Jeopardy! as a Research Site

by Thomas Linneman

I never planned on studying uptalk (when people use the upward inflection of a question? When they’re making declarative statements?). I was interested in it as a pet peeve, noticing a rise in uptalk among my students over the years, particularly women students. To my older man ears, this made them sound unsure of their oral presentations. But I figured the phenomenon of uptalk was under the purview of sociolinguists, so I let it remain just a peeve.
Linneman_blogimage2However, one day I was watching my favorite game show Jeopardy! and I noticed some interesting variation. While contestants are required to give their responses in the form of a question, most of them used a flat, declarative intonation when answering (Alex Trebek: “This is the best journal ever.” Contestant: “What is Gender & Society.”). Around a third of the time, though, contestants gave their responses using uptalk (“What is Gender & Society?”), and it was typically women contestants who did so. I simply couldn’t resist the impulse to study this variation. It turned out to be my favorite project of my career (thus far), and certainly the most successful in terms of popular interest (The New York Times, Smithsonian Magazine, Women’s Health, Wired, GQ, NPR, Jezebel, and…Fox News!). Yet as fun and interesting as this project was to conduct, I did encounter some challenges along the way.

Challenge One: I wasn’t a sociolinguist. I had absolutely no training in sociolinguistics, nor had I read much in the field. I had conducted several content analysis projects over the years, so I developed a content analysis of episodes of Jeopardy!. I developed my variables, ranging from the demographic (gender, age, race) to the contextual (At the moment of response, how ahead or behind was the contestant? What was the gender of the contestant’s opponents?). I also started to read up on sociolinguists’ fascinating discoveries about gender and language.

Challenge Two: My attempt at garnering self-reported demographic data from the Sony Corporation (which owns Jeopardy!) proved unsuccessful. To the credit of the Sony executive to whom I spoke, I have never been told “no” so nicely in my life. To address this problem, I turned to a fun fact: nearly every television show has a ridiculously detailed fan website. For Jeopardy!, it is J!Archive, which catalogs every clue and response, and provides photos of the contestants. I distributed these photos to ten coders, who provided their estimates of contestants’ ages and races. I was then able to combine these into overall demographic measures. While these estimates are not ideal, they were the closest I could get. In retrospect, in keeping with feminist research methods, I also should have asked the coders to estimate gender, as my coding of this variable is my estimate alone.

Challenge Three: This typically happens to me whenever I am conducting a content analysis: I get a good way into coding my data when I inductively realize I neglected to code for an important variable. The emotional mix is intense. Yay: I have another potential finding. Boo: I have to go back and recode all the episodes I had already coded. Around fifteen episodes into coding Jeopardy!, I noticed an interesting phenomenon concerning correcting moments (when a contestant answers incorrectly, and then another contestant rings in and answers correctly). It seemed that men seldom used uptalk when correcting other men, but they frequently used it if they were correcting women contestants. Though a hassle to recode these episodes, it proved completely worthwhile to go back and add this variable.

Challenge Four: While the resulting analyses all proved interesting, and corroborated earlier research on uptalk (women use it more than men, young women use it more than older women, white women use it more than Black women, and people in general use it when they are uncertain), there were two context-related findings that stood out. The first I mentioned just above: men are much more likely to use uptalk if they are correcting women. The second: the further ahead a man was when responding, the less he used uptalk, but the further ahead a woman was, the more she used uptalk. The challenge here is one that is common for content analyses: how do you assign meaning to such findings? Based on my review of the literature, I made some educated speculations. For example, I argued that because the literature has shown that confident, aggressive women are unfortunately penalized in the workplace, the women contestants might be attempting to repair their gender performance by seeming less threatening (I’m kicking the pants off you, but I’ll use uptalk so that I sound unsure as I do it). But unless I talk to the contestants themselves, this remains merely a hunch.

Challenge Five: I thought having nearly 5,500 cases was enough. But then I developed more empirical questions that my data simply couldn’t address. Has the use of uptalk increased over time, and among whom? What if an older man lawyer is correcting a young bartendress? Should I have kept coding? Should I return to coding anew? And what should I say when someone with thirty years of Jeopardy!-laden VHS tapes contacts me? I’ll face that challenge when I come to it.

Thomas J. Linneman is associate professor of sociology at the College of William & Mary. He is the author of “Gender in Jeopardy!: Intonation Variation on a Television Game Show,” published in the February 2013 issue of Gender & Society.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s