In this study, over 60 Chinese loanwords in Vietnamese are claimed to have been borrowed during the East Han or West Jin Dynasties. These Early Sino-Vietnamese (esv) words are identified via a combination of linguistic, historical, archaeological, and ethnological data sources and frameworks. Such an interdisciplinary method helps to confirm or refute these words’ status as loanwords and as belonging to this specific historical period. The combined linguistic and extralinguistic data also leads to hypotheses about possible phonological changes in Chinese from the Old Chinese (oc) to Middle Chinese (mc) periods. In particular, while Sino-Vietnamese words from the mc period have expected qusheng tones for Chinese qusheng loanwords, oc-era esv words have either shangsheng or, unexpectedly, pingsheng tones. It is hypothesized that esv words with shangsheng tones for oc qusheng words were borrowed earliest, while esv items with pingsheng represent a later stage in oc in which final *-s was in the process of being lost in the first few centuries ce.
本文重點討論東漢或西晉時代借自漢語的60多個早期漢越語詞，通過語言學、歷史學、考古學、民族學等多個學科，綜合使用各種資料加以論證。這種跨學科的方法可用來解答它們是否屬於外來詞、這些詞是否屬於此一歷史時期的問題。此外，語言學材料和非語言學材料的結合還使我們能夠就上古到中古可能存在的語音變化提出假設；具體說來，源自中古的漢越語借詞本該以去聲對應漢語去聲，然而在源自上古時代的早期漢越語中卻可以是上聲甚至是出人意料的平聲。我們因此可以假設：對應古音去聲的早期漢越語上聲出現得比較早，而對應上古音去聲的早期漢越語平聲，應該出自*-s逐漸消亡的晚期上古音階段，即剬元紀年的頭幾個世紀。(This article is in English.)
1 The Significance of Extralinguistic Data in Sino-Vietnamese Historical Linguistics
Early Sino-Vietnamese vocabulary2 (esv hereafter) is the layer of Chinese vocabulary borrowed into Vietnamese, and in some instances into Proto-Vietic,3 from the Han to Tang Dynasties, that is, from around 200 bce to the mid-first millennium ce (cf. Alves 2009, 2016). This is in contrast with the later layer of Literary Sino-Vietnamese (lsv hereafter), which is a standardized set of Vietnamese pronunciations of Chinese characters from Middle Chinese from roughly the Tang to Song Dynasties (618–1279 ce). Thus, esv overlaps the periods of Old Chinese (oc hereafter) and the early stages of Middle Chinese (mc hereafter). As this study is focused on words borrowed two thousand years ago, it is necessary to combine extralinguistic evidence with historical linguistic data to support claims of timing of the loanwords in question. In some cases, historical documents and/or archaeological data are the only sources of clarifying information as some of the historical linguistic data does not provide absolute criteria by which to determine their age. In other cases, slight phonological irregularities cast doubt on Vietnamese words as Chinese in origin, but the weight of the historical, archaeological, and/or ethnological evidence is strong and supports the status of these words as loans.
In this section, an example of the proposed esv word for ‘iron’ is discussed in terms of both linguistic and extralinguistic factors. Subsequently, the overall goals and sections of the paper are delineated.
1.1 Is Proto-Vietic *k-rac ‘iron’ from Old Chinese *l̥ˤik?
How much should archaeology influence historical linguistics? Consider the Proto-Vietic word *k-rac ‘iron’. Is the source Old Chinese (oc hereafter) *l̥ˤik ‘iron’, which has a similar word shape but also some phonological differences? The Southeast Asian archaeologist Charles Higham has claimed the following about the region of modern-day northern Vietnam in the mid-first millennium bce: “At present, the direct Chinese contact with Dong Son people is the most likely means whereby knowledge of iron reached Bac Bo” (Higham 2014:209). If true, this means that the Iron Age in the region of northern Vietnam began with the arrival of Chinese cultural presence. For historical linguists, this means that speakers of oc were in contact with speakers of Vietic in that period of the transmission of metallurgy practices, but before large-scale southward migrations of Chinese groups.
Thus, iron metallurgy—and the word for ‘iron’—were likely borrowed by Vietic groups at that time. Considering the data in Table 1, it is clear that Vietnamese does have a variety of words for metals of Chinese origin. Chinese words for metal have also entered at the proto-language level in Hmong-Mien and Tai. Thus, almost the entire semantic domain of metals is seen throughout the language families of Southern China and northern Southeast Asia, thereby supporting evidence of the regional spread of Chinese metallurgical practices.
While such words in Proto-Tai and Proto-Hmong-Mien seem almost unquestionably to be Chinese loanwords in light of the consistent phonological correspondences, and the Vietnamese word gang ‘steel’ appears as well to be a Chinese loanword, the Vietnamese words sắt from Proto-Vietic *k-rac and especially thép 4 ‘steel’ (no proto-Vietic available reconstruction) have phonological disparities with the oc item. Focusing on Proto-Vietic *k-rac, while the meaning is the same as in Chinese, the initial *kr and final *c of the Vietic items are not consistent with oc * l̥ and *-k, in contrast with the more consistent forms of Proto-Tai and Proto-Mien. However, in support of the Sinitic-Vietic5 connection of the word, the Proto-Vietic medial *-r- is comparable to that of Proto-Mien (pm). It is also possible for this to be a Chinese loanword borrowed through non-Chinese speech groups. Overall, the linguistic evidence leaves considerable uncertainty as to whether the item can be considered a Chinese loanword or whether it is chance coincidence.
I argue that, in this case, we must assume that the Vietic word *k-rac is, despite slight phonological differences, very likely a Chinese loanword from the Western Han Dynasty (206 bce – 9 ce). The following points support this claim.
Semantic domains: Table 1 shows clear evidence not of the borrowing single words but rather entire semantic domains of words employed throughout the broader region of southeast China and northern Vietnam. Indeed, while not impossible, it would be strange for the word for ‘iron’ to be excluded at a time when words for silver, gold, and possibly bronze were all borrowed.
Archaeological data: Considering the weight of evidence of the influx of both transported and locally-produced iron objects in this period, the relatively small phonological inconsistencies cannot be considered strong counter-evidence against the claim that the word is of Chinese origin. Additional archaeological evidence is presented in section 5.1.
Limitations of reconstructions: Reconstructions of borrowed words rely on lexical data, and the highly conservative but rather understudied minor Vietic languages may lack key pieces of lexical data. Moreover, there may be gaps for other reasons (e.g., transmission via small, socioculturally isolated speech communities).
Variation in Lexical Borrowing: Words borrowed sporadically by monolinguals (as in early in Sino-Vietnamese contact) rather than systematically by bilingual communities (as in later Sino-Vietnamese history) are more likely to vary phonologically. Moreover, it is possible that Sinitic at that time may have had phonological differences among its speech communities.
While some of these scenarios are speculative and ultimately non-disprovable, the weight of the archaeological evidence of sociocultural contact and influence in this place and time cannot be ignored, though such information must be applied with discretion. This position is taken as a primary perspective in the rest of this paper.
1.2 Goals of this Paper
It is the purpose of this paper to show instances of esv words which have corroborating extralinguistic data to identify, with more certainty, these lexical items as Chinese in origin and to show that they were likely borrowed in the Han Dynasty. In this way, historical, archaeological, and ethnological evidence can serve as tools for phonological reconstructions, and such historical reconstructions become another means by which to assess ethnolinguistic scenarios of early history or pre-history. This paper thus becomes an exercise in ethno-historical linguistics using a blend of disciplines to increase overall understanding of human sociocultural history in the region of modern day northern Vietnam and bordering areas of southeast China.
The identification of Chinese loanwords in Vietnamese that belong to the pre-Tang period has become increasingly solid with recent thorough oc reconstructions (e.g., Starostin 1998–2003, Schuessler 2007, Baxter and Sagart 2014, etc.) and other comparative resources (e.g., Proto-Tai (Li 1977 and Pittayaporn 2009), Proto-Hlai (Norquest 2007), Proto-Hmong-Mien (Ratliff 2010), Proto-Austronesian (Blust and Trussel), etc.). Tryon (1979) and Phan (2013) reviewed many posited Early Sino-Vietnamese (esv hereafter) words noted by Wang Li, Pulleyblank (1981, 1984), among others, and Starostin (1998–2003) also incorporated many esv in his oc online dictionary. Baxter and Sagart (Ibid.) have extensively utilized esv and relevant Proto-Vietic data for many aspects of their oc reconstruction. However, linguistic data sometimes has incomplete datasets and incomplete phonological patterns, and thus such data cannot necessarily identify historical periods in which words were borrowed.
Thus, to verify and better understand the oc loanwords in esv and the early Sino-Vietnamese language contact situation, additional sources of information are required. For instance, providing historical support, Phan (2013) proposed a large community of Chinese speaking Jin Dynasty immigrants in the early 300s ce. Phan provides some phonological criteria to distinguish Han era from Jin era loanwords. However, beyond general statements of waves of Chinese immigrants, there is also sufficient historical and archaeological data to provide more detail about specific items and their associated words. Thus, it is reasonable to attempt to draw on the combination of linguistic, historical, and archaeological data to prove that Vietnamese words of likely Chinese origin can be connected to events millennia in the past.
In subsequent sections, the following aspects of Han Dynasty era oc loans in esv are covered: (a) phonological characteristics of esv, (b) the uses and caveats of historical and archaeological data, (c) the sociolinguistic circumstances in which esv was transmitted, and (d) examples of esv vocabulary with supporting historical, archaeological, and ethnological evidence.
2 Phonological Criteria for esv and Degree of Certainty9
The phonological correspondences that identify esv have been studied since the mid-1900s. Primary works include those of Wang Li (1948, 1958), Haudricourt (1954a), Đào (1979), Tryon (1979), Pulleyblank (1981 and 1984), Phan (2013), and Baxter and Sagart (2014), among others. Phan claims that while some esv words are from the Han Dynasty, others belong to the Western Jin Dynasty (265–420 ce), but in many cases, there is no clear distinction based on phonology alone (Phan 103:77–79). This author has gathered an unpublished list of about 450 likely esv items of medium to high certainty based on phonological criteria, and of these, several dozen have sufficient phonological and, in many cases, corroborating historical and archaeological evidence to connect them to the Han Dynasty. Another 160 items of the 450 are identified broadly as coming from Old Chinese, but without sufficient data to determine a more specific era of borrowing (e.g., Western Han, Eastern Han, or Western Jin dynasties).
A number of language-internal phonological changes help to identify esv. The phonological patterns can only be viewed within the context of Vietic historical phonological changes and the place of Vietnamese in Austroasiatic (e.g., Haudricourt 1954a, 1954b, and 1965, Ferlus 1982, Nguyễn T.C. 1995, Shorto 2006, etc.). As an Austroasiatic language with Mon-Khmer typology, it must be assumed that, like the highly conservative minor Vietic languages, the Vietic ancestor of Vietnamese was a non-tonal polysyllabic language with prefixes, infixes, and initial consonant clusters. Modern Vietnamese typology thus represents evidence of a massive typological shift including the following: (a) reduction of clusters and presyllables (thus leading to monosyllabicity) and subsequent development of initial spirants, (b) loss of final the glottal stop and fricatives *-s and *-h, followed by the development of tones, (c) diphthongization of earlier /a/, and (d) shifting of voicing and quality (e.g., nasalization or implosives) of initials. These changes from Austroasiatic thus necessarily apply to both Austroasiatic etyma and oc loanwords.
Based on such phonological changes, degree of certainty must be considered: the more frequently phonological patterns occur, the higher the assumed certainty is that such words are from oc. The items in this study are considered to be high certainty or, in a few cases, medium certainty when there are either slight phonological inconsistencies or there is a lack of additional instances of words with the same change. Moreover, oc words were borrowed at different times and possibly from differing varieties of Sinitic, and there will necessarily be gaps in the examples. In addition, as noted in section 1, small inconsistencies may occur in lexical borrowing of single words by non-native speakers. For instance, in some cases, while the tone category, segments, and semantics are consistent, and the sociocultural information supports the borrowing of a word, the tone height is not consistent (e.g., ‘crossbow’ in § 5.2). Such instances are of medium rather than high certainty Han Dynasty Chinese loanwords. And therefore, the extralinguistic data becomes essential in clarifying the situation.
2.1 Tones in esv
The tone categories in esv are strong evidence of the oc period of Chinese. In light of early studies (Haudricourt 1954b, Mei 1970, Baxter 1992, Baxter and Sagart 2014), it is now widely assumed that modern Chinese, Vietnamese, Tai-Kadai, and Hmong-Mien were originally toneless (Haudricourt 1954b, Matisoff 1970, Ratliff 2010, Delancey 2010) and that the tones emerged in the language groups regionally, a kind of typological convergence (Ratliff 2010:185–193). Tones in this region are the end result of the loss of final glottal stops and fricatives as well as changes in the voicing of initials (see Haudricourt 1954a and Thurgood 2002 on Vietnamese tonogenesis). The presence of comparable final glottal stops and final fricatives reconstructed in the four language groups allowed for the highly consistent maintenance of tone categories across a large geographic region.
The four main tone categories and their representative tones in oc, esv, and lsv are presented in Table 2.10 Crucially, there is a reversal in shangsheng and qusheng categories between esv and lsv. In the esv period, Old Chinese and Vietic still had final glottals and fricatives, while in the lsv period, Chinese and Vietnamese had lost them, leaving open syllables (i.e., syllables without final consonants or final nasals), but with tonal categories instead. This shifting of tone categories thus represents doublets borrowed in different periods separated by several centuries. Also noted in the chart are differences based on the yin-yang/high-low tone distinctions within the categories. Finally, under row C, the qusheng category, esv has one other sub-type, namely, the pingsheng tones, the Vietnamese tones ngang and bằng, for qusheng words as discussed below. The properties of Vietnamese tones are briefly described in Appendix 2.
The following statements characterize the application of these tone categories to the identification of esv vocabulary and distinguishing general periods of lexical borrowing.
The shangsheng-qusheng reversal between esv and lsv vocabulary is evidence of the oldest layer and can even be connected to the Western Han Dynasty.
The pingsheng-qusheng reversal between esv and lsv vocabulary is evidence of an older layer within esv, though it may extend only from the later period of Eastern Han to the Western Jin dynasties.
Pingsheng and rusheng words (i.e., words with final /p/, /t/, or /k/ stops) maintain the same status in esv and lsv and thus can help to identify esv but cannot further distinguish historical periods.
Pingsheng words with initial sonorants have the low tone bằng in esv, while in lsv, such words have the upper tone ngang. However, while this identifies esv, it does not clearly distinguish oc from early mc era loans.
Samples of the reversal of the tone categories between esv and lsv are in Table 3. Only three of the possible four patterns are available in the lexical data: esv
yinshang for lsv
yinqu, yangshang for yangqu, and yinqu for yinshang. There is no instance of a yangqu word for the yangshang category. The lexical samples in Table 3 are not discussed in subsequent sections as there is no corroborating historical and/or archaeological evidence for them. Instead, these words must be considered Han Dynasty oc loanwords solely based on comparative linguistic evidence, including both tones and segments. Additional discussion of segmental evidence is provided in sections 2.2 and 2.3.
On the matter of esv pingsheng tones for Chinese qusheng, these have been noted but not explained (Wang Li 1948:76–77, Phan 2013:107 and Baxter and Sagart (2014:100, 137, 177, 191)). Wang Li (1948:Ibid) considers these instances of ‘Vietnamized’ tones. However, as Phan notes, a Chinese administrator and scholar of the sixth century wrote of the perceived confusion of pingsheng and qusheng words among southerners (Phan: Ibid). Zhu has hypothesized that the final *-h/-ɦ was in the process of being lost by the Wei-Jin Dynasties, and the data in this study offer support for this claim (Zhu 2009: Section 7). As to be discussed in subsequent sections, words with this tone category are associated with items that may have been borrowed in the first century ce, somewhat earlier than Zhu (Ibid.) proposes, as suggested by historical and archaeological data. Other esv words beyond those presented in this paper have pingsheng for qusheng, but the linguistic and extralinguistic data cannot connect those words with the Han Dynasty and could also be loans from the later Jin Dynasty (265–420 ce). Additional study and data will be required to determine whether and how many of those additional words not included in this paper are from the Han or Jin Dynasties.
The hypothesis in this study is that the oc final *-s, and thus Chinese qusheng, is realized in esv as shangsheng in words borrowed as early as the Western Han Dynasty, while in the Eastern Han and into the Jin Dynasty, the final *-s was lost in Chinese varieties in the southern region prior to 500 ce and thus was realized as esv
pingsheng. This is a phonological historical hypothesis that needs further evaluation, but for now, the archaeological and historical data is sufficient to provide reasonably solid evidence of when such Chinese goods were in northern Vietnam. Moreover, there are no instances of Austroasiatic final fricatives leaving level tones, so the first and most reasonable assumption is that Chinese at the time of borrowing lacked the final *-s.
2.2 Initials in esv
Regarding esv initials, the two major categories of initials are those that distinguish eras of loanwords and those that do not. esv words that retain oc initials, primarily the Vietnamese stops ‘b’, ‘đ’, and ‘c’ or the nasals ‘m’ and ‘ng’ (see Appendix 2 for a list of Vietnamese consonants and their corresponding sounds in ipa), provide little or no clarification of timing of the borrowing. Instead, for words with such initials, only tone categories and rhymes can be considered in determining the approximate time of borrowing. In contrast, initial spirants in esv, such as ‘v’ and ‘gi’, when they are distinct from the lsv layer (as lsv words can have these initials following mc phonological categories) and when they correspond to oc words with pre-initials and consonant clusters, can serve to identify words originally from the oc period. This pattern of spirantization of complex initials is similarly seen in the evolution of initials from Proto-Vietic to Vietnamese through comparison with modern Minor Vietic languages, such as Ruc and Arem (Ferlus 1982, Nguyễn T.C. 1995:285–298, Shorto 2006). In Table 5, the Vietnamese orthography Quốc Ngữ letters are shown with ipa next to them.
Types of oc clusters that are the sources of esv initial spirants are numerous, making it difficult to provide highly specific historical phonological rules. For example, esv initial /v/ occurs when oc words had complex initials containing labials, but both the types of labials and the types of pre-initials vary considerably. In Table 6, the oc words had initials including *b, *p, *p, *ph, and even a labialized consonant *kw.
In a separate category, a final note is needed for esv ‘r’, directly matching the oc source. It is hypothesized that when an oc word had initial *r and currently has /r/ in esv, it is likely from the Han period. In contrast, esv words with /l/ for oc *r are likely later borrowings (e.g., esv
lìa ‘to leave’, lsv
li, oc 離 *[r]aj), from the Jin Dynasty or later. esv words with initial /r/ also exemplify esv low tones on words with sonorant initials, in contrast with lsv high tones with sonorants. Finally, in Table 7, ‘smelt’ is an instance of an esv low pingsheng word for an lsv
2.3 Rhymes and Vowels in esv
In most cases, oc and esv general syllable structure are parallel: open syllables (i.e., lacking final obstruents) to open syllables, final nasals to final nasals, or final stops to final stops. The maintenance of finals applies as well to the oc final *-s and *-ʔ, though in such cases, the equivalent tone category is maintained in modern open syllables. However, two items of note are words considered to be possible esv items: tươi ‘fresh’ noted by Wang Li (1958:359) and ngòi ‘spring/source’, that has the meaning ‘canal’ in Vietnamese, which is noted in Starostin’s Old Chinese database.11 Both words in oc are reconstructed with final *-r. (Baxter and Sagart 2014:252–267) provide support for reconstructing final *-r in oc, which is typically realized as final /-n/ in Chinese. In Vietnamese, final /-n/ is almost always the reflex for oc final *-r in both lsv and esv. However, Proto-Austroasiatic and Vietic *-r are generally realized as a final ‘-i/-y’ offglide, not ‘-n’ (e.g., ‘two’ paa *ɓaar, pv haːr, Vietnamese hai ‘to fly’ paa *cpiir, pv pər, Vietnamese bay). Thus, it is assumed that esv with final ‘-n’ were borrowed after the change to final /n/ in Chinese, while in contrast, the two words ‘fresh’ and ‘spring/canal’ would appear to be quite early loanwords from a period in which oc *-r had not been nasalized, probably in the Western Han era.
Regarding vowels, in many cases, modern Vietnamese vowels (see Appendix 2) are directly comparable to oc vowels (e.g., oc *a > esv ‘a’, oc *e > esv ‘e’, etc.). In identifying features of vowels that identify Han Dynasty loanwords, Phan notes preservation in esv of low oc vowels (e.g., ‘grave’ and ‘register’) and identifies the preservation of final ‘-i/-y’ offglide in a number word words in esv (e.g., ‘to grind’ and ‘moth’) (Phan 2013:81). Another common feature of esv words is diphthongization, just as Vietnamese words of Austroasiatic origin have, resulting in ‘ua’ or ‘uô’, but also back unrounded ‘ưa’. Both diphthongs tend to stem from oc words with *a. However, Vietnamese ‘ưa’ occurs most often in syllables with ‘plain’ Type B initials (in contrast with pharyngealized Type A (i.e., Category iii) initials) and oc medial *-r-, though bừa ‘rake’ would seem an exception despite other support for this as an esv item (cf. Alves 2014).
Overall, regular phonological patterns—along with semantic consistency—support the claims of the several dozen lexical items in question (see Appendix 1). They also serve to distinguish these words from later periods of borrowing. The next matter is how much archaeological and historical data can support these claims, or provide grounds for excluding them despite the purported linguistic data.
3 On Using Historical and Archaeological Sources 12
To study oc loanwords in esv, combining linguistic, historical, archaeological, and ethnological frameworks and data can provide insights since applying each discipline without combined reference has limitations. Historical linguistic studies can show very clear linguistic structural patterns, and yet to offer valid historical claims, such research requires extra-linguistic evidence. Studies of historical documents provide information about historical events and their timing, but such research depends on the amount of information and detail in such documents and the degree of objectivity of authors of written records. Archaeological studies can provide approximate dates of material objects, a very powerful tool, but behaviors, activities, and cultural context can only be inferred. Finally, in the area of ethnology, knowledge of the spread of cultural practices, studies of semantic domains, and correlative sociocultural factors provide analytical frameworks for considering possible past situations, but ethnohistorical claims rely heavily on data from other disciplines. Clearly, combined application of these disciplines is the most effective method due to—and in spite of—the increased range of parameters.
3.1 Applications and Caveats of Historical Evidence in Historical Linguistic Data
Large numbers of complete Chinese historical documents from the Han and post-Han eras are available, and now in text-searchable formats (cf. The Chinese Text Project). However, these documents can be problematic in a number of ways.
Limited detail: Ancient Chinese writings are often very terse and lack clarifying details.
Mixed factual accuracy: These writings consist of a mixture of fact with overgeneralization or error.
Writer subjectivity: There can be writer bias and thus potential mischaracterization of situations and details.
Uncertain sources: Ancient Chinese histories are displaced from original events they describe by centuries or more often with no clear source of that information.
Therefore, whatever information is contained in such writings, one must keep these various caveats in mind and proceed tentatively with claims of borrowing of concepts or items and their associated words. Nevertheless, ignoring such data is counter to principles of scientific data analysis, particularly when the weight of such evidence is so robust that counter-claims themselves require substantial evidence.
3.2 Applications and Caveats of Ethno-archaeological Evidence in Historical Linguistic Data
As for archaeological data, archaeological studies provide concrete instances of cultural contact along with carbon dating to provide approximate ages of archaeological items and thus some point of reference for words for those items. As for historical ethnographic studies, research in macrosociology has shown numerous statistical correlations between types of sociopolitical structures (e.g., tribes, chiefdoms, states) and sociocultural practices, such as craft specialization, ancestor worship, warfare, marriage practices, and so on (e.g., Nolan and Lenski 2006:63, 114, 115, 116). Nevertheless, there are weaknesses in the application of these data sources and analytical frameworks. These gains and limitations are presented in Table 10.
Thus, the benefits of archaeological data are clear, and they include ethnological matters such as an overall cultural package inferable from both the archaeological and linguistic data. For instance, the claim that the Vietnamese verb rèn ‘to forge’ is from oc of the Han Dynasty is in large part due to the phonological pattern (see §5.1). However, this claim makes even more sense in light of the combined archaeological evidence of the introduction of iron, silver, and gold metallurgy and the representative words for these metals. Borrowing of the verb rèn ‘to forge’ is a logical extension of the semantic domain of metals and metallurgy, in addition to the phonological evidence.
Still, Table 10 shows that while archaeological studies deal with material culture and provide approximate dates and ethnological assumptions, there are challenges and weaknesses in applying these to historical linguistic studies of lexical data. There can be gaps in the archaeological record, such as chance losses of items. In some cases, words may exist without archaeological evidence of such items (e.g., esv cưa ‘saw’, which this author cannot find in lists of objects from the period). Instead, corroborating evidence comes from the large number of other esv items in the semantic domain of tools for manufacturing and architecture (e.g., hammer, awl, scissors, pliers, etc.). In other instances, the reverse is sometimes true. Many ploughshares were found at the Cổ Loa site, and these are associated with Chinese contact (Higham 2014), but the Vietnamese word cày ‘plough’ does not appear related to any oc word for plough, but rather to an Austroasiatic etymon *lngal/*ŋgal and Proto-Vietic *gal. Conversely, bronze metallurgy was practiced in the Red River Delta in the early first millennium bce (Higham 2014:168–169), nearly a thousand years before the Han took administrative control of the region, and yet the Vietnamese word for bronze/copper is of Chinese origin (see § 1.1). It is also important to note that when there are native Vietnamese words synonymous with borrowed Chinese words, this means that the Chinese word was introduced, not that an entirely new concept was introduced (see § 5.1 and the example of ‘to forge’).
As for inferences of cultural practices based on material culture, these are naturally speculative. This makes it difficult to certify that verbs, adjectives (especially generic verbs and adjectives which are not associated with any specific physical items), and abstract nouns are actually from oc. And still, physical objects with carbon dating are powerful tools in interpreting ethnological situations in early history, and they can provide very persuasive evidence that we cannot obtain through linguistic matters alone, as will be exemplified below. In the next section, the sociolinguistic circumstances of Sinitic-Vietic contact are described.
3.3 The Sinitic-Vietic Language Contact Situation
After considering the applications and potential pitfalls of linguistic, historical, and archaeological data sets, one can make some assumptions, or at least posit hypotheses, about the sociocultural circumstances of Sinitic-Vietic sociolinguistic contact in the first millennium bce. When the short-lived Qin (221–206 bce) and subsequent Han Dynasty (206 bce–220 ce) expanded into modern-day northern Vietnam, peoples of the Bronze Age Đông Sơn (Chinese 東山 Dongsan) Culture (circa 500 bce to 200 ce) had a long history of employing metallurgy, developed agrarian practices, and evolved a sociopolitical system of a complexity to have built and managed the Cổ Loa site (cf. O’Harrow 1978, Reinecke 2009:29–31, Kim et.al. 2010:1012, Higham 2014:210–211). In the mid-first millennium bce, the Đông Sơn culture was centered in northern Vietnam in the Red River Delta region, or Bắc Bộ (Chinese 北部 beibu) region, as illustrated in Map 1, with Cổ Loa as an emerging urban center. Đông Sơn Culture is generally considered to be the source of Vietnamese culture, and if it is so, we can assume it consisted of a Proto-Vietic speech community within an emerging state-level society at Cổ Loa. It would thus appear that by 200 bce, when the Han Dynasty commenced its expansion, there was an emerging state in the region, likely with early Vietic speakers as a significant part of the population.
In terms of linguistic characteristics, Vietic was presumably a typologically Mon-Khmer-style Austroasiatic language or language group. The language(s) they spoke would have been non-tonal but likely with phonation, polysyllabicity, affixation with prefixes and infixes, but lacking suffixes, subject-verb-object sentence type, and right-branching noun phrase structure. oc was similar in most respects, as listed in Table 11.
In sum, around 200 bce, Sinitic and Vietic were both non-tonal, polysyllabic languages or language groups that had more complex syllable templates than today. Both were svo languages, but their noun phrase structures differed. As for sociocultural factors, while the Chinese groups had long developed state-level societies and complex agricultural, economic, and literary practices and had developed iron metallurgy technology, the Đông Sơn peoples had recently reached an Iron Age state-level society in at least the Red River region, with developed agrarian practices (e.g., metal plows and developed water management). However, there is insufficient data to determine whether it had developed complex economic and administrative practices. Considering these circumstances, historical evidence of esv can be presented.
4 Historical Evidence of oc Loans in esv
Understanding the historical context requires a brief summary of the history of Sinitic-Vietic contact. The earliest contact between early Vietic and Sinitic groups is difficult to pinpoint. While Chinese administrative control began at the end of the second century bce, prominent figures from the north were influential in political developments in the regions of modern-day Vietnam and Guangdong even towards the end of the Eastern Zhou Dynasty. In the mid-third century bce, a prince reportedly from the Shu 蜀 state conquered the state based at Cổ Loa by enlisting other groups in the region. Additional historical information can be inferred from archaeological data, as discussed in Section 5. Beginning in 111 bce, the Western Han Dynasty established administrative districts in Vietnam from the north to the upper central region of modern-day Vietnam: 交阯 / Jiaozhi / Giao Chỉ around the Red River Delta near modern-day Hanoi; 九真 / Jiuzhen / Cửu Chân extending a few Vietnamese provinces southward, and 日南 / Rinan / Nhật Nam in the area of central Vietnam near modern-day Huế. These are noted in Chinese historical texts such as 漢書 Hanshu ‘the Book of Han’, 後漢書 Houhanshu ‘the Book of the Later Han’, 三國志 Sanguozhi ‘the Records of the Three Kingdoms’, among others.
However, the Houhanshu excerpt refers to details four centuries prior and is largely taken from and revised as a court document for an emperor. It has literary Chinese conciseness compounded by the author’s perspective and bias, so it must be interpreted rather than taken at face value. The Jiuzhen governor 任延/Renyan from modern-day Henan province is recorded as having ordered adults in Jiuzhen to choose a person to marry through formal ceremony (Taylor 1983:34). And yet, reconstructed terms for kinship terms in Austroasiatic (e.g., *waʔ ‘parent’s sibling’, *məy ‘mother’s sister’, ‘*kuuɲ ‘father, mother’s brother’, etc. (Shorto 2006)) clearly shows that Governor Xiguang must have brought Chinese-style marriage practices rather than introducing marriage. As for agriculture, Đông Sơn groups had been practicing wet-rice agriculture for centuries. One must assume again that Chinese-style agricultural practices were shared, and indeed, some words for oc-era implements, such as ‘harrow’ (cf. Alves 2014) appear to be part of this early era. One word of note from the excerpt is Chinese 義 yì ‘duty; justice’, for which Vietnamese nghĩa (lsv nghị, oc *ŋ(r)aj-s) is a likely Han-era borrowing. Finally, for perspective, while some 20,000 Chinese soldier-settlers were sent by the Chinese general 馬援/Mayuan to the region in the first century ce (Taylor 1983:49), this population was in contrast with a census population of nearly 750,000 in Jiaozhi, over 160,000 in Jiuzhen, and about 70,000 in Rinan, nearly a million in total (Taylor 1983:55–56). It is thus not surprising that, while a good number of notable oc words were borrowed, this was not a situation in which large-scale linguistic influence could have occurred.
“When Emperor Guangwu regained power, Xiguang served as administrator to Jiaozhi and Renyan to Jiuzhen. Thereupon, they taught them farming (plowing and sowing), mandated clothing (hats and shoes), initiated matchmaking processes, let them know of marriage, established schools, and guided them in proper behavior (ritual and righteousness).” (Translation by this author)
Overall, while such information does not provide certain evidence of those items as oc-era loanwords, it is a necessary and valuable part of the accumulated inter-disciplinary data. In the next sub-sections, words from the above-mentioned semantic domains (e.g., agriculture, literacy, household items, etc.) are discussed in terms of combined historical linguistic and actual historical information.
4.1 esv Terms Related to Agriculture
Explicit mention in Chinese historical records of the transmission of agricultural practices provides support for the timing of the esv items below. As noted, Chinese records suggest that the Jiuzhen district to the south of the Red River Valley had a less developed set of practices. However, the Đông Sơn Culture centered at Cổ Loa was a fully agricultural one and thus had ample agricultural practices and technology to provide for the large population in that region. Also, there are native, or at least non-Chinese, Vietnamese words as evidence of agricultural implements prior to Chinese arrival (cf. Alves 2015). Therefore, these esv items are not evidence of the introduction of agriculture, but rather the introduction of certain Chinese-style practices. Each word is briefly discussed.
‘To Sow’ – esv gieo ‘to sow’, noted by Lê (1967), has a reasonable initial and word shape, though the vowel is less certain as there are no corroborating examples which show /ɛw/ for oc *a. The level tone for a qusheng word suggests this is a first-century ce item. This specific word appears in the Houhanshu excerpt in section 4, and thus, despite its questionable vowel, it has reasonable overall support to suggest it is possibly an Eastern Han Dynasty loanword.
‘Well’ – esv giếng ‘well’ has expected segments for an oc loanword: a palatalized initial, a final velar corresponding to the later palatalized nasal final, and a qusheng tone for a shangsheng Chinese word. Though the excerpt in section 4 mentions agriculture in general but not wells specifically, wells are expected attributes of both agricultural areas and households. Though it is difficult to locate archaeological studies showing Chinese-style wells specifically in northern Vietnam at that time, that wells were common nearby in southern China from the second century bce is supported by the numerous models of wells found in Guangdong from the Western Han period (Jiao 2008). Thus, the borrowing of the word could certainly stem from this period.
‘Enclosure/stall/pen’ – The esv word ràn ‘enclosure/stall/pen’ is quite similar to the oc reconstruction and has /r/ for oc *r, making it a likely Han Dynasty loanword. The governor of southern Jiuzhen 任延/Renyan is recorded in the Houhanshu as having noted that those in Jiuzhen primarily practiced hunting and did not employ animals for farming (九真俗以射獵為業，不知牛耕) and then required the use of farming practices and tools (延乃令鑄作田器，教之墾闢). This indirect yet relevant information supports the claim that ràn could have been from that period.
4.2 esv Terms Related to Marriage Practices
Two of the esv terms in this category have some phonological problems, but the historical information about mandated marriage and match-making makes them reasonable candidates for status as Han-era loanwords. As noted above, it is assumed marriage-based kinship terms in proto-Austroasiatic demonstrate that these words represent the mandated practice of Chinese-style marriage, not the introduction of marriage overall.
‘Wife’ – As noted by Wang Li, esv vợ ‘wife’ has an expected esv tone and initial /v/, linking it with the Han Dynasty (Wang 1958:375). It is reasonable to assume this was borrowed in the situation of mandated marriage.
‘Husband’ – The esv word chồng ‘husband’ has a reasonable initial and overall word shape, though the vowel /o/ is harder to account for. Moreover, the tone category is problematic; this is a shangsheng word, but the esv word has a pingsheng tone. It would make sense for this to have been borrowed in parallel with ‘wife’, but the tone category and vowel make this a weak candidate for esv.
‘Match-maker’ – Regarding esv mai ‘match-maker’, while the tone and initial are non-distinctive, telling little about the timing of borrowing, the lower vowel suggests that this could be from the Han era. And according to the historical excerpt from the Houhanshu, this practice was explicitly spread in Jiaozhi and Jiuzhen.
4.3 esv Terms Related to Literacy
The Houhanshu excerpt notes the introduction of Chinese schools in northern Vietnam, and the lexical data bears this out.
‘Character’ – Haudricourt noted esv chữ ‘a written character’ (Haudricourt 1954a:74). The initial, vowel, and tone all support this as Han Dynasty era item. This author is not aware of any Chinese records referring to the establishment of Chinese schools during the earlier Western Han Dynasty (206 bce-25 ce). However, it is reasonable to assume that Chinese writing was in general introduced in that period.
‘Paper’ – Both Wang Li and Haudricourt noted esv giấy ‘paper’ (Wang 1958:376 and Haudricourt 1954a:77). The palatalized initial gi and tone category (i.e., qusheng for shangsheng) enable a dating of this term to the Han Dynasty. The history of paper, however, is somewhat more complicated. The Chinese eunuch 蔡倫/Cai Lun (50–121 ce) is credited with inventing papermaking, but samples of paper dating possibly to the second century bce have been found (Tsien 1985:2). It was not until the first century ce that Chinese paper began to be used for writing, gradually replacing over a period of centuries the use of bamboo and wooden tablets (Tsien 1985:1–2). It is difficult to find studies showing preserved archaeological evidence of paper from this period in northern Vietnam, where such material easily disintegrates in the tropical climate. This makes it impossible to date the loanword precisely other than to say it could date at least to the first century ce, but possibly earlier.
‘Writing brush’ – Vietnamese viết ‘to write’ is a possible Chinese loanword. Its initial, word shape, and tone are reasonable, with a ‘v’ initial suggesting a Han Dynasty borrowing. The rusheng tone category is consistent, though the tone alone cannot identify the time period. It is, however, a verb in Vietnamese. One can speculate, but not prove, that a Mon-Khmer style nominalizer prefix was used and later merged to ‘v’, as was the case for oc presyllables. Finally, within the historical context of introduced Chinese schools, this term may be associated with culturally-adapted loan word from that period.
‘To draw’ – Wang Li identified esv vạch ‘to draw’ (Wang 1958:374). This is a reasonable assumption considering the overall word shape, initial ‘v’, and a rusheng tone. It is unusual, however, to see a final palatal ‘ch’ in both the esv word and its lsv counterpart. Still, considering the historical information and overall semantic domain, this verb represents part of a literary tradition that was introduced at that time.
‘Drawing’ – As noted by Wang Li, esv vẽ ‘a drawing’ has an initial ‘v’ (Ibid.). It furthermore has a shangsheng tone for a qusheng word, dating this item potentially to the first century ce. It is a natural correlate of the action of drawing and thus the word ‘to draw’.
4.3 esv Terms Related to Households
Other esv words likely from this period include a variety of personal possessions or household objects. All four items below have been posited by scholars as being in the esv category: ‘hat’ by Haudricourt 1954a:75, ‘chopsticks’ by Wang Li (1948:67), ‘garden’ by Đào (1979:74), and ‘household’ by Wang Li (1948:78).
‘Hat’ – esv mũ ‘hat’ is a reasonable candidate for the Han Dynasty in all aspects: an expected initial, vowel, and tone. The shang-for-qu tone puts this word at least in the Han period. Furthermore, clothing in general, with explicit mention of headwear, is noted in the Houhanshu excerpt as a mandated item (though the fifth century excerpt contains the specific word 冠 (Md guān) ‘hat/crown’). Finally, this word has a comparable item in Proto-Tai *muak D1 (Li 1977:75), though the timing of the borrowing appears to differ from the Vietnamese form as it lacks a final stop, while Proto-Tai has a final *-k.
‘Chopsticks’ – The esv word đũa for ‘chopsticks’ is most likely from the archaic Chinese word 箸 (Md zhù). The initial ‘đ’ /ɗ/ matches the oc initial, and the diphthongization and tone category are both expected. The author cannot find any explicit written records related to the introduction of cooking, but with mandated aspects of Chinese lifestyle, required marriages, and a sizeable population of Chinese settlers and intermarriage, it is quite reasonable to assume this was the general period in which this word was borrowed. Furthermore, the Proto-Tai reconstruction *thɯ B2 (Pittayaporn 2009:82) and Proto-Hmong-Mien *drouH (Ratliff 2010:282), both likely from the same Chinese source, further highlight the transmission of this item in that region in the first few centuries ce.
‘Garden’ – It is difficult to know precisely what esv vườn ‘garden’ refers to—whether it is a garden of a wealthy home, a rural homestead, a farm, or another reference. Regardless, the initial, word shape, and tone have reasonable correspondences. Though there is no oc medial *-r-, the diphthongized ‘ươ’ is reasonable, as Austroasiatic *a sometimes has ‘ươ’ as a reflex in Vietnamese (e.g., Vietnamese nước ‘water’ for proto-Austroasiatic *ɗaak). As for the situation in which this cultural practice was transmitted, wealthy homeowners or other classes of Chinese settlers may have brought this Chinese cultural practice.
‘Household’ – esv họ ‘household’ has less distinctive phonological features to place it in the Han Dynasty. The tone corresponds to the final glottal stop of oc, and the vowel suggests a pre-Tang era item. However, neither the initial nor the single monophthong can distinguish the era of borrowing. However, Chinese censuses from mid-first century ce for population management and the collecting of taxes both require the concept of, and likely the term, for ‘household’. While the word could have been borrowed later, the first century seems a reasonable possibility.
5 Archaeological Evidence of oc Loans in esv
In this section, words related to metallurgy (‘iron’, ‘gold’, ‘silver’, and ‘to smelt/forge’), weapons (‘sword’, ‘knife’, and ‘crossbow’), and other objects (‘tomb’, ‘mirror’, ‘bowl’, and ‘roof tile’) are discussed in terms of known archaeological evidence. Additional corroborating historical evidence is added when available.
5.1 esv Terms Related to Metallurgy
As noted in Table 1, Chinese terms for the metals silver, gold, bronze, and iron have spread into Tai, Hmong-Mien, and Vietic. The archaeological record shows that gold objects are common in Bắc Bộ only from the Han era (Reinecke 2015). As for linguistic patterns, semantic shifts account for the relationship between Chinese for ‘yellow’ and Vietnamese for ‘gold’ and similarly between Chinese ‘white’ and ‘silver’ in Vietnamese.
‘Gold’ – esv vàng ‘gold’ is evidently derived from the Chinese word for ‘yellow’. The tone is non-distinctive, and the rhyme, while matching, is not informative. The initial ‘v’ is the sole distinguishing feature that identifies this as a likely Han era esv word. Archaeological studies of gold in the Đông Sơn era make this word if not an Eastern Han, then a Western Han loanword.
‘Silver’ – In a parallel fashion, esv bạc ‘silver’ appears to be derived from the Chinese word for ‘white’. However, little can be gained from the phonological shape: the initial, rhyme, and tone are non-distinctive, with only the final velar versus the lsv palatalized nasal indicating an early borrowing. Nevertheless, as with gold, silver is primarily a Han Dynasty era metal, making bạc a potential Han Dynasty borrowing.13
‘Iron’ – esv sắt ‘iron’, related to Proto-Vietic *k-rac (cf. Cuoi kʰrat7), has a reasonably comparable word shape and tone type, but the vowel and final do not fall into a recurring pattern of correspondences with oc *l̥ˤik. However, historical and archaeological evidence clearly shows the impact of Chinese contact in the region. The Iron Age in Mainland Southeast Asia is generally considered to begin around 500 bce. Higham notes the excavations of over 100 gravesites in Guangxi province with many Chu iron products from the Warring States period, and he then suggests this as a path iron metallurgy could have made its way to the Red River Valley (Higham 2014:197). Taylor notes that in Chinese historical documents, mention is made of a ban on iron and cattle to Nanyue by the Han administration in 185 bce (Taylor 1983:35). This makes it clear that iron had been in demand in the region at least in the third century bce, long before large Chinese populations had settled in northern Vietnam. Furthermore, Renyan is said to have mandated the use of iron implements in the central Jiuzhen district (Taylor 1983:34). Therefore, whether sắt is directly derived from Chinese may be questioned, but it remains difficult to argue that it is mere chance phonological similarity. The better explanation is that it was borrowed very early in the West Han Dynasty or even before, and any irregularity is due to how it was borrowed, such as being borrowed informally through trade and not necessarily via native Chinese speakers.
‘To smelt/forge’ – esv rèn ‘to smelt/forge’ is clearly a correlate of the metals discussed above. The initial ‘r’ for oc *r makes it a possible Han Dynasty item, though the pingsheng for a qusheng word suggests it is from the first century or later, not earlier. Vietnamese does have other words with the same sense: dọt ‘to forge/strike’ and gò ‘to forge/hammer’, though these do not have Proto-Vietic reconstructions and are of uncertain time depth.14 This matches the fact that the Bronze Age had entered the Red River Delta several centuries prior to the Han Dynasty. Regardless, the borrowing of Chinese ‘smelt/forge’ at this time, when metals were borrowed and Chinese cultural practices were mandated, along with the phonological evidence, makes this a highly probable ESV item.
5.2 esv Terms for Weapons
Although bronze knives and axes were part of Đông Sơn Culture several centuries prior to Chinese presence and cultural influence, many metal implements, such as swords, are Chinese in origin (e.g., Yang 2013:236–237, Higham 2014:207). Correspondingly, a significant percentage of words for metal tools and weapons in Vietnamese are of Chinese origin (Alves 2015). Longswords can be clearly traced to Chinese imports, while the metal crossbow mechanism has been debated regarding the direction of borrowing (i.e., whether from China to msea or the reverse). Regarding swords, while the Chinese 刀 dāo was traditionally a single-edged sword, the term dāo is now generally considered a knife, the Chinese 劍 jiàn was more specifically a double-edged sword and today refers to swords in general.
‘Sword (double edged)’ – esv gươm has a lenited initial for the oc initial cluster, it has medial -ư- for oc *-r-, and it has preserved final oc *-m. The pingsheng tone for oc qusheng makes this more likely a 1st century ce or later borrowing. It is definitely an object brought with the arrival of the Chinese.
‘Sword/knife (single edged)’ – The esv word dao ‘knife’ has a lenited initial ‘d’ /z/ to indicate an early time of borrowing, but the rhyme and tone do not indicate any specific timing. However, considering that swords in the region do extend into the Western Han Dynasty, and gươm is more likely from the Eastern Han, it is quite possible that dao was borrowed even earlier.
‘Crossbow’ – The Vietnamese term ná for ‘crossbow’ is quite possibly an ancient loanword considering its lower vowel ‘a’ and a qusheng tone for a shangsheng word. This word is reconstructable in Sinitic, Tibeto-Burman, Tai, and Austroasiatic languages of Mainland Southeast Asia (cf. Schuessler 2007:404), all with a pre-initial consonant (e.g., Proto-Austroasiatic *snaʔ Shorto 2006). In archaeological records, bronze and iron crossbow bolts dating to the 5th or even 6th centuries bce have been found in China (Wagner 1993:157–158, Blench 2015:7–8), making it likely much older than in Southeast Asia. In Vietnam, ten thousand crossbow bolts, including casting molds, were found at the Cổ Loa site from the 3rd century bce, with apparent Chinese influence, though less direct (Kim et. al. 2010:1020, Higham 2014:2004). There is a legend of supposed Chinese prince ruler of Cổ Loa, Thục Phán (Chinese 蜀泮/Shupan) creating a magical crossbow trigger (Taylor 1980:25). It is impossible to precisely interpret to which historical event this passage refers, if any, but it highlights the innovation of the crossbow and suggests its perceived importance in the region. Finally, while wooden crossbows may have been innovated in Mainland Southeast Asia, the metal crossbow mechanism that was developed in China does appear to be, based on archaeological data, an item shared southward.
5.3 esv Terms for Other Objects
A few other Han-era esv words for culturally-transmitted objects found in archaeological literature include ‘tomb’, ‘bowl’, ‘mirror’, and ‘(roof) tile’. All four forms were posited by Wang Li (1948, 1958) as belonging to 古漢越語 ‘Old Sino-Vietnamese’, corresponding to the term esv in this article.
‘Tomb’ – esv ‘tomb/grave’ in Vietnamese is complicated by having a set of three phonological forms (esv mả and mồ and lsv mộ), but the set provides a means of determining approximate age. Phan considers esv mả ‘tomb’ to be a Han Dynasty borrowing based on the lower vowel /a/ compared to later lsv /o/ (Phan 2013:78). The lsv reading of Chinese墓 (mn. mò, oc *C.mˤak-s, mc muH) is mộ, for which esv mồ ‘grave’ is an obvious related form. The pingsheng tone of mồ identifies it as an esv reading, but the shared vowel suggests a later time of borrowing than the Han Dynasty. esv words with the pingsheng for Chinese qusheng, but otherwise the same segments, are more likely to be part of the layer of Chinese loanwords starting from the 4th century Jin Dynasty, but possibly approaching the early part of the Tang Dynasty. Moreover, while Proto-Vietic has an apparent indigenous proto-word for ‘tomb/grave’ *g-roʔ/g-roh, the alternate Proto-Vietic form *-mah is further evidence of borrowing from a very early time depth. The numerous Han-style brick tombs (e.g., Taylor 1983:50, Higham 2014:207) leave little question that the first-century ce of the Eastern Han Dynasty would have been the latest time at which this practice entered the region and apparently the associated word in Vietic broadly.
‘Bowl’ – The esv term chén ‘cup/bowl’ has most of the phonological features expected of Han era words in terms of both segments and tone category. Having the qusheng tone for lsv shangsheng suggests an early period. In archaeological literature, Chinese glass bowls uncovered in the Red River valley date back to the around the beginning of the Common Era (Borell 2012:70–77), thus bordering the Western and Eastern Han Dynasties. This is arguably less persuasive evidence as the esv term is not as easily connected with a more generic item as, say, swords and crossbows. Still, regarding timing, the final esv ‘n’ (not ‘j’, as noted in Section 2.3) for oc *-r suggests it was borrowed after the nasalization of that segment occurred, so if it is tied to the Han, it is the Eastern Han.
‘Mirror’ – esv gương has excellent phonological correspondences: voiced spirant velar initial, diphthongized vowel with Vietnamese medial -ư- for oc *-r-, and a final velar for the palatalized final nasal. Finally, the pingsheng tone for a qusheng category may put this word in the Eastern Han. Archaeological finds in northern Vietnam Chinese from the first century ce include bronze mirrors (Higham 2014:207). I have found no other mention in archaeological studies of mirrors in msea prior to this period, suggesting the likely import of this from the north.
‘Tile’ – Regarding the esv word ngói ‘tile’, while having reasonable correspondences in initial, tone, and overall word shape, the final off-glide does not match comparative evidence. No documented modern varieties of Chinese (cf. Starostin 1998–2003) have the off-glide, nor does Proto-Hmong-Mien *ŋʷæX or varieties of Tai (e.g., Central Tai: Lei Ping vaa3; Lungming vaa6; Lungchow vaa6; Ning Ming ŋwaa6 (Hudak 2008:189)). According to Laurent Sagart (pc), the main support for the oc *-j off-glide is one instance of its rhyme in the Odes: it rhymes with 罹 lí ‘sorrow’ oc *[r]aj. An alternative explanation lies in the hypothesis that the 歌 rhyme group had a final *-r (Haudricourt 1954b:363 and Li 1974:265–266), resulting in Li Fang-Kuei’s hypothesized reconstruction of *ngwrarx (Li 1974:265). As noted in section 2.3, Austroasiatic *-r eventually developed into -j in modern Vietnamese, which could account for the off-glide in this word. This is parallel to the hypothesized connection between 源 ‘spring/source’ oc *N-ɢʷar and esv ngòi ‘canal’, similarly with oc *-r realized as final -j in Vietnamese. Thus, the linguistic data regarding esv ‘tile’ is reasonable but of mixed quality.
Historical records for that period do not explicitly mention roof tiles. However, archaeological studies show excavation of thousands of Chinese-style ceramic roof tiles at the northern Vietnam Cổ Loa site (Kim 2010:269–274) and (Kim et. al. 2010:1019–1020). Carbon dating demonstrates these are from as early as the beginning of the second century bce, that is, before the establishment of the Han administration in the region. Large numbers of these tiles at this site, compounded with the fact that they are not to be found elsewhere in northern Vietnam at this time, indicate emulation of, or possible direct borrowing, from other groups. Thus, archaeological evidence not only increases the likelihood that Vietnamese ngói is a Sinitic borrowing but also that it could have been borrowed as early as the second century bce in the Western Han Dynasty, or even possibly the end of the Eastern Zhou. If so, the final -j is modern evidence of a very ancient feature lost in Sinitic or Sino-xenic languages, and this word could possibly be among the oldest Chinese loanwords in Vietnamese.
6 Concluding Thoughts
Based on linguistic, historical, archaeological, and ethnological data, during the first four centuries of Sinitic-Vietic contact, Chinese culture appears to have had more of a broader cultural than linguistic impact. The fact that most of the esv are from the first century ce, while only a few can be tied to the earlier Western Han period (e.g., ‘iron’ and ‘roof tile’), suggests that the impact of language contact in the first two centuries was rather small. Chinese individuals and later Chinese administrations did steer the Đông Sơn people towards the Iron Age, simultaneous to the usage of precious metals and coins that modified economic practices. Chinese administrators also attempted to modify the Vietic people’s cultural customs in terms of literacy, marriage, and the household, though this appears to have taken hold more slowly, and to a significant extent only after the arrival of the Jin Dynasty Chinese refugees. Many unanswered questions about the sociolinguistic situation will require further historical study. Regardless, the number of words is much smaller than in later periods, but the words themselves and the broader semantic domains of borrowed vocabulary represent profound changes in early Vietic society.
The application of multiple datasets and analytical approaches has challenges and limitations, but it nevertheless provides clearer understanding of unrecorded aspects of sociocultural history in the region and can provide a means by which to continue such investigation. It is assumed that some of the ideas in this paper will be improved with further study, but hopefully, some of the ideas herein will be of use to those interested in sociocultural history in this border region of Southern China and Mainland Southeast Asia.
Appendix 1: List of Han or Jin Dynasty esv Words
Appendix 2: Vietnamese Quốc Ngữ consonants and ipa
Blench Roger . 2015. Ethnographic and archaeological correlates for a mainland southeast asia linguistic area. Paper for a volume from the conference ‘Beyond the Sanskrit Cosmopolis’ November. Singapore: Institute of Southeast Asian Studies.
Calo Ambra . 2008. Heger I bronze drums and the relationships between Dian and Dong Sun cultures. In Interpreting Southeast Asia’s Past: Monument Image and Text eds. Sharrock Peter Glover Ian C. Bacus Elizabeth A 208–224. Singapore: National University of Singapore Press.
Court Christopher trans. 1972. Two-way and Three-way Splitting of Tonal Systems in Some Far Eastern Languages. In Tai Phonetics and Phonology ed. Harris Jimmy G. and Noss Richard B. 58–86. Salaya: Central Institute of English language, Office of State Universities, Faculty of Science, Mahidol University.
SEALANGMon-Khmer Etymological Dictionary. http://www.sealang.net/monkhmer/dictionary/ (accessed 15 April 2015).
Zhu Xiaonong 朱曉農. 2009. Shēngdiào qǐ yīn yú fā shēng—jiān lùn Hànyǔ Sìshēng de fāmíng, 聲調起因于發聲——兼論漢語四聲的發明 (Phonation as the phonetic cause of tonogenesis: With special reference to the origin of Chinese tones). Yǔyán yánjiū jíkān 語言研究集刊 (第六輯)6:1–29. Shànghǎi 上海: Shànghǎi císhū 上海辭書出版社.
Yang Yong . 2013. New archaeological discoveries of the Bronze and early Iron Age in the Yunnan-Guizhou plateau and some related problems. In Crossing Borders: Selected Papers from the 13th International Conference of the European Association of Southeast Asia Archaeologists Vol. 1 eds. Tjoa-Bonatz Mai Lin Reinecke Andreas and Bonatz Dominik 230–238. Singapore: National University of Singapore Press.
I wish to thank Nam Kim for comments on archaeological issues in the paper, Liam Kelley on historical issues, and a reviewer for additional suggestions.
The term ‘Early Sino-Vietnamese’ was proposed by Phan in contrast with ‘Old Sino-Vietnamese’, from Wang Li’s term 古漢越語 The rationale is that ‘Old X’ often refers to a proto-language reconstruction, as in ‘Old Chinese’. Phan’s term is adapted throughout this article.
Vietic is the sub-branch of Austroasiatic that includes Vietnamese, varieties of Muong, and two dozen highly conservative varieties of related languages that are bisyllabic and with either four-tone systems or phonation/register systems. See the Glottolog entry for Vietic (http://glottolog.org/resource/languoid/id/viet1250) for more studies and also Vietic lexical data online at http://lacito.vjf.cnrs.fr/pangloss/index_en.htm.
The final labial makes this word particularly hard to consider to be related to the Chinese item. Thus, for now, it is listed with a question mark (?) in Table 1 to indicate its uncertain status.
The terms Sinitic and Vietic are used in this paper to refer respectively to (a) Old Chinese but potentially other closely related speech communities and (b) the predecessor to Vietnamese, varieties of Muong, and the various Minor Vietic groups. The terms Sinitic and Vietic are thus used to refer broadly to those two speech communities two thousand or more years ago.
Proto-Hlai clearly has borrowed ‘silver’ and ‘copper’ from Chinese (Norquest 2007:401–402). However, there are no reconstructions for ‘iron’, ‘steel’, or ‘gold’. Thus, Hlai is not included in this dataset.
The sources for Table 1 include Baxter and Sagart 2014 for oc and mc, Proto-Vietic of Ferlus on the SEALANG Mon-Khmer Etymological Dictionary, Proto-Tai of Li 1977, and Proto-Hmong-Mien of Ratliff 2010.
Note that the Sino-Vietnamese readings vary in the degree of usage in spoken or written Vietnamese and in their status as full words or elements in Chinese-type two-syllable compounds. When they refer to metals, the forms kim, ngân, cương, and thiết are essentially bound morphemes, while đồng is a free morpheme.
Appendix 2 has charts of the Vietnamese Quốc Ngữ alphabet and the ipa sounds for consonants and vowels. For tones, the numeric system is used to indicate contour (1 is the lowest, while 5 indicates the highest). Also, glottalization and differences in pitch based on syllable shape (i.e., open, final nasals, or final stops).
Also see Court’s added comparative tone chart in his 1972 English translation of Haudricourt’s 1961 article. That chart includes Chinese traditional categories, Tai script categories, Vietnamese (i.e., esv), Sino-Vietnamese, and Austroasiatic and Proto-Miao.
Starostin, Sergei, "Chinese Characters” in The Tower of Babel: An Etymological Database Project (1998–2003). Web. http://starling.rinet.ru/cgi-bin/query.cgi?root=config&morpho=0&basename=\data\china\bigchina.
The lexical data discussed in the sections on historical and archaeological evidence is focused on words for which there is corroborating extralinguistic data, whether direct or inferential evidence. Words lacking such evidence are presented largely in the section on phonological data, and in these instances, some words can be considered oc loanwords based solely on their shared semantic properties and phonological patterns. It is not the goal of this study to suggest that a lack of corroborating extralinguistic data need not be a reason to exclude such words from consideration, but rather to highlight the additional strength of claims when there are additional venues of support for these borrowings approximately two thousand years ago.
Moreover, this esv item could be the source of a widespread form seen in various languages of msea Austroasiatic, all identified on the Mon-Khmer Etymological Dictionary: Proto-West-Bahnaric *prak; Proto-Katuic *praʔ or *prak; Khmer prak, Nyah Kur (Southern) prák. They have preserved the initial /pr/ cluster. Another complex issue is the possible alternative source, Malay perak ‘silver’.
Admittedly, these two words may be related to more general verbs meaning ‘to hit’ and thus are derivative rather than basic words for ‘to forge’. It may be impossible to determine this with certainty.