Linguistics 001 Fall 2001 Final Project Due Mo 12/17
In plain language: explain something about how a piece of talk works.
More exactly: analyze the communicative effects of some aspects of one or more linguistic performances, attending to at least two different levels of linguistic analysis.
"Analyze" means "explain on the basis of specific examples and general principles."
"Communicative effects" means "the (likely) effects on a real or hypothetical interlocutor or listener." This can be construed very broadly: the interpretation of the literal meaning; the perception of dialect; the relationship to social stereotypes or cultural norms; the generation of an emotional response; the control of turn-taking; or many other things.
"(Some aspects of) one or more linguistic performances" means (in the simplest case) a recording of a conversation, speech, reading or monologue. You can also compare two such recordings, or look at properties of a large set of such recordings (though this last is probably too ambitious for a project of this size). Songs are permitted -- but your analysis should be primarily linguistic, not musical. In principle, you can start with a written text (such as a transcript) rather than a recording, but a recording is better. You might also use a set of written texts for background analysis, with a more detailed analysis of a specific passage you look at in spoken form. (Note that "performance" doesn't mean "acting", it is just being used as a term to cover everything from ordinary conversation to a formal political speech.)
"Attending to at last two different levels of linguistic analysis" means that you should discuss two or more of the standard levels of linguistic analysis: phonetics, phonology, morphology, syntax, semantics, pragmatics. The two levels might be connected to a single communicative effect, for example if you looked at the use of pitch range and sentence structure in quoted dialogue; or the relationship of pronunciation and lexical choice to dialect perception; or the rhetorical use of timing and parallelism. However, you can also consider several linguistic aspects of a passage with respect to several different kinds of effect.
All in all, this is a very loose requirement. The main constraints are: deal with (at least one) specific spoken passage, and look in detail at the relations between linguistic form and communicative effect. There should be some sort of overall point: it's not enough (for example) to break a passage down into morphemes and also show its syntactic structure.
Q and A
Q: When is the project due?
A: By the scheduled time of the final exam, which is December 17. You can hand it in before then, and we will be grateful if you do so.
Q: Can we do joint projects?
A: No, everyone should hand in an individual project report.
Q: Can you give some examples of plausible projects?
Q: Does the linguistic performance have to be in English?
A: No. But it needs to be in some language that the instructor or one of the TAs knows well, or else the analysis needs to be compelling and very clearly explained.
Q: What should I actually hand in as my final project report?
A: An ordinary printed document is fine, though you might choose to create a multimedia document (say in html). In the latter case, it will probably be easier for us if you can put it on line (e.g. in http://www.sas.upenn.edu/~yourname/foo.html). In any case, you should include a complete transcript (at least of the relevant sections). If your analysis focuses on aspects of the sound, then we should have access to the sound (preferably in digital form).
Q: How long should my project report be?
A: The short (though not very helpful) answer is "long enough to make your point." Keep in mind that this project is substituting (relative to last year's course structure) for two homeworks and half the final exam, so you should not expect to be able to do a good job in an hour or two.
Q: How should I approach the project?
A: Start with a selection of material and a short but clear statement (for your own purposes) of the main point of the analysis. Do the detailed analysis before you start to write. A transcript is a good place to start, with bits of linguistic analysis keyed to pieces of it. In most cases, you should expect to spend more time doing the analysis than writing it up. You don't have to limit yourself to linguistic analysis -- an interesting discussion will almost certainly bring in non-linguistic facts and themes -- but the linguistic analysis has to be there.
Q: What are some things to avoid?
A: Don't wait until the last minute to start the analysis (then discovering that you can't figure out how to download an audio file, or that you can't get a transcription program to work for you, or that you don't understand how to do a particular type of analysis). Don't spend too much time messing around with audio formats and speech analysis programs (unless you like to do that for your own recreational purposes), and too little time doing your analysis.
Q: What help is available?
A: For guidance or feedback in picking examples, checking analyses, etc., contact your TA. For help with audio file formats and programs, contact your ITA. As a last resort, send email to Mark Liberman or come to his office hours.
Audio formats and programs
These examples use a variety of formats. In most cases, you should be able to play these by double-clicking on them, and download them in the usual way (e.g. right-click>Save Target As in a browser under Windows). For more careful scrutiny, open formats such as .wav, .mp3, .au are accessible to programs such as GoldWave, CoolEdit, Transcriber and WaveSurfer. To make this possible, first install such a program, then download the audio selection in question to your own machine, and then open the chosen selection in the chosen program.
The free software program Transcriber (available in Windows, Mac and Linux versions) is particularly recommended for producing basic transcriptions. If you want to measure durations, look at pitch contours etc., contact Mark Liberman before the end of classes and we will set up a tutorial session.
There are various proprietary formats from RealMedia and Microsoft that may be more difficult to access conveniently other than by playing or streaming from an on-line site. As a result, you will probably want to avoid material that is only available as RealAudio or similar, because you'll have a hard time even making a transcript, and will not be able to make any phonetic measurements (such as pause length) without using techniques that violate the DMCA. For audio format questions, please ask your local ITA first.
Feel free to find (or record) your own examples. Links to a few somewhat random examples are given below, and more will be added later. You can find many spoken-language CDs in the library, or buy them at stores or online.
Short selections of
20th Century U.S. Presidents from the Vincent Voice Library.
Basil Rathbone reading Poe's The Raven.