Models in professional knowledge

Unexpected countenance of change - The Boston Globe
SOMEONE, a psychiatrist, discouragingly once said that people don't change very much, but the little ways they change, when they do change, are enormous. It seemed a dour but accurate assessment.

It was practical wisdom -- the kind someone would acquire after being in the business many admirable years. Just recently, a patient proved to me that it is utterly wrong.
...
People change vastly, and for many reasons. It makes life worth living.

This is a good example of two conflicting models of a particular aspect of human nature. In one, exemplified by large-scale data, people change only a little, in another exemplified by one specific (and emotionally salient) case, they change radically. There are two points of interest here:

  1. How is this particular kind of expert (or even academic) knowledge constructed, maintained and applied? (Goffman's frames or Lakoff's models have a lot to contribute to the answer here).
  2. Can conflicting models such as these be reconciled through some formal heuristic? The answer is most likely no. But hopefully reference to some of the above mentioned research might help with the reconciliation.