Social Biases in Speech Variation

I read Voice-Specific effects in semantic association and Effects of phonetically-cued talker variation on semantic encoding as well as the TIC. The first paper found that speaker-specific phonetic cues affected the interpretation of spoken words. Listeners responded differently to the same words spoken by different people. The speakers varied in age, race, gender, and dialect background. The second paper investigated phonetic variation on semantic encoding. This focused on how variation in speech affected how well the words were encoded in the listener's memory.  Clear speech with focus on certain letter pronunciations led to better encoding. Words were equally recognizable even when the frequency of them differed. General American listeners did tasks while listening to General American speakers, British English speakers, and New York City speakers. The general american and british english showed strong semantic priming and low false recall rates as opposed to NYC speakers.

In Sumner's social weight of spoken words, she talks about how speech is highly variable, yet when well encoded, it is well understood. This speech influences our biases - our goals, desires, and social experiments. Listening to a speaker exhibiting a certain variation in speech influences how we think about the speaker, what they are talking about, etc. That is what the other studies are about - how well words are encoded depending on speakers and how the specific reactions people get when listening to the same words uttered by different speakers.

For me, this is all about the social biases when it comes to speech variations. For example, in the movie Good Will Hunting, the Boston Southside accent easily gave away some of the characters in the movie. By the Harvard students, they were judged for this speech variation. Of course they were completely understood, but based on the variation some other people thought less of them. This goes for all kinds of speech variations - like how a lisp can be considered by some as an indicator of intellect, which is false. Overcoming these variations in speech can be done - especially as shown by Sumner that accents do not have a large impact on encoding of words. What is difficult to overcome is the different reactions people get from different speakers saying the same words. This is the bias.