An adequate theory about how some knowledge is learned must have three main parts. The first necessity is a theory of the domain of that knowledge. Second is a description of the neophyte's state of relative ignorance. Finally, a learning theory must explain both how learning is possible and why it is non-trivial. Let the phonics approach to reading serve as an example. Reading is considered a process of translating a string of graphic symbols to an aural representation, which is then comprehended as all heard utterances are. The neophyte, the pre-reader, is held to be ignorant of letter-sound correspondences. Learning to read is thus a matter of learning these correspondences and the process of 'blending' sounds into words. There are two difficulties that make learning to read hard. The phonetic irregularity of English orthography is first mentioned. Secondly, the modification of sounds by each other when said together is a mighty confusion, i.e. 'blending' is not easy to understand. I do not consider this theory of learning to read entirely wrong, but I do hold it to be inadequate. For the theory is a partial one, leaving out of the learning process one critical factor: the learner as an active agent. I hope to establish a place for thinking in learning to read.
When we teach reading to five and six years old children, are we trying to stuff a lot of 'knowledge' into all those empty heads? What's learned must be assimilated to whatever is already known. No head is empty, though the contents may differ from what we first expect. If you can not see education as implanting new 'knowledge' but must see it as modifying what is in somebody's mind by the refinement of his conceptions, where do you start? Must you not find out what a person thinks, even if he be such a small person as a child? How else can you reinforce sound intuitions or highlight problems with erroneous ideas? How else can we help a child 'catch on' to a way of understanding language we believe is very important?
This experiment is an investigation of children's linguistic awareness - not merely with respect to which units of speech the child can manipulate - but as an inquiry into what the child thinks about the parts of what he says.