Pattern ortografici o unita’ lessicali? Effetti di priming morfologico in bambini di scuola primaria

[Autom. eng. transl.] INTRODUCTION The role of morphology in the acquisition of reading has been verified in several languages (see Verhoeven & Perfetti, 2011), but a wide debate is underway on the level of elaboration in which the morphemes would be identified. According to some authors (Rastle & Davis, 2008), this would be an automatic process, implemented based on the detection of specific sequences of graphemes, which would occur in the early stages of stimulus processing; according to others, instead, the recognition of the morphemic constituents would occur through the activation of lexical representations, organized on the basis not only of formal characteristics, but also of semantics (Feldman et al., 2009). To verify the role of the orthographic structure and the lexical-semantic components in the reading process in school age, Quémart, Casalis & Colé (2011) proposed to children of 8 and 12 years a lexical decision task with masked prime, finding confirmation , in French, of the fundamental role of orthographic components; on the contrary, Beyersmann, Castles & Coltheart (2012), in English, with the same paradigm, found clear effects related to semantic-lexical components. Faced with the heterogeneity of the results emerged in the literature, it seems interesting the contribution that the study of a transparent spelling language, like Italian, can provide, in order to better discriminate the role of the orthographic structure from that of lexical and semantic information, in the use of the morphemic constituents during the acquisition of the reading. METHOD Participants. 128 children attending classes III, IV and V of primary school (51% male; average age = 112 months). Stimuli. 120 pairs of prime-target words were selected, 40 for each of the following conditions: a) morphological relationship (eg, floury-FLOUR); b) pseudoderivation (eg, violence-PURPLE); c) orthographic similarity (eg, costume- COST). Each target could be preceded by a word in relation to it, according to one of the three conditions listed above, or by a control word (eg, fearful-FLOUR). The target words were matched for frequency of use in elementary vocabulary, length in letters and number of spelling neighbors. Procedure. Children were asked for a lexical decision task in a masked priming paradigm. EPrime 2.0 software was used for data collection. RESULTS The analyzes carried out on the reaction times showed the expected effects of the class attended, the frequency and the length of the word. As far as the priming effect is concerned, this was found only in the case in which prime and target were linked by an actual morphological relation (words in morphological relation: M = 1300 ms, DS = 444.4; control: M = 1379, DS = 442.3). No facilitating effects emerged either for the pseudo derivation (M = 1391, DS = 446.0; control: 1379, DS = 412.1), nor for the pure orthographic similarity (M = 1370, DS = 430.9; control: M = 1399, DS = 431.4). DISCUSSION The data seem to confirm the hypothesis that during the reading learning process, the association between orthographic form and meaning is essential for the identification of the morphemic constituents. On the contrary, there are no elements that manifest the use of a morphological decomposition on a purely orthographic basis, not even in a language with transparent spelling such as Italian.