THE MOST DEMANDING SCIENCE.

Allan Metcalf has a nice appreciation of Eric P. Hamp and of his field, historical linguistics, in Lingua Franca:

Indo-European linguists like Hamp compare the modern languages with one another to reconstruct the common ancestor spoken some thousands of years ago, long before any language was recorded. That means observing patterns of relationships among hundreds of current languages. To do this properly means studying those hundreds of languages. Hamp has done this, not only with written languages but also with personal fieldwork throughout Europe and parts of Asia to learn lesser-known languages and dialects.

He quotes some nice bits from Hamp’s articles in the latest issue of Comments on Etymology, e.g. “Welsh illustrates with its normal set of numeral terms how a sophisticated and notably artistic and musical culture can evolve a set of terms at the same time traditionally systematic yet so complex that it would tire out and lose any of their neighbors if they ever took the trouble to learn to read their genuinely gorgeous poetry.” Hamp is 92 and still going strong, and reading things like this makes me wish I’d stuck it out in what was once my field as well.

Comments

  1. Although I write, I’m not an expert on language like everyone else here, so I find myself wanting to see posted here someday a “top ten” list of the major changes, in general (if such a thing exists) in language from thousands of years ago compared to now.

  2. Trond Engen says:

    Shelley: You might enjoy Guy Deutscher’s The unfolding of language, a well-written, enjoyable account of the processes that work on language now and how those same processes made it become language in the first place. For as long back as we can see, these processes have been in broad balance. While every language is constantly changing, there’s no change in language as such. (Or so goes the standard view. There’s a case to be made that professional bureaucracies and writing has led to the development of more complex syntactical constructions even in ordinary language.)

  3. Hat: I stuck it out, and more and more I think I made a mistake doing so: my academic employment (at several different institutions in different cities) has mostly involved teaching (remedial high school-level) L2 French, something neither my dissertation nor the seminars I took prepared me for at all. Increasingly I suspect they served no purpose whatsoever.
    Paradoxically, this extensive teaching experience is what strengthens my impression that getting a doctorate was a mistake: as long as I was pursuing my piled-higher-and-deeper degree I could convince myself that my department was atypical in its vicious political wars, petty vindictiveness, political commissar-like professors regurgitating the same theory they themselves were force-fed with, and rampant anti-intellectual atmosphere.
    Today I realize that said department was utterly typical: indeed linguistics departments are better than language departments, in that they lack the extreme snobbery (quite beyond caricature) which seems obligatory in the latter.
    Shelley: I’m not sure I understand what you mean with your “top ten” request. Individual languages have undergone various changes, of course, but language change as a whole is no different today from what it was in the oldest known languages: indeed it can be amusing to show that some individual change taking place today in, say, English is identical in nature to one which took place far away in time and space in some other language.

  4. Now I feel better about my life choices; thanks!

  5. des von bladet says:

    I’m not sure I understand: If I am a neighbour of Wales and I learn to read their genuinely gorgeous poetry I will – as a direct consequence? – be tired out and lost by the complexity of their numeral terms?
    I’m not sure if I should be pleased I no longer live in Bristol, for sure.

  6. marie-lucie says:

    I could have said the same thing as Etienne: my academic employment … has mostly involved teaching (remedial high school-level) L2 French, something neither my dissertation nor the seminars I took prepared me for at all.
    Unfortunately at this time it is hard to recommend a PhD in linguistics as a career move, let alone in historical linguistics which to many linguists seems to be a parent pauvre – a poor, struggling relative. Some linguistics departments do not offer even one course in the subject! Yet when you ask students at the start of an introductory course what they find most interesting about language, the majority answer “the history of words” or something similar. One could build a course on this interest, adding linguistic concepts little by little, but no modern textbook takes this approach.
    The image that comes to my mind is an iceberg, or rather an island: each language is like an island, which is the visible, above water tip of a mountain, a mass of earth connected to the bottom of the sea, and often to other islands and to continents. Of this enormous underwater mass the average inhabitant of an island knows nothing. For most purposes it is enough to know the parts above water or just below the surface, but in order to understand the structure and formation of the island and its relationship to other islands and to the nearest continent one needs geology. Similarly, ordinary users and students of a language, including descriptive linguists, do fine without knowing much about its history and its relationships with other languages, and those topics are the business of historical linguistics. Currently fashionable “theoretical linguistics” arose as an offshoot of descriptive linguistics, and has largely ignored the historical foundation.

  7. That’s a brilliant metaphor!

  8. marie-lucie says:

    Thanks, LH. I tend to think in images, even though I was told in school that I had no imagination whatsoever.

  9. I’ve probably said this before, but when I returned to university in the Eighties (after 20 years) I was glad to discover, after eight months of my old pleasure in languages and historical linguistics, that I couldn’t afford to go any further into debt to improve my academic standing, because I had had several glimpses into the petty politics and the anti-intellectualism Etienne refers to.
    And m-l, your metaphor rings true to an islander who knows something of geological structure.

  10. “Thanks, LH. I tend to think in images, even though I was told in school that I had no imagination whatsoever.”
    I see what you did there.
    “Currently fashionable “theoretical linguistics” arose as an offshoot of descriptive linguistics, and has largely ignored the historical foundation.”
    Which ignorance is apparent form some of the bizarre expalnations they offer for things which are quite simple if viewed in thier historical context.
    Etienne, your story reminds me of scholars in China who lived between dynasties when there were no imperial examinations. They transmitted the literature and culture of previous ages and a love for them even when therose things were of no career use. But they set their families up to succeed in later generations with their effort. We are in fact just coming to the end of such a period in our own lifetimes.

  11. “One could build a course on this interest, adding linguistic concepts little by little, but no modern textbook takes this approach.”
    Could one of those historical linguists turned unwilling remedial french teachers try writing a french (or whatever language they’ve been reduced to teaching) textbook with furtive linguistic lessons slipped in? It would be a huge improvement on the textbooks written at a toddler’s comprehension level to recreate natural language learning or whatever the writers think they’re doing, and would be providing at least some benefit for humanity…

  12. re:”Some linguistics departments do not offer even one course in the subject!” – in Aus, that’s ‘most linguistics departments’….

  13. All: I don’t wish to convey the impression that I am especially bitter about my own choices. I feel some wistful sadness, perhaps, but no real bitterness.
    Marie-Lucie: the problem is that teaching a course or two in historical linguistics needn’t imply that a department will actually hire a historical linguist! They’ll have sessionals teach such a class, often lending them some professor’s class notes, or some young freshly-hired non-tenured professor will have their arm twis…err, I mean, will be convinced to teach this course which is unrelated to their research.
    The core problem, it seems to me, relates to the fact that Academia is a highly globalized activity: trends and fashions are global, so that if something goes out of fashion it goes out of fashion *everywhere*. Meaning that research funds dry up *everywhere*, positions are unavailable *everywhere*…this is an especially serious state of affairs with some data-oriented subfields: language documentation and description are as unfashionable as historical linguistics today, for example, but should they ever become fashionable a generation hence, a huge number of languages will have gone extinct, unrecorded, in the meantime.
    Jim: unlike you, I don’t think we’re coming to the end of a normal historical cycle. I think that we’re coming to the end of a bubble known as industrial civilization, and that the next few centuries will be much more reminescent of the Middle Ages than of the twentieth century.
    Sometimes I wonder whether a thousand years from now some future linguist will write a history of the language families that will have arisen out of the diversification of English or Spanish or other single languages spoken today. If so I hope some knowledge of historical linguistics makes it to this future era: a comparative study of, say, the “neo-English” languages of the future and the “neo-Latin” languages of today would make a wonderful scholarly book. A pity none of us will be there to read it.
    s/o: actually, in some of my language courses I did slip in some historical linguistics. Whenever my students asked me why French had such-and-such a feature I told them what its origin was. Indeed, in an English course I taught to francophones a long time ago I was once asked why English has strong versus weak verbs, and I gave them a quick introduction to comparative Germanic that had them glued to their seats.
    As for writing a language textbook, I admit the thought has crossed my mind. One of the most popular French courses I have ever taught involved teaching students how to predict grammatical gender, something that is much more predictable than most French teachers realize, and a textbook introducing this would definitely fill a need. Since making these predictions does require some linguistic knowledge (i.e. pronunciation versus spelling, plus knowledge of the morphological make-up of the noun), putting a little more linguistics in there than is strictly required would not be too difficult…

  14. Historical linguistics is still puzzlingly unfashionable, sadly (just the other day I heard a colleague disparaging “19th-century linguistics”.) But in one respect, the situation is rather better than Etienne suggests: since the late 1990s/early 2000s, language documentation and description have become rather trendy, and a good deal of funding for them has become available, both in Europe and in the US (though that’s threatened by the current economic situation).
    If you’re looking for a historically-based language learners’ textbook, the best thing I can recommend is The Loom of Language, a splendidly opinionated old-school British socialist’s effort to show you how to take advantage of historical linguistics to minimise the effort involved in learning any Germanic or Romance language, written during World War II. It could use some updating and expanding, but it’s a great starting point.

  15. I admit that despite some early interest in historical linguistics I quickly turned away at university because (1) Transformational Grammar and synchronic linguistics in general were creating much more buzz and (2) I was very language-learning oriented, so learning about language wasn’t as interesting as learning languages themselves. Even structural linguistics was interesting for language-learning rather than for theoretical reasons. The only teacher of historical linguistics I can remember from university days was Chris Hauri of the German Department, whose lectures I didn’t really get into. And now I feel envious when m-l and Etienne rattle on about technical issues of sound change that I failed to familiarise myself when I had a chance. Concepts of sound change (in Chinese, for instance) go right over my head because I haven’t got either the framework or the habit of thinking that way.

  16. language documentation and description are as unfashionable as historical linguistics today
    Put those together and you’ve basically got my idea of (real) linguistics, which is why (for instance) Language Log, valuable as it is, usually doesn’t interest me very much. Theoretical linguistics, as practiced for the last half century, strikes me much like theoretical everything else (modern lit crit, for example, and any field where they use words like “problematizing”): more about intellectual one-upmanship than anything to do with reality. (No offense to theoretical linguists—I’m describing my own biases, which I do not confuse with an objective account.)

  17. Etienne – I had one german teacher who pointed out every unobvious english cognate, and wrote out the patterns of irregular verb stem vowel changes and said, figure out which verbs fit into which pattern. The other dozen just handed out long lists to memorize. I suspect the displaced linguists could easily do a better job teaching languages than the current batch of foreign language teachers. (but that’s a very small consolation, i’m sure.)
    Lameen – thanks for the recommendation, it sounds like a hoot if nothing else. I’ve requested it from the library.

  18. Yet when you ask students at the start of an introductory course what they find most interesting about language, the majority answer “the history of words” or something similar.
    This was me before I started university – in fact, I had planned to study linguistics. Then when I got there I discovered that there was little to nothing in the way of historical linguistics, and I promptly switched to political science (for those wondering, this was in the North American system, which generally allows that sort of change).

  19. “Jim: unlike you, I don’t think we’re coming to the end of a normal historical cycle. I think that we’re coming to the end of a bubble known as industrial civilization, and that the next few centuries will be much more reminescent of the Middle Ages than of the twentieth century.”
    Well there’s one cycle completed for you.

  20. marie-lucie says:

    Here are two more recommendations for accessible historical linguistics books: Jean Aitchison: Language change: progress or decay?, now in its 3rd edition. Among other books, she also wrote The articulate mammal, a general introduction to linguistics. The other one is Larry Trask: Historical linguistics which is more of a textbook but easily accessible to readers with basic linguistics. Trask was an American who became a topnotch Basque specialist, and he uses a lot of Basque examples in the book.

  21. J.W. Brewer says:

    I remember reading with pleasure The Loom of Language, which was floating around my local public library circa 1980, although I don’t know how much I’d get out of it if I read it now actually knowing more about the subject. I found the Deutscher “Unfolding” book rather frustrating; I wanted to like it more than I actually did, although ymmv.
    There’s perhaps a question also of who goes into historical linguistics because they’re interested in language as such versus because they’re interested in history or particular cultures illuminated by a particular language family. Thus, e.g., a recent biographical blurb for a lecture given by the woman who was my undergraduate teacher in historical linguistics (way back in ’85 or ’86) reads: “Stephanie Jamison is a professor of Asian Languages and Cultures and head of the program in Indo-European Studies at UCLA. She was trained as a historical and Indo-European linguist (PhD Yale 1977), but for many years she has concentrated on Indo-Iranian, especially (Vedic) Sanskrit and Middle Indo-Aryan languages and textual materials. She works not only on language and linguistics, but also literature and poetics, religion and law, mythology and ritual, and gender studies in these languages, and she is also interested in comparative mythology and poetics, especially with Greek materials.” Did she shift/drift away from Grimm’s Law to “gender studies” because of academic fashion, or did she always have interests tending in that direction but when she started out you couldn’t major in “South Asian Mythopoetic Gender Studies” and if you wanted to learn Sanskrit at all you had to go deal with the crotchety old historical linguistics people?

  22. Stephanie was a classmate of mine (I should have gotten my PhD in ’77 but settled for an MPhil instead); I don’t remember an interest in gender studies in those days, but I probably wouldn’t anyway—it wouldn’t have come up much. In grad school you’re pretty narrowly focused.

  23. marie-lucie says:

    JWB: The well-rounded person has many interests.

  24. @languagehat:

    Put those together and you’ve basically got my idea of (real) linguistics

    Just out of curiosity: wher would you put things like applied linguistics, sociolinguistics and linguistic anthropology? They’re not theoretical (certainly not formal), but they don’t fit neatly as language documentation and description either.

  25. They’re great, but they’re not what I think of when I think of the field and what got me interested in it. This has nothing to do with objective evaluation—I’m simply talking about my own private linguistics.

  26. marie-lucie says:

    Alon: applied linguistics, sociolinguistics and linguistic anthropology
    Plain “linguistics” deals with language in general, and with individual languages as instances of language. Applied linguistics uses a knowledge of linguistics for specific purposes, most often language teaching (although some people would place language documentation and description under the “applied” label). Sociolinguistics looks at languages as reflecting societies. Linguistic anthropology looks at languages from the point of view of anthropology, and anthropological linguistics looks at anthropology through the lens of linguistics (I don’t think there is much difference between these two specialties – it is more a matter of emphasizing one aspect or the other). In every case, people engaging in those other specialties should have at least a basic knowledge of linguistics in addition to training in anthropology, etc.

  27. @marie-lucie: thanks, though I wasn’t asking for the textbook explanation; I work in applied linguistics myself.
    I was just wondering what was the underlying criterion for lh’s evaluation, the point being that much current work in linguistics is decidedly empirical and safe from Chomskyan mumbo-jumbo.
    As for “plain linguistics”, I don’t think there can be such a thing. One can describe language and language use in terms of either the individual cognitive activity of speakers, or in terms of the social and historical dynamics of interaction (at a range of scales, from adjacency pairs in single conversations to the selective diffusion of linguistic innovations that drives language change). Ideally, a comprehensive description should be able to coordinate both elements: the cognitive structures of inclination and expectation, and the social structures of evaluation and classification. I don’t see the need to posit anything else.

  28. David Marjanović says:

    the extreme snobbery (quite beyond caricature) which seems obligatory in the latter

    Examples, please! :-)

    Jim: unlike you, I don’t think we’re coming to the end of a normal historical cycle. I think that we’re coming to the end of a bubble known as industrial civilization, and that the next few centuries will be much more reminescent of the Middle Ages than of the twentieth century.

    I agree there’s no cycle here, we’re in something qualitatively new. It’s definitely a possibility that industrial civilization will come to an end; that seems to depend mostly on how quickly we manage to get off of cheap oil. The best-case scenario is that everything continues with little change over the rest of the century (well, that’s already more than I dare make any predictions about), powered by such things as the huge advance in the efficiency of photovoltaic cells that was announced this week (on a lab scale, of course); the worst-case scenario is widespread famine and collapse while people try to evacuate, among other places, Bangladesh. Modern agriculture is “oil into potatoes”.

    One of the most popular French courses I have ever taught involved teaching students how to predict grammatical gender, something that is much more predictable than most French teachers realize, and a textbook introducing this would definitely fill a need.

    !!!
    !!!
    !!!
    That’s one of the bigger understatements I’ve read so far!
    All I know is 1) -e is often f. (like in German, and equally unreliable; of course from Latin -a except when it’s not); 2) -age, -ège, -ige, -au (but not eau), -ment, -isme are masculine; 3) m. is a sort of default (because the Latin n. merged into it); 4) sometimes I know Latin/Romance cognates (in which case la dent trips me up); 5) sometimes my native-speaker intuition gets me somewhere (fairly often non-cognates with about the same meaning have about the same gender… I think).

    If you’re looking for a historically-based language learners’ textbook, the best thing I can recommend is The Loom of Language, a splendidly opinionated old-school British socialist’s effort to show you how to take advantage of historical linguistics to minimise the effort involved in learning any Germanic or Romance language

    I’ve read the German translation (which replaces the chapter about German with one about English) at least twice. Socialist? Definitely old-school, as Frederick Bodmer had no compunctions about calling inflection “ballast” that languages should (!) get rid of, or calling the Cyrillic alphabet a cultural brake. It’s very comprehensive, though, in that it (briefly) presents many of the big language families of the world.
    It’s not a textbook for learning languages, though; it just tries to give you a broad base in order “to show you how to take advantage of historical linguistics to minimise the effort involved in learning any Germanic or Romance language”.

    I had one german teacher who pointed out every unobvious english cognate, and wrote out the patterns of irregular verb stem vowel changes and said, figure out which verbs fit into which pattern. The other dozen just handed out long lists to memorize.

    Rote memorization is how English irregular verbs are taught (to 11-year-olds) where I come from. Many have obvious German cognates, and often these are irregular in the same ways, but there are way too many exceptions for this to be reliable. Unobvious cognates were never pointed out; I found out only two days ago that teach is a cognate of zeigen “point/show”, and that only because I read it on Wikipedia. I hadn’t noticed in almost 20 years.

    I found the Deutscher “Unfolding” book rather frustrating

    Surprises me; I loved it (the German edition anyway).

  29. marie-lucie says:

    DM: -age, -ège, -ige, -au (but not eau), -ment, -isme are masculine
    There are some exceptions: the following come to mind:
    -age: cage; image; nage ‘swimming’
    -ige: pige ‘payment for a free-lance writer’; tige ‘stem’ of a plant, ‘thin bar’ esp of metal; and of course neige ‘snow’.
    -ment: jument ‘mare’
    teach is a cognate of zeigen “point/show”
    … and also of the “deic” in “deictic”, the technical linguistic term meaning approximately ‘demonstrative’, eg English this, that.

  30. David: welcome back!
    Examples of extreme snobbery…hmm, there was the specialist in Modern French literature who refused to believe that Modern spoken French had anything in common, typologically, with Algonquian languages. French is something you learn to become a Superior Person, you see, whereas Algonquian Languages are useless for that purpose, therefore French and Algonquian Languages cannot have anything in common. QED.
    I had enough horse sense NOT to quip that basic logic did not appear required for attaining Superior Person status. I said nothing, either, when I was asked by another “colleague” whether I had ever considered a post-modern approach to my work on historical linguistics, which as I had just told him with a straight face consisted of examing the impact upon Middle Provençal vocabulary of the infamous Welsh and Mongolian invasions (I couldn’t resist, I couldn’t resist, I just couldn’t! To quote a French comic strip character, “Je me terrifie moi-même quand j’ai des idées comme celle qui vient de faire gaonng dans mon cerveau traumatisé, mais si l’envie bête d’y résister me prenait, je suis sûr que ça me rendrait très malade.”)
    As for the fate of industrial civilization: much of my pessimism derives from my being a daily reader (though I have yet to comment) of THE OIL DRUM, an excellent blog on oil and energy issues, whose participants (scientists, engineers, retired oil industry specialists mostly) basically agree: we’re toast. Key notions I have learned there: EROEI (Energy Return On Energy Invested) and Jeffrey Browne’s Export Land Model. I recommend it unreservedly.
    As for French grammatical gender: frankly, it doesn’t strike me as more opaque than German grammatical gender assignment, and a textbook introducing these rules/principles for German L2 learners would also be a DESIDERATUM, I imagine.
    Some of the exceptions brought up by Marie-Lucie to the rules you knew can be accounted for by taking morphology into account: nouns ending in -age, for example, are always masculine IF -age is a separate morpheme, either a nominalizing element used to form nouns from verbs (“tapage”, from the verb “taper”, for example) or a derivational suffix forming abstract nouns from more concrete ones, whatever the gender of the base noun is (“dosage”, formed from “dose”, itself a feminine noun, for example).
    Finally: another set of cognates for “Teach/Zeigen” and the Greek root “deik” (ah, I can still remember learning how to conjugate it: deik-ny-mi “I point”, deik-s-o “I will point”…) is Latin “dicere”: Oscan “deikom” (with a different infinitival ending) preserves the Indo-European diphthong, but both Latin and Oscan share the innovation of having shifted the meaning to “to say”, still preserved in those Romance languages (most of them) which still have a reflex of this verb.

  31. marie-lucie says:

    Etienne, the specialist in Modern French literature who refused to believe that Modern spoken French had anything in common, typologically, with Algonquian languages
    It is vanishingly rare that literature specialists have any interest in linguistics (unless they had to learn some of it, perhaps like the Sanskrit/PIE specialist mentioned not too long ago, who publishes mostly on literatures in Sanskrit and related ancient languages). “Typology” doesn’t mean anything to them.
    I don’t know much about Algonquian, but some years ago someone wrote an article titled something like “The Chinook structure of French” (referring to the actual language Chinook, formerly spoken along the Columbia River, not the Chinook Jargon which started from a simplified form of Chinook). As traditionally transcribed (eg by Boas), Chinook seems to have very long verb forms, but this is because the verb stem is preceded by other morphemes which are uttered together with the stem, in a very similar manner to the order in French structures such as je te le donne ‘I give it to you’ (literally ‘I you it give’), which if written as most commonly spoken would be chteldonn or jetledonn. Similarly, possessed nouns work exactly as in Italian il mio padre ‘my father’ (literally ‘the my father’). Of course, these are typological resemblances only, the actual morphemes and the vocabulary are completely different from those of Romance languages but similar to those of some other languages of Western North America.

  32. It is vanishingly rare that literature specialists have any interest in linguistics … “Typology” doesn’t mean anything to them.
    Well, sure, but the proper response to being told something new to you is “Interesting, I’ll have to look into that,” not “Nonsense, I don’t already know that so I don’t believe it.” I’m pretty sure Etienne wasn’t expecting the professor to already have considered the likeness but was objecting to his rejecting it out of hand.

  33. marie-lucie says:

    LH, I agree with your interpretation. I meant that most literature people are totally uninterested in linguistics (or in languages other than the one they teach) and don’t want to hear anything about the topic, especially if they have to encounter a technical term or concept that they haven’t already run into. Adding to this negative attitude toward a different discipline the scholar is not familiar with is the element of cultural snobbery and prejudice. So of course, a linguist’s comment about resemblances of a very general nature between French and Algonquian would be rejected out of hand as unworthy of consideration by the snobbish professor ignorant of linguistics and of any aboriginal language.
    Alon, my comment was not directed only at you but also to other readers. By “plain linguistics” I mean what is also called “core linguistics”, dealing with the structures which are peculiar to language (sounds, morphemes, words, sentences etc). Without a grounding in these basic features of language, attention to the cognitive or social aspects can only remain vague.

  34. “whether I had ever considered a post-modern approach-”
    I am so going to save this for the next time I need to make a serious comment on someone else’s total nonsense. It might be trickier to find the right occasion in finance, though. (I’ve seen some creative approaches to accounting, but i’m not sure if they qualify as Post-Modern.)
    Just out of curiousity, what would a post-modern approach to the influence of welsh and mongolian on middle provencal look like? Would you even need the supposed invasions to have occured?

  35. Of course not! It is precisely the absence of such invasions that leaves its traces everywhere!

  36. Hat: you’re right: to this professor the notion that linguistic structure and linguistic prestige are separate issues simply wasn’t conceivable. And inasmuch as he recognized that Algonquian languages are indeed *languages* and not some bruttish simplistic subhuman jargon, he was quite ahead of the pack, actually.
    To s/o: As for what a post-modern study of the linguistic impact of the Welsh and Mongolian invasions on Middle Provençal would look like: I didn’t ask, and if one were thrust into my hands I am not sure I would read it: my sense of absurdist humor has its limits.
    Happily, the professor in question is now the Department chair, so I doubt he’ll have the time to produce any such study.
    (I’m a fan of the comic strip DILBERT, and the above does seem covered by “The Dilbert Principle”).

  37. You were right not to ask: you might get an answer. I’m sure his days are happily filled with department paperwork.

  38. @marie-lucie:

    By “plain linguistics” I mean what is also called “core linguistics”, dealing with the structures which are peculiar to language (sounds, morphemes, words, sentences etc).

    Thing is, I don’t think any of these structures can be described other than in cognitive or social terms.
    Impromptu spoken language, for example, does not really have anything resembling sentences; it has more-or-less structured periods, with much looser grammatical constraints. I’d argue that the concept of sentence is the result of a specific social technology (writing) and the ways in which it has been used, and I think it’s seriously misleading to consider it a linguistic primitive.
    Phonemes provide another example. While the structuralist approach of Trubetzkoy is a very useful approximation, there is no way to reduce phonology to its articulatory aspect. Any model of our knowledge of sound events requires the inclusion of compositional mental representations of these events, and is therefore essentially cognitive.

  39. Alon – what about formal spoken language – rhetoric and oral poetry or story-telling?

  40. marie-lucie says:

    Alon: I’d argue that the concept of sentence is the result of a specific social technology (writing)
    Have you read any transcriptions from hitherto unwritten languages? grammars of such languages? There are plenty of sentences there. And contrary to what some linguists say, most people do speak in sentences, even if they don’t write very much.
    there is no way to reduce phonology to its articulatory aspect
    True, you can’t reduce a sound system to the actual sound production, described by phonetics, but you do need to take those sounds into account.

  41. marie-lucie says:

    Back to French word endings usually indicating gender, with exceptions when age is part of the stem:
    I can’t believe I forgot la page ‘the page (of a book)’. But there is also le page ‘the page (basically a young “go-fer”)’.

  42. @s/o: formal spoken language of that kind is modelled on written language. It’s true, as marie-lucie says, that people do speak in sentences; that is, some people do some of the time, when they consciously monitor their speech production to accommodate social constraints that equate articulate expression with strict grammatical well-formedness.
    The point is, if you look at language empirically and not through the filter of this-goes-against-linguistic-theory, you can’t consider grammatical well-formedness to be a basic concept (a primitive notion, in axiomatic terms). Most language isn’t well-formed. Now, it is certainly a fact that we monitor our own and each other’s grammar, but I don’t see any well-founded reason to describe this in fundamentally different terms than our monitoring of r-dropping or develarization of final nasals. Both are variable, contextual, and probabilistic.
    @marie-lucie:

    you do need to take those sounds into account

    If I gave the impression I was arguing otherwise, my apologies. I would certainly agree that need to describe the acoustics of it, insofar as the characteristics of our perceptual and articulatory resources are key constraints on phonology. But this does not change the fact that phonological representations are cognitive constructs, not objective acoustic properties.
    You argue that “without a grounding in these basic features of language, attention to the cognitive or social aspects can only remain vague”. I’d be inclined to turn that on its head: without being informed by what we do know about human cognition and interaction, models of language structure can only remain rigid, artificial and inaccurate— much like Chomsky’s ideal speaker with infinite cognitive resources and a perfectly homogeneous community.

  43. David Marjanović says:
    DM: -age, -ège, -ige, -au (but not eau), -ment, -isme are masculine

    There are some exceptions: the following come to mind:
    Yes, the rule only holds for actual morphological endings (assuming one can recognize them, but that’s usually feasible).

    “Je me terrifie moi-même quand j’ai des idées comme celle qui vient de faire gaonng dans mon cerveau traumatisé, mais si l’envie bête d’y résister me prenait, je suis sûr que ça me rendrait très malade.”

    :-) :-) :-)

    As for French grammatical gender: frankly, it doesn’t strike me as more opaque than German grammatical gender assignment

    Heh. I never said it was! :-) Gender is one of the reasons I’m so glad I’m a native speaker of German and therefore don’t need to learn it.

    Impromptu spoken language, for example, does not really have anything resembling sentences; it has more-or-less structured periods, with much looser grammatical constraints. I’d argue that the concept of sentence is the result of a specific social technology (writing) and the ways in which it has been used, and I think it’s seriously misleading to consider it a linguistic primitive.

    Heh. Spoken German may not really have sentences, but it most definitely has clauses and distinguishes different kinds of them by word order. Where within or around a consonant cluster it puts its syllable boundaries, on the other hand, or whether that’s even a concept that makes sense, is another question – I’m often tempted to consider syllables an evil Sumerian invention by which they tried to make the rest of us feel guilty for having languages with identifiable relatives.
    And indeed, the Serbocroatian spelling rules (as of 1960 or so) leave the choice to the writer of whether to separate ze-mlja or zem-lja (lj being considered a single letter and representing a single sound).

    I don’t see any well-founded reason to describe this in fundamentally different terms than our monitoring of r-dropping or develarization of final nasals. Both are variable, contextual, and probabilistic.

    …You seem to be talking about American sociolects. There’s nothing variable about my r-dropping.

  44. David, does Austria have a policy on Eszett? I’m reading a book that says that Germany is kind of deciding on it by region, that many people have gone back to using ‘ss’.

Speak Your Mind

*