The project included two studies, which addressed how sound-letter mappings affect speech production and perception. Study 1 focused on the initial stages of L2 learning, and Study 2 addressed highly proficient bilinguals who learned their second language before they started to learn how to read.
For Study 1, 75 L1-Spanish monolinguals learned novel vowels. Participants were divided into 3 groups who learned the vowels 1) paired with letters which map onto different sounds in L1; 2) paired with novel letters; 3) without letters. All participants came to the laboratory three times. During all sessions, participants were exposed to the novel vowels; groups 1 and 2 also saw the vowels’ written forms. At the end of the third session, I assessed their production and perception of the novel vowels. Results showed that participants in the first group, with orthographic mismatches between L1 and L2, produced and perceived the newly learned vowels in a less accurately: Their production and perception were influenced by the L1 vowel represented by the same letter as novel vowel. Learning the novel vowels paired with novel letters did not have a detectable influence on participants’ production and perception. For an exploratory analysis, I also tested participants’ L1 vowel production and perception before and after learning the novel vowels.
For Study 2, I tested 60 highly proficient Spanish-Basque bilinguals on their production and perception of Basque words containing the letter ‘z’, which corresponds to different sounds in Spanish and Basque. Half of the participants acquired Basque as L2; the other half acquired Basque as L1. Importantly, all participants started learning their L2 in early childhood before they learned to read. Results showed no orthographic effects on speech production, but on speech perception: L2- and L1-Basque speakers alike judged Basque words as correct when the letter ‘z’ they contained was mispronounced as the Spanish sound it corresponds to.
Taken together, these results show the robust effect that mismatched sound-letter mappings between L1 and L2 have on L2 speech production and perception. These effects were not only found in the initial learning stages but even in highly proficient bilinguals who learned their L2 in early childhood before learning to read. Unexpectedly, orthographic mismatches between L1 and L2 also affected L1 speech perception in these early bilinguals.
To further investigate the effect of orthographic forms on speech production and perception, I collaborated with other researchers in five studies. One research line investigated how orthography affects L1 and L2 production in highly proficient bilinguals who learned their L2 as teenagers or young adults. Preliminary results support strong orthographic effects on both L1 and L2. A second research line investigated how children, when they are learning to read, are affected by orthographic inconsistencies in their L1 (e.g. the sound /k/ can be spelled , or in Spanish). Results suggest that orthographic inconsistencies make it harder for children to process spoken language. A third research line (2 studies) tested the effects of orthographic inconsistency within the L1 in beginning readers (children) and experienced readers (adults), by comparing Spanish and French speakers. The results show that both beginning and experienced readers are affected by orthographic inconsistency in spoken language perception but not production, thus paralleling the findings of this project’s Study 2.
