I am absolutely shocked to learn from cosmicore that Prof. Gong Hwang-cherng passed away on September 11. Academia Sinica has created an extensive memorial site for Prof. Gong. The eulogies contain a lot of biographic information.

I first learned of him in 1996 when I started studying Tangut and had the honor of taking a class from him on Sino-Tibetan at the LSA Linguistic Insititute at Cornell the following year. I wrote a paper for him attempting to find correlations between his reconstructions of Tangut retroflex vowels and rhotic consonants in other Sino-Tibetan languages. It was extremely primitive and will remain unpublished. I'm amazed he humored me.

In 1999, I had the pleasure of meeting him again at Academia Sinica. He spoke to me in fluent Japanese, the language of his education in Taiwan during Japanese colonial rule. His ability to use and analyze languages is surprisingly rare among linguists. Two years earlier, he had told me that I needed to study Russian if I was to work on Tangut. I took his advice and started a Russian course a few weeks later when I returned to the University of Hawaii.

W. South Coblin describes Prof. Gong's language background:

Professor Gong was born in 1934 in Bĕigăng, Yúnlín County, Taiwan. After completion of his secondary schooling, by which time he knew three languages, i.e., Taiwanese, standard Mandarin, and Japanese, he matriculated at National Taiwan Normal University, where he took his BA in a fourth language, English, in 1958. He then served for nine years as an English instructor at Dàtóng High School in Taipei. During this period he received for the 1961-62 academic year a Fulbright Fellowship, which he used for advanced study of English in California. In 1966 he was awarded a Goethe Institute Fellowship, whereupon he proceeded to Munich for a one-year course in language instruction. Having become fluent in German, he remained in Munich and matriculated in 1967 at the Ludwig-Maxmilians-Universität, taking his Ph. D. in 1974. During this period he also served as a lecturer in the Institute of East Asian Culture at the University. Professor Gong s fascination with linguistic matters began during his youth, when, as indicated above, he had already learned three different languages. At Munich he pursued these interests, first undertaking a full course of study in Indo-European comparative philology and then moving to the East Asian sphere, where he focused on Chinese historical phonology.

Eventually Prof. Gong's interests expanded to include Tangutology, which Coblin justly describes as

[...] a truly arcane and daunting field. Entry into it required mastery not only of its own immensely complex ancient textual corpus but also of a range of modern languages in which it had been studied by specialists in various parts of the world. Professor Gong already commanded a number of these, but at least one, Russian, was new to him. With characteristic application and aplomb he quickly mastered this difficult language, so that in his work he had full control of all requisite publications on Tangut produced everywhere in the world. And as soon as he had absorbed all of these he began to work on the reconstruction of the pronunciation of this language, assessing earlier achievements and adding improvements wherever he felt they were warranted. Then, as his Sino-Tibetan comparative work advanced, he factored Tangut into the equation, something that no one else had so far done or even been competent to do.

I reviewed a collection of Prof. Gong's papers on Sino-Tibetan for Cahiers de linguistique - Asie orientale in 2004. Prof. Gong then mailed me a copy of his collection of papers on Tangut. I have been using both books regularly for the past few years.

This blog is heavily indebted to Prof. Gong. My ideas about Tangut are to a large extent outgrowths of his: e.g., my reconstruction is a modification of his. I had hoped that one day I could refine my hypotheses into presentable form so that I could share them with Prof. Gong and show him how far I had come since my class with him many years ago. Alas, that will never be. I can never repay him for all he has taught me.

The American holiday of Thanksgiving begins in an hour. It is already Thanksgiving in Taiwan. Perhaps it is fate that I learned of Prof. Gong's passing tonight. Thank you and goodbye, Prof. Gong. I will always remember you.

10.11.23.23:59: GAPS AS GUIDES (PART 6)

In my last entry I asked,

How many tones did Ancient Thai have [in syllables ending in sonorants]? Five, three, or some other number? Can you determine the number by looking at Thai spelling?

In the Thai spelling examples I gave, the tones of Ancient Thai *paa, *hnaa, and *naa were indicated with

- no superscript marker

- a superscript resembling 1

- a superscript resembling 2

This implies a system with three phonemic tones. Each tonal phoneme had allophones conditioned by initial classes:

Once initials merged (e.g., high *hn- and low *n- merged into n-), the allophones became phonemic, and collapsed into a five-tone system in modern standard Thai:

Note that modern standard Thai falling tone can originate from two Ancient Thai tones (1 and 2). Hence only orthographic and/or comparative evidence can determine the Ancient Thai tone of a modern standard Thai falling tone syllable with an ambiguous initial (e.g., ph-, f-, m- which may be either primary or secondary).

11.24.00:45: Nonstandard Thai dialects may still have distinct tonal reflexes of Ancient Thai tones 1^low, 2^high, and 2^mid: e.g., Nakhon (Brown 1985: 203):

10.11.22.23:07: GAPS AS GUIDES (PART 5)

In my last entry, I asked,

What do you think the difference (if any) between the two spellings [for sonorants] was in ancient Thai?

The two sets originally represented voiceless and voiced sonorants. The prefixed letter ห h- indicated voicelessness:

If we spllit the sonorants into two classes (an original m-class and a hm-class that became a secondary m-class), we can see that the tonal distribution of the hm- and m-classes parallels that of the ph-/f- and b-/v-classes:

In traditional Thai terminology, the various initial classes are merged into three (mid/high/low):

Ignoring (rare) and counting the number of tones per initial, the table can be simplified even further as:

How many tones did Ancient Thai have*? Five, three, or some other number? Can you determine the number by looking at Thai spelling?

*For simplicity, I am only asking about tones in 'live' syllables ending in sonorants, not 'dead' syllables ending in obstruents.

In my last entry, I asked,

What does this [two sets of letters for fricatives in modern Thai] imply about the phonetic values of those letters in ancient Thai?

The two sets originally represented voiceless and voiced fricatives with one exception:

The second h was created to write borrowings with *h that had tones normally associated with voiced initials.

If we spllit the f-class into two classes (an original f-class and a v-class that became a secondary f-class), we can see that the tonal distribution of the f- and v-classes parallels that of the ph- and b-classes:

Ignoring (rare) and counting the number of tones per initial, the table can be simplified as:

Now only the sonorant classes have five tones. Each sonorant has at least two spellings in modern Thai:

(This is not a comprehensive list.) What do you think the difference (if any) between the two spellings was in ancient Thai?

11.22.00:50: Hint: Compare the spellings and tones of

หมา maa R 'dog'

มา maa M 'come'

หนา naa R 'thick'

นา naa M 'paddy field'

หลา laa R 'yard (linear measure)'

ลา laa M 'donkey'