Today, I got up (way too) early to go to a conference being held by the Sorbonne on music sociology. It was apparently the 12th International Colloquy of Art Sociology, but as is the case with most conferences, it wasn’t particularly international. It was at least 90% French scholars doing very French sociology.
Anyway, I went to the conference because this year’s theme was “25 years of Music Sociology in France”. I had looked at the schedule online and there were a least one or two papers that I was interested in, plus a plenary session by Simon Frith (a very famous popular music scholar from the UK).
I had been under the impression that French sociology (and most continental sociology) was far less statistics-driven and scientistic about their work, leaning heavily on Critical Theory and working on cultural texts in the way Anglophone anthropology works with ethnography / interviews. Boy, was I in for a surprise. Here is a archetype of the sorts of papers I heard today:
“I gave a survey to fans of rap music about their tastes in rap music. Here is the number of people I interviewed, here are the questions I asked, and here are a few tables with lots of numbers on it. People of low social status preferred one set of rap groups and people of high social status preferred another set of rap groups. Therefore, social class reproduces itself within musical genres that are commonly thought of as generally ‘popular’ or ‘working-class.’”
So, in other words, they are exactly like Anglophone sociologists.
Don’t get me wrong; there are certainly benefits to quantitative data and statistics. But there are also some risks and gaps in that approach, such as mistaking correlation for causality. For example: just because lower-class folks tend to like one set of rap artists doesn’t mean they like it because they’re low-class; there can be a whole array of reasons for why you get these sorts of results. Some of this can be mitigated by employing more nuanced surveys, but this can also be achieved by talking in less structured ways with people directly involved in the phenomenon and reading their responses closely. Also, hard-structured surveys and other forms of quantitative data collection presumes that the best way for people to tell their stories is through a frame that you’ve imposed; in other words, you may want to know how their taste in rap is determined by their class status, and they may want to tell you about how rap plays a part of their everyday social life. One may inform the other, but the information is ignored if the isn’t a box for it on the survey form.
Anyway, there are different ways to create knowledge about people and statistical data has its uses, but it’s certainly not the kind of work I’m interested in doing. I did actually see two papers that were interesting—one of them survey-based and the other a rare theoretical paper—but the majority of my day was spent quietly listening to people make banal observations of the obvious.
On the upside, the reception at the end of the day had delicious hors d’oeuvres and an open bar with champagne. Yay.