While attending Thanksgiving dinner at a professor’s house (she was kind enough to invite me over as I’m stuck on campus finishing up my apps!) I had a long and very interesting conversation with a linguist!
She was asking me about econ, and I thought she might have heard of a paper I posted about earlier (The Effect of Language on Economic Behavior). It turns out she had heard of the paper, but she wasn’t nearly as enthused about it as I was. And apparently, the linguist-researcher community doesn’t like the paper much, either.
Here is a post from the Language Log (a popular linguist blog) about the paper. They take issue with several of the author’s arguments.
The first issue is that the author’s classification of a language as strong- or weak-FTR is not the be-all-end-all of that language’s use of tenses. The post cites English as an example: though we are strong FTR, we often use the present tense to denote the future (my plane ‘arrives’ at 8:30pm tonight, instead of ‘will arrive’).
However, anecdotal usages that go against a language’s entire classification don’t seem like the type of error that ruins the author’s arguments. I’m sure there are cases in all languages where patterns of speaking can be found that do not fall under the language’s larger classification. The author, to his credit, went with the EUROTYP Project‘s classifications of language, instead of coming up with his own classifications, and these are perhaps the most legitimate, true-to-linguistics categorizations that can be used.
The second leg of the post provides a counterargument: the Piraha Indians, who are very weak-FTR, and exhibit the exact behaviors that Chen would predict for a strong-FTR language. Here, our response should naturally be: anecdotal observations cannot dismantle an argument based on data in the aggregate. Outliers can be found for even the highest t-stats.
The author of the post then starts to worry that it might be easy to find other correlations between linguistic characteristics and other life outcomes. I don’t see how this detracts from the validity of the paper.
We finally arrive upon what’s really at the heart of the issue: the judgment. Chen’s work apparently inspired some journalists to summarize his findings as ‘some languages make you healthier/better than others’. Linguists hate that normativity, and the idea that some languages are better or worse than others. I think that’s really the key issue!
Another Language Log author created a post pointing out the idea that correlation does not equal causation (along with some very cool visuals to aid his point.) This argument is a more legitimate one, grounded in the fact that languages are highly diffusive, as are other traits about people.
Then Chen steps in to defend his paper! And he does a good job, especially on the correlation versus causation point. Of course, as discussed before, he can’t go as far back as the beginning of time and genetics to address the question that genetics creates languages AND behaviors.
I think this discourse is pretty interesting, and I’m liking the changing levels of doubt that you can track in the comments section of each post. I think Chen’s paper still holds water, as he addresses both linguists’ criticism and defends the causal relationship central to his paper. Perhaps it’ll lead to some cool, interdisciplinary work between linguists and economists in the future!