effet du nombre des graphèmes en Anglais - Aix Marseille Université

Appendice I 231Chumbley, 1984). In the MROM, we postulate three processes underlying a speeded binary lexical decisionresponse. Two of the processes use intra-lexical information to generate a "yes" response, and the third usesextra-lexical information to generate a "no" response. The two intra-lexical sources of information are : i)the overall (global) activity in the orthographic lexicon, operationalized in the simulation model as the sumof the activation levels of all word units, hereafter referred to as s, and ii) the (local) activity of functionalunits within the lexicon, operationalized as the activation level of individual word units, or µ. The extralexicalsource of information is time (t) from stimulus onset. In the MROM, a criterion value set on each ofthe three information dimensions determines the type (yes/no) and speed of a response. The criterion on the(local) µ dimension is referred to as M, the criterion on the (global) s dimension as ∑ and the temporal deadlineas T. Figure 3 illustrates how these three response criteria combine to determine the type and the speedof a response in the LDT.Figure 3. Application of the multiple read-out model to the lexical decision task. Three response criteria M, ∑,and T are set on three information dimensions : i) unit activity in the mental lexicon (µ), ii) summed lexical activity(s), and iii) time (t). Increases in µ and s over time follow the sigmoid function of an interactive activationnetwork (McClelland & Rumelhart, 1981). In general, word recognition is said to occur when the M criterion isreached, whereas a positive lexical decision response can be triggered when either the M or the ∑ criterion isreached before the T criterion. A negative lexical decision response is given in the converse situation.If either the local M or the global ∑ response criteria are reached before the T criterion then a positiveresponse is given, otherwise a negative response is given. Errors to word stimuli (false negatives) thereforearise when the T criterion is set too low and/or both the M and ∑ response criteria are set too high. Errorsto nonword stimuli (false positives) arise in exactly the opposite circumstances (high T criterion and/or lowM criterion or low ∑ criterion). In the example given in Figure 3, both the M and the ∑ response criteriaare reached before the T criterion giving rise to a positive lexical decision response. The speed of this responseis determined by the earliest moment in time that either the M criterion is reached (i.e., a specificword has been identified), or the ∑ criterion is reached (i.e., a fast guess has occurred). Response time for anegative response is simply given by the value of the T criterion.THE MROM-P. The starting point for the coding scheme of the MROM-P is the V-type (boxological)model of orthographic-phonological processing by Grainger and Ferrand (1994 ; see also Ferrand &Grainger, 1996). This model (see Figure 4) was empirically motivated by results from a series of maskedpriming studies (Ferrand & Grainger, 1992 ; 1993 ; 1994), and represents the simplest possible (global)phonological coding scheme within an IA-type architecture, that includes sublexical phonological structure.However, as is typically the case with V-type models, Grainger and Ferrand did not specify the nature of thephonological processing units.According to the principles of canonical and nested modeling (Grainger & Jacobs, present volume), westarted the construction of the MROM-P with the original structure, processing assumptions, and parametersof the MROM. These elements had already been kept constant in our previous "English" and "French" extensionsof the IAM, the semistochastic interactive activation model, or SIAM (Jacobs & Grainger, 1992),the letter-frequency model (Grainger & Jacobs, 1993), the dual read-out model, or DROM (Grainger &Jacobs, 1994), the semistochastic interactive activation model for the fragmentation task, the SIAM-FRAG(Ziegler, Rey, & Jacobs, in press d), and the MROM (Grainger & Jacobs, 1996).MULTILENGTH LEXICON AND THE CODING OF LETTER-IN-WORD POSITION. In our previousIA models, we used the simplification of a length-specific lexicon representing a single word length (eitherfour or five letters). Given the absence of an isomorphism between the size of orthographic andphonological representations (i.e., grapheme and phoneme units), the present "English" MROM-P isequipped with a much richer lexicon (albeit still a very simplified one) including all 3-5 letter, monosyllabic6 English words extracted from the CELEX database (Baayen, Piepenbrock & van Rijn, 1993). This led6 Focusing on monosyllabic words is a simplification that has been adopted by the majority of experimental andmodeling studies in the field. In future work, it will have to be revised.