A nonconnectionist schema theory of understanding surprise‐ending stories

Abstract

This paper compares two different ways of thinking about mental schemas that people use to comprehend and remember discourse. A parallel distributed processing (PDP) schema theory postulates that knowledge schemas are stored as connection weight patterns among units in an associative network. A nonconnectionist biofunctional schema theory, on the other hand, assumes that the brain stores no connection information. All knowledge is live knowledge created by the ongoing activity of the brain. Distributed constellations of specialized brain microsystems create and uphold schemas much in the same way, by analogy, that burning constellations of color‐coded light bulbs create and uphold colorful patterns of light. Three experiments are reported testing the predictions of the two theories using surprise‐ending stories, which have the special property of permitting two alternative interpretations of the same physical text. PDP schema theory implies that the two interpretations are built on two separate sets of units connected by mutual inhibition paths. Biofunctional schema theory, on the other hand, implies that the same constellation of brain microsystems participates in both interpretations. In Experiment 1, subjects read a story and rated the congruity of inference statements with the schema in mind. As predicted by biofunctional schema theory, the two schemas involved in the comprehension of the surprise‐ending story were rated as mutually incompatible. The results of Experiments 2 and 3 supported the prediction, also derived from biofunctional theory, that the two interpretations of the surprise‐ending story are incompatible to the extent they must be upheld by the same set of (shared) brain microsystems, and not because they are built on two separate subnetworks of units connected by mutual inhibition paths, as suggested by PDP schema theory.