Key points are not available for this paper at this time.
One way of investigating what happens to the language learning capacity at puberty is to examine what the essential contributions of formal instruction are: since formal instruction is so useful for the adult and non-essential for the child language learner, it is probably the case that the main contributions formal instruction makes are in just those areas where the LAD is affected at puberty. A feature-type comparison of language teaching methods known to be successful in helping adults learn language reveals that the universal and presumably crucial ingredients of formal instruction are (1) the isolation of rules and lexical items of the target language, and (2) the possibility of error detection or correction. This hypothesis predicts that any new system that produces significant increases in second language proficiency for adults will contain these two features. It also predicts that adults who seem to be able to learn second languages in informal linguistic environments have some means of approaching rules and lexical items one at a time and are getting feedback. The analysis also points out direction for determining which values of non-universal features might be most useful for adult language learning instruction.
Krashen et al. (Sun,) studied this question.