David Robson, a journalist linked here a number of times before (e.g.), writes about “how thinking in a foreign language improves decision-making”:
As Vladimir Nabokov revised his autobiography, Speak, Memory, he found himself in a strange psychological state. He had first written the book in English, published in 1951. A few years later, a New York publisher asked him to translate it back into Russian for the émigré community. The use of his mother tongue brought back a flood of new details from his childhood, which he converted into his adopted language for a final edition, published in 1966.
“This re-Englishing of a Russian re-version of what had been an English re-telling of Russian memories in the first place, proved to be a diabolical task,” he wrote. “But some consolation was given me by the thought that such multiple metamorphosis, familiar to butterflies, had not been tried by any human before.”
Over the past decade, psychologists have become increasingly interested in using such mental metamorphoses. Besides altering the quality of our memories, switching between languages can influence people’s financial decision-making and their appraisal of moral dilemmas. By speaking a second language, we can even become more rational, more open-minded and better equipped to deal with uncertainty. This phenomenon is known as the “foreign language effect” and the benefits may be an inspiration for anyone who would like to enrich their mind with the words of another tongue. […]
The foreign language effect does not depend on the particular features of the language that someone speaks; instead, it is concerned with the general experience of moving from a first to a second language. How will my thinking change, for example, when I move from English, which I absorbed in my cot, to Italian, which I have painstakingly studied as an adult? […]
For Prof Boaz Keysar at the University of Chicago, who pioneered this research, the inspiration was personal. He grew up in Israel, and has now lived in the US for more than three decades, but still finds that Hebrew has more emotional resonance than English. Driving home from work one day, he started to wonder whether this might influence our moral reasoning, which is often driven by our gut feelings rather than logical reasoning. […]
In a preliminary experiment, Keysar’s team asked participants who had learned Spanish as a second language to consider [the “trolley problem”] in either their native or adopted tongue. As he had hypothesised, they were far more likely to make the utilitarian choice when they used Spanish compared with English. The effect was so big that Keysar delayed publishing the results. “I just didn’t believe the data,” he says. A later collaboration with Albert Costa at the Pompeu Fabra University in Barcelona, however, documented the same results across diverse participants from the US, Europe and Asia. In one sample, participants were twice as likely to choose the utilitarian option when speaking a second language.
In the meantime, Keysar had turned his attention to the classic cognitive biases such as “myopic loss aversion”. Previous experiments showed that people are unwilling to risk losing a small amount of money for the chance of winning a bigger sum, even if the odds are tipped in their favour. If you give people a 50% chance of winning £2.50, for example, or offer them the opportunity to take £1 for sure, most people go for the guaranteed money. This may pay off in the short term, but when you have made many such decisions, it severely limits your profits.
Myopic loss aversion is a serious problem for investors, but Keysar found that this tendency was much less pronounced when the bets were presented in a foreign language. “People just hate the prospect of losing,” says Keysar. “But they hate it less in a foreign language.” Further experiments have shown that speaking a foreign language can also reduce the framing effect, which is our tendency to be swayed by the particular phrasing of information, and the sunk cost effect, which is our reluctance to leave a failing endeavour.
One of the most intriguing recent studies tested the “bias blind spot” – the expectation that we are less susceptible to error than the average person. “We believe that other people are stupid, and we are not,” explains Michał Białek, an associate professor at the University of Wrocław, Poland. In line with Keysar’s findings, he found that speaking a foreign language punctures this egotistical way of thinking.
It is now 11 years since Keysar published his first paper on the foreign language effect. Over the same period, many psychological findings have crumbled under scrutiny – but not the foreign language effect. “It’s consistent and replicable,” says Simone Sulpizio, an associate professor at the University of Milano-Bicocca, Italy, who conducted a recent meta-analysis examining the evidence to date, though both he and Białek emphasise that we need more research into the mechanisms behind the effect.
Psychologists have also started to explore how the foreign language effect may touch many other areas of our psyche. Consider memory, the topic of Nabokov’s musings. We encode our recollections with verbal narratives, particular words triggering specific associations attached to an event. As a result, a childhood memory may seem less vivid when someone recalls it using a language that was acquired later in life. That may be a disadvantage for someone writing a memoir in a foreign tongue. On the plus side, it could help people to process painful events. People who recalled a childhood trauma in a second language showed less emotional distress than those recounting the event in their native tongue.
Keysar’s latest study, led by his PhD student Leigh Grant, has shown that the foreign language effect can discourage the creation of false memories. In one experiment, the participants were first given a list of connected words, such as “dream”, “snooze”, “bed” and “rest”. When their recall was tested later, many people mistakenly remembered hearing the word sleep, which was not on the list. This is a standard technique to test people’s tendency to form false memories, and it was considerably less likely to happen if they performed the task in their foreign language. […]
Keysar suggests that people employ more careful and deliberative thinking when using their foreign language: “You need to make sure that what you say, and how you understand things, is correct.” And they apply the same attention to the monitoring of their memories, leading them to question the accuracy of their recollections rather than simply recounting the first thing that comes into their heads.
The foreign language effect may even stretch to elements of our personality. Silvia Purpuri at the University of Trento, Italy has examined “tolerance of ambiguity”, which concerns people’s appreciation of uncertainty, and their willingness to enter unfamiliar situations. This can influence our behaviour in a changing workplace. “Those with high levels of tolerance of ambiguity display more flexibility and adaptability, leading to a more harmonious working environment and overall successful performance,” explains Purpuri. By allowing people to entertain new ideas, tolerance of ambiguity can improve people’s creative problem-solving. Recruiting Italian-English bilingual speakers, Purpuri and her colleagues found that people naturally score more highly on this trait when primed to use their second language.
More material, and links, in the article. (We discussed Keysar and the foreign language effect back in 2012.) Just to be clear, I don’t believe any of this, not in the sense I believe that the sun will come up tomorrow or that languages are constantly changing; this field is full of underbaked studies and irreproducibility. Still, it’s interesting stuff, and who knows, maybe there’s something to some of it… (Thanks, Trevor!)
Forgot the link?
https://complit.fas.harvard.edu/article/how-thinking-in-a-foreign-language-improves-decision-making/
D’oh! Thanks, I added the one I intended; yours is a different, more compressed, one.
The effect was so big that Keysar delayed publishing the results. “I just didn’t believe the data,” he says
So: working on a problem in a foreign language can slow you down enough to make it less likely that you would simply jump to a conclusion based on gut feeling. Extraordinary! No wonder he didn’t believe it.
Perhaps even counting to ten before you speak might prove to have similar counterintuitive benefits? Or sleeping on an important decision? Seems unlikely, I know: but …
If you give people a 50% chance of winning £2.50, for example, or offer them the opportunity to take £1 for sure, most people go for the guaranteed money. This may pay off in the short term, but when you have made many such decisions, it severely limits your profits.
a very small point, but i’m inclined to be skeptical of any conclusions about “rationality” that are based on assumptions like the one underlying the last sentence here. the very specific goal of profit-maximalization through iteration is not synonymous with rationality, and not choosing to pursue that specific goal is neither “myopic” nor necessarily related in any way to “loss aversion”. the same is true of utilitarianism in any other form.
(perhaps relatedly: just how self-centered does a person have to be to believe that “such multiple metamorphosis, familiar to butterflies, had not been tried by any human before”? cyclic self-translation is hardly novel! to give just one example from nabokov’s twenties: an-sky’s Tsvishn Tsvey Veltn (better known as Der Dibek / The Dybbuk), probably the most famous yiddish play among non-yiddish-speakers, is famously the product of writing process that moved between russian and yiddish enough that the scholarly consensus is that neither language can claim to hold the first version.)
Extraordinary! No wonder he didn’t believe it.
I’m with you on this, DE – though I’d have to think it through in French to reach a more reliable judgement. Not only does slower processing with a less familiar language give us salutary pause, our deliberation is more likely to be explicit (I almost wrote “syllogistic”) in its detail, giving the undertow of cognitive biases less traction.
How does all this fit with the European use of Latin as a second language for serious thought and communication over the centuries, and its loss over the last? How does it fit with the rise of computer “languages” as secondary codes in which a small but perhaps disproportionately internet-active communicators are now fluent?
I’d have to think it through in French to reach a more reliable judgement
Indeed, indeed …
Of course, this means that the French belief that the French language is supremely logical is in fact perfectly true … but only if you’re not French.
This all seems to presuppose people with some competence in an L2 but where it’s still fairly easy to tell which is the “native” language and which the “foreign” one. I’d be curious to know whether they’ve run the same sorts of tests on people who’ve been fluently bilingual since early childhood. Of course, for many such people, the languages in which they’re fluent are characteristically used in different domains rather than in random free variation, so you could experiment with having them try to do a task in the non-obvious language for the domain with which it would be associated.
Or you could e.g. (assuming time-travel etc.) ask Samuel Becket to do a French translation of the English translation of one of the works he composed in French (but without giving him access to a copy of his French original) and see what happened.
I think Pierre Menard would be a more appropriate guinea pig.
Separately, the Nabokov thing seems a bit maybe off-point, because I don’t think the other examples involve answering the question or analyzing the problem or telling the story in language A and then translating that answer/analysis/story into language B. For Nabokov you would somehow need to come up with let’s say six anecdotes from his childhood that he’d never put down in written form before and ask him to write out three of them in English and the other three in Russian and see if any interesting differences emerged. Or if you could get him to orally tell the story about the time such-and-such happened in his childhood in both languages, but with the occasions separated by enough days/weeks/months that the exact wording he’d used in language A was not necessarily in his mind when he was asked to do it in language B but was more likely as it were to be doing a separate original composition on the same underlying theme.
I’m not surprised that Nabokov had some different thoughts on his childhood a few years after publishing his autobiography. Some of these different thoughts may indeed be due to writing in a different language, but I would expect that anyone, no matter whether they knew one, two, or more languages at whatever times, would have some different thoughts just because, as we grow, we interpret things differently. We also remember different things, depending on all kinds of triggers. We may even make up or embroider stories (consciously or unconsciously).
I would not be surprised if (some) genuinely bilingual people had subtly different “personalities” in their different L1s. As everyone but Chomskyites knows, language and culture are intimately linked, and only fictional characters have single consistant “personalities”; real people deploy different personality traits in different social environments (far from this being necessarily hypocritical, it’s a precondition of being able to function as a normal human being.)
This is quite a different phenomenon from the supposed cognitive benefits of being able to use more than one language, and conflating these very different things is just going to lead to nonsense.
As with many children of immigrant families, there were aspects of my (Taiwanese-American) wife’s California childhood that might seem unusual-to-bizarre from the standpoint of a median American family of the time and place but were boringly commonplace for immigrant families of the same ethnic background. While this is impressionistic rather than data from a rigorously controlled experiment, she has remarked that if she describes things like that in English they somehow “feel” weirder as she’s talking about them than if she were describing the exact same thing in Taiwanese or Mandarin. To the extent that’s a personality shift akin to what David E. mentions, I wonder if it’s less a change in her personality-as-such but because the different languages have different prototypical assumed listeners, who in one case will presumptively be familiar with the relevant cultural context and in the other case will presumptively not be.
BTW, that Pierre Menard is a name of uncertain referent is itself a very Borgesian touch: https://en.wikipedia.org/wiki/Pierre_Menard_(disambiguation)
@JWB, yes, I remember the same feeling (when I was about to disccuss here something that I and others usually don’t discuss in English), but only once.
“We believe that other people are stupid, and we are not,”
What a brilliant idea ! Wish I had thought of it.
Has LH discussed the idea that hard-to-read fonts promote greater retention? Google “Sans Forgetica”
“Two years after its release and having received a great deal of publicity, the first peer-reviewed study demonstrated that Sans Forgetica was not effective for enhancing memory.” Sounds like one of those fabulous (in all senses) Ariely/Gino ideas. This one simple trick will solve your problem! It’s Science!!
Has LH discussed the idea
But of course, in 2010. At least this time Sans Forgetica was actually designed to be hard to read, unlike the supposedly-harder fonts in the 2010 study.
@rozele:
Not taking small gambles with positive expected value is indeed completely unrelated to rationality or myopia. It is also not necessarily related to loss aversion, but in fairness loss aversion is a plausible and parsimonious way of explaining it.
More interesting: I don’t think I’ve cited here before what might be the economics article with maximum Hattic interest, by no less central a figure in the field than Paul Samuelson: Why we should not make mean log of wealth big though years to act are long.
Previous experiments showed that people are unwilling to risk losing a small amount of money for the chance of winning a bigger sum, even if the odds are tipped in their favour. If you give people a 50% chance of winning £2.50, for example, or offer them the opportunity to take £1 for sure, most people go for the guaranteed money. This may pay off in the short term, but when you have made many such decisions, it severely limits your profits
Apart from anything else, this formulation omits absolutely crucial information: do you get to play more than once? If so, how many times? If you only get to play once, do you have to play as an individual, or can groups play as a team and pool their winnings? If so, how big can the group be?
Anybody who doesn’t even notice that they have seriously underspecified the conditions of the test should probably not be accusing others of cognitive bias.
DE, I am pretty sure that it describes not any real experiment, but what these experiments generally measure. Of course, there might be an experiment like that. It’s hard to imagine that it will involve people assembling into groups. What would it measure? How much people are inclined to argue with they fists when money are involved?
@Giacomo what might be the economics article with maximum Hattic interest, by no less central a figure in the field than Paul Samuelson: …
The main Hattic interest I extracted was that the guy is almost illiterate — a lot less articulate than the average Economics theorist, I hasten to add. (I was brought up on Lipsey ‘An Introduction to Positive Economics’ as the authoritative barely-post-Keynesian text. Samuelson was disregarded as hopelessly fuddy-duddy. If I read any as an undergrad, neither the observation nor the execrable style has ‘stuck’.) I had to read that title three times and then the whole article and then circle back to figure out what it was trying to say.
There are much better treatments of his hypothesis within econometrics: look for amortised gains from repeated rounds of ‘Prisoners’ Dilemma’ for example.
@DE this formulation omits absolutely crucial information: do you get to play more than once? If so, how many times? …
Exactly. “Provided that the “shadow of the future” is big enough, these factors encourage co-operative behaviour” @D.O. the amortised game theory approach (an abstract experiment, but there were small real-life trials) attempts to model exactly those sorts of issues.
If you steal the winnings from a competitor, they won’t play next round: you both lose in the long run. (This is the outcome of the unsophisticated ‘Prisoners’ Dilemma’.)
Myopic loss aversion is a serious problem for investors, …
That’s all right. Hand over your life savings to one of the wide-braces boys in the City who have no aversion to (your) loss (what they speak is tantamount to a foreign language) and will back a three-legged nag in any race, book their commission and vamoose.
the guy is almost illiterate
He wants to keep his words short: he says so right at the end, and the way he does that is bad. Sure, he could do that with less pain for us, but he is not so skilled with words. Why should he be? It is not his field, and I think he does not want to try hard.
There is a Danish popular story/joke/parable about a boy who is regarded as dim-witted, and people will offer him the choice between a quarter and a dime and be amused when he picks the dime. But when asked if he knows the quarter is worth more, his reply is “of course, but if I start taking the quarters, people will stop giving me money.”
I’m sure this exists in other cultures as well, (I’m pretty sure I’ve seen an English version), and I only mention it to show that vulgus is well aware that the assumptions of naive game-theoretic economics are not even close.
What I want to know is whether I should shop for a new phone in Swedish or in Spanish. (I have a feeling that I know English too well, and maybe Swedish is too close to Danish so it’s somehow just may native thinking language in disguise).
The main Hattic interest I extracted was that the guy is almost illiterate — a lot less articulate than the average Economics theorist, I hasten to add.[…] I had to read that title three times and then the whole article and then circle back to figure out what it was trying to say.
Is this po-faced humor, or did you not read the last line?
I’m pretty sure I’ve seen an English version
I definitely have.
What I want to know is whether I should shop for a new phone in Swedish or in Spanish.
Perhaps you should split the difference and shop in Swadesh, although there is no word for ‘phone’ there.
… or did you not read the last line?
I did see that last line. I did not get it that one-shot words was the sole point of his game — so far as to play the game with the most ill-will for the sense or ease-to-read.
Then it is no more a piece on some theme (buy-sell-know-stuff) — as it claims — than on the one-shot game at large.
It is no more a “economics article” as per @GP’s claim than a gob-shite game. That it is from so “central a figure in the field” is no part of the point. I see no “maximum Hattic interest”; I see no Hattic point one bit. All fools can write in such a way — on some theme.
@JC he could do that with less pain for us, but he is not so skilled with words.
Else-wise word-stuff from Samuelson shows not so bad skill. He goes out of his way there.
The one important thing I can contribute to this thread is that somebunny has solved the trolley problem – in a quite language-independent manner. All I can do is quote Huxley from when he had finished reading the Origin: “How stupid of me not to have thought of this myself!”
Yeah. There are reasons why I’ve never invested in Bitcoin.
Interesting indeed. I also wonder if English as the global language of science in recent decades has been having similar effects.
And indeed I do this less than other people, and indeed I’m less able to function as a normal human being than most.
I see what you did there.