When was the last time you heard someone speak in an “accent”? What did you think about that person? Did their accent tell you something about them? For many of us, the answer to that last question is “yes.” It’s no secret that no two people who speak a language speak it identically, but the difference between two people’s speech can range from quite similar to so different that one might as well be speaking in a different dialect. In our globalized and internet-penetrated world, we’ve grown used to hearing and interacting with people who speak our language in a markedly different way, which has led to the construction of stereotypes about people who speak in what we’d consider to be “accents.” Accent is in quotes here because it’s generally used to describe a collection of audible speech variables that is different from that of the speaker. This definition is somewhat misleading: Midwestern Americans may say Brits have a British accent, but those same Brits would likely say the Americans have an American accent. Hence, every version of a language can be considered an accent.
Knowing this linguistic reality doesn’t change the fact that many still consider “accented” speech as “different from mine,” a mindset which promotes stereotypes and develops from a young age. If we want to reduce voice-based discrimination, it helps to understand how children develop these discriminatory patterns, and to do that, we need to know when children begin to differentiate between accents and what social judgements they make on them. Some linguistics researchers have began to explore this developmental trend; Kinzler et al. and Creel in 2009 and 2016, respectively, found that five-year-old children would rather be friends with others their age who speak the same version of English as them, a factor that even outweighed skin color in importance to the predominantly white children studied. These studies concluded that this preference was likely due to the children’s preferences for familiarity, and Kinzler et al. continued to suggest that such accent preferences are prioritized because ancient humans differed much more in accent between groups than in skin color.
While we could accept these findings and agree that, at minimum, children must prefer people who sound familiar to them, these studies’ methodologies (Creel’s experiment is a slightly modified recreation of Kinzler et al.’s) have a poignant flaw: in each, researchers present children with two differently accented voice recordings side-by-side and ask them which speaker they’d rather be friends with. This type of test is known as a forced-choice test, and it leaves the children’s feelings toward the unselected speaker up to interpretation. Do they dislike the differently accented speakers? Do they dislike both but like the natively accented speaker slightly more? Is there something other than accent guiding their preferences? Are they indifferent? These questions are important because with the current data, we don’t know how prejudice and preference are weighed in these children’s friendship decisions. If someone were to design an educational intervention to, say, make children more willing to be friends with those who speak differently than them, the designer would not know whether to deconstruct preference of familiarity, prejudice toward unfamiliarity, or some mixture of both. A successful intervention would promote acceptance of “accented” speakers throughout society; we all have accents, and there’s nothing wrong with speaking one way or another, regardless of how familiar that speech is.
In 2010, linguists Whitely and Kite critiqued the commonly used “Preschool Racial Attitude Measure” (PRAM) test, a forced-choice racial bias test developed by John Williams and J. Kenneth Morland (1976) that has similar pitfalls to the forced-choice accent-preference test used by Kinzler et al. and Creel. In the PRAM, tested children look at two drawings identical in all features but skin color, one of a black-skinned boy and one of a white-skinned boy, and are either told to choose which one is a “good boy” or choose which one is a “bad boy.” Whitely and Kite argue that even if children reliably associate one skin color with positive adjectives and the other with negative adjectives, there is not enough evidence to confirm those decisions are due to racial bias because arbitrary factors may be at play (i.e. the participant may say, “I chose that child because he has friendlier eyes than the other child).
Whitley and Kite proposed two solutions that could allow an evolution of the PRAM to more effectively assess racial attitudes without leaving the presence of racial bias up to assumption. One of these is to use a continuous measure of prejudice to assess such attitudes; the authors claim that questions like “Would you feel comfortable living next door to a black family?” rated on a multi-option scale from “No, definitely” to “Yes, definitely” would not force children to choose one skin color over the other and provide a clearer look at how they felt about different races.
The same logic applies to the forced-choice test used in Kinzler et al.’s 2009 study and Creel’s 2016 study. Instead of a design that leaves part of children’s language attitudes up to assumption, I suggest modifying Creel’s methodology (which itself was an evolution of Kinzler et al.’s) to use a continuous measurement of listener attitude for each voice heard. In Creel’s study, researchers showed participants pictures of two children’s faces on a computer screen, had them listen to a sentence of the same words in a different accent form each, and asked them which they would rather be friends with. In my variation, researchers would present participants with one picture and one voice (either in their native English accent or a foreign one) at a time and have them rate their friendship preference along a five-step scale of emojis (big smile, small smile, neutral face, small frown, big frown) before moving onto another picture with the other accented voice. For the sake of simplicity, I’ll refer to these emojis as numbers 1 through 5, with 1 being a big smile, 3 being a neutral face, and 5 being a big frown.
This modification uses one of Whitley and Kite’s suggested alternatives to forced-choice testing, allowing researchers to better understand the true nature of young children’s attitudes toward familiar and unfamiliar varieties of English. Results from this modified study could either affirm and refine previous findings, or reject and contest them. In the former scenario, a statistically significant number of participants would indicate that they prefer to be friends with children who sound similar over foreign-accented ones. This level of preference could range from a 1 for the local speaker and a 5 for the foreign one to a much more subtle difference, like a 1 for the local speaker and a 2 for the foreign one. Any result like this affirms existing research and refines it by revealing tangible differences in participants’ feelings toward both speakers. Results that reject existing research would show that participants mostly felt the same about each speaker or actually preferred the foreign-accented one; in either case, the new results would contest existing ones, which could be due to methodological flaws in the forced-choice method.
In either case, with more accurate language attitude data, educators could more effectively design educational interventions to decrease voice-based and other forms of discrimination. If such interventions are successful, we may see positive changes in inclusiveness and tolerance trends throughout society. Different doesn’t have to mean worse, and this research puts us one step closer to that reality.
Works cited (APA):
Creel, S. C. (2016). Accent detection and social cognition: evidence of protracted learning. Developmental Science, 21(2), 1-12. https://onlinelibrary.wiley.com/doi/abs/10.1111/desc.12524
Kinzler, K. D., Shutts, K., Dejesus, J., & Spelke, E. S. (2009). Accent trumps race in guiding children’s social preferences. Social cognition, 27(4), 623–634. https://guilfordjournals.com/doi/abs/10.1521/soco.2009.27.4.623
Whitley, B. E., & Kite, M. E. (2010). Psychology of prejudice and discrimination. (pp. 285-287). Boston, MA: Cengage Learning.