by

Predicting Prudence From Tense Marking?

munich in the rain800PX-~1

Neuhauser Steet, in Munich. Do Germans save more than English speakers because of grammar?

Keith Chen, an economist at the Yale School of Management, recently gave a TED talk about his claim (in a forthcoming paper in American Economic Review) that certain of people’s prudence-related behaviors are attributable to the grammar of future time-reference in the languages they speak.

English speakers say “It will rain tomorrow” (with the future-marking modal auxiliary will), where German speakers would say Morgen regnet es (literally “Tomorrow rains it”), and Germans turn out to save significantly more of their income than English speakers do. This, Chen claims, is no coincidence: Across the world, speakers of languages with grammatically obligatory future marking tend (according to their own questionnaire responses) to save less for their retirement, eat more junk food, smoke more, and have more unsafe sex.

Chen’s story about this is causal. He thinks that “grammatically separating the future and the present leads speakers to disassociate the future from the present.” In fact when speaking to BBC business news he suggested (preposterously, in my view) that “encouraging yourself to think in the present tense makes it a little bit easier to engage in self-control.”

However, when he put up a recent blog post about his work, he slipped up and stated the opposite of one of his results. He wrote: “a 20 percentage point increase in the frequency of future tenses results in 1% more of GDP saved.” He meant (and later corrected the post to say): “1% less of GDP saved.”

The slip seemed to me telling: Ever since I first heard about Chen’s results (see this Language Log post) I have felt that his causal intuition is no more plausible than the opposite one. That is, these two conflicting conjectures seem equally plausible, prima facie:

  • Dubious Intuition 1:
    Speaking a language that grammatically emphasizes the distinctness of the future from the present encourages you to think of the future as another country where financial and health matters could be unpleasantly different, so you’ll prepare yourself by being prudent.
  • Dubious Intuition 2:
    Speaking a language that grammatically emphasizes the distinctness of the future from the present encourages you to think of the future as another country that right now you needn’t worry about, so you won’t bother to be prudent.

I’m quite sure that an effective presenter could convince a TED audience—or any audience—of the soundness of either (especially given the uncritical affection people seem to have for claims about language affecting thoughts and perceptions). Indeed, I think you could also convince an audience to accept either outcome from the absence of strong grammatical future-marking in a language rather than its presence:

  • Dubious Intuition 3:
    Speaking a language that uses simple present tense for future time references encourages you to see the future as continuous with the present, not as some distant and hypothetical realm, so you’ll tend to plan for future wealth and health with all the seriousness that you apply to promoting solvency and health right now.
  • Dubious Intuition 4:
    Speaking a language that uses simple present tense for future time references encourages you to see the future as continuous with the present, so you will feel no need to change your wealth- or health-related behaviors from what they are now.

Again, at first sight neither of these seems significantly less credible than the other.

Notice, I am not denying Chen’s correlation or its robustness. I am simply asking why we should join him in favoring his choice of causal intuitions. If the facts had been diametrically otherwise, he could have written a paper presenting those correlations instead, but with prose matching intuitions 1 and 4, instead of the 2 and 3 that he assumes given the actual facts of the world. He could present a causal rationale for his quantitative results either way. That should worry him: an explanation for every possible outcome is no explanation at all.

However, when we turn from Chen’s causal intuitions to the language/behavior correlation itself, there is a lot more to worry about. There are some new results that seem to me very bad news for Chen. I don’t have space to cover them here, but (may I use the present tense to refer to future time?) they are to be the subject of my post on Monday, March 4.

[Image by Andrew Bossi, Creative Commons]

Return to Top