Edge of the Knife.

Catherine Porter reports for the NY Times on what sounds like a very worthy promising project, Canada’s first Haida-language feature film, Edge of the Knife:

With an entirely Haida cast, and a script written in a largely forgotten language, the film reflects a resurgence of indigenous art and culture taking place across Canada. It is spurred in part by efforts at reconciliation for the horrors suffered at those government-funded residential schools, the last of which closed only in 1996. […]

Fewer than 20 fluent speakers of Haida are left in the world, according to local counts. For the Haida themselves, the destruction of their language is profoundly tied to a loss of identity.

“The secrets of who we are are wrapped up in our language,” said Gwaai Edenshaw, a co-director of the film, who like most of the cast and crew grew up learning some Haida in school but spoke English at home. […]

Mr. Edenshaw was a co-writer of the script for the 1.8 million Canadian dollar ($1.3 million) film, which is set in Haida Gwaii — an archipelago of forested islands off the west coast of Canada — during the 1800s. It tells an iconic Haida story of the “wildman,” a man who is lost and becomes feral living in the forest. In this version, the wildman loses his mind after the death of a child, and is forcibly returned to the fold of his community in a healing ceremony.

The script was translated into two remaining, distinct dialects of the language: Xaad Kil and Xaayda Kil. None of the stars are conversant in either dialect. The crew held a two-week language boot camp in April so cast members, who also have little or no acting experience, could learn to pronounce their lines before filming started in May. […]

The film would seem cripplingly ambitious if not for the record of the executive producer, the Inuit director Zacharias Kunuk. He made his name with “Atanarjuat” (“The Fast Runner”), which depicted an Inuit folk epic and starred untrained Inuit actors speaking their traditional language, Inuktitut.

That film won the Caméra d’Or at the Cannes Film Festival in 2001, and is still considered one of the best Canadian films of all time.

I saw Atanarjuat and can confirm that it is absolutely terrific, so I have high hopes for this one. We discussed Haida poetry here (with Robert Bringhurst himself appearing in the comment thread) and the language here (where marie-lucie said “Sapir thought that Haida belonged to [Na-Dene], but more recently this has been considered very unlikely, and Vajda’s work makes it even more improbable”). Thanks, Eric!

Comments

  1. Christopher S says:

    Inspired by this post, I just went to check out the trailer for Atanarjuat. I’m sitting in a coffee house with my laptop, but the content looked innocent enough. Then the camera panned to the runner’s exposed cock. Oops…

  2. Heh.

  3. Bathrobe says:

    @ Christopher S. Maybe it was your eyes that were panning. I saw only a full-frontal nude.

  4. SFReader says:

    -Xaad Kil

    I wonder if Kil is related to Finnish kieli and Mongolian khel

  5. kieli
    Võro kiil comes even closer.

  6. Bathrobe says:

    Ooops, maybe your eyes were zooming, not panning.

  7. …a very worthy project…

    I was a bit taken aback by your apparent lack of enthusiasm for the film until I read your final paragraph and realized that you were using worthy in a positive sense.

    OED calls the negative sense of worthy “chiefly Brit.,” but I wonder how many other readers would also have the same reaction that I did.

  8. Stephen C. Carlson says:

    OED calls the negative sense of worthy “chiefly Brit.,” but I wonder how many other readers would also have the same reaction that I did.
    As for this American, it never occurred to me that worthy had a negative sense.

  9. Yeah, I’m not familiar with a negative sense of worthy at all.

  10. January First-of-May says:

    I can’t recall having ever encountered a negative sense of worthy either.

  11. Lars (the original one) says:

    From the examples in the online Oxford it is in the category of damning with faint praise — an undertaking of which you can say nothing more exciting is, well, boring. Based on common prejudices I can well believe that to be a more common reading in Britain. Also ‘village worthies’ has a ring of faintly comical self-assigned importance to it.

    I can’t access the OED so maybe it has British examples where the ‘boring’ sense of worthy is extended in some more negative direction.

  12. I am familiar with this ironic use of worthy ‘showing good intent, ostentatiously correct (but tedious or clumsily executed)’, as in “the film is a little dull and worthy” (which is one evolutionary step from “worthy but dull”).

  13. OED:

    d. Chiefly Brit. In a negative sense. Characterized by good or noble intentions but lacking in humour, imagination, excitement, etc.

    Frequently with reference to literary or other artistic works.

    1930 New Statesman 29 Mar. 809/2 One feels a little of that particular weariness that comes from too much worthiness… One feels that it [sc. a book] itself is so worthy; the author is so worthy; that one is so worthy oneself to have [read it].

    1959 Guardian 4 Dec. 6/4 Dresses, as I have said, are terribly ladylike and could be terribly worthy unless you were careful.

    1970 Daily Tel. 12 Oct. 11 That rather unfortunate category among television documentaries—worthy but not terribly watchable.

    1998 Times 21 Mar. (Mag. section) 29/1 We’re probably a bit square… We always were, a bit lefty, a bit worthy.

    2010 T. Huddleston in J. Pym Time Out Film Guide 2011 957/2 The film draws on timely issues of class struggle and criminal violence… It’s all rather worthy.

  14. January First-of-May says:

    I feel that the 1970 citation, in particular, is just straight-up ironic, in the “damning with faint praise” way.
    And I don’t see any negativity in the 1930 citation (the negativity that is there comes from the “too much” part).

    I guess I’m just not enough of a British English speaker.

  15. As for this American, it never occurred to me that worthy had a negative sense.

    Same here! Divided by a common language, and all that.

  16. I changed it for the sake of my UK readers; I don’t want anybody to be put off.

  17. As a British(ish) reader, I roughly second Piotr’s understanding: the ‘literal’ meaning of worthy is still in itself positive in BrE, but it’s quite strongly associated with ironic/damning-with-faint-praise uses, to the extent that in a fairly neutral context like here, that can be the most natural reading. In speech, of course, tone of voice will usually disambiguate it clearly.

  18. As an Australian, I’ve never heard of “worthy” used in a negative sense.

  19. I’m fairly familiar with worthy person as negative in precisely this way, though COCA doesn’t back me up here. There is an implied “but …” in it.

  20. Here are a few present-day examples from the British corner of the Internet:

    Political pop and message music is usually embarrassingly worthy and dull. [Express]

    And then the opposite tack is if an artist is a bit too dry and heavy and kind of like, “Oh, it’s so bloody worthy,” the work will go (a) unsold and (b) unvisited probably – because it’s a bit dull. [Financial Times]

    Veggie eateries often end up as unbearably worthy sort of places. … If veggie food had a slightly less worthy image it would be a whole lot more popular. … But persevere – all things considered, it has a splendidly low worthiness rating. … At the Place Below the cooking is of a high standard, prices are commendably low and you hardly notice any worthiness at all. [Evening Standard]

    When did “worthy” become a pejorative term? … Literature worthy of our attention, it seems, must not commit the sin of being worthy. The phrase “worthy but dull” seems to have been conflated into one epithet. … Before the current fashion for “political” work, “worthy” was often used to reject work deemed to be issue-based. (etc.) [Guardian]

  21. marie-lucie says:

    I did not know that “worthy” had gone from positive to sarcastic to downright negative during my lifetime.

  22. A negative sense of “worthy” would never have occurred to me either. And now I wonder: Does the ironic meaning also apply to the less common noun form?

  23. marie-lucie says:

    Brett: The phrase “village worthies” is certainly ironic or sarcastic: people who may feel important but within a context that is much too limited to justify that sense of importance.

  24. This shift of worthy towards a “faint praise” sense reminds me of what happened with BrEng quite. (I’m getting a headache trying to figure out what “quite worthy” would mean.)

  25. Brett,

    Not to worth, but as example #3 shows, you can have pejorative worthiness.

  26. Marja Erwin says:

    AmE speaker. “Worthy” is decidedly positive for me, “worthies” usually ironic, and “betters” usually when those who claim they’re the “betters” are insulting other people.

  27. This shift in worthy reminds me of what happened with BrEng quite.

    Or Am.Eng. awsome. Of course no longer exclusively American.
    https://separatedbyacommonlanguage.blogspot.com/2014/12/2014-us-to-uk-co-word-of-year-awesome.html

  28. Bathrobe says:

    Is it akin to the slightly mocking tone of “earnest”?

  29. The original meaning of nice upon its adoption from Old French in the 13th c. was ‘ignorant, stupid, silly, frivolous’ (from Lat. nescius). It evolved into ‘fussy, dandy, delicate’ some two or three generations later, and underwent semantic amelioration to ‘refined, elegant’ by the Elizabethan era, eventually developing a wide range of definitely positive meanings. Now it’s often used sarcastically (Onslow [Geoffrey Hughes]: Oh, nice!), so who knows what may happen to it next.

  30. It would be… nice… if it somehow reverted to meaning ‘ignorant.’

  31. January First-of-May says:

    As exemplified in this famous quote on language change:

    Ye knowe eek, that in forme of speche is chaunge
    With-inne a thousand yeer, and wordes tho
    That hadden prys, now wonder nyce and straunge
    Us thinketh hem; and yet they spake hem so,
    And spedde as wel in love as men now do…

    (Geoffrey Chaucer, Troilus and Criseyde, published 1385)

    [EDIT: I’m not sure if the meaning “stupid, silly, frivolous” or “fussy, dandy, delicate” is intended here; but either way it’s not very positive]

  32. In sondry londes*), sondry ben usages.

    *) E.g. Britain, America.

  33. marie-lucie says:

    Piotr: The original meaning of nice upon its adoption from Old French in the 13th c. was ‘ignorant, stupid, silly, frivolous’ (from Lat. nescius)

    I suddenly realized something I have occasionally been wondering about for decades. As a teenager I was given a book on la Gascogne, about the region neighbouring my mother’s parents’ birthplace in le Languedoc. Among descriptions of traditional customs and anecdotes were scattered words and sayings in the local variety of Occitan, which was close to my grandparents’ own. I have always remembered one word which I didn’t quite understand, which was obviously used in local French because it did not have an exact French equivalent: nèci (I think that was the spelling), which seemed to mean something like ‘lovably silly’ but I could not be sure. Here I have finally found an answer!

  34. Lars (the original one) says:

    [wordes] straunge us thinketh hem — what construction is that? I see one plural noun and two plural oblique pronouns as arguments of a verb in what I thought was 3p singular, so I assume I am misidentifying something.

    *googling happens* Right, Chaucer was a Londoner so thinketh can be 3p plural too. Learn and live. But what’s with hem?

  35. marie-lucie says:

    SFR: I wonder if Kil is related to Finnish kieli and Mongolian khel
    juha: Võro kiil comes even closer.

    I don’t know Haida except for having read Swanton’s grammar (there is a much more recent one by John Enrico), and Finnish, Mongolian etc even less, but since starting to study NorthWestern languages and cultures 40 years ago I have always felt that there could still be traceable language links between the two sides of the North Pacific Rim. I know that this has been a topic of interest in Russia, but not so much in North America (where until recently the topic would have been practically taboo).

    That said, it is well-known that k type sounds (velars) are extremely likely to change according to the vowels they are in contact with, especially so when preceding front vowels such as i and e, so even if the strong resemblances between the words above were indicative of at least borrowing if not actually genetic relationship, the passage of hundreds if not thousands of years (even if not so ancient as the end of the Ice Ages) would probably have obscured the original resemblances. Witness for instance German Kind, English child (all the sounds changed except d), among hundreds of similar examples in languages from the various continents. So unless a number of such resemblances including Haida could be found, these are probably only coincidences.

  36. Talking of silly, there was a time when silly (ME sēlī) meant ‘blessed, worthy (no, no, no irony intended), noble, lucky’. Chaucer was one of the first writers to use it also in the sense ‘foolish’ (beside the older positive meaning). The pivot of the change was the collocation a silly child, where the originally intended meaning was ‘innocent’ (i.e. blessed with innocence).

  37. Us thynketh hem means ‘we think them (to be)’. The construction is like methinks (me thynketh).

  38. David Marjanović says:

    From the examples in the online Oxford it is in the category of damning with faint praise — an undertaking of which you can say nothing more exciting is, well, boring.

    That reminds me: I’ve been told (by one of the participants) about a conversation in Texas where “he’s interesting” was meant to express “I know next to nothing about him, he seems kinda boring” – the next step on the euphemism treadmill after “bless his heart”, perhaps –, but was misunderstood as “I’m in love with him or nearly so”.

  39. Trond Engen says:

    Deponent verb. Oblique subject. Pick your analysis.

  40. marie-lucie says:

    Lars (the original one): [wordes] straunge us thinketh hem

    = ‘We think them [= these words] strange.’

    The original construction with “think” was an impersonal one where the subject (normally it) was not mentioned. For instance, it was “me thinks” not “I think”. To “think” was not something one did, but something that affected one. It corresponded to French il me semble ‘it seems to me’ (the latter probably a calque on French), not to je pense ‘I think’. So here “us thinketh” (later ‘we think’) corresponds to “me thinks” (later ‘I think’).

  41. marie-lucie says:

    Sorry, Piotr, I just saw your post above (at 3:08), You thought faster than me, posting while I was writing, but I am glad we agree.

  42. David Marjanović says:

    German Kind, English child (all the sounds changed except d)

    The d is changed, too. In English it is in fact [d], reliably voiced, except before a pause. It’s voiceless in all kinds of German, and an outright [t] north of the White-Sausage Equator.

    How close Haida is to Na-Dené remains to be figured out (it has plenty of Tlingit loanwords for a start). Morphological features that seem hard to borrow do, however, suggest that it belongs to the Dené-Caucasian languages more generally speaking – which are not the ones that have k-front vowel-l words for “tongue, language”.

    there was a time when silly (ME sēlī) meant ‘blessed, worthy (no, no, no irony intended), noble, lucky’.

    German selig “extremely happy”, “Blessed” (as opposed to “Saint”).

  43. David Marjanović says:

    For instance, it was “me thinks” not “I think”.

    Those were actually separate verbs, one the causative-or-something of the other. German: ich denke, obsolete mich dünkt = mich deucht.

  44. They were separate in Old English: þynċan (pret. þūhte) ‘seem, appear’ (often used in the “methinks” construction with no explicit subject and an indirect object) and þenċan (þōhte) ‘think (abouth sth), imagine, remember’ (transitive). They were conflated in Middle English, hence the mixed construction with the expressed subject of þynċan reinterpreted as its object, accompanied by a predicative expression.

    “The book [subject] me thinks interesting” -> “Me thinks the book [object] (to be) interesting”.

  45. Doesn’t ċ mean ch? If so, why is it “think” and not “thinch” (or “thench”)?

  46. I wonder if Kil is related to Finnish kieli and Mongolian khel
    They surely look close enough, but once we rewind the former back to Proto-Uralic *kälə (compare e.g. Moksha /kælʲ/), the resemblance starts being a bit less obvious.

    And accidental matches are not that hard to find: compare also PU *-ta- ‘verbalizing / causative suffix’ and Haida -taa- ‘causative suffix’. The similarity here is demonstrable as accidental, but I leave the details for interested readers to work out 😉

    That said, it is well-known that k type sounds (velars) are extremely likely to change according to the vowels they are in contact with, especially so when preceding front vowels such as i and e

    That depends. Only two of the 40-ish Uralic language groups do anything of the sort (namely Votic and Proto-Samoyedic), and in fact both labialization (e.g. *ko- > *kʷo- > Udmurt kua-) or retraction (e.g. *ka- > *qa- > *χa- > Hungarian ha-) are more common. I imagine this is related to the fact that Uralic languages already have pretty sizable inherited sibilant inventories, whereas Indo-European languages started off with just *s and have been doing successive rounds of palatalization to fill in further members.

  47. Apparently generalised from 2/3 sg. -ncst, -ncþ, where affrication was blocked in clusters produced by the early syncope of *-i-. With þynċan, 3sg. þyncþ was by far the most common form. An epenthetic vowel was often re-inserted (þynceþ), but by the time this happened velar palatalisation was no longer productive. The Anglian tendency to restore unpalatalised velars (usually blamed on the Vikings) may also have played a role.

  48. David Marjanović says:

    and have been doing successive rounds of palatalization to fill in further members.

    Except German, which hasn’t had a single round of palatalization since PIE, but suddenly decided to create two new sibilants from plosives anyway and ended up with a total of four in Middle and probably Rather Late Old High German. (Then they quickly merged to two or more commonly three, in slightly different ways in different places… oh, wait, some varieties are back up to four.)

    I still don’t understand how [sk] > [sx] > [ʃ(ː)] is supposed to have worked, BTW. The second step is straightforward assimilation, but the first? [sk] is supposed to have not been aspirated.

  49. Apparently generalised from 2/3 sg. -ncst, -ncþ

    Þankþ, þat makeþ þenþ.

  50. Lars (the original one) says:

    mixed construction with the expressed subject of þynċan reinterpreted as its object — that was the part that confused me. So thinketh is actually singular here, in a sentence with implicit subject?

    In Danish the dative experiencer argument of synes, drømme, tykke, … got promoted to nominative subject _after_ the old nominative subject with predicate construction was replaced by one with a subordinate phrase, sidestepping the problem of what to do with the original subject and leaving me nothing to model the English construction on.

    (In Swedish the end result at least was the same, and there the word is still very much alive. Jag tycker att… = ‘My opinion is that…’, as opposed to tänker which is used for thinking in general).

  51. I still don’t understand how [sk] > [sx] > [ʃ(ː)] is supposed to have worked

    I think I have seen retention after Grimm’s Law as a possible explanation: i.e. first *sp *st *sk > *spʰ *stʰ *skʰ > *sɸ *sθ *sx after all, followed by fortition back to *sp *st *sk to resolve the awkward two-fricative consonant clusters. There has after all definitely been similar spirantization/dissimilation tug-of-war with Late PIE *ks > PG *xs > pre-High German *ks > OHG /xs/ > ModG /ks/ chs. I’ve also seem claims of evidence for /pʰs kʰs/ > /fs xs/ as the ancient > medieval values of Greek ψ ξ (in some varieties?), now again reverted to /ps ks/ in Modern Greek… someone else reading this is more likely to know more.

  52. Stephen C. Carlson says:

    For what it’s worth (ahem), the second definition of “worthy” from Google is “characterized by good intent but lacking in humour or imagination. worthy but tedious advice” and in the on-line Cambridge dictionary for British English “FORMAL Something that is worthy is not very interesting but should be admired for its good and useful qualities: a worthy book“.

    The American Merriam-Webster and the American English of the Cambridge dictionary only have positive definitions.

  53. Eli Nelson says:

    @j & David: Perhaps *sp *st *sx is related to how velars seem to be more prone to (retaining?) fricative and approximant realizations in other contexts in certain Germanic languages? E.g. reconstructions of Proto-Germanic, or at least Proto West Germanic, tend to have word-initial non-plosive [ɣ] alongside plosive [b] and [d]; in English, non-initial [ɣ], but not [v] or [ð], was further lenited to become a semivowel, [w] or [j], and I have read that in many varieties of German medial /g/ is realized as an approximant or fricative [ɣ] or [j], devoiced word-finally to [x~χ] and [ç], as far as I know without a comparable “soft” realization in most dialects of medial and final /d/ and /b/. I realize this isn’t really an explanation, but maybe there is some connection between these phenomena.

  54. accidental matches are not that hard to find
    With some being more eerie than others:

    unohtaa/unhottaa – unutmak (Turkish) “forget”
    suur/suuri – zur/zor (various Turkic) “big, great; mighty; difficult”
    sokea – sokur (Turkish) “blind”

  55. David Marjanović says:

    The trick about German is that *sp *st *sk survived the High German consonant shift unchanged, show up as sp st sc/sk/sg (occasionally also sb sd) in OHG, and then the velar one spontaneously changed into sch, not only in High German but even in Dutch and at least parts of Frisian, where it’s pronounced [sx ~ sχ] to this day even though /k/ in other environments is not even aspirated (and probably never was).

    It’s true that differences between places of articulation are common. [g] is objectively more difficult than [b] or [d], because the farther back in the mouth the closure is, the harder it is to build up enough pressure for a plosive while at the same time keeping the airstream flowing so the vocal cords can vibrate; this explains why [g] is a gap in the system in Arabic, for example, and why [ɢ] is so much rarer than [q]. I just don’t know how any such effect explains [sk] spontaneously becoming [sx] while [sp] and [st] are unaffected.

    I have read that in many varieties of German medial /g/ is realized as an approximant or fricative [ɣ] or [j], devoiced word-finally to [x~χ] and [ç], as far as I know without a comparable “soft” realization in most dialects of medial and final /d/ and /b/.

    Part of what’s going on is a substrate effect: /g/ is [ɣ~ʁ] by default in much of Low German and [j] by default in the (endangered to moribund) dialects of places like Cologne and Berlin. However, that’s at most partially related to the canned voice on the bus here in Berlin turning the intervocalic /b/ and /g/ in Liebe Fahrgäste “dear passengers” into approximants so evanescent they might as well not be there at all (*Liee Faheste). I don’t think I’ve encountered this with /d/, though it’s reliably voiced in intervocalic /nd/ and /ld/, which is not the case elsewhere.

    Late PIE *ks > PG *xs > pre-High German *ks > OHG /xs/ > ModG /ks/ chs.

    What evidence is there for “pre-High German *ks”? It remained hs in Old Saxon and Old Dutch, IIRC. Old English clearly had [ks], as shown by its enormous bidirectional confusion between x and sc, but the x of Old Norse could just have been a spelling convention – apparently [xs] is usual in Icelandic.

  56. this explains why [g] is a gap in the system in Arabic

    Many Arabic varieties pronounce Classical [q] as [g], hence “Moammar Gadhafi” as he spelled himself in Latin letters, and of course [dʒ] > [g] in Egyptian (if indeed [g] is not a retention rather than an innovation).

  57. David Marjanović says:

    “(and probably never was)” was supposed to be a link to the interesting chapter that starts on p. 223 of this book.

    BTW, p. 261 is where the chapter on the OHG unstressed vowel system starts that I’ve alluded to recently.

  58. marie-lucie says:

    (m-l) German Kind, English child (all the sounds changed except d) – David M: The d is changed, too. In English it is in fact [d], reliably voiced, except before a pause. It’s voiceless in all kinds of German, and an outright [t] north of the White-Sausage Equator.

    I did not mention the final d > [t] because it is an entirely predictable adaptation having to do with the phonology of German, unlike the changes in the other sounds. If the final d had “changed” to t, then the plural of [kint] would probably be [kinter]. The “change” also applies to the other plosives, while the English palatalization only applies to Germanic k.

    As for English d becoming unvoiced “before a pause”, I don’t think I have ever heard such a thing from native English speakers, or even read a description of it.

    How close Haida is to Na-Dené remains to be figured out (it has plenty of Tlingit loanwords for a start).

    Surely loanwords by definition are not proof of relationship! Geographical closeness is enough of an explanation. Haida also has a number of Tsimshianic loanwords, for the same reason.

    Morphological features that seem hard to borrow do, however, suggest that it belongs to the Dené-Caucasian languages more generally speaking

    I rely on the testimony of Michael Krauss, one of the prominent Dené scholars, who wrote something like “As soon as I looked at Haida I became convinced that it was completely different from Na-Dené”. As for “Dené-Caucasian”, I don’t know anyone who supports this hypothesis. Dené-Yeniseian, on the other hand, is now considered at least plausible (even by the originally skeptical Krauss) as a result of work by Edward Vajda, relying on morphology as well as phonology.

  59. David Marjanović says:

    if indeed [g] is not a retention rather than an innovation

    Somewhere on Jabal al-Lughat there’s a post that explains why, surprisingly enough, it’s probably an innovation. I can’t find it now; I found a comment from 2012 saying I couldn’t find it, and Lameen himself couldn’t remember what the evidence was.

    There are lots of “damned if you do, damned if you don’t” situations in historical phonology. Certain sounds are bad, but the gaps they leave when they disappear are also bad, so the gaps are sometimes filled again. Celtic got rid of /p/, and then P-Celtic promptly turned /kʷ/ into a new /p/. Slavic worked tirelessly for centuries to eliminate all syllable codas, and then the yers fell, and it was all for naught…

  60. David Marjanović says:

    I did not mention the final d > [t] because it is an entirely predictable adaptation having to do with the phonology of German

    No, that’s another thing! Syllable-final fortition is limited to “north of the White-Sausage Equator“. I’m from south of there and don’t have it; I find it very noticeable here in Berlin. And yet, /b d g/ are voiceless in all environments despite staying distinct from /p t k/.

    As for English d becoming unvoiced “before a pause”, I am not sure I have ever heard it

    Actually, I’m not sure if /d/ does it either. Fricatives do.

    Surely loanwords by definition are not proof of relationship!

    Of course. I just mean that contact can make things difficult to sort out.

    “As soon as I looked at Haida I became convinced that it was completely different from Na-Dené”

    That’s no proof of lack of relationship either! 🙂

    Dené-Yeniseian, on the other hand, is now considered at least plausible (even by the originally skeptical Krauss) as a result of work by Edward Vajda.

    Yes, but (pdf of conference presentation, 16 slides).

  61. SFReader says:

    @marie-lucie

    Browsing Dictionary of Alaskan Haida, I learned something extremely interesting. The word ‘kil’ means “language, voice, speech, words”.

    There is also a term Kílaad or Kílaad Xaat’áay – which apparently means “people of the word” (comp. with Slovene – Slavs, from slovo ‘word’).

    One would expect that this term applies to people who speak similar language, but no, the Haida term “people of the word” means Tsimshian people.

    As for Haida (Xaadas – Haida people, Xaad kil – Haida language), the term is apparently derived from Xáad – one’s father, one’s paternal uncle (father’s brother), husband of one’s maternal aunt (mother’s sister)
    NOTE: This term is only used in reference to a female’s father or uncle. For a male’s father or uncle, see çung.

    Based on this information, it seems that the Haida are a matrilineal society and that the Haida women originally spoke a Tsimshian language, but switched to language of their husbands – now known as Haida.

  62. marie-lucie says:

    velar evolution
    (m-l) it is well-known that k type sounds (velars) are extremely likely to change according to the vowels they are in contact with, especially so when preceding front vowels such as i and e
    (j.) That depends. Only two of the 40-ish Uralic language groups do anything of the sort (namely Votic and Proto-Samoyedic), and in fact both labialization (e.g. *ko- > *kʷo- > Udmurt kua-) or retraction (e.g. *ka- > *qa- > *χa- > Hungarian ha-) are more common.

    Perhaps only the two languages in question have “palatalization” of k before front vowels, but “labialization” of the velar as in *ko- > *kʷo- is indeed a change conditioned by the following “rounded” vowel o. “Retraction” of the velar also seems conditioned by the following “low” vowel a.

    JC: [dʒ] > [g] in Egyptian (if indeed [g] is not a retention rather than an innovation).

    I don’t see [dʒ] > [g] as a plausible phonetic evolution, while the opposite is quite common (with intervening palatalization), so the Egyptian case does look like a retention.

    Most linguists consider that in a case like this one, where one language out of several close relatives does not follow the correspondence pattern shared by the others, this language is likely to be the innovative than the retentive one, but I see no reason to take this hypothesis as an absolute rule when the “innovation” goes against everything that is known about the directions of phonetic change.

  63. marie-lucie says:

    David M: Somewhere on Jabal al-Lughat there’s a post that explains why, surprisingly enough, it’s probably an innovation

    Too bad the post cannot be found!

    I was going to add to my post: Retentions usually occur in isolated or marginal areas, innovations in geographically or culturally central locations which attract new people, so if the Egyptian case is a retention, it is difficult to explain.

    However, if it is an innovation, it could be a socially and culturally determined, deliberate counter-retention of a known archaic feature, perhaps religiously conditioned (a return to an older pronunciation of the Qur’an, for instance, as taught by men originating from a country that had not yet fully adopted palatalization).

  64. Right-on,’ used attributively, has the same disdainful tone as ‘worthy’.

  65. That’s even more bizarre to me, since on this side of the pond the phrase (as far as I know) is only a distant memory/mockery of the sillier side of the ’60s.

  66. In Paris about 5 years ago I met a Californian woman born c.1970 for whom “right on” was the discourse particle for “yeah; I see; OK”. The fact that I remember evinces how unusual I found it.

  67. Somewhere on Jabal al-Lughat there’s a post that explains why, surprisingly enough, it’s probably an innovation. I can’t find it now; I found a comment from 2012 saying I couldn’t find it, and Lameen himself couldn’t remember what the evidence was.

    The memory of the post may have been confused with this comment on a Languagehat post, citing Studies in Modern Semitic Languages, Izre’el and Raz (ed), which makes quite a decent argument for it.

  68. and then the velar one spontaneously changed into sch, not only in High German but even in Dutch and at least parts of Frisian, where it’s pronounced [sx ~ sχ] to this day

    Something similar happened in Old English too, where [sk] > [sc] > [sç] > [ʃ(ː)], eventually even before back vowels and /r/ (there are a few lexical exceptions, the most important of them being āscian ‘ask’). Spellings like sceort, bisceop, fisceas (with “palatal diphthongisation”) in LOE prove that the cluster was palatal by that time (in these cases, before originally back vowels). On the other hand, in Late West Saxon metathetic variants like fixas (with [çs ~ cs]) for fiscas or tūx for tūsc ‘tusk’ (ME tush ~ tusk ~ tux) were still common, suggesting that the coalescence of the clustered consonants into a long palatoalveolar was not yet complete. The ax variant of ask (OE ācsian ~ āxian) is as old as the hills. Inherited [xs] was “hardened” to [ks] rather late — certainly later than the breaking of front vowels (*waxsa- ‘wax’ > [wæxs] > [wæɑxs] > LWS weax [wæɑks]) and probably later than Anglian smoothing (A wæx) — i.e. in the 9th century at the earliest. In Early Old English the grapheme x stood for [xs ~ çs], but also for etymological *k + s (produced by vowel syncope), as in rīxian ~ rīcsian ‘rule’. I doubt if there was a real pronunciation difference here, since orthographic variants like rīhsian, āhsian ‘ask’ also occur.

  69. David Marjanović says:

    The memory of the post may have been confused with

    That must be it, thank you!!!

    [sk] > [sc] > [sç] > [ʃ(ː)]

    I always assumed it was more like in Italian: [sk] + front vowel > [stɕ] > [ɕtɕ] > [ɕː] > [ʃː], so that Middle English sch could really have been s followed by ch… but of course that doesn’t begin to explain how it spread to back-vowel environments either.

    I doubt if there was a real pronunciation difference here, since orthographic variants like rīhsian, āhsian ‘ask’ also occur.

    Ah, so |ks| > /xs/ was still active as a surface filter… makes sense.

    …Is it possible that the palatality in sceort and perhaps bisceop was leveled in from forms with umlaut from elsewhere in the paradigm (comparative, plural – the German plural of Bischof is mercilessly Bischöfe)? Of course bishops are most often encountered one at a time, and the whole idea doesn’t work for fisceas, but that word, like bisceop, at least has a front vowel on the other side…

  70. dainichi says:

    @Lars:

    > the dative experiencer argument of […] tykke

    Did you mean “tykkes” here, or am I missing something? In my understanding the development went from “Mig tykkes” (dative experiencer) -> “Jeg tykkes” (nominative experiencer).

    I’m confused, though. If “tykkes” comes from Proto-Germanic þunkijaną (“seem”, with I assume a dative experiencer), then why the need to add the mediopassive (if that’s right term) -s in Danish? I’d have thunk (haha!) that it became dative-experiencer by adding the -s to an originally nominative-experiencer verb.

    Going off on a tangent: online sources seem to disagree on whether “tykkes” should be pronounced with an /ø/ or /y/ sound. It’s obsolete and so not in my active vocabulary, but in the derivation “samtykke” (consent) I definitely use /y/. Also, “tøs” is the common word corresponding to Standard “synes” in at least my father’s Sønderjysk dialect. (With short ø. Nominative experiencer as in modern synes).

  71. marie-lucie says:

    SFR: Some comments on your post on Haida.

    Browsing Dictionary of Alaskan Haida, I learned something extremely interesting.

    Thanks for giving the reference. I just went through a lot of pages of this book. It is relevant that it deals with Alaskan Haida, not the two dialects still spoken a bit on Haida Gwaii (the original island home of the people, some of whom migrated to Kaigani in Alaska only a few generations ago).

    The word ‘kil’ means “language, voice, speech, words … There is also a term Kílaad or Kílaad Xaat’áay – which apparently means “people of the word” … One would expect that this term applies to people who speak similar language, but no, the Haida term “people of the word” means Tsimshian people.

    In the Tsimshianic languages (which include 4 varieties), each of the varieties calls itself “real/true language” (Coast Tsimshian sm’algyax, Nisqa’a and Gitksan both sim’algax_) (in each the final x is [X]) although the fourth variety, Southern Tsimshian,, recently extinct, can also be called sgüüx_s as well as sm’algyax_. From historical evidence it seems that the (“Coast”)Tsimshian, living just across Hecate Strait from Haida Gwaii, had a strong trading relationship with them in which the Tsimshian seem to have had the upper hand: the Haida had to learn Tsimshian while the Tsimshian – claiming to speak the only true language – did not learn Haida (according to a missionary who worked among both peoples). In pre-contact times the balance of power must have fluctuated back and forth in favour of one people or the other, and there are many local accounts of Haida raids on Tsimshianic-speaking communities.

    As for Haida (Xaadas – Haida people, Xaad kil – Haida language), the term is apparently derived from Xáad – one’s father, ….

    This must be a folk etymology which simply cannot be true about the origin of the Haida self-naming. Why the i in the name Haida, impossible to justify from Xaad? Why the y in the older spelling Hyda as in Hydaburg in Alaska? Why the final h in the (also older) alternate Hydah?

    It is very relevant that the Nisqa’a equivalent is Háy’dax_, where the y’ is a glottalized y, realized before consonant or finally as ?i (with unvoiced ). As often happens with loanwords, the original sounds are preserved in the borrowing language even though they have been changed or even disappeared from the original language. (The initial h was not original, but a replacement for X: except for Southern Tsimshian the Tsimshianic varieties do not use [X] as a prevocalic initial consonant and replace it with [h] in loanwords).

    (meaning of Xáad)
    Xáad – one’s father, …. one’s paternal uncle (father’s brother), husband of one’s maternal aunt (mother’s sister)- NOTE: This term is only used in reference to a female’s father or uncle.
    Based on this information, it seems that the Haida are a matrilineal society and that the Haida women originally spoke a Tsimshian language, but switched to language of their husbands – now known as Haida.

    A matrilineal society, yes. But a difference of language between the sexes is not quite right. There are legends about Tsimshianic-speaking women married to Haida men. In at least one case I know of, the woman is a “princess” who marries a Haida prince and has several children, who are taunted by the Haida children because they “don’t have (maternal) uncles” from whom they could inherit status and titles. So the woman decides to leave with her children to return home on the continent to join her brothers. She does this with the full accord of her Haida husband, who is sorry to see them leave but knows that his children will never have any status among the Haida and need to be with their maternal uncles in order to inherit from them the status he could not give them (his own status passing on to one of his nephews).

    This concern is mostly for the boys: girls may have had closer relationships with their Haida fathers, since one term of address by Nisqa’a women to their fathers used to be Hadiy’ (Haida Xáad + -iy’ ‘first sg pronoun).

  72. The Haida–Tsimshian relationship sounds a little like the Roman–Greek one in classical times. Romans took Greek slaves, not vice versa, but Romans looked up to Greek culture as superior to their own, and even had a tradition that their law code (one of the most Roman things about them) came from Greece originally.

  73. the Nisqa’a equivalent is Háy’dax̠, where the y’ is a glottalized y, realized before consonant or finally as ?i (with unvoiced i). As often happens with loanwords, the original sounds must have been preserved in the borrowing language even though they have been changed or even disappeared from the original language.

    Are there any other examples of loanwords into Tsimshian with a y’? I am asking, because in Georgian, plosives in Russian loanwords come off as ejectives. In other words, the glottalized y in the borrowed word does not necessarily mean the sound was glottalized in the source.

  74. marie-lucie says:

    Y: I can’t think of a loanword with glottalized y (which does not mean that none exist), but according to the grammatical sketch which begins the Alaskan Haida dictionary, the two Alaskan dialects differ in that one of them has lost some (back) consonants still present in the other. I am not aware either of any plosives in borrowings ending up as Nisqa’a glottalized consonants. In any case, whether the glottalization of y is original or not, the presence of a palatal approximant does not seem compatible with the lack of such a sound in the modern Haida names.

  75. Lars (the original one) says:

    @dainichi, from the very laconic list of forms in ODS it looks like ‘active’ mig tykker was the original construction, as still in Swedish (with the nominative now), and the mediopassive is very much a bleached ‘middle voice’ without any real semantic difference here. Cp. synes where it does have a passive sense, but may also have influenced tykkes.

    The close vowel is probably a spelling pronunciation, the mid-close is older.

  76. Is it possible that the palatality in sceort and perhaps bisceop was leveled in from forms with umlaut from elsewhere in the paradigm (comparative, plural – the German plural of Bischof is mercilessly Bischöfe)?

    No. First, neither sċ(e)ort not bisċ(e)op have any umlaut-inducing case endings — both are ordinary a-stems in OE. Secondly, OE velar palatalisation (as opposed to the simplification of ) is an old change — older than i-umlaut, syncope, apocope and all that jazz. Because of that, umlauted vowels fail to trigger velar palatalisation. That’s why we have king, keep, geese (OE cyning, cēpan, gēs, with [eː] < [øː]), not *!*ching, cheep, yeese.

  77. David Marjanović says:

    If “tykkes” comes from Proto-Germanic þunkijaną (“seem”, with I assume a dative experiencer),

    Accusative in German. (Unlike its replacement, mir scheint = “it seems to me”.)

    then why the need to add the mediopassive (if that’s right term) -s in Danish?

    Perhaps to justify the oddity of an accusative experiencer.

    Why the final h in the (also older) alternate Hydah?

    To clarify that the word ends in /a/ rather than /ə/. But of course there’s no vowel at all at the end of Xaad.

    Are there any other examples of loanwords into Tsimshian with a y’? I am asking, because in Georgian, plosives in Russian loanwords come off as ejectives.

    That has an easily identifiable reason, though. In Georgian, and in Caucasus-area sound systems generally, all plosives and affricates are either voiced or aspirated or ejective (or long in some languages); there are no plain ones. The ones closest to plain are the ejectives; they’re lenes (very much unlike in North America), so they’re the closest thing to the Russian unaspirated voiceless plosives (or the Greek ones for that matter), even though those are fortes.

    Because of that, umlauted vowels fail to trigger velar palatalisation.

    Oh, that’s why! I had been wondering if the roundedness blocked the change somehow, which it hasn’t in comparable cases in other languages…

  78. Lars (the original one) says:

    Accusative in German, but dative in Old Norse. ON also used the mediopassive with a nominative subject (and predicate) to mean ‘consider oneself’ — arguably it’s the experiencer that appears in the original reflexive suffix (the distinction between accusative and dative reflexives in verb forms disappeared very early), but once grammaticalized as a mediopassive the experiencer role would probably move to the subject — and conflation with the active use seems likely.

  79. marie-lucie says:

    Haida again
    (m-l) the presence of a palatal approximant does not seem compatible with the lack of such a sound in the modern Haida names.

    In my focus on the difference between Xaad ‘father” and Hay’dax_ I forgot the probably older, now conservative Haida form Xaayda for the ethnonym, which still has the y, although not the glottalization. It is likely that glottalized y’ was in the original Haida word, since there is no reason for the Nisqa’a form to have glottalized an original y in this borrowed word, and the sequence ay is common in the language, especially in borrowings from Coast Tsimshian where the sequence is very common.

  80. marie-lucie says:

    Egyptian retention?

    AK, thank you for mentioning the relevant comment. I looked it up and discovered that not only I had read it, I had also commented, along the same lines as here.

  81. marie-lucie says:

    David M: (m-l) Why the final h in the (also older) alternate Hydah? – (DM) To clarify that the word ends in /a/ rather than /ə/.

    That’s a very good reason, but it could also be that there was some aspiration as the result of the weakening of an original X as attested in Nisqa’a Hay’dax_. Since the languages of the area have “unEnglish” sounds, and they were written down by non-linguists over a period of time, those people’s spelling conventions tend to vary quite a bit (there are at least 20 recorded ways of spelling “Tsimshian”).

    But of course there’s no vowel at all at the end of Xaad.

Speak Your Mind

*