1) You seem to include any kind of processing in "finding patterns". I wanted to both get your definition for "finding patterns" as well as the proof that this is how humans learn.
2) In the example of reading a book in class you fail to include the past history of experiences/learnings that takes place outside of class and which helps you to put the words in context. This might very well be a major differentiating factor compared to the first example.
To give a silly example: If I kick a wall and get a pain response through my toes which teaches me to not do that again, did I learn this by observing a pattern? Is that really a useful model for how the learning took place? Surely there was a pattern, perhaps even several ones, no denying that, but I highly question the accuracy and thus general usefulness of this model for learning.
Depending on definition, finding patterns could mean just about anything, thus meaning nothing.