This study explores the developmental pattern observed when Japanese adult learners acquire
the locality condition on English reflexives. Experimental tasks were designed specifically to deal
with the methodological problems of earlier research and then administered to Japanese learners
of English at five proficiency levels (n = 411) as well as English and Japanese control
groups (n = 40). Results from the learner groups indicate that the locality condition is
acquired significantly better with sentences containing embedded that-clauses (type
E-1) than with sentences containing embedded infinitival clauses (type E-2). This asymmetry
exists even at beginning stages of learning and persists through later stages. For type E-2 clauses,
there is an appreciable percentage of advanced learners (about 35% in this study) who failed to
acquire the locality condition, which, I argue, is extremely difficult to account for within the UG
models proposed thus far.