Sunday, November 19, 2006

Bransford and McCarrell, "Cognitive Approach to Comprehension"

Bransford, John D., and Nancy S. McCarrell. "A Sketch of a Cognitive Approach to Comprehension: Some Thoughts about Understanding What It Means to Comprehend." Cognition and the Symbolic Processes. Walter Weimer and David Palermo, eds. New York: Lawrence Erlbaum, 1974. 189-229.

Bransford and McCarrell set out to "raise particular questions about comprehension that may provoke further discussion and research" (189). How does one comprehend? What is the relation of object perception to comprehensible event? What is the ratio of perceptual experience to recall or memory in the meaningful event? Bransford and McCarrell key on the notion of a "click of comprehension." The "click" rests at the crisis point between aporia and epiphany: a moment of meaning-detonation. Bransford and McCarrell work through contextual constraints on meaning (missing or missed cues). They also explain the crucial role of relations among objects (195). That is, isolates (whether words or "brute things") do not mean in quite the same way when their relationality is diminished or altogether ignored. Their strong emphasis on comprehension as an orchestration of relations (perception, new knowledge, existing knowledge) matches well with Gibson's discussion of affordances (and also Fuller's discussion of Gibson's affordances).

In one example, Bransford and McCarrell speculate about what they term "abstract invariances" (197). They refer to walking as a relative class of events which, in specific cases, might vary. The "different particulars" involved with walking, then, might make is seem a novel event; however, there remains an abstract class of invariable information--abstract invariances--that apply to all (most!) instances of walking (or that function to distinguish walking from non-walking at the level of comprehension). The article goes on to consider relations and "entities involved" (an environmental or ecological consideration) and a few of the ways these ideas might be generalized to linguistic comprehension (201). That which is "directly expressed" in sentences (204) can only be comprehended in concert with whatever other information is available, including what Bransford and McCarrell "alinguistic information" (204). They also consider the "instigating force" of an unnamed, unseen entity that impacts comprehension despite its absence from the context, be it syntactic or scenic (211).

Other strong examples include the function of "submerged" in a paragraph about a man abandoning his car and walking toward the city (214). When the word is missing from the paragraph, comprehension is much more difficult. When added, however, the entire sequence of events as well as their motivations is comprehendible. A second example involves ambiguity in a confusing word, like "nog." "[The] ambiguous sentence The boy was found by the nog can be disambiguated as a function of our knowledge of nogs. If nog is assumed to refer to a monument in Central Park the sentence will be understood as a paraphrase of The boy was found near the nog; but if it refers to a furry animal with a good nose for tracking, the sentence will be understood to be a paraphrase of The nog found the boy" (219).

Sum: "Our approach to comprehension focuses on the comprehender's ability to use his general knowledge to create situations that permit the relations specified in input sentences to be realized, or to postulate situations (e.g., instigating forces) that allow perceptual events to be understood. In short, the ability to create some level of semantic content sufficient to achieve a click of comprehension depends upon the comprehender's ability to think" (220).

Key terms: click of comprehension (189, 200, 210, 215), meaningful entities (191), brute things (191), Piaget's "assimilation" (192), Bartlett's "effort after meaning" (192), affordances (193), abstract invariances (197), ecological niche (200), grasping of relations (200), sufficient alinguistic information (204), special assumption sentences and self-contained sentences (208), elaboratives (209), constraints (210), instigating force (211), categories of information (216), single lexical equivalents (216), labels for relations (218), nog (219), storehouse of images (220).

"Similarly, our perception of the world is rarely confined to identification of an individual object in isolation, but instead includes perception of an object's role in events" (190).

"Perception affords more than information about the characteristics of individual objects; it affords information about the spatio-temporal relations among entities that characterize the dynamic perceptual events (cf. E.J. Gibson, 1969; J.J. Gibson, 1966)" (191).

"That physical properties may have meaningful implications is important for consideration of perceptual learning, because it suggests that relational information that allows objects to become meaningful also affects what perceptual characteristics are learned" (194).

"The preceding discussion suggests that knowledge of entities arises from information about their relations to other knowledge, and that knowledge of relations distinguishes a meaningful object from a 'brute thing'" (195).

"Isolated objects cannot be taken as the basic unit of analysis when one seeks to understand how they become meaningful. Objects become meaningful by virtue of their interrelations with other objects (including the knowing organism); and objects are not always identified as mere objects" (197).

"Knowledge of entities and relations also interact to allow the comprehender to understand implicational significances of events which involve more information than is momentarily present" (199). ^Like Bruner?

"These examples illustrate how relational information about objects and information about abstract invariants characteristic of events interact to affect one's ability to comprehend novel situations" (199).

"The basic paradigm [for "a detailed analysis of relational information derived from perception"] is the 'ecological niche.' It consists simply of a film of a set of artificial entities that can be made meaningful to an organism as a function of his perception of their interactions in the perceptual mini-world" (200).

"These studies suggest that information 'directly expressed' by sentences cannot always be equated with the information available to the comprehender. Comprehenders do not simply store the information underlying sentences, but instead use linguistic inputs in conjunction with other information to update their general knowledge of the world" (204).

"Reasonable evidence suggests that the comprehender must frequently do considerable work to create situations that allow him to grasp the relations specified in input sentences, and that at least some specifications are necessary for the click of comprehension to occur" (210).

"We have proposed that knowledge of language might fruitfully be conceptualized as knowledge of abstract cues or instructions that guide the comprehender. The semantic content of a particular linguistic message is created only as the comprehender, guided by the linguistic cues, specifies conditions under which the abstract relations can be realized given his knowledge of the world" (215).

"Organisms must have information about an entity's relations to other aspects of his knowledge system to understand it. It follows that an image of a word's referent cannot be equated with its meaning, and similarly, the meaning of a whole sentence like The man made a touchdown cannot be equated with an image of a man crossing the goal line" (221).

Bookmark and Share Posted by at November 19, 2006 5:31 PM to Tools-in-use
Comments