Indirect archeological evidence for the spread and exchange of languages in medieval Asia

« previous post | next post »

The title of this article about the Belitung shipwreck (ca. 830 AD) is somewhat misleading (e.g., there is no direct evidence of Malayalam being spoken by any of the protagonists, but it is broadly informative, richly illustrated, and well presented.

"Mongols speaking Malayalam – What a sunken ship says about South India & China’s medieval ties

The silent ceramic objects that survive from medieval Indian Ocean trade carry incredible stories of a time when South Asia had the upper hand over China."

Anirudh Kanisetti

The Print (8 September, 2022)

It's intriguing, at least to me, that the author identifies himself as a "public historian".  He is the author of Lords of the Deccan, a new history of medieval South India.

The Belitung shipwreck played a key role in The True History of Tea (Thames and Hudson, 2009), by myself and Erling Hoh.  That is because, out of the roughly 60,000 artifacts retrieved from the wreck, one bowl bore this inscription:  chá zhǎnzi 茶盞子, actually tú zhǎnzi 荼盞子.  As demonstrated in Appendix C of The True History of Tea, the orthography, phonology, and morphology (note the noun suffix) of the inscription all characterize it as Middle Vernacular Sinitic — which fits with the date of the wreck and the history of tea.


Chinese emperors sent multiple embassies to South India from the 13th century onwards. At the same time, the archaeological record shows a sudden increase in ceramic sherds found along the coast. Just as South Indian kings and merchants recognised the value of commerce, Professor Tansen Sen shows that so did the Chinese — especially the rulers of the Yuan Dynasty (1271–1368), who were of Mongol descent.

Seeking both military and commercial expansion, Yuan ruler Kubilai Khan sent no less than fourteen missions to India in his reign. The objective of these missions was to secure embassies from the Pandya kingdom as well as from smaller trading cities such as Kozhikode and Kollam on the Malabar coast; interestingly, we know that Kollam’s local Christian, Jewish and Muslim merchants also sent embassies to the Yuan court, joining its ambassadors on their return voyage. In one case, a merchant from Gujarat arrived at Kollam to make submissions to the Chinese embassies. All of these were presented to Kubilai Khan’s court as proof that distant polities acknowledged his supremacy, thus establishing him as a worthy successor to his grandfather Genghis Khan.

Along with these courtly interactions, we see a deepening of the networks of merchants. Chinese traders also began to play a role in trade; they are mentioned travelling to India in large ships, even setting up a large pagoda in the Tamil port of Nagapattinam. The Yuan dynasty’s encouragement of trade was so successful that they began to panic about the outflow of metallic currency being used to buy Indian luxuries like pearls and kingfisher feathers. Despite bans, trade and merchant movements continued to grow; by the time of the Ming dynasty in the 15th century, some Indian kingdoms were even hiring immigrant Chinese merchants to lead their embassies to China.

Unlike the Yuan rulers, who saw the Indian Ocean trade as a means to profit and prestige, the Ming emperor Yongle (r. 1402–1424) was interested in establishing China as a “civilising” power in the Indian Ocean — the culmination of diplomatic and commercial ties built up over centuries. The expeditions sent by his admiral Zheng He to Southern India were a result of its emergence as an important trading partner in previous centuries. Strikingly, the Ming dynasty also interfered in the affairs of the Bengal Sultanate in the 15th century — we’ll explore these and their motivations in a future column.

There's much more to the story.  Stay tuned.


Selected readings and viewings

[h.t. William Triplett]


  1. David Marjanović said,

    September 15, 2022 @ 2:52 am

    chá zhǎnzi 茶盞子, actually tú zhǎnzi 荼盞子

    That made me curious!

    The Wiktionary article on 荼 gives it, when pronounced tú*, the following meanings (I haven't copied the links):

    "1. a type of bitter vegetable, of Sonchus or Lactuca
    2. (figuratively) pain; suffering
    3. white flower of cogon grass or reed

    Usage notes

    荼 can refer to one of several types of bitter[-]tasting vegetables. Early Chinese texts are rather vague in their description, which makes it difficult to determine which species it originally referred to."

    It looks like the prehistory of tea is contained in the history of this character! I would like to speculate that 荼 originally referred to bitter vegetables in general, and that the character 茶 was consciously derived from 荼 to write the name of a previously unknown bitter plant, i.e. a new and special kind of 荼…

    * It has no less than three other pronunciations: chá when it is considered a variant form of 茶, Yé as a surname, which is clearly irrelevant here, and -shū in the equally inapplicable name of the deity Shēnshū ~ Shénshū.

  2. Victor Mair said,

    September 15, 2022 @ 5:16 am


    Read The True History of Tea for more, especially Appendix C.

  3. Chris Button said,

    September 15, 2022 @ 5:18 am

    @ David Marjanović

    荼 and 茶 represent two forms of the same OC word *laɣː [laɰː] . Evidence for the velar coda is found in things like Burmese *lak. The word represented by 茶 is often reconstructed with a medial -r- to account for the Middle Chinese reflex with retroflex onset and a similarly conditioned vowel, but some largely ignored related proposals by Pulleyblank (not specifically on this character) show the variation to be the result of analogy caused by some secondary vocalic lengthening in certain environments like this one.

  4. Victor Mair said,

    September 15, 2022 @ 6:39 am

    @Chris Button


    And Burma is right in the direction of the botanical homeland of tea.

    Note that tea has deep roots in Wa (a Mon-Khmer language) present in southernmost Yunnan (where the oldest and tallest tea TREES in what is now "China" are to be found).

    The t- and ch- words for "tea" in Sinitic ultimately derived from l- words in Austroasiatic.

  5. Jonathan Smith said,

    September 15, 2022 @ 7:36 am

    chá ‘tea' and 'bitter vegetable (?)' are of course etymological doublets in some sense… but the latter is really a lexicographical artifact in Mandarin, and similar issues affect early sources like the medieval rhyme books. Needless to say there is more than one historical phonological way to skin this cat…

    The idea that Chinese 'tea' goes back to lateral-onset southwestern 'leaf' words (Lolo-Burmese… AA…) is at least as old as Sagart (1999)… older? And no "medial -r-" in chunks of Div. II is increasingly standard… no -r- in 'tea' on e.g. OC systems like Schuessler's (2007)… though note reference to "Old Chinese" might be entirely spurious if 'tea' is a later southernism.

    The head word~character actually *defined* as 'tea [leaf/tree?]' in at least Guangyun is incidentally neither "荼" nor "茶" (a 'vulgar' variant') but "", I suppose unsurprisingly found connected there to the old southwest 巴南…

  6. Jonathan Smith said,

    September 15, 2022 @ 7:37 am

    Haha, new unicode… should be 木+荼

  7. Chris Button said,

    September 15, 2022 @ 8:50 am

    @ Jonathan Smith

    You need to account for the retroflexion in the division II form somehow. Diacritics omitted, Schuessler (2009) has *r-la becoming *dra, and schuessler (2007) has *d-la seemingly on the assumption that the -l- functions as -r-. Neither seems to work very well.

  8. Jonathan Smith said,

    September 15, 2022 @ 9:50 am

    I mean, given "*alra", "*r-la", "*d-la", "*laɣː", etc., and all possible associated stories explicit or im-… the really odd thing would be coming to regard one's own idiosyncratic account as plainly superior. And again — the item hardly need be "Old Chinese" — likely isn't, really. Maybe this is way Baxter & Sagart (2014) now seem to exclude it…

  9. Chris Button said,

    September 15, 2022 @ 10:17 am

    @ Jonathan Smith

    The difference is that all those liquids and other coronals assume that 荼 and 茶 represent two originally distinct forms. The idea that we have all these unspecified affixes with pseudo morphological functions (here we need some kind of liquid infix and velar suffix to give something like *l-r-a-k) starts getting silly. Why not just look at the evolution in terms of comparative phonology and phonetics?

  10. Jonathan Smith said,

    September 15, 2022 @ 12:33 pm

    @Chris Button I'm not seeing any "unspecified affixes…" in past proposals; my interpretation is that everyone (you included it seems) regards these as doublets of some kind…

    E.g. the first 'tea' form I listed above was supposed to be Sagart's (1999) *lra Type A, where cf. the "*lˤra" given in B&S's more recent database — which is found married to the character "荼" and the meaning '[name of a plant]'. So, "doublets", devil is in the details, agree phonology/phonetics may be involved :D

  11. Chris Button said,

    September 15, 2022 @ 6:00 pm

    @ Jonathan Smith

    By "unspecified affixes", I mean affixes that are being proposed solely to account for later reflexes in Middle Chinese without any convincing explanation as to why those affixes are there in the first place. I actually meant is a general comment about common approaches to Old Chinese reconstruction, including the case in hand.

    FWIW, Sagart (1999) suggests the medial -r- to either be an infix "in reference to the granular aspect of tea leaves" or that a Tibeto-Burman source had a retroflex lateral that was rendered in Chinese as lr-.

  12. Chris Button said,

    September 16, 2022 @ 9:11 am

    @ Victor Mair

    Thanks for the comment about Wa. Regarding the Mon-Khmer link and the final -k, I just dug up this from one of my earlier posts on the topic in LLog:

    Regarding a final -k, I speculated on the earlier "Caucasian Words for Tea" post (cited above) that the -k in Written Burmese "lak(phak)" (now pronounced /lə(pʰɛʔ)/) "tea(leaf)" might be connected to the -ʔ in the Mon-Khmer source *slaʔ "leaf". Although Old Burmese did have a glottalic final that became a tone category (ultimately from the same *-s as Chinese qu-sheng), its form as a subscript glottal suggest that already at the time of borrowing it might have been a glottalic co-articulation (modern day creaky tone) rather than an individual segment. As a result, perhaps WB -k was the next closest thing to MK -ʔ. It is notable that Mon-Khmer *slaʔ "spleen" gives Burmese sarak(rwak) (now pronounced /θəjɛʔ(jwɛʔ)/) "spleen" also with a *-k although in this case the *s- in the onset is also retained

  13. Jonathan Smith said,

    September 16, 2022 @ 3:17 pm

    @ Chris Button Your insistence re: others' use of "affixes" is bewildering. Even Sagart (1999), while pointing to his *-r-, doesn't explicitly use it in the manner you claim, i.e., to distinguish cha2 vs. "tu2". So, again… there is deep abiding agreement with you that this difference is likely NOT morphological in nature. Again… doublet pair. Again… lots of possible ways to account for.

  14. Chris Button said,

    September 16, 2022 @ 5:07 pm

    @ Jonathan Smith

    I’m not sure I’m following you.

    I’m reconstructing one form for 荼 and 茶 that diverged in pronunciation. Think of a Brit saying “bath” and an American saying “bath”. Others seem to think some kind of affix (standard would be infixed -r-) caused the alternation, and so reconstruct different forms for 荼 and 茶. That would be like saying that the British vocalism in “bath” resulted from some morphological affix rather than a phonological evolution attested elsewhere in the lexicon too, albeit not entirely consistently (much like the Pulleyblank proposal in fact). To me, it’s all about natural language on sound phonetic/phonological principles rather than algebra.

RSS feed for comments on this post