The insight gain studied by Wagner et al. initially revealed itself in an implicit manner.
Then, through a slow process (enhanced by sleep), it emerged from the nondeclarative into the declar- ative realms as a fully assembled insight into the task structure.
The authors proposed that insight gain is not a pro- cedural learning process, since the reac- tion time data did not become faster with learning, which is the hallmark of procedural learning.
Instead, all participants who gained insight (regardless of whether they were in the sleep or wake groups) actually showed a slowing in reaction time just prior to insight gain compared with participants who did not gain insight.
“Specifically, the slow- ing of reaction time in solvers appears to reflect the presence of an incipient representation of the rule overlapping with that required for implicit task per- formance.”
Nonsolvers’ reaction time continued to decrease as their proce- dural skills improved without declara- tive awareness.
Solvers, on the other hand, showed a slowing in response while insight was emerging from non- declarative information into a declara- tive reportable result. The insight may, therefore, develop in a nondeclarative fashion, but the moment of ‘aha’ indi- cates the solidification of the informa- tion into a declarative thought. Thus, insight gain appears to contain aspects of both declarative and nondeclarative learning.This sounds a lot like Wallas' ideas of "incubation" and "enlightnment" come to light again.