An Updated Overview of the Austroasiatic Components of Vietnamese

Abstract

This article presents an updated view of the language history of Vietnamese from its native Austroasiatic roots, including key historical phonological, morphological, and syntactic features and developments; a characterization of its Austroasiatic etyma; and the context of this information in Vietnamese linguistic ethnohistory. It is now possible to make better supported claims and more precise characterizations due to improved understanding of the history of Austroasiatic and Vietic and their reconstructions, the nature and effect of language contact with Chinese, and the process of typological convergence of the ancestral language of Vietnamese. This study shows that, while Vietnamese is not a typologically characteristic Austroasiatic language, the Austroasiatic components of the Vietnamese lexicon and linguistic structure are more prominent than previously supposed.

1. Introduction

This study highlights the Austroasiatic lexical and structural elements (phonology, morphology, syntax) in modern Vietnamese and characterizes Vietnamese language history in terms of its Austroasiatic roots. This can be accomplished via the use of modern digital resources and tools, substantive comparative data from both modern languages and proto-language reconstructions, and recent ethnohistorical research in the region.

The Austroasiatic origin of Vietnamese has long been recognized, but there was considerable debate in the first half of the 20th century. Uncertainty about the linguistic affiliation of Vietnamese was due to (a) the limited amount of available comparative data, (b) what is now understood to be major changes in phonological word structure from Austroasiatic to Vietnamese, (c) the complex language contact situation in Mainland Southeast Asia over past millennia, (d) the intense lexical influence of Chinese on Vietnamese and neighboring languages, and (e) the typological resemblance of Vietnamese to Chinese and Tai languages, among other factors. Even though the Austroasiatic origin of Vietnamese was widely accepted, there has been considerable focus in linguistic literature on the Chinese influence on Vietnamese and less detail about the Austroasiatic elements. It is only in the 21st century that enough digital lexical resources and proto-language reconstructions (and crucially, advances in reconstructions of Chinese, Proto-Austroasiatic, and Proto-Vietic), combined with ethnohistorical and archaeological information, have become available to create a clearer picture of both the retentions of and changes to the original Austroasiatic components of Vietnamese.

The Chinese typological and lexical features of Vietnamese are a highly visible aspect of Vietnamese. Vietnamese has significant typological similarity with Chinese in its tone system, isolating morphology, vocabulary, and disyllabic lexical compounding. Also, Chinese loanwords in Vietnamese, which number in the thousands, are an essential part of Vietnamese etymological study and are seen in its literary tradition and cultural matters. However, despite the effect of two thousand years of language contact with Sinitic, the similarities should not be overstated. First, the typological convergence leading to the modern template happened over a period stretching from the beginning of the 1st millennium CE to the 1800s: this was not a sudden event. Next, many native Austroasiatic and Vietic features have been retained, while conversely, many features seen among Chinese languages did not become part of Vietnamese typology (e.g., the A-not-A question pattern, pre-nominal adjectival modifiers, post-nominal possessive markers, etc.). Also, though the lexical impact of Chinese is statistically significant, Vietnamese still has a robust core native Austroasiatic and Vietic component. However, the latter aspect has been difficult to assess, as previously, knowledge of Austroasiatic language history was less developed, and no reliable Proto-Austroasiatic (hereafter pAA) reconstructions were available.

The primary goal of this article is to characterize the historical changes from Austroasiatic to Vietnamese at primary linguistic levels: phonological, morphological, and syntactic. An attempt is made to identify features that are considered retentions of Austroasiatic, while putting aside those which may arguably be considered the result of areal contact. Despite the Chinese-like features of Vietnamese, Vietnamese still has core Austroasiatic elements at all linguistic levels, some of which can only be seen through reconstructions, but others which are much like pAA or any other modern Austroasiatic language.

2. Overview of Vietnamese Language History

To provide context for the historical linguistic hypotheses in this study, this section summarizes past stages of the language situations of Vietnamese supported by ethnohistorical and archaeological information.

Table 1. Tentative stages from Austroasiatic to modern Vietnamese.

Period	Chronological Points of Reference	Hypothesized Typological Traits
Proto-Austroasiatic	c. 2000 BCE, pre-Metal Age, dispersal of Austroasiatic peoples in Mainland Southeast Asia	Disyllabic, nontonal, CCVC syllables
Proto-Vietic	c. 1000 BCE, Metal Age, settled Austroasiatic groups	Disyllabic, nontonal, CCVC syllables, emerging phonation2 (?)
Heavy Sinitic–Vietic contact	First half of 1st mill. CE, Chinese administration	Disyllabic, nontonal, CCVC syllables, emerging north–south speciation in Vietic (?)
Proto-Viet-Muong	By 1000 CE, end of Chinese administration, start of local administrations	Beginning of reduction in disyllabicity (?), CCVC syllables, tonal (pre-register)
Archaic Vietnamese	Early 2nd mill. CE, use of nativized Nôm script	Reduced disyllabicity, CCVC syllables, tonal (post-register (?))
Middle Vietnamese	Mid-2nd mill. CE to the early 1800s, increased regional trade and contact	Monosyllabic, CCVC with progressive loss of onset clusters, tonal
Modern Vietnamese	Late 2nd mill. CE, Late Colonial period	Monosyllabic, CGVC, tonal

Note: (?) indicates that a feature cannot be demonstrated with high certainty based on comparative data but is feasible based on the larger context.

presents the posited historical linguistic periods, their approximate timing, and hypothesized stages of typological characterizations from Austroasiatic to modern Vietnamese.

Table 1 begins with the Austroasiatic dispersal,1 over 4000 years ago, at which point the ancestral language of Vietnamese would have been a typical Austroasiatic language, with sesquisyllabic word structure (i.e., disyllabic words with an iambic stress pattern) and onset clusters with medial -l- and -r-, but no tones. At some point after the dispersal, the Vietic group speciated but kept the Austroasiatic morphophonological template, like other modern conservative Vietic languages. The beginning of the Metal Age in the region in c. 1000 BCE is noted as a chronological point of reference, as it was long after the dispersal, enough for linguistic differentiation, but at a time when presumably there was contact with other groups and potential sociocultural developments and corresponding linguistic changes. After contact with the incoming Sinitic ethnolinguistic community around the beginning of the Common Era, there may have been increasing linguistic distinctions between the northeast and southern Vietic speech communities, as suggested by the many early Chinese cultural loanwords present in Vietnamese but not Vietic groups to the south (Sidwell and Alves 2021, p. 184).

The hypothesized lexical differentiation of Vietic sub-branches, with associated cultural implications, precedes further speciation of the northeast group consisting today of Vietnamese and a few dozen lects collectively called Muong.3 The Viet–Muong languages share lexical and phonological features, including tone system patterns. The timing of tonogenesis in this group is unknown, but Alves (2018, p. 17) has posited that it happened by the end of the 1st millennium CE in Vietnamese, though whether this is also the case in Muong lects has not yet been determined. Also, no analysis has shown a precise phylogenetic tree of this sub-branch, and it has been posited that, considering the lack of shared phonological innovations, no precise tree is even possible (Phan 2012).

Next, at the start of the 2nd millennium, a variety of Chinese of a hypothesized Chinese ethnolinguistic group merged with the Viet–Muong language(s) (Phan 2013). The speech community after that merging can be said to have spoken Archaic Vietnamese. At this point, thousands of Late Middle Chinese words—in both literary and vernacular Vietnamese—became part of Vietnamese and Muong, which in Vietnamese is typically called từ Hán-Việt or ‘Sino-Vietnamese vocabulary’.

As for the phonological word at that time, it is difficult to reliably determine the Proto-Viet–Muong template given the transitional nature of the language, but there was likely retention of presyllables into the early 2nd millennium. Two sets of evidence for this are (a) lenited onsets in Vietnamese that softened in intervocalic position (discussed in Section 3 before Table 5) and (b) textual evidence from the 1200s suggesting that Archaic Vietnamese still retained presyllables.4 The Chinese-based Nôm orthography at that time offers evidence of a lingering number of disyllabic words (Shimizu 2015). Regarding the timing of the process of monosyllabification, we can only speculate that there may have been decreasing percentages of disyllabic words over the centuries, but there are insufficient data to offer details. However, by the Middle Vietnamese stage, as exemplified by the 1651 Romanized Vietnamese–Portuguese–Latin dictionary (Rhodes [1651] 1991), the process was complete, as Vietnamese had no prosodic disyllabic words. However, it retained some onset clusters with -l- at that time. It is only by the mid-1800s that mainstream Vietnamese had its complete modern template with only medial [-w-] (in CGVC, G refers to a ‘glide’) (though some rural Vietnamese dialects in north–central Vietnam have been reported to have/tl/onset clusters (Nguyễn 1999)).

Thus, based on available evidence, in this hypothesized scenario, the Vietic ancestor of Vietnamese retained its Austroasiatic template into the late 1st millennium CE, though it had developed tones by that time. Then, at some point in the first half of the 2nd millennium, it completely lost presyllables. Finally, in the 19th century, Vietnamese largely reached its modern largely CVC syllable canonical shape, which resembles the typology that Chinese had developed by the period of Early Middle Chinese in the first half of the 1st millennium CE. Thus, it can be said that only in the past half millennium did Vietnamese take only a truly Chinese-like appearance, and that prior to that, it was still very much an Austroasiatic-like language.

The next sections review the phonological, morphological, and syntactic matters in more detail and connect the Austroasiatic origins to modern Vietnamese features.

3. Historical Phonological Aspects

Thus, Vietnamese has been substantially restructured from the original Austroasiatic template, and today, the phonological typology of Vietnamese resembles that of Chinese in many ways. However, as noted, this situation was only fully reached in the last several hundred years. The Vietnamese syllable is most often a simple CV(C) structure, with only [-w-] as a medial in at most 10 percent of Vietnamese syllables (Kirby and Alves 2022), or as little as 5.2 percent (Trần and Vallée 2009, p. 232). Vietnamese has a complex tone system (i.e., six contour tones with phonation features, while two tones exhibit phonetic distinctions in syllables with p/t/k codas) that maps directly onto the system of the Chinese historical phonological tradition (i.e., the four ping/shang/qu/ru tone categories divided into eight tones by height according to onset voicing). In early 20th century research, this situation created uncertainty regarding its linguistic affiliation, with speculation of Tai or Chinese origins (see Gage 1985 and Alves 2006 for a summary). However, language contact and typological convergence extends beyond Chinese. Vietnamese is in a larger sphere of language contact, sharing the typology of Kra–Dai and Hmong–Mien languages of southern China and Mainland Southeast Asia. These languages are also largely monosyllabic and have typologically comparable tone systems.

Despite linguistic influence due to language contact, we can still identify features that are retentions of Austroasiatic features. Among the pAA etyma in Vietnamese, pAA vowels *i, *a, *u and common pAA consonants *m/n/ɲ/ŋ, * k, and *l/r in onsets of main syllables are readily found retained in modern Vietnamese reflexes. Indeed, some modern Vietnamese words closely or completely resemble their original proto-language source forms (main syllables in cases of pAA disyllabic words), including both pAA and Proto-Vietic (pV hereafter), as in

Table 2. Vietnamese words with retentions of pAA main syllable forms.

N. Vietnamese	pV	pAA	Gloss
cɔː5 (chó)	*cɔʔ	*cɔʔ (AA030)	dog
ciːm1 (chim)	*ciːm	*ciːm (AA016)	bird
kaː5 (cá)	*ka:ʔ	*kaʔ (AA009)	fish (n)
kɔːn1 (con)	*kɔːn	*koan (AA213)	child
mat6 (mặt)	*mat	*matˀ (AA024)	eye
muːn6 (mụn)	*mu:nʔ	*muːnˀ (AA070)	pimple
ɲaː2 (nhà)	*ɲaː	*ɲaːˀ (AA110)	house
laː5 (lá)	*sə.laːʔ	*sə.‘laʔ (AA010)	leaf
ɽuːŋ6 (rụng)	*ruŋʔ	*ruŋˀ (AA444)	fall, drop (v)
ɽuː5 (rú)	*bə.ruːʔ	*bə.‘ruʔ (AA322)	hill, mountain
ɽuəj2 (ruồi)	*Cə.rɔ:j	*roaj (AA028)	fly (n)

. Basic vocabulary items such as ‘dog’, ‘bird’, ‘fish’, and others are clearly connected to their pAA counterparts, which have 4000-plus-year histories.

Following are notes on data presentation in this article. Subsequent tables with data contain (a) neutral IPA transcriptions of Vietnamese words (i.e., not matching a dialect but rather based on the Vietnamese orthography, following the system of phonetization of (Kirby 2008), followed by Vietnamese Quốc Ngữ romanized orthographic spelling in parentheses, (b) pV reconstructions based on those of Ferlus as in the SEALang Mon-Khmer Etymological Dictionary but modified to parallel the format of the pAA reconstructions, and (c) the pAA reconstructions of Sidwell’s (2024) 500-word list with their specified indicators (e.g., AA030) after them. Some potential pAA reconstructions (mostly by Sidwell in the spreadsheet assembled by Sidwell and Alves) that currently lack identifying numbers are marked ‘AAP’, while others are from Shorto’s (2006) reconstructions (marked with # and the reference number) when they are attested in at least several Austroasiatic branches, indicating deep history in Austroasiatic but not necessarily back to the pAA stage.

Naturally, many other pAA segments have changed over time in Vietnamese, but the phonological slots have been retained. The main-syllable onset system has undergone restructuring not only in voicing but also manner of articulation, with some becoming the implosives /ɓ/ and /ɗ/, as in

Table 3. Onset voicing of pAA and corresponding tone height in Vietnamese.

Vietnamese	pV	pAA	AA Gloss
ɓɛː3 (bẻ)	*pɛh	*pɛːh (AA188)	break in two
ɗaːn1 (đan)	*taːɲ	*taːɲ (AA005)	weave (v)
ciːn5 (chín)	*ciːnʔ	*ciːnʔ (AA017)	cooked, ripe
ʔɔːj5 (ói)	*ʔoːlʔ	*ʔo:lˀ, *ʔuəl (AA213)	choke, vomit (v.i.)
ɓuːŋ6 (bụng)	*buŋʔ	*buŋ, *buuŋʔ (Shorto #624a.b)	belly
ɗoːj2 (đồi)	*doːl	*doːl, *duəl (AA423)	hill, mound
caj6 (chạy)	*ɟalʔ	*ɟarˀ (AAP)	run (v)
kam2 (cằm) ‘chin’	*gam (tent.)	*-ga:m (AAP)	molar tooth

, or fricatives such as /v/, /z/, or /ɣ/, as in Table 5, as discussed below.

Changes in onset voicing, which have occurred in many languages in the region, is seen in pAA and pV etyma in Vietnamese, though the details of this change in Vietnamese are not well understood. Onset voicing is also related to tone height in Vietnamese, as in other neighboring tonal languages: original pAA voiceless stops *p/t/c/k occur in syllables with upper register tones in Vietnamese (1 = the ngang tone, 3 = the hỏi tone, 5 = the sắc tone), while proto-language voiced onsets *b/d/g occur in those with lower register tones (2 = the huyền tone, 4 = the ngã tone, 6 = the nặng tone). For example, pAA *ciːnʔ—‘cooked, ripe’—is the source of a modern Vietnamese word with the upper register tone 5, while *ɟarˀ—‘run (v)’—is the source of a modern word with the lower register tone 6. While counterexamples exist, and not all details are accounted for, the pattern is reasonably consistent.

Some changes in the onset system have resulted in some original sounds being lost. For instance, pAA implosive onsets have become Vietnamese nasals instead (Ferlus 1992, p. 117), as in

Table 4. Nasal onsets from pAA implosive *ɓ/ɗ.

Vietnamese	pV	pAA	Gloss
maŋ1 (măng)	*tə.ɓaŋ	*tə.’ɓaŋ (AA023)	bamboo shoots
mɔː3 (mỏ)	*kə.ɓɔh	*cə.’ɓɔh, *cə.’ɓu:ˀ (AA442)	beak
nɨək5 (nước)	*ɗaːk	*ɗaːkˀ (AA039)	water
nuːŋ1 (nung)	*ɗuŋ	*ɗɨŋ, *ɗuŋ (AA492)	cook (v)
nɔːŋ1 (nong)	*ɗoːŋ; *ɗoːŋʔ	*kə.’ɗo:ŋ (AA214)	winnowing basket
mɔːk5 (móc)	NA	*ɓɔkˀ, *ɓuəkˀ (AA392)	hook (v)
muːk5 (múc)	*ɓuːk	*ɓəkˀ (Shorto #367)	draw water

. And yet, though not retained from a proto-language stage, the implosives /ɓ/and/ɗ/ have been revived in the modern Vietnamese phonological inventory by developing from voiceless plosives, as seen in Table 3. Implosives are common in Austroasiatic languages, but the WALS database (Maddieson 2013) shows that such sounds are not as common in languages in other language families in Southern China and Southeast Asia. Thus, we can speculate that there has been a regional typological pressure towards retaining or redeveloping these. In Vietnamese, implosive onsets are distinct from plosive onsets in that (a) they tend to correspond to upper register tones 1, 3, and 5, unlike plosive voiced onsets, which correspond to lower register tones 2, 4, and 6, and (b) they do not undergo lenition to become voiced fricatives in intervocalic position, unlike plosive consonants, as in Table 4.

One particularly notable change that makes Vietnamese typologically less like other Austroasiatic or even other Vietic languages is its series of voiced onset fricatives: /v/, /z/, and /ɣ/ (represented in Vietnamese orthography by ‘v’, ‘d’ and ‘gi’ for /z/, and ‘g/gh’ respectively).5 These developed from main syllable onsets in intervocalic position (see Ferlus 1992, p. 113 and updates by Alves 2024), as in the examples of Table 5, in which all the reconstructed forms are disyllabic.

Table 5. Voiced fricative onsets in intervocalic position in pAA presyllables.

Vietnamese	pV	pAA	Gloss
vɨən6 (vượn)	*kə.waɲʔ	*kwaaɲʔ (Shorto #934)	gibbon
voːj1 (vôi)	*kə.puːr	*knpur (Shorto #1636)	lime (mineral)
zaːm1 (dam) ‘field crab’	*kə.ta:m	*kə.’taːm (AA004)	crab
zɔː5 (gió)	*kə.jɔːʔ	*kə.’jaːl; *kə.’jɔʔ (AA075)	wind, air (n)
ɣəw5 (gấu)	*cə.guːʔ; cə.kuːʔ	*cguːʔ; ckuːʔ (Shorto #1817)	bear (n)
ɣoːj5 (gối)	*tə.kuːlʔ	*Cə.’kuːl (AA326)	knee
ɣaːw6 (gạo)	*rə.koːʔ	*rəŋ.’koːˀ (AA066)	husked rice

Table 5. Voiced fricative onsets in intervocalic position in pAA presyllables.

Vietnamese	pV	pAA	Gloss
vɨən6 (vượn)	*kə.waɲʔ	*kwaaɲʔ (Shorto #934)	gibbon
voːj1 (vôi)	*kə.puːr	*knpur (Shorto #1636)	lime (mineral)
zaːm1 (dam) ‘field crab’	*kə.ta:m	*kə.’taːm (AA004)	crab
zɔː5 (gió)	*kə.jɔːʔ	*kə.’jaːl; *kə.’jɔʔ (AA075)	wind, air (n)
ɣəw5 (gấu)	*cə.guːʔ; cə.kuːʔ	*cguːʔ; ckuːʔ (Shorto #1817)	bear (n)
ɣoːj5 (gối)	*tə.kuːlʔ	*Cə.’kuːl (AA326)	knee
ɣaːw6 (gạo)	*rə.koːʔ	*rəŋ.’koːˀ (AA066)	husked rice

While many instances retroflex onsets in Vietnamese words are retentions of retroflex onsets in Middle Chinese loanwords (Ferlus 1992), in native Vietnamese words, such sounds often correspond to earlier onset clusters, as in

Table 6. Retroflex onsets from onset clusters.

Vietnamese	pV	pAA	Gloss
ʂaːŋ5 ‘sáng’	*plaːŋʔ	*plaaŋʔ (NA)	shine (v)
ʂɔːk5 ‘sóc’	*prɔːk	*pə.’rɔːkˀ (AA091)	squirrel
ʂaː5 ‘sá’	*kraːʔ	*kraʔ (AA162)	path, road
ʈajŋ1 ‘tranh’	*plaŋ	*pə.’laŋ (AA060)	thatch-grass
ʈaːj5 ‘trái’	*pə.leːʔ > *pleːʔ	*pə.’leʔ (AA043)	fruit

. In a development distinct from the just noted lenition, some of these onset clusters appear to have come from previous presyllables that at some point lost syllable status. There is a regional process in Austroasiatic of presyllable reduction due to the iambic stress pattern weakening presyllables, leading to a variety of manifestations of the so-called sesquisyllabic typology (Pittayaporn 2015).

The Vietnamese tone system has served as a key example of the historical process of tonogenesis. Haudricourt’s (1953) seminal study of the origins of Vietnamese tones as a seminal study, and that theoretical model has been applied to the study of the emergence of tones in Chinese, Kra–Dai, Hmong–Mien, and Tibeto-Burman languages, as well as other Vietic languages. Among cognates in other Austroasiatic languages, there are correspondences between Austroasiatic coda fricatives *-h/-ɕ and Vietnamese tones 3 or 4 and between final voiceless stops *-p/-t/-k/-ʔ and tones 5 or 6, as exemplified in

Table 7. Tone categories corresponding to pAA coda types.

Vietnamese	pV	pAA	Gloss
taj1 (tay)	*siː	*siːˀ (AAP)	hand, arm
sɨəŋ1 (xương)	*tʃ.ʔaːŋ	*cə.’ʔaːŋ (AA012)	bone
ɗoːj2 (đồi)	*doːl	*doːl, *duəl (AA423)	hill, mound
ruəj2 (ruồi)	*mə.rɔːj	*roaj (AA028)	fly (n)
ɗoː3 (đổ)	*toh	*toh (AAP)	pour
ɓaj3 (bảy)	*pəs	*pɔh, *pəɕ (AA209)	seven
muːj4 (mũi)	*muːs	*muːh, *muːɕ (AA003)	nose
ɣɔ:4 (gõ)	*C.kɔ:h ;*gɔ:h	*guəh ‘knock (v)’ (Shorto #1978)	knock (v)
caw5 (cháu)	*cuːʔ	*cuʔ (AA096)	grandchild
muːn6 (mụn)	*mu:nʔ	*muːnˀ (AA070)	pimple
kat5 (cắt)	*kac	*katˀ (AA186)	cut (v)
ʂɔːk5 (sóc)	*pə.rɔːk	*pə.’rɔːkˀ (AA091)	squirrel

. Crucially, Vietnamese has lost final *-h/ɕ and *-ʔ, while reflexes of these segments are found among many Austroasiatic languages. The rephonologization of previous codas as tones in Proto-Viet–Muong was likely stimulated by contact with Chinese. However, tonogenesis in Viet–Muong may also be seen as the result of a larger trajectory predating Sinitic–Vietic language contact. Pharyngeal and registral features and phonation have been seen as stimulants of suprasegmental features in Austroasiatic languages (e.g., Gehrmann 2022), such that tonogenesis could have occurred regardless of contact with Chinese. Nevertheless, the borrowing of numerous Chinese words with matching tones allows for the influence of the Chinese tone system on the Vietnamese one.

Overall, while various pAA consonants and vowels have changed in Vietnamese, many have still been retained in their original forms. The later stage of pV phonology also shows examples of changes from pAA phonology, but in many cases, the reconstructions are very close to or the same as the pAA reconstructions, so they also help demonstrate the ultimate pAA origin of these speech sounds and words.

Regarding the influence of Chinese, despite the evidence of its impact on Vietnamese phonology, the native elements of Vietnamese phonology are still evident. In a study of some 8000 Vietnamese phonemically distinct monosyllabic words (with the number including distinctions due to tones), Kirby and Alves (2022, pp. 122–23) determined that, while native words in Vietnamese have statistically prominent phonological features identifying them, Sino-Vietnamese loanwords (i.e., Late Middle Chinese and the system of readings of Chinese characters) mostly do not. For instance, the onset /ɣ/ ‘g/gh’ occurs only in native Vietnamese words (and very early Chinese loanwords, at least several centuries before the Sino-Vietnamese layer), not in Sino-Vietnamese Chinese character pronunciations. Another example is the diphthong /uə/ ‘uô/ua’, which is found in only nine Sino-Vietnamese words but in 279 words of native Vietnamese (again, the latter including some early Chinese loanwords). Regarding other phonemes of Vietnamese, there are consistently higher ratios of those sounds in native words than among Sino-Vietnamese vocabulary such that no phonological features specifically mark syllables as Chinese loanwords.

4. Native Lexicon: The Vietic and Austroasiatic Components

We now consider the native portion of the Vietnamese lexicon in terms of quantitative data and semantic/cultural domains of words. The Vietnamese lexicon is well known for its Chinese component, though in the past, inflated numbers (e.g., 70 percent) of the portion of Chinese loanwords have been mentioned without reference to a specific study or research method that determined such numbers. More recently, numbers based on focused counts of entries in dictionaries (Hùng et al. 2018, pp. 348–51) and selected lexical sets (Alves 2009, p. 619) show a range of 25 to 35 percent.6 Thus, according to these recent studies, the bulk of the Vietnamese lexicon—some two-thirds to three-quarters—is native, not borrowed. Based on data assembled for this study, of the native portion of the Vietnamese lexicon, (a) nearly two hundred words are pAA etyma or at least early Austroasiatic items, (b) at least several hundred are likely pV items,7 (c) some can be reconstructed only to the Proto-Viet–Muong stage, and (d) an uncertain number but likely thousands of words can only be found in Vietnamese.8

Which strata words belong to can be determined by collating lexical data for Vietic languages, determining which are widespread in Vietic and which are restricted to Vietnamese and Muong lects, identifying those which are likely Austroasiatic etyma, and excluding those of likely external origin (primarily Chinese, other Austroasiatic languages, Tai, and Indic).9 Ferlus’s Vietic reconstructions contained in the Mon–Khmer Etymological Dictionary, combined with a dozen other sources of Vietic languages, constitute a spreadsheet with over 1200 reconstructions that I have assembled and to which I have added reconstructions. Vietic reconstructions have then been considered for their potential relationship to other Austroasiatic etyma, though this has previously been a problem as a developed set of pAA reconstructions was lacking.

An early statistical study of the lexical relationship of Vietnamese to other Austroasiatic languages is that of Thomas and Headley (1970), who used 100-word lists of basic vocabulary. Their study showed that inter-branch lexical relatedness of the languages mostly ranged from 20 to 30 percent, and such was the case for Vietnamese and Muong, making them as close to Austroasiatic languages as any others. As for studies that propose pAA etyma of Vietnamese, two are by Hô (1992) and Alves (2020a). Hồ’s work, listing about 200 words, does not explicitly refer to pAA reconstructions, making those claims tentative. Alves (Ibid.) similarly identified about 200 possible Austroasiatic items (many of which are not the same as Hồ’s) and another 500 pV etyma in Vietnamese. For the Austroasiatic words, he referred to Shorto’s (2006) 2200 Mon–Khmer reconstructions, but as Shorto’s work should be regarded as a work in progress, Alves also considered the geography of Shorto’s reconstructions and utilized only items with distribution widespread enough to validate a claim of proto-language status.

However, only recently has Sidwell (2024) provided a list of pAA etyma reconstructions based on criteria to focus on words which have demonstrable historical depth. Applying Sidwell and Alves’s (2023) approach in determining which words are reconstructable to the time of the pAA dispersal over 4000 years ago, Sidwell’s reconstructions consist only of items with attestations in Austroasiatic branches both inside and outside of Mainland Southeast Asia, the latter consisting of Munda, Nicobarese, Aslian, and Khasian. This geographic distribution is strong evidence for the posited deep-history time-depth. In Sidwell & Alves’s working spreadsheet, besides the 500 pAA reconstructions, a few hundred more have been marked potential pAA items. Another few hundred items are found only in Mainland Southeast Asia, which means they do not meet Sidwell & Alves’s criterion to be high-confidence pAA etyma, but they still may have substantial time-depth in Austroasiatic language history. Thus, both of these latter categories are also considered in the current study.

The results of checking these data are that 146 Vietnamese words are modern reflexes of pAA etyma (Appendix B), and an additional 46 items can be regarded as potential pAA etyma or at least words early in Austroasiatic history (Appendix C), a total of 192 words. While the number of Austroasiatic words in Vietnamese is modest, those words include a wide range of basic vocabulary (e.g., body parts, common animals, natural phenomena, basic actions, etc.). Moreover, other words provide somewhat more specific information about early Austroasiatic activities (e.g., to cook, to dream, to meet, to gather/associate, to comb, to hunt, etc.) and families (e.g., grandchild, child, mother). Crucially, many of the words match archaeological portrayals of the daily life of settled agriculturalists in this region (e.g., Higham 2021). Some of these words are related to farming practices in the Neolithic period of Austroasiatic history (e.g., rice, bran/husk, mortar, pestle, winnowing basket, to plant seed, etc.). The 13 function words in the lists include number terms, units of time, and deictic terms. It is notable that, while neighboring Kra–Dai languages have mostly lost their native numeral word systems, Vietnamese has retained an Austroasiatic set, in spite of other significant lexical borrowing from Chinese.

Table 8. Semantic domains of Austroasiatic vocabulary in Vietnamese.

Semantic Domains	No. of Words	Examples of Senses (Complete Data in Appendix B and Appendix C)
Actions of general nature	36	to comb, to hunt, to open, to cover, to scatter/sprinkle, etc.
Body terms (nouns and verbs)	32	eye, gall/bile, to open one’s mouth, hair, to choke/vomit
Animals	28	dog, crab, fish, python, horn, etc.
Actions for motion	16	to carry on one’s back, to pour, to fall/drip, to wade/swim, to run, etc.
Agricultural and vegetation	14	husked rice, bran/husk, taro, fruit, to plant seed, etc.
Food and related actions	7	cooked/ripe, to suck(le), to cook, etc.
Function, time, and numeral words	13	this, that, you, year, day, four, nine, etc.
Natural environment	11	hill, mountain, rain, cloud, gravel, etc.
Items and related actions	11	house, pestle, mortar, to weave, to scoop/bale, etc.
Cultural aspects (cognition, sense, kinship, society, and speech)	10	grandchild, mother, to dream, to meet/associate, to ask, to call out, etc.
Spatial	11	distant, deep, supine, side, etc.
Sense	8	sour, bitter, to rub/stroke, warm (adj), etc.

lists the various categories and provides examples taken from Appendix A and Appendix B.

Altogether, despite the seeming modest number, these nearly 200 words form a core of Austroasiatic etyma in Vietnamese that can hardly be considered trivial. Indeed, these words highlight the Austroasiatic essence of Vietnamese.

5. Morphology and Syntax

Identifying the Austroasiatic elements of Vietnamese morphology and syntax requires comparison with both modern languages and earlier stages of Austroasiatic and Vietic. While there are more grammatical descriptions of Austroasiatic languages than ever, many of them are short treatments, with only some more thorough descriptions in articles and dissertations. Regarding pAA syntax, recent advances in understanding have been made, but the claims in this area are still general and somewhat speculative. As for morphology, some affixes are well represented among Austroasiatic languages, such that categories of derivational morphology can be reconstructed. Austroasiatic-style reduplicative can be assumed to have been an operational word formation strategy, though no reduplicative words have been reconstructed. The connections of these matters to Vietnamese are less direct than for the lexicon and speech sounds but still demonstrable.

5.1. Syntactic Structures

One impediment to developing hypotheses of Vietnamese diachronic syntax is that some features are shared by many neighboring languages in the region regardless of language family, such as SVO/AVP clause structure and post-nominal modifiers in noun phrases. Thus, it is difficult to distinguish retention from contact-induced typological convergence. Still, according to recent work, we can posit that pAA had head-initial phrase structure in both clauses and noun phrases, while modern Vietnamese has a blend of this due to later contact-induced factors. These issues are discussed below.

In the past decade, pAA clause structure has been hypothesized to be verb-initial (Jenny et al. 2015; Jenny 2020). Jenny (2020) provides supporting evidence for this claim from multiple sources: (a) main clauses in some Khasian, Nicobarese, and Aslian languages, (b) embedded clauses of some Palaungic languages, and (c) noun compounds of Munda languages. If this hypothesis is valid, one distinct possibility is that a shift to verb-medial main clause structures happened among many Austroasiatic languages in the region due to language contact, though this cannot as yet be determined. As for pAA noun-phrase structure, no specific studies have presented a complete argument for a reconstruction, but a head-initial type is strongly indicated. Noun phrases with strictly post-nominal modifiers occur in Old Khmer texts from the pre-Angkorian period of the early 1st millennium CE, and head-medial order in Vietic appears to be later a development due to the spread of classifier phrases at least partly due to language contact with Chinese (Alves 2020c, p. 63).

For Vietic clause structure, Alves (2020c, p. 69) reconstructs topic–comment structures (with subjects/agents morphologically indistinguishable from topics), verb-medial clauses, middle-passive constructions (lacking any explicit lexical means of marking the passive), and a right-branching VP with objects/patients, other complements, and adjuncts after verbs. These hypothesized proto-language features are feasible considering the most general properties of Vietic languages and such features in early Vietnamese texts.

Alves (2020c, p. 75) posits the following syntactic structures in pV, in which clauses are verb-medial and noun phrases are head-initial. There is no language-internal evidence to hypothesize an earlier verb-initial stage like that of Jenny’s claim for pAA. In modern Vietnamese, clauses are similarly verb-medial.

The Early Vietic Clause *[topic-verb-complement]

Modern Vietnamese [topic-verb-complement] nó   bắt   cá 3s    catch fish ‘He/She/It… catches fish.’

Similarly, Modern Vietnamese noun phrase structure has a head-medial structure with elements both before and after head nouns. In Vietnamese, quantification (i.e., phrases consisting of quantity words followed by measure words or classifiers) precedes head nouns, while all other elements (i.e., modifiers, determiners, and possessives) follow them. In a reconstruction of the pV noun phrase, Alves (2020b, p. 67) posits that it matched the modern Khmer structure of head-initial, with all modifiers following a head noun. Additional support comes from Vietnamese textual evidence from the early 2nd millennium CE in which there was substantial flexibility of the position of quantification. Quantification commonly occurred varyingly before and after head nouns. The following early text sample from the mid-2nd millennium shows a quantification phrase (e.g., ‘one unit’) following the head noun (e.g., ‘letter’).

情	書	蔑	幅	封	群	謹
tình	thư	một	bức	phong	còn	kín
love	letter	One	cls	seal	still	tight
‘A love letter still sealed tight.’ (From the poem “Ba Tiêu” in “Quốc Âm Thi Tập”)

This was hypothetically a transitional stage leading from the earlier post-nominal position to the modern stage with only pre-noun position quantification. While the adoption of Chinese classifiers and measure words may have influenced the position of quantification in the Vietnamese noun phrase, the position of other modifying elements, including adjectival elements, possessive nouns, and determiners, was unaffected.

The Early Vietic Noun Phrase *[noun–modifier–quantity]

Modern Vietnamese [quantity–noun–modifier] một   con   cá   lớn  ấy one    cls    fish big distal ‘That one large fish.’

If the pAA verb-initial hypothesis holds, at best, we can say that Vietic participated in the early regional change to topic-comment clauses and that Vietnamese shares with pAA a post-verbal object. As for noun phrases, the post-nominal position of modifiers is a retention from pAA. A summary of clause and noun-phrase structures at different stages is provided in

Table 9. Clause and noun-phrase structures in pAA, pV, and Vietnamese.

Structural Aspect	Stage	Pattern
Clauses	pAA	verb-initial
	pV	verb-medial
	Viet.	verb-medial
Noun phrases	pAA	head-initial
	pV	head-initial
	Viet.	head-medial

.

5.2. Word Formation Strategies

As for morphology, pAA has been reconstructed with derivational prefixes and infixes (Sidwell 2008), and lexical reduplication involving alliteration, riming, and ablaut is widespread enough to consider a word formation strategy in pAA. A more challenging matter is to determine how much phonesthemes played a part in pAA word formation, but some evidence allows for that possibility. The Austroasiatic-style reduplication is robustly attested in Vietnamese, and some evidence of pAA phonesthemes is evident. As for Austroasiatic derivational morphology, there is some evidence of this in the remaining monosyllabic roots, though most of them have been completely lost with no means of recovering them.

5.2.1. Reconstructing Lost Affixation

The complete loss of presyllables in Vietnamese has made recovery of past prefixes and infixes challenging at best and impossible in many instances. Reconstructing the presence of presyllables alone is difficult, let alone their precise phonological material. However, one type of retention for which there is some persuasive evidence is pAA nominalizing nasal infixes, a widespread word formation strategy among Austroasiatic branches.

Table 10. Nominalizing infixes in Austroasiatic languages.

Language (Branch)	Source Verbs	Derived Nouns
May (Vietic)	pi3 ‘to carry on one’s back’	pani3 ‘basket’
Maleng Bro (Vietic)	tęʔ ‘to urinate’	trnęʔ ‘urine’
Buxing (Mangic)	tain ‘to weave’	tɣm.lain ‘a braid’
Car (Nicobaric)	kahúl ‘to cook’	kamhṹl ‘a cook’
Mlabri (Khumic)	tɛk ‘to hit’	trnɛk ‘hammer’
Santali (Munda)	ʤɔk ‘to sweep’	ʤɔnɔk ‘broom’
Stieng (Bahnaric)	pu:s ‘to sweep’	pənus ‘broom’

shows examples of source verbs and derived nouns with vowel-plus-nasal sequences in several branches of Austroasiatic. This type of derivational infix is even seen in Munda and Nicobaric, demonstrating that these date to the pAA stage.

While derivational morphology is mostly fossilized even in conservative Vietic languages, some examples can be provided, as in the first two instances in Table 10. The possibility then exists for Vietnamese in a previous stage to have lost the presyllables of these derived nouns, leaving main syllables with nasal onsets.

Ferlus (1977, pp. 53–54) provides such instances in Vietnamese of pairs of verbs and nouns that share rhymes (i.e., vowel, coda, and tone), and the actions apply to the related nouns, as in

Table 11. Possible instances of retained nominalizing infixes in Vietnamese.

Source Verbs	Derived Nouns
ɗaːn1 (đan) ‘weave’	naːn1 (nan) ‘bamboo slat’
ɗap5 (đắp) ‘cover’	nap5 (nắp) ‘lid, stopper’
ɗəː6 (đợ) ‘pawn; pledge’	nəː6 (nợ) ‘debt’
ɗuːt5 (đút) ‘insert’	nuːt5 (nút) ‘stopper, plug (n)’
ceːm1 (chêm) ‘wedge’	neːm1 (nêm) ‘wedge (n)’
ceːp5 seːp5 (chếp xếp) ‘fold’	neːp5 (nếp) ‘wrinkle (in nếp nhăn)’
cɔːk6 (chọc) ‘thrust, poke’	nɔːk6 (nọc) ‘stake, picket’
ɽiːt6 (rịt) ‘tie up’	niːt6 (nịt) ‘belt’
taːj3 (tải) ‘bag (n)’	naːj3 (nải) ‘sack, bag’
ciə4 (chĩa) ‘pitchfork; fish with a fish lance’	niə4 (nĩa) ‘fork’

. While Ferlus notes that a couple of these pairs have cognates in Khmu in the Khmuic branch of Austroasiatic, he provides no instances in other Vietic languages.10

While nominalizing infixes and causative prefixes are common among Austroasiatic languages, affixes with other functions have developed among Austroasiatic languages. We can thus speculate that other affixes may have existed as well in earlier stages of Vietic. As noted, onset clusters *kl in pAA and pV have become [ʈ] ‘tr’ in modern Vietnamese, and in some cases, such phonological material may have come from presyllables.

Table 12. Vietnamese source verbs and derived verbs that may have had prefixes.

Source Verbs	Derived Verbs	17th Cent.
leːn1 (lên) ‘ascend’	ʈeːn1 (trên) ‘atop’	tlên
lɛːw1 (leo) ‘climb’	ʈɛːw1 (treo) ‘hang, suspend’ (cf. ʈɛːw2 (trèo) ‘climb’)	tleo
loːj1 (lôi) ‘pull, drag’	ʈoːj1 (trôi) ‘drift, float’	tlôi
lɨət5 (lướt) ‘glide’	ʈɨət6 (trượt) ‘slip, skid’	tlợt
NA (cf. pAA *lun, *lɔːn ‘round, roll (v)’)	ʈɔːn2 (tròn) ‘round’	tlòn
ɽaːj3 (rải) ‘spread, scatter’	ʈaːj3 (trải) ‘spread, lay’	klảj

contains instances of potential pairs of verbs and either stative or active verbs in which the posited source words have [l] onsets while the derived verbs have [ʈ] onsets. Also in Table 12 are the words spelled in early romanized script from de Rhodes’ 1651 dictionary, in which these words had [tl] cluster onsets instead of the possible [ʈ], which also existed at that time. These must be regarded as tentative until additional comparative evidence can confirm or refute them (e.g., similar evidence in other Vietic languages, early Vietnamese textual data showing evidence of presyllables, etc.), but they at least show early onset clusters and thus possible previous presyllables that represent earlier stages of Vietic with a more Austroasiatic-like template.

5.2.2. Austroasiatic-Style Reduplication

Another Austroasiatic-like feature of Vietnamese is reduplication. Austroasiatic languages exhibit a type of reduplication in which templates of monosyllabic or disyllabic words are copied, but typically one or more segment alternates from the root. This results in patterns of alliteration, rhyming, or ablaut, as exemplified in

Table 13. Examples of reduplication with alternating segments in Austroasiatic languages.

Types	Examples
Alliteration	Vietnamese: lɔːŋ1-lajŋ1 (long lanh) ‘sparkling’ Bru (Katuic): tamaj-tami:ŋ ‘new’ Sre (Bahnaric): crih-cra:j ‘strange, wonderful’
Ablauting	Vietnamese: muːm4-miːm4 (mũm mĩm) ‘chubby, plump’ Cuoi (Vietic): hɔŋ⁴⁴ˀ haŋ⁴⁴ˀ ‘empty’ Pacoh (Katuic): pu:c-pa:c ‘to flutter’
Rhyming	Vietnamese: boːj5-roːj5 (bối rối) ‘embarrassed; puzzled’ Mundari (Munda): cali-bali ‘dressing up nicely’ Phong (Khmuic): kup-plup ‘butterfly’

. While precise reduplicative forms cannot be reconstructed to pAA or even pV, they are widespread enough among Austroasiatic languages (e.g., note the instance in Mundari in India, as shown in Table 13) to regard as a word formation strategy at the proto-language stage.

While Vietnamese has instances of full reduplication (e.g., thɯəŋ²-thɯəŋ² (thường) ‘sometimes’), a type of reduplication involving alternating segments is much more common, with thousands of words formed in this way, as in the Từ Điển Từ Láy Tiếng Việt [Dictionary of Vietnamese Reduplicative Words] (Viện Ngôn Ngữ Học 1995). This type of reduplication is distinct from the prominent types of reduplication in Chinese languages (e.g., full reduplication with pragmatic functions, partial reduplication such as the A-not-A questions, an ABB pattern of copied second syllables, etc.).11 As expected, it is also a feature of conservative Vietic languages, such as the instance of Cuoi in Table 13.

5.2.3. Phonestheme Onsets

While Vietnamese lacks presyllables—and thus lacks prefixes or infixes—it does exhibit some phonesthemic word formation patterns with ideophonic features related to oral actions or senses. In some instances, these patterns are shared by other Austroasiatic languages, including some that can be seen in pAA etyma. One can cast an excessively wide net and find chance similarities or even more common typological patterns connected to sound symbolism. Nevertheless, the ideophonic sound symbolism was undoubtedly a factor in the original creation of the words and patterns, and their widespread distribution suggests a deep history in the language family, especially when neighboring language families (e.g., Kra–Dai, Sinitic and Tibeto-Burman, etc.) show different patterns that cannot be considered part of a real phenomenon. The proposed phonestheme onsets in Vietnamese and other Austroasiatic languages include (a) nasal onsets for words related to oral activities (snoring, choking, sucking) and taste (sweetness), (b) /h/ onsets in words referring to inhaling, singing, sniffing, or singing, and (c) /l/ onsets for words expressing licking or sticking out a tongue.

Nasal Onsets

pAA etyma in this category include *ɲɔːkˀ ‘suck’ (AA483), *ɲɔːtˀ ‘drink, suck (v)’ (AAP), and *ŋaːm ‘sweet’ (AA113), but Vietnamese does not have cognates of these. Instead, several pV etyma in Vietnamese with senses related to oral and nasal activities similarly have nasal onsets, mostly engma, as shown in

Table 14. Vietnamese words with nasal onsets expressing oral and nasal activities.

Vietnamese	Gloss	PV
ɲa:j1 (nhai)	chew	*tə.ɲaːj
ŋəm6 (ngậm ‘hold in mouth’)	suck	*ŋəmʔ
ŋɛːn6 (nghɛːn)	choke (while eating)	*ŋɛːnʔ
ŋɔːt6 (ngọt)	sweet	*tə.ŋɔːc
ŋaj5 (ngáy)	snore	*tə.ŋarʔ
ŋɨːj4 (ngửi)	sniff, smell (v)	*tə.ŋəs

. It is not clear what the purpose is of the presyllables on these items.

/h/ Onsets

Another set of words has/h/onsets and semantics of verbs to denote inhaling or breathing. The number of words with this feature suggests that this onomatopoetic feature became part of a modestly productive word formation pattern. pAA has only *huːtˀ ‘drink, sip (v)’ (AA233), but pV and Vietnamese have additional relevant instances, as in

Table 15. Vietnamese words with /h/ onsets expressing oral activities.

Gloss	AA	pV	Viet.
suck/inhale	*huːtˀ	*hu:c	huːt5 (hút)
inhale, sniff, snuffle	*hɨːtˀ (AA402a)	*hiːt (tent.)	hiːt5 (hít)
open (mouth) (v)	*haʔ (AAP)	*haːʔ	haː3 (hả)
kiss/sniff (v)	NA (Proto-Katuic *huuɲ, *huoɲ smell, sniff’)	*huːɲ	hoːn1 (hôn)
sing	NA	*haːt	haːt5 (hát)
cough (v)	NA	*hɔː	hɔː1 (ho)
suck in	NA	NA	huːp5 (húp)

.

/l/ Onsets

Words with /l/ onsets that are related to the mouth can be readily found in languages outside of Austroasiatic, for example, Proto-Indo-European *leigh ‘lick’, Proto-Sino-Tibetan *liek ‘lick, tongue’, and Proto-Tai (Pittayaporn 2009) *C̬.lwiəA ‘to lick’ and *li:nC ‘tongue’. Thus, the ideophonic status of such words seems likely. Regardless, in pAA, similar categories of words with /l/ onsets have been reconstructed, such as *lən.’taːkˀ ‘tongue’ (AA054) and *liəmˀ, *liːmˀ ‘lick (v)’ (AA1h19) (and also *liətˀ and *liəpˀ ‘lick (v)’ (AAP)). Of these, only *liəmˀ, *liːmˀ has a reflex in Vietnamese, but Vietnamese has other such words with an /l/ onset, which are pV etyma, and in other Austroasiatic languages, there are also instances of words (both cognates and non-cognates) with /l/ onsets having senses, as in

Table 16. Vietnamese and cognate or non-cognate Austroasiatic words with /l/ onsets and senses related to the mouth.

Viet	pV	Other Austroasiatic languages
liəm5 (liếm) ‘lick (v)’	*-lɛːmʔ	Proto-Katuic *laas, *lɨas ’tongue, lick’ Proto-Khmuic *-lɛːl ’to lick, taste’
lɯəj4 (lưỡi) ‘tongue’	*laːs	Bolyu (Pakanic) lɔ31 ljiːm⁵3 ’tongue’ Bolyu (Pakanic) ljim31 ’to lick’ Munda (not reconstructed but seen in a widespread form) laːŋ ‘tongue’
lɛː1 (le) ‘stick out tongue’	*t.lɛːl	Proto-Bahnaric *liər ~ *lɛːr ‘to stick out tongue’ Kensiu (Aslian) lʌhleh ’to wiggle/stick out tongue’

.

Overall, while words in this section can be considered part of an ideophonic feature in other languages, the number of Vietnamese words with cognates in Austroasiatic proto-language and modern Austroasiatic languages altogether suggest an ideophone phonestheme pattern of word formation in Austroasiatic that has been retained in Vietnamese.

6. Implications in Vietnamese Linguistic Ethnohistory

Despite the impact of Chinese on Vietnamese, especially in its lexicon, but also its phonology (e.g., syllable structure and tones) and noun phrase structure (i.e., movement of quantification to before head nouns), there are many retentions of Austroasiatic features at all linguistic levels. In addition, Vietnamese maintained its Austroasiatic typology for many centuries after the start of Sinitic–Vietic language contact; the linguistic restructuring was not an instance of rapid transformation. While tones in Vietnamese may have emerged in the later part of the 1st millennium, disyllabic words were retained into the early centuries of the 2nd millennium (Shimizu 2015), and onset clusters lasted into the early 1800s (Vu 2019). Also, while hundreds of early Chinese loanwords were borrowed in the 1st millennium, the large-scale borrowing of Chinese words did not take place until the beginning of the 2nd millennium. This was the time of the hypothesized shift of a local Chinese community to Archaic Vietnamese and the contemporaneous spread of Chinese rime dictionaries and a tradition of Chinese-character reading.

Altogether, despite the Chinese administration in northern Vietnam for a millennium and significant impact on the Vietnamese lexicon and some elements of its phonology, these changes did not result in a language that was unrecognizable as an Austroasiatic language. Thus, we can speculate that the Vietic community in northern Vietnam in that period held considerable sociocultural status. There was even retention of a native numeral system (in contrast with replacement of the numeral systems of Tai languages with the Chinese one), despite an influential resident Chinese community in which early Vietic speakers intermarried, while borrowing many other categories of words. While details of the sociocultural situation in northern Vietnam in the 1st millennium CE are not available, the sociolinguistic status of Viet–Muong and, later, Vietnamese, was apparently significant enough to stave off many of the potential effects of intense language contact.