Skip to content
 

Do regression structures affect research capital? The case of pronoun drop. (also an opportunity to quote Bertrand Russell: This is one of those views which are so absurd that only very learned men could possibly adopt them.)

A linguist pointed me with incredulity to this article by Horst Feldmann, “Do Linguistic Structures Affect Human Capital? The Case of Pronoun Drop,” which begins:

This paper empirically studies the human capital effects of grammatical rules that permit speakers to drop a personal pronoun when used as a subject of a sentence. By de‐emphasizing the significance of the individual, such languages may perpetuate ancient values and norms that give primacy to the collective, inducing governments and families to invest relatively little in education because education usually increases the individual’s independence from both the state and the family and may thus reduce the individual’s commitment to these institutions. Carrying out both an individual‐level and a country‐level analysis, the paper indeed finds negative effects of pronoun‐drop languages. The individual‐level analysis uses data on 114,894 individuals from 75 countries over 1999‐2014. It establishes that speakers of such languages have a lower probability of having completed secondary or tertiary education, compared with speakers of languages that do not allow pronoun drop. The country‐level analysis uses data from 101 countries over 1972‐2012. Consistent with the individual‐level analysis, it finds that countries where the dominant languages permit pronoun drop have lower secondary school enrollment rates. In both cases, the magnitude of the effect is substantial, particularly among females.

Another linguist saw this paper and asked if it was a prank.

I don’t think it’s a prank. I think it’s serious.

It would be easy, and indeed reasonable, to just laugh at this one and move on, to file it along other cross-country comparisons such as this—but I thought it could be instructive instead to take the paper seriously and see what went wrong.

I’m hoping these steps can be useful to students when trying to understand published research. Or, for that matter, when trying to understand their own regression.

So how can we figure out what’s really going on in this article?

To start with, the claimed effect is within-person (speaking a certain type of language affects your behavior) and within-country (speaking a certain type of language affects national values and norms), but all the data are observational and all the comparisons are between people and between countries. Thus, any causal interpretations are tenuous at best.

So we can start by rewriting the above abstract in descriptive terms. I’ll just repeat the empirical parts, and for convenience I’ll put my changes in bold

This paper empirically studies the correlation of human capital with grammatical rules that permit speakers to drop a personal pronoun when used as a subject of a sentence. . . Carrying out both an individual‐level and a country‐level analysis, the paper indeed finds negative correlations of pronoun‐drop languages with outcomes of interest after adjusting for various demographic variables. . . . speakers of such languages have a lower probability of having completed secondary or tertiary education, compared with speakers of languages that do not allow pronoun drop. The country‐level analysis uses data from 101 countries over 1972‐2012. Consistent with the individual‐level analysis, it finds that countries where the dominant languages permit pronoun drop have lower secondary school enrollment rates. In both cases, the magnitude of the correlation is substantial, particularly among females.

OK, that helps a little.

Now we have to dig in a bit more. First, what’s a pronoun-drop language? Or, more to the point, which languages have pronoun drop and which don’t? I looked through the paper for a list of these languages ora map of where they are spoken. I didn’t see such a list or map, so I went to wikipedia and found this:

Among major languages, two of which might be called a pro-drop language are Japanese and Korean (featuring pronoun deletion not only for subjects, but for practically all grammatical contexts). Chinese, Slavic languages, and American Sign Language also exhibit frequent pro-drop features. In contrast, non-pro-drop is an areal feature of many northern European languages (see Standard Average European), including French, (standard) German, and English. . . . Most Romance languages (with the notable exception of French) are often categorised as pro-drop too, most of them only in the case of subject pronouns . . . Among the Indo-European and Dravidian languages of India, pro-drop is the general rule . . . Outside of northern Europe, most Niger–Congo languages, Khoisan languages of Southern Africa and Austronesian languages of the Western Pacific, pro-drop is the usual pattern in almost all linguistic regions of the world. . . . In many non-pro-drop Niger–Congo or Austronesian languages, like Igbo, Samoan and Fijian, however, subject pronouns do not occur in the same position as a nominal subject and are obligatory, even when the latter is present. . . .

Hmmmm, now things don’t seem so clear. Much will depend on how the languages are categorized.

The next thing we need, after we have a handle on the data, is a scatterplot. Actually a bunch of scatterplots. A scatterplot for each within-country analysis and a scatterplot for the between-country analysis. Outcome of interest on y-axis, predictor of interest on x-axis. OK, the within-country data will have to be plotted in a different way because the predictor and outcome are discrete, but something can be done there.

The point is, we need to see what’s going on. In the within-country analysis, where do we see this correlation and where do we not see it? In the between-country analysis, what countries are driving the correlation?

Again, the analysis is all descriptive, and that’s fine, but the point is we need to understand what we’re describing.

I have no idea if the causal claims in this paper are true—given what I’ve seen so far, I see no particular reason to believe the claims. But, in any case, if these patterns are interesting—and I have no idea on that either—then they’re worth understanding. The regression won’t give us understanding; it just chews up the data and gives meaningless claims such as “we find that the magnitude of the effect is substantial and slightly larger for women. Specifically, women who speak a pronoun drop language are 9‐11 percentage points less likely to have completed secondary or tertiary education than women who speak a non‐pronoun drop language. For men, the probability is 8‐10 percentage points.” That way lies madness. We—Science—can do better.

P.S. I scrolled down to the end of the paper and found this sentence which begins the final footnote:

Pronoun drop rules are not perfect measures of ancient collectivism.

Ya think? In all seriousness, who could think that pronoun drop rules are any sort of measure of “ancient collectivism” at all? As Bertrand Russell said, this is one of those views which are so absurd that only very learned men could possibly adopt them.

69 Comments

  1. David says:

    This paper aside, I don’t think it’s so preposterous that the way we talk about something influences how we think about it. That’s not easy to analyze and disentangle from other causal mechanisms (in this case, I’d bet on causality going in the other direction), but I don’t think we shouldn’t study it just because it’s hard. Consider this paper in the American Economic Review on the effect of language on savings and health behavior: https://www.aeaweb.org/articles?id=10.1257/aer.103.2.690

    “I test the hypothesis that the 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.”

    Which is to say: one can imagine many possible alternative theories, but they can also be tested and perhaps ruled out. From what I recall, the author looks at multilingual countries (e.g. French and German speakers in Switzerland, French and English speakers in Canada) and children of bilingual parents. Once you’ve ruled out the most likely alternative explanations, the effect of language becomes the leading contender.

    We wouldn’t expect a single paper to provide definite proof, but a series of good papers showing how language correlates with behavior should make us take it seriously. Otherwise, we have no hope of ever learning anything in the social sciences…

    • Andrew says:

      David:

      Just to be clear: I agree that there could be something to the theory in that paper, even if the data shown do not make a clear case. The sentence that I thought was completely laughable was this: “Pronoun drop rules are not perfect measures of ancient collectivism.” Maybe in my next article I’ll let loose with a sentence such as, “My cat is not a perfect surrogate for Jesus.” Or, perhaps, “Cornell students are not perfect at ESP.”

    • Joe says:

      I’d be careful about citing that particular paper. There are numerous problems with it from a linguistic viewpoint, ranging from the dataset he used to the variables that he tried to control for (he’s not a linguist, and it really shows in that paper). Ars Technica had a pretty nice discussion on the issue:

      https://arstechnica.com/science/2015/06/louder-vowels-wont-get-you-laid-and-other-tales-of-spurious-correlation-2/

      I think the problem with statements like “how we talk about something influences how we think about it” is that they are too ill-formed to do research on, since there are lots of different ways to talk about something, and lots of different ways of thinking about it. I don’t see any reason at all to think that speakers of languages where there isn’t a future tense verb form show more forward-thinking behaviour because the lack of a grammatical future tense mentally connects the present and future (that was Chen’s explanation).

      English is supposedly one of those languages that have a future tense on the verb form (I don’t agree with that analysis, but let’s grant it for now). Think of all the ways we could talk about the (verb forms are placed within *): Tomorrow I *leave* for the airport, I*’m leaving* next week, he’*ll get* tenure within the next few years”, etc. Would you really say that what allows speakers to distinguish the future from the present is the verb form?

      (Then there’s the problem that most of the people in the world speak more than one language).

      I think that a point that has often been made on this blog is that observational studies that aren’t based on sound theory are really in danger of just chasing noise, since you can always invent a theory that explains the data you got. And the theory should actually be the one that is the subject of investigation and not the “headline” account (as I said, language affect thought isn’t a real research question, it’s an implication). At the very least, I think one has to show that verb tense is the predominant marker of future time before you could say anything else about the influence of verb tense on behaviour, and I think that’s already a pretty highly implausible claim).

    • David J. Littleboy says:

      “This paper aside, I don’t think it’s so preposterous that the way we talk about something influences how we think about it.”

      I think that this is right, but only at inanely low levels. For example, in English “brotherhood” and “sisterhood” are (metaphors for) _EQUALITY_ relations, and in Japanese they are extremely strong _INEQUALITY_ relations. Who grew up being the older, and who grew up being the younger sister/brother is the first thing the Japanese want to know. (The actual interactions, relationships, conflicts involved aren’t particularly different, it’s just that we sweep them under the rug and the Japanese beat you over the head with them.) There are lots of other examples, the other usual metaphors are all different, greetings are different, etc.

      The problem with this low-level lexical approach, though, is that it’s hard to say anything that’s sensible and generic about things. The Sapir-Whorf hypothesis is widely (and correctly) derided. Ditto for anything at any higher level. It would be nuts to, for example, argue that the necessity of specifying the syntactic singular vs. plural in English (as opposed to just not bothering with it at all (Japanese and Chinese)) makes us bad at thinking sensibly about group theory. (Hint: the singular vs. plural distinction is a horrific bear in English, and it’s actually seriously illogical and innumerate. For native speakers, it seems natural and logical and obvious. It’s not.) Sure, language and culture are closely intertwined, but as soon as you know enough about some second language (I spent 25 years translating Japanese to English) you find that (a) it just works, and (b) any time someone makes a generic claim about that language you either laugh, cry, go ballistic, or all three. Usually all three.

      • Terry says:

        “the singular vs. plural distinction is a horrific bear in English, and it’s actually seriously illogical and innumerate. For native speakers, it seems natural and logical and obvious. It’s not.”

        Say what?

        You have to provide more explanation or a good link for this. I’m not saying I don’t believe you, you know much more about this than I do. As a native English speaker, though, this just never occurred to me. It’s like you are a creature from another dimension.

        • David J. Littleboy says:

          “It’s like you are a creature from another dimension.”

          Well, I have degrees in both humanities and engineering (making me somewhat other dimensional), but I’m a native English speaker from Boston (MA), mom was an English major, Radcliffe ’38!.

          And before I learned Japanese, I though singular/plural made sense, too. It turns out it makes no more sense than masculine/feminine in German and French. Really. It’s a syntactic distinction that may have mapped the real world in original principle (maybe), but doesn’t work when you try to come up with consistent logical rules to describe it.

          Sorry, I should have references for this, since it’s a favorite pet peeve. Presumably the ESL folks would have something to say on this…

          • Joe says:

            I’d be curious which examples you were thinking if. I mean, there is something arbitrary (but probably not capricious) about the fact that “luggage” is singular-only and “groceries” is plural-only, but what makes contrast between singular and plural in “innumerate?”

            • Terry says:

              I was wondering this too. I can see that sometimes the plural is vestigial (“groceries”), but other times it is meaningful (“I crashed the car” versus “I crashed the cars”).

            • David J. Littleboy says:

              The special cases (water vs. waters) are, of course, perverse, but all natural languages have perverse special cases.

              The most blatant craziness, off hand, is that zero is plural. Now, zero is only a recent invention in mathematics, but it ain’t logically plural. It’s part of the “not singular” space, so has to be plural. But it ain’t pretty.

              Plurals are also complicated. Japanese doesn’t have syntactic number, but it’s got lots of ways of talking about counting things. And they’re way more useful than English plural. Take “ra” for example. The Japanese-English dictionaries will tell you that it’s a pluralizing suffix, but that’s not the whole story. It means “the noun marked and the other things in that group”. Huh? you are thinking. Tsuma means wife, but tsumara doesn’t mean “wives” it means “the particular wife under discussion and the people associated with her in that discussion”. So in a newspaper article about the victims in the OJ Simpson case, “tsumara” means “His wife and the other victim”. But machine translation programs will tell you that OJ is a bigamist…

              So baking plurals into the syntax is way more hairy than a monolingual native English speaker could ever imagine. And if these statistic infatuated psych types tried, they could easily prove that Japanese 7th graders are way better at group theory than US 7th graders.

        • Martha (Smith) says:

          David said,
          “For example, in English “brotherhood” and “sisterhood” are (metaphors for) _EQUALITY_ relations”

          I don’t think of them as (metaphors for) equality relations — but more as expressions that we should support each other and not deride or exclude others in the same group (men, women, people).

      • Terry says:

        I agree with your take on the idea that language influences thought significantly.

        At first, it seems cool and entirely possible that this would be true. But the more you think about it (and the more you read critiques of the examples of it), the less compelling it gets.

        Take the notion that language X does not have a word for Y. At first, you go, “cool man”. Then you think: Ok they don’t have A SINGLE word for it, but can’t they say the same thing using two, or three, or five words? Inuit’s have 50 words for snow, but can’t we be just as descriptive by adding adjectives to say things like “fluffy snow”, “icy snow”, “drifted snow”, “yellow snow”, or “that snow over that looks a little pink in the sunset”? Sure, Jamaicans don’t have a traditional word for snow, but if it started snowing regularly in Jamaica, they would very quickly come up with a lot of words for snow. Its not like Jamaicans are permanently unable to think about snow because they didn’t have a word for it once upon a time.

        • David J. Littleboy says:

          “Its not like Jamaicans are permanently unable to think about snow because they didn’t have a word for it once upon a time.”

          Sheesh. I wish more people were able to think this logically. Life would be much easier. This is exactly right.

          These sorts of theories are so insulting to (and oblivious to the amazingness of) human intelligence. If you learn a language and use it every day, you find that it just works. Without problem, for anything you need to do. Languages is really kewl like that.

          Anyway, borrowing vocabulary items is one of the most common linguistic phenomenon. Pretty much everyone does it, even the French. English didn’t have a word for schadenfreude, but now it does. If you do language family analysis by vocabulary similarity, Japanese would be a close dialect of Chinese. It’s not.

  2. Jon Baron says:

    Another thing going on here might be this. The country analysis is based on 75 countries. In a regression model, this would have 75 df minus the number of predictors. But the “true” df should have more to do with language families, like “romance” “northern European”, “African” … There may be really only 5 families here.

    It would be a mistake to test across 114,894 individuals without considering the countries they are from. And it would be the same sort of mistake to consider countries without considering the language groups they are in. The trouble is that we have no neat definition of language groups.

    I think that this kind of problem is present in a lot of research that deals with “country” as the unit of analysis.

    • Martha (Smith) says:

      I (possibly channelling Andrew?) suggest that the influence of language groups might be taken into account using multilevel models.

      • Jon Baron says:

        This assumes that you know what the groups are.

        Presumably, by looking at individuals, you could include “language group” in the model without assuming that countries are nested within language groups. (Probably they aren’t. Many countries have 2 or more commonly spoken languages.)

        But you still cannot deal with the similarities of the language groups. Two or three groups may behave as if they were really one group, because of (say) common historical antecedents.

        This situation arises with other studies of countries (and many other things). “Poor African countries” may not behave as a group but rather as points in a multi-dimensional space that also includes other countries. And you usually have no way to define such a space.

        I’m not saying that we can’t learn anything from these studies. But I’m not sure how they are all that much better than “anecdotal” comparisons, e.g., “the 2 or 3 European countries that have working presumed-consent laws for organ donation seem to get more donors than those that do not”.

    • Terry says:

      An important point.

      Investment in education is highly correlated with economic development, and economic development has definite geographic patterns. So there are going to be a zillion variables all correlated with each other and economic development (religion, hair color, shoe style, favorite spices, type of music, donkeys per capita, alcohol consumption, etc.)

      It is hard to believe that pronouns are a uniquely powerful driver of educational investment when there are so many other factors at work.

    • David J. Littleboy says:

      “I think that this kind of problem is present in a lot of research that deals with “country” as the unit of analysis.”

      And then there’s India, with 22 recognized languages from at least 5 families.

      By the way, the definition of “language” informally accepted by linguists is “a dialect with an army”.

  3. Martha (Smith) says:

    “Chinese, Slavic languages, and American Sign Language also exhibit frequent pro-drop features”

    I don’t claim any expertise in ASL, but my understanding is that pronouns in ASL are often dealt with by contextual “position assignments” (eg., Andrew is referred to by pointing in a specific direction, Bob in another specific direction, and Corey in a third )– which has the advantage over spoken English of not having to make the choice between being unclear (e.g., “He gave him his address,” with each pronoun referring to a different individual) and being lengthy (spelling out each name manually).

  4. Terry says:

    such languages may perpetuate ancient values and norms that give primacy to the collective, inducing governments and families to invest relatively little in education because education usually increases the individual’s independence from both the state and the family and may thus reduce the individual’s commitment to these institutions.

    I always find passages like this unsettling. The passage gives the eerie feeling that languages are semi-sentient disembodied entities that roam the countryside malevolently (they always seem to act malevolently) oppressing the innocent and powerless. In short, they seem to have agency. Languages “perpetuate ancient values” that “give primacy to the collective” etc. On the other hand, actual, sentient people with agency (and their governments) are “induced” to “invest relatively little” in education, which reduces individual independence etc.

    Sure, this can be interpreted as a causative story where languages don’t actually have agency, but the verbiage encourages the reader to picture them as being malevolent anthropomorphic entities with agency.

    The problem is that this encourages lazy thinking. It encourages you to ignore the real causal links in favor of a morality play about evil languages. Thinking in terms of morality plays is easy. Understanding actual causal links is a lot harder.

    • Martha (Smith) says:

      Terry said, “I always find passages like this unsettling. The passage gives the eerie feeling that languages are semi-sentient disembodied entities that roam the countryside malevolently (they always seem to act malevolently) oppressing the innocent and powerless. In short, they seem to have agency.”

      Wow, this seems like reading a lot into the passages.

      The way I see it, languages often have a tendency to influence people’s worldview, simply because people so often tend to think of things in terms of their primary language — for example, if two ideas are somewhat similar and are labeled by the same word in language A, but are denoted by very different words in language B, someone who only knows language A is more likely to see them as “essentially the same” than someone who only knows language B.

      This is why I think that people who speak more than one language tend to see more distinctions than people who speak only one language.

      • Terry says:

        To be fair to myself, I wasn’t saying this is what the author was actually saying. I was just saying this is the feeling I get from reading these passages. It is an easy way of thinking to fall into and therefore must be guarded against.

        Humans do this quite a bit. Forces of nature are often transformed into spirits with agency and good or bad intentions. It is particularly easy to fall into this trap when you don’t really understand how something actually works.

      • The subject of my very first university essay https://en.wikipedia.org/wiki/Linguistic_relativity

        Not sure if things have changed (a quick look at RJB’s link below suggests it hasn’t), but it seems that the effects are slight and largely fleeting. Some of the experiments I read that offer support upon careful reading of results actually brought it into question. The analysis of research and peer review of it was really poor back then. Maybe that hasn’t changed ;-)

      • Terry says:

        On a related note, some people think a composer’s language affects the type of music they write. I find this believable. French composers sound different than German composers. Chinese music is definitely very different.

        • There’s no way to deconvolve the role of culture and language here. Do native chinese speakers who grow up in Germany compose music that sounds like what?? it’s not even possible to imagine teasing this apart, as the treatment of enforcing native chinese speaking while growing up in Germany automatically makes you culturally non-German…

          • Terry says:

            The reason I find it believable is that language is so closely associated with music.

            Start with opera. Since opera is sung in a language, it is almost tautological that French, German, Italian, and Chinese language operas will sound different.

            The question that is not obvious is whether this difference carries over into instrumental music. I also find that believable, but much less certain.

            A good musicologist might shed light on this. Musical analysis can be quite technical.

            • I have no problem with *song*, it’s obviously constrained by the phonemes of the language. It’s far from clear though what to make of instrumental music.

              • Martha (Smith) says:

                +1

              • David J. Littleboy says:

                Instrumental music is sometimes a specific cultural thing (Chinese Opera, Japanese “classical” music (koto, shakuachi, shamisen), Chicago blues) and radically unique in the extreme, and sometimes a shared intellectual endeavor (CPP Clasical, 12-tone, Bebop (although Bebop is essentially CPP harmony in a different cultural context).

                Among the various European nationalities in the common practice period, though, many of the composers moved around and composed for the audience (and orchestras and singers and language) at hand. The rules of the harmonic games being played were the same. The only thing new in town was when, towards the end of the period, the French discovered the minor seventh flat five chord, which the beboppers took over with great enthusiasm. (My understanding is that the minor seventh flat five chord, which is a perfectly obvious chord within classical harmony, was ignored until Debussy.) It’s a core tool in jazz.

              • Martha (Smith) says:

                Pardon my ignorance, but what’s CPP?

              • David J. Littleboy says:

                Replying to Martha (Smith) below:

                CPP is common practice period. (Sorry, I should have defined that.)

              • Martha (Smith) says:

                Thanks

          • Terry says:

            Very relevant is the “speech to song illusion.” If you repeat a phrase enough times, it becomes music.

            Try these examples: http://deutsch.ucsd.edu/psychology/pages.php?i=212

            Quite eerie.

            So if phrases sound different in different languages, perhaps their music will sound different as a result.

            • Martha (Smith) says:

              Interesting link. Thanks.

              Also, the topic and link bring up the question of tonal languages (such as Chinese), and remind me of the following incident:

              One day a number of years ago I was sitting in my office with the door open and gradually became aware of two people ( a man and a woman) in the hall speaking what I guessed was a Chinese language. I don’t speak or understand any Chinese language, so I wondered what about the conversation had distracted me. I then realized that the man’s speech was not as strongly tonal as the woman’s. The thought came to me that the man might be the graduate student whose parents had emigrated to Canada from China and had grown up bilingual. I peaked out of my door and, indeed, he was the man speaking. I was facing the woman, so asked her, “Does he speak Chinese with an accent?” and she replied, “Yes”. He seemed surprised, so I explained how I had guessed.

  5. Tom says:

    On the statistical problems associated with this kind of exercise, you might be interested in this. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3321347

  6. RJB says:

    Language log has a great related archive: No word for X [in Language Y]

  7. Steven Edward Armstrong says:

    Having trouble understanding what a pronoun-drop language is. English definitely isn’t. What do you think?

    • Bill Spight says:

      Well, your first sentence has a pronoun drop. And it’s English.

      Spanish: Hoy soy muy triste. (I’m very sad today.) No pronoun, yo (I). But the verb, soy, implies yo. So why say that the yo is dropped?

      Japanese: Tsukareta. (I’m tired.) Tsukareta? (Are you tired?) The subject is inferred. Good case here for pronoun drop.

      • Steve says:

        Yes, it was a joke. I dropped three pronoun. English is not suppose to be a pronoun drop language according to linguists of this school, but every language can drop pronouns when the context makes it clear. So, unclear where the line between one no drop and pronoun drop languages is.

        • Martha (Smith) says:

          Nicely done.

        • Joe says:

          In fairness, just about any linguist talking about English would note a number of different exceptions that occur in both spoken and written (e.g., diary) language.

          Your last example (unclear where . . . ), however, might show a different analysis, since you’re dropping both the pronoun and the verb (although that sounds to me more like Yoda speech than anything). One analysis would be that English has a drop weak initial syllable rule, so it’s a phonological rule rather than a syntactic one. Think about the following: “This book any good?”, “Woman over there wants to talk to you”, “You come here often?” What gets dropped in this context are just not pronouns but auxiliary verb (is), determinative (the), and then another auxiliary (do). In each case, they are phonologically weak. So the situation In English differs from that of syntactic pro-drop languages, even though they superficially look the same.

          • Steve says:

            All the linguists will now note the exceptions. The problem was the generative grammar folks set out to find the fundamental structure of language not just group languages based on similarities. Look at the goal of this paper, it was suppose to show that the structure on a language can lead one to think and act in a certain way — ancient collectivism. But, if people are capable of changing the “basic” structure based on context, why should we think of that structure as basic. Maybe, the arrow of causation is the other way, and we don’t drop the pronouns because we are individualistic, but we will start dropping them once AOC becomes president (a joke, not an attack on AOC). Or maybe, there is no relationship. Do the counterexamples I gave mean the generative linguists are wrong. Probably not. But, that is because their claims are not well-defined enough to tell whether the theory is true of false.

            • Joe says:

              I’m not in any shape or form a generative linguist, but I don’t think argumentation by parody is a valid approach.

              All I said is that every linguist who has ever looked at pro-drop languages in any kind of detail will note examples like you’ve shown and attempted to deal with them. And as I said, I personally think the examples in English you’ve given are part of a more general phonological phenomenon, so they are not examples of pro-drop, which is syntactic. That might explain why “Should go the gym tomorrow” is grammatical (although restricted in where it occurs) but “? Should go to the gym tomorrow?” is dubious.

              Somewhat simplifying, the problem the “generative linguistics people” set out to answer is how human beings acquire language. The larger debate between generative linguists and others is whether general human cognitive abilities are sufficient, or whether there is a specific language acquisition “device” that humans are genetically endowed with. Generative linguists want to figure out the nature of that device, which must be pretty abstract given the diversity of languages in the world. I doubt whether anyone can answer that question definitively one way or other given current state of knowledge, so I get tired of these arguments really quickly. But I see the question itself as an entirely valid one.

              The article cited above was not written by linguists but by economists, using a dataset that was not designed for the purpose which they are putting it to. Personally, I don’t understand why economists keep doing “research” on questions (or at least data) pretty much outside their field. I guess they think of it as just applied math.

          • Anonymous says:

            In common speech, people do all sorts of things are aren’t “right”. We do it for convenience, or out of laziness.
            But we know what the right way to do it is. Just because rules are often broken doesn’t mean there is no rule.

  8. Jonathan says:

    When my oldest was learning Chinese, she’d complain it was so contextual that if you didn’t know what was being said, you often didn’t know what was being said. Under what model does that become collective behavior? It seems more a localization of language to a shared personal space, like the way sign language works in person (disclosure: my wife taught the deaf to speak). There may be cultural reasons for this or it may be an accident of history (which looks like a cultural reason, if there’s a difference).

    If indeed Japanese or Chinese fit this structure, then how does one argue they invest less in human capital? One of the remarkable features of modern Asia is its investment in human capital, particularly in education. When my oldest spent most of 10th grade in Xian, the extent of investment in people by people and the government was striking. Everyone was learning English. The government worked to identify talented students and provided them with extra instruction levels (all the way up to ‘encouraging’ the family to send the kid to a special center). One can explain that as a cultural legacy: the Empire allowed for a rare, talented child from a village to rise through effort and state-sponsored exams to very high levels. Is it language too? My Chinese friends argue it’s the calligraphy and the care, detail and nuance of expressing the imagery within the ‘words’ which are actually conceptual renderings. This is similar, interestingly, to my Russian friends view of why Russian expresses nuance and how that fits more closely to French thinking (reflexivity versus an ideal), as compared to English/Germanic method of mental organization. If you haven’t heard Russians argue that this is why they’re so good at math and so terrible at other things, you haven’t an argument about why Russians can’t even say yes without an inflection of no because every yes is qualified by some degree of no.

    • When my oldest was learning Chinese, she’d complain it was so contextual that if you didn’t know what was being said, you often didn’t know what was being said.

      English is the same. Without context and world knowledge, we’re lost in even simple stories.

      Under what model does that become collective behavior?

      Human cognition and the need to communicate over a thin channel. It forces us to balance brevity with clarity. That means leaning heavily on shared understanding and on context. In any language.

  9. Bill Harris says:

    Is it time to pull out the /Tractatus Logico-Philosophicus/?

    • You’ll get more sense on this topic from the later Wittgenstein in Philosophical Investigations, which contains an extensive discussion on meaning and definition.

      You’d be even better off with the collected works of Susan Gelman. She’s published pretty extensively around the cognitive side of how children learn concepts. I think the right angle for all this is cognitive more than philosophical or anthropological, though all these fields overlap at this juncture, which is partly what makes it so much fun.

      I miss teaching philosophy of language and hanging out with psycholinguists.

  10. Terry says:

    Meta alert!

    Doesn’t this phenomenon have a name? The Sapir-Whorf hypothesis? Does it mean something that no one has used that term here? Would it change our ability to talk about this if we called it by it’s name?

    • Steve says:

      “Does it mean something that no one has used that term here?”

      It means that the people who think it’s not such a bad idea have never read any of the (many) criticisms of it.

  11. Terry says:

    I understand that Japanese is different at a very fundamental level, and pronoun drop is a pretty minor part of that difference (something having to do with subject-verb-object order?). So I don’t see why pronoun drop would be a uniquely powerful effect of speaking Japanese.

    • David J. Littleboy says:

      Wikipedia, in its article on pronoun dropping languages, give Japanese as the first example…

      In Japanese, anything and everything may be dropped, and often is.

      • Terry says:

        Doesn’t the passive voice in English accomplish the same thing as pronoun drop, i.e., not specifying the subject? Sometimes you don’t care who/what the subject is so you just drop it. Other times, you want to intentionally elide who/what the subject is.

        • David J. Littleboy says:

          I was going to say “Yes and no.”, but on second thought, no. Passive is different.

          Dropping things in Japanese is more about not needing to say them than about not wanting, or not being able to say. (It’s a Gricean sort of thing.) Japanese has a passive voice for pulling the target of the action into the subject position and announcing that you want to not make a big deal of who did it.

          Kare ha korosareta da yo! (He was killed, I tell you!) (Passive: ha is a subject marker)

          Would be handy when trying to persuade the police that there actually was a crime committed.

          But the pronoun drop version would probably be taken by the police to be saying that you yourself killed him. Oops. (It strongly implies that the subject has already been introduced into the conversation, since it wouldn’t be used if the subject hadn’t been. And you are the only subject around.)

          Kare wo koroshita da yo! ( killed him, I tell you!) (Pronoun drop. Active: wo is a direct object marker, subject dropped.)

          Would be natural in a discussion about the accused. Said accused is the subject of the discussion and there’s no need to mention him/her again. English requires a pronoun.

          So the English and Japanese passives seem very similar, and pronoun drop in Japanese would rarely, if ever, correspond to a passive in English.

          • jflanag2011@gmail.com says:

            I haven’t studied Japanese, but I kinda recall that the prototypical subject in Japanese corresponds to the topic of the discourse (rather than in English, where it prototypically refers to the agent of an action). Is that correct? (And again,I’m just talking about prototypical uses: you could make an argument that passive clauses in English put the topic in the subject position even though it isn’t an agent, and there are numerous examples where the subject in English indicates neither an agent nor a topic.

            • David J. Littleboy says:

              Well, sort of. It’s complicated, but the basic idea is that the topic is marked by wa and the subject by ga. That is, there are two subjects. There was a famous popular book on this (that I haven’t read) called “Zo wa hana ga nagai”, (Elephant THEME Nose SUBJECT Long), or The Elephant has a Long Nose. A painful literal translation would be “We’re discussing the Elephant, and the nose on said beasts is long.”

              (Why is elephant singular and beasts plural? The Japanese gets the set and part-to-whole relationships right without have to play any plural games. It.Just.Works.)

              Hmm. Subject as the agent. I’d think subject would be a much wider thing. You’ll need to argue with a real linguist on that one…

              (Adjectives function as nearly full-fledged verbs in Japanese and conjugate for tense and adverbial usage. But I digress.)

              Speaking of digressing, I was thinking of getting back to the topic of this thread and ranting about how friggin insane the proposed phenomenon in the paper here is, but I see that Andrew understands that quite nicely, thank you.

        • Joe says:

          Hi,

          The passive is probably the most mischaracterised linguistic phenomenon in English. Except in very specified conditions, passive sentences in English do have a subject. If I say “Your letter was received on the 5 March”, “your letter” is the subject, in the case with “Mistakes were made, ” “mistakes” is the subject.

          A very good description of the passive in English (written for a lay audience) is found here:

          http://languagelog.ldc.upenn.edu/nll/?p=2922

Leave a Reply