YWrite an analysis of intrastate variation
Your speaker: A speaker of your choice whose speech is publicly available in a controlled setting, like Oprah (a politician, talk show host or other public figure)
What you will do:
Formulate a question on hypothesis about this individual’s style-shifting. What social or situational factors do you think might motivate them to style-shift? (Some examples: the race or age of their interlocutor, the context (to the camera vs. to the audience), the topic (fashion vs. race relations)). I expect you to motivate your question or hypothesis using our readings and discussion about style-shifting. The factors you identify are your independent variables.
Select examples of the individual’s speech that allow you to test your hypothesis (search for videos on YouTube or another public platform). Remember that you need to isolate the social variables you are testing. This means you want two excerpts of speech that are as similar as possible and differ only with respect to the one social factor you have identified in your question. In other words, if you are interested in testing the effect of the listener’s race on the talker’s use of language, you would ideally select videos with listeners who are from two different racial backgrounds, but differ minimally on other dimensions like age, gender, socioeconomic class, familiarity to the research etc. Of course, this will be difficult to control if your talker is speaking to a large audience. Do your best to describe the differences in the audiences.
I recommend selecting about 10 minutes of material so that you have a nice dataset to analyze. This would mean that for a single independent variable like age of interlocutor, you would have five minutes of your speaker talking to a young person and five minutes of them talking to an older person. It’s ok if they talk more or less in different excerpts – just shoot for a balance of speech samples for your independent variable.
How to proceed once you have established your question:
Transcribe the talker’s speech. You will create an orthographic transcriiption of your speech samples. You should focus on presenting the content and form of your selected talker’s speech. You may also adopt transcriiption conventions and include a key at the top of your transcriipts (you can Google search “transcriiption conventions” for additional examples). It’s up to you if you want to transcribe the interlocutors’ speech – this is not required.
Notate. Notate your transcriipts for features of the talker’s speech that you observe to be shifting. These features will be your dependent variables. This task can be accomplished with no prior research on specific features you may expect your talker to use, although it may be easier to have some linguistic variables in mind before you begin. For examples of features of African American English, you should reference papers like Rickford and McNair-Knox (1994) and Hay et al. (1999). You might also consider looking at classic variables like (ING) (Fischer 1958, Campbell-Kibler 2007), t-release (Eckert 2008), or (r) (Labov (1966). I suggest making a first pass where you consider a range of variables – phonetic, prosodic, discursive, and morphosyntactic – whatever jumps out at you. Your final transcriipts can note only the variable you analyze, but you are free to note additional features that you do not quantify.
Quantify at least one dependent linguistic variable that you hypothesize demonstrates style-shifting. Present a summary of your quantitative analysis in a similar fashion as for Project 1 (i.e. at the very least, a cross-tabulation table and/or graphic that summarizes the relationship between dependent and independent variables. Statistical testing is not required).
If you like, note additional features of the talker’s speech or style-shifting behavior that you found interesting, if applicable. (Here, you have an opportunity to go beyond the basic quantitative analysis, if you like, and talk about additional differences in their speech that you find interesting).
Write-up your research following the same format and guidelines from Project 1. The paper should be about (2-3 pages, double-spaced with 11-12 point font and 1 inch margins) and should be submitted on Canvas as a .doc(x) or PDF by the due date. Append your transcriipts to your paper and include links to the clips you analyzed.
In addition to the basic analysis above, your paper is expected to engage the following questions:
How do your findings relate to the literature on intra-speaker variation and models of style-shifting? Do you adopt a particular model in formulating your hypothesis or in conducting your analysis? (*I expect to see you reference at least one approach to modeling intra-speaker variation in your paper, and discuss how your data relate to this approach.) Like for Project #1, strong papers will begin with a motivation that comes from the literature on intraspeaker variation that we have read and theories we have discussed.
What aspects of the talker as a speaker or social actor (their background, their current demographic identity (age, etc.), their current stylistic presentation) may impact their shifts? I recommend including some basic information on the talker’s background in your introduction or methods section (you can use Wikipedia for this purpose), but this should not be the majority of your discussion. This is a paper about style-shifting, not a biography of your selected talker.