Tocharo-Sinica
« previous post | next post »
Language Log has been fortunate to have had several guest posts and numerous comments by Douglas Adams, doyen of Tocharian studies in America (see "Selected readings" for a sampling). Now, stimulated by the recent post on Chinese chariotry, he has written the following ruminations in response.
I read with interest the material on early Chinese chariotry. It was far outside my competence to judge. As you knew, I was most interested in the comment that was looking to the possibility of Tocharian > Chinese lexical borrowings. As you also know, it has long been my suspicion that there was more west > east influence on Chinese language and culture than is generally realized. And the "westerners" involved were most likely to have been Tocharians of one sort or another ("Tocharian D"?). It's probably not only PIE pigs and honey that, via Tocharian, show up in Chinese.
It's a pity that the ancient Chinese, like the ancient Greeks, were so totally uninterested in the "barbarian" languages that were their neighbors. We have our single sentence example of the Jie language recorded, not in a difficult Greek alphabet transcription, but rather in an inscrutable Chinese character transcription. (Inscrutable because reconstructing pre-Middle-Chinese Chinese phonology makes reconstructing PIE phonology look like child's play.) Adding to the problem of course is that Chinese phonology, at all times, renders that incorporation of foreign words possible only at the cost of (considerable) deformation. Look, for instance, at how poor Buddhacinga's name is rendered (Fótúchéng 佛圖澄 [many people used to mispronounce that Fótúdèng]) (ca. 232–348 AD) . Who would have guessed?
I'm still suspicious that the name given by the Chinese to the Kuchean royal house, Bai, may be connected to the homophonous designation of the "barbarian" rulers surrounding the nascent Chinese state on the North China Plain. And, speaking of royalty, is it possible that wang 'king' might be from pre-Tocharian *wnatke (TchA nātäk 'lord'), which in turn might be cognate with Greek (w)anakt– '[Mycenean] king.' (TchA nāśi 'lady' would equally be the equivalent of Greek (w)anassa 'queen' from pre-Greek *wanakyā-, both irregularly related to 'king/lord' [where's the *-t-?]. The latter word survives in Modern Greek in pant-anassa 'all-queen,' an epithet of the Virgin Mary.) But, even if true, who's going to believe it?
This is but a taste of what is to come. Doug is preparing a paper that touches on one of these subjects at greater length. It is tentatively titled "Resurrecting an Etymology: Greek (w)ánax ‘king’ and Tocharian A nātäk ‘lord’" and will probably appear in Sino-Platonic Papers sometime this summer.
Selected readings
- "Tocharian words for oil" (6/22/22)
- "The origins and affinities of Tocharian" (8/20/23) — lengthy, classified bibliography
- "Tocharian C: its discovery and implications" (4/2/19)
- "The geographical, archeological, genetic, and linguistic origins of Tocharian" (7/14/20) — with a comprehensive bibliography
- Hajni Elias, "The Southwest Silk Road: artistic exchange and transmission in early China", published online by Cambridge University Press: 07 March 2024; Bulletin of the School of Oriental and African Studies, First View, pp. 1 – 26. This article has impressed me to such a degree that I have rechristened the road she wrote about as "The Southwest Bronze Road".
- "From Chariot to Carriage" (5/5/24)
- "An early fourth century AD historical puzzle involving a Caucasian people in North China" (1/25/19)
Wolfgang Behr said,
May 13, 2024 @ 3:40 pm
The Gk. ἄναξ, -κτος 'lord, ruler' ( OC 王 'king' idea is an old one, mentioned e.g. by TT Chang (1988, in SPP No. 7, p. 23). Given that uvular reconstructions of OC now posit *ɢʷaŋ (> MC hjwang) it looks a bit less plausible than 36 years ago, and the implied contact pre-Gk/OC scenario still seems mindboggling. OC 王 occurs since the earliest layers of oracle bone inscriptions, so it would imply a borrowing during the middle of the second millennium BCE. On the Chinese side, 王 is usually thought to belong to an OC word family including 往 (OC *ɢʷaŋʔ, 'to go to'; maybe also 皇 (OC *[ɢ]ʷˤaŋ) 'august; sovereign'), because the Shang kings clearly were "peripatetic" rulers. To plausibilize the proposed connection the first step would be to show that this word family is isolated within Sino-Tibetan. Also, why would OC have rendered -nakt (or -nak, since -kt is younger, as per Beekes Etym. Dict. s.v., p.99) by *-aŋ rather than by a rusheng word ending in *-ak; and why would the Gk. disyllable not have at least ended up in a sesquisyllabic equivalent in OC (type: *ɢʷə-nak)?
Lucas Christopoulos said,
May 13, 2024 @ 7:00 pm
Earlier than the famous 甲骨 (used again always as a single perpetual genitor of the Chinese characters and language), there are the older Erlitou pottery scripts, exactly similar to the Dispilio and the Tărtăria tablet (5500–5300 BC). Most likely an early Indo-European writing system coming from the Black Sea area.
Myceneans/Minoans/Andronovo are genetically connected, and the scripts also have the same 王. Even the Linear A/B has it somehow..so it is very possible that a Wanax was riding his war chariot there in Northern China.
Erlitou pottery scripts:
https://www.degruyter.com/document/doi/10.1515/CHAR.2005.5.1.183/html?lang=en
martin schwartz said,
May 13, 2024 @ 7:26 pm
Of course looking forward to DQA's further thoughts.
Was I naïve in thinking that Toch. A nātäk is merely < Skt. nātha-
'leader'?A priori, relicta Myc *wanakt- and Toch. A do(es) seem a
stretch….
Martin Schwartz
martin schwartz said,
May 13, 2024 @ 7:40 pm
p.s I do see online Gerd Carling, Sanskrit and Tocharian, take
nātäk via a Middle Indic form with *-aka- from the Skt.
Martin Schwartz
Chris Button said,
May 13, 2024 @ 9:16 pm
The Tocharian B form ñakte actually seems closer to the Greek than Tocharian A does.
Regarding Chinese, I approach uvulars rather differently from B&S. I would reconstruct 王 as ʁaŋ, which at least removes B&S's labialization but doesn't get us much closer otherwise.
Chris Button said,
May 13, 2024 @ 9:27 pm
Regarding silver, I would reconstruct 銀 internally as Old Chinese ᵑgrən to give Early Middle Chinese ŋin. That chimes well with Tocharian nkiñc ~ ñkante. The OC medial -r- is needed to account for EMC ŋin rather than ŋɨn, but it may well be spurious if the EMC vocalism is reflective of the Tocharian loan instead of an OC pedigree.
David Marjanović said,
May 14, 2024 @ 2:53 pm
This is a very large number of claims that need to be substantiated.
Myceneans and Andronovo, yes: they both have ancestry from people buried by the Pit Grave (yamnaya) Culture.
The Minoans lacked that.
"Even"? Linear A is (the linear version of) the Minoan writing system, and Linear B is its Mycenean adaptation (for a very different language – Greek).
Which sign do you mean by 王 in Linear A and B?
You'd think a shape as simple as 王 would occur in Linear A and/or B just by chance, but I can't find it in the lists of Linear B glyphs that had, or may have had, sound values in this article; unfortunately I'd need a font for the logograms, so I can't see them.
Linear A does have it – but most likely it simply stood for the syllable ne and had nothing to do with royalty.
On the Chinese side, 王 shows remarkably little change over the millennia.
Lucas Christopoulos said,
May 14, 2024 @ 6:29 pm
@Marjanovic
Nope…Andronovo people are DIRECTLY connected with the Myceneans
https://brill.com/display/book/edcoll/9789047420712/Bej.9789004160545.i-763_029.xml
Jonathan Smith said,
May 14, 2024 @ 7:31 pm
re: "the famous 甲骨 (used again always as a single perpetual genitor of the Chinese characters and language), there are the older Erlitou pottery scripts, exactly similar to the Dispilio and the Tărtăria tablet"
I agree it is useful to step away from the assumption that Jiaguwen are the sole font of all subsequent Chinese writing… but it's not going to be productive to compare much earlier pottery marks etc. (of which there are no shortage) to material from Neolithic eastern Europe… it hasn't even been productive (so far!) to compare the Chinese Neolithic marks to OBI, for heaven's sake.
And earliest "王" often looks like 'upside-down Y with strikethrough and underscore' :P; see at center here… and yeah it would hardly be surprising or significant even if it did always look like some symbol in say Linear A.
Lucas chrisitopoulos said,
May 15, 2024 @ 1:43 am
"it would hardly be surprising or significant even if it did always look like some symbol in say Linear A."
, wa-na-ka, in Linear B (Mycenean)
Lucas Christopoulos said,
May 15, 2024 @ 1:49 am
Wa-na-ka in Linear B (Mycenean)
https://www.palaeolexicon.com/Word/Show/16631
Philip Taylor said,
May 15, 2024 @ 7:42 am
David — " I'd need a font for the logograms, so I can't see them" — I just installed https://fonts.google.com/noto/specimen/Noto+Sans+Linear+B, whereupon the logograms became immediately visible.
Georges-Jean Pinault said,
May 15, 2024 @ 12:02 pm
I concur with Martin Schwartz and Wolfgang Behr. Thanks for their comments. I may mention that in the Dictionary and Thesaurus of Tocharian A, authored by Gerd Carling and myself, Wiesbaden, Harrassowitz, 2023, TA nâtäk is taken as derived from Middle Indic nâthaka-, based on Sanskrit, as it has been surmised for long. This is the simplest solution, since Skt; nâtha- masc. (based on a neuter meaning 'refuge, help') was currently used from Epic Sanskrit onwards in the sense 'protector, possessor, owner, lord' (Monier-Williams, 1999: 534c). The connection of the TA word with Gk. (w)anaks, (w)anakt-os has been already proposed by Werner Winter (1970). This fancy idea faces severe difficulties as for consonantism and vocalism. Actually, one cannot reconstruct a unique PIE noun 'king', which does not mean that this notion was foreign to IE-speaking peoples. The Gk. word, which is very probably shared by Phrygian, belongs probably to non-IE substrate. It any case, it cannot receive a PIE etymology, despite speculative accounts in the more or less recent past. All Greek words for 'king' are either borrowed or from relatively recent coinage in Greece and neighborhood itself.
David Marjanović said,
May 15, 2024 @ 1:43 pm
I don't understand what you mean. They have the common ancestry I mentioned; what else?
Thank you. That's not a logogram, it's three syllable signs; the word they spell might theoretically be related to wáng (though see above for reasons for strong skepticism), but the signs themselves have nothing in common with the sign 王.
Lucas Christopoulos said,
May 15, 2024 @ 5:47 pm
-for the Hellenic/Mycenean—>Andronovo similar DNA. Mathieson I et. al 2018; Chang Chao et al 2019. Mycenean–>Minoan, Lazaridis et al, supplementary information: pp 2-3. For the meander on their potteries (S. Zharkinova 1985)
Other words of Greek origins in Tocharian either from the Mycenean period or Hellenized Bactria (The problem of Tocharian origins SPP 259. J. P Mallory).
"All Greek words for 'king' are either borrowed or from relatively recent coinage in Greece and neighborhood itself." Definitely not…
Lucas Christopoulos said,
May 15, 2024 @ 5:51 pm
"I don't understand what you mean. They have the common ancestry I mentioned; what else?" Not only distant, check also the cultural and technological similarities (meander designs, war-chariot system/spokes, bronze warfare technology, society organization etc.)
Chris Button said,
May 17, 2024 @ 6:38 am
If 往 is etymologically related to 于 (the presence/lack of velar nasal being due to a particular conditioning environment rather than a suffix–note Pulleblank's velar -ɣ/ɰ coda in 于), then this is probably even more of a stretch than it seems just semantically.
David Marjanović said,
May 18, 2024 @ 2:38 pm
All that is general Indo-European, exactly as distant as the common ancestry.
I don't know about the meanders, but expect they were invented numerous times independently.