Answer by Marc Ettlinger, Ph.D. in linguistics, research neuroscientist and the Deptartment of Veterans Affairs:
Recent work by economist claims that language can indeed have a big impact on culture: He argues that languages that explicitly mark the future tense pay more attention to the future, and therefore have lower rates of obesity, better rates of saving, better pension plans, lower smoking rates, etc.
Statistically, the correlations are high, as seen in some nice charts and graphs, and the reasoning makes sense: If you need to know to mark the future tense grammatically, it's on your mind more, and if it's on your mind more, you're more likely to factor it into your decisions.
So case closed, right?
The problem is that the precise opposite is true. I flipped Chen's argument on its head to make a point. In fact, languages that do not explicitly mark the future tense have better rates of saving, etc. The actual rationale presented by Chen is that languages that grammatically differentiate future and present tense do not appreciate the way the future and present are related; they think of them as different things. That just-so story works just as well as the one I presented at the beginning.
Things get even messier when you realize that three of the top five savers are countries that use Scandinavian languages, which though from different family trees (Finnish is actually more closely related to Hungarian), share anand have very unique social environments related to their Scandinavian-ness. Also, languages that don't mark the future tense can mark future tense on objects in some cases (e.g., Finnish).
Here's the crux of the problem: There are way too many just-so stories that can be written about these kinds of data, but there's no good way to control for the proliferation of possible hypotheses. Sure, you can make up some hypothesis about strict German rules or romantic-sounding French, but those linguistic impressions are very much driven by your a priori cultural presumptions and are not necessarily a property of the languages themselves.
As noted in this discussion on, it's way too easy to find spurious correlations, like between siestas and verb inflection. This sort of work has sampling bias, multiple comparison problems, file-drawer problems, and so on, written all over it, so extra care is needed in analyzing the results. A single correlation is not enough. You need behavioral studies, correlations across multiple variables that aren't otherwise correlated with each other and an appropriately selected sample group.
That is but one example that I thought it was important to highlight. More broadly, there are a number of examples of culture interacting with language detailed here.
In particular, the notion ofis an important concept wherein language may influence implicit cultural attitudes toward a gender, race, or other named entity. Language can also affect how stories are reported and may even influence the art of a society. Crucially, these effects are direct and unambiguous in terms of scientific prediction. That's the key difference.
We can also juxtapose this with the idea that, which there is ample evidence for.
What is going on? How do we reconcile these seemingly disparate accounts on the relationship between language, culture, and thought? A possible factor is that culture has a long historical arc, spanning hundreds of years. So, indirect, tangential effects, like what Chen is talking about, are unlikely to have any real effects, especially given the important of context, etc. That is, if something seems too much of a stretch, too good to be true, then it probably is.
More questions on Quora: