Book Review: Talk on the Wild Side

Talk on the Wild Side: Why Language Can't  Be Tamed
Lane Greene
Linguistics

I'd probably like Talk on the Wild Side better if I could pinpoint what it's about. Nominally, the topic is something like "how do people react when confronted with the ambiguity of natural language"? If that sounds a bit unfocused, you've put your finger on the issue.

It begins tamely enough, with a couple chapters lightly bashing linguistic prescriptivists--material that will be familiar to readers of Stephen Pinker and John McWhorter. There's a long digression where Greene takes pot shots at one particular grammar loon, one who's apparently a big noise among the kind of critics who conflate crankiness with eloquence. It's amusing enough. There's a more sympathetic pass at the people who have tried to create genuinely logical artificial languages (Esperanto, Lojban). That's interesting, although the only point seems to be that not many people speak these tongues.

Then we start getting into the tangled history of machine translation, symbolic AI, and deep learning. Okay, that's a sidestep, but I see the connection. Natural language is chaotic and ambiguous and hasn't been "tamed" by rule-based methods (current translation algorithms are, loosely speaking, statistical in nature).

Then we turn to vowel shifts, pronunciation, and the fact that words drift in meaning over time. Has someone been disputing that there have been changes in pronunciation? But, sure, there are probably people out there proclaiming that the changes must stop here and now. The next chapter, though, is about the politicization of languages--minority languages, language nationalism, and so forth. It's starting to get a bit ambiguous as to who's doing the taming here. Are the Quebecois, who have succeeded in making Quebec markedly more francophone, really domesticating English? It's more like they're just replacing it. This drifts into a discussion of different registers of political speech. 

Oh, and somewhere in there Greene talked about code-switching (shifting into and out of standard English as the occasion demands; he's for it), language acquisition in children, what constitutes good writing . . .

It's all pretty interesting, taken piecewise. It's entertainingly written, too. What the whole thing adds up to, though, is harder to discern--other than an appreciation of "wild" language. Greene works at The Economist; his book might have worked better as a series of connected articles in that publication.