Prescriptive vs. Descriptive

I saw an excellent talk by Joseph Reagle today. Joe’s dissertation is on Wikipedia, and he presented a chapter called “Encyclopedic Anxiety,” in which he argues that reference works “often serve as a flashpoint for larger social anxieties.” This was very timely as I am half-way through David Foster Wallace’s essay “Authority and American Usage” from Consider the Lobster, in which Wallace riffs on class and power issues in American lexicography and grammar in general.

To make a very long story very short, Wallace outlines the debate between proponents of descriptive dictionary building (dictionaries should describe how people use the language, including terms considered non-standard or “wrong” like ain’t and irregardless) and prescriptive dictionary building (dictionaries should codify and enforce the rules of standard grammar). Clearly, what’s “right” and “wrong” are social judgments that usually reflect the interests of the class of people building the dictionary. Wallace calls these people SMOOTS; they are the copy editors of the world, often disparaged as “grammar Nazis.”

Reading this essay proved two things to me.
1) I was never formally taught grammar at any of the many educational institutions I attended, and therefore am woefully ignorant AND
2) Contentious debates about participatory culture, Web 2.0, content contribution etc. are just part of a very long struggle over Who Decides the Rules, Who Decides Whether Or Not There Should Be Rules and Who Is Allowed to Bend the Rules.

Joe places Wikipedia at the end of a long debate over reference works ranging from the first edition of Britannica (which encountered controversy for its entry on midwifery, which included illustrations of a fetus and the female pelvis), to the French Enlightenment text the Encyclopedie (which caused no end of trouble including censure, confiscation, imprisonment, and the torture and execution of a young nobleman purportedly under its influence), to Webster’s Third, a version of the unabridged dictionary seen as overly permissive and dismissed as “dogma that far transcends the limits of lexicography” (source: Wikipedia’s page on Webster’s Third). But what fascinates me is this idea of descriptive vs. prescriptive.

Descriptive vs. prescriptive is a debate that extends far beyond reference works. Any time a point of view is taken, there is a normative stance associated with it. For instance, if I write for a magazine for teenage girls that includes long articles explaining plastic surgery and bikini waxing, it is very likely that I will be seen as endorsing those practices, even if the articles themselves are neutral.

In my status project, I am examining what I’m calling status affordances: software mechanisms that enable users to be ranked or placed in some sort of social hierarchy or order. These include features like MySpace’s Top 8, Yelp’s Elite status and LiveJournal’s public Friends list. Reputation mechanisms, like eBay Feedback, Amazon reviewer rankings, and Digg’s “made popular” attribution often function as status affordances as well. Most of these mechanisms aim to be descriptive: They show which reviewer has written the most reviews, or how many friends a person has. But they are also prescriptive, in that they define what a user must do to improve his or her reputation or social status. These mechanisms operationalize whatever they’re measuring, whether it’s “trustworthiness,” “usefulness” or “coolness,”, which in turn creates a metric for that feature that makes it clear what must be done to increase one’s rank.

This means that status affordances both define and prescribe intended behavior. They create a determination of “quality” for a site’s users or contributions and determine what is appropriate or inappropriate behavior. For instance, the rank and order built into Amazon’s review system encourages people to submit thousands of reviews. Shay David and Trevor Pinch (I know I’ve cited this study before, but it’s so interesting) used custom-built software to parse existing reviews and found that hundreds were exact duplicates, modified slightly for different products. This explains how top Amazon reviewers can have more than five thousand reviews, which naturally calls into question their quality. If quantity was not a reputational mechanism, this type of behavior might not happen. Thus, descriptive mechanisms prescribe how to “game” the system.

Powered by ScribeFire.