Implicit Bias and the Teaching of Writing

In a recent article in the Chronicle of Higher Education (Yes, You have Implicit Biases, Too), David Gooblar challenges us to become aware of those latent prejudices that have inevitably developed within ourselves as we’ve been exposed, over the course of our lives, to harmful stereotypes. Gooblar highlights the work of psychologist Patricia Devine, who argues that stereotypes emerge as we form associations over our experiences. These stereotypes then become a lens that we bring to interacting with the world, as we act in unconscious ways to uphold the stereotype.

Writing is a particularly pernicious area for implicit biases to play out in. Multiple choice tests, quizzes, and other means of assessing student learning each bring their own possibilities for bias, but at least scoring of them is relatively straightforward. Writing is incredibly subjective. Scholars in different disciplines have wildly divergent ideas of what constitutes “good” writing, at every level from constructing sentences to constructing arguments. Moreover, each of us has our own, highly idiosyncratic list of pet peeves and things to reward.

Even error itself is more subjective than we tend to think of it. Compositionist John Bean writes in Engaging Ideas: The Professor’s Guide to Integrating Writing, Critical Thinking, and Active Learning in the Classroom:

Students’ prose contains fewer mistakes than teachers sometimes perceive… Almost every essay contains more grammatically correct sentences than incorrect ones. Instructors, however, tend to remember the error-filled sentences more clearly. Furthermore, research has shown that different readers will notice different errors and that some “grammatical errors”—like wordiness or passive voice—are really stylistic choices.

It can be challenging to describe in concrete terms what we value in writing. We tend to all agree that we like ‘clear prose’ and ‘well-organized ideas’, but as a list of operational criteria that someone unfamiliar with those terms might use to measure a piece of writing, what do those things actually look like? Writing is, in a word, mushy. Much of our reaction to a particular piece of writing (the way it makes us feel, the impression we form of the author) is at an unconscious level, and is therefore particularly prone to bias.

Yet more challenging is the fact that writing is mediated through language. Language is a minefield for a group of unconscious biases that sociolinguists refer to as language attitudes: ideas that purport to be about language, but are in fact reflections of individual and social ideas about groups of people. The coffee be cold and I might could do that are sentences that would be considered by many to be objectively wrong, but neither violates any logical principle of language. The first reflects a highly regular, rule-based aspect distinction (that the coldness of the coffee is an on-going state rather than a one-time occurrence, as in the coffee cold). The second, which has two modal auxiliaries, is a sentence the gloss of which would be considered acceptable in many other languages, including Dutch, German, and Chinese. Their so-called incorrectness isn’t a function of any inherent linguistic issue, but of bias: the former construction is associated with African-American Vernacular English, the latter with Appalachian English, speech varieties that have been (and continue to be) disparaged in the United States because of societal attitudes towards their speakers.

All of us have implicit biases, writing is a particularly fraught area for them, and language tends to amplify them. How, given all of this, can we hope to minimize the effect of implicit bias on our own work with students and their writing? There are no easy answers here, and we have a long way to go in higher education in resolving this issue—but we have two very powerful leads. To minimize the effect of implicit bias in your own work with student writers:

Be explicit about what you value in writing. When our ideas about writing are nebulous and implicit, it’s easy for them to become fertile ground for implicit biases to play out. When they are concrete and explicit, not only do we reduce the potential field in which prejudice can operate, we also make it easier for students to meet our expectations. Grading rubrics (particularly rubrics that you co-develop with your students) are one very powerful way to make your standards for assessing writing as objective, fair, and accessible as possible.

Know your own biases. All of us have implicit biases; they’re the inevitable result of coming of age within imperfect human societies. Only by becoming aware of our own biases, Devine and Gooblar argue, can we take positive action to counteract them in our work with students.

Interested in exploring your own biases, so that you can take positive action to correct for them? Project Implicit at Harvard University offers a free online tool that you can use to learn more about your own biases.