As a result of his experiences, Schmidt challenges Krashen's notion that 'learned knowledge' is of little help in developing a second language. His Noticing Hypothesis claims that explicit knowledge in fact contributes to the implicit acquisition process because it helps us to notice features in the input that we might otherwise have been oblivious to, even if that noticing is largely unintentional.

Later Schmidt (1990) establishes from a review of research in cognitive psychology, that:

• Implicit learning (ie learning which is not intentional, and of which we may not be strongly aware) is possible but relies on our ability to attend to key features from the mass of information to which we are exposed.
• Second language learning can certainly be implicit but implicit processes may have limitations: our attention is likely to be drawn to salient and frequently-occurring features of linguistic input.

Salient and frequently occurring features are likely to be lexical items in the first instance, and as was established in section 2.1.2, lexical items are the mainstay of early SLA. You may remember that Pienemann also proposed that operations involving the beginnings and ends of grammatical units are more salient and therefore likely to be processed before those which involve movement into the middle of a grammatical unit (see section 2.1.4.4). In implicit learning generally, our attention is also likely to be drawn to information which, for whatever reason, has personal significance, or links in to existing frameworks of knowledge, or which corresponds to a perceived need. This provides a psycholinguistic foundation for the emphasis in the Communicative Approach on language which is meaningful to the learner.

But Schmidt's key point for language teachers is that learners are unlikely to attend implicitly to all the key elements of a target language. For this reason, a purely implicit process (ie Krashen's acquisition) is likely to lead to a fossilized grammar. This connects with the point made by Skehan (section 2.3.3.1) that we don't focus detailed attention on form when we try to understand, and McLaughlin et al's point (section 2.2.1.4) about the importance of restructuring the knowledge base for development of a complex, cognitive skill.

Schmidt claims that teaching can help the internal grammar grow by drawing attention to various features of the target language. Thus Krashenian 'learning' can potentially speed up the acquisition process, so long as learners also have opportunities for meaningful interaction in the target language, and attention is not diverted by complex metalinguistic information. This was the case for Schmidt himself acquiring the imperfect: he suggests that he would not have noticed it in the input, had it not been for his language class.