Thursday, May 3, 2007

Pemberton, "Modeling Theory and Composing Process Models."

Pemberton, Michael A. "Modeling Theory and Composing Process Models." CCC 44.1 (1993): 40-58.

Pemberton examines models in composition studies by taking into account Flower and Hayes' cognitive process model of writing and its mixed reception. In one sense, then, this essay is an account of uptake. But there're more. Pemberton opens up many generative, provocative questions about the status of modeling in rhetoric and composition. He works through a stasis of definition (what are models?) and explains some of the givens (models simplify) and assumptions (models are mechanistic and positivistic; they are "partial isomorphs" of any complex phenomena) that have inhibited the production and circulation of models in the field.

To demonstrate the range of possible critiques of models, Pemberton cites Duhem, who argued against models because they weren't positivistic enough; on the other end of a spectrum, he refers to critics of Flower and Hayes' models who disclaimed them because they were too positivistic. This would indicate a full range of critical and evaluative treatments that is not explicitly tied to the activity of inventing models. Such critiques, perhaps, are more common when models are scarce or when their persuasive viability is undermined by their hybridity (as they often mix the discursive and non-discursive, the visual and the abstract, bridge the theoretical and its basis in data).

The essay is divided into seven sections: Opening, Models as Conceptual Frameworks (42), What Constitutes a Model? (44), Models as Simplifications (47), Models as Misleading Representations (48), Implications for Theory and Composition Discourse (52), Conclusions (54). Even though this is a follow-up to Flower and Hayes' model (addressing, very generally: what has come of the modeling of writing behaviors in the wake of Flower and Hayes?), it expands well beyond that moment by adding a layer (turning to the meta of modeling practices, modeling theory). Where models are treated as critical frameworks, Pemberton provides the following illustration:

Data - - - Models - - - Theories - - - Paradigms

His point with this is that each of the elements are "hierarchical," "interdependent," and "contiguous." Of course, even as they potentially bridge data and theories, models (when they are scarce and monumental, as with Flower and Hayes') are easy targets for critique. Simplification and misrepresentation are hazards (and exceedingly common bases for critique), as Pemberton rightly points out, but these should not prevent us from learning to make models, from using models to persuade and to mobilize (as Latour mentions).

Returns: terminological confusion related to "models" (44b), subject and source for a model as relates to Kuhn's 'preferred analogy' (45b), the principle of selection (research is always reductive and limiting (48)) (46b), Emig's inquiry paradigm (model as... or method as...) (54c).

Also work through Berthoff's critique of reductionism. How can visual models be abstract? General vs. abstract // study vs. sting (Barthes)...power of expansion and third meaning? (47b)

Pemberton ends the essay with a series of questions that, should we take up the work of modeling, we ought to sort through, address, etc.

Phrases: positivism (40), composing processes (41), paradigms (41), empirical scholarship (41), theory-building (41), modeling theory (42), conceptual frameworks (42), distillation of data (44), 'possibility' proofs (45), Kuhn's 'preferred analogy' (45), partial isomorphs (45), mechanistic (46), simplifications (47), incompleteness (53).

"To Duhem, meaningful understanding was intimately linked to scientific rigor, mathematical exactitude, and representational precision; since models were simplifications, their descriptions were unreliable and their utility questionable at best. In an age when positivism had not yet been supplanted as the dominant ideology guiding scientific inquiry, Duhem criticized models for their failure to be positivistic enough" (40)."

"Comparatively little attention has been paid, however, to the issue of modeling in composition studies, despite its central role in the interpretation of research data and the sheer number of models which exist to describe writing behaviors" (42).

"Before we can accurately interpret, evaluate, or employ any model of composing processes--or fully understand how several such models can coexist--we must be thoroughly informed with the knowledge of exactly what a model is, how it can be used effectively, and what its limitations are" (42). Significant here is Pemberton's mention of thresholds for coexisting models. How many can we have? Why not x+1? How many are too many? When they are dynamic and abundant rather than static and scarce, how is their intervention (or bridging between data and theory) different?

"The interdependence of these conceptual frameworks is reciprocal, operating in both a 'bottom-up' and 'top-down' fashion" (42).

"The terms 'subject' and 'source' can therefore be used to characterize the nature of the modeling relationship. We can assert, for instance, that any subject we wish to model--be it a tangible artifact or an intangible process--has a finite set of properties whose precise number is bounded, in part, by our ability to perceive and identify them" (45). Finitude?

"In addition, the model itself--or more properly speaking, the preferred analogy which is used to shape the model--will embody a number of intrinsic properties that do not properly belong to the subject being modeled" (45). See Wood, The Power of Maps, c. 5 and Monmonier, How to Lie with Maps, on generalization (rel. to abstraction).

"We must be careful, therefore, to guard against the urge to dismiss, preemptively, the value of a model merely because it contains imperfections" (46).

"The moment we decide what we want to investigate and how we want to conduct our research, we automatically delimit our field of inquiry and define its boundaries" (48).

"As I have already discussed the nature of such critiques, I will not belabor the issue further than to reiterate the point that incompleteness is an unavoidable epistemological weakness common to all models and all methods of data collection" (53).

"Researchers need to address questions such as: What are my methodological assumptions? What factors are likely to be included or excluded by my mode of inquiry? What assumptions shape the way I make my observations and interpret data? How are my representations likely to simplify writing processes, and how are they likely to misinterpret them? How to the epistemic tenets which ground my model compare with or connect to the tenets that ground the models of others?" (55).

Related reading:
Black, Max. Models and Metaphors. Ithaca: Cornell UP, 1962.
Emig, Janet. "Inquiry Paradigms and Writing." CCC 33 (Feb. 1982): 64-75.
Lauer, Janice. "Heuristics and Composition." CCC 23 (Dec. 1970): 396-404.

Bookmark and Share Posted by at May 3, 2007 10:38 PM to Models