But according to McLaughlin et al (1983), mere repetitive practice is not enough for developing a complex, cognitive skill such as language. For sustained development to occur, we have to repeatedly reorganize or restructure the way in which we represent the information required for the skill:

[…] there is more to learning a complex cognitive skill than automatizing subskills. The learner needs to impose organization and to structure the information that has been acquired. As more learning occurs, internalized, cognitive representations change and are restructured.
(McLaughlin et al, 1983:136)

Repetitive practice can only bring about automatization of the very specific subskill being practised (DeKeyser, 2001: 131). So, for example, repeating phrases in a foreign language (what we referred to earlier as chunks - see section 2.1.5.1) will certainly help you to utter those phrases, but not necessarily in the range of appropriate contexts. Nor will such practice alone help you to segment the grammar of the phrase. Automatization provides the potential for attention to be redirected to 'imposing organization' and 'structuring information', but it does not guarantee that this will happen.

Sometimes features of the environment in which we perform the skill may be strong enough to trigger restructuring of the knowledge we draw on to perform the skill (see McLaughlin, 1990). For example, the experience of not being understood in a second language may cause us to listen more carefully to native speakers and start to segment phrases we may earlier have automatized. We may start hearing 'new' features in the language input and work these into our internal grammar. (This is the process Schmidt (1990) refers to as 'noticing': see Reflective task 19 and section 2.3.3.4). This restructuring does not have to be an explicit, intentional process: it can often happen without us being strongly aware of it. But we may also need explicit feedback of some kind, for example, someone telling us what to do, helping to give us a new understanding of our task, and prompting us to pay attention to new aspects of our skill.

But what kind of feedback helps? Part of the problem with explicit feedback on a complex skill is the difficulty of 'explaining' the skill in the first place. Taking explanations of grammar as an example, rules often appear to learners more complex than the particular language operation they are designed to explain. Explanations of the behaviours required to perform a complex skill rarely translate easily into effective (or improved) performance of that skill. This may explain why many of us get so frustrated with the manuals written for the use of computer software!