An Outbreak of Greenglish.

Helena Smith reports for the Observer on yet another moral panic about foreign incursion on precious linguistic purity:

Usually, Professor Georgios Babiniotis would take pride in the fact that the Greek word “pandemic” – previously hardly ever uttered – had become the word on everyone’s lips. After all, the term that conjures the scourge of our times offers cast-iron proof of the legacy of Europe’s oldest language. Wholly Greek in derivation – pan means all, demos means people – its usage shot up by more than 57,000% last year according to Oxford English Dictionary lexicographers.

But these days, Greece’s foremost linguist is less mindful of how the language has enriched global vocabulary, and more concerned about the corrosive effects of coronavirus closer to home. The sheer scale of the pandemic and the terminology spawned by its pervasiveness have produced fertile ground for verbal incursions on his mother tongue that Babiniotis thought he would never see. “We have been deluged by new terms and definitions in a very short space of time,” he told the Observer. “Far too many of them are entering spoken and written Greek. On the television you hear phrases such as ‘rapid tests are being conducted via drive-through’, and almost all the words are English. It’s as if suddenly I’m hearing Creole.”

With nine dictionaries to his name, the octogenarian is the first to say that language evolves. The advent of the internet also posed challenges, he concedes, but he has never opposed adding new words that translated and conveyed technological advances. “I included them in the Lexicon,” he says of his magisterial 2,500-page dictionary of modern Greek language. “But where possible, I also insisted that if they could be replaced by Greek words they should. I came up with the word diadiktyo for the internet and am glad to say it has stuck.” […] “For Greeks, language has always been a sensitive issue,” says Babiniotis. “I know what I say troubles some, but it is the duty of a linguist to speak out.” Babiniotis’s protestations have been fodder for cartoonists and the butt of debate. But he is not alone.

The emergence of “Greenglish” – Greek written with English letters – as an unofficial e-language since the arrival of the internet has also sparked alarm. Facebook groups have emerged, deploring the phenomenon. “A lot of youngsters use it to message one another because they think it’s easier,” says Susanna Tsouvala at the Polyglot Bookstore, which specialises in foreign language textbooks in central Athens. “Spelling’s easier and they don’t have to use the accents required in Greek, but ultimately it’s going to be our language’s loss.”

For many, book publishers have become the last line of defence. At Patakis, one of the country’s most established publishers, inclusion of foreign words in any work is carefully monitored. “Books are guardians of the language,” insists Elena Pataki, whose family-run firm publishes books for all ages. “We recently published a business book about family-owned enterprises and made a conscious choice to limit references to foreign terms.”

OK, in the first place, “Greece’s foremost linguist” is absurd. Not only should English journos not be in the business of anointing foreign scholars as “foremost,” Babiniotis isn’t a linguist at all — he’s a philologist, lexicographer, and former Minister of Education and Religious Affairs (God save the mark). Actual linguists don’t go around pontificating about the horror of loanwords. (Artemis Alexiadou is an example of an actual Greek linguist; she doesn’t practice my kind of linguistics, but she’s definitely in the field.) In the second place, book publishers carefully monitoring inclusion of foreign words is ridiculous: you’re going to wind up producing books in a language no one speaks, just like in the bad old days of katharevousa. And in the third place, there’s nothing wrong with loanwords, amigo. They’re just words. (Thanks, Trevor!)

Comments

  1. David Eddyshaw says

    Europe’s oldest language

    I think we’ve established that that should be Welsh. (That was on the BBC and must be true.) All other European languages are of Dravidian origin (especially Basque), except the Scandinavian languages, which are an independent branch of KONGO, and were brought to Europe during the Viking era.

    [Admittedly, this is slightly less stupid than some such claims. Still stupid …]

  2. Quite so.

  3. Maybe Greenglish sounds better in Greek than in English. I find the implied /iŋg/ impossible according to American English phonotactics.

  4. I have to say, when I saw the title I was looking forward to Greenlandic-flavored English.

  5. David Eddyshaw says

    Yes, so was I. (Ideally, polysynthetic.)

  6. There’s a politics to loan words, and I look forward to reading about this someday. I’ve been in situations where my conversation partner has peppered his speech with English loan words, unnaturally so, to convey his attitude to American culture, and signal to me just how sophisticated and well-educated he is; and how very, very impressed I should be.
    And the other way happens also, where someone avoids loanwords, for various reasons.

  7. J.W. Brewer says

    Didn’t they already get rid of the accents (or maybe just most of them) in the orthography of Modern Greek when katherevousa fell from power?

  8. Third to Y and David, but also did a spit-take when I encountered previously hardly ever uttered. I’m sorry but pandemic was hardly an obscure word; a quick Google Ngram query registers it as consistently more frequent (before 2019) than, for example, nightlife.

  9. J.W. Brewer says

    Relatedly, here’s a just-published anti-loanwords piece from a different direction, complaining that a particular loanword-heavy variety of AmEng should be changed because the variety’s sociolinguistic effects (promoting in-group solidarity and marking outsiders as outsiders) are inconsistent with what the relevant group’s attitude toward welcoming and assimilating outsiders (or at least a certain target group of outsiders) ought, in the author’s opinion, to be. That “ought” is based on a non-linguistic view about non-linguistic goals and priorities, but that said the discussion of how the language variety works seems reasonably free of gross misunderstanding of linguistics. (DISCLOSURE: I know some young people who know the author and I I’ve met him once or twice, maybe five years back.)

    https://www.tabletmag.com/sections/community/articles/politics-and-the-yeshivish-language

  10. PlasticPaddy says

    Grenglish (also known as gringlés) is the language gringos speak.

  11. David Eddyshaw says

    In Kusaal, all Europeans and Americans speak Nasaal “Whitish.” Finer distinctions exist, but are naturally only of interest to specialists and geeks.

  12. Yeshivish at LH: 2016, 2020.

  13. Matthew_M says

    Similarly, in the parlance of older/rural Thais, all Europeans and Americans speak Farang. These days, Phasa Farang refers to English by default, not French as you might expect by etymology.

  14. Didn’t they already get rid of the accents (or maybe just most of them) in the orthography of Modern Greek when katherevousa fell from power?

    They got rid of the polytonic orthography, so there is just one accent (still called a “tonos”) in use, but it is quite widespread. Every polysyllabic word has one and some monosyllabic ones do as well.

    What I find a little disturbing is that apparently you need two key strokes to enter an accented letter on a Greek keyboard. Or at least this is the case I get after installing Greek keyboard driver on an originally non-Greek computer. When I write in Czech I have no problems with accents whatsoever, because the accented letters (except the rarest ones) have their own keys, so using accents does not slow me down. Needing an extra stroke in Greek is a bit annoying.

    On the other hand, many Czechs write without accents as a habit, even if it does not save them any time or effort. This used to be explained by the fact that in the olden days many computers did not have accented letters available, but apparently today the non-accenters include young people who cannot remember those times.

  15. Athel Cornish-Bowden says

    OK, in the first place, “Greece’s foremost linguist” is absurd. Not only should English journos not be in the business of anointing foreign scholars as “foremost,”

    The Observer was guilty of something else similar today, when I read about “How early humans’ quest for food stoked the flames of evolution”, which struck me as handwaving speculation typical of popular science writers whose study of evolution is limited to what they’ve read in Scientific American (at best). Nonetheless, according to the newspaper the authors are “two leading academics”. A quick search hasn’t revealed anything to support that description.

  16. Athel Cornish-Bowden says

    many Czechs write without accents as a habit,

    When email first became widespread I noticed that all messages in Spanish that I saw were written without accents. (I think my wife still doesn’t bother with accents.) There was, however, a glaring exception: absolutely no one was willing to write ñ as n. Some wrote it as ny (OK), others as ni (horrible in my opinion), and I think there was a third that I can’t think of at this moment). As an anglophone person I tended then to think of ñ as n with a ~ over it, but apparently people whose native language is Spanish see it as a separate letter in its own right (as it is, of course) quite different from n.

  17. Athel Cornish-Bowden says

    In Kusaal, all Europeans and Americans speak Nasaal “Whitish.” Finer distinctions exist, but are naturally only of interest to specialists and geeks.

    When I went to an archaeological site in Pueblo in 2003 the notices were in Spanish, Nahuatl and English. To my eyes Nahuatl looks very different from any European language — even Basque. However, there was a young Japanese couple there who asked me if the Nahuatl was French. I was a bit taken aback by the question, but afterwards I thought it was reasonable, as for a Japanese speaker English, French and Nahuatl may well look equally alien.

  18. From early-90’s Usenet newsgroups I also remember the Portuguese-style ‘nh’ as another option for Spanish ñ, but as you say no one tried to do anything to replace the acutes. An interesting challenge was coping with French with all the accented characters simply deleted on my system and having to be guessed (as people were insisting on using extended character sets when there was no guarantee they would be transmitted given the Internet of the time) – a double space therefore meant à with the actual word deleted. (One reader in Japan was complaining that for him each French accented letter was systematically replaced by a kanji… for me, that would have been an actual improvement over what I was seeing). In 1995-6 I was teaching in a Czech school where the school student newspaper decided to print the (fascinating) accounts of some of their students on an exchange program in the US – but as received via unaccented ASCII email rather than properly retyped, which made it just a little more challenging for a foreign reader.

  19. “Feliz ano nuevo”…

  20. David Marjanović says

    Not only should English journos not be in the business of anointing foreign scholars as “foremost,”

    Journos should, after 150 years, finally let go of the view of science as a hierarchy of respect where people worship each other’s brilliance, and somehow lead each other, according to everyone’s place in the pecking order.

    Nasaal

    Standard Average European.

    apparently you need two key strokes to enter an accented letter on a Greek keyboard

    Also in Spanish. On the German keyboard layout, ´ and ^ are likewise keys of their own, and ` is Shift+´. (And the degree sign, °, is ^+Shift.)

    ä, ö, ü, ß have their own keys, and their traditional workarounds ae, oe, ue, ss are hardly used in e-mails or the like anymore; sz seems to have died out in any use in the 1990s.

  21. I was a bit taken aback by the question, but afterwards I thought it was reasonable, as for a Japanese speaker English, French and Nahuatl may well look equally alien.

    Similarly, what percentage of anglophones would be able to tell the difference between written Hindi, Punjabi, Bengali, Telegu, and Tamil?

  22. A. Sasportas says

    “There was, however, a glaring exception: absolutely no one was willing to write ñ as n.”

    That’s because ano is the Spanish for ‘anus’. I am sure that “absolutely no one was willing….” was limited to the word for ‘year’ (aňo).

    It soon became a convention, in the Spanish-speaking world when computers became possible to write ni if you had no way of writing ň.

    Some years ago I saw an advertisement aimed at Hispanophones that said (I’m translating) “With our software you’ll be able to spell aňo the right way” and I’m sure every reader who was then facing the problem of not being able to write ň took the hint.

    P.S. Whenever speaking Latin, be sure to pronounce the second word with a geminate: anus annus.

    :An interesting challenge was coping with French with all the accented characters.”

    Vietnamese spelling requires even more diacritics than French spelling.

  23. When my daughter was little, she had an electronic keyboard with all the letters of the alphabet, and it had both English and Spanish modes. I could do several things, like say the name of each letter pressed, or a word that started with that letter (in the appropriate language), or play a tune for each letter. The songs included “Row, Row, Row Your Boat,” “Red River Valley,”* and similar standards. The alphabetic keyboard had twenty-seven letter, because it included an ñ, but the button did nothing when the language was set to English. In Spanish mode, the tune played by that button was “La Cucaracha.”

    * That tune is actually more associated in my mind when the these alternative lyrics by Woody Guthrie., about the Abraham Lincoln Brigade who fought for the Republicans in the Spanish Civil War. (My father told me that in 1972, when he had a summer job crunching numbers for the Army, between graduate school and medical school, the Lincoln Brigade was among the organizations that having belonged to would have precluded him from getting the low-level security clearance needed for the position. Another one he remembered was the Dante Alighieri Society, which was pro-Mussolini in the 1930s, but which has been a legitimate cultural group since at least 1948. I used to walk by their chapter headquarters on the way from MIT to North Cambridge all the time.)

  24. Terry K. says

    Maybe Greenglish sounds better in Greek than in English. I find the implied /iŋg/ impossible according to American English phonotactics.

    Depends whose American English phonotactics. For me, “English”, (along with words like sing, bing, ring) has same vowel as “Greek”. (The vowel of “beat”, not the vowel of “bit”).

    My difficulty with Greenglish is, seeing it written, it looks like it starts with “green”, green-glish.

  25. It should be greeklish.

  26. And in fact Greeklish is considerably more common.

  27. Similarly, what percentage of anglophones would be able to tell the difference between written Hindi, Punjabi, Bengali, Telegu, and Tamil?

    That would be actually easier (than differentiating French from Nahuatl) since the scripts are not the same. The differences between Devanagari, Gurmukhi and Bengali scripts could be compared to the differences between Latin, Greek and Cyrillic (or perhaps Antiqua, Fraktur and Gaelic uncial). Tamil and Telugu scripts are even further away and don’t even look similar to the North Indian scripts. So the identification can be purely graphical. I can recognise written Tamil even though I cannot read it.

    A better analogy may be an English speaker telling the difference between written Hindi and one of the Sino-Tibetan languages that use Devanagari in their written form.

    On the other hand, Japanese speakers usually know Romaji much better than Anglophones know Brahmic scripts, which spoils the symmetry a bit.

  28. January First-of-May says

    A better analogy may be an English speaker telling the difference between written Hindi and one of the Sino-Tibetan languages that use Devanagari in their written form.

    Hebrew and Yiddish, perhaps (regularly confused, though admittedly also a less common script). I think I’ve heard of a few cases where written Kazakh was confused for Russian.

    Maybe the (nearly) ideal analogy would be someone seeing a text written in Sorabe and asking if it was Farsi…

  29. PlasticPaddy says

    https://en.m.wikipedia.org/wiki/Kazakh_alphabets
    has a sample of Kazakh written in Cyrillic script. If you are familiar with Russian you notice:
    1. additional letters, e.g. i
    2. adapted letters, e.g. q is Russian k with a “tail”
    3. Impossible phonetics, e.g. qy in “quqyqtary”
    But many words (9/24 in the sample) look like possible Russian words.

  30. David Eddyshaw says

    Hebrew vs Yiddish is a very good parallel, especially given that the two languages are genetically unrelated*; though it would be even harder for anyone who didn’t know Hebrew or Aramaic to identify the Aramaic passages in the Bible, even if they knew the individual letters.** (You actually could do this without knowing the languages, for example by noticing that a whole lot more words suddenly seemed to end in aleph, but it would be much more difficult than telling Yiddish from Hebrew.)

    *Moreover, French and Nahuatl (unlike Hebrew and Yiddish) are both polysynthetic, so confusion is natural. These typologically exotic languages are hard to keep apart in one’s mind, especially when your native tongue is a well-behaved left-branching human-bog-standard agglutinative SOV type.

    https://www.academia.edu/2000636/Grammaticalization_of_polysynthesis_with_special_reference_to_Spoken_French_?email_work_card=view-paper

    **They might be the one person in the world who knew Ladino but couldn’t recognise Hebrew when they saw it. I mean, it’s logically possible, right?

  31. J.W. Brewer says

    I’m belatedly a bit confused by Brett’s point about English phonotactics blocking “Greenglish.” The most hard and fast rule I can find is that the ŋg cluster never appears in word-final (or word-initial) position. It sometimes appears intervocalically and other times not, as witness pairs like anger/hanger and finger/singer that do not rhyme with each other. Is the claim that it cannot follow certain vowels but can follow others, such that whether FLEECE or KIT is the vowel assumed for the first syllable of “Greenglish” is crucial?

  32. David Eddyshaw says

    There’s David Greenglass, who changed his testimony to implicate his sister Ethel Rosenberg rather than his own wife, who probably did the actual typing that Ethel Rosenberg was executed for.

    But there’s a morpheme break in there, and moreover the /ŋ/ has developed secondarily by assimilation to the following /g/. Offhand, I can’t think of any English words with /i:ŋ/ or /ijŋ/ (as opposed to /ɪŋ/) where that isn’t so, and I think Brett is right that there’s an actual phonotactic constraint there, for all that the pronunciation entails no particular difficulty. We don’t need no steenking examples.

  33. J.W. Brewer says

    Since “English” is pronounced as if “Inglish,” I see no reason why “Greenglish” should not be pronounced as if “Gringlish,” which presents no potential phonotactical difficulties. Rhymes with “Singlish,” for example.

  34. *Moreover, French and Nahuatl (unlike Hebrew and Yiddish) are both polysynthetic, so confusion is natural. These typologically exotic languages are hard to keep apart in one’s mind, especially when your native tongue is a well-behaved left-branching human-bog-standard agglutinative SOV type.

    A related discussion.

  35. PlasticPaddy says

    @de
    Is any long vowel apart from long a (e.g., anger, dangle) allowed to precede either of the two versions of n+hard g in the English language? I think the a is even long only in some accents.

  36. That would be actually easier (than differentiating French from Nahuatl) since the scripts are not the same.

    Very true, but unfortunately I would still bet that the percentage of non-South Asian anglophones able to tell the difference is below 1%. I suspect the percentage of anglophones able to differentiate Korean, Japanese, and Chinese would be much higher, but certainly below 5%.

  37. David Eddyshaw says

    @Fancua:

    Thanks!

    The clitic-versus-affix discussion there reminds me of the classic and much-cited Pullum/Zwicky paper on English “n’t”, which is similarly both discombobulating and convincing:

    https://www.researchgate.net/publication/244421136_Cliticization_vs_Inflection_English_N'T

  38. David Eddyshaw says

    @Plastic:

    I think it’s the /ŋ/ itself rather than the clusters /ŋk/ or /ŋg/ as such which triggers the phonotactic rule (if rule it be.)
    I think you may be right that other “long vowels” are also not found before /ŋ/. I don’t have your proposed exception /ɑ(:)ŋ/ myself either, but introspection is a particularly unsafe guide for me in this matter as my idiolect is non-standard wrt vowel length. (Yr Alban am byth!)

    I’m not sure how far this is a synchronic rule (if rule it be) rather than the contingent outcome of the history of the language, though. How do Australians pronounce “boong” for example (not that they should of course. But how would they pronounce it if they did?)

  39. unfortunately I would still bet that the percentage of non-South Asian anglophones able to tell the difference is below 1%

    Perhaps this may be different in the UK with its history and sizable Indian minorities, but from much of Europe, I guess, India appears to be rather monolithic. I’d be surprised if the percentage of Europeans who know that Telugu language exists would be higher than low single digits. Recognising the script is the next level.

    (Although a language and geography nerd myself, I always confuse Telugu, Kannada and Malayalam scripts. Tamil script is a different thing though, because it is ugly.)

  40. David Marjanović says

    I’m not sure how far this is a synchronic rule (if rule it be) rather than the contingent outcome of the history of the language, though.

    Both: formerly long vowels don’t occur before former consonant clusters, and every /ŋ/ is in a current or former consonant cluster; and there hasn’t yet been much opportunity to fill this gap with loans or new developments (I’m curious, too, how anyone pronounces “boong”…!).

    (…It’s strange, actually, that /ng/ didn’t count as homorganic in the pre-Orrmulum lengthening of vowels before homorganic clusters in monosyllabic words, the phenomenon that has brought us mild, wild, child, grind, wind and the like – the other wind was later extracted from windy and windmill which weren’t monosyllabic words and therefore escaped the lengthening.)

    The exception are the accents in the handwavy-northeastern US where most vowels are raised before velar consonants, /ŋ/ included, so that egg ends up with FACE instead of DRESS, angry ends up with FACE instead of TRAP, dog and strong end up with THOUGHT instead of LOT, and sing, bing, ring, English end up with FLEECE instead of KIT. (I didn’t know about that last one, but it makes perfect sense.)

  41. David Eddyshaw says

    Ah, that explains what PP meant about angry. I misunderstood.

  42. David Eddyshaw says

    Regarding the reanalysis of French clitic pronouns as affixes, I’m reminded of a similar tendency among typologists and comparativists looking at West African languages, who have sometimes decided that the clitic subject pronouns and preverbal particles typical of the region must actually be flexions, largely (I think) because the corresponding morphemes are flexional prefixes in Bantu, rather than because of anything about the West African languages themselves.

    To be fair, the question often ends up being more a matter of definition than description. “Word” can be a very slippery concept.

  43. Giacomo Ponzetto says

    Arkadiev’s paper on French polysynthesis is far beyond my depth, but Jacques feels a bit more accessible and makes me wonder to what extent this is meant to be a French-specific argument.

    One the one hand, French is special among Romance languages for requiring subject pronouns. On the other hand, the numbered examples seem to work rather analogously for Catalan, Italian and Spanish. Though the more I think of it, the less confident I am I can tell in any language, even Italian, whether some particularly abstruse combinations of clitics are merely unidiomatic or altogether ungrammatical.

    In fact, my Spanish grammar (Manual of the NGLE) says that:

    El apoyo formal que necesitan los pronombres átonos por parte del verbo al que se adjuntan los asimila en alguna medida a los afijos. La asimilación no puede ser total porque las posiciones que ocupan los pronombres átonos no son análogas a las que permiten los afijos, y también porque estos no desempeñan funciones sintácticas.

  44. Since John Cowan linked to this at another thread, I’ll link to it here.

    It could be seen as an argument (roundabout, it’s true) for developing local/native vocabulary rather than relying heavily on loanwords. At the very least, it throws a dash of cold water on “the more loanwords the merrier” school of lexical development.

    Why having no word for X can matter.

  45. John Cowan says

    In any case, KIT/FLEECE is always neutralized before eng. I first learned that some people use FLEECE rather that KIT when I wrote “We don’t need no steenkeeng badges!” on the Conlang list, and someone questioned the need for the second “ee”.

    handwavy-northeastern US

    The Northern Cities Shift is in effect westward from western New York State along the line of the Great Lakes as far west as Cedar Rapids, Iowa. It also includes the Lower Peninsula of Michigan which is sandwiched between Lakes Huron and Michigan, a single body hydrologically). There are two southward arms, one on the western end as far as St. Louis, one on the eastern end as far as Allentown, PA. Erie, PA is an exception; it is unshifted despite being on the lakeshore between Buffalo in the east and Cleveland in the west. In no universe can any of this be called the Northeast, even handwavily.

    Around 1965, when I went to visit my cousins in a suburb of Detroit, I already noticed how different their accent was from mine, though it was an impressionistic rather than a detailed view.

    The Greek word for Greeklish/Greenglish is φραγκοβλάχικα lit. ‘Frankish-Welsh’.

  46. In no universe can any of this be called the Northeast, even handwavily.
    If you’d divide the contiguous U.S. into four quadrants, most if not all of that area would be in the North-Eastern quadrant. It’s not the North-Eastern end of the U.S. and maybe not what Americans would normally call North-East, but “in no universe” is too strong, and handwavily it’s okay. 🙂

  47. David Marjanović says

    My impression wasn’t that that’s part of the Northern Cities Shift (which is not limited to following velars). But maybe my impression was wrong…

  48. David Marjanović says

    I don’t think this specifically prevelar raising is part of the Northern Cities Shift. In particular, Bernie “Northeast” Sanders has THOUGHT in strong, as you pointed out to me, actually.

  49. January First-of-May says

    In no universe can any of this be called the Northeast, even handwavily.

    Well, even assuming that the area description is correct, it is perhaps the east of the north rather than the north of the east (which is what “the Northeast” usually means in a US context), but either way it’s pretty northeastish; and, for that matter, if you start in Kansas (where the US center is typically located) and go northeast you’d end up somewhere around that area.
    Certainly any of the other seven directions would fit even worse. (North-northeast, maybe.)

  50. John Emerson says

    The Midwest has traditionally been thought to start with Ohio. The East thus gets far less than half the space but a much larger proportion of the population. I’ve always thought “East / Midwest / West / Far West, with the West beginning with the Dakotas. (It takes nine or so Western states to match the population of Illinois, so p pop populationwise The West shouldn’t even be regarded as a major section. .)

  51. David Eddyshaw says

    populationwise the West shouldn’t even be regarded as a major section.

    In view of the geographical mismatches, I would suggest “Bush” rather than “West”; or if this is felt to be too political, perhaps “Outback.”

  52. January First-of-May says

    IIRC “flyover states” is the usual relevant descriptor. I don’t recall offhand if there is another.

  53. J.W. Brewer says

    The current U.S. “Midwest” is mostly the “Old Northwest,” which was the “Northwest” from the point of view of the 1780’s when the nation’s western boundary was still the Mississippi and the center of population had not yet made it across the Appalachians. As the actual “West” shifted across the Mississippi en route to the Pacific, the “Mid-” qualification got added, but for some reason the Old Southwest (Alabama, Mississippi, maybe parts of Tennessee) just got lumped in with the “South” without further qualification, with the new Southwest starting maybe in western Arkansas.

    Folks from places with the NCS like Buffalo and Rochester typically do not self-identify as Midwesterners, although that isn’t to say they self-identify as “Northeasterers,” which is a less coherent concept anyway. But Allentown isn’t not in the Northeast by most plausible definitions of Northeasternness. (The exception would be those approaches that break the country into 8 or 10 regions rather than 4 or 5, such that “MIddle Atlantic” becomes its own category contrasted with the Northeast.)

  54. John Emerson says

    Friends of my son’s from Buffalo discussed this question and decided they should self-identify as Canadians.

  55. “Flyover states” isn’t the same as “The West”. Speaking as one who lives in one but not the other.

  56. @J.W. Brewer: I find that the old Northwest Territory is sometimes a useful way of referring to a specific group of states. However, that description relies on a certain base of assumed knowledge, and a lot of Americans have forgotten (or never learned) where the territory was. I imagine that people from the states in question might be more likely to know about the Northwest Territory, especially as one gets closer and closer to Evanston, Illinois, where Northwestern University is located.

  57. David Marjanović says

    IIRC “flyover states” is the usual relevant descriptor. I don’t recall offhand if there is another.

    “The Heartland” of course. Guess which political orientations the two terms line up with.

  58. January First-of-May says

    the old Northwest Territory

    …not to be confused with the old North-Western Territory, or its modern descendant, the Northwest Territories.

  59. J.W. Brewer says

    @Brett. As the holder of a degree from Northwestern University, I have hands-on experience explaining to my children why it would be called that even though they don’t think of its location as being in the ‘Northwest.” The college in Memphis formerly known as Southwestern Presbyterian is now Rhodes College, w/ no compass-point heading specified, eliminating the need for similar explanations.

    The pejorative “flyover country” can cover everything in between the western edge of the NYC-or-DC suburbs and the eastern edge of the LA-or-SF suburbs, so is probably wider in scope than most of the other descriptors referenced above.

  60. ktschwarz says

    As David M said, he’s not talking about the Northern Cities Shift, which affects vowels in all environments, not just pre-velar. But pre-velar raising—especially, pre-eng—isn’t northeastern, either. If it’s any region, it’s Western American. Some phonetics blogs that identify it as either Western or California:

    Literal-Minded: Engma Enigma (Neal Whitman grew up in Texas, though he doesn’t say so in the post)
    John Wells: iːŋ
    Dialect Blog: What’s with the Western US and Velars?
    Ace Linguist: Pre-Velar Raising

    But I wonder if it’s more widespread as an individual variation. In the History of English Podcast, I keep hearing about “keengs” of England. I hear /i/ in his linguist, distinct, and wing as well, though not before g, only before ng. The host Kevin Stroud is from eastern North Carolina—several listeners have pegged him by his accent—but as far as I know, that’s not part of a Southern accent. (I do hear a little Southernism in some other vowels: his type sounds more like “tahp”, now more like “nay-ow”.)

  61. David Marjanović says

    Curiouser and curiouser. (And of course I commented on John Wells’s thread that I completely forgot afterwards – though my comment wasn’t about the original topic.)

    now more like “nay-ow”

    My colleague from Nova Scotia does that: two syllables, [ˈneːaʊ̯].

  62. John Cowan says

    To suppose that the U.S. Northeast is the north-eastern quadrant of the contiguous 48 states is the etymological fallacy. In the UK the part of Britain where there is no FOOT-STRUT split is “oop North”, not “oop Middle”.

  63. David Marjanović says

    It’s the north of England. Scotland is hyperborean.

  64. The Literal Minded post that ktschwarz linked to in their comment brings out for me the distinction between phonemes versus the actual pronounced sound. If the sound is in between the prototypes of the two phonemes, then there certainly might be some randomness in what phoneme people think of it as. Two people in the same speech community who say it the same might think of it differently.

    I think my actual sound in bang maybe closer to that of ben than that of bane (and definitely not the vowel of ban). I think of it as the vowel of bane, but I recognize it’s different than the a of bane. (And, note, bag does not rhyme with vague for me. No general velar raising.)

  65. David Eddyshaw says

    Scotland is hyperborean

    Enough of this Anglocentrism.
    The correct view is that England is hypoborean.

  66. J.W. Brewer says

    Thus the well known alternative Spanish name for England, “Baja Escocia.”

  67. David Eddyshaw says

    “Y Dwyrain Pell” is also acceptable.

  68. Per Howard, the dry bed of the Baltic Sea is Hyperborea, west of ancient Cimmeria and Asgard. More southerly realms conquered by the Baltic indigenes became too the shortened name Hyborea.

    Per Smith, Hyperborea was farther north, but was obliterated in the Pleistocene, as foretold by the White Sybil of Polarion.

  69. Who spoke Polari, of course.

  70. Andrej Bjelaković says

    @David Marjanović

    I think dog and strong go with THOUGHT in all American accents, nothing to do with the Midwestern egg, bag etc.

  71. David Eddyshaw says

    Who spoke Polari, of course.

    Bold!

  72. January First-of-May says

    Enough of this Anglocentrism.
    The correct view is that England is hypoborean.

    Let’s be honest here, England is hyperborean and Scotland is ultraborean.

    Regarding the actual Hyperborea, it seems to be 1) north of the Riphean Mountains, and 2) warm for its latitude. The Riphean Mountains are “the source of <…> several large rivers (the Dnieper, the Don, and the Volga)”, which would put them in the Valdai Hills, in which case Hyperborea roughly corresponds to the Republic of Novgorod (an area that does get significantly warmed by the Gulf Stream), and/or perhaps the general vicinity of what is now Murmansk (even more warmed).

  73. David Marjanović says

    I’ve come across the idea that Hyperborea was beyond Siberia – but not actually north of it: China.

    I think dog and strong go with THOUGHT in all American accents

    That would make the “dawg” eye-dialect spellings superfluous.

  74. @David Marjanović: There is a long tradition of using eye dialect to represent standard pronunciations by rural, uneducated, or otherwise disdained speakers. It has been discussed numerous time as Language Log (e. g. here, here, and here).

    For another example, in this song “American Dawg”, the eye dialect is explicit in the lyrics, but the pronunciation is standard for dog. Neal Gladstone stretches out the vowel a bit in the lead vocals, but Barbara Gladstone and Audrey Perkins, singing backup, don’t even really do that.

  75. J.W. Brewer says

    Similarly, this dictionary gives “hawg” as a variant American spelling of hog, calling it a “pronunciation spelling,” which I guess is like a spelling pronunciation in reverse. https://www.collinsdictionary.com/us/dictionary/english/hawg

  76. J.W. Brewer says

    BTW, the ancient Greeks may have started with skewed expectations about the relationship between latitude and climate, because e.g. Athens has notably milder winters than most cities around the world at the same approximate latitude (out of an intentionally varied sample of six non-Mediterranean alternatives I checked, Stockton, Cal. was the only one that wasn’t dramatically colder on an average January day, although).

  77. David Eddyshaw says

    skewed expectations about the relationship between latitude and climate

    Britain is (more or less) at the latitude of Hudson’s Bay and Kamchatka.
    (However, this is unlikely to impress a Russian, given that St Petersburg is at the latitude of Shetland.)
    And as for all you Scandinavians …

  78. Lars Mathiesen says

    What about us? I’m at about the same latitude as Glasgow, and the weather is worse.

  79. David Eddyshaw says

    and the weather is worse

    Impossible, surely? All my childhood memories are of rain (except when it was snowing, of course.) And chilblains.
    (Young people of today, never seen a chilblain …)

  80. Don’t tell them, they would refuse to fight against climate change.

  81. John Cowan says

    φραγκοωλάχικα

    Arrgh, annoying Greek keyboard: that should have been φραγκοβλάχικα, i.e. frangovlakhika. Pkease fix.

    Friends of my son’s from Buffalo discussed this question and decided they should self-identify as Canadians.

    I would be happy to self-identify as a Canadian, but Canadians wouldn’t agree with me.

    Hebrew vs Yiddish is a very good parallel

    I don’t have much problem distinguishing them given enough running text, despite not knowing either language. Hebrew is either vowel-free or fully voweled. Yiddish is voweled very systematically and is in fact a true alphabet: a mark on aleph, vav, or yod means it’s a vowel, the absence of a mark means it’s not, and peh and feh have a distinguishing mark each. However, Hebrew borrowings in Yiddish are fully voweled, so mostly there are just occasional marks until at some point you see a word with plenty of marks.

    That would make the “dawg” eye-dialect spellings superfluous.

    For me dog is unique among -og words in being CLOTH=THOUGHT, all the rest are LOT=PALM. Of course, half the country and all of Canada have the THOUGHT=PALM merger, so it makes no difference for them: it’s more palmy in the West and more thoughtful in Eastern Mass and Canada. ~~ snicker ~~ Indeed, Eastern and Southern Americans plus AAVE-speakers are among the few worldwide for whom CLOTH is actually an important lexical set any more. However, even those with THOUGHT=PALM have awwww! ‘pity, ironic pity’ distinct from ahhhhh! ‘satisfaction, comprehension’, so they can hear the difference even if they don’t use it.

  82. Yiddish is often written unpointed these days. The Yiddish Wikipedia entry on Yiddish has some paragraphs with points, others without. That includes Hebrew borrowings.

    Yiddish has much more of the double glyphs וו and יי. Yiddish has the single-letter article א while Hebrew has no one-letter words (except abbreviations, which are marked with a geresh, as א׳.) Beyond that there are various affixes characteristic to each, but then you need to start learning to recognize the letters.

  83. David Marjanović says

    it’s more palmy in the West and more thoughtful in Eastern Mass and Canada.

    Depends on where in Canada, I guess – purely palmy for the mentioned Nova Scotian.

    (The farther you get away from the actual palm trees…?)

  84. J.W. Brewer says

    It’s notably warmer in Vancouver in January (Hyperboreanly so?) than it is in Halifax, despite the former being five degrees of latitude farther north. There are indeed palm trees in Vancouver although apparently from a few specific species that are hardier than average because Vancouver winters are still non-tropical even if mild compared to most of the rest of Canada.

    (In fact if you go due east from Vancouver you cut across Newfoundland on your way to the Atlantic, and not right along the south coast either.)

  85. Lars Mathiesen says

    There are a few palm trees in Copenhagen too, even outside of the Palm House at the Botanical Gardens. The municipal gardeners heap gravel around the bases and wrap the crowns in felt for wintering, so it’s intensive gardening, but they’ve been there since 2015 or so with no sign of dying off. (These are not your equatorial palms, though, but Chinese windmill palms which are a highland species).

    Glasgow may have had one of the rare dry occasions when I checked the other day, in average it’s twice as wet as Copenhagen year round — but about 5C warmer in winter. The palms would probably be happy in Glasgow too.

  86. Attalea princeps, the most famous palm tree in Russian literature.

    And in looking it up, I learn that it’s inspired a violin concerto.

  87. John Emerson says

    It’s warmer in Petersburg Alaska than it is in Minnesota. 2 friends of mine moved ther for the balmy climate and became fisherman.

    America’s N Pacific coast is balmy, but unbearably gloomy and drizzly. I can’t remember whether I remember this from Trask’s anthology or just made it up, but the traditional poetry of the area mostly consists of complaints about the gloom.

  88. January First-of-May says

    the most famous palm tree in Russian literature

    Surely that title goes to the partner of Lermontov’s pine (apparently both borrowed from Heine).

  89. PlasticPaddy says

    @je
    My life is cold, and dark, and dreary;
    It rains, and the wind is never weary;
    My thoughts still cling to the mouldering Past,
    But the hopes of youth fall thick in the blast,
    And the days are dark and dreary.

    Be still, sad heart! and cease repining;
    Behind the clouds is the sun still shining;
    Thy fate is the common fate of all,
    Into each life some rain must fall,
    Some days must be dark and dreary …

    Happy St. Patrick’s day (not at all dark or dreary here)

  90. Surely that title goes to the partner of Lermontov’s pine

    Maybe? But surely not “surely”; the Garshin story is very famous.

  91. John Emerson says

    20 degrees F, snow, and sunshine is fun, and blindingly bright (literally bonding). 32 degrees, and nimbostratus, and drizzle is miserable, especially after the first full week.

  92. John Emerson says

    Literally blinding.

  93. John Cowan says

    A good day to stay inside in NYC, at least if you have my temperament, though the usual drunken idiots should not be so much in evidence in this second year of the plague.

  94. David Eddyshaw says

    New Yorkers are quite right to celebrate a great Welshman. It shows a proper spirit, in my opinion.

  95. David Eddyshaw says

    Gŵyl San Padrig Hapus i Ddinas Efrog Newydd!

  96. John Cowan says

    In principle, yes. But it’s how they do it. I on the other hand am lying on suitably green sheets while I type this, as I work best lying down. I am, as is well known, half Milesian and half Theodiscan: I may not be a Waugh (which is why I had to look up dinas), but at least I am not a Sayce.

    That lookup led me to discover that not only is the Irish cognate dún, which was easy to see in hindsight, but that the Proto-Celtic word was borrowed twice into Germanic: once post-Grimm to give the English town (and the Germans Zaun ‘fence’), but also pre-Grimm to give them down and, after a side-trip through French, dune.

  97. David Eddyshaw says

    In principle, yes.

    But he was not a Welshman, but an Armenian. And it was not in New York, but in Smolensk, at the Collective Farm Sports Festival. And they were not celebrating his feast day, but stealing his bicycle.

  98. David Eddyshaw says

    Dinas is also of interest in having swapped gender since the Middle Ages. But then, haven’t we all?

  99. But he was not a Welshman, but an Armenian. And it was not in New York, but in Smolensk, at the Collective Farm Sports Festival. And they were not celebrating his feast day, but stealing his bicycle.

    This made me laugh quite heartily. Radio Erevan and St. Patrick: two great tastes that go great together!

  100. John Cowan says

    I had hoped that someone would concoct a Radio Yerevan joke, and here it is!

    I forgot to link Sayce, a name (as Tolkien remarked) well-known especially in the annals of philology. He taught himself Latin and Greek and later Hebrew, Sanskrit, Egyptian, and Akkadian. At Oxford, his circle included John Rhys, Ruskin, and Max Müller, with whom he eventually shared the Professorship of Comparative Philology (he later became the first Professor of Assyriology). Middle Eastern archaeology and linguistics became his life’s work, though he lived to see most of it superseded: more lastingly perhaps, he wrote an 1880 book on the comparative method that emphasized the role of analogy as an essential counterpart to it.

  101. Funny, I just read about Sayce. He was one of the ones who contemptuously denounced the Shapira scrolls as forgeries. A recent analysis (here, at academia.edu; here, an article in the NYT) shows (very persuasively, to my ignorant eye) that they were likely not forgeries, and indeed the oldest known biblical text.

  102. the Proto-Celtic word was borrowed twice into Germanic: once post-Grimm to give the English town (and the Germans Zaun ‘fence’), but also pre-Grimm to give them down
    Surely the other way round?

  103. John Emerson says

    “If you enjoy seeing people puking on themselves, St. Patrick’s Day is a wonderful day,” says my ex-barmaid sister in law. My ex-bartender brother theorizes that many Irish Americans who don’t normally drink much feel obliged to tie one on on this day.

  104. marie-lucie says

    Hello from Halifax, where it is indeed colder than Vancouver, where I lived some years ago. I don’t think I have seen palm trees of any kind in Halifax, and the wind from the sea can be fierce, while Vancouver is protected by mountains to the North and a large island to the West.

    “French and Nahuatl (unlike Hebrew and Yiddish) are both polysynthetic”

    I was quite surprised to read this about French, but the reference to G. Jacques’ blog (to which a number of Hattics responded!) explained what was meant. I recalled that some years ago an American linguist (I don’t remember who) wrote about “the Chinook structure of French”. The Chinookan languages (Penutians of the Columbia estuary and valley) also have pre-verbal affixes, personal and others.

    Lately I have been reading a fair amount of Italian for relaxation – a language I never studied formally but that I can read with fair understanding using resemblances with French, Spanish and Latin. As G. Ponzetto wrote above, Italian also uses such affixes – except they are after the verb, and written together with it, without a break. Sometimes I think I am encountering an unknown word before realizing that the letters at the end are actually pronominal suffixes.

  105. David Eddyshaw says

    I recall reading some years ago (things may well have changed since) that the per capita consumption of alcohol in Ireland was much the same as in Great Britain; but that what this did not capture was that there are a lot more people in Ireland who never drink alcohol at all.

  106. A long time ago someone told me a kind of folk-nugget, about how the four (principal) peoples of the British Isles stereotypically get when drunk: The Irish get happy, the Scots get belligerent, the Welsh get maudlin, and I don’t remember about the English (maybe quiet), and I probably got the stereotypes in the wrong order, too. It wasn’t until now that I realized these behaviors correspond to the four classic temperaments.

  107. David Marjanović says

    He was one of the ones who contemptuously denounced the Shapira scrolls as forgeries.

    Quoted on p. 9 of the book:

    It is really demanding too much of Western credulity to ask us to believe that in a damp climate like that of Palestine any sheepskins could have lasted for nearly 3,000 years, either above ground or under ground, even though they may have been abundantly salted with asphalte [sic] from the Vale of Siddim itself.

    Damp climate?!?

    And he wrote this sitting in England!?!

    Anyway, I’ll read on to the Moabite Pottery Scandal. That’s a great name for a scandal at least.

  108. J.W. Brewer says

    Obligatory response to David M.: “I think I saw Moabite Pottery Scandal opening for [NAME OF ONCE-COOL BAND] at [NAME OF LONG-SINCE-DEFUNCT NIGHTCLUB] in 198_.”

  109. David Eddyshaw says

    The Moabite Pottery Scandal is surely a lost Sherlock Holmes story.

  110. I don’t think I have told this story here before: During my first and second years of graduate school, one of the other applied math students with whom I shared an office (Peter McNamara) was from Ireland. There was also an undergraduate who liked to hang around with the graduate students, in the departmental lounges and sometimes in our offices. On around March 12, the undergraduate came into our office, and this following (paraphrased) conversation about Saint Patrick’s Day ensued:

    Undergraduate: Peter, do you want to go out with me to celebrate Saint Paddy’s Day?
    Peter: I don’t know. Americans mostly don’t seem to take observing Saint Patrick’s Day seriously. They just treat it as an excuse to get really drunk.
    Undergraduate: Well, how do you know I’m not going to take it more seriously?
    Me: Once you’ve called it “Saint Paddy’s Day,” that doesn’t seem very likely, does it?

  111. Y, thanks so much for the link to the book on Shapira’s manuscripts. Boy, forget the Moabite pottery, the whole story is tailor-made for an intellectual thriller (we should hope that the genre will move on from Dan Brown).

  112. They just treat it as an excuse to get really drunk.

    That’s the adult version. Kids treat it as an excuse to pinch kids who are not wearing anything green.

  113. D.O.: It is a good story. Last I heard about it it was just an old forgery story and that is that. Maybe Hat would want to make a whole posting out of it.

    Dershowitz, judging by his work on academia.edu, is a regular iconoclast, but a very careful one, and maybe he’s generally right. His other papers argue, among other things, that Noah’s story was originally about a drought, not a flood, and that Jewish law implicitly tolerated male homosexuality. His arguments for all of these things are measured and not outlandish.

  114. Stu Clayton says

    Tread softly when carrying a big shtick.

    Observe due measure, moderation is best in all things, especially wackiness.

  115. PlasticPaddy says

    Is there a word for someone who often or even habitually lies, so that he or she is (or would be) disbelieved even when telling the truth? This would be very useful when discussing purveyors of art, antiquities, Rolexes, Gucci bags, etc.

  116. Salesman of the year.

  117. PlasticPaddy says

    @dm
    re “damp” it depends what is meant
    Egypt
    avg humidity: 56%
    avg year precipitation: 1 in

    Southern England
    avg humidity: ~70%
    avg year precipitation: 25-30 in

    So England and Egypt (perhaps also Palestine) are not that far apart in humidity.

  118. David Eddyshaw says

    Is there a word for someone who often or even habitually lies, so that he or she is (or would be) disbelieved even when telling the truth?

    “Matilda”

    http://holyjoe.org/poetry/belloc4.htm

  119. And of course the classic Boy Who Cried “Wolf.”

  120. Once you’ve called it “Saint Paddy’s Day,

    That is at least acceptable to Irish Americans, the true owners of the holiday.

    Just last week we were offered a “Saint Patty’s Day” menu by some American caterers, not of Irish heritage. My Boston friend scoffed, “do they think he was a woman?”

    Food was excellent though.

  121. Saint Patties are what they put on holy hamburgers.

  122. @PlasticSaint:

    some guy from crete, but i can’t remember which one.

    (i’ll see myself out)

  123. David Eddyshaw says

    Epimenides:

    https://en.wikipedia.org/wiki/Epimenides_paradox

    According to him, you can’t but not refuse to decline to believe a word he says. But can you?

  124. Maybe Hat would want to make a whole posting out of it.

    Done!

  125. John Cowan says

    The Irish get happy, the Scots get belligerent, the Welsh get maudlin, and I don’t remember about the English (maybe quiet)

    The traditional four stages of drunkenness are jocose, bellicose, lachrymose, and comatose. I don’t drink and never have, but Gale tells me she always skipped bellicose somehow and would stop drinking (on that occasion) because lachrymose was too unpleasant to go on with.

  126. David Marjanović says

    There are no stages. Before people get comatose or at least vomitose, they take one personality trait they have and let it shine, says my experience watching people get seriously drunk (mostly in Poland).

  127. Am I the only one who remembers (or maybe was the right age, in the right part of the country, to be surrounded by) the “Four Stages of Tequila” shirts?

    ???????? I’m rich
    ???????????????? I’m good looking
    ???????????????????????? I’m bullet proof
    ???????????????????????????????? I’m invisible

    For David Marjanović: After a 1993 visit to the Prater in Vienna, where we experienced a roller-coaster with terrifyingly inadequate safety accommodations, several of my friends and I agreed that the last stage should be changed to, “I can ride the Hochbahn.”

  128. David Marjanović says

    Oh dear.

  129. @Marie-Lucie (upthread, March 17, 7:32):

    “I recalled that some years ago an American linguist (I don’t remember who) wrote about “the Chinook structure of French”. The Chinookan languages (Penutians of the Columbia estuary and valley) also have pre-verbal affixes, personal and others”.

    I think I have found the exact reference: it is not an American linguist, but the French writer Raymond Queneau. He thought that spoken and written French could be considered separate languages, with the former but not the latter bearing a strong typological similarity to Chinook. He also believed that spoken French needed to be standardized as a separate language (which he called “néo-français”): he was a reader of the French linguist Joseph Vendryes (A fine linguist, who deserves to be much more widely read than he is today: I still have a copy of his dissertation on Latin loanwords in Old Irish, and found that his grammar of Old Irish and Thurneysen’s complement one another very nicely) and, during a trip to Greece, had been made aware of the opposition between Demotic and Katharevousa (which probably heightened his belief that spoken and written French could be considered separate languages). So, the reference:

    Queneau, Raymond. 1946 « Connaissez-vous le chinook ? » (p. 57-60). Les Lettres françaises 109, 24 mai.

  130. Great find! It doesn’t seem to be online, but there’s most of Imke Bröhl’s “Le néo-français vu par la linguistique et dans les romans de Raymond Queneau.”

  131. Stu Clayton says

    According to the bibliography in this, the essay appears in Bâtons, chiffres et lettres:

    # Queneau, Raymond (1965): „Connaissez-vous le chinook“, in: ders.: Bâtons, chiffres et lettres.
    Paris: Gallimard : 57–59. #

    I could flip through (some of?) the pages of the book at Gallimard, but I didn’t find the essay – if essay it be.

    I feel just like Zazie: Elle ne verra pas le métro avant de quitter Paris, mais lorsque sa mère lui demandera ce qu’elle a fait, “j’ai vieilli”, répondra tout simplement la petite Zazie.

  132. David Eddyshaw says

    Dans ses écrits, Raymond Queneau parle d’une vraie catastrophe quant à l’écart entre l’écrit et le parlé. Dans la Conversation avec Georges Ribemont-Dessaignes, Queneau dit que les Français sont « à plaindre » parce qu’ils n’ont pas « le droit d’écrire comme ils parlent ».

    This has been a real issue with Welsh, where the traditional literary language is much more distant from any spoken dialect than in the case of French, and there is no standard spoken language.

  133. Stu Clayton says

    That conversation with Ribemont-Dessaignes is to be found in Bâtons, chiffres et lettres.

  134. As I recall, it’s an issue with Tamil as well.

  135. David Eddyshaw: Hmm, is the gap between Literary Welsh and the spoken varieties of Welsh greater or lesser than the gap between spoken and written French? I remember something Martinet once wrote: that a language could be called difficult if it is challenging for its native speakers to adequately master its standard written form: and on that basis, he wrote that French may be the most difficult language in the world.

    In the Welsh case the chief difference (correct me if I am wrong) lies in verb morphology, where Literary Welsh seems to be a separate language altogether from any form of colloquial Welsh spoken today. At first glance this is far more radical than the difference between spoken and written French in terms of verbal (or indeed nominal) morphology.

    In the French case, however, the gap between the written versus the spoken morphology is huge: even native speakers of the most bookish, standard-like French need to be taught what is practically a foreign grammar in order to learn when to write “aimer” versus “aimez” versus “aimé” versus “aimés” versus “aimée” versus “aimés” versus “aimés” versus “aimées” versus “aimai” (and, for speakers of non-standard varieties which lack the distinction between front mid-high and mid-low vowel phonemes, we can add “aimais”, “aimait” and “aimaient” to the mix). The same applies to number agreement in the noun phrase: the opposition between “Une mère et une fille charmante” (A mother and a charming daughter) versus “Une mère et une fille charmantes” (A mother and daughter, both charming) is something that French-speaking children need to learn consciously, as it corresponds to nothing whatsoever in spoken French, no matter how bookish.

    For speakers of colloquial forms of French, where the gap in vocabulary and syntax between their speech and the standard written language is even deeper, learning the standard is even more of a challenge.

    Furthermore, the spelling of individual words is very opaque to L1 francophones, whereas my understanding is that it is the structure, and not the sound/spelling rules, of Literary Welsh which constitutes the chief obstacle for Welsh L1 speakers wishing to master the standard written language.

    Incidentally, do you or indeed does anybody know of any good comparative study on the gap between the colloquial and written forms of different languages?

  136. PlasticPaddy says

    Etienne, It seems to me his ideas for orthography cast doubt on his linguistic nous (his reformed spelling has arbitrary bad word divisions and does not allow any scope for other francophone accents), unless he is joking:
    « Jérlu toudsuit lé kat lign sidsu, jépapu manpéché de mmaré ».
    Source: https://fr.m.wikipedia.org/wiki/N%C3%A9o-fran%C3%A7ais
    Other authors in the Wikipedia page seem to have better approaches, e.g.
    « Nou som le mèrkredi douz novanbr, il é dizneûveur vin sink, dizneûveur trant, par la, on va atandr juska dizneûveur karant, on atan ke toulmond ariv é sinstal é an atandan…»
    (the only thing here I see is that on/an/in are separate but not an/en).

  137. David Eddyshaw says

    @Etienne:

    I wasn’t taking spelling into account at all. The difficulties posed by written Welsh there are indeed relatively minor: Welsh has easily the most transparent orthography among the indigenous languages of the British Isles (a low bar) and certainly more straightforward than French too.

    Verbal morphology, specifically, is indeed the big challenge for a Welsh speaker learning Literary Welsh. The original present tense has mostly become a future in spoken Welsh; the pluperfect and the subjunctive have disappeared, and the imperfect still exists but has modal rather than tense use. Spoken Welsh uses periphrastic constructions for present and imperfect (and unlike English, it doesn’t make a distinction between dynamic and stative verbs in this.) The actual flexional forms are substantially more complex and irregular in Literary Welsh. Literary Welsh is pro-drop, and the actual forms of many pronouns are different. Spoken Welsh has also simplified the initial-mutation system a fair bit, although it still plays a major role in the language.

    Literary: Caraf di. “I love you.”
    Spoken: Wy’n dy garu di.

    In practice, most written Welsh is nowadays a sort of compromise.

    Hardcore Literary Welsh regarded the language of the 1588 Bible as normative (mostly: the Bible still had the default SVO word order of Middle Welsh, which is now called the “Abnormal Order” and is no more even in the most hifalutin LW.) The language of the 1588 Bible was already archaic, and the resulting situation is much as if the standard for modern written English was Chaucer. In the Welsh original of Simon Evans’ Middle Welsh grammar, he normally cites Middle Welsh examples without any translation into Modern Welsh.

  138. Catharêfwsa, as it were.

  139. J.W. Brewer says

    The “four stages” meme Brett references looks to be an abridgement of the classic “Ten Stages of Drunkenness” first propounded by Dan Jenkins (1928-2019) in his 1981 novel Baja Oklahoma and reprised here: https://www.golfdigest.com/story/the-comedy-issue-the-story-behind-the-10-stages-of-drunkenness

    Much subtlety and nuance is lost in the abridgement.

  140. marie-lucie says

    Etienne: “it is not an American linguist, but the French writer Raymond Queneau.”

    In this case, I did not encounter this in Queneau”s work, but as filtered through an American linguist.

    Queneau’s quasi-phonetic spelling has many imperfections. Among other things, it presumes a pronunciation which ignores a variety of regional or social differences. As well, even speakers with very limited education are constantly exposed to standard writing: the latter shows up every day, whether in posters, newpaper and magazine titles, names of products in a supermarket, directions in a city or on the road, and such short instances rarely considered as texts. Even in such minimal instances of writing, all the traditional “e muets” are usually shown, even if they would rarely be pronounced (and when they are, the “rules” for their pronunciation can differ from one region to another). Leaving out the “e muets” in order to reproduce an author’s pronunciation also distorts the appearance of a word and is more disturbing to the reader than leaving them in. As for putting together several words spoken without a pause, as in Queneau’s “jérlu”, the reader needs to say it aloud in order to recognize “j’ai relu” (I reread).

  141. David Eddyshaw says

    I was just reading Queneau’s Wikipedia page (and realising that his floruit was actually at least a decade earlier than I thought. A sign of age …)

    Interesting that he translated The Palm-Wine Drinkard. I can see why it would appeal to him.
    Also that excellent film Certains l’aiment chaud.

  142. Marie-Lucie, is this your reference? (J. Vendryès, 1921, Le langage: introduction linguistique à l’histoire. pp. 102–103.) Vendryès must have been the source for Queneau:

    Le cas inverse est fourni par certaines langues américaines dans lesquelles les morphèmes et les sémantèmes sont conçus et exprimés séparément. On réunit d’avance au début de la phrase les indications morphologiques, on donne en quelque sorte un résumé algébrique de la pensée; tout y est, moins les représentations des objets, qui ne viennent qu’ensuite. Pour dire : l’homme a tué la femme avec un couteau, la phrase sera du type: lui elle cela avectuer homme femme couteau (langue chinook) (1). Tout ce qui précède le tiret que nous avons introduit dans la phrase ne comprend que les indications grammaticales, les morphèmes; les sémantèmes sont donnés après.

    Ne nous étonnons pas trop d’une structure aussi singulière. Le français parlé connaît des tours qui sont très voisins de celui-là. On entend dire dans le peuple: Elle n’y a encore pasvoyagé, la cousine, en Afrique ou Il l’a-ti jamaisattrapé, le gendarme, son voleur? Tout ce qui précède les tirets ne contient aussi que des morphèmes: indications du sujet, du régime (direct ou indirect), du genre, du nombre, du temps, du caractère négatif ou interrogatif de la phrase: nous avons là, avant de savoir de qui et de quoi il s’agit, tous les éléments grammaticaux de la phrase. Pour compléter cette dernière, il n’y a plus qu’à désigner les personnages et l’action à laquelle ils prennent part, en un mot les faits de la cause; les données abstraites sont mises en tête, et, en queue, les données concrètes.

    (Couldn’t he find a nicer example sentence?)

    I also found a 2000 article by Haase, Das Französische als exotische Sprache, FWIW.

  143. Another apropos reference is Cécile De Cat, French subject clitics are not agreement markers, Lingua 115 (2005), 1195–1219. It quotes Vendryès and a 1976 paper by Givón.

  144. David Marjanović says

    Verbal morphology, specifically, is indeed the big challenge for a Welsh speaker learning Literary Welsh. The original present tense has mostly become a future in spoken Welsh; the pluperfect and the subjunctive have disappeared, and the imperfect still exists but has modal rather than tense use. Spoken Welsh uses periphrastic constructions for present and imperfect

    That’s more than the difference between Standard German and my dialect, but not by a lot. The present is still the present, but it occasionally gets “do”-support for rarer verbs (which are by no means less regular), and the three umlaut classes of OHG/MHG have merged into two in a different way from the standard.* The future is generally indicated by adding “then” to the present; the construction that forms the future in the standard exists, but indicates that something is probably the case right now.** The pluperfect was analytic to begin with, but has been made even more analytic (and rarer). That’s because the synthetic past (sometimes called “imperfect” for no reason other than the existence of that term) is gone except for two verbs, and “have” is not one of them.*** The disappearance of the synthetic past has had the side effect of freeing the synthetic, uh, conditional (“past subjunctive”, Konjunktiv II) of regular verbs from the possibility of confusion, so that category is alive and well, though regularizations abound, as do double markings, and it isn’t rare for the same person to use two or three forms interchangeably for the same verb. The “present subjunctive” (Konjunktiv I), however, has disappeared; it has almost disappeared from the standard within the last 100 years, too, except that it remains in very common use for reported speech e.g. among journalists, who need it to make clear that they neither agree with the person whose speech they report (indicative) nor disagree (conditional).

    * (1) no umlaut in the singular; (2) umlaut in the 2nd & 3rd person, but not in the 1st, and not in the imperative either; (3) umlaut in all three. The standard has merged (3) into (2) except for the imperative, the Bavarian dialects have instead merged (2) into (1). An example of (2) is fahre, fährst, fährt; fahr, an example of (3) is gebe, gibst, gibt; gib < OHG synchronically predictable gibu, gibis(t), gibit; ?gib…i?.
    ** Marginally, this exists in English with going to.
    *** And so, ich hatte getan “I had done” becomes ich habe getan gehabt. Bafflingly, this Ultra-Perfekt (as a baffled prescriptivist has called it) has been spreading lately into the area where the synthetic past is in daily use; it even reached the sea maybe 10 years ago.

  145. David Marjanović says

    (Couldn’t he find a nicer example sentence?)

    “Things really got interesting when I tried to add arguments back in, and to further inflect the gerund. In such times, one turns naturally to death and crocodiles for unambiguous agents and transitivity. The conversation went something like this.”

    Spoiler: there’s an instrument in there, too!

    On entend dire dans le peuple: Elle n’y a encore pas

    Nowadays that would be elle a pas encore – not that it would change the point, though (plus ça change…).

    Il l’a-ti jamais ‖ attrapé

    So that’s the famous question marker ti? I haven’t encountered it. Questions and statements differ only in intonation in the vast majority of cases, quoi.

  146. (And it should be ta cousine, not la cousine. I fixed some OCR errors, but not that one.)

  147. except that it remains in very common use for reported speech e.g. among journalists, who need it to make clear that they neither agree with the person whose speech they report (indicative) nor disagree (conditional).
    That’s what they usually report in descriptions of current Standard German, but in my experience of listening to and reading news reports, even in such cases the indicative is used regularly now, and cases of the “distancing” present conjunctive have become rare. I started noticing that already in the 90s, and I’d be interested in whether you notice that as well. It would be an example of descriptions of standard language that are behind the development curve by a generation, due to descriptions being based on quoting previous descriptions, instead of fresh observation.

  148. David Marjanović says

    Interesting. I haven’t noticed that, but I haven’t read that much news in German in recent years, and I wouldn’t necessarily notice a decrease by half. I’ll try to pay attention from now on.

    Geography may be a factor. In Swiss Standard German, welch- as the relative pronoun is alive and well, unlike Standard German elsewhere where it’s become quite rare, and unlike Swiss or other dialects where there doesn’t seem to be any trace of it.

  149. David Marjanović:

    1-On the subject of the interrogative particle “ti”, which I have never heard used by any European either*, allow me to quote myself (from this thread: http://languagehat.com/quebec-french/)

    “The Quebec interrogative particle “tu” is an innovation, which for a time coexisted with an interrogative particle “ti” which used to be widespread in non-standard European French, as I mentioned above. Now, whereas “tu” is alive and kicking in Quebec French, “ti” is nearly extinct in both Europe and Quebec. As a result, whereas a century ago spoken French on both sides of the Atlantic had an interrogative particle “ti” (with a variant “tu” in Quebec), and thus did not differ much from one another, today the interrogative particle is non-existent in Europe, with Quebec on the other hand having generalized “tu”: as a result there now exists a much more pronounced difference between Quebec and European French than there did a century or so ago.”

    Incidentally, the “ti” particle was borrowed by some forms of Occitan in France, and, more surprisingly, by Plains Cree (i.e. the westernmost variety of the Cree dialect continuum), which got it from Canadian French. Thus, the distribution of reflexes of “ti” today is an excellent example of the survival, at the periphery, of an older feature which went extinct in more central areas.

    *However, I have encountered it in older French literature: In a number of Arsène Lupin novels (Maurice Leblanc, the author, died in 1941), uneducated workers/peasants are depicted as using interrogative “ti” in speech.

    2-About the example sentence-

    “Elle n’y a encore pas voyagé, ta cousine, en Afrique” versus what you call contemporary “Elle a encore pas voyagé, ta cousine, en Afrique”: Actually, in my own speech, the two sentences are not quite semantically equivalent.

    Neither has a negator “ne”, but the first one, with “y”, implies that Africa is the destination of the trip, with your cousin living outside the continent, whereas the second, without “y”, implies that your cousin is already within Africa and that she has yet to travel within the continent.

    Which, now that I think about it, is very similar to the way object clitics are almost obligatory when the object noun is definite (Je le vois, le/ce/mon chien) and almost wholly disallowed when the object noun is indefinite (Je vois un chien, not *Je le vois un chien).

    3-Your remarks on the use of the Konjunktiv I by journalists to disassociate themselves from a statement sounds a lot like a special use of the conditional in French, known as the “conditionnel dubitatif”, which is likewise only really alive in the speech (and occasionally writing) of journalists.

  150. Stu Clayton says

    @Hans: That’s what they usually report in descriptions of current Standard German, but in my experience of listening to and reading news reports, even in such cases the indicative is used regularly now, and cases of the “distancing” present conjunctive have become rare.

    Like David, I do not read that much news in German here. But I do wage a lonely battle preserving that use of the subjunctive. It’s because I distance myself practically every day from crazy-ass anonymous third-party opinions (on technical Java subjects) that have been dragged into emails to which I must reply.

    It’s a tool I would not like to give up.

    Example from this week: a young developer gets it into his head to oppose me on the subject of explicit use of class loaders. He does this by citing half-baked, confused remarks by an unnamed contributor to a github FAQ. I replied by pointing out that citing such remarks are not an acceptable substitute for himself understanding the subject. If he understood the subject he would not need to cite github.

  151. I wish you success in your battle! 🙂

  152. Stu Clayton says

    The distancing present subjunctive allows me to introduce ungestraft a certain snarkiness – it’s all distanced and I’m not quoting !

  153. Can you give a simple example for us curious heathens?

  154. Stu Clayton says

    Well, let me go into the kitchen and see what leftovers I could warm up at short notice. Mostly I do technical haute cuisine not easily digestible by curious heathens. Much like y’all’s thousand-year polysynthetic eggs.

  155. You want I should give you some subjunctive?

  156. Stu Clayton says

    That’s a good example right there !

  157. David Eddyshaw says

    @Y:

    God forbid!

  158. Stu Clayton says

    Clearly one is not dealing with everyday heathens here.

  159. Stu Clayton says

    Y fragt, ob er einem Meister des distanzierenden Konjunktivs I noch etwas Konjunktiv geben solle !

    Yes, my little chickadees, nobody writes quite like that anymore. I usually add a pinch of jokey snark to make it more palatable, and to deflect the charge of being a pedantic old fart.

  160. David Marjanović says

    what you call contemporary “Elle a encore pas voyagé, ta cousine, en Afrique”

    My point, other than the loss of ne, was just that it’s pas encore, not encore pas. That I dropped the y was an error, however.

    “conditionnel dubitatif”

    Ah yes.

  161. Stu Clayton says

    Is that what’s going on in je crains qu’il ne rate notre rendez-vous ? (hope I got that right) Je crains qu’il ne vienne en retard.

    That ne is a mind-blower.

  162. David Eddyshaw says

    Says WP:

    Herbie is an antithetical hero: short, obese, unstylish, and young. Deriving some of his powers from genetics and some from magical lollipops from “the Unknown”, Herbie can talk to animals and sometimes even inanimate objects (who all know him by name), fly at high enough speeds to quickly travel to other galaxies (by walking through the sky), become invisible, cast spells and summon spirits from other dimensions, quickly dispatch his enemies with apparent ease, and travel through time. Herbie is emotionless, terse, irresistible to women, consulted by world leaders, nearly omnipotent, and more powerful than Satan.

  163. Stu Clayton says

    A rôle model for our times ! Too bad I am already settled in my ways.

  164. And — he wields the power of the subjunctive.

  165. David Eddyshaw says

    From which all else surely follows.

  166. Stu Clayton says

    I wonder whether Herbie could explain, in 25 words or less, what “polysynthetic” means.

  167. marie-lucie says

    Y: “Marie-Lucie, is this your reference? (J. Vendryès, 1921, Le langage: introduction linguistique à l’histoire. pp. 102–103.) Vendryès must have been the source for Queneau:”

    I only know the name of Vendryès, I have never read anything by him.

    “Le cas inverse est fourni par certaines langues américaines dans lesquelles les morphèmes et les sémantèmes sont conçus et exprimés séparément. On réunit d’avance au début de la phrase les indications morphologiques, on donne en quelque sorte un résumé algébrique de la pensée; tout y est, moins les représentations des objets, qui ne viennent qu’ensuite. Pour dire : l’homme a tué la femme avec un couteau, la phrase sera du type: lui elle cela avec ‖ tuer homme femme couteau (langue chinook) (1). Tout ce qui précède le tiret que nous avons introduit dans la phrase ne comprend que les indications grammaticales, les morphèmes; les sémantèmes sont donnés après.”

    Vendryès must have consulted Boas’ grammar of Chinook (1911, bound together with those of several other languages). I own a copy and was looking for it in order to find the above example, and perhaps a less gruesome one too, but I couldn’t find my book (I know I have it, as I consulted it perhaps 2 weeks ago). This separation of grammatical morphemes first, lexical item last, also occurs in Kalapuyan languages (once spoken along the Willamette river, which runs perpendicular to the Columbia, home of the Chinooks) and perhaps others in the same geographical area.

    “Ne nous étonnons pas trop d’une structure aussi singulière. Le français parlé connaît des tours qui sont très voisins de celui-là. On entend dire dans le peuple: Elle n’y a encore pas ‖ voyagé, la cousine, en Afrique …

    “Elle n’y a encore pas ‖ voyagé, la cousine, en Afrique”

    A person who would say this would not have “n'” (ne) here! I have two other remarks:
    – “la cousine” vs “ta cousine”: Both phrases are about a woman who is known as someone’s cousin; “ta cousine” ‘your cousin’ is a relative of the listener; but “la cousine” can also imply that the woman in question is ‘someone else’s’ cousin, that the two conversationalists both are used to hearing about.
    – “Elle y a encore pas voyagé en Afrique” vs “Elle (n’) y a pas encore …”
    The first sentence suggests to me that the cousin has been planning to go to Africa and that she has let all her friends and relatives know of these plans, but that she keeps “not going” there, becoming an object of derision. The second sentence only means that she has not gone there yet .

    – *Couldn’t he find a nicer example sentence?” That’s what I was hoping to find in the grammar, or in a book of Chinook legends that I also have, but such legends are often quite realistic. ALso, the woman in question could be a legendary animal rather than a truly human character.

    David M; “l l’a-ti jamais ‖ attrapé, le gendarme, son voleur?”

    If the French speakers you know are mostly academics, indeed you will not hear this type of sentence from them. The famous “ti” of uneducated speech is from “…-t-il” with the intrusive -t- which copies the -t of “irregular” verbs, attaching it to verb forms ending in a vowel rather than in the old 3sg -t ending. So for instance “a-t-il” on the model of “est-il” or “vient-il”. The final -l of “il” is not usually pronounced in colloquial speech (as in the first word of the example sentence), hence just “ti” which became a question marker even in sentences including another question marker. Trying to think of another example I might have heard in my own family – years ago! – my father (from the Parisian lower-middle-class ) would often ask “Y a-t-i (du café, du pain, etc)” but he never used the uneducated “ti”.

    – “l l’a-ti jamais ‖ attrapé, le gendarme, son voleur?”

    Here the more formal (though still colloquial) equivalent would be “L’a-t-il jamais attrapé, le gendarme, son voleur” ‘Has the policeman ever caught his thief?’ (where the initial “L'” is the 3sg Object pronoun) but the written equivalent would be “Le gendarme a-t-il jamais attrapé son voleur?’

  168. David Marjanović says

    If the French speakers you know are mostly academics, indeed you will not hear this type of sentence from them.

    On the one hand, yes, they’re almost all academics or at least education-adjacent. But on the other, 30 young people from all over the country spending weeks on a dig site in 2004… that’s not the kind of environment where people strive for literary expression. That’s where I learned vachement, quoi “like, totally”, y a d’la flotte “there’s water in there”… in short, they sounded more like Titeuf.

    (I also encountered residual regionalisms there, like gobelet vs. verre for a transparent plastic cup, or je sais point “I have no idea at all”.)

  169. marie-lucie says

    These words and phrases are quite familiar to me and have been all my life. I would not consider “un gobelet” to be be a regionalism, and I would not call a plastic cup “un verre”. If I say “un verre d’eau” it refers to the contents, not the container.

    I had vaguely heard of “Titeuf” but consulted Wikipedia for more information. Among extra information were titles in which “pô” meant “pas” (the negative particle). My father’s mother, who spent her entire life in Paris (in the XVe arrondissement), often pronounced “pas” that way (almost), especially as short for “n’est-ce pas”.

  170. does “you want i should” appear in non-yiddishized english? (i’ve always understood it as a calque.)

    i don’t think i’ve ever encountered it otherwise, but i’m hardly a reliable test case – i live in new york city, where most english is at least a little yiddishized.

  171. I always think of it as nothing but, especially in mid–20th century comic books, especially ones from the New York area, and especially ones written by Jewish people such as Richard E. Hughes (né Leo Rosenbaum).

  172. does “you want i should” appear in non-yiddishized english? (i’ve always understood it as a calque.)

    I don’t think so, and it’s definitely a calque.

  173. Je crains qu’il ne vienne en retard.

    “I fear lest he come late.”

  174. J.W. Brewer says

    Some have noted a not-obviously-Yiddish-influenced instance of the want/should construction (or at least a close relation) in a famous bit of correspondence, viz. the letter sent in 1860 by Grace Bedell to Abraham Lincoln in which she suggested he grow a beard: “I am a little girl only 11 years old, but want you should be President of the United States very much so I hope you wont think me very bold to write to such a great man as you are.”

  175. J.W. Brewer says

    Another 19th century instance from Howells’ “The Rise of Silas Lapham” (published 1885):

    Lapham spoke to the woman again. “Do you want I should call a hack, or do you want I should call an officer?”

    Does the presence of the auxiliary “do” distinguish this from the yiddishism?

  176. David Eddyshaw says

    It makes it a Balkan Sprachbundism …

    “You want me to call a hack?” would be perfectly fine informal (if unYiddish) English for the first clause, so the ismicity does not reside in the absence of “do.”

  177. Lapham spoke to the woman again. “Do you want I should call a hack, or do you want I should call an officer?”

    I wonder if it was a Germanism back then? There were a lot of German-Americans in the 19th century.

  178. J.W. Brewer says

    No doubt there were Germanisms floating around some varieties of 19th century AmEng, but the Lapham character is supposed to be a rustic Yankee from rural Vermont who ends up a wealthy Bostonian without the social polish and posh background of the Brahmins he interacts with, so a particularly unlikely fellow to have Germanisms in his idiolect.

  179. Stu Clayton says

    @Rodger C: “I fear lest he come late.”

    Thanks, that somewhat lessens the mind-blowing effect on me of that-there ne. “Lest” has exactly the same function. I have no idea where it comes from.

    I would otherwise have rendered it into English as “I am afraid he might/will come late”.

  180. David Eddyshaw says

    I wonder whether Herbie could explain, in 25 words or less, what “polysynthetic” means.

    Polysynthesis-QUOT mean-NMLZ-explain-APPL-DES-SBJV-1sg>2sg?

    https://en.wikipedia.org/wiki/List_of_glossing_abbreviations

  181. Trond Engen says

    I was just going to say that if he understands polysynthesis he can explain it in one.

  182. A man of few words and many morphs.

  183. What the Vendryès examples show is not just a transition from analytic to synthetic, but at the same time also a transition from SVO to VSO word order. Are there other examples of this kind of process?

  184. Trond Engen says

    Let me have a go:

    Polysynthesis-QUOT = “”polysynthesis””, i.e. “the word ‘polysynthesis'”

    mean-NMLZ = “meaning”

    explain-DES = “will explain”
    explain-DES-SBJV “would (like to) explain”
    explain-DES-SBJV-1sg>2sg “would like me to explain to you”
    explain-APPL-DES-SBJV-1sg>2sg “you would like me to explain to you”

    “Would you like me to explain the meaning of the word polysynthesis to you?”

  185. Giacomo Ponzetto says

    The Vendryès examples puzzle me further.

    I feel I could say those sentences perfectly naturally in Italian: “Non c’è ancora andata, tua cugina, in Africa” or “L’ha mai acciuffato, il carabiniere, quel ladro?” I don’t even perceive either sentence as particularly colloquial.

    In fact Italian can even reproduce exactly the gruesome Chinook example: “Ce l’ha ammazzata, la vittima, col coltello, il criminale” (i.e., with it her he did || kill the victim with the knife the criminal), although postponing all three parts of the sentence definitely doesn’t come natural to me.

    Should I realize this is all polysinthetic and exotic and innovative?

    In my linguistic ignorance, I’d have rather suspected it was a pan-Romance heritage of the word-order freedom Latin enjoys thanks to its cases.

  186. Vendryès was talking about word order, not polysynthesis really. In your Italian examples, Non c’è ancora andata and Ce l’ha ammazzata are respectively four and three phonolological words, aren’t they?

  187. Trond Engen says

    It’s both. The point of this exercise is exactly to show under which conditions polysynthetics arise, It’s when a syntactic pattern like this becomes obligatory,

  188. David Eddyshaw says

    @Trond:

    A good (if somewhat stilted) rendering; more idiomatically it would go “You want I should explain you the meaning of this here ‘polysynthesis’?”

  189. William Burroughs, Exterminator!:

    During the war I worked for A. J. Cohen Exterminators ground floor office dead end street by the river. An old Jew with cold gray fish eyes and a cigar was the oldest of four brothers. Marv was the youngest wore wind breakers had three kids. There was a smooth well dressed college trained brother. The fourth brother burly and muscular looked like an old time hoofer could bellow a leather lunged “Mammy” and you hope he won’t do it. Every night at closing time these two brothers would get in a heated argument from nowhere I could see the older brother would take the cigar out of his mouth and move across the floor with short sliding steps advancing on the vaudeville brother.

    “You vant I should spit right in your face!? You vant!? You vant? You vant!?”

  190. marie-lucie says

    Y: ‘What the Vendryès examples show is not just a transition from analytic to synthetic, but at the same time also a transition from SVO to VSO word order”

    V’s examples are VSO, but they don’t have to be, as in “Ta cousine, son voyage en Afrique, elle l’a déjà fait?” ou bien “Ce voyage en Afrique, elle l’a déjà fait, ta cousine?’ The point is that the grammatical roles are obligatorily occupied by clitic pronouns in fixed positions next to the verb, while the lexical referents of the pronouns can be almost freely distributed in other places in the sentence. Whether the basic order is SVO, VSO or something else does not seem to be affected since the S and O are fixed-place clitics.

    English is much more restricted than French, Italian or Spanish (at least, but these are the ones I know best) in its SVO order, whatever the nature of S and O (pronouns or larger “adjuncts”). The Romance freedoms must indeed be due to survivals from Latin. So must be the use of ne in je crains qu’il ne vienne and similar constructions.

  191. Je crains qu’il ne vienne en retard

    It’s even worse in Latin:

    Timeo ne veniat. “I fear that he may come.”
    Timeo ut veniat. “I fear that he may not come.” (with nary a negator in sight)

  192. Giacomo Ponzetto says

    @Y: the counting of phonological words is beyond my (non-existent) linguistics.

    Those clitics are called “atonic pronouns” for a reason. I suppose it’s fair to say that “Non c’è andata” and “L’ha ammazzata” have one stress each and don’t read any different than *nonciandata and *lammazzata.

    Perhaps the thesis of polysinthetic spoken Romance languages is that those are single phonological words?

    Not sure how that fares with insertions like ancora, though. Does it go to the bitter end of reading *nonciancorandata as a single word with secondary stress on co and primary on da? That seems unhelpful to me, but what do I know.

  193. marie-lucie says

    GP, These pronouns are “atonic” because they are not normally stressed in the verb phrases that they belong to. This occurs whether the vowel of such a pronoun is sounded or not, especially in fast speech (very slow speech may have more vowels, but it is an artificial, deliberate form of speech).

    As for words like ancora (Fr encore), they (and a few others) are allowed to be part of the verb complex or phrase.

    The long “words” that you quote are not single words, but phrases or even sentences. Phonologically they are “utterances” rather than words, but from the point of view of morphology and syntax, whether they have one stress or more is irrelevant, and similarly whether they are spoken fast or slowly.

  194. There is no clear-cut definition of a word, either phonological or syntactic, and in fact reexamining the concept of a word a hot topic right now.
    For defining polysynthesis the syntactic word is more often used: do the various morphemes have to be in the same order? Can they be separated by a relatively ill-defined combination of other morphemes? For example, can ancora be replaced or augmented by more adverbs? For phonological wordhood, a single primary stress is one diagnostic criterion. Another is, can you separate the “word” in question into smaller ones by inserting pauses, in natural but careful speech? Is it more natural to say “Non—c’è—ancora—andata” than “Elle—n’y—a—encore—pas—voyagé”?

  195. Giacomo Ponzetto says

    Thanks, Marie-Lucie! That makes sense to me.

    So the polysynthetic point is that to rearrange the sentence arbitrarily we can no longer shuffle the verb alone like Latin, but rather need a verb complex that includes some pronouns?

    In Italian probably one atonic pronoun alone, with extra ones becoming optional and colloquial. Which may also be why French is considered more polysynthetic?

  196. Giacomo Ponzetto says

    Thanks, Y!

    Tentatively I’d venture the only non-separable unit there is non c’è. At least I lack the fantasy to think of ways of breaking it.

    On the other hand, you can certainly replace, qualify, or move around the adverb: non c’è proprio andata, non ci è forse ancora andata, ancora non ci è andata. But probably not quite with an ill-defined combination of words, only with suitable adverbs.

    I suspect — but I have to wonder if I’m stretching my mother tongue improperly — you could even move the participle around for emphasis: “Andata non c’è ancora, in Africa, tua cugina, ma …” Although I’d be more comfortable with “Andarci, non c’è ancora andata …”

    Still, it seems necessary to keep that verb phrase together as a unit, even as I try to manipulate its sub-components. Not quite sure where that leaves me on the polysynthesis scale.

  197. Stu Clayton says

    Progress towards an elucidation of “polysynthesis” has been rather slow so far. But then I think of the Celerity Deathmatch between the Tortoise and the Hare, and we all know who won that one.

  198. Timeo ne veniat. / Timeo ut veniat.
    I’m stumped. How do they work? Why do they mean the opposite of what they mean?

  199. very similar to the way object clitics are almost obligatory when the object noun is definite (Je le vois, le/ce/mon chien) and almost wholly disallowed when the object noun is indefinite (Je vois un chien, not *Je le vois un chien).

    Looks similar to clitic doubling and definite conjugation in Hungarian.

  200. Why do they mean the opposite of what they mean?

    They presumably didn’t to a native Latin speaker. Rather the translation of “Timeo” as “I fear” is not quite accurate. “Timere” must have had the sense of “I want the opposite result but don’t have confidence in it”. So “timeo ne veniat” in a Latin mind is “I want/hope he does not come but don’t believe that is the likely outcome”.

    Originally it may have developed from a paratactic construction. So “timeo” = “I fear it is not the case but” : “Ne veniat” = “may he not come.”

  201. “Timere” must have had the sense of “I want the opposite result but don’t have confidence in it”. So “timeo ne veniat” in a Latin mind is “I want/hope he does not come but don’t believe that is the likely outcome”.

    I got lost in the negations. If “timeo X” means “I want not X but think that probably X” (which is still a good match to “I fear X” in my opinion), then in “timeo ne veniat” X would be “he does not come” and together it would mean “I want that he comes but think that probably he doesn’t”. Which is not the purported seemingly paradoxical meaning “I fear that he may come”. Am I missing something?

    Also timeo can take nominal objects in pretty much the same way as “fear”. Timeo Danaos.

  202. I think you’re missing that “veniat” is in the optative subjunctive. “Ne veniat” means “may he not come!”, not “he doesn’t come/isn’t coming.” Hence the reasoning that the Latin constructions developed from some earlier parallel construction. “Ne veniat” – I hope he doesn’t come! + “timeo” – but I fear he will anyway.

  203. Giacomo Ponzetto says

    I’d venture that Latin verbs of fearing (not limited to timeo) construct like English wish. Not unlike timeo ne veniat, “I wish he wouldn’t come” suggests to me a presumption he will in fact come. But perhaps it doesn’t to native speakers.

  204. PlasticPaddy says

    “I wish he wouldn’t come” is an unfortunate sentence, as it needs an ending to be repeatable in polite society ????, maybe “…to the house (so often).” The (so often) is because for me “wish” + “wouldn’t” means a repeated action. For a one-off, i would say “I hope he won’t / doesn’t come” (even here “…to the house/party/etc” is a bit better for me)

  205. It all gets more fun when you negate timere and have to replace ut with ne non. Non timeo, ne non veniat, “I don’t fear that he will not come”.

  206. In Mongolian, ‘doubt’ and ‘suspect’ are the same word, if that puts your mind at ease.

  207. *Be anxious* seems like a good parallel, if you then render the subjunctive with *should*: I’m anxious that he should come = I fear he might not come.

  208. David Marjanović says

    I’m stumped. How do they work? Why do they mean the opposite of what they mean?

    The way I manage to remember it is that wishing works: je crains qu’il ne vienne / timeo ne veniat is “I fear in order that he won’t come”, “I fear lest he come”. That last one has the proper present subjunctive.

    Contrast German, where I don’t think that the indicative in ich fürchte, (dass) er kommt can be replaced by the present subjunctive (komme) even if you climb some literary Parnassus.

    Would is past subjunctive.

    Perhaps the thesis of polysinthetic spoken Romance languages is that those are single phonological words?

    The more innovative accents of French (the ones that have made [ɛ ɔ] vs. [e o] completely predictable from closed vs. open syllables) probably don’t even have phonological words. Stress in French is “utterance-final”, basically prepausal. The claim that certain “pronouns” and “adverbs” aren’t clitics but prefixes (of therefore polysynthetic verbs) in sufficiently colloquial French comes from the facts that they’re obligatorily present in the sentence even when the corresponding unreduced words are also present (moi, je pense…), they’re obligatorily stuck to (mostly finite) verbs, they have a fixed order, and nothing else can be inserted between them.

    Concerning the fixed order, check out this conference handout or whatever it is: on p. 3 there’s a verb template with nine prefix positions – all nine will hardly ever be filled at the same time, but the order of the ones that are present is determined by the template, and no slot can be doubly filled.

    Contrast southern German dialects, where reduced forms of pronouns can form pretty long chains suffixed to verbs if you play it right, some even take their forms from verb endings rather than the corresponding unreduced pronouns, and some engage in clitic doubling or perhaps tripling – but their order isn’t completely fixed, and they attach to the first word in the clause, no matter if that’s a verb, a conjunction or in some dialects even a noun. Verdict: not polysynthetic. (But noun-incorporating to an ever greater degree.)

  209. Catalan weak pronouns are pretty impressive too.

  210. One of the references in that WP article is El catalán sin esfuerzo.

  211. Giacomo Ponzetto says

    The claim that certain “pronouns” and “adverbs” aren’t clitics but prefixes (of therefore polysynthetic verbs) in sufficiently colloquial French comes from the facts that they’re obligatorily present in the sentence even when the corresponding unreduced words are also present (moi, je pense …), they’re obligatorily stuck to (mostly finite) verbs, they have a fixed order, and nothing else can be inserted between them. Concerning the fixed order, check out this conference handout or whatever it is …

    On this basis, I find the claim more puzzling than enlightening.

    All Romance languages I know have clitic pronouns that get obligatorily stuck to verbs in a fixed order. If anything, their order may be the more strictly fixed the less colloquial the language.

    No Romance language I know makes all pronouns obligatory. As I was taught them, French requires the subject and Spanish the indirect object. Catalan, Italian and Portuguese are so enthusiastically pro-drop they are more likely to forbid than mandate pronouns. I understand that optional and even forbidden pronouns become more prevalent the more colloquial the language. But is there truly any variety of spoken French that’s sufficiently colloquial to make j’ai donné le livre à ton frère jarringly unidiomatic?

    In Italian, I don’t need many pronouns even to rearrange word order. I feel perfectly natural saying:
    Ho dato il libro a tuo fratello
    but also:
    A tuo fratello, ho dato il libro
    I only need a duplicate pronoun to emphasize the direct object:
    Il libro, l’ho dato a tuo fratello
    L’ho dato a tuo fratello, il libro
    Il libro, a tuo fratello, l’ho dato
    A tuo fratello, il libro, l’ho dato

    I suddenly realize I perceive the very same sentences as allowable in English too. I gave the book to your brother. To your brother, I gave the book. The book, I gave it to your brother. I gave it to your brother, the book. The book, to your brother, I gave it. To your brother, the book, I gave it. The last two sound less idiomatic than in Italian, to the limited extent I can tell; but they aren’t outright ungrammatical, are they?

    Anyway, in Italian I can certainly add the always optional subject and the often disallowed duplicate indirect object. But I’m not saying the same thing if I say Il libro, l’ho dato a tuo fratello or Il libro, gliel’ho dato, a tuo fratello. Doesn’t French allow the same distinction? It’s especially useful in colloquial spoken language, because I can hardly imagine an instance of the latter utterance in which I’m not complaining about your brother …

  212. PlasticPaddy says

    @gp
    The first two of your orderings are acceptable to me. For the others:
    *the book, to your brother, i gave it.
    For me if the book comes first. I has to come second, i.e., The book(?)–I gave it to your brother
    *to your brother, the book I gave it
    For me if the brother comes first, I has to come second AND “it” must be deleted, i.e., To your brother I gave the book. This I would not say, only write.

  213. David Eddyshaw says

    Old French does the thing with non-negative negatives in clauses introduced by que, even when they are not complements of verbs of fearing: thus, not only

    Crient que la vieille n’oublit.
    “He fears that the old woman may forget.”

    but even

    Plus fresche que n’est rose.
    “Fresher than a rose is.”

    Old Occitan does it after “verbs of inherently negative meaning” (like “fear”), and also after negated verbs (Paden p259):

    Eu no-m puesc sofrir
    Qu’eu no-us fasa lausar a tota gen.

    “I cannot keep from making everyone praise you.”

  214. Giacomo Ponzetto says

    but even Plus fresche que n’est rose.

    Funny that this should seem extreme, because it’s the only one Italian does as a matter of course. Normally Più fresca di una rosa, but with a verb Più fresca di quanto non sia una rosa. The things you don’t notice in your own language …

  215. I feel like I’ve seen something like the Plus fresche que n’est rose construction in modern French too, but don’t have an example.

    Hebrew has an odd “non-negative” use of the negator in general indefinite statements of the ‑ever type (a.k.a. “unconditionals”):

    ma she-lo yihiye what that-not be.fut.3.m.sg “whatever happens”
    le’an she-lo telekh where that-not go.fut.2.m.sg “wherever you go”

    I don’t know if this usage is calqued from some European language.

  216. Giacomo Ponzetto says

    @PlasticPaddy

    Thanks! I wouldn’t actually try to say those two in English. I guess I’ll just have to downgrade them from unidiomatic to ungrammatical then.

    However, the easy-rolling single-inversion statements don’t convey correctly the admittedly rare nuance of meaning those alternatives provide.

    If I say, in Italian but I’d suppose also in English: “the book, I gave it to your brother;” then I’m just explaining what happened to the book. If I say, and I’m liable to say it as well as writing it: “to your brother, I gave the book;” then I must be explaining what items I gave to a series of people.

    But suppose we’re exchanging Christmas gifts at a large family gathering. We both know I’ve gotten your brother a book. What I’m trying to tell you is that, while you may not have noticed in the festive confusion, I’ve already given it to him. Neither statement above is fit for this purpose. I’m going to tell you instead: “Il libro, a tuo fratello, l’ho dato” if we’re counting the presents I’ve already delivered; or “A tuo fratello, il libro, l’ho dato” if we’re rather counting the people I’ve already given a present to.

    My best attempt at these in English might be: “Your brother’s book I’ve given;” and more confidently: “Your brother I’ve given the book to.” What else would be more idiomatic? Surely English cannot lack a way of saying this exactly.

  217. In most of these cases, maybe you could argue that what looks like a negator is better called an irrealis marker.

  218. I don’t know if this usage is calqued from some European language.

    I looked this up in the Historical Grammar of Modern Hebrew… Stam. There is no such book.

  219. David Eddyshaw says

    what looks like a negator is better called an irrealis marker

    Yes, I think that’s the key. A sort of semantic shift from “not (at all)” to “not (just at present/in the circumstances).”

  220. Those typologists. They have thought of everything:
    Lakey, H. 2016. The grammar of fear: morphosyntactic metaphor in fear constructions, U. Oregon dissertation, here.
    Lakey, H. 2015. The Grammaticalization of Latin + Subjunctive Constructions. J. Latin Linguistics 14(1):65, here.

  221. Irrealis does seem like a good characterization of the Hebrew usage. I don’t see that it applies to Plus fresche que n’est rose, though; in that case maybe the negator is due to some kind of contamination from a paraphrase like Even a rose is not so fresh.

    @Giacomo: what you’re talking about is topic-comment structure (a.k.a. topic-focus structure, a.k.a. lots of other things). In all of your Italian examples the topic constituents come first. English is much less nimble than Romance languages when it comes to marking multiple or complex topics by means of word order; I think your last two examples would most normally be expressed simply by I’ve given your brother his book, with the main sentence stress falling on given to show it is the focus.

  222. John Cowan says

    Originally it may have developed from a paratactic construction. So “timeo” = “I fear it is not the case but” : “Ne veniat” = “may he not come.”

    That is indeed the usual view of Latinists, from what I can make out, but in simpler and even more paratactic form: ‘I am afraid! May he not come!’ > ‘I am afraid he will come’. This also explains the use of ut in the negative: ‘I am afraid! May he come!’ > ‘I am afraid he will not come’. This is not specific to timeo but applies to all verbs of fearing.

  223. It is the usual view, but it’s always struck me as a bit just-so. Did anyone really ever speak that kind of caveman Latin?

    The English subordinator that is generally explained with the same kind of paratactic origin, from the demonstrative (I fear that: he may come!); this somehow feels more intuitively plausible to me, but I vaguely recall it’s been challenged recently on the basis of corpus studies. Maybe someone here will know the article.

  224. @TR: There is at least a consensus that hypotaxis in the IE languages is a development that happened in the individual families; the only conjunctions that can be reconstructed safely for PIE are paratactic clitics (*-kwe “and”, *-we “or”). It seems that for subordination and for establishing the logical relationship between actions, PIE rather relied on participal constructions and on particles (like Greek men, de, ge etc.)

  225. J.W. Brewer says

    Can I split the difference by suggesting that “the book to your brother I gave it” is grammatical in the “Poetic License” variety of English (i.e. it wouldn’t raise my eyebrow very far if it occurred in a poem or song lyric) but not grammatical otherwise. The “Poetic License” variety has a different grammar that relaxes some but not all constraints of the regular grammar, either because deviating from those constraints makes it easier to find rhymes and fit the meter or because those deviations create a certain surprising effect that is thought aesthetically pleasing while not rendering the text incoherent or unparsable. (Or both.)

  226. Sorry, “the book to your brother I gave it” doesn’t work for me even in the “Poetic License” variety; it would have to be punctuated “the book? to your brother I gave it” to be even marginally acceptable. As a standalone clause it’s just not English.

  227. J.W. Brewer says

    Well, my “Poetic License” assumption is that it would be a standalone *line* in a poem or song but probably not the first one, so the previous line(s) would have sort of set up the situation. On further reflection (hat may or may not agree) “the book to your brother I gave” might be smoother in PLEng, but the gratuitous final “it” seems parallel (modulo word order) to other PLEng constructions such as where you might start a cowboy-type song with “The day it was hot but the gun it was cold” (is that some sort of cleft?) even though you have a perfectly grammatical sentence without the instances of “it.”

  228. Nice try, but no sale!

  229. For “the book to your brother I gave it” to work for poetry, you have to imagine the right meter. It’s an amphibrach. Think:

    There once was a man from Nantucket

    similarly:

    The book to your brother I gave it

    Now, whether it would work outside a limerick, perhaps not.

  230. you might start a cowboy-type song with “The day it was hot but the gun it was cold”

    A couple of excellent discussions at Language Log:
    Left dislocation (2008). Mark Liberman: “… its use in formal written English has been decreasing since about 1500, and is now either informal or archaic.” First comment: “So left dislocation is not as common as it used to be. Can we say that the times they are a-changing?”
    Presidential left dislocation (2012). Mark Liberman: “I’m confident that American presidents at least since Truman have used this construction, and probably all of them going back to Washington would exhibit it, if we had access to samples of their more informal speech.” Languagehat comment: “Me, I’d never use it.”

    So, folk songs, yup. From an unforgettable review of Sandy Denny’s first solo album (written before her death): “After a few tracks of such relentless syntactical fidgets, the listener’s patience, it is exhausted.” (The reviewer is passionately frustrated with her “merely becoming a British rock queen, instead of nurturing a world-class songwriting talent into the revolutionary force it once bade fair to be.”)

  231. J.W. Brewer says

    I guess we can’t know the counterfactual, but I am satisfied to think that Miss Denny’s tragically early death was more to be mourned for the loss of her singing voice (and her reasonably good judgment in what material to sing with it) than her potential-but-unrealized talents as a lyric-writer. But that left-dislocation folksy “fidget” is certainly used by her quondam bandmate Richard Thompson who is very much critically-lauded as a lyric-writer, e.g. the opening line “My father he rides with your sheriffs.”

    I can see why Clive James would focus on her lyrics to “Fotheringay” because they seem all artsy/literary but I quite like, among her original compositions, the rather minor “Crazy Lady Blues,” which is effective imho because it does not seem ambitious yet works very well.

  232. The book to your brother I gave it

    Works without “it,” but not with it.

  233. @Giacomo Ponzetto:

    Regarding your statement that-

    “All Romance languages I know have clitic pronouns that get obligatorily stuck to verbs in a fixed order. If anything, their order may be the more strictly fixed the less colloquial the language”

    -it is important to note that French is unusual within Romance not only because of its obligatory subject clitic pronouns, but also because its object pronouns are (almost!) always in a fixed PRE-VERBAL order, even when the verb is non-finite. Furthermore, a finite verb cannot receive the object clitic of a non-finite verb: “Je lui parle” versus “Je veux lui parler”, where the second sentence cannot become *”Je lui veux parler” (The last is just barely possible in some very pedandic registers of written French, but is so marginal as to practically be extinct).
    The contrast with, say, Italian (“Gli parlo” versus “Voglio parlargli” or “Gli volgio parlare”, with *”Voglio gli parlare” being excluded) is stark. And means that, to my mind, French clitic pronouns are much more easily analyzable as a verbal prefix system than other Romance object clitic systems (or than the clitic systems of Southern German dialects, as David Marjanović pointed out in his March 23: 10:05 pm comment above).

    In answer to your other question-

    “But is there truly any variety of spoken French that’s sufficiently colloquial to make ‘j’ai donné le livre à ton frère’ jarringly unidiomatic?”

    -my own impression is that in colloquial Montreal French the above sentence, without the pre-verbal “lui” (realized as /i/), is indeed unidiomatic/unnatural. Its standard word order contributes something to that impression, I must say.

    Finally, a relevant anecdote: when I first started teaching (Standard, obviously) French in the Canadian West I had a talk with a student who had just returned from an exchange program: she had spent a few months in Quebec city. She told me something that left me puzzled at the time: she said the she could make herself understood and understand others when they were speaking to her, but when listening in to conversations between L1 French speakers she had the impression that she suffered from some exotic kind of aphasia: she recognized most of the words, but the sentences made no sense to her. In retrospect, I think her problem was simple: the free word order of colloquial French, where you need to pay close attention to the pre-verbal morphemes (whether we consider them clitics or prefixes), was too different from the standard/bookish French she had been taught for her to make the leap (doubly so since object clitic pronouns are a feature anglophone learners of French find especially difficult to acquire fully).

    If I ever get the chance to create an introductory colloquial French course, in fact, I think I will begin with the clitic system: in a very real sense it is the core of the language.

  234. ktschwarz says

    I’d guess that if Clive James were asked “Why is it OK when Richard Thompson does it?”, he’d retort “I said relentless.” Thompson’s song has four left dislocations spread over four verses, interspersed with other constructions and the chorus, whereas Denny in the bits he quotes is piling up too many too close together, as well as—this is the bigger problem—using imagery that’s so trite that there’s nothing to notice except the syntax.

  235. J.W. Brewer says

    Well, I’m certainly not going to dispute that RT was on average a better lyric writer than SD, and maybe being judicious and measured in the use of certain Ye-Olde-Folksong syntactic tics like left dislocation is part of his talent. The separate notion that SD could have been better at writing than she was if she’d only tried harder or been better-advised seems non-falsifiable. But given that Clive James in the same piece managed to write the extraordinarily bonkers noun phrase “the college-educated rock-writing tradition which in America had airily come up with John Sebastian and Randy Newman” his opinions do not appear to be worth giving any particular weight at all, because he doesn’t appear to know what he’s talking about. I appreciate that in the particular milieu in which he became a well-paid celebrity critic, celebrity critics were not actually expected to consistently know what they were talking about, only to appear clever and witty to a mass audience too ignorant to know when they were being bullshitted. But still.

  236. PlasticPaddy says

    https://m.youtube.com/watch?v=HApy-Xoix-g
    Maybe it should have been “a rare thing was she”. Naaah.

  237. Giacomo Ponzetto says

    @Etienne:

    Thanks!

    I know Romance clitics move around in funny ways. I believe Italian, Catalan and Spanish are on the same page: li parlo and li vull parlar or vull parlar-li; le hablo and le quiero hablar or quiero hablarle. My Portuguese is rudimentary, but I reckon the European norm is very enclitic: falo-lhe and quero-lhe falar or quero falar-lhe; not to mention the mind-boggling mesoclitic future falar-lhe-ei.

    Do I finally understand correctly that a reason, or perhaps the reason, to consider French alone polysynthetic is that its clitics don’t move around anymore like they used to?

    On mandatory pronoun inclusion, I find it interesting that spoken Québecois has switched to mandating the redundant indirect object, like Spanish. I wonder if colloquial European French does that too: I’d most likely have failed to notice anyway. Does this count as polysynthetic even though the direct object is still missing?

    Or perhaps, is the root cause of my puzzlement that polysynthesis isn’t a binary feature, but rather a property best measured on a scale, with English having no polysynthetic features, all Romance languages a few, French more than others, and Chinook a whole lot?

  238. Stu Clayton says

    a property best measured on a scale

    English has been weighed, and found wanting.

    Perhaps “oligosynthesis” would be a less tendentious term. Of course this proposal may itself be tendentious. You just can’t win.

  239. John Cowan says

    Oligosynthetic languages are ones that have, let us say, 100 morphemes or less, each very short, and form their words by synthesis of them. Whorf thought that Hopi fell into this class, an idea now quite generally rejected. The conlang /auːiː/ (conventionally aUI) is definitely oligosynthetic; the complete list of its morphemes is:

    /a/ ‘space’, /e/ ‘movement’, /i/ ‘light’, /u/ ‘person’, /o/, ‘life’, /y/ ‘negation’ /œ/ ‘conditional’;

    /aː/ ‘time’, /eː/ ‘matter’, /iː/ ‘sound’, /uː/ ‘mind, spirit’, /oː/ ‘feeling’ (it is unclear whether [ː] is truly length or rather tenseness;

    /b/ ‘together’, /d/ ‘through, by means of’, /f/ ‘this’, /g/ ‘inside, inner’, /h/ ‘what, question marker/, /k/ ‘above, top, upper’, /l/ ’round’, /m/ ‘quality; (when final) adjective marker; /n/ ‘quantity; plural marker’, /p/ ‘before, (in) front’, /r/ ‘positive, good’, /s/ ‘thing’ /ʃ/ ‘be, exist’, /t/ ‘toward, direction, tendency’, /w/ ‘power, strength, work; (when final [v]) verb marker’, /z/ ‘part, divide’, /ʒ/ ‘equal, even, horizontal’.

    The simple numbers are nasal vowels: /ỹ ~ ã ~ ẽ ~ ĩ ~ ũ ~ õ ~ ãː ~ ẽː ~ ĩː ~ ũː ~ õː/ ‘0-10’.

    Thus the name of the language is /a/ ‘space’ + /uː/ ‘mind, spirit’ + /iː/ ‘sound’, or ‘language of space’, it being intended for communication with ETs, particularly those from Zeta Reticuli. ‘God’ is /kuː/ whereas ‘Satan’ is /yrkuː/, and the sequence /os/, /bos/, /ukbos/, /bõzvos/, /waubos/ are respectively ‘life-thing, animal’, ‘together-life-thing, domestic animal’, ‘human-above-together-life-thing; horse’, ‘together-five-part-active-life-thing, cat’ (I suppose because cats are said to be awake only 20% of the time); ‘power-space-human-together-life-thing, dog’. As DE says, it’s all obvious — which is why there are no known natural oligosynthetic languages.

  240. Trond Engen says

    Plus fresche que n’est rose. “Fresher than a rose is.”

    Should this be read it as “fresher than any rose” = “no rose is fresher”? If so, isn’t it a different use of the negator as in the examples with craint?

  241. David Eddyshaw says

    I took the example and the rendering direct from Einhorn’s grammar, which makes no further comment.

    Still, your idea would be in line with Y’s excellent suggestion that ne in these non-negative uses has shifted from a negation marker to an irrealis marker.

    A shift from a “not at present” tense meaning to irrealis is very common with so-called “discontinous past” forms

    https://www.academia.edu/1761770/Towards_a_typology_of_discontinuous_past_marking

  242. David Eddyshaw says

    Negation and interrogation often pattern together, too: a familiar example is the distribution of English “some” versus “any”:

    I can see some linguists
    I can’t see any linguists.
    Can you see any linguists?

    They also pattern together in Kusaal, where clauses with a negated VP and questions share the feature that the normal loss of word-final short vowels is blocked clause-finally:

    Li anɛ gbigim. “It is a lion.”
    Li ka’ gbigimnɛ. “It is not a lion.”
    Li anɛ gbigimnɛɛ? “Is it a lion?”

    (Cf Mampruli gbigimni “lion.”)

  243. Stu Clayton says

    Oligosynthetic languages are ones that have, let us say, 100 morphemes or less, each very short, and form their words by synthesis of them.

    That being so, “polysynthesis” is even less appropriate than I thought to describe clitic bunching up-front.

    I now suggest “fronted cliticoglomeration”.

  244. David Eddyshaw says

    I agree that “polysynthetic” is a vague term; indeed, all attempts to make it precise seem doomed.

    Basically it describes a language where there are typically a lot of morphemes per word, but (quite apart from where you draw the line for “a lot”), there are a number of problems with this definition.

    On problem (the one which has arisen here) is that “word” is itself a problematic concept. There are a lot of useful criteria for wordhood, like phonological stresses/tones and junctures, potential for interruptability in the middle, capacity to stand alone as an utterance, limitations on ordering or selection of subconstituents and so forth: the trouble is that these useful criteria very often give different results instead of neatly aligning with one another, and you end up having to distinguish (as a minimum) “phonological words” from “syntactic words”, and also with neither fish-nor-fowl “clitics” whose status as affixes, words or word fragments can be not so much a matter of objective fact as how you happen to like your pretty model to look.*

    Even if you can come up with a good definition of “word” for your purposes, there are still a lot of quite different ways that a word may end up containing lots of morphemes, eg. stem compounding, exuberant flexional systems, exuberant derivational systems, and it’s not hard to find languages which do only one of these things and not the others; so lumping all cases together as “polysynthetic” doesn’t seem particularly insightful.

    Mark Baker tried to concoct a Chomsky-inspired “rigorous” definition of polysynthesis in which noun incorporation was the key ingredient, but this never really caught on, and his associated idea that there was a single “polysynthesis parameter” at the back of all this is almost embarrassingly easy to refute.

    *My Kusaal grammar describes several “clitics” which have no actual segmental form at all. I am not ashamed. Oh, no.

  245. ‘God’ is /kuː/

    Kin-dza-dza!

  246. David Marjanović says

    Do I finally understand correctly that a reason, or perhaps the reason, to consider French alone polysynthetic is that its clitics don’t move around anymore like they used to?

    Yes. They’re each stuck to one particular slot in the verb template, like prefixes.

    The question that remains, as you have noted, is whether they’re all obligatory enough.

    you happen to like your pretty model to look

    I’ve recently encountered one that is particularly pretty because it looks like it should really cover all edge cases…

  247. David Eddyshaw says

    Yimas, which is usually regarded as polysynthetic (indeed, archetypically so, except, presumably, in the opinion of strict Bakerites), has a complicated system of verb affixes agreeing with subjects and objects by person and by umpteen genders – and which are not obligatory.

  248. David Marjanović says

    The model I mean is in this book (in German, and not the easy kind). I can’t do it justice right now, though; I’ll try later.

  249. John Cowan says

    the 1588 Bible […] still had the default SVO word order of Middle Welsh, which is now called the “Abnormal Order” and is no more even in the most hifalutin LW

    Indeed. But where did the Normal Order come from? Old Welsh word order is ambiguous: the Surrexit Memorandum (written in the 8C but a copy of a 6C-7C text) begins with a VSO sentence, but the V in question is Latin! Gaulish was SVO, Middle Welsh is pretty much all Abnormal Order, and Cornish and Breton are SVO to this day. In fact, the books of the earliest drafts of the Bible show Normal Order chiefly in Psalms, Proverbs, and Isaiah. Is it conceivable that the Normal Order got into Welsh directly from the Hebrew, and then spread downmarket by elite dominance?

    What’s also interesting is that in SE Wales (the part closest to Cornwall) the Abnormal Order still appeared in Spoken Welsh right up to the 20C in certain constructions.

  250. David Eddyshaw says

    Is it conceivable that the Normal Order got into Welsh directly from the Hebrew, and then spread downmarket by elite dominance?

    But the standard Bible version right up until these degenerate days of Y Beibl Cymraeg Newydd* (NT 1975) was (essentially) Bishop Morgan’s, which overwhelmingly uses the Abnormal Order. The reversion to VSO took place long before 1975.

    There’s also the consideration that although the Abnormal Order was unmarked functionally, it is not unmarked structurally: in positive statements it’s really just the same shape as a (Modern Welsh) clause with a fronted subject. This is the case not only with the construction Subject + (relative particle) a + Verb, but also with the verb “to be”, which has the distinctively relative 3rd person form sydd.

    Williams’ Elfennau Gramadeg Cymraeg [a grammar of Literary Welsh for hapless Welsh secondary-school children and college students] (p126) says “Ceir sydd mewn brawddeg annormal, lle daw’r goddrych yn gyntaf, er nad oes pwyslais arno” (“Sydd is found in the abnormal sentence, where the subject comes first, though there is no emphasis on it”), and cites, for example Psalm 100:5 Canys da yw yr Arglwydd: ei drugaredd sydd yn dragywydd “For the Lord is good: his mercy is eternal.”

    So the question is really one of markedness (as DM rightly pointed out some time back.)

    With regard to the language prior to Middle Welsh, note that Old Welsh still had the verbal distinction between absolute and conjunct forms that we know and love from Old Irish**; this distinction doesn’t make a lot of sense unless it correlated with VSO word order at some point.

    * We do not speak of beibl.net.

    ** Precariously preserved to this day in a few proverbs, e.g. Tyfid maban, ni thyf ei gadachan “The baby grows, his swaddling clothes do not grow.”

  251. David Eddyshaw says

    [Evidently cadachan “swaddling clothes” is YET ANOTHER instantiation of the Universal Root of All Things *kat, btw]

  252. Oh, I don’t deny that Proto-Celtic must have been VSO, but here’s what I’m suggesting: Pre-Proto-Brythonic VSO > Proto-Brythonic through Middle Welsh SVO > Bishop Morgan SVO except in the most poetical books, where he uses VSO > Modern Welsh VSO. Now is this the reappearance of aboriginal VSO (perhaps preserved in really old poetry now lost?), or is it an entirely new VSO? And if the latter, is it just barely possible that it’s calqued on Hebrew?

  253. David Eddyshaw says

    Highly poetical and pious though the Welsh people may be, it seems a bit of a stretch to me …

    I suspect that the VSO of Modern Welsh is more to do with the change of preverbal pronoun subjects into mere affirmative particles, which then got dropped in turn. Modern Welsh has fe and mi (historically, “he” and “I”) as preverbal particles, and they now have nothing to do with person at all, but just mark the clause as independent positive affirmative. (As a clear sign of its Welsh substratum, Kusaal has a tone neutralisation on the main verb in just such clauses.)

Speak Your Mind

*