This paper draws on the philosophy of Karl Popper to present a descriptive evolutionary epistemology that offers philosophical solutions to the following related problems: "What happens when learning takes place?" and "What happens in human learning?" It provides a detailed analysis of how learning takes place without any direct transfer of information from the environment to the learner, and it significantly extends the author's earlier published work on this topic. She proposes that learning should be construed as a special case of "problem solving" and as a fundamentally critical and creative process in which learning organisms develop "expectations" that are not purely an outcome of genetic inheritance or random mutation. Human learning is then characterised with reference to: objectified knowledge; descriptive and argumentative language; theoretical problems; the search for error and specific limitation. If the author's evolutionary analysis of learning is valid, it would suggest that we should, if we wish to promote learning, be wary of corralling children and older students in environments that inhibit autonomous activity, that discourage criticality and creativity and generally limit opportunities for trial and error-elimination. But education institutions, particularly those for older children and adolescents, are very often environments of this constraining kind. Traditionally, educationists have vastly underestimated the human potential for imaginative criticism--because in general they have not recognised the extent to which it lies at the heart of what humans, including the youngest children, do in order to succeed at even the most basic tasks.