Language Processing: Construction of Mental Models or More?

Abstract Language comprehension is often seen as the incremental update of a mental model of the situation described in the text. With every word a reader or listener processes, the model is adjusted to fit the linguistic input. This conception of language comprehension sounds very plausible at first sight but its scientific utility strongly depends on the definition of a mental model. In this paper, we will discuss different definitions of mental models for text comprehension, as well as the way these models may have an impact on comprehension processes. A second distinction we will discuss is an “eliminative” view of mental modeling compared to a “hybrid” view. An eliminative view denies the relevance of more linguistic levels of processing below the construction of a mental model; claiming that most if not all processing phenomena can be explained without reference to syntactic or semantic levels of representation. From a more hybrid perspective, inspection of a mental model is one among several factors influencing linguistic processing. We will argue for this latter perspective. (We are, of course, not arguing against every possible variety of an eliminative view, but just against some that have been and still are very prominent in the psycholinguistic literature. So there surely are conceivable variants that are compatible with our objections.)