Speakers, Semantic Association, and Social factors

Continuing from the class discussion, this week’s articles are centered around sociolinguistics, discrimination based on speech, and how social factors can affect comprehension. Sumner’s article talks about how different vocal cues that indicate things like class and accent can affect how individuals are understood. This is called social weighting. Reading this article is a great introduction into the Sumner/Kataoka and King/Sumner articles.

Sumner and Kataoka did a study where they examined 3 different dialects: General American (GA), Southern Standard British English (BE), and New York City (NYC). They aimed to see how well listeners recognized words when listening to non-GA accents. Also, their experiments investigated “the semantic priming of visual targets by related and unrelated primes produced by our three talkers for GA listeners.” In the end, it was found that when the subjects were listening to GA or BE, they had similar recall and recognition trends; NYC had a higher false recall rate. If targets were preceded by primes that were semantically similar, they had a faster RT than targets preceded by unrelated primes.

In the King and Sumner article, it examines speakers M and J. Speaker M is a white American female in her late 30s, and speaker J is an African American male in his early 80s. The goal is to see if words are interpreted differently depending on the speaker. The example they use is that the word “yeast” spoken by a man might trigger the word “bread”, but, if spoken by a woman, might trigger the word “infection”. From the first experiment, it was found that each speaker triggered a different semantic association. Another result is that listening to speaker M, the white woman, had more consistent and less varied results than speaker J. In Experiment 2, which examined voice matching, response time was not dependent on the speaker.

In all of these articles, there is sufficient evidence to suggest that speakers, whether the difference be between dialects, genders, or race, do play a role in semantic interpretation and comprehension. These findings, though, go against the Lev-Ari and Keysar study we examined in class. They concluded that voice characters do not influence comprehension. Personally, I agree more with these readings than the Lev-Ari and Keysar findings. It is very difficult to believe that the way someone talks and sounds does not affect comprehension.

I have experienced this many times before. Because English is not my parents’ first language, I often have to explain what my parents are trying to say to English speakers; however, their English is completely understandable. There was a time when my mother was asking a sales associate a question and the associate kept insisting that my mom wasn’t making any sense. Although I felt her request could easily be understood, I reiterated exactly what she said and the associate instantly understood. I believe she assumed that my mom was talking nonsense just because of her accent. This shows how vocal cues really do influence comprehension in the real world.