Tuesday, October 9, 2012

Gendered pronouns: the options

Given that I think using the generic masculine at least a factor contributing to gender stereotyping and bias in society, and that it makes men a more salient interpretation of the referent, I want to try another strategy.

I have three major options:
  1.     Use "she" as the pronoun for all examples.
  2.     Alternate between "he" and "she."
  3.     Use the gender neutral "they" for all examples.
For each of these options, I would like to know:
  1. What effect it has on the reader (if any) in terms of gender stereotype/bias/salience etc.
  2. What effect it has on the reader (if any) in their evaluation of my writing or understanding of the content.
I'll take Option Two (alternating pronouns) first.
There are two ways of alternating pronouns: using "he or she" together, or alternating sentences with "he" and "she." In "Alternating Between Masculine and Feminine Pronouns: Does Essay Topic Affect Readers’ Perceptions?" Sex Roles, Vol. 54, Nos. 3/4, February 2006, Madson and Shoda argue that readers overestimate the number of feminine pronouns in a text which alternates between masculine and feminine pronouns in sentences. There are a number of possible reasons for this, but the effect diminishes when a feminine pronouns accompany a stereotypically feminine topic. For example, if I were to consistently say "she" when talking about nurses, and "he" when talking about mechanics, readers would not overestimate the number of feminine pronouns as much. However, they still do to an extent. Troublingly, readers also rate texts with alternating sentence pronouns as being worse in terms of quality. So if I write an academic paper alternating pronouns, I run the risk of readers unconsciously considering it a lower quality piece of writing than if I didn't alternate pronouns.

Using what Madson and Shoda call the "paired condition" ("he or she") results in higher assessments of quality, and less of a perception of gender bias. What is striking is that Madson and Shoda manipulated the experiments so that there were equal numbers of masculine and feminine pronouns in both cases. When readers thought that there were more feminine pronouns, they rated the text more gender biased--they were more aware of the gendered nature of the construction. One theory is that alternating sentences causes readers to imagine a male and female subject in turn, whereas paired pronouns are more easily read as a generic.

The authors conclude,
If the author is trying to persuade a conservative audience, paired pronouns may be his or her best option. If, on the other hand, the author is at least as interested in changing the world as in persuading the audience, alternating pronouns may be used precisely because they are jarring to the reader. Alternating pronouns might make readers’ perceptions of the text somewhat less positive, but they might also motivate readers’ to think differently about sexism in language and in general. The authors need to decide whether they are willing to take this risk.
This study, of course, says absolutely nothing about whether alternating pronouns impacts gender bias in the world. It tells me, tentatively, something about how my readers might characterize my writing.

And this is the problem I've found with studies about pronoun use and its effects on gender-based bias. For example, from the abstract of a 1981 study:
Sixty-six female and 66 male undergraduates read six job descriptions referring to the jobholder(s) either as "he," "he or she," or "they." Following each description subjects indicated their interest both in the job and a related job category, and estimated how difficult the fob would be to get for a (1) Black person, (2) Mexican-American person, (3) woman, (4) person over age 60, (5) handicapped person, and (6) man. There were no main effects for pronoun condition, but several sex of subject × pronoun interactions. Significant effects for females generally supported the notion that use of masculine pronouns to refer to people in general narrows the reader's attitudes toward the topic in question, while corresponding results for males did not. (Stericker, "Does this `He or She' Business Really Make a Difference?" Sex Roles, Vol 7, No 6, 1981)
The author of this study was trying to determine whether jobs identified with various pronouns would be more or less attractive to different groups. Would men be less likely to be interest in a job described by using "she/her" pronouns? The study did not find strong correlations of the kind one would anticipate, were it true that the generic masculine narrows the reader's perceptions to male persons. However, the study was limited in scope and only had a few pronouns in the job descriptions. Perhaps other studies would turn up different results? In fact, a 1983 study of the same phenomenon did find a decrease in the attraction to a job on the part of women who had read about it with a generic masculine pronoun, and an increase when they read about it with a gender neutral pronoun (Option Three above).

In a 1988 study in the same journal, Sex Roles, Wilson and Hung Ng presented ambiguous male and female faces to undergraduates, accompanied with a sentence containing the masculine or feminine generic pronoun. They predicted and found that the generic pronouns influenced the direction that viewers resolved the images (masculine generics predicted resolving ambiguous faces as male, etc.). However, just what their findings entail for ethically objectionable actions outside of the experimental setting is not self-evident. The implication is that using a feminine generic pronoun possibly primes readers to conceive of a female subject. Does this mean that using feminine generic pronouns instead of gender-neutral pronouns (Option One above) could counteract societal biases? This is unclear.

There are lots of fascinating studies relating grammatical gender and psychological processes. One study (Jakobson 1996) had Russian speakers select male or female-sounding voices to personify the day of the week. They chose speakers whose biological sex matched the grammatical gender of the weekday. Another study had students read passages in English or Spanish/French, and then measured their sexism via a test. The English readers had less sexism than the others, possibly due to the fact that gender was more salient in a language which distinguishes between the grammatical genders of its nouns.

Studies involving children do seem to show a strong priming effect with pronouns. Quoting from a summary:
Hyde (1984) found that when children were asked to write a story in response to the prompt “When a kid goes to school, [he/they/he or she] often feels excited on the first day,” (p. 699) only 12% and 18% wrote about female characters when he and they were used, respectively, whereas 42% wrote about a female character when he or she was used. (Prewitt-Freilino, Caswell, Laakso "The Gendering of Language" Sex Roles, Vol 66, 2012)
However, there are no definitive studies at present that I'm aware of which could give me sufficient reason to choose one option over the other. Thus I've made my decision based on mostly pragmatic reasons.

Tentative Conclusions and Reasons for Option One

Given the data that I've encountered (and I'll readily admit that I am no expert in this area), I have chosen to use the feminine generic throughout my academic papers for several reasons.

1. Employing the alternating option (two) requires that I keep track of pronoun use in order to keep it roughly equivalent. It also means that I would want to pay attention to the stereotypes employed by my choices of examples. If I am writing a paper and I use "she" to refer to an imaginary proponent of an argument, I want to ensure I don't too often use the feminine pronoun when describing an opponent whose view I think is weak or wrong. I want to ensure that I don't too often make the female actor in a thought experiment a stereotypically feminine person. And etc. Further, it seems that regardless of whether I've carefully paid attention to these features, my reader will likely overestimate the number of times I use the feminine pronoun.

2. Employing the gender-neutral pronoun is attractive to me for a few reasons. First, it recognizes the reality that not all human beings are comfortable within what sociologists call "the gender binary." There are people who would eschew traditional gender pronouns and be called by new ones ("ze"and "hir" for example). Second, it seems like it might take gender entirely out of the equation. Saying of a mechanic that "they should pay attention to detail" sounds like it would be a way to universalize without assuming most mechanics are men, or should be men, or etc. However, there are grammatical prescriptivists who find "they" repugnant, despite its long history as a gender-neutral singular pronoun (dating back to the 15th century, I believe). Further there is evidence---though I can't put my hands on the study right now---that "they" doesn't diminish the bias towards reading its referent as a male. Against the backdrop of my culture, "they" isn't taken to refer to men and women, but primarily to men.

3. Employing the feminine pronoun means that I don't have to keep track of pronoun use. It means that I don't have to give up (apparent) grammaticality. It means that for any example, I will have a feminine generic which can be read to encompass both men and women but, more likely, will be taken to refer to a female. While this, at first, seems like it simply replicates the original problem in reverse, I think this is not the case. When we get to a place in our society where, statistically speaking, women are not the subject of inequality, I think criticizing a feminine generic might be appropriate. Certainly if we got to a place where women dominated and oppressed men, a feminine generic would be a problem. But that is not the way the world is at the moment.

I realize that if I use the feminine generic throughout my papers, I risk some readers viewing me as overly "political" or as unnecessarily divisive, when I could simply adhere to tradition and use the generic masculine. I'm not under the illusion that my use of this pronoun will immediately change gender bias in philosophy or the world at large. I don't have the evidence for that. However, given the state of academic philosophy, I am content to be seen as rabidly feminist by my use of the feminine generic. If it means that it makes female subjects more salient to my readers rather than functioning as a "true generic", I am also content, as I doubt that philosophy will be harmed by this effect, but rather the reverse.