PROLOG and models of reasoning in science

Abstract

When we learn to program a computer in BASIC, in LOGO, or in FORTRAN, we soon learn that we must pay minute attention to the order in which things are done. The intended effect of a program is something different from how that effect is achieved. Thus, looking up an item in a database needs a routine which searches the database. Adding something to a number until it reaches a certain value requires a routine to check whether the value has been reached and to add what is wanted if it has not.This has come to seem so natural and inevitable a constraint that it gets elevated into an educational virtue, notably amongst enthusiasts for LOGO, and-I have to say-in my own thinking about (for example) the Dynamical Modelling System we produced for the revised Nuffield Advanced Physics course (Ogborn and Wong 1984, Ogborn 1987). There is something in the idea, no doubt. The logic of the way calculations of first a force, then an acceleration, then a velocity and finally a displacement have to be fitted together can help one to see how Newtonian mechanics works, particularly that forces affect positions indirectly through changing velocities, which runs counter to our normal experience of moving objects about.