Faced with an apparent conflict between two approaches to the teaching of deaf children : (i) that we should teach deaf children a language so that they can think, and (ii) that we should teach deaf children to think so that they can then acquire a language - I have examined the assumptions about thinking assumed by these two schools of thought. Reductionists hold that thinking is nothing but such things as inner speech (they identify thinking with its expression). Duplicationists argue that this is an inadequate explication of the concept of thinking (that it is only half the story) and they argue that thinking is something else as well as its expression. If successful Duplicationism becomes an objection to Reductionism. Unfortunately it results in an infinite regress. A third alternative account of thinking (Ryle's Adverbial account) regards thinking as an adverbial characterization: thinking is the way or circumstances in which we perform certain diverse and neutral (vis-a-vis thinking) activities. By such an account the elements of thinking which Duplicationists accuse Reductionist of ignoring become conditional dispositions. I argue that they should be regarded as categorical dispositional ascriptions. Additionally Ryle assumes a "process" account of thinking when in point of fact an "episodic" account is required. The thesis concludes by arguing that we need an ontology sufficiently large to take in all the aspects of thinking and that in turn this will generate not one precept but a matrix of precepts for the education of the deaf.