Let us look at each of these sources apart.
Our insights, like our perceptions, though intuitive in form, are obviously determined by previous experience, association, and habit.
Hence, on its passive side, an illusion of insight may be described as a wrong interpretation of a new or exceptional case.

For example, having associated the representation of a slight feeling of astonishment with uplifted eyebrows, we irresistibly tend to see a face in which this is a constant feature as expressing this particular shade of emotion.

In this way we sometimes fall into grotesque errors as to mental traits.

And the most practised physiognomist may not unfrequently err by importing the results of his special circle of experiences into new and unlike cases.
Much the same thing occurs in language.

Our timbre of voice, our articulation, and our vocabulary, like our physiognomy, have about them something individual, and error often arises from overlooking this, and hastily reading common interpretations into exceptional cases.