Deep Structure of Language

In English, we use “have” to represent past experience, and “will” to represent future — “I have skied before”, “I will ski tomorrow”.  Since English is full of homonyms, we seldom think about the fact that “have” is also used to indicate posession, and “will” is used to represent desire (“I have some money.  I will that it be given to my nephew when I die.”)  It’s just a coincidence, right?

Actually, Chinese has the same construct, but without homonyms.  In Chinese, the word “you” (pronounced “yo”) is used to signify both posession and past action (exactly like “have”) in English.  And the word “yao” (pronounced “yow”) is used to signify both wanting and future action.  In Chinese, there are plenty of homonyms too, but every word has a different character, so you can tell the difference between homonyms in written text.  In this case, the “you”s and “yao”s are exactly the same character.  There is no ambiguity — wanting is future action, and posession is past action.

It is remarkable that two languages so distantly separated use exactly the same constructs to represent the most basic concepts of past and future.  These are fundamental to human thought.  Since English has many homonyms, and “will” has fallen out of use as a word for “want”, it’s understandable that most English speakers don’t make the connection.  In fact, the use of inflected verb forms makes the “have” and “will” somewhat unnecessary, and many English speakers omit them.  But it’s interesting that most Chinese speakers don’t notice it either.  In part, I think it is because modern Chinese has an alternative form to “you” that can be used to indicate past action (similar to English -ed).  In Taiwan, people still say “wo you kan shu” (“I have read a book”), but in the north, they prefer “wo kan le shu”.  The “you” form is considered stiff and old-fashioned.  But even Chinese speakers I’ve talked to have assumed that this is just some accident of history.

The fact that exactly the same structure exists in both languages has got to be more than just a coincidence.  Having, wanting, past, and future are four of the most fundamental concepts for enabling human thought.  I am sure that some linguists have remarked on this before, but I couldn’t find anything about it.

So, why is posession so tightly tied with past, and want so tightly tied with future, in our deep structure?  Wanting is somewhat understandable, but for many people, the past is about loss.  It’s also interesting that Chinese and English have somewhat deprecated the opposite poles of this relationship — English has mostly de-linked want/future, while Chinese has started to deprecate posession/past.  While I think that sapir-whorf is a bunch of B.S., I wonder if this indicates anything about relative cultural priorities.  I would be interested to read more about this.


10 thoughts on “Deep Structure of Language”

  1. If you achieved something in the past, you “have” it in your experience. In order to achieve something in the future, you “will” that it be done. It’s always been interesting to me that many languages have developed more than one past tense form of verbs (in addition to or in combination with constructs like “have”), but very few languages have a pure future tense — they almost always use a present or past tense in combination with other words (or even context) to indicate that the action occurs in the future.

  2. Interesting. I still think that is pretty abstract (“having in your experience” — since you have the present in your experience to a degree much greater than memories).

    My theory was something like this: I think having and wanting were the very first concepts in human language (or close to it). You either posess something, in which case you defend it from other monkeys; or you want something, in which case you try to take it away from other monkeys. So the concept of “me”, “mine”, and property/posession were the most important concepts. Time and time-coding probably came after (communication about time is necessary for more planned type activities, but not for stealing food from other bands of monkeys). When it became necessary to communicate about time concepts, the monkeys just re-used the two basic concepts that the language had — “having” completes “wanting”; it moves wanting to the past.

    Your point about context is true — since Chinese does not have inflected verbs, past or future are very often indicated by context. But the pure structure of both languages is have/want.

    I also find it interesting that Chinese represents chronological order using up/down instead of behind/front as in English, and the up is past while down is future. Each language seems backwards to speakers of the other; Chinese often mix up last/next, and English often mix up shang/xia. I have seen research which shows that this does not actually affect the thinking of either culture, and that it apparently doesn’t bias the mind of someone speaking her native tongue. But it tells me that *positional* modifiers for time came *after* posession/wanting, since these show more divergence. I think this shows that time as position is a newer concept than time as posession.

  3. Ni hao! Wo xing Mindslant. I should really get back into Chinese. Also Russian. Also Esperanto. My French isn’t too horrible yet. Spanish is getting better because I live in Texas and I’m hungry so whaddya going do? I’ve neglected a lot of my language studies.

    But that’s not the point here. Have you ever studied Noam Chomsky. Not his political stuff, his I/E Language theories. I think it has some real bearing on your observation. Nim Chimpsky…that still makes me laugh.

    It proposes language is a universal function of the brain in Humans. If chimps do have language, they have a innately different language because they have innately different brain structures.

    So mostly I’m saying:
    *Noam Chomsky
    *I/E Language theories

    Be careful though, there’s early incarnation Chomsky and late incarnation Chomsky. It’s all pretty good, I mean he did drive the coffin nail/knell in the Behaviorist Model.

  4. Yeah, I liked Chomsky’s early stuff, which is basically textbook material now (though some has been replaced by more modern theories). I think he was one of the first to talk about deep structure (don’t remember if he called it that).

    I have read some more recent stuff looking for neurological/physiological basis for semantics and cognition; and it is all pretty interesting. Basically at the deepest level, how some thought patterns get processed by the machinery that lets us throw a rock at a moving target, and so on. I believe the two are codependent variables — certain thought activities can alter the physiology of the brain, and certain physiology traits of the brain can impact thought.

  5. I have several comments to make on your post. I should state at the beginning that the background assumptions that I make about language are different than those you make. You are using the terminology of a “generativist” (which I use to refer to any Chomskian theory as well as other recent theories that still assume a LAD and/or universal grammar), whereas I prefer to adopt a “functionalist” perspective (see e.g. the work of T. Givon, William Croft, Ron Langacker, Leonard Talmy, George Lakoff, Matthew Dryer, Paul Hopper, or Sandra Thompson; I am including ‘cognitive linguistics’ under the rubric of ‘functionalism’ here). I am very concerned with treating the study of language as a science, and I feel that this emphasis on taking a scientific approach is woefully neglected by generativists. The way I think about the problems you discuss is different than the way you do, and so it is no surprise that my conclusions are different as well.

    First, the way most linguists analyze ‘have’ as part of a verbal phrase such as ‘have skied’ is not that it represents ‘past experience’, but rather that it encodes relevance to the present situation (i.e. the time of the speech act). Notice the past tense marker on the verb ‘ski’ in ‘I have skied before’, which does express information about when on the time line the action occurred. Both ‘I skied there many years ago and ‘I have skied there’ could refer to the same event in the past, and so it’s clearly not the ‘have’ that’s telling us that it took place in the past. Instead, it’s claimed that the ‘have’ is used when the issue of whether one has ever skied (or has ever skied in a particular location, etc.) is relevant to the current (moment of speech) situation. Think of these two scenarios:
    1) We are hiking in the mountains in the winter, and our friend Jerry takes a fall and needs emergency help. Jerry has brought his skis with him, but he’s in no condition to ski back to the lodge. You or I must go, but you’ve never been on skis before, and you don’t know if I have, either. You could ask me ‘Have you ever skied before?’, but there’s really no natural way to ask the question with
    *‘Did you ski before/yesterday/this route?’ (I use an asterisk here to represent an ungrammatical, unnatural, or pragmatically unlikely sentence). This is because whether I have skied before, and therefore presumably have some ability to ski that is better than yours, is extremely relevant at the moment of speech.
    2) You go for a vacation in Colorado, and we meet for coffee several days after you return. You want to tell me about what you did while on vacation. You could say ‘I went skiing at (place)’, or ‘I skied almost every day of the trip’ but not *’I have skied on my vacation’. This is a less ‘marked’ situation than the one above in (1), and the point of the story is to inform me of something, not to state that your having skied (rather than scuba dived) is in some way particularly relevant in defining or changing the present (again, at the time of the speech act) situation. Notice that, if you have always wanted to be the kind of person who has skied all over the world, and what you’re doing is checking off “Colorado” on your list, and therefore you are changed into a new person by having skied there, and you want to let me know this, you could say ‘Well, I’ve now skied Colorado.” The relevance to your present definition of yourself allows this.

    I hope that I have convinced you with this informal explanation. There is some (fairly limited, to my recollection) empirical evidence in the literature, but I won’t attempt to gather and then simplify and present it here and now.

    Second: while it’s true that Chinese and English have a similar correspondence in their grammatical expression of the future tense and their word for desire, this is just the beginning of the story. For instance, there are many other languages that share the same characteristic as well. So it’s an even stronger link than you thought, which shows that it reveals something profound about human thought, right? Wrong. There are several different reasons why such a link could be found in multiple languages. One is that, what we want, we don’t have (or else we wouldn’t want it), and we often try to obtain/attain it, and if we do obtain it, it will inevitably be at some future time (as compared to the wanting). Therefore, there is a natural connection between desire and future time. Additionally, since if we want something, we necessarily don’t have it, wanting is closely linked to the ‘irrealis’ (which is a term linguists and philosophers have used for a long time –since Aristotle, though his was obviously a Greek translational equivalent – as one of the four epistemic modalities, expressing something like ‘possible truth’ – see e.g. Carnap 1947 “Meaning and Necessity”). The irrealis is known to be linked to the expression of future tense (i.e. it is common for languages to use one form, or historically related forms, to express the two meanings). Why would this be the case? Since future events are ‘possible truths’, much as the world we set up in our minds when we dream of ourselves driving home in that new sportscar we want, there is a simple connection between the lack of reality of the thing we want and the lack of reality of the future. So there are at least two possible explanations that are based on our experience in the world (of lack of reality and of desire and time).

    One problem, in my view, with generativists, is that they generally don’t know much about typology (essentially meaning they lack knowledge of a wide variety of languages, including common patterns). Another is that they don’t look for common-sense solutions to linguistic puzzles based on our experience in the real-world, preferring instead to quickly jump to conclusions about ‘the deep structure of language’. Any good scientist follows the rule known as Occam’s Razor, which is one of the oldest and most fundamental scientific principals, and states that simplicity in explanation is preferable to complexity. To look for common-sense understanding of the human experience and use this understanding is far preferable to positing some underlying connection hard-wired into our brains before birth. Let’s save the hard-wired stuff for those parts of language that just don’t have a common-sense real world explanation, including psychological, neurological, anatomical, social, cultural, and historical explanations (if there are any linguistic phenomena that cannot be explained or motivated using one or more of these perspectives). Linguists’ job, as I see it, is to search for these simple explanations before postulating the existence of “deep structure”. Let’s exhaust the possibilities before doing something so drastically un-scientific as simply saying ‘there must be a link somewhere built into our DNA’ or ‘they must be fundamental to human thought’. Especially when we’re basing it on (flawed) data from just two languages.

    Finally, I wonder how you know that ‘most Chinese speakers don’t notice (the connection between the meanings ‘possess and ‘Past’ ). Did you take some survey or do an experiment? Please, let’s leave broad generalizations like that to people who have done their homework, so that they are meaningful. Even if you are a native speaker of Mandarin (which you may be, for all I know), it is scientifically irresponsible to state this as if you had good evidence for it (which you may have, for all I know, but there’s no mention of it in the post). Concern for the quality of data is another way that generative and functional linguistics differ; generativists almost exclusively rely on their own intuition, whereas (good) functionalists collect data that is more valid and reliable (two essential qualities for good data based on the commonly-accepted methodology of science). While it is true that some functionalists are content with intuition-derived data, the majority are unable to publish anything without more solid data (such as those derived from corpora, psychological experiments, or non-linguist native speakers’ discourse).

    Additionally, as a small and somewhat picky point, your explanation for this ‘fact’ fails to take into account that mainland Mandarin speakers still use the form ‘you mei you’ (‘have not-have’) to ask questions about past events, and that they answer in the negative by using ‘mei you’ (‘don’t have’). So a fairly high frequency of actual uses of the word ‘you’ (‘have’) are used with the sense of ‘past’, and it’s easy to think that mainland Mandarin speakers may still make the connection (though, as you said, it now sounds odd to hear someone use the positive ‘you’ to mean ‘past tense’ in a positive declarative sentence on the mainland).

  6. Victor, it sounds like you are agreeing with me, actually. I try not to identify too strongly with a particular “camp”, but in this specific case, I think the very simple explanation makes sense. Chinese and English split well before the Indo-European branch, so it is rational to assume that any commonality between them represent the oldest parts of human language that developed. I also think it is rational to assume that have/want were some of the very first useful concepts that humans codified into language, and so it is reasonable to expect some cross-language consistency. As I explained earlier, I think the reason have/want were used for past/future, is simply because the early humans were dealing with a very small set of symbols, and it was convenient to re-use.

    If you have followed this blog over time, you will see that one of the most repeated themes is that of language as man-made tool — I believe that humans created language to serve ourselves, and that too often people allow themselves to be enslaved by the tool of their own creation. So perhaps that is more “functionalist”. But by the same token, there is plenty of evidence that there is a codependency between language and physiology; so I think it is too simplistic to hew to any extreme.

    Now, to be honest, the discussion of inflected verb forms does not convince me. I think it’s clear that these forms developed relatively late in the evolution of language (as did gender). It may be true that many past/future concepts can be expressed today without have/want, but that just proves the point — the have/want constructs are very old and are slowly being replaced.

    Regarding “not seeing the connection”, it is true that I make my comments about native Chinese speakers purely based on speaking with many native speakers, and not based on a “scientific” survey. You are welcome to believe what you want, or hold off pending some survey at some unspecified point in the future, but I am relatively confident of this fact and continually do due dilligence. I am as suspicious of confirmation bias as any healthy scientist. Even when coached a bit, they fail to see it. Here is a test you can do: ask a native speaker to explain the difference between yao/want and yao/will, for example — pretend you are confused about the difference. Say, “Chinese seems too limited — there is no way to distinguish between ‘I want to go to Beijing’ and ‘I will go to Beijing’ — ‘wo *yao* qu Beijing’ and ‘wo yao qu Beijing'”. You would be surprised how many highly educated native speakers will tell you there is essentially no difference. If you press, depending on the person, they will either tell you that you need more context, or that you need to stress a sylable, or whatever. Then ask about putting it in writing — that sentence (wo yao qu Beijing) naked and alone is ambiguous. In any case, you can expect confusion and rationalization instead of the connection I made above. Even when I have made the exact connection for people, they have *always* reacted with surprise. And I’ve had at least three people exclaim that, “If I had recognized this, it would have helped me to speak better English”. Native Chinese speakers tend to screw up tenses in English in large part because they do not recognize the connection.

    Again, you are welcome to do your own experiements, but I am not exactly an ignorant orientalist basing my comments on a visit to a Chinese restaurant. If such experiements haven’t been done, that simply means that my blog post will be recognized as ground-breaking work when science verifies my theory. 🙂

    Also, regarding colloquial use of “you meiyou”, this is widespread in Taiwan, as well as areas like Fujian which also speak Taiwanese. It is also not uncommon in other southern provinces. This again tells me that it is a very old construct; since it is common across many dialects, but the “le” form shows more variance across dialects. As I said, the “you meiyou” form for past is considered rather queer in Beijing — it will get you marked as a priss or a bumpkin. Finally, I will point out that we do this in English, especially in interrogative scenarios — “have you or have you not stopped beating your wife?”. We still do that sometimes, yet oddly enough nobody makes the connection.

    As I said, I believe that the have/want linkage is of diminishing importance to today’s human condition, and therefore you can see it being weeded out from the language and being replaced by -le, -ed, and so on. It is kind of like the coccyx — a vestigial marker of past history. However, I think it is obvious that this is one of the oldest characteristics of human language and it *was* very important at one time. That is the common-sense explanation.

  7. Plus, I think this is important, since human language has evolved rather more quickly than physiology. Regardless of what you feel about deep structure, it is common-sense that having multiple ways to communicate the same concept can result in more cusomized communication. Hammarskjold would argue that the deepest human ethic is to “employ words with a deep heart-felt love of the truth”, and that to use words ambiguously is to “poison the wells”. Therefore, higher-precision language, such as inflected verb forms, multiple verb tenses, and so on — are very ethical. You could argue that early humans smushed together have and past simply because they did not have as precise tools as we have today (and I would agree with this).

    But there is a flip-side, which deal with communication that is pre-logical. You will notice that children respond more readily to vivid demonstrations than to clearly-formulated logic; in part because their minds have not been equipped with the tools of logic. But it goes beyond this: often it is the case that pre-logical communication mechanisms are the only way to get past a person’s gestalt filters and defensive mechanisms of neurotic rationalizations. Since logic is a tool employed by men, it is often the case that men use logic to create impenetrable defensives to rationalize impoverished decision behaviors. If you are in the business of helping people enrich their mental models, you need to be able to employ non logo-based techniques. There are plenty of ways to express concepts without words, and in fact using words that have ambiguity and are linked to deeper/older conceptual formations can be a very useful way to bypass elaborate logical defenses.

    As an example, you could imagine a scenario where you need to get someone to realize that a problem state is in the past, and get them to stop allowing it to interfere with their present. If you are familiar with these scenarios, you know that logic rarely works. One strategy might be to construct a story/metaphor which talks about desire turning to posession, and desire fading away — then posession fading (“what you had is not what you have with you right now; it is stored away in a safe place, you no longer have it with you”). I am not saying to make it abstract or vague (although that can work). Maybe you talk about some past event of shared relevance, and tell the story in a way that makes these connections. Or maybe you consturct a new fitional story. Or maybe you orchestrate events to cause the specific story to happen, as a vivid demonstration. But the point here is that you never talk about past and future explicitly, you never say “the past is over, you idiot, stop chosing to let past events ruin your present!”. You talk about something different, and allow the conecpt to be embedded without opposition. I am fairly confident that this linkage can work, because it is incredibly old and fits with the way our species evolved — and because there is enough ambiguity which still exists that you can count on some semantic bleed.

  8. If it sounds like I’m agreeing with you, then you didn’t understand my post.

    I’m saying that your assertion that there’s a link between ‘have’ and the concept PAST is incorrect, as analyzed by just about every linguist I know, and that your conclusions are jumps of faith rather than based on science.

    I have provided an alternate way to understand the link between the concepts DESIRE and FUTURE that *is* based on science, that doesn’t invoke any “black box” or “deep structure”.

    The terminology you’re using is that of the generativists, whether you realize it or not, and that leads to seeing the question and possible answers through that perspective. If you think that you are doing semantics without a theory, I’d tell you that such a thing is impossible, unless of course you want to create your own theory (in which case you need to be explicit about everything from why you’re dong it to defining the atomic elements of your theory to how it connects to other theories) or else it’s meaningless to the community of scholars. It isn’t possible to practice linguistic analysis without a theory, either explicit or implicit. By adopting generative terminology, you’re defining your understanding in their terms, which is what it means to be a generativist.

    As far as the quality of data, you may not be an “ignorant orientalist basing my comments on a visit to a Chinese restaurant”, but it is important to differentiate between your intuitive conjectures and solid data-backed theories and facts.

    And finally, your assertion that Chinese and English are historically related and diverged at some point does not fit within the realm of science – it’s pure speculation. There is not a single legitimate scientist in the world that would agree with you, as there’s no evidence for such a connection. In fact, there cannot be any evidence for such a connection. We’re lucky to be able to go back 4 or 5 thousand years with solid science-based historical linguistics, as we can do with Proto-Indo European, due to the diversity, completness, extent and depth of the evidence (Sanskrit, Ancient Greek, Latin, etc.), but anything beyond that is fantasy. I might recommend Lyle Campbell’s book ‘Historical Linguistics’ as a good introduction to the science of historical linguistics. As for the untenable assumptions of generativists, I recommend starting with George Lakoff’s “Women, Fire, and Dangerous Things”, though it really only covers the basics.

    I don’t have a problem with what you’re doing, as long as it is placed firmly within the realm of philosophy. Once you start to claim that you’re doing science, I feel compelled to step in and explain the differences between your musings and real science. Linguistics can, and should, be a science, and the irresponsible practice of linguistics is something that gets my goat.

  9. OK, now you are just trolling. I realize that “deep structure” is a very “generativist” term. But the explanation I offered is in no way “generativist”, and as I said I think it is silly to hew religiously to either label. It’s as if you have locked yourself in the ghetto of your own terminology any aren’t capable of communicating with another individual who hasn’t done the same.

    Regarding, “not a single legitimate scientist”, I’m stunned that you would say that. Either the languages all “sprung up” independently in various population groups when the earth’s population was much, much smaller. Or the rudiments of language developed early, and language has morphed over time. We have tons of evidence of language changing locally and evolving from a single source (language mutates and evolves far more rapidly than our phenotypes do). So it is utterly insane to posit that there is no connection between Chinese and English. Especially when we have such a deep and fundamental similarity as I point out here.

    Seriously, for you to posit that the two are not linked historically (at least from the time that men had words to talk about past and future), is not at all credible.

    And it really doesn’t matter how you bluster and appeal to authority; it’s ridiculous. I suppose you are one who would not believe the merits of biological evolutionary theory until carbon daing and DNA analysis “proved” it. Ignore the fact that it’s common sense and fits occam’s razor; you seem to be arguing that you won’t accept any theory until you have hard evidence via whichever measuring tool happens to be popular with today’s scientists.

    So far you have offered nothing to rebut my comments except some obviously incorrect comments about regional Chinese speech patterns, a frankly irrational discussion of “-ed” endings, and now strident repetition that “it’s not science!” and “languages do not evolve into different langauges over time”.

    And arguing about something that happened 10,000 years ago by saying “prove it” is just trolling. You can’t even prove that gravity won’t reverse itself tomorrow. You seem to be locked into the cult of reason, and have confused science with blind faith. Science is about questioning theories, science never claims to have absolute truth. If you are so interested in absolute truth and appeals to authority, you should join a church. Every scientist in the world would agree with me (and if they don’t, they are apostate)

Leave a Reply

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