Bad at planning for the future? You might be able to blame your language. Differences in the way various languages talk about the present and future could help explain why Germans urge free-spending Greeks to adopt their fiscal discipline, and why Americans are baffled by China’s low consumption and high savings rates, according to research published in the American Economic Review in April
Keith Chen, a behavioral economist at the University of California-Los Angeles, researches intertemporal decision-making, or how people make choices when the consequences of those choices are spread out over time. Do you spend your money on a fancy sports car, or save up for retirement? Spend a spare hour working out, or relaxing in front of the television? He noticed that a group of countries whose thrifty behavior baffled economists had also attracted attention from linguists. “Every country economists thought of as an outlier in savings behavior was also an outlier in grammar,” Chen says.
Specifically, they had a way of talking about the future that would sound odd to English speakers. English is a “strong-future” language, which means it forces speakers to distinguish between the present—“It’s raining”—and future—“It will rain.”
But other, weak-future tongues treat present and future the same, relying on context instead. Economists have found that people typically need extra compensation to sacrifice present pleasure for future rewards—they’d rather have things now, not later, Chen wrote in his paper. Talking about the future as though it were the present might make it feel closer, he wrote, while being forced to make a linguistic distinction could make it feel farther away, making the speakers of strong-future languages less likely to wait for future payoffs.
Members of an Aboriginal community in Australia who speak a language with no words for left and right are better at perceiving cardinal directions than English speakers. They think of timelines as moving not left to right, but east to west.
Using data from the World-Values Survey, which asked individuals in 76 countries between 1994 and 2007 about culture and values, Chen tested whether people who speak strong-future languages, like English, Greek, Portuguese, Italian, and Spanish, were more or less likely to save than speakers of weak-future languages, like Japanese, German, and Mandarin. In a regression analysis holding individual and cultural characteristics constant, people who spoke strong-future languages were consistently about 45 percent less likely to report saving money in a given year.
In a smaller sample of countries where both language types are spoken, including Switzerland, Singapore, and Nigeria, Chen found similar correlation between language and likeliness to save among citizens of each country.
The relationship seemed to hold at the country level as well. Of the 16 member countries of the Organisation for Economic Co-Operation and Development (OECD) with the lowest average total savings rate as a percentage of GDP between 1985 and 2010, just two spoke weak-future languages. Regression analysis found a strong correlation between national savings rates and languages’ treatment of the future as well, even after controlling for demographic characteristics.
Chen performed similar regression analysis on retired households’ net worth, using data from the Survey of Health, Ageing, and Retirement in Europe (SHARE), a panel survey measuring socioeconomic status and health among retirees in 13 European countries, and also found a strong correlation between how languages treat the future and how much households have saved up for retirement, divided by that country’s average disposable income, with strong-future speakers saving around 39 percent less.
Since the SHARE survey and MEASURE DHS project, which conducts surveys in developing countries on behalf of USAID, also ask about health quality and behaviors, Chen also tested whether speaking a language that makes the future feel closer is associated with taking time for healthy habits that might be costly in the present, but improve long-term health. Speaking a strong-future language was correlated with a nearly 30 percent lower probability of being physically active and a 13 percent greater risk of obesity among individuals in SHARE survey countries, and 20 percent and 17 percent respectively for individuals in countries in the DHS survey.
The results were dramatic, and according to several linguists, a little too dramatic (see a few responses here and here). Even Chen thought his findings—both the strength of the association between language and behavior and the magnitude of the average differences—were “disturbingly strong.”
But no matter how he restructured his experiments, Chen says, the results remained the same. When he compared speakers of different languages in the same country (so taxes and other economic policies affecting saving would be consistent, and culture would be similar) weak-future speakers still saved more. And while language didn’t appear to affect beliefs about the importance of saving, even among people who responded in a global survey that they thought it was important to teach their kids to save, strong-future speakers saved less.
“The data are most consistent with a world where it’s unconscious and every time you speak, there’s a subtle distinction that influences you,” Chen says. “There is nothing suggesting this isn’t the case.”
Surprising as it sounds, Chen’s findings fit nicely with a linguistics theory called Whorfianism, which argues that words don’t just describe our thoughts; they also shape them.
Members of an Aboriginal community in Australia who speak a language with no words for left and right, for example, are better at perceiving cardinal directions than English speakers. They think of timelines as moving not left to right, but east to west, according to a 2010 study in the journalPsychological Science.
And a 2007 study in the National Academy of Sciences journal found that Russian speakers, who have separate words for darker and lighter shades of blue, are better at distinguishing the colors than English speakers, who can use one word for all shades.
John McWhorter, a linguist and professor at Columbia University, said it’s an appealing idea, but one that has limited use outside of the lab. In the Russian study, for instance, the difference in the speed with which Russian and English speakers distinguished hues was measured in milliseconds—too small a difference to affect how people see the world, McWhorter says.
“There is a small effect, but the question is whether those small priming effects have the real-world effects they claimed,” McWhorter says. “The tiny differences are real, but what makes Germans and Russians different is not their grammar.”
But those arguments don’t apply so neatly to Chen’s work, which links language to the sort of decisions and long-term effects McWhorter argues are impossible.
In Chen’s case, McWhorter says the Russian language was wrongly classified as distinguishing between present and future when it doesn’t have tenses, as English does. Chen says he consulted with other linguists who said that while Russian may not use tenses, it does have other ways of distinguishing between present and future that have the same key effect: forcing speakers to separate the future from the present each time they speak.
McWhorter also questions why individuals who speak languages that share common roots and have the same grammar for talking about present and future, such as Slavic languages, can have such different savings rates. Chen says there are many factors that affect a country’s savings rate, and that his findings showed that, on average, countries with mostly strong-future speakers saved less than similar countries with weak-future speakers.
More concerning, Chen says, is the possibility that what affects savings and other future-oriented decisions isn’t language, but another, unknown factor closely tied to how an individual’s language deals with time.
The obvious candidate is culture. But even if it’s all about culture, not language, Chen says he finds the results equally fascinating. Since the language-saving correlation held even among speakers of different languages within a single country, whatever the cultural effect might be, it must be intimately connected to how languages developed tenses, and more fundamental than nationality. “Even if that were what’s driving this,” he says, “it would still be kind of amazing.”
In the months since the paper was published, Chen has begun working with other researchers investigating the topic. He’s also exploring other features of languages, such as whether speaking a language that forces you to step into someone else’s shoes when talking about what’s near and far helps you more easily look beyond your own perspective.
He says he’s excited his research hit a nerve and gave him the chance to dig deeper, and he’s as curious as anyone else to see where it takes him.
“In five years, this will either be deeply enshrined in how we think about savings behavior, or I’ll have to stick my head in the sand every time I go to a conference,” Chen says.
