Hand/Arm across Europe.

I can’t believe I’ve never linked to mapologies before, since they specialize in maps showing the words for things in different languages (the latest is “The names of Donald Duck’s nephews”), but better late than never; what drove me to post was Hand and arm in several languages (“An etymology map with a handful of words”), which shows words for HAND in bold black type and words for ARM beneath them in gray, with language families divided by color. In the box on the upper right explaining the colors, they give proto-forms for each, e.g. “Proto-Nakh *ko.” Very cool. And the only comment so far is “In dagestan in avar language Kwer and Rougk,” which is downright helpful.

Comments

  1. I think i’ve written before about the sites arbitrariness & inaccuracies when it comes to language boundaries equating to political boundaries. 🙁

    Some information is downright wrong, eg. the map about toasting, about the colour orange and the one about capital cities. Some is just dubious eg. the map about eight & night. Some is fun, like the districts of Prague map.

    Slightly related to the topic: the Croatian word for shoulder, rame, is cognate to the English word arm.

  2. David Eddyshaw says

    Nice map. Bears out what Plastic was saying about lámh in Irish, too (not that I ever doubted him.)
    Scots Gaelic seems to be different, which is interesting, if correct.

    Breton dorn must be the same etymon as Welsh dwrn “fist.”

  3. David Marjanović says

    The Bavarian “arm” word is a plain lie: the word, and the concept, are absent. “Leg” is absent, too, unless you resort to “thigh of pork” as a dysphemism.

  4. Really nice maps. Of course, there are quibbles. In this particular map, it’s not clear which languages use the same word for ‘hand’ and ‘arm’. The Hebrew transliteration is inaccurate in some cases (and enough with all the apostrophes). Tapuz ‘orange’ comes from tapuakh/> ‘apple’, plus zahav, not paz (either meaning ‘gold’). The Hebrew name for Jerusalem is set from left to right, and its etymology is folk-. Other words are set backwards too. The Arabic word for ‘red’ does not come from a root meaning ‘donkey’. Many Hebrew words are marked as ‘unknown etymology’, where their etmologies are well-known.

    The Spanish vocabulary maps are not new, but fascinating. The word palta ‘avocado’, from Quechua, is understandably prevalent in Andean and neighboring countries, but why also in Guatemala (along with aguacate)?

    the Armenian word for ‘blue’ comes from an Iranian word means ‘pigeon’. Is Russian golub’ an Iranian calque?

  5. January First-of-May says

    Is Russian golub’ an Iranian calque?

    IIRC the bird name (with its Proto-Slavic nasal vowel) had been connected to Latin columba, though (again IIRC) the phonetics imply a borrowing rather than a cognate. I suspect that the color was named for the bird rather than vice versa.

  6. The Bavarian “arm” word is a plain lie: the word, and the concept, are absent.

    Everyone knows Bavarians don’t have arms.

  7. They’re just not allowed to bare them. (old joke, I know)

  8. In Sanskrit there are eight words for ‘arm’.

  9. A. Sasportas says

    For the Indo-European languages, much of the material given on the maps is also available, though in a different form, in Carl Darling Buck’s A Dictionary of Selected Synonyms in the Principal Indo-European Languages: A Contribution to the History of Ideas (1949).

    For example, ‘arm’ is treated in section 4.31 and ‘hand’, in 4:33, with cross-references in both directions.

  10. “In this particular map, it’s not clear which languages use the same word for ‘hand’ and ‘arm’. ”

    Yes. How people distinguish between these two is an interesting (more interesting, actually) isogloss per se, and I expected to see it on the map. But it is not easy to formulate questions about semantic fields. Same with vagina and vulva etc: in WOLD* Kossmann gave two Riffian word for the former and then words glossed as “clitoris” and “lips” for the latter. These organs can be divided in constituent parts in many ways and the respective words can be distrubuted across contexts differently – exactly this makes the question so interesting.

    In Russian you take an apple with your hand/arm, and you hug a person with your hands/arms.
    But there is that weird plecho “shoulder” that was incredebly confusing for me as a child.

    This time a person has wide shoulders.
    That time it is that horizontal surface where your parrot sits.
    And then in school anatomy it is that part of your hand/arm between that horizontal surface and your elbow.
    And then in physics it is arm.

    I was utterly confused.


    *https://wold.clld.org/semanticfield/4, WOLD, the body.

    P.S. and no, for a Russian mind “hand” and “arm” are the same thing. It is not polysemy:)

  11. Just like to English speakers the front and back of the part below the head are both ‘neck’.
    I haven’t been able to find a language which has distinct word for the opening of the mouth (as in ‘open mouth’) and the oral cavity (as in ‘full mouth’). Surely some exist though.

  12. I remembered those people (Indochina – PNG – Australia I think?) with base-15 counting, who would count with gestures, using their fingers for 1-10 and then touching nose, forehead, shoulder… for 11-15.

    (I think Russian kids would make some jokes about base-21 counting here. “21st finger” is an euphemism I heard only in school)

  13. David Eddyshaw says

    We had some discussion earlier about the range of meaning of words for arm/hand:

    http://languagehat.com/trasks-historical-linguistics-online/#comment-3933562

  14. The map implies that in Czech hand and arm are the same (ruka), which is true colloquially, but in the literary language (or if disambiguation is needed) the word paže (whose etymology seems to be pretty much unclear) can be used for arm.

  15. PlasticPaddy says

    Re pazhe, for “paw” resp. “hand” there is French patte and German Patsche, both of which have no clear etymology and are possibly onomatopoetic (e.g., if the original meaning of Patsche is a slap, compare Livornese dialect patta).

  16. My Czech etymological dictionary (by J. Rejzek, 2001) says paže is possibly from Proto-Slavic *paz-ja, which may be alteration of *paz-ucha, itself unclear. Then it mentions Slovene pazduha (armpit) and speculates that the latter part *-duch may be continuation of PIE *dous (unknown to Wiktionary) and cognate to Old Irish doë [sic]. Polish pacha (armpit) and Avestan zasta (hand) are mentioned as possible cognates of unclear precise relation.

    No mention of patte or Patsche, although these formations are known to the author, who lists them as possible analogues of Czech pac(ka) (paw), which is indeed supposed to be onomatopoeic.

  17. Athel Cornish-Bowden says

    I’ve always said “computador”, but I’ve met “computadora” and “ordenador”. However, I hadn’t realized that “computador” was specific to Chile and Colombia. People in Spain understand what I mean when I say “computador”, and I understand them when they say “ordenador” (which is almost the French word). Mostly the maps agree with what I expect in Chile, but my wife says “piyama”, not “pijama”. I had thought that “palta” was more specifically Andean than the map says.

    The case of “night”/”eight” is very striking.

  18. Re night/eight: Is there something more to this than just a random PIE similarity?

  19. No, but it’s a cute similarity.

  20. David Marjanović says

    German Patsche

    Never encountered that, but read Patschhand once. Synchronically transparently onomatopoetic, in that *patsch* is the sound effect of various slaps and smacks.

    But French patte “paw”, and German Pfote “paw”, and Middle Irish patu ~ pata “hare”, and the common West Germanic path and pad words form an intriguing Gordian knot that includes but is not limited to what seems to be a Kluge mess.

    a random PIE similarity

    And not even that similar, because “8” has *ḱ, but “night” has *k.

  21. Athel: I’ve heard Chilean “la compu”.

  22. David Eddyshaw says

    The night/eight one works a bit better with Middle Welsh oeth “eight” (a variant of wyth) and henoeth “tonight” (where the he- element is the same as in heddiw “today.”)

    GPC says that the Welsh nos “night” itself is from a derivative in *-stu, which looks a bit ad hoc.

  23. re. Patsche, the 9th edition of Hermann Paul’s Deutsches Wörterbuch lists the following meanings:

    1 ‘klatschender Schlag o. Sturz’; 2 (16. Jh.) ‘flaches Werkzeug zum Schlagen’, vgl. Fliegenp.; 3 ‘die klatschende Hand’; 4 (Anf. 18. Jh.) ‘Straßenschmutz, -kot’, hierzu Anf. 18. Jh. in der P. sein, sitzen, stecken ‘in Not, Schwierigkeiten, Verlegenheit sein’

    Apart from the compound Fliegenpatsche, the saying cited last (in der Patsche stecken) seems to be the only really current usage, at least the only usage of Patsche I have ever heard or read.

  24. Angus Macdonald says

    @David Eddyshaw says: “Scots Gaelic seems to be different, which is interesting, if correct.” …

    …which it is not! For two reasons: firstly, it is misspelled and uses the wrong accent (acute). The accent used in Scottish Gaelic, according to the current standard orthography, is the grave accent. The correct word and correct spelling is làmh. The map has the correct word and correct spelling for the word for the arm (gàirdean). Cas, in Scottish Gaelic, however, is the word for both leg and foot.

  25. David Eddyshaw says

    I’m prepared to overlook spelling mistakes; it seems the map was at least right about Scots Gaelic not conflating “arm” and “hand”, which seems to be different from Irish. The Irish situation could easily be conservative, given that the Brythonic “arm” words are borrowed from Latin. (Come to that, so are the words for “leg”, as opposed to “foot.”)

    I suspect that the map may not take full account of the fact that there may be perfectly distinct simple words for “arm” and “hand” in some languages, but that making the distinction is not forced on the speaker the way it is in English, so that, for example, you could perfectly well say “I picked up the dropped coin with my arm/hand”, but would still be able to specify “please take the blood from my arm/not-hand.” Kusaal (for one language) works like that: nu’ug “arm/hand”, kukpauŋ “arm (not hand.)”

    Moreover, this may well not be an all-or-nothing thing: languages may differ in the specific linguistic contexts in which it is normal to distinguish “arm” and “hand”, perhaps (for example) by preserving older set expressions in which the “hand” word also means “arm”, but forcing greater specification in everyday speech.

  26. Jen in Edinburgh says

    I wouldn’t have hauled gairdean out of my memory, but lamh as far as I’m aware is restricted to hands.

    Breton ‘dorn’ is interesting – that’s Gaelic for fist.

    I’d have expected ‘airm’ in Scots, although possibly pronounced erm, which looks silly written. Interesting echo in Frisian.

  27. Stu Clayton says

    (in der Patsche stecken) seems to be the only really current usage,

    Along with its counterpart aus der Patsche helfen.

  28. The Croatian for armpit is pazuho (neuter noun). This is more often heard in the genitive “ispod pazuha” (“in the armpit” literally, under the armpit)fromwhich is derived the masculine noun “pazuh”.

    According to the Croatian encyclopedic dictionary, the etymology is preslavic pazuxo.

    Croatian also uses the same word noga for English leg and foot. eg. Udari nogom = Hit/kick with the foot. Croatian does have a word for foot though.

  29. January First-of-May says

    Croatian also uses the same word noga for English leg and foot. eg. Udari nogom = Hit/kick with the foot. Croatian does have a word for foot though.

    Russian doesn’t quite have separate leg/foot words either; when translated child-oriented texts need to put something on the “foot” label, they tend to use ступня, which I would translate as “sole”.
    The usual specific term for “hand” appears to be кисть руки, lit. “brush of the arm” (I’m not actually sure which meaning is secondary here). I think I’ve also seen ладонь “palm” used in contexts where кисть would sound too scientific, but I’m not very confident.

  30. Y: In Sanskrit there are eight words for ‘arm’.

    Many-armed Nataraja and Kali….

  31. Finländare says

    Re night/eight: Is there something more to this than just a random PIE similarity?

    Obviously the Noble Ancient Indo-Europeans conceived of the night as being eight hours long, hence why they were always so well-rested and therefore able to conquer so many lands.

    Joking aside, I have no idea why they have included “handleggur” and “armur” under Finnish. The usual word for arm is käsivarsi, while olkavarsi is only the brachium.

  32. David Eddyshaw says

    It must be all those Icelanders who’ve moved to Helsinki.

  33. Student assistant: “But … isn’t Iceland a synonym for Finland?”

  34. It must be all those Icelanders who’ve moved to Helsinki.
    Not implausible.
    During Euro 2016 1/5 of them were in France. Two years later 1/10 of them were in Moscow.

    P.S. eightland, nightland.

  35. John Emerson says

    Iceland plays a remarkably large role in Western culture considering that it’s only been populated for a bit more than 1000 years and has about the same population as the 150th largest metropolitan statistical area in the US, Fort Collins, CO.

  36. David Marjanović says

    “Western” indeed. The reason I can type ý on my keyboard by just hitting the accent key (next to Backspace) and then the Y key must be that Iceland has been a NATO member since NATO was founded in 1949, which in turn must be due to its important strategic position in the geography of WWII and the early Cold War.

    Þ & ð are not on the German keyboard layout, but they are in the ISO-8859-1 set, no doubt for the same reason.

    For comparison, I cannot write ć the same way; I need to copy & paste it from the character map. It’s in ISO-8859-2, the post-Cold-War one. And the interesting thing is that just typing it as accent + C does work on the Czech keyboard layout even though Czech and Slovak lack the letter or a particularly pressing need to write Polish names.

  37. Iceland received a special deal when NATO was established. The country does not have the same kinds of defense spending targets as most other members, because its strategic location was considered too important to forgo. It was also considered important to offer membership to as many democratic countries in the North Atlantic region as possible, even ones that were not expected to make meaningful military commitments (like Luxembourg, which also is not asked to maintain a military proportional to its population).

  38. Trond Engen says

    Davis M.: For comparison, I cannot write ć the same way;

    That one annoys me out of my mind. There’s no reason why accent + any letter shouldn’t work.

  39. My tablet lets me type in accented letters by “attaching” the accented letter to the base letter.

    By this i mean, when you press down the Y button, another button for Ý pops up.
    Ditto for þ (located on the t button) and đ (on the d button).

    The mobile phone works the same way. Funny how desktop keyboards dont support that.

  40. John Cowan says

    common West Germanic path and pad words

    Thought to be borrowed from Iranian, which has both unshifted /p/ and /θ/ < /t/ or /t

    Synchronically transparently onomatopoetic, in that *patsch* is the sound effect of various slaps and smacks.

    A new toy store opens for its first day, and hundreds of parents and children who have been waiting on line for hours rush in. Within five minutes every toy in the store large enough not to be in a package is being played with, and all the kids are screaming with delight despite their parents’ attempts to calm them down. The store manager doesn’t know what to do: his stock is rapidly being rendered unsaleable.

    At this point, a mild-looking fellow approaches him. “I happen to be a child psychologist, and I think I can help you … for a fee, of course.” The manager instantly agrees. So the psychologist picks one child, approaches him, and whispers something in his ear. It’s magic: the kid instantly calms down, lets go of the toy, and quietly goes off to stand beside his parents. The next kid also gets a message in her ear and reacts the same way. And the next, and the next, until the room is completely quiet.

    “My God,” says the manager. “It’s a miracle.” He pays the psychologist twice the agreed fee out of his own pocket. “But what did you say?”

    “Simple,” says the psychologist, smiling. “I said, ‘If you don’t stop that noise at once and go over and stand quietly by your tateh-mama, I’ll give you such a potch on the tukhus you won’t sit down for a week.'”

  41. John Cowan says

    When the 8859-x series was first created, there were four encodings: -1 for the West (including Ireland, which is not a NATO member to this day), -2 for the East, -3 for the South (Turkish and Maltese) and -4 for the North (Lithuanian, Latvian, Estonian, Greenlandic, and Latin-script Sami). But it turned out that Turks didn’t particularly want or need to exchange text with Maltese-speakers, even with Esperantists thrown in, so the Turkish goverment created a dialect of 8859-1 with the rarely-used Icelandic characters replaced by Turkish-specific ones, and ISO blessed this as 8859-9. (-5, -6, -7, and -8 were already in use for ASCII + Cyrillic, Greek, Arabic, and Hebrew scripts).

    Later, -10 was designed to handle all the languages of the Nordic countries, but again it was more important to interchange with the Southlands that used -1 than with people using computers in Sami or Greenlandic. -13 was meant for the Baltic languages, -14 for the Celtic languages, -15 as an upgrade to -1 that squeezed out some less-used symbols in favor of Š/š and Ž/ž for Finnish loanwords and Œ/œ and Ÿ for French (which had been squeezed out for -1), and finally -16 for the European Southeast, especially Romanian. (-11 encoded ASCII + Thai, and -12 was abandoned before being finalized.) Finally all this was ditched in favor of UTF-8, with the sole exception of 8859-1, which is barely surviving.

  42. Ý may remind of Icelandic, but one shouldn’t forget the vile abomination of the current Turkmen orthography where ý stands for /j/, while y stands for /ɯ/. As good as one could expect from a mad post-Soviet dictator, though.

    And the interesting thing is that just typing it as accent + C does work on the Czech keyboard layout even though Czech and Slovak lack the letter or a particularly pressing need to write Polish names.

    The need to write Polish names was not pressing enough to have a way to type ł, ż, ą or ę. More likely, the Czech keyboard layout designers thought that when the acute accent key is already there, then they could as well make it work in every possible combination. (That is, combinations they were aware of. There is apparently a ǵ in Unicode which cannot be written on the Czech keyboard. I have no idea what language uses that – Latin transcription of Macedonian perhaps?)

    Also, is there a w with acute accent in Welsh? It may be super rare, but I don’t see a reason why it should not exists. Yet I cannot find it in the character map.

  43. David Eddyshaw says

    Also, is there a w with acute accent in Welsh?

    No. Welsh limits its accents to the humble circumflex.

  44. As good as one could expect from a mad post-Soviet dictator, though.

    The previous version with Y y, ¥, ÿ was by the Academy of Sciences. He closed the Academy in 1998, but I can’t say that the modification from 1999* is too mad:)

    $ and ¢ for upper and lower case [ʃ] are cute.


    *WIkipedia

    Also:
    https://www.academia.edu/6385563/Emblems_of_independence_script_choice_in_post_Soviet_Turkmenistan

  45. Welsh limits its accents to the humble circumflex.

    I’ve looked up the Wikipedia article on Welsh orthography and it mentions words mẁg and gẃraidd. So is this something invented by crazy Wikipedians? Note that both words have their Wiktionary entries with accents as well, but it proves nothing. The entries may have been written by the same Wikipedians, after all.

    The previous version with Y y, ¥, ÿ was by the Academy of Sciences.

    Ah, I see. I will have to adjust my ideas about what is normal in Turkmenistan. Türkmenbaşy was perhaps a pretty unremarkable member of his nation.

    $ and ¢ for upper and lower case [ʃ] are cute.

    £ and ſ for [ʒ] is even more funky. If it wasn’t an Islamic country I would suppose they were drunk at the Academy. (The idea of having exchange rates between consonants is sort of cool however.)

    But still, was there a particular reason why they did not just adopt Turkish orthography?

  46. Currency signs obviously came from this set.

    The question is why not TWO alphabets (unless there was a second one which Wiki is unaware about): one for typographies and those computers that would have a new code page (which they would design and implement and distribute) and one with currency signs as a temporary substitute.

  47. But ſ is not part of it, or did they use the upper half of the integral sign for that?

  48. David Eddyshaw says

    @prase:

    Yes, you’re right. I don’t think graves and acutes appear very much in real life, but they aren’t confined to crazy Wipikedians, e.g.

    https://www.decymru-tan.gov.uk/gwacau/

    One hopes that the fire service is not a haunt of crazy Wikipedians, anyhow.

  49. Also I still don’t uderstand their current script. What kind of thinking leads to choosing a cedilla to post-alveolarise s and a háček to do the same job for z? Well, Wikipedia says that Turkmen z is not /z/ but /ð/, but still, why two different accent marks? In the same time they have no problem using háček over ň for a different kind of articulation shift. And why ç and no c?

  50. @David Eddyshaw:

    It may also be a Dravidian plot to discredit the language. I mean, would you trust a fireman with a grave accent?

  51. Türkmenbaşy

    Not to be outdone by Elbasy, I guess.

  52. Speaking of hands, it seems that beforehand is confined to English, Dutch, the Nordics, Spanish, and Portuguese:

    Etymology
    From Middle English biforhand, biforhond, beforehonde, bifornhand, equivalent to before +‎ hand. [13th century. After Old French avant main]

    Danish: på forhånd
    Dutch: op voorhand
    Finnish: etukäteen
    Norwegian Bokmål: på forhånd
    Portuguese: em antemão
    Spanish: de antemano
    Swedish: på förhand

    https://en.wiktionary.org/wiki/beforehand

  53. David Eddyshaw says

    @prase:

    I’ve just now realised that I have in fact been using circumflexes when I ought to have been using acutes (e.g. gwacâu instead of gwacáu. I am ashamed.

    I don’t recall seeing graves outside technical discussions of vowel length, e.g. càlon “heart”, where you’d have expected the vowel to be (half-)long, instead of short, as it is in Proper (i.e. Literary) Welsh. (The word is a bit of a mystery, I think: it had a double /l/ in Middle Welsh (not /ɬ/), hence the short vowel; but the doubling seems not to be original.)

  54. What kind of thinking leads to choosing a cedilla to post-alveolarise s and a háček to do the same job for z?

    I do not know what it looked like from the Caspian sea.

    As a Russian I immediately recognize and process ž and š. These are… native for me, somehow. I do not know why: I know them since early age, but I did not meet them too often in that early age.

    Meanwhile scholarly Arabic transcriptions (apart of š) keep making me stumble and swear very, very badly all the time – and I read them a lot.

    ş in turn is used in all most* orthographies for Oghuz languages.


    *I meant ‘most’. But I never realized that almost is all most.

  55. Arabic transcriptions

    ǧ is maybe the most annoying letter for me. Ǧermany. Brr.

  56. January First-of-May says

    There is apparently a ǵ in Unicode which cannot be written on the Czech keyboard. I have no idea what language uses that – Latin transcription of Macedonian perhaps?

    Proto-Indo-European.

    Less jokingly, Wikipedia mentions Macedonian, Pashto, Karakalpak, a few even more exotic contexts, and “Cantonese Yale” [sic].
    The relevant romanization of Karakalpak is fairly recent, and Pashto isn’t exactly a commonly transcribed language; comparing with the list for ḱ, which is also used in both PIE and Macedonian, I suspect the culprit was in fact the Yale romanization of Cantonese, where (apparently) syllabic ng with a rising tone ends up written .

     
    EDIT: turns out you were probably right after all; the letter next to it in Unicode is the Macedonian digraph dz, which suggests that ǵ is also Macedonian.
    I was (yet) unable to find any reason why ǵ and ḱ ended up in different ranges, however (Latin Extended-B and Latin Extended Additional respectively). It looks like both were added in Unicode 1.1.

    For that matter, Yale Cantonese would (probably) have included g̀ (with a grave accent) as well as ǵ, and the former is (as far as I’m aware) not in Unicode to this day.

     
    EDIT 2:

    As a Russian I immediately recognize and process ž and š. These are… native for me, somehow. I do not know why

    Seconded (particularly for š), and I don’t actually recall where I found them originally either.

  57. I’ve just now realised that I have in fact been using circumflexes when I ought to have been using acutes (e.g. gwacâu instead of gwacáu.

    I’m starting to have serious doubts about your Welshitude.

  58. If it wasn’t an Islamic country I would suppose they were drunk at the Academy.
    My experience in the region is that they’re more post-Soviet than Islamic. Most people drink, all though I know some who drop it during Ramadan. So inebriatedness cannot be excluded.
    But still, was there a particular reason why they did not just adopt Turkish orthography?
    Because they’re not Turks, and the world needs to know?

  59. On Welsh graves: I have come across some examples of what Wikipedia describes “to mark vowels that are short when a long vowel would normally be expected”. (Thus acting as the opposite of the much more common circumflex which indicates long vowels where the spelling rules would otherwise indicate a short vowel; the acute is mainly to show unexpected stresses on the final syllable. Both vowel length and stress are normally predictable enough not to need explicit markers. )
    I see even the Geiriadur Prifysgol Cymru is happy with mẁg (“mug”, as opposed to mwg “smoke” with long vowel), and also with clòs (“close”) which is what I first thought of as an example. It does tend to be more common in borrowings from English which some people might disapprove of in principle anyway…

    (I’m writing this with the UK extended keyboard which, as one might have hoped, handles Welsh fine).

  60. My experience in the region is that they’re more post-Soviet than Islamic.
    True.
    What is also true is that they make wine in Shiraz and drink wine in Shiraz.

    For the wine once produced in the Iranian city of Shiraz, see Shiraz wine. (WIkipedia)
    Bullshit. About “once produced”. Because they make it and drink it:)

  61. Both vowel length and stress are normally predictable enough not to need explicit markers.

    My impression is that “predictable” in case of Welsh vowel length means “exhibiting a set of rules designed to get rid of the need for accents, convoluted enough to be hard to learn, yet still not reliable”. But I guess it still counts as “predictable” when compared to Saesneg.

    I wonder whether vowel shortness is the reason for doubling of r or n in words like ennill (but enillaf with a single n because the syllable is not stressed and cannot get long?). Obviously the same trick cannot be done with l or d since the digraphs are already taken. Why there are apparently no double ms is a mystery for me.

    Edit: I realise now that this is already the second time recently that I am drawing the discussion towards peculiarities of Welsh vowel orthography. Not sure what to think about that.

  62. David Eddyshaw says

    No, the consonant length is real, but lost word-finally, so that the vowel length, which would otherwise have been (mostly) predictable, becomes contrastive. In Middle Welsh, words like pen “head”, which has a short vowel, still appear as penn etc, whereas e.g. dyn “man, person”, which has a long vowel, is written dyn, as in Modern Welsh.

    Double consonants also become single before the stress, but that’s an actual real change, not a way of indicating vowel length. These sort of changes had still not happened at the stage vowel length was determined, and don’t figure in Middle Welsh orthography (which is, however, not very consistent in writing double consonants in general.)

    M is always double (subject to the simplifications I’ve just mentioned); single m became f /v/, via earlier /ṽ/ (still written m in Old Welsh orthography.) The voiceless stops and /s/ are also always geminate word-internally between vowels; the original single consonants have become b d g h. This is the word-internal equivalent of the famous Soft Mutation. The consonants p t c s are still (redundantly) routinely written double between vowels in the 1588 Bible.

    The rule, which goes all the way back to Brythonic following the shift of original vowel length into quality differences, was in principle pretty simple: vowels in open syllables were long, others short. The rule either precedes the loss of final syllables, or applied in such a way that single word-final consonants were extrasyllabic. Subsequent developments like the loss of word-final gemination and vowel fusion after loss of intervocalic */ɣ/ and */h/ have created exceptions, but simple rules still account for the great majority of cases*. The rules differ somewhat between the modern dialects when it comes to vowels before non-geminate consonant clusters, so that gwallt “head hair”, for example, has a long vowel in the north, but short in the south.

  63. David Eddyshaw says

    Inaccuracy (well, OK, error) above: the change *s -> h is not parallel to the Soft Mutation: it affected all single /s/, including word-initially, as in hen “old”, for example. But it’s similar to the extent that remaining s always represents original *ss (or *st.)

    I omitted to mention that the old Bible orthography also always writes mm for m between vowels (except in foreign names), again redundantly.

    I am also deeply gratified to notice that the 1588 Bible writes -(h)âu for the verbal nouns of verbs derived with -ha- (like gwaca-, which is from *gwag-ha-.) Vindicated! My spelling has the highest possible corroboration. I may be dysorthographic, but at least I’m saintly.

    [Apologies for the orphaned asterisk in the previous post; I was going to add a note about the long /e/ in brenin “king” but ran out of time, and have subsequently realised that I don’t know why it’s brenin and not *brennin.]

  64. vowels in open syllables were long, others short

    The reading rules are not difficult in Swedish as far as the spellings usually correspond to the actual pronunciation. The basic principle is based on the syllabic balance. The syllabic balance is a common Scandinavian feature which stands for obligatory lengthening of stressed syllables. It means that a long vowel can only occur before a single short consonant, as well as a short vowel – only before a cluster or a single doubled consonant. Doubled consonants are said twice as long as their single equivalents. Thus, a vowel is long if:

    it is the last sound in the syllable (veta /`ve:ta/, läsa /`lɛ:sa/);
    it occurs before one consonant (tid /’ti:d/, ord /’u:ɖ/).
    A vowel is short:

    before a consonant cluster in which the first consonant is said a bit lengthened (eld /’eld/, aska /`aska/);
    before a doubled long consonant (Anna /`an:a/, rinna /`rin:a/).

    https://www.conservapedia.com/Swedish_language

    Sieves and Herrings: For Distinctive Vowel Length in Swedish
    https://ling.auf.net/lingbuzz/003136/current.pdf

  65. David Eddyshaw says

    Yet a further correction: my copy of the 1588 BIble actually writes intervocalic s single, not double (I was thinking of MIddle Welsh); it’s true that it’s always derived from a cluster in words other than loans, though.

    The 1588 Bible does write all intervocalic p t c m double, however. What I tell you three times is true.

  66. David Marjanović says

    The syllabic balance is a common Scandinavian feature

    It was almost a common Germanic feature at some point around the end of the Middle Ages, but before it had finished spreading, English had lost consonant length, and so has most of Continental West Germanic…

    I’ll stop here for now 🙂

  67. J.W. Brewer says

    I just saw someone on Facebook post the comparable map for the names of Donald Duck’s nephews, which seems to be attracted general interest (probably more so than this anatomy one would) despite the failure of the Swadesh List to include them. (The longer 207-word edition of the Swadesh List has some kinship terms, but “nephew” doesn’t make the cut.)

  68. David Eddyshaw says

    @juha:

    Something like that could be at the back of brenin “king.”. In Modern Welsh, it makes the plural brenhinoedd, has the derived adjective brenhinol and so forth; the loss of the h after the stress is regular, so no problem there. Middle Welsh has brenhin.

    However, what is peculiar is that the -nh- derives from *nt, so that the outcome should be -nnh, which would be expected to become nh before the stress in Modern Welsh, but nn after the stress, as in cant “hundred”, plural cannoedd. The first vowel of brenin is actually the result of contraction after loss of Old Welsh /ɣ/; the word comes from *brigantīnos; the form breenhin actually still occurs in early Middle Welsh. So it would be neat if the -nn- had got degeminated because the preceding vowel was long as the result of contraction.

    I’m not altogether sold on this idea, though. The whole system of consonant alternation in these cases looks like it’s been subject to a lot of levelling and analogy, e,g cenhedloedd, plural of cenedl “nation”, which is from *kenetl-, with nary a sniff of a medial -nt-.

  69. David Marjanović says

    Thought to be borrowed from Iranian

    That runs into timing issues, though. The kind of Iranian that would fit would be Old Iranian, so you’d seem to need Old Iranian to be in contact with Pre-Germanic – so either Grimm’s law happened oddly early and the geography gets a bit iffy, too, or you need to assume an oddly conservative Iranian variety. Conversely, for an Iranian word that shows up in West but not North Germanic, you’d naturally blame the Alans, but it’s not likely they had kept [θ] the whole time (and then lost it right after). Here’s a brief inquiry into the issues on the Germanic and the Iranian side, plus into the question if Finnic could have mediated (verdict: probably not), and in parentheses an intriguing proposal for a Celtic etymology is cited. (Hint: Celtic *gʷ > *b before the loss of aspiration, so why not before Grimm’s law, too. Admittedly, we’d then have to assume loss in North Germanic.)

    The /θ/ ~ /t:/ ~ /d:/ variation, if indeed these words all belong together, looks like a Kluge mess. That would need some time to form, so the Alans probably came too late.

    Oh, BTW, if you’re not confused enough yet, how about some pot? A very short and smokey discussion of that word, in its original meaning, starts here.

  70. John Cowan says

    Latin Extended-B and Latin Extended Additional respectively

    Latin Extended-B letters come from various non-European Latin character sets plus further IPA characters. Latin Extended Additional letters are base+diacritic(s) characters plus mediaeval manuscript letters. So g with acute was a member of some pre-existing character set, whereas k with acute wasn’t, even though they are both used in Macedonian transliteration.

  71. January First-of-May says

    So g with acute was a member of some pre-existing character set, whereas k with acute wasn’t

    Indeed, and I’m still not sure why. In particular, I’m not sure if Pashto, Yale Cantonese, or (somehow) Macedonian all along is more plausible, or for that matter whether it was in fact something else entirely.

  72. Mostly the maps agree with what I expect in Chile, but my wife says “piyama”, not “pijama”.

    Interesting. I find their version of Argentinean Spanish often suspect.

    One example: they list sartén as indistinctly masculine or feminine, but my intuition is that the masculine lingers only marginally. In CREA, only 1 in 60 examples used it, and while the ratio is a bit higher in Corpus del Español, it was still 1 to 20.

    In the opposite direction, they list only pollera for ‘skirt’. Falda is admittedly less common, but the ratio is roughly 3:5, much closer than in the above case.

    Cara o ceca I haven’t heard from anyone under my grandparents’ age; it’s been cara o cruz since my childhood in the remote 70s.

    I suspect this is one of those cases where national boundaries are quite different from isoglosses, though.

  73. Iceland plays a remarkably large role in Western culture

    Out of the Nordics, it is Sweden that has a reputation as a musical superpower for some reasons (as in, indeed some of them); but again per capita, I once calculated that already a handful of notable acts (Björk, Sigur Rós, múm, Yagya, Hilmar Örn Hilmarsson, Ólafur Arnalds, etc.) suffice to give Iceland the status of having produced easily the largest proportion of critically acclaimed musical artists over the last 3–4 decades, out of any country or even subnational division in the world.

Speak Your Mind

*