What considerations make propositional models of sentence memory more attractive than linguistic models?
This is an interesting way to phrase a question. Let us assume that you mean by propositional models models that model sentence memory as uniquely a memory of the propositional content, and by 'linguistic models' models that model sentence memory uniquely as the memory of aspects of the linguistic form. Then first of all I do not know any purely linguistic models.
The interesing cognitive models usually model information about surface form as well as information about propositional content. Complex interactions, also interacting with with genre and rhetorical goal, determine the relative richness of these components and their relations. For example, Rolf Zwaan, already in his dissertation, showed that reader represent more of the surface structure when thay think to read littereary work than when they think to read a new paper article (look at it, very smart research design!!). Later Zwaan worked a lot on this issue (as did many others) so his work may be a good entrance for further references. Also a lot of work has been done on difference (and similarities) between semiotoc codes. Look at the work of Ray Jackendoff.
- Badges
- Science topic
- Similar topics
- Linguistics