Whorfian Economics

Via Mark Thoma

Languages di?er widely in the ways they partition time. In this paper I test the hypothesis that
languages which grammatically distinguish between present and future events (what linguists call strong-
FTR languages) lead their speakers to take fewer future-oriented actions. First, I show how this prediction
arises naturally when well-documented e?ects of language on cognition are merged with models of decision
making over time. Then, I show that consistent with this hypothesis, speakers of strong-FTR languages
save less, hold less retirement wealth, smoke more, are more likely to be obese, and su?er worse long-
run health. This is true in every major region of the world and holds even when comparing only
demographically similar individuals born and living in the same country. While not conclusive, the
evidence does not seem to support the most obvious forms of common causation. Implications of these
?ndings for theories of intertemporal choice are discussed.

Unforunately I read that just before bed, so it bugged me all night. Whorfism – the belief that characteristics of a language alter the neurology, thought or behaviour of its speakers. Whorfism,  the linguistic equivalent of Austrian Business Cycle Theory ; seductive rubbish readily ensnaring the lazy minded.

Back when I did my own work using language as an explanatory variable I was at pains to point out that the characteristics of the language were of no account. When I thought the idea was arising  elsewhere on Troppo I intervened. In that last case I was even using a Future Tense hypothesis like Chen uses, only to mock the idea.

Still Chen (the author) does not appear to be lazy or feeble minded. He’s certainly aware of the esteem in which the Whorf Hypothesis is held. So I think it’d be wise to treat this as legitimate, but in my mind probably spurious research. I still think the economics of language, or linguoeconomics is ripe for more research. If this opens up the doors a little, it only does good.

At Language Log, Geoffrey Pullum critiques Chen on the linguistics. Interestingly:

When I engage in amateur reflection on how language might affect thought, I find that I might just as well be convinced that a language with grammatical future tense marking would have speakers who paid MORE attention to worrying about the future. After all, they use a linguistic device that explicitly picks it out. Chen’s hypothesis is that instead they would naturally pay LESS attention to what the future might hold in store. Which hypothesis is right? Why is it Chen’s favorite that is right?

The alternate hypothesis Pullum is actually the one I used to mock Whorfism here, the reverse relationship Chen proposes would never have occurred to me [fn1]. Earlier he uses the Pirahãas a counterexample to Chens hypothesis when I used as an example of false correlation (the “small tribal group”) .

His Stable mate Mark Liberman merely focuses on the potential for spurious correlation. Chen graciously responds on the same blog, but I think their skepticism is well warranted…..even if they appealed to my preconceptions.

[fn1] Which weakens the findings in a Bayesian light I guess.



About Richard Tsukamasa Green

Richard Tsukamasa Green is an economist. Public employment means he can't post on policy much anymore. Also found at @RHTGreen on twitter.
This entry was posted in Uncategorized. Bookmark the permalink.

14 Responses to Whorfian Economics

  1. conrad says:

    There are some rather marginal areas where you can find Whorfian effects (e.g., thinking about time with timelines). I prefer to look at it the other way — do major differences across languages like word order make any difference to the way we think about things in general or even explain things to others non-verbally ? The answer to that is no — and there’s an interesting paper here, looking at it.

  2. Mike Pepperday says:

    Can we think if we don’t have the words? The beginning of understanding is naming things. Technologists have to learn the technical terms and then they are in a position to discuss and think about their subject.

    The doctor informs you that tests show that you have (say) Gribble’s disease. He tells you no one knows the cause or a cure. You have learnt nothing and yet you feel like you know something. You can go home and tell your family what’s wrong with you.

    It is often remarked that English has no equivalent of the German word, “Schadenfreude” (and what does that reveal about THEM?!!).

    I have just found out that German has no word for “connive” or “collude.” They have one for “conspire” or “plot” (sich verschwören) but its use is confined to the serious meaning relating to such things as treason and betrayal. Connivance and collusion are missing.

    They can use a phrase, as we can for Schadenfreude, but mustn’t it mean that the concept does not come readily to mind, that it is not discussed and not readily imputed to others? As a result, wouldn’t German social relations be different from English social relations?

  3. conrad says:

    “but mustn’t it mean that the concept does not come readily to mind, that it is not discussed and not readily imputed to others”

    Mike, if you’re worried about the efficiency of various grammatical patterns versus simply being able to describe the same thing (even from person to person — does your connive have exactly the same meaning as as mine? Seems unlikely as mine is connected to meanings at least in part derived from a different set of sneaky people than yours) then there are innumerable differences across and within languages. For example, the probable reason that grammar-pedants hate split infinitives (and e.g., double negatives), apart from that’s what the Chicago Manual of Style told them to dislike, is that they create an extra demand on short-term memory. So, for example, if I measured how long you took to understand “He shook the drink up”, it would be slightly slower than “He shook up the drink”. However, I doubt that would change the meaning after you had understood it.

    There are in fact a small set of words that do have special properties such as capturing your attention very easily (e.g., your name, rude words), and perhaps if Schadenfreude was one of those then there really would be a difference in processing compared to “laughing at your misery”, since I could shout “Schadenfreude” and it would be a bit like shouting your name, which is clearly different and would get a different reaction to shouting e.g., “peanuts”. Whether these sorts of small things are very meaningful in terms of overall language usage is another question, which is not to say that they’re not interesting by themselves.

  4. Mike Pepperday says:

    Your connive and mine are for this purpose the same – the English meaning. The Germans don’t have anything. I lived there for years, worked, played, read, and listened to radio and TV like everyone else and never noticed the concept was missing. I only found out a couple of weeks ago and was most surprised to find the dictionaries struggling to interpret it.

    I can only think the matter must seldom arise there. The fact that we even want to distinguish connive from collude and that these are both perfectly ordinary words must mean that the concept looms larger to us than for those who don’t have any word. It must mean we can and do think of what they stand for more readily and that we can talk about them more readily because we don’t need a circumlocution.

    I once taught a mathematical cartography subject to technical students and our main reference used Greek letters. I issued copies of the alphabet and told the students to learn it but they didn’t. So people were talking about “that squiggly thing divided by the one that looks like n.” It was hopeless. I gave notice of a test and after that things went smoothly – just because they had the words. Once you have a word for something, you can move on: relate it, compare it, build on it, qualify it.

  5. Tel says:

    …English has no equivalent of the German word, “Schadenfreude”…

    Well, no equivalent other than “schadenfreude” of course, which is the main thing keeping English one step ahead of every other language.

  6. Mike Pepperday says:

    Tel, I think that is an ethnocentric conceit.

    Conrad, I just remembered something which brings the occurrence of lacking concepts into higher relief. We distinguish horse, steed, and nag. So, quite precisely, do the Germans: Pferd, Ross, and Gaul. It’s not easy to say what the distinctions between these three creatures are and yet they must be, or once have been, important in both cultures.

  7. conrad says:


    if you’re thinking about the Whorfian stuff, then the idea is that causation runs in the opposite direction to your example — what you are showing is that if something is important to us, we invent different words for it. That works sometimes (e.g., types of beer), and not other times, even when we want it to (e.g., Eskimos have words for snow). Experts do this all the time — they invent their own jargon to describe things better. However, all this shows is that once you know something well, you can specify it better and more easily and that by specifying things better, you can converse more easily and not blow out your memory using examples such as you gave before. This is just standard language change.

    To get support for the Whorfian argument (i.e., causation in the opposite direction), basically you need two cultures that know about horses enough such that they can happily distinguish the features of 3 types of horses (horses, steeds, & nags). However, in one of those cultures, you would only have one word for all three types of horse and the other you would have 3. You would then need to show that because of the difference in the number and scope of words, people process the features of horses differently when doing some non-linguistic task. Perhaps one type of feature really is impossible to learn without the word for it. You can actually do this with containers, which Malt et al. did here, since languages have lots of different names for containers whose scopes of use differ vastly. However, this doesn’t really affect how people classify those containers non-linguistically. Thus, whilst people might talk about containers quite differently, that doesn’t affect how people conceive them non-linguistically. Of course, it might well be easier to describe some type of container in one language than another (or type of horse), but I don’t think that counts as evidence for the Whorfian stuff — this just makes it easier to use and learn about something that certainly wouldn’t be impossible even if the words and grammar you used were slightly more inefficient.

  8. Tel says:

    Mike, I think it’s lack of conceit actually that has resulted in an eclectic tendency to rapidly adapt and assimilate useful concepts and technology from arbitrary origin (i.e. the diametric opposite of “not made here” syndrome). The English language is littered with examples, and so is British culture, US culture and even Australian culture.

    Of course, from a historic point of view, regularly engaging in warfare (with approximately equal wins as losses) does tend to jumble up a language a bit. I’d be looking at language as the chart recorder of a people’s approach to the world, rather than the engine of that approach.

    By the way, in the particular case of “Schadenfreude”, it is wrong to say there is no English equivalent. The literal translation from the German is “shady joy” and it is only written as a single word in the German for the simple reason that the linguistic convention in German is to adjoin adjectives to nouns making compound words that are written as single words. English does not follow the convention so we write adjectives before the nouns, and it makes essentially no difference which convention is in effect (other than the rather arbitrary distinction of exactly what make a “word”).

    If you use “shady joy” in an appropriate sentence, most English speakers would catch the drift, or similar pairs such as “spiteful pleasure”, etc.

  9. Mike Pepperday says:

    Thanks for pointing me at Malt et al. Have to accept their findings but I was thinking as I read it that distinguishing jar from bottle from container was pretty trivial and then I read that the authors think these are complex. Perhaps they are, compared with colours, but I am left wondering how they would experiment with collusion and connivance. Re causal direction – I can’t get worked up about it for surely behaviour and thoughts are mutually reinforcing.

    The conceit is yours, Tel, not that of the language. If a Frenchman were to expatiate on the glories of English eclecticism it would be different.

    Schaden means damage or harm. The rest of that paragraph is also incorrect. Both German and English are known to philologists as agglutinating tongues (skylight, railway) and whether compounds have spaces between words is immaterial (railway carriage, computer screen). We can say “Danube Steamship Company Captain” just as easily as the Germans. French can’t say any of these compounds but none of this is really relevant to Sapir-Whorf.

  10. Tel says:

    “… whether compounds have spaces between words is immaterial …”

    Agreed. Thus English does have many close equivalent, and equally expressive compounds to the German “Schadenfreude”, so nitpicking details aside, I guess we agree on the main point.

    By the way, “skylight” and “railway” are not counter examples to what I said earlier, given that neither “sky” nor “rail” is an adjective (but hey, that’s just nitpicking). But at any rate, English speakers are far less enthusiastic about the practice, but not purist enough to abandon it completely. The only statements you can make about English are statistical statements after all.

  11. Mike Pepperday says:

    Schaden is a noun. As far as I know, all languages use adjectives to qualify nouns. Some languages – so-called agglutinating ones – can combine noun with noun. Hence Schadenfreude, hence railway.

    Generally, neither German nor English combines adjectives with nouns to make a single word. (There are isolated instances: red-wine, Weissbrot. )

    That French minister is (bravely, resignedly) facing up to the pervasiveness of English. He is not recognising any supposed superior eclecticism. English is not dominant because it has some special intrinsic quality (It has none that I know of.) but because it had the most powerful military. With your link to the French minister, you are mistaking dominance for virtue – and that’s no longer conceit; that’s hubris.

  12. FDB says:

    “Generally, neither German nor English combines adjectives with nouns to make a single word.”

    No, in German this is absolutely de rigeur.

  13. john says:

    Meaning is mostly at a much higher level of grain than the level of simple string syntax. Otherwise translation of things like War and Peace or the transposing of ‘ode to joy’ e to ring tone piece would create a totally unrecognizable ‘new’ thing.

    However there are things like nonsense poetry, contemporary political jargon and cultural theory that are essentially syntax without semantics; Therefore Whorfisum might apply in these cases??

Leave a Reply

Your email address will not be published. Required fields are marked *

Notify me of followup comments via e-mail. You can also subscribe without commenting.