Improving second language reading through visual attention cues to corpus-based patterns

Abstract

The patterns inherent to written text often remain opaque to second language learners due to the considerable cognitive demands that reading places on working memory. Learners must attend to the meaning of unknown words, the grammatical structure of sentences, and the meaning of the text as a whole – and this all simultaneously. One solution for helping learners to better attend to existing form, function, and frequency patterns within texts is through systematic visual attention cues, which may offload some of the burden on working memory. Lex-See is a Chrome browser extension that highlights words within a user-supplied text in a variety of shades and colors based on underlying corpus-based data about frequency and word class, and also provides further information about forms, definitions, and phonetic similarity, on mouse-over. Currently Lex-See is optimized for Czech, a less-commonly taught, morphologically rich language with a clear need for easily accessible corpus-informed language learning tools, but it is designed to work with any language for which lemma frequency, form, dictionary, and phonetic data can be supplied.