This is the second post in a series exploring a possible cognitive theory of genre. The first, appearing here, covered some concepts from contemporary cognitive science. You’ll find this post easier if you read the previous one.
Not all the developments in cognitive science that are relevant to a cognitive theory of genre post-date the “social turn” in genre theory signaled by Miller (1984). In fact, two important earlier developments have been cited in the genre literature, albeit in ways that suggest they are misunderstood or misapplied or were not attended to in a fashion that exploits their utility for genre study. These developments are Rosch’s prototype theory (Rosch & Mervis, 1975) and a variety of brands of “script theory,” typified by Schank and Abelson (1977).
Eleanor Rosch and her collaborators (e.g., Rosch & Mervis, 1975) sought to explain how humans understand categories and their members. Previous work had shown that some categories recognized by human agents could not be described as “logical bounded entities” with binary or digital characteristics (573). In other words, there are categories that cannot be described by necessary and sufficient logical conditions; rather, they are described with reference to more-or-less prototypical examples, and there might not be any single element or elements common to all the members of a category. For example, an American is more likely to regard a robin than an ostrich as a prototypical member of the category “bird” and is more likely to regard a chair than a radio set (the research took place when there were such things) as a prototypical member of the category “furniture.” Such prototype preferences prove reliable across multiple speakers and predict performance in response times for subjects determining category membership. So far, so good: and Swales (1990) refers to this research (pp. 51-52), though he spends twice as much time on Wittgenstein’s musings on the concept of “family resemblance”—a name that Rosch and Mervis borrow for the kind of category membership they describe—than he does on Rosch’s work. But Rosch and Mervis (1975) are also interested in a different question: “what principles govern the formation of category prototypes and gradients of category membership?” (574) This aspect of Rosch’s work is ignored in Swales (1990), and Rosch’s work is not even mentioned in Miller (1984), Berkenkotter and Huckin (1994), or Bhatia (1993). Worse, Amy Devitt (2004) misapprehends Rosch’s theory when she writes:
There is a danger… in providing samples of genres. Students easily turn samples into models, and models easily turn writing into formula. People commonly select one instance of a category as a prototype of that category, as more representative of that category than other examples (p. 209; citing a 1957(!?) work of Rosch and the work of George Lakoff).
This type of categorization is not the result of the processes that Rosch describes. Similarly, Bawarshi and Reiff (2010) attribute to Brian Paltridge the view that “prototype theory describes how people categorize objects according to a prototypical image they have conditioned in their minds by socio-cultural factors” (p. 39). We’ll further discuss in an upcoming post how the work of Rosch and her successors can provide insights into a cognitive theory of rhetoric.
On the theory of “scripts,” Swales (1990) and Bhatia (1993) cite the work of Roger Schank and Robert Abelson (1977). Schank and Abelson define a “script” as “a structure that describes appropriate sequences of events in particular context”; “made up of slots and requirements about what can fill those slots.” (Other theorists have used the term “frames” or “schemata.”) They argue that the script is a cognitive work-saver, in that it allows the agent to commit to memory only the script and, with regard to a given episode, the manner in which it varies from the script. (As we’ll see in the discussion of “Relevance” below, saving cognitive resources is a key concept for the cognitive theory of genre.)
Among our researchers, only Swales (1990) and Bhatia (1993) cite the work of Schank and Abelson (or others working in the same general space). Swales (1990) devotes Chapter 5 to discussing “genres, schemata, and acquisition.” Unfortunately, he pigeon-holes the cognitive angle: “the schema theorist’s emphasis on cognition has tended to isolate the text from its communicative purpose and from its environment” (88). Writing in 1990, Swales’ may have been influenced in his judgment by the Carnegie-Mellon model of cognitivism. It’s hard to believe a cognitive scientist employing contemporary theory could see a cognitive theory of writing or reading as being removed from communicative purpose. Bhatia (1993) also briefly takes up Schank and Abelson, properly noting that they are interested in the structuring of script-knowledge in the individual agent, based on his/her episodic experience; not in “social knowledge available to a particular discourse or professional community” (21). This will prove important to my proposed cognitive theory, as my intuition (yet to be fully worked out) is that there is no such thing as social knowledge—instead, I believe knowledge belongs to individual agents. “Social knowledge” if it can be said to exist in a “community” or “group,” is a central tendency in the knowledge of the individual members of the community or group. In other words, social knowledge exists as the territory of overlapping individual knowledges. Another way to think about this is to view social knowledge as the “family resemblance” among the individual knowledges of members of whatever community the knowledge is being attributed to. More on implications of script theory in a coming post.
Relevance: Communication as cognitive/pragmatic enterprise
Neither have all the developments relevant to a cognitive theory of genre had their principal homes in cognitive science. Relevance theory evolves out of linguistic pragmatics, a field focused on understanding how language works as used. Nevertheless, the movement I’m interested in has a distinctly cognitive flavor: Wilson and Sperber’s relevance theory.
The human mind evolved as a toolbox, a set of capacities, to survive in certain environments. Among the important implements in our cognitive toolboxes are heuristics, “quick and dirty” techniques for drawing conclusions or inferences about our environments that are not necessarily rational in any substantive sense. For example, Gigerenzer and Brighton (2009) identify several key heuristics, including the recognition heuristic (if one of two alternatives is recognized, it is preferred); and the fluency heuristic (if two alternatives are recognized, but one is recognized more quickly, it is preferred). Gigerenzer and Brighton’s recognition heuristic is akin to the concept of the affordance proposed by Gibson (1979). I’ve seen this illustrated by Dr. Carol Berkenkotter with a picture of a wood duck house. It’s a wooden box mounted on a pole or a tree a certain number of feet above the ground with a hole within a narrow range of circumferences. A wood duck confronted by a surface with such a hole at such a height actually perceives it as a potential home; in other words, it is perceived according to its utility to the wood duck as agent, not in terms of its separate characteristics (height, circumference, etc.). This makes good evolutionary/cognitive sense: it’s much more efficient to develop a heuristic for recognizing certain patterns of characteristics according to their utility than to have to perform a compositional analysis of separate characteristics to come to the same conclusion each time a new environment is acquired. (It does not always work well, of course, and Kahneman, 2003, and others have pointed to some of the poor results that can arise from heuristic thinking.) Such efforts to minimize cognitive effort are outgrowths of what Herbert Simon called “bounded rationality”—there’s only so much cognitive effort that a brain can bring to bear on survival problems.
Nevertheless, the effort of compositional analysis—conscious problem-solving—requires a lot of cognitive energy, and the human mind has evolved to minimize cognitive effort, just as the human body evolved to minimize physical effort. When determining whether to break from the heuristic/default approach, the human agent balances the effort required against the likely benefit. It is this ratio that Wilson and Sperber (2006) refer to as “relevance” in the context of human communication: The greater the utility and the lesser the effort to comprehend, the greater the relevance. They argue that we maximize relevance because of the way our brains are evolved (p. 610), and that “the search for relevance is a basic feature of human cognition, which communicators may exploit” (p. 608).
According to Wilson and Sperber (2006), “[t]he central claim of relevance theory is that the expectations of relevance raised by an utterance are precise and predictable enough to guide the hearer toward the speaker’s meaning” (p. 607). “The universal cognitive tendency to maximize relevance makes it possible (to some extent) to predict and manipulate the mental states of others” (p. 611). Their theory is more fully developed in extended works on the subject (Sperber & Wilson, 1996; Wilson & Sperber, 2012). But of central importance for a cognitive theory of genre is their proposed relevance-theoretic comprehension procedure, which calls on the listener to take the “path of least effort in computing” meaning, to “test interpretive hypotheses (disambiguations, reference resolutions, implicatures, etc.) in order of accessibility,” and to “[s]top when your expectations of relevance are satisfied (or abandoned)” (p. 612). In short, the easy interpretation should be the first; and by definition—at least from the heuristic standpoint—it will be the best (though it might not be correct). In my view, the easy interpretation will be the one that takes advantage of the contextual representations activated in the listener’s cognition at the time of the utterance: in short, it will depend on the listener’s expectations given the situation.
Thus the communicative agent’s success in achieving her objectives is closely tied to her ability to situate her utterance within the expectations of her audience. All communicative knowledge, including genre knowledge, is therefore an assessment by the agent of her audience’s expectations, given the situation. The agent must “make the case” of increased utility to the listener, if the agent requires the listener to expend precious cognitive energy in comprehension. Relevance theory, then, requires activation of the cognitive capabilities we have already discussed: the recognition of any prototypical characteristics or script of a communicative interaction based upon deeply situated knowledge.
Importantly, relevance theory is contemplated as an empirical theory and has motivated studies in cognitive linguistics and psychology. But despite its interest in rhetorical subject matter, it has received very little uptake in the field of rhetoric, and none of the texts on this reading list refer to or cite Wilson and Sperber’s work. Strangely, Rhetoric Society Quarterly chose to print Sperber (2007), describing it as an important work in cognitive rhetoric. The work, originally published in 1975 in French, predates Sperber’s work with Deirdre Wilson. In his introduction to this 2007 translation, Harris (2007) quotes Wilson and Sperber: “If relevance theory is right…. then rhetoric has no proprietary subject matter to study, or to teach.” This is a provocative claim, but the 1975 article describes a theory of meaning-making that is radically different from, and much less developed than, Sperber’s later collaboration with Wilson; and it bears hardly any resemblance to the form of relevance theory that prompted the provocative quote. It think it is perhaps because of this old article’s recent appearance in RSQ that there has been relatively little uptake of Wilson and Sperber’s work in rhetoric. The old article’s approach is better adapted for analytical philosophy than for rhetoric: reducing its relevance (in a relevance-theoretic sense) to readers in the discipline of rhetoric. Wilson and Sperber’s work has not suffered that fate in pragmatics, cognitive linguistics, and psychology: As of March 7, 2013, Google Scholar lists nearly 11,000 citations to Sperber and Wilson (1996).
Not your parents’ cognitive science
As we have seen, contemporary genre theory has taken an unproductive stance toward cognitive theory and methods, demonstrating an apparent belief that cognitive science is frozen in the form it held in 1981. Thus, contemporary genre theory has ignored theories of situated cognition; it has misapplied or misconstrued the findings of prototype theory; and it has entirely ignored relevance theory. To the extent cognitive issues have been raised in genre theory, it appears to have been a trend in the 1990s that has since died out. Bawarshi and Reiff’s (2010) survey of genre theory barely makes any mention of psychological or cognitive issues (but see references to them in the context of pedagogy below). The following two posts explain the ways that a cognitive theory of genre can address central questions in genre theory with techniques and theory from cognitive science.
-Brian