How letter order is encoded in bilingual reading? The role of deviant-letter position in cognate word recognition

How letter order is encoded in bilingual reading? The role of deviant-letter position in cognate word recognition

First Author: Dr. Montserrat Comesaña -- CIPsi, School of Psychology, University of Minho
Additional authors/chairs: 
Coelho, R.
Keywords: Bilingualism, cognitive processes, Lexical Decision, Orthographic Knowledge, visual word recognition
Abstract / Summary: 

An issue receiving increasing attention in the monolingual literature has to do with the way letter order is encoded in reading. However, to the best of our knowledge only one unpublished study in the bilingual domain has attempted to address this issue (Font, 2001). According to Font, the recognition of cognate words (translation equivalents sharing their form) is affected by the position of the deviant letter (the position of the letter that changes in a cognate translation pair, such as the final letter in paper-papel [papel is the Portuguese word for paper]). This calls the validity of the input-coding scheme of the most relevant computational model of bilingual word recognition (The Bilingual Interactive Activation Plus Model [BIA+; Dijkstra & van Heuven, 2002; Dijkstra, Miwa, Brummelhuis, Sappelli, & Baayen, 2010]) into question, since it holds that letter positions are perfectly encoded. The aim of the present research was to test whether the front-end of the BIA+ model needs to be amended by manipulating the deviant-letter position of European Portuguese (EP)-English (EN) cognate words: at the beginning (e.g., coala-koala), or at the end (e.g., matriz-matrix). A third condition varying both at the beginning and at the end (e.g., escala-scale) was also included to test how the level of cross-language overlap modulates the results. Two hundred and twenty-eight stimuli (72 cognates, 72 noncognates and 144 pseudowords) were selected from the P-PAL (Soares et al., 2014) and the N-WATCH (Davis, 2005) databases. Logarithmic frequency per million words, word length, summed logarithmic bigram frequency, mean logarithmic bigram frequency, phonological and orthographic neighbors were controlled for within and across-languages. Twenty-eight proficient EP-EN bilinguals and twenty EN native-speaker controls performed a masked priming lexical decision task in EN. The results are discussed attending to more current letter position coding schemes.