Linguistics 220: Language and Society

Fall 2020



Style Assignment (22.5%)

Part 1: Project description and data links (2.5%) Due to Canvas by 11:59 p.m. Thursday October 22

Part 2: Write-up (20%) Due to Canvas by 11:59 p.m. Thursday, November 5

This project will investigate stylistic variation in one speaker (intra-speaker variation). You will analyze the same speaker in two different stylistic contexts. Data will come from existing public recordings -- you will select a public figure (a Youtube personality, a celebrity, a politician, etc.) and analyze how some aspect of their speech varies between the two recordings/clips, connecting your findings to the differences in the context or personae the speaker is aiming to portray in the different settings.


Data and Analysis

The Data: You will select one public figure whose speech has been recorded in multiple settings and is publicly available for viewing. These will likely include celebrities, Youtube personalities, politicians, etc.

Select two recordings, with at least 10 full minutes of speech from your speaker in each recording. You should select recordings that you think capture the speaker in different speech settings, or in which they are portraying different personae or stances. You will be looking at how these differences in context affect linguistic and extra-linguistic differences in style. Please do not select trained actors who are explicitly portraying characters in TV or movies for your clips. We will be discussing parody and high performance later in the quarter, but this is not the purpose of the present assignment.


Quantitative Analysis: Select at least one particular linguistic feature that you'd like to examine quantitatively across the two contexts. This can include particular segments (consonants, vowels), voice quality (creak, falsetto), intonation patterns, morpho-syntactic forms, lexical variables, discourse markers (like, ya know, filled pauses like um/uh) etc. Click here for an overview of frequently analyzed variables for this project, along with explanations of the variants for each, some notes, and examples from sociolinguistic literature. This document also includes variables we do not recommend for this project. You may select a variable not on this list, but please review the tips for selecting a variable in the document - it's important that you pick a quantifiable variable that you are comfortable counting auditorily. Depending on the length of the recordings, sound-related variables will likely be easier to quantify, since you will most likely get more instances of each variable. You can conduct your analysis by ear, reporting in a similar way to the quantitative analysis assignment (include a table with proportions and raw total Ns, and visualization (chart/graph) of your quantitative results in your paper).


Qualitative Analysis: Observable differences between the two contexts (visual, auditory, historical, etc.) that might not be quantifiable are just as important to an analysis of style and personae. In addition to your quantitative analysis of a linguistic variable, you will need to describe aspects of the linguistic and/or non-linguistic styles that might not be "countable," but appear to contribute to differences in what type of persona the speaker is projecting in each clip. This could be something about the content or discourse context, something about visual or gestural differences between the two, etc. You should support this with specific and detailed examples in your write-up.


Select your recordings and submit the links to Canvas by 11:59 p.m. Thursday October 22nd with a paragraph description of the linguistic feature you plan to analyze quantitatively, as well as the "non-countable" qualitative difference(s) you plan to analyze. Use the frequently analyzed variable document for help selecting a variable. You will receive feedback and suggestions prior to completing the write-up.


Paper

You will write up a short research paper on your findings, approximately 5 pages, double spaced. The paper should contain the following sections:

  1. Introduction: Describe the speaker you chose and any background about the speaker that's relevant to your analysis, the recordings you chose and why you chose them, and the features you chose to examine (the quantitative variable and variants, as well as the qualitative aspects).

  2. Methods: Describe the methods you used to conduct your quantitative and qualitative analysis.

  3. Results: Present your findings. Include a table with proportions and raw total Ns and a chart (visualization/graph) that plots proportions, for the quantitative portion. Include specific examples with timestamps and/or detailed descriptions of your qualitative features for your qualitative portion.

  4. Discussion/conclusions: Interpet your findings. Why do you think you see differences (or no differences) in the two different contexts? How do your findings relate to the theoretical discussions we've had about styles and personas? Draw upon readings and concepts we've discussed in class to frame your claims. If you refer to readings, cite them in the text e.g. (Labov, 1972) and include a references section at the end of your paper. Use any citation style (APA, MLA, etc.) as long as it's consistent.

As always, in your interpretations, be sure to only make claims that you can substantiate with the data you have. You are welcome to describe the additional information or data required to confirm potential interpretations as well.

Submit your finished paper via Canvas by the deadline as a PDF or Word doc.