Thanks for the recommendation.

It does seem that Johnson-Laird is good at telling convincing-sounding stories and short on proving/predicting much (and probably his field).

So I'll read it for the collection of interesting experiments and over-confident speculation.