ABSTRACT
There is no doubt that the representation built by the reader is multidimensional (see chapter 4, §4.2). From a theoretical point of view, Gernsbacher (1990) stipulated that a correct mapping between information currently being processed and information processed previously relies on the coherence of information along five dimensions: referential, temporal, spatial, causal, and structural. Zwaan, Magliano, and Graesser (1995) also underlined, through their Event-Indexing Model (see Zwaan, Langston, & Graesser, 1995), that events conveyed in narratives are connected in memory along the dimensions of time, space, protagonist, causality, and intentionality (Zwaan, Langston, & Graesser, 1995; Zwaan, Magliano, & Graesser, 1995; Zwaan & Radvansky, 1998; Zwaan, Radvansky, Hilliard, & Curiel, 1998). Those two theoretical frameworks are based on the postulate of situational continuity, according to which the processing time of two events is facilitated if those events are situationally continuous along one of the aforementioned dimensions. The comprehension process should therefore be promoted when an incoming event continues on from another event on each of those five dimensions. Another underlying assumption is that the likelihood that events are connected in long-term memory is a function of their relatedness on each of these dimensions, with the strength of the relation between two events being a function of the number of situational cues they share.
