Tuesday, November 1, 2011

Tea vs Chai, the Tekka Centre and my last name (II)

In yesterday's post I talked about the correspondence between Hokkien 't' and Mandarin 'zh' (a retroflex sound produced with the tongue slightly further back than the sound represented by 'ch' in 'chunk' and without the puff of air). Both sounds are descended from an earlier 'tr' cluster in Early Middle Chinese, as reconstructed by historical linguists.

What does this have to do with the word for 'tea'?

People who know Hindi, may laugh surreptitiously when they hear people order a 'chai tea', since चय chay means 'tea' in Hindi, so the order is basically for a 'tea tea'. In English though, 'chai tea' is perfectly acceptable because the word 'chai' has been borrowed to designate what one would call मसाला चय masaalaa chai 'spiced tea' in India.

The Hindi word for tea is चय chay is much closer to the Mandarin cha (the 'ch' sound here is pronounced like the retoflex 'zh', the only difference is that it is accompanied by a puff of air). Other Indo-Aryan languages like Nepali have चिया chiyā. Within Indo-European, we also have Russian чай chay. The Japanese also use cha. In contrast, English has tea, French thé and Malay teh. Hebrew too uses תה te (I was taught that תה נענע te nana is '(spear)mint tea' in Hebrew). These languages all have a word for 'tea' that's closer to the Hokkien / Minnan word te (tone not given).

The reason for this difference is that languages like English borrowed (whether directly or indirectly) the word from one of the Minnan dialects / languages, while languages like Russian and Hindi borrowed the word from other Chinese languages like Mandarin or Cantonese. The Wikipedia article explains this in greater detail and gives more examples from other languages.

Etymologically though, Mandarin cha and Hokkien te share the same origin. Pulleyblank (1991) gives the reconstructed forms draɨ /drɛ (Early Middle Chinese) and trɦa: (Late Middle Chinese). Again, we see the correspondence between the Mandarin retroflex sounds (written in pinyin as 'zh' and 'ch') and Minnan 't',

So voilà, it took me two posts to do it, but there you have it - the common thread linking my last name, the name of the Tekka Centre and the name of one of the most consumed beverages on the planet.

[This post was inspired by 3 separate conversations I've had in the last month about each of these topics. Tomorrow I'm off to the great tea-growing state of Assam in NE India. The word in Assamese চাহ (transliterated as chah) is clearly related to the non-Minnan form of the word, but is now pronounced 'sah' in Assamese. Something for me to get used to saying again!]

8 comments:

  1. oh, and Min is not a descendent of Middle Chinese. Min split off much earlier. However, as you've shown, there is still a lot of correspondences between Southern Min and other Sinitic languages.

    ReplyDelete
  2. (oh it seems like that my preceding post hasn't been published... here I'll do it again.)
    By the way, the usual understanding is that by the time of Early Middle Chinese, tr was a single segment [ʈ] and not a cluster [tr]. See, e.g., page 7 of Pulleyblank (1991) that you lovelily provided a link of :).

    ReplyDelete
  3. Ah, thanks for clearing that up. But yeah, despite the time gap, there are still quite a number of regular sound correspondences.

    You're right, Pulleyblank says that [tr] was not "a single segment and not biphonematic cluster", but I think he thought it was a phonetic affricate, not a retroflex stop [ʈ] (although I have seen the symbol used for Tibetan retroflex consonants, which I've always heard as affricates.)

    ReplyDelete
  4. Oh interesting. Missed that bit where he says /ʈ/ was an affricate (I read too quickly). But then at the same time there were also minimal pairs between the /ʈ/ series and the /tʂ/ series in his Early Middle Chinese. In fact there were minimal sets between the /ʈ/ series, the /tʂ/ series, the /ts/ series, and the /tɕ/ series (e.g. 張莊將章 respectively in Early Middle Chinese). I suppose if /ʈ/ is phonetically an affricate... it would have to be... perhaps [tʃ], given that /tʂ/ was presumably already [tʂ] and the other coronal places are also taken.

    ReplyDelete
  5. I think the release phase of the affricate might have been a voiceless rhotic. One example he gives is English 'train' which I think is phonetically [tʃ] + voiceless rhotic.

    ReplyDelete
  6. Wah :D cool. A phonemic contrast between /ʈ/[tʂ] and /ʈʂ/[ʈʂ].

    OK, grant you that the /ʈ/-series did merge into the /ʈʂ/-series later in Late Middle Chinese (probably after the merger/confusion between /dʐ/ and /ʐ/). However, what I was saying was that in Early Middle Chinese, his characterisation of /ʈ/ being phonetically an affricate at that stage (without having merged into /ʈʂ/) was problematic to me. There was a phonemic contrast between /ʈ/ and /ʈʂ/ at that stage, but he also claims that /ʈ/ was phonetically an affricate, and he gives Southern Vietnamese tr [ʈʂ] as an example (next to the English 'train' example) of what /ʈ/ might sound like. Given that /ʈ/ was retroflex, and it has to be distinct from /ʈʂ/ (and also from /tɕ/,/ts/ and /t/ in his Northern Early Middle Chinese), one wonders what phonetic value /ʈ/ could have had if it was an affricate, and at the same time distinct form /ʈʂ/. (A distinction of [tʂ] and [ʈʂ] seems unlikely, but perhaps not impossible.)

    ReplyDelete
  7. Hmmm, I was thinking even that /ʈ/ vs /ʈʂ/ was simply a contrast betweennaspirated and unaspirated [ʈʂ], only because when I try to produce [ʈɻ] with a fricated release, it sounds almost like [ʈʂ] but with a lot more aspiration.

    But that has other problems too, because I think he mentions a distinction between /tr/ and /trh/ as well I believe...

    ReplyDelete
  8. Yes, messy indeed (fun though). There were voiceless unaspirated, voiceless aspirated and voiced for all the plosives and affricates...

    ReplyDelete