Discussion on Nguyen Tai Can’s hypotheses about early Tai loanwords in Vietic Mark J. Alves (Note: This is a prepublication English version of the Vietnamese-language chapter “Thảo luận về một vài điểm trong giả thuyết của Nguyễn Tài Cẩn về các từ mượn tai ở Vietic” in the book Nguyễn Tài Cẩn Tư Tưởng, Tác Phẩm Và Kỷ Niệm [Nguyen Tai Can - Thoughts, Works, and Memories], ed. by Nguyễn Hồng Cổn, Trần Trí Dõi, Trịnh Cẩm Lan, and Dương Xuân Quang, 222-251. Hanoi National University Press. A Vietnamese-language video is available at https://youtu.be/7Vu3_7K7jH8.) Contents 1. Introduction ............................................................................................................................................... 1 2. Preliminary re-evaluation of the 40 items ................................................................................................. 2 2.1 Austroasiatic ....................................................................................................................................... 5 2.2 Chinese / Sinitic .................................................................................................................................. 6 2.3 Regional words ................................................................................................................................... 7 2.4 Words with unknown etymological origins ........................................................................................ 9 2.5 Vietic ................................................................................................................................................. 10 2.6 Tai Loanwords .................................................................................................................................. 10 3. Other Tai-Vietic words ........................................................................................................................... 11 5. Summary of a new perspective ............................................................................................................... 13 References ................................................................................................................................................... 14 Abstract: Nguyen Tai Can’s 1995 book “Giáo Trình Lịch Sử Ngữ Âm Tiếng Việt “A Textbook of Vietnamese Historical Phonology” contributed a systematic overview of the history of Vietnamese phonemes and other aspects of Vietnamese language history. In his book (1995:321-323), he listed about 40 words which he proposed are ancient Tai loanwords in Vietnamese. He hypothesized a sociocultural scenario in which Vietic peoples learned agricultural practices—including wet-rice production, animal husbandry, and others—from Tai groups. 25 years later, we can now review these words as well as relevant archaeological literature. The words he proposed as loanwords have several different origins, including five Proto-Tai words, but also various native Austroasiatic and Vietic words, early Chinese loanwords, regional Southeast Asian words, and words of unknown origin. We have identified twelve Tai loanwords that still support early language contact between Vietic and Tai. The lexical and archaeological data show that Tai and Vietic groups were in sociocultural contact, did trade, and shared some agricultural practices. However, Austroasiatic groups—and thus Vietic groups—were rice producers, and proto-Vietic has a rich rice-producing vocabulary (over 20 Proto-Vietic relevant terms in Alves 2020:xxxii). As a result, in some cases, the direction of borrowing of Tai and Vietic words cannot always be determined. Key words: loanwords, language contact, Vietic, Tai 1. Introduction Professor Nguyễn Tài Cẩn published Giáo Trình Lịch Sử Ngữ Âm Tiếng Việt “A Textbook of Vietnamese Historical Phonology” in 1995, over 25 years ago. This book is a very thorough study of historical linguistic origins and developments of Vietnamese phonemes (i.e., onsets, vowels, codas, and tones). He utilized a tremendous amount of comparative Vietic data (Vietnamese, Muong, Pong, Chut, and other Vietic languages) that was available at that time, in addition to Chinese historical phonological data, since he had authored the seminal work Nguồn Gốc và Quá Trình Hình Thành Các Đọc Hán Việt (The Origins and Process of Formation of Sino-Vietnamese Readings) (1979). His books contain consistent, rigorous historical linguistic methodology that allow further research (cf. Nguyễn Văn Lợi 2016). Near the end of the book (1995:321-323), he listed about 40 Vietnamese words which are mostly related to agriculture, and he hypothesized that these are Tai-Kadai loanwords. These are listed in Table 1 (English glosses are provided to specify semantics). He suggested that Tai-Kadai groups experienced cultural developments earlier than Vietic groups did due to prior cultural contact with groups from north in China. He also hypothesized that Vietic groups moved from mountainous areas to the Red River Delta after Tai groups. He considered both linguistic data and ancient historical legends to suggest that Tai groups shared agricultural practices with Vietic groups. These ideas present an interesting hypothesis of sociocultural borrowing by Viet-Muong groups from Tai groups, including food production methods, animal husbandry, tools, and produce. Table 1: NTC’s hypothesized Tai loanwords in Vietic to be re-evaluated in this paper Animal vịt ‘duck’; bò ‘cow’; gà ‘chicken’; ngan ‘goose’; bồ câu ‘pigeon’; con đực ‘male husbandry animal’; (con) mái ‘hen’; gáy ‘to crow’ Farming mương ‘ditch’; đồng (ruộng) ‘wet rice field’; đâm (gạo) ‘to pound (rice)’; phai ‘to fade (of material or fragrance)’; rẫy ‘dry field’; bừa ‘rake’; cuốc ‘hoe’; vải ‘cloth’; bứt ‘pluck’; cỏ ‘grass’; gạo ‘rice; uncooked’; phân ‘dung’; cái nong ‘type of basket’; bánh ‘pastry’; mùa ‘season’ Produce mận ‘plum’; cam ‘orange’; tre pheo ‘bamboo’; quít ‘tangerine’; tỏi ‘garlic’; ngọn ‘top; peak’; nụ ‘bud’; mền ‘blanket’; bửa ‘to cleave (a tree)’ Lifestyle of bè ‘raft’; ao ‘pool’; dầm ‘to soak’; sông ‘river’; chèo ‘paddle’; bơi ‘to swim’; lội ‘to lowland fields wade’; lụt ‘flood’; đỉa ‘leech’ However, 25 years after that publication, there is a much larger foundation of both linguistic and archaeological data. In the last 15 years, several massive digital lexical databases and reconstructions have become available for Austroasiatic, Proto-Tai, Old and Middle Chinese reconstructions, and other large resources on regional languages (Tibeto-Burman, Hmong-Mien, and Austronesian) (see Alves 2017 for a list and explanation). Moreover, archaeological information of the Red River Delta has progressed tremendously, especially in the last decade. Therefore, we can now re-evaluate these words’ etymological histories based on comparative lexical data and more understanding of historical phonology of Vietic to which Professor NTC contributed. 2. Preliminary re-evaluation of the 40 items NTC listed the Vietnamese words, but he did not provide comparative lexical evidence, such as Proto-Tai reconstructed words or data from specific Tai languages, reconstructions of Old or Middle Chinese, or other languages in the region. At the time of his publishing, few digital resources were available. This situation changed in the first decade of the 21st century, and it is essential to use all available resources to the fullest extent. In order to evaluate the items NTC’s proposed to be Tai loanwords, I checked several sources. The sources listed in Table 2 are used throughout this paper. Moreover, I have my own databases of several likely hundred early Chinese loanwords and several hundred Vietic reconstructions. The early Chinese loanwords are words borrowed before the era of Late Middle Chinese and Sino-Vietnamese. This was still during the late Vietic period but before Proto-Viet-Muong. Secondary resources include Tibeto-Burman in STEDT and Thurgood’s Proto-Chamic reconstructions. Thurgood (1999:307-334) notes over one hundred Chamic etyma that were probably borrowed from Bahnaric and/or Katuic into Chamic. Some of these are Austroasiatic etyma and therefore are also in Vietic. Those words which are related to this study include *buc ‘to pluck’, *luay ‘to swim’, *kruac ‘citrus’, *kaduŋ ‘pocket/pouch’, *krɔːŋ ‘river’, and *kapa:s ‘cotton’. We here list these probable loanwords into Chamic to prevent future confusion. I have only occasionally checked reconstructions of Proto Proto-Hmong-Mien (Ratliff 2010) for evidence of regional words since Hmong-Mien speakers were not in the region two thousand years ago. As for the The Xiaoxuetang (Tiểu Học Đường) Character Study Database, I consulted it for Chinese words lacking Old Chinese reconstructions. Finally, I occasionally checked for Chinese words in the massive online Chinese Text Project to see how earlier they appeared in ancient Chinese texts and whether they appeared in this early period. Table 2: Databases of Comparative Linguistic Evidence • The Mon-Khmer Etymological Dictionary (MKED) • The Munda Etymological Dictionary • The Old and Middle Chinese reconstructions of Baxter and Sagart (2014) or Schuessler (2006) • Proto-Tai of Li (1977) and Pittayaporn (2009) • The Sino-Tibetan Etymological Dictionary (STEDT) • Proto-Austroasiatic reconstructions of Shorto (2006) • Proto-Vietic reconstructions of Ferlus (2007, available in the MKED) • 小學堂文字學資料庫 [The Xiaoxuetang Character Study Database] • The Chinese Text Project • Digital Dictionaries of South Asia The phonological histories of Sinitic, Tai, and Vietic are certainly complex. Combining these histories to study loanwords in ancient periods is obviously difficult and prone to confusion. However, for this study, we focus as much as possible on proto-language reconstructions, which increases the degree of certainty of hypotheses. It also simplifies the process of evaluation of etymologies, regarding both inclusion and exclusion of hypothesized loanwords. The minimum requirement is that all onsets, vowels, codas, and tones in words either have the same phonemes or expected phonological patterns. In this study, I frequently refer to these issues in NTC’s book. Because the modern words are all monosyllabic and mostly with CVC syllable structure (except some onset clusters in Tai languages), the possibility of chance similarity is high. However, Vietic and Tai reconstructions often have complex onsets and some presyllables, which provide crucial comparative data. Also, these groups were in sociocultural contact in the past, and we can consider ethnohistorical issues when evaluating the possibility of words being borrowed. I will not present the entire system of sound changes in Vietnamese that NTC presented in his work (as has Michel Ferlus 1992, 1997, 2014a, etc.). In this paper, when segments and tone categories are the same, I will simply mention that is the case. When they differ, I will refer to recurring sound changes that NTC describes in his book. One instance of a historical phonological pattern is Vietnamese ‘ưa’ coming from earlier *a (NTC 1995:180-183). Examples of this sound change in this paper is seen in bửa ‘to cleave/split,’ a Vietic etymon, and bừa ‘rake/harrow’, an early Chinese loanword, both of which are reconstructed with *a. An important tendency is for earlier presyllables to become single consonants ‘g/gh’, ‘v’, ‘d’, ‘gi’, and ‘r’ (Ferlus 1982). These presyllables occurred in both Austroasiatic etyma and Old Chinese words as reconstructed by Baxter and Sagart (2014). This feature helps us to determine one etymological origin and exclude other possibilities. One crucial factor in etymological studies in Vietnamese is tone correspondences. The origins of the four basic tone categories are primarily based on the final segments of syllables, as per Table 3. For a summary of these matters, see Alves 2018. Tones A and D do not provide information about the timing of borrowing of words. They could be from any period. Consonants and vowels are useful in determining timing of borrowing but with less precision. One useful exception is that Sino-Vietnamese words with sonorant initials (i.e. nasals and liquids) have the upper register ngang tone (NTC 1979:292), while early Chinese loanwords with sonorant initials have the expected low register huyền tone. Tones B and C indicate early borrowing at a time when final *-ʔ and *-s/h were retained in both Old Chinese (or possibly even into early Middle Chinese tones had developed) and Vietic. The precise timing is not certain, but at the very least, loanwords with these tones could have been borrowed before the late 1st millennium CE, or even the BCE era. These tone categories are generally consistent among all the languages in this study, but there are instances of irregularities. We accept that in the process of borrowing words, second-language factors can result in phonological changes and also sometimes phonological irregularities. Also, lexical borrowing can be affected by dialectal variation of the various languages involved and sound changes in process (e.g. the process of the loss of final laryngeal sounds, etc.). Therefore, some loanwords may have unexpected phonological and/or semantic features, but we still look for patterns as much as possible. Table 3: Tone categories among the languages Language A Old Chinese *-ø or *final sonorants Vietic *-ø or *final sonorants Tai A Middle Chinese 平 píng Early Chinese loanwords ngang/huyền in Vietnamese Sino-Vietnamese ngang (voiceless onsets & sonorant onsets) huyền (voiced nonsonorant onsets) B *-ʔ *-ʔ C 上 shǎng sắc/nặng hỏi-ngã C *-s/h *-s/h B 去 qù Old Chinese: hỏi/ngã Middle Chinese: ngang/huyền sắc/nặng D *-p/t/k *-p/t/k D 入 rù sắc/nặng sắc/nặng The following are the primary results of this study. They are discussed in detail in following sections. There may be other individual Tai languages with words that are similar to Vietnamese ones, but we have not yet found them. Regardless, words in single languages—which could be modern loanwords from Vietnamese into minority group languages—cannot be used to identify ancient loanwords in Vietnamese. • Tai loanwords: Only five of the forty words NTC proposed to be Tai loanwords in Vietnamese appear to be probable loanwords from Proto-Tai. These five are the bolded words in Table 1, and they are discussed in Section 2.6. To these, we add eleven more possibilities in Section 3. As per NTC’s hypothesis, the semantic domains involve agriculture and trade. We have also identified two words implying intermarriage, but these two words are in other Vietic languages, not Vietnamese. • Native etyma (Austroasiatic and Vietic): About a dozen of the words are native retentions (seven Austroasiatic etyma in Section 2.1 and six Vietic etyma in Section 2.5). We found no evidence of related proto-Tai etyma or words that are widespread in Tai. • Early Chinese loanwords: Ten of the words are likely early Chinese loanwords (i.e., before Late Middle Chinese) in both Tai and late Vietic. These are discussed in Section 2.2. • Regional terms: Six words are regional in Southeast Asia and Southern China (three or more language families) and either of uncertain origin or from Sanskrit/Pali. These are discussed in Section 2.3. • Uncertain origins: For several other words, in Section 2.4, available data does not support a Tai origin or other origin, so these must be considered Vietnamese words of unknown origin. In the subsequent sections, based on data and methods described above, I discuss each of the etymological origins of the words that NTC noted. Each section has lists of the words with comparative data and notes. The comparative notes are not exhaustive since the topics would all require much more extensive research. For now, the goal is to provide brief statements with acceptable hypotheses that can be further evaluated. 2.1 Austroasiatic Seven of NTC’s proposed words are Austroasiatic etyma. These are listed in Table 4. The rich comparative data in the online Mon-Khmer Etymological Dictionary clearly identifies these as Austroasiatic etyma. Except for rẫy ‘dry field’, all the words in this category also have Vietic reconstructions. As for rẫy, it is not widespread in Vietic, but because it is widespread in Austroasiatic, and it is in Vietnamese, it should be considered a native word, unless data suggests another hypothesis. The Austroasiatic language family has 13 branches, including Vietic. For the words bửa ‘to cleave (a tree)’, bứt ‘pluck’, gạo ‘rice (uncooked)’, lội ‘to wade’, rẫy ‘dry field’, the Austroasiatic etyma are attested in 8 to 11 branches of Austroasiatic, often including Munda. Thus, these words have deep histories and should not be considered Tai loanwords into Vietic. In fact, some of these words could be loanwords into Tai in an early period. Sagart (2021) has come to a similar conclusion about the borrowing of Austroasiatic gạo ‘rice (uncooked)’and rẫy ‘dry field’ into Tai. We can add the Austroasiatic word for ‘citrus’, as discussed in Section 2.2. The following are additional explanations of the words in Table 4. • For bửa ‘to cleave,’ there is no Tai reconstruction, while the Proto-Austroasiatic and Proto-Vietic forms are reasonable sources for the Vietnamese word. As mentioned above, this is a case of Vietnamese ‘ưa’ from Vietic *a, as NTC (1995:180-183) described. • For bứt ‘to pluck,’ there is no Tai form, but there is a Proto-South-Western Tai reconstruction *ɓitD ‘to squeeze, twist’ (Jonsson). However, the initial implosive *ɓ would have become /m/ in Vietnamese (cf. NTC 1995:19-23), and the semantics are close but rather general. This is likely an example of chance partial similarity. • The word đỉa ‘leech’ only has a related reconstruction in Proto-Katuic, but both the Vietic and Katuic reconstructions share final *-h, and Vietnamese ‘ia’ comes from *ɛ (cf. NTC 1995:152-153), as is in the Proto-Katuic reconstruction. Moreover, the reconstruction of Proto-Tai ‘leech’ is clearly unrelated. • Whether lụt ‘flood’ is related to the Austroasiatic etymon is not certain, but the Katuic reconstruction appears related to the Vietic one. Both reconstructions are completely different from the Proto-Tai word. • The Austroasiatic etymon for gạo ‘unhusked rice’ has a *r- sesquisyllable (leading to the fricative ‘g’ in Vietnamese, as noted in Section 2). Li (1977) did not reconstruct the Proto-Tai etymon with a presyllable, while Pittayaporn (2009) did, but not a specific segment. This is very possibly a loanword into Tai; in fact, it is not impossible that Vietic was the source of this word in Tai. However, this is a major question that will require careful archaeological research to answer, but it may be unanswerable. • The Austroasiatic etymon for lội ‘to wade’ looks similar to the Tai form. However, the tone category is very different: a final glottal stop in the etyma in Austroasiatic and Vietic, which is connected to a Type B tone in Vietnamese. However, the Tai word has a Type A tone, meaning it had no final glottal stop. Whether there is a relationship between the Tai and Austroasiatic etyma is uncertain, but it should be considered an Austroasiatic etymon in Vietnamese, not a loanword. • rẫy ‘dry field’ has no reconstructable Vietic word because it is not widespread in Vietic. This Vietnamese word has a tone that matches the Tai tone category, not *-ʔ of the Austroasiatic word. However, this word is very widespread in Austroasiatic and is even in Munda, which suggests it is quite old in Austroasiatic. The Proto-Austroasiatic reconstruction has an onset cluster, as it does in most Austroasiatic languages, so it could be older than the Tai etymon. It hypothetically could have been borrowed into Tai because Nam Á groups were rice-growers (Higham 2017), and the Phung Nguyen and Dong Dau cultures of the Red River region have remains of cultivated rice (Hán 2009:177). In modern times, many Austroasiatic groups practice dry-rice agriculture. This is a major anthropological issue and cannot be fully resolved in this study. Table 4: Austroasiatic etyma in Vietnamese with Comparative Tai Data Word Austroasiatic Vietic Tai bửa ‘to cleave *buuh ‘to split’ (in 9 branches) *bah/pah NA (a tree)’ bứt ‘pluck’ *pic, *piic, *piəc, *pəc ‘to pluck’ *pəc NA (in 7-9 branches) đỉa ‘leech’ NONE (cf. Proto-Katuic ‘leech’ *l-dɛh / l-tɛh *pliŋA (Li 1977) *ʔadeh, *deh ‘water leech’) gạo ‘rice *rk[aw]ʔ ’husked rice’ (in 11 *r-koːʔ *C̬.qawC ‘rice’ (Pittayaporn (uncooked)’ branches, including Munda) 2009); *xəuC (Li) lội ‘to wade’ *lujʔ ‘to wade, to swim’ (in 9 #lo:jʔ *lo:jA ‘to swim’ (Pittayaporn branches) 2009) lụt ‘flood’ *laac, *ləc, *ləəc ‘to be flooded, #lu:t *thuemC (Li 1977) to sink’ (in only 4 branches, but cf. Proto-Katuic *luut) rẫy ‘dry field’ *sreʔ ‘field’ (in 9 branches, NA *rɤjB (Pittayaporn 1977) including Munda and Nicobaric) 2.2 Chinese / Sinitic Haudricourt (1954) noted shared early Chinese loanwords in both Tai and Vietnamese. NTC himself mentioned in a footnote (NTC 1995:322) that some of the words could be Chinese loanwords, while still claiming these were borrowed into Viet-Muong through Tai-Kadai. This scenario is not impossible, and many of the same early Chinese loanwords are in both Tai and Vietnamese. However, based on available lexical data, I have observed that many of the types of early Chinese loanwords into Vietnamese and Tai differ significantly. Moreover, hundreds of early Chinese loanwords in Vietnamese are not in Proto-Tai or even in other Tai languages. That suggests that, even though Vietic had language contact with Tai in the pre-Qin period, Tai and Vietic were distinct speech communities when they were in contact with Sinitic. Unless clear evidence shows otherwise, we must assume that these words were borrowed directly from Sinitic. The following are notes on the words in Table 5. In all cases (with the exception chèo ‘oar/to row’ which has no available reconstruction), the phonological correspondences between the early Chinese loanwords in Vietnamese match those of both Chinese reconstructions and later Sino-Vietnamese pronunciations. • In bừa ‘rake/harrow,’ *a becoming ưa is a common sound change in Vietnamese historical phonology, as NTC notes (1995:180-183). This matter indicates this word was borrowed from the Chinese word with *a, not via Tai with a different vowel. See Alves (2014) for discussion of this loanword. • We speculate that chèo ‘oar/to row’ is an early Chinese loanword, but its history is not completely certain. It first appears in Western Han era texts at the assumed time of early Sinitic contact first with Tai and later with Vietic groups. NTC notes that Vietnamese ‘e’ comes from *ɛ (NTC 1995:152-153), as in the Tai reconstruction, not *a, as in most modern Chinese languages, but with ɛ in a few dialects in Guangdong. However, there is no Old Chinese reconstruction for the word, so we cannot yet assume the word had *a in Sinitic. Also, the palatal initial ‘ch’ in Vietnamese matches the palatal onsets in Yue than the Proto-Tai aspirated dental sound (cf. NTC 1995:46-48). While the features are not perfect, we consider this an early Chinese loanword for now. • đực ‘male (animal)’ has an onset ‘đ’ that matches the Chinese reconstruction *d better than the ProtoTai dental aspirate. NTC (1995:41-46) notes that Vietnamese ‘đ’ comes from *d or *t, not an aspirated consonant. • We propose that mền ‘blanket’ is related to the Chinese word for ‘cotton’ or ‘cotton-padded’. We admit that the semantic shift does not yet have concrete evidence, but the word patterns • • • phonologically with expected early Chinese loanwords (i.e. the huyền tone instead of ngang with ‘m’, and ‘ê’ compared to Sino-Vietnamese ‘iê’). The association between ‘cotton’ and ‘blanket/quilt’of 棉 mián is part of the semantic range of this word in Chinese. Moreover, there is no Proto-Tai reconstruction for ‘blanket’ to be the source in Vietnamese. A semantically related word is vải ‘cotton’, as discussed in Section 2.3 on regional words. Ngan ‘goose’ has a complex and vague history in the region of China and Southeast Asia (see Alves 2015). However, in general, this word has the expected phonological traits of an early Chinese loanword in Vietnamese. The Tai reconstruction with initial *h shows a different path of borrowing from Chinese. Phân ‘feces’ clearly has a Middle Chinese syllable shape, the same as the standard Sino-Vietnamese pronunciation, but an early Chinese tone. quít ‘tangerine’ vaguely resembles the Austroasiatic reconstruction *kruuc ‘citrus’, but it matches the Old Chinese reconstruction better. As NTC notes (1995:108-114), *kr would result in Vietnamese ‘s’ As NTC notes (1995:108-114), *kr would result in Vietnamese ‘s’, meaning quýt is not an Austroasiatic etymon. An interesting matter is that the Proto-Tai form appears similar to the Austroasiatic reconstruction. The direction of borrowing is uncertain, but the word is attested in 7 branches of Austroasiatic. It is in Nicobaric, which indicates that it has a deep history in Austroasiatic and could be a loan from Austroasiatic into Tai. Hopefully, archaeological information can help answer this question (e.g. Fuller et al. 2018). Table 5: Early Chinese loanwords in Vietnamese and Tai Vietnamese LSV Chinese OC bánh ‘pastry’ bính *peŋʔ 餅 bǐng bừa ‘rake’ bà *[b]ˤra 耙 pá chèo ‘oar/to trạo NONE ([tsau] or 棹 zhào row’ [tʃau] in Yue 小學 堂) h cuốc ‘hoe’ quắc 鐝/钁 jué NONE ([k uak] in Yue 小學堂) đực ‘male đặc *dˤək 特 tè animal’ mền ‘blanket’ miên NONE ([min] in 棉 mián Yue 小學堂) ‘cotton’ ngan ‘goose’ nhạn *C.[ŋ]ˤrar-s 雁 yàn MC pjiengX bae NONE Tai *pi̥ ɛŋC (Li 1977) *fɯeA (Li 1977) *thɛuA (Li 1977) NONE NA dok *thɯkD (Li 1977) NONE NA ngaenH *ha:nB (Pittayaporn 2009) NA phân ‘feces’ phấn/phẫn 糞 fèn *p[u]r-s pjunH quýt ‘tangerine’ quất 橘 jú *[s.k]ʷi[t] kjwit tỏi ‘garlic’ toán 蒜 suàn *[s]ˤor-s swanH *kruutD ’citrus, a kind of’ (Li 1977) NA 2.3 Regional words Six of the words appear in languages throughout Southeast Asia and southern China and therefore cannot be considered Tai loanwords. It is extremely difficult or even impossible to describe the histories of these words. Five of these words have no clear origin, while the word vải ‘cloth’ has a Sankskit/Pali origin, but the precise history is not yet understood (Headley et al. 1977 notes this origin in Khmer). • Bè ‘raft’ (typically of bamboo) can be tentatively reconstructed in Vietic as #bɛ: (# indicates my own reconstruction) based on data in several Vietic languages. It has a comparable Chinese word, 簰 • • • • • pái/bēi ‘bamboo raft’, but this character does not appear in Chinese texts until the Tang Dynasty, the Middle Chinese period, and cannot be reconstructed in Old Chinese. The word is scattered through Tai, Austroasiatic, and Tibeto-Burman. With a very simple syllable shape, it is not possible to exclude chance similarity, but we assume these are related. Archaeological data is needed, but I have not found such research. Cam ‘orange’ also has a complicated history. Schuessler (2006:249) speculates an Austroasiatic origin, but in the MKED, this form is only in Vietic, so that hypothesis is not supported. This is a standard Sino-Vietnamese reading of the Chinese character 柑 gān. The Proto-Southwest Tai reconstruction is *khwaamA (Jonsson 1991) According to Fuller et al. (2018:33-24), the appearance of that word in ancient Chinese texts is in the Han Dynasty, and the geographic range of this word is in southern China and northern Vietnam. Thus, it is reasonable to consider this a word from Tai or even Vietic that was borrowed into Sinitic, but there is no way to confirm the direction of borrowing. For reference, see the note about quít ‘citrus’ in Section 2.2. For (bồ) câu ‘pigeon’, there are reports of pigeon domestication noted in Chinese texts from the Eastern Han Dynasty (Lu 2014). Thus, it is possible that the Chinese people are the earliest domisticators of pigeon domestication in the region. However, such birds are throughout the region, I have not found archaeological studies of pigeon domestication in the region, and the word is onomatopoeatic, meaning chance similarity is more likely. All of this further complicates the etymological origin. Hopefully, archaeological data can clarify the matter in the future. Nong ‘basket’ is another word with a complex situation. The Austroasiatic reconstruction has a presyllable, suggesting a longer history than the monosyllabic Tai reconstruction. The Proto-Tai word has tone C, which suggests a final pharyngeal sound that is lacking in the Austroasiatic form, and also nong in Vietnamese, with a Type A tone. Thus, the direction of borrowing between Tai and Austroasiatic cannot be determined conclusively. However, we speculate that the Austroasiatic word is older and that Austroasiatic groups had a well-developed weaving tradition. This is noted in archaeological literature such as a Đồng Đấu era site (Hoàng 2003:172-180), and of course, the word đan ‘to weave’ is deep in Austroasiatic since *t1aaɲ is seen in all 13 branches. Regardless, nong is not only an issue for Vietnamese and Vietic but also of Austroasiatic in general, so it is certainly not a Tai loanword into Vietic specifically. sông ‘river’ has been claimed to be an instance of an Austroasiatic word borrowed into Chinese (e.g. Norman and Mei 1976:280), but there is no evidence that clearly supports this claim (see notes by Schuessler 2007:306). It has a regional geographic distribution (e.g. STEDT #2322 PTB *kl(y)u(ŋ/k) VALLEY / RIVER; https://stedt.berkeley.edu/~stedt-cgi/rootcanal.pl/etymon/2322), as Blench has noted (2015:6). It is seen in many Tibeto-Burman languages and even Proto-Chamic. It is not reconstructed in Proto-Tai. The OC and Tibeto-Burman forms have onset clusters. Thus, it is possible that these are older etyma. However, there are too many complications, and unless more data makes the situation clearer, this word must b considered a regional word with no certain source. vải ‘cloth’ ultimately comes from Sanskrit kārpāsa ‘made of cotton’, but perhaps not a direct loanword. The Vietic reconstruction is *k-pas, and the Austroasiatic reconstruction *kpaas is attested in 7 branches, including Munda. It appears that Sanskrit ‘cotton’ shifted to the meaning ‘cloth’ in vải, which has also happened in other Austroasiatic languages, such as Surin Khmer. In Section 2.1, we hypothesize that the Chinese word for ‘cotton’ is the source of mền ‘blanket,’ and Vietnamese also has the native word bông ‘cotton/flower.’ It appears these words related to cotton have all undergone some amount of semantic shifting within this small semantic domain. Table 6: Vietnamese words of regional origin in Southeast Asia and southern China Vietnamese Comparative Notes bè ‘raft’ • Tai *be:A (Pittayaporn 2009) • Chinese 簰 pái/bēi • Tibeto-Burman (Loloish, Qiangic, etc.) • Austroasiatic #bɛː/pʰɛː (in 7 branches) cam ‘orange’ • Vietic #ka:m • Late Han Chinese *kɑm (Schuessler 2007:249) • Southwestern Tai ‘tamarind’ *khwaamA (Jonsson 1991) (bồ) câu ‘pigeon’ • A common form [ku] in Old Chinese *ku (鳩 jiū, HV cưu), Tai languages, and Hmong-Mien (Alves 2015:10) nong ‘type of basket’ • Tai *ɗoŋC (Pittayaporn 2009) • Austroasiatic #C.ɗoːŋ (cf. Proto-Bahnaric *-ɗoːŋ; Proto-Katuic *kɗoŋ; Proto-Khasic *pduŋ; Proto-Vietic *ɗoːŋ) • Chamic *kaduŋ ‘pocket/pouch’ sông ‘river’ • Old Chinese *kˤroŋ (江 jiāng, HV giang) • Austroasiatic *ruuŋ, *ruŋ, *ruəŋ ‘channel, river’ (initial /k/ in Mon) • Tai (Lao kʰɔ̆ːŋ ‘the Mekong River’; Thai kʰǒoŋ ‘the Mekong river’) • Chamic *krɔːŋ ‘river’ vải ‘cloth’ • Sanskrit kārpāsa ‘made of cotton’ (Apte 1957-1959:563) • Austroasiatic *kpaas • Tai *faiC ‘cotton’ (Li 1977) • Chamic *kapa:s ‘cotton’ • Tibeto-Burman (various languages in STEDT) 2.4 Words with unknown etymological origins Because NTC did not provide comparative Tai evidence, I am sometimes uncertain which words in Tai he thought were the source of the Vietnamese words. In six instances, I have found no sources in ProtoVietic, Austroasiatic languages, or Old or Middle Chinese. As for Tai data, the I have checked the SEALANG Thai and Lao dictionaries as well. I have provided notes in the table. Until comparable words are located, these must be considered Vietnamese words without other etymological sources. Table 7: Words of unknown origin in Vietnamese Vietnamese Comparative Notes ao ‘pool’ no relevant data was found dầm ‘to soak’ cf. Proto-Tai *čumB ‘to soak’ (Li 1977) (problems with tone and vowel) (con) mái ‘hen’ cf. Lao mɛ̄ ː ‘female’ (insufficient data) mùa ‘season / 10th cf. Proto-Tai *mɯeB ‘time’ (Li 1977) (problems with tone, month harvest’ vowel, and semantics) nụ ‘bud’ no relevant data was found phai ‘to fade (of no relevant data was found material or fragrance)’ 2.5 Vietic Six of NTC’s proposed loanwords are reconstructed in Vietic and have no viable sources in Tai. In two instances, the words are only in Viet-Muong and Cuoi, so the words might not have a deep history. Of five of these, there are no comparable Proto-Tai reconstructions. Two partially similar Proto-Tai words have major phonological problems. • gà ‘chicken’ vaguely resembles the Proto-Tai etymon. However, that is a Chinese loanword which spread into Tai and Hmong-Mien (see Alves 2015a). The Vietic reconstruction has a presyllable, as attested by both Vietic languages and Khmer, which borrowed this word from Vietic when borrowing words for the zodiac calendar (Ferlus 2014b). Moreover, the vowel *a should come from *a (NTC 1995:168-170), not Old Chinese *e, Middle Chinese *i, or Tai *aj. • As for bơi ‘to swim’, I have found no Tai words that are similar. • Cỏ ‘grass’ is vaguely similar to the Proto-Tai word, but it has a different tone category. The vowel does not match since as NTC notes (1995:131-134), Vietnamese ‘o’ generally comes from *u or *o, not *a. This is a good example of chance similarity. • Gáy ‘to crow’ has a reconstruction with a presyllable. Besides the presyllable, which the Proto-Tai form does not have, the Vietnamese word has Tone C, while the Tai word has a Tone A category. The two reconstructions are too different to be considered related. • Ngọn ‘top peak’ is reconstructed in Vietic, but only Viet-Muong and Cuoi. However, there is no related Tai reconstruction, and I could not find comparable words in Tai languages. • As for pheo ‘type of bamboo’, the Proto-Tai reconstruction *phəiB ‘bamboo’ does not match the vowel or tone category of Vietic. Table 8: Vietic words in Vietnamese Vietnamese Proto-Vietic gà ‘chicken’ *r-ka: bơi ‘to swim’ cỏ ‘grass’ gáy ‘to crow’ ngọn ‘top peak’ (tre) pheo ‘type of bamboo’ #pɤ:j *kɔh *t-karʔ *ŋɔːnʔ (Viet-Muong and Cuoi) *p-hɛːw (Viet-Muong and Cuoi) Proto-Tai *kajB (Pittayaporn 2009) Old Chinese *kˤe, Middle Chinese ki (雞 jī) NA *ɣaA ‘grass’ (unrelated) – unrelated *χalA (Pittayaporn 2009); *xanA (Li 1977) NA *phəiB ‘bamboo’ (Li 1977) – unrelated 2.6 Tai Loanwords After excluding the other words, only five of the 40 words that NTC proposed have some supporting evidence, but not all of them have perfect sound correspondences. Nevertheless, they are considered reasonable hypotheses for the present time until additional data or new ideas can provide an alternate hypothesis. Of these, only mận ‘plum’ has a tone which demonstrates that this word must have been borrowed in a very early period. • Đâm ‘to pound (rice)’ has a syllable shape, tone, and semantics that are similar to the Proto-Tai word. However, in Tai, it means only ‘to pound’, while ‘to pound rice’ is a different etymon *zo:mC (Pittayaporn 2009). Why wasn’t that word borrowed into Vietic? Another weakness is that Vietnamese ‘â’, a short vowel, does not come from *a, as NTC notes (1995:177-180). Also, both the Vietic and Tai words have an onomatopoetic sound of hitting. It is also worth noting that Vietnamese has (at least) two other relevant words: giã ‘to pound’, which has no identifiable etymological source, and đập ‘to pound’, which is listed in Section 3 as possible additional Tai loanwords. We consider this a possible Tai loanword for now, but this could be chance similarity. • Đồng ‘field’ is problematic. It has the expected semantics and syllable shape, but the Tai tone B should correspond to a ngã tone, not a huyền tone. Also, there are other Vietnamese words for agricultural fields (e.g. bãi ‘field (general)’, ruộng ‘rice field’, nương ‘terrace field’, rẫy ‘mountain • • • field’ (see Section 2.1), etc.), so đồng might have the right semantics and be borrowable in addition to other similar words. Mận ‘plum’ is most likely a Tai loanword in Vietnamese. The semantics, syllable shape, and tone all have reasonable correspondences. There are scattered possible instances of this word in Austroasiatic languages (e.g. Palaung, Mang, Khsing-Mul, etc.), but its distribution is limited and thus a likely Tai loanword in these languages. I have found no similarly word forms in other language groups in the region (i.e. Tibeto-Burman, Proto-Hmong-Mien, Proto-Chamic). Plums were already domesticated by Chinese in the Han Dynasty (Wang 1982: 53, 206). Vietnamese has an early Chinese loanword mơ (cf. SV mai) from Old Chinese *C.mˤə (梅 méi). I have found no studies of the history of the practice among Tai groups in southeastern China, but based on the tone, this word could have been borrowed in the pre-Qin period. Mương ‘ditch/canal’ has clear semantic and phonetic correspondences with the Tai form. In vịt ‘duck’, the [v] onset could be the result of a palatal feature of the Tai word, perhaps due to the main vowel *i. This is not absolutely certain from the reconstructions. Still, there is a little archaeological evidence of duck domestication in southern China, potentially where Tai groups were (see Alves 2015). However, I currently believe we cannot know what direction the borrowing went. While Li (1997) reconstructs *v in the Proto-Tai phoneme system, Pittayaporn (2009) does not (but his Proto-Southwestern Tai does (Pittayaporn 2008)). If Proto-Tai had *v, it could have borrowed Vietic *vi:t with *v, but if it did not, a *p could be the result. Table 9: Words that NTC proposed as loanwords that have supporting evidence Vietnamese Vietic Tai đâm (gạo) ‘to pound (rice)’ *təm *tamA ‘to pound’ (Pittayaporn 2009) đồng (ruộng) ‘wet rice field’ #to:ŋ *doŋB (Pittayaporn 2009) mận ‘plum’ #mən *manC (Pittayaporn 2009) mương ‘ditch/canal’ *-mɨəŋ *ʰmɯəŋA (Pittayaporn 2009) vịt ‘duck’ *vi:t *pitD (Pittayaporn 2009); *pi̥ etD (Li 1977) 3. Other Tai-Vietic words Besides NTC’s proposed Tai loanwords in Vietic, we propose eleven additional items that indicate early Tai-Vietic language contact. Only eight of these have Proto-Vietic reconstructions, the other three are only in Vietnamese, and two of the Proto-Vietic words are not in Vietnamese. In a few cases, the direction of borrowing is uncertain, and we hope that more linguistic and ethnohistorical data can clarify these issues. The following are notes on the items in Table 10. • Kèn ‘reed flute/khene’ is culturally specific, and the khene instrument is widely considered to be a Tai cultural item. It is scattered through Austroasiatic languages in the region (Bahnaric, Katuic, Khmer, Khmuic), apparently a widespread loanword. However, it is not possible to determine when this word was borrowed. A Dong Son figurine of a man playing of flute (Bảo Tàng Lịch Sử Quốc Gia 2014:182) is interesting, but more linguistic and ethnohistorical data is needed to clarify the situation. • In available data, lam ‘kind of sticky rice’ is seen only in Vietnamese, not other Vietic languages. The semantics and phonological form match the Proto-Tai form, and it is a feasible cultural borrowing. It is not possible to identify the time of borrowing, whether it is an ancient or recent loanword. • Máng ‘water pipe of bamboo’ has a good phonological match, including the tone category. A small matter is that the Proto-Tai initial is not imploded, as it is in the Proto-Vietic reconstruction. This is significant as Vietnamese /m/ should come from an imploded *ɓ (NTC 1995:19-23). However, it is otherwise a reasonable match. The tone category makes it possible to be a very early loanword. • Mường ‘mountain village’ is a strong candidate for its phonological, semantic, and culturally specific features. It is an ethnonym for the Muong peoples and represents sociocultural contact. This matter is very large and beyond the scope of this paper. • • • • • • Muống ‘water spinach’ has very good phonological and semantic match with the Proto-Tai counterpart. The tone is expected, and the initial /m/ in Vietnamese stems to implosive *ɓ. Què ‘crippled / lame’ and my tentative Proto-Vietic reconstruction are very similar to the Tai reconstruction. The word appears in a few other branches of Austroasiatic (e.g. Bahnar kwɛɛ, Surin Khmer khe:, Mon kheˀ, etc.), but it has a limited geographic distribution regarding timing, which suggests a loanword, not a native etymon. The tone is not informative, so we cannot know when the borrowing happened. Quét ‘to sweep’ is not a perfect match. There is no Vietic reconstruction for it. The vowel is not expected in comparison with the Tai reconstruction. However, as it is generally similar, we will consider it a Tai loanword until more evidence is available. Rựa / rạ ‘knife / bush-knife’ is a strong candidate for loanword status in terms of phonology, semantics, and sociocultural matters. The Proto-Tai and Proto-Vietic forms match, as does a tentative Proto-Austroasiatic reconstruction #mraaʔ. In Austroasiatic, it is seen in six branches in the region , but not Munda or Nicobaric, which is reasonable if this is a Tai loanword in a later period (see Alves 2014). Trống ‘drum’ is a very important word because it could be related to the Dong Son bronze drums of the 1st millennium BCE. The word forms in Tai and Vietic have been noted by Alves (2015c:42) and by Churchman (2016:33). Churchman (Ibid.) supposed the [kl] onset of Tai was and indication of a Tai loan into Vietic, but this is not accurate as the Vietic reconstruction also has [kl], as in various Vietic languages. However, while the Tai and Vietic reconstructions have the same consonants and vowels in Pittayporn’s reconstruction, the vowel is only partial related in the Li reconstruction (i.e., *o versus *ɔ), and the tone category differs in both (i.e., type A in Tai versus C in Vietic). Also, onomatopoeia may be involved, but it is interesting that we find no other Austroasiatic languages with this word form in the MKED (only different words for ‘drum’), only Vietic and Tai. If we assume that Tai is primarily associated with the Lingnan region of southeast China and Vietic language groups are associated with the Dong Son culture, we can say that both language groups have a long tradition of bronze drums. This archaeological data is complex, but broadly speaking, Heger I drums are found widely in northern Vietnam, Yunnan, and throughout Southeast Asia (cf. Kim 2015:27), while Heger II drums are found widely in southeastern China (cf. Churchman 2016:7). While the history of the bronze drums is not entirely clear, (a) Heger I drums spread geographically, while Heger II to IV drums are largely localized, (b) comparative archaeological evidence suggests a connection between Heger I drums and the Dong Son culture, and (c) dating of the bronze drums suggests Heger I drums were developed earlier than the other types (Calo 2009:4-6). Thus, we cannot say with certainty what the direction of borrowing is, but it is possible that it is from Vietic into Tai. Proto-Vietic *pa:C ‘aunt’ (in Muong, Malieng, and Thavung) and *bəɯC ‘daughter-in-law’ (in Muong, Pong, and Chut) are not in Vietnamese, but they are in various other Vietic languages. The syllable shapes and tone categories match. They indicate intermarriage between the two groups. Correspondingly, genetic data from the Núi Nấp archaeological site suggest Tai and Vietic genetic mixing (Lipson et al. 2018:2). There are other words with partial similarities, but these words have too problems such that we must reject them unless solid evidence can be obtained. cụt ‘cut off; lopped’ matches the Proto-Southwestern Tai word well. However, it is a basic word that is less likely to be borrowed. It can also be considered onomatopoeia (notice it resembles English ‘cut’). Dây ‘rope/string’ seems similar to Proto-Tai *sa:jA ‘string’ (Pittayaporn 2009), but we cannot reconstruct a Proto-Vietic word, and Vietnamese ‘d’ generally comes from *d or *t, not *s. Đập ‘to thresh rice’ is similar to *tɤpD ‘to slap’ (Pittayaporn 2009). It is reconstructed as *dəp in Vietic, though only seen in the Viet-Muong and Pong-Cuoi groups, which suggests the possibility of borrowing. However, the semantic features are not consistent: the Proto-Tai word is not about ‘rice’. Moreover, it would be better if both items the same initials, either both *d or *t. Finally, the element of onomatopoeia is again a factor that makes the situation less certain. Undoubtedly, other researchers will find more of such words. We should not to be too flexible with semantics and phonology, especially for words that are basic and not culturally specific. Historical linguists do not want to throw away interesting data, but it is important for us to exclude problematic instances as much as possible. Table 10: Proposed shared Tai-Vietic words and possible Tai loanwords Vietnamese Vietic Tai kèn ‘reed flute / khene’ *gɛ:n ‘khene / reeded wind *gɛnA (Li 1977) instrument’ lam ‘kind of sticky rice’ NR *hlamA (Li 1977) máng ‘water pipe of bamboo’ mường ‘mountain village’ *ɓaːŋʔ ‘water pipe of bamboo’ NR *baŋB/C ‘tube, bamboo’ (Pittayaporn 2009) *mɯəŋA (Pittayaporn 2009) muống ‘water spinach’ *ɓɔːŋʔ ‘water spinach’ *ɓuŋC (Pittayaporn 2009) què ‘crippled / lame’ #gwɛ: *gwɛA (Li 1977) quét ‘to sweep’ NR *kwa:tD (Pittayaporn 2009) rạ / rựa ‘knife / bush-knife’ *m-raːʔ *ɟm̩ .ra:C (Pittayaporn 2009); *vraC sword (Li 1977) *klɔŋA (Li 1977) trống ‘drum (n)’ #kloːŋʔ NA ‘aunt’ *paːʔ ‘fath’s elder broth’s wife / *pa:C (Li 1977) ‘aunt’ fath/moth’s elder sister’ (Muong, Malieng, Thavung) *bəːʔ ‘daughter-in-law’ *bəɯC (Li 1977) NA ‘daughter-in-law’ 5. Summary of a new perspective To review, of the 40 words NTC hypothesized were Tai loanwords in Vietnamese, only 5 are supported by comparative data. We have added 12 more potential Tai loanwords, some of which are still in the domains of agriculture and trade, but also two kinship terms. However, in several cases, some Austroasiatic words were likely borrowed into Tai, and for some other words in Vietnamese, the direction of borrowing is unclear. Regardless, these early shared Tai-Vietic words indicate early sociocultural contact and lexical exchange. We must put this number in context. It is much smaller than over 600 early Chinese loanwords in Vietnamese (many of which are also in Muong, but I have not yet checked completely). I have also assembled over 150 possible early Chinese loanwords in Proto-Tai from various studies (cf. Manomaivibool 1975, Pittayaporn 2014, Alves 2015b) and have checked many of these words for phonological patterns (but this is still in process). This is also a much larger number than the number of Tai-Vietic words from an early period. Therefore, it appears that the language contact situation between Tai and Vietic was much less intense than either Tai or Vietic with Sinitic. Even if the number of Tai loanwords doubles in the future, this hypothesized situation of language contact will not change. Another issue with the small number of early loanwords is determining the timing of borrowing. In most cases, there is not enough linguistic or archaeological evidence to estimate with certainty when the words were borrowed. Only a few have tones which that indicate borrowing before tonogenesis. As for early Chinese loanwords, as there are hundreds of such words, we have enough phonological patterns to estimate approximate time of borrowing. With only 17 Tai loanwords, this is not possible. The small number of loanwords also makes it difficult to evaluate the sociocultural contact. Nevertheless, a few words appear to be borrowed before tonogenesis (e.g. rựa ‘knife / bushknife’, muống ‘water spinach’, mận ‘plum’, máng ‘water pipe of bamboo’). Thus, we can tentatively accept that more of these hypothesized Tai words were borrowed in the Dong Son period. If so, according to archaeological studies, at the time of borrowing, the Austroasiatic (and thus Vietic) peoples already grew rice and millet (Higham 2017), including the early Phùng Nguyên culture (Hán 2009). The Đồng Đậu (c. 1,500-1,000 BCE) and Đông Sơn (c. 500 BCE-100 CE) cultures were already of the Bronze Age (cf. Trịnh Sinh 2003:162-172). Since archaeogenetic data from the Màn Bạc site in this region before those cultures contains Austroasiatic genetic material (McColl et al. 2018), there is reason to assume that Vietic speakers were part of the Bronze Age in this region at the time of Tai-Vietic language contact. Considering the linguistic and archeological data, we must present a different picture from NTC’s. Various historical and archeaological descriptions show that the Lingnan region and the Red River Delta were part of a larger region of trade and cultural contact (e.g. Demandt 2020, etc.). Tai and Vietic groups, both rice-producing Bronze Age groups, must have interacted for purposes of trade, including some exchange of agricultural products and practices. At some point, there was intermarriage as demonstrated by archaeogenetic research and the borrowing of a few kinship terms. However, it appears that the incoming Sinitic groups quickly developed sociocultural status in both Tai and Vietic regions. These are general and tentative ideas, and they must be considered further as additional data and analyses are available. This is a short paper that cannot answer all the questions. Instead, it develops the ideas that NTC put forth, using many of his own claims about the historical phonology of Vietic. References Alves, Mark J. 2014. A note on the early Sino-Vietnamese loanword for 'rake/harrow'. Cahiers de Linguistique Asie Orientale 43: 32–38. Alves, Mark J. 2015a. Etyma for ‘Chicken’, ‘Duck’, & ‘Goose’ among Language Phyla in China & Southeast Asia. Journal of the Southeast Asian Linguistics Society 8:39-55. Alves, Mark J. 2015b. Grammatical Sino-Tai Vocabulary and Implications for Ancient Sino-Tai Contact. Presentation at the 48th International Conference of Sino-Tibetan Languages and Linguistics. 21-23 August 2015 at the University of California, Santa Barbara. Handout. Alves, Mark J. 2015c. Historical notes on words for knives, swords, and other metal implements in early Southern China and Mainland Southeast Asia. Mon-Khmer Studies 44: 39-56. Alves, Mark J. 2016. Identifying Early Sino-Vietnamese Vocabulary via Linguistic, Historical, Archaeological, and Ethnological Data. Bulletin of Chinese Linguistics 9: 264–295. Alves, Mark J. 2017. Etymological research on vietnamese with databases and other digital resources (Nghiên Cứu Từ Nguyên Tiếng Việt Bằng Cơ Sở Dữ Liệu Và Tài Nguyện Điển Tử Khác). Ngôn Ngữ Học ở Việt Nam: 30 Năm Đổi Mới và Phát Triển (Kỷ yếu Hội thảo Khoa học Quốc tế) [The Linguistics of Vietnam: 30 Years of Renovation and Development (International Conference)], Viện Ngôn Ngữ Học, Hanoi: Nhà Xuất Bản Khoa Học Xã Hội, 2017, pp. 183-211. Alves, Mark J. 2018. Early Sino-Vietnamese Lexical Data and the Relative Chronology of Tonogenesis in Chinese and Vietnamese. Bulletin of Chinese Linguistics 11.1-2:3-33. Apte, Vaman Shivaram. Revised and enlarged edition of Prin. V. S. Apte's the practical Sanskrit-English dictionary. Poona: Prasad Prakashan, 1957-1959. Baxter, William H. and Laurent Sagart. 2014. Old Chinese: A New Reconstruction. Oxford: Oxford University Press. Blench, Roger. 2014. Reconstructing Austroasiatic prehistory. Unpublished. Calo, Ambra. 2009. Distribution of Bronze Drums in Early Southeast Asia. British Archaeological Reports. Churchman, Catherine. 2016. The People between the Rivers: The Rise and Fall of a Bronze Drum Culture, 200-750 CE. New York: Rowman and Littlefield. Demandt, Michèle H.S. 2020. Reaching the Southern Wilderness: Expansion and the Formation of the Lingnan Transportation Network during the Qin and Han Dynasties. Journal of the Economic and Social History of the Orient 63 (2020):157-194. The Chinese Text Project. https://ctext.org/. Accessed 15 April 2021. Ferlus, Michel. 1982. Spirantisation des obstruantes médiales et formation du système consonantique du vietnamien. Cahiers de Linguistique Asie Orientale 11.1: 83-106. Ferlus, Michel. 1992. Histoire Abregée De L’évolution des Consonnes Initiales du Vietnamien. MonKhmer Studies Journal 20: 111-127. Ferlus, Michel. 1997. Le Maleng Brô Et Le Vietnamien. Mon-Khmer Studies Journal 27: 55-66. Ferlus, Michel. 2007. Proto-Vietic reconstructions (unpublished, available in the Mon-Khmer Etymological Dictionary). Ferlus, Michel. 2014a. Proto viet-muong (Proto-Vietic). halshs-02490370. Ferlus, Michel. 2014b. The sexagesimal cycle, from China to Southeast Asia. 23rd Annual Conference of the Southeast Asian Linguistics Society, May 2013, Bangkok, Thailand. halshs-0922842v2. Fuller, Dorian Q., Cristina Castillo, Eleanor Kingwell-Banham, Ling Qin and Alison Weisskopf. Charred pummelo peel, historical linguistics and other tree crops: approaches to framing the historical context of early Citrus cultivation in East, South and Southeast Asia. in Archaeology and History of Citrus Fruit in the Mediterranean: Acclimatization, Diversifications, Uses, ed. by Véronique Zech-Matterne and Girolamo Fiorentino. CNRS, Collection du Centre Jean Bérard, 48:29-48. Hán, Văn Khẩn. 2009. Xóm Rền: Một di tích khảo cổ đặc biệt quan trọng của thời đại đồ đồng Việt Nam [Xom Ren: an especially important archaeological relic of the Bronze Age in Vietname]. Nhà Xuất Bản Đại Học Quốc Gia Hà Nội. Haudricourt, André G. 1954b. De l’origine de la ton de Vietnamien. Journal Asiatique 242:69–82. Headley, Robert K., Jr., Kylin Chhor, Lam Kheng Lim, Lim Hak Kheang and Chen Chun. 1977. Cambodian--English Dictionary. xxvii, 1495 pp., 2 vols. Catholic University Press. Higham, Charles F. W. 2017. First farmers in mainland Southeast Asia. Journal of Indo-Pacific Archaeology 41:13-21. Hoàng, Văn Khoán. 2003. Nghề đàn của người Đồng Đậu (qua các dấu đan in trên đồ gốm) [The craft of the Dong Dau people (through imprints of weavings on pottery)]. Văn Hóa Đồng Đậu: 40 Năm Phát Hiện và Nghiên Cứu (1962-2002). Nhà Xuất Bản Khoa Học Xã Hội. 172-180. Jonsson, Nanna L. (1991) Proto Southwestern Tai. Ph.D. dissertation, available from UMI. Kim, Nam C. 2015. The Origins of Ancient Vietnam. Oxford University Press. Li, Fang-Kuei. 1977. A Handbook of Comparative Tai. Oceanic Linguistics Special Publications No. 15. Honolulu, HI: The University of Hawaii Press. Lipson, Mark, et al. 2018. Ancient genomes document multiple waves of migration in Southeast Asian history. Science. 10.1126/science.aat3188. 1-10. Lu, Liu. 2014. Time flies through eras of ups and downs. ChinaDaily.com. Updated 1 December 2014. http://www.chinadaily.com.cn/sunday/2014-01/12/content_17230834.htm. Accessed 14 April 2021. Manomaivibool, Prapin. 1975. A Study of Sino-Thai Lexical Correspondences. PhD thesis. University of Washington. McColl, Hugh, et al. 2018. Ancient Genomics Reveals Four Prehistoric Migration Waves into Southeast Asia. Preprint, Science, doi:10.1126/science.aat3628. Nguyễn, Tài Cẩn. 1979. Nguồn Gốc và Quá Trình Hình Thành Các Đọc Hán Việt [The Origins and Formation of Sino-Vietnamese Readings]. Hà Nội: Nhà Xuất Bản Khoa Học Xã Hội. Nguyễn, Tài Cẩn. 1995. Giáo trình lịch sử ngữ âm tiếng Việt (sơ thảo) [A textbook of Vietnamese historical phonology (a draft)]. Hanoi: Nhà xuất bản Giáo dục. Nguyễn Văn Lợi. 2016. Giaó trình lịch sử ngữ âm tiếng Việt của giáo sư Nguyễn Tài Cẩn: Thành tựu và những điều gợi mở. https://nvloi.wordpress.com/. Norman, Jerry and Tsu-lin Mei. 1976. The Austroasiatics in ancient South China: Some lexical evidence. Monumenta Serica 32 (1976): 274-301. Ratliff, Martha. 2010. Hmong-Mien Language History. Canberra, Australia: Pacific Linguistics. Pittayaporn, Pittayawat. 2009. The Phonology of Proto-Tai. Ph.D. Doctoral Dissertation. Cornell University. Pittayaporn, Pittayawat. 2014. Layers of Chinese loanwords in Proto-Southwestern Tai as evidence for the dating of the spread of southwestern Tai. MANUSYA: Journal of Humanities, Special Issue 20 (2014): 47-68. Sagart, Laurent. 2021. Language families of Southeast Asia. The Oxford Handbook of Southeast Asian Archaeology. In press. hal-03099922. Schuessler, Axel. 2007. ABC Etymological Dictionary of Old Chinese. University of Hawaii Press. Shorto. 2006. SEAlang Mon-Khmer Etymological Dictionary, accessed 15 April 2015, http://www.sealang.net/monkhmer/dictionary/. Shorto, Harry L. 2006. A Mon–Khmer Comparative Dictionary. eds. Paul Sidwell, Doug Cooper and Christian Bauer. Canberra: Australian National University. The Sino-Tibetan Etymological Dictionary and Thesaurus. STEDT. The University of California, Berkeley. http://stedt.berkeley.edu/~stedt-cgi/rootcanal.pl. (Accessed 15 April 2015). Thurgood, Graham. 1999. From Ancient Cham to Modern Dialects: Two Thousand Years of Language Contact and Change: With an Appendix of Chamic Reconstructions and LoanwordsAuthor(s). Oceanic Linguistics Special Publications, No. 28. University of Hawai’i Press. Trịnh, Sinh. 2003. Đồng Đậu – một bước nhảy vọt của nghề đúc đồng. Văn Hóa Đồng Đậu: 40 Năm Phát Hiện và Nghiên Cứu (1962-2002). Nhà Xuất Bản Khoa Học Xã Hội. 162-171. Wang, Zhongshu. 1982. Han Civilization (translated by K.C. Chang and Collaborators). New Haven, Connecticut: Yale University Press. 小學堂文字學資料庫 [The Xiaoxuetang Character Study Database]. https://xiaoxue.iis.sinica.edu.tw/. Accessed: 17 April 2021.