You Are How You Speak (in a court of law)

 This week’s readings have gotten me particularly further interested in the concept of how the way in which one speaks can influence others’ perceptions of them in major ways.  Specifically, all the three readings (the 3 Sumner articles) all kept making me think about previous issues of identity perception and judgment through speech that have been discussed in class, particularly with regards to the issue of the court testimony that we talked about in class, and, upon reading these articles, just many factors can influence the effects of one’s speech beyond the actual content, that all add up to have a major effect on how one is perceived and understood on the whole.

The Sumner Trends in Cognitive Sciences provides the perfect open-ended backdrop against which to begin analyzing these issues. It discusses the social function of words and how spoken sound patterns can indicate to the speaker, even unconsciously, of social tendencies and biases, and thus greatly affect the way we map those words consciously and draw linguistic connections. 

Voice Specific Effects in Semantic Association, by Sumner and King, continues along this idea with the more specific exploration of how we as humans understand and make connections from the speech of others based on the specifics of their voices.  Using the free association and semantic priming tasks, it showed that there are relationships between the semantic effects of a word and the speaker, and uses this for its argument that the social information and resulting effects of voice details of a speaker need to be considered when studying the models of word recognition that currently exist.  This made me think about, as the TiCS suggested, that these associations hold powerful social effects, which can be effective for quick processing, but also make me think about the possible negative effects of such assumptions. Such as in the court testimony example, I wonder how much of what the witness remembers hearing or understood from the situation was due in her memory to such social associations with semantics. Likewise, what kind of issue might similarly arise in the court’s understanding, trust, and biases of the testimony based on the witness’s own semantic variations. What kind of associations might they draw from her voice that would be different if she spoke like, say, a young white man, or an elderly Hispanic woman?

Sumner and Kataoka’s 2013 article Effects of phonetically-cued talker variation on semantic encoding provided even more evidence to me for such concern. Here the effects of frequency of hearing certain dialects were tested in tasks of semantic priming and false memory. The chosen forms of English were a General American (GA), New York City (NYC), and a Southern Standard British English (BE). Speakers of each of these types provided spoken word lists that were then listened to by GA-speaking students who were then tested. Here, what most struck me was the issue of false memory, particularly the higher rate of false memory in NYC speakers in the test’s results. In thinking about the same issue of the effects of speaker type on social biases and perception, this adds another dimension of concern to the previous issue. Now we must think, in the example of the court case, not only what unconscious stereotypes may be taken from the speaker’s patterns, and what kind of associations might be made based on their specific voice, but what is the possibility of false memory or perception due to the dialect of the speaker? This creates a serious concern, especially in criminal cases and similar situations. It makes me wonder if it would be possible/more likely for a false witness testimony to result simply due to a difference in frequency of hearing the accused’s dialect? Or what conclusions may be drawn about them or their intentions based on speaker attributes? And what if the same were to affect the credibility/effectiveness of the witness themselves?

There are a lot of issues at stake in situations such as the court case discussed in class, where ambiguity of testimony facilitated a major point for the prosecution. But beyond this, and even only in the specific realm of linguistic issues (aside from other major social and technical concerns) there proves to be an incredible complexity and potential for error that adds up to how we perceive a person. This should be considered with regards to a defendant, victim, or witness.