Anglish?

Bathrobe sent me Isham Cook’s recent post Anglish and English: Why our language is 750 and not 1,500 years old; both the confrontational title and the fact that the page features a misspelled label “Miscellania” suggested crackpottery to me, but I thought I’d pass it along for discussion. The basic thesis is in this passage:

The answer to the riddle of how the Anglo-Saxons were able to force the Britons in England to assimilate to their language and culture so rapidly after 449 is likewise shockingly simple: the indigenous Britons in southern and eastern England were not Celts but were ethnically kin to the invaders and spoke a related language. “Anglo-Saxon,” as we will see, is not quite the right term for this language but refers to the invaders’ tongues. It could simply be called “English,” though it is distinct and foreign to the English spoken today. According to recent computational research (Forster and Renfrew), this archaic English, which I call “Anglish,” broke off from Common Germanic during earlier waves of migration to England hundreds or thousands of years before the Anglo-Saxon invasion of 449, and forms its own branch on the Germanic tree. The conventional view is that English descends directly from the West Germanic family of languages (Frisian, Dutch, and German); in the revised view, English is more closely related to the eastern branch of North Germanic (Danish and Swedish).

The post is very long, but you can skim it (as I did) and get a good idea of what’s going on. Have at it, and throw your brickbats or laurel wreaths in the comments below.

Comments

  1. minus273 says

    If this summary is to be believed, then that recent computational research is nothing but Garbage in, garbage out sprinkled with statistics magical powder. Said magical powder actually do work in biology and quantitative social sciences, because the sheer magnitude of the data makes consistent but weak pattern visible to the method. But no way with cognacy tests & Swadesh lists.

  2. SFReader says

    If it existed thousands of years before 449 AD, then it makes no sense to insist on migration TO England, since Germanic languages broke up sometime in the second half of 1st millenium BC.

    Surely it would be more patriotic to suggest that Germanic languages originated in England and all continental Germanic languages are a result of migrations FROM England.

  3. Marja Erwin says

    The population of Roman Brittania is usually estimated at around 3,000,000 people. The population of post-Roman England is usually estimated at less than 1,000,000 people. The Burghal Hidage is a useful clue, since it indicates the strength of the militia, although it only covers Greater Wessex. I suspect – following Michael Jones – that this demographic collapse was due to late-Roman soil erosion, and famine.

    Does social collapse make language replacement more likely?

  4. Yvy tyvy says

    When I saw the word “Anglish,” I thought this was going to be about modern English without borrowings. I am saddened.

  5. Trond Engen says

    Rather, it’s the two centuries between 1100-1300 that something momentous happened to English, more change packed into a shorter time than any other major language in the historical record experienced.

    The kind of transformation Mainland Scandinavian went through around 1300-1500 is surely comparable. Also, though I can’t date it, Dutch and Low German. This is surely no coincidence, but rather a result of koinéization processes in the increasingly mobile zone around the North Sea. Because Old English started the transformation to Modern North Sea Germanic earlier, I actually find it easier to read than Old Norse of the same age.

  6. I don’t have time to find the reference, but didn’t Julius Caesar write that the people on both sides of the English Channel spoke the same language? And the Romans said that the inhabitants of Gaul were Celts. And whatever Roman records say about the British (not much I think) seems to indicate Celts. I don’t know of any evidence that there were Germanic speakers in Britain before 400.

    Unfortunately there’s not much documentation of what happened in Britain after the Romans left. There’s the Arthurian material but it’s gone through a lot of transformations that make it hard to pick out what could be genuine history. Some people have made an attempt but there’s no real way to verify it.

    I believe that a comparable case exists in modern-day Turkey, where the genetic evidence indicates that the population has been pretty stable for thousands of years. But those people have been Hittites, Luwians, Greeks and now Turks. Why did all those Greeks become culturally Turkish? Presumably because there was some advantage doing so. But other Greeks remained culturally Greek and were forced out after WWI.

    The English were conquered by the Normans even though the number of Normans was pretty small. But in the end the English language became a French-English hybrid.

    On the other hand the Manchus conquered China but Manchu culture was swamped by Chinese culture. And I suppose that you could say that although Rome was conquered by barbarians, in the end Roman culture won out in the form of the Roman Church.

    It seems that things could tip either way.

    Britain was Roman a lot longer than Romania was, but yet Britain didn’t adopt a Romance language, but Romania did.

    Complicated questions. I hope some of the experts here can contribute.

  7. Surely it would be more patriotic to suggest that Germanic languages originated in England and all continental Germanic languages are a result of migrations FROM England.

    I like it! Start a blog to propound the theory; I’m sure you’ll get a following.

  8. David Marjanović says

    I had a long comment in moderation. Did I make it disappear by clicking “Cancel” (not “Delete”)?

  9. David Marjanović says

    Said magical powder actually do work in biology and quantitative social sciences, because the sheer magnitude of the data makes consistent but weak pattern visible to the method. But no way with cognacy tests & Swadesh lists.

    The dataset consists of just 56 words… that’s far too little.

    *reading on*

    They assumed a fixed mutation rate. I’m out of words.

  10. I had a long comment in moderation. Did I make it disappear by clicking “Cancel” (not “Delete”)?

    There’s nothing in moderation now. I assumed that when a comment was in moderation you just waited until it got approved; do you have to click something?

  11. David Marjanović says

    There was a link in it, and I edited it to fix a borkquote. Editing a comment with a link in it puts it into moderation. Strangely, the problem didn’t go away. So I clicked “edit” again, found I had done the right thing, assumed I needed to refresh or something and clicked “cancel”. And the comment was gone.

    I also submitted the comment (adapted to a different audience) at Cook’s blog, where it went into moderation, presumably just because I’m a first-time commenter there. I wrote another comment, apologizing for different formatting issues (the blockquotes are really bombastic there). This, too, went into moderation. Now I’ve checked back, the second comment is still in moderation, the first is gone. o.O Anyway, I don’t have it anymore and would need to write it from scratch, which I might do later, but of course I doubt I have any special insights to offer to present company.

  12. Britain was Roman a lot longer than Romania was, but yet Britain didn’t adopt a Romance language, but Romania did.

    Britain did adopt Latin and developed it’s own variety of vernacular Latin (strongly influenced by the Brittonic substrate) which survived the withdrawal of the imperial administration and the Roman troops in the 4th century. But it was probably spoken mainly in the urban centres of southeastern England, the first to be overridden and linguistically assimilated by the Anglo-Saxons. Some refugees from that area moved to Wales, Cornwall and Strathclyde, taking their spoken British Latin with them, but they were outnumbered by Old Brittonic speakers in those parts, and Latin no longer had an army and a navy to back it up, so it died out by the end of the 7th century.

    The situation in Romania was similar during the existence of Dacia Traiana: Latin was the language of the townsfolk and the administration, whereas the indigenous Dacian language (related to or identical with Proto-Albanian) continued to be spoken in the countryside. It isn’t clear what happened to the Romanised Dacians after the withdrawal of the Romans in the 270s. Some towns survived and Roman influence was still strong despite the general collapse of the administrative system and the devastation caused by the migrating Germani and the Carpi. Quite certainly many Daco-Romans escaped to the southern Bank of the Danube, into Moesia, and their descendants returned to present-day Romania centuries later. We don’t know if Daco-Romanian developed in situ (the continuity theory) or was re-imported from the Balkans at the end of the Great Migrations (the immigrationist theory). Or both, perhaps.

  13. The national personification of the Anglish speakers is a jovial Anglishman called John Bullshit.

  14. Piotr: In yn alltr bethisad, lla linghedig Ladin di lla Ysl Prydein rhen mherif; sa subrwiwaf e punef lla “P-Celtaidd”, gwiwen sul i lla Arvoreg.

  15. “Britain was Roman a lot longer than Romania was, but yet Britain didn’t adopt a Romance language, but Romania did.”

    If the Romanian Urheimat is on the other side of the Danube (around Niš is considered a good candidate), that area was Latin-speaking for a fairly long time, definitely longer than the Dacia that Romanian nationalists want to trace themselves back. Romanian and Albanian were closely linked in prehistory, and Albanian shows multiple layers of Latin loans, suggesting that contact went on for centuries.

  16. per incuriam says

    Surely it would be more patriotic to suggest that Germanic languages originated in England and all continental Germanic languages are a result of migrations FROM England

    Piltdown English.

    I don’t know of any evidence that there were Germanic speakers in Britain before 400

    This was discussed here back in 2006.

  17. John, Polish never developed in the Bethisad universe, so I am not there at all.

  18. Romanian and Albanian were closely linked in prehistory, and Albanian shows multiple layers of Latin loans, suggesting that contact went on for centuries.

    And of course there is a layer of “Dacian” substrate words in Romanian, many of which are clearly Proto-Albanian. I emphasise the “proto” part because they reflect a stage hardly distinguishable from reconstructed Proto-Albanian. Cf. Romanian cioară : Albanian sorrë ‘crow’, Daco-Romanian barzë ‘stork’, Aromanian bardzu ‘white’ (of horses) : Albanian bardhë ‘white’, etc. Aromanian and Istro-Romanian show the same substratal influence, so it’s a common Romanian feature, not necessarily pointing to Dacia. Romanian and Albanian look like descendants of the same partly Romanised bilingual speech community, re-segregated into a Latin-influenced cluster of Albanian dialects and an Albanian-influenced cluster of Romance dialects.

  19. Self-correction: barzë -> barză (same schwa, a different spelling).

  20. Piotr: Polak remańe Polak, wól sie woka “Wened” u parla lęgwa łacina.

  21. David Eddyshaw says

    “earlier waves of migration to England hundreds or thousands of years before the Anglo-Saxon invasion of 449”

    Beallucas.

  22. David Marjanović says

    Some refugees from that area moved to Wales, Cornwall and Strathclyde, taking their spoken British Latin with them, but they were outnumbered by Old Brittonic speakers in those parts, and Latin no longer had an army and a navy to back it up, so it died out by the end of the 7th century.

    That late? What’s the evidence for this?

    There is evidence for the refugees: Peter Schrijver has ascribed the seemingly paradoxical Latin features of Welsh, which look like substrate influence, to those people. (Loads of superstrate loanwords came in earlier.)

  23. David Eddyshaw says

    Just happened to be reading a book by an actual scholar (albeit of some time back), Kenneth Jackson: “Language and History in Early Britain.” Apart from the fact that (as hardly needs saying) real scholarship of this sort as opposed to pseudoscientific fantasy makes it abundantly clear to those of us who don’t believe in flying saucers that the inhabitants of Britain were speaking Celtic languages when the Romans came, Jackson’s book taught me something new. Brythonic languages are full of Latin loanwords dating from Roman times, which is hardly surprising, but what is odd is that while they can hardly all be learned words, they don’t show some features of contemporary Vulgar Latin (like the lowering of short i u to e o, and the loss of the old quantity distinction in vowels) and this despite the fact that there is abundant evidence for widespread actual Latin speech in Roman Britain, in towns at any rate. In fact, people who presumably actually spoke British almost always used Latin if called upon to write anything. Jackson’s idea is that outside the towns and in the less central areas which mostly weathered the Roman collapse better in the end, the Latin that has influenced Brythonic was the Latin of the (British) landowners and their families, which was learnt as a second language (albeit early) and much more influenced by the prescriptions of the normative grammarians than the authentically Vulgar Latin of yer Gaulish peasants.
    It all seems a bit contrived, but there’s certainly a real issue needing to be explained (and one which I never noticed), which is interesting.

  24. David Eddyshaw says

    Dipping into Tacitus’ bio of his father-in-law (which is actually not all that forthcoming on ethnography) I found

    Caelum crebris imbribus ac nebulis foedum.

    Got that right, mehercle. Can see it out of my window right now …

  25. It all seems a bit contrived

    It seems pretty reasonable to me. (Certainly more reasonable than the crackpottery I linked to.)

  26. Lars (the original one) says

    comment was gone — I’m sure I’ve pressed cancel on the edit box on occasion when I realized I had already been clever enough, and with the intended result. I.e., the unedited comment remained, so your action should not have deleted yours. But there might be a bug when it’s in moderation.

  27. David Eddyshaw says

    I expect you’re right. Jackson is a sort of anticrackpot. Abundant documentation of his evidence and clear exposition of his reasoning. Worthwhile even if you end up unconvinced, and you’d need to do as good a job as him to justify differing from his views.

    AFAIK much of his work hasn’t really been advanced on, even after all this time, but I’m far from up to speed with it all. David M?

  28. SFReader: Your idea of England as the Urheimat of Proto-Germanic should be combined with the proposal (which some scholars actually made in the nineteenth century) that Welsh and Cornish spread to Wales and Cornwall from Brittany after the fall of the (Western) Roman Empire, and thus that Brittany was where Proto-Brythonic was originally spoken. This proposal combines patriotism (i.e., English is more of an indigenous language in the British Isles than Cornish or Welsh) and scholarship, as the presence of Proto-Brythonic outside of the British Isles would nicely and elegantly explain why no stratum of non-Old English Germanic loanwords can be found in Proto-Brythonic. We should write a grant proposal together to explore this piece of insane crackpott– err, I mean, this new transgressive counter-hegemonic socio-historical and linguistic paradigm.

    David Eddyshaw: Jackson’s claim, that the Latin loanwords found in British Celtic are unusually conservative, has been to my mind quite thoroughly refuted: see Gratwick, A. S., 1982. “Latinitas Britannica: Was British Latin Archaic?”, in N. Brooks (ed.), Latin and the Vernacular Languages in Early Medieval Britain, (Leicester 1982), pages 1–79.

    Piotr: the Romanian-Albanian similarities extend far beyond vocabulary, and indeed the convergence found between the two languages is very suggestive of large-scale bilingualism: it was pointed out long ago that Romanian AM “I have” is phonologically inexplicable as a regular reflex of Latin HABEO, and that mutual Romanian-Albanian influence is the only way to account for the similarity of this form with Albanian KAM. Which merely strengthens the point you are making, that Romanian and Albanian must for a time have been close geographical and social neighbors.

    About your claim that British Latin/Romance was strongly influenced by the Brythonic substrate: do you have a source for this other than Peter Schrijver?

  29. Lars (the original one) says

    comment was gone — something fishy is going on. I wrote a comment about how cancelling your edit shouldn’t make a comment vanish, and while it was still in editable state, I made a further comment for testing purposes that I put enough links into to force moderation, then saving and cancelling a few edits with no ill effects. But when I then deleted the testing comment to reduce clutter, it ate the non-moderated non-testing comment before it as well!

    So something’s buggy. Always backup your valuable thoughts to a scratch monkey before hitting send or save.

  30. I got the following under “Please moderate” in my inbox:

    Author: Lars (the original one)
    Comment:Testing — this comments should go into moderation because it links to language and philosophy on Wikipedia.

  31. Lars (the original one) says

    That’s expected. Wonder what would happen if you approved it after i deleted it from the edit interface.

    It’s very likely that the author of the moderation functionality never considered the possibility of an edit plugin changing things in the database, and vice versa. And having poked at the innards of WordPress plugins on occasion, I can only say that I understand them.

    So when editing comments that are in moderation, expect the unexpected and save a copy.

  32. David Eddyshaw says

    @Etienne:

    Jackson’s claim, that the Latin loanwords found in British Celtic are unusually conservative, has been to my mind quite thoroughly refuted: see Gratwick, A. S., 1982. “Latinitas Britannica: Was British Latin Archaic?”, in N. Brooks (ed.), Latin and the Vernacular Languages in Early Medieval Britain, (Leicester 1982), pages 1–79.

    Very interesting, but sadly inaccessible. Does it lend itself to a précis? Is Jackson’s error misidentification of the actual source of the loans, or oversimplification of the variety of contemporary Latin speech, or is it something more interesting yet?

    My own feeling of unease about his explanation was based on the fact that I find it difficult to imagine how L2 Latin at a time when there were plenty of L1 speakers about in Britain, let alone elsewhere, could maintain an essentially distinct phonological system, no matter how uppity the prescriptivists might get. I would have thought it was a good deal easier to bully your pupil into writing bucca not bocca than to correct his saying [o] instead of [ʊ], especially as it seems pretty clear that the Classical values of the letters had got pretty stretched latterly, to the point where the errant schoolboy might well just feel that bucca was just the right way to write [bok:a] in any case. A fortiori for things like vowel length. And if educated native speakers actually normally did say [bok:a] and write it as bucca, it seems improbable that the lad would maintain a weird spelling pronunciation not only in this one word but right across the whole system of the language. His Gaulish girlfriend would laugh at him …

    I await enlightenment …

  33. That late? What’s the evidence for this?

    Latin inscriptions from Wales and Northern Britain, showing vernacular Vulgar Latin features.

  34. David Marjanović says

    David M?

    No idea.

    About your claim that British Latin/Romance was strongly influenced by the Brythonic substrate: do you have a source for this other than Peter Schrijver?

    Uh, I’m the one who brought up Schrijver – for saying the opposite, namely that the oddly Latin-like grammatical features of Welsh (I only remember the pluperfect, but there’s more) come from a Latin substrate brought in by fleeing Romans who would rather live with Christian barbarians than die at the hands of heathen barbarians. The Latin loanwords which e.g. distinguish a from ā came in much earlier in this scenario, back when Latin was a superstrate as you’d expect.

    something fishy is going on. I wrote a comment about how cancelling your edit shouldn’t make a comment vanish, and while it was still in editable state, I made a further comment for testing purposes that I put enough links into to force moderation, then saving and cancelling a few edits with no ill effects. But when I then deleted the testing comment to reduce clutter, it ate the non-moderated non-testing comment before it as well!

    It has resurfaced.

    Latin inscriptions from Wales and Northern Britain, showing vernacular Vulgar Latin features.

    Awesome. Next question: some features of Old High German, in particular a lot of more or less optional vowel epenthesis and the complete lack of a distinction between the vowel systems of stressed and unstressed syllables*, have been blamed on a Romance substrate; why aren’t there any such features in Old English? Or are there?

    * Reversed with a vengeance at the beginning of Middle High German. All the epenthetic vowels were purged, almost all other unstressed vowels were merged, and consonant clusters and overlong stressed syllables have been accumulating ever since.

  35. David Marjanović says

    Cook has let all my comments out of moderation and replied to them! I’ll celebrate by going to sleep early. ^_^

  36. David: The comment of Piotr’s I was asking about was this one (10:09 AM) :

    “Britain did adopt Latin and developed it’s own variety of vernacular Latin (strongly influenced by the Brittonic substrate) which survived the withdrawal of the imperial administration and the Roman troops in the 4th century. But it was probably spoken mainly in the urban centres of southeastern England, the first to be overridden and linguistically assimilated by the Anglo-Saxons.”

  37. Stephen C. Carlson says

    @David Eddyshaw: Very interesting, but sadly inaccessible. Does it lend itself to a précis? Is Jackson’s error misidentification of the actual source of the loans, or oversimplification of the variety of contemporary Latin speech, or is it something more interesting yet?

    I was able to find a scan of the book somewhere online. Gratwick has a brief summarization of the “Pervasive flaws in Jackson’s case” on pp. 62-63 as follows:

    “Jackson’s argumentation as a whole is weakened by his failure to allow for the fact that loans, once established as native words, fossilize traits of the phonology of the original language. He takes too simple a view of what is ‘Classical’ and what is ‘Vulgar’ Latin, of the certainty of the absolute dating of phenomena in the development of the latter, and of their generality throughout the Empire. He assumes wrongly that the earliest loans in British must date from the time of Agricola. In fact we must reckon in principle which the possible entry of some loanwords already before the time of Christ.”

  38. David Eddyshaw says

    @Stephen C. Carlson

    Much obliged!
    The brief summarisation looks pretty unconvincing (especially given the fact that Jackson is nothing if not thorough), but that’s hardly the way to form a proper judgment, of course. I’ll see if I can emulate your impressive Google-fu.
    The basic point seems to be a combination of my two guesses.
    I’m delighted to find that the question of Latin loans in Brythonic is so much more complicated than I thought.

  39. Andrew Smith adopted for the purposes of Brithenig the “classicizing Latin” theory, as you can see at the Grand Master Plan, which specifies the sound-shifts in brief. (Not near as complex as the Wenedyk GMP, I’m glad to say, which tracks Polish all the way from Proto-Slavic.)

  40. Lars (the original one) says

    It has resurfaced — yes, it was probably never gone at all, I just wasn’t prepared for two new comments appearing in the few minutes it took to do the experiment _and_ for the act of deleting the experiment post to reorder the page so it disappeared from view.

  41. Totally unrelated, but has anyone come across holy smoked, as in:

    “We got holy-smoked that there’s possibly a tree that’s leaning toward the equator wherever it grows,” says Ritter.

    Cook pine

  42. David, that makes me wonder if MHG as we know it is perhaps not the descendant of OHG as we know it. Some of the “Duke of York” changes between OE and ME just reflect the fact that OE is standardized Wessex (which had the original changes like vowel breaking) whereas ME is a later standardization of West Midlands (which did not have those changes).

  43. David Marjanović says

    that makes me wonder if MHG as we know it is perhaps not the descendant of OHG as we know it

    I doubt that. OHG didn’t have any standard, what little was written was always in the local dialect as far as anyone has been able to tell. While MHG did at some point acquire a prestige dialect that all the troubadour literature is written in, some diversity is known; and various changes in the sound system and the grammar can be tracked through time within both OHG and MHG. Finally, all High German dialects except Highest Alemannic can be derived entirely or practically entirely from MHG as we know it, but there was neither any large-scale migration nor any rise of a particularly prestigious urban center or something between OHG and MHG.

    Also, the sound system of unstressed syllables in OHG is simply a retention, conspicuous only because all the other Germanic languages of the time had already simplified it; and the epenthesis was the culmination of a trend that began deep in Proto-Germanic prehistory – apparently things like *miluk “milk”, with the same epenthetic *u as the one that was used to “resolve” the syllabic sonorants, can be reconstructed all the way down.

  44. Trond Engen says

    How much of the Balkan Sprachbund is Albanian substrate?

  45. Etienne,

    What I had in mind was mainly the kind of indirect evidence analysed by Schrijver. There’s little doubt that most features of vernacular British Latin were shared with the neighbouring continental varieties, and that areal innovations crossed the Channel easily. Anyway, since early Brittonic was not yet very different from Gaulish, the substratal influence in Gaul and in Roman Britannia would have produced similar effects. Therefore, the British Latin of the surviving short inscriptions just shows the expected traits of northwestern Vulgar Latin. Some of the important Latin documents from Britain (such as the Vindolanda tablets) do not represent British Latin at all. One could expect specifically insular innovations to accumulate in the sub-Roman period, with Latin and Brittonic influencing each other (as well as interacting with Old Irish) and Britain growing more isolated, but unfortunately this stage is very poorly documented. From the late Roman period, we have the Bath curse tablets, which do represent the local variety of Latin and show a few possible regionalisms (lexical as well as phonological), discussed by Adams (1992, British Latin: The text, interpretation and language of the Bath curse tablets, Britannia 23, 1–26). They are not necessarily attributable to Brittonic influence, however.

  46. Trond Engen says

    David M.: there was neither any large-scale migration nor any rise of a particularly prestigious urban center or something between OHG and MHG

    One major change around 1000 AD is the emergence of Germany as a political entity, with a unified German Kingdom as the German-speaking core of the Holy Roman Empire. This came with increased mobilty, with busy market towns giving merchants and skilled artisans the opportunity to move out and take up shop in other parts. The Ostsiedlung was just this energetic urban mobility working out on Slavic lands, Even (or maybe especially) without an established prestige variety you would expect a great deal of linguistic levelling between dialects.

  47. Totally unrelated, but has anyone come across holy smoked

    Not I, and if that’s the only place you’ve seen it I’d guess it was a nonce creation.

  48. However, there is a definition of holy smoke at MW.

  49. Oh, sure, “holy smoke” is an exclamation of long standing — I’ve used it myself (in jest). I’m talking about the participial form exhibited in “We got holy-smoked.”

  50. David Marjanović says

    Even (or maybe especially) without an established prestige variety you would expect a great deal of linguistic levelling between dialects.

    That’s what we see in the new east, but apparently not elsewhere. A bunch of isoglosses are older than that, like the whole Rhenish Fan and the shortening of long consonants in overlong syllables in Franconia and north of there; they show up in OHG spelling.

    And the gap between OHG and MHG is not so much around 1000 as around 1150.

  51. Trond Engen says

    I mainly responded to this:

    Next question: some features of Old High German, in particular a lot of more or less optional vowel epenthesis and the complete lack of a distinction between the vowel systems of stressed and unstressed syllables*, […]

    * Reversed with a vengeance at the beginning of Middle High German. All the epenthetic vowels were purged, almost all other unstressed vowels were merged, and consonant clusters and overlong stressed syllables have been accumulating ever since.

    What I meant to suggest was that the Germanic main syllable stress (or the phonological results of it) finally spread to High German with the increased mobility. The change becoming visible around 1150 doesn’t mean it didn’t start earlier, just that it took time before the old monasterial spelling traditions became untenable.

  52. “How much of the Balkan Sprachbund is Albanian substrate?”

    Brian Joseph’s 1983 monograph “The Synchrony And Diachrony of the Balkan Infinitive” concluded that at least the loss of the infinitive was a result of Greek influence on South Slavic, Romanian, and (Tosk) Albanian.

  53. David Marjanović says

    Oh, the stress was already there, and increased mobility between commercial centers doesn’t get you into alpine valleys. Little enough was written in OHG that there wasn’t much of a spelling tradition; there’s a lot of variation in how the same phonemes were represented.

  54. My god, I knew Brian Joseph (briefly) over forty years ago! Small world, especially in historical linguistics.

  55. “got holy-smoked” if you Google it in quotes and with the hyphen shows up ONLY in the Cook pine context. There are a few other instances of “was holy-smoked” etc. (One from 1898!) Not a bad expression — let’s make it happen!

  56. Okay. “The altar boy holy-smoked the congregation until they were all coughing and wheezing except Robert Samuel Joseph Alfred Kaffiraza, Esq.”

    This reminds me of Dorothy Parker’s attempt to use “opium” in a sentence: “I opium mother is having a good time.”

  57. January First-of-May says

    Supposedly the Cook pines in Peradeniya Botanical Gardens, Sri Lanka, don’t actually consistently lean anywhere, and in fact some trees are so crooked they can’t be said to lean in any particular direction.

    I’ve seen an article about Peradeniya that claimed they actually lean in the direction of prevailing wind. That doesn’t seem very likely, however (prevailing winds would probably mostly not be oriented on a north-south axis).
    OTOH, a “turn towards the sun” pattern averaged out over days and years would provide exactly the observed result (slight slant in the direction of the equator).

    (The stock photo in the New Scientist article is from Peradeniya, by the way.)

  58. gnaa! Will people please stop rewriting the history of 1500 years ago. It’s hard enough keeping up with today’s history. (And now I’ve got to go find out about Basque-Icelandic pidgin.)

    So “England/English” comes from “Anglo” in Anglo-Saxon comes from (now called) “Angeln” in Jutland. So far I knew. But the people-from-Angeln didn’t speak Scandinavian at the time (says David M); rather something very close to Saxon.

    If Anglo and Saxon are much the same language, why didn’t all those inflections/cases just merge? Why did so many get lost/replaced by prepositions? Why did the language change to strongly word-ordered? Seems over much of a drama.

    Are the Angeln’s the bunch (“Vikings”) that turned up in longboats, raped/pillaged the North-East as fabulated by the venemous Bede, established the Danelaw? How were they (and their language) related to the “Vikings” that arrived in Northern (now called) France/became Normans? Where do the Norwegians in the North Atlantic come into it? What language did they speak at the time?

    I could go on …

  59. There was a germanic system of plowing heavy soils, using community resources, that stands opposed to the celtic independent farmer system, which worked better with lighter soils, easily wasted out. Saxon immigrants to Britain could farm rich wet bottomland without shoving aside the celts, This could be background to a hon-heroic Saxon “invasion” that may have begun long before 449 ad. For a time, especially with late Roman governance smoothing everything, it may have been an experiment that kept going on its own steam.
    Saxons involved in the experiment may well have communicated with their mainland relatives before, during and after the Roman withdrawal. These would not have been so divorced by time as to have language problems. Christian Saxons from Britain were used as missionaries to Saxony around 800 ad. This would not have been possible if they spoke a language too many centuries divorced from mainland German.
    I substitute my deep-plowing germanics in place of too-early Belgae, whatever they were, certainly incapable of converting Saxon pagans to Christianity.

  60. Bathrobe says

    So I gather that the verdict on the Cook article is… inconclusive?

  61. Lars (the original one) says

    gnaa! — well, if the evidence would stop changing maybe the interpretation would too… Right now DNA data are forcing a lot of upheavals in mainstream views, but new discoveries have always been changing what the ‘best’ theory is.

    As to Jutes, Angles and Saxons — my impression was that most of the loss of endings and consequent simplification of grammar in OE happened a few hundred years after the migrations. Is dialect / language mixture actually seen as the major cause of that? It’s a natural consequence of the common Germanic first syllable stress, and North Germanic was much further along at the time I think. (And how High German managed to resist I don’t know — Icelandic and Faeroese has isolation as their explanation).

  62. Even in Wulfila’s Gothic the loss/syncretism of inflectional endings was already underway for purely phonological reasons (loss of unstressed *a, *i in final syllables, shortenings and mergers affecting unstressed long vowels).

  63. David Marjanović says

    It’s hard enough keeping up with today’s history.

    True!

    And now I’ve got to go find out about Basque-Icelandic pidgin.

    It’s all on Wikipedia. 🙂

    Are the Angeln’s the bunch (“Vikings”) that turned up in longboats, raped/pillaged the North-East as fabulated by the venemous Bede, established the Danelaw?

    No, no! That was a later invasion of actual Norse-speaking Vikings. And when they settled in northern England in large numbers around 300 years after Bede, that seems to have caused Late Old English to make compromises with Old Norse, which then spread north-to-south in Middle English.

    Saxon immigrants to Britain could farm rich wet bottomland without shoving aside the celts, This could be background to a hon-heroic Saxon “invasion” that may have begun long before 449 ad.

    Very interesting point.

    Christian Saxons from Britain were used as missionaries to Saxony around 800 ad. This would not have been possible if they spoke a language too many centuries divorced from mainland German.

    Of course not. Old English and Old Saxon aren’t that far apart; distort the vowels a bit, and it’s pretty much the same.

    So I gather that the verdict on the Cook article is… inconclusive?

    The article makes several claims, and the verdicts on them are all different. 🙂

    The title claim isn’t a claim so much as a matter of nomenclature, of convention: do we privilege cladogenesis, i.e. do we call it English as soon as English and Frisian have split, or do we privilege anagenesis, i.e. do we call it English as soon as it’s easily recognizable as English, after the profound changes of the 13th century or so?

    And how High German managed to resist I don’t know

    It didn’t resist that much. Nowadays, it’s rather less inflected than Icelandic (I don’t know about Faeroese); many endings are still there, but they’ve all merged into very few distinct ones, as expected from a stress-timed language whose stress isn’t word-final.

    What I said about vowel epenthesis in OHG is the culmination of another Germanic trend: an ever-greater emphasis on syllables over words to bring syllables closer to the CV ideal. Germanic not only kept Sievers’s law from PIE as a persistent constraint, but even developed a converse of it. Even such things as the West Germanic consonant lengthening can be explained this way (I’ll look for my source later): -VC.jV- is a “bad syllable contact” due to its rise in sonority instead of the drop of the -VCV- ideal; reinterpreting it as -V.CjV- is not an option because it would change a heavy syllable into a light one; changing it into -Vː.CjV-, as Norse eventually did, has a wonderful sonority drop, but perhaps such a change in vowel length in grammatical derivation (*-ja- being a common suffix) was too confusing; West Germanic went for the Italian option, -VC.ːjV-, which preserves syllable and vowel length and at least doesn’t have a sonority rise anymore. And the output of this procedure was promptly subjected to Sievers’ law again!

    Eventually, this trend conflicted with the consequences of first-syllable stress and was abandoned, indeed reversed into an emphasis on words over syllables.

  64. Lars (the original one) says

    So if I got that right, language contact may indeed have sped up the loss of flexion in English but it was contact / mixing with Old Norse, not between the Ingvaeonic dialects of the 5th century influx.

    On a more general level, has the tendency to see creolization wherever language evolution is a bit obscure abated by now?

  65. How it usually works mathematically.

    Densely populated region with area of 30-40 thousand sq. km and population, say, close to half a million, experiences invasion, war, total breakdown and civilizational collapse.

    As a result population drops fivefold – to 100 thousand.

    20 thousand Anglo-Saxon Barbarians move in. All adult males.

    They take wives from the conquered population – about 20% of remaining population which is equal to all women of childbearing age.

    Anglo-Saxons reproduce, Britons don’t (because all their women are taken by Anglo-Saxons).

    Voila, 100% population replacement within a generation.

  66. J.W. Brewer says

    I guess “Germanic speakers already in Britain in substantial numbers in Roman times” is one of those things we can’t *completely* rule out on the basis of it not being mentioned in any actual historical sources because we know those sources are patchy and incomplete. But they are probably sufficiently patchy and incomplete that various other stories one could concoct that seem more far-fetched than that could likewise not be ruled out simply because the surviving written sources do not mention them.

    I am puzzled by Phil Jennings’ point re English missionaries to the heathen Continental Saxons circa A.D. 800 — just because I don’t see how that poses any problem for the conventional timeline (i.e. speakers of what became Anglo-Saxon don’t arrive in Britain until after the Romans are gone, but start being Christianized circa A.D. 597, with that process mostly completed by the time Bede is born approx 75 years later) that gets solved by the unconventional timeline proposed by Cook.

  67. Phil Jennings says

    In answer to Brewer, I offer my slightly early Saxons in place of the improbable Belgae. The Romans were always trying to bring the good life to the parts of their empire north of the olives-and-wine zone by bringing in improved breeds of sheep, etc. and may have seen Saxons as an improved breed of farmer. The heroic invasion circa 449ad becomes an episode in a process rather than a shock out of the blue. The process may have begun as late as the Fourth Century. Time brings estrangement so my guess is not earlier.

  68. @SFReader:

    > Anglo-Saxons reproduce, Britons don’t (because all their women are taken by Anglo-Saxons).

    1. Doesn’t this mean that Britons are reproducing, as well as Anglo-Saxons (namely, the Briton women are reproducing)?

    2. Don’t babies generally acquire the conquered language, from their mothers? (For example, in colonial Brazil native Tupi was the predominant language, even among the children of Portuguese conquerors and their descendants; we have documents where priests bitterly complain that white children would only ever learn Portuguese in school. It took waves of new European migrations a couple centuries later, as well as explicit anti-linguistic repression programs and a generous dash of genocide, to finally get rid of Tupi as the general language. Paraguay, a mirror image in many ways, did not suffer through the same degree of repression, and as a result still routinely speaks Guarani as well as Spanish.)

  69. I see some arguments in the thread that phonetic indistinctness of German suffixes drove the simplification of morphological cases. It is my understanding that one can’t explain the Romance reductions in this way. Not only Romance languages aren’t “stress-timed” (a concept, I might add, that’s not quite unproblematic), but they’ve all uniformly lost certain endings, like -ōrum, which had a lot of phonetic prominence. Phonetics seem to play no role in determining which suffixes survived (or else -ōrum, -ibus etc. would be high on the list.)

    I think morphological simplification is too general a phenomenon to be attributed simply to an artifact of phonological reduction. Consider that loss of cases occur throughout Indo-European history in about all branches, regardless of how “Germanic-like” are they; and consider that similar reductions occur widely in other language families, e.g. Japonic (compare the old past/perfective tense/aspect/mood Japanese suffixes with the modern ones), or Tupian (Old Tupi to Nheengatu), and so on. Consider that creoles have no inflections, that inflections are acquired late by infants (wherefore the old saying “Norwegian is Icelandic as spoken by three-year-olds”), and that inflections are notoriously hard to acquire by second-language speakers—all of which are true regardless of how phonetically distinct are the inflections.

  70. Interesting point. I guess it falls under the general heading of “language is always simultaneously simplifying and complicating itself.”

  71. SFReader says

    @Leonardo Boiko

    1. Germanic people like most Indo-Europeans define ethnicity by father.

    2. We have several interesting cases in this regard. Pitcairn island English is probably the best known. After bloodbath among Bounty mutineers, the island was left with one adult male (English speaker), nine adult women (all Tahitian speakers) and 19 small children. English prevailed. Prestige language (spoken by males) almost always wins.

    Another similar case is Niniltchik Russian (village founded by three ethnic Russian adult males, three Kodiak Eskimo women and their children, surrounded by Tanaina Indians. result – north Russian speech without a trace of Kodiak Alutiiq language, with about a dozen loanwords from Tanaina language)

    Regarding Guarani- the Jesuits had a system which gathered Indians into crowded missions carefully isolated from the white population. And they deliberately promoted Guarani to ensure further isolation from the white world and its sins. That’s how Paraguai ended up speaking Guarani (if Jesuit system lasted till independence, probably they wouldn’t even speak Spanish).

  72. David Eddyshaw says

    I’ve read that in Paraguay the indigenes, seeing how the land lay, deliberately adopted the policy of marrying their daughters to the invaders, and that it was these daughters who perpetuated the language. I gather that the language has, correspondingly, survived much better than the pre-conquest culture.

    Copper Island Medny Aleut has it both ways, and it certainly isn’t Russian. Similarly with Michif in Canada. Neither represents the paternal language ousting the maternal. Admittedly these are pretty odd cases, though.

    The various ethnic groups where I used to live in Ghana define ethnicity by father, but this doesn’t stop language being transmitted by predominantly by mothers. In the Bawku area, for example, there are many ethnic Mamprussi who speak Kusaal and can’t understand Mampruli (I used to work with some.) Still, I guess that’s more like the Norman Conquest scenario in Britain, with a small group of predominantly male incomers getting assimilated to a much larger group of original inhabitants, and not like the Pitcairn scenario.

    There must actually be genetic data bearing on this question in Britain, looking at Y chromosomes versus mitochondrial DNA. I know that the genetic evidence refutes the traditional picture of a wholesale slaughter and replacement of the Britons by the ravening Saxon hordes, but I don’t know what it says about paternity. I seem to recall a study suggesting that with the later Norse settlement in the north there is indeed evidence that Olaf married Ethel instead of importing Ingrid.

  73. David Eddyshaw says

    There’s always the Vaupes in Amazonia, where your language is your father’s language, but your mother tongue is always different from that, because marrying someone who speaks the same language as you is basically incest.

  74. I don’t know that Japonic is really a good poster blob for simplification of endings motivated by things other than phonological reduction, is it? What sort of things are you thinking of specifically? (Certainly there were endings that fell out of use or were replaced rather than being worn away, but given the level of agglutinativity involved that’s less remarkable than everyone just giving up on -orum.)

  75. Olaf married Ethel instead of importing Ingrid

    Ingrid was often named Aileen, Maria, or Anastasia anyway. The Norse were big on marriage-by-rape even at home.

  76. ” Saxon immigrants to Britain could farm rich wet bottomland without shoving aside the celts,”

    Phil, this is the pattern in Alsace and that area and also in Belgium, with the boundary between Walloons and Flemings set by terrain. There’s a similar pattern in SE Asia, with specific groups at specific elevations throughout the region.

    “in the revised view, English is more closely related to the eastern branch of North Germanic (Danish and Swedish). ”

    There’s nothing to preclude North Germanic settlements around the Rhine Delta or in coastal Britain quite early. There is a lot of that kind of long range migration/settlement in Europe. Those ships show up in petroglyphs from the Bronze Age. I think there’s genetic evidence of flow between Scandinavia and Britain dating well before the Viking era.

    The big unknowable is what the Belgae spoke. The Romans called them Celts and to us that is defined by Celtic languages. It’s speculation to assume that’s what it meant to the Romans.

  77. Lars (the original one) says

    There is a large technological step up from coast-hugging boats (that do allow long-range migrations) to the sort of ‘Norway to Ireland and back while the barley grows’ raiding that the Vikings did. (Not to speak of colonizing New Foundland). But Calais-Dover is just 20 miles and can be navigated by sight, so Danish genes in England are quite possible without Viking ships, and thus even North Germanic speakers in Roman Britain though a very good reason would have to be come up with to explain their invisibility to contemporary sources.

    But to seriously believe that a method that purports to detect a closer relation of 1st century anything to Eastern rather than Western Scandinavian is not fundamentally flawed when those branches only separated in the latter half of the first millennium, is that a sign of willful or congenital ignorance?

    (IIRC, glottochronology Don Ringe style simply constrains the trees to respect known historical facts, such as known ancestors of English and Danish being very different and attested at the same time, to avoid this kind of silliness).

  78. “But to seriously believe that a method that purports to detect a closer relation of 1st century anything to Eastern rather than Western Scandinavian is not fundamentally flawed when those branches only separated in the latter half of the first millennium, is that a sign of willful or congenital ignorance?”

    No one is even talking about a separation between any branch of North Germanic; surely the question is a separation between North and West Germanic. The distance between East and West Germanic was significant when Gothic and Anglo-Saxon were recorded, surely more than five centuries would allow for. It’s not a set rate, and not knowable, but if you compare the Alemannic dialects with Franconian dialects the distance after a thousand years is less over comparable or even greater distances (if you accept that water connects and land divides populations.)

  79. Lars (the original one) says

    No one is even talking — Cook is, according to the sentence you quoted.

  80. David Marjanović says

    Oh, it’s clear that phonological erosion isn’t the only agent of grammatical change!

    Keep in mind, though, that there are cascade effects. If phonological erosion eliminates a case in the singular, it’s going to be lost in the plural very quickly. The Upper German loss of the simple past tense is obviously due to one particular apocope for regular verbs, but for irregular verbs this doesn’t work – there’s a plausible suggestion that the analogical extension to the irregular verbs was helped along by the preference for SVOV word order which was bolstered by the loss of the simple past of regular verbs.

  81. In Late Old and Early Middle English the phonetically motivated merger of -e, -u and -a as [-ə] and the merger of -en, -an, -on and -um as [-ən] (with a further tendency to drop the nasal) immediately played havoc with the inflectional system. Of course it wasn’t the only reason why the system collapsed, but in some paradigms (e.g. the strong feminine nouns) practically everything was levelled away: ġiefu, ġiefe, ġiefa, ġiefum all merged as yeve (vel sim.) ‘gift’. Very soon the syncretic declensional ending -e in those nouns which had a consonant-final nominative singular was reinterpreted as a “universtal post-prepositional” marker with such a low functional load that the pressure to use it consistently it was relaxed. On the other hand, the endings that resisted loss for phonological reasons (gen.sg. or nom.pl. -es, verbal -est, -eth and so on) were not only retained but also generalised.

  82. marie-lucie says

    Lars (the original one): colonizing New Foundland

    In decades of living in English Canada I have never run across this analysis of Newfoundland, which is always treated as a single word. At best it would have to be Newfound Land ‘newly discovered land’, which I have never encountered either. The French name is Terre-Neuve literally ‘New Land’.

  83. The OED’s first quotation is dated 1504: “Item [Also] to the merchauntes of bristoll that have bene in the newe founde launde”, showing that it was felt as an open compound rather than a proper name at the time. John Donne’s Elegy XX, probably written in the 1590s but not published until after Donne’s death, contains the iambic pentameter line “O, my America, my Newfoundland”. Or at least so modern editions print it, but the 17C editions have “new-found-land” or “new-found land”, which suggests something more vague than the specific island off the coast of Labrador.

  84. marie-lucie says

    Thanks JC for the history. Of course the current name started as an open compound, with “new” and “found” as the original compound, not “found” and “land”. My comment was about Lars’s use of “New Foundland” as a reference to the specific island, instead of the current name “Newfoundland”.

  85. January First-of-May says

    Regarding Newfoundland, I’m reminded of a phrase in the original Strange Maps post on the Atlas of True Names:
    “and Newfoundland… remains ‘Newfoundland’, one of remarkably few place names with an etymology recent enough for us to take the toponym literally.”
    (Other such names include Salt Lake City and, to an extent, Oxford; but of course it depends on which language you’re taking it literally in, obviously. From a Russian perspective, it’s Novaya Zemlya that has a literal etymology.)

    (Incidentally, this combination of Russian island names and literal etymology reminds me of another Arctic archipelago name, the New Siberian islands – the Russian is “Novosibirskiye ostrova”, which makes them sound like they have something to do with the town of Novosibirsk, but in fact it’s an independent derivation, and they’re basically about as far away from Novosibirsk as one can get and still be in Siberia – or even a bit farther.)

  86. Marie-Lucie, would terre nouvelle be grammatical to you? If so, how would you interpret it differently from terre neuve?

  87. Lars (the original one) says

    I blame finger macros.

  88. David Marjanović says

    Bizarrely, “New Foundland” is how it’s stressed in German (in the middle), even though the usual logic of compound nouns predicts first-syllable stress.

    Edit 10 seconds after posting and 10 or 20 years after first wondering: analogy from New Zealand could explain it.

  89. marie-lucie says

    Y: would terre nouvelle be grammatical to you? If so, how would you interpret it differently from terre neuve?

    A newly discovered land would probably not be called Terre-Neuve nowadays, this usage is archaic as neuf/neuve now means ‘brand new’. The most likely phrase would be Nouvelle Terre, which is not very modern either: see for instance the names of territories such as la Nouvelle-Calédonie and cities like la Nouvelle-Orléans. Terre nouvelle would not be ungrammatical, but nouveau/nouvel/nouvelle is one of the adjectives that is normally placed before a noun unless there is a reason to attract particular attention to it (while neuf/neuve always comes after the noun in Modern French).

    I could write a sentence such as L’île de Terre-Neuve s’appelle ainsi depuis le 16ème siècle parce que c’était une terre nouvelle pour les Européens ‘The island of Newfoundland has been called this since the 16th century because it was a new land for Europeans’. In this sentence, using une nouvelle terre could imply that it was an additional land, ‘yet another land’.

  90. Thanks for that; I love seeing these subtleties of usage clarified.

  91. Here’s one I’m fond of from Italian: postposed adjectives are restrictive, preposed adjectives are non-restrictive (modulo the usual lexical exceptions). Thus we call Dante’s work La divina commedia by way of compliment to it, not to distinguish it from earthly comedies (though that works too). He himself simply called it his Commedia, and it was Boccaccio who first applied the adjective.

  92. I have it on the authority of a colleague — a linguist from Memorial University of Newfoundland, St. John’s — that in the local accent both NewfoundLAND and LabraDOR have primary stress on the final syllable.

  93. I knew about NewfoundLAND but not LabraDOR.

  94. Thanks, m.-l.! I’d remembered postposed nouveau used for things like new potatoes.

  95. I recall that NewfoundLAND sounded extra peculiar to me because the vowel in -found- is not reduced, as it is in the US pronunciation of the name.

  96. newfinLAND preserves the stress pattern of “new-found land” as in “We went to a new-found land” anyway, so this pronunciation is conservative. labraDOR also reflects the name of the Portuguese explorer for whom it is named, João Fernandes Lavrador [ʒuˈɐ̃w̃ fɨɾˈnɐ̃ðɨʃ lɐvɾɐˈðoɾ].

  97. January First-of-May says

    For what it’s worth, Russian has second-syllable stress in Ньюфаундленд, but final stress in Лабрадор.

    (Though in Russia they’re much more famous as dog breeds than as Canadian regions, anyway. But then they wouldn’t have a capital letter. Same stress though.)

  98. From Don Marquis, A Temperance Tract:

    Buns use up dough; what my fun did,
    Were it refunded one day,
    Would fund the Banks of Newfoundland
    And float the Bay of Fundy.

    Another witness of final stress (and of shameless wordplay).

  99. J.W. Brewer says

    I see no reason why “my fun did” couldn’t be pronounced as if a single three-syllable word with initial stress, with MY-f’n-did thus slant-rhyming with NEW-f’n-lund. You’ve also gotta mess a bit with the stress pattern you’d expect for “one day” to make it match up with Fundy, which might cast further doubt on the reliability of the evidence …

  100. Maybe. I just assumed everything was strict iambic (except for the catalectic endings), like the other 15 verses of the poem.

  101. marie-lucie says

    Piotr: I have it on the authority of a colleague — a linguist from Memorial University of Newfoundland, St. John’s — that in the local accent both NewfoundLAND and LabraDOR have primary stress on the final syllable.

    This must be true, but in most of Canada the stress is on the initial syllable.

    In a previous discussion of this topic I remember saying I had never heard “New-FOUND-land”, but I have met someone from New Brunswick who uses this pronunciation, unlike anyone else I know.

    JC: newfinLAND preserves the stress pattern of “new-found land” as in “We went to a new-found land” anyway, so this pronunciation is conservative.

    Newfoundland English has many peculiarities, some of them archaisms, because of its long isolation. The island only joined Canada in 1949, after being a separate, rather neglected British colony for centuries. The move was approved by a tiny majority of the population and many older people have a nostalgia for the old days when they had a distinctive identity as Newfoundlanders, instead of being blended with the rest of Canada, although the move brought a measure of prosperity to the island.

  102. Then there’s the question of the province’s legal name: it used to be simply Newfoundland (with maps sometimes specifying Island of Newfoundland for its primary part), but in 2001 it was changed to Newfoundland and Labrador. On the one hand the province’s name became clunkier, but on the other hand the island’s became more intuitive.

    Oh, and another fun thing is the never-ending border dispute with Quebec. In the 1920s the Coast of Labrador (at the time was a colonial dependency of Newfoundland, sort of like the Caymans were to Jamaica) was expanded from a mere strip to a more substantial territory (big enough to co-opt the name “Labrador” from the broader peninsula), but there’s a sizable chunk in the south that has never been definitively figured out. Maps in Quebec show a naturalistic border in its favor, but maps in N&L and elsewhere show a straight-line border in its favor.

  103. January First-of-May says

    And then there’s Killiniq Island, just off the northern end of the (eastern) Labrador peninsula, of which a small eastern strip belongs to (Newfoundland and) Labrador, and the remaining part… to Nunavut, as part of Hudson Bay (Quebec only owns the mainland parts – at least in this particular region).

  104. I know it will be one day,
    maybe Monday, maybe not.
    Still I’m sure it will be some day,
    maybe Tuesday will be my good news day.

    I’ll wish for Sunday on the Isle in Fundy.

    … and so all else above
    the highest tidal range in the world.

  105. David Marjanović says

    I wrote on 10 June 2017:

    Next question: some features of Old High German, in particular a lot of more or less optional vowel epenthesis and the complete lack of a distinction between the vowel systems of stressed and unstressed syllables*, have been blamed on a Romance substrate; why aren’t there any such features in Old English? Or are there?

    * Reversed with a vengeance at the beginning of Middle High German. All the epenthetic vowels were purged, almost all other unstressed vowels were merged, and consonant clusters and overlong stressed syllables have been accumulating ever since.

    Not in fact at the beginning of MHG, I have recently learned, but earlier, and gradually: in the south, the length distinction for unstressed vowels was lost during the OHG period, while their qualities remained, in Switzerland, distinct well into MHG (and are preserved in the more extreme Walser dialects to this day), while in the north the length was kept longer but the qualities of the short unstressed vowels merged earlier. Also, a round of syncope increased the number of heavy stressed syllables shortly after the beginning of MHG.

    Also, it has been argued that when epenthetic vowels show up in writing, as they do in OHG and in one particular stage of Runic Norse, that’s actually a sign of widespread vowel reduction, so that speakers can no longer tell if a phonetically reduced vowel is epenthetic or inherited. I find that very convincing.

    and on June 12th:

    West Germanic went for the Italian option, -VC.ːjV-, which preserves syllable and vowel length and at least doesn’t have a sonority rise anymore. And the output of this procedure was promptly subjected to Sievers’ law again!

    Nope. Sievers’s law somehow ceased to operate around the time of Proto-Northwest Germanic, and superheavy syllables became allowed again. This is demonstrated by a bunch of loanwords in Finnic and by the operation of WGmc. consonant lengthening in some words where the consonant follows a heavy syllable nucleus, which would not have occurred if Sievers’s law had still been in operation.

  106. ‘Newfoundland’, one of remarkably few place names with an etymology recent enough for us to take the toponym literally.”

    Taking literally, Netherlands are in hell.

  107. I remember the feeling of revelation when, at age 13, I realized that “Netherlands” and “Low Countries” were a doublet.

  108. John Cowan says

    Polak remańe Polak, wól sie woka “Wened” u parla lęgwa łacina.

    Here is a 2008 (I think) article about Wenedyk written in Polish, with some explanations of how the language works. In particular, it compares Polish:

    Ojcze nasz, któryś jest w niebie, święć się imię twoje

    with Wenedyk:

    Potrze nostry, kwały jesz en czałór, sąciewkaty si twej numię

    and that in turn with:

    Pater noster, qui es in caelis, sanctificetur nomen tuum.

    So I was reading the Google Translate version of the article, which rendered the first sentence with its more-or-less-KJV equivalent:

    Our Father who art in heaven, hallowed be thy name

    and left the third sentence appropriately alone, but the second became:

    I need nostrils, you eat knots, you are very curious about your numias.

  109. So? The first and third are perfectly natural, and who hasn’t had knots in their stomach?

Speak Your Mind

*