THIS STUDY is concerned with two of the major problems involved in interpreting the nature and un derlying mechanism of proactive inhibition in the forgetting of meaningful school material, namely , the effect of recent prior learning of interfering (subsuming) material, and the role of discrim ina bility of the learning from the subsuming material. Proactive inhibition is an import ant and ever present factor in school learning because most new ideational materials that pupils encounter in a school setting are significantly related to a previous ly learned background of meaningful ideas and infor mation. In fact the curriculum is deliberately or ganized in this fashion to provide for the gradual and untraumatic introduction of new facts and concepts. Hence it seems reasonable to suppose that interac tion between a meaningful new learning task and the existing, cumulativity established ideational sys tems which the individual brings with him to the learning situation would have greater influence on forgetting than the immediate interfering effects of material interpolated between learning and retention. This view--that the forgetting of meaningful ma terial is primarily reflective of p r o a c t ive rather than of retroactive inhibition?has been confirmed by the findings of a recently reported study (1). Ret roactive inhibition is a significant determining fac tor in forgetting only in instances of rote learning; since relatively little interaction between rote ma terials and established ideational foci occurs under these conditions, interference from s ubsequently learned rote materials assumes greater importance. Also, as Underwood (2) has convincingly demon strated, the retroactive inhibition paradigm of for getting does not account for all cases of r o te for getting. When the experimental arrangem e nts are such that interfering rote processes precede rather than follow the learning task, most of the forgetting can be attributed to proactive inhibition (2). There are good reasons for believing, therefore, that entirely different mechanisms underlie rote forgetting and the forgetting of meaningful materials. In rote forgetting, interference proceeds largely from other relatively current rote processes -ei ther proactively or retroactively--on the basis of situational or content similarity. Degree of proac tive inhibition, for example, in the forgetting of lists of paired adjectives is directly proportional to the number of previous lists learned (2). Apart from such interference, rote mat e r i a Is are also vulnerable to forgetting simply because their lack of meaningfulness obviates the possibility of anchorage to established ideational systems. In the forgetting of meaningful material, on the other hand, most of the interference is p r o a c tive in di rection and proceeds from stable, relevant concep tual foci acquired in the course of cumulative, long term learning experience. Hence, our first hypo thesis in the present study states that the operation of proactive inhibition in the forgetting of meaning ful materials is not dependent upon recent prior learning of interfering materials. At the very most, such prior learning could be expected to mobilize or bring already established subsuming systems into sharper relief. The model that we have proposed for the forget ting of meaningful material assumes the existence of a cognitive structure that is hierarchically or ganized in terms of highly stable and inclusive con ceptual clusters under which are subsumed less stable and more specific illustrative data. As new meaningful materials enter the cognitive field, they interact with and are appropriately subsumed under relevant, more inclusive ideational foci--since this is the most orderly and efficient way of organizing large quantities of material for ready availability . This initial stage of subsumption is purely a cata loguing operation. Hence at first, for a variable period of time, the recently catalogued data can be dissociated from their subsuming concepts and are elicitable as individually identifiable entities. Eventually, however, because it is more econ omical and less burdensome to retain a single in clusive concept (that incorporates in itself much supportive information) than to remember a large battery of more specific items, the import of the latter is assimilated by the generalized meaning of the former.