruction, and Evaluation of Arguments

ARENA is a project envisioned by the Department of Logic and the Methodology of Sciences at Comenius University in Bratislava, Faculty of Arts (Slovakia). Funded by the Slovak Research and Development Agency, the project started on August 1 2018 and will span four years, finishing in 2022.

The project’s acronym stands for “Analysis, Reconstruction, and Evaluation of Arguments” and neatly represents the dialogical, conflictual nature of the process of argumentation. This is also reflected in the project’s logo, an ancient colosseum.

Project abstract

The project focuses on the analysis, reconstruction and evaluation of arguments, of both deductive and non-deductive form, whose evaluation (as valid, plausible, defeasible, probabilistically valid, etc.) is inconclusive either a) within a given communication context; or b) with respect to different theoretical frameworks of argument reconstruction and evaluation.

The project presupposes a specific definition of argument as an inferential relation between a set of premises and a conclusion such that it may be used to attain some methodological aim. The object of our reconstruction and evaluation will be arguments which i) pertain to expert public discourse (e.g., law, ethics, sociology, economics, political science, environmental sciences, etc.); and are either ii) evaluated inconclusively within a given expert discourse; or iii) evaluated non-equivalently when reconstructed in different theoretical frameworks (including deductive and inductive logics, argumentation theory, or formal epistemology).

The project takes up a hotly debated issue of competing arguments (and their reconstructions) and their analysis from the point of view of logic, argumentation theory and formal epistemology.

Goals

The goal of the project is to present and test a sufficiently fine-grained classification scheme of arguments which will provide a logico-methodologico-pragmatic (LMP-) profile of arguments that will enable the conclusive evaluation of problematic arguments (or competing arguments).

The Classification Proposed

Although we acknowledge the possibility of further elaboration and modification of our current proposal in the later phases of research, we believe that the core of our theoretical proposal can be expressed by means of the following eight basic logical, methodological, and pragmatic parameters that acquire various values in particular cases:

  1. Argument structure:
    1. simple argument (a set of one or more premises supporting a single conclusion)
    2. complex argument:
      1. chain argument (a system of several arguments in which the conclusion of an argument is a premise of another argument)
      2. convergent argument (a system of arguments in which different sets of premises support the same conclusion)
      3. divergent argument (a system of arguments in a set of premises supports different conclusions)
      4. combinations of the above
  2. The kind of support supplied by the premises to the conclusion – with regard to a theoretical (logical) system S:
    1. logical (in)validity in Si
    2. the relation of defeasible (revisable) support with regard to a current knowledge base B (and system Si) – (cases of non-deductive kinds of defeasible arguments such as abduction, analogy, statistical generalization etc.)
    3. the relation of inductive (confirmational) support with regard to system Si
  3. The kind of evidential status of premises:
    1. premises accepted as true ∕ false (on the basis of evidence E)
    2. premises postulated as true ∕ false (hypothetical premises)
    3. premises modeled as degrees of belief expressed using a concept of probability
    4. premises accepted as defeasibly plausible
    5. premises proven as true
    6. unsupported premises
  4. Specification of the modal nature of premises and conclusion:
    1. logically possible ∕ impossible ∕ necessary
    2. nomologically possible ∕ impossible ∕ necessary
    3. conceptually possible ∕ impossible ∕ necessary
    4. deontically possible ∕ impossible ∕ necessary
    5. metaphysically possible ∕ impossible ∕ necessary
    6. practically (technologically) possible ∕ impossible ∕ necessary
    7. other modality
  5. The methodological goal (potential) of the use of argument:
    1. objective support (justification) for thesis T (or its challenging)
    2. change of database of beliefs (of an agent)
    3. explanation
    4. prediction
    5. retrodiction
    6. justification of (the motivation of ∕ need for) action
  6. The kind of information in the premises ∕ conclusion:
    1. logical truth
    2. conceptual information ∕ meaning postulate
    3. convention
    4. empirical information (about a state of affairs)
    5. normative information
  7. Context of occurrence and the form of argument:
    1. monological
    2. dialogical
      1. proponent
      2. addressee ∕ opponent
    3. multilateral
  8. Pragmatic goal ∕ purpose of use of argument:
    1. cognitive (persuasion or challenging of the addressee, change of their information base etc.)
    2. conative (influencing the actions of the addressee etc.)
    3. evaluative (influencing the addressee’s value, i.e. non-doxastic, attitudes etc.)
    4. psychological (throwing the addressee off balance, manipulation etc.)

Methods

In the representation of logical and certain methodological or pragmatic components of arguments, we draw on the rich and established tradition of a variety of complementary theoretical approaches. Besides classical logic and certain systems of intensional and hyperintensional logic, we draw on the results of a multitude of other (semi-)formal approaches:

  • the pragma-dialectical approach of the Dutch school, Walton’s approach based on argumentation schemes, as well as logic-based approaches to argument formalization

  • Reiter’s pioneering work on nonmonotonic logic

  • John Pollock’s theory of defeasible arguments, the so-called AGM theory of belief revision, the theory of ranking functions

  • Bayesian epistemology, systems of inductive and probabilistic logic