Saturday, June 30, 2007

iTunes U: Martin Lewis on historical linguistics

This blog seems to reach an unsettlingly hip demographic, so you probably already know about iTunes U (if not, see here), basically 'enhanced podcasts' on educational topics. (These have podcast audio, with some visuals, slides of the PowerPoint sort, no video; see this.) Heck, you're probably listening to one right now on your new iPhone as you read this blog on another iPhone. Well, maybe our audience is more the healthy kind of hip than rich and trendy.

Stanford University has a particular presence on iTunes U, including a course on the Geography of World Cultures, by Martin W. Lewis. I knew about iTunes U but hadn't listened to /watched any of them until a colleague told me that this particular course included a number of lectures on historical linguistics with some nice maps. If you like maps and historical linguistics, you owe it to yourself to check these out.

This series shows the great potential of enhanced podcasts. At the same time, it underscores the need for real expertise in both of the fields at hand here: As far as I can see (= as a non-geographer), he does a great job on the geography, like the spread and areal distribution of languages. Historical and cultural geographers seem to be masters of sketching the broad strokes of history nicely, including here archeology and such, and tying it to space. Needless to say, the maps are great, for the most part (a few are fuzzy, etc.)

While Lewis knows a ton about languages, the thing is shot through with claims that will make historical linguists jump, like this: "Languages don't change as much as they used to." I don't know of any good evidence to back up such a claim and am not sure how you'd get it. Looks like just a naked assertion.

Lewis seems to rely heavily on popular works. In particular, the various references to Nicholas Wade's most recent book, Before the Dawn (2006), will outrage some linguists, given the widespread griping about Wade's reporting on language for the NYT. He draws on John McWhorter's Power of Babel as well. McWhorter is (or was, before he became a conservative think-tank guy) a professional linguist and he knows his stuff, but I'd argue that you need a tremendous amount of background to understand what he's actually doing in his popular and professional work — there's a lot of provocation in his writing that a non-specialist will probably read right past, and agendas that you might not pick up on all the time.

Sadly, there are a few howlers in the lecture. Two examples from pretty early on (with an indication of roughly where they come in the hour and a half lecture):
  • around 18:30: "How did English actually become a language?" He says point blank that it was the Norman invasion. He goes on to declare Modern English grammar heavily Romance (as well as heavily Germanic). Ouch.
  • around 21:30: Greenberg's take on genetic relationships among languages (basically, the belief in Proto-World) is presented as a minority view but not as utterly insane. There's not even mention at the pervasive and fundamental patterns of errors: students need to know that the most basic data Greenberg presents just can't be trusted at all.
  • And you could write a long post on his treatment of clicks (it runs through much of the lecture), but that's something for another day ... .
Anyway, I feel bad saying negative stuff about what's basically a nice piece of work. Again, enhanced podcasts represent a powerful tool, but I really wish he'd done these lectures together with a card-carrying historical linguist, or at least checked the content with one.

Doff of the summerweight tuque to Professor V. (no relation) for alerting me to this series.


Joe said...

I'm convinced that almost everything of real interest in our field at this point is coming from collaboration, and having just listened to that lecture, I'm impressed by how much lingusitics he DOES know. In a sense, it's not so much Lewis's shortcoming but maybe a matter of him trying to do more than we can expect from a single person.

For example, it's really hard to make people who aren't trained in historical linguistics understand why Proto-World is so crazy and the idea seems like it's just too tempting for people like him to reject as the lunacy that it is. And for the record, I dismantled the very example he gives of Proto-World, namely *tik:

Salmons, J. 1992. A look at the data for a Proto-World etymology: *tik 'finger'. Explanation in Historical Linguistics, ed. Garry Davis and Gregory K. Iverson. Amsterdam: Benjamins. (Current Issues in Linguistic Theory, 84.) 207-228.

Mr. Verb said...

Oh, I'm definitely planning to recommend that my students listen to or watch (what do you say for an enhanced podcast?) this, though I might well give them a few notes on points that strike me as problematic.