There's been quite a bit of internet buzz this past week about the work of Keith Chen, a behavioural economist at Yale. You can see his TEDx talk here: Could language affect your way to save money? as well as an attention-grabbing and rather misleading BBC article about his work: Why speaking English can make you poor when you retire. Chen's central hypothesis is that the language you speak may affect your savings behaviour, depending on how your language grammatically encodes statements about the future.
(If you're also interested, Chen's working paper can be downloaded here. The Language Log blog actually discussed his work a year ago in two posts: Keith Chen, Whorfian economist, where Geoff Pullum critically examines the linguistics behind the proposal; and Cultural diffusion and the Whorfian hypothesis, where Mark Liberman questions the interpretation of the statistics. Chen was also invited to write a response to those raised concerns, which can be read here: Whorfian economics.)
First of all, I should state that even though I still have a lot of doubts regarding this study, I absolutely love the cross-disciplinary nature of the work. This is the sort of work that needs to be actively encouraged. It doesn't matter if Chen himself is not a linguist - inter-disciplinary work will often lead one into territory that one is unfamiliar with, and it doesn't help if the other side is aggressively marking their territory. He has clearly done a lot of research, and is not basing his claims simply on personal observations and anecdotes. All the media attention should also not detract from the fact that Chen has put in a lot of thought and effort into the study. And given that much of academia these days is being starved of funding, a little media attention really shouldn't hurt.
With regards to this study, there are three broad areas that I think would be fruitful topics of discussion: (1) the link between language and thought; (2) the idea of 'futureless' languages; (3) the data that was used in this study and the problem of a 'monolingual' mindset in linguistics. In this post, I'd just like to start off with the first topic.
Language and thought
A very common assumption that I've come across is that if Language X has a word for a particular concept (usually one's own language), but Language Y doesn't have a word for it, then that concept must not exist in Language Y. One argument against this is that it doesn't matter if there's no single word translation if a full sentence can capture the same meaning. Others may argue that it's just not the translation in Language Y just doesn't have the exact same meaning as in Language X. Some people may then put this difference down not to language, but to culture. This may then lead them to try and tease language from culture, while others argue that it can't be done. What we have then, is a hypothesis that cannot be proven or falsified either way.
Thankfully, this assumption is not the premise for Chen's research. Despite the unfortunate wording on the TEDx talk page, he is not making the point that just because a language doesn't have future tense marking, its speakers don't have a concept of the future. Rather, he is tapping into a growing body of work that suggests that the language you speak subtly influences what you pay attention to in the world. Much of this work is built around Roman Jakobson's observation that “Languages differ essentially in what they must convey and not in what they may convey”, a quote that Chen cites in his research.
Bear in mind that within linguistics, the debate about the relationship between language and thought is by no means over. The Economist website held an online debate in December 2010 on just this topic. Suffice it to say, there is even less consensus on the effects language might have on behaviour, or the mechanism by which this might occur. Even if Chen's findings are robust enough to show a strong correlation between language and saving behaviour, there is a lot that needs to be done to show that it is a causal relationship, i.e. it is the language causing a change to saving behaviour.
However, my main concern with Chen's work is the actual application of this to tense systems of the world's languages. This will be the subject of my next post.
[Guy Deustcher's article in the New York Times (26/08/10): Does Your Language Shape How You Think? provides great coverage of the research in linguistics to a non-specialist reader.]
(If you're also interested, Chen's working paper can be downloaded here. The Language Log blog actually discussed his work a year ago in two posts: Keith Chen, Whorfian economist, where Geoff Pullum critically examines the linguistics behind the proposal; and Cultural diffusion and the Whorfian hypothesis, where Mark Liberman questions the interpretation of the statistics. Chen was also invited to write a response to those raised concerns, which can be read here: Whorfian economics.)
First of all, I should state that even though I still have a lot of doubts regarding this study, I absolutely love the cross-disciplinary nature of the work. This is the sort of work that needs to be actively encouraged. It doesn't matter if Chen himself is not a linguist - inter-disciplinary work will often lead one into territory that one is unfamiliar with, and it doesn't help if the other side is aggressively marking their territory. He has clearly done a lot of research, and is not basing his claims simply on personal observations and anecdotes. All the media attention should also not detract from the fact that Chen has put in a lot of thought and effort into the study. And given that much of academia these days is being starved of funding, a little media attention really shouldn't hurt.
With regards to this study, there are three broad areas that I think would be fruitful topics of discussion: (1) the link between language and thought; (2) the idea of 'futureless' languages; (3) the data that was used in this study and the problem of a 'monolingual' mindset in linguistics. In this post, I'd just like to start off with the first topic.
Language and thought
A very common assumption that I've come across is that if Language X has a word for a particular concept (usually one's own language), but Language Y doesn't have a word for it, then that concept must not exist in Language Y. One argument against this is that it doesn't matter if there's no single word translation if a full sentence can capture the same meaning. Others may argue that it's just not the translation in Language Y just doesn't have the exact same meaning as in Language X. Some people may then put this difference down not to language, but to culture. This may then lead them to try and tease language from culture, while others argue that it can't be done. What we have then, is a hypothesis that cannot be proven or falsified either way.
Thankfully, this assumption is not the premise for Chen's research. Despite the unfortunate wording on the TEDx talk page, he is not making the point that just because a language doesn't have future tense marking, its speakers don't have a concept of the future. Rather, he is tapping into a growing body of work that suggests that the language you speak subtly influences what you pay attention to in the world. Much of this work is built around Roman Jakobson's observation that “Languages differ essentially in what they must convey and not in what they may convey”, a quote that Chen cites in his research.
The point here is not about whether Language X has a word or a tense that Language Y has, but whether Language X obliges speakers of that language to talk about something in a way that Language Y does not. One widely cited example is that of Guugu Yimithirr, a language spoken in Australia, whose speakers are forced to use cardinal points to describe the location of a person or thing in relation to another one, given the lack of what we would consider to be terms like 'left' and 'right' in English, i.e. speakers have to say something like 'He is sitting to my north'. It has been found that speakers of this language are much more aware of their cardinal orientation than speakers of languages that aren't forced to habitually make reference to cardinal points. (This is grossly simplified, but you can read more about this here and in Levinson's 1997 paper). The point here is not that English lacks the resources to say 'He is sitting to my north'. It may sound odd, but it is perfectly possible. Rather, the point is that speakers of English are not obliged to habitually note their cardinal orientation, while speakers of Guugu Yimithirr are.
In his TEDx talk, Chen highlights the problem of translating 'uncle' into Mandarin Chinese, since you are forced to specify whether the person is your father's sibling or your mother's, and whether they are older or younger than your father or mother. In English, you could certainly use the phrase 'maternal uncle who is older than my mother' but you are not forced to. Therefore, one might argue that a speaker of Mandarin has to be constantly aware of such relationships much more than an English speaker.
Here, Chen's example is not the best. I come from a Chinese family myself, and many of these relationships and the terms to address various extended family members were 'calculated' by my parents when I was pretty young. From then on, the name got associated with that person, and I'm not actually forced to habitually make these calculations anymore. A better example would have looked at information that a language compels its speakers to specify, suggesting that they have to habitually pay attention to such information, e.g. how Russian obligatorily forces speakers to say whether they walked or travelled by vehicle; or how Sherpa obligatorily forces speakers to say whether an action was done on purpose or not.
Here, Chen's example is not the best. I come from a Chinese family myself, and many of these relationships and the terms to address various extended family members were 'calculated' by my parents when I was pretty young. From then on, the name got associated with that person, and I'm not actually forced to habitually make these calculations anymore. A better example would have looked at information that a language compels its speakers to specify, suggesting that they have to habitually pay attention to such information, e.g. how Russian obligatorily forces speakers to say whether they walked or travelled by vehicle; or how Sherpa obligatorily forces speakers to say whether an action was done on purpose or not.
In general, I would say that Chen's work is informed by current linguistic research into the relationship between language and thought. This can be seen in his working paper:
[...] English forces its speakers to habitually divide time between the present and future in a way that Mandarin (which has no tenses) does not. Of course, this does not mean that Mandarin speakers are unable (or even less able) to understand the difference between the present and future, only that they are not required to attend to it every time they speak. This difference, in the obligatory marking of future events is a central characteristic of the weak vs strong FTR classification (Thieroff 2000), and is the difference between languages I exploit in my study of savings behaviors.
Bear in mind that within linguistics, the debate about the relationship between language and thought is by no means over. The Economist website held an online debate in December 2010 on just this topic. Suffice it to say, there is even less consensus on the effects language might have on behaviour, or the mechanism by which this might occur. Even if Chen's findings are robust enough to show a strong correlation between language and saving behaviour, there is a lot that needs to be done to show that it is a causal relationship, i.e. it is the language causing a change to saving behaviour.
However, my main concern with Chen's work is the actual application of this to tense systems of the world's languages. This will be the subject of my next post.
[Guy Deustcher's article in the New York Times (26/08/10): Does Your Language Shape How You Think? provides great coverage of the research in linguistics to a non-specialist reader.]
