In “The Social Weight of Spoken Words,” Dr. Sumner talks about the ways social representations triggered by variations in accent are activated early on during speech processing. These variations can have a subtle effect on the speed with which we understand words--although not necessarily because of the variations themselves. As demonstrated by Dr. Sumner and Reiko Kataoka, the social representations (or ‘prototypes’) triggered by different accents can result in different speeds of understanding, presence of false memories, and lack of/presence of priming. This is because different social representations command different levels of attention from us when we’re processing speech. In a parallel to social biases of today, accents that are associated with more marginalized social groups are given less attention than accents that are considered more prestigious.

The article “Condoleeza Rice and the socio-phonetic construction of identity acknowledges these issues and investigates the way individuals navigate them through examining the speech patterns of Condoleeza Rice. As a member of a diverse array of groups with characteristic speech patterns (Western, Southern, Politician, African-American, and Conservative), Condoleeza Rice faces the issue of which groups to align herself with when constructing her identity. This study showed that her while Rice’s speech patterns drew a bit from each group, they aligned strongly with those of the educated and conservative, perhaps demonstrating a subconscious (or conscious) recognition on her part of the weight each dialect holds and an effort to exploit these differences to her advantage.

In contrast to the Lev-Ari and Keysar study in 2010, this series of articles argued that the extent to which accented speech is attended to is not dependent on the ‘heaviness’ of the accent but the social representations with which the accent is associated with. Instead of a model where social bias results from difficulty understanding, these studies suggest that social bias may in fact cause this difficulty. One real-word implication I wish I knew more about was piqued by Sumner’s study (handed out in class) about how female speakers were rated as less trustworthy than men when they were speaking together. In light of the recent heated presidential race, I’m curious as to whether this effect can also be seen during the presidential debates when both a male and female candidate were speaking together--especially given that fact checking is such an important issue.