The logical process of model-based reasoning

Abstract

Standard bivalent propositional and predicate logics are described as the theory of correct reasoning. However, the concept of model-based reasoning (MBR) developed by Magnani and Nersessian rejects the limitations of implicit or explicit dependence on abstract propositional, truth-functional logics or their modal variants. In support of this advance toward a coherent framework for reasoning, my paper suggests that complex reasoning processes, especially MBR, involve a novel logic of and in reality. At MBR04, I described a new kind of logical system, grounded in quantum mechanics (now designated as logic in reality; LIR), which postulates a foundational dynamic dualism inherent in energy and accordingly in causal relations throughout nature, including cognitive and social levels of reality. This logic of real phenomena provides a framework for analysis of physical interactions as well as theories, including the relations that constitute MBR, in which both models and reasoning are complex, partly non-linguistic processes. Here, I further delineate the logical aspects of MBR as a real process and the relation between it and its target domains. LIR describes 1) the relation between model theory - models and modeling - and scientific reasoning and theory; and 2) the dynamic, interactive aspects of reasoning, not captured in standard logics. MBR and its critical relations, e.g., between internal and external cognitive phenomena, are thus not "extra-logical" in the LIR interpretation. Several concepts of representations from an LIR standpoint are discussed and the position taken that the concept may be otiose for understanding of mental processes, including MBR. In LIR, one moves essentially from abduction as used by Magnani to explain processes such as scientific conceptual change to a form of inference implied by physical reality and applicable to it. Issues in reasoning involving computational and sociological models are discussed that illustrate the utility of the LIR logical approach. © 2010 Springer-Verlag Berlin Heidelberg.