In the first version of the Sino-Austronesian theory the Tibeto-Burman (TB) languages were incorrectly excluded from the relationship. In the current version (Sagart 2005), the Sino-Tibetan family (Chinese, TB) is a valid taxon and Sino-Tibetan is a sister group to Austronesian. The homeland is thought to have been in NE China in the 5th millennium BCE when domesticated foxtail millet and rice appear together in Neolithic sites.

The published evidence for Sino-Tibetan-Austronesian (STAN) consists of sound correspondences on 61 and 14 items of, respectively, basic and cultural vocabulary. In addition several morphological processes are shared. The sound correspondences relate the last syllable of Austronesian, and the unique syllable of Chinese and TB words, as in PAN *punuq ‘brain,’ OC 腦 nˁuʔ ‘brain,’ PTB *s-nuk ‘brain,’ where the correspondences are n-n-n, u-u-u and q-ʔ-k. Initial pu- is not accounted for.

Assuming PSTAN had words of 1, 2 or 3 syllables with final stress, it is reasonable to expect that evolution to monosyllabism in ST would have led to the preservation of final syllables and to the reduction or loss of non-final vowels and consonants. However, initial consonants might be preserved, for instance when they formed natural clusters with medial consonants. Indeed penultimate syllable initial consonants are retained in Tibeto-Burman when the Austronesian medial consonant is *-R-. Compare the following forms:

A variety of situations in which PSTAN penultimate syllable initial consonants were preserved in ST will be discussed.

References
Matisoff, J. 2003. Handbook of Proto-Tibeto-Burman. Berkeley, Los Angeles, London: University of California Press.
Sagart, L. 2005. Sino-Tibetan-Austronesian: An updated and improved argument. In The peopling of East Asia: Putting
together archaeology, linguistics and genetics, ed. by L. Sagart, R. Blench and A. Sanchez-Mazas, 161-176.
London: Routledge.

In this paper, I will discuss the linguistic interaction in South China with concentration on the selected vocabulary items shared by Chinese, Tai and Miao-Yao. Although the three language groups have been sometimes put together as a branch of the grand Sino-Tibetan stock, especially in mainland China, scholars nowadays prefer to leave them in separate families. Leaving aside the issue of genetic relationship, one finds a strikingly large number of shared lexical items among these languages. I will discuss how these shared items provide us with very useful evidence of ancient sounds, phonological development as well as historical contact. The nature of their relationship and the ancient world and societies in which they used to live together, as reflected by the languages, will be explored.

The discussion will take into account the recent progress made in historical Chinese linguistics as well as the new systems of Proto-Tai and Proto-Miao-Yao as reconstructed by the author.

In this presentation I will survey the main obstacles that the recent presentations of the broader ‘Altaic’ hypothesis (Robbeets 2005, 2010), and the narrower Korean hypothesis (Unger 2009) are facing. My main objections to these hypotheses can be summarized as follows:

The partisans of these hypotheses pay lip service to the regularity of phonetic correspondences, but in actual presentation the irregularities rule supreme to the extent that every single etymology has to be salvaged in its own unique way.

There are also significant gaps, e.g., there are no good Japonic-Korean etymologies for PJ *y-, *w-, or *n-. Gaps are typical for contact relationships, but not for true genetic ones.

One of the greatest vices of long-range comparativists, namely the failure to explain unaccounted for segments is carried by Robbeets ad absurdum in her proposal that only the first CVC sequence in a given word is sufficient for proving a genetic relationship, and all the rest can be ignored (Robbeets 2005:422). This results in an ad hoc morphological segmentation that introduces non-existent affixes and roots.

Given a clear tendency to rely on dictionaries and on the secondary literature rather than on texts and Japanese philological tradition, there are, not surprisingly, a great number of ghosts and miscitations. In other words, on many occasions the said hypotheses are built on unreliable data.

References
Robbeets, M. I. 2005. Is Japanese related to Korean, Mongolic, Tungusic, and Turkic? Turcologica 64. Wiesbaden:
Harrassowitz.
Robbeets, M. I. 2010. Trans-Eurasian: Can verbal morphology end the controversy? In Transeurasian verbal morphology
in a comparative perspective: Genealogy, contact, chance. Turcologica 78, ed. by L. Johanson and M.
Robbeets, 81-114. Wiesbaden: Harrassowitz.
Unger, J. M. 2009. The role of contact in the origins of the Japanese and Korean languages. Honolulu: University of Hawai‘i
Press.

At present, no one has offered a definitive demonstration that the Japanese language is related to Korean, Manchu, Tamil, or any other language to which it has been compared. Thus, technically speaking, Japanese is an isolate, a language whose ancestors and all their other progeny now appear to be extinct. But to say that Japanese is an isolate does not answer specific questions, such as when and from where the earliest speakers of Japanese came to their present range. On the contrary, to claim that it does so is very different from merely giving reasons to reject this or that purported answer.

Most linguists would agree that isolates in the provisional sense (e.g., Basque, Burushaski, Sumerian, Ainu) are few compared to languages with at least one living or attested relative. We have reasonably good estimates of the age of several large, highly diverse families of languages (e.g., Indo-European, Austronesian, Sinitic), and none is more than about 5,000 years old. Therefore, to say that Japanese is an isolate in the strong sense is to claim that it must be extraordinarily old, and that fate has annihilated every one of the other descendants of its ultimate ancestor over its immense lifespan. It is easy to imagine different prehistories in which such a situation might have arisen, but none gets support from either non-linguistic or linguistic data, which are reviewed in detail in this paper. In the case of Japanese, pertinent non-linguistic evidence is copious and particularly useful in limiting the scope of realistic linguistic hypotheses.

The genealogical relationship of Japanese with the Transeurasian languages is among the most disputed issues of historical comparative linguistics. The term “Transeurasian” is used in reference to a large group of geographically adjacent languages that share a significant number of linguistic properties and include at most five families: Japonic, Koreanic, Tungusic, Mongolic and Turkic. The major objections raised against the genealogical relatedness of these languages are first, that they do not have enough bound morphology in common, and second, that all similarities can be attributed to borrowing. In this presentation both objections will be examined, using the traditional comparative method as a basic tool.

The primary controversy about the relationship of Japanese with Korean and other Transeurasian languages boils down to the question of whether all shared forms are induced by language contact or whether some are the residue of inheritance. Given the nature of the disagreement, binary comparison of two adjacent families, for instance Japonic and Koreanic or, Turkic and Mongolic, will necessarily lead to confusion. Since adjacent languages usually stand in a high-contact relationship, borrowings will be misinterpreted as cognates and cognates will be mistaken as borrowings. Starting from a hypothesis that includes low-contact languages provides not only the best hope for solving the long-standing debate for the Transeurasian languages, it will also shed light on the nature of the relationship between Japanese and Korean.

Vovin (2010) attempts to refute the Martin (1966)-Whitman (1985) reconstructions of Proto-Japanese-Korean, proposing that the forms were actually copied from Korean into Japanese in early historical times. Whereas Unger (2009) rehabilitates the Martin-Whitman corpus by using the semantics of lexical items as an indication against borrowing, the approach taken here is to argue that common verb morphemes simultaneously shared by Japanese, Korean, Tungusic, Mongolic and Turkic are unlikely to be accounted for by borrowing.

References
Martin, S. E. 1966. Lexical evidence relating Korean to Japanese. Language 42(2):185-251.
Unger, J. M. 2009. The role of contact in the origins of the Japanese and Korean languages. Honolulu: University of Hawai‘i
Press.
Vovin, A. 2010. Koreo-Japonica: A re-evaluation of a common genetic origin. Honolulu: University of Hawai‘i Press and
Center for Korean Studies, University of Hawai‘i.
Whitman, J. B. 1985. The phonological basis for the comparison of Japanese and Korean. Ph.D. dissertation, Harvard
University.

This presentation will briefly review the current state of Japanese-Ryukyuan and Korean internal reconstruction. Internal reconstruction shows Proto-Japanese-Ryukyuan (PJ-R) and Proto-Korean (PK) to have had extremely similar phonological inventories, with no voicing contrast among consonants and a system of 6 to 8 vowels, depending upon the assumptions of the reconstruction. The presentation then reviews previous lexical comparisons proposed by Martin (1966) and Whitman (1985), reviewed by Vovin (2010). It identifies a small but reliable set of cognates between PJ-R and PK, including approximately 15 items on the standard Swadesh 100 word list, smaller than Martin and Whitman's earlier sets, but somewhat larger than Vovin's. It then introduces a set of cognate inflectional morphemes, including the root suffixes *-i ‘infinitive/converb,’ *-a ‘infinitive/irrealis’, *-or ‘adnominal/nonpast,’ and *-ko ‘gerund.’ The presentation further investigates the possibility that the tone systems found in varieties of both languages reflect the loss of one distinction in Korean (voicing) and two in Japanese (voicing and coda consonants).

The paucity of cognate vocabulary in Japanese and Korean is sometimes considered, by those who acknowledge the possibility of a genetic affiliation, to indicate a great time depth of separation. But this presentation argues that the relatively basic nature of the cognates, once loans are stripped away, and the amount of shared grammatical material argue for a scenario favored by many researchers in Japan: introduction of new vocabulary due to contact. On this scenario the material shared with Korean has substrate status. It is impossible at our present level of knowledge to identify a superstrate source, but Austro-Tai (Ostapirat 2005) remains a promising possibility.

References
Martin, S. E. 1966. Lexical evidence relating Korean to Japanese. Language 42(2):185-251.
Ostapirat, W. 2005. Kra-Dai and Austronesian: Notes on phonological correspondences and vocabulary distribution.
In The peopling of East Asia: Putting together archaeology, linguistics and genetics, ed. by L. Sagart, R. Blench
and A. Sanchez-Mazas, 107-131. London: Routledge Curzon.
Whitman, J. B. 1985. The phonological basis for the comparison of Japanese and Korean. Ph.D. dissertation, Harvard
University.
Vovin, A. 2010. Koreo-Japonica: A re-evaluation of a common genetic origin. Honolulu: University of Hawai‘i Press and
Center for Korean Studies, University of Hawai‘i.

Japanese is often referred to as a language that has not been convincingly demonstrated to be related to any other language, but this view considers neither the Ryukyuan languages nor the language of Hachijō Island whose relationship with Japanese is beyond any doubt, to the point that they have often been treated as Japanese dialects. Though the recognition of several “Japonic” languages does not solve the problem of their relationships with other languages, it opens new perspectives.

A comparison of Japanese and Ryukyuan reveals that they exhibit mutually exclusive innovations, and thus form two sister branches. Moreover, though the Ryukyuan languages are very innovative in some aspects and some evidence indicates that the settlement of the Ryukyu Islands happened quite recently, Ryukyuan exhibits many lexical, phonological and grammatical features that can only be explained if we assume Ryukyuan and Japanese split before the eighth century.

If Ryukyuan is not a daughter of Old Japanese but a sister branch to it, the logical conclusion is that the Ryukyuan data is at least as important as Old Japanese for Japanese historical linguistics. Traditionally, most studies have relied exclusively on philology and internal reconstruction based on Old Japanese, and the vast data mine of the Ryukyuan languages remains largely untapped. It thus seems that a proper comparative reconstruction of Proto-Japonic taking into account data from the whole family is needed more than ever as a preliminary step for further research on the prehistory of the Japanese language.

Indeed, a thorough comparison of Ryukyuan and Japanese leads to modifications in the reconstruction of Proto-Japonic phonology, morphology and lexicon, which bears importance in the search for external relatives of Japonic.