I’m continuing my reading and response to two textbooks on cognitive poetics. The topic in the third chapter of both books, prototype analysis, was more challenging than figure & ground because I had not encountered it explicitly before, and my response here feels less concise. At the same time, writing this response was more rewarding. I have become more aware of what I understand and, more importantly, what I don’t understand.
Prototype analysis is an alternative to the classical method of categorization that dates back to Aristotle. Classically, categories are understood as a set of properties shared by all members in the category. This is the method taught as grade school biology: an organism is part of the plant or animal kingdom if it has all the properties of plant or of animal, an organism is a mammal if it has all the properties of animal and mammal. For example, if the organism is a marsupial, then it has all the properties of animal (cells without cell walls) and mammal (fur, warm blooded, etc.) and also carries its offspring in a pouch. From the classical view, categories are nested and an object can belong to more than one category; but an object either is or is not a member of a category. Even in grade school, I felt a rub with classical categories. I understand how it works and how it can be useful, but I’ve never felt it accurately described how I carve up the world. Prototype analysis allows for gradations of membership in a category.
Instead of being nested, prototypes have a radial structure with good examples clustered in the center and poor examples located in the periphery. Psychological studies are used to determine what people consider good and poor examples. For most people, chair is a good example of furniture. Apple and banana are good examples of fruit. Ottoman and lychee are poor examples of these categories, even though they meet the definitional requirements for membership.
Two ideas central to prototype analysis are family resemblances and basic level categories. Family resemblances refers to the idea that membership in a category does not entail an essential common feature. Things can be related by multiple features so that members of a category do not need to share all the features. Basic level categories are the not too specific, not too general level of category that people tend to think in most readily. It is easier to imagine a dog than it is to imagine either a border collie or an animal . To complicate matters further, prototype categorization is context dependent and dynamic.
As I mentioned previously, I’ve never felt comfortable with classical categorization. It always seemed to be a type of thought control (at worst) or a useful fiction (at best). Prototype analysis seems no less a fiction, but because prototype theory comes from cognitive psychology, I take it as descriptive rather than proscriptive. As a literary tool, prototypes can explain personal literary pantheons without turning them into a canon.
In Cognitive Poetics: An Introduction, Stockwell introduces prototype analysis generally and demonstrates a couple of ways of using prototypes for literary analysis. Applying prototype analysis with a broad view, Stockwell considers literary genres and how we come to understand a text as belonging to one genre or another. Considering genre as prototypes enriches the experience of reading and writing. He also considers prototypical sentences. A basic level statement has both topic and subject as agent. Variations from the basic level clause structure necessitate more interpretation by the reader. Writing that uses only the basic level clause structure meets reader’s expectations but allows limited reader interpretation.
The chapter in Cognitive Poetics in Practice on prototype analysis describes how prototypes are used on-the-fly to make meaning. Gibbs writes that we do not use a pre-existing structured knowledge base but construct a database as we read. Readers use embodied information to creatively compose meaning in the moment. Gibbs introduces two high-level processing mechanism,: memory organization p and thematic organization packets, used to create meaning from a text. Memory organization packets relate new information with expectations and goals to make predictions. Thematic organization packets help identify similarities between different events. He argues that meaning is not built from abstract concepts but from concrete examples.
Stockwell’s overview provides the basic understanding needed to follow Gibbs’ chapter, but I had to read Gibb’s several times before I could grasp his argument. Partly, the multiple readings were necessitated by his writing style. He wrote detailed descriptions of what prototype analysis is not, so that I had to understand two ideas at once: his conception of prototype analysis and the misconception of prototype analysis. Prototype analysis is complex because it describes a complex process: all the different ways we make meaning in all kinds of situations. The more accurate the description of that process, the more complex it is. I find that I had to resort to the same technique of describing what prototype analysis is not before I can articulate what it is.
The detailed description of dynamic meaning making given by Gibbs suggests a rich potential for meaning making. Readers create meaning fluidly as they read a story or a scene or a sentence. Gibbs provides literary examples from Anne Dillard and Anais Nin. Meaning is created using memory organization packets and thematic organization packets, so that readers understand the text without having experienced what is being written about. Tracking how meaning is made and re-made from the beginning to the end of a sentence is like doing finger exercises as a musician. The practice strengthens the linguistic muscles that are used to create.
The classical method of categorization has been used for such a long time because it is facile and efficacious. School children are taught this way of categorization fairly young. And in George Lackoff’s Where Mathematics Comes From, the foundation of mathematics is linked to container schemas and classical categorization. On the other hand, prototype analysis is complicated and ambiguous. Prototypes may be a more accurate description of the meaning making activity, but it has a different utility from classical categorization. As a tool of cognitive psychology, prototype analysis doesn’t attempt to describe ontological categories but the thinking that goes into categorization.