Subsumption Theory (David P. Ausubel – 1962)

One of the strongest criticisms of the information processing model was that it did not account for variation in the effort necessary to acquire knowledge of different content types or by different learners.[1] The prominent champion of this view was David P. Ausubel ( 1918-2008), who dedicated much of his professional career to defining and promoting the idea of meaningful cognitive learning. His theory was first presented in A Subsumption Theory of Meaningful Learning and Retention (Ausubel, 1962) and The Psychology of Meaningful Verbal Learning (Ausubel, 1963). It was later expanded in two editions of Educational Psychology: A Cognitive View (Ausubel, Novak, & Hanesian, 1968; Ausubel, Novak, & Hanesian, 1978). The 1978 edition was a revision of Ausubel’s ideas based on research and feedback from students and colleagues. It was this edition that was selected as the primary text to review in the present study for local principles of learning according to Ausubel’s theory, along with A Subsumption Theory of Meaningful Learning and Retention (Ausubel, 1962).

In1978 Ausubel was formally referring to his theory as assimilation theory in order to “emphasize a major characteristic; the important interactive role that existing cognitive structures play in the process of new learning” (Ausubel et al., 1978, p. v). To contextualize his theory, Ausubel et al. distinguished between two types of learning, rote and meaningful, and argued that—contrary to some popular claims[2]—most school learning was not rote, but meaningful:

The rote learning of lists of nonsense syllables and arbitrarily paired adjectives is representative of few defensible learning tasks in modern classrooms. It is also difficult to find supportive evidence for Underwood’s assertion that “much of our educational effort is devoted to making relatively meaningless verbal units meaningful” (11, p. 111). Brute memorization of representational equivalents (e.g., lists of vocabulary in foreign language study, the values of various constants in mathematics and science) tends to form a very small portion of the curriculum, especially beyond the elementary school years, once children have mastered the basic letter and number symbols. Meaningful learning of verbally presented materials constitutes the principal means of augmenting the learner’s store of knowledge, both within and outside the classroom. Hence, no research program purporting to advance this objective can avoid coming to grips with the fundamental variables involved in meaningful learning.(Ausubel, 1962, p. 215)

According to Ausubel et al. (1978) both rote and meaningful learning could occur in two different modes, reception and discovery. Though not completely against them, Ausubel et al. felt that “discovery methods of teaching hardly constitute an efficient primary means of transmitting the content of an academic discipline” (p. 26). This inefficiency was due to the extra effort required by the learner. Where in reception learning “the entire content of what is to be learned is presented to the learner in its final form” (Ausubel, 1961, p. 16), discovery learning requires a much greater effort in which learners must “rearrange a given array of information, integrate it with existing cognitive structure, and reorganize or transform the integrated combination in such a way as to create a desired end product or discover the missing means-end relationship” (p. 17).  In the end “the discovered content is internalized just as in reception learning” (p. 17).

In verbal reception learning, presented material is merely “internalized,” i.e., made available (functionally reproducible) for future use…Reception learning is meaningful provided that the learner adopts a set to relate the material to cognitive structure, and that the material itself is logically, i.e., non-arbitrarily, relatable thereto. In other words, pupils do not independently have to discover concepts or generalizations before they can understand or use them meaningfully. (Ausubel, 1962, p. 213)

Ausubel assumed a model of cognitive organization that supposed “the existence of a cognitive structure that is hierarchically organized in terms of highly inclusive conceptual traces under which are subsumed traces of less inclusive sub-concepts as well as traces of specific informational data” (1962, p. 216). Ausubel et al. (1978) later described three learning processes by which new knowledge is assimilated into existing cognitive structure:

1. Subordinate learning (there are two types):

In derivative subsumption, new information is linked to superordinate idea A and represents another case or extension of A. The critical attributes of the concept A are not changed, but new examples are recognized as relevant. (p. 68)


In correlative subsumption, new information y is linked to idea X, but is an extension, modification, or qualification of X. The critical attributes of the subsuming concept may be extended or modified with the new correlative subsumption. (p. 68)

2. Superordinate learning:

In superordinate learning, established ideas a1, a2, and a3 are recognized as more specific examples of new idea A and become linked to A. Superordinate idea A is defined by a new set of critical attributes that encompass the subordinate ideas. (p. 68)

3. Combinatorial learning:

In combinatorial learning new idea A is seen as related to existing ideas B, C, and D but is neither more inclusive nor more specific than ideas B, C, and D. In this case, new idea A is seen to have some criterial attributes in common with preexisting ideas. (p. 68)

In all three types of assimilation learning “new information is linked to relevant, preexisting aspects of cognitive structure and both the newly acquired information and the preexisting structure are modified in the process” (Ausubel et al., 1978, p. 68). The “major principle of organization” that makes this possible is one of “progressive differentiation” (p. 62), whereby “the most general and inclusive ideas of the discipline are presented first. Then they are progressively differentiated in terms of detail and specificity” (pp. 189-190). Three variables determine the extent to which assimilation through progressive differentiation is possible: (a) available subsumers, (b) discriminability, and (c) stability and clarity of subsumers. This was described as follows:

One important variable affecting the incorporability and longevity of new meaningful material is the availability in cognitive structure of relevant subsuming concepts at an appropriately proximate level of inclusiveness to provide optimal anchorage. If appropriately relevant and proximate subsumers are not present, the learner tends to utilize the most relevant and proximate ones that are available. But since the latter subsumers do not provide optimal anchorage, and since it is highly unlikely that the most relevant and proximate subsuming concepts are typically available to learners in most learning situations, it would seem desirable to introduce the appropriate subsumers and make them part of cognitive structure prior to the actual presentation of the learning task. The introduced subsumers would thus constitute efficient advance “organizers” or anchoring foci for the reception of new material.

A second important factor presumably affecting the retention of a meaningful learning task is the extent to which it is discriminable from the established conceptual systems that subsume it. A reasonable assumption here, borne out by preliminary investigation (6), would be that if the distinguishing features of the new learning material were not originally salient and clearly discriminable from stable subsuming foci, they could be adequately represented by the latter for memorial purposes, and would not persist as dissociable entities identifiable in their own right. In other words, only discriminable categorical variants of more inclusive concepts would have long-term retention value. The discriminability of new materials could be enhanced by repetition or by explicitly pointing out similarities and differences between them and their presumed subsumers in cognitive structure.

Lastly, the longevity of new meaningful material in memory has been shown to be a function of the stability and clarity of its subsumers (6). Ambiguous and unstable subsumers not only provide weak anchorage for related new materials, but also cannot easily be discriminated from them. Factors probably influencing the clarity and stability of subsuming concepts include repetition, their relative age, the use of exemplars, and multi-contextual exposure.

(Ausubel, 1962, pp. 219-220; see also the restatement of these ideas in Ausubel et al., 1978, pp. 168-169)[3]

In describing subsumption theory in 1962, Ausubel provided an interesting and very plausible explanation for the phenomenon of forgetting. His explanation centers on the concept of memorial reduction, which is, the least common denominator capable of representing cumulative prior experience:

Although the stability of meaningful material is initially enhanced by anchorage to relevant conceptual foci in the learner’s cognitive structure, such material is gradually subjected to the erosive influence of the conceptualizing trend in cognitive organization. Because it is more economical and less burdensome to retain a single inclusive concept than to remember a large number of more specific items, the import of the latter tends to be incorporated by the generalized meaning of the former. When this second or obliterative stage of subsumption begins, the specific items become progressively less dissociable as entities in their own right until they are no longer available and are said to be forgotten.

This process of memorial reduction to the least common denominator capable of representing cumulative prior experience is very similar to the reduction process characterizing concept formation. A single abstract concept is more manipulable for cognitive purposes than the dozen diverse instances from which its commonality is abstracted; and similarly, the memorial residue of ideational experience is also more functional for future learning and problem-solving occasions when stripped of its tangential modifiers, particularized connotations, and less clear and discriminable implication. Hence, barring repetition or some other special reason [e.g., primacy, uniqueness, enhanced discriminability, or the availability of a specially relevant and stable subsumer (see below)] for the perpetuation of dissociability, specific items of meaningful experience that are supportive of or correlative to an established conceptual entity tend gradually to undergo obliterative subsumption. (pp. 217-218)

Ausubel (1962) further explained that the common factor between learning and forgetting is that they both represent a “change in the availability or future reproducibility of the learning material,” with learning representing an “increment in availability” and forgetting representing a “decrement in availability” (p. 218).

In addition to the principles of learning found in his exposition of assimilation theory, Ausubel et al. (1978) provided a fairly comprehensive discussion of several other ideas regarding the nature of learning. These include statements—some based only on theory, others based on empirical evidence—regarding cognitive aspects of learning such as: concept formation, criteria attributes, multi-contextual learning, stages in concept acquisition, integrative reconciliation, sequential organization, transferability, the role and significance of practice and drill in a non-stereotypical sense, the role of feedback, the nature of practice, frequency of practice, early vs. delayed review, prompting and guidance, autonomous unguided discovery, differential practice schedules, and the context of practice vs. context of performance.  Although these are far too numerous to discuss here, they have been coded according to the method outlined in chapter two.

[1] Individual differences in retention and acquisition, as well as differences based on different content types was noted at least as early as 1913 by Ebbinghaus (1913, p. 3).

[2] Ausubel specifically cites Underwood, B. J. (1959) Verbal learning in the educative processes. Harvard Educational Review., 29, 107-117.

[3] Note the reference here to advance “organizers.” Though Ausubel supposed one could supply the necessary anchoring points just in time for instruction, (Anderson, Spiro, & Anderson, 1978, p. 439) cite Barnes & Clawson (1975) as stating that the research to support this claim “has proven inconclusive.” Anderson et al., further state that “it is difficult to see why outlining subsequent material in abstract, inclusive terms should help readers…when the reader does not possess relevant schemata, there is no good reason to suppose that they can be acquired from a few abstractly worded sentences.”

Leave a Comment