Sapir's armchair

« previous post | next post »

Yesterday we discussed this puzzling passage from Ange Mlinko's 9/7/2010 review in The Nation of Guy Deutscher's Through the Language Glass:

Edward Sapir, Whorf's teacher, was an armchair linguist influenced by Bertrand Russell and Ludvig [sic] Wittgenstein's work on the limits of language.

Where in the world, I wondered, did Ms. Mlinko get the bizarre idea that Edward Sapir was an "armchair linguist"? Well, now we know.

Guy Deutscher writes:

I suspect the reviewer's confusion came from reading, a little too quickly perhaps, the following paragraph in my book:

"Edward Sapir (1884-1939), like Humboldt a century before him, started his linguistic career far from the open vistas of American languages. His studies at Columbia concentrated on Germanic philology, and consisted of things rather reminiscent of the pedantic collections of obscure verbal forms in ancient tongues which he derided in the passage I quoted earlier. Sapir credited his conversion from the dusty armchair of Germanic philology to the great outdoors of Indian languages to the influence of Franz Boas, the charismatic Professor of Anthropology at Columbia…"

And given this lead, I was also able to track down the source of the almost equally puzzling idea that Sapir was influenced by Bertrand Russell and Ludvig [sic] Wittgenstein.  A quick search in Deutscher's book turns up this passage (p. 138-9):

In addition to the exhilaration of discovering weird and exotic grammars, there was something else in the air that pushed Sapir toward the formulation of his linguistic relatively principle. This was the radical trend in the philosophy of the early twentieth century. At the time, philosophers such as Bertrand Russell and Ludwig Wittgenstein were busy decrying the pernicious influences of language on the metaphysics of the past. Russell wrote in 1924: "Language misleads us both by its vocabulary and by its syntax. We must be on our guard in both respects if our logic is not to lead to a false metaphysic."

Sapir translated the claims about language's influence on philosophical ideas into an argument about the influence of the mother tongue on everyday thoughts and perceptions. He started talking about the "tyrannical hold that linguistic form has upon our orientation in the world," and as opposed to anyone before him, he went on to inject such slogans with actual content.

This "something … in the air" way of putting it is reasonable enough. There is indeed a zeitgeist-y similarity between aspects of the "linguistic turn" in philosophy and some of Sapir's ideas about language and thought. Deutscher references the Russell quote to his 1924 essay "Logical Atomism", and observes (p. 267, endnote to p. 139) that

Sapir was introduced to such ideas by the book The Meaning of Meaning: A Study in the Influence of Language upon Thought, by Ogden and Richards (1923).

And here we can finally find some evidence of more than a zeitgeist-y intersection of auras. Sapir's Selected Works, which (as I noted) completely lacks any relevant references to Wittgenstein or Russell,  does yield a reference to Ogden and Richards (in "The Grammarian and his Language", p. 156-7, originally published in The American Mercury, 1924):

There is no doubt that the critical study of language may also be of the most curious and unexpected helpfulness to philosophy. Few philosophers have deigned to look into the morphologies of primitive languages nor have they given the structural peculiarities of their own speech more than a passing and perfunctory attention. When one has the riddle of the universe on his hands, such pursuits seem trivial enough, yet when it begins to be suspected that at least some solution of the great riddle are elaborately roundabout applications of the rules of Latin or German or English grammar, the triviality of linguistic analysis becomes less certain. To a far greater extent than the philosopher has realized, he is likely to become the dupe of his speech-forms, which is equivalent to saying that the mould of his thought, which is typically a linguistic mould, is apt to be projected onto his conception of the world. […] In their recently published work on "The Meaning of Meaning" Messrs. Ogden and Richards have done philosophy a signal service in indicating how readily the most hardheaded thinkers have allowed themselves to be cajoled by the formal slant of their habitual mode of expression.

Just to keep things in balance, I should note that this same essay is the source (on p. 153) of a quote that I featured a few years ago ("Edward Sapir and the formal completeness of language", 11/20/2003):

The outstanding fact about any language is its formal completeness. This is as true of a primitive language, like Eskimo or Hottentot, as of the carefully recorded and standardized language of our great cultures. […] [W]e may say that a language is so constructed that no matter what any speaker of it may desire to communicate, no matter how original or bizarre his idea or his fancy, the language is prepared to do his work. […] The world of linguistic forms, held within the framework of a given language, is a complete system of reference, very much as a number system is a complete system of quantitative reference or as a set of geometrical axes of coordinates is a complete system of reference to all points of a given space.

Sapir goes on demolish at length the idea that it could possibly be true that language limits thought, in the sense that speakers of some particular language might simply have no way to express some particular concept. This "no word for X" idea, so common in modern pop-Sapir-Whorfianism, was completely and explicitly alien to Sapir's way of thinking.


  1. michael ramscar said,

    September 12, 2010 @ 4:54 pm

    Dan Slobin described the division that languages like Turkish and Russian make in English blue by getting me to think of the line where the sea meets the sky. It seems to illustrate of Sapir's "complete system of reference" idea nicely. But then again, the difference between the sky and the sea depends a lot on which sea — we weren't at the beach when we had this conversation, and I think we'd both have been more confident that I understood exactly what he had in mind if we both spoke a language that had conventionalized the two different 'blues.'

    Which suggests, to me at least, that whether language is a "complete system of reference" is going to depend a lot on what you mean by "complete" "system" and "reference" (the last one being, of course, the mother of all worm cans). The same goes for whether language affects thought: it all depends on what one means by "language" and "thought." From reading back through your posts (which I've enjoyed immensely, thanks!) it seems like Sapir was more than aware of this.

    So it seem that Saphir might say "that a language is so constructed that no matter what any speaker of it may desire to communicate… the language is prepared to do his work" while leaving open the question of whether ideas might be easier to communicate in some languages, or whether the differences between languages might make that communication more or less certain (better or worse, even).

    Which I think goes back to your "zeitgeist-y" observation. Like Wittgenstein, Saphir seems to have been more that usually aware of the danger that the way we talk about language can affecting our thinking about that:

    "One keeps hearing the remark that philosophy really makes no progress, that the same philosophical problems that had occupied the Greeks are still occupying us. … The reason [why this is true] is that our language has remained the same and seduces us into asking the same questions over and over. As long as there is a verb ‘to be’ which seems to function like ‘to eat’ and ‘to drink,’ as long as there are adjectives like ‘identical,’ ‘true,’ ‘false,’ [and] ‘possible’, as long as one talks about a flow of time and an expanse of space, etc., etc., humans will continue to bump up against the same mysterious difficulties, and stare at something that no explanation seems able to remove." (Wittgenstein)

    It's an important message. It's a shame that it has gotten lost in the noise of arguments about whether "language affects thought" more generally.

  2. Andrew Garrett said,

    September 13, 2010 @ 2:24 am

    I disapprove of the putative armchair dustiness ("the dusty armchair of Germanic philology"). Armchair linguists deserving of their pipes and tweed elbow patches are constantly checking texts, dictionaries, grammars, journal articles, and now online resources, while walking from desk to bookshelf and back, and and are far too active for dust to settle on their chairs.

  3. jan wohlgemuth said,

    September 13, 2010 @ 4:59 am

    When I read this article's headline, I was briefly expecting an actual picture of an, or _the_, armchair.

  4. maidhc said,

    September 13, 2010 @ 5:05 am

    I have a question about the "formal completeness" concept.

    I've read that one of the contributions of Alfred the Great was that he had a lot of Latin books translated into Anglo-Saxon. Particularly philosophy books required the development of an entirely new Anglo-Saxon vocabulary which made it possible to discuss complex intellectual topics in Anglo-Saxon, which had not been possible before.

    I suppose one could say that what he did was more of a cultural change, which made people want to talk about those topics. Had they wanted to discuss those topics previously, they could have constructed the vocabulary from their existing wordhoard. Which is exactly what the translators had to do.

    I wonder if there are similar examples?

    [(myl) More from Sapir 1923:

    Formal completeness has nothing to do with the richness or the poverty of the vocabulary. It is sometimes convenient or, for practical reasons, necessary for the speakers of a language to borrow words from foreign sources as the range of their experience widens. They may extend the meanings of words which they already possess, create new words out of native resources on the analogy of existing terms, or take over from another people terms to apply to the new conception which they are introducing. None of these processes affects the form of the language […]

    Sapir considers at length the problem of translating Kant into Hottentot or Eskimo, and the more specific problem of creating a word to express the notion of "causation", and concludes that

    The unsophisticated natives, having no occasion to speculate on the nature of causation, have probably no word that adequately translates our philosophic term "causation," but this shortcoming is purely and simply a matter of vocabulary and of no interest whatever from the standpoint of linguistic form. […] All languages are set to do al the symboic and expressive work that language is good for, either actually or potentially. The formal technique of this work is the secret of each language.


  5. maidhc said,

    September 13, 2010 @ 5:27 am

    Colour terms—colour has both hue and saturation. In English the colour words mostly describe hue. So "red" describes a hue regardless of saturation. It could be dull red, light red or bright red.

    In Irish, saturation is much more important. "Rua" is a dull red but "dearg" is a bright red. "Glas" can be a dull blue, green or grey.

    Essentially the colour space can be chopped up into pieces and given names in different ways.

    I don't know that this demonstrates any deep truths about how people view reality or not.

    In Irish you don't say "my food" or "my money", you say "my part of the food" or "my part of the money" ("mo chuid airgid"). Some people say this shows that Ireland is at heart an egalitarian society.

    In English "you" is the same for singular and plural, so obviously English speakers don't know how to count.

  6. marie-lucie said,

    September 13, 2010 @ 9:04 am

    the dusty armchair

    If the armchair gets dusty, it means that no one is sitting in it, therefore the potential occupant has better things to do than just sit.

    In English "you" is the same for singular and plural, so obviously English speakers don't know how to count.

    Similarly, English speakers are always saying "we have to sit down and talk": this must mean that the language demands such concentration that speakers cannot properly talk unless they first sit down.

  7. marie-lucie said,

    September 13, 2010 @ 9:39 am

    So it seem that Saphir might say "that a language is so constructed that no matter what any speaker of it may desire to communicate… the language is prepared to do his work" while leaving open the question of whether ideas might be easier to communicate in some languages, or whether the differences between languages might make that communication more or less certain (better or worse, even).

    Sapir also wrote (I quote from memory) that in matters of language, Confucius walks with the savage warrior of Assam, and Plato with the swineherd of Macedonia. In other words, just because a language is spoken by people with presumably no interest in philosophy does not mean that this language does not have the potential for communicating philosophical ideas. The Greeks had been speaking their language for centuries before the great philosophers (free from tilling the soil or herding animals) had the leisure to begin their speculations.

    On the other hand, some philosophical problems are created by features of language: not every language needs a "verb to be" which creates an expectation that "being" can be defined just like "eating" or other verbs with meanings relating to actual actions or states. But I am sure that philosophically inclined Russians are not barred by their language from discussing ideas.

    In the 18th century a Frenchman won a prize by praising the French language for its "clarity", claiming that English or German can be unclear, but not French. It is relevant to know that the organization awarding the prize was largely composed of French speakers (Huguenot refugees in Berlin). In fact, like most languages, French can be very clear on certain things and unclear on some others, and we can usually express ourselves and understand others most clearly in our own language.

  8. marie-lucie said,

    September 13, 2010 @ 9:44 am

    philsophy and Anglo-Saxon

    The new words were not borrowings from Latin, but coined using existing Anglo-Sason constituents.

  9. J.W. Brewer said,

    September 13, 2010 @ 9:54 am

    I'm trying to figure out if the formal completeness point is vacuous or not. It blurs the line between actual existing languages, with whatever stock of syntactic constructions and lexical resources they may have as of right now, and some indefinite range of hypothesized potential languages that the currently-existing languages could perhaps evolve into. How would the claim that "ok, right now, it's not possible to have a conversation about Kant or quantum mechanics in language X but it definitely would be if enough speakers of that language wanted to have such conversations, because the necessary resources could and thus would then be created through any of a number of historically-attested processes (at which point, of course, X would arguably be a somewhat different language, with the usual line-drawing issues)" be falsified? I mean, obviously it's not falsified because in a given instance the members of that speech community interested in such topics are already bilingual in English or Spanish or some other more culturally dominant language and simply use that langauge for such discussions, because that can happen for all sorts of social/cultural reasons that show nothing about what capabilities language X could have developed but did not. So what would falsify it? Or is it an essentially theological claim?

  10. Nathan said,

    September 13, 2010 @ 11:20 am

    @J.W. Brewer: Did Kant write his works in some previously nonexistent form of German? Did the first quantum physcists have to create new languages to publish their first papers?

    I think the point of talking about formal completeness is to combat the sort of linguistic chauvinism that rears its head all the time, implicit in the suggestion that your hypothetical speech community needs to use a "more culturally dominant language" to discuss new ideas.

    I suppose falsification might start with something like your informant saying, "I'm sorry, can we talk about this in English? We can't add that idea to my language." Yeah, it is a pretty silly think to imagine.

  11. marie-lucie said,

    September 13, 2010 @ 11:58 am

    Discussing a philosophical, scientific, etc topic originally propounded in one language is not the same as using this language to carry on the discussion. Most current specialists in Ancient Greek philosophy do not hold congresses where the language is Ancient Greek (even in Greece). Instead, the vocabulary for many key concepts have been borrowed or adapted from Ancient Greek (eg the word philosophy) into a vareity of languages and are used like regular vocabulary items in whatever languages the concepts are discussed. Where borrowing is rare, speakers make up new words, often with considerable ingenuity (as for instance in Chinese).

    I remember reading something by a linguist or anthropologist who was with a native family in the Northwest Territories (Canada) as they watched the first moon landing on TV. The old grandfather did not speak or understand English, so his bilingual son kept up a running commentary in the native language. No doubt there were some technical words for which the language did not have an equivalent, but that would not have been a massive impediment. Borrowing technical words is not always necessary if it is possible to freely form new words meaning "the thing that does X" or "that looks like an X" in more or less detail.

  12. Ran Ari-Gur said,

    September 13, 2010 @ 12:02 pm

    @michael ramscar, regarding "I think we'd both have been more confident that I understood exactly what he had in mind if we both spoke a language that had conventionalized the two different 'blues.'": Maybe, but on the other hand, native speakers often disagree, sometimes vehemently, about what color something is. This happens a lot at the boundaries between color-regions — is a yellow-green building "yellow", or is it "green"? — but it can even happen in one speaker's "middle": I remember once seeing a purple sweater, royal purple, purple as purple can be, only to find that another person present somehow considered it blue. Google for phrases like "it's not red, it's orange", and you'll find no shortage of further examples.

  13. Jerry Friedman said,

    September 13, 2010 @ 12:14 pm

    @Nathan: I suppose falsification might start with something like your informant saying, "I'm sorry, can we talk about this in English? We can't add that idea to my language." Yeah, it is a pretty silly think to imagine.

    I've heard it the other way, something like "You'd understand this if we could talk about it in Navajo, but there's no way to say it in English." The person spoke English better than Navajo.

    @Ran Ari-Gur: I totally agree with you about the boundaries between color regions, but in the case of that purple sweater, there's another possible explanation: seeing purple as blue is typical of us "color-blind" people.

  14. Jim said,

    September 13, 2010 @ 12:14 pm

    "I wonder if there are similar examples?"

    The invention of masses of petrochemical terms in Arabic? Probably not really very diferent from the invention of masses of petrochemical terms in English other than happening in a much shorter time.

  15. Memphis wally said,

    September 13, 2010 @ 12:33 pm

    I've got the '"I think we'd both have been more confident that I understood exactly what he had in mind if we both spoke a language that had conventionalized the two different 'blues.'"'

    Yeah, I've got the '"I think we'd both have been more confident that I understood exactly what he had in mind if we both spoke a language that had conventionalized the two different 'blues.'"'

    And I'm gonna get out my color wheel and check out some of those hues

  16. Xmun said,

    September 13, 2010 @ 12:39 pm

    On a recent visit to Rarotonga we happened to get talking with the Cook Islands prime minister at our farewell dinner. He told us the debates in the Cook Islands parliament are conducted in English, because — although everyone also spoke Cook Islands Maori — English was "more precise". That's to say, it has the vocabulary required for such debates. In Cook Islands Maori one would have to use a clumsy periphrasis to express such a concept as (say) "stakeholder needs analysis".

  17. Akunin said,

    September 13, 2010 @ 12:47 pm

    There seems to be a resurgence of neo-Whorfism, and as always, it captivates the public's imagination. If you are going to talk about Guy Deutscher's piece, maybe it would be worthwhile to comment on this Radiolab episode:

  18. Nathan Myers said,

    September 13, 2010 @ 2:17 pm

    I'm reminded of practices by flight crews such as at Korean Airlines, who conduct cockpit discussions in English because safety requires an egalitarian frame of mind said to be impossible when speaking Korean. Whether it is or isn't, KAL's accident rate plummeted, (according to the not infallible Malcolm Gladwell) after they instituted the rule.

  19. marie-lucie said,

    September 13, 2010 @ 2:31 pm

    In Cook Islands Maori one would have to use a clumsy periphrasis to express such a concept as (say) "stakeholder needs analysis".

    This is a problem of translating bureaucratese. English or German compounds (whether the components are written separately or together), which abound in such discourse, are also difficult to translate into Romance languages, which also resort to periphrases. But those compounds are not always precise, depending how one understands the relationship of the components to each other (where one could draw "trees" to dispel the potential ambiguities – an elementary exercise in syntax).

  20. marie-lucie said,

    September 13, 2010 @ 2:33 pm

    safety requires an egalitarian frame of mind said to be impossible when speaking Korean.

    I suppose this is about translating "you" when unable to determine the proper degree of politeness to be used with one's interlocutor. It is not directly related to the structure or vocabulary of the language itself.

  21. J.W. Brewer said,

    September 13, 2010 @ 2:45 pm

    Nathan: my concern is in fact that the default culture of anthropology and linguistics departments in the modern academy is so committed to "combat[ting] . . . linguistic chauvinism" that certain anti-chauvinist premises might be taken to be true in some moral or quasi-theological sense rather than being treated as, you know, empirical claims about language that might or might not be supported by evidence. Saying that one shouldn't really compare language A and language B as they actually exist but instead as they could each in principle be potentially transformed in the future to address new needs seems like a convenient way of keeping these egalitarian precommitments from being contaminated by empiricism.

  22. George said,

    September 13, 2010 @ 2:47 pm

    It has been claimed that what is significant is not what we CAN say, but what we MUST say with our inflections. An example is a language with evidentials. In English, if we wish, we can express our degree of certainty, how we acquired the knowledge, etc. But, in some languages, this is obligatory. So, presumably, the speaker must consider this in making any statement.

    Another example is comparatives. Presumably, an English speaker routinely notices relative sizes, weights, etc. where an Aymara speaker may not even notice as there are no comparatives in Aymara.

  23. George said,

    September 13, 2010 @ 3:16 pm

    I suspect that the use of English by Korean flight crews is the result of cultural norms that exist independent of any language constraints. Does Gladwell suggest that the cultural norms are dispensed with when the same people switch to English? Hmm.

    It is strange that Koreans seem perfectly capable of commercial enterprises like construction and manufacturing, in maintaining a military, a government and many modern heirarichal activities. Yet, they can't safely fly airplanes using their native language. Hmm.

    Also, I wonder how much English lingua franca in airline travel might have had an effect.

  24. Nathan Myers said,

    September 13, 2010 @ 5:37 pm

    George: Reportedly, KAL switched to English in the cockpit as one way to correct an execrable safety record.

    marie-lucie: I mentioned the KAL case specifically because it seemed to demonstrate an example of a structural difference between Korean (like many languages) and English, one not addressable by vocabulary additions. In the cockpit there is rarely any ambiguity about the relative social positions of the flight officers: that's the problem. The copilot cannot, in Korean, say to the pilot things that must nonetheless be said. The opposite problem, exemplified by the Japanese friends, professor and student, who must converse in English because neither can correctly address the other in Japanese, seems equally resistant to vocabulary infusions.

    Could KAL's safety problems have been solved by some means other than switching their flight decks to English? We can't run the experiment, and I think I'm glad they didn't.

  25. Sir K said,

    September 13, 2010 @ 6:45 pm

    I lived in Iceland for a while where the coinage of "new" words seems to be a national sport. The local papers published long letters from the readers, in which they would discuss in great detail how to best translate technical, philosophical and other terms that didn’t exist in Icelandic so far.

    Sometimes different letter writers prefer different new terms. This can be either newly coined words or sometimes archaic words that are not in everday use which are attributed a new or expanded meaning. Especially among teachers and university professors this seems to be quite popular and somewhat competitive (at least judging from the passionate defenses of different Icelandic coinages for English words in said papers).

    As an aside, Icelanders are extremely proud that their language “hasn’t changed since the time of the sagas” – at least that is the common and much-loved myth about a language the locals perceive to be very pure (but that’s another discussion). Foreign loan words are frowned upon and as soon as a new term becomes more widely used, the hunt for a “proper” Icelandic word begins

  26. Nat said,

    September 13, 2010 @ 8:16 pm

    I'd like to register a concern about the cogency of Sapir's appeal to "formal completeness" here. This means something very specific in logic. Does Sapir literally mean that all languages have the resources of first-order predicate logic? That claim seems reasonable enough. But it doesn't establish his argument. "Completeness" doesn't mean "able to express every possible concept". First order logic can't express the difference between countably infinite and uncountably infinite (v. Lowenheim-Skolem). Nor the difference between red and blue, for that matter.
    If he doesn't mean "formal completeness" in his technical sense I have a hard time seeing how this claim could be plausibly established. Well, this is the point at which I should read his text, I suppose.

    As an aside, morality, theology, and the logical properties of languages are three distinct topics. That said, figuring out the kind of evidence that could be used to support Sapir's claims might help to make it clearer just what he means.

  27. marie-lucie said,

    September 13, 2010 @ 8:21 pm

    … one shouldn't really compare language A and language B as they actually exist but instead as they could each in principle be potentially transformed in the future …

    Is anyone actually saying that? Nobody can predict the future of a language any more than the future of a person or a nation. Of course there are differences between languages spoken by millions of people and with a long literary and scholarly tradition (including the study of other languages) and those spoken by a small, isolated group of people, but in many cases the sisolated group may have a smaller vocabulary (especially of things or concepts unknown or irrelevant to their culture – not in things important to the culture) but it often has a more complex phonology and morphology. "Large" languages on the other hand tend to accumulate vocabulary from various sources, but the many threads of intercommunication between different regions, classes, professions, etc tend to reduce internal structural diversity in phonology and grammar. Also, in "large" languages which have a standard form (usually that of the capital city) and a number of dialectal forms, which dialect emerged and developed its vocabulary through exposure to internal diversity, foreign adaptations, new coinages, etc is not a matter of the features of that dialect but simply the fact that it was that of the region or city which was chosen as the seat of government (often a place already well known for reasons having nothing to do with the features of the local dialect). It is simply not the case that a dialect is chosen as having "more potential" than another, unlike the choice of a capital city or cultural centre. In cases where language planning takes place in order to develop a national language (eg in Indonesia or the Philippines), the choice of the base dialect is political and cultural, not strictly linguistic. The point is that any language can be expanded, transformed, etc by its speakers if circumstances arise to make such changes indispensable.

  28. marie-lucie said,

    September 13, 2010 @ 8:32 pm

    (Korean and japanese problems)

    The copilot cannot, in Korean, say to the pilot things that must nonetheless be said. The opposite problem, exemplified by the Japanese friends, professor and student, who must converse in English because neither can correctly address the other in Japanese

    These are instances of new sociolinguistic norms needed for new, non-traditional situations, not features of the structure of Korean or Japanese. It is not that things cannot be said at all, it is that the social situations require them to be said in certain ways which contradict the traditional social interactions and therefore create awkwardness. What do Koreans and Japanese ignorant of English do? English conveniently comes to the rescue for those who know it well enough, but it does not help people who are largely monolingual, and who will eventually determine the new norms.

  29. George said,

    September 13, 2010 @ 9:21 pm


    Are you saying that Koreans when speaking English can violate Korean cultural norms?

  30. J.W. Brewer said,

    September 13, 2010 @ 9:22 pm

    m-l: I was responding/reacting to Safir's statement quoted by myl above that "All languages are set to do all the symbolic and expressive work that language is good for, either actually or potentially." I mean, maybe they are, but my concern is that I don't know if the "potentially" part of the claim is falsifiable (and thus if the sweeping claim as a whole is falsifiable), and it somehow strikes me as like saying every child has the potential to grow up to be president. It feels sort of motivational-speaker / self-esteem-boosterish.

  31. marie-lucie said,

    September 13, 2010 @ 9:41 pm

    Are you saying that Koreans when speaking English can violate Korean cultural norms?

    No, it is just that some of those norms are irrelevant when speaking English. I don't know Korean or Japanese, but I have heard of similar situations in Asian cultures where the use of English (for those who have learned it) can be used to bridge unconventional situations.

    Let me give you an example I am very familiar with, from French (and other European languages). In French a person can be addressed with the pronoun "vous" (neutral with a fellow adult stranger, respectful with an older person or superior) or "tu" (children, classmates, etc)(simplifying the subtleties). These are times where you are not sure which of those two pronouns to use, but if you speak English, the distinction is irrelevant since "you" is completely neutral in those respects. On the other hand, in English you might wonder whether to address a certain person as "Sir" or Doctor X" or "John" or "Hey, you!" – but none of these is obligatory in English – you don't have to use a term of address at all. If there was a different pronoun to go with each of those terms (not just the ubiquitous "you") and you had to settle on one of these (because the language made it obligatory to choose the appropiate pronoun even if you left out the actual term of address), and if the other person might be greatly offended, or at least very embarrassed, if you used the wrong pronoun, then you might be in the situation that I imagine to be that of the Korean pilots.

  32. marie-lucie said,

    September 13, 2010 @ 9:55 pm

    it somehow strikes me as like saying every child has the potential to grow up to be president.

    No, because each individual child has their own gifts and personality, and the child's development is affected by genetic, physical, cultural and emotional factors. But languages are not biological organisms, they arise from communication between people under circumstances which vary, and they are not born anew with each generation but carry within themselves the heritage of countless generations. Children grow up into adulthood, and later fade into old age, regardless of the society they live in. Some children have outstanding gifts in spite of being born and raised under very unfavorable circumstances, and some are mediocre in spite of the best circumstances. The evolution of languages cannot be described in any of those terms.

  33. Diane said,

    September 13, 2010 @ 11:14 pm

    I have run into quite a few people who say that it is easier to express a lot of concepts related to modern science and technology in English than in their native language. These have all been Asian people, mostly Indians, whose higher education was conducted in English, and they bring up this perceived deficiency in their native language to justify the wisdom of conducting higher education in English in their countries.

    I find this saddening. It seems to me that if their native language really is deficient at expressing modern concepts, then it must be *because* so much higher education is conducted in English. Educated people in their countries must be learning how to express modern concepts in English, and not being given the chance to develop a way to express it in their native languages. If they don't want their language to become irrelevant, they ought to stop conducting higher education in English forthwith.

    So far, I have convinced no one.

    Still, it strikes me as really sad. I always assumed you could express any concept in any language, but I didn't allow for the possibility that the speakers of some languages could not be bothered to figure out a way to do so.

  34. marie-lucie said,

    September 14, 2010 @ 12:21 am

    Diane, I quite agree with you.

    One reason that English has the words for scientific concepts is that it (along with other European languages) has borrowed a lot of vocabulary from Latin and Greek along with the concepts, and then added to this vocabulary by constructing new words with bits and pieces of the Latin and Greek words. Many of those "English" words are therefore of international origin, not peculiar English coinages.

  35. D.O. said,

    September 14, 2010 @ 12:41 am

    It might be insufferable nitpicking on my part, but then maybe I am missing something? Why the tags for this post are Ignorance of Linguistics and Language and gender. Ignorance of linguistics is OK (though it is closer to Ignorance of the History of Linguistics), but gender?

    [(myl) It's because "Language and gender" is next to "Language and culture" on the list, and I was using an unreliable pointing device in a hurry, and clicked on the wrong thing by mistake, and didn't proofread the choice. It's fixed now.]

    Ange Mlinko is a woman while Guy Deutscher, Edward Sapir, Benjamin Whorf, Bertrand Russell, Ludvig (or Ludwig, whatever) Wittgenstein, Alexander von Humboldt, Franz Boas, Charles Ogden, Ivor Richards and even Mark Liberman are men. Does it bear any significance on the discussion?

    [(myl) No. See above. One problem here is that at some point in the past, in a fit of taxonomic enthusiasm, someone added more than 100 categories to our list. Most of them have never been used, but they still clutter up the list of choices and make it easy to make mistakes of this kind.]

  36. groki said,

    September 14, 2010 @ 2:38 am

    @Diane, @marie-lucie: I sympathize with your sadness, but let me expand on what marie-lucie brings up.

    in Europe over the last several hundred years, Latin was the language that learnéd experts–including natural philosophers, scientists, and inventors–carried out their studies in, whatever native languages they spoke. in fact (like Diane's non-English speakers with English) they lavished praise on Latin for its aptness and excellence as the language of the mind.

    one eventual result, as marie-lucie notes, was that English and other languages were able to utilize Latin (and Greek) to amass a treasure trove of very useful scientific and technical terms. it is now the richness of that very trove that, among the non-English speakers, contributes to English being so in demand for higher learning.

    so if the Latin-to-English history is any guide (though I realize that today ceteris are certainly not paribus), then instead of those other languages being left behind, maybe English will wind up the dead language and those speakers (or their descendants) will eventually mine English to enrich their own languages' vocabularies.

  37. Nathan Myers said,

    September 14, 2010 @ 3:01 am

    What do […] Japanese ignorant of English do?

    As I understand it, (1) in such a circumstance, they typically do not speak; and (2) this consequence is more or less intentional in the design of the Japanese language. I don't know Japanese, so I must rely on reports of those who do.

  38. George said,

    September 14, 2010 @ 6:24 am

    Thanks for the explanation. In fact, I had thought of the French tu-vous variants. I too am not familiar with Korean, but I understand they have an elaborate system of honorifics.

    However, I am highly skeptical that a switch to English (and particularly when speaking with other Koreans) allows them to fly airplanes more safely. Koreans seem to have no trouble operating other aspects of a modern society, including an efficient military, which entails social cooperation and coordination.

    I also know that we often apply our native-language pragmatics even when speaking a second language. I suspect this would be even more so when speaking with other speakers of the same native-language in a second language.

  39. David J. Littleboy said,

    September 14, 2010 @ 8:27 am

    What do […] Japanese ignorant of English do?

    Uh, they do just fine, thank you. Every day, day in, day out.

    In general, the senior party in the conversation speaks informally, and the junior party speaks respectfully. It's really not all that different in the US, except that in the US, the professor gets to set the tone and the students follow in kind. At least at the institutions I've been at (only two, and east coast US ones), the professors are very much in charge, even in informal situations.

    I was heading home from bowling the other night, and observed a pair of punk rockers (blue mohawks, studded jeans) conversing with an older serious/professional looking woman and her assistant. The punk rockers were speaking the most amazingly polite Japanese I'd heard in a long time. I found it more than a tad incongruous, but the principles involved were not having any problems whatsoever. (I guessed that the women were promoters/organizers of some sort and the kids were from a band, since this was at a venue that holds concerts.)

  40. John Cowan said,

    September 14, 2010 @ 11:51 am

    I think what makes very hierarchical languages difficult to use in the airplane context is that it is a situation where subordinates must in safety-related situations give orders to their superiors, and that in the plainest possible language: "Let go of that knob, it's the 'jettison fuel' knob!" Using a foreign language imports to some extent the attitudes of the speakers of that language, though not directly for structural reasons. In all languages with highly elaborated registers, it is the plain register that people learn first and (I should think) the one to which they will react fastest.

    In Javanese, there are six status-related registers in use, distinguished almost exclusively by the lexicon. Each of them is appropriate to use, symmetrically or asymmetrically, in particular face-to-face circumstances (though not everyone learns all of them). This makes it very hard to give a political speech in the language, because one does not know what register to use to address a large group of people of diverse social statuses. So politicians speak Indonesian, which functions effectively as a seventh register of Javanese appropriate for certain non-traditional situations.

  41. J. W. Brewer said,

    September 14, 2010 @ 11:57 am

    What should India do as an alternative to having an Anglophone educated elite? Conduct higher education in 17-or-more separate languages, thus creating a bunch of isolated elites who would be unprepared to pursue professional opportunities outside their native part of the country? Or conduct it nationwide in Hindi, thus privileging one indigenous group over the others (as the role of English was downgraded post-independence in Ceylon / Sri Lanka, with both the purpose and effect of keeping the Tamils down — for the history of pro-English rioting in Madras see here: )? The promotion of a semi-artificial national lingua franca that was not merely a relabeled Hindi (parallel perhaps to what was done in the Phillipines and Indonesia) would be unlikely to bridge the gap between the various IE languages and the Dravidian ones, to the almost certain disadvantage of speakers of the latter. If history had taken a different turn, maybe Sanskrit could be used for scholarship the way Latin was in Europe until the 19th century; if the Moghuls had not been overthrown it might still be Persian; if something else had happened there might be a local analogue to Swahili in East Africa, reasonably closely related to the various contending languages without being identified too closely with any one significant ethnic faction. But they've got to have something.

  42. Tadeusz said,

    September 14, 2010 @ 12:24 pm

    I think that your question is quite interesting: what would falsify the hypothesis about formal completeness of a language. By the way, in you can find more details about it. While I do not remember clearly Sapir's line of reasoning, I take it that he meant that any language has a complete range of formal means to produce new form/meaning units. Which, in turn, means that you can produce a unit with a new sense, or a completely new unit, i.e., a new form with a new meaning. The simplest unit is the word (or a free meaningful form in languages which do not have words), a new phrase, a new clause, etc. So if you find a language that: a. does not allow its users to produce new meanings, b. does not allow its users to produce new forms (either one factor or both), that means that the hypothesis has been falsified.

  43. Tadeusz said,

    September 14, 2010 @ 12:25 pm

    I am sorry: J. W. Brewer.

  44. marie-lucie said,

    September 14, 2010 @ 1:03 pm

    George: I also know that we often apply our native-language pragmatics even when speaking a second language. I suspect this would be even more so when speaking with other speakers of the same native-language in a second language.

    But in a second language, the obligatory pragmatics (such as status markers) may not exist, as in English: "you" is OK for everybody, so you don't have to worry about what pronoun to use to express the precise relationship between you and the other person. There have been descriptions of English learners or second-language speakers in Asian countries conversing in English for precisely that reason (practicing one's English with other learners is socially liberating).

  45. George said,

    September 14, 2010 @ 2:09 pm

    I would bet that if 'tu/vous' were an impediment to French crew communications, they would devise a system in which either 'tu' or 'vous' is consistently used, or titles like captain, co-pilot, navigator, etc. I would expect the same of Korean honorifics.

    Somehow the French and Korean militaries, police, emergency rooms, operating rooms, etc. seems to function, presumably in their native languages.

  46. Nathan Myers said,

    September 14, 2010 @ 2:54 pm

    George: You can bet all you like, but what actually worked for KAL was switching to English. Emergency and operating rooms that "seem to function" in the U.S. cause tens of thousands of unnecessary deaths every year. Those deaths would be prevented by the methods employed to prevent unnecessary deaths in aviation. Spontaneous pragmatics suffice when the stakes are low.

  47. George said,

    September 14, 2010 @ 3:17 pm


    Maybe these dysfunctional U.S. operating rooms should follow the Korean airline example and switch to English. (~smile~)

  48. David J. Littleboy said,

    September 14, 2010 @ 4:19 pm

    "I think what makes very hierarchical languages difficult to use in the airplane context…"

    My opinion here is that some twit amateur linguist (Gladwell*) confused a culture problem (Korean airlines was full of ex-military types who didn't like taking orders from civilian pilots) with an also present linguistic problem (English is the language used for talking to control towers, and if your second language is poor English, this is bad news from a safety standpoint). Switching languages slows people down, especially if they are not really bilingual and weak in the second language. If you do everything in English, then by the time an emergency comes around, you'll be a lot faster in getting to the words you need.

    "Recently an internal report leaked onto the internet alleged that an authoritarian cockpit culture, inadequate English and pilot error were compromising safety.

    One of the concerns is that civilian and former military pilots working for the airline do not gell together."


  49. marie-lucie said,

    September 14, 2010 @ 4:51 pm

    If the KAL pilots (presumably Korean) are switching to English while in the cockpit, a) they must have a pretty good command of English already, and b) they have presumably been trained by English speakers, probably Americans. Having learned the appropriate technical terms and common phrases in English, they may be more at home in English than in Korean in this professional situation, in addition to avoiding the sociolinguistic pitfalls of normal Korean conversation in the course of daily life.

    I gave the French tu/vous dichotomy as an example of obligatory choice (as opposed to the neutrality of English you), not as something that currently causes problems in communication.

  50. Nathan Myers said,

    September 14, 2010 @ 5:08 pm

    George: Category error.

  51. George said,

    September 14, 2010 @ 5:11 pm

    David J. Littleboy:

    The reasons given in the BBC article you linked makes more sense to me than language inadequacies.

    As I recall, "authoritarian cockpit culture" has been cited as an issue with American airline safety as well.

  52. marie-lucie said,

    September 14, 2010 @ 5:16 pm

    David Littleboy: I had not seen your comment when I wrote my latest one, but it makes a lot of sense.

  53. peterv said,

    September 14, 2010 @ 7:31 pm

    Having worked in Korea and with Korean business-people, I would say that the main problem with using Korean in the cockpit is not with norms of hierarchy (since those were mostly been broken by the long Japanese occupation and the disruption of the invasion by the North), but with a traditional culture of indirection. This culture makes it most uncommon for Korean managers to give clear and unambiguous instructions to subordinates. A standing joke among expatriates in Seoul is of the office staff meeting where the boss speaks only elliptically and poetically (eg, "The sky is dark, but the moon is bright!"). When he leaves the room, his subordinates spend the rest of the meeting arguing over what exactly he meant for them to do.

  54. Diane said,

    September 14, 2010 @ 8:11 pm

    @ J.W. Brewer

    You make a good point about the need for English in a country with as many diverse languages as India. I do realize that this is why I am not going to convince anyone to give up English-medium education.

    My point is simply that many Indians *also* argue that English is just better for expressing science and technology concepts, and I think they don't necessarily recognize that this could be the result of their educational system and not some inherent superiority of English.

  55. George said,

    September 14, 2010 @ 8:29 pm

    I have encountered a strong affinity for English among Arab elites. I suspect it is related to making a social distinction between themselves (educated, urbane) and ordinary people.

  56. Nathan Myers said,

    September 14, 2010 @ 10:48 pm

    If the KAL pilots … are switching to English while in the cockpit, a) they must have a pretty good command of English already

    No. Major airline cockpit procedures are not decided by whim. The changes instituted at KAL were a major investment. Flight crew had intensive, required language training after the airline made the decision to switch. Previously, cockpit operations had been conducted in Korean.

  57. marie-lucie said,

    September 15, 2010 @ 1:38 am

    Do you mean then that the pilots were taught English just so they could use it in the cockpit, and that worked better than when they used their own language? did the improvement occur on flights within Korea as well as on international flights? I suppose that the Korean air controllers already knew enough English if they were dealing with international flights, so the pilots had to learn English to communicate with the air controllers too (so that those controllers would not be switching back and forth between Korean and English). What a nightmare! And we must be talking about South Korea only – I doubt that North Korean pilots have much opportunity to learn English.

  58. Nathan Myers said,

    September 15, 2010 @ 6:13 pm

    KAL is a South Korean airline. KAL has never had any say in North Korean flight operations.

    There is no need to suppose about any of this. You can read all about KAL's switch to English yourself. Pilots already knew limited English for tower and controller communication, but not enough for flight deck operations. Flight crews got intensive training so they could use English among themselves in the cockpit to the exclusion of Korean. To use different flight deck procedures for domestic and international flights would be silly; if what the international crew does is better, there's no reason not to do it on all flights.

    KAL had numerous fatalities blamed on communication problems before the switch, and none (of any kind) since. I count that as improvement.

  59. marie-lucie said,

    September 15, 2010 @ 10:00 pm

    Well, if it works for them, then it works for them (strange as it seems).

    About North and South Korean airlines, of course they have nothing to do with each other right now, but at some point in the future the two halves of the country are bound to be reunited and there will be interesting communication problems.

  60. Nathan Myers said,

    September 15, 2010 @ 11:01 pm

    marie-lucie: If it seems strange to you that switching to a second language could improve communication to the point of eliminating previously-frequent fatal accidents, maybe that means you should make the effort to learn more about it.

    Does North Korea have what we would call an airline? Integrating what they do have into KAL seems among the least of the problems of a Korean re-unification. Is re-unification really bound to occur? What do you see as the prospects for Romania's and Moldova's re-unification? How about (parts of) Belgium with France?

  61. Martin Ellison said,

    September 16, 2010 @ 4:14 am

    Indonesian invented a new second-person pronoun (anda).

  62. pm said,

    September 16, 2010 @ 7:53 am

    marie-lucie said (September 15, 2010 @ 1:38 am)

    "I doubt that North Korean pilots have much opportunity to learn English."

    If the Government of the DPRK (aka North Korea) wanted their pilots to learn English, this could no doubt be arranged. The DPRK used to send its diplomatic and foreign trade personnel to learn English in Harare, in special classes conducted by the University of Zimbabwe.

  63. marie-lucie said,

    September 16, 2010 @ 10:29 am

    NM: I am quite aware that North and South Korea do not share airlines or much else at this time. But for centuries they were one country, and on both sides the people speak Korean and consider themselves Korean. Similarly, for several decades there were two Germanies, both peopled by Germans, but they eventually reunified. Those two cases are of formerly unified countries which were split into two through external events and foreign intervention, not through the voluntary secession of one distinctive region. The cases of French and Belgium, or Romania and Moldava, are not the same at all.

  64. Tracy W said,

    October 25, 2011 @ 3:53 pm

    This is a year late I know, but Ive been browsing the archives. Precise visual description strikes me as something that English can't handle, possibly. There is a whole field of technical drawing with a long history. Nowadays we have computers, photocopiers and faxes so copying and distributing images is cheap, but in the 19th century I understand that places like engineering firms used to maintain teams of draftsmen and scientific books and journals would add images at considerable expense. If any English speaker could sufficiently precisely specify anything they could think in English, wouldn't it have been cheaper to just describe things verbally?

RSS feed for comments on this post