Professor leads NSF-funded study of double negative use among American dialects

UNIVERSITY PARK, Pa. — Grammarians have long maligned double negatives as an incorrect use of the English language. But a new study led by Frances Blanchette, associate research professor of psychology and assistant director of the Center for Language Science at Penn State, seeks to bring some new perspective to the argument.

Blanchette and three fellow researchers recently received a $384,272 National Science Foundation collaborative research grant to fund their study, “A Microcomparative Study of Negation in Dialectal Variation,” a continuation of Blanchette’s longtime research on linguistic diversity among American English dialects.

Linguistic variation — the use of different words, sounds or grammatical structures to communicate the same meaning — is a deeply ingrained part of American spoken English, Blanchette said. For instance, depending on where you live, a sweetened carbonated beverage might be referred to as “soda,” “pop” or “coke" (regardless of whether or not the beverage in question is a Coca-Cola). 

The same linguistic diversity applies to the use of double negatives, or Negative Concord, as they’re technically known. While one speaker of American English might say, “I didn’t eat anything this morning,” to express that they skipped breakfast, another might say, “I didn’t eat nothing,” to convey the same thing.

While most people can understand what the second person is saying, the use of double negatives is often looked upon as improper speech, carrying with it a significant social stigma, said Blanchette, who has devoted substantial research to double negatives and in 2020 gave the TEDxPSU talk, “The Grammar of Double Negatives.”

“When someone says ‘I didn’t eat nothing,’ other people may respond with “You don’t sound intelligent; that’s the wrong way to speak,’” said Blanchette, the study’s principal investigator. “Language is one of the last domains where overt discrimination is actually acceptable, because people don’t see it as discrimination. However, double negatives are actually very common in American English and most of the Englishes in the world. Historically, we see it in Shakespearean English."

Blanchette explained that the language evolved over the centuries, people in higher social classes stopped using double negatives, and they became stigmatized. To this day, though, they are still used in most colloquial Englishes.

“My previous research has shown that even if you say you don’t use double negatives and think they’re totally unnatural and unacceptable, you can understand them perfectly well in context,” Blanchette said. “People more easily interpret ‘I didn’t eat nothing’ to mean ‘I ate nothing,’ as opposed to ‘I ate something,’ even if they say they don’t use it and say it’s bad English.”

For the study, Blanchette and her collaborators — Cynthia Lukyanenko of George Mason University, Paul Reed of the University of Alabama and Jessi Grieser of the University of Michigan — will collect data from speakers of three varieties of American English: African American English and Appalachian English, where double negatives are common; and Mainstream English, where they are not.

Through comparison, the researchers will seek to find areas where the three groups overlap in their linguistic knowledge, as well as pinpoint any differences that exist. Specifically, they’ll examine prediction of upcoming words or structures and the association of social information about contexts in which a structure is likely to appear. Their methods will include phoneme detection, which is a measure of prediction, and eye tracking, which can measure real-time comprehension and association, including social knowledge.

“What happens when you do the same experiments with three groups of people — two groups who use double negatives, and one group who doesn’t and says they’re really bad?” Blanchette asked. “Does this finding that people nevertheless understand them hold and do the patterns look similar across the three groups? Or do some differences emerge? It speaks to the bigger question of what is really happening in a context where you have different varieties spoken. We can usually understand each other pretty well, even if we don’t speak the same variety. What are the differences we notice or don’t notice, and what do they mean for how language exists in our minds?”

By working with participants from African American and Appalachian communities, Blanchette and her colleagues will collect much-needed data on speakers who are typically not included in experimental research on language, the researchers said. In addition, the project will offer undergraduate and community researchers from underrepresented minority groups the chance to gain hands-on experience.

Blanchette said the group should be ready to present the study’s results to the public in about three years. The expectation, she said, is that the research will demonstrate that the three groups of speakers have much more in common linguistically than what is conventionally assumed.

“We’re just starting to use experimental methods to look at issues of linguistic diversity and what it means to speak more than one variety of a language,” Blanchette said. “This project will try to get more information that will help us understand linguistic diversity and appreciate it, and we think that makes it a very worthwhile project.”