Sumner and Podesva

I was particularly struck by the studies we discussed during class in which “Native speakers were rated as more truthful than non-native speakers” and then “Native and ‘mild’ speakers were rated as more truthful than ‘heavy’ non-native speakers.” I wondering if these results can be extrapolated to the case where a General American (GA) speaker is compared to any other accent, whether American or foreign. For example, would a General American speaker find a speaker with a New York accent to be less trustworthy? 

I was fascinated by Sumner’s discussion in “Effects of phonetically-cued talker variation on semantic encoding” because it described effects beyond trustworthiness, such as memory effects and association affects. For example, in the false memory tests, when people heard New York speakers say all the words associated with “sleep,”  “listeners falsely recalled more lures for NYC sets than for GA or BE sets.” This leads me to concluded that something about the NYC accent causes people to only catch the gist of what is said and not to remember as well as when they heard GA or BE speech. When you only remember the gist of what was said because of somebody’s accent, do you feel that the speaker is less trustworthy? 

I am wondering if this is why, in Podesva’a paper, he describes how Condolezza Rice has essentially de-emphasized regionality in her speech. He states, “it is perhaps to Rice’s advantage to project a linguistic and symbolic identity that de-emphasizes regionally marked features in favor of features that are ideologically linked to ‘neutrality’ and standard language.” Has she altered the way she speaks in order to have a more General American accent that would therefore sound more trustworthy? Interestingly, Podesva found that Rice’s linguistic features “work together to present a coherent picture of a professional African American public figure.” Her speech is “largely aregional,” “the only Southern vowel feature she exhibited was the fronting of the back vowels, which…is characteristic of a number of dialect regions in the United States, including the West,” her vowels are “the picture of standardness,” she “releases word-final voiceless obstruents at rates exceeding those of other speakers when reading” and there is “a connection between this feature and literacy and education,” and she uses a backed BAT vowel which “appears to have a more general indexical meaning in U.S. English: “correct, educated, and sophisticated.” So, in sum, Rice has decided to speak in a way that benefits her position the most and conveys trustworthiness. I have concluded this because of Sumner’s argument in her paper “The social weight of spoken words” in which she states, “Listeners readily map sound patterns in speech to social representations. This mapping introduces social biases on the recognition and encoding of sound patterns produced by different groups and individuals.” Therefore, Rice’s language is most likely to be mapped to produce the least social biases, regardless of people’s already-held beliefs about her actions as a politician.