General Principles underlying the Design of Cognitive Models: Fuschl Conversatio …

IFSR Newsletter 1998 Vol 17 No. 2 June
Arne Collen, Eliano Pessa, Nicholas Paritsis M.P. Penna
The general plan of the investigation was:

  • Introduction
  • The need for cognitive models in a number of different domains
    (psychology, psychiatry, human – computer interactions, organizations, etc.)

  • The need for general principles underlying the design of cognitive models
  • Problems of definition
    • Different definitions of knowledge
    • Problems with different approaches
    • Need for considering the different approaches as complementary rather than opposite
  • General Principles
    • The models should be presented in a wider context
    • The model could be improved by considering higher and lower levels of system organization, by taking are of the interfaces between levels, so as to constrain the model.
    • Attempt is to be made to relate microscopic and macroscopic aspects of
      model in clear way. (hybrid models)

    • Clarify the relationships between traditional informational content (amount) and knowledge content of the model
    • Consider the relationships between cognitive functions taken consideration and non-cognitive aspects, such effects, emotions, etc.
    • Use of interdisciplinary and transdisciplinary approach which can be related
      to social and psychological approaches

    • Use as much as possible systems concepts and language
    • Find isomorphisms between information – processing systems and other living, and
      artificial, organizational systems

    • Consider the purpose of the model in order to decide the contents and the tools of modeling (e.g, see Jackson’s classification)
    • Try to find a description as economic as possible
    • mainly those involved in circular causality processes
    • quote the technical methods to eliminate variables in loops
    • Examples of cognitive models which could become frameworks for cognitive
      modeling

    • Model by Paritsis
    • Model by Anderson (ACT)
    • Artificial life Model
    • Relation to general principles
    • Use as few variables as possible (only the most important ones) to explain and
      represent the complexity of the cognitive processes.

    • Pay attention to circular causal loops that have to be preserved in the model
      because they usually play an important
      role.
| Category: IFSR NEWS