Brain Plasticity in Learning Visual Words

This study used event-related brain potentials and performance to trace changes in the underlying brain circuitry of undergraduates who spent 5 weeks learning a miniature artificial language. A reaction time task involving visual matching showed that words in the new language were processed like nonsense material before training, and like English words at the end of the 5 weeks of training. Scalp electrical recordings were used to explore the underlying basis for the change due to learning. Results of the ERPs were consistent with brain imaging studies showing posterior areas related to visual orthography and more widespread left lateral frontal and temporal areas related to semantic access. A posterior component at about 200 ms proved sensitive to differences in the orthography but did not change over the course of 5 weeks of training. A later ERP component at about 300 ms was sensitive to semantic task demands and underwent changes over the 5 weeks that were congruent with training-related changes observed in subjects’ matching task performance.