Exploring inferences across media: an application of Kintsch’s framework of inference processes

ABSTRACT

Walter Kintsch proposed a very different framework of inferences than what is reflected in other classification systems. His framework focused on the extent that processes are automatic or controlled and whether inferences involve the retrieval of existing propositions or the generation of new propositions. In contrast, other classification systems of inferences focused on specifying types of inferences (goal, causal antecedent) or function (bridging, elaboration). In this commentary, I show how Kintsch’s inference framework can be used to formulate testable hypotheses concerning how the affordances of media differentially affect inference processes. These hypotheses demonstrate the framework’s promise in helping formulate a unified theory of inference generation that can explain what is common and different about it across media such as text, sequential picture stories, and film.

Acknowledgments

I would like to thank Laura Allen, Lingfei Luan, Heather Ness-Maddox, Puren Öncel, and Shu Hu for their contributions to my understanding of the affordances of different media for inference generation and comprehension. Our collaborations have provided a foundation for the ideas presented in this commentary.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. Graesser et al.’s (1994) constructionist theory of inference generation, which did specify production rules that specified conditions that lead to generation different inferences, could arguably be formalized in a computational model, but to my knowledge no one has ever done so.