Development of production skills in the absence of precise phonolexical representations

The study investigates the relationship between lexical encoding and production in order to establish whether learners are able to produce a difficult contrast in words that they merged in their mental lexicon. Forty American English learners of Russian were tested on their production and lexical encoding of familiar and highly-frequent words with the plain/palatalized contrast in second language (L2) Russian. Results suggest that the relationship between phonolexical encoding and production is less straightforward than a simple mirror image and is strongly affected by the prosodic position of the target consonants. In word-final position, learners did not lexically encode the difference between plain and palatalized consonants but they strived to produce it, although not very successfully. In intervocalic position, learners’ ability to encode and produce words with the plain/palatalized contrast was more accurate than in word-final position, which was attributed to the ‘spelling trap’ effect. Since Russian orthography employs vowel graphemes to mark the plain/palatalized status of preceding consonants, it appears that learners relied on these assumed vowel differences to articulate complex palatalization gestures. Thus, the findings of this study suggest that L2 learners can produce a contrast that they have not yet lexically encoded.