Dung Times

« previous post | next post »

There's a roundly execrated publication of the CCP called Global Times in English.  The Chinese name is Huánqiú shíbào 环球时报.  Associated with the People's Daily, it is infamous for its extreme, provocative, anti-Indian, anti-Japanese, anti-Western (especially anti-American) editorials and articles.

Now it seems that some Indian Tweeps are referring to the Global Times as "Gobar Times", using Hindi  gobar गोबर ("cow-dung") to mimic the sound and the sentiment the name evokes. A tweet by Donald Clarke calls our attention to this fecal phenomenon.

Here it is in use.

Notes from Jichang Lulu:

The word seems to come from Sanskrit गोर्वर gorvara / गोवर govara, referring to the pulverised kind.

The more common Sanskrit word for cow-dung seems to be गोमय gomaya (e.g., it's the first one given for 'cow-dung' in Āpte and Monier-Williams' English-Sanskrit dictionaries).

My tweets on this important matter.

Gobar cakes, called गोबर का उपला gobar ka upla (or "ooplah", found in Hobson-Jobson) is used as fuel. (Hindi upla is seemingly from Sanskrit utpala.) Here is some for sale.

Dung of various animals is certainly used as fuel all around the world (in Mongolian it's called argal).

"Gobar Times" can be confusing because there actually is a publication unironically called that, but genuine references to the GT can be found.

This is entirely harmonious with a popular Chinese name for the rag: Huánqiú shǐ bào 环球屎报 ("Global Excrement News").

It's delightful that Chinese and Indian netizens independently came up with faecal names for the same publication. They might dislike the Globule for different reasons (it's a highly dislikable thing), but somehow add to a history of Sino-Indian conceptual contact and resonance that is almost two millennia older than the Party-state behind the rag.

I (VHM) certainly recall gobar from the days when I was a Peace Corps volunteer in Nepal (1965-67).  Not only was it used for fuel and fertilizer, it was also used as a waterproof coating for walls.  Above all (perhaps I should say below all), it was used to plaster the floors of houses.  Believe it or not, after my landlady carefully spread gobar on the floors of my rooms, they smelled fresher and felt cooler than before.

Another vivid memory was sleeping in the world's highest roofed structure, a yak shed near Everest base camp, with a Peace Corps buddy (Brian Cooke) and two Sherpas.  We made a fire with yak dung cakes to boil water for our meal.

Remember "Pig Sanskrit"?



38 Comments

  1. Ambarish Sridharanarayanan said,

    March 14, 2018 @ 11:50 am

    Sanskrit 'go' seems to be cognate with the English 'cow'.

    https://en.wiktionary.org/wiki/Reconstruction:Proto-Indo-European/g%CA%B7%E1%B9%93ws

  2. Ursa Major said,

    March 14, 2018 @ 12:34 pm

    I'm struck by the apparent similarity of shǐ 屎 and the English word meaning the same thing, which I assume is coincidence (Wiktionary gives a history of the Chinese so it's not a loan). Is there a real similarity, and do people recognise and use it?

  3. KevinM said,

    March 14, 2018 @ 2:15 pm

    "using Hindi gobar गोबर ("cow-dung") to mimic the sound …"
    Answering the (Monty Python, of course) question, "What's brown and sounds like a bell?"

  4. Jichang Lulu said,

    March 14, 2018 @ 4:05 pm

    As usual when dealing with political punning on the Innernets, China Digital Times (Zhōngguó shùzì shídài 中国数字时代) is a great source on the Globule / Dung Times. (It has been covered on Language Log, most recently here.)

    The Chinese version of CDT has three pages worth of pages tagged with the rag's faecal name cited in the OP (Huánqiú shǐ bào 环球屎报). E.g., this post by Sandra Severdia tells of the Gobar's beloved editor Hu Xijin 胡锡进 being himself a victim of censorship.

    The CDT's English-language "Grass-Mud Horse Lexicon" records further variations: Muddled Shit Times (Hùnqiú shǐ bào 混球屎报), Horrid Ball Shit Times (Huàiqiú shǐ bào 坏球屎报) and Horrid Servile Shit Times (Huàiqiú shǐ bào 坏俅屎报), the latter used in a Dung-honchal portrait by renowned artist Rebel Pepper (Biantai Lajiao 变态辣椒).

  5. Michael said,

    March 14, 2018 @ 7:24 pm

    Does English "hooplah" (hype/bullshit) come from upla?

  6. Rhona Fenwick said,

    March 14, 2018 @ 11:51 pm

    @Ambarish Sridharanarayanan:

    Sanskrit 'go' seems to be cognate with the English 'cow'.

    Indeed. And even beyond the rich IE heritage (virtually every Indo-European branch preserves cognates of PIE *gʷṓws), this is one of those terms that also has similar forms in other unrelated Eurasian languages, perhaps indicating wide cultural borrowing (if the terms aren't simply imitative, which is of course a possibility): Tajik also has гов "cow", Ubykh gʷəmá "cow", Kabardian гуу "bull", Egyptian gw "bull", and Sumerian gu₄ ~ gud "bull", perhaps best known through the mythological Great Bull of Heaven, gu₄.gal.an.na.

  7. Andreas Johansson said,

    March 15, 2018 @ 1:32 am

    @Rhona Fenwick

    Tajik is, of course, Indo-European (and relatively close to Sanskrit, both being in the Indo-Iranian branch), so isn't gov just another descendant of *gsup>wows?

  8. Andreas Johansson said,

    March 15, 2018 @ 1:33 am

    With apologies for messed up formating in the previous post …

  9. B.Ma said,

    March 15, 2018 @ 4:34 am

    @Ursa Major

    Apart from the coincidental resemblance between the Pinyin of 屎 and 'shit', they are not particularly similar. Shit is a noun, verb, adjective and interjection, and considered rude; 屎 is mainly a noun and not rude, though as a Japanese Kanji pronounced くそ I believe it can be used to mean "Damn!" In Cantonese I have seen it used to form the plural of fan (of a celebrity) as in "FAN屎", no other examples of this use come to mind at present.

  10. Ursa Major said,

    March 15, 2018 @ 7:03 am

    Thanks B.Ma, that's interesting. Multilingual or etymological puns and playing with words that have similar forms is something I like, in fact I derived this user name in that way from my real name via 2 or 3 languages.

  11. Victor Mair said,

    March 15, 2018 @ 9:36 am

    Thank you, Rhona Fenwick.

    Colleagues following this thread may be interested in the long discussion going on over at the "Of dogs and Old Sinitic reconstructions" post, where the o.p and commenters also look into connections within and beyond IE and before PIE.

    http://languagelog.ldc.upenn.edu/nll/?p=36996

  12. Jichang Lulu said,

    March 15, 2018 @ 10:47 am

    Further to the comparative spirit of the 'dog' and other recent posts, shǐ 屎 'excrement' has been seen as related to Tibetan lci[-ba] ལྕི་བ (< *hly-, cf. Jacques) as well as Burmese, Tangut and other items. This was mentioned on LL by Tsu-Lin Mei. I wonder what Chris Button and others think about the Sino-Tibetan reconstruction of this item, as seen from the point of view of different reconstructions and specifically Pulleyblank-Button's.

  13. Chau said,

    March 15, 2018 @ 2:26 pm

    @ Ambarish Sridharanarayanan. Corresponding to Sanskrit go 'cow' is gû 牛 'cow/bull/cattle' in vernacular Taiwanese. The literary Taiwanese reading of 牛 is giû which is used only in reading Literary Sinitic (文言文) and in reciting Tang poems. Literary Taiwanese is said to be close to the speech of the Tang court. Japanese imported giu 牛 and spells gyu.

    @ Ursa Major, regarding "the apparent similarity of shĭ and the English word meaning the same thing." Parallel to this is another apparent similarity between Taiwanese (vernacular) sái 屎 and German Scheiss (Scheiß) 'excrement'. Taiwanese lacks the sh sound. (It can be shown through a regular sound correspondence that Tw. sái 屎 is related to Common Germanic *skīt-/*skit- 'excrement, feces'.)

  14. Eidolon said,

    March 15, 2018 @ 5:20 pm

    Since the topic of Indo-European loans in Old Chinese seems to be a trend in recent days on Language Log, what do readers think of the recent thesis from Jan Zidek, titled "Tocharian Loanwords in Chinese"? The author seems to have ended up categorically refuting many of the proposed Tocharian or Indo-European loans into Chinese "to his own disappointment." Mentions are given to 牛, 犬, and 狗 from the other thread. For people less familiar with the field, the references for the histories of these proposals are fairly useful. Obviously, the debate has been going on for quite a while, and the end does not seem in sight.

  15. Chris Button said,

    March 15, 2018 @ 9:17 pm

    @ Jichang Lulu

    Further to the comparative spirit of the 'dog' and other recent posts, shǐ 屎 'excrement' has been seen as related to Tibetan lci[-ba] ལྕི་བ (< *hly-, cf. Jacques) as well as Burmese, Tangut and other items. This was mentioned on LL by Tsu-Lin Mei. I wonder what Chris Button and others think about the Sino-Tibetan reconstruction of this item, as seen from the point of view of different reconstructions and specifically Pulleyblank-Button's.

    Old Chinese 屎 *ʰlə̀jʔ is a perfect correlate down to the tone with Old Burmese *kʰlɨj2. The Old Burmese form just has a k- before the lateral.

    @ Eidolon, Rhona Fenwick, Chau and others…

    Old Chinese 牛 *ŋʷə̀ɣ (Baxter & Sagart *[ŋ]ʷə] reconstructed on the basis of its Shijing rhyme would actually have vacillated with *ŋə̀w (the velar onset maintaining the labial feature in *-wə̀ɣ for longer before a complete merger with the rhyme *-ə̀w). As such the similarity with PIE *gʷow- is striking, although technically *ŋʷə̀ɣ would be prior to *ŋə̀w such that there must have been some degree of analogical leveling in terms of the vacillation between the two forms. Disregarding a possible minor issue between OC ə/a versus PIE e/o (discussed over on the "dog" thread – although some words in OC do seem to have alternated between ə and a internally anyway), the only major difference of note is the OC nasal. However, it is notoriously difficult to retain the voicing of a velar obstruent (as indeed to a lesser degree with any obstruent) such that pre-nasalisation would have been very possible here such that *gʷ > *ᵑgʷ > *ŋʷ. (Incidentally, this is why I am very suspicious of theories regarding a nasal vocing prefix in Old Chinese since the nasalisation of voiced obstruents being cited as evidence in loanwords from OC seems far more likely to result from the basic use of nasalisation as an articulatory mechanism to retain voicing)

  16. Eidolon said,

    March 16, 2018 @ 12:44 am

    @ Chris Button

    How specific is the PIE phonetics? As Rhona Fenwick mentioned the word is widespread in Eurasia; perhaps most importantly, it is found in Sumerian, in the region of the oldest domesticated cattle. Presuming the word spread from Sumerian, doesn't it make the argument that it is "notoriously difficult to retain the voicing of a velar obstruent," moot? Most of the downstream languages seems to have retained a voiced velar obstruent without trouble. Or is this a property specific to OC? Are there any other downstream languages with a *gʷ > *ᵑgʷ > *ŋʷ transformation?

  17. Chris Button said,

    March 16, 2018 @ 9:32 am

    @ Eidolon

    Obstruents are by their nature voiceless such that any voicing that may occur is relatively hard to maintain (even in English, the so-called "voiced" obstruents are only reliably voiced intervocalically otherwise they tend not to be fully voiced or even voiced at all). Pre-nasalisation of voiced obstruents is often used as an articulatory mechanism to retain voicing since nasals are sonorants and by their nature voiced. There's a good paper on this by John Ohala (1983)

    Pre-nasalisation can often simply remain as surface phonetic phenomenon (e.g. varieties of Japanese where the phoneme /g/ may surface as [ᵑg] or [ŋ] in certain environments but still maps phonemically onto /g/). However it may also become phonemic (e.g. phonemic /g/ in the Kuki-Chin language Tedim has completely merged with phonemic /ŋ/ in the closely related language Sizang).

    When a word is loaned, even if the phoneme remains as /g/, a surface realisation of [ᵑg] or [ŋ] may be what is transferred since that is what is heard.

  18. Eidolon said,

    March 16, 2018 @ 1:13 pm

    @ Chris Button

    Thanks. Is there any evidence of an IE or PIE /g/ loan surfacing as [ŋ] in a neighboring language? Particularly Tibeto-Burman, since you brought up Kuki-Chin. I admit that I have substantial skepticism towards direct PIE loans in any Sinitic language. For one, it doesn't fit the age of expansion, since by 3,000 BC PIE had already disintegrated into daughter languages and technologies such as bronze, wheeled vehicles, or domesticated cattle did not reach China until well after that time. For another, it ignores the presence of intermediate languages between the PIE Urheimat – and the Urheimats of descendant languages like Indo-Aryan, Indo-Iranian, and Tocharian – and the Sinitic Urheimat in the North China Plains. For example, the early Bronze Age culture of Qijia in Gansu served as a corridor by which influences from the Afanasevo and Andronovo reached inner China, but this culture is not likely to have been Sinitic, and the same applies to the other steppe routes through Mongolia. The challenge of showing PIE or early IE loans in OC could very well be because they never reached OC directly from PIE or IE, but were transmitted by languages from other families through layers of borrowing.

  19. Chau said,

    March 16, 2018 @ 7:08 pm

    @ Eidolon

    I am very interested in your viewpoint expressed in the last comment. Have you published it anywhere?

  20. Chris Button said,

    March 16, 2018 @ 7:49 pm

    @ Eidolon

    All I can say for certain in that regard is I don't believe there is any real evidence showing that PIE and OC shared a common origin as Edwin Pulleyblank suggested. These seem to be individual lexical borrowings (if they are indeed even that), and the broader structural association of PIE e/o with OC ə/a presumably reflects a broader linguistic universal of sorts. As for the specific nature of the borrowings, the best we can really do from a linguistic perspective is reconstruct the underlying phonology and then see how plausible surface reflexes of that phonology might match up (ideally supported by archaeological evidence). As I've stated many times before, that is why I prefer a reconstruction of OC that looks for deep phonological structure allowing for a variety of surface reflexes rather than the more standard rigid shallower reconstructions that limit surface (i.e. dialect/accent) variability.

  21. Chris Button said,

    March 17, 2018 @ 8:52 am

    @ Eidolon

    By the way -thanks for the link to Jan Židek's "Tocharian Loanwords in Chinese". It's useful to have such a large list with a variety of sources cited.

    The word 麥 "wheat" is notably absent although perhaps that is due to no convincing Tocharian connection outside of the broader Proto-Indo-European one? Václav Blažek's (2013) recent proposal (mentioned on an earlier LL thread) is convincing in terms of the PIE connection there.

  22. James Wimberley said,

    March 17, 2018 @ 8:12 pm

    Is this now the Year of the Poodle?

  23. Chris Button said,

    March 18, 2018 @ 10:27 pm

    @ Jichang Lulu

    Since Baxter & Sagart (2014:285-6) pay particular attention to 屎, I thought I would add some additional comments to respectfully challenge their reconstruction.

    First, a minor point is that to be consistent with my reconstructions of the etymologically related words 齒 *kɬə̀ɣʔ and 杵 *kɬàɣʔ on the "dog" posting, I should really have reconstructed 屎 as *ɬə̀jʔ with a lateral fricative rather than *ʰl-, but this is a minor detail.

    B&S reconstruct 屎 *ɬə̀jʔ as *[qʰ]ijʔ. Their *-ij corresponds with my *-əj (Schuessler has *-i for B&S *-ij presumably since *-ij would be tantamount to assuming distinctive vowel length relative to other open syllables). Their tentative uvular *qʰ appears to be reconstructed on the basis of an onomatopoeic usage of 屎 as "moan", but I would put this down to the instability regarding (partial) devoicing of sonorants (the inverse of what I mentioned in an above post regarding the instability of voicing obstruents).

    B&S also separate 尸 *ɬə̀j from 屎 *ɬə̀jʔ through an alternative reconstruction of 尸 that would represent *ɬə̀l in my reconstruction using Pulleyblank's lateral coda. The justification for this is the use of 尸 *ɬə̀j as a loangraph for 夷 *lə̀l in the bronze inscriptions. However, the confusion of the rhymes *-ə̀l and *-ə̀j, which weren't consistently distinguished in the Shijing, would already have been in play by then. B&S further suggest that the oracle-bone forms for 夷 "evidently represent a person" which they want to associate with the squatting person shown by 尸, but Takashima (2010:132) observes that 夷 is "usually distinguishable" from 人 'person' which "becomes unclear in Period V" (i.e. very late in the inscriptions). You can see the clearly differentiated oracle-bone form of 夷 on the "ctext" site where it is mistakenly listed under 尸 (the graph 尸 is not distinguished in the oracle-bone inscriptions and 屎 simply has 人 in the oracle-bones where it now has 尸), yet B&S do not appear to note this distinct form.

  24. Eidolon said,

    March 20, 2018 @ 5:58 pm

    @Chris Button

    Have to agree about Pulleybank. As a general comment, I believe that the key to tracing PIE-OC connections lie in unraveling the layers of borrowing reaching across the entire Eurasian steppe. Many of the earlier languages of the region might be extinct without a trace, but presuming IE influences arrived during specific periods of history, there should be regularities in the pattern of borrowing even accounting for the idea that they were loaned through intermediaries. Languages belonging to the Tibeto-Burman family and the "Altaic" group are of particular interest, in this respect, as are Paleo-Siberian languages of significant antiquity, as such languages may contain the last traces of ancient tongues once spoken in the vast space between the homelands of IE and OC, and which should, therefore, have been in contact with IE even earlier and more directly.

    @ Chau

    Unfortunately, no, as synthesizing the many threads of evidence would take a tremendous amount of time, and so I am left with only broad strokes. But I will say that I approach these issues from the perspective of anthropology and archaeology more than linguistics, and that much of what I wrote above is drawn from the contemporary literature of those fields, in case you're interested.

  25. Chris Button said,

    March 21, 2018 @ 7:52 am

    Regarding the "Eurasian Steppe", I was just looking at Pulleyblank's 1999 ("The Peoples of the Steppe Frontier in Early Chinese Sources") article and noted the following comment on p.38:

    Hua 華… which also means "flower", and Xia 夏… which read in the departing tone… means "summer", the season of flowering, are quite likely etymologically related words"

    It reminded me of a later similar proposal by Beckwith that was discussed on this LLog post:

    http://languagelog.ldc.upenn.edu/nll/?p=34247

  26. David Marjanović said,

    March 21, 2018 @ 8:54 pm

    Obstruents are by their nature voiceless such that any voicing that may occur is relatively hard to maintain (even in English, the so-called "voiced" obstruents are only reliably voiced intervocalically otherwise they tend not to be fully voiced or even voiced at all). Pre-nasalisation of voiced obstruents is often used as an articulatory mechanism to retain voicing since nasals are sonorants and by their nature voiced. There's a good paper on this by John Ohala (1983)

    I really wouldn't put "even" and "English" so close together. The English fortes (as those of most other Germanic varieties) are redundantly marked by aspiration wherever that is easy to maintain; therefore, the lenes don't need to be all voiced all the time, and therefore they aren't. In French or Russian or mainstream Dutch, where aspiration is lacking, the lenes are reliably voiced.

    Many instances of the Japanese voiced/prenasalized lenes actually come from consonant clusters beginning with a nasal, though perhaps not all of them.

  27. Chris Button said,

    March 21, 2018 @ 10:46 pm

    @ David Marjanović

    I really wouldn't put "even" and "English" so close together. The English fortes (as those of most other Germanic varieties) are redundantly marked by aspiration wherever that is easy to maintain; therefore, the lenes don't need to be all voiced all the time, and therefore they aren't. In French or Russian or mainstream Dutch, where aspiration is lacking, the lenes are reliably voiced.

    Also in coda position you get fortis clipping such that vowel length can be used to distinguish "feet" and "feed" rather than different codas. As a result, English does not really need to maintain its voicing there either. Not all languages use pre-nasalisation as a means to maintain voicing; it is just a very common means employed to maintain it.

    Many instances of the Japanese voiced/prenasalized lenes actually come from consonant clusters beginning with a nasal, though perhaps not all of them.

    Only velar [ᵑg] in variation with [ŋ] and [g] is attested in Modern Japanese (e.g. hage "bald" surfacing as something closer to haŋe for many speakers). I would suggest that the issue is not about [ᵑg] being treated as a direct continuation from Old Japanese intervocalically (following a preceding nasal syllable), but rather that it only pertains to the velar obstruent since voicing is harder to maintain on velars than on coronals or bilabials etc.

    In terms of Old Chinese, a nasal prefix may indeed have triggered the voicing of obstruent onsets. However, without any solid evidence supporting it, I would favor the explanation that the pre-nasalisation attested in loans from Chinese resulted from a well-attested surface articulatory phenomenon instead. By the way, Kuki-Chin evidence from Lamkang suggests an earlier fortis-lenis distinction rather than voiceless-voiced (where voicing emerged as surface reflex of lenis obstruent onsets in the same way length emerged as a surface realisation of fortis sonorant codas) which may have some interesting implications for PTB/PST as a whole.

    @ Eidolon

    Speaking of words with external connections that have clearly gone through a variety of different languages, the word 桂 "cinnamon, cassia" is an interesting one.

    Based on its xiesheng series, I would reconstruct it as *χájs with a voiceless uvular fricative. However, this could regularly be derived from an earlier form *χʲáts which provides a very suggestive association with Hebrew qetsia "cassia"

  28. Chris Button said,

    March 21, 2018 @ 11:50 pm

    Minor correction: apparently non-velar prenasalisation is attested/retained in some varieties of Japanese, but the big alternation is between g and ŋ

  29. Chris Button said,

    March 23, 2018 @ 11:35 am

    @ Eidolon

    Is there any evidence of an IE or PIE /g/ loan surfacing as [ŋ] in a neighboring language?

    Pulleyblank (1989, 1995) compares OC 雁 *ŋráns "wild goose" with PIE *gʲʰans- "goose". A close correspondence with 鵝 *ŋál "goose" is presumably what lies behind Baxter & Sagart's reconstruction of a dialect shifted *-r coda in both words. The comparison is included in Židek's list, although Pulleyblank is not cited.

  30. Eidolon said,

    March 23, 2018 @ 7:56 pm

    @ Chris Button

    Zidek doesn't seem to think that this is a plausible loan into OC, though, so in a sense we are trying to establish the validity of an example with another example whose validity still needs to be established. Zidek does mention "this might indicate a chain of borrowing and gradual shift in meaning or a simple case of onomatopoeic base." It's curious to me that reconstructed proto-languages all around the world often have such similar words for animals – the "cow" example above being a prime example. In this case, it's even over two physically different animals – the Chinese goose which are derived from the wild swan goose A. cygnoides and have a large knob at the base of the bill, and the West Asian and European goose which are derived from the greylag goose A. anser and historically lacked the knob. A recent genetics study of domesticated goose in China indicates that virtually all breeds of goose in China are derived from A. cygnoides, except for a breed in Xinjiang that came from A. anser, perhaps reflecting an ancient Tarim Basin lineage, or a more recent import. It's strange, when the physical evidence and the linguistics do not match.

  31. Chris Button said,

    March 23, 2018 @ 9:28 pm

    In spite of its somewhat problematic reconstruction in PIE, the other oft-cited animal comparison between Indo-European and Old Chinese is PIE *marko- "horse" with OC 馬 *mráɣʔ as mentioned by Prof. Mair in his original post regarding "dog".

    What's interesting about 犬 "dog", 牛 "cow", 雁 "goose" and 馬 "horse" is that they all seem to be relatively isolated forms in Old Chinese. This contrasts with 車 *kɬàɣ which has developed a word-family of its own in Chinese.

  32. David Marjanović said,

    March 25, 2018 @ 4:49 pm

    In terms of Old Chinese, a nasal prefix may indeed have triggered the voicing of obstruent onsets. However, without any solid evidence supporting it, I would favor the explanation that the pre-nasalisation attested in loans from Chinese resulted from a well-attested surface articulatory phenomenon instead.

    But why so haphazardly, then? Wouldn't that be the least regular sound change ever?

    By the way, Kuki-Chin evidence from Lamkang suggests an earlier fortis-lenis distinction rather than voiceless-voiced (where voicing emerged as surface reflex of lenis obstruent onsets in the same way length emerged as a surface realisation of fortis sonorant codas) which may have some interesting implications for PTB/PST as a whole.

    What do you mean by "fortis sonorant"? I'm not used to any usage of "fortis" that could make sense here.

    PIE *marko- "horse"

    Why do you call it PIE at all? It's only found in Celtic and Germanic, and by containing a pesky *a it doesn't even sound like a PIE stem. Of course the Germanic version could be a (pre-Grimm) loan from the Celtic one, in which case the Celtic one could be a zero-grade like *(H)m(H)rk(ʲ)-, but I can't find any suitable candidate for that in the appropriate Wiktionary appendix, for whatever that's worth (perhaps not all that much).

  33. David Marjanović said,

    March 25, 2018 @ 4:50 pm

    Sorry if I sound snarky – I'm genuinely interested.

  34. Chris Button said,

    March 25, 2018 @ 10:27 pm

    But why so haphazardly, then? Wouldn't that be the least regular sound change ever?

    The degree of phonetic nasalisation with a phonemic voiced obstruent can vary from overt (sometimes ending up as complete nasalisation) through to not really discernible outside of acoustic analysis (what is generally considered just a plain voiced stop). Even if there were not this variation over time and dialect, loanwords are not candidates for beautifully regular sound change anyway since there are many variables at play.

    Not being a specialist in any of the languages within which this pre-nasal correlation is attested, I would love to know how exact the correspondence is between "intransitive voicing" in OC (i.e. clearly a derived verb) and pre-nasalisation in a loan from OC, as opposed to regular voicing (i.e. no evidence in OC for any derivational process) and no pre-nasalisation. One could very easily slap a nasal prefix on every voiced obstruent in Old Chinese that has a pre-nasalised correlate in a loanword in another language regardless of any evidence for a derivational process in OC, but that would not be very good linguistics in my opinion.

    What do you mean by "fortis sonorant"? I'm not used to any usage of "fortis" that could make sense here

    Zapotec languages have some really good examples of this.

    Why do you call it PIE at all? It's only found in Celtic and Germanic, and by containing a pesky *a it doesn't even sound like a PIE stem.

    That's why I referred to *marko- as "somewhat problematic". There are a variety of proposals out there in the literature on how to fill in the gaps here. By the way, the weird *a also occurs in "goose" and I think *-h₂e- is often reconstructed instead.

  35. Chris Button said,

    March 26, 2018 @ 5:29 am

    Apologies for the messed up formatting at the bottom – the last 4 lines should not be in "blockquote"

    Also, just to be clear, by…

    One could very easily slap a nasal prefix on every voiced obstruent in Old Chinese that has a pre-nasalised correlate…

    … I mean one could very easily slap a nasal prefix to condition voicing on every obstruent onset in OC that has a pre-nasalised correlate….

  36. David Marjanović said,

    March 26, 2018 @ 2:33 pm

    Quoth Wikipedia:

    "In Zapotec languages, fortis typically corresponds to voicelessness and extra length in obstruents and extra length in sonorants. Lenis corresponds to voicing and less length in obstruents and less length in sonorants. In addition, stressed vowels before lenis consonants may be longer than those before fortis consonants.[19]"

    So, it's simply length. I suppose that makes sense.

    Here is a complex attempt to explain *marko- as an originally specialized term, not a general word for "horse", which involves the Germanic version definitely being a Celtic loan and the Celtic version possibly derived from an Iranian one via a pretty drastic shift in meaning.

  37. Eidolon said,

    March 27, 2018 @ 8:46 pm

    "What's interesting about 犬 "dog", 牛 "cow", 雁 "goose" and 馬 "horse" is that they all seem to be relatively isolated forms in Old Chinese. This contrasts with 車 *kɬàɣ which has developed a word-family of its own in Chinese."

    This is easy to explain, though: 車 is a disruptive technological development, while the other examples are naturally occurring animals, domesticated or not. 車 is a known import, besides, so the fact that it developed a word-family of its own likely reflects the revolutionary nature of the technology, rather than deep linguistics.

    As a side comment, what I find particularly interesting is that there is a general lack of metallurgical cognates between IE languages and OC. It's curious that of the lists produced by various comparative analyses over the years, IE metallurgical terms so rarely feature among them. Part of the reason is that PIE has few shared words relating to metallurgy, indicating that the descendant languages of IE split before the Bronze Age arrived. But even the daughter languages of IE do not share much in common with China with respect to words for bronze, and Zidek's list tellingly features no candidates for such words borrowed from Tocharian. Thus, by the time chariots arrived from the Andronovo expansion, bronze was well-established in China already, which would explain the lack of shared bronze terms.

    In that case, we must assume that the cultures of China had contact with West Eurasia through intermediaries besides IE, since the Bronze Age in China still should be a foreign import in its initial phase. Yet the Near East had access to bronze a thousand years before China, so the process of transfer must have been slow, passing through many local cultures until eventually reaching East Asia. What languages these cultures spoke are, perhaps, lost to history; but it is entirely possible that they played a similar role in the transfer of domesticated animals during the middle and late Neolithic.

  38. Chris Button said,

    March 28, 2018 @ 9:52 am

    Thanks for the additional insights Eidolon. In particular, your comment regarding 車 is a great observation. I had always struggled with the notion that since 車 has a decent word-family of its own in Chinese it could not a priori have originally been a loanword.

RSS feed for comments on this post