§77. Chinese enters history as the language written onto cattle or sheep scapulae and turtle plastrons by the diviners of an archaeological culture centred at 小屯 Xiǎotún (c.1200–1050 bce). The decipherment of these oracle bone inscriptions permits the identification of this culture with the 商 Shāng dynasty of traditional Chinese historiography (K. Chang 1980). Bone inscriptions record questions concerning weather, crops, warfare, and the regulation of the Shāng’s complex ritual life. Although they serve historians as invaluable sources (Keightley 1978), these short formulaic texts are difficult to profitably use in historical linguistics. Linguistic research on the language of the oracle bone inscriptions focuses primarily on syntax (Takashima 2000).

§1. Tibetan originated as the language spoken in the Yarlung valley, the cradle of the Tibetan empire (Takeuchi 2012a: 4). Together with the troops of this empire the Old Tibetan language colonized the entire Tibetan plateau, extinguishing almost all of the languages formerly spoken across that territory (Takeuchi 2012a: 6). Evidence is available for three such languages. Most famous is Źaṅ-źuṅ, the language of a pre-existing polity in West Tibet and the sacred tongue of the Bon faith. Źaṅ-źuṅ is preserved in one bilingual cosmological text, the Mdzod phug, and a number of short passages in Bon texts (cf. Martin 2010). The closest living relative of Źaṅ-źuṅ is the Darma language of Uttarakhand state in India (Martin 2010: 17–21, 2013). Aside from Źaṅ-źuṅ, samples of two Trans-Himalayan languages are preserved among the collection of documents found at Dunhuang. F. W. Thomas, who first published the manuscripts containing these two languages, confusingly dubs them ‘Źaṅ-źuṅ’ (Thomas 2011) and ‘Nam’ (Thomas 1948); there is no evidence to accept these identifications (Martin 2010: 10, 2013).

§44. Prior to the appearance of Burmese the Pyu and Mon languages had already been spoken and written in what is now Burma for several centuries (Krech 2012: 120–3, Bauer 1990). Burmese emerged as the language spoken by the Burman population in Pagan at the time of the Pagan dynasty (1113–1287 ce). The Burmese-speaking population entered Burma from the north; this is clear from the distribution of the Burmish languages, namely that they are all in the north, and further the Burmese word toṅ ‘mountain’ means also ‘south’, which suggests that at one point the Burmans lived to the north of the mountains. The oldest document in Burmese is the Myazedi inscription of 1113 ce (cf. Nishida 1955, Yabu 2006). Essentially all documents in Old Burmese are stone inscriptions recording land grants to Bu ddhist establishments (cf. Frasch 1996: 1–16).

§205. The preceding chapters trace the attested forms of Tibetan, Burmese, and Chinese backwards in time to the greatest extent that currently appears possible. This final chapter compares the results of these three exercises. If all developments in the three languages followed exceptionless phonological patterns with no interference from analogy and lost morphology a few scant remarks would suffice to point out that the backward projection of each of the three languages leads to the self-same result. The true situation is far less elegant. The Trans-Himalayan family is an ancient and ramified one; the three languages studied here offer only fragmentary glimpses of the proto-language.

The goal of this work is to present the sound laws relating Tibetan, Burmese, and Chinese, and to reconstruct the linguistic unity from which these three languages descend, so far as current knowledge permits. Tracing the development of etyma from their primitive origins into the living tongues of today would bring the narrative satisfaction of accompanying a hero through his struggles, but it is dishonest to present historical phonology as the trials of reconstructed forms progressing through sound changes towards an ultimate destiny in history. The end of this journey, the attested corpus of related languages, is fixed, but the original linguistic unity is the protean and mercurial product of research. Rather than presenting reconstructions picked out of the air and discussing their development, I present sound changes in reverse chronological order. I subsequently reiterate these sound changes in chronological order, so that, after seeing how the reconstructions are arrived at, one can see how it is that the reconstructed forms become the attested forms.