Edward L. Thorndike, answered from the texts and cited to the page.
The old faith of the schools was that the mind is a set of general faculties — memory, reasoning, attention — and that hard subjects like Latin and geometry, however useless in themselves, strengthen these faculties as exercise strengthens a muscle, so that the training carries over to everything.1 Woodworth and I put this faith to the test.
We trained people at one task — estimating areas, cancelling letters, judging weights — and measured whether they improved at other, related tasks. They did not, except where the two tasks shared identical elements — the same specific habits, methods, or bits of content. Improvement is specific; it transfers only so far as the new situation contains the very same components that were trained.
There is no general strengthening of a faculty of memory or reason; there are only particular connections, and each must be formed. This cut the ground from under the classical curriculum's chief defense: Latin is worth teaching for what Latin itself is worth, not because it disciplines a mind that will then do everything better.234
They do, however, at least put out of court the old doctrine of a very wide spread of a very large percentage of the special improvement. Possibly nobody ever really believed that the improvement made in reasoning about Latin syntax would spread equally, or almost equally, to all or nearly all varieties of reasoning; but men wrote as if they believed substantially this.Educational Psychology Vol 2 The Psychology of Learning, pp. 417–418
We have studied the means of preserving or eliminating particular instincts and capacities, of forming particular habits of attention, interest, thought and feeling, of helping pupils to acquire particular connections of motor response, of selecting and abstracting particular elements, of guiding pupils in particular inductions and particular deductions.The Principles of Teaching Based on Psychology, pp. 253–254
They assumed the last function only after they had become useless in village economy, and after more popular games had become available. Man has a veritable passion for keeping up habits merely because he has them; there are men who would rather beat a sick child than write 'thru.Educational Psychology Vol 2 The Psychology of Learning, pp. 424–425
The one thing of which a teacher can be sure is the particular information, the particular habits and powers, the particular interests and ideals which his training gives directly; he may fairly expect improvement, but less in amount, in abilities closely like that trained; he may hope for some in more remote abilities, but for less and less and finally for none as the ability has less and less kinship with the…The Principles of Teaching Based on Psychology, pp. 260–262
This is among Thorndike's most secure and influential findings. The transfer experiments demolished the doctrine of formal discipline and permanently changed how education justifies its subjects, shifting the curriculum away from "mental training" toward specific useful content and skills. The identical-elements theory has been refined rather than overturned: modern research finds that transfer does occur but is often narrow and depends on shared structure, principles, and strategies (not only Thorndike's specific identical elements), so that the picture is more generous to transfer than his austere formulation while confirming his central destruction of the formal-discipline myth. This is Thorndike near his strongest.