Experimental Phonetics, Trinity Term 2021

Course outline

Convenor: Prof. J. S. Coleman
Thursdays, 2–3 p.m., weeks 1, 3, 5, 6 and 8

Target audience: First year MPhil students interested in offering the experimental phonetics option paper B (vii) in Trinity Term 2022, and second year MPhil students who started it in their first year and are working towards completing it before the week 5 deadline this term.

These classes will be conducted entirely online. As usual, I shall provide course materials (handouts and presentations) via my website, as I usually do, and to arrange the group meetings at the usual meeting time of Thursday 23 p.m. in weeks 1, 3, 5, 6 and 8, using Microsoft Teams for videoconferencing. In weeks 1 to 3 I shall also post three pre-recorded lectures on selection of topics in experimental phonetics on Canvas, for you to watch at your convenience. These do not relate very closely to the group experiment, but the works presented there provide good examples of the conduct of different kinds of experiment, and good models of how to write up the results in a report.

The rubric for assessment of paper B(vii) Experimental Phonetics is by submission of "A written report of between 5,000 and 7,500 words on the design and execution of an original research project." "Original" here means that the experiment or study has not been done before in quite the way we do it; in fact we shall be taking note of two prior studies of the exact same topic. "Original" does not mean that every candidate has to design and carry out an entirely separate project from everyone else (that is permissible, but I do not encourage it because of supervision considerations); it is not, after all, a second MPhil thesis! Rather, in the meetings we shall work together collaboratively as a research group; I shall suggest the topic of the investigation, and provide research training in the methods required to carry it out.

In some years, the classes focus on design of the experiment, and may get little further than setting up a pilot with a few subjects. In other years, we use data that is already available, having been collected previously, and the classes focus on questions of measurement and analysis in order to test one or more hypotheses. This year, because it is still not possible to invite subjects into the lab to collect data, we shall use pre-existing data: we shall again work on the acoustic-phonetic correlates of stress in Modern Spoken Welsh, using audio data that we found in a corpus last year, and made a start with analysing.

After planning what we need to measure in order to address the hypothesis/es, we'll divide up the work of making the measurements between everyone in the class (me included). By sharing it out, we can work on more data than one person could do individually; we will then pool our measurements together, and I will guide us through the statistical analysis of the data. Upon that guidance, students will carry out the statistical analysis themselves, independently: this is because the write-up of your own analysis of the data will be an important part of the report (assuming that you decide to do it as an assessed option). If you do decide to take the option, you could then work on the write-up during the Long Vacation. I can give individual feedback on one pre-final draft of the report, before you then polish it up and prepare it for submission. Since I shall be on research leave from October onwards, it is advisable for you to have completed your first draft before the beginning of Michaelmas Term.

The programme for the course is roughly this:

I shall be posting the video presentations on Panopto, accessible via the rather minimalist Canvas page for this course (https://canvas.ox.ac.uk/courses/75865)