2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 2
Presentation Time: 8:15 AM

THE SEMIOTIC CRITIQUE OF EARTH-SURFACE PROCESS MODELING


BAKER, Victor R., Hydrology and Water Resources, Univ. of Arizona, Tucson, AZ 85721-0011, baker@hwr.arizona.edu

Advances in computational technologies are providing ever-expanding opportunities for the creative formulation and application of mathematical Earth-surface process models. But there are also stinging criticisms: (1) inadequate understanding of model assumptions, (2) the lack of proper measurements on the real-world phenomena to which the models should correspond, (3) the logically unverifiable structure of realistic models, and (4) the damage to public policy by reliance on model predictions. As a resolution to this paradox I suggest that many of the perceived shortcomings of models derive from an overly narrow view of their representational essence. Models function to signify for human conceptualization those complex realities for which understanding is sought. Unfortunately, this is commonly envisioned via a two-valued semiology, in which the model acts as signifier for a signified natural process. Too much emphasis is placed on the relationship (usually via “prediction”) between signifier (model) and signified (reality). Models are more appropriately conceptualized via a three-valued semiotics, in which the model acts as a sign in an inseparable relationship both to its object (“reality”) and to the “interpretant” that is entailed by the sign-object relationship. This triadic relationship means that the sign (model) is inseparable from its context (object and interpretant), including various assumptions and consequences of application. Examples include (1) the formulation of the “system” that the model portrays, (2) the relationship of that “system” to reality, including the model assumptions, and (3) the differences among engineering, geophysical, and geological modeling. Particularly important are these differing criteria for model evaluation: (1) correspondence between the real-world entities and the deduced consequences of the model (predictions, simulations, model representation), and (2) ubrity (fruitfulness) of scientific inquiry facilitated by the model in terms of its coherence, consistency, and consilience with the real world.