Futurity in Chinese and English and Its Supposed Economic Consequences
A guest post from Victor H. Mair, Professor of Chinese Language and Literature, Department of East Asian Languages and Civilizations, University of Pennsylvania.
The linguists at Language Log have several times discussed the work of economist Keith Chen on the alleged economic impact of the way the future is conveyed in different languages:
Keith Chen, Whorfian economist
Thought experiments on language and thought
Keith Chen at TED
And Chen himself has presented his views on the same forum:
Now that Chen’s work has been featured in Scientific American (How Your Language Affects Your Wealth and Health), I believe that it is time for another look at Chen’s theories.
Since the LLoggers have done a good job of exploring Chen’s general linguistic theories, while no one has yet examined the way the future is expressed in Chinese, I thought it would be worth focusing on that aspect of his argument.
Chen’s research cited by Scientific American is presented in this paper (pdf) that is forthcoming at the American Economic Review:
The Effect of Language on Economic Behavior: Evidence from Savings Rates, Health Behaviors, and Retirement Assets
M. Keith Chen
Yale University, School of Management and Cowles Foundation
American Economic ReviewAbstract
Languages differ widely in the ways they encode time. I test the hypothesis that languages that grammatically associate the future and the present, foster future-oriented behavior. This prediction arises naturally when well-documented effects of language structure are merged with models of intertemporal choice. Empirically, I find that speakers of such languages: save more, retire with more wealth, smoke less, practice safer sex, and are less obese. This holds both across countries and within countries when comparing demographically similar native households. The evidence does not support the most obvious forms of common causation. I discuss implications for theories of intertemporal choice.
Here’s the only Mandarin sentence that Chen cites (on p. 3 of his paper) and his complete analysis of it:
For example, if I wanted to explain to an English-speaking colleague why I can’t attend a meeting later today, I could not say ‘I go to a seminar’. English grammar would oblige me to say ‘I (will go, am going, have to go) to a seminar’. If on the other hand I were speaking Mandarin, it would be quite natural for me to omit any marker of future time and say
Wǒ qù tīng jiǎngzuò
I go.PRS listen seminar
“I am going to listen to a seminar”.
with no reference to future time, since the context leaves little room for misunderstanding. 7
Footnote 7: In this and all subsequent examples I follow the Leipzig glossing rules, where fut and prs indicate future and present morphemes. See Croft (2003) for details
The situation regarding the expression of futurity in Chinese is much more complicated than that. Any decent grammar of Mandarin (e.g., that by Y. R. Chao or the one by Charles Li and Sandra Thompson) will show in great detail how time (including the future!) is indicated in Mandarin, and several other Chinese languages also have respectable grammars (e.g., that by Stephen Matthews and Virginia Yip for Cantonese) which treat these matters.
First of all, let us dispose of Chen’s reference to “future and present morphemes”, since there is no such thing in his sample sentence. Furthermore, neither the invocation of the Leipzig glossing rules, nor Croft 2003 shed light on what Chen is trying to prove.
It’s very unusual in discussions of Chinese grammar to speak in terms of “future morphemes” or “present morphemes.” That would work for languages like French or Spanish that have tense and indicate some (though not all) tenses through morphemes (more commonly called “verb endings”) suffixed to the end of a verb, but the consensus is that Chinese doesn’t even have grammatical tense. By labelling qù (“go”) as a “present morpheme“, Chen completely vitiates his own argument.
The tricky part here is that the Leipzig rules have been devised to facilitate computer searching of glosses. The question is whether qù ought to be glossed go.PRS in this context rather than FUT, or not at all. Since qù is unmarked, the PRS is inferred, not overt. Also, the whole sentence might mean that the speaker is right at that moment on his/her way to the seminar room, but it might also mean that he will go there several hours later. To distinguish between the two possibilities, Chinese has very precise means for specifying present and imminent action versus future action. I will discuss some of them below.
Now that we have disposed of Chen’s fallacious “present morpheme,” let us move on to a dissection of the sentence itself. All native speakers of Mandarin whom I consulted agree that “Wǒ qù tīng jiǎngzuò” is a perfectly acceptable sentence, BUT they also pointed out numerous problems with Chen’s handling of this short sentence.
First of all, jiǎngzuò 讲座 is better translated as “lecture” than “seminar”, which would normally be the translation of yántǎo huì 研讨会.
In Taiwan the terminology is slightly different, so for “I’m going to a lecture” they would say wǒ qù tīng yǎnjiǎng 我去聽演講.
Chen’s sample sentence could also mean “I went to attend a lecture” or “I am on my way to attend a lecture”. Rendering tīng as “to listen to” reveals clumsy unfamiliarity with the meaning of the Chinese as one would express it in English.
At first glance, one might interpret the sentence to mean “I’m going to the / a lecture”, as in, “Okay, I’m heading out now.” Of course, given the right context (“Yesterday, I got to work late. I went to the lecture. By the time that was done, I had no time to finish the project.”), the sentence could also be interpreted as past tense.
Chen’s sentence only indicates the future in certain, specific contexts and under certain conditions. For instance, if A meets B on the road and A asks B:
Nǐ qù nǎlǐ ya? 你去哪里呀? OR Nǐ shàng nǎ’er qù a? 你上哪儿去啊
(“Where are you going?”)
B can reply:
Wǒ qù tīng jiǎngzuò 我去听讲座
(“I’m going to [attend] a lecture”)
Without the context provided by the above question, we can specify very clearly when the action takes place, but we do it by other means:
Wǒ xiàwǔ yào qù tīng jiǎngzuò 我下午要去听讲座 (“This afternoon I’ll go to a lecture”) — note the auxiliary verb yào 要 indicating futurity and the time word xiàwǔ 下午 (“afternoon”) indicating when the attendance will take place
Wǒ xiànzài yào qù tīng jiǎngzuò 我现在要去听讲座 (“I’m going to the lecture now”) — again we have the auxiliary verb indicating present or imminent action and the time word xiànzài 现在 (“now”)
Wǒ gāngcái qù tīng jiǎngzuò le 我刚才去听讲座了 — here we have a sentence final particle indicating completed action and the time word telling us that it happened in the recent past
This is how Chinese people really speak, not in some disembodied, timeless arena of being that Chen would have us imagine.
If somebody does, in fact, speak Chen’s sentence, it would almost always be in response to a question that specifies the time. For examples:
Nǐ xiàwǔ xiǎng zuò shénme? 你下午想做什么? (“What are you intending to do this afternoon?”)
Nǐ (xiàwǔ) xiǎng qù tīng yīnyuè huì háishì qù tīng jiǎngzuò? 你(下午)想去听音乐会还是去听讲座? (“This afternoon, are you intending to go to a concert or a lecture?”)
It is only under such circumstances that a person can reply with Chen’s sentence:
Wǒ qù tīng jiǎngzuò 我去听讲座(“I’m going to [attend] a / the lecture”)
Note that the questions supply the time with an auxiliary verb and a time word, which enables the respondent to omit them. The time when the action is going to take place is explicit in the question and it is clearly implied in the reply. Thus futurity is just as evident in Chinese discourse as it is in English or French or German; the only difference is that the mechanisms for expressing it are different.
Some more examples of how this works:
- Question: Nǐ xiàwǔ yào zuò shénme? 你下午要做什么? (“What will you do this afternoon?”) — yào 要 is the auxiliary verb, indicating future
Answer: Wǒ (yào) qù tīng jiǎngzuò 我(要)去听讲座 (“I’m going to a lecture”) — yào 要 is optionally omitted, but the sentence means that the respondent will go to a lecture, because the information about the time has already been given in the question
- Question: Nǐ zuótiān wǎnshàng qù nǎ’er le? 你昨天晚上去哪儿了? (“Where did you go yesterday evening?”) — 了 is the verb particle, indicating past
Answer: Wǒ qù tīng jiǎngzuò (le) 我去听讲座(了) (“I went to a lecture”) — 了 is optionally omitted, but the sentence means that the respondent went to a lecture last night, because the information about the time has already been given in the question
- Question: Nǐ dàihuì yào qù nǎ? 你待会要去哪？(“Where will you be soon be going?”) — dàihuì yào 待会要 indicates very near future
Answer: Wǒ qù tīng jiǎngzuò 我去听讲座 (“I’ll [soon] be going to a lecture”) — omitting the dàihuì yào 待会要 for the near future, which is nonetheless still implied
- Question: Nǐ zuótiān zěnme méi lái shàngkè? 你昨天怎么没来上课? (“Why didn’t you come to class yesterday?”) — zuótiān 昨天 (“yesterday”) is the time word and méi 没 (“had / have not”) is the negative past auxiliary
Answer: Wǒ qù tīng jiǎngzuò 我去听讲座 (“I went to a lecture”) — omitting the time word, which is nonetheless understood.
Without the specific temporal context supplied by the questions, and optionally also provided by the answer, these replies would make no more sense in Chinese than a sentence devoid of tense would in English.
The best friends of my wife and me were a Chinese couple from the Mainland and Taiwan. The wife had a B.A. in English from a college in Taiwan and an M.A. in the same subject from an American university, whereas the husband had a B.A. in statistics from a college in Taiwan and an M.A. in the same subject from an American university. They would say sentences like this:
Husband: Yesterday I go see movie.
Wife: Yesterday I went to see a movie.
Husband: Tomorrow I go see movie.
Wife: Tomorrow I’ll go to see a movie.
Although the husband’s grammar never improved in 40 years, he — like all Chinese with whom I’m familiar — knew quite well how to indicate the past, present, and future. The wife did it like a native English speaker; the husband did it like someone speaking English with Chinese grammar and syntax.
If Chen’s analysis of the only Mandarin sentence in his paper is so open to question and refinement as I have shown it to be in the above paragraphs, then we need to look very carefully at the sample sentences of all other languages that he cites and analyzes. After all, other than English, Mandarin is supposedly the language that Chen knows best.
Anyway, the fundamental premises and methodology of Chen’s paper are so shaky that we really cannot take his sweeping conclusions seriously at all. Language does not CAUSE behavior, as Chen claims. So why are so many in the media hyping him this way? I think that it is because they don’t know anything about Chinese, much less linguistics and economics. We often encounter pseudo-wisdom about Chinese, both made-up proverbs and phony etymologies.
Legions of pop economists have deluded the public with their pompous proclamations about how “the Chinese character for ‘crisis’ is made up of ‘danger’ + ‘opportunity'”. Keith Chen’s assertions about the lack of futurity in Chinese are equally delusory.
Astute comments by colleagues who have seen a draft of this post:
1. Randy Alexander:
You might add that English does not have a future tense either!!!
English deals with future time in basically the same way Chinese does, either through modal verbs, or using present tense (if in Chinese you can even call it that) with context clues and adverbial modifiers (时间状语).
2. Jeremy Goldkorn:
Has Keith Chen ever actually been to China? There are more smoking gamblers having unprotected sex on every city block here than in the average American state….
[Thanks to W. South Coblin, Neil Kubler, Jim Unger, Bob Sanders, John Rohsenow, David Moser, David Branner, Perry Link, Anne Henochowitz, Cynthia Ning, Yunong Zhou, Maiheng Dietrich, Grace Wu, Zheng-sheng Zhang, Melvin Lee, Liwei Jiao, Sophie Wei, Cao Lin, Cheng Fangyi, and Jing Wen]
You make a great critique of Chen’s argument. Chen really does “cherrypick” his sentence and divorce it from context to make an untenable argument. I have a minor critique/point from your own wording, namely the use of 要 as “futurity.” The Chinese auxillary verb express “intentionality” more than “futurity” which is to say that it refers to an imminent action that one has a desire or will to do, which almost always applies to the future but is not necessarily temporal. Other aux. verbs fall into this same category like 想， 謀， 將， 會. Historically speaking the two concepts are distinct and become more blended with time and the changes in the language. I make the point more to help clarify to other readers that this aux. term is not necessarily a “future” type conjugation, but a concept closer related to the “intention” of the speaker.
The historical background is interesting; English is similar, since English “will” also indicates “volition” or “intentionality”. But I would note that in modern Chinese, yao can be used as an auxiliary verb expressing future time not only for people but also for inanimate nouns (Feiji kuai yao jiangluole “The plan will land soon” or Xiawu yao xiayu le “It will rain in the afternoon”). I don’t think the concept of “intentionality” works for airplanes or meteorological phenomena.
I think Christopher’s makes a useful point, and I’m not at all sure that meteorological phenomena or airplanes can be written off as far as ‘intentionality’ goes.
The use of the ‘dummy subject’ in, “It will rain in the afternoon,” seems to me to suggest a reluctance on the part of English speakers to do with out an agent. The weather, or the heavens, might just be viewed as having intention. Obviously, Chinese doesn’t require a pronoun in such situations.
Similarly, might, “The plane will land soon,” be understood in some situations as a kind of conventionalized personification, or even a transfer of agency? The pilot will land the plane soon. We will reach our destination soon. We will land soon. It feels as though there is something of a sliding scale between the predictive and the personified but, in some cases, it seems awfully close to attributing agency to the machine – a tendency we’re familiar enough with, e.g. “Argh, it won’t let me log in.”
English and Chinese are very close in this regard.
Randy, thank you for saying that out loud. I was baffled when I read some version of Chen’s work that the reviewers (and Chen himself) all apparently failed to notice that two of the three English sentences he gives have exactly the same properties the Chinese one does with regard to present/future time–and that those two, “I am going/have to go to a seminar” are better responses to “Why aren’t you going to the meeting this afternoon?” than “I will go to a seminar.” Maybe they get so caught up in the odd-sounding English translation of the Mandarin sentence, which they’ve been told is supposed to be odd, that they miss it? Or are most Americans are such grammar-phobes that they’ll willingly believe anything that sounds faintly like grammar class? Sorry, personal rant that has nothing to do with Chinese. I’ll give it a rest now.
I’m in the camp of those who think Chen’s methodology and conclusions are seriously flawed. But I do think that some of the specific criticism that has been aimed at him is off the mark. Chen is not making any claims about whether Chinese speakers have the communicative means to be clear about relative time (i.e. past, present, future), whether by making use of context, of optative/auxiliary verbs, or lexical means (such as time adverbs). Indeed, I’m sure he would readily agree that Chinese speakers are as certain as English speakers about whether they are talking about past, present, or future events. Chen is making a much more specific claim, namely that Chinese does not encode tense grammatically. Or, as he puts it, that Chinese “grammatically associate[s] the future and the present”. This is in fact true — there are no distinct present- and future-tense forms of verbs in Chinese, nor indeed is there any other tense distinction encoded grammatically (as opposed to lexically or contextually). What Chen — who is not a linguist — has done is to reference language descriptions that allow him to categorize languages as grammatically associating the future and the present or as NOT grammatically associating the future and the present. He claims to have found a degree of correlation that is greater than chance probability linking this feature of languages with certain cultural behaviors.
There are many ways to attack his methods and conclusions. For example, determining whether a language has grammatically identical forms for present and future is not always a simple matter. Chinese does and Japanese does, yes. But English is sort of a mixed case, since sometimes we use grammatically distinct forms (“goes” vs “will go”) but sometimes use identical forms (“is going”). The vagaries of how his data source categorizes English will affect his results. Then there are the arguments over whether the English future with “will” is truly a grammatical tense or if it is a periphrastic construction with an optative auxiliary verb. There are questions about Chen’s sample sizes of languages and cultures, about how he deals with historical changes (in both savings rates and grammatical structures), etc. etc.
But I would argue that simply demonstrating that Chinese speakers are not confused about whether they are talking about past, present, or future events is irrelevant to the claim Chen is making.
If Chen’s claim were true — and as I said, I don’t believe it — it would raise interesting questions about the cognitive role of grammatical categories on non-linguistic behaviors. That’s certainly an area worthy of additional research.
If, as largely seems to be the case, Chen’s distinction concerns languages which do vs. do not encode tense grammatically, why the concern at this example with showing that speakers of Mandarin, etc., may dispense even with such (syntactic) means as they do possess of indicating relative time?
The answer may relate to Randy’s and Katie’s comments: given that, in many cases, English also employs syntactic means of encoding time information (‘will’, ‘going to’, etc.), might we find here an urge to maintain a C/E distinction by reference rather to “optional” (the auxiliary verb yao4, etc.) vs. “required” (whatever ‘will’ is) elements? (Highly unfortunate) shades here of Alfred Bloom (1981) on Chinese argument structure…
So Chen might ought to have stuck to the obvious fact of the single Mandarin word qu4, etc…. but this would have left English in a rather ambivalent, analytic-ish position.
It seems that Östen Dahl’s comment on Chen’s LL post settle’s matters as far as categorizing future time preferences goes. Chen appears to have found a strong correlation between something and something else. But quite what that something is, I wouldn’t like to guess. I hope he’s clearer on the something else.
Still, I’m always happy to see discussion of the conceptualization of time. As Christopher and Neil have noted, there is quite a lot of common ground between modern Chinese and modern English.
I think Sima’s summary has it exactly right: “Chen appears to have found a strong correlation between something and something else.” Since it’s not clear what those somethings are, and since it’s even less clear if they are meaningful somethings, it’s hard to conclude that the correlation involves a meaningful causality.
@JS: I’m having a hard time understanding exactly what you are asking about. The way I see it, every language has a different set of categories that are encoded in the grammar. Such categories include number (for nouns), tense (for verbs), degree (for adjectives), evidentiality (for verbs), etc. etc. If a language does not encode a category in its grammar, it will always have a means of expressing the same semantic content. But because the category isn’t mandatorily expressed, speakers will also tend to omit expression of the semantic content if it is either irrelevant or understood from context.
I think it’s worth reiterating that whatever the similarities and differences between English “will” and Chinese yào, one difference is clear: “will” is incompatible with past-time references, and yào is not. This is one reason that “will” is interpreted as a tense marker in English. Compare:
1a. Tomorrow the plane will land at 3:00.
1b. *Yesterday the plane will land at 3:00.
2a. Míngtiān fēijī kuàiyào jiàngluò de shíhou …
2b. Zuótiān fēijī kuàiyào jiàngluò de shíhou …
3a. Tomorrow when the plane is about to land …
3b. Yesterday when the plane was about to land …
It’s clear that constructions like yào and kuàiyào do not encode tense — that is, do not encode time of action relative to time of speech act — but like English “about to” encode other aspects such as imminence, intentionality, and so on.