The Mental Lexicon of Children Aged 6 to 11. Presentation of the Manulex Lexical Bases and Manulex-infra.

Abstract

To study the mental lexicon, it is necessary to develop lexical databases to estimate the nature of the lexical units represented in the lexical memory of individuals. The aim of our intervention will be to present the French lexical bases MANULEX and MANULEX-INFRA which quantify the written lexicon addressed to children aged 6 to 11 in reading manuals. Using these bases, we will also attempt to estimate the size of the vocabulary (i.e., the size of the lexicon used) of a primary school child and to describe its development. MANULEX (Lété, Sprenger-Charolles and Colé, 2004; see also Lété, 2003, 2004) is a lexical database which provides the frequencies of word occurrences calculated from a corpus of 54 school textbooks (1.9 million words). The word frequency lists are provided for three levels of reading expertise: CP (6 years) where the child's lexicon is built on the basis of phonological mediation, CEI (7 years) where it is built the orthographic lexicon by progressive automation of recognition of the written word and cycle 3 (CE2-CM2, 8-11 years) where the lexical stock is consolidated and enriched by repeated exposure to writing. A fourth level is made up of the grouping of all the manuals from CP to CM2. MANULEX is made up of two lexicons for the 4 levels considered: a lexicon of 48,900 different spelling forms encountered in textbooks (songs, chantait, boat, boats, etc.) and a lexicon of 23,900 lemmas for which the frequency corresponds to the sum of the frequencies of the orthographic forms (chanter = lets sing + would sing, boat = boat + boats, ...). MANULEX-INFRA (Peereman, Lété, Sprenger-Charolles, 2007) describes the lexicon addressed to children in primary school at an infra-lexical level (grapheme-phoneme associations, syllables, bigrams, trigrams, etc.). In particular, we have developed a metric for the consistency of grapho-phonological and phono-graphic relationships at each age level for all the words in the lexicon of MANLILEX orthographic forms. The consistency of each word is measured on a scale of 1 to 100 at each level because this can vary depending on the lexicon addressed to the child at a particular stage of their learning.