The artificial reality model.

Once a satisfactory macrotheory (T1) is developed, the initial artificial reality (R1) can be modified to provide a more realistic conflict situation (R2), such as by converting something that was held constant in R1,into a variable. Efforts can then be made to generalize the earlier theory (T1) so that it applies to R2. The output is more general macrotheory T2 of which T1 is a special case. T2 is tested against history generated by experimentation with R2. This procedure is continued, and hopefully produces a sequence of successively more general macrotheories: T1, T2, . . . Tn.

As this set of theories expands, it can be analyzed to find principles that explain how the theories must be generalized in order to apply to more realistic artificial realities. That is, a metatheory is sought, one that yields a procedure for generatlng either t(n+1), givent1, t2, . . . tn, or T(n+1) given T1, T2, . . . Tn. These can then be tested in a new artificial reality R(n+1) which is a modification of Rn. The development of such a metatheory makes possible larger jumps toward theories of real conflicts hopefully to a theory that applies to reality in all its complexity. (PS 234-235)

(see Figure 2)

This page was last updated on July 24, 1996, by Rob Sable.