THE MOST INTERESTING LANGUAGE.

Over at linguaphiles, ein_wunderkind asks: “What is the most interesting language you know of and why? I’m bored and need some reading material.” The mischievous Anatoly, from whom I got the link, answers “English” (“Just about the most weird-ass grammar out there, a vocabulary that reads like a Wal-Mart shopping list, and don’t even get me started on spelling, ‘cuz maaaaan!”), but there is discussion of the intricacies of languages like Navajo, Hebrew, and Tibetan (“This is how you can have, for instance, words that are spelled bka’ brgyud or ‘bras spungs and pronounced, respectively, [kacy] and [tʂɛpuŋ]“). Makes me want to start learning another language.

Comments

  1. Arizonabren says:

    What about OLDE English? It can be a tongue twister with ancestry…

  2. Modern Hebrew is just about as boring as a language can get, probably as it’s a “common denominator” of so many contributing tongues: no special consonants, except maybe the ‘r’; standard five-vowel system; run-of-the-mill phonology; no syntactic features to speak of; only two Cases; and a dwindling ancient-Hebrew-based morphology which is hardly productive, as opposed to Arabic for example.

  3. The original question is “what is the most interesting language”, but on his Russian blog Anatoly says English is the most difficult (самый тяжелый). I’m not sure I’d say most difficult is the most interesting. I might vote for Japanese as the language that provides the most intellectual fun (although it’s a near thing), but I think both Arabic or Chinese are more difficult than Japanese for native English speakers.

  4. I pulled this quote from the comments at linguaphile to make the descriptivists smile -”I’d say English grammar is so poorly taught, at least in the United States, that many non-native speakers have a better grasp of grammar than native ones.”

  5. Vanya, why do you think that “Chinese [is] more difficult than Japanese for native English speakers” ? I speak only a little Mandarin, and even less Japanese, but my understanding is that the rules for knowing how to express something politely in Japanese are very much more complex than the same concept in Chinese.

  6. “but my understanding is that the rules for knowing how to express something politely in Japanese are very much more complex than the same concept in Chinese.”
    The way to sound polite in Chinese is lexical, so there aren’t a lot of rules. You use a lot of set phrases, try to sound as Classical as possible, and hope they guess you’re foreigner and smile rather than sniff at you.

  7. >I pulled this quote from the comments at linguaphile to make the descriptivists smile -”I’d say English grammar is so poorly taught, at least in the United States, that many non-native speakers have a better grasp of grammar than native ones.”
    As a descriptivist I have no quarrel with that assertion. It’s quite often that you hear from speakers of any language that they didn’t understand the principles of syntax, morphology et al at all until they formally studied some foreign language. Likewise if you ask most native speakers to explain some construction or to provide guidelines for what is grammatical or not, they usually will have no analytical insight for you, short of repeating different permutations of the phrase and then saying ‘yeah, you can say this, but not that’.
    Unless of course you were responding to the ‘grammar is poorly taught’ part, as a descriptivist sentiment, in which case I agree! Though I would argue that in the general case the state of English grammar instruction in America is not a case of descriptivism vs. prescriptivism as much as it is a case of none at all, or at best a set of blanket proscriptions, versus even the most cursory explanation of the basic functions of English.

  8. Phillip, I think the rules for expressing things politely are actually subtle and complex in most languages. Japanese is probably more complex than most, but that’s only one aspect of the language. In my experience it’s much easier to achieve basic reading fluency in Japanese than in Chinese. Japanese pronunciation is also easier for Westerners, even ignoring the whole issue of tones. And I don’t find Japanese grammar that daunting – at least the rules are fairly regular. Learning Russian or Arabic grammar is much harder.

  9. ZD, I interpreted that statement, in its full context, as meaning foreigners speak “better” English because they know the explicit grammatical rules and can produce textbook correct sentences. Do I have a deeper understanding of German than a native because I can use the genitive correctly and many native German speakers don’t? That seems like a nonsensical statement to me. I suppose it hinges on whether one takes grammar to mean a) the actual morphology and syntax of the language, or b) the set of formalized rules describing those features. I agree that a non-native speaker in many cases will have a better understanding of b) than a native speaker.

  10. Vanya, I would wager that di_glossia was indeed talking about b). After all, they go on to say ‘Exactly. What Azza’s going on about are approaches to learning and understanding the English language that a native speaker wouldn’t necessarily know.’ and ‘The average English speaker, in the US at least, would not be able to identify it as such. To the average speaker, English has one article.’
    The latter assertion, in particular, while I don’t really agree with it (I think that if I went outside and asked ten people how many articles English has, I would get ten blank looks), seems more likely to indicate that native speakers probably don’t understand how many articles one should say the English language has, rather than, you know, repeatedly fail to use English articles correctly.

  11. A traveller’s handbook called “Culture Shock: Japan” (part of a series) once told me that the only people capable of mastering Japanese politeness forms are Indonesians.
    I thin that the problem people have with American politeness rules is that a lot of ostentatious informality is mixed in with the various American taboos, so someone might conclude that there are no rules.

  12. (especially because they’re talking about ‘the’ plural and ‘the’ singular, which are of course impossible to mix up in speech. Their interlocutor points out that the two articles are ‘a’ and ‘the’, not ‘the’ and ‘the’, but I think di_glossia’s lapse on that count still indicates that they’re talking about an inability to recite parts of speech and the like.
    I’ve just myself been reading Coulmas on writing, and he’s done a good enough job of convincing me that linguistics as it is practiced is generally quite scriptist and thus inconsistent in its claimed scope, such that today I can’t find myself even wanting to come down concretely on how many articles English does have, or if ‘article’ is a meaningful word, or if there are such things as words at all.

  13. (especially because they’re talking about ‘the’ plural and ‘the’ singular, which are of course impossible to mix up in speech. Their interlocutor points out that the two articles are ‘a’ and ‘the’, not ‘the’ and ‘the’, but I think di_glossia’s lapse on that count still indicates that they’re talking about an inability to recite parts of speech and the like. )
    I’ve just myself been reading Coulmas on writing, and he’s done a good enough job of convincing me that linguistics as it is practiced is generally quite scriptist and thus inconsistent in its claimed scope, such that today I can’t find myself even wanting to come down concretely on how many articles English does have, or if ‘article’ is a meaningful word, or if there are such things as words at all.

  14. Anne Collins says:

    I notice that non native Americans got a better English grammar.

  15. komfo,amonan says:

    The paucity of OSV languages makes them interesting to me. I once undertook to count them, and discovered twelve. All are tropical except one. That would be Haida. Maybe I should learn it; it only has 50 speakers.

  16. marie-lucie says:

    many non-native speakers have a better grasp of grammar than native ones.
    I think that this is likely to be true with languages that have several registers: foreigners are taught the high register, which natives have had to study in school while they themselves speak in a colloquial manner or with frowned-upon regional features. So foreigners might speak “too well”, meaning in a register that is too high for the occasion. Or they might speak in an archaic style, if they have been studying the literature of past centuries but have had little contact with actual live speakers in natural situations. But if you are teaching your own language it is hard to avoid using a high register, because you have to speak more slowly and deliberately than usual and that triggers features of the high register. For instance, in teaching ESL you might articulate “I am going to …” while you normally say “I’m gonna …”, because for one thing you have been taught the full form, and that is what the students will encounter in writing, and for another, you cannot very well say “I’m gonna” very slowly and still sound natural.

  17. David Marjanović says:

    The paucity of OSV languages makes them interesting to me. I once undertook to count them, and discovered twelve. All are tropical except one. That would be Haida. Maybe I should learn it; it only has 50 speakers.

    In principle it is now feasible to learn it at home, I suppose, because it’s very well documented (a huge dictionary came out in 2004). Of course, it has full-on polysynthetic grammar, full-on Pacific Northwest sound system (not as extreme as Tlingit, but still), and so on and so forth. If you like a challenge… :-)

  18. the only people capable of mastering Japanese politeness forms are Indonesians.
    And not all Indonesians either – they meant the Javanese, who also have a bewildering number of politness levels.
    The dirty secret is that even most modern Japanese don’t “master” all the politeness forms, and sometimes have to go to special classes to learn to speak keigo. This is why I think foreigners can make too big a deal of how complex Japanese politeness forms are. In everyday life for most Japanese it’s not as big a deal as is often portrayed. As a foreigner you probably won’t get every nuance correct even if you marry a Japanese national and live there for 20 years. But you could say the same about Russian humor, religious allusions in Arab or Chinese chéngyŭ. I think this more a cultural issue than linguistic.

  19. A few miscellaneous facts about Hebrew to defend it as “interesting”, pace Yuval.
    A small set of consonants, probably historically pharyngeals, tend to a-color adjacent vowels. These days some of them have merged with non-pharyngeal partners, and are now distinguishable only by the a-coloring.
    The old velar/postvelar contrast between /k/ and /q/ has been lost, except in the Yemenite dialect. However, /k/ lenits to /x/ postvocalically, and /q/ doesn’t. There is another /x/, from an old pharyngeal, that has merged with lenited /k/, but it triggers a-coloring and lenited /k/ doesn’t.
    Many transitive verb-stems take a derivational prefix /hit-/ to form reflexives or reciprocals. The /t/ metathesizes with a small class of (also alveolar) initial consonants, so /hit/+/tsa’er/ gives /hits-ta’er/.
    Hebrew spelling is tricky, since it’s deficient in some vowels, and represents the language before some lenitions (so postvocalic P is sometimes pronounced [f], and so on). According to some of the rules, [v] is often spelled WW, to distinguish it from [u], spelled W. The name “Yuval” ought to be spelled YWWWL, which would be charming, but alas, an adjustment rule reduces triple letters to doubles, so I think it’s just spelled YWWL.
    Hebrew has the most aggressive gender-agreement rules I know. Not only the third-person pronouns, but also the second-person ones, must agree with the gender of the referent; with person, gender, and number agreement there are ten pronoun sets. (Only ten, because even Hebrew has only an epicene first-person pronoun.)
    The verb agreement is different in the three major tenses. In the present tense, verbs agree with the subject only in gender and number; the four possibilities are neatly marked with four suffixes. But in the past tense, there are ten suffixes to match the whole pronoun repertoire, and in the future there are ten circumfixes (which can almost be factored into semantic components). In the present tense (but only there and in the imperative), verbs must agree with even first-person subjects.
    There is a preposition for marking objects of transitive verbs, but only when they are definite. Indefinite objects are not thus marked.
    Instead of a genitive case for marking the possessum, Hebrew (sometimes) puts the possessor in a special “ungenitive” case, traditionally called the “constructive”. (/bajit/ “house”, /lexem/ “bread”, /bejt lexem/ “bakery, Bethlehem”). When the possessor is a pronoun, it glues onto the constructive and changes form radically, producing ten possessive suffixes.

  20. Coulmas sounds cool.
    Another book that should have been written years ago (you can read that as “that I should have written years ago but got beaten to it” if you want).

  21. I need hardly state that I agree with ACW. Hebrew has had a great many phoneme splits and mergers, and since the root-and-pattern system means that almost every inflectional paradigm exposes almost every sound to almost every possible environment, many of those splits and mergers result in weird patterns of halfway-allophony. (As for the a-coloring that ACW mentions, I like to point out that there are three verbs /kara/, all of them perfectly regular, all of them having different infinitives: /likro/, /likrot/, and /likroa/. Fortunately the verbs are spelled differently, at least.)
    But really, lots of languages are interesting. Take French, for example, where there’s a long tradition of fiddling with the external sandhi in whatever way sounds best to you and most frustrates people from the generation before you. ;-)

  22. marie-lucie says:

    Ran: Take French, for example, where there’s a long tradition of fiddling with the external sandhi in whatever way sounds best to you and most frustrates people from the generation before you.
    Tantalizing. How about some examples?

  23. Well, actual examples make it sound prosaic; where’s the fun in that? I really just meant the steady decline in use of liaison, except that when people are trying to speak in a higher register, they’ll sometimes insert it in places where it doesn’t fit (even going to such extremes as */mwa.zo.si/ for “moi aussi”, at least if Wikipédia is to be believed; I haven’t heard that one myself, though I have heard some fairly similar ones).

  24. marie-lucie says:

    Ran, what else is new? your examples have been developing for generations, so it is not as if each generation is finding new ways of “fiddling with the external sandhi” in order to “frustrate people from the generation before you”.

  25. It doesn’t have to be new, only interesting! Though I will grant that if there’s a language whose sandhi changes in an entirely new direction every generation or so, then that would probably be more interesting than the situation in French.

  26. What is the most interesting language you know of and why?
    What does “interesting” mean in this question?
    Are there people smart enough ‘to know’ languages which are not “interesting”?

  27. Sorry for the self-reference, but in fact it is not just that as I have quoted there the opinion of someone else on the most entertaining language of the world.

  28. Marie-Lucie : do people really say “I’m gonna” outside of Western/”Cowboy” films ? I had a young Chinese correspondent who used to drive me to distraction by using exactly that form in his written e-mails, but I assumed (perhaps naively) that this was more an attempt to sound “cool” than an accurate reflection of his (or others’) pronunciation. On the other hand, I would not use “I am going to” in a TEFL situation except when dealing with complete beginnings : “I’m going to” is, for me at least, appropriate in all situations.

  29. Philip, I think almost every American regularly contracts “I’m going to” in normal speech even if they don’t recognize they’re doing it. “Cowboy” speech to me is “I’m gunna” with a clear drawn out vowel after the g. In normal educated northeast American that vowel becomes an almost inaudible schwa – I’mgna.

  30. do people really say “I’m gonna” outside of Western/”Cowboy” films ?
    Sure thing! In colloquial Texas since at least the 60s and still going strong.

  31. I think almost every American regularly contracts “I’m going to” in normal speech
    That was also my memory of things, but I didn’t want to risk the claim beyond Texas.
    In normal educated northeast American that vowel becomes an almost inaudible schwa – I’mgna.
    I’mgna is rather fast. In Texas it can be much slower: I’M – GUHNna. This is performed tempo quasi drawlando, ma senza gli vocali di cowboy.

  32. Imunna or Ahmunna (silent g) is also common.

  33. “do people really say “I’m gonna” outside of Western/”Cowboy” films ?” I thought they said “Ah ma gonna”. No?

  34. Very common in Australia, pronounced the same as “gunner” (non-rhotic).
    However, I think people write it much less than they say it. The Chinese student picked up a good colloquial expression and overused it in writing, which is why it stuck out like a sore thumb.

  35. When I useta chair the admissions committee for our PhD program, I would sometimes be startled by “wanna” for “want to” in emails from students in Korea and (or?) China expressing interest in applying to the program. I assumed that they picked up the expression on the internet and had no clue how informal it was.

  36. There’s two very common—pretty much obligatory—shortenings of ‘I am going to’ in American English. The first one is, “I’m gonna”, pronounced /aɪm gənə/ (I honestly can’t tell which of the first two syllables takes primary stress, so it might wander depending on context. But the final schwa is always unstressed). The second vowel is rarely omitted entirely, except in hurried speech, but it is normally quite short. If the speaker is going to shorten it further they’re more likely to omit the ‘I’ than the ‘gon’, leading to “‘mgonna”. “I’mgna”, I can’t work out any natural reconciliation of that with the spoken vernacular.
    I would say the second form is a continuum between “Imunna or Ahmunna (silent g)” and “I’m'a”. I’m'a, especially, is a noted feature of Urban or AAVE, and even though it’s quite truncated it’s not necessarily associated with hurried speech. It is considered substandard, though, while ‘I’m gonna’ can only be considered widespread in every register of American English. I would imagine that outside of specific circumstances (eg speech-giving) I would hear “I am going to” as a marked utterance, a spelling pronunciation to indicate deliberation and emphasis, akin to to pronouncing ‘the’ as ‘thee’ or ‘a’ as ‘ay’.

  37. marie-lucie says:

    ZD: “I’m gonna”, pronounced /aɪm gənə/
    Are you sure about the [aɪm]? I don’t think I ever hear a diphthong, or indeed a stress: “‘mgonna” is more like it. [aɪm] would fit in more with the full form “going to”. Of course, I am not familiar with many regional varieties, even in Canada, let alone the US.
    Chinese, etc students using “gonna, wanna”: the internet is not the only place where one would see this, much written dialogue in novels also uses it. And the students may have been taught to say those shortened forms, and to recognize them in movies, for instance. But they probably were not taught to identify different registers.

  38. It’s a fair point on the diphthong. I actually do think that I hear it, though, sometimes; indeed I think it’s a testament to just how standard ‘gonna’ is that in measured, even speech you would keep your ‘gonna’ even as you pronounced “I’m” in full.
    Moving down the hurried/informal/vernacular line, I do hear other diphthongs as well: /aɪ/ as /əɪ/. It probably shakes out to a continuum from /ai/ to (/a/ or /əɪ/) to /ə/ to /ø/.
    As for stress, I was ready to give it to the ‘guh’ but I am hearing counterexamples. For instance, in the common sequence, “Yeah, I’m gonna…” Pronounced, roughly, “Y:ahm gonna”, “Yeah-I’m” takes the stress and both syllables of “gonna” are unstressed. Which is dicey—but in that construction you can also drop the “Yeah”, but maintain the / – - stress pattern, for a similar conversational tone.
    For instance, if you and I were at a board meeting and you just gave a talk about your opinions of the market, and I thought you were full of shit, I might muster up all my politeness, raise a hand, and say “I’m gonna disagree’ (or “I’m gonna hafta disagree”), stress on the “I’m” (which is a diphthong if not with a full /a/). I would argue that this is different from standard topicalizing stress, because the notion that someone else would be disagreeing is not really a possibility.

  39. I pulled this quote from the comments at linguaphile to make the descriptivists smile -”I’d say English grammar is so poorly taught, at least in the United States, that many non-native speakers have a better grasp of grammar than native ones
    When my father taught French in Australia just after WWII, he said he had to teach his students English grammar before they could understand French grammar. Plus ca change …
    Middle Europeans who came to the UK before the start of WWII mostly spoke notably grammatical and beautifully accented English, though there was usually a tiny trace of accent – or perhaps their precision – that identified them as not native speakers.

  40. “do people really say “I’m gonna” outside of Western/”Cowboy” films ?” I thought they said “Ah ma gonna”. No?
    You betcha. They also say “didja” for “did you”, and a lot of other things. “Ah ma gonna” I think is “I’m a-gonna” which I associate with southern or hillbilly talk. Instead of “I’m going to” they also might say “I’m fixing to” or “Ahm a-fixin’ tuh.” What about the phrase “he went thataway.” Not sure why “a-” gets tacked to the front of words like that–it’s not my dialect. You can hear the late Mary Travers do something similar here at about 1:58, singing “Lord I can’t go a-home this a-away.” I play this for my students when we study numbers, but I never know what to tell them about “go-a”. (For those discussing how to pronounce “I’m”, she sings it several times throughout the song–she grew up in New York.)

  41. Facebook friends have reported that their kids think that “ginormous” is a normal word meaning “very big”.

  42. komfo,amonan says:

    When my father taught French in Australia just after WWII, he said he had to teach his students English grammar before they could understand French grammar. Plus ca change …
    My experience was different. I learned French (& French grammar) before I formally learned English grammar. It was a helpful sequence for me. I was perhaps a freakish child.
    I’m trying to think of words like “ginormous” that have become mainstream. But I don’t know what I mean by “words like ‘ginormous’”. Humorous coinages? Humorous coinages made by splicing two “normal” words together?

  43. Well, you’re talking about pormanteaus of course, which are many. But what makes a portmanteau humorous, rather than regular? There does seem to be a line between ludic coinages and natural coinages like ‘Spanglish’ or ‘Gerrymander’ (both examples from Wikipedia).
    The first thing to note is that both Spanglish and Gerrymander are themselves slightly humorous, and I think it’s fair to surmise that ‘Gerrymander’ started life even more humorously, when people knew what it was a reference to. It’s quite possibly fair to say that every portmanteau is going to sound a little humorous as long as it’s still recognized as such.
    Beyond that, though, I might propose that one thing that makes a certain blend more humorous would be a lack of utility. In both the cases of Spanglish and Gerrymander, the words are (at least initially) recognized as blends of two distinct components, whose dual and distinct meanings comprise (up to the entirety of) the meaning of the new word. Spanglish, or the rather brutal ‘skort’ and ‘shacket’, all succintly and transparently indicate that this new thing is a direct mixture of two components—Spanish and English, skirt and short(s), shirt and jacket.
    ‘Ginormous’ on the other hand more immediately conveys a ludic approach. It’s a silly sounding word, but you can’t sound much sillier than ‘skort’ (modern fashion may also be an area of the language more amenable than most to new coinages, in parallel to new styles). I would propose, then, one possibility is that unlike Spanglish, the two components of ‘ginormous’ overlap heavily in meaning, if they aren’t synonyms. Thus there’s no apparent necessity or work done by the coinage, aside from poetic variation and non-literal emphasis (to say that something is both enormous and gigantic and therefore BIGGER for the synonymy is not really how the words work, but it makes sense on an aesthetic level). That the portmanteau has no apparent logical reason for existing may go a long way towards outlining it and its fellows’ place in the firmament.

  44. David Marjanović says:

    This seems to be the best place to mention how close French has (so far) come towards being a topic-and-comment language like Chinese, Japanese, Quechua and many others outside Europe.

  45. David Marjanović says:

    “Ginormous” is weaksauce. Try gihugrongous. That’s what the people who work on the really big dinosaurs use.

  46. michael farris says:

    “Not sure why “a-” gets tacked to the front of words like that–it’s not my dialect.”
    It is part of my dialect (not crucial to it, but it’s there) and sometimes does change the meaning though it’s hard to explain how. Compare:
    He was just sittin’ there. (not doing anything else)
    He was just a’ sittin’ there. (and having a fine time doing so).
    I have the idea (though I might be wrong) that the a’ that occurs before some -ing words was at one time the preposition ‘at’ so that modern a’ sitting at one time was ‘at sitting’ and that this might be a calque borrowing from some Celtic language or other.
    I hope and trust that if anyone here knows better they will enlighten us all.

  47. marie-lucie says:

    ginormous
    This is a new word to me. Has it replaced “humongous”?
    He is a-sitting, Froggy went a-courting, etc
    I don’t have a reference at hand, but I believe that he prefix a- in such cases is not from at, but from on. Similarly words like aboard once started with on. The two “on”s were not the same, any more than they are in “what’s going on” versus “it’s on the table”.

  48. Re: a- before gerund-participles: The OED Online seems to agree with Marie-Lucie. At least, it seems to cover it in A, prep.1 (section II, “With a verbal noun or gerund, forming part of a verbal expression. (Now usually written with a hyphen or as one word with the verbal noun.)”, with two senses and a slew of subsenses), whose etymology begins, “Variant of ON prep. with loss of the final consonant -n, reflecting an unstressed pronunciation of the word in proclitic use; compare an, variant of ON prep., and also o, variant of ON prep.” Though I must admit, I find the array of a- prefixes and prepositions quite bewildering. The mind boggles at how the OED people sorted it all out.

  49. ginormous has been around for over a decade if not longer – same meaning as humungous
    gunna is a feature of normal colloquial spoken NZ english – outside of literature you’re unlikely to see it written
    When I was teaching, I found it interesting seeing what spellings students came up with when they didn’t know how to spell a word. One of the classics was quilty for quality (which probably indicates the speed of spoken newzild englsih more than anything else).

  50. shacket: lubricious noise made while living together

  51. He is a-sitting, Froggy went a-courting, etc
    I don’t have a reference at hand, but I believe that he prefix a- in such cases is not from at, but from on.
    This whole thing was explained in a comment by an erudite person, very recently. Unfortunately I can’t remember who or when.

  52. “ginormous has been around for over a decade if not longer – same meaning as humungous”
    I’ve heard this word since my childhood (in London) in the ’70s – used frequently by my mother, so I’m sure it goes back quite some time before that. Despite its longevity, it still doesn’t seem to have the status of Real Word (at least, not in certain circles…).

  53. marie-lucie says:

    Well, perhaps ginormous hasn’t made it to Canada! (unless Canadian readers pop up with examples). Google apparently has 229,000 citations for the word, and it seems to date (officially) from 1948, but World Wide Words says it arose in WW2 military slang, so it is definitely older than humongous.
    I lead a sheltered life.

  54. Yo Z.D. Smith: I’m really happy for you, and I’m'a let you finish, but woo’t is the the best contraction of all time. Of All Time!
    … Well *someone* had to go there…

  55. Interesting language for wunderkind might be one of the Salishan languages – exotic alignment and other syntactic features, alarming phonologies, fun lexicon and derivative processes and so far the tourists haven’t found them and ruined everything.

  56. marie-lucie says:

    Their neighbours the Wakashan languages have similar structural features and even more alarming phonologies. Unfortunately all these languages are on the severely endangered list.

  57. perhaps ginormous hasn’t made it to Canada!
    Not Wobegon either, maybe it’s British.

  58. I once met a grad student who did fieldwork on an Athabaskan language, and when I asked her what had drawn her to Athabaskan linguistics, she lyrically answered: “How can you not love a language with twenty slot positions for affixes, phonemic tone, and glottalized affricates?”. I could see her point: these languages are as interesting/exotic (=typologically similar to) as Salish or Wakashan, and as a bonus, several are not immediately endangered.

  59. marie-lucie says:

    Etienne, are you suggesting that the Athabaskan languages are typologically similar to the Salishan and Wakashan ones? or do you mean that they are typologically similar to each other within the family? I find your sentence ambiguous.

  60. Ginormous can definitely be found in America, if you know where to look. It’s just, as mentioned, something of a ludic usage, so it’s, I’m sure, more directly pegged to certain demographics than others.

  61. The most interesting had better be the one you’re working on, if you’re to get anywhere with it.

  62. marie-lucie says:

    Panglott, as you work on it, it usually gets more and more interesting. If not, you are probably not doing it right.

  63. “Unfortunately all these languages are on the severely endangered list. ”
    So one more speaker would be a real gain. And an outsider learning the language woudl set apowerful example to the young’ns.
    I think Etienne just meant they are equally exotic and intricate. For me, I have to admit that I have a lot more interest in Salishan and Wakashan because the cultures interest me more and I like the landscapes better. The Skagit Valley is a nice place ot be studying anything.

  64. marie-lucie says:

    Jim: an outsider learning the language woudl set apowerful example to the young’ns.
    But also powerful feelings of envy and resentment from some, both young and old.
    But go ahead, Jim! I don’t want to discourage you, and some of the old people will be pleased.

  65. Marie-Lucie: I meant typologically similar to Salish and Wakashan, and thereby quite “exotic”, i.e. unlike European languages.

  66. marie-lucie says:

    Etienne, to my knowledge, Salishan and Wakashan are typologically similar, but Athabaskan languages are quite distinctive.

  67. Marie-Lucie: Athabaskan languages are treated as being (to some degree) part of the Pacific Northwest SPRACHBUND in the reference books I have consulted. Hence I take it for granted that Athabaskan, Salish and Wakashan are closer to one another, typologically, than any of them are to European languages. I accept your statement that Salish and Wakashan are closer to one another than either is to Athabaskan, but that does not negate my earlier statement.

  68. marie-lucie says:

    Etienne, I am very surprised to see the Athabaksan languages considered as part of the Pacific Northwest Sprachbund. Certainly they are not close at all to European languages (and they have recently be shown to be related to the Yeniseian languages of Siberia, the last living exponent of which is Ket)(see recent work by Edward Vajda). But the bulk of the Athabaskan languages are in Alaska and the Northwest Territories, with a few isolated spots along the Pacific Coast until another cluster of languages in the Southwest US (Navaho, Apache, etc). Everywhere they occur, they are characterized by a very distinctive overall structure, although they may have borrowed (or exchanged) a few features from neighbouring languages. I know of no case where there is a dispute about whether a certain language belongs to the group or not, as there is with some other large hypothesized groups. The areal commonalities described for instance in Mithun 1999 (p. 312, 314-315) are fairly superficial and do not affect the general structural pattern of these languages.

  69. Marie-Lucie: my source was Thomson & Kinkade’s article “Languages”, in the volume on the Pacific Northwest of the HANDBOOK OF NATIVE NORTH AMERICAN LANGUAGS (the common features of the area are listed on pages 42-44 of the article).
    The common structural features (polysynthesis, widespread use of complex morphophonemic alternations) are fairly general in nature (meaning that the Athabaskan/Pacific Northwest similarities could be coincidental), I agree. (Though they do stand out if you take European languages as your basic template).
    On the other hand, the consonant system of Athabaskan (with its distinctive use of glottalization with affricates and lateral consonants) certainly is too similar to its neighbours’ and too typologically unusual for coincidence to be a reasonable explanation. As I understand it, Proto-Athabaskan had a “Pacific Northwest”-like consonant system: the Proto-Athabaskan URHEIMAT, if Kraus (1979) is to be believed, was in Alaska, i.e. away from the center of the Pacific Northwest SPRACHBUND today.
    The discontinuous spread of Athabaskan suggests that this family must once have had a more widespread distribution along the entire North American Pacific coast. Indeed, an Athabaskan “enclave language”, Nicola, once wholly surrounded by Salish languages, only disappeared comparatively recently.
    Thus, if (inter alia) Salish expanded at the expense of Athabaskan, I wonder whether the phonological features of the Pacific Northwest SPRACHBUND languages might not be due to an “Athabaskanization” of their phonologies (and of other subsystems?) through substrate and/or adstrate transfer (such “Athabaskan” features could thence have spread to other languages/language families of the area not in direct contact with Athabaskan).
    Just my two cents.

  70. “The discontinuous spread of Athabaskan suggests that this family must once have had a more widespread distribution along the entire North American Pacific coast.”
    That’s a negatory, good buddy. These languages are intrusive everywhere south of BC and Alberta. Along the West Coast south of the Tlingit they lived on all the crappiest land, generally a sign of late arrival. The early arrivals, the Chumash for instance, lived in Santa Barbara. Need I say more. In california the Athapaskan groups lived in the deepest darkest redwood forests upriver from everyone on the Eel River – beautiful but hardly anything to eat. Really the last place anyone is going to settle. To this day even the whites won’t live there.
    And their late arrival in the Southwest was a matter of evidence in a court case the Hopi won aginst the Navajo.
    “Thus, if (inter alia) Salish expanded at the expense of Athabaskan, ”
    If Salishan did expand at anyone’s expense, it was into Oregon at the expense of Penutian groups – the Tillamook were were either intrusive or stupid enough to bypass the mouth of the Columbia or else were shoved aside by the Chinook and others – unlikely since a newly arriving gorup is going to be small and weak. So no, not likely.
    And since it is unlikley that anyone has expanded at the expnse of Athabaskan, that leaves the question of similairities in their phonologies. Glottalization is nothing unusual west of the Rockies, and that’s a big, big area, but as you point the issue is glottalization on affricates. As for the laterals, Athapaskan languages affricate everything they can. Who else on the continent has t-theta afrricates? So back to the glottalization of affricates – dunno, is my answer. But if borrowing and language shift follow prestige, the flow is going to be from rich, populous Salishans to half-starved Athapaskans stumbling in off the muskeg or down out of the woods.

  71. Man, I learn a lot from you guys (using “guys” as gender-neutral).

  72. marie-lucie says:

    From what I know of Pacific Coast languages (having worked on some of them in their own environment and on others from written documents), I am with Jim on the topic.
    (Non-linguists, be warned that there will be a fair amount of jargon in what follows).
    The Athabaskan languages (or more precisely their speakers) are considered to be the last to have entered North America before the more recent (archeologically datable) spread of the Eskimo languages in the circumpolar area (together with technology adapted to permanently cold conditions). The distant relationship of the Athabaskan family with the Yeniseian family (which once covered a large area of Siberia as shown by place and river names) was suspected for a long time until it was put on a more solid comparative basis about 2 years ago by Edward Vajda. This relationship, which is based on both distinctive morphological structure and regular phonological correspondances (like other established relationships such as those within the Indo-European family), tallies with a relatively recent entry of the Athabaskan languages into the New World, since a much more ancient split between the two language families (and within the Athabaskan family also) would probably be have left the languages much more different from one another by now.
    In North America the Athabaskan territories only rarely include coastal and riverine areas with rich fishing grounds (such as in the heart of the thriving Salishan language area), places where people can survive and maintain their culture on a relatively small piece of land as long as they can make a living from the sea (classical European examples are the Basques and the Celtic speakers, who are, or were until recently, located on the Westernmost fringes of Europe, along the Atlantic). In an area of (present or former) multiple language groups, or successive language shifts (as for instance in France, formerly Gaul, where the Celts were in turn superimposed on a more ancient population) there are linguistic ways to determine, for instance, which languages the names of rivers come from (since it is rare for newcomers to rename rivers), or which language group is likely to have borrowed technological items from another. So the (relatively) recent migrations which brought Athabaskan speakers South, mostly inland but also on a few coastal ares (where they adopted local technology) are not just an unconfirmable hypothesis.
    As far as general typological resemblances go, “polysynthesis” can mean different things to different people, and once one considers the actual details of the “polysynthetic” languages in the Pacific Coast area (which are not the majority), there is wide variety in structure, which suggests that the local developments are independent. Nor do we find, for instance, that languages which are neighbours of Athabaskan languages display more, or more complex, polysynthesis than those of the same family which are located farther away, as one might expect if there had been structural borrowings from Athabaskan. In fact the Athabaskan languages are so complex and distinctive in structure that it would be very difficult for other languages (unless they were actually related to them like Eyak and Tlingit) to be influenced by them except for borrowing a few words.
    As for phonological resemblances such as the occurrence of glottalized affricates, they are not at all probative. Glottalization plays a morphological role in a number of languages of the area (for instance helping to form verb tenses). The occurrence of glottalized affricates does not mean that original affricates have undergone glottalization: it could be that original glottalized stops (or stop sequences) have become affricated, since in general affricates are known to result from the evolution of stops or of consonant sequences. The many affricates currently reconstructed for Proto-Athabaskan (which, like any “proto-language”, is a hypothesis subject to confirmation or modification) are unusually numerous, but they are not in themselves unusual, and most of them, including affricate laterals (which are admittedly rarer than most other affricates) exist, glottalized or not, in several other languages of the area, where they can often be demonstrated to derive from stops or sequences of stops.

  73. Jim, Marie-Lucie:
    I agree with you both and stand by my original two cents. That is to say, I entirely agree that Athabaskan (or, more broadly, Na-Dene) is a latecomer, certainly the last language family to enter the Pacific Northwest.
    1-However, to argue from the present-day geographical marginality of Athabaskan speakers that their languages *always* lacked prestige strikes me as unjustified. Marie-Lucie mentioned the “Celtic fringe” in Europe. Well, on the basis of the Modern Celtic languages and their (lack of) prestige it would seem quite ridiculous to claim that Latin/Early Romance or Germanic absorbed a number of Celtic loanwords…err, except, of course, that they both did.
    2-There is also a matter of chronology here: Paul Kroeber claimed that Proto-Salish was circa 3000 years old, which is older than Athabaskan (circa 2000 years old) but younger than Na-Dene (4000-6000 years, according to different sources).
    3-More broadly, I find it implausible in the extreme to imagine that the Pacific Northwest ethnolinguistic map before the Na-Dene + Athabaskan expansion was the same as today’s, minus Na-Dene/Athabaskan. Nor do I see any reason to assume the pattern of contact/language shift to have been simple or straightforward.
    (Again, the Celtic fringe in Europe is a useful comparandum: while Celtic has, over the past two millenia, globally lost ground to Romance and Germanic, there are two modern Celtic languages (Breton and Scottish Gaelic) which have expanded at the expense of Romance and Germanic, respectively).
    4-Thus, Salish may well have expanded, not so much at the expense of Athabaskan as at the expense of Na-Dene languages (the reconstruction of whose proto-language is on much shakier ground than that of proto-Athabaskan, alas) as a whole: quite possibly the expansion of Athabaskan (and subsequent assimilation of Athabaskan by Salish speakers) was but the last chapter of a long process of “Na-Dene”-ization of Salish and other Pacific Northwest languages.
    5-Marie-Lucie: if indeed, outside Athabaskan, glottalized affricates can be shown to be a recent feature, then my two cents are stronger. For Proto-Athabaskan is reconstructed on the basis of all Athabaskan languages, be they of the Pacific Northwest or not. Hence the presence in the proto-language of glottalized affricates does seem a solid datum. This makes diffusion from Athabaskan (and other Na-Dene languages?) to the Pacific Northwest far likelier than I originally thought.

  74. marie-lucie says:

    Etienne,
    I entirely agree that Athabaskan (or, more broadly, Na-Dene) is a latecomer, certainly the last language family to enter the Pacific Northwest.
    Actually, if the peopling of the Americas did not exclusively come through the then-dry Bering Strait, then other languages (including Athabaskan!) could have been carried by later immigrants arriving by sea (still a controversial idea, but not as shocking as it used to be).
    1- on the basis of the Modern Celtic languages and their (lack of) prestige it would seem quite ridiculous to claim that Latin/Early Romance or Germanic absorbed a number of Celtic loanwords…err, except, of course, that they both did.
    True, but conditions were different two thousand years ago when Celtic was still spoken over a very wide area (including Ireland, where the language survived as that of the general population for another 1500 years or so), and some early Celtic loanwords are still recognizable in present-day Romance and Germanic.
    I see your point that Athabaskan languages could have similarly influenced other language families if they formerly occupied a much wider area than nowadays (which we cannot prove historically at this point), but there is not much evidence for Athabaskan loanwords in languages which are not now in close proximity with the Athabaskan languages themselves, and Athabaskan morphological structure is not of a type that is easily borrowed, even in parts (see Mithun 361-366 for a brief description).
    2-There is also a matter of chronology here: Paul Kroeber claimed that Proto-Salish was circa 3000 years old, which is older than Athabaskan (circa 2000 years old) but younger than Na-Dene (4000-6000 years, according to different sources).
    Speaking personally, I am very wary of suggested dates for this or that proto-language in the absence of convincing supporting evidence from other sources (eg archeology). I find it very strange that Proto-Salish and especially Na-Dene should be pushed that far back into the past, when the languages in each family are still very obviously related to each other. I think that the date for Proto-Salish has been picked because it tallies with some archeological finds in the Fraser valley, but there is no way of proving which languages were spoken in the area at that time: Proto-Salish is only a guess.
    3-More broadly, I find it implausible in the extreme to imagine that the Pacific Northwest ethnolinguistic map before the Na-Dene + Athabaskan expansion was the same as today’s, minus Na-Dene/Athabaskan.
    Agreed.
    (Again, the Celtic fringe in Europe is a useful comparandum: while Celtic has, over the past two millenia, globally lost ground to Romance and Germanic, there are two modern Celtic languages (Breton and Scottish Gaelic) which have expanded at the expense of Romance and Germanic, respectively).
    I doubt that Breton expanded at the expense of Romance. This would suggest that Brittany was thoroughly romanized at the time of the emigration from Britain to Brittany, which is doubtful. It is more plausible that the Britons coming from Great Britain (especially Cornwall) crossed the sea to join a depleted Gaulish-speaking community with which they already had long-standing contacts and which had not been too deeply influenced by Roman culture. The two Celtic languages were probably largely mutually intelligible.
    In Scotland, do you mean the Viking settlements? I think that as in Normandy, the Vikings must have mostly married local women, and their children and grandchildren therefore went from initial bilingualism to monolingualism in the Celtic language of their mothers and grandmothers.
    4-Thus, Salish may well have expanded, not so much at the expense of Athabaskan as at the expense of Na-Dene languages
    … as a whole:

    This is quite possible, as there is at least one other example further north (upper Skeena valley, to my knowledge), but that does not mean that there was an enormous Athabaskan-speaking area which gradually became depleted as speakers switched to other languages: why would they?
    quite possibly the expansion of Athabaskan (and subsequent assimilation of Athabaskan by Salish speakers) was but the last chapter of a long process of “Na-Dene”-ization of Salish and other Pacific Northwest languages.
    Here I can’t follow you with either large-scale absorption of Athabaskan speakers or a “long process of Na-Dene-ization” which would have to be substantiated by specific linguistic facts which have not thus far been apparent.
    5- if indeed, outside Athabaskan, glottalized affricates can be shown to be a recent feature, then my two cents are stronger. For Proto-Athabaskan is reconstructed on the basis of all Athabaskan languages, be they of the Pacific Northwest or not. Hence the presence in the proto-language of glottalized affricates does seem a solid datum. This makes diffusion from Athabaskan (and other Na-Dene languages?) to the Pacific Northwest far likelier than I originally thought.
    I don’t see that it follows. Affrication is a very common phonetic process, as you well know, and does not need an appeal to diffusion to justify its existence in a particular language. That it should be combined with glottalization in an area of the world where glottalization is common (including as a morphological process) is not at all surprising. Again, diffusion of this feature would be more plausible if it occurred more frequently or generally in languages close to the existing Athabaskan areas than in others, but glottalization is pervasive throughout Western North America (at least).
    As I said above, the presence of so many affricates (glottalized or not) in reconstructed Proto-Athabaskan suggests to me that the reconstruction can be pushed farther back in time, since affricates generally have other origins (and therefore both PA and Proto-Eyak-Tlingit-Athabaskan might eventually look different from what they are assumed to have been).

  75. Marie-Lucie–
    On Breton: actually, it is very clear that it spread at the expense of Romance. Celtic toponyms in Brittany show the effects of Western Romance sound changes, *not* those of the sound changes which turned Proto-Celtic/British Celtic into Breton. Hence the most reasonable assumption is that the Celtic-speaking population of Brittany had been romanized before the arrival of Breton speakers. This is explained in more detail in Joseph Lot’s book LES MOTS LATINS DANS LES LANGUES BRITTONIQUES.
    I strongly suspect that claims that Gaulish survived in Brittany are a product of Breton nationalism more than of scholarship. Parenthetically, similar such claims have been made for Scottish Gaelic, i.e. that the language of Irish invaders was mixed/fused with the indigenous Celtic language of Scotland. Kenneth Jackson, in a short monograph (COMMON GAELIC), quite thoroughly demolished the idea.
    There is, incidentally, similar evidence indicating that Albanian and Basque both expanded (at around the same time) at the expense of Romance: I can supply some references, should you or anyone else be interested.
    The key point is: the fact that there are Celtic toponyms in an area where a Celtic language (Breton) is found does *not* mean that the area has always been Celtic-speaking without interruption. I see no reason to doubt that similarly such complex linguistic history can be found in other parts of the world, including the Pacific Northwest.
    Finally: on the lack of Athabaskan loanwords in neighboring languages. Diffusion of sounds across a language area needn’t imply shared vovabulary: the area of Europe with front rounded vowels (French, German, Dutch, continental Scandinavian, Finnish, Hungarian) forms a geographically contiguous whole, but does not have any exclusive shared vocabulary. So a sound type could have spread (from Athabaskan to neighboring languages, perhaps…) without accompanying vocabulary.

  76. “Paul Kroeber claimed that Proto-Salish was circa 3000 years old,”
    That was a good guess along, long time ago. It can’t possible be valid. People have been living in the Salish Sea region for more than 9,000 years, it is THE desirable place to live north of say the Mendocino Coast, and so once someone setleds there, they are going to cling like barnacles. That means the Salishan peoples have been there for all that time. So 3,000 BCE for proto-Salishan has probelems. The only plausible way the langugae could have arived there and then stayed uniform to reamin one language is if they were somehow bottled up. It’s up to the person making that claim to identify that mechanism.
    I’m not really comfortable with comparing the Celtic fringe with Salishan or the spread of the Athapaskan languages because the terrain and terrain effects are so different. You can comapre Athapaskan with Turkic maybe. There is really nowhere in Western Europe that is as rich and desirable as the Salish Sea region, and that matters when it comes to language shift and population movments.
    Here is a question about language shift for you both: is it likelier in areas that have already shifted language once, where the population is not so identified emotionally with its current languages as to make language shift a matter of ethnic identity? In Brittany was a shift to Breton easier because whatever form of Latin people spoke was equally foreign to them? after all, it had been only a few centuries since they shifted to it. In Scotland, was Gaelic just one more contender among Brythonic and whatever forms of Germanic speech may have been present, and whatever else? I wonder if this may have happened in Mexico in the spread of Spanish, where other imperial langauges may have spread before Spanish arrived.

  77. Jim: the fact that Salish speakers today inhabit highly desirable real estate does not mean that language shift did not take place (I suspect repeatedly) among the dwellers in said real estate. Language spread typically takes place without the intrusive language being spoken by a demographically more dynamic group: rather, the language, spoken initially by a minority, is sufficiently prestigious for people to switch to it, until it becomes a majority language. It has long been noted there there is, in Europe, a hopeless mismatch between population genetics and language affiliation: Northern France is Romance-speaking but is genetically closer to The Netherlands than to Italy or Spain, Bulgarians and Serbians are genetically closer to Greeks than to Poles or Russians…in like fashion, I see no reason to doubt that a similar mismatch exists in the Pacific Northwest.
    As for your second question: I doubt frequency of language shift in a given area has much of an impact as to how language shift is subsequently accepted. The reason is that in pre-modern times language shift of a population on its home territory was a very gradual process, taking centuries to come to completion. After which it takes very little time for the very existence of the earlier language to be wholly forgotten.

  78. marie-lucie says:

    Etienne: obviously you have read more about Breton, etc than I have, so I won’t try to argue with you about it. I would indeed be interested in some references about Albanian and Basque vs Romance.
    However, about the spread of phonetic/phonological features:
    the area of Europe with front rounded vowels (French, German, Dutch, continental Scandinavian, Finnish, Hungarian) forms a geographically contiguous whole, but does not have any exclusive shared vocabulary. So a sound type could have spread (from Athabaskan to neighboring languages, perhaps…) without accompanying vocabulary.
    It is true that the occurrence of front rounded vowels seems to be an areal feature in Europe, and the example of Scots English can be added to the list (and the feature is still spreading in English), but I am not aware that this feature (wherever it may come from) is attributed to the earlier influence of a single substrate language which may have affected the ancestors of the present languages, as you suppose Athabaskan to have been with respect to the Pacific Coast languages.
    About Hungarian and Finnish, their front rounded vowels are not due to the spread of a Germanic feature but caused by the feature of vowel harmony which both languages have inherited from a distant common ancestor spoken before the westward migration of the people. One might attribute Germanic “umlaut” to a version of vowel harmony (since Germanic speakers were part of the westward migration too), but that explanation does not seem possible for French and Occitan (which has the high front rounded vowel).

  79. marie-lucie says:

    Jim, this time I agree with Etienne about the potential antiquity of Salishan. Even if there has been continuous human occupation in the area of the Salish Sea for more than 9000 years, it does not mean that the same language has been spoken (with continuous change) by the inhabitants, even assuming that the same ethnic group has been living there all that time. There is a small residual language family, Chimakuan, which has managed to survive in two small corners of the border area between Wakashan and Salishan, and it is most likely that at least Salishan has expanded at the expense of Chimakuan, if not of other language families which may have disappeared from the same area.

  80. Marie-Lucie:
    1) On the expansion of Basque at the expense of Romance in the present-day Spanish Basque country, see Schmoll, Ulrich. 1959. DIE SPRACHEN DER VORKELTISCHEN INDOGERMANEN HISPANIENS UND DAS KELTIBERISCHE. Wiesbaden: Otto Harrasowitz.
    2) On Albanian expansion at the expense of Romance: Norbert Jokl 1928, in his BALKANLATEINISCHE STUDIEN (BALKAN ARCHIV, 4, p. 195-217), had pointed out that Albanian toponyms often contain /o/ for Latin short /u/, whereas Albanian loans from Latin have /u/ as the reflex of Latin short /u/. I *think* Eric P.Hamp used this as one of several pieces of evidence to postulate a more northern URHEIMAT (Montenegro? Southern Serbia?) for proto-Albanian.
    Its spread at the expense of some form of Romance in present-day Albania is alleged to have taken place after the fall of the (Western) Empire.
    3-I expressed myself poorly earlier: I certainly didn’t mean to imply that I thought Na-Dene/Athabaskan influence could explain all the shared features of the Pacific Northwest language area: just *some* such features, with glottalized affricates being a possible example.
    4-Granted that the front-rounded vowel area of Europe does not owe its phonological feature to a single substrate language. But I had written that a feature of ultimate Na-Dene/Athabaskan origin needn’t have diffused directly. Instead it could have spread indirectly, spreading from adjacent language to adjacent language. So the two phonological SPRACHBUND features might be more alike than seems to be the case at first glance: and the lack of common vocabulary items in the first *may* be parallel to the same lack in the second.
    5-To Hat, and also to whomever has had the fortitude/masochism/intellectual curiosity/whatever to read this far: a thousand thanks for your patience.

  81. marie-lucie says:

    Etienne: I certainly didn’t mean to imply that I thought Na-Dene/Athabaskan influence could explain all the shared features of the Pacific Northwest language area: just *some* such features, with glottalized affricates being a possible example.
    That’s the way I understood it, but you seemed to stress the existence of glottalized affricates. Perhaps it is because I am so used to them by now, but I don’t see anything particularly striking about their presence in the languages of the area, given that these languages also have glottalized stops, and some of them have glottalized sonorants (which are more rare than glottalized affricates). Did you have in mind a situation similar to the one that spread clicks to the originally click-less Bantu languages in South Africa? (or, alternately, that preserved clicks from the substrate language(s) whose speakers adopted Bantu languages).

  82. Marie-Lucie–
    Yes, the spread of click consonants from Khoisan to (some) Bantu languages is another example of the sort of process I had in mind. I checked the WORLD ATLAS OF LANGUAGE STRUCTURES, and unfortunately they have nothing on glottalized affricates, so it isn’t clear whether a contact explanation is required in order to explain their widespread presence in the Pacific Northwest.

  83. Marie-Lucie,
    Salishan has certainly expanded at the expense of Chimakuan, but no one claims that chimakuan was ever the spoken over the entire region i.e that salish is intrusive in the area. The entire area where Chimakuan is spoken, the north shore of the Olympic Peninsula, is ecologicaly and therefore economically marginal. There is no sign of any Chimakuan presence in the Salishan core area: the Puget Sound, Fraser Delta, Straits of Georgia, etc.
    For the length of the West Coast the best lands were invariably populated with people who spoke the most isolate languages – either relict or simply branched so far up as to be hard to align with anyone else – and were therefore probably the oldest settlements. Examples: the Chumash, in Santa Barbara and Vneture Counties(!); the Pomo on the Russian River and adjoining lands(! again), the Achomawi in the Pit River country, asnd I almost forgot, the Yuki in Round valley and the Wappo in Napa Valley. To this day these are the most desirable places to live. Newcomers got the shitty places.
    “rather, the language, spoken initially by a minority, is sufficiently prestigious for people to switch to it, ”
    Yes, Etienne, that is exactly my point – you have to show how Athapaskan could ever posibly have been the high prestige langauge in an area where they are the poorest of the poor. Or else you must show how incoming Salishan speakers, poor and weak as new settlers, could somehow become so numerous and predominant that anyone would want t adopt their speech. I think you may be uncosciously applying a model in which small but powerful groups, such as Europeans or Turks in imperial mode, could swing the language vote their despite their lack of numbers. Not even the Aztecs pulled that one off, much less a few (for the sake of argument) late arrival canoe-loads of Salish people.
    But here is a bonafide example of who had the cultural upper hand in the region. Adstrate effects are the thin edge of the wedge when it comes to language shift, or at least jack lang seemed to think. Salishan languages are rgidly verb-initial and the entire syntax seesm set up for this. Nuu Chah Nulth, and probably all the others in Wakashan, is also verb intial, but if you look at the verbs, you find all the TAM stuff suffixed, where it would normally not be. So how did it get there? Wrong question, wrong way around – the TAM plumbing was always there, it’s the verbs that have moved to the front of the clause. That is after all an easier move. And where that influence came from is pretty obvious.
    I have heard elsewhere that Basque expnanded at the expense of romance in Spain; what mechanism caused that, do you think? same for Albanian, I suppose; why were these Romance-speaking populations so ready to shift?
    M-L, I understand your point about the spread of Salishan at the expense of what may have once been there, but it begs the question of what mechanism could have occasioned that spread. As I understand situations like the Salish Sea in other parts of the world, what is typical is that the language situation is stable over dozens of centuries.
    And I have some doubts about dating splits based on some guesstimate clock of shared and non-shared innovations, and then from that deriving earliest possible dates for a proto-language. Salishan languages have stayed in close contact, even when slightly removed as in the case of Tillamook and Tsamosan, the entire time, so much so that innovations could easily have spread and standardized over these centuries. The Great Vowel Shift in English occurred under less likely conditions.
    And I second Etienne on thanking Hat for his forbearance, although I suspect he has just gotten bored and wandered off.

  84. Oh, no, I find this fascinating; I just have nothing to contribute!

  85. That never stops most of us.

  86. That never stops most of us
    Right on, Meg! All day long at work I can afford to open my mouth only after judicious consideration. It is rather tiring. What better way to relax of an evening than to whirl around on the rides at hat’s bemusement park?

  87. Some of the best threads here are the ones where I can only listen. That’s not cause and effect, mind you. Sometimes it’s like visiting the faculty lounge at a much-better-than-average research university.

  88. Jim: regarding the spread of Basque, Albanian and Breton at the expense of Romance. It is significant that in all three instances the spread took place in the Early Middle Ages (which, incidentally, is also when West Germanic and South Slavic expanded at the expense of Romance), i.e. after the fall of the Western Roman Empire. Life became intensely local as a consequence. During earlier centuries Romance had the weight and prestige of an Empire behind it: but after the fall, speakers of a Romance vernacular might well shift to a non-Romance one if speakers of the latter, *locally*, enjoyed higher prestige for some reason.
    And indeed, in the aftermath of the fall of the Empire, ethnic groups whose participation in the broader imperial economy had been marginal may well have been better able to cope with the aftermath of its collapse and build viable local societies: so viable that they (and their languages) subsequently expanded.
    An interesting clue as to how such sociolinguistic reversals of fortune can take place is found in (Cham) Albanian, which has REMER as a reflex of ROMANUS, with the meaning having shifted to …”Shepherd”. SIC TRANSIT GLORIA MUNDI indeed…
    As for language shift in the Pacific Northwest: the simple fact is that ALL of the languages of the area would be isolates (or at the very least not obviously related) if what we had there today was merely a continuation of the original settlement pattern of 9000 years ago. The reality is that language shift must have taken place to account for the spread of all (easily) identifiable language families in the area.
    HOW this shift took place is a more difficult question. Considering how many uncertainties there are surrounding the spread of Indo-European, I doubt we’ll get a clear-cut answer to the question as to how the major language families of the Americas spread anytime soon.
    And I am delighted that Jim, Marie-Lucie and myself are not the only ones still reading this thread. Rest assured that I have likewise learned volumes from threads here that I contributed nothing to.
    I disagree with John Emerson on one point however: I doubt there is a faculty lounge anywhere which produces discussions that are consistently as erudite, interesting, and just plain fun, in an atmosphere of such mutual respect, anywhere on the planet.
    (And if I’m wrong, and there is such a University, please tell me they have an opening…)

  89. David Marjanović says:

    Someone please fill me in on the linguistic history of Scotland. Gaelic comes from Ireland, and before that, there were the Picts, who spoke either a Brythonic language or something utterly exotic (or perhaps both). So far, so good; that much is textbook wisdom. But Germanic? At that time? ~:-|

    Athabaskan morphological structure is not of a type that is easily borrowed, even in parts (see Mithun 361-366 for a brief description)

    …or just start reading the Wikipedia article on the Navajo language and wait till your brain overheats.

    affricates generally have other origins (and therefore both PA and Proto-Eyak-Tlingit-Athabaskan might eventually look different from what they are assumed to have been)

    On the other hand, Vajda found evidence of a kentum-satəm division in his work on Yeniseian and Na-Dené.
    Large numbers of ejective affricates can be found elsewhere, like in the Caucasus…

    the area of Europe with front rounded vowels (French, German, Dutch, continental Scandinavian, Finnish, Hungarian) forms a geographically contiguous whole

    Not just continental Scandinavian. Icelandic has them just as well. And it’s not just Finnish and Hungarian either, but all the way back to Proto-Uralic… French, West Germanic (lost in English, now sort of coming back), and North Germanic look interesting in this respect, though.

    Perhaps it is because I am so used to them by now, but I don’t see anything particularly striking about their presence in the languages of the area, given that these languages also have glottalized stops, and some of them have glottalized sonorants (which are more rare than glottalized affricates).

    Standard Average European is relatively rare in affricates, and rather unusual in how it tends to treat affricates and plosives separately. Compare the Chinese affricates, which come in aspirated/unaspirated pairs or aspirated/stiff-voiced/slack-voiced triplets just like the plosives, with English with its (rather weakly but audibly) aspirated plosives but barely, when at all, aspirated affricate.

    Europeans or Turks in imperial mode

    :-)

    And indeed, in the aftermath of the fall of the Empire, ethnic groups whose participation in the broader imperial economy had been marginal may well have been better able to cope with the aftermath of its collapse and build viable local societies: so viable that they (and their languages) subsequently expanded.

    While the Romance-speaking population contracted, moving out of mountain valleys (or even dying out).
    Austria is a special case, where Odo(w)akar actually ordered the Romance population to move to Italy. Many did, taking the bones of St Severin with them, and leaving empty countryside behind that was settled by Bavarians in the northwest and Slavs in the southeast.

    REMER

    Römer ;-)

  90. David Marjanović says:

    An important factor in the fall of the Western Roman Empire were the immense taxes. The simplest way to stop paying taxes was to become a barbarian…

  91. marie-lucie says:

    Etienne: about the end of the Roman empire and the expansion of the local languages in some areas, could it be that the Romance-speaking, Romanized population (many members of which were probably bilingual) was mostly limited to the cities, while the rural population (including some descendants of the former elites) had continued to speak their own language, so that with the severance of the official links to Rome (except in the Church), there was little advantage in continuing to learn and speak Latin? A similar example is the reemergence of English after its kings no longer had possessions in both England and France, and immigration from France had therefore ceased, so that the originally foreign dominant class eventually became identified with the local population and French was no longer indispensable or even useful for social advancement. A highly-placed person (such as an ambassador) visiting England in the 14th century and associating with other highly-placed persons could have described the country as French-speaking (except of course for the peasants), while it would have been more accurate to say that in the cities there were large numbers of bilingual people, and the countryside remained English-speaking.

  92. Of course it’s fascinating, even for those of us who read it with the google toolbar. I certainly appreciate those who take the time to post remarks that are knowledgeable and concise.

  93. marie-lucie says:

    jim: [in order to support an Athabaskan-to-Salishan language shift] you must show how incoming Salishan speakers, poor and weak as new settlers, could somehow become so numerous and predominant that anyone would want t adopt their speech.
    Since the heart (or probably earliest place of settlement) of Salishan country is along a strait with plentiful marine resources, it is possible that the Salishan settlers had better fishing or fish-processing technology than the locals, or some other claim to fame, or were just better-fed (thanks to that technology) and therefore better able to survive, and that this was a source of economic and cultural superiority. This would be the case, for instance, if they knew how to catch, process and preserve salmon, while the locals were mosly foraging for seafood on the beach, or knew how to catch salmon but not how to preserve it for long term storage.
    … the spread of Salishan at the expense of what may have once been there … begs the question of what mechanism could have occasioned that spread.
    As one gets farther away from the Salish Sea, along other coastal areas and along rivers leading into the Interior (therefore along avenues of trade), the language territories become larger. This probably indicates that peripheral peoples living in those (less populated and resource-rich) areas must have adopted the prestigious language. Alternately, Salishan speakers may have migrated into those areas in large numbers, but since their core area was much richer in food resources, language shift on the part of other groups is more likely.
    Etienne: As for language shift in the Pacific Northwest: the simple fact is that ALL of the languages of the area would be isolates (or at the very least not obviously related) if what we had there today was merely a continuation of the original settlement pattern of 9000 years ago.
    This would mean that the original settlers had multiple origins, and probably came at different times.
    The reality is that language shift must have taken place to account for the spread of all (easily) identifiable language families in the area.
    I am not sure I follow. Also, the currently recognized language families (easily identifiable, because their component languages are morphologically and lexically similar) are so numerous because little serious and credible work has been devoted to finding genetic links between them.
    David: affricates generally have other origins … – On the other hand, Vajda found evidence of a kentum-satəm division in his work on Yeniseian and Na-Dené.
    Affricates usually derive from stops under certain conditions, especially velar stops (k, g), or from some consonant clusters. On the other hand, affricates, once formed, often evolve into plain fricatives. So a sequence of changes k > ts > s is not particularly remarkable, with k remaining in some areas or environments, changing to ts elsewhere, and later ts becoming s (as happened between Latin and Old French, for instance). So whether the k ~ s correspondence occurs in “kentum-satem” (’100′ in Latin and Sanskrit respectively, used as an exemple of the correspondence in these languages) or in Yeniseian vs Na-Dene, is immaterial, as it is due to a general phonetic tendency.
    LH and all: thank you for the encouragement!

  94. David: in the days of the Roman Empire Scotland spoke Pictish. In the ninth-tenth century or thereabouts Gaelic expanded from Ireland into the Scottish Highlands, Old Norse (from Scandinavia) expanded on the Northernmost Scottish islands, and Anglo-Saxon (later to become Scots) had earlier expanded in the Scottish lowlands.
    Subsequently the Southern languages expanded northwards: Scots at the expense of Gaelic, Gaelic at the expense of Old Norse. The latter linguistic expansion of Gaelic being what I was talking about when referring to the expansion of Celtic at the expense of Germanic.
    Marie-Lucie: actually, there is little to indicate that there ever was a lasting urban/rural divide between Latin/non-Latin languages in the Empire. One major factor in Romanization, after all, was the custom of giving army veterans a large plot of land after they had served twenty years in the legions: this meant that Latin speakers were to be found, salted in the countryside, wherever good land was found.
    Significantly, Basque, Albanian and British Celtic were all spoken in remote mountainous areas whose agricultural value was limited, to say the least.
    Returning to the Pacific northwest: whether the original settlers (9000 years ago) spoke one or many languages is immaterial to the point I am making. Even if they had spoken a single language, its daughter languages today would be as distinct as Bengali and English, if not more so. This is not the case for languages belonging to families of the area such as Salishan, whose spread is therefore far more recent. Hence this spread postdates human settlement in the area, and hence must have involved language shift.
    Whether the shift to Salishan was because of the military superiority of Salishan speakers, their better food production techniques, or cultural prestige, or some combination thereof…well, frankly, we may never know.

  95. marie-lucie says:

    Etienne: Significantly, Basque, Albanian and British Celtic were all spoken in remote mountainous areas whose agricultural value was limited, to say the least.
    Significantly indeed. There would have been very few Romans permanently implanted in those regions, as opposed to temporarily stationed there, so the vast majority of people would have spoken their original languages, with some degree of bilingualism in towns or cities where the Romans were concentrated, and only the actual Romans there (soldiers, administrators) would have spoken Latin exclusively.
    Returning to the Pacific Northwest, perhaps I misunderstood what you said earlier, because I agree that there must have been language shift, therefore spread, in many cases. I would not compare the spread of Salishan, a compact language family, with that of Indo-European. The geographical scale is very different, as is the obvious relationship between the languages, even in the case of Nuxalk which is geographically separated from the others and has been affected by Wakashan to some extent (mostly syntactically and lexically).

  96. Marie-Lucie: what I meant (and I apologize, I really haven’t mastered the art of being clear and brief) was that even if it turns out that a number of language families in the area are related to one another, the distribution of these languages/language families is not a straightforward continuation of the original situation.
    Another Old World analogy: Europe today is dominated by Indo-European languages, but a majority of Romance, Germanic and Slavic languages today have replaced other languages (including Indo-European ones) over the course of their own expansion, which postdates that of Indo-European. In like fashion, even if Salishan is related to some of its neighbours within a larger language family (“Proto-West coast”, let’s call it), its geographical spread implies that Proto-Salishan (as opposed to its sisters) expanded well after the break-up of “Proto-West Coast”.

  97. David Marjanović says:

    “kentum-satem” (’100′ in Latin and Sanskrit respectively

    Satəm in Avestan, śatám in Sanskrit. Anyway, my point was that this correspondence shows that the affricates were already there in Proto-Na-Dene… but actually, I forgot which side has the plosives and which the affricates, and I also forgot a whole lot of other stuff about that extended conference abstract. I’ll have to look it up sometime.

    In the ninth-tenth century or thereabouts Gaelic expanded from Ireland into the Scottish Highlands

    Oh yeah, I forgot it was that late…

  98. “Since the heart (or probably earliest place of settlement) of Salishan country is along a strait with plentiful marine resources, it is possible that the Salishan settlers had better fishing or fish-processing technology than the locals, or some other claim to fame, or were just better-fed (thanks to that technology) and therefore better able to survive,”
    OK, M-L, that is plausible. In fact that appears to have happened in California, when the Wintu or the Maidu groups moved in from Oregon. And specifically it was salmon technology and then later home-grown acorn technology that powered their spread.(Why not the Miwokan and Yokutsan groups, who came in earlier? Because they probably came from Oregon by way of Nevada.)
    “even if Salishan is related to some of its neighbours within a larger language family (“Proto-West coast”, let’s call it), its geographical spread implies that Proto-Salishan (as opposed to its sisters) expanded well after the break-up of “Proto-West Coast”.
    I see what you’re saying, but since there is not even a vestigial sign of a Proto-West Coast but rather separate settlements by unrelated groups – nevermind; your point is the possibility fo speread agianst other groups. Possible, but those other groups held their own in their own eqivalent high-value terrains. The groups I have heard Salishan linked to are inland, that is Kutenai and then further on, Algonkian.
    “Even if they had spoken a single language, its daughter languages today would be as distinct as Bengali and English, if not more so. ”
    “I would not compare the spread of Salishan, a compact language family, with that of Indo-European. ”
    Let’s take these together. Of course over a period of centuries, let alone millenia, there is going to be change. But if these languages remain in contact, and in the case of Salishan they have, then there is no reason these changes are not going to leak back and forth throught the same period. That’s one of the defining characteristics of a Sprachbund, after all. It’s the linguistic equivalent of genetic drift in a polulation. Among other things, Bengali and English are so different because they are so distant – I mean physically distant.
    So then that raises another question – if these languages remained in contact, how did they differentiate into separate lanagauges at all? Simple answer – because people consciously decide to differentiate; at least socially they differentate into separate communiites, and they may very well use language as the marker of distinctness. That’s what drives class-based dialect variations. Apparently specifically in the case of Coastal salishan languages, word taboos were very active, one more mechanism.

  99. I would suggest Italian. It is extreemely rich, has an intricate syntax and an extreemely complicated so-called “verbal group” that often resambles the structure of predicates in languages like Navaho, comprizing much more than what is traditionally called “verb”.
    And it is a languge of a rich culture.
    Itlian is easy to master on a “tourist level” but when you try to get a little past that level, you will see how complicated and intricate it really is.
    Finnish is also an interesting language, but not very complicated.

Speak Your Mind

*