TED talk: how a language's grammar affects economic behavior

[Gray Richardson]

Linguistics was in the spotlight over at the TED talks website today.

Behavioral economist Keith Chen posted a TED talk in which he posits that people who speak "futureless" languages save more because their notion of the reward of savings is more immediate, while those people who speak languages with a future tense experience the future as more remote—and thus save much less of their income.

blog post with links here: http://blog.ted.com/2013/02/19/saving-f ... -behavior/

video here: http://www.ted.com/talks/keith_chen_cou ... money.html

Additionally, Jessica Gross blogs about other ways that our language affects how we organize the world: http://blog.ted.com/2013/02/19/5-exampl ... -we-think/

[GBR]

[note, I didn't check the links yet]

Smells dangerously of Sapir-Whorfianism to me... Let alone the questionable aspects of TED in general.

[clawgrip]

Not sure what I think of it. Some of it seems suspiciously nonsensical, for example:

"English speakers were much more likely to remember who accidentally popped balloons, broke eggs, or spilled drinks in a video than Spanish or Japanese speakers. (Guilt alert!) Not only that, but there’s a correlation between a focus on agents in English and our criminal-justice bent toward punishing transgressors rather than restituting victims, Boroditsky argues."

That's quite a leap from "much more likely to remember who" to "bent on punishing transgressors."

"Russian speakers, on the other hand, have separate words for light blue (goluboy) and dark blue (siniy). According to a 2007 study, they’re better than English speakers at picking out blues close to the goluboy/siniy threshold."

So, Russian speakers are better than English speakers at identifying an arbitrary colour threshold determined by the Russian language.

[GBR]

[I really should watch the video before I comment any more, but]
Sounds to me like causal lines are being drawn where they don't necessarily exist. The notion that one distinguishes colours better if they're distinguished in ones language is, IIRC, fairly well attested, but it's a huge leap from that to 'no future tense = Japanese people think of the future as being less remote'. Are they taking into account that Japan is a generally more conservative nation? It seems to me a nation's judiciary and economic practices, and their language are much more likely to be informed by culture than for the culture, economy and justice system to be influenced by something as wishy-washy and changey-wangey as language.

[clawgrip]

Yeah, I also wonder how they are able to separate the posited influences of language from the influences of culture.

As for the colour thing, I guess I see their point, but it's pretty badly phrased.

[Astraios]

Lol rubbish. He's barely even got a noticeable trend on his graph of futureless-language-countries being better savers. Also, he's not a linguist, so. Also also, TED gave a standing ovation to that Aquatic Ape woman for saying something along the lines of "Why should we believe what the majority of scientists say?", so.

[Gray Richardson]

[note, I didn't check the links yet]

Smells dangerously of Sapir-Whorfianism to me... Let alone the questionable aspects of TED in general.

Reallly? The video is 12 minutes long. You couldn't wait 12 minutes to listen to what he had to say—you couldn't even spend a few seconds to click on the links to peruse the text of his blog post, before posting something negative? Frankly, I find your own behavior questionable.

Further, you use the term "Sapir-Whorfianism" as a code word to imply some sort of danger. How do you feel that his theories are "Sapir-Whorfian," and why do you view that as a bad thing, or dangerous?

[Pthagnar]

i thought he used the word "dangerously" as a code word to imply danger

[Melteor]

So, Russian speakers are better than English speakers at identifying an arbitrary colour threshold determined by the Russian language.

This distinction is true of any culture with a taxonomizing tradition. The words make it easier to ask questions, search for data, etc.

[linguoboy]

I didn't watch the video. I looked at the accompanying graph, which shows a very high incidence of "future tense use" among English-speakers and none at all among German-speakers. This is completely nonsensical: linguists treat both languages as having only two tenses, past and non-past, with certain modal constructions employed to express future action. If German werden + INF doesn't count as a "future tense", then neither should English will + INF. The sizable discrepancy between Catalan and Spanish suggests to me that he's looking only at the synthetic future in these languages (Catalan makes very little use of the anar a + INF construction due to the likelihood of confusion with the anar + INF periphrastic preterite). So, again, if that's your working definition, the English belongs on the bottom rung with German and Dutch.

Edit: Language Log weighed in with healthy scepticism on Chen's paper a year ago, when it was yet unpublished. Pullum takes him to task for his uncritical acceptance of EUROTYP classifications and Liberman faults him for not taking into account the significance of cultural diffusion.

[patiku]

Once in a Ted talk someone proved that the Indus Valley Civilization spoke a Dravidian language because one of their glyphs had seven points on it or something.

[linguoboy]

Once in a Ted talk someone proved that the Indus Valley Civilization spoke a Dravidian language because one of their glyphs had seven points on it or something.

That's nothing: Once on the Internet, someone posted something that was untrue. So now I don't believe anything at all that I read on the Internet.

[patiku]

No need to get nasty, Ted!

[Aurora Rossa]

Once in a Ted talk someone proved that the Indus Valley Civilization spoke a Dravidian language because one of their glyphs had seven points on it or something.

I remember reading something along those lines on a websites for ancient scripts, maybe www.ancientscripts.com or something.

[GBR]

[note, I didn't check the links yet]

Smells dangerously of Sapir-Whorfianism to me... Let alone the questionable aspects of TED in general.

Reallly? The video is 12 minutes long. You couldn't wait 12 minutes to listen to what he had to say—you couldn't even spend a few seconds to click on the links to peruse the text of his blog post, before posting something negative? Frankly, I find your own behavior questionable.

Further, you use the term "Sapir-Whorfianism" as a code word to imply some sort of danger. How do you feel that his theories are "Sapir-Whorfian," and why do you view that as a bad thing, or dangerous?

I didn't watch the video because I didn't have time, but I thought I should let people know I hadn't in case your summary of it was wrong. His stuff sounds Sapir-Whorfian because he's implying language exerts heavy control on what we will and wont do. I don't think the Sapi-Whorf thoery, at least in it's strong form, holds any water, so I said so. I think it could be construed as dangerous because it provide a ground for racist or nationalistic thought. I wouldn't say it's on a par with other factors, but still, it's there.
I have watched the video now, and I still think he's on the wrong track. Everyone else has said pretty much all I would say on the topic. What I might add is that, I assume he groups people by country so he can control for culture, but that's ludicrous, since culture varys widely in one country, especially where several languages are spoken.

[Gray Richardson]

I didn't watch the video because I didn't have time, but I thought I should let people know I hadn't in case your summary of it was wrong. His stuff sounds Sapir-Whorfian because he's implying language exerts heavy control on what we will and wont do. I don't think the Sapi-Whorf thoery, at least in it's strong form, holds any water, so I said so. I think it could be construed as dangerous because it provide a ground for racist or nationalistic thought. I wouldn't say it's on a par with other factors, but still, it's there.

Oh, I applaud your honesty in admitting that you hadn't watched the video. I'm just questioning why, if you didn't have time, that you couldn't just wait, think about it, and post a more reasoned and informed post when you had more time. It's not a race. The first one who says "Whorf" doesn't win anything.

As for Sapir-Whorf, I don't think anyone believes the strong version holds any water, that idea has long been discredited. Even Whorf never espoused it—that was a mischaracterization of his work by other scholars. However, the weak version has become widely accepted in recent years. I don't think it's controversial to say that the language you speak has some impact on the way you think. That's in the forefront of research that cognitive linguists are doing these days.

Sapir and Whorf were not racists, were not nationalists. They were very forward thinking, talented, concerned, caring and noble linguists. I just hate to see someone try to use their names as some kind of slur. They deserve better than that.

[Gray Richardson]

I didn't watch the video. I looked at the accompanying graph, which shows a very high incidence of "future tense use" among English-speakers and none at all among German-speakers. This is completely nonsensical: linguists treat both languages as having only two tenses, past and non-past, with certain modal constructions employed to express future action. If German werden + INF doesn't count as a "future tense", then neither should English will + INF. The sizable discrepancy between Catalan and Spanish suggests to me that he's looking only at the synthetic future in these languages (Catalan makes very little use of the anar a + INF construction due to the likelihood of confusion with the anar + INF periphrastic preterite). So, again, if that's your working definition, the English belongs on the bottom rung with German and Dutch.

Well, that's at least better than the guy up above who didn't look at anything before posting something negative. I applaud you for partially informing yourself before jumping to conclusions.

Now, I don't necessarily vouch for Mr. Chen, it just seemed pretty interesting to me, so I posted it here to see if it might spark some interesting discussion. I have no opinion, really, on what languages count as futureless and which ones do not. But I believe, if you were to watch the video, and really listen to his speech, he might just address some of your questions.

As I understand his explanation, he was saying that English requires you to think about whether an action is in the present, past, or future. You must decide on and parse a tense each time you use a verb. It's not optional. Whereas the supposed "futureless" languages he is talking about have a default that is neutral with respect to time, at least between the present and the future (non-past). Those languages may all have methods of expressing future tense, it's just that it's optional. The language doesn't force you to constantly think about it, or decide each time you use a verb—unless the time is germane to the conversation.

And it's this grammatical ambivalence as to future time that he submits makes those speakers more apt to save for the future, because the future feels more immediate to them, less distinguishable from the present.

Whereas those speakers of languages who are constantly forced to think about the future—required to draw a distinction between the future and present each time they utter a verb—those speakers have a sense that the future is more remote. They have less of a sense that the gains of saving will ever be realized in the here and now, so on average they save less.

In his speech he talks about some of the ways he collected and analyzed the data in order to be able to exclude any other factors. Maybe it's not conclusive, but I think he makes a compelling case.

Edit: Language Log weighed in with healthy scepticism on Chen's paper a year ago, when it was yet unpublished. Pullum takes him to task for his uncritical acceptance of EUROTYP classifications and Liberman faults him for not taking into account the significance of cultural diffusion.

I find it charming that, rather than just watch the fascinating and very short speech to form your own opinion, you went out of your way to research what other scholars have said that is negative about Mr. Chen. Healthy skepticism is one thing, but it sounds more as if you are doing opposition research!

Regardless of the merits, if you are a conlanger, the ideas in his speech might help you think of ways your conglang's grammar might influence your conculture—or conversely, how the ways in which your conculture views the world might be reflected in their language.

For example, just off the top of my head, what if a conlang had no grammatical way to express negative polarity? Sure they would come up with periphrastic or elaborate, roundabout ways of expressing negation. But how might the fact that their language does not require them to ordinarily think about negation, and perhaps even places mental hurdles in their way for expressing negation (say because they have to think just a little bit harder when they want to express negation in a roundabout way), how would that affect how they view the world? Negotiate, trade, and conduct business? Discipline their children? Or interact with outsiders? Or even criticize or argue with one another? How might that affect violence in their culture? Or the criminal justice system?

[GBR]

Oh, I applaud your honesty in admitting that you hadn't watched the video. I'm just questioning why, if you didn't have time, that you couldn't just wait, think about it, and post a more reasoned and informed post when you had more time. It's not a race. The first one who says "Whorf" doesn't win anything.

NB: Putting winks at the end of your sentences doesn't make them sound any less condescending. I wasn't racing to be the first person to say Whorf, I was reading the boards, saw your post and gave my initial reaction to it. I made it completely clear ("not read", "seems") that I wasn't giving an informed rebuttal of the research (and couldn't from just watching a TED video anyway), but that I thought the premise sounded unreasonable. I used 'Sapir-Whorf' as shorthand for 'strong version of the Sapir-Whorfian theory'. I guess I should have been clearer.

As for Sapir-Whorf, I don't think anyone believes the strong version holds any water, that idea has long been discredited. Even Whorf never espoused it—that was a mischaracterization of his work by other scholars. However, the weak version has become widely accepted in recent years. I don't think it's controversial to say that the language you speak has some impact on the way you think. That's in the forefront of research that cognitive linguists are doing these days.

No academics or amateur linguists believe in the the strong version, I'm sure, but I've encountered a good deal of similar thinking in lay-people - the sort of people to whom TED's talks are targeted. The sort of people that includes Economists that might form a policy without fully understanding the intricacies and assumptions that went into the original research.

Sapir and Whorf were not racists, were not nationalists. They were very forward thinking, talented, concerned, caring and noble linguists. I just hate to see someone try to use their names as some kind of slur. They deserve better than that.

Just as well no one is saying or doing that then, isn't it?

[linguoboy]

Now, I don't necessarily vouch for Mr. Chen, it just seemed pretty interesting to me, so I posted it here to see if it might spark some interesting discussion. I have no opinion, really, on what languages count as futureless and which ones do not. But I believe, if you were to watch the video, and really listen to his speech, he might just address some of your questions.

So what questions of mine does he address in the video which he didn't address in this response of his to Pullum and Liberman? (Bonus points if you read down to the comments, where one of the authors of the EUROTYP data Chen used to classify the languages in his study complains about his distortion of their terminology and misconstrual of the data.)

As I understand his explanation, he was saying that English requires you to think about whether an action is in the present, past, or future. You must decide on and parse a tense each time you use a verb. It's not optional. Whereas the supposed "futureless" languages he is talking about have a default that is neutral with respect to time, at least between the present and the future (non-past). Those languages may all have methods of expressing future tense, it's just that it's optional. The language doesn't force you to constantly think about it, or decide each time you use a verb—unless the time is germane to the conversation.

Don't you still have to decide each time whether time is "germane to the conversation"?

It's hard to come up with examples in German where the future construction is obligatory. It's not, however, difficult to find examples where it is more natural. That is, you could avoid it, but it would be unidiomatic to do so. However, that's not what the EUROTYP researchers set out to determine. They didn't analyse conversation, they analysed weather reports. Which do you think is more indicative of how ordinary people think--or (crucial to Chen's conclusion) has more influence on their savings-related behaviours?

I find it charming that, rather than just watch the fascinating and very short speech to form your own opinion, you went out of your way to research what other scholars have said that is negative about Mr. Chen. Healthy skepticism is one thing, but it sounds more as if you are doing opposition research!

I find it less than charming that you're too cognitively-challenged to assume I had some unnefarious reason for this behaviour, namely that I was someplace where it was inconvenient to watch videos but it was convenient to read texts. (Moreover, when given the option, I prefer reading to watching in any case. And I didn't go to Language Log to do "opposition research"--whatever that even means. I went because I read it regularly and vaguely recalled Chen's ideas having been discussed there before.)

I don't blame you for this shortsightedness, by the way; I lay the fault at your native language, whose grammaticalisation of transitivity and causality regularly obligates speakers to leap to false conclusions about agency and motive.

For example, just off the top of my head, what if a conlang had no grammatical way to express negative polarity?

I've seen some attempts to create conlangs of this sort, but they all just end up kicking the can.

[Gray Richardson]

So what questions of mine does he address in the video which he didn't address in this response of his to Pullum and Liberman?

Wait, you refuse to watch his short video, but you want me to read your opposition research, do a comparison between it and the speech, and then break it down for you point by point? Dude, I'm just linking to the post because I thought it was interesting. If you've got questions, watch it. If you don't want to watch it, don't watch it.

I'm not trying to prove the guy is right; whether his theory is right or wrong is not even the point. Science is all about coming up with theories and trying to test them. We learn as much from the failures as successes. Good on Mr. Chen for going out and trying to add to human knowledge.

The takeaway from his speech is the ideas. Right or wrong, there's some interesting stuff there that could be useful for conlangers.

Don't you still have to decide each time whether time is "germane to the conversation"?

No. It might be clearer if you watched his speech.

In English, do you have to decide whether a couch, a car, a bridge, a desk or most nouns are masculine or feminine each time you utter a noun? No. Because English doesn't require it. It's there, it can be expressed; English has mechanisms with which you can add gender to those nouns if you like. But it's optional. Whereas in French and Spanish and many other languages it is required. You have to learn it and think about gender every time you speak.

Some Mayan tongues apparently have special quantifiers that are used with count nouns that require you to think about the shape of the objects you are counting (whether they are flat or coiled or skinny like a stick) so you can pick the proper quantifier that goes with them. It's mandatory. You have to think about shapes. You can't not do it, or it will sound ungrammatical. What would you say if a Mayan asked you: "Don't you have to decide each time whether shape is germane to the conversation?"

Apparently, in Chen's "futureless" languages, if tense is not required, you don't have to think about it. There's a one-size fits all conjugation that works for, at least, both the present and the future. Or so I gather.

I don't blame you for this shortsightedness, by the way; I lay the fault at your native language, whose grammaticalisation of transitivity and causality regularly obligates speakers to leap to false conclusions about agency and motive.

That's actually a good one. That was funny.

I've seen some attempts to create conlangs of this sort, but they all just end up kicking the can.

Oh, I'm not advocating this one in particular. It was just something off the top of my head. I was just trying to illustrate the thought. But, say you were writing a science fiction story, and you create some aliens that have some radically different feature about their language, how would it affect how they view and and function in their world? How would it affect their contact with outsiders? How would it make them more or less able to adapt to a particular change? It might be something that could drive a good story.

[zompist]

Gray, you need to turn it down a notch. Posting stuff you find interesting is fine; being defensive and hostile on behalf of third parties is not helpful. People won't always react to things with the same enthusiasm you do; that's life.

Language Log is always a very good place to check for this sort of thing. If Chen's work isn't going over well there, don't bet the farm on it.

[clawgrip]

I think there are two dubious assumptions made in his hypothesis:

The first: I am doubtful of his method of dividing languages into futured and futureless. He mentions languages like German and Chinese as being futureless, but Linguoboy pointed out that some of these futureless languages do in fact have future indicators, even if they may be used somewhat less commonly than in English. The presenter's own example of Chinese uses the word 明天 "tomorrow" to explicitly mark the action as a future event, thus "cleaving it from the present." Maybe Chinese doesn't "force us to constantly" divide up the time spectrum, as he says, but the inclusion of the word "tomorrow" shows that the speaker is quite conscious of the need to clarify that it is "not now". He also claims that English "requires us to" use future tense, ignoring the fact that we use present tenses for future all the time, e.g. "I'm meeting my friend there," "My flight leaves March 23rd," "I'm going to be a doctor when I grow up," "If you see him, say hi for me," "I hope I'm as healthy as you when I'm 70."

The second: His family thing is, I think, true. But I wonder if it is as big a burden as he claims. Are we really constantly having to think about these things? Do we have to "decide" to use the future tense or not? Do French speakers really "decide" if a car is feminine? Do Mayan speakers "decide" if an object is long? They seem to occur pretty automatically to me. Maybe it still influences us. Maybe someone knows of experiments that show, for example, that Chinese speakers are poorer at estimating because Chinese lacks plurals, or that Japanese speakers are poor at identifying referents in discourse because the language has no definite articles and pronouns are optional (or are they better, because the language relies more on context to convey unspoken information than English?).

Also, if the futured/futureless distinction has this much of an influence on our way of thinking, shouldn't we see its effects in other areas, like say, punctuality? Shouldn't speakers of futureless languages be less concerned with being on time for future plans, because when you arrive, you arrive? The few things I know about punctuality among cultures seems to suggest language has no influence.

[linguoboy]

So what questions of mine does he address in the video which he didn't address in this response of his to Pullum and Liberman?

Wait, you refuse to watch his short video, but you want me to read your opposition research, do a comparison between it and the speech, and then break it down for you point by point?

Um, no, If you had actually followed that link, you'd've seen that I was linking to Chen's own work. As I said, it's a "response of his"--i.e. his response to their criticisms. I've also looked at the actual published paper (faculty.som.yale.edu/keithchen/papers/LanguageWorkingPaper.pdf). My question to you is: what is in the video which is missing from those sources, thus requiring me to view it in order to understand his argument?

You still haven't clarified what you mean by "opposition research", by the way. It's not like Pullum and Liberman belong to a different ideological camp than Chen. They're simply raising issues regarding the evidence and methodology undergirding his bold and novel claim. That's how science works, remember? You make a claim and provide evidence for it, then everyone else goes through your arguments with a critical eye, looking for reasons why you haven't actually proven what you say you have. Most of the time, they're right.

Language Log is always a very good place to check for this sort of thing.

In particular, Mark Liberman is, unlike me, a linguist with a solid understanding of statistical analysis. That makes him my go-to guy for breaking down claims which require a knowledge of both.

[Salmoneus]

If you're not with them, linguoboy, you're against them. If people disagree with them, they can ignore them, because anyone who disagrees with them is just 'the opposition'. They're just there to do them down and find fault, damnit...


I was going to jump in and remind people how little English cares about the future/present distinction, but I see it's been done already. Except that when I say 'I see', I actually mean 'I saw', and when I say 'when I say' and 'I actually mean', I actually meant 'when I said' and 'when I meant', because it turns out English doesn't care too much about the present/past distinction either, in practice (oh, look (unmarked future tense!), I even mixed present and past verbs in that sentence). In English, a huge number of 'technically' future tense things are unmarked, and even many past tenses. And, conversely, some things are marked as future tense even when they 'shouldn't' be. You hear people saying things like "I'll just point out that..." - no sir, you ARE pointing that out, right now! That's not a real future! "That'll be the postman knocking" - no, that IS the postman knocking. "I'll say!" - no, you are saying. "I'll wait here" - you are waiting there.

This is because English has no future tense. English instead makes extensive use of a large number of auxiliary verbs, many of which have some future-tense element to their semantics. But using those auxiliaries has primarily modal and pragmatic purposes, rather than tense purposes: the difference between "my plane'll leave in an hour" and "my plane leaves in an hour" is not that one is future and the other is not. Nor 'my plane's going to leave in an hour', nor 'my plane should leave in an hour', and so on and so forth.

[Vuvuzela]

He also claims that English "requires us to" use future tense, ignoring the fact that we use present tenses for future all the time, e.g. "I'm meeting my friend there,"

This could in fact be either present or future, your other examples however,

"My flight leaves March 23rd,"

This mentions a future time, which wouldn't make it a bad example on it's own, but let's take it away; "My flight leaves." seems like an awkward present tense construction to me. Better to say "My flight is leaving.", which also works in your example.

"I'm going to be a doctor when I grow up,"

The auxiliary verb "be going to" is a common English marker of, if not simple future, something very close.

[ "If you see him, say hi for me,"

A command qualified by a conditional statement. No one commands people to do things at the exact moment of utterance, that would be silly, so the structure still does betray a time of reference.

"I hope I'm as healthy as you when I'm 70."

A little better, but I still think "hope" puts the verb sentence in the future.