PACIFIC LINGUISTICS

School of culture, History and Language

College of Asia and the Pacific
THE AUSTRALIAN NATIONAL UNIVERSITY
__________________________________________________________

Out of print books in pdf format

This PDF document may be copied, printed and stored for use
in libraries and for personal use. It may not be reproduced for
sale or distribution.

Pacific Linguistics Home Page: http://www.pacling.com

613 Hmong-Mien language history

Martha Ratliff

This book presents a new reconstruction of Proto

Hmong-Mien, the ancestor language of the modern
Hmong-Mien (Miao-Yao) languages of southern
China and northern Southeast Asia. It also contains
discussion of selected topics in the history of Hmong-
Mien: phonological change, tonogenesis and tone
development, ancient morphology, numerals and
pronouns, language contact, and the ancient Hmong-
Mien world.

2010 ISBN 9789858836150 308 pp

Prices: Australia AUD $115.50
(incl. GST, hardback)
International AUD $105.00
To the memory of

Paul K. Benedict (1912–1997)

and

Wang Fushi 王輔世 (1919–2001)

 Table of contents

Acknowledgements................................................................................................................xi

1 Introduction .....................................................................................................................1

1.1 Scope and purpose ........................................................................................................1

1.2 The Hmong-Mien language family .............................................................................1

1.3 Scholarship on Hmong-Mien language history .........................................................4

1.4 Reconstruction methodology .......................................................................................6

1.5 Organization of the book .............................................................................................7

1.6 Future research..............................................................................................................8

2 Proto Hmong-Mien........................................................................................................10

2.1 Introduction to the reconstruction............................................................................10

2.1.1 Phonological characteristics of present-day languages .............................................10
2.1.2 Issues in the reconstruction of onsets ........................................................................12
2.1.2.1 Pre-initials ........................................................................................................................... 12
2.1.2.2 Initials 14
2.1.2.3 Medials 18
2.1.3 Issues in the reconstruction of rimes .........................................................................19
2.1.3.1 Rime mergers ...................................................................................................................... 19
2.1.3.2 Vowel length in Mienic....................................................................................................... 26
2.1.3.3 Nasal codas in Hmongic...................................................................................................... 28
2.1.3.4 Vowels or glides? ................................................................................................................ 29
2.1.3.5 Concluding remarks ............................................................................................................ 29
2.1.4 Issues in the reconstruction of tones..........................................................................30
2.2 Proto Hmong-Mien onsets............................................................................................31
2.3 Proto Hmong-Mien rimes...........................................................................................108
3 Tonogenesis and tone development............................................................................183

3.1 Introduction ..............................................................................................................183

3.2 Relative timing of Chinese and Hmong-Mien tonogenesis ...................................185

3.2.1 Tone assignment in different contact situations ......................................................186
3.2.2 Sinospheric languages developed tones together ....................................................191
vii
viii

3.3 Phonation-to-tone development .............................................................................. 192

3.3.1 Distribution of low-pitch tones in modern languages ............................................. 193
3.3.2 Residual phonation contrasts in modern languages ................................................ 194
3.3.3 Breathy voice and vowel quality splits ................................................................... 195
3.3.4 Complex contour tones as phonation ‘gear switches’............................................. 195
3.3.5 Summary ................................................................................................................. 196
4 Morphology.................................................................................................................. 198

4.1 Introduction .............................................................................................................. 198

4.2 Nominal prefixes....................................................................................................... 199

4.2.1 Hmong-Mien nominal prefixes today ..................................................................... 199
4.2.2 Cross-linguistic prefix variation.............................................................................. 200
4.2.3 Prefix pre-emption................................................................................................... 202
4.2.4 Nominal prefixes and Proto Hmong-Mien disyllabism .......................................... 204
4.3 Prenasalization ......................................................................................................... 206

4.4 Traces of ancient affixes .......................................................................................... 208

4.4.1 Odd/even tonal pairs within the same tone category............................................... 209
4.4.2 Derivation by tone change....................................................................................... 211
5 Numerals, personal pronouns, and demonstratives................................................. 213

5.1 Introduction .............................................................................................................. 213

5.2 Numerals ................................................................................................................... 213

5.3 Personal pronouns.................................................................................................... 217

5.4 Demonstratives ......................................................................................................... 220

6 Language contact......................................................................................................... 223

6.1 Introduction .............................................................................................................. 223

6.2 Chinese ...................................................................................................................... 224

6.2.1 Phonology................................................................................................................ 224
6.2.2 Lexicon 226
6.2.3 Syntax 226
6.2.4 The classifier construction ...................................................................................... 227
6.3 Lexical connections to other language families ..................................................... 233
6.3.1 Tibeto-Burman ........................................................................................................ 233
6.3.2 Tai-Kadai................................................................................................................. 234
6.3.3 Mon-Khmer............................................................................................................. 235
6.3.4 Austronesian............................................................................................................ 237
 ix

7 The ancient Hmong-Mien world ................................................................................238

7.1 Introduction ..............................................................................................................238

7.2 Homeland ..................................................................................................................239

7.3 Hunting ......................................................................................................................241

7.4 Agriculture ................................................................................................................241

7.5 Reckoning and measuring .......................................................................................242

7.6 Material culture ........................................................................................................242

7.7 Non-material culture ................................................................................................243

Index of proto forms by English.....................................................................................245

Index of proto forms by White Hmong .........................................................................273

References .......................................................................................................................287
 Acknowledgments

Paul Benedict and I were originally planning to write this book together. Unfortunately, we
never got past early discussions about what we wanted the book to be. He was an
inspiration to me—not only because he could think (and almost speak) in protolanguages,
but also because he took such joy in his work.
I would like to thank my colleague Liu Haiyong for bringing me the best gift I have
ever received: a copy of Wang Fushi and Mao Zongwu’s 1995 reconstruction of Proto
Miao-Yao [Proto Hmong-Mien]. This book has been my constant companion over the past
fourteen years: the pages are now falling out, and are covered with notes that have been
written, erased, and re-written countless times. As a result, I have felt a special kinship
with the late Wang Fushi, whom I had the pleasure to meet only once, at a conference on
the minority languages of China held in Santa Barbara, California in 1986.
The Old Chinese reconstructions in this book are from work in progress by the eminent
sinologists William H. Baxter and Laurent Sagart. Over the years they have both been
extremely generous with their time and advice as I have struggled to interpret the Chinese
element within Hmong-Mien. I am especially grateful that they were willing to let me use
their provisional reconstructions before they had been published. I look forward to our
continued collaboration on issues of language contact and language prehistory in China
and Southeast Asia.
I would like to thank my editor Paul Sidwell for his patience in formatting my tables
and putting up with all of my corrections, and my daughter Judy Kaplan for working
through the proofs. Finally, I am indebted to my fellow linguists and good friends Ljiljana
Progovac and Paul Newman for their helpful advice and unfailing support through every
stage of this process.

Martha Ratliff
Detroit, Michigan U.S.A.
April 2010

xi
1 Introduction

1.1 Scope and purpose

The purpose of this book is to present a new reconstruction of Proto Hmong-Mien, the
ancestral language of modern-day speakers of the Hmongic (Miao) and Mienic (Yao)
languages of southern China and Southeast Asia, and a set of discussions on topics relevant
to the historical development of these languages. Some of the discussions argue for
particular aspects of historical development in some detail (for example, the development
of tone, the origin of the classifier construction, the innovation of vowel length), while
others merely sketch out a field of inquiry for future work. This book brings together in
one place ideas the author first presented in various publications and conference
presentations, as well as a number of new analyses.

1.2 The Hmong-Mien language family

Languages of the Hmong-Mien family are spoken in southern China, and northern
Vietnam, Laos, and Thailand. Most Hmong-Mien speakers belong to the Miao and Yao
“nationalities”—minority ethnic groups—in China, although not all Miao or Yao people
speak a Hmong-Mien language; many now speak local varieties of Chinese instead
(Wurm et al. 1988). Hmong-Mien speakers in China primarily inhabit Guizhou, Hunan,
and Yunnan provinces and the Guangxi Zhuang Autonomous Region; smaller numbers
live in Sichuan, Guangdong, Hubei, and Jiangxi provinces, and on the island of Hainan.
Under pressure from the expanding Han Chinese population, waves of Hmong and
Mien speakers migrated to Vietnam, Laos, and Thailand during the 19th and early 20th
centuries (Jenks 1994, Culas and Michaud 2004). Another wave of migration followed the
end of the Vietnam War in the 1970s, when tens of thousands of Southeast Asian Hmong
and Mien emigrated to the United States, Canada, France, French Guiana, and Australia.1
Traditionally, Chinese scholars have claimed that Hmong-Mien belongs to the Sino-
Tibetan language family, along with Chinese, Tibeto-Burman (which includes Tibetan,
Burmese, Karen, and many smaller languages of southern and western Asia), and Tai-
Kadai (which includes Thai, Lao, Shan, Zhuang, and many smaller languages of Southeast
Asia). Most recent Chinese linguistic scholarship supports this view as well (Wang 1986,
Pan 2006). Although a genetic relationship between Chinese and Tibeto-Burman (together,
Sino-Tibetan) is generally accepted, the belief that the Sino-Tibetan family also includes

1
See valuable information on late 20th century immigration of Southeast Asians to the United States on
the website of the Southeast Asia Resource Action Center (http://www.searac.org/), and information
specifically on Hmong immigration on the website of the Hmong Studies Internet Resource Center
(http://www.hmongstudies.com/).
1
2 Chapter 1

Hmong-Mien and Tai-Kadai is not widely shared by linguists outside China. Despite
massive numbers of Chinese loanwords in Hmong-Mien languages, differences in basic
vocabulary raise serious doubt that Chinese and Hmong-Mien are related. Furthermore,
most linguists reject typological similarities as evidence of genetic relationship; similarities
in grammar, word structure, and phonological systems between Chinese and Hmong-Mien
languages can be explained by the power of Chinese influence in the area and widespread
bilingualism.

Figure 1: Distribution of Hmong-Mien languages

Other family connections have been proposed. Forrest (1973 [1948]:93-103), Downer
(1963), Haudricourt (1966), and Peiros (1998:155-160) favor the possibility of a family
relationship with Mon-Khmer, while Benedict linked Hmong-Mien to Austronesian and
Tai-Kadai as part of “Austro-Tai” (Benedict 1942, 1975). Neither of these proposals has
gained general acceptance among scholars. Until a careful separation of layers of Chinese
 Introduction 3

borrowings from native Hmong-Mien vocabulary has been completed and the remaining
core has been systematically compared to these other families, the question of wider
relationship cannot be resolved. The most prudent position to take in the meantime is that
Hmong-Mien constitutes an independent family of languages (see chapter 6).
Within the family, two main sub-families have been identified: the Hmongic sub-family
(c. 5 million speakers) and the Mienic sub-family (c. 1.3 million speakers). The Hmongic
sub-family is an internally diverse group that includes the following mutually unintelligible
languages (some with mutually unintelligible dialects): Hmong (1,800,000 speakers,
primarily in Guizhou province, Yunnan province, and Southeast Asia), Hmu (1,100,000
speakers, primarily in Guizhou province), Qo Xiong (770,000 speakers, primarily in
Hunan), Bunu (258,000 speakers in the Guangxi Zhuang Autonomous Region), A-Hmao
(200,000 speakers in northwestern Guizhou and Yunnan), and Ho Ne, known in China as
the “She” language (900 speakers in Guangdong), among others.2 The Mienic sub-family
is smaller and less internally diverse, but is represented over the entire geographical area,
from Yunnan province in the southwest to Guangdong province in the southeast, and into
northern Southeast Asia. It includes the languages Iu Mien, Mun, Biao Min, and Zao Min.
A provisional family tree structure is presented below:

Although the highest-level two-branch structure is not in doubt given the obvious
lexical, phonological, and grammatical differences between languages belonging to the
two sub-families (Ratliff 1998), more work needs to be done to refine the internal structure
of each sub-family (see section 1.4 below).
The designation “Miao-Yao”, both the name of these ethnic groups and an alternate
name for the language family, is of Chinese origin. It represents the concept of
“nationality”, which is not a purely linguistic classification but also takes into account
cultural practices, politics, and self-identification (Sun 1992). For example, speakers of
Mien together with speakers of Bunu, a Hmongic language, and Lakkia, a Tai-Kadai
language, are classified in China as members of the Yao nationality (Mao, Meng, and
Zheng 1982). Conversely, speakers of the Mun language on Hainan Island are classified as

2
Numbers of speakers taken from Gordon 2005, which in turn is based largely on Wurm et al. 1988. An
overview of the individual languages of this family, the locations where they are spoken, and their
numbers of speakers can be found in Niederer 1998. When speakers of Hmong-Mien languages in
Southeast Asia and in the west are included, the size of the family will undoubtedly prove to be larger
than 6.5 million.
4 Chapter 1

members of the Miao nationality despite the fact that their language is Mienic (Shintani
and Yang 1990). To avoid confusion of ethnic and linguistic categories, many Western
linguists have adopted the name “Hmong-Mien” to refer to this language family, and this
practice is followed here. Western anthropologists, however, continue to use the name
“Miao-Yao” in reference to the language family insofar as it is the generally accepted
name for these communities of speakers in China, and does not arbitrarily elevate the
names of two representative languages to the name of the group as a whole. However,
naming a language family after two languages from different branches with either wide
geographical representation or a large number speakers is one common way language
families in the area have been named—compare “Tibeto-Burman” (a language family that
contains many more languages than Tibetan and Burmese) and “Mon-Khmer” (a language
family that contains many more languages than Mon and Khmer). The name in no way
suggests that the Hmong and Mien languages are more significant than the other languages
of the family. Furthermore, the use of the designation “Hmong-Mien” in reference to
languages does not preclude the use of the traditional designation “Miao-Yao” in reference
to peoples.

1.3 Scholarship on Hmong-Mien language history

The present work depends in innumerable ways on the work of other scholars who have
published studies on the history of the Hmong-Mien languages. The following is a brief
overview of those individuals who have done full and partial reconstructions and have
written influential large-scale historical studies on the family.
The earliest serious reconstruction work was undertaken by Kun Chang (Chang 1947,
1953, 1966, 1972)—a series of articles on the history of Hmong-Mien tones that remains
the definitive body of work on Hmong-Mien tonology today—and by André G.
Haudricourt (1954a), a comparative study of 131 etyma in 8 Hmongic languages, Mien,
and Mun. Chang also published a substantial article on Proto Hmongic initials (Chang
1976).
The first reconstruction of Proto Hmong-Mien was the Cornell University PhD
dissertation of Herbert C. Purnell (1970). It was a full reconstruction of the protolanguage
and the intermediate-level Hmongic and Mienic protolanguages based on an analysis of
1069 sets of cognates. It also included useful synchronic descriptions of the 20 languages
upon which it was based. Although much more material is available on individual
languages now making this reconstruction less useful today than it once was, for years it
was the standard to which others referred.
In 1979, at the 12th International Conference on Sino-Tibetan Languages and
Linguistics in Paris, Fushi Wang distributed his reconstruction of the Proto Miao (Proto
Hmongic) language family (Wang 1979). His handout at this conference was actually a full
manuscript of a book that was later, with some changes, to be published in Tokyo (Wang
1994). The following year, Wang and Zongwu Mao published their reconstruction of Proto
Miao-Yao (Proto Hmong-Mien) based on 23 sampling sites—14 Hmongic and 9 Mienic—
chosen to represent the diversity within the family (Wang and Mao 1995). This is
unquestionably the most important contribution to Hmong-Mien historical studies to date:
it pulls together 829 groups of likely cognate forms, and places them within
correspondence sets. Its publication has made further historical work on this family
possible; one cannot overstate its importance. For example, the data for the present
 Introduction 5

reconstruction comes primarily from Wang 1994 and Wang and Mao 1995. However, the
reconstruction does not look like a natural language, nor do Wang and Mao claim that it
should: their reconstructed protolanguage has 260 initials (single consonants and
consonant clusters) and 210 finals. Although each correspondence set is assigned a
phonetic value, many of these are clearly “place-holders”, awaiting further work.
An influential unpublished paper on problems in the reconstruction of the rimes of
Proto Hmong-Mien was delivered by Gordon B. Downer at the 15th International
Conference on Sino-Tibetan Languages and Linguistics in Beijing (Downer 1982).3 This
was by no means a full reconstruction, but in this paper Downer presented his ideas about
how a reconstruction should proceed, and proposed a number of reconstructed forms based
on a comparison of four languages, drawing on data from Wang 1979 and on West
Hmongic data he had collected himself. In lesser hands, a conference paper would not be
worth special mention, but everything Downer wrote about the history of this language
family has proved prescient (see also Downer 1963, 1967, 1971, 1973, 1978, 1979).
In 1993, Theraphan L-Thongkum published a masterful and extremely useful
reconstruction of Proto Mienic (L-Thongkum 1993). This work was based exclusively on
fieldwork she conducted in Thailand and in Guangxi Zhuang Autonomous Region in
China, and is thus valuable both for the analysis and for the data upon which it is based. Of
special significance for language historians is the inclusion of data on “Northern Mien”
spoken in Longwei village, Yangshou County (Guangxi), which, like Biao Min, has
preserved consonant clusters lost elsewhere not only in Mienic, but in the family as a
whole. A separate reconstruction of Proto Mienic is not included in this work, because L-
Thongkum’s reconstruction fits quite neatly into the larger framework of Proto Hmong-
Mien presented here. In fact, one aspect of her reconstruction has been deliberately
borrowed for words attested in Mienic alone: the representation of the Proto Hmong-Mien
series *NT-, *NTH-, and *ND- as Proto Mienic *ʔD- *DH- and *ND- respectively (where
the capital letters “N”, “T”, and “D” stand for stops at all places of articulation). This very
economically captures pre-nasalization and voicing facts that have distinctive voicing and
tone register outcomes in Mienic. For examples, see onset 1.4 in chapter 2, section 2.2, and
the discussion that follows.
Finally, in 1998 Barbara Niederer published an indispensable reference tool for linguists
working on Hmong-Mien language history: a compilation of all published work on
Hmong-Mien languages, including all Chinese scholarship, up to 1994 (Niederer 1998).
The book also provides concise descriptions of 40 Hmong-Mien languages, a useful
comparison of the reconstructions by the scholars mentioned above, and a record of
dialect/language name equivalences that allows one to track the same dialect/language
across publications. Niederer has also performed extensive fieldwork on Hmong-Mien
languages in China and Vietnam, most notably on two dialects of Pa-Hng, which she has
brought into her discussion of historical topics (1997, 1999, 2004).
This is certainly not intended to represent an exhaustive list of those who have
published work on Hmong-Mien language history, however. Scholars including Michael
Johnson, Yunbing Li, David Mortensen, Ilia Peiros, David Solnit, David Strecker, and
Yoshihisa Taguchi have also made important contributions that will be mentioned in the
pages to come.

3
Reconstruction work at a lower level was also presented in Downer 1979.
6 Chapter 1

1.4 Reconstruction methodology

The present reconstruction differs in a number of important ways from earlier
reconstructions. The first three points below reflect foundational ideas about the language
contact situation in southern China and the internal structure of this particular family which
have had important consequences for the shape of this reconstruction.
1) A major working assumption of the present reconstruction is that Hmong-Mien is not
a sub-family of Sino-Tibetan, and that the vast majority of those words which Hmong-
Mien and Chinese share are Chinese loanwords, even at the deepest level. This is based on
the nature of the words shared by Chinese and Hmong-Mien, which for the most part are
cultural rather than basic vocabulary. Identifying and setting aside Chinese loanwords
gives us a better picture of the phonological and semantic nature of the native component
of the Hmong-Mien lexicon (chapter 7), and also yields insights into the nature of the
contact between the two language groups (chapters 3 and 6). No previous reconstruction of
Proto Hmong-Mien, Proto Hmongic, or Proto Mienic has attempted to separate out the
Chinese element from the native element.4
2) The reconstruction of the onsets in this work reflects the assumption that modern
monosyllabic-word Hmong-Mien languages descended from a language which had words
with a more complex structure. Disyllabic or sesquisyllabic word structure is preserved to
some extent in the complex set of Hmong-Mien initial clusters; in some phonologically
unusual, but quite regular, initial correspondences sets; and in the segmental and
suprasegmental traces of old morphology. It is also retained and refreshed in the prefix-
root structure of nouns in specific daughter languages. For discussion and examples, see
chapter 2, section 2.1.2.1 on ancient “pre-initials” and chapter 4 on morphology.
3) The dramatic merger of a large number of Proto Hmong-Mien rimes into a small
number of Proto Hmongic rimes represents a major Hmongic innovation. It has been used
here as a guide in the reconstruction of rimes at both levels. As will be explained in some
detail in the introduction to the reconstruction, chapter 2, section 2.1.3, the fact that over
125 Proto Hmong-Mien rimes collapsed into only 28 rimes in the development of Proto
Hmongic provides us with a powerful tool in reconstruction. In order for a set of rimes to
merge, they must have shared certain phonological features. Therefore, the merger
provides a constraint on the reconstruction of sets of Proto Hmong-Mien rimes; such sets
must be “merge-able”. On the other side of the equation, the reconstruction of the 28 Proto
Hmongic rimes is facilitated by “looking above” to the Proto Hmong-Mien rimes that feed
into each one. The number of Proto Hmongic rimes is small, but the reflexes of each one
are strikingly divergent. The Proto Hmongic rime correspondence 3, for example, here
reconstructed as *æw, is represented by the following reflexes: a, ɯ, e, i, a, æ, i, e, ei, o, a.
Wang (1994) reconstructs *æ for this set, without regard to the Mienic side of the family.
But the Mienic cognates of words in set 3 belong to five sets that must be reconstructed
with a back vowel followed by a rounded element, here *æu, *əu, *au, *ɔu, and *uw. The
merger to *æw is not implausible, and helps account for the rounded or back reflexes in the
most conservative languages in the Hmongic sub-family. No other reconstruction of Proto
Hmong-Mien has taken advantage of the rime merger in this way.
The last two points reflect general assumptions about reconstruction methodology:

4
Wang and Mao (1995:19) admit that some of the etyma used in their reconstruction might be ancient
loanwords from Chinese. But since they also admit the possibility that they might be cognates, they are
incorporated into their data set without special distinction.
 Introduction 7

4) Generalizing from point (3) above, it is the case that higher-level patterns are often
easier to discern than lower-level patterns. In this relatively small family, where the fact of
relationship is not in question, it has made sense to work on the reconstruction at the level
of Hmong-Mien and at the level of each sub-family simultaneously—in advance of
detailed work on subgrouping within Hmongic and Mienic and lower-level
reconstructions. If a traditional “bottom-up” approach were to be taken, one that holds that
each subgroup must be identified and reconstructed before taking the next step back in
time, it would be easy to be fooled by local convergence and contact effects that have quite
a different historical status from the effects of normal language transmission. It would also
be wasteful, in that it would necessitate constant correction of lower-level reconstructions
determined without considering all the relevant evidence.
5) Although an artifact, the reification of a hundred intersecting hypotheses, a
protolanguage should be natural-looking, and it should be possible to imagine how its
descendants could have developed from it through natural processes. Accordingly, efforts
were made (1) not to over-reconstruct (the correspondence sets in Wang and Mao 1995
were combined wherever possible), (2) not to “stuff” proto-forms (Matisoff 1990) with
segments reflecting every feature of every reflex, and (3) not to reconstruct highly
unnatural phoneme inventories (presented in the context of onset and rime inventories in
chapter 2, sections 2.2 and 2.3). Typology also served as a guide in another way: wherever
possible, simple proto-segments and clusters were reconstructed for well-represented
correspondence sets, on the assumption that common segments (defined cross-
linguistically) would have been represented in the lexicon with greater frequency than
unusual segments.

1.5 Organization of the book

The major component of the book, chapter two, presents the reconstruction. It includes:
• an introduction to the reconstruction, which includes a brief overview of the
phonology of present-day Hmong-Mien languages and a discussion of selected issues
in the reconstruction of onsets, rimes and tones;
• the Proto Hmong-Mien onset inventory and the onset correspondence sets with
discussion of individual sets and notes on individual etyma;
• the Proto Hmong-Mien and Proto Hmongic rime inventories and the rime
correspondence sets with discussion of individual sets and notes on individual etyma.
Chapters three through five address selected topics concerning aspects of the
protolanguage and the historical development of Hmong-Mien. The historical development
of tone is discussed in chapter three. Hmong-Mien tonogenesis is discussed in relation to
tonogenesis in Chinese, and it is proposed that that neither group of languages was tonal at
the time the first set of Chinese loanwords were borrowed. This chapter also refines the
account of tonogenesis and split within the Hmong-Mien family by proposing that two
layers of phonation contrasts preceded the development of “tone proper”, or tone perceived
as distinctive pitch differences. Chapter four examines traces of ancient morphology in the
present-day languages. Three cases involve prefixation of increasing age: (1) still-evident
nominal prefixes, (2) prenasalized onsets as possible fused prefixes, (3) and prefixes
reflected indirectly as an initial voicing contrast in a small number of doublets. There are
8 Chapter 1

also a few doublets that suggest that there may have been an ancient “derivation by tone
change” process. Chapter five addresses the reconstruction of three closed classes:
numerals, personal pronouns, and demonstratives.
The sixth chapter presents information on language contact and some brief speculation
on distant relationships between Hmong-Mien and other language families. This subject is
vast, and the summary of observations presented in this chapter is of necessity incomplete
and colored by the author’s focus on the Hmong-Mien family. The intent is merely to add
another voice to the on-going discussion on language contact and genetic relationships in
this linguistically crowded and complex part of the world. The chapter begins with a
consideration of the effects of Chinese contact on Hmong-Mien, a topic which has long
been of interest to sinologists for the potential light Hmong-Mien may shed on the history
of Chinese. This section addresses aspects of Sino-Hmong-Mien phonology, reviews the
extent and nature of lexical borrowings, and provides an argument that the classifier
construction was borrowed from Chinese. The chapter closes with lists of Hmong-Mien
words that bear both a phonological and semantic likeness to words in Tibeto-Burman,
Tai-Kadai, Mon-Khmer, and Austronesian, along with some speculative remarks about
how we might interpret these similarities.
The short seventh chapter pulls together words by semantic field—flora and fauna,
hunting, agriculture, reckoning, material culture, non-material culture—which do not show
a connection to Chinese, and may thus be useful in reconstructing a picture of the
homeland and daily life of Proto Hmong-Mien speakers some 2500 years ago.
The book concludes with a references section and two indices: an English index to the
reconstructed forms and a White Hmong index to the reconstructed forms. The second
index is included in the hope that this book will be of interest to Hmong scholars and
students.

1.6 Future research

Important research projects for the future fall into two areas: further work on the internal
history of the language family, and a full exploration of the external connections of this
family to language groups in East and Southeast Asia. In the first area of research, a new
family tree based on an explicit subgrouping methodology that weighs the relative
importance of phonological, grammatical, and lexical innovations needs to be carefully
worked out. Patterns that emerged in the process of working on this reconstruction (see
especially the discussion of Pa-Hng and Qo Xiong in chapter 2, section 2.1.3) will provide
a new starting point for this work. After the subgrouping issue has been addressed, lower-
level reconstructions can be undertaken that will certainly necessitate adjustments in the
reconstruction proposed here. Finally, an etymological dictionary on the model of the
Sino-Tibetan Etymological Dictionary and Thesaurus (http://stedt.berkeley.edu/) will
provide detail about language, culture, and history that goes far beyond a reconstruction
based on a limited number of etyma.
Determination of how Hmong-Mien is connected to its neighbors through contact—and
how it may be related to its neighbors by descent from a common ancestor—represents the
greatest and most exciting challenge for future research. Work on strata of loanwords from
different Chinese languages on the model of Downer 1973 will be best carried out by those
who are specialists in the history of Chinese. And more evidence of common ground
between Hmong-Mien and each of the other language families of the south—Mon-Khmer,
 Introduction 9

Tai-Kadai, Tibeto-Burman, and Austronesian—needs to be added to what has been

collected so far, and this evidence needs to be subjected to various tests designed to help
discriminate between inheritance and contact (Thomason and Kaufman 1988, Ratliff and
Holst 2005). Although the Hmong-Mien language family is comparatively small in size, its
central geographic position ensures that this work will be crucial to our understanding of
the prehistory of southern China and northern Southeast Asia.
2 Proto Hmong-Mien

2.1 Introduction to the reconstruction

2.1.1 Phonological characteristics of present-day languages

The syllable in Hmong-Mien languages is the typical Southeast Asian type:

T
(C)C {j/w/l} ｛ i̯ /u̯ } (V)V (C)
ONSET RIME

The initial consonant or consonants (the “initial”) plus the medial approximant (the
“medial”) make up the “onset”. The “final” consists of the vocalic nucleus, which may
include a final consonant (the “coda”). The final together with an on-glide, if present, make
up the “rime”. A particularly difficult problem of both synchronic and diachronic
phonology is the assignment of syllable-internal glides to either the medial position (part of
the onset) or to the on-glide position (part of the rime). From a historical perspective, this
problem is confounded by the fact that over time the allegiance of a syllable-internal glide
can change from onset to rime, or from rime to onset. A lexical tone is associated with the
nucleus of the rime in all but a few minor morphemes. The two halves of the syllable have
been reconstructed here, the onset and the rime, rather than individual consonant and
vowel segments, following the custom in work on Asian languages. This usefully
reconstructs phonotactics along with phonemes; it is an easy matter to abstract an
inventory of individual consonants and vowels from the inventories of onsets and rimes.
In present-day languages, initial obstruents usually show a contrast in aspiration rather
than a contrast in voice. In addition, initial stops in Hmongic may be contrastively
prenasalized. Voiced obstruents, which must be reconstructed for the protolanguage, are
preserved in a few Hmongic languages, and are a secondary development from
prenasalized stops in Mienic. Sonorant consonants (nasals and liquids) may show a
voiceless/voiced contrast in languages on both sides of the family. Affricates are common
across the family (tθ, ts, dz, pʐ, tʃ, dʒ, ʨ, ʥ, etc.), as are palatal consonants. Medial
consonants in present-day languages include the approximants -j-, -w-, and -l-. A “back l”
[-ɭ-] exists in many West Hmongic languages as both a simple onset and a medial
consonant (Johnson 2002), and retroflex consonants are also common in Hmongic. Several
Hmongic languages also show a contrast between velar and uvular consonants.
Mienic languages display many more rime contrasts than Hmongic languages. Mun
languages show length contrasts in several vowels; Mien shows a length contrast in /a/.

10
 Proto Hmong-Mien 11

Furthermore, Mienic languages may include up to six final consonants in the coda
position: -m, -n, -ŋ, -p, -t, -ʔ.5 In Hmongic languages, there are fewer possible rimes, no
vowel length contrasts, and a maximum of two final consonants in the coda: -n and -ŋ.
All Hmong-Mien languages are tonal, and the primary function of tone is lexical
discrimination. For the most part, tones are associated with heavy syllables of the
“Mandarin” type rather than light syllables of the “Shanghai” type (Duanmu 1999): thus
the tones stay with the syllables to which they belong and rarely spread to neighboring
syllables (see Ratliff 1992b and 1992c for a discussion of a tone language typology, and
the place of Hmong-Mien languages within the typology). There is a limited paradigmatic
(replacement) type of tone sandhi in some West Hmongic languages, which appears to be a
relic of an older, phonologically-driven tone sandhi process (see Downer 1967 and Ratliff
1992a, chapter 2 for discussion). When it occurs, tone sandhi operates from left to right in
West Hmongic languages (i.e., the trigger tone is on the left, and the changed tone is on the
right), and from right to left in North Hmongic and Mienic languages. Tone inventories are
relatively large: five of the eight Shidongkou Hmu (“Black Miao”) tones are level tones at
different pitches, a world-record number of pitch level distinctions (Kwan 1966), and
Zongdi Hmong has twelve tonal contrasts including both level and contour tones (Wang
and Mao 1995). Tones in some languages are characterized by distinctive phonation types
(breathy voice, creaky voice, or both) as well as pitch contrasts (Andruski and Ratliff
2000).
This sketch of the phonological characteristics of Hmong-Mien languages naturally
leaves out an enormous amount of detail. For a thorough introduction to the segmental
phonology of present-day Hmong-Mien languages, consult Niederer 1998. One may also
see the bulk of this chapter, sections 2.2 and 2.3, to review representative word forms.
Given the greater number of onsets in Hmongic (uvular consonants, prenasalized
consonants, obstruent-liquid clusters) and the greater number of rimes in Mienic (vowel
length, coda consonants), a common description of the relationship between the two sub-
families is that Hmongic better preserves the onset of the Proto Hmong-Mien syllable and
Mienic better preserves the rime (Purnell 1970:180; Downer 1982). This is true in broad
outline, but the details prove to be more complex. For example, Mienic preserves a trace of
a Proto Hmong-Mien pre-initial voicing element (here reconstructed as *N-) which leaves
no trace in Hmongic. Biao Min (Solnit 1996) and the Yangshou dialect of Northern Mien
recorded by L-Thongkum (1993) also show medial -l- after labial and velar initials. The
medial -l- of the *kl- clusters in ‘horn’, ‘insect/worm’, ‘road/way’, etc. (onset 5.31) have
not been preserved in Hmongic; they can only be reconstructed on the basis of the Mienic
evidence. By the same token, it is not always the case that the Proto Hmong-Mien rime can
be equated with the Proto Mienic rime. In a number of cases mergers have taken place in
Mienic so that evidence from the Hmongic side needs to be brought in to determine the
best Hmong-Mien reconstruction (see 2.1.3.1 below).
These and other general issues and challenges in the reconstruction of Hmong-Mien
onsets, rimes, and tones will be presented below, with explanations for the decisions taken
here. Discussion of problems relating to individual correspondences will be addressed in
the reconstruction itself, sections 2.2 (Proto Hmong-Mien onsets) and 2.3 (Proto Hmong-
Mien rimes).

5
Final -k appears on a few loanwords, but the regular development of Hmong-Mien *-k in Mienic is /-ʔ/.
12 Chapter 2

2.1.2 Issues in the reconstruction of onsets

2.1.2.1 Pre-initials
One of the differences between earlier reconstructions and the one presented here is the
assumption of greater disyllabism in the protolanguage. In the four areas taken up below,
matter to the left of the main syllable is acknowledged in this reconstruction.

Tight and loose clusters with *N-

Reminiscent of the notion of fused and loosely attached prefixes as reconstructed for Old
Chinese by Sagart (1999:13ff.), both tight *NC- and loose *N-C clusters are here
reconstructed for the Hmong-Mien protolanguage. This is due in part to the fact that there
are two different correspondence patterns for which an initial nasal element is a reasonable
reconstruction. Hmong-Mien prenasalized obstruents appear as such in Hmongic, but yield
voiced obstruents in Mienic. For example, Hmong-Mien *ntəut ‘navel’ (2.4/13):6

Hmongic Mienic
tu7 ntu7 ntaɯ7 ntə7a nʔtoD ntu7 - dut7 duːt7’ din7 -

Loose *N-C clusters also yield voiced obstruents in Mienic (whether from original
voiceless or voiced obstruents, as reflected by the tonal register), but leave no trace
whatsoever in Hmongic. For example, Hmong-Mien *N-top ‘skin’ (2.1/13):7

Hmongic Mienic
tu3 tə3 taɯ 3
tə 3a
to B
- - dup7 dup7 din7 dip7

The question of whether or not these pre-initial elements may have played a morphological
role in the protolanguage is addressed in chapter 4, section 3.

Voiceless sonorants
Voiceless sonorant consonants are here reconstructed as *hC- rather than as *C̥-. Not only
does this make voiceless nasals and liquids comparable to those for which an initial glottal
element must be reconstructed (*ʔC-), but it also deliberately suggests that whatever
caused the sonorant to de-voice was a pre-initial element, presumably an *S- . External
support for this idea comes from borrowings from Tibeto-Burman languages for which an
*s- has been reconstructed. According to Benedict, “[p]refixed *s- is the work-horse of the
TB/Karen prefixial apparatus, generally directive/causative/intensive with verbal roots and

6
Hmong-Mien, Hmongic, and Mienic reconstructions are followed by numbers that indicate the onset and
rime correspondence sets to which they belong. Hence *ntəut ‘navel’ can be located under both onset
correspondence set 2.4 (*nt-) in 2.2 of this chapter, and under rime correspondence set 13 (13f: *-əut) in
2.3.
7
Taguchi (2005) has discovered that Hmongic Pa-na has /tl- / for ‘skin’, and attributes the voiced initials in
Mienic to this medial -l-. This works for the dental/alveolar series, but there is no evidence for a medial -
l- in the “loose N-C” pattern at other places of articulation.
 Proto Hmong-Mien 13

playing the role of ‘animal prefix’ or ‘body-part prefix’ (< *śɑ ‘animal, flesh’) with
nominal roots but appearing frequently elsewhere.” (Benedict 1987b:44).
For example, the following Tibeto-Burman borrowings (Matisoff 2003) form a tight
semantic set:

moon/month Hmong-Mien *hlaH (2.41/4)

Tibeto-Burman *s/g-la ‘moon/month’

sun/day Hmongic *hnɛŋA (2.8/22); Mienic *hnu̯ ɔiA (2.8/11)

Tibeto-Burman *s-nəy ‘sun’

night Hmong-Mien *hməŋH (1.8/21)

Tibeto-Burman *muːŋ ‘cloudy; dark’
cf. Burmese hmuìŋ ‘very dark’ (Benedict 1972:78)

And three Hmong-Mien verbs seem to reflect the Tibeto-Burman *s- prefix with a
“directive/causative/intensive” function:

to slice Hmong-Mien *hlep (2.41/10)

Tibeto-Burman *s-lep ‘to slice’

to extend tongue8 Yanghao l̥ hei3 , Jiwei l̥ ha3, Bunu ɬe3; Mien ɬɛ5, Biao Min ɬia5
Tibeto-Burman *s-lyak ‘to lick (causative)’

to sniff at Hmong-Mien *hmji̯ əmH (1.23/18)

Tibeto-Burman *s-nam (tr.)/*m-nam (intr.) ‘to smell’

More examples can be found in Ratliff 2001a. Although not all Hmong-Mien voiceless
sonorants are Tibeto-Burman borrowings, of course, this particular development suggests a
possible phonological origin of all sounds of this type, and motivates the choice of the
*hC- notation.

Prefix pre-emption
There are certain nouns that show very strange onset correspondences: in fact, these are not
“correspondences” at all, since no two pattern alike, but reflect the unpredictable
absorption of one of several possible nominal prefixes in one or more of the languages in
the set. For example, Hmong-Mien *-bɔuX (1.3/3) ‘hand/arm’, perfectly regular in both
rime and tone, has Hmongic reflexes with labial, coronal, and velar onsets. The Mienic
reflexes consistently show a labial onset, which allows the reconstruction given above. All
that can be reconstructed of the first syllable is a prefix position, captured by use of the
left-edge hyphen. For more discussion and examples, see chapter 4, section 2.

8
Forms from The Office of Miao-Yao Research Lexicon (1987). See also White Hmong /hl̥ e3/ (hlev), ‘to
extend (tongue)’. Due to the irregular rime correspondence, no reconstruction is attempted.
14 Chapter 2

Other unusual initial correspondences

There are two other types of unusual initial correspondences that point to disyllabic words
in the protolanguage. The first involves nasal onsets. There is a clear *m- correspondence
(1.9), wherein every language has either a reflex with /m-/ or an understandable deviation
from /m-/, and there is a clear *n- correspondence (2.9), wherein every language has a
reflex with /n-/. But there are also two words that show /m-/ across the board except in
Mun, which has an inexplicable /n-/, and two which have /n-/ across the board except in
Pa-Hng, which has an inexplicable /m-/. In these four cases, the “minor” nasal has been
reconstructed as a loosely adjoined nasal pre-initial:

to have Hmong-Mien *n-mɛj (1.9/11)

to go Hmong-Mien *n-mʉŋ(X) (1.9/30)
bird Hmong-Mien *m-nɔk (2.9/29)
rain Hmongic *m-noŋC (2.9/28)

In the case of ‘to have’, a Hmong-Mien root *mɛj having to do with possession appears in
a family of four words, so the pre-initial may have been a stative prefix (see chapter 4,
section 4). The word ‘bird’ is to be equated with Proto Malayo-Polynesian *manuk
‘bird/fowl’ (ACD) and Proto Tai *n-lok ‘bird’ (Li 1977), which clearly shows that the Pa-
Hng form has retained a trace of the first syllable of an original disyllabic word.
In another case, in onset series 2.55 (*ʔr-), 2.56 (*hr-), and 2.57 (*r-), all reflexes point
to a simple onset except the Hmongic language Jiongnai. Jiongnai has irregular reflexes in
these three sets, but frequently shows clusters with a prenasalized velar consonant /ŋkj-/
with an upper-register tone in 2.55, /ŋkh-/ or /ŋkj-/ with an upper-register tone in 2.56, and
/ŋkj-/ with a lower register tone in 2.57. The repeated occurrence of this reflex led Wang
and Mao to the reasonable reconstruction of *ŋkl-/*ŋkhl-/*ŋgl- for this series in Hmong-
Mien, rather than a simple liquid. But the fact that a prenasalized velar only appears in one
language argues that this is another pre-initial, not part of the root, and that Jiongnai retains
a trace of the first syllable of an original disyllabic word. The difficulty here is that we
would not expect to see the same pre-initial in each case unless this were a trace of old
morphology, which is highly unlikely given the heterogeneous nature of the group of
words in these sets (‘liver’, ‘high/tall’, ‘sound/noise’, ‘fern’, etc.).

2.1.2.2 Initials

Prenasalized stops
Prenasalized stops and affricates characterize both the present-day Hmongic languages and
the Hmong-Mien protolanguage. In the native component of the lexicon, they do not seem
to reflect ancient ‘fused’ morphology of any sort since prenasalization appears on words of
all lexical categories (see chapter 4, section 3). These initial nasal elements, therefore, are
here reconstructed as phonological elements.
Prenasalization must be reconstructed for both voiceless or voiced obstruents, however
typologically odd. The area-wide tone split conditioned by voicing in the initial is
conditioned solely by the voicing of the oral stop or affricate to which the nasal is attached:
in other words, since we have syllables of the type /mpo1/ and /mpo2/ in the daughter
 Proto Hmong-Mien 15

languages, we must reconstruct both *mpoA and *mboA, or clusters that have both identical
and mixed values for voicing.

Fricatives
A noteworthy aspect of the reconstructed consonant inventory of Proto Hmong-Mien is the
relative scarcity of fricatives. Only four fricatives can be reconstructed for native words,
here *s-, *sj-, *ɕ-, and *h-, and given the words in these sets, only *h- “feels” native. Half
of the words in the first three sets are either clearly Chinese borrowings or are likely to be
Chinese borrowings, which throws suspicion on the rest of the words in these sets, none of
which belongs to core vocabulary. Some fricatives that exist in the daughter languages
derive from affricates or aspirated stops; others derive from glides via fortition: *w- > /v-/,
*j- > /ʑ/.

Aspirated stops
Aspirated stops must be reconstructed for the protolanguage, but reconstructed etyma with
aspirated stops are not numerous: they seem to have carried a lower functional load than
plain voiceless and voiced stops in the protolanguage, and suggest that aspiration may have
been a secondary development of some sort. There are correspondence sets of aspirated
stops in Wang and Mao 1995 made up entirely of Chinese borrowings: for example, their
sets 2 (‘CLF–sheets’, ‘to push aside’, ‘to split’), 14 (‘to dry in the sun’, ‘to carry in two
hands’, ‘to sweep’), 17 (‘to sprinkle’), 38 (‘to splash’), 134 (‘thousand’, ‘lacquer’), 137
(‘to insert’), 206 (‘bucket’, ‘bellows’, ‘hoop’, ‘charcoal’), 254 (‘to pull out’, ‘to insert’),
and 350 (‘to fry’, ‘bed’, ‘otter’, ‘grain’), etc. Thus any etymon for which an aspirated
initial must be reconstructed is a potential loanword. But a number of Hmong-Mien words
with complex aspirated onsets have no obvious Chinese connection and must be taken as
native (see chapter 7): Hmong-Mien *phlei ‘shell’ (1.32/12), and Hmongic *mphjeD
‘daughter/girl’ (1.20/10), *mphleA ‘finger ring’ (1.35/10), *nthaŋA ‘attic’ (2.5/24),
*nthrɔŋA ‘puttees’ (2.50/29).

Coronal co-occurrence restrictions in clusters

Not every medial is reconstructed for every coronal category. The following is a statement
of these restrictions:

Plain dental/alveolar (“T”) series: no oral stops or fricatives with -l-

Dental/alveolar affricate (“TS”) series: no oral stops or fricatives with -l- or -r-
Palatal (“C”) series: no oral stops or fricatives with -j-, -l- or -r-

No coronal series is reconstructed with medial -l-, but this does not mean that such clusters
do not appear in the daughter languages: see Lyman 1974:35 on Green Hmong /tl-/ as a
variant pronunciation of /kl-/, and Taguchi 2005 on /tl-/ and /dl-/ clusters in Pa-na.
16 Chapter 2

Velars and uvulars

Both velars and uvulars appear in Hmongic languages; only velars appear in Mienic
languages. All reconstructions to date, including this one, have taken the velar/uvular
contrast back to Proto Hmong-Mien and posit a merger of the two in Mienic. However,
there is good evidence that some original velars have “fallen back” and become uvulars on
the Hmongic side of the family, making it appear that there were many more uvular-initial
words in the protolanguage than there actually were.
In support of the idea that velar retraction operated in Hmongic, note first that Chinese
borrowings with velar initials frequently appear with uvular initials in Hmongic languages:

甘 ‘sweet’ OC *[k]ˤam > MC kam > Man. gān

HM *Kam > H *qaŋA (5-6.1/24)
h
槀 ‘dry/withered’ OC *[k ]ˤawʔ > MC khawX > Man. gǎo
HM *KhæwX > H *qhæB (5-6.2/5)
雞 ‘chicken’ OC *kˤe > MC kej > Man. jī
HM *Kəi > H *qeA (5-6.1/10)

If these were loans from Middle Chinese, this may reflect the fact that velar retraction was
an active rule at the time of borrowing. Or if these words were loans from Old Chinese, it
may be that pharyngealized *k- was borrowed as *q- in Proto Hmong-Mien. In either case,
this correspondence argues for the source of some Hmongic uvulars in old velars.
Second, velar retraction has been reported for other languages of the area: Matisoff
(2003:20) observes that “[p]ostvelars are generally secondary developments of the TB
[Tibeto-Burman] *velar series, as in Black Lahu, where they regularly descend from
simple *velars that are not followed by a glide”. Solnit (1996:13-14) suggests that this may
have also happened in Hmong-Mien, with Hmong-Mien velars becoming uvulars except
before medial -r- and back -l-, and “under some other condition, perhaps involving front
vowels/glides”.
Finally, and most persuasively, this “other condition” has been discovered: it is not a
front vowel or glide, but a back rounded vowel or on-glide (u, o, ʉ, u̯ ) that blocked velar
retraction. Note the reconstruction of rimes following *k-, as opposed to those following
*K—either *k- or *q-, it is impossible to tell—in the words below, and their White Hmong
reflexes:

5.1 *k- ( > k- ) 5-6.1 *K- ( > q- )

*ku̯ eiX ‘younger brother’ > kɯ3 *KæŋX ‘frog’ > qa3
*kowD ‘to gnaw’ > kaɯ7 *N-KəjX ‘excrement’ > qua3
*kuwD ‘bamboo hat’ > kau7 *N-Kaŋ ‘back (vs. front)’ > qa1

Back rounded vowels and liquids make a more natural class than front vowels and liquids,
and makes the split in *k- understandable. This “retracted” class of sounds prevented a
velar from retracting to uvular, a dissimilatory constraint: it was important that the onset be
clearly perceived against the following element in the syllable.
Thus some Hmongic words with q- go back to *q-, other Hmongic words with q- go
back to *k- . Hmongic words with k- ( < *k-) represent the special case: *k- did not retract
before liquids and back rounded vowels.
 Proto Hmong-Mien 17

*q- *k-

q- q- k-

Therefore an unequivocal *k- or a *q- is reconstructed here only before liquids and back
rounded vowels. The two dorsals contrast before these segments. But for words with /q/
reflexes in Hmongic, only *K- can be reconstructed before all other vowels: due to the
merger, it is impossible in this environment to distinguish /q/s that come from *k- from
/q/s that come from *q-.
This account can explain an interesting asymmetry in the reconstruction of Wang and
Mao (1995). They reconstruct any correspondence set with q- reflexes in Hmongic by
something in the *q- series, and thus have far more words with typologically marked *q-
(Maddieson 1984b:32) than with the typologically unmarked *k-. The numbers below
indicate the numbers of words in each correspondence set:

*k- 5 *q- 13
*kh- 2 *qh- 5
*g- 1 *G- 7
*nk- 1 *nq- 5
*nkh- 1 *nqh- 1
*ng- 3 *nG- 8

This asymmetry can be easily explained if many of the uvulars were originally velars
before unrounded, non-back vowels. The reconstruction of the rimes for each word in these
sets supports this hypothesis.

Liquids
Wang and Mao (1995) and Solnit (1996) have taken the contrast between West Hmongic
front and back “l” ([l] vs. [ɭ]) as a conservative feature, and have projected the contrast
back to the Hmong-Mien protolanguage.9 But Wang and Mao took words with Proto
Hmongic *r- (as reconstructed in Wang 1994) back to their interesting/problematic *ŋkl-
series (here onsets 2.55, 2.56, 2.57), and thus do not reconstruct *r- at all at the Proto
Hmong-Mien level. They reconstruct Proto Hmong-Mien as a two-liquid language, but the
two liquids are the somewhat unnatural pair *l- and *ɭ- , with no *r- .10
In this reconstruction, only *r- and *l- are posited for Proto Hmong-Mien, in both initial
and medial positions. The “back l-” [ɭ-] of some West Hmongic languages is taken to be a
reflex of *lj- (*ʔlj-/*hlj-/*lj-; 2.40.1–2.42.1). For support, note the Chinese loanwords in
these sets: Hmong-Mien *ljim (2.42.1/18) ‘sickle’ (which has a Zongdi reflex of /ɭæin2/)
from 鎌 (Middle Chinese ljem); Hmongic *ljɛŋA (2.42.1/22) ‘to measure’ (which has a
9
Johnson (2002) takes the contrast back to “Proto-Far-West-Hmongic”, a well-justified low-level
reconstruction.
10
Based on Mienic correspondences rather than West Hmongic [ɭ], Purnell (1970) reconstructed three
liquids for Proto Hmong-Mien (*l1-, *l2-, and *r-). Benedict (1975:164) interpreted Purnell's *l2- as * ɭ- ,
and used the reconstruction of these two laterals in Hmong-Mien to help connect the family to
Austronesian, calling it “a highly distinctive AT [Austro-Tai] feature”.
18 Chapter 2

Zongdi reflex of /ɭoŋ2/) from 量 (Middle Chinese ljang); and Mienic *lji̯ oŋB (2.42.1/26)
‘tael (40 grams)’ from 兩 (Middle Chinese ljangX).11 Those correspondence sets for which
Wang in 1994 reconstructed Proto Hmongic *ʔr-/*r̥ -/*r- and in 1995 reconstructed Proto
Hmong-Mien *ŋkl-/*ŋkhl-/*ŋgl- are here reconstructed with *(-)r- at both levels (*ʔr-/*hr-
/*r-; 2.55–2.57). See also Taguchi 2006, in which convincing arguments for *r- at the
Proto Hmong-Mien level are presented. Support for this analysis comes from the phonetic
character of the reflexes—a velar fricative in East Hmongic, retroflex fricatives in North
and West Hmongic, and an /l-/ in Biao Min.

2.1.2.3 Medials

Liquids
The decision to reconstruct only two liquids, defended in the section on initial liquids
above, also extends to the reconstruction of medial liquids: only two liquids are here
reconstructed in medial position, and they are *-l- and *-r- rather than *-l- and *-ɭ- . The
complex medial *-lj- is also reconstructed. The difficulty in accounting for the array of
different correspondences involving dorsal-liquid clusters (see Solnit 1996 and Niederer
1999)—for which a third liquid would come in handy—is made somewhat easier here by
the decision to reconstruct these complex clusters as well as a *k-/*q- contrast.

Glides
A palatal glide as a medial is reconstructed here, but not a labial glide. Medial -w- in the
daughter languages can be explained as one of three secondary developments. It appears
(1) in correspondence patterns that consistently show the association of the rounded
element with the rime rather than the onset, (2) after a palatalized labial consonant and
before /a/ in certain Mienic languages, and (3) as the labial portion of an “unpacked”
labiovelar initial in a Chinese loanword.
For the first case of medial -w- development, see for example Wang and Mao 1995,
rime set 242 (in this reconstruction, rime set 8b). All of the words in this set show a medial
-w- in Mienic languages: for example, in Jiangdi Mien /mwei2/ ‘2sg (‘you’), /twei3/ ‘tail’,
/ʨwei3/ ‘snail’, /pwei5/ ‘to sleep, /tswei5/ ‘smelly’. Wang and Mao place each of these
words in an onset group reconstructed with a medial -w- (*mw-, *tw-, *pw-, etc.), but
clearly the -w- consistently patterns with the rime (here reconstructed as *-u̯ ei). Proper
attribution of the appearance of a medial -w- to a rounded element in the rime rather than
to the onset also accounts for Jiangdi and Xiangnan Mien words with -w- under Wang and
Mao rime sets 39, 75, 77, 207, etc.
Another source of medial -w- in Luoxiang Mien (8) and occasionally in Lanjin Mun (9)
is the environment Pj __ /a/ , where “P” stands for any labial consonant. The syllable-
internal palatal glide dropped in these Mienic languages, and a -w- developed in its place.
The sequence of events is not clear, but the palatal glide appears to be a key feature of the
conditioning environment: Luoxiang Mien also has words with a labial next to /a/ in which
no medial -w- develops: /m̥ aŋ3/ ‘night’ from Hmong-Mien *hməŋH (1.8/21), /ma8/

11
One Tibeto-Burman loanword also shows evidence of *lj-: Hmongic *ljeD (which has a Zongdi reflex of
/ɭe8/) and Mienic *ljəpD ‘lightning flash’ can both be related to Tibeto-Burman *(s-)lyap ‘flash; lightning’
(Matisoff 2003).
 Proto Hmong-Mien 19

‘thick/dense’ from Mienic *makD (1.9/4) /pa6/ ‘nose’ from Hmong-Mien *mbruiH
(1.51/8), etc. Although this medial -w- is restricted only to some dialects of Mien and Mun,
Wang and Mao reconstruct it back to the Mienic or Hmong-Mien protolanguage in each
case.

Mienic Luoxiang Mien Lanjin Mun

*bhi̯ ɛkD ‘chaff/husk’ (1.5/1) bwa7
*hmjaB ‘grass’ (1.23/4) m̥ wa3 wa3
*pjatD ‘wooden basin’ (1.16/4) pwat7
*mjænA ‘person/Mien’ (1.24/19) mwan2
*hmjænB ‘spirit/ghost’ (1.23/19) m̥ wan3 ŋwaːn3
*bi̯ əŋA ‘flower’ (1.3/24) pwaŋ2
*mjəŋA ‘wife of mother’s brother’ (1.24/29) mwaŋ2 ŋwaŋ2

Finally, a third source of medial -w- is an “unpacked” labiovelar initial of a Chinese

loanword. For example:

Luoxiang Mien Lanjin Mun Biao Min

瓦 tile *ŋʷæX (5’.9/5) ŋwa4 ŋwa4 ŋɔ4
斤 catty *kʷjan (5’.16/20) ʨwan1 san1 ȶwan1
瓜 cucumber *Kʷa (5-6’.1/4) kwa1 kwa1 kwa1

2.1.3 Issues in the reconstruction of rimes

2.1.3.1 Rime mergers12

Major rime mergers in Hmongic

The most dramatic story in the phonological history of the languages of this family
concerns the large-scale mergers of ancient rimes in the development of Hmongic. One of
the main differences between this reconstruction and those of Purnell (1970) and Wang
and Mao (1995) is that rime mergers in Hmongic are used to aid in the reconstruction of
Proto Hmongic and Proto Hmong-Mien vowel qualities.
Rimes in Hmongic languages average 14 contrasts, and rimes in Mienic languages
average 60 contrasts, based on the 40 profiled languages in Niederer (1998:211–212). This
is because Mienic languages retain many more features of Proto Hmong-Mien rimes, most
notably -m, -n, -p, -t and complex combinations of vowel, on-glide, and off-glide.13 Mienic
languages have also innovated length contrasts (see section 2.1.3.2). Wang (1994) only
established 30 rime correspondences for Proto Hmongic, while Wang and Mao (1995)
established 210 rime correspondences for Proto Hmong-Mien. In this reconstruction, the
number of rime correspondences in Proto Hmongic is reduced from Wang’s 30 to 28 (his
sets 12 and 17 are combined, and 26 contains only loanwords). The original numbering of
12
This section is based on Ratliff 2002.
13
Final -n in some Hmongic languages is a development internal to Hmongic: Proto Hmongic only had one
final nasal (see discussion in 2.1.3.3 below). Proto Hmong-Mien final *-k has weakened to -ʔ in Mienic.
20 Chapter 2

Wang’s 30 categories are used to facilitate comparison to this reconstruction; these

numbers organize section 2.3 of this chapter into large Hmongic “merger targets”.
Additionally, the number of rime correspondences in Proto Hmong-Mien is here reduced
from Wang and Mao’s 210 to 127, of which 9 are represented by Chinese loanwords only.
In principle, the methodology used here is straightforward. Since we know that a group
of different Proto Hmong-Mien rimes (reflected fairly directly by the Mienic rime
correspondences) merged into each Hmongic rime, we must reconstruct these Proto
Hmong-Mien rimes with enough common features to show how a merger to the “target”
Proto Hmongic rime would have made sense.
For example, for rime set 3, the reconstruction for Proto Hmongic is *æw. The Mienic
correspondence sets reflecting different Proto Hmong-Mien rimes that merged into Proto
Hmongic 3 are the following (rime correspondence set numbers from Wang and Mao 1995
are given at the end of each line):

H-M Mienic reflexes

*au aːu/au a/au aːu aːu/a aːu aːu/au au ɔu/a u (76, 95)
*ɔu wo wə u u u u au/u ɔu u (77, 244)
*əu au au au au au au au/a ɔu u (96)
*æu ou əu eu a/əu u/ou ɔu au æ/ɔu u (113, 259)
*uw u u o u o u u u u (115)

Knowing that these five sets reflect rimes that merged in Hmongic, each was reconstructed
with a back or low vowel as the first element and a rounded second element, either a *-u or
a *-w, depending on whether the reflexes were diphthongs or monophthongs (this will be
discussed further in section 2.1.3.3 below). Of course, each particular reconstruction is also
constrained by values assigned to other correspondence sets, and the phonetic value of
each reconstruction can only be suggestive. The point is that these reconstructed values
could all have yielded these particular reflexes, and, at the same time, this set of
reconstructed values could have merged to Proto Hmongic *æw quite easily. Wang and
Mao had eight sets rather than five, and for them reconstructed *æːu, *aːu, *æu, *ui, *au,
*Ai, *ə, and *Au—a less homogeneous set of values; a set less likely to merge.
But beyond this, in practice the method involves “triangulation” rather than a
straightforward comparison between a firm reconstruction of a value for a Hmongic
correspondence and that of a number of Mienic correspondences. This is because although
simple in terms of the total number of correspondences, the reflexes of each Hmongic rime
correspondence are strikingly diverse, making the reconstruction of Hmongic rimes more
difficult than one might expect. For example, for rime set 3 the Hmongic reflexes given
below do not suggest an obvious reconstruction:

a ɯ e i a æ æ i i e ei o a a

For this very well-represented set (43 words display this correspondence in the
reconstruction), Wang (1994) reconstructed *æ and Downer (1982), for the subset of West
Hmongic reflexes /e i a æ æ i i e ei/, reconstructed *e. But neither reconstruction can
account for the North Hmongic /ɯ/ or the Pa-Hng /o/. It becomes clear that a rounded off-
glide element must be reconstructed for Proto Hmongic only when one makes reference to
 Proto Hmong-Mien 21

the Proto Hmong-Mien rimes that merged into this one rime (see above) and when one
studies the cross-correspondence tendency for both North Hmongic and Pa-Hng to be
conservative in their preservation of the end of the Proto Hmong-Mien rime (here a
[+back, +high, +round] element). The conservative nature of North Hmongic and Pa-Hng
as reflected in their rimes will be discussed further below.
The position adopted here is that in the matter of Hmong-Mien rimes, top-down and
bottom-up constraints should used in tandem. As Anttila observed, “It is natural that such
triangulation from above and below will often lead to a correct solution that would
otherwise have taken much more labor, if it were possible at all.” (1972:346). Some
historical linguists may prefer a strict “bottom-up” approach to reconstruction that does not
look across (which in the matter of Hmong-Mien rimes, given the conservatism of Mienic,
is in effect “above”) in order to find the best intermediate reconstruction. This
methodology is dangerous in one potential application: to justify new proposals for distant
genetic relationship. When the question of relationship is open, it is too easy to let the
desire to demonstrate the relationship guide the choice of possible reconstructed values,
whether consciously or unconsciously. Thus, one should not let the rime value of a
reconstructed lookalike Mon-Khmer word subtly shape the reconstruction of a Proto
Hmong-Mien rime since this will have the effect of making one’s own case for Hmong-
Mien–Mon-Khmer stronger. In the case of the Hmong-Mien family, however, which is
small and compact compared to the other large language families of the area—Mon-
Khmer, Tibeto-Burman, Tai-Kadai, Austronesian—the genetic relationship of the member
languages has never been in question: even upon superficial examination, the lexical
similarities and interwoven phonological correspondences make the family relationship
obvious. There is no need to hypothesize about rime mergers in Hmongic—it is quite clear
that they took place—so there is no danger that we might reconstruct particular values in
order to bolster a particular hypothesis. And not only are there no negative consequences
of using a triangulation methodology, in this case there are negative consequences of not
doing so. From the Hmongic side, without a willingness to see how Proto Hmongic might
best fit into the larger picture of the family as one reconstructs, one will make mistakes that
will only have to be corrected at a later stage of the work. From the Mienic side, assuming
that these reflexes more closely reflect Proto Hmong-Mien, without a willingness to take
the historical fact of the Hmongic mergers into account, one will “hug the ground” too
closely in projecting the exact values of the Mienic reflexes back in time, leading to an
unconstrained, over-reconstructed Proto Hmong-Mien.

Minor rime mergers in Mienic

Given the fact that Mienic preserves many of the Proto Hmong-Mien rimes intact or
changed in only minor ways, Downer (1982) correctly points out that Mienic frequently
provides key evidence for Proto Hmong-Mien reconstruction. He provides the following
example: in Hmongic, ‘liver’ and ‘high/tall’ are homophonous in every language (for
example, in White Hmong both are /ʃia1/) and are thus identical at the Proto Hmongic level
(*hrinA). But the two are clearly distinguished in Mienic by a coda consonant contrast that
is lost in Hmongic: at the Hmong-Mien level they must be reconstructed as *-hri̯ ən
(2.56/18f) ‘liver’ vs. *-hri̯ əŋ (2.56/18g) ‘high/tall’.
But Hmongic is not the only side of the family to have lost key information in the rime.
On several occasions, mergers have taken place in Mienic, too, and the only way to arrive
22 Chapter 2

at the Proto Hmong-Mien reconstruction is to use information preserved in Hmongic. One

full example of this is presented below, followed by a list of other Mienic rime mergers.
There are two sets of words in Mienic which can only reasonably be reconstructed as
*ai in Proto Mienic, as L-Thongkum (1993) has done, since the reflexes do not vary
significantly:

Mienic
have maːi2 ma2 maːi2 maːi2 naːi2 naːi2 ma2 mai2 -
買 buy maːi4 ma4 maːi4 mai4 maːi4 maːi 4 - mai4 mai4
賣 sell maːi6 ma6 maːi6 mai6 maːi6 maːi6 ma6 mai6 mai6
come taːi2 ta2 taːi2 taːi2 taːi2 taːi2 ta2 - tɛi2
waist ʨaːi3 ka3 klaːi3 klaːi3 klaːi3 klaːi3 kla3 lai3 lai3

These five words have different Hmongic correspondences, however, and cannot all be
reconstructed at the Proto Hmong-Mien level with *ai:

have/buy/sell ɛ e ua a u əŋ oŋ a u ɔŋ ə ɪ ɔ ɔ *u̯ ɛ
come/waist a ɑ ua a o u a a a ɔ uɔ a a a *u̯ a

I have reconstructed Proto Hmong-Mien *ɛj for the first set, and Proto Hmong-Mien *aj
for the second set. The merger of two very similar Proto Hmong-Mien rimes to Mienic *ai
is thus not surprising.

Rime Hmongic Hmong-Mien Mienic

a) 11c *u̯ ɛ < *ɛj > *ai
15d *u̯ a < *aj > *ai

Other Mienic rime mergers appear below:

Rime Hmongic Hmong-Mien Mienic

b) 1c *i < *i̯ æn > *æn
19b *æn < *æn > *æn

c) 3c *æw < *əu > *əu

7d *o < *əw > *əu

d) 3d *æw < *au > *au

5c *æ < *æw > *au

e) 3f *æw < *uw > *u

7c *o < *u̯ o > *u
16c *u < *u > *u

f) 9d *uw < *ut > *ut

13f *ow < *əut > *ut

g) 9e *uw < *uk > *ok

 Proto Hmong-Mien 23

13e *ow < *ok > *ok

h) 18g *in < *i̯ əŋ > *i̯ əŋ

24d *aŋ < *i̯ aŋ > *i̯ əŋ

i) 20e *en < *eŋ > *eŋ

29j *ɔŋ < *u̯ eŋ > *eŋ

j) 21d *əŋ < *əŋ > *u̯ əŋ

29e *ɔŋ < *u̯ ɔŋ > *u̯ əŋ

k) 21g *əŋ < *əuŋ > *uŋ

28b *oŋ < *uŋ > *uŋ

l) 21h *əŋ < *əaŋ > *aŋ

24c *aŋ < *aŋ > *aŋ

m) 22c *ɛŋ < *ɛŋ > *əŋ

29h *ɔŋ < *u̯ əŋ > *əŋ

New patterns
To the extent that new cross-correspondence patterns of innovation, retention, and merger
come to light, they lend support to the values reconstructed for correspondences using the
“triangulation” method of reconstruction described above. As a bonus, once these patterns
have been independently established, they can help decide how other problematic sets
should be reconstructed. Described below are patterns across correspondence sets that have
emerged in the course of work on this project.

Hmongic rounding of rimes ending in -i and -j

The development of Proto Hmongic *u̯ ɛ and *u̯ a from Proto Hmong-Mien *ɛj and *aj
described above may have seemed odd. It is clearly necessary to do something to explain
rounded reflexes in Hmongic forms that correspond to a number of Mienic rimes that show
no sign of rounding, such as the following:

Hmongic Mienic
12/17c u ei ou au u o ɔ ou ou a ja i i i ei ei e ei ei ei i i ɛi
15c a ɑ ua a o u a a a ɔ uɔ a a a ai ai ai ai ai ai ai ai ai

One could reconstruct a rounded element in Proto Hmong-Mien which was subsequently
lost in Mienic. However, such a solution would ignore the pattern evident in the example
sets above: in each case, the Mienic reflexes end in /i/. The decision taken here is to
attribute the rounded element to a Hmongic innovation, the insertion of a -u̯ - on-glide
when the Hmong-Mien final ends in either *-i or *-j:
24 Chapter 2

Hmong-Mien Hmongic
11 *ɛj, *u̯̯ ɛj > *u̯ ɛ
12 *ei, *æi, *u̯ ei, *u̯ əi > *u̯ ei
15 *əj ,*aj, *u̯ əj > *u̯ a

This solution is preferred because not only are rime innovations more common in Hmongic
than in Mienic, there is a good parallel development in an unrelated language, arguing that
this is a kind of natural dissimilatory process. An identical phonological innovation took
place in the development of Old French -ei to Modern French -wa (Posner 1997:252-
262):14

Latin lēgem > Early Old French lei > Modern French loi
Latin mē > Early Old French mei > Modern French moi

Hmongic preserves the first part of complex Hmong-Mien rimes, but Pa-Hng and Qo
Xiong often preserve the second part
Not surprisingly, given the better retention of onsets in Hmongic, if Hmongic retains only
part of a complex Hmong-Mien rime, it will be the first part.15 Simplification of complex
Hmong-Mien rimes in these cases involves loss or reduction (to a glide) of non-initial
vocalic elements:

Hmong-Mien Hmongic
1 *i, *i̯ æn, *i̯ əu, *i̯ ɛk > *i
2 *ɨ, *i̯ eu, *eu, *ik, *ek > *ɨ
3 *æu, *əu, *au, *ɔu > *æw
9 *uj, *up, *ut, *uk, *u̯ ɛt, *u̯ ɛk > *uw
10 *e, *ej, *ep, *et > *e

In contrast to the pattern described above, Pa-Hng shows a distinct development across
correspondence sets, which, along with conservative aspects of its initial consonantism
(Benedict 1986; Strecker 1987a, b; Li and Mao 1997, Niederer 2004), make it appear as
the first language to separate from the Hmongic branch. The generalization that Hmongic
preserves the first part of the Hmong-Mien rime above can then more precisely be said to
apply to “Hmongic proper”, or to Hmongic languages other than Pa-Hng. A new discovery
of this reconstruction project is that Qo Xiong (representing Northern Hmongic) often
patterns with Pa-Hng in this regard, as will be shown below. The archaism of Qo Xiong
suggests that the structure of the Hmong-Mien family tree needs to be re-calculated from
the beginning, with no preconceptions about Hmongic language sub-grouping.
First, in three sets Pa-Hng and Qo Xiong appear to conserve the second part of Proto
Hmong-Mien diphthongs, against the usual pattern for Hmongic, which, if it conserves
only part, will conserve the first part (the reflex in parentheses goes against this pattern):

14
I am grateful to Geoff Nathan (p.c. 2002) for pointing out the French parallel.
15
Other possible outcomes are retention of the complete Hmong-Mien rime, or a “compromise” rime, a
merger of two Hmong-Mien vowel qualities. What one does not find is loss of the first part and
preservation of the last part of the Hmong-Mien rime.
 Proto Hmong-Mien 25

Hmong-Mien Pa-Hng Qo Xiong

3 *æu, *əu, *au, *ɔu > o ɯ, o, u
11 *ɛj, *u̯̯ ɛj > ɪ e
12/17 *ei, *æi, *u̯ əi > i, ɪ i, ei, (ɔ)

Second, in two sets, Pa-Hng and/or Qo Xiong conserve a trace of the coda consonant by
making a distinction in vowel quality depending on whether the Hmong-Mien syllable was
open or closed.

Hmong-Mien Pa-Hng Qo Xiong

4 *at > e, i ei, i
*a > a ɑ

7 *əp, *ət, *u̯ ət > a

*o, *u̯ o, *əw, *i̯ ou > o

In a third set, Qo Xiong conserves a trace of the coda consonant by making a tone
distinction according to whether the Hmong-Mien syllable was open or closed. Recall that
in other Hmongic languages (the languages of “Hmongic proper”), it did not matter
whether the Hmong-Mien rime was open or closed: the two types of syllable simply
merged into one.

Hmong-Mien Qo Xiong
13 tone 7 (< -p, -t, -k) > u
tones 1, 3, 4, 5, 6 > ə

In a fourth set, Qo Xiong conserves a trace of a Hmong-Mien rounded on-glide by making

a distinction in vowel quality on the basis of its presence or absence. Wang (1994) divided
this set into two sets (12, 17) solely on the basis of these two different Qo Xiong reflexes:
better reconstructions of the Hmong-Mien rimes that merge into this set reveal the
conditioning factor.

Hmong-Mien Qo Xiong
12/17 *ei, *æi > ei, i
*u̯ əi > ɔ

The delineation of a new family tree for Hmong-Mien on all relevant criteria is beyond the
scope of the present study, but it is clear that these patterns will help us do a better job of
sub-grouping than has been done to date. Any new tree that is proposed, however, will be
inadequate unless it includes a re-examination of the position of the North Hmongic
branch.

Chinese borrowings
If a reconstruction of Hmong-Mien is done well, independent support for the values
reconstructed should come from what is known about the form of Chinese words borrowed
by Hmong-Mien in the various rime categories. The reconstructed shape of obvious
Chinese borrowings should ideally dovetail with those proposed here for Hmong-Mien; to
26 Chapter 2

fall in with a particular rime correspondence set, a given Chinese borrowing must have
shared phonological properties with a particular phonologically-defined set of native words
(whether Proto Hmong-Mien, Proto Hmongic, or Proto Mienic).
The shape of Middle Chinese (c. 500 CE) is fairly well known (Pulleyblank 1991,
Baxter 1992), and progress is being made on the reconstruction of Old Chinese (c. 1500
BCE: Baxter 1992, Sagart 1999, Baxter and Sagart, 2009). The most difficult problems in
the reconstruction of Old Chinese involve the onsets rather than the rimes, so support for
the reconstruction of Hmong-Mien rimes from Chinese borrowings is accordingly more
valuable than support for the reconstruction of Hmong-Mien onsets from Chinese
borrowings (in the case of onsets, it may be that Hmong-Mien will be in a better position
to help clarify questions of Chinese historical phonology). For each rime correspondence
set in section 2.3 below, Chinese borrowings that pattern in the same way as the other
words of that set have been included, along with both the Hmong-Mien reconstruction and
the most recent reconstruction from either Baxter 2000 for Middle Chinese or Baxter and
Sagart 2009 for Old Chinese, whichever the reconstructed Hmong-Mien form resembles
most.

2.1.3.2 Vowel length in Mienic16

No language in the Hmongic branch has contrastive vowel length, nor do the Biao Min and
Zao Min languages of the Mienic branch. On the basis of vowel length contrasts in dialects
of the Mien and Mun languages alone, however, the majority view is that vowel length
should be reconstructed for Proto Hmong-Mien. Purnell (1970) and Downer (1982)
reconstruct only /a/–/aː/ for Proto Hmong-Mien. Wang and Mao reconstruct contrastive
vowel length for all vowels, but as they point out in the introduction (1995:15), only
dialects of Mun exhibit length contrasts in vowels other than /a/.
Downer held out the hope that reconstructing additional length contrasts could help
clear up some of the difficulties in linking Mienic rimes with the severely reduced number
of Hmongic rimes: “Since it is necessary to project the Yao length distinction back into
PMY [Proto Miao-Yao] (but not PM [Proto Miao]) for the low vowels, a further
assumption might be made: that PMY possessed similar length distinctions with other
vowels. Such an assumption could then explain some other cases where a single Miao rime
corresponds to two different Yao rimes …” (1982:5). But there is no evidence that Proto
Hmong-Mien vowel length played a role in the patterns of merger into Proto Hmongic: the
quality of nuclear and peripheral vocalic elements alone seem to have determined the
outcome of these mergers, as the present reconstruction makes clear.
Reconstructing vowel length is a reasonable idea, especially since Mienic typically
conserves rime contrasts which have merged in Hmongic. But Purnell is right: the
convenience of setting up Proto Hmong-Mien rimes as equivalent to Proto Mienic rimes
(including the length contrasts) “…is merely an attempt to organize the large number of
PMY final correspondences in a way that would facilitate further investigation…[it is] not
meant to imply that PY has preserved the PMY final system…intact” (1970:183).
Nonetheless, most have proceeded under the assumption that it has.
An exception, however, is L-Thongkum, who does not reconstruct vowel length in her
1993 reconstruction of Proto Mienic. She accounts for the development of a phonemic
vowel length contrast in /a/ as the result of a Proto Mienic contrast in vowel quality (*-ə >

16
This section is based on Ratliff 2007.
 Proto Hmong-Mien 27

-a; *-a > -aː).17 Under her account, other length contrasts, manifested sporadically, have
developed through contact with neighboring languages that have phonemic vowel length
contrasts. Interpreting the length contrast as one of vowel quality at the Proto Mienic level
would clearly remove motivation for the reconstruction of length at the even earlier stage
of Proto Hmong-Mien.
L-Thongkum’s contact/diffusion explanation for vowel length in Mienic seems
superior, for the following reasons (for further discussion, see Ratliff 2007):
1) Mien and Mun speakers are in contact with speakers of vowel-length languages.
Mien speakers in Thailand are in contact with vowel-length languages Standard and
Northern Thai (Purnell 1965:3 and L-Thongkum 1993:193). In the provinces of Hunan,
Yunnan, Guangdong, Guizhou, Jiangxi, and the Guangxi Zhuang Autonomous Region,
Mien and Mun speakers live among Tai vowel-length language speakers (most notably
Zhuang) and Chinese. On Hainan Island, Mun speakers are in a contact situation with
(among others) speakers of two languages with contrastive vowel length, Hlai (Li) and
Cantonese. Furthermore, Mien is not alone in showing a length contrast only in /a/. This is
also true of Shan, Tai Lɯ, Wuming Zhuang (Tai-Kadai), Cantonese (Sinitic), Chamic
(Austronesian), and undoubtedly other languages of the area.
2) Loanwords with long vowels outnumber native words with long vowels (Kosaka
2002).
3) Although stimulation from contact with vowel-length languages is important for the
development of vowel length in non-vowel-length languages, we know from studying
tonogenesis that languages must also have an “internal readiness” to develop the prosodic
features of neighboring languages. And there is an internal, phonetic reason why an
emergent phonemic length contrast should start with [a], as in Mien, and then be extended
to other vowels, as in Mun. Gordon (2002:73) has explained the connection between the
“well-documented tendency for low vowels to be crosslinguistically longer than high
vowels” and the development of phonemic vowel length. The internal subphonemic
variability lying ready for exploitation must have been the natural tendency of low vowels
to manifest a wider range of length differences than other vowels—but significantly, wide
durational differences in [a] are only found in languages which do not already have
contrastive vowel length: “In virtually all cases, languages without phonemic vowel length
display greater durational differences between vowels of different qualities… In languages
with phonemic length contrasts there is less room for the intrinsically longer low vowels to
enhance their inherent length by undergoing additional lengthening, because additional
subphonemic lengthening would potentially lead to neutralization of phonemic length
distinctions” (Gordon 2002:72-73). The historical stages in the development of vowel
length contrasts must then have been the following:

(a) at first no length contrasts, and low vowels show great variability in length;
(b) this length variability is then exploited (phonologized) in low vowels;
(c) and is then extended by analogy to other vowels.

17
In a synchronic study of Mien phonology, Purnell (1965:78ff.) also analyzes short [a] as /ə/, and
long [aː] as /a/.
28 Chapter 2

4) Two independent descriptions of Mien (Downer 1961, Purnell 1965) and a note on Mun
in Shintani and Yang (1990) show these two vowel length languages are characterized by a
prosody that alternates short and long (or light and heavy) syllables. Downer (1961:539)
describes Mien (“Highland Yao”) as follows: “[Reduced syllables] are of two kinds—
regular reduced syllables, and reduced syllables in -a. The two kinds agree in having
weaker stress and shorter duration than the following full syllables so that a strong iambic
rhythm is imparted to disyllabic words …”. Given the iambic rhythm of these languages,
not only must non-phrase-final syllables be light, but phrase-final syllables (syntactic
heads) must maintain a certain gravity.18 This may have been a contributing factor in the
development of phonemic length contrasts.
Since prosody is the most easily diffused aspect of phonology (Matisoff 2001), it is
therefore plausible that Mien and Mun became vowel-length languages in contact with
other languages of this type, in the first instance by exploiting the significant inherent
length peculiar to the vowel [a] that only occurs in languages that have had no previous
vowel length contrasts.

2.1.3.3 Nasal codas in Hmongic

There is no need to reconstruct an *-n/*-ŋ contrast for Proto Hmongic (see also Downer
1982 and Niederer 1998:276). The two final nasals occurred in complementary distribution
in Proto Hmongic, with *-n only after front vowels, and *-ŋ elsewhere:

in ɨŋ ʉŋ uŋ
en, ein oŋ
ɛŋ əŋ ɔŋ
æn aŋ

This differs from the Proto Hmongic reconstruction of Wang 1994, where four pairs of
rimes contrast *-n and *-ŋ:

in uŋ
en, eŋ on, oŋ
ən, əŋ ɔn, ɔŋ
æn aŋ ɑŋ

No evidence exists in the Hmongic reflexes for the reconstruction of these four minimal
rime pairs above. Comparison with Mienic does not support this reconstruction either,
since each rime corresponds either to all three possible nasals (-m, -n, -ŋ), or to a
haphazard subset of the three.
There are two possible ways to reflect the lack of contrast of final nasals in Proto
Hmongic: reconstruct a single *-N, the value of which is spelled out by the preceding
vowel, or reconstruct allophony, as presented here. Either way works: the latter course is
chosen here to make the reconstruction more directly informative and to provide a clearer

18
135 sentences illustrating basic syntactic types in the Shintani and Yang Hainan Mun dictionary (1990)
show a strong correlation between vowel length and phrase-final position (location of the syntactic head
of the phrase).
 Proto Hmong-Mien 29

link to Proto Mienic, but this should not be taken to mean that there were two contrastive
nasal codas in Proto Hmongic.19

2.1.3.4 Vowels or glides?

Another difficult problem in the reconstruction of rimes is to decide, given a particular
vowel quality one needs to reconstruct, whether that element should be reconstructed as a
full vowel or as a glide. Given the working rule that a coda can contain either an off-glide
or a consonant but not both, the following inventory of rime types in the protolanguage is
assumed:

monophthong nuclei V VC VG GV GVC GVG

diphthong nuclei VV VVC - GVV - -

Many possible rime types are not typical of the present-day languages nor are they
reconstructed. This suggests that there is a certain maximum weight to a Hmong-Mien
rime, wherein only the full vowels and the off-glide would count toward the calculation of
syllable weight. This would exclude syllable types VVG and GVVG as being too heavy.
By this calculation, a GVVC syllable would be theoretically possible, but has not been
needed in this reconstruction.
Beyond this, in a fairly impressionistic way, if the reflexes in a particular set show
stable diphthongs, a diphthong is reconstructed, but if the reflexes show a mix of
monophthongs and diphthongs, a GV or VG is usually reconstructed, with the more
lightweight glide assigned to the less stable element across the set.

2.1.3.5 Concluding remarks

A particular reconstructed Proto Hmong-Mien rime can be evaluated in three primary
ways:

(1) the Mienic reflexes are derivable from it in a straightforward way;

(2) it fits into a phonologically natural set with the other rimes that have merged
into the same Hmongic rime; and
(3) borrowed Chinese words that pattern the same way in Hmong-Mien were
phonologically similar to the reconstructed form at an earlier stage.

If these three independent lines converge, we can be fairly confident of the reconstruction;
and we will feel even better if cross-correspondence patterns begin to emerge given these
reconstructions. Yet there are remaining problems. First, the over-reconstruction of rimes
is still troublesome: it will be important to find conditioning factors to reduce the bloated
Proto Hmong-Mien rime inventory. Wang and Mao’s reconstruction of every possible low

19
Some languages have -n and -ŋ in complementary distribution (-n after front vowels and -ŋ after back
vowels). Others, like Jiongnai, have developed a true contrast between the two nasals: /an~ai/ from *in
and *æn, and /aŋ/ from *əŋ, *ɛŋ, or *ɔŋ.
30 Chapter 2

vowel on the IPA chart (*æ, *a, *A, *ɐ, *ɑ, *ɒ)20 appears to be a clear example of “over-
reconstruction”, but one could also say that this reconstruction is just as unnatural: what
present-day language makes a distinction between /ei/ and /ej/, between /æi/ and /æj/, or
worse, among /ɔu/, /ou/, and /ow/— rimes claimed here to have been distinctive in Proto
Hmong-Mien? It is clear that these values are abstractions, and only capture the true nature
of the protolanguage in a shadowy fashion. Second, several decisions were fairly arbitrary
and may well need to be changed in the light of more evidence or better analysis. For
example, the decision to reconstruct Proto Hmongic correspondence 21 as *əŋ and 22 as
*ɛŋ was based on very subtle differences in the reflexes of these two proto-phonemes.
Also, given secondary rounding in Hmongic, discriminating among the rounded nasal
rimes 27–30 at the Proto Hmongic level (*uŋ, *oŋ, *ɔŋ, *ʉŋ) was quite difficult.

2.1.4 Issues in the reconstruction of tones

Hmong-Mien tones are reconstructed as abstract historical tone categories rather than as
ancient tones with particular values (high, low, rising, etc.). Given the great variation in the
phonetic reflexes of a particular category, few common tone features emerge for any
particular category. The categories themselves, however, are remarkably stable, which
makes this aspect of the reconstruction the most straightforward.21 A glance at a page of
the rime tables in section 2.3 shows that cognate words typically belong to the same tone
category, as represented by superscript numbers from 1 through 8.22 The occasional form
with a non-corresponding tone, whether due to the “promotion” of a tone sandhi variant or
for some other reason, does not generally obscure the proper assignment of a word to a
particular tone category and thus, to a particular syllable type in the protolanguage. See
chapter 3 for a discussion of tonogenesis in Hmong-Mien and the interpretation of
historical tone categories.
The major difficulty with the reconstruction of tones (i.e., tone categories) involves
those sets of forms—on other measures cognate—that show two different tone reflexes.
Often the distribution of tones follows the division between Hmongic and Mienic. For
example:

Hmong-Mien Hmongic Mienic

to eat ?? > *nuŋA *ɲənC
finger ?? > *ntaB *ʔdokD
crossbow (*hnək) > *hnænB *hnəkD
3SG (he/she/it) *ni̯ æn(X) > *niB *nænA
to go *n-mʉŋ(X) > *n-mʉŋB *n-mɨŋA

Although ‘to eat’, ‘finger’ and ‘crossbow’ are undoubtedly Proto Hmong-Mien words, they
are listed separately in this work, since they fall into different correspondence sets. In some
words that begin with a nasal, like ‘crossbow’, secondary nasalization in the rime makes

20
[A] is not a common IPA symbol. It indicates “a fully open central unrounded vowel” (Pullum and
Ladusaw 1996).
21
Kun Chang (1947, 1953, 1966, 1972) was the first to elucidate the history of tones in Hmong-Mien.
22
Or by superscript letters in the case of protolanguages and Fuyuan, a Hmongic language that did not split
the original 4 tones into 8 upon merger of voiced and voiceless consonants.
 Proto Hmong-Mien 31

Hmongic and Mienic difficult to reconcile at the higher level. But when the tone is the only
ambiguous element, as in ‘he/she/it’ and ‘to go’, it is reconstructed, with indication (by use
of parentheses) that variant forms probably existed in the protolanguage. In the event of a
tone mismatch across the Hmongic/Mienic divide, therefore, decisions about whether or
not to reconstruct, and how to reconstruct, at the Proto Hmong-Mien level have been made
on a case-by-case basis; others may choose to handle these cases differently.
One special case is the mismatch of Mienic tone 7 to Hmongic tone 5, or Mienic tone 8
to Hmongic tone 6. In this case we can confidently reconstruct to the Proto Hmong-Mien
level: there is ample evidence to show that there was a merger of Proto Hmong-Mien
words with final -k (category D) to the corresponding register of category C in Hmongic
(Downer 1967, Chang 1972).23 This merger did not affect Proto Hmong-Mien words that
ended in -p or -t. For example:

Hmong-Mien Hmongic Mienic

bird *m-nɔk > *m-nɔŋC *nɔkD
guest (MC khæk) *Khæk > *qhæC *khækD
hundred (MC pæk) *pæk > *pæC *pækD
strength (OC *kə.rək) *-rək > *-roC *khləkD
six (TB *k-ruk) *kruk > *kruwC *krokD

2.2 Proto Hmong-Mien onsets

1 2 3 4 5 6 7
1. Stop voiceless p t ts c k q/(qʷ) ʔ
2. Aspirated ph th tsh ch kh
3. Voiced b d dz ɟ g ɢ
4. Pre-nasalized stop voiceless mp nt nts ɲc ŋk ɴq
5. Aspirated mph nth ntsh ɲch ŋkh
6. Voiced mb nd ndz ɲɟ ŋɡ ɴɢ
7. Nasal pre-glottalized ʔm ʔn ʔɲ
8. Aspirated hm hn hɲ
9. Voiced m n ɲ/(ɲʷ) (ŋ)/(ŋʷ)
10. Glide pre-glottalized ʔw ʔj
11. Aspirated (hw) hj
12. Voiced w j
13. Fricative voiceless s ɕ h
14. Aspirated
15. Voiced (ɣ) (ɦ)

-j- 1 2 3 4 5 6 7
16. Stop voiceless pj tj tsj kj/(kʷj)
17. Aspirated phj (thj) tshj khj
18. Voiced bj dj dzj gj/(gʷj)
19. Pre-nasalized stop voiceless (mpj) ntj ntsj ŋkj
20. Aspirated mphj nthj ntshj
21. Voiced mbj (ndj) ndzj ŋgj

23
For an explanation of “tone C”, “tone D”, and “register”, see chapter 3.
32 Chapter 2

22. Nasal pre-glottalized

23. Aspirated hmj hnj
24. Voiced mj (ŋj)
25. Glide pre-glottalized
26. Aspirated
27. Voiced wj
28. Fricative voiceless sj (xj)
29. Aspirated
30. Voiced (ɣj)

l-, -l- 1 2 3 4 5 6 7
31. Stop voiceless pl/plj kl/(klj) ql
32. Aspirated phl khl
33. Voiced bl/blj gl/glj ɢl/(ɢlj)
34. Pre-nasalized stop voiceless mpl/mplj ŋkl/ŋklj
35. Aspirated mphl
36. Voiced mbl/mblj ŋɡl/ŋɡlj ɴɢl
37. Nasal pre-glottalized
38. Aspirated
39. Voiced ml
40. Liquid pre-glottalized ʔl/ʔlj
41. Aspirated hl/hlj
42. Voiced l/lj

r-, -r- 1 2 3 4 5 6 7
46. Stop voiceless pr tr kr qr
47. Aspirated phr (thr)
48. Voiced br dr gr ɢr
49. Pre-nasalized stop voiceless mpr ntr ŋkr
50. Aspirated nthr
51. Voiced mbr ndr ŋɡr
52. Nasal pre-glottalized
53. Aspirated hnr
54. Voiced mr
55. Liquid pre-glottalized ʔr
56. Aspirated hr
57. Voiced r
 Proto Hmong-Mien 33

Correspondences sets
The correspondence sets are organized by place of articulation category. The first sets
contain no medial consonant. They are followed by correspondence sets for the same place
of articulation in combination with each of the medial consonants (-j-, -l-, -r-). Within each
set, those etyma that can be reconstructed to Proto Hmong-Mien (PHM) are presented first,
followed by etyma that can be reconstructed only to Proto Hmongic (PH), and then etyma
that can be reconstructed only to Proto Mienic (PM). Within each subset, etyma with no
initial matter are presented before etyma with reconstructed initial matter. Within each of
these smaller groups, etyma are ordered first by tone category.
Onsets that contain two medial approximants (e.g. *hlj-) will follow after their single
medial approximant counterparts (hence *pl- is 1.31 and *plj- is 1.31.1). The onsets with a
third numerical subdivision will all be onsets of this type.
Obvious Chinese loanwords are included in the correspondence sets to which they
belong. The entries appear at the end of each list. If a connection to a particular Chinese
form is possible but the connection is not straightforward due to a serious mismatch of one
or more features (tone, initial, final, or meaning), the form is given in the notes. If a
correspondence set contains only Chinese loanwords, this marks the reconstructed onset as
belonging to loan phonology, and it appears in parentheses in the chart above. Loan
phonology is discussed in chapter 6 on language contact.

Languages
Eleven maximally distinct languages were chosen to represent the 20-30 languages of the
Hmong-Mien family. Since Hmongic is more internally diverse than Mienic, seven of the
eleven are Hmongic languages. If a form was taken from Wang and Mao 1995, Wang
1994, or the first (primary) source listed, no mention is made of the source of the form.24 If
a form was taken from the second or third source listed, or if another closely related dialect
was substituted for the one listed here, the source and dialect will be given in the notes.
The eleven criterial languages will be referred to in discussion by the underlined names.

1. East Hmongic (Qiandong): Northern vernacular: Guizhou, Kaili County, Yanghao

Township, Yanghao
Sources: Wang and Mao 1995, language 1 = Wang 1994, language 1
2. North Hmongic (Xiangxi): Western vernacular: Hunan, Huayuan County, Jiwei
Township, Jiwei
Sources: Wang and Mao 1995, language 2 = Wang 1994, language 2
3. West Hmongic (Chuanqiandian): CQD subdialect, #1 vernacular: Laos/Thailand, White
Hmong
Sources: Heimbach 1979, Bertrais 1979 [1964]
4. West Hmongic (Chuanqiandian): Mashan subdialect, Central vernacular: Guizhou
Province, Ziyun Miao/Buyi A.C., Zongdi Township, Zongdi
Sources: Wang and Mao 1995, language 7 = Wang 1994, language 7 (tones 1a,
3a, 5a, 7a < voiceless stop, preglottalized stop, and preglottalized sonorant onsets;

24
For some reason, not all forms from Wang 1994 were included in Wang and Mao 1995. Unless obvious
Chinese loans, they have all been incorporated here.
34 Chapter 2

tones 1b, 3b, 5b, 7b < voiceless aspirated stop, voiceless fricative, and voiceless
sonorant onsets)
5. West Hmongic (Chuanqiandian): Luopohe subdialect, #2 vernacular: Guizhou Province,
Fuyuan County, Fuyuan
Sources: Wang and Mao 1995 language 8 = Wang 1994, language 8 (no tone
split)
6. Hmongic: Guangxi/Zhuang A. R., Jinxiu Yao A. C., Changdong Township, Jiongnai
Sources: Wang and Mao 1995, language 13; Mao and Li 2001
7. Hmongic: Guangxi/Zhuang A. R., Miao A.C., Rongshui District, Baiyun Pa-Hng
Source: Niederer 1997
8. Mienic: Mien, Luoxiang vernacular: Guangxi/Zhuang A. R., Jinxiu Yao A.C., Luoxiang
Township, Mien
Sources: Wang and Mao 1995, language 17; L-Thongkum 1993 “JX” (tone 5 <
voiceless stop, preglottalized stop, and preglottalized sonorant onsets; tone 5’ <
voiceless aspirated stop, voiceless fricative, and voiceless sonorant onsets)
9. Mienic: Jinmen: Guangxi/Zhuang A. R., Lingyun County, Lanjin Township, Mun
Source: Wang and Mao 1995, language 20 (tones 1, 3, 5, 7 < voiceless stop,
preglottalized stop, and preglottalized sonorant onsets; tones 1’, 3’, 5’, 7’ <
voiceless aspirated stop, voiceless fricative, and voiceless sonorant onsets)
10. Mienic: Biaomin: Guangxi/Zhuang A. R., Quanzhou county, Dongshan Yao Township,
Biao Min
Sources: Wang and Mao 1995, language 21; Solnit 1982
11. Mienic: Zaomin: Guangdong Province, Liannan Yao A. C., Daping Township,
Zao Min
Source: Wang and Mao 1995, language 23

The names for these languages used in the text (underlined above) are those most
commonly used in the literature on Hmong-Mien languages. They are therefore not
consistent: they are a mix of four Chinese place-names (Yanghao, Jiwei, Zongdi, Fuyuan),
one English translation of a Hmong autonym, based on the color of formal dress (White
Hmong), and six autonyms in their most common spellings (Jiongnai, Pa-Hng, Mien, Mun,
Biao Min, Zao Min). Data from these eleven languages appear beneath the eleven
numbered columns in the following correspondence sets.
 Proto Hmong-Mien 35

1. Labials

1.1 *p-
PHM *p- 1 2 3 4 5 6 7 8 9 10 11
1. to know *pei pu1 - pau1 pɔ1a - pe1 pɪ1 pei1 pei1 pəi1 bɛi1
2. molar tooth *pæ - pa2 pua1 - - - - - - - ba1
3. full *pu̯ ɛŋX pɛ3 pe3 pu3 poŋ3a paŋB paŋ3 pɔ̃3 pwəŋ3 pɔŋ3 pɔŋ3 baŋ3
4. to shoot *pənX paŋ3 pɑŋ3 pɔ3 poŋ3a poŋB poŋ3 pɔ̃3 pwan3 fan3 - bun3
5. to sleep/lie down
pi5 pə5 pɯ5 pu5a puC pau5 pɛ5 pwei5’ fei5 - bui5
*pu̯ eiH
6. 1PL (we/us) *N-pɔu pi1 pɯ1 pe1 pæ1a peiA pa1 pɯ5 - pu1 - bu1
7. husband *N-poX - po3 - - - - - - - bu3 -
8. soybean *N-peiX pu3 - pau3 - - - - - bei3 bi3 mɛi3
9. 百 hundred *pæk pa5 pa5 pua5 pa5a piC pa5 pe5 pɛ7 pe7 pɛ7 ba7

PH *p- 1 2 3 4 5 6 7
10. to fall *pʉŋA - - pɔŋ1 paŋ1a poŋA - -
11. to feed *pɛŋA pɛ1 - pu1 - - - -
12. thigh *pæA pa1 pa1 pua1 pa1a - - pe1
13. to open (door)
pu7 pu7 - - poD - pɔ7
*powD
14. 風 air *poŋC poŋ5 - pa5 paŋ5a poŋC - -

PM *p- 8 9 10 11
15. to burn *pɔuB pu3 pu3 pau3 bu3
16. 鞭 whip *pinA biŋ1 pin1 pin1 -
17. 斧 axe *pouB pou3 pɔu3 bəu3 pu3
18. hide oneself *piŋC
piːŋ5 - pjɛ5 bɔŋ5
(Cant. /pɛːŋ5/)
19. 放 to let go *puŋC puŋ5’ puŋ5 pə5 bɔŋ5

Notes
1. Possibly from Chinese 別 ‘to separate, distinguish’ (OC *N-pret > MC bjet > Man. bié).
2. In White Hmong, ‘jaw, chin’.
3. Compare Malayo-Polynesian *penuq ‘full (as a container)’ (ACD).
4. Compare Malayo-Polynesian *panaq ‘to shoot’, Mon-Khmer *paɲʔ ‘to shoot’ (Shorto
#905). In White Hmong, ‘to throw’.
5. The tone subcategory in Luoxiang Mien (8) indicates possibility of aspirated initial.
6. This word is used as a plural suffix in Mien, suggesting an original meaning of ‘group’.
In Hainan Mun, it is a 1PL exclusive pronoun. The voiced initials in Mienic suggest a pre-
initial voicing element.
7. Compare Chinese 父 ‘father’ (OC *[b](r)aʔ > MC bjuX > Man. fù). See also Hmong-
Mien ‘father/male’ *pjaX (1.16/4).
10. Possibly from Chinese 崩 ‘to fall, collapse (of mountain)’ (OC *Cə.pˤəŋ > MC pong >
Man. bēng).
36 Chapter 2

11. In Yanghao (1), ‘to allow’; in White Hmong, ‘to give’. Yanghao (1), Xuyong /po1/ ‘to
feed’, Shimen /pu1/ ‘to feed’ from Office of Miao-Yao Research Lexicon (1987).
18. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).

1.2 *ph-
PHM *ph- 1 2 3 4 5 6 7 8 9 10 11
1. 片 CLF-quilts *phəan phaŋ1 - phɔ1 - phoŋA - phɯ5 phaːn1 - phəu1 -
2. 劈 to split/chop *phek pha1 pha1 tshi1 - phoA - - phi7 - - -
5 5 5 5b C 5 5 5 5’ 5
3. 破 to cut open *phajH pha pha phua pa pha pha pha phaːi paːi pha -

PH *ph- 1 2 3 4 5 6 7
4. to push earth aside *phʉA phɛ1 - phua1 pou1b - phau1 phɪ1
5. thunder *S-phoA ho1 so1 sɔ1 so1b suA xu1 (mo1)
6. thread *S-phoB fhə3 so3 sɔ3 so3b suB - (mo3)
7. to twist/rub *S-phu̯ aA fha1 - sua1 sa1b saA va1’ -

PM *ph- 8 9 10 11
8. 奉 to carry in 2 hands
phwəŋ3 pɔŋ3 phɔŋ3 -
*phu̯ ɛŋB
9. 拂 to sweep *phu̯ ətD phut7 - - -

Discussion
‘Thunder’ here is probably related to Hmong-Mien *mpuə ‘thunder’ (1.4/16), and the Pa-
Hng form for ‘thread’ suggests a similar pattern. A *ph- is reconstructed for ‘thunder’,
‘thread’ and ‘to twist/rub’ on the basis of ph- onsets in other locations: Fengxiang (West
Hmongic), Qibainong Bunu, and Yaoli Bunu (9, 10, 11 in Wang and Mao 1995). The Pa-
Hng forms in parentheses above pattern with Mienic, and suggest the reconstruction of an
initial nasal element, but since the nasal element shows no trace in the other languages
above, it may be that the nasal was “detachable”—that is, it was more likely to have been
morphological than phonological. Finally, the reconstruction of an aspirated initial for
these three words is tentative: Zongdi (4) voiceless “b” tones indicate an original [spread
glottis] initial, but an old [spread glottis] initial may have been either an aspirated stop, a
voiceless fricative, or a voiceless sonorant. A hypothetical S- prefix might have had the
same effect on Zondgi tones as an aspirated initial, thus an original *S-p- for these three
words (with secondary aspiration in Fengxiang, Qibainong Bunu, and Yaoli Bunu) is
possible. See also ‘chaff/husk’ for a similar split (Hmongic *S-phjæC (1.17/5) and
Mienic/Pa-Hng *mphi̯ ək (1.5/1).

Notes
1. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). In White
Hmong, this word serves as a classifier for sheets of skin or leather.
2. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). This
word is pan-Asian: compare Proto Tai *phreːk ‘split open’; Chinese 劈 ‘split, chop, cleave’
(OC *(m-)pʰˤek > MC phek > Man. pī). These forms are probably from Chinese because of
the non-correspondence in tones—they suggest an earlier borrowing in Mienic (from a
form in * -k), and a later borrowing in Hmongic (after D > A).
 Proto Hmong-Mien 37

1.3 *b-
PHM *b- 1 2 3 4 5 6 7 8 9 10 11
1. flower *bi̯ aŋ paŋ2 pei2 pa2 pua2 venA pen2 pɛ2 pwaŋ2 faŋ2 pjaŋ2 pjaŋ2
2. to see *bu̯ ət poŋ8 - pɔ8 po8 mpuD pu8 pa8 pwat8 fat8 phi8 -
3. 伏 to hatch/embrace
pə6 - pua6 pəa6 vaC pu6 po6 pu6 pu6 pu6 pu6
*buəH

PH *b- 1 2 3 4 5 6 7
4. female/woman *boA - - pɔ2 - vuA - -
5. to repay *bu̯ eiA pə2 pi2 pau2 pɔ2 vuA - -
6. face *bowB - - - pɯ4 - - -
7. mountain *bæwB pi4 - pe8 pe4 veiB - -
8. bracelet *bɔC - pɔ6 pau6 po6 vuC - po6
9. pus *bu̯ eiC pu6 pɔ6 pau6 po6 vuB pei6 pɪ6
10. shoulder *bʉC - pə6 pɯ6 - vuC - -
11. 腐 bad/spoiled *bu̯ aB pa4 pɑ4 pua4 pəa4 vuB - -

PM *b- 8 9 10 11
12. to hold w/2 hands 2 2
pou pɔu - -
*bouA
13. 平 level *beŋA pɛŋ2 peːŋ2 pɛ2 pɛŋ2
14. 耙 to rake *baA pa2 pa2 - pa2
15. 薄 thin *bi̯ ɛkD pwa8 fa6 - pjɛ8
16. 白 white *bækD pɛ8 pe6 phɛ8 pa8

Discussion
Proto Hmong-Mien ‘hand/arm’ and Proto Hmongic ‘thorn’ are best reconstructed with this
initial, although the correspondence is obscured by the process of “prefix pre-emption”
(Matisoff 1997). See discussion in chapter 2, section 2.1.2 and chapter 4, section 4.2.3.

1 2 3 4 5 6 7 8 9 10 11
hand/arm *-bɔuX pi4 tɯ4 te4 ʂe4 weiB tʃa4 tɛ-pɯ4 pu4 pu4 pau4 pu4
thorn *-boB pə4 to4 pɔ4 pɯ4 vuB - -

In addition to the forms above, Hmongic Ho Ne (She) forms belonging to these sets have
corresponding rimes and tones, but velar initials (‘hand’ /khwa4/ and ‘thorn’ /khɤ4/).

Notes
1. The -w- in Luoxiang Mien (8) is secondary: -j- dropped, then -w- developed between a
labial and /a/. Compare Austronesian *buŋa ‘blossom of fructifying tree or bush’ (ACD),
Written Burmese /pwɑŋ1/ ‘blossom’, and Laqua (Tai-Kadai) /puŋ/ ‘flower’ (Benedict
1975:295).
3. Compare the semantics of Chinese 抱 bào ‘hold in arms, embrace; hatch, brood’. The
meaning ‘to hatch’ is attested across the family; the additional meaning ‘to embrace’ is
attested only in Hmongic.
4. Xuyong /po2/, Shimen /po2/ (Office of Miao-Yao Research Lexicon 1987). Perhaps
borrowed from Chinese 婦 ‘wife, married woman’ (MC bjuwX > Man. fù), although the
tones do not correspond.
38 Chapter 2

6. Also Shimen /bey4I/, Qingyan /pau4/, Gaopo /pə4/, Fengxiang /pɔ4/ (Wang and Mao
1995). Xuyong /bleu4/ with medial -l- may also be cognate (in Xuyong /bleu4 plho7/,
Office of Miao-Yao Research Lexicon [1987]; for the second element, compare White
Hmong /phlu5/‘cheek’).
7. White Hmong ‘up there on the slope’. The tone in White Hmong is the result of a recent
morphological class formation by tone-shift (geographical feature > locative) (Ratliff
1992a:104-112).
8. Green Mong has been substituted for White Hmong (3).
12. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
15. The -w- in Luoxiang Mien (8) is secondary: see note 1 above.

1.4 *mp-
PHM *mp- 1 2 3 4 5 6 7 8 9 10 11
1. thunder *mpuə - - - - - - mo1 bu1 bu5 - bjau1
2. to dream *mpeiH pu5 mpei5 mpau5 mpɔ5a mʔpuC mpe5 mi5 bei5 bei5 bəi5 bɛi5
3. name *mpɔuH pi5 mpu5 mpe5 mpæ5a mʔpeiC mpa5 mo5 bu5’ bu5 bau5 bu5
4. 沸 to boil (INTR)
- - mpau5 mpɔ5a mʔpuC mpei5 mɪ5 bwei5’ vei5 - bui5
*mpu̯ æiH

PH *mp- 1 2 3 4 5 6 7
5. to hold in mouth *mpæB - - mpua3 mpa3a mʔpziB - -
6. pig *mpæC pa5 mpa5 mpua5 mpa5a mʔpeiC mpei5 me5
7. to throw on (clothes)
pa5 - mpua5 mpa5a - - ma5
*mpu̯ aC

PM *ʔb- (< *mp- ) 8 9 10 11

8. loud/sound *ʔbujA bui1 buːi1’ bəi1 bai1
9. bat (animal) *ʔbujA bui1 buːi1’ bəi1 bai1
10. monkey *ʔbiŋA biŋ1 biːŋ1 - bjaŋ1
11. grey *ʔbɔuB bu2 bu3 bau3 -
12. rock *ʔbeŋC bɛŋ5' - bɛ5 bɛŋ5
13. 報 to tell *ʔboC bu5' bu5 bu5 bɔu3

Discussion
In this and all subsequent prenasalized onset sets, following L-Thongkum (1993),
Hmong-Mien (*NT-) > Mienic (*ʔD-)
Hmong-Mien (*NTH-) > Mienic (*DH-)
Hmong-Mien (*ND-) > Mienic (*ND-)
Without reconstructing more material than absolutely necessary, this gives enough
information to account for (1) the voiced onsets that regularly correspond to Hmong-Mien
prenasalization in each case, and (2) the correct tonal register: 1, 3, 5, and 7 after
glottalized and aspirated onsets (*ʔD-, *DH-), and 2, 4, 6, and 8 after voiced onsets
(*ND-).

Notes
1. This may be related to Hmongic ‘thunder’ *S-phoA. See discussion under 1.2 above.
2. Jinxiu Mien (8) from L-Thongkum 1993:205.
 Proto Hmong-Mien 39

6. Compare Tibeto-Burman *pʷak ‘pig’ (Matisoff 2003).

7. In White Hmong, ‘to wrap around the body’. Possibly from Chinese 被 ‘to cover self’
(OC *m-p(r)ajʔ > MC bjeX > Man. bèi).
8. The Mun (9) tone subcategorization suggests aspirated initial; L-Thongkum reconstructs
Proto Mienic *bhui A (1993:215).
10. The Hmongic word for ‘monkey’ (*ʔlinA 2.40/18) has the same tone and rime as this
word, but begins with a different initial. To join the two at the Hmong-Mien level,
Benedict (1975:338) reconstructed a cluster: *ʔbliŋ.
11. In Hainan Mun, the cognate means ‘blue’ and the word for ‘grey’ is the compound
‘white-blue’; in Biao Min, it also means ‘mildew, mold’ (Solnit 1982).

1.5 *mph-
PHM *mph- 1 2 3 4 5 6 7 8 9 10 11
1. chaff/husk *mphi̯ ɛk - - - - - - m̥ e5 bwa7 va7’ bja7 bjɛ7
2. 噴 to scatter/
- mphu5 mphɔŋ5 - mʔphoŋC - - bwan5 ban5 - -
sprinkle *mphu̯ ənH

PH *mph- 1 2 3 4 5 6 7
3. ant *mphæB phen3 mpha3 - - - mphai3 m̥ je3

PM *bh- (<*mph- ) 8 9 10 11
4. 拍 to clap *bhækD bɛ7 be4 bɛ7 -

Notes
1. See Hmongic *S-phjæC ‘chaff/husk’ (1.17/5), and discussion of words that pattern
similarly under 1.2. The -w- in Luoxiang Mien (8) is secondary: -j- dropped, then -w-
developed between a labial and /a/.
2. Changping Mien (Wang and Mao 18) has been substituted for Luoxiang Mien (8).
4. This is probably the source for Hmongic *mbæA ‘to clap’ (1.6/5) as well.

1.6 *mb-
PH *mb- 1 2 3 4 5 6 7
1. to cover *mboC mə6 - mpɔ6 mpu6 mpuC - pho5
2. 拍 to clap *mbæA ma2 ma2 mpua2 mpa2 mpziA - -

PM *mb- 8 9 10 11
3. 婦 daughter-in-
bwəŋ4 bɔŋ4 bɔŋ4 -
law/bride *mbu̯ ɛŋB
4. 浮 to float *mbi̯ əuA bjeu2 bjɔu2 bjau2 bɛu2

Notes
2. This is probably the source for Mienic *bhækD ‘to clap’ (1.5/5) as well.
3. Although this word probably comes from Chinese 婦 ‘woman/wife’ (OC *Cə.[b]əʔ >
MC bjuwX > Man. fù), Chinese has no final nasal.
40 Chapter 2

1.7 *ʔm-
PHM *ʔm- 1 2 3 4 5 6 7 8 9 10 11
1. illness/pain *ʔmun moŋ1 moŋ1 mɔ1 məŋ1a ʔmoŋA mɐŋ1 mæ̃ 1 mun1 mun1 mən1 man1

PH *ʔm- 1 2 3 4 5 6 7
2. to grasp *ʔmu̯ ɛA mɛ1 me1 mua1 muɑ1a - - -

PM *ʔm- 8 9 10 11
3. green/blue *ʔmeŋA mɛŋ1 meːŋ1 mɛ1 mɛŋ1

Notes
2. In Zongdi, ‘to grasp’ has narrowed to mean ‘to take a wife’. There is a morphological
relationship between ‘to grasp’ and the more widespread Hmong-Mien word *n-mɛj ‘to
have’ (1.9/11) that can be attributed to either a voiceless causative prefix on ‘to grasp’ (of
which the glottal element is a trace) or a voiced stative prefix on ‘to have’ (which would
account for the /n-/ initial in Mun). For more discussion, see chapter 4.

1 2 3 4 5 6 7 8 9 10 11
to grasp H *ʔmu̯ ɛA
mɛ1 me1 mua1 muɑ1a - - -
(<*ʔ-mɛj)
2 2 2 A 2
to have HM *n-mɛj mɛ me mua - ma mɔ mɪ2̃ maːi2 naːi2 ma2 -

3. Hmongic *mpruA ‘green/blue’ (1.49/16) may be related.

1.8 *hm-
PHM *hm- 1 2 3 4 5 6 7 8 9 10 11
1. flea *hmiəŋ m̥ hen1 - hm̥ ɔ1 mɔŋ1b m̥ enA mɔ1 n̩ 1 mɛŋ1 muŋ1’ m̥ au1 mɔŋ1
2. vine *hmein - ɕi1 hm̥ a1 ma1b m̥ oŋA m̥ e1’ m̥ a1 m̥ ei1 mei1’ m̥ əi1 mɛi1
3. to taste/try *hmeiH m̥ hu5 - - - - - - - - m̥ i5 -
4. night *hməŋH m̥ haŋ5 m̥ hɑŋ5 hm̥ ɔ5 mɔ5b m̥ oŋC - mæ̃ 4 m̥ aŋ3 - m̥ ɔŋ5 mɔŋ5

PH *hm- 1 2 3 4 5 6 7
5. Hmong *hmʉŋA m̥ hu1 ɕoŋ1 hm̥ ɔŋ1 maŋ1b m̥ joA - ŋ̥ŋ1
6. wild dog *hmaŋC m̥ hɑŋ5 - hm̥ a5 ma5b - - -

PM *hm- 8 9 10 11
7. animal fat/oil *hmejA m̥ ei1 mei1’ m̥ əi1 mi1
8. 面 face *hmienA mjen1 min1’ m̥ in1 mɛn1
9. 米 husked rice *hmeiB m̥ ei3 mei3 m̥ i3 mɛi3

Notes
2. Possibly the same as Chinese 蔓 ‘creeping plant’ (OC *ma[n]-s > MC mjonH > Man.
wàn).
3. Both ‘to taste/try’ and ‘to sniff at’ (Hmong-Mien *hmji̯ əmH, 1.23/18) mean to taste or
smell intentionally. Different words exist for ‘to taste’ and ‘to smell’ unintentionally.
Possibly related to Chinese 味 ‘taste (n.)’ (OC *m[ə][t]-s > MC mjɨjH > Man. wèi).
 Proto Hmong-Mien 41

4. Chinese 晚 ‘evening’ (Man. wǎn) > Mien /muən1/ seems to be different. More likely
connections are to Tibeto-Burman *s-muːŋ ‘dark’ (Matisoff 2003) or Mon-Khmer
*m[h][ɯə]h ‘evening, night’ (Shorto #264) or *maŋ ‘night, evening’ (Shorto #638).
5. ‘Mien’ and its variants is probably from Chinese 民 ‘people’ (OC *mi[n] > MC mjin >
Man. mín).
6. This word refers variously to ‘wolf’, ‘fox’, ‘jackal’, etc. Compare Proto Tai *hma1 ‘dog’
(Li 1977).
7. Compare Austronesian *SimaR ‘grease/oil/fat’ (ACD).
9. Same as Chinese 米 ‘husked rice’ (OC *[m]ˤijʔ > MC mejX > Man. mǐ) and Tibeto-
Burman *mey or *may (Matisoff 2003), but the immediate source is unclear because of the
voiceless initial. Compare AN *Semay ‘cooked rice’ (ACD).

1.9 *m-
PHM *m- 1 2 3 4 5 6 7 8 9 10 11
1. 2SG (you) *mu̯ ei moŋ2 mɯ2 - - - mɐŋ2 mɯ6 mwei2 mei2 məi2 mui2
2. 2PL (you) *mi̯ əu maŋ2 me2 ne2 mein2 menA - mɯ5 - njou2 - -
3. fine (flour)
moŋ4 mɑŋ4 mɔ4 - - - mæ̃ 4 mwən6 - mun6 mun6
*mənX/H
4. eye *mu̯ ɛjH mɛ6 me6 mua6 moŋ6 maC mɔ6 mɪ6̃ mwei6 ŋwei6 mi6 mai2
2
5. to have *n-mɛj mɛ me2 mua2 - maA mɔ2 mɪ2̃ maːi2 naːi2 ma2 -
6. to go *n-mʉŋ(X) moŋ4 moŋ4 mu4 məŋ4 muB ŋ̥ŋ4 nɪ2̃ miŋ2 niŋ2 - mi2
7. 買 to buy *mɛjX mɛ4 - mua4 - maB mɔ1 mɪ6̃ maːi4 maːi4 - mai4
8. 賣 to sell *mɛjH mɛ4 me6 mua6 moŋ6 maC mɔ4 mɪ6̃ maːi6 maːi6 ma6 mai6
9. 蠓 fly (N.) *məuŋX - mɑŋ4 mɔ4 moŋ4 moŋB mɐŋ4 mæ̃ 4 - - mə4 mɔŋ4
10. 望 to look at
- - - - - - - - ŋwaːŋ6 mə6 mɔŋ6
*maŋH

PH *m- 1 2 3 4 5 6 7
11. persimmon *minB mi4 mɛ4 - - menB - -
12. wheat *muŋC maŋ4 - mɔ6 məŋ6 muC - -
13. female *minD mi8 - nia8 mi8 menC mai8 mɛ7̃

PM *m- 8 9 10 11
14. bee *mu̯ eiB mwei4 ŋwei4 mi4 mui4
15. brother *mɔuB mai4 - mau4 mu4
16. carrying pole
moŋ5 muŋ4 mɔŋ4 -
*mu̯ əŋB
17. to return home
mu4 mu4 mu4 mu4
*muB
18. thick/dense *makD ma8 ma6 - mɔu8
19. 霧 fog *mowC mou6 - - mu6
20. 廟 temple *miuC miu6 miu6 - miu6

Notes
1. Similar to the Austronesian 2PL root *-mu (ACD); the 2PL root was extended to the 2SG
in Austronesian. Western Hmongic has a different root: *gʉA ‘you’ (5.3/8).
2. The Hmongic forms except for Pa-Hng (7) are from Wang 1994: the Pa-Hng is from
Niederer 1997. The Mun form (9) is from Shintani and Yang 1990.
4. Possibly from Chinese 目 ‘eye’ (OC *[m][u]k > MC mjuwk > Man. mù).
42 Chapter 2

5. On the relationship of ‘to grasp’ and ‘to have’, see note 2 under 1.7 above.
5 & 6. A pre-initial *n- is tentatively reconstructed for these two words on the basis of
initial n- in Mun.
10. Hmongic cognates of this word may also be Ho Ne (She) /mɔ8/ and Shimen (West
Hmongic) /maɯ5/, so it is given here as Proto Hmong-Mien.
13. In White Hmong, ‘mother’.
19. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
Compare Cantonese /mou6/.
20. There is a Bunu cognate /mjɔ6/ ‘temple’, but this does not give enough evidence to
place this word in one of the major rime sets.

1.10 *ʔw-
PHM *ʔw- 1 2 3 4 5 6 7 8 9 10 11
1. son-in-law *ʔweiX - - vau3 wɔ3a ʔwuB ve3 vi3 vei3 - - vɛi3
2. 冠 crest/comb *ʔwi̯ æn - - ʔi1 - ʔweA - - - - - vɔn1

PH *ʔw- 1 2 3 4 5 6 7
3. cherry *ʔwaA va1 wɑ1 - ʑi1a - - -
4. winnowing fan *ʔwaŋA vaŋ1 wei1 va1 wua1a ʔwenA ven1 vɛ1
5. to cover self *ʔwoB - - vɔ3 wo3a ʔwuB - va3
6. wife/elder sister
vi3 - vi3 - - va3 vo3
*ʔwæwB

PM *ʔw- 8 9 10 11
7. that *ʔweB wa3 va3 wə3 vɛi3
8. 碗 bowl *ʔwənB van3 van3 wan3 vjɛn3
9. 抉 to dig *ʔwetD vet7 veːt7’ wun7 vɛt7

Notes
1. Changping Mien (Wang and Mao 18) has been substituted for Luoxiang Mien (8).
Compare Tibeto-Burman *krwəy ‘son-in-law’ (Matisoff 2003). This connection is
strengthened by the parallel ‘sister-in-law/daughter-in-law’ *ʔɲam (4.7/24) which closely
resembles Tibeto-Burman *nam ‘daughter-in-law’ (Benedict 1987a).
2. See also /vi1/ (West Hmongic Gaopo, Fengxiang) and /vai1/ (West Hmongic Bunu). This
is from Chinese 冠 ‘cap’ (OC *[k]ˤon > MC kwan > Man. guān, as in jī guān
‘cockscomb’).
6. In White Hmong (3), ‘elder sister’.

1.11 *hw-
1 2 3 4 5 6 7 8 9 10 11
1. 園 garden *hwunA hun1 vin1’ hun1 von1

Notes
1. The same word was borrowed separately into Hmongic as *waŋA (1.12/24).
 Proto Hmong-Mien 43

1.12 *w-
PHM *w- 1 2 3 4 5 6 7 8 9 10 11
1. sky/heaven *wɛŋ vɛ2 - - - - - vɔ2̃ - - - vaŋ2
2. 芋 taro *wouH vu6 wə6 vaɯ6 wɯ6 woC vau6 vɔ6 hou6 hɔu6 - vu6

PH *w- 1 2 3 4 5 6 7
3. 1SG (I/me) *wɛŋB vi4 we4 - - - va4 vɔ̃4
4. 園 garden *waŋA vaŋ2 - va2 wua2 - vi2 ɦje2
5. 萬 ten thousand *wi̯ aŋC vaŋ6 - va8 waŋ6 waŋC - -

PM *w- 8 9 10 11
6. 皇 yellow *wi̯ əŋA waŋ2 vaŋ2 waŋ2 vjaŋ2

Notes
1. This is part of a very irregular set with Hmong-Mien *ɴɢɛuŋ ‘sky/heaven’ (6.6/22) and
Hmong-Mien *ndɛuŋ ‘sky/heaven’ (2.6/22).
4. The same word was borrowed separately into Mienic as *hwunA (1.11/27).
6. Hmongic borrowed ‘yellow’ from Chinese 黃 (see 5-6’.3/24).

1.16 *pj-
PHM *pj- 1 2 3 4 5 6 7 8 9 10 11
1. three *pjɔu pi1 pu1 pe1 pæ1a pziA pa1 po1 pu1 pu1 pau1 bu5
2. bedbug *pji - - - - - pi1 pɪ5 pi1 pi1 pi1 bɛi1
3. father/male *pjaX pa3 pɑ3 tsi3 pi3a paB pe3 pa7 pwa3 fa3 - -
4. fruit *pji̯ əuX tsen3 pi3 tsi3 pei3a pzeB pi3 pje3 pjeu3 pjɔu3 pjau3 bɛu3

PH *pj- 1 2 3 4 5 6 7
5. crop of bird *pjinB pi3 - tsia5 pjein5a puB pai3 pẽ1
6. to bear fruit *pjiC tsen5 - tsi5 pei5a pzeC - -

PM *pj- 8 9 10 11
7. wooden basin *pjatD pwat7 - - -

Notes
1. In addition to Fuyuan (5), forms suggesting -j- are Shimen /tsi1/, Fengxiang /tsi1/.
2. Forms with -j- appear in Ho Ne /pji3/, Jiangdi Mien /pje1/, and Xiangnan Mien /pje1/.
3. Compare Chinese 父 ‘father’ (OC *[b](r)aʔ > MC bjuX > Man. fù). See also Hmong-
Mien ‘husband’ *N-poX (1.1/7).
4. See also West Hmongic Xuyong /tsɿ5/, Shimen /tsi5/ and Bunu /pi5/ (Office of Miao-Yao
Research Lexicon 1987). Compare the Hmongic derivative *pjiC ‘to bear fruit’ (1.16/1).
6. ‘To bear fruit’ is a C-tone derivative of Hmong-Mien *pji̯ əuX ‘fruit’ (1.16/1); see
chapter 4.
7. *pj- is reconstructed on the basis of Mien forms /pjet7/ and /pje7/ (Wang and Mao 15 and
16).
44 Chapter 2

1.17 *phj-
PH *phj- 1 2 3 4 5 6 7
1. burn (TR) *phjæwB phi3 - - - - v̥ a3 pho3
2. half *phjeC - - - pje7b phjeC - -
3. chaff/husk *S-phjæC fha5 sa5 sua5 sa5b siB v̥ ei5 (m̥ e5)

Notes
3. Compare Hmong-Mien *mphi̯ ɛk ‘chaff/husk’ (1.5/1) and see discussion of words that
pattern similarly under 1.2.

1.18 *bj-
PHM *bj- 1 2 3 4 5 6 7 8 9 10 11
1. to enter *bju̯ ɛk pə6 pɔ5 - - - pɔ6 pɯ6 pja8 - phja8 pjɛ8

1.19 *mpj-
PHM *mpj- 1 2 3 4 5 6 7 8 9 10 11
1. 補 to repair/mend
pu3 mpɑ3 ntsi3 mpi3a mʔpaB mpan3 ma3 bwa3 va3 bja3 bjɛ3
*mpjaX

1.20 *mphj-
PH *mphj- 1 2 3 4 5 6 7
1. daughter/girl
phi7 mphɑ7 ntshai7 mpje7b mʔphjeD phai7 phe7
*mphjeD

1.21 *mbj-
PHM *mbj- 1 2 3 4 5 6 7 8 9 10 11
1. 辮 to braid/a braid
mi4 - ntsia4 mpjein4 mpen4 pjen6 ma1 bin4/bin5 bin4 - pin4/bjɛn6
*mbjinX

PM *mbj- 8 9 10 11
2. young man *mbjauA bjaːu2 baːu2 - -

Notes
1. A number of languages (including Mien (8) and Zao Min (11) above) have differentiated
this word based on the verb/noun distinction (they appear in this order), or, more likely,
have borrowed the verb and noun separately from Chinese. In most locations, the same
word is used for both noun and verb, or only one or the other is attested.
2. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
 Proto Hmong-Mien 45

1.23 *hmj-
PHM *hmj- 1 2 3 4 5 6 7 8 9 10 11
1. footprint/track
m̥ ha3 - hn̥ ɛŋ3 mein3a m̥ aB - - m̥ wan3 ŋwaːn3’ m̥ jɛn3 -
*hmjænX
2. tooth *hmjinX m̥ hi3 ɕɛ3 hn̥ ia3 mi3b m̥ jenB m̥ ai3 mɪ3̃ - - mjen3 -
3. to sniff at *hmji̯ əmH m̥ hi5 - hn̥ ia5 mi5b m̥ jenC m̥ ai5 mi5 hom5 hɔːm5’ m̥ ja5 -

PM *hmj- 8 9 10 11
4. grass *hmjaB m̥ wa3 wa3 m̥ ja3 mjɛ3
5. spirit/ghost *hmjænB m̥ wan3 ŋwaːn3 m̥ jɛn3 mjɛn3

Notes
1. Compare Tibeto-Burman *naŋ ‘to follow’, *s-naŋ ‘to follow/with’ (Matisoff 2003),
Kachin /mənaŋ/ ‘companion’, Lai /neʔ-hnɑŋ/ ‘footprint’ (Kenneth Van Bik, p.c.).
2. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10).
3. Both ‘to sniff at’ and ‘to taste/try’ (Hmong-Mien *hmeiH, 1.8/12) mean to smell or taste
intentionally. Different words exist for ‘to smell’ and ‘to taste’ unintentionally. Compare
Tibeto-Burman *s-nam ‘to smell’ (‘directive, causative, intensive’ s- prefix; see Matisoff
2003:99 ff.).
5. The -w- in Luoxiang Mien (8) and Lanjin Mun (9) is secondary: -j- dropped, then -w-
developed between a labial and /a/.

1.24 *mj-
PH *mj- 1 2 3 4 5 6 7
1. horse *mjænB ma4 - nɛŋ4 mein4 maB me4 mɪ4̃

PM *mj- 8 9 10 11
2. person/Mien 2 2 2
mwan mun min min2
*mjænA
3. wife of mother’s
mwaŋ2 ŋwaŋ2 mjaŋ2 mjaŋ2
brother *mjəŋA

Notes
1. Although Hmongic ‘horse’ is ultimately related to Chinese 馬 (OC *mˤraʔ > MC mæX
> Man. mǎ), these forms suggest a more direct connection to some Tibeto-Burman
language (TB *mraŋ, Matisoff 2003). Furthermore, the rime matches Hmong-Mien
*hmjænX ‘footprint/track’ (1.23/19), which is probably also a Tibeto-Burman borrowing.
Mienic words for ‘horse’ are all borrowings from Chinese.
2. See also Mien dialects Jiangdi (Wang and Mao 15) /mjen2/ and Xiangnan (Wang and
Mao 16) /mjəŋ2/. Pa-Hng /mjɛ2/ patterns with Mienic here as it does in a number of other
instances (see ‘chaff/husk’ and ‘thunder’). The probable source of the Mienic autonym,
which takes the variant forms ‘Mien’, ‘Mun’, and ‘Min’, is Chinese 民 ‘people’ (OC
*mi[ŋ] > MC mjin > Man. mín). Wang and Mao link these Mienic forms in m- that mean
both ‘person’ and ‘group’ to the Hmongic *nænA ‘person’ (2.9/19) since the rimes and
tones correspond. The Hmongic word for ‘person’ is entered separately here for the time
being.
46 Chapter 2

2 & 3. The -w- in Luoxiang Mien (8) and Lanjin Mun (9) is secondary: -j- dropped, then
-w- developed between a labial and /a/.

1.27 *wj-

PH *wj- 1 2 3 4 5 6 7
1. cooking pot/pan 4 4 4 4 B 4
vi wɛ jia ʑein wen van vɪ4̃
*wjinB

Discussion
The only word in this set patterns like *j- in White Hmong (3) and Zongdi (4), and like
*w- elsewhere.

1.31 *pl-
PHM *pl- 1 2 3 4 5 6 7 8 9 10 11
1. four *plei l̥ u1 pʐei1 plau1 plɔ1a plouA ple1 pi1 pje1 pjei1 pləi1 pɛi5

PH *pl- 1 2 3 4 5 6 7
2. stomach *plaŋA - - pla1 - - - -
3. flour *plu̯ eiB - - - plɔ5a plouB - -
4. owl *ploŋD l̥ ioŋ5 - pla7 - - - -

PM *pl- 8 9 10 11
5. to bury *plopD plop7 plɔp7 - -

Notes
1. From Tibeto-Burman *b-ləy (Matisoff 2003). See chapter 5.
3. Compare Proto Malayo-Polynesian *palu ‘sago flour’ (ACD).
4. Yanghao (1) from Office of Miao-Yao Research Lexicon (1987).
5. Hainan Mun (Shintani and Yang 1990) has been substituted for Lanjin Mun (9).

1.31.1 *plj-
PHM *plj- 1 2 3 4 5 6 7 8 9 10 11
1. hair *pljei ȴ̥u1 pi1 plau1 pɭɔ1a plouA ple1 pi1 pje1 pjei1 pli1 pɛi1

PH *plj- 1 2 3 4 5 6 7
2. heart *pljowB ȴ̥u3 - plaɯ3 pɭə3b ploB - -
3. wildcat *pljɨD ȴ̥aŋ7 - pli7 pɭei5a pleC - -

Notes
3. Perhaps the same as Chinese 狸 ‘wildcat’ (OC *[m]ə.rə > MC li > Man. lí) from a
variant bù lái indicating an initial *pə- (Sagart 1999:88). In Hmong-Mien the word would
have had a final -p or -t, so this may be a loan from Hmong-Mien to Chinese.
 Proto Hmong-Mien 47

1.32 *phl-
PHM *phl- 1 2 3 4 5 6 7 8 9 10 11
1. shell/covering 7 1 C
l̥ hu - phlau - phlau - - - - - fɔu1
*phlei

PH *phl- 1 2 3 4 5 6 7
2. cheek/face *phl-C - phɑ5 phlu5 - - - -
3. to stroke *phlʉC l̥ hɛ5 phʐə5 phlɯ7 plou5b phluC - -

PM *phl- 8 9 10 11
4. to stroke *phlunA phjun1 - phlən1 -
5. to play (flute)
pwəm3 pjɔm3 - -
*phlu̯ əmB

Notes
1. Yanghao (1) from Office of Miao-Yao Research Lexicon (1987). Most other West
Hmongic cognates have tone 1. In both Yanghao (1) and Fuyuan (5), the meaning has
narrowed to ‘soft membrane inside shell’. Compare Proto Tai *pl-k ‘shell, bark’ (Li 1977).
3 & 4. Although the rimes and tones do not correspond, these are probably the same word.
Hmongic forms from Wang 1994 except White Hmong (3) from Heimbach 1979; Mien (8)
from Office of Miao-Yao Research Lexicon (1987), and Biao Min (10) from Solnit 1982.
Compare Chinese 撫 ‘to caress’ (OC *[pʰ](r)aʔ > MC phjuX > Man. fǔ).
5. Changping Mien (Wang and Mao 18) /phləm3/. Compare the similar Hmongic *phroC
‘to play (flute)’ (1.47/7).

1.33 *bl-
PHM *bl- 1 2 3 4 5 6 7 8 9 10 11
1. forehead *bl-A - - plia2 - - - - pɔŋ2 plɔŋ2 plə4 paŋ2

PH *bl- 1 2 3 4 5 6 7
2. mussel/clam *blinC l̥ iə6 pɹɛ4 plia6 - - - -

Notes
1. Jinxiu Mien (8) from L-Thongkum 1993. Biao Min form (10) from Office of Miao-Yao
Research Lexicon (1987). Compare Proto Tai *ph(r/l)aak (Li 1977).
2. Yanghao (1) and Jiwei (2) from Office of Miao-Yao Research Lexicon (1987).

1.33.1 *blj-
PH *blj- 1 2 3 4 5 6 7
1. soul/spirit *bljaA ȴ̥u2 pjə2 pli6 pɭɪ6 vloA - pjɔ2

Notes
1. Compare Tibeto-Burman *b-la ‘demon, soul’ > Nung /phəla/ (Benedict 1972). Tone 6
forms can be explained as sandhi form promotions, but the rime is highly irregular.
48 Chapter 2

1.34 *mpl-
PHM *mpl- 1 2 3 4 5 6 7 8 9 10 11
1. butterfly *mpleuH - - - mpei5a mʔpleC - - bjeu5’ blɔu5’ bja4 -

PM *ʔbl- (<*mpl- ) 8 9 10 11
2. blind *ʔblɔuB bu3 bu4 - -

Notes
2. The medial *-l- is reconstructed on the basis of Hainan Mun /blou4/ (Shintani and Yang
1990). This could also be *mbl- since both upper and lower register tones are attested.

1.34.1 *mplj-
PH *mplj- 1 2 3 4 5 6 7
1. twin *mpljeA pi1 - ntsai1 mpje1a mʔpleA - -

1.35 *mphl-
PH *mphl- 1 2 3 4 5 6 7
1. finger ring *mphleA - - mphlai1 - - mʔphleA -

1.36 *mbl-
PHM *mbl- 1 2 3 4 5 6 7 8 9 10 11
1. rice plant/paddy
na2 nɯ2 mple2 mplæ2 - mpla2 mjo2 blau2 blau2 blau2 bjau2
*mbləu
2. smooth/glossy
- mjɛ6 mpla6 mplein6 mplenC - wɪ6̃ - bjaŋ6 - -
*mbli̯ əŋH
3. glutinous /sticky
nə8 nu8 mplau8 mplu8 mploD - - but8 blut8 blun8 bit8
*mblut
4. tongue *mblet ȵi8 mjɑ8 mplai6 mple8 mpleD mpli8 mje8 bjet8 bjet8 blin8 bɛt8

PH *mbl- 1 2 3 4 5 6 7
5. leaf *mblɔŋA nə2 nu2 mplɔŋ2 mplaŋ2 mploŋA mplɔŋ2 mjɔ̃2
6. whip *mblowD - - mpləɯ8 mplə8 mʔploD - -

PM *mbl- 8 9 10 11
7. to shed leaves/drop
blwei6 blei6 - -
*mblu̯ eiC
8. rain *mbluŋC bluŋ6 buŋ6 blə6 biŋ6

Notes
2. In White Hmong, ‘brilliant’.
3. In Hmongic, this is used for sticky rice, and in Mienic it means both ‘glutinous’ and
‘snot’. Compare Proto Malayo-Polynesian *belit/*bulit ‘viscous, sticky’ (ACD) and
Chinese 秫 ‘glutinous millet’ (OC *m.lut > MC zywit > Man. shú) (Sagart 1999:79).
4. Probably from Chinese 舌 ‘tongue’ (OC *m.lat > MC zyet > Man. shé (Sagart 1999:79);
see parallel with ‘glutinous’ above.
 Proto Hmong-Mien 49

5. Compare Malayo-Polynesian *buluŋ ‘medicinal herbs’ which has come to be used as

‘leaf’ in many Austronesian languages (ACD). The Hmongic word for ‘leaf’ may be
related to Mienic *nɔmA (2.9/29), given the match in rime and tone.
8. Compare Khmer /phliəŋ/ ‘rain, to rain’. Possibly related to Hmongic *m-noŋC ‘rain’
(2.9/28), since they belong to the same rime set.

1.36.1 *mblj-
PHM *mblj- 1 2 3 4 5 6 7 8 9 10 11
1. fan *mbljæp zen8 mjɑ7 ntsua8 mpja8 mpziD mpli8 mi8 bjap8 bjaːp8 bjɛn8 dzjɛp8

PH *mblj- 1 2 3 4 5 6 7
2. crazy *mbljɔŋA - - - mpjaŋ2 mpzoŋA - -
3. to have food w/rice
- - ntsua6 mpje6 mpjiC - -
*mbljæC

Notes
2. A medial -l- appears in Gaopo /mploŋ2/ (West Hmongic, Wang and Mao 6).
3. A medial -l- appears in Gaopo /mplɑ6/ (West Hmongic, Wang and Mao 6).

1.39 *ml-
PHM *ml- 1 2 3 4 5 6 7 8 9 10 11
1. soft *mlu̯ ɛjH mɛ6 ne6 mua6 noŋ6 mjaC mɔ6 mɪ6̃ mwei6 ŋwei6 mi6 mai6

Notes
1. Medial *-l- reconstructed on the basis of Gaopo /mləŋ6/ (West Hmongic, Wang and Mao
6). Compare Proto Malayo-Polynesian *ma-lumu ‘soft, tender, gentle’ (ACD).

1.46 *pr-
PHM *pr- 1 2 3 4 5 6 7 8 9 10 11
1. five *pra tsa1 pʐɑ1 tʂi1 pʐɪ1a pjaA pui1 pja1 pla1 pja1 pla1 pjɛ5
2. house *prəuX tsɛ3 pʐɯ3 tʂe3 pʐæ3a peiB pja3 pjo3 pjau3 pjau3 pla3 pju3
3. stick *praX - pʐɑ3 - - - - pja3 pla3 pja3 - -

PH *pr- 1 2 3 4 5 6 7
4. to sow *præwA - pʐo1 tʂe1 pʐæ1a pjiA - -
5. to winnow *prɔŋB - pʐu3 tʂɔŋ3 pʐaŋ3b pjoŋB pjɔŋ3 -
6. to pour *praC tsɑ5 pʐa5 tʂi5 - - - -
7. to wring *proC - pʐo5 - pʐo5a pjuC - pja5
8. dark *pruwD tsə7 pʐu7 tʂau7 pʐu7a pjoD pjɔu7 -

Notes
1. From Tibeto-Burman *b-r-ŋa (Benedict 1987a). See chapter 5.
6. Forms taken from Office of Miao-Yao Research Lexicon (1987).
50 Chapter 2

1.47 *phr-
PHM *phr- 1 2 3 4 5 6 7 8 9 10 11
1. head *S-phreiX fhu3 pʐei3 hau3 wɔ5b huB - pɦi3’ pje3 pjei3 pli3 pɛi3

PH *phr- 1 2 3 4 5 6 7
2. to blow *phruA tsho1 phʐo1 tʂhua1 pʐa1b phjaA phu1 phjo1
3. play (flute) *phroB - - tʂhɔ3 pʐo3b - - -
4. tree *S-phruA fhu3 - - wa1b - - -

Notes
1. Extended semantically in White Hmong to ‘base, root, origin’ as well as ‘summit’ (tone
3) and ‘leader, headman’ (tone 5). Mienic and Hmongic words have the same rime and
tone, but slightly different initials, presumably the effect of the prefix in Hmongic.
2 & 3. These are clearly similar in form and meaning, but their rimes and tones differ.
3. Compare the similar Mienic *phlu̯ əmB ‘to play (flute)’ (1.32/29).
4. This word is primarily found in West Hmongic and patterns like ‘head’ above. The
Yanghao (1) form is a classifier for vegetation (Office of Miao-Yao Research Lexicon
1987), as is the Shimen (West Hmongic) word /faɯ1/.

1.48 *br-
PH *br- 1 2 3 4 5 6 7
1. noose/trap *briB - - tʂi4 - - - -
2. rat *bru̯ aC - - tʂua6 pʐəa6 vjaB - -

Discussion
Both of these words have a medial -l- in Gaopo (West Hmongic, Wang and Mao 6): /pli4/
‘noose/trap’ and /plu6/ ‘rat’.

Notes
1. Xianjin Hmong (Wang and Mao 3) has been substituted for White Hmong (3). Other
forms appear in Wang 1994.

1.49 *mpr-
PH *mpr- 1 2 3 4 5 6 7
1. green/blue *mpruA zo2 mʐo1 ɳtʂua1 mpʐa1a mʔpjaA - mjo1
2. lung *mprʉC zɛ6 mʐə5 ɳtʂɯ7 mpʐou5a mʔpjuC - mɔ5

1.51 *mbr-
PHM *mbr- 1 2 3 4 5 6 7 8 9 10 11
1. ear *mbræu zɛ2 mʐɯ2 ɳtʂe2 mpʐæ2 mpjiA mpja2 mjo2 - - blau2 bju2
2. fish *mbrəuX zɛ4 mʐɯ4 ɳtʂe4 mpʐe4 mpjiB mpja4 mjo4 bjau4 bjau4 bla4 bju4
3. bamboo shoot/
za6 mʐɑ6 ɳtʂua6 mpʐəa6 mpjaC mpja6 mje6 bje6 bjai6 blai6 bɛi6
orphan *mbrəjH
4. nose *mbruiH zɛ6 mʐə6 ɳtʂɯ6 mpʐu6 mpjuC mpjau6 mjɔ6 pa6 - bli6 -
5. spicy *mbrat za8 mʐei8 ɳtʂi8 mpʐɪ8 mpjaD mpwai8 mje8 blaːt8 bjat8 blan8 bjɛt8
 Proto Hmong-Mien 51

PH *mbr- 1 2 3 4 5 6 7
6. to groan/
zaŋ2 mʐei2 ɳtʂa2 mpʐua2 mpenA mpwen2 -
moan *mbraŋA
7. ant *mbrɔD - - ɳtʂau 8
mpʐɔ 8
mpju D
- -

Notes
2. Compare Proto Tai *pla A1 (Li 1977), Proto Kam-Sui *mprai 3 (Thurgood 1988).
3. This root also means ‘orphan, widow(er)’ in all Hmongic languages except Jiongnai, Ho
Ne and Pa-Hng. It may be that ‘bamboo shoot’ is basic, and ‘orphan, widow(er)’ is a
derived meaning since the distribution of ‘bamboo shoot’ is wider. However, given Proto
Tai *br[a] ‘orphan’ (Li 1977), perhaps ‘orphan/widow(er)’ is the basic meaning.
4. Compare Chinese 鼻 ‘nose’ (OC *m-[b]i[t]-s > MC bjijH > Man. bí).
7. Compare Mon-Khmer *srmuuc (Shorto #873), Proto Tai *mu̯ it D2S (Li 1977), Malay
/semut/.

1.54 *mr-
PHM *mr- 1 2 3 4 5 6 7 8 9 10 11
1. to listen
- - mlɔŋ6 mʐəŋ6 - - mæ̃ 5 moŋ5 muŋ5’ m̥ ɔŋ5 maŋ5
*(S-)mru̯ ɔŋH

Notes
1. Biao Min (10) from Office of Miao-Yao Research Lexicon (1987). Mienic and Pa Hng
tone 5, the Mun tone subcategory, and the Biao Min voiceless nasal suggest a voiceless
prefix that devoiced the initial in some languages.
52 Chapter 2

2. Dentals

2.1 *t-
PHM *t- 1 2 3 4 5 6 7 8 9 10 11
1. to answer *tau ta1 tɯ1 te1 tæ1a tiA ta1 no1 - tau1 - du1
2. son/boy *tu̯ ɛn tɛ1 te1 tu1 toŋ1a taŋA - tɔ̃1 - tɔːn1 twan1 dan1
3. body louse *tɛmX tɛ3 te3 tu3 toŋ3a taŋB taŋ3 nɔ̃3 tam3 tam3 dan3 dam3
4. rib *tamX taŋ3 - ta3 tua3a tenB - - - taːm3 - -
5. tail *tu̯ eiX tɛ3 tə3 tɯ5 tou3a tuB ðau3 tɛ3 twei3 tei3 dwai3 dui3
6. to kill *təjH ta5 tɑ5 tua5 ta5a - ta5 ta5 tai5' tai5 tai5 dai5
7. to snap *tɛŋH tɛ5 te5 tu5 toŋ5a taŋC taŋ5 tɔ̃5 - - taŋ5 daŋ5
8. earth *N-təu ta1 tɯ1 te1 tæ1a tiA ta1 to1 dau1 - - -
9. deep *N-tu̯ o to1 to1 tɔ1 to1a tuA ðu1 to1 do1 du1 du1 -
10. skin *N-top tu3 tə3 taɯ3 tə3a toB - - dup7 dup7 din7 dip7
11. wing *N-tat ta7 tei7 ti7 ti7a taD ðe7 te7 daːt7 daːt7 - dɔt7
12. 擣 to husk/pound rice
to3 - tua3 ta3a taB tu3 to3 - tau3 tau3 -
*tuX
13. 得 to get/gain *təuk tɛ5 tɔ5 tau5 tɔ5a tuC tei5 tɯ5 tu7 tu7 tu7 -

PH *t- 1 2 3 4 5 6 7
14. thick *tæA ta1 ta1 tua1 ta1a tiA tei1 te1
15. gourd *tuwA - tɔ1 tau1 tu1a toA - -
16. skirt *tinA tei1 tɛ1 tia1 tæin1a tenA ðai1 tĩ1
17. to hold/grasp hand
- - tua3 ta3a tiB - -
*tæB
18. frost *tæwC ta5 tɯ5 te5 tæ5a tiC ða5 no5
19. axe *tu̯ eiC to5 tɔ5 tau7 tɔ5a tuC ðei5 ti5
20. foot *towC tu5 tə5 taɯ5 tə5a toC - -
21. to ignite/light *towD tu7 - taɯ7 te7a toD - -
22. to pick up/clamp *teD - tɑ7 tai7 te7a teD - næ̃ 6

PM *t- 8 9 10 11
23. 酒 liquor *tiuB tiu3 tiu3 tiu3 diu3
24. 碓 tilt-hammer/pestle
tɔi5’ tɔːi5 twai5 dui5
*tu̯ ɔiC
25. 凳 stool/bench *təŋC taŋ5’ taŋ5 taŋ5 daŋ5

Discussion
‘Earth’, ‘deep’, ‘skin’, and ‘wing’ show voiced stops in Mien and Mun. Since the source of
voiced stops in Mien and Mun is usually prenasalization in Hmongic, I have proposed a
“loosely-adjoined” voicing element for these words (*N-), one which did not survive in
Hmongic (these pre-initials have been reconstructed for words with other initials as well:
see 2.3, 3.1, 3.6, 4.1, etc.). These four words also consistently show a voiced fricative
initial in Jiongnai (6), a conservative Hmongic language. On the basis of the presence of
Jiongnai voiced fricative initials in ‘skirt’, ‘frost’, and ‘axe’, it seems reasonable to think
that there may have been a loosely adjoined voicing element in in these words as well,
although they are not reconstructed here. Fieldwork on the West Hmongic language Pa-na
 Proto Hmong-Mien 53

language reported in Tagachi 2005 suggests that a medial -l- may account for the voiced
initials in Mienic; however, there is no evidence that a medial -l- can account for words
with this pattern at other places of articulation.

Notes
6. Yanghao (1) from Office of Miao-Yao Research Lexicon (1987). This word is important
in the reconstruction of morphology in Hmong-Mien, as it bears a morphological
relationship to ‘to die’ (*dəjH, 2.3/15), see chapter 4. These two words are also to be
related in some fashion to proto-Austronesian *pa-aCay ‘to kill’ and *ma-aCay ‘to die’
(ACD).
8. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
23. The Ho Ne cognate /tiu3/ ‘liquor’ alone does not give enough information to allow us
to place this word in one of the major rime sets. See Hmongic *cowB (4.1/13), borrowed
independently from the same word.

2.2 *th-
PHM *th- 1 2 3 4 5 6 7 8 9 10 11
1. 桶 bucket *thɔŋ(X) - - thɔŋ1 - thoŋA - thɔ̃7 thoŋ3 tɔŋ3 dɔŋ3 tuŋ5
2. 炭 charcoal *thanH thɛ5 the5 thɛŋ5 tæin5b thaŋC thɔu3 thæ̃ 5 thaːn5 taːn5’ than5 tan3

PH *th- 1 2 3 4 5 6 7
3. to boil (TR) *thoA tho1 - - to1b thuA - -
4. bellows *thoŋB thoŋ3 thɑŋ3 - thaŋ3b thoŋB - tho3
5. hoop *thɨC tha5 - thi5 tei5b theC - -

Notes
4. This word bears a slight resemblance to Chinese 橐 ‘sack, bellows’ (MC thak > Man.
tuó).
5. Possibly from Chinese 帶 ‘belt, girdle’ (MC tajH > Man. dài), although the Chinese
initial is not aspirated.

2.3 *d-
PHM *d- 1 2 3 4 5 6 7 8 9 10 11
1. elder brother *da ta2 - ti2 ti2 ðaA - - - ta2 - -
2. to come *daj ta2 - tua2 ta2 ðaA - te2 taːi2 taːi 2 ta2 tɛi2
3. fire *douX tu4 tə4 taɯ4 tɯ4 ðoB tau4 tɤ4 tou4 tɔu4 təu4 tu4
4. to explode *douH tu6 tə6 taɯ6 tɯ6 ðoC - tɤ6 - tɔ6 - -
5. to die *dəjH ta6 tɑ6 tua6 təa6 ðaC ta6 te6 tai6 tai6 tai6 tai6
6. tortoise *duH - - - - - - - to6 tu6 - -
7. to bite *dəp tə8 to8 tɔ8 to8 ðuD - ta8 tap8 tap8 than8 -
8. to put on/wear
- - - - - tei8 - taːp8 taːp8 - -
(shoes) *dap
9. half (of day) *N-
taŋ4 - ta4 - - - - daːm2 daːm2 - -
dam(X)
10. 蹄 hoof *dej - - - - - tei2 - tei2 tei2 təi2 tɛi2
11. 銅 copper *dɔŋ tə2 toŋ2 tɔŋ2 taŋ2 ðoŋA twaŋ2 tɔ̃2 toŋ2 tɔŋ2 tɔŋ2 tuŋ2
12. 荅 bean *dup tə8 - tau8 tu8 noD tɔ8 tɯ8 top8 tɔp8 thən8 tup8
54 Chapter 2

PH *d- 1 2 3 4 5 6 7
13. level *dinA - - tia2 tæin2 - - -
14. to sink *dəŋA taŋ2 - tɔ6 toŋ6 toŋA - tæ̃ 4
15. bowl *deB - - tai4 te4 - - ta4
16. CLF-horses *dɛŋB tɛ4 - tu4 toŋ4 ðaŋB - -
17. to wait *dəŋB taŋ4 tɑŋ4 tɔ4 toŋ4 ðoŋB - te4
18. which *dʉC tei6 - tɯ6 tu6 ðuB - ti1
19. 蹋 to step on/tread
ten8 tɑ8 tua8 ti8 - - -
*dæD

PM *d- 8 9 10 11
20. to soak *doA tu2 - ta2 tɔu2
21. to put on/wear
taŋ2 taŋ2 - -
(jewelry) *dəŋA
22. back (of body)
tan4 tan4 - -
*danB
23. pig *duŋB tuŋ4 tuŋ4 twə4 tiŋ4
24. 頹 to fall (of rain)
tui2 - tui2 -
*dujA
25. 袋 bag *diC ti6 ti6 ti6 -
26. 地 non-irrigated
tei6 tei6 təi6 ti6
land *dejC

Notes
1. Compare Chinese 弟 dì ‘younger brother’, used in the compound 弟兄 dìxiong to mean
‘brothers’.
2. In White Hmong, ‘to come back to a place other than one’s home’. Compare Hmong-
Mien *ləwX (2.42/7): ‘to come back to one’s home’.
3. Also ‘firewood’ in White Hmong.
3 & 4. ‘Fire’ and ‘to explode’ are morphologically related; see chapter 4.
5. This word is important in the reconstruction of morphology in Hmong-Mien, as it bears
a morphological relationship to ‘to kill’ (*təjH, 2.1/15); see chapter 4. These two words are
also to be related in some fashion to proto-Austronesian *ma-aCay ‘die’ and *pa-aCay
‘kill’ (ACD).
6. Reconstructed to Proto Hmong-Mien on the basis of its appearance in Mienic and
Qibainong Bunu /tu6/ (Wang and Mao 10).
7. Compare Malayo-Polynesian *ketep ‘to bite’ (ACD).
8. Compare Chinese 踏 ‘to step into or on to’ (MC thop > Man. tà).
19. In White Hmong, ‘to kick backward with foot (as of a horse)’.
20. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
23. Perhaps from Chinese 豚 ‘young pig’ (OC *[d]ˤu[n] > MC dwon > Man. tún), although
the tones and final nasals do not correspond.
 Proto Hmong-Mien 55

2.4 *nt-
PHM *nt- 1 2 3 4 5 6 7 8 9 10 11
1. fragrant *ntaŋ - - - - - nten1 nɛ1 daːŋ1 daːŋ1 daŋ1 dɔŋ1
2. wet *ntu̯ ɛn - nte1 ntu1 ntoŋ1a nʔtaŋA ntaŋ1 tɔ4̃ - dɔːn1 - -
3. cloth *ntei - ntei1 ntau1 ntɔ1a nʔtuA nte1 ni1 di1 - di1 dɛi1
4. long *ntauX ta3 ntɯ3 nte3 ntæ3a nʔtiB ða3 to3 daːu3 daːu3 da3 du3
5. to chop *ntəwX to3 - ntɔ3 nto3a nʔtuB ntu3 no3 - - dau3 dau3
6. to flip with finger
- - nti3 ntei5a - ntai5 na5 dit7 - - -
*ntik
7. to warm self by fire
ta5 ntɯ5 nte5 ntæ5a nʔtiC nta5 no5 - daːu5 dau5 du5
*ntauH
8. to put on/wear (cap)
tə5 ntu5 ntɔŋ5 ntaŋ5a nʔtoŋC ntwaŋ5 no5 doŋ5’ dɔŋ5 dɔŋ5 dɔŋ5
*ntɔŋH
9. to weave *ntət to7 nto7 ntɔ7 - - ntu7 na7 dat7 dat7 dan7 dat7
10. navel *ntəut tu7 ntu7 ntaɯ7 ntə7a nʔtoD ntu7 - dut7 duːt7’ din7 -
11. 擔 to carry on
- - - - - tã1 tæ̃ 5 daːm1 daːm1 da1 dɔm1
shoulder *ntam

PH *nt- 1 2 3 4 5 6 7
12. to float *ntaŋA - nten1 nta1 - - - -
13. finger *ntaB ta3 ntɑ3 nti3 nti3a nʔtaB - -
14. paper *ntowB tu3 ntə3 ntaɯ3 ntə3a nʔtoB ntau3 nɔ3
15. many/much *ntɔC nɛ5 - ntau5 ntɔ5a nʔtuC ntei5 -
16. to beat (someone)
- - ntau7 ntu7a - - ta7
*ntuwD

PM *ʔd- (< *nt- ) 8 9 10 11

17. deaf *ʔduŋA - duŋ1 dwə1 dɔŋ1
18. to ladle *ʔdəmB dam3 - - da3
19. to fly *ʔdəiC dai5 dai5’ dai5 -
20. finger *ʔdokD do7 dɔ7 dɔ7 -
21. 底 lower level
di3 di3 ti3 di3
*ʔdiB
22. 脫 to peel off/
- dut7’ dun7 -
escape *ʔdutD

Notes
1. Compare Kam /taaŋ1/ and Mak /taaŋ1/ ‘fragrant’ (Downer 1978).
6. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
8. Compare ‘tree’ (2.19/29), which is exactly the same in Hmongic (Hmongic *ntɔŋC), but
differs in onset at the higher level (Hmong-Mien *nt- ‘to put on/wear (cap)’ vs. Hmong-
Mien *ntj- ‘tree’).
13 & 20. These two sets of forms for ‘finger’ may be related, but they differ in rime and
tone, and thus are listed separately.
15. Compare Mun /duːŋ5/ ‘many/much’.
18. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). Possibly
from Chinese 斟 ‘to ladle out’ (OC *t.[q][ə]m > MC tsyim > Man. zhēn).
56 Chapter 2

2.5 *nth-
PH *nth- 1 2 3 4 5 6 7
1. attic *nthaŋA - - ntha1 ntua1b nʔthenA - -
2. to open out/undo
tha3 nthɑ3 nthua3 - nʔthaB - -
*nthu̯ aB
3. to weed/hoe *nthuC - ntho5 nthua5 - - - -

2.6 *nd-
PHM *nd- 1 2 3 4 5 6 7 8 9 10 11
1. sky/heaven *ndɛuŋ - - ntu2 ntoŋ2 - - - - - - -
2. sweet potato
na2 - - - - - - du2 dɔːi2 dwai2 dai2
*ndu̯ əj
3. to lay eggs *ndəuH na6 - nte6 nte6 ntiC tei5 no6 - - dau6 du6
4. ramie/hemp *nduH no6 no6 ntua6 ntəa6 ntaC ntu6 mjo6 do6 du6 du6 -

PH *nd- 1 2 3 4 5 6 7
5. long sword *ndaŋA - - nta2 ntua2 ntenA - -

PM *nd- 8 9 10 11
6. river *ndaiA - - da2 dɔi2
7. table *ndɔuA du2 nu2 - -
8. in front *ndaŋC daːŋ6 daːŋ6 - -

Notes
1. Part of a very irregular set with Hmong-Mien *wɛŋ ‘sky/heaven’ (1.12/22) and Hmong-
Mien *ɴɢɛuŋ ‘sky/heaven’ (6.6/22). Representing Mienic, see Changping Mien /ðuŋ2/
(Wang and Mao 18).
4. The labial initial in Pa-Hng (7) suggests an ancient prefix or disyllabic form.

2.7 *ʔn-
PHM *ʔn- 1 2 3 4 5 6 7 8 9 10 11
1. snake *ʔnaŋ naŋ1 nei1 na1 na1a ʔnenA nen1 nẽ1 naːŋ1 naːŋ1 naŋ1 nɔŋ1
2. this *ʔneinX noŋ3 nei3 nɔ5 na3a ʔnoŋB ne3 n̩ 3 nei3 nei3 na3 na3
3. cold *ʔnu̯ ɔmH - noŋ5 nɔ5 nɔ5a - nɐŋ5 ȵɔ̃5 nam5’ nam5 nan5 nam5

PM *ʔn- 8 9 10 11
4. CLF-bowls/houses
nɔm1 nɔ1 nɔ1 na1
*ʔnɛɔmA
5. short *ʔnəŋB naŋ3 niŋ3 naŋ3 naŋ3
6. to swallow *ʔnəkD na7 na7 na7 -
7. 拗 to break *ʔnəuB nau3 nau3 - -

Notes
3. In Mienic, the meaning is ‘cool (water)’.
4. Compare Hmongic *ʔlɛŋA ‘CLF-bowls/houses’ (2.40/22) and see discussion in chapter 6,
section 6.2.1.
 Proto Hmong-Mien 57

5. Compare Hmongic *ʔlɛŋB ‘short’ (2.40/22) and see discussion in chapter 6, section
6.2.1.
7. Compare Hmongic *ʔloB from the same Chinese word under (2.40/7) and see discussion
in chapter 6, section 6.2.1.

2.8 *hn-
PHM *hn- 1 2 3 4 5 6 7 8 9 10 11
1. grain head/bag *hnɔn n̥ haŋ1 n̥ hei1 hn̥ a1 na1b n̥ enA n̥ en1’ n̥ ẽ1 - nɔ1 nan1 -
2. to hear *hnəumX n̥ haŋ3 n̥ hɑŋ3 hn̥ ɔ3 nɔ3b n̥ uB n̥ ɐŋ3 nɔ̥ 3̃ nom3 num3 n̥ ən3 -
3. to put on/wear
naŋ4 n̥ hei3 hn̥ a3 na3b - nen4 nẽ3 - - - nɔŋ3
(clothes) *(h)naŋX
4. to cough *hnɔp ŋo4 - hn̥ ɔŋ7 naŋ7 n̥ oD - - n̥ op7 - n̥ ən7 -

PH *hn- 1 2 3 4 5 6 7
5. sun/day *hnɛŋA n̥ hɛ1 n̥ he1 hn̥ u1 noŋ1b n̥ aA n̥ ɔ1’ n̥ ẽ1
6. crossbow *hnænB n̥ hen3 - hn̥ ɛŋ3 nein3b n̥ aB - n̥ ẽ3
7. to forget *hnuŋB n̥ hoŋ1 noŋ3 hn̥ ɔ3 na3a n̥ oŋA n̥ ɐŋ3 nõ3
8. perilla (sū má)
n̥ haŋ3 n̥ hen3 hn̥ a3 - n̥ enB - -
*hnaŋB

PM *hn- 8 9 10 11
9. sun/day *hnu̯ ɔiA n̥ ɔi1 nɔːi1’ n̥ wai1 nai1
10. to resemble *hnəŋB n̥ aŋ2 naŋ3’ - -
11. to lift *hnɨŋC niŋ5 niŋ5’ - -
12. crossbow *hnəkD n̥ ak5 na7’ - -
13. 泥 mud *hniA ni1 ni1’ n̥ i1 nɛi1
14. to think of
- lam3 - -
*hnəmB~*hləmB

Discussion
The reconstructions for two words in this set—‘sun/day’ and ‘crossbow’—are entered
separately for Hmongic and Mienic, even though these forms undoubtedly go back to
single words in Proto Hmong-Mien. Their rimes are difficult to reconcile due to secondary
nasalization in Hmongic. The oral rime/nasal rime pair for ‘to cough’ is easier to
reconstruct at the Hmong-Mien level given the appearance of both in rime set 29. ‘To
forget’ behaves much the same way, but since its Mienic counterpart has a palatal initial
(*hɲouB), it is entered under 4.8 below.
The apostrophe after the tone number in Mun (9) marks a tone subcategory conditioned by
aspiration in the initial.

Notes
1. Chinese 囊 ‘bag’ (OC *nˤaŋ > MC nang > Man. náng) may have been borrowed from
Hmong-Mien, since it shows only the derivative meaning (see chapter 7). Yet compare
Tibeto-Burman *s-nam ‘ear of grain’ (Matisoff 2003).
3. Since tones 3 and 4 are both attested, the initial h- is placed in parentheses. Chang
(1972:563) thought this tonal alternation could be traced back to an initial voicing contrast
that might in turn have been caused by an old prefix (see chapter 4).
4. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
8. Perilla is a medicinal herb in the mint family (also known as “shiso”).
58 Chapter 2

10. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
11. Despite the tone mismatch, this is probably the same word as Chinese 拎 ‘to carry, lift’
(Cant. /nɪŋ1/, Man. līng).
12. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
14. This is a borrowing from Cantonese /nɐm3/ (~lɐm3/) (Downer 1973:19). The variation
between n- and l- in Cantonese is reflected in Mienic (Jiangdi Mien /n̥ am3/, Luoxiang Mien
/lam3/, Changping Mien /lam3/).

2.9 *n-
PHM *n- 1 2 3 4 5 6 7 8 9 10 11
1. 3SG (he/she/it)
nen2 - nɯ4 ȵi4 nenB nɐŋ4 nɯ4 nan2 nen2 nin2 -
*ni̯ æn(X)
2. mother’s brother
nen4 ne4 nu4 noŋ4 ȵiB nɔ4 nɯ4 nau4 nau4 nau4 nuŋ4
*neuŋX
3. bird *m-nɔk nə6 nu6 nɔŋ6 nəŋ6 noC nwaŋ6 mõ6 nu8 nɔ6 n̥ ɔ8 nɔu8

PH *n- 1 2 3 4 5 6 7
4. to eat *nuŋA naŋ2 noŋ2 nɔ2 nɔ2 noŋA nɐŋ2 nɔ2
5. person *nænA nɛ2 ne2 nɛŋ2 - naA nai2 -
6. mouse/rat *naŋB naŋ4 nei4 na4 - - nen4 nĩ4
7. to ask *nɛŋC nɛ6 ne6 nu6 noŋ6 naC nɔ6 nɪ6̃
8. rain *m-noŋC noŋ6 noŋ6 na6 naŋ6 noŋC nɐŋ6 mõ6

PM *n- 8 9 10 11
9. leaf *nɔmA nɔm2 nɔːm2 nan2 num2
10. mouse/rat *nauB naːu4 - nɔ4 -
11. to ask *nu̯ aiC naːi6 naːi6 na6 nɔi6
12. 膿 pus *nɔŋC noŋ6 nuŋ6 nɔŋ6 nɔu6

Notes
3. The final -k of ‘bird’ is reflected in a velar nasal in Hmongic (the shift to a final nasal
conditioned by the nasal initial). Since the Hmong-Mien rimes are organized by the large
sets of merged rimes in Hmongic, this word appears under set 29, a nasal rime set.
Compare Proto Malayo-Polynesian *manuk ‘bird/chicken’ (ACD) and Proto Tai *n-lok
‘bird’ (Li 1977).
4. Mienic *ɲənC (4.9/27) is probably related to these forms for ‘to eat’.
5. Wang and Mao link these Hmongic forms in n- that mean only ‘person’ to the Mienic
forms in m- that mean both ‘person’ and serve as autonyms for the different Mienic
groups. This Mienic word *mjænA (1.24/19) is entered separately here. Pa-Hng /mjɛ2/ with
a labial initial patterns with Mienic, as it does in a number of other instances (see
‘chaff/husk’ and ‘thunder’).
6 & 10. Forms for ‘mouse/rat’ are entered separately under Hmongic and Mienic due to
differences in rime even though the two sets probably represent a single word.
7 & 11. Forms for ‘to ask’ are entered separately under Hmongic and Mienic even though
these forms probably represent a single word.
8. Possibly related to Mienic *mbluŋC ‘rain’ (1.36/28), since they belong to the same rime
set.
 Proto Hmong-Mien 59

9. The Mienic word for ‘leaf’ may be related to Hmongic *mblɔŋA (1.36/29), given the
match in rime and tone.
10. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8), and
Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min (10).

2.13 *s-
PHM *s- 1 2 3 4 5 6 7 8 9 10 11
1. first day of month 1 1b A
- - sia sæin sen - - - - - hɛŋ1
*seŋ
2. sour *suj ɕhu1 ɕɔ1 - su1b soA θjɔ1 sɯ1 ɕui1 tθui1’ swəi1 si1
3. 送 to send/deliver
shoŋ5 - sa5 saŋ5b soŋC θɐŋ5 sõ5 θuŋ5 tθuŋ5 swə5 hiŋ5
*suŋH

PH *s- 1 2 3 4 5 6 7
4. to shut (door) *sow shu7
D
- - ʂə7 - - -

PM *s- 8 9 10 11
5. stomach *siA ɕi1 tθi1’ - sɛi1
6. quilt *su̯ ɛŋC ɕwəŋ5 tθɔŋ5’ sɔŋ5 suŋ5
7. thread *sujC ɕui5 tθuːi5’ sui5 si5
8. centipede *səpD ɕap7 tθap7’ sən7 tsap7
9. near *sətD θat7 tθat7 - -
10. straw sandals
ɕu7 tθu7’ ȶu7 -
*səukD
11. 髟 beard *su̯ amA ɕwam1 tθɔːm1’ - tsam1
12. 心 heart *simA θim1 tθim1 ɕɛn1 -
13. 箴 needle *simA ɕim1 tθim1’ ʨɛn1 tsum1
14. 雙 pair *suŋA ɕuŋ1 tθuŋ1’ swə1 hiŋ1
15. 聲 sound *siŋA ɕiŋ1 tθiːŋ1’ ɕɛ1 -
16. 醒 to sober up
- tθiŋ3 ɕɛ3 -
*siŋB
17. 細 small/fine *sæiC - - - hɛi5
18. 赤 red *sekD ɕi7 tθi7’ ɕi7 sjɛ7
19. 刷 to wipe *sotD ɕot7 - - -

Discussion
The Mun (9) reflexes of voiceless fricatives fall into two large sets: those with tθ- and
those with s-. The words with a Mun reflex of tθ- have been placed together here and have
been reconstructed at both the Proto Hmong-Mien and Proto Mienic level as *s- . The
words with a Mun reflex of s- have been reconstructed at the Proto Hmong-Mien level as
either *sj- (2.28) or *ɕ- (4.13), merging at the Proto Mienic level as *ɕ-. The distinction
between Hmong-Mien *sj- and *ɕ- is based on developments in Western Hmongic: *sj- >
ʂ- and *ɕ- > ɕ-.
The messiness of the fricative correspondences led Wang and Mao to reconstruct eleven
voiceless fricatives for the protolanguage, but since the daughter languages are not
characterized by large numbers of fricatives, it seems reasonable to assume that the
protolanguage did not have many fricatives, either. There are many Chinese loanwords in
these sets, some with unique patterns, that obscure the native correspondences.
60 Chapter 2

Notes
2. This may come from Chinese 酸 ‘sour’ (OC *[s]ˤo[r] > MC swan > Man. suān).
4. This word patterns like no other: the retroflex fricative in Zongdi seems to be an
irregularity.
8. There are similar forms in both Tai-Kadai (Lao /khep7/) and Mon-Khmer (Khmer
/kʔæp/).

2.16 *tj-
PHM *tj- 1 2 3 4 5 6 7 8 9 10 11
1. 種 to grow *n-tju̯ ɛŋH - - - - - - ȵɤ5 ʨwəŋ5 sɔŋ5 ȶɔŋ5 tsuŋ5

PM *tj- 8 9 10 11
2. to wait *tjɔuB ʨu3 ȶu3 ȶu3 -
3. 粢 millet *tjəiA ʨai1 ȶai1 ȶai1 tɛi1
4. 煮 to boil (TR) *tjouB ʨou3 ȶɔu5 ȶəu3 tsu3
5. 紙 paper *tjejB ʨei3 ȶei3 ȶəi3 tsi3
6. 粥 porridge *tju̯ okD ʨwo7 su7 - -
7. 捉 to take *tjɔkD - ȶɔ7 - -

Notes
2. See also Jiangdi Mien (Wang and Mao 15) /tswo3/ for evidence supporting the
reconstruction of both this onset and this rime.

2.17 *thj-
PHM *thj- 1 2 3 4 5 6 7 8 9 10 11
1. 漆 lacquer *thjet shei7 tshei7 - se7b tsheC ʃai7 sɤ7 θjet7 - tshan7 tat7

Notes
1. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
Although not well represented in this small set of languages, the medial -j- is widely
represented elsewhere in Mienic.

2.18 *dj-
PHM *dj- 1 2 3 4 5 6 7 8 9 10 11
1. seven *djuŋH ɕoŋ6 tɕoŋ6 ɕa5 tɕaŋ6 zoŋC ʃɐŋ6 tɕɤ6 sje6 - - -

PM *dj- 8 9 10 11
2. thigh *djujA ʨui2 ȶuːi2 - si2
3. 遲 late *djæiA ʨai2 - ȶai2 -
4. 是 to be *djejB ʨei4 ȶei4 ȶəi4 sɛi4
5. 柿 persimmon
ʨai4 ȶai6 - -
*djəiB
6. 灼 to burn *dji̯ ɛkD ʨa8 sa8 ȶha8 sa8
7. 熟 cooked/ripe
ʨwo8 - ȶhɔ8 -
*dju̯ okD
 Proto Hmong-Mien 61

Notes
1. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). The onset
seems to tie the Mien form to the Hmongic set, but the rime does not agree. Mienic *ŋjiC
(5.24/1) may be related, but the correspondence is difficult. Hmong-Mien ‘seven’ is from
Tibeto-Burman *s-nis ‘seven’ according to Benedict (1987a:13); see chapter 5.

2.19 *ntj-
PHM *ntj- 1 2 3 4 5 6 7 8 9 10 11
1. tree *ntju̯ əŋH tə5 ntu5 ntɔŋ5 - nʔtoŋC ntwaŋ5 - djaŋ5’ gjaŋ5 djaŋ5 djaŋ5

PM *ʔdj- (< ntj- ) 8 9 10 11

2. to drip *ʔdjɛpD djep7 - djan7 dɛp7

Notes
1. Compare ‘to put on/wear (cap)’ (2.4/29), which is exactly the same in Hmongic
(*ntɔŋC), but differs in onset at the Hmong-Mien level (*nt- ‘to put on/wear (cap)’ vs. *ntj-
‘tree’). Perhaps the same as Chinese 樹 ‘tree’ (OC *m-toʔ-s > MC dzyuH > Man. shù), but
compare Mon-Khmer *t2ʔɔɔŋ ‘tree’ (Shorto #491).

2.20 *nthj-
PHM *nthj- 1 2 3 4 5 6 7 8 9 10 11
1. 秤 balance 5 5 5’ 5
- - - - - - ȵɔ
̥ ̃ ʥaŋ ȡaŋ ȡaŋ dzaŋ5
*nthju̯ əŋH

PM *dhj- (< *nthj-) 8 9 10 11

2. to sharpen/grind 5 5’ 5
ʥeu gjɔu ȡau dziu5
*dhjæuC

Notes
2. Compare Chinese 剉 ‘to file’ (MC tshwaH > Man. cuò).

2.21 *ndj-
PM *ndj- 8 9 10 11
1. 慈 mother/aunt
ʥi4 - ȡai4 -
*ndjiB

2.23 *hnj-
PHM *hnj- 1 2 3 4 5 6 7 8 9 10 11
1. heavy *hnjeinX ȵ̥hoŋ3 hei3 hȵ̥a3 ȵa3b ȵ̥oŋB ȵ̥e3 ȵ̥a3 ȵ̥i3 ni3 n̥ i3 nɛi3

Notes
1. This word is reconstructed with *hnj- instead of *hɲ- (4.8) because most reflexes in
Mienic do not show palatal nasals.
62 Chapter 2

2.28 *sj-
PHM *sj- 1 2 3 4 5 6 7 8 9 10 11
1. light(weight) *-sji̯ eu fha1 ɕa1 ʂi1 sei1b seA v̥ uai1’ fɤ1 ɕeu1 sɔu1’ hjau1 -
2. to stand/get up
ɕhu3 ɕə3 ʂaɯ3 sə3b soB θjau3 sɤ3 ɕou3 - səu3 fu3
*sjouX
3. 暑 warm *sji̯ ouX ɕhə3 ɕo3 ʂɔ3 so3b suB - - - sɔu3 sau3 sju3

PH *sj- 1 2 3 4 5 6 7
4. thin *sjɔŋA - ɕu1 ʂɔŋ1 saŋ1b soŋA θjɔŋ1 -
5. on top of *sjuA - - - sa1b saA - -
6. sinew *sju̯ eiB ɕhu3 ɕi3 - sɔ3b suB - si3
7. cooked/ripe *sjenB ɕhaŋ3 ɕɛ3 ʂia3 sæin3b senB θeŋ3 sɪ3̃
8. midday meal *sjɛŋC - - ʂu5 soŋ5b saŋC - -
9. to rest *sjoC ʨhə5 ɕo5 ʂɔ5 so5b suC θjeu5 -
10. to wipe *sjəŋC ɕhaŋ5 ɕɑŋ5 ʂɔ5 soŋ5b soŋC - -
11. 廝 RECIPROCAL
ɕi5 - ʂi1 - siA-D - -
*sjɨA-D
12. 升 measure for
ɕhen1 - ʂia1 sæin1b senA - ʨhɔ1̃
grain *sjinA
13. 收 to receive/
ɕhu1 - ʂau5 - suA - -
gather in *sjuwA

Discussion
The Mun (9) reflexes of voiceless fricatives fall into two large sets: those with tθ- and
those with s-. The words with a Mun reflex of tθ- have been placed together and have been
reconstructed at both the Proto Hmong-Mien and Proto Mienic level as *s- (2.13). The
words with a Mun reflex of s- have been reconstructed at the Proto Hmong-Mien level as
either *sj- (here) or *ɕ- (4.13), merging at the Proto Mienic level as *ɕ-. The distinction
between Hmong-Mien *sj- and *ɕ- is based on developments in Western Hmongic: *sj- >
ʂ- and *ɕ- > ɕ-.
The messiness of the fricative correspondences led Wang and Mao to reconstruct eleven
voiceless fricatives for the protolanguage, but since the daughter languages are not
characterized by large numbers of fricatives, it seems reasonable to assume that the
protolanguage did not have many fricatives, either. There are many Chinese loanwords in
these sets, some with unique patterns, that obscure the native correspondences.

Notes
1. The labiodental initials in Yanghao, Jiongnai, and Pa-Hng are difficult to explain. Wang
and Mao reconstructed a *phs- initial for their set 59, in which this word appears, although
this word is the only one to show only labial fricatives rather than labial stops.
5. There are retroflex onsets for ‘on top of’ in West Hmongic Xianjin and Shimen (Wang
and Mao 3 & 4). It may be that White Hmong saum /ʂau8/ ‘above’ is related, although
neither the rime nor tone corresponds.
 Proto Hmong-Mien 63

2.40 *ʔl-
PH *ʔl- 1 2 3 4 5 6 7
1. CLF-bowls/houses
lɛ1 le1 lu1 noŋ1a ʔlaŋA xoŋ1 lɔ1̃
*ʔlɛŋA
2. monkey *ʔlinA lei1 - lia1 læin1a - lai1 -
3. red *ʔlinA - - lia1 læin1a ʔlenA - -
4. short *ʔlɛŋB lɛ3 le3 lu3 loŋ3a ʔlaŋB laŋ3 lɔ3̃
5. to squeeze/pinch
la5 lɑ5 li5 lei5a - l̥ i5 la5
*ʔlɨC
6. CLF-mouthfuls *ʔloC lo5 - lɔ5 lo5a - lu5 -
7. 拗 to break *ʔloB lo3 - lɔ3 lo3a ʔluB lu3 ljõ3’

PM *ʔl- 8 9 10 11
8. CLF-fields *ʔli̯ əuC lau5’ - - -

Notes
1. Compare Mienic *ʔnɛɔmA ‘clf-bowls/houses’ (2.7/22) and see discussion in chapter 6,
section 6.2.1.
2. The Mienic word for ‘monkey’ (*ʔbiŋA 1.4/18) shares the same tone and rime with this
word, but begins with a different initial, suggesting the possibility of an ancient disyllabic
source. Compare Proto Tai *liŋ ‘monkey’ (Li 1977).
2 & 3. These may be the same word, if the monkey was referred to as ‘the red one’. How
the connection between ‘monkey’ and ‘red’ may relate to the connection suggested above
between the Hmongic and Mienic words for ‘monkey’ is unclear. Probably one connection
or the other is correct, but not both.
4. Compare Mienic *ʔnəŋB ‘short’ (2.7/22) and see discussion in chapter 6, section 6.2.1.
7. Compare Mienic *ʔnəuB from the same Chinese word (2.7/7) and see discussion in
chapter 6, section 6.2.1.
8. Also Jiangdi Mien /ljou5/ and Xiangnan Mien /ljəu5/. Compare Gaopo (Hmongic) /loŋ5/
‘field’ and Tibeto-Burman *low ‘field’ (Matisoff 2003). The tone subcategory of Luoxiang
Mien indicates that this word should be reconstructed with *hl-, but given the evidence
from these other forms, it is provisionally placed here.

2.40.1 *ʔlj-
PH *ʔlj- 1 2 3 4 5 6 7
1. to pour *ʔlju̯ eiA ȴu1 - lau1 ɭo1a ʔluA - -
2. rabbit *ʔljuB - - lua3 ɭa3a ʔlaB - -
3. male *ʔljuwB - - lau5 ɭu3a ʔloB - -

Notes
2. In White Hmong, the words for ‘rabbit’ and ‘donkey’ are homophonous: luav /lua3/. A
luav-nees ‘donkey-horse’ is a mule. Another word for ‘mule’ (White Hmong luj /lu2/,
attested across the family) is clearly a loanword from Chinese 騾 ‘mule’ (Man. luó).
Connections among these are unclear.
3. In White Hmong, lau /lau5/ is male of birds, while laug /lau6/is male of sheep, goat, cow.
64 Chapter 2

2.41 *hl-
PHM *hl- 1 2 3 4 5 6 7 8 9 10 11
1. bamboo *hləwX l̥ ho3 l̥ ho3 hl̥ ɔ3 - - - - l̥ au3 lɔu3 ɬau3 lau3
2. plum *hli̯ əŋX - - - - - - le3 gaŋ3 - ɬaŋ3 ȵɔŋ3
3. moon/month *hlaH l̥ ha5 l̥ hɑ5 hl̥ i5 li5b l̥ aC l̥ e5 l̥ a5 la5 la5 ɬa5 lɔu5
4. to slice *hlep l̥ hei7 l̥ hɑ7 hl̥ ai7 le7b - l̥ ai7 - le7 - - -

PH *hl- 1 2 3 4 5 6 7
5. marrow/brains 1 1 1 1b A
l̥ hɛ l̥ hə hl̥ ɯ lou l̥ u - -
*hlʉA
6. bridge *hlæA - - - la1b - - -
7. rope *hlæC l̥ ha5 l̥ ha5 hl̥ ua5 la5b l̥ iC l̥ ei5 l̥ e5
8. 湯 to scald *hlaŋA - - hl̥ a1 lua1b l̥ enA - la6
9. 鐵 iron *hluwC l̥ hə5 l̥ hɔ5 hl̥ au5 lu5b l̥ oC l̥ ɔ5 l̥ ɯ5

PM *hl- 8 9 10 11
10. petal *hlemC ɬim5 leːm5 - ɬiŋ5
11. which *[hl]aiC laːi5 - ha5 -

Notes
2. This may be from Chinese 李 ‘plum’ (MC liX > Man. lǐ). Hmong-Mien *l- > g- in
Luoxiang Mien and Mun before high vowels and glides (Solnit 1996:9–10).
4. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). Compare
Tibeto-Burman *(s-)lep ‘to slice’ (Matisoff 2003).
5. According to Sagart (1999:67), this is the same as Chinese 髓 ‘marrow’ (OC *s-lojʔ >
MC sjweX > Man. suǐ).
6. See also Shimen /l̥ a1/, Qingyan /l̥ o1/, and Gaopo /l̥ hɑ1/ ‘bridge’ (all West Hmongic).
Compare Tibeto-Burman *(s-)ley ‘bridge, ladder’ (Matisoff 2003), which is a closer
semantic match than Chinese 梯 ‘wooden steps, stairs’ (OC *l̥ [ə]j > MC thej > Man. tī).
9. See Mienic *hrɛkD ‘iron’ (2.56/9). These were borrowed separately from Old Chinese
*l̥ ˤik.
10. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
11. The placement of Mienic ‘which’ in this set is tentative. The reflexes show either l- or
h-; Wang and Mao (1995) reconstruct the word with an *h-.

2.41.1 *hlj-
PHM *hlj- 1 2 3 4 5 6 7 8 9 10 11
1. big *hljo ȴ̥hə1 - hl̥ ɔ1 ɭo1b l̥ uA l̥ jeu1’ ȴo1 lo1 lu1’ ɬu1 lɔu1
2. sash/cord/rope
l̥ haŋ1 - hl̥ a1 ɭua1b - - - l̥ aːŋ1 laːŋ1’ ɬaŋ1 -
*hljaŋ

PH *hlj- 1 2 3 4 5 6 7
3. young *hljæD - - hl̥ ua7 ɭa7b - - -
4. 燎 to burn
- - hl̥ aɯ3 ɭe3b - - -
(mountain) *hljowB
 Proto Hmong-Mien 65

Discussion
Luoxiang Mien and Lanjin Mun do not show initial g- in ‘big’ and ‘sash/cord/rope’ as
would be expected since the regular development of *l- before a high vowel or glide is g-
in these dialects (Solnit 1996:9–10). It may be that *lj was not a liquid-glide cluster, but
rather a palatal lateral.

Notes
1. In White Hmong, hlob /hl̥ ɔ1/ is ‘to grow; elder; great in volume, proud’. The simple
word for ‘big’ is loj /lɔ2/, which appears to have the same root. The tonal and onset
difference may be attributed to a prefix that devoiced the initial of ‘to grow’: *hlj- yielded
upper-register tone 1, while *lj- yielded lower-register tone 2 (see chapter 4).
2. The meaning ‘sash/cord/belt’ is associated with this word in Hmongic; the meaning
‘rope’ is associated with this word in Mienic. Hmongic forms are from Wang 1994.
Possibly from Chinese 繩 ‘string, cord’ (OC *m.ləŋ > MC zying > Man. shéng).

2.42 *l-
PHM *l- 1 2 3 4 5 6 7 8 9 10 11
1. long (time) *ləu la2 lɯ2 le2 læ2 liA la2 lo2 lau2 lau2 - lu2
2. CLF-people *lan lɛ2 le1 lɛŋ2 læ2 - laŋ1 læ̃ 2 laːn2 laːn2 - na1
3. to come back *ləwX lo4 lo4 lɔ4 lu4 luB lɔu4 lo4 lau4 lau4 lau4 lau4

PH *l- 1 2 3 4 5 6 7
4. to rot *lʉA la2 - lɯ2 lou2 lenA - -
5. old *lu̯ eiB lu4 - lau4 lo4 lauB - -
6. late *liC - - li6 li6 - - lɦe6
7. 聾 deaf *loŋA l̥ oŋ2 - la6 - loŋA - lõ2
8. 兩 tael (40 grams)
laŋ4 - la4 - laŋB - ljɔ6̃
*li̯ aŋB
9. 露 dew *lʉC - - lɯ6 lu6 - - -

PM *l- 8 9 10 11
10. bellows *louA lou2 lɔu2 ləu2 lu2
11. 累 lazy *lu̯ eiC lwei6 lei6 lwəi6 -

Notes
1. The Green Mong form meaning ‘a long time ago’ has been substituted for White Hmong
(3).
2. Compare Chinese láng 郎 ‘young man’ (in Min dialects a homophonous word means
‘person’).
3. In White Hmong, ‘to come back to one’s home’; compare 2.3/15 ‘to come (back to a
place other than one’s home)’.
5. The individual forms look very much like Chinese 老 ‘old’ (MC lawX > Man. lǎo), but
at the Hmongic level the rime does not correspond.
6. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
8. See Mienic *lji̯ oŋB ‘tael’ (2.42.1/26) from the same source.
10. Possibly from Chinese 蘆 ‘reed’ (MC lu > Man. lú), since Hmong and Mien bellows
are made from a cylinder with a pump inside (Lemoine 1972:130–131).
66 Chapter 2

2.42.1 *lj-
PHM *lj- 1 2 3 4 5 6 7 8 9 10 11
1. field *ljiŋ ȴi2 - lia2 ɭæin2 lenA leŋ2 lĩ2 giŋ2 giːŋ2 ljɛ2 ljaŋ2
2. 鎌 sickle *ljim ȴen2 - lia6 ɭæin2 lenA ljen2 - lim4 - ljɛn2 dzjam2
3. 里 half kilometer
ȴi4 - - - liB - - - - ljaŋ4 li4
*ljɨX

PH *lj- 1 2 3 4 5 6 7
4. to bury *ljəŋB ȴaŋ4 ȴɑŋ4 lɔ4 ɭoŋ4 loŋB - læ̃ 4
5. to be used to *lju̯ aC ȴa6 ȴɑ6 - ɭəa6 laC - -
6. lightning flash *ljeD ȴi8 ȴɑ8 lai8 ɭe8 leD - -
7. 量 to measure (rice)
ȴi2 - lu2 ɭoŋ2 laŋA - -
*ljɛŋA

PM *lj- 8 9 10 11
8. lightning flash *ljəpD gap7 - - dzjɛp8
9. to sew *lju̯ ɛnA gwən2 gɔn2 - -
10. 兩 tael (40 grams)
guŋ4 guŋ4 lwə4 liŋ4
*lji̯ oŋB

Notes
1. Compare the almost identical Chinese 田 ‘field; to hunt’ (OC *lˤiŋ > MC den > Man.
tián) and Tibeto-Burman *liŋ ‘field’ (Matisoff 2003). Both Haudricourt and Strecker
(1991) and Sagart (1999:183–184) entertain the possibility that this was a loan from
Hmong-Mien to Chinese.
4. This may be from Chinese 斂 ‘to dress a corpse’ (Man. liàn), although the tones do not
correspond.
6 & 8. The Hmongic word would have ended in either -p or -t (Hmong-Mien tone 8 words
ending in -k merged with tone 6 in Hmongic), so these two words are clearly related.
Compare to Tibeto-Burman *(s-)lyap ‘flash; lightning’ (Matisoff 2003).
10. See Hmongic *li̯ aŋB ‘tael’ (2.42/26) from the same source.

2.46 *tr-
PH *tr- 1 2 3 4 5 6 7
1. to return *trəŋB ȶaŋ3 - ʈɔ3 ʈoŋ3a tʂoŋB - -
2. houttuynia cordata
ȶu7 ʈu7 ʈau7 ʈu7a tʂoD - -
*truwD
3. 張 CLF-tools *traŋA ȶaŋ1 ʈei1 ʈa1 ʈua1a tʂenA tʃen1 -
4. 蹢 hoof/claw/nail *tru̯ eiC - ʈa5 ʈau5 ʈɔ5a - - -
5. 著 to put on/wear
ȶi5 ʈɔ5 ʈau5 ʈɔ5a tʂuC - tɕɯ5
(shoes) *trɔC

PM *tr- 8 9 10 11
6. 張 CLF-tools *truŋA - ȶuŋ1 ȶwə5 -

Notes
2. Also known as 魚腥草 (Man. yú xīng cǎo), this is a plant used primarily for medicinal
purposes (see chapter 7).
 Proto Hmong-Mien 67

3 & 6. This classifier for tools has been independently borrowed from the same source in
the two sub-families.
5. Compare Hmongic *drɔC ‘to hit target’ (2.48/6) from a different form of the same
Chinese root (chapter 4, section 4).

2.47 *thr-
PHM *thr- 1 2 3 4 5 6 7 8 9 10 11
1. 插 to insert *threp ȶhi7 - ʈhai7 ʈe7 tʂheD ʃai7 - θip7 tip7 tɕhɛn7 hɛp7

PH *thr- 1 2 3 4 5 6 7
2. 抽 to pull out *throC ȶhə5 ʈho5 ʈho5 ʈo5b tʂhuC - -

Notes
1. Occasional voiced initials in Mienic (not represented here) suggest the presence of a
voiced pre-initial.

2.48 *dr-
PHM *dr- 1 2 3 4 5 6 7 8 9 10 11
1. 箸 chopsticks *drouH ȶu6 ʈɯ6 ʈaɯ6 ʈɯ6 ʐoC ʃau6 tɕɔ6 tɕou6 ȶɔu6 ȶəu6 tau6

PH *dr- 1 2 3 4 5 6 7
2. firm/durable *druA - - ʈua2 ʈa2 - - -
3. step *droA ȶə2 ʈo2 ʈɔ2 ʈo2 ʐuA tʃa2 tɕa6
4. table/bench *drɔŋA - - ʈɔŋ2 ʈaŋ2 ʐoŋA - tɕe2
5. bamboo *drowD - - ʈaɯ8 ʈə8 - - -
6. 筒 flute/tube *droŋA ȶoŋ2 ʈoŋ2 ʈa2 - ʐoŋA - -
7. 著 to hit target *drɔC - ʈɔ6 ʈau6 ʈo6 - - tɕɯ6

Notes
3. The Jiwei (2) form is from Office of Miao-Yao Research Lexicon (1987).
7. Compare Hmongic *trɔC ‘to put on/wear (shoes)’ (2.46/6) from a different form of the
same Chinese root (chapter 4, section 4).

2.49 *ntr-
PHM *ntr- 1 2 3 4 5 6 7 8 9 10 11
1. 中 center/middle *ntroŋ ȶoŋ1 ɳʈoŋ1 ɳʈa1 ɳʈaŋ1a ɳʔtʂoŋA - ȵtɕy1 toŋ5’ tɔŋ5 tɔŋ5 -

PH *ntr- 1 2 3 4 5 6 7
2. breast/chest *ntrəŋA - - ɳʈɔ1 ɳʈoŋ1a - - -
3. to be wedged in *ntrinC - ɳʈhɑ7 - ɳʈei5b - - -
4. to butt (of bulls)
ȶu5 - ɳʈau5 ɳʈu5a ɳʔtʂoC - -
*ntruwC

Notes
1. The Hmongic and Mienic tones do not correspond. They appear to have been borrowed
from different forms of Chinese 中.
68 Chapter 2

4. This is probably from Chinese 鬥 ‘fight, wrangle’ (MC tuwH > Man. dòu).

2.50 *nthr-
PH *nthr- 1 2 3 4 5 6 7
1. puttees *nthrɔŋA ȶhu1 ɳʈhu1 ɳʈhɔŋ1 ɳʈaŋ1b ɳʔtʂhoŋA ɳtʃwaŋ1’ ȵɔ 1
̥ ̃

Notes
1. Tan Trinh Pa-Hng (Niederer 1997) has been substituted for Baiyun Pa-Hng (7).

2.51 *ndr-
PHM *ndr- 1 2 3 4 5 6 7 8 9 10 11
1. drum *ndruX ȵə4 ɳɑŋ4 ɳʈua4 ɳʈəa4 ɳtʂaB ɳtʃu4 mjo4 dʑo4 ȡu4 du4 -

PH *ndr- 1 2 3 4 5 6 7
2. plain (level ground)
- - ɳʈa4 - ɳtʂoŋB - -
*ndroŋB
3. interior *ndrəŋC ȵaŋ6 ɳɑŋ6 ɳʈɔ6 - - - -
4. to track *ndroC - ɳɑŋ6 ɳʈɔ6 - - - -
5. young man *ndru̯ eiC ȵo6 - ɳʈau6 ɳʈo6 ɳtʂuC ɳtʃi6 -
6. 摘 to strip leaves
ȵu4 ɳei4 - ŋko4 ɳtʂuB - -
*ndru̯ eiC

Notes
4. White Hmong nrog means ‘to accompany; with’ (‘to track’ > ‘to follow’ > ‘to
accompany’ > ‘with’).

2.53 *hnr-
PHM *hnr- 1 2 3 4 5 6 7 8 9 10 11
1. seed *hnru̯ em ȵ̥hu1 ɳu1 nɔŋ1 ɳaŋ1a n̥ oA nwaŋ1 ȵõ1 ȵem1 ŋjim1’ n̥ jɛn1 num1
2. cooked rice *hnrəaŋH - - hn̥ ɔ5 ɳɔ5b - - - naːŋ5 naŋ5’ n̥ aŋ5 nɔŋ5

Notes
1. Downer (1973) noted the similarity between this word and Cantonese /ŋan/ ‘seed’.
2. Possibly from Chinese 饟 ‘bring food to workers in the field’ > ‘food brought to
workers in the field’ (OC *n̥ aŋ(ʔ)-s > MC syangH > Man. xiǎng).

2.55 *ʔr-
PHM *ʔr- 1 2 3 4 5 6 7 8 9 10 11
1. stone *-ʔrəu ɣi1 ʐɯ1 ʐe1 ʐæ1a ʔwjiA ŋkja1 jo1 gau1 gjau1 lau1 dzu1
2. tender *ʔrəunH ɣaŋ5 ʐɑŋ5 - - - ɳtʃoŋ5 - gun5’ - lun5 in5
3. good *-ʔrɔŋH ɣu5 ʐu5 ʐɔŋ5 ʐaŋ5a ʔwjoŋC ŋwaŋ5 jɔ5̃ gwəŋ5’ gɔŋ5 lɔŋ5 dzɔŋ5
4. vegetable *ʔræi ɣu1 ʐei1 ʐau1 ʐɔ1a ʔwjuA ji1 ɦji1 gai1 gjai1 lai1 ɛi1

PH *ʔr- 1 2 3 4 5 6 7
5. to dry in sun *ʔrinA - - ʐia1 ʐæin1a - - -
6. honey *ʔraA va1 - ʐi1 ʐɪ1a ʔwjaA - -
7. forest *-ʔrɔŋB ɣu3 ʐu3 ʐɔŋ3 - ʔwjoŋB ŋkjɔŋ3 jɔ3̃
 Proto Hmong-Mien 69

8. to look after *ʔroB ɣə3 - ʐɔ3 ʐo3a ʔwjuB - -

9. near *-ʔræwC ɣi5 ʐɯ5 ʐe5 ʐæ5a ʔwjiC ŋkja5 jo5
10. to conceal *ʔreD ɣi7 ʐɑ7 ʐai7 ʐe7a ʔwjeD vwei7 ɛ7

PM *ʔr- 8 9 10 11
11. clothes *ʔrujA gui1 gui1 lwəi1 -

Discussion
The *r- initials 2.55, 2.56, 2.57 are historially very interesting, and the interpretation of the
patterns is challenging. On the basis of prenasalized velar stops in Jiongnai (6), Wang and
Mao reasonably reconstruct clusters with velar stops for this series: *ŋkl- , *ŋkhl- , *ŋgl- .
In this reconstruction, following the pattern of other correspondence sets where there is
evidence for loosely-adjoined initial material, this material has not been explicitly
reconstructed. But unlike the case of Hmong-Mien *-bɔuX ‘hand/arm’ above (1.3/3), the
evidence does not point to a variety of prefixes which have obscured the original onset
correspondence in the root syllable, since in every case the Jiongnai forms point to an
ancient dorsal (the velar initials in Mienic are a secondary development of the original
liquid). Reasonable alternate reconstructions would be *K-r-, *Kh-r, *G-r-.

Notes
1. Also ‘grindstone’ in White Hmong.
4. Perhaps Chinese 卉 ‘herbs, vegetation’ (MC xjwɨjH > Man. huì). See correspondence to
‘to conceal’, this set.
7. Jiongnai (6) from Mao and Li 2001.
10. Perhaps Chinese 諱 ‘taboo; to conceal’ (MC xjwɨjH > Man. huì). See correspondence
to ‘vegetable’, this set.

2.56 *hr-
PHM *hr- 1 2 3 4 5 6 7 8 9 10 11
1. liver *-hri̯ ən xhi1 ʂɛ1 ʂia1 ʐæin1b senA ŋkheŋ1’ hɪ1̃ gan1 gjen1’ ɬan1 -
2. high/tall *-hri̯ əŋ xhi1 ʂɛ1 ʂia1 ʐæin1b senA ŋkheŋ1’ hɪ1̃ gaŋ1 gjaŋ1’ ɬaŋ1 dzɔŋ1

PH *hr- 1 2 3 4 5 6 7
3. sound/noise *-hruA xhə1 ʂo1 ʂua1 ʐa1b saA ŋkjeu1’ saŋ1
4. fern *-hruA xhə1 ʂo1 ʂua1 ʐa1b saA ŋkjeu1’ -
5. artemisia *hrɛŋB xhi3 - - ʐoŋ3b saŋB - -
6. quick *hreC xhi5 ʂɑŋ5 ʂai5 - seC v̥ wei5 -
7. to write *hru̯ eiC xho5 ʂei5 ʂau5 ʐɔ5b - - -

PM *hr- 8 9 10 11
8. to measure (rice)
gaːu1 gaːu1’ ɬa1 dzau1
*hrauA
9. 鐵 iron *hrɛkD gja7 gja7’ ɬja7 ljɛ7

Discussion
See discussion of this series under 2.55.
70 Chapter 2

Notes
7. Compare Chinese 鏤 ‘to incise’ (OC *[r]ˤo-s > MC luwH > Man. lòu).
9. See Hmongic *hluwC ‘iron’ (2.41/9). These were borrowed independently from Old
Chinese *l̥ ˤik.

2.57 *r-
PHM *r- 1 2 3 4 5 6 7 8 9 10 11
1. nest *-rəuX ɣi4 ʐɯ4 ʐe4 ʐe4 wjiB ŋkja4 - gau4 gjau4 la4 -
2. urine *-raX va4 ʐɑ4 ʐi4 ʐɪ4 wjaB ŋkwe4 ve4 wa4 va4 lɔ4 vjɛ4
3. pangolin *rɔiH ɣo6 ʐɔ6 ʐau6 ʐo6 wjoŋA - ɦje6 gai6 gjai6 - -
4. sharp *-rajH ɣa6 ʐɑ6 ʐua6 - wjaC ŋkja6 - gai6 gjai6 lai6 hɛi6
5. 龍 dragon *-roŋ ɣoŋ2 ʐoŋ2 ʐa2 ʐaŋ2 wjoŋA ŋkjaŋ2 jõ2 - - ljuŋ2 luŋ2
6. 梨 pear *rəj ɣa2 ʐɑ2 - ʐa2 wjaA - ɦje2 gei2 gei4 lai2 dzai2
7. 里 village *rəŋX ɣaŋ4 ʐɑŋ4 ʐɔ4 ʐoŋ4 wjoB - - gaŋ4 gjaŋ4 laŋ4 dzɔŋ4
8. 廩 barn/granary
- ʐe4 - ʐoŋ4 wjaŋB - - gam4 gjam4 lan4 dzum4
*rɛmX
9. 力 strength *-rək ɣə6 ʐo6 ʐɔ6 ʐu6 wjuC ŋkjeu6 ja6 [kha7 - khla7] dzja6

PH *r- 1 2 3 4 5 6 7
10. comb *ræC ɣa6 ʐa6 ʐua6 ʐəa6 wjiC vi6 ɦji6
11. grind (grain) *roD - ʐo8 ʐɔ8 ʐo8 wjuD - -

PM *r- 8 9 10 11
12. pair *reŋC - geːŋ6 lɛ6 dzaŋ6
13. 淋 to water *rəmA gjem2 - ljen2 dzum2
14. 流 to flow ri̯ əuC gjeu6 gjɔu6 ljɔu6 -
15. 立 to erect ri̯ ɛpD gjep8 gjap8 - -
16. 笠 bamboo hat
gap8 gjap8 ɬan8 dzup8
rəpD

Discussion
See discussion of this series under 2.55.

Notes
2. Although the initials in Mien (8) and Mun (9) are irregular, it is not clear that a different
initial should be reconstructed for this word.
6. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10).
9. Xiangnan Mien (Wang and Mao 16) has been substituted for Luoxiang Mien (8). The
preinitial element in ‘strength’ must have been aspirated, unlike the other preinitial
elements in this set, to account for the upper register Mien and Biao Min forms (M
*khləkD, 5.32/7).
13 & 14. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao
Min (10).
 Proto Hmong-Mien 71

3. Dental affricates

3.1 *ts-
PHM *ts- 1 2 3 4 5 6 7 8 9 10 11
1. 竈 stove *N-tsoH so5 - tsɔ7 so5b - tsu7 ʨɔ1 do5’ du5 tsu5 tɔu5
2. 接 to receive/
sei7 - tsai7 se7a tseD tʃei7 - - tθip7 ʨin7 tɛp7
borrow *tsep

PH *ts- 1 2 3 4 5 6 7
3. billhook *tsæC sa5 - tsua7 sa7a tsiC - -
4. to join *tsæD sen7 tshɑ7 tsua7 sa7a - - -
5. 騣 mane *tsʉŋA soŋ1 tsoŋ3 tsɔŋ1 saŋ1a pjoŋA - -

PM *ts- 8 9 10 11
6. grave/tomb *tsouB θou3 tθɔu3 tsəu3 -
7. to dry (clothes) - si7 ʨi7 tsaŋ1
*tsekD
8. bland/tasteless θaːm3 tθaːm3 tsan3 tɔm3
*tsamB
9. 隻 CLF-birds *tsi̯ ɛkD - sa7’ ȶa7 tsa7
10. 識 to know/ - sa1 ȶa7 tsa7
recognize *tsi̯ ɛkD
11. 織 to weave *tsi̯ ɛkD ʨa7 sa7 ȶa7 -

Discussion
There are many Chinese loanwords in this series. Once all the etymologies have been
thoroughly researched, it may be possible to assign dental affricates to the “loan
phonology” component of the inventory.

Notes
1. Since irregular, probably borrowed at different times.
3. A ‘billhook’ is a tool with a curved blade used for heavy cutting work.
8. From an Early Min word that means ‘tasteless, insipid’ (/tsjäm/ in the Jíyùn dictionary:
Norman 1988:232).

3.2 *tsh-
PHM *tsh- 1 2 3 4 5 6 7 8 9 10 11
1. bone *tshuŋX shoŋ3 soŋ3 tsha5 saŋ3b tshoŋB θɐŋ3 sõ3 θuŋ3 tθuŋ3 swə3 hiŋ3
2. 瘡 sore/boil/blister
shaŋ1 - tsha1 sua1b tshenA - - - - saŋ1 -
*tshaŋ
3. 千 thousand *tshi̯ en shaŋ1 tshɛ1 tshia1 ʨein1b tshenA ʃen1 ɕɛ5 θin1 tin1’ ʨhin1 hun1
72 Chapter 2

PH *tsh- 1 2 3 4 5 6 7
4. steel *tshæwA sha1 sɯ1 - sæ1b - - ʨo1
5. husked rice/millet
shɛ3 se3 tshu3 soŋ3b tshaŋB - -
*tshɛŋB
6. tin *tshæD - - tshua7 sa7b - - -

Notes
6. In White Hmong, ‘lead metal’.

3.3 *dz-
PHM *dz- 1 2 3 4 5 6 7 8 9 10 11
1. 錢 money *dzi̯ en saŋ2 - tsia2 sæin2 zenA ʃiŋ2 ʨɪ2̃ θin2 tθin2 tsən2 hɛn2
2. 匠 craftsman
ɕaŋ6 ʨɑŋ6 - saŋ6 zaŋC - - θɛŋ6 tθaːŋ6 tsaŋ6 -
*dzi̯ ɔŋH
3. 漬 to soak *dzeiH - - - - - - ʨi6 θei6 tθei6 - hɛi6
4. 鑿 chisel *dzəuk so6 - tsau6 so6 zuC ʃu8 ʨɪ4̃ θu8 tθu6 tshu8 hiu8

PH *dz- 1 2 3 4 5 6 7
5. able *dzowA su2 - tsaɯ2 sə2 - - -
6. cool (water) *dzinB sei4 tsɛ4 tsia4 sein4 - tʃai4 ʨɦĩ4
7. 慚 ashamed *dzaŋA - tsei2 tsa2 sua2 zenA - ʨe2
8. 造 to reach/arrive at
so6 - tsɔ6 - zuC - θu5
*dzoC

PM *dz- 8 9 10 11
9. firewood *dzaŋA θaːŋ2 tθaːŋ2 tsaŋ2 hɔŋ2
10. 槽 trough *dzuA θou2 - tsu2 -
11. 坐 to sit *dzu̯ eiB θwei4 tθei4 tswei4 hɛi4
12. 緻 beautiful *dzejC - - tsəi6 -

Notes
9. Downer (1973:21) says this is a loanword from Chinese 柴 ‘firewood’ (MC dzrɛ > Man.
chái), one of a number that show a relationship between final -i or -u in Middle Chinese
and final -ŋ in Mien.
12. Other Mienic forms for ‘beautiful’ have voiced initials, so this word probably had a
voiced pre-initial.

3.4 *nts-
PHM *nts- 1 2 3 4 5 6 7 8 9 10 11
1. 早 early *ntsi̯ ouX so3 ntso3 ntsɔ3 ntso3a nʔtsuB ɳtʃu3 ȵo3 djeu3 gjɔu3 djəu3 du3
2. 澡 wash (hands)
sa3 ntsa3 ntsua3 ntsa3a nʔtsiB ɳtʃei3 ȵe3 daːu5 daːu5’ dza3 dɔu3
*ntsæwX

PH *nts- 1 2 3 4 5 6 7
3. husked rice *ntsuwC - ntsɔ5 - - - ɳtʃei5 ȵɯ5
4. grave/tomb *ntsaŋC saŋ5 ntsei5 ntsa5 ntsua5a - - ȵe5
5. to suck *ntseD - - ntsai7 - - - -
 Proto Hmong-Mien 73

PM *ʔdz- (<*nts-) 8 9 10 11
6. boat/ship *ʔdzaŋB daːŋ3 daːŋ3 dzaŋ3 dɔŋ3
7. 騣 mane *ʔdzɔŋA θoŋ1 dɔŋ1 tsɔŋ1 dzəŋ1

Notes
4. In Yanghao (1), ‘to store a body before burial’ (Wang 1994).
5. This word is also found in West Hmongic Shimen /ntsai7/, Gaopo /nze7/, and Fengxiang
/zɛ7/ (Wang and Mao 4, 6, 9).

3.5 *ntsh-
PHM *ntsh- 1 2 3 4 5 6 7 8 9 10 11
1. to launder *ntshuH sho5 ntsho5 ntshua5 ntsa5b nʔtshaC ɳtʃu5 ȵo5 do5 du5’ dzu5 du5

PH *ntsh- 1 2 3 4 5 6 7
2. clean *ntshɨA sha1 ntsha1 - ntsei1b nʔtsheA θei1 -
3. rough/coarse
sha1 ntshɑ1 ntshi1 ntsi1b nʔtshaA - -
*ntshaA
4. steep *ntshoŋA shoŋ1 ntshoŋ1 ntsha1 ntsaŋ1b nʔtshoŋA ɳtʃɐŋ3 -
5. elephant *ntshʉC shɛ5 - ntshɯ5 - nʔtshuC - -

PM *dzh- (<*ntsh-) 8 9 10 11
6. to disperse *dzhanC daːn5 daːn5’ dzan5 -

Notes
1. In White Hmong, ‘to scrub’.
3. Compare Chinese 粗 ‘coarse’ (OC *s.[r̥ ]ˤa > MC tshu > Man. cū).

3.6 *ndz-
PHM *ndz- 1 2 3 4 5 6 7 8 9 10 11
1. to play *N-dzæwH - tsa4 - - - - - dzjaːu6 dzaːu6 dza6 -

PH *ndz- 1 2 3 4 5 6 7
2. thin (person) *ndzu̯ eiC su5 ntsei5 ntsau6 ntso6 - ɳtʃe6 nʑi6

PM *ndz- 8 9 10 11
3. root *ndzuŋA duŋ2 duːŋ2 - -
4. word/character *ndzaŋC - daŋ6 dzaŋ6 -
5. 鹺 salty *ndzaiA daːi2 daːi2 dza2 dɔi2
6. 淨 clean *ndzəŋC daŋ6 dðaŋ6 - -

Notes
1. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
3. In Jiangdi Mien and Xiangnan Mien (Wang and Mao 15 & 16), the initial is dz- . This
may be related to Hmongic *ɟoŋA ‘root’ (4.3/28), but if so, there must have been a pre-
initial that caused voicing in Mienic (the Hmongic onset is not prenasalized).
74 Chapter 2

4. This is probably from Chinese 字 ‘word, character’ (OC *Cə-[dz]ə-s > MC dziH > Man.
zì), despite the difference in rime. Downer (1973:21) cites this as one of a number of
examples where Middle Chinese -i or -u corresponds to Mienic -ŋ.
6. In Jiangdi Mien and Xiangnan Mien (Wang and Mao 15 & 16), the initial is dz- .

3.16 *tsj-
PHM *tsj- 1 2 3 4 5 6 7 8 9 10 11
1. gall bladder *tsji̯ əu ɕen1 ʨi1 tʂi1 sei1a tseA - ʨe1 - - - tɛu1
2. malodorous *tsju̯ eiH - ʨə5 tʂɯ5 sou5a tsuC tʃau5 ʨɔ5 θwei5’ tθei5 tswəi5 ti5
3. 甑 rice steamer *tsjɛŋH ɕi5 ʨe5 tʂu5 soŋ5a tsaŋC tʃjaŋ5 ʨɔ̃1 θaŋ5’ tθaŋ5 tsaŋ5 taŋ5

PH *tsj- 1 2 3 4 5 6 7
4. what *tsjɨB ɕi3 - tʂi5 si7b tsiB - -
5. to let go *tsjəŋC ɕaŋ5 ʨɑŋ5 tʂɔ5 - - - ʨõ5
6. armpit *tsjoC/D ɕə5 ʨo5 tʂɔ7 so7a tsuC - ʨa5
7. pheasant *tsjowD - - tʂaɯ7 sə7a - - -
8. 祖 tiger *tsjoB ɕə3 ʨo3 tʂɔ3 so3a tsuB tʃu3 -

Notes
8. This is a borrowing from Chinese 祖 ‘ancestor’ (MC tsuX > Man. zǔ). There is a custom
in Southeast Asia of addressing the tiger as one’s grandfather or ancestor (Wessing 1986).

3.17 *tshj-
PHM *tshj- 1 2 3 4 5 6 7 8 9 10 11
1. hungry *tshje - ɕi1 tʂhai1 se1b tsheA - sa1 ɕa1 sa1’ - -
2. new *tshji̯ əŋ xhi1 ɕɛ1 tʂhia1 sæin1b senA ŋkheŋ1’ sɪ1̃ ɕaŋ2 saŋ2 saŋ1 sjaŋ1
3. ashes *tshju̯ əiX ɕhu3 ɕi3 tʂhau3 sɔ3b tshuB θe3 ɕi3 ɕwai3 saːi3 swai3 sɔi3

PH *tshj- 1 2 3 4 5 6 7
4. spinning wheel *tshju̯ aA - - tʂhua1 - tshaA - -
5. to sieve *tshjɔC - ɕɔ5 tʂhau5 - - - -

Notes
2. The velar initial in Jiongnai (6) is not figured into the reconstruction of this onset. It may
reflect more than the absorption of a prefix, however; see initial 2.55 ff., where a
prenasalized velar is the regular reflex of the *rj- series in Jiongnai. This may be from a
variant of Chinese 清 ‘clear’ (OC *N-s-r̥ eŋ > MC tshjeng > Man. qīng), given its close
resemblance to pre-nasalized *ntshji̯ əŋ ‘clear’ (3.20/18); see chapter 4.
3. Chinese 灰 ‘ashes’ (MC xwoj > Man. huī) has a very different onset and a different
tone, but the rime corresponds.
4. The meaning of this word in White Hmong is ‘machine’. It is probably from Chinese 車
‘cart, chariot, thing with wheel’ (MC tsyhæ > Man. chē).

3.18 *dzj-
PHM *dzj- 1 2 3 4 5 6 7 8 9 10 11
1. few/lack *dzjok ɕu6 - tʂaɯ6 - zoC ʃau6 ʨɤ6 θo8 tθɔ6 tshɔ8 hu8
2. porcupine *N-dzjeiH ɕen6 ʨi6 tʂau6 so6 zuC ʃa6 ʨi6 dei6 dei6 - hɛi6
 Proto Hmong-Mien 75

PH *dzj- 1 2 3 4 5 6 7
3. bland/tasteless *dzjæC ɕen6 - tʂua6 səa6 ziC ʃi6 ʨi6
4. chin/lower jaw *dzjeD ɕi1 ʨɑ8 tʂai8 se8 zeD - -
5. to wake *dzjaD - - tʂi8 si8 zaD - ʨi8

Notes
3. Tan Trinh Pa-Hng (Niederer 1997) has been substituted for Baiyun Pa-Hng (7).

3.19 *ntsj-
PHM *ntsj- 1 2 3 4 5 6 7 8 9 10 11
1. salt *ntsjəuX ɕi3 ȵʨɯ3 ɳtʂe3 ntsæ3a nʔtsiB - ȵo3 dau3 3 3
dau dza -
2. to choke *ntsjeŋH - - - - - ɳtʃei5 ni5 ʥɛːŋ5 - ʥɛ5 jaŋ5
3. to stop up/plug
- ntshu7 ɳtʂaɯ7 ntsəu5a nʔtsuD ɳtʃau7 sɔ7 tsot7 - - -
*ntsjot
4. 眨 to wink *ntsjep shei 7
- ɳtʂai 7
ntse7
nʔtse D
- ta 7
dzjop 7
- hin 7
-

PH *ntsj- 1 2 3 4 5 6 7
5. to drive in (nail)
- ȵʨɛ5 ɳtʂia5 - nʔtsenC - ni5
*ntsjinC
6. to soak *ntsjæwC - ȵʨɯ5 ɳtʂe5 ntsæ5a nʔtsiC ɳtʃa5 ȵʨo5
7. sorghum *ntsjuC - - ɳtʂua7 ntsa7a - - -
8. 畟 sharp *ntsjæwC - - ɳtʂe5 ntsæ5a - - -

Notes
2. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
3. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). Compare
Cantonese /tsat/, Hainan Mun /tjat7/ ‘to plug (a hole)’.
4. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
6. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
7. Upon inspection of more cognates, it becomes clear that tone 7 in these two forms for
‘sorghum’ are tone sandhi forms derived from tone 5.
8. Probably from Chinese 畟 ‘sharp-cutting’ (OC *[tsʰ]rək > MC tsrhik > Man. cè), given
the parallel correspondence between Hmongic *ndzjæwC ‘vertical/upright’ (3.21/3) and
Chinese 直 ‘straight/vertical’ (OC *N-t<r>ək > MC drik > Man. zhí).

3.20 *ntshj-
PHM *ntshj- 1 2 3 4 5 6 7 8 9 10 11
1. blood *ntshjamX ɕhaŋ3 ȵʨhi3 ɳtʂha3 ntsua3b nʔtshenB θi3 ȵ̥e3 ɕam3 saːm3 san3 dzjɛm3
2. head louse *ntshjeiX ɕhu3 ȵʨhi3 ɳtʂhau3 ntsɔ3b nʔtshuB - ʨhi3 ɕei3 tθei3 - dzɛi3
3. to fear/be afraid
ɕhi1 ȵʨha5 ɳtʂhai5 ntse5b nʔtsheC ɳtʃei1 ȵ̥ɛ5 ʥa5 ȡa5’ ȡa5 dzjɛ5
*ntshjeH
4. 清 clear *ntshji̯ əŋ ɕhi1 - ɳtʂhia1 ntsæin1b nʔtsheA - nɪ1̃ dzaŋ1 - - -

PH *ntshj- 1 2 3 4 5 6 7
5. otter *ntshju̯ aA ɕha1 ȵʨhɑ1 ɳtʂhua1 ntsa1b nʔtshaA θja1 -
76 Chapter 2

Notes
1. Compare Mon-Khmer *jhaam ‘blood’ (Shorto #1430).
2. Tan Trinh Pa-Hng (Niederer 1997) has been substituted for Baiyun Pa-Hng (7).
4. Xiangnan Mien (Wang and Mao 16) has been substituted for Luoxiang Mien (8). This
may also have been borrowed as Hmong-Mien *tshji̯ əŋ ‘new’ (3.17/18); see chapter 4.

3.21 *ndzj-
PH *ndzj- 1 2 3 4 5 6 7
1. to stumble *ndzjowD - - - - - - -
2. 直 vertical/upright
- - ɳtʂe6 - - - -
*ndzjæwC

Notes
1. This word is only represented by two West Hmongic languages in Wang 1994: Xianjin
/ɳtʂeu8/ and Shimen /ɳdʐɦey8II/.
2. This word is only represented by two other West Hmongic languages in Wang 1994:
Xianjin /ɳtʂe6/ and Shimen /ɳdʐə6II/.
 Proto Hmong-Mien 77

4. Palatals

4.1 *c-
PHM *c- 1 2 3 4 5 6 7 8 9 10 11
1. beeswax *N-ci̯ æŋ ʨaŋ1 ʨi1 cia1 ʨua1a ʨenA - - - - ʥjɛ1 -
2. earthworm *N-cuŋ ʨoŋ1 coŋ1 ca1 ʨaŋ1a ʨoŋA jɐŋ1 jõ1 duŋ1 - hwjə1 dziŋ1
3. wind *N-cæwH ʨen5 ci5 cua5 - ʨiC tʃi5 ʨi5 ʥau5 ȡaːu5 hja5 dzau5
4. 捷 quick *cu̯ ɛp - - - - - - ʨe7 tɕep7 - - -

PH *c- 1 2 3 4 5 6 7
5. to lead *coŋA - ʨoŋ1 ca1 ʨaŋ1a ʨoŋA - -
6. thrush (huà méi
ʨu1 ʨu1 cɔŋ1 ʨaŋ1a ʨoŋA tʃwaŋ1 ʨɔ̃1
niǎo) *cɔŋA
7. to trap/ensnare *cuA - - cua1 ʨæ1a ʨaA - -
8. body/trunk *cæwB ʨi3 ʨɯ3 ce3 ʨæ3a ʨiB - -
9. to carry on shoulder
ʨi5 - - ʨæ5a - - -
*cæwC
10. 酒 liquor *cowB ʨu3 ʨɯ3 caɯ3 ʨə3a ʨoB tʃau3 ʨɤ3
11. 炙 to bake/toast
ʨen5 - ci5 ʨi5a ʨiC - -
*ciC
12. 嚼 to chew *cu̯ aC ʨu 5
ca 5
cua 5
ʨa5a
ʨa C
tʃa 5
ʨa5

PM *c- 8 9 10 11
13. crop (of bird)
ʨiŋ3 ȡiŋ3 - -
/gizzard *ciŋB
14. to bark *cuŋC ʨuŋ5’ ȶuŋ5 ȡwə5 -
15. 獐 deer *cuŋA ʨuŋ1 ȶuːŋ1 - -

Discussion
In this series most of the Hmongic languages have palatal affricates and some Mienic
languages have velar stops. Since palatal stops tend to be released with some degree of
friction given the wide area of contact between the tongue and hard palate, the
development from *c- to /ʨ-/ is quite natural (as is the development from *c- to /k-/,
another way to resolve an unstable *c-). Developments of *ʨ- to either /c-/ or /k-/ would
be more unusual.

Notes
3. Jiwei (2) is taken from Office of Miao-Yao Research Lexicon (1987).
8. This word appears to be morphologically related to Hmongic *ɟæwB ‘leg/branch’
(4.3/3). The difference in initial voicing may have been due to a prefix; see chapter 4.
10. See Mienic *tiuB (2.1/A), borrowed independently from the same word.
11. Morphologically related to Hmongic *ɟiC ‘to burn/be alight’ (4.3/1); see chapter 4.
13. Downer (1973) gives Cantonese /khan/ ‘gizzard’ as the source of Mien /kien1/,
although this is not obviously the same word.
78 Chapter 2

4.2 *ch-
PHM *ch- 1 2 3 4 5 6 7 8 9 10 11
1. 穿 to thread *chu̯ en ʨhaŋ1 ʨhɑŋ1 chɔ1 ʨoŋ1b ʨhoŋA tʃui1 - ɕwən1 sen1’ ȶhwən1 tsui1
2. 處 bed/place
ʨhu5 - chaɯ5 - ʨhoC - ʨhõ2 ɕou5 ȶɔu5’ ȶhəu5 fu5
*chouH

PH *ch- 1 2 3 4 5 6 7
3. to sweep *chæwA ʨhi1 - che1 ʨæ1b ʨhiA - -
4. belly *chu̯ eiA ʨhu1 ʨhi1 chau1 ʨɔ1b ʨhuA the1 -
5. CLF-rooms *choŋB ʨhoŋ3 - cha3 ʨaŋ3b khoŋB - -
6. to peel/shave *cheD - - chai7 tɕe7b tɕheD - -

PM *ch- 8 9 10 11
7. 炒 to fry *chauB ɕaːu3 ȶaːu3 ȶha3 hau3

Notes
2. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
3. Mienic *ʔɟæuC ‘to sweep’ (4.4/3) has the same place of articulation and rime, but differs
in tone and aspiration.
4. In White Hmong, cognate /chau1/ means ‘to crawl on one’s belly’ (compare the verbal
use of ‘belly’ in English: ‘to belly up to the table’).
5. This bears a very slight similarity to Chinese 間 ‘room’ (MC kɛŋ > Man. jiān).

4.3 *ɟ-
PHM *ɟ- 1 2 3 4 5 6 7 8 9 10 11
1. male(human) *ɟi̯ aŋH ʨaŋ6 - - ʨua6 - ɳtʃi6 - ʨaŋ2 - - kjaŋ2
2. knee *ɟu̯ əiH ʨu6 ʨɔ6 cau6 ʨo6 ʑuC - ʨi6 ʨwai6 - ȶwai6 -
3. nine *N-ɟuə ʨə2 ʨo2 cua2 ʨa2 ʑaA tʃu2 ko5 du2 du2 ju2 ku2
4. 蕎 buckwheat *ɟæu - - ce2 ʨæ2 ʑiA - - ʨou2 - ȶɔ4 -
5. 茄 eggplant *ɟa ʨa2 - - ʨəa6 - - - ʨe2 - ȶa2 kjɛ2
6. 橋 bridge *ɟow ʨu2 - chɔ2 - - tʃi2 ʨe6 ʨou2 ȶɔu2 ȶəu2 ku2
7. 騎 to ride *ɟej ʨi2 - cai2 ʨe2 ʑeA tʃei2 ʨhi2 ʨei2 ȶei2 ʨi2 ki2

PH *ɟ- 1 2 3 4 5 6 7
8. alive *ɟinA - - cia2 ʨein2 ʑiA - -
9. root *ɟoŋA ʨoŋ2 ʨoŋ2 ca6 ʨaŋ6 ʑoŋA - kõ2
10. to finish *ɟɔŋB ʨu4 ʨu4 - ʨəŋ4 - - -
11. to grow (TR) *ɟiB ʨen4 - - - ʑeB tei4 -
12. leg/branch *ɟæwB ʨi4 kɯ4 ce6 ʨe4 ʑaB - -
13. to meet *ɟu̯ aC ʨa6 - cua6 ʨəa6 ʑaC - -
14. to plant *ɟəŋC - ʨɑŋ6 cɔ6 ʨoŋ6 - - -
15. 臼 mortar/treadmill
ʨə4 ʨo4 cɔ4 ʨu4 ʑiB - -
*ɟoB
16. 炙 to burn/be alight
ʨen6 - ci6 ʨi6 ʑeC - -
*ɟiC

PM *ɟ- 8 9 10 11
17. 茶 tea *ɟaA ʨa2 ȶa2 ȶa2 ta2
18. 韭 Allium *ɟiuB ʨiu4 - - kiu4
 Proto Hmong-Mien 79

Notes
1. In White Hmong, the cognate may be /ca6/ in the phrase ob cag /ʔɔ1 ca6/ ‘others’.
3. Benedict (1987a:14) attributed the voiced initial in Hmong-Mien ‘nine’ to a Tibeto-
Burman coronal prefix (*d-kəw). The use of a pre-element *N- here, as elsewhere in the
reconstruction, does not account for the earliest voicing of the root initial; a voiced initial
must be reconstructed for Proto Hmong-Mien to account for the lower-register tone. *N-
merely indicates the loose adjunction of something that voiced the initials in Mienic at
some later point, after the early voicing contrast had been lost. This pre-initial element may
have been something other than a nasal consonant.
4. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
5. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
7. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
9. This may be related to Mienic *ndzuŋA ‘root’ (3.6/28), but if so, there must have been a
pre-initial that caused voicing in Mienic (the Hmongic initial is not prenasalized).
12. This word appears to be morphologically related to ‘body/trunk’ above: *cæwB (4.1/3).
The difference in initial voicing alone may have been due to a prefix; see chapter 4.
16. Morphologically related to Hmongic *ciC ‘to bake/toast’ above (4.1/1); see chapter 4.
18. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).

4.4 *ɲc-
PHM *ɲc- 1 2 3 4 5 6 7 8 9 10 11
1. pillow *ɲcu̯ əmH - ȵʨɯ5 ɲcɔŋ5 ȵʨaŋ5a ȵʔʨoŋC ntjoŋ5 ȵɔ5 dʑum6 ȡam5 - -
2. 竹 bamboo strip
ʨen5 ȵʨɔ5 ɲcau5 ȵʨɔ5a - ntjeu5 - dʑu7 ȡu7 ȡu7 dziu7
*ɲcəuk

PH *ɲc- 1 2 3 4 5 6 7
3. to climb *ɲcæwC ʨi5 ȵʨɯ5 ɲce5 ȵʨæ5a ȵʔʨiC ntja5 -
4. to remember *ɲcoC - - ɲcɔ5 ȵʨo5a ȵʔʨuC - -
5. to peck/dig *ɲcowD ʨu7 ȵʨu7 ɲcaɯ7 ȵʨə7a ȵʔʨoD - ʨɔ1
6. 正 straight *ɲci̯ aŋA - - ɲca1 ȵʨaŋ1a ȵʔʨaŋA ntjeu1 -

PM *ʔɟ- (<*ɲc- ) 8 9 10 11
7. thorn *ʔɟimB jim3 ȡim3 - -
8. to sweep *ʔɟæuC - ȡɔu5 - -
9. 夾 to pick up *ʔɟəpD dʑap7 - kan7 -

Notes
1. Perhaps from a C-tone variant of Chinese 枕 ‘pillow’ (OC *[t.q][ə]mʔ > MC tsyimX >
Man. zhěn).
5. There is a similar word meaning ‘to peck’ in Mienic: Jiangdi Mien /dzo7/ (Wang and
Mao 15) and Xiangnan Mien /dzəu7/ (Wang and Mao 16). However, the rimes do not
correspond. The word is listed twice in Wang and Mao, once as ‘to peck’ and once as ‘to
dig’ because in Yanghao (1) ‘to peck’ has tone 5 and ‘to dig’ has tone 7. The forms in all
other locations with these meanings are identical, which suggests the differentiation is
limited to Yanghao. It is also used in White Hmong to mean ‘to hoe/clean the ground’.
Compare Chinese 撅 ‘to dig’ (OC *[g]ot > MC gjwot > Man. jué).
80 Chapter 2

6. In White Hmong (3),‘to straighten’. ‘To be straight’ in both this language and in Xianjin
Hmong (Wang and Mao 3) is the same word with tone 2, indicating either a morphological
contrast in Chinese (and borrowing of both forms) or an innovated morphological contast
within the parent of these two closely-related languages: it is impossible to tell which is
more likely (see chapter 4).
8. Hmongic *chæwA ‘to sweep’ (4.2/3) has the same place of articulation and rime, but
differs in tone and aspiration.
9. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).

4.5 *ɲch-
PHM *ɲch- 1 2 3 4 5 6 7 8 9 10 11
1. cooking smoke 5 5 5b B 5 5’
- ȵʨho ɲchɔ ȵʨo ȵʔʨhu - - ɕeu sɔu - -
*ɲchi̯ ouH

PH *ɲch- 1 2 3 4 5 6 7
2. to sprinkle/spill 3 3 3b B
- ȵʨho ɲchua ȵʨa ȵʔʨha - -
*ɲchuB

PM *ɟh (< *ɲch-) 8 9 10 11

3. ant *ɟhi̯ əuB ɕeu3 sɔu3 sau3 dziu3

Notes
1. Also ‘to emit smoke’ or ‘steam’.
2. Also ‘to pour out’.

4.6 *ɲɟ-
PHM *ɲɟ- 1 2 3 4 5 6 7 8 9 10 11
1. mouth/beak *ɲɟuj ȵu2 ȵɔ2 ɲcau2 ȵʨu2 ȵʨoA ntjɔ2 ȵʑɯ2 dzuːi2 - ʨi2 dzi2
2. pillar *ɲɟæu ȵi2 ȵɯ2 ɲce2 ȵʨæ2 ȵʨiA - ŋo2 - - ʨɔu2 -

PM *ɲɟ- 8 9 10 11
3. to water *ɲɟunA dzun2 - - -
4. sword *ɲɟəukD dʑu8 ȡu6 ȡu8 dziu8

Notes
1. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8), and
Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min (10).
Perhaps from Chinese 嘴 ‘beak’ (OC *[ts]ojʔ > MC tsjweX > Man. zuǐ).
2. Yanghao (1) is from Office of Miao-Yao Research Lexicon (1987). Sunjiang Biao Min
(Wang and Mao 22) has been substituted for Dongshan Biao Min (10).
3. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).

4.7 *ʔɲ-
PHM *ʔɲ- 1 2 3 4 5 6 7 8 9 10 11
1. sister-in-law *ʔɲam ȵaŋ1 ȵi1 ȵa1 ȵa1a ʔȵenA ȵi1 - - ȵaːm1 ȵan1 ȵɛm1
2. to weep *ʔɲæmX ȵaŋ1 ȵɛ3 ȵia3 ȵi3a ʔȵenB ȵiŋ1 - ȵem3 ŋjim3 ȵan3 ȵɛm3
 Proto Hmong-Mien 81

PH *ʔɲ- 1 2 3 4 5 6 7
3. to be at/live *ʔɲəŋA ȵaŋ1 ȵi1 ȵɔ1 ȵɔ1a ʔȵoŋA ȵɐŋ1 ȵõ1

PM *ʔɲ- 8 9 10 11
4. POSSESSIVE *ʔɲi̯ əŋA ȵaŋ1 ȵaŋ1 nin2 -
5. to crawl *ʔɲɛɔŋA ȵɔŋ1 ȵɔːŋ1 - -
6. shy *ʔɲæiA/B ȵai3 ȵai1 - -
7. claw/talon *ʔɲauB ȵau1 ȵaːu3 - -

Discussion
Although prepalatal nasals have been recorded for almost all locations for this series,
simple palatal nasals have been reconstructed here, in keeping with the reconstructions for
all initials in the palatal series. The goal of the reconstruction is not to establish precise
phonetic characteristics of these nasals, or of any other segments in the protolanguage; it is
rather to define contrastive places of articulation.

Notes
1. Also ‘daughter-in-law; bride’. Compare Tibeto-Burman *s-nam ‘daughter-in-law’
(Benedict 1987a:18).
3. Mienic *ʔjəmA (4.10/21) ‘to be at/live’ is probably related to this word, but the absence
of the nasal initial is hard to explain (Wang and Mao explain it as function word
reduction); see chapter 4.
5. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
6. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
7. Downer (1973:30) relates this word to Amoy /jiaù/ ‘claw’.

4.8 *hɲ-
PHM *hɲ- 1 2 3 4 5 6 7 8 9 10 11
1. intestines *hɲeuŋX - ɕe3 ȵ̥u3 ȵoŋ3b ȵ̥enB ȵ̥ɔ3 ŋ̥ŋ3 ȵ̥eu3 ȵɔu3 - -
2. year *hɲu̯ əŋH ȵ̥hu5 - - - - ȵ̥waŋ5 - ȵaŋ5 ȵaŋ5’ ȵ̥aŋ5 ȵaŋ5

PH *hɲ- 1 2 3 4 5 6 7
3. forehead *hɲenA ȵ̥haŋ1 ɕɛ1 - ȵi1b - - nɪ1̃

PM *hɲ- 8 9 10 11
4. to forget *hɲouB - ȵɔu3 n̥ au3 nau3
5. to carry on back
ȵa5’ ȵa5 ȵa3 -
*hɲaC
6. to squeeze *hɲaiC ȵai 5’
ȵai5
- -
7. to winnow *hɲumC ȵum5 ȵum5’ n̥ ən5 num5

Notes
2. Since the tones and rimes correspond exactly, this is likely to be the same as Hmongic
‘year’ *ɕɔŋC (4.13/29), perhaps both from Tibeto-Burman *s-niŋ ‘year’, with prefix pre-
emption in Hmongic (Mortensen 2002). Compare Chinese 年 ‘year’ (OC *[n]ˤi[n] > MC
nen > Man. nián).
4. Compare Hmongic *hnuŋB ‘to forget’ (2.8/27).
5, 6, & 7. The tone subcategory indicated by an apostrophe suggests that these words
should be reconstructed with *hɲ- rather than *ʔɲ-, as it appears in Wang and Mao 1995.
82 Chapter 2

4.9 *ɲ-
PHM *ɲ- 1 2 3 4 5 6 7 8 9 10 11
1. raw/unripe *ɲu̯ emX ȵu4 ȵu4 ȵɔŋ4 - ȵoB ȵwaŋ4 ȵõ4 - ŋjim4 - -
2. to steal *ɲemH ȵaŋ6 ȵɛ6 ȵia6 ȵi6 ȵenC ȵiŋ6 ȵɪ6̃ nim6 neːm6 - ȵɛm6

PH *ɲ- 1 2 3 4 5 6 7
3. shallow *ɲenB ȵi4 - - ȵi4 - - ȵɦɪ4
4. 姌 thin *ɲenB ȵaŋ4 ȵɛ4 ȵia4 ȵi4 ȵenB ȵiŋ4 nɪ4̃
5. 髯 beard *ɲaŋC ȵaŋ6 ȵi6 - ȵa6 ȵenC ȵi6 nɪ6̃

PM *ɲ- 8 9 10 11
6. to dye *ɲumC ȵum6 ȵum6 ȵin6 ȵam6
7. to eat *ɲənC ȵen6 ŋjin4 ȵin6 ȵan6

Notes
3. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
6. This word for ‘to dye’ is attested in both the Tai (Lao /njɔm/) and Kam-Sui (Mak
/njum/) branches of Tai-Kadai (Downer 1978). Also compare Vietnamese nhuộm. All are
probably from Chinese 染 ‘to dye’ (MC nyemX > Man. rǎn).
7. Hmongic *nuŋA ‘to eat’ (2.9/27) is probably related.

4’.9 *ɲʷ-
PHM *ɲʷ- 1 2 3 4 5 6 7 8 9 10 11
1. 銀 silver *ɲʷi̯ ən ȵi2 ŋoŋ2 ȵia2 ȵi2 ʔȵenA ȵiŋ2 ŋ̩2 ȵwan2 ȵaːn2 ȵan2 ȵan2

4.10 *ʔj-
PH *ʔj- 1 2 3 4 5 6 7
1. mosquito *ʔjɔŋB ʑu3 - jɔŋ3 ʑaŋ3a ʔʑoŋB - jõ3
2. to fly *ʔjaŋC ʑaŋ5 ʑi3 ja5 ʑua5a ʔʑenC i5 je5
3. 秧 seedling *ʔjɛŋA ʑi1 ʑɑŋ1 ju1 ʑoŋ1a ʔʑaŋA jaŋ1 jɔ̃1
4. 要 to ask for/want *ʔjuB ə3 - jua3 ʑa3a - wjaŋ3 ja3
5. 幼 small/young *ʔjuwC ʑu5 ʑɔ5 jau5 ʑu5a ʔʑoC ju5 jɯ5
PM *ʔj- 8 9 10 11

6. to be at/live *ʔjəmA jem1 jam1 jan1 -

7. 1SG (I/me) *ʔjaA ja1 ja1 - -
8. to spill *(ʔ)jamC ɕaːm5 jaːm6 ja6 -
9. 秧 seedling *ʔjaŋA jwaŋ1 jaːŋ1 jɔ1 -

Notes
2. Also ‘to run’ in Pa-Hng (Niederer 1997:120).
3 & 9. Since the rimes don’t exactly correspond, these were probably separate borrowings
from Chinese 秧 ‘seedling’ (MC ʔjang > Man. yāng).
4. The Chinese and Hmongic tones do not correspond.
6. Hmongic *ʔɲəŋA ‘to be at/live’ (4.7/21) is probably related to this word; see chapter 4.
 Proto Hmong-Mien 83

8. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). This
word has reflexes with both tone 5 and tone 6; it could equally well be reconstructed
*jamC.

4.11 *hj-
PHM *hj- 1 2 3 4 5 6 7 8 9 10 11
1. to shrink *hjuk xhu7 - - - - - - ɕo7 - - -

Discussion
It is not clear whether *hj- should be listed as a voiceless palatal glide (as here), or as an h-
onset with a -j- medial. Given the prominence of voiceless sonorant initials in Hmong-
Mien it has been placed here, but the decision is fairly arbitrary.

Notes
1. See also Xianjin Miao /xau7/, Jiangdi Mien /su7/, and Changping Mien /ðut7/. The final -
t in Changping Mien and the Hmongic tones suggest that this word ended in -t, not -k. But
the word is otherwise similar to Chinese 縮 ‘to shrink’ (OC *[s]ruk > MC srjuwk > Man.
suō), so a -k is tentatively reconstructed here.

4.12 *j-
PHM *j- 1 2 3 4 5 6 7 8 9 10 11
1. uncle *jæuX ʑu5 ʑo4 - - - ja2 jo4 jou4 jɔu4 - -
2. to lick *-jep ʑi8 ʑɑ8 jai8 ʑe8 - ŋklai8 jɛ8 - - - dzjɛp8
3. eight *jat ʑa8 ʑi8 ji8 ʑi8 ʑaD je8 ji8 jat8 jet8 hjɛn8 dzjat8
4. 羊 sheep/goat *juŋ ʑoŋ2 ʑoŋ2 ja2 ʑaŋ2 ʑiA jɐŋ2 - juŋ2 juːŋ2 wjə2 dziŋ2

PH *j- 1 2 3 4 5 6 7
5. tree of heaven *jɛŋA - - ju2 ʑoŋ2 ʑaŋA - -
6. cow/ox *joB - ʑu4 - - ʑuB - -
7. whorl at top of head
ʑi6 - ji4 ʑi4 weB - -
*jɨB
8. to be *jəŋC - - jɔ6 ʑoŋ6 ʑoŋC - ȵɯ6
9. 煬 to dissolve/melt
ʑaŋ2 ʑi2 ja2 ʑua2 - ȵi2 je2
*jaŋA
10. 楊 poplar tree *ji̯ aŋA - - - ʑaŋ2 - - -
11. 養 to raise (chickens)
ʑi6 - ju6 ʑoŋ6 ʑaŋC jɐŋ6 -
*jɛŋC

PM *j- 8 9 10 11
12. armspan *ju̯ amA jwam2 jɔm2 jaŋ2 dzjam2
13. to walk *jəŋA jaŋ2 jaŋ2 ȵaŋ2 dzaŋ2
14. to dissolve *jəukD ju8 ju6 - -
15. 煬 to dissolve/melt
- - wjə2 dziŋ2
*juŋA
16. 養 to raise (chickens)
juŋ4 juŋ4 wjə4 dziŋ4
*juŋB
84 Chapter 2

Notes
1. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
2. The complex onset in Jiongnai (6) resembles the Jiongnai onsets in sets 2.55, 2.56, 2.57
(*ʔr-, *hr-, *r-). As with certain words in those sets, the hyphen represents the presence of
pre-initial matter.
3. From Tibeto-Burman *-rjat ‘eight’ (Matisoff 2003) See chapter 5.
5. Ailanthus altissima.
7. In White Hmong, ‘cowlick’ (an errant patch of hair growing anywhere).
9 & 15. The same word borrowed independently into Hmongic and Mienic. Compare #14
above, which is similar.
11 & 16. Hmongic from Chinese yàng ‘to keep, support’ and Mienic from Chinese yǎng
‘to nourish, raise’ (same character).
12. There is a word for ‘span from thumb to index finger’ with a palatal nasal that also
seems to have been borrowed from this word.

4.13 *ɕ-
PHM *ɕ- 1 2 3 4 5 6 7 8 9 10 11
1. to stretch *ɕoŋ ɕhoŋ1 - ɕa1 - - - ɕõ1 ɕoŋ1 - - -

PH *ɕ- 1 2 3 4 5 6 7
2. to breed/propagate
ɕhaŋ5 - ɕa5 - - - -
*ɕaŋC
3. year *ɕɔŋC - ʨu5 ɕɔŋ5 ʨaŋ5a - - ʨɔ̃5

PM *ɕ- 8 9 10 11
4. winnowing fan *ɕi̯ əŋA ɕaŋ1 saŋ1’ ȶaŋ1 -
5. to twist/rub *ɕætD ɕat7 saːt7’ saːt7’ sjɛt7
6. 媳 daughter/girl *ɕi̯ ɛkD ɕa7 sa7’ sa7 sa7

Discussion
The Mun (9) reflexes of voiceless fricatives fall into two large sets: those with tθ- and
those with s-. The words with a Mun reflex of tθ- have been placed together and have been
reconstructed at both the Proto Hmong-Mien and Proto Mienic level as *s- (2.13). The
words with a Mun reflex of s- have been reconstructed at the Proto Hmong-Mien level as
either *sj- (2.28) or *ɕ- (here), merging at the Proto Mienic level as *ɕ-. The distinction
between Hmong-Mien *sj- and *ɕ- is based on developments in Western Hmongic: *sj- >
ʂ- and *ɕ- > ɕ-.
The messiness of the fricative correspondences led Wang and Mao to reconstruct eleven
voiceless fricatives for the protolanguage, but since the daughter languages are not
characterized by large numbers of fricatives, it seems reasonable to assume that the
protolanguage did not have many fricatives, either. There are many Chinese loanwords in
these sets, some with unique patterns, that obscure the native correspondences.

Notes
3. The onset of this Hmongic word for ‘year’ is highly irregular: rather than showing either
clear fricative reflexes of *ɕ- or stop/affricate reflexes of *c- , it shows a mixture of the
two. It may be from the same source as the Hmong-Mien *hɲu̯ əŋH (4.8/29) ‘year’, perhaps
 Proto Hmong-Mien 85

both from Tibeto-Burman *s-niŋ ‘year’, with prefix pre-emption in this set of forms
(Mortensen 2002). Compare Chinese 年 ‘year’ (OC *[n]ˤi[n] > MC nen > Man. nián).
4. Probably from Chinese 扇 ‘fan’ (MC syenH > Man. shàn).
5. Hainan Mun (Shintani and Yang 1990) has been substituted for Lanjin Mun (9). This
word is probably the same as Hmongic *S-phu̯ aA ‘to twist/rub’ (1.2/15).
86 Chapter 2

5. Velars

5.1 *k-
PHM *k- 1 2 3 4 5 6 7 8 9 10 11
1. younger brother *ku̯ eiX - kɯ3 kɯ3 kou3a kuB - - - - ki3 kui3

PH *k- 1 2 3 4 5 6 7
2. ditch *koŋA koŋ1 - - kaŋ1a koŋA kjɐŋ1 -
3. 1SG (I/me) *kɛŋB - - ku3 koŋ3a kaŋB - -
4. bamboo hat *kuwD - ku7 kau7 ku7a koD - kɯ7
5. to gnaw *kowD ki7 ku7 kaɯ7 kə7a koD - -

PM *k- 8 9 10 11
6. to cover *komB kom3 kɔm3 - -

Discussion
Hmong-Mien *velars merged with uvulars in Hmongic unless followed by back rounded
vowels, -l- , or -r-, and in Mienic, all Hmong-Mien *uvulars merged with velars (see this
chapter, section 1.2).25 Since words in the *k- and *q- series proper only contrast in these
three restricted environments, a number of correspondence “sets” in series 5 and 6 will
only contain one or two words. Words reconstructed with a dorsal consonant before any
vowel other than a back rounded vowel are placed in the 5-6 series, which represents an
undifferentiated dorsal (here written with a capital letter: *K- *KH-, *G-, etc.).

Notes
1. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Quanzhou Biao Min
(10).
2. Compare Biao Min /khlɛ1/ ‘ditch, ravine, mountain stream’ (Solnit 1982). Zao Min (11)
/ku1/ appears to have been borrowed from Chinese 溝 ‘ditch, irrigation canal’ (MC kuw >
Man. gōu).
3. This is the Hmong-Mien first person singular pronoun with the widest distribution; see
chapter 3. Although it appears to be cognate to Dongshan Biao Min /kəu3/ ‘I/me’ (Wang
and Mao 21) and Sunjiang Biao Min /kɔ3/ ‘I/me’ (Wang and Mao 22), it is not possible to
place these forms into any existing Hmong-Mien rime set. The Proto Hmong-Mien form
must have had an initial *k- and a back rounded vowel, however, since the *k- did not
retract to q- in Hmongic; see chapter 2, section 1.2.
6. This word for ‘to cover’, as well as Mienic *hi̯ əmB/C ‘to cover (with tile)’ (7.13/18),
appear to be recent borrowings from Tai-Kadai. Li (1977) reconstructs a number of similar
words with this general meaning: *xr[um]B1 ‘to cover, veil; *ɣ[um]B2 ‘to cover, protect’;
*homC1 ‘to cover up’. See also the Mienic forms under ‘to cover’ with initial g- in L-
Thongkum’s reconstruction: Mien /gom3/, Mun /gəm3/, etc. (1993:204).

25
Medial -j- also prevented a *velar from retracting in Hmongic, but since there are no contrasting *kj- and
*qj- correspondence sets, it is assumed that there was no *qj- series in Proto Hmong-Mien.
 Proto Hmong-Mien 87

5.2 *kh-
PH *kh- 1 2 3 4 5 6 7
1. plum *khowA - - khaɯ1 kə1b khoA - -
2. to pick up *khowD ʨhə1 - khaɯ7 kɯ7b khoD - -

PM *kh- 8 9 10 11
3. good/tasty *khuB - ku3 - kɔm3
4. 開 to open *khu̯ ɔiA khɔi1 kɔːi1’ khwai1 -
5. 堀 hole *khotD khot7 kɔt7 khwan7 -

5.3 *g-
PH *g- 1 2 3 4 5 6 7
1. 2SG (you) *gʉA - - kɔ2 - - - -
2. 蝗 grasshopper *gʉŋA ku - kɔŋ kaŋ ɣoŋ - -

Notes
1. Also attested in West Hmongic Xianjin /kau2/, Shimen /ʨy6/ and Bunu /kau2/.
1 & 2. It is clear that central rounded /ʉ/ patterns as a back rounded vowel, since velars
before vowels other than back rounded vowels retracted to uvulars in Hmongic.

5.4 *ŋk-
PH *ŋk- 1 2 3 4 5 6 7
1. muddy *ŋkoD - - ŋkɔ7 ŋko7a - - ŋa7

Notes
1. Hainan Mun /glɔɔ6/ ‘muddy’ looks similar (Shintani and Yang 1990), and also fits with
the forms under Mienic *ŋglokD ‘turbid’ (5.36/13). This word may involve expressive
phonology.

5.5 *ŋkh-
PH *ŋkh- 1 2 3 4 5 6 7
1. soot *ŋkhowA ʨhu1 - ŋkhaɯ1 ŋkə1b ŋʔkhoA - -
2. crooked *ŋkhuwD - ŋkhu7 ŋkhau7 ŋko7b ŋʔkhaD - ŋɔ1

Notes
2. Compare Chinese 曲 ‘bend, crooked’ (MC khjowk > Man. qū), Tibeto-Burman *kuk
(Matisoff 2003), and Malayo-Polynesian *buŋkuk (ACD). Tone 7 in Hmongic points to a
final *-p or *-t, rather than *-k, however.

5.6 *ŋg-
PH *ŋg- 1 2 3 4 5 6 7
1. river deer *ŋgu̯ eiB - - - ŋko4 ŋkuB - -
88 Chapter 2

5.9 *ŋ-
PHM *ŋ- 1 2 3 4 5 6 7 8 9 10 11
1. 牛 water buffalo/
ȵen2 ȵe2 ȵu2 ŋoŋ2 ȵenA ȵɔ2 ŋ̩2 ŋoŋ2 ŋɔːŋ2 ȵuŋ2 ŋ̩2
cow *ŋiuŋ

Notes
1. Downer (1973:21) gives Chinese 牛 ‘ox/cow’ (OC *[ŋ]ʷə > MC ngjuw > Man. niú) as
the source, despite the fact that the word ends in a nasal consonant in Hmong-Mien.

5’.9 *ŋʷ-
PHM *ŋʷ- 1 2 3 4 5 6 7 8 9 10 11
1. 瓦 tile *ŋʷæX - - vua4 waə4 waB - wa4 ŋwa4 ŋwa4 ŋɔ4 ŋa4

5.13 *ɣ-
PM *ɣ- 8 9 10 11
1. 嫌 to dislike *ɣi̯ emA jem2 ȡiːm2 gjɛn2 -
2. 下 to descend *ɣaC ja8 ȡa6 ȡa6 ga6

5.16 *kj-
PHM *kj- 1 2 3 4 5 6 7 8 9 10 11
1. to lift *kjeŋ - - - - - - tɕõ1 ʨɛŋ1 ȶeːŋ5 kjɛ1 kaŋ1
2. medicine *N-kjaj tɕa1 ŋkɑ1 - ka1a kaA ðja1 ȵa1 - gjaːi1 hja1 tsɛi1
3. 金 gold *kjeəm tɕen1 ɲce1 ku1 koŋ1a tɕenA tʃeŋ1 tɕĩ1 tɕem1 sam1 ȶan1 kɛm1

PH *kj- 1 2 3 4 5 6 7
4. needle *kjɔŋA tɕu1 tɕu1 kɔŋ1 kaŋ1a koŋA tʃwaŋ1 kɔ1̃

PM *kj- 8 9 10 11
5. cold (day) *kju̯ ɛŋB tɕwəŋ3 sɔŋ3 ȶɔŋ3 kuŋ3
6. to remember *kji̯ əŋC ʨjaŋ5’ saŋ5 ȶaŋ5 kɛŋ5
7. 緊 tight *kjənB tɕen3 sen3 tɕin3 -
8. 句 CLF-sentences
ʨou5’ ȶɔu5 ȶəu5 kui5
*kjouC

Discussion
The palatal medial does not belong to the same phonological class as the other sounds
which block retraction to uvular (“back” sounds: back rounded vowels, -l-, and -r-).
Nonetheless, none of the correspondence sets suggests a contrast between a *kj- series and
a *qj- series. Palatal onsets are disproportionately common in the Chinese element of
Hmong-Mien: perhaps most (or all) of the words in this series were borrowed from
Chinese words after the retraction process had applied, or perhaps retraction only applied
to simple onsets for some reason.

Notes
4. Compare Chinese 箴 ‘needle’ (OC *t.q[ə]m > MC tsyim > Man. zhēn).
 Proto Hmong-Mien 89

5’.16 *kʷj-
PHM *kʷj- 1 2 3 4 5 6 7 8 9 10 11
1. 斤 catty *kʷjan ʨaŋ1 kɑŋ1 - ʨein1a ʨenA tʃoŋ1 kõ1 ʨwan1 san1 ȶwan1 tsan1

PM *kʷj- 8 9 10 11
2. sore/boil/blister
ʨwei2 sei2 - -
*kʷjejA
3. child *kʷjeiB ʨwei3
sei 3
hwjəi 7
-
4. 筋 sinew *kʷjanA ʨwan1 saːn1 ȶwan1 tsan1
5. 蕨 fern *kʷjətD ʨwət7 set7 ȶwan7 kɔt7

Discussion
The initial *kʷj- is most clearly reflected in a dialect of Mien not included above:
Changping Mien (Wang and Mao 18): ‘catty’ /kwjan1/, ‘child’ /kwjei3/, ‘fern’ /kwət7/.

Notes
3. Compare Pucheng (Min) /kiãi3/ ‘child’ (Norman 1988:243). This word may also be
reflected in the otherwise opaque second half of the White Hmong compound meaning
‘children’: tub-ki /tu1-ki5/ ‘son-(child)’.

5.17 *khj-
PHM *khj- 1 2 3 4 5 6 7 8 9 10 11
1. itch(y)/scratch(y)
tɕhu5 ɕi5 khau7 ko7b khoD ʃɔ7 ŋɯ7 ɕet7 set7’ hin7 kɛt7
*khju̯ ɛt

PH *khj- 1 2 3 4 5 6 7
2. shoes *khjɔC - ɕɔ5 khau5 kɔ5b khuC ʃu5 ŋkʉ5

PM *khj- 8 9 10 11
3. 薑 ginger *khjuŋA ɕuŋ1 tθuŋ1’ - kɔŋ1

Discussion
Pa-Hng (7) has a nasal initial for ‘itch(y)/scratch(y)’ and ‘shoes’. It is not clear whether
this is a retention or a secondary development.

Notes
2. In Jiongnai (6) and Pa-Hng (7), ‘straw sandals’. Ho Ne (Wang and Mao 14) gives
evidence of the medial -j- : /khju5/. Perhaps from Chinese 屨 ‘sandals, straw shoes’ (OC
*kro-s > MC kjuH > Man. jù), although the Chinese word is not aspirated. Compare
Mienic *səukD ‘straw sandals’ (2.13/6).
90 Chapter 2

5.18 *gj-
PHM *gj- 1 2 3 4 5 6 7 8 9 10 11
1. ten *gju̯ ɛp tɕu8 ku8 kau8 ku8 ɣoD tʃɔ8 kɯ8 ɕep8 sap8 ȶhan8 sjɛp8

PH *gj- 1 2 3 4 5 6 7
2. tea *gjiB tɕen4 ci4 - tɕi4 ʑiB tʃi4 tɕi4

PM *gj- 8 9 10 11
3. bracelet *gjəmA tɕem2 sam2 - -
4. steep *gjujB tɕui4 ȶui4 - ki4
5. 裙 skirt *gjunA - ȶun2 ȶwən2 kɔn2

Notes
1. A borrowing from either Tibeto-Burman *g(j)ip ‘ten’ or Chinese 十 ‘ten’ (OC *[g][i]p >
MC dzyip > Man. shí); see chapter 5, section 2.
2. Also West Hmongic /ki4/ (Qinyan, Gaopo, Fengxiang = Wang and Mao 5, 6, 9) and Ho
Ne /khji4/ (Wang and Mao 14).
3. Also Changping Mien /kjəm2/ (Wang and Mao 18). Compare Lungchou (Tai-Kadai)
/kiim2/ (Li 1977).

5’.18 *gʷj-
PHM *gʷj- 1 2 3 4 5 6 7 8 9 10 11
1. to get up *gʷjaX - - - - - tʃe4 - kja4 kwe4 ȶɔ4 kjɛ4

5.19 *ŋkj-
PHM *ŋkj- 1 2 3 4 5 6 7 8 9 10 11
1. mushroom *ŋkjæu tɕi1 ŋkɯ1 ɲce1 ȵtɕæ1a ȵʔtɕiA ɳtʃa1 ŋo1 tɕeu1 sɔu1 ȶau1 ku1
2. 粔 rice cake *ŋkjuəX tɕə3 - ɲcua3 ȵtɕa3a ŋʔkaB - ŋo3 ju3 ȡu3 ȡu3 gu3

PM *ʔgj- (< *ŋkj-) 8 9 10 11

3. 含 to hold in mouth
gɔm1 gjɔːm1 gan1 -
*ʔgjɔmA

Notes
1. Onsets in the Mienic languages are irregular: they should be voiced, reflecting Hmongic
prenasalization. Compare Chinese 菇 ‘mushroom’ (OC *mə.kˤa > MC ku > Man. gū).
 Proto Hmong-Mien 91

5.21 *ŋgj-
PH *ŋgj- 1 2 3 4 5 6 7
1. boat/ship *ŋgjəŋA ȵaŋ2 ŋɑŋ2 ŋkɔ2 ŋkoŋ2 ŋkoŋA ɳtʃoŋ2 ŋɯ2
2. to crawl *ŋgjoŋC ȵoŋ6 - ŋka6 ȵtɕaŋ6 - - -
3. thick (liquid) *ŋgjɔŋC ȵu6 - ŋkɔŋ6 ŋkəŋ6 ŋkoŋC - ŋa8
4. pair *ŋgjowD ȵu8 ŋoŋ8 ŋkaɯ8 ŋkə8 ŋkoD ɳtʃau8 ŋɤ8

PM *ŋgj- 8 9 10 11
5. asleep *ŋgjɔmA gɔm2 - - -
6. to snore *ŋgjanA gaŋ2 daːn2 - -

Notes
1. Compare Central Malayo-Polynesian *waŋka[ŋ] ‘canoe’ (ACD).
5. Also Hainan Mun /gjoːm2/.
6. Xiangnan Mien (Wang and Mao 16) has been substituted for Luoxiang Mien (8).
Compare Chinese 鼾 ‘to snore’ (MC xan > Man. hān).

5.24 *ŋj-
PM *ŋj- 8 9 10 11
1. seven *ŋjiC ȵi6 ŋji6 ni6 ȵi6
2. 硬 hard/solid *ŋjeŋC ȵɛŋ6 ŋjeŋ6 - -

Notes
1. This is from Tibeto-Burman *s-nis (Benedict 1987a:13). Hmong-Mien *djuŋH (2.18/28)
may be from the same source, but the correspondence is difficult. See chapter 5, section
5.2.

5.28 *xj-
PM *xj- 8 9 10 11
1. 寫 to write *xjaB θja3 kja3 ɕa3 sjɛ3

5.30 *ɣj-
PM *ɣj- 8 9 10 11
1. 象 elephant *ɣji̯ ɔŋB θɛŋ4 kjaːŋ4 - sjaŋ4

5.31 *kl-
PHM *kl- 1 2 3 4 5 6 7 8 9 10 11
1. to fry *kleu ka1 ca1 ki1 tɕi1a tɕeA - - - [tjiː1] khla1 -
2. horn *klɛɔŋ ki1 ce1 ku1 koŋ1a kaŋA kjaŋ1 kɔ̃1 kɔŋ1 kjɔːŋ1 klɔ1 kɔu1
3. insect/worm *klæŋ kaŋ1 ci1 ka1 kua1a tɕenA kjen1 kɛ1 kɛŋ1 kjeːŋ1 klɛ1 tsaŋ1
4. road/way *kləuX ki3 kɯ3 ke3 kæ3a tɕiB kja3 ko3 kjau3 kjau3 kla3 tsu3
5. liquid/soup *klæwX - ca3 kua5 ka3a tɕiB - - - - kla3 -
6. to cut *klæp ken7 ɴqɑ7 - - - tʃei7 - kap7 kjap7 klan7 kɛp7
7. grasshopper *klup - - - - - - ka-jɔ7 tɕop7 - klin7 -
92 Chapter 2

PH *kl- 1 2 3 4 5 6 7
8. maggot *klaŋC kaŋ5 - ka7 kua7a tɕenC - kɛ1-lõ7

PM *kl- 8 9 10 11
9. pullet/hen *klanC kaːn5 kjaːn5 - -
10. egg *kləuC kjau5’ kjau5 klau5 tsu5
11. 蝸 snail *klu̯ eiA/B kwei1 kui3 kjau3 -

Discussion
The medial -l- from *kl- is preserved only in Biao Min. However, the disyllabic form of
‘grasshopper’ and ‘maggot’ in Pa-Hng (7) may also reflect the -l-, and is one more witness
to the conservative nature of this language; see chapter 4.

Notes
1. Possibly from Chinese 攪 ‘to disturb, stir’ (OC *kˤruʔ > MC kæwX > Man. jiǎo), but the
tones do not correspond. Biao Min /khla1/ (Solnit 1982) means ‘to turn over, stir’ (of
drying grain, stir frying), hence this reconstruction with medial -l-, despite the problem
with aspiration. In Hainan Mun ‘to fry in oil or fat’ is /tjiː1/. The rime does not
correspond—it is probably a late borrowing from 煎 ‘to fry’ (Man. jiān).
2. The forms for Zongdi (4) and Fuyuan (5) come from Wang 1994. These forms were
entered inconsistently in Wang and Mao 1995. Although the tones do not correspond, this
is probably the widespread word for ‘horn’, Chinese 角 (OC *k.rˤok > MC kæwk > Man.
jiǎo).
3 & 8. Hmong-Mien ‘insect’ and Hmongic ‘maggot’ have the same root. The medial -l-
may be present in the Pa-Hng (7) disyllabic form. ‘Maggot’ seems to have been derived
from ‘insect’ by C-tone derivation; see chapter 4. Compare Austronesian */qulej/ ‘maggot’
and the following note by Blust: “Most problematic is the evidence from Puyuma, Western
Bukidnon Manobo and Tagabili that PAN */qulej/ meant not only ‘maggot’, but also
‘insect’” (ACD). Forms with tone 7 are secondary tones derived by tone sandhi.
5. Also Shimen /ka5/ and Bunu /cai3’/ (Office of Miao-Yao Research Lexicon 1987). In
Biao Min (10), this means ‘dew’.
6. Compare Cantonese /kip/. The Jiwei form /ɴqɑ7/ and the Jiongnai form /tʃei7/ are
irregular.
7. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
10. This is no doubt the same as Hmongic 5-6.1/3 *qæwC, since the tones and rimes
correspond. The problem is that the medial -l- in Biao Min should correspond to velars
rather than uvulars in Hmongic if from *kl- (see 5.31) or should correspond to uvulars with
medial -l- in Hmongic if from *ql- (see 6.31). Although Hmong-Mien *qlj- is a possiblity,
they are conservatively entered separately.
11. For medial -l-, see Sunjiang Biao Min /kle1/ (Wang and Mao 22). Compare Hmongic
*ɢʉA ‘snail’ (6.3/8) in the same rime category, and Mienic *kʷlejA ‘snail’ (5-6’.31/10).

5.31.1 *klj-
PM *klj- 8 9 10 11
1. 龍 dragon *kljuŋA kuŋ1 kuːŋ1 kljɔŋ1 -
 Proto Hmong-Mien 93

Notes
1. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10). This word represents a second borrowing from Chinese ‘dragon’: see also Hmong-
Mien *-roŋ (2.57/28).

5.32 *khl-
PHM *khl- 1 2 3 4 5 6 7 8 9 10 11
1. hot (water) *k(h)lɛɔm khi1 ce1 ku1 koŋ1a kaŋA khjaŋ1 kɔ̃1 kɔm1 kjɔːm1 klœn1 tsam1

PH *khl- 1 2 3 4 5 6 7
2. centipede *khluwD khu7 - - - - khjɔ5 -

PM *khl- 8 9 10 11
3. 力 strength *khlək D
kha7 - khla7 -

Notes
1. Jiwei (2) from Office of Miao-Yao Research Lexicon (1987). Sunjiang Biao Min (Wang
and Mao 22) has been substituted for Dongshan Biao Min (10). Aspiration has been
tentatively reconstructed for Hmong-Mien because of the aspirated initials in Hmongic.
2. Compare the Central Tai and Southwest Tai forms: Lao /(khii3-)khep7/, Thai /takhaap7/
(Downer 1978). See also Proto Mon-Khmer *kʔaip > Khmer /kʔæp/ (Shorto #1226).
3. Xiangnan Mien (Wang and Mao 16) has been substituted for Luoxiang Mien (8).
Hmong-Mien *-rək ‘strength’ (2.57/7) is from the same source.

5.33 *gl-
PM *gl- 8 9 10 11
1. mountain (range) *gli̯ emA kem2 kiːm2 kleŋ2 -
2. thin *glæiC kje6 kjai6 klai6 kɛi6

Notes
1. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10).
2. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10). A word meaning ‘thin (person)’ with matching rime and tone exists in Hmongic, but
its initial is *ndz- (3.6).

5.33.1 *glj-
PHM *glj- 1 2 3 4 5 6 7 8 9 10 11
1. to fool/humor 1 4 4
- - - - - kle - - kja kljɔu kɛi4
(children)*glje(X)

Notes
1. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10). The meaning of this word is given as hǒng as in hǒng xiǎo há ‘to handle children’ in
Wang and Mao 1995.
94 Chapter 2

5.34 *ŋkl-
PHM *ŋkl- 1 2 3 4 5 6 7 8 9 10 11
1. shallow *ŋkli̯ əŋX - - ŋkla3 - - - - glaŋ3 gjaŋ3 - lɛŋ3

Notes
1. Green Mong has been substituted for White Hmong (3).

5.34.1 *ŋklj-
PM *ʔglj- (< *ŋklj-) 8 9 10 11
1. scissors *ʔglji̯ euB jeu3 ȡiːu3 gjau3 gɛu3

Notes
1. Also Sunjiang Biao Min /kljɔu3/ (Wang and Mao 22).

5.36 *ŋgl-
PHM *ŋgl- 1 2 3 4 5 6 7 8 9 10 11
1. cattle pen *ŋgluə ŋə2 - ŋkua2 ŋka2 ŋkaA - tɕõ2 - gu2 glu2 dzu2
2. 藍 indigo *ŋglam ȵi2 ȵi2 ŋka2 ŋkua2 ȵʨenA ɳtʃen2 ŋɤ2 gaːm2 gaːm2 klan2 -

PH *ŋgl- 1 2 3 4 5 6 7
3. lazy *ŋglænB ŋi4 ɲe4 ŋkɛŋ4 ȵtɕein4 ŋkaŋB - ŋɛ4

PM *ŋgl- 8 9 10 11
4. turbid *ŋglokD glo8 glɔ6 - -

Notes
1. Biao Min (10) is from Solnit 1982. This may be the same as Chinese 圉 ‘prison, pen’
(OC *m-q(ʰ)(r)aʔ > MC *ngjoX > Man. yǔ).
2. The palatal nasal in Yanghao (1) suggests that this might have had a medial -r- (*ŋgram)
instead of a medial -l-. Since native words in *gr- pattern differently, however (see 5.48
below), the word has provisionally been placed here. Sunjiang Biao Min (Wang and Mao
22) has been substituted for Dongshan Biao Min (10).
3. Although a medial -l- is not attested, it is reconstructed for this word. Given the
reconstructed front vowel, it would have become a uvular if not protected by a following
liquid. Compare Chinese 懶 ‘lazy’ (OC *[r]ˤanʔ > MC lanX > Man. lǎn).
4. Compare Hmongic *ŋkroB/*ŋgroB ‘turbid’(5.49/7). This word may involve expressive
phonology. Hainan Mun /glɔː6/ means ‘muddy’ (compare Hmongic *ŋkoD ‘muddy’, 5.4/7).

5.36.1 *ŋglj-
PM *ŋglj- 8 9 10 11
1. tiger *ŋgljænA ʥan2 gjaːn2 jɛn2 kjɛn2
 Proto Hmong-Mien 95

Notes
1. This is a difficult word to reconstruct; none other patterns like it. It is tentatively placed
in this “free” slot with a medial -l- due to its resemblance to Mon-Khmer *klaʔ ‘tiger’
(Shorto #197).

5.46 *kr-
PHM *kr- 1 2 3 4 5 6 7 8 9 10 11
1. to laugh *krət ȶə7 ʈo7 ʈɔ7 ʂo7a tʂuD tsu7 tɕa7 kjat7 kjet7 klan7 tut7
2. six *kruk ȶu5 ʈɔ5 ʈau5 ʂu5a tʂoC tʃɔ5 tɕɯ5 kwo7 kjɔ7 klɔ7 tɔu7

PH *kr- 1 2 3 4 5 6 7
3. to burn (TR) *kru̯ eiB ȶu3 - ʈau3 ʂɔ3a tʂuB - -

Notes
1. Also Shimen /tl̥ o7/ and Ho Ne /kɤ7/.
2. A borrowing from Tibeto-Burman *k-ruk (Matisoff 2003); see chapter 5, section 5.2.

5.48 *gr-
PHM *gr- 1 2 3 4 5 6 7 8 9 10 11
1. door/gate *gru̯ eŋ ȶu2 ʈu2 ʈɔŋ2 ʂaŋ2 ʐoŋA ʃwaŋ2 tɕɔ̃2 kɛŋ2 keːŋ2 klɛ2 -
2. to be fat *grəunH ȶaŋ6 ʈɑŋ6 ʈɔ6 ʂoŋ6 ʐoŋC ʃoŋ6 tɕõ6 kun6 kun6 klin6 tin6

PH *gr- 1 2 3 4 5 6 7
3. animal fat/oil *grəŋA ȶaŋ2 - ʈɔ2 ʂoŋ2 ʐoŋA - -

Notes
2 & 3. Although only attested in Hmongic, ‘animal fat/oil’ clearly involves the same root
as ‘to be fat’. It appears that the widespread ‘to be fat’ derived from the now more
restricted ‘animal fat/oil’ via C-tone derivation; see chapter 4, section 4.4.2.

5.49 *ŋkr-
PH *ŋkr- 1 2 3 4 5 6 7
1. turbid *ŋkroB/ŋgroB - ɳʈo3 ɳʈɔ5 ɳʈo3a ɳʔtʂuB ɳtʃu4 ȵɔ̃3

Notes
1. Also Ho Ne /kɤ4/ (Wang and Mao 14). It is impossible to tell whether this word belongs
to 5.49 or 5.51, since upper and lower register tones are represented equally. Compare
Mienic *ŋglokD ‘turbid’ (5.36/13), reflexes of which have tones 6 and 8, and a similar
initial. Expressive phonology.

5.51 *ŋgr-
PH *ŋgr- 1 2 3 4 5 6 7
1. to lean (on stick) *ŋgraC ȵa6 ɳɑ6 ɳʈi6 ɳʈe4 ɳtʂaC - ȵa2
2. to seek shelter *ŋgreD - - ɳʈai8 ɳʈe8 ɳtʂeD - -
3. 漏 to drip *ŋgroC ȵə6 - ɳʈɔ6 ɳʈu6 ɳtʂuC - -
96 Chapter 2

Notes
1. Also Ho Ne /ki4/ (Wang and Mao 14). In White Hmong (3) also ‘a stick, cane’.
2. In White Hmong (3), ‘to hide’.
 Proto Hmong-Mien 97

5-6. Dorsals (undifferentiated velar/uvular)

5-6.1 *K-
PHM *K- 1 2 3 4 5 6 7 8 9 10 11
1. frog *KæŋX qaŋ3 - qa3 hua3a qenB ken3 kɤ7 - ȶeːŋ3 kwa3 -
2. to crow *KajH qa5 qɑ5 qua5 ha5a qaC ka5 ŋa5 kaːi5’ kaːi5 - kɔi5
3. back (vs. front) *N-Kaŋ qaŋ1 - qa1 hua1a - - - daːŋ1 daːŋ1 - kɔŋ1
4. excrement*N-KəjX qa3 qɑ3 qua3 ha3a qaB ka3 ka3 dai3 dai3 kai3 kai3
5. 雞 chicken *Kəi qei1 qa1 qai1 he1a qeA kai1 kɛ1 tɕai1 ȶai1 kai1 kui1
6. 甘 sweet *Kam qaŋ1 - qa1 hua1a qenA ken1 kɛ1 kaːm1 kaːm1 kan1 -
7. 假 to borrow *KaX - qɑ3 qe3 hɪ3a - ku3 ka3 ka3 ka3 kɔ3 kɔu3

PH *q- (< *K-) 1 2 3 4 5 6 7

8. crops *qɔŋA qə1 - qɔŋ1 haŋ1a - - -
9. star *qɛŋA qɛ1 qe1 qu1 hoŋ1a qaŋA ka1 kɔ1̃
10. grape *qinB qei3 qɛ3 - hæin3a qenB - ji3
11. to scold/curse *qeC - - - he5a qeC - -
12. egg *qæwC ki5 - qe5 hæ5a qwjiC kja5 ko5

PM *k- (< *K-) 8 9 10 11

13. 鴿 pigeon *kɔpD kop7 kɔp7 - ku7
14. 割 to cut *katD kaːt7 kaːt7’ kwan7 kɔt7

Notes
1. Compare Proto Austronesian *keŋkeŋ ‘frog; hollow sound’ (ACD).
2. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10). Possibly from Chinese 歌 ‘to sing; a song’ (OC *[k]ˤaj > MC ka > Man. gē), although
this word is not used for the song of animals, and the tones do not correspond.
3 & 4. For ‘back’, see also Mien (Wang and Mao 16) /gaŋ1/. For both words, Mien (8) and
Mun (9) show unusual initials in d- for which there is no satisfactory explanation; since the
rimes and tones correspond, however, they probably belong here. The voiced initials in
Mienic are accounted for by the reconstruction of an *N-, a nasal element which, unlike a
prenasalized stop, left no trace in Hmongic.
5. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10). This is an areal word: see also Proto Tai *kəi (Li 1977).
12. This is no doubt the same as Mienic *kləuC ‘egg’ (5.31/3), since the tones and rimes
correspond. The problem is that the medial -l- in Biao Min should correspond to velars
rather than uvulars in Hmongic if from *kl- (see 5.31), or should correspond to uvulars
with medial -l- in Hmongic if from *ql- (see 6.31). Although Hmong-Mien *qlj- is a
possibility, they are conservatively entered separately.
98 Chapter 2

5-6’.1 *Kʷ-
PHM *Kʷ- 1 2 3 4 5 6 7 8 9 10 11
1. 瓜 cucumber *Kʷa fa1 kwɑ1 di1 hwɪ1a qwaA kwe1 qwa1 kwa1 kwa1 kwa1 ka1
2. 廣 wide *Kʷi̯ aŋX faŋ3 kwei3 da3 hwua3a - kwan3 kwɤ3 kwaŋ3 kwaŋ3 kwaŋ3 kjaŋ3
3. 過 to cross (river)
fa5 kwɑ5 dhau5 - qwaC kwa5 kwa5 kwɔi5’ kwaːi5 kwa5 kɛi5
*KʷajH

Discussion
As with the clearly velar and uvular labiovelars (5’ and 6’), it is likely that all words in the
*Kʷ series have been borrowed from Chinese. The rounding of the secondary articulation
did not have the inhibitory effect of an initial rounded element in the rime: all of the dorsal
initials in this series retracted to uvular in Hmongic (see discussion note under 5.1 above).

Notes
3. There is evidence of original aspiration both in White Hmong (3) and in the tone
subcategory of Luoxiang Mien (8).

5-6.2 *Kh-
PHM *Kh- 1 2 3 4 5 6 7 8 9 10 11
1. 槀 dry/withered 3 3 3b 3 3 3 3
- qha qhua ha - khei khe - kaːu khɔ -
*KhæwX
2. 客 guest *Khæk qha5 qha5 qhua5 ha5b qheiC - - khɛ7 - khiɛ7 -

PH *qh- (< *Kh-) 1 2 3 4 5 6 7

3. to tie/cross (legs)
qhei1 - qhai1 he1b qheA khai1 -
*qheA
4. to pack/wrap *qhʉB qhɛ 3
- qhɯ 3
hou 3b
qhu B
khau 3
-
5. ginger *qhinB khi3 - qhia3 hæin3b qhwjenB khai3 -
6. 孔 hole *qhəŋB qhaŋ3 qhu3 qhɔ3 hoŋ3b qhoŋB khoŋ3 khõ3

PM *kh- (< *Kh-) 8 9 10 11

7. bamboo bowl 7
khɔ - - -
*khɔkD
8. 牽 to lead *khienA - kiːn1’ khən1 -

Notes
2. The Mien (8) and Biao Min (10) forms are from Office of Miao-Yao Research Lexicon
(1987).
4. Possibly from Chinese 裹 ‘to wrap’ (OC *[k]ˤo[j]ʔ > MC kwaX > Man. guǒ).
7. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
 Proto Hmong-Mien 99

5-6.3 *G-
PHM *G- 1 2 3 4 5 6 7 8 9 10 11
1. 號 sing/cry out *Gæw ken2 - qua2 ha2 ʁiA - - haːu2 haːu2 hjau3 -
2. 下 low/short *GaX ka4 - qe4 hɪ4 ʁaB tʃe4 ŋe4 ha4 ha4 hɔ4 -

PH *ɢ- (< *G-) 1 2 3 4 5 6 7

3. bamboo pipes *ɢænA ki2 - qen2 hæin2 ʁwjaŋA - -
4. to slant/slanting *ɢeA - qa2 qai2 - ʁeA - -
5. 葫 garlic *ɢaA qa2 - qe2 hɪ2 ʁaA - -

PM *g- (< *G-) 8 9 10 11

6. oneself *gənA kan2 kan2 kan4 -
7. thick (porridge) *gətD - kat8 - -

Notes
2. A related word within Chinese was also borrowed by Hmong-Mien: Chinese 下 ‘to
descend’ appears as Hmongic *ɴɢaB (5-6.6/4) and, in a separate borrowing, as Mienic *ɣaC
(5.13/4).
4. In White Hmong also ‘to lean’.
6. ‘Oneself’ is technically Hmong-Mien, since this word also appears in Ho Ne (Hmongic):
/kan4/. The form is irregular, however, and looks like a loanword from some Mienic
language.
7. Also Jiangdi Mien /ʥat8/ and Xiangnan Mien /jə8/. These forms suggest *gj-, but ‘thick’
does not pattern with the other words in 6.18. Compare Proto Malayo-Polynesian *buket
‘thick, viscous’ from a Proto Austronesian root */-keC/; Proto Malayo-Polynesian */-ket/
‘adhesive, sticky’ (ACD).

5-6’.3 *Gʷ-
PH *ɢʷ- (< *Gʷ- ) 1 2 3 4 5 6 7
1. to escape *ɢʷaD fa8 qwei8 di8 hwɪ8 ʁwaD - -
2. 黃 yellow *ɢʷaŋA faŋ2 kwei2 da2 hwua2 ʁwenA kwan2 kwɛ2
3. 黃 bright/light *ɢʷaŋA faŋ2 - ka2 hwua2 ʁwenA kui2 -

Notes
1. The labialized velar marks this as a loanword. According to Sagart (1999:123–124), it is
a loanword from 脫 ‘to peel off, escape’ (OC *lˤot > MC dwat > Man. tuō). Mortensen
(2002) proposed Tibeto-Burman *g-lwat ‘free, release’ as the source, which works better
since it has a dorsal consonant (for Sagart, the dorsal consonant is a secondary
development of a hypothetical intermediate stage *dl- in Chinese). Note that this Chinese
word with a voiceless initial is taken here as the source of Mienic *ʔdutD (< *N-t-) ‘peel
off/escape’ (2.4/9).
2 & 3. It is of interest that ‘yellow’ and ‘bright/light’ are not identical in White Hmong (3)
and Jiongnai (6), as one would expect. It may be that the word with the meaning
‘bright/light’ was a later borrowing in these dialects. In Mienic, ‘yellow’ was probably
borrowed from Chinese 皇 ‘yellow and white spotted’ (OC *[ɢ]ʷˤaŋ > MC hwang > Man.
huáng) as *wi̯ əŋA (1.12/24).
100 Chapter 2

5-6.4 *NK-
PHM *NK- 1 2 3 4 5 6 7 8 9 10 11
1. cogongrass *NKan qɛ1 - ɴqɛŋ1 ŋkæin1a - ŋkan1 ŋɛ1 gaːn1 gaːn1 gwan1 gɔn1

PH *ɴq- (< *NK- ) 1 2 3 4 5 6 7

2. to low/bellow *ɴqəŋB - - ɴqɔ3 ŋkoŋ3a ɴʔqoŋB - -
3. 價 price *ɴqaC qa5 ɴqɑ5 ɴqi5 ŋkɪ5a ɴʔqaC ɳtʃe5 ŋa5
4. 鉤 hook *ɴqæwC qa5 - ɴqe5 ŋkæ5a - - -

Notes
3. In White Hmong (3) also pronounced /ɴqe5/.

5-6’.4 *NKʷ-
PHM *NKʷ- 1 2 3 4 5 6 7 8 9 10 11
1. 缺 to have a gap
qei5 ɴqwɑ3 ntai3 - - - - gwa7 gu7 - -
*NKʷet

PM *ʔgʷ- (< *NKʷ- ) 8 9 10 11

2. 黃 bright *ʔgʷi̯ əŋA gwaŋ1 gwaŋ1 gwaŋ1 vjaŋ1

Notes
1. The Chinese source has an aspirated initial (MC khwɛt > Man. quē). In Chinese the
word means ‘to be short of, to lack, defective, incomplete’ which is better reflected in the
White Hmong (3) meaning ‘broken off’. The development of apparently irregular White
Hmong (3) nt- must have been from *NKʷet - > ɴqʷ- > ɴql- > nd- > nt- (d- is a regular
development from kl- or ql- in White Hmong, but the onset nd- does not occur; nt- is the
closest equivalent). Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang
Mien (8).
2. See discussion of Chinese ‘yellow’ as the source of both ‘yellow’ and ‘bright/light’
under 5-6’.3 above.

5-6.5 *NKh-
PHM *NKh- 1 2 3 4 5 6 7 8 9 10 11
1. dry/thirsty *NKhæj qha1 - ɴqhua1 ŋka1b ɴʔqheiA ŋkhei1 - gaːi1 gaːi1’ - gɔi1
2. 渴 thirsty *NKhat - ɴqhe7 ɴqhi7 - - - - gaːt7 gaːt7’ - gɔt7

Notes
1. Expressive phonology.
2. In White Hmong (3) also pronounced /ɴqhe7/.
 Proto Hmong-Mien 101

5-6.6 *NG-
PHM *NG- 1 2 3 4 5 6 7 8 9 10 11
1. meat *NGej ŋi2 ɲa2 ɴqai2 ŋke2 ɴqeA ŋkai2 ŋɛ2 - - dʑi2 hɔi2
2. diligent *NGəjH ŋa6 ŋɑ6 ɴqua6 ŋkəa6 ɴqaC ŋka6 ŋe6 - - gwai6 -

PH *ɴɢ- (< *NG-) 1 2 3 4 5 6 7

3. shuttle *ɴɢəŋB ŋaŋ4 nɑŋ4 ɴqɔ4 ŋkoŋ4 ɴqoŋB ŋwaŋ4 -
4. to swallow *ɴɢəŋB ŋaŋ4 ŋu4 - - ɴqoŋB - -
5. 下 to descend *ɴɢaB ŋa4 - ɴqe4 ŋkɪ4 ɴqaB - ŋe4
6. 狹 narrow *ɴɢeD ŋi8 ŋɑ8 ɴqai8 ŋke8 ɴqeD ŋkai8 ŋɛ8

Notes
3. Mienic *ɴɢlowB ‘shuttle’ (6.36/16) suggests a rounded vowel, which might make this
initial uvular with a medial /-l-/. It is not clear that they reflect the same word, however,
since there is no final nasal consonant in Mienic and the rime correspondence is not
attested elsewhere.
4. Expressive phonology.
6. Hmongic and Mienic both borrowed this word from Chinese 狹 ‘narrow’ (OC *N-
kˤ<r>ep > MC hɛp > Man. xiá), but Hmongic *ɴɢeD was an earlier borrowing from OC,
and Mienic *hepD (7.13/10) was a later borrowing from MC.

5-6’.31 *Kʷl-
PM *kʷl- (< *Kʷl- ) 8 9 10 11
1. 蝸 snail *kʷlejA kwei1 gwei1 kli3 ki1

Notes
1. Compare Hmongic *ɢʉA ‘snail’ (6.3/8) in the same rime category, and Mienic *klu̯ eiA/B
‘snail’ (5.31/8).
102 Chapter 2

6. Uvulars

6.1 *q-
PHM *q- 1 2 3 4 5 6 7 8 9 10 11
1. old *qu̯ oH qo5 qɔ5 qu1 - - ku5 ko5 ko5’ ku5 ku5 ku5

PH *q- 1 2 3 4 5 6 7
2. neck *qoŋB qoŋ3 qoŋ3 - haŋ3 - kɐŋ3 -
3. marry (a woman) *qu̯ aC qha5 - - ha5a kaC - -

Discussion
Hmong-Mien *velars merged with uvulars in Hmongic unless followed by back rounded
vowels, -l- , or -r- — and in Mienic, all Hmong-Mien *uvulars merged with velars (see this
chapter. section 1.2).26 Since words in the *k- and *q- series proper only contrast in these
three restricted environments, a number of correspondence “sets” in series 5 and 6 will
only contain one or two words. Words reconstructed with a dorsal consonant before any
vowel other than a back rounded vowel are placed in the 5-6 series, which simply
represents an undifferentiated dorsal (here written with a capital letter: *K- *KH-, *G-,
etc.).

Notes
1. Wang and Mao (1995:327–328) list ‘old’ in the sense of ‘aged’, and ‘old’ as the
opposite of ‘new’ separately because of the presence of two forms in Ho Ne (She) that
differentiate these senses of ‘old’ (kɤ5 and kɔ5, respectively). Since the forms are otherwise
identical, they are presented together here. This is probably from Chinese 故 ‘old (not
new)’ (OC *kˤa(ʔ)-s > MC kuH > Man. gù), especially since in White Hmong it appears
before the noun it modifies, as in Chinese, rather than in the customary position for
adjectives, after the noun.
2. Compare Chinese 頸 ‘neck’ (OC *keŋʔ > MC kjiengX > Man. jǐng).

6’.1 *qʷ-
PHM *qʷ- 1 2 3 4 5 6 7 8 9 10 11
1. far *qʷuw - qɯ1 de1 hwæ1a qweiA kwa1 ko1 ko1 ku1 ku1 kɔu1

Notes
1. Compare Chinese 迂 ‘bend, deflect; astray, far from’ (OC *qw(r)a > MC ʔju > Man.
yū).

26
Medial -j- also prevented a *velar from retracting in Hmongic, but since there are no contrasting *kj- and
*qj- correspondence sets, it is assumed that there was no *qj- series in Proto Hmong-Mien.
 Proto Hmong-Mien 103

6.3 *ɢ-
PH *ɢ- 1 2 3 4 5 6 7
1. cuckoo *ɢʉB - qu4 - hu4 - - -
2. duck *ɢu̯ aC ka6 - - - ʁwaC - -
3. to fall (over) *ɢu̯ eiC ku6 qɔ6 qau6 ho6 ʁuC - -
4. back (of body) *ɢuwD kə8 - qau8 hu8 ʁoD kɔ8 kɯ8
5. 蝸 snail *ɢʉA ki1 qə1 qɯ2 hu2 ʁwjuA - -

Notes
2. Also Fengxiang (West Hmongic) /qa6/ (Wang and Mao 9). Expressive phonology.
4. In White Hmong, this refers to the upper back. Tan Trinh Pa-Hng (Niederer 1997) has
been substituted for Baiyun Pa-Hng (7).
5. Compare Mienic *kʷlejA ‘snail’ (5-6’.31/10) and *klu̯ eiA/B ‘snail’ (5.31/8).

6.4 *ɴq-
PH *ɴq- 1 2 3 4 5 6 7
1. to swallow *ɴqowC - - - ŋkə5a - ŋkau5 ŋɛ5
2. 鳩 pigeon *ɴquA qo1 ɴqo1 ɴqua1 ŋka1a ɴʔqaA ŋku1 ŋo1

Notes
1. Expressive phonology.

6.6 *ɴɢ-
PHM *ɴɢ- 1 2 3 4 5 6 7 8 9 10 11
1. sky/heaven *ɴɢɛuŋ - - - - ɴqwaŋA ŋkuŋ2 ŋwɔ2 guŋ2 guːŋ2 - -
2. 滑 smooth/slippery
- - - - - ŋkɔ8 ŋgɯ8 gut8 gɔt8 gwan8 gut8
*ɴɢu̯ at

Notes
1. Part of a very irregular set with Hmong-Mien *wɛŋ ‘sky/heaven’ (1.12/22) and Hmong-
Mien *ndɛuŋ ‘sky/heaven’ (2.6/22).
1 & 2. Tan Trinh Pa-Hng (Niederer 1997) has been substituted for Baiyun Pa-Hng (7).
104 Chapter 2

6.31 *ql-
PHM *ql- 1 2 3 4 5 6 7 8 9 10 11
1. neck *qlaŋ - - da1 - - - - klaːŋ1 klaːŋ1 klaŋ1 kan1
ka-
2. dog *qluwX l̥ a3 qwɯ3 de3 l̥ æ3a qleiB kla3 klo3 klu3 klu3 ku3
ljɔ̃7
3. waist *qlajX l̥ a3 qwɑ3 dua3 l̥ a3a qlaB kla3 la3 klaːi3 klaːi3 kla3 lai3
4. eagle/hawk *qlaŋX l̥ aŋ3 qwei3 da3 l̥ ua3a - klen3 kwɤ3 klaːŋ3 klaːŋ3 klaŋ3 kjan3

PH *ql- 1 2 3 4 5 6 7
5. white *qlowA l̥ u1 qwə1 daɯ1 l̥ ə1a qloA klau1 kwɔ1
6. to roll down *qləŋB ȴ̥aŋ3 cɑŋ3 dɔ3 - qloŋB - -
7. walnut *qlowC - - - l̥ ə5a qloC - -

PM *ql- 8 9 10 11
8. to repay *qlauB - klaːu3 klau3 kau3

Discussion
The *ql- series is differentiated from the *kl- series by the relative persistence of -l- in
languages across the family, and the occasional retention of a uvular stop.

Notes
1. Compare Chinese 頸 ‘neck’ (OC *keŋʔ > MC kjiengX > Man. jǐng). Wang and Mao
(1995) give two different forms for Zao Min ‘neck’, and note that both are irregular in
rime: /kan1/ (p. 341) and /kuŋ1/ (p. 500). It is not clear whether these are variants, or one
form was entered in error. /kuŋ1/ is arbitrarily entered here.
2. This word for ‘dog’ appears in languages across the area: compare Chinese 狗 ‘dog’
(OC *-[k]ˤ(r)oʔ > MC kuwX > Man. gǒu) and see Benedict 1996. The disyllabic Pa-Hng
(7) form /ka-ljɔ̃7/ is especially interesting for what it may suggest about the origin of some
Hmong-Mien consonant clusters (see chapter 4, section 4.2.4 and Niederer 2004, Ratliff
2006).
4. Compare Mon-Khmer *klaŋ ‘hawk, large raptor’ (Shorto #714) and Tibeto-Burman
*(g)laŋ ‘eagle’ (Matisoff 2003).

6.33 *ɢl-
PHM *ɢl- 1 2 3 4 5 6 7 8 9 10 11
1. round *ɢlun - - - - - - ka1-lɔ̃4 klun2 klun2 klin2 -
2. 桃 peach *ɢlæw l̥ en2 qwa2 dua2 l̥ a2 ʁleiA - kwi2 klaːu2 klau2 kla2 kɔu2

PH *ɢl- 1 2 3 4 5 6 7
3. river *ɢlæwA - - de2 - ʁleiA - -
4. Allium *ɢləŋB - - dɔ4 l̥ oŋ4 ʁloŋB - -
5. mountain pass *ɢlowD - - daɯ4 l̥ ə4 - - -

Notes
1. Compare Chinese 輪 ‘wheel; round’ (MC lwin > Man. lún); see chapter 4 on the
significance of Pa-Hng disyllabic form for interpreting Hmong-Mien clusters.
 Proto Hmong-Mien 105

5. This may be the same as Chinese 峽 ‘gorge between two mountains’ (OC *N-kˤ<r>ep >
MC hɛp > Man. xiá; the same word as 狹 ‘narrow’), but the Chinese word does not have a
rounded vowel.

6.33.1 *ɢlj-
PM *ɢlj- 8 9 10 11
1. 腸 intestines *ɢljaŋ A
klaːŋ2 klaːŋ2 klaŋ2 kjaŋ2

6.36 *ɴɢl-
PH *ɴɢl- 1 2 3 4 5 6 7
1. front surface *ɴɢlinA - - ntia2 - ɴqlenA - -
2. to flow *ɴɢlʉB - - ntɯ4 - qluB ljɔu4 -
3. ragged/shabby *ɴɢluC nei6 - ntua6 - qlaC - -

PM *ɴɢl- 8 9 10 11
4. shuttle *ɴɢlowB gou4 glɔu4 - -

Notes
1. There are also Hmongic forms that match this word in meaning, rime, and tone but have
an *mbl- initial: Gaopo /mplɛ2/, Zongdi /mplæin2/, and Fengxiang /mplen2/ (Wang 1994).
3. Reflexes in Xianjin (/ɳʈua6/) and Qingyan (/ɳʈo6/) have retroflex initials, so this word
might better be reconstructed as *ɴɢruC. It is possibly from a prefixed form of Chinese 陋
‘humble, mean’ (OC *[r]ˤo-s > MC luwH > Man. lòu).
4. These forms and Hmongic *ɴɢəŋB ‘shuttle’ (5-6.6/21), may be related, although there is
no trace of -l- in Hmongic, and no final nasal in Mienic.

6.46 *qr-
PHM *qr- 1 2 3 4 5 6 7 8 9 10 11
1. bear *qrep ȴ̥i7 - dai7 l̥ e7a - - - kjep7 kjaːp7 kljɛ7 -

PH *qr- 1 2 3 4 5 6 7
2. black *qrɛŋA l̥ ɛ1 qwe1 du1 l̥ oŋ1a qlaŋA klaŋ1 -
3. spirit/ghost *qraŋA ȴ̥aŋ1 qwei1 da1 l̥ ua1a qlenA kli1 kwɛ1
4. trough *qroŋA ȴ̥oŋ1 coŋ1 da1 l̥ aŋ1a qloŋA kjɐŋ1 ljõ1
5. to tear *qru̯ aC ȴ̥i5 - dua5 ɭ̥a5a qlaC - kwɛ3
6. span, finger *qroC l̥ o5 - dɔ7 l̥ o5a qlouC - -
7. salty *qrowC ȴ̥u5 - daɯ5 l̥ ə5a - - -
8. ice *qru̯ eiD ȴ̥u7 - dau7 - - - -

PM *kr- (< *qr- ) 8 9 10 11

9. black *qri̯ ɛkD kje7 kja7 kja7 kjɛ7

Notes
1. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10).
106 Chapter 2

2 & 9. These two words for ‘black’ are undoubtedly related, but the rime and tone
correspondences are irregular. In Mienic, the medial liquid is reflected in Sunjiang Biao
Min /klja7/.
4. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
5. Tan Trinh Pa-Hng (Niederer 1997) has been substituted for Baiyun Pa-Hng (7).
6. In White Hmong (3), this is the span of thumb to middle finger.

6.48 *ɢr-
PH *ɢr- 1 2 3 4 5 6 7
1. armspan *ɢraŋA ȴ̥aŋ2 ci2 da2 l̥ ua2 ʁleiA - -

Notes
1. In White Hmong (3), this measure is for a double arms’ length (fingertip to fingertip).

7. Laryngeals

7.1 *ʔ-
PHM *ʔ- 1 2 3 4 5 6 7 8 9 10 11
1. one *ʔɨ i1 ɑ3 i1 ei1a - i3 - - - i1 a1
2. two *ʔu̯ i o1 ɯ1 ɔ1 ɔ1a uA u1 va1 vi1 i1 wəi1 vi5
3. bitter *ʔim i1 ɛ1 ia1 æin1a - an1 jɪ1̃ im1 im1 in1 jɛm1
4. mountain pass *ʔæ(C) - - - - - e7 - ɛ7 e1 ɛ7 a1
5. water *ʔu̯ əm ə1 u1 - aŋ1a oŋA waŋ1 ʔɔ1̃ wəm1 wɔm1 ən1 m̩ 3
6. to do/work *ʔəjH ɛ5 - ua5 aŋ5a aC - ʔɪ5̃ - ai5 wəi5 ai5
7. to swell *ʔəumH aŋ5 ɑŋ5 ɔ5 oŋ5a oŋC oŋ5 ʔõ5 om5’ ɔm5 ən5 -
8. 媼 wife *ʔəuX a3 - - - - - - au3 au3 kau3 -
9. 鴨 duck *ʔap - - ɔ7 o7a - ai7 a7 aːp7 aːp7’ an7 ap7

PH *ʔ- 1 2 3 4 5 6 7
10. that/the former *ʔɨB i3 a3 i5 i3a eB - -
11. cloud/haze *ʔɔŋC en5 - ɔŋ5 aŋ5a - - -
12. 烏 crow (bird) *ʔuA - - ua1 a7a - - -

PM *ʔ- 8 9 10 11
13. meat *ʔaB a3 a3 - -
14. 愛 to love *ʔu̯ ɔiC - ɔːi5 - i5

Notes
1. ‘One’ is highly irregular, which is to be expected since it is both short and frequently
used. The onset is not in question, however. It is the same as Chinese 一 ‘one’ (OC *ʔi[t] >
MC ʔjit > Man. yī); see chapter 5. The Jiongnai (6) form is from Mao and Li 2001.
8. In Yanghao (1), ‘elder sister’. In Dongshan Biao Min (10), /u3/ means ‘elder sister’ and
/au3/ means ‘wife’; they are presumably both taken from this source.
 Proto Hmong-Mien 107

7.13 *h-
PHM *h- 1 2 3 4 5 6 7 8 9 10 11
1. 欱 to drink/smoke *hup hə7 hu7 hau7 ɦo7b hoD xɔ7 hɔ7 hop7 hɔp7’ hən7 hup7

PH *h- 1 2 3 4 5 6 7
2. fog/cloud *huA ho1 ho1 hua1 ɦa1b haA - ho1
3. to weave/plait *hinA hei1 hɛ1 hia1 ɦæin1b henA xan1 -
4. clay pot/earthen jar
- - hu1 ɦoŋ1b haŋA - -
*hɛŋA
5. to sharpen/grind *hoB xhə3 ho3 hɔ3 ho3b huB xu3 ho3
6. to boil (TR) *hu̯ eiC hu5 hɔ3 hau5 - - - -
7. to ladle out *heD hei7 - hai7 ɦe7b heD - fɛ7
8. to rob *hæD - - - ɦa7b hiD - -
9. 腐 bean curd *hʉB - - - ɦou3b huB - -

PM *h- 8 9 10 11
10. to cover (with tile) 6 3
hom hɔːm - hum3
*hi̯ əmB/C
11. to scold/curse *hi̯ emC heːm5 - - -
12. 狹 narrow *hepD hep8 heːp8 hjɛn8 hɛp8

Notes
9. This must be from Chinese 腐 ‘rotten’ (MC bjuX > Man. fǔ ‘curd’); a late borrowing
into West Hmongic. See also Hmongic *bu̯ aB 'bad/spoiled' (1.3/15), an earlier borrowing
from the same word.
10. These forms for ‘to cover’, as well as Mienic *komB (5.1/21), appear to be borrowings
from Tai-Kadai. Li (1977) reconstructs a number of similar words with this general
meaning: *xr[um]B1 ‘to cover, veil; *ɣ[um]B2 ‘to cover, protect’; *homC1 ‘to cover up’. See
also the Mienic forms under ‘to cover’ with initial g- in L-Thongkum’s reconstruction:
Mien /gom3/, Mun /gəm3/, etc. (1993:204). The Mienic reflexes in both tone 3 and tone 6
suggest an initial voicing alternation as well as a tone alternation, but the tones may simply
be “loan-tones” (see chapter 3), chosen to reflect the tone of the donor language(s) as
closely as possible.
11. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). This
may belong with Hmongic *qeC (5-6.1/10) ‘to scold/curse’, as represented in Wang and
Mao 1995 (p. 327).
12. Hmongic and Mienic both borrowed this word from Chinese 狹 ‘narrow’ (OC *N-
kˤ<r>ep > MC hɛp > Man. xiá), but Hmongic *ɴɢeD was an earlier borrowing from OC,
and Mienic *hepD (7.13/10) was a later borrowing from MC.

7.15 *ɦ-
PM *ɦ- 8 9 10 11
1. 厚 thick *ɦɔuB ho4 hu4 hau4 hu4
108 Chapter 2

2.3 Proto Hmong-Mien rimes

V u̯ - i̯ - u̯ - i̯ - u̯ - i̯ -
-ø i u̯ i ɨ e ɛ æ
(1) (14) (2) (10) (5)
-j ej ɛj u̯ ɛj æj
(10) (11) (11) (5)
-w æw
(5)
-p ep i̯ ɛp u̯ ɛp æp
(10) (1) (9) (5)
-t et u̯ ɛt
(10) (9)
-k ik ek i̯ ɛk u̯ ɛk æk
(2) (2) (1) (9) (5)
-m im em u̯ em ɛm æm
(18) (20) (29) (22) (20)
-n in i̯ en u̯ en u̯ ɛn æn i̯ æn
(18) (20) (21) (22) (19) (1)
-ŋ iŋ eŋ u̯ eŋ ɛŋ u̯ ɛŋ æŋ i̯ æŋ
(18) (20) (29) (22) (22) (24) (24)

V i̯ - u̯ - i̯ - u̯ - i̯ - u̯ - u̯ -
-ø ə a ɔ o u̯ o ʉ u
(4) (7) (7) (16)
-j əj u̯ əj aj uj
(15) (15) (15) (9)
-w əw ow uw
(7) (16) (3)
-p əp ɔp op up
(7) (29) (13) (9)
-t ət u̯ ət at u̯ at ot ut
(7) (7) (4) (9) (13) (9)
-k ək ɔk ok uk
(7) (29) (13) (9)
-m i̯ əm u̯ əm am u̯ ɔm
(18) (29) (24) (27)
-n ən i̯ ən u̯ ən an ɔn un
(21) (18) (29) (19) (24) (27)
-ŋ əŋ i̯ əŋ u̯ əŋ aŋ i̯ aŋ ɔŋ i̯ ɔŋ u̯ ɔŋ oŋ ʉŋ uŋ
(21) (18) (29) (24) (24) (29) (26) (29) (28) (30) (28)
 Proto Hmong-Mien 109

VV u̯ - i̯ - u̯ -
-ø ei u̯ ei eu (2) i̯ eu æi u̯ æi æu
(12) (8) (2) (12) (12) (3)
-k
-t
-m eəm ɛɔm
(23) (22)
-n ein
(25)
-ŋ iəŋ iuŋ euŋ ɛɔŋ ɛuŋ
(23) (23) (23) (22) (22)

VV u̯ - i̯ - i̯ -
-ø əi u̯ əi əu i̯ əu au (3) ɔi ɔu (3) ou i̯ ou ui (8) uə
(10) (12) (3) (1) (6) (13) (7) (16)
-k əuk
(6)
-t əut
(13)
-m əum
(21)
-n əan əun
(21) (21)
-ŋ əaŋ əuŋ
(21) (21)

Proto Hmongic rimes

Oral Nasal
i (1) ɨ (2) ʉ (8) u (16) in (18) ɨŋ (23) ʉŋ (30) uŋ (27)
e (10) o (7) en (20)/ein (25) oŋ (28)
ɔ (6) ɛŋ (22) əŋ (21) ɔŋ (29)
æ (5) a (4) æn (19) aŋ (24)/i̯ aŋ (26)

uw (9) u̯ u (14)
ow (13) u̯ ei (12/17)
æw (3) u̯ ɛ (11)
u̯ a (15)
Rime tables
The rimes of Mienic languages are much more complex than the rimes of Hmongic
languages: not only do Mienic languages have final stop consonants and final -m
(Hmongic languages only have -n and -ŋ at most), they also show more diphthongs and on-
glides.27 The protolanguage is assumed to have looked more like Mienic in its rime
inventory, and this assumption is reflected in the tables above. Chinese loanwords have
introduced even more rimes: these rimes appear in the tables in italics. Further research
may reveal that other rimes should be also be attributed to the influence of Chinese, which
would simplify the picture above.

Correspondence sets
The correspondence sets below follow the numbering of the original 30 Proto Hmongic
rime correspondence sets in Wang 1994. In total, however, there are only 28 sets: set 17
has been combined with 12, and set 26 contains only Chinese loanwords.28 The reason for
using the simpler Hmongic rime correspondences to organize the more complex Hmong-
Mien rime correspondences is given fully in the introduction to the reconstruction above
(section 2.1.3.1). Briefly, this organization of the data makes it possible to see at a glance
that in the development of Hmongic, each Hmong-Mien rime merged into one of only a
small number of rimes. Thus, each of the super-sets is divided into sub-sets according to
differences in the Mienic correspondences which more directly represent the Proto
Hmong-Mien rime values. Given such a transparent presentation of the data, the reader
will be able to evaluate the reconstruction of each set of Proto Hmong-Mien rimes by two
primary measures:
(1) The reconstruction of each Hmong-Mien sub-set must reflect its more conservative
Mienic reflexes in a direct and natural way29
(2) Hmong-Mien sub-set reconstructions must share certain phonological features
which would have made their merger into a single Hmongic rime natural.
Information on the Chinese loanwords that belong to each sub-set have also been included.
Middle Chinese forms are the transcriptions of Baxter (2000), and Old Chinese
reconstructions represent work in progress by Baxter and Sagart (2009). These provide yet
a third measure for evaluating the reconstruction:
(3) Loanwords must have fallen into sets of native words that were phonologically
similar.
The first (“a”) sub-set within each super-set will contain words attested in Hmongic alone,
and will illustrate the merger target pattern. The next sub-sets will contain words attested
in both Hmongic and Mienic, each with a different array of reflexes in Mienic, requiring a
different Hmong-Mien reconstruction. Words attested in Mienic alone will be included if
they correspond to Mienic words that have cognates in Hmongic and thus clearly belong in

27
Vowel-length contrasts in Mun and Mien increase the complexity of these rime inventories. Vowel length
has not been reconstructed for Hmong-Mien, however: see section 2.1.3.2 for discussion.
28
And set 14 consists of only one word, ‘two’.
29
Although Mienic rimes are in general closer to those of Proto Hmong-Mien, there have been a number of
mergers into Mienic as well, as described in the introduction to the reconstruction, section 2.1.3.1.
 Proto Hmong-Mien 109

a particular sub-set, or, in some instances, if the reconstructed Mienic rime is closely
related to the other rimes in a particular super-set. At the end of this section, three small
Mienic sets are presented that display correspondences not clearly represented in any of the
super-sets that bridge the Hmong-Mien divide.
The words illustrating the rime correspondences are from Wang and Mao 1995 and
Wang 1994, with the addition of a few others. The main objective in this work is not to
present more data, but to organize it in such a way that the rime correspondence patterns
and the justification for the reconstructions can be clearly understood, evaluated, and
improved by others.

Languages
For reasons of space and ease of use, the same eleven maximally distinct languages (seven
Hmongic, four Mienic) used in section 2.2 on Hmong-Mien onsets were chosen to
represent all the languages of the Hmong-Mien family. However, given the greater
variation in vowel reflexes, using such a small sample makes it difficult to see the rationale
behind some of the rime reconstructions; the reader may wish to consult Wang and Mao
1995, which includes forms from 23 sampling locations. If a form was taken from Wang
and Mao 1995, Wang 1994,30 or one of the primary sources listed, no mention is made of
the source of the form. If a form was taken from another source, or if another closely
related dialect was substituted for the one listed here, the source and dialect will be given
in the notes. The eleven criterial languages will be referred to in discussion by the
underlined names.

1. East Hmongic (Qiandong): Northern vernacular: Guizhou, Kaili County, Yanghao

Township, Yanghao
Sources: Wang and Mao 1995, language 1 = Wang 1994, language 1
2. North Hmongic (Xiangxi): Western vernacular: Hunan, Huayuan County, Jiwei
Township, Jiwei
Sources: Wang and Mao 1995, language 2 = Wang 1994, language 2
3. West Hmongic (Chuanqiandian): CQD subdialect, #1 vernacular: Laos/Thailand,
White Hmong
Sources: Heimbach 1979, Bertrais 1979 [1964]
4. West Hmongic (Chuanqiandian): Mashan subdialect, Central vernacular: Guizhou
Province, Ziyun Miao/Buyi A.C., Zongdi Township, Zongdi
Sources: Wang and Mao 1995, language 7 = Wang 1994, language 7 (tones 1a,
3a, 5a, 7a < voiceless stop, preglottalized stop, and preglottalized sonorant onsets;
tones 1b, 3b, 5b, 7b < voiceless aspirated stop, voiceless fricative, and voiceless
sonorant onsets)
5. West Hmongic (Chuanqiandian): Luopohe subdialect, #2 vernacular: Guizhou Province,
Fuyuan County, Fuyuan
Sources: Wang and Mao 1995 language 8 = Wang 1994, language 8 (no tone
split)

30
For some reason, not all forms from Wang 1994 were included in Wang and Mao 1995. Unless obvious
Chinese loans, they have all been incorporated here.
110 Chapter 2

6. Hmongic: Guangxi/Zhuang A. R., Jinxiu Yao A. C., Changdong Township, Jiongnai

Sources: Wang and Mao 1995, language 13; Mao and Li 2001
7. Hmongic: Guangxi/Zhuang A. R., Miao A.C., Rongshui District, Baiyun Pa-Hng
Source: Niederer 1997
8. Mienic: Mien, Luoxiang vernacular: Guangxi/Zhuang A. R., Jinxiu Yao A.C., Luoxiang
Township, Mien
Sources: Wang and Mao 1995, language 17; L-Thongkum 1993 “JX”
(tone 5 < voiceless stop, preglottalized stop, and preglottalized sonorant onsets;
tone 5’ < voiceless aspirated stop, voiceless fricative, and voiceless sonorant
onsets)
9. Mienic: Jinmen: Guangxi/Zhuang A. R., Lingyun County, Lanjin Township, Mun
Source: Wang and Mao 1995, language 20 (tones 1, 3, 5, 7 < voiceless stop,
preglottalized stop, and preglottalized sonorant onsets; tones 1’, 3’, 5’, 7’ <
voiceless aspirated stop, voiceless fricative, and voiceless sonorant onsets)
10. Mienic: Biaomin: Guangxi/Zhuang A. R., Quanzhou county, Dongshan Yao Township,
Biao Min
Sources: Wang and Mao 1995, language 21; Solnit 1982
11. Mienic: Zaomin: Guangdong Province, Liannan Yao A. C., Daping Township,
Zao Min
Source: Wang and Mao 1995, language 23
 Proto Hmong-Mien 111

1
1a. H *i
1 2 3 4 5 6 7 H
1. to grow (TR) tɕen4 – – – ʑeB tei4 – *ɟiB
2. noose/trap – – tʂi4 – – – – *briB
3. tea tɕen4 ci4 – tɕi4 ʑiB tʃi4 tɕi4 *gjiB
4. to bear fruit tsen5 – tsi5 pei5a pzeC – – *pjiC
5. late – – li6 li6 – – lɦe6 *liC

Loan in set 1a:

tsyek (MC) 炙 zhì ‘to roast/broil’ > H *ciC ‘to bake/toast’
> H *ɟiC ‘to burn/be alight’

1b. HM *i (1, 18)31

H *i : HM *i : M *i
1 2 3 4 5 6 7 8 9 10 11 HM/M
6. bedbug – – – – – pi1 pɪ5 pi1 pi1 pi1 bɛi1 *pji
7. stomach ɕi1 tθi1’ – sɛi1 *siA
8. seven ȵi6 ŋji6 ni6 ȵi6 *ŋjiC

Loans in set 1b:

nej (MC) 泥 ní ‘mud’ > M *hniA
tejX (MC) 底 dǐ ‘bottom/base’ > M *ʔdiB ‘lower level’
dzi (MC) 慈 cí ‘love/mother’ > M *ndjiB ‘mother/aunt’
dojH (MC) 袋 dài ‘bag’ > M *diC

1c. HM *i̯ æn (81)

H *i : HM *i̯ æn : M *æn
1 2 3 4 5 6 7 8 9 10 11 HM
9. 3SG (he/she/it) nen2 – nɯ4 ȵi4 nenB nɐŋ4 nɯ4 nan2 nen2 nin2 – *ni̯ æn(X)

Loan in set 1c:

*[k]ˤon (OC) > kwan (MC) 冠 guān ‘crest/comb’ > HM *ʔwi̯ æn

1d. HM *i̯ əu (210)

H *i : HM *i̯ əu : M *i̯ əu
1 2 3 4 5 6 7 8 9 10 11 HM/M
10. gall bladder ɕen1 tɕi1 tʂi1 sei1a tseA – tɕe1 – – – tɛu1 *tsji̯ əu
11. 2PL (you) maŋ2 me2 ne2 mein2 menA – – – njou2 – – *mi̯ əu
12. fruit tsen3 pi3 tsi3 pei3a pzeB pi3 pje3 pjeu3 pjɔu3 pjau3 bɛu3 *pji̯ əuX
13. ant ɕeu3 sɔu3 sau3 dziu3 *ɟhi̯ əuB
14. CLF-fields lau5’ – – – *ʔli̯ əuC

31
The numbers following each rime reconstruction represent the rime sets in Wang and Mao 1995 to which
these words are assigned.
112 Chapter 2

Loans in set 1d:

*m.b(r)u (OC) > bjuw (MC) 浮 fú ‘to float’ > M *mbi̯ əuA
*[r]u (OC) > ljuw (MC) 流 liú ‘to flow’ > M *ri̯ əuC

1e. HM *i̯ ɛk (148, 110)

H *i : HM *i̯ ɛk : M *i̯ ɛk
1 2 3 4 5 6 7 8 9 10 11 HM/M
15. chaff/husk – – – – – – m̥ e5 bwa7 va7’ bja7 bjɛ7 *mphi̯ ɛk
16. black kje7 kja7 kja7 kjɛ7 *qri̯ ɛkD

Loans in set 1e:

tsyak (MC) 灼 zhuó ‘to burn’ > M *dji̯ ɛkD
tsyek (MC) 隻 zhī ‘CLF-birds’ > M *tsi̯ ɛkD
syik (MC) 識 shí ‘to know/recognize’ > M *tsi̯ ɛkD
tsyik (MC) 織 zhī ‘to weave’ > M *tsi̯ ɛkD
sik (MC) 媳 sī ‘daughter/girl’ > M *ɕi̯ ɛkD
bak (MC) 薄 bó ‘thin’ > M *bi̯ ɛkD

Discussion
In this and in subsequent correspondences, the initial element of the Hmong-Mien rime is
continued into Hmongic. This is in keeping with the general tendency for Hmongic to
retain the beginning of the Hmong-Mien syllable and Mienic to retain the end of the
Hmong-Mien syllable. For many of the words in rime 1, one could have chosen to
reconstruct the medial high front element as -j- , part of the onset, instead of an on-glide to
the vowel, part of the rime. Given the fact that words in this set merged to Hmongic *-i
whereas other words with medial -j- did not, these words are reconstructed with the high
front element already associated more closely with the vowel.
Another rime which fits the pattern above is Hmong-Mien *-i̯ ɛp, represented only by
Mienic *ri̯ ɛpD (2.57/1) ‘to erect’ from Chinese 立 ‘to stand’ (OC *(kə.)rəp > MC lip >
Man. lì).

Notes
2. Xianjin Hmong (Wang and Mao 3) has been substituted for White Hmong (3). Other
forms appear in Wang 1994.
4. ‘To bear fruit’ is a C-tone derivative of ‘fruit’ (rime 1d); see chapter 4.
5. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
8. This is from Tibeto-Burman *s-nis (Benedict 1987a:13). HM *djuŋH (2.18/28) may be
from the same source, but the correspondence is difficult. See chapter 5, section 2.
9. The 3SG pronoun is irregular in rime and in tone (tone 4 in Hmongic, tone 2 in Mienic).
The placement of this word in rime 1 is based on the Hmongic reflexes; given this analysis,
the nasal codas that appear in Hmongic must be taken as secondary.
11. The Hmongic forms are from Wang 1994 and the Mun form (9) is from Shintani and
Yang 1990.
12. Compare derivative ‘to bear fruit’ in set 1a above.
14. Also Jiangdi Mien /ljou5/ and Xiangnan Mien /ljəu5/. Compare Gaopo (Hmongic) /loŋ5/
‘field’ and Tibeto-Burman *low ‘field’ (Matisoff 2003).
 Proto Hmong-Mien 113

15. This word is probably related to Hmongic *S-phjæC ‘chaff/husk’ (1.17/5). The -w- in
Luoxiang Mien (8) is secondary: -j- dropped, then -w- developed between a labial and /a/.
16. This is no doubt the same as Hmongic *qrɛŋA ‘black’ (6.46/22).
114 Chapter 2

2
2a. H *ɨ
1 2 3 4 5 6 7 H
1. clean sha1 ntsha1 – ntsei1b nʔtsheA θei1 – *ntshɨA
2. that/the former i3 a3 i5 i3a eB – – *ʔɨB
3. what ɕi3 – tʂi5 si7b tsiB – – *tsjɨB
4. whorl at top of
ʑi6 – ji4 ʑi4 weB – – *jɨB
head
5. hoop tha5 – thi5 tei5b theC – – *thɨC
6. to squeeze/pinch la5 lɑ5 li5 lei5a – l̥ i5 la5 *ʔlɨC
7. wildcat ȴ̥aŋ7 – pli7 pɭei5a pleC – – *pljɨD

Loan in set 2a:

sje (MC) ‘mutually’ 廝 sī ‘with each other’ > HM *sjɨA-D RECIPROCAL

2b. HM *ɨ (35)
H *ɨ : HM *ɨ : M *ɨ
1 2 3 4 5 6 7 8 9 10 11 Initial HM
8. one i1 ɑ3 i1 ei1a – i3 – – – i1 a1 7.1 *ʔɨ

Loan in set 2b:

liX (MC) 里 lǐ ‘half kilometer’ > HM *ljɨX

2c. HM *eu (35)

H *ɨ : HM *eu : M *eu
1 2 3 4 5 6 7 8 9 10 11 HM
9. to fry ka1 ca1 ki1 tɕi1a tɕeA – – – [tjiː1] khla1 – *kleu
10. butterfly – – – mpei5a mʔpleC – – bjeu5’ blɔu5’ bja4 – *mpleuH

2d. HM *i̯ eu (2, 35)

H *ɨ : HM *i̯ eu : M *i̯ eu
1 2 3 4 5 6 7 8 9 10 11 HM/M
11. lightweight fha1 ɕa1 ʂi1 sei1b seA v̥ uai1’ fɤ1 ɕeu1 sɔu1’ hjau1 – *-sji̯ eu
12. scissors jeu3 ȡiːu3 gjau3 gɛu3 *ʔglji̯ euB

2e. HM *ik (15)

H *ɨ : HM *ik : M *it
1 2 3 4 5 6 7 8 9 10 11 HM
13. flip with finger – – nti3 ntei5a – ntai5 na5 dit7 – – – *ntik

2f. HM *ek (17, 34)

H *ɨ : HM *ek : M *ek
1 2 3 4 5 6 7 8 9 10 11 M
14. to dry (clothes) – si7 ʨi7 tsaŋ1 *tsekD
 Proto Hmong-Mien 115

Loans in set 2f:

phek (MC) 劈 pī ‘to split/chop’ > HM *phek
tsyhek (MC) 赤 chì ‘red’ > M *sekD

Discussion
In set 2c above, *ik is reconstructed for Hmong-Mien, despite the reflexes in -it in Mienic,
because the Hmongic reflexes have tone 5: Hmong-Mien D-tone syllables ending in -k
regularly go to the corresponding register of the C-tone in Hmongic.

Notes
3. The variability of tone may be due to this word rarely being used in isolation. For
example, in White Hmong (3) it is part of the unanalyzable word dabtsi ‘what’, and tone
sandhi rules would explain a change from tone 3 to tone 5.
4. In White Hmong, this means ‘cowlick’ (an errant patch of hair growing anywhere).
5. Possibly from Chinese 帶 ‘belt, girdle’ (MC tajH > Man. dài), although the Chinese
initial is not aspirated.
7. Perhaps the same as Chinese 狸 ‘wildcat’ (OC *[m]ə.rə > MC li > Man. lí) from a
variant bù lái indicating an initial *pə- (Sagart 1999:88). In Hmong-Mien the word would
have had a final -p or -t, so this may be a loan from Hmong-Mien to Chinese.
8. ‘One’ is highly irregular, which is to be expected since it is both short and frequently
used. It is the same as Chinese 一 ‘one’ (OC *ʔi[t] > MC ʔjit > Man. yī); see chapter 5. The
Jiongnai (6) form is from Mao and Li 2001.
9. Possibly from Chinese 攪 ‘disturb, stir’ (OC *kˤruʔ > MC kæwX > Man. jiǎo), but the
tones do not correspond. Biao Min /khla1/ (Solnit 1982) means ‘turn over, stir’ (of drying
grain, stir frying), hence this reconstruction with medial -l-, despite the problem with
aspiration. In Hainan Mun ‘to fry in oil or fat’ is /tjiː1/. The rime does not correspond—it is
probably a late borrowing from 煎 ‘to fry’ (Man. jiān).
13. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
Although some Mienic forms end in -t, the tone correspondence 7 Mienic to 5 Hmongic is
usually indicative of final -k.
 116 Chapter 2

3
3a. H *æw
1 2 3 4 5 6 7 H
1. to sow – pʐo1 tʂe1 pʐæ1a pjiA – – *præwA
2. steel sha1 sɯ1 – sæ1b – – tɕo1 *tshæwA
3. to sweep ʨhi1 – che1 ʨæ1b ʨhiA – – *chæwA
4. river – – de2 – ʁleiA – – *ɢlæwA
5. body/trunk ʨi3 ʨɯ3 ce3 ʨæ3a ʨiB – – *cæwB
6. to burn (TR) phi3 – – – – v̥ a3 pho3 *phjæwB
7. wife/elder sister vi3 – vi3 – – va3 vo3 *ʔwæwB
8. leg/branch ʨi4 kɯ4 ce6 ʨe4 ʑaB – – *ɟæwB
9. mountain pi4 – pe8 pe4 veiB – – *bæwB
10. to carry on
ʨi5 – – ʨæ5a – – – *cæwC
shoulder
11. to climb ʨi5 ȵʨɯ5 ɲce5 ȵʨæ5a ȵʔʨiC ntja5 – *ɲcæwC
12. egg ki5 – qe5 hæ5a qwjiC kja5 ko5 *qæwC
13. frost ta5 tɯ5 te5 tæ5a tiC ða5 no5 *tæwC
14. near ɣi5 ʐɯ5 ʐe5 ʐæ5a ʔwjiC ŋkja5 jo5 *-ʔræwC
15. to soak – ȵʨɯ5 ɳtʂe5 ntsæ5a nʔʦiC ɳʧa5 ȵʨo5 *ntsjæwC

Loans in set 3a:

*mə-qˤ(r)o (OC) > kuw (MC) 鉤 gōu ‘hook’ > H *ɴqæwC
*[tsʰ]rək (OC) > tsrhik (MC) 畟 cè ‘sharp-cutting’ > H *ntsjæwC ‘sharp’
*N-t<r>ək (OC) > drik (MC) 直 zhí ‘straight/vertical’ > H *ndzjæwC ‘vertical/upright’32

3b. HM *æu (113, 259)

H *æw : HM *æu : M *æu
1 2 3 4 5 6 7 8 9 10 11 HM/M
16. ear zɛ2 mʐɯ2 ɳtʂe2 mpʐæ2 mpjiA mpja2 mjo2 – – blau2 bju2 *mbræu
17. mushroom ʨi1 ŋkɯ1 ɲce1 ȵʨæ1a ȵʔʨiA ɳtʃa1 ŋo1 ʨeu1 sɔu1 ȶau1 ku1 *ŋkjæu
18. pillar ȵi2 ȵɯ2 ɲce2 ȵʨæ2 ȵʨiA – ŋo2 – – ʨɔu2 – *ɲɟæu
19. uncle ʑu5 ʑo4 – – – ja2 jo4 – jɔu4 – – *jæuX
20. to sharpen
ʥeu5 gjɔu5’ ȡau5 dziu5 *dhjæuC
/grind
21. to sweep – ȡɔu5 – – *ʔɟæuC

Loan in set 3b:

gjew (MC) 蕎 qiáo ‘buckwheat’ > HM *ɟæu

32
‘Upright/erect’ is attested only in Xianjin and Shimen (Wang 1994) and in White Hmong (ntseg
‘erect/vertical/steep’), closely related West Hmongic dialects.
 Proto Hmong-Mien 117

3c. HM *əu (96)

H *æw : HM *əu : M *əu
1 2 3 4 5 6 7 8 9 10 11 HM/M
22. earth ta1 tɯ1 te1 tæ1a tiA ta1 to1 dau1 – – – *N-təu
23. stone ɣi1 ʐɯ1 ʐe1 ʐæ1a ʔwjiA ŋkja1 jo1 gau1 gjau1 lau1 dzu1 *-ʔrəu
24. long (time) la2 lɯ2 le2 læ2 liA la2 lo2 lau2 lau2 – lu2 *ləu
25. riceplant
na2 nɯ2 mple2 mplæ2 – mpla2 mjo2 blau2 blau2 blau2 bjau2 *mbləu
/paddy
26. house tsɛ3 pʐɯ3 tʂe3 pʐæ3a peiB pja3 pjo3 pjau3 pjau3 pla3 pju3 *prəuX
27. road/way ki3 kɯ3 ke3 kæ3a ʨiB kja3 ko3 kjau3 kjau3 kla3 tsu3 *kləuX
28. salt ɕi3 ȵʨɯ3 ɳtʂe3 ntsæ3a nʔʦiB – ȵo3 dau3 dau3 dza3 – *ntsjəuX
29. fish zɛ4 mʐɯ4 ɳtʂe4 mpʐe4 mpjiB mpja4 mjo4 bjau4 bjau4 bla4 bju4 *mbrəuX
30. nest ɣi4 ʐɯ4 ʐe4 ʐe4 wjiB ŋkja4 – gau4 gjau4 la4 – *-rəuX
31. to lay eggs na6 – nte6 nte6 ntiC tei5 no6 – – dau6 du6 *ndəuH
32. egg kjau5’ kjau5 klau5 tsu5 *kləuC

Loan in set 3c:

*ʔˤuʔ (OC) > ʔawX (MC) 媼 ǎo ‘old woman’ > HM *ʔəuX ‘elder sister/wife’

3d. HM *au (76, 95, 177)

H *æw : HM *au : M *au
1 2 3 4 5 6 7 8 9 10 11 HM/M
33. to answer ta1 tɯ1 te1 tæ1a tiA ta1 no1 – tau1 – du1 *tau
34. long ta3 ntɯ3 nte3 ntæ3a nʔtiB ða3 to3 daːu3 daːu3 da3 du3 *ntauX
35. to warm self
ta5 ntɯ5 nte5 ntæ5a nʔtiC nta5 no5 – daːu5 dau5 du5 *ntauH
by fire
4
36. mouse/rat naːu – – – *nauB

3e. HM *ɔu (77, 244)

H *æw : HM *ɔu : M *ɔu
1 2 3 4 5 6 7 8 9 10 11 HM/M
37. three pi1 pu1 pe1 pæ1a pziA pa1 po1 pu1 pu1 pau1 bu5 *pjɔu
38. 1PL
pi1 pɯ1 pe1 pæ1a peiA pa1 pɯ5 – pu1 – bu1 *N-pɔu
(we/us)
tɛ-
39. hand/arm pi4 tɯ4 te4 ʂe4 weiB tʃa4 pu4 pu4 pau4 pu4 *-bɔuX
pɯ4
40. name pi5 mpu5 mpe5 mpæ5a mʔpeiC mpa5 mo5 bu5’ bu5 bau5 bu5 *mpɔuH
41. table du2 nu2 – – *ndɔuA
42. blind bu3 bu4 – – *ʔblɔuB
43. to burn pu3 pu3 pau3 bu3 *pɔuB
44. grey bu2 bu3 bau3 – *ʔbɔuB
45. to wait ʨu3 ȶu3 ȶu3 – *tjɔuB
46. brother mai4 – mau4 mu4 *mɔuB

Loan in set 3e:

huwX (MC) 厚 hòu ‘thick’ > M *ɦɔuB
118 Chapter 2

3f. HM *uw (115)

H *æw : HM *uw : M *u
1 2 3 4 5 6 7 8 9 10 11 HM/M
47. far – qɯ1 de1 hwæ1a qweiA kwa1 ko1 ko1 ku1 ku1 kɔu1 *qʷuw
48. dog l̥ a3 qwɯ3 de3 l̥ æ3a qleiB kla3 ka-ljɔ̃ klo3
7
klu3 klu3 ku3 *qluwX
49. good/tasty – ku3 – kɔm3 *khuB

Discussion
The details of the phonetic values assigned to the subsets of set 3 are approximate: the
general pattern of a back vowel followed by a back rounded vowel is what they all share,
however, and what allowed them to merge into Hmongic *-æw.

Notes
3 & 21. These two sets of forms meaning ‘to sweep’ have the same rime and place of
articulation, but differ in tone and aspiration.
5 & 8. These two words appear to be morphologically related. The difference in initial
voicing may have been due to a prefix; see chapter 4.
7. ‘Elder sister’ in White Hmong (3).
9. White Hmong ‘up there on the slope’. The tone in White Hmong is the result of a recent
morphological class formation by tone-shift (geographical feature > locative) (Ratliff
1992a:104-112).
12 & 32. These two words for ‘egg’ are no doubt the same; although there are difficulties
with the onsets, their rimes and tones correspond.
15. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
17. Compare Chinese 菇 ‘mushroom’ (OC *mə.kˤa > MC ku > Man. gū).
18. The Yanghao form (1) is from Office of Miao-Yao Research Lexicon 1987. Sunjiang
Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min (10).
20. Perhaps from Chinese 剉 ‘to file’ (MC tshwaH > Man. cuò).
21. This may be from Chinese 帚 ‘broom’ (MC tsyuwX > Man. zhǒu), although the initial
voicing suggests *ɲc-, and the tones are different.
22. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
23. Also ‘grindstone’ in White Hmong.
24. The Green Mong form meaning ‘a long time ago’ has been substituted for White
Hmong (3).
29. Compare Proto Tai *pla A1 (Li 1977), Proto Kam-Sui *mprai 3 (Thurgood 1988).
36. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
Hmongic *naŋB ‘mouse/rat’ (2.9/24) is probably the same.
37 & 38. See discussion of the similarity between‘three’ and ‘we/us’ in chapter 5. ‘We/us’
is used as a plural suffix in Mien, suggesting an original meaning of ‘group’. In Hainan
Mun, it is a 1PL exclusive pronoun.
44. In Hainan Mun, the cognate means ‘blue’ and the word for ‘grey’ is the compound
‘white-blue’; in Biao Min, it also means ‘mildew, mold’ (Solnit 1982).
47. Compare Chinese 迂 ‘bend, deflect; astray, far from’ (OC *qw(r)a > MC ʔju > Man.
yū).
48. This word for ‘dog’ appears in languages across the area: compare Chinese 狗 ‘dog’
(OC *-[k]ˤ(r)oʔ > MC kuwX > Man. gǒu) and see Benedict 1996.
 Proto Hmong-Mien 119

4
4a. H *a
1 2 3 4 5 6 7 H
1. cherry va1 wɑ1 – ʑi1a – – – *ʔwaA
2. honey va1 – ʐi1 ʐɪ1a ʔwjaA – – *ʔraA
3. rough/coarse sha1 ntshɑ1 ntshi1 ntsi1b nʔtshaA – – *ntshaA
4. soul/spirit ȴ̥u2 pjə2 pli6 pɭɪ6 vloA – pjɔ2 *bljaA
5. finger ta3 ntɑ3 nti3 nti3a nʔtaB – – *ntaB
6. to pour tsɑ5 pʐa5 tʂi5 – – – – *praC
7. to lean (on stick) ȵa6 ɳɑ6 ɳʈi6 ɳʈe4 ɳtʂaC – ȵa2 *ŋgraC
8. to escape fa8 qwei8 di8 hwɪ8 ʁwaD – – *ɢʷaD
9. to wake – – tʂi8 si8 zaD – ʨi8 *dzjaD

Loans in set 4a:

*[g]ˤa ‘gourd’ (OC) > hu (MC) 葫 hú ‘garlic’ > H *ɢaA ‘garlic’
*mə.qˤ<r>aʔ-s (OC) > kæH (MC) 價 jià ‘price’ > H *ɴqaC

4b. HM *a (92, 130, 149, 187)

H *a : HM *a : M *a
1 2 3 4 5 6 7 8 9 10 11 HM/M
10. five tsa1 pʐɑ1 tʂi1 pʐɪ1a pjaA pui1 pja1 pla1 pja1 pla1 pjɛ5 *pra
11. father/male pa3 pɑ3 tsi3 pi3a paB pe3 pa7 pwa3 fa3 – – *pjaX
12. stick – pʐɑ3 – – – – pja3 pla3 pja3 – – *praX
13. to get up – – – – – tʃe4 – kja4 kwe4 ȶɔ4 kjɛ4 *gʷjaX
14. moon/month l̥ ha5 l̥ hɑ5 hl̥ i5 li5b l̥ aC l̥ e5 l̥ a5 la5 la5 ɬa5 lɔu5 *hlaH
15. elder brother ta2 – ti2 ti2 ðaA – – – ta2 – – *da
16. urine va4 ʐɑ4 ʐi4 ʐɪ4 wjaB ŋkwe4 ve4 wa4 va4 lɔ4 vjɛ4 *-raX
17. 1SG (I/me) ja1 ja1 – – *ʔjaA
18. grass m̥ wa3 wa3 m̥ ja3 mjɛ3 *hmjaB
19. meat a3 a3 – – *ʔaB
20. to carry on back ȵa5’ ȵa5 ȵa3 – *hɲaC

Loans in set 4b:

*kʷˤra (OC) > kwæ (MC) 瓜 guā ‘melon/gourd’ > HM *Kʷa ‘cucumber’
gja (MC) 茄 qié ‘eggplant’ > HM *ɟa
*Cə.kˤraʔ (OC) > kæX (MC) 假 jiǎ ‘to borrow’ > HM *KaX
*(mə)-pˤaʔ (OC) > puX (MC) 補 bǔ ‘repair/mend’ > HM *mpjaX
*[g]ˤraʔ ‘down’ (OC) > hæX (MC) 下 xià ‘low/short’ > HM *GaX
*[g]ˤraʔ ‘down’ (OC) > hæX (MC) 下 xià ‘to descend’ > H *ɴɢaB
*mə-[g]ˤraʔ-s ‘to descend’ (OC) > hæH (MC) 下 xià ‘to descend’ > M *ɣaC
bæ (MC) 耙 pá ‘to rake’ > M *baA
dræ (MC) 茶 chá ‘tea’ > M *ɟaA
sjæX (MC) 寫 xiě ‘to write’ > M *xjaB
120 Chapter 2

4c. HM *at (88, 107, 127 and 221)

H *a : HM *at : M *at
1 2 3 4 5 6 7 8 9 10 11 HM/M
21. wing ta7 tei7 ti7 ti7a taD ðe7 te7 daːt7 daːt7 – dɔt7 *N-tat
22. spicy za8 mʐei8 ɳtʂi8 mpʐɪ8 mpjaD mpwai8 mje8 blaːt8 bjat8 blan8 bjɛt8 *mbrat
23. eight ʑa8 ʑi8 ji8 ʑi8 ʑaD je8 ji8 jat8 jet8 hjɛn8 dzjat8 *jat
24. wooden
pwat7 – – – *pjatD
basin

Loans in set 4c:

*Nə-qʰˤat (OC) > khat (MC) 渴 kě ‘thirsty’ > HM *NKhat
*mə-kˤat (OC) > kat (MC) 割 gē ‘to cut’ > M *katD

4d. M *ak (184)

1 2 3 4 5 6 7 H HM M 8 9 10 11 Initial M
25. thick/dense *ak ma8 ma6 – mɔu8 1.9 *makD

Discussion
The most interesting feature of this rime set is the way that Jiwei (Qo Xiong) (2) and Pa-
Hng (12) pattern. They show a rime difference that reflects whether or not the Hmong-
Mien rimes of this set (as reflected in the Mienic forms) were open or closed. Open
Hmong-Mien rime 4b yields Jiwei /ɑ/ and Pa-Hng /a/, whereas the closed Hmong-Mien
rime 4c yields Jiwei /ei, i/ and Pa-Hng /e, i/. The reflection of the open/closed Hmong-
Mien contrast as a vowel quality difference can also be seen in Jiwei in rimes 12/17 and
13, and in Pa Hng in rime 7. This suggests that Jiwei and Pa-Hng may have split off the
main Hmongic branch at an early stage (see Strecker 1987a,b on the conservatism of “Na-
e”, or Pa-Hng).
The effect of the medial *-r- in ‘spicy’ and initial *j- in ‘eight’ on the rime *-at account for
the irregularities in set 4c.

Notes
3. Compare Chinese 粗 ‘coarse’ (OC *s.[r̥ ]ˤa > MC tshu > Man. cū).
4. Compare Tibeto-Burman *b-la ‘demon, soul’ > Nung /phəla/ (Benedict 1972). Tone 6
forms can be explained as sandhi form promotions, but the rime is highly irregular.
5. Mienic *ʔdokD ‘finger’ (2.4/13) may be related.
6. Forms taken from Office of Miao-Yao Research Lexicon (1987). The details of the low
vowels in languages 1 and 2 do not correspond, but the word seems to fit in this set given
the White Hmong cognate /tʂi5/ ‘to pour’.
7. Also Ho Ne /ki4/. White Hmong (3) also ‘a stick, cane’.
8. The labialized velar marks this as a loanword. According to Sagart (1999:123–124), it is
a loanword from 脫 ‘to peel off, escape’ (OC *lˤot > MC dwat > Man. tuō). Mortensen
(2002) proposed Tibeto-Burman *g-lwat ‘free, release’ as the source, which works better
since it has a dorsal consonant (for Sagart, the dorsal consonant is a secondary
development of a hypothetical intermediate stage *dl- in Chinese). Note that this Chinese
word with a voiceless initial is taken here as the source of Mienic *ʔdutD (< *N-t-) ‘peel
off/escape’ (2.4/9).
 Proto Hmong-Mien 121

10. From Tibeto-Burman *b-r-ŋa ‘five’ (Benedict 1987a). See chapter 5.

11. Compare Chinese 父 ‘father’ (OC *[b](r)aʔ > MC bjuX > Man. fù). See also Hmong-
Mien ‘husband’ *N-poX (1.1/7).
15. Compare Chinese 弟 dì ‘younger brother’, used in the compound 弟兄 dìxiong to mean
‘brothers’.
18. The -w- in Luoxiang Mien (8) and Lanjin Mun (9) is secondary: -j- dropped, then -w-
developed between a labial and /a/.
23. From Tibeto-Burman *-rjat ‘eight’ (Matisoff 2003). See chapter 5.
24. *pj- is reconstructed on the basis of Mien forms /pjet7/ and /pje7/ (Wang and Mao 15 &
16). The -w- in Luoxiang Mien (8) is secondary: -j- dropped, then -w- developed between
a labial and /a/.
122 Chapter 2

5
5a. H *æ
1 2 3 4 5 6 7 H
1. bridge – – – la1b – – – *hlæA
2. thick ta1 ta1 tua1 ta1a tiA tei1 te1 *tæA
3. thigh pa1 pa1 pua1 pa1a – – pe1 *pæA
4. ant phen3 mpha3 – – – mphai3 m̥ je3 *mphæB
5. to hold/grasp
– – tua3 ta3a tiB – – *tæB
hand
6. to hold in mouth – – mpua3 mpa3a mʔpziB – – *mpæB
7. billhook sa5 – tsua7 sa7a tsiC – – *tsæC
8. chaff/husk fha5 sa5 sua5 sa5b siB v̥ ei5 m̥ e5 *S-phjæC
9. pig pa5 mpa5 mpua5 mpa5a mʔpeiC mpei5 me5 *mpæC
10. rope l̥ ha5 l̥ ha5 hl̥ ua5 la5b l̥ iC l̥ ei5 l̥ e5 *hlæC
11. bland/tasteless ɕen6 – tʂua6 səa6 ziC ʃi6 tɕi6 *dzjæC
12. comb ɣa6 ʐa6 ʐua6 ʐəa6 wjiC vi6 ɦji6 *ræC
13. to have food
– – ntsua6 mpje6 mpjiC – – *mbljæC
w/rice
14. to join sen7 tshɑ7 tsua7 sa7a – – – *tsæD
15. to rob – – – ɦa7b hiD – – *hæD
16. tin – – tshua7 sa7b – – – *tshæD
17. young – – hl̥ ua7 ɭa7b – – – *hljæD

Loans in set 5a:

*mə-pʰˤrak (OC) > phæk (MC) 拍 pāi ‘to strike’ > H *mbæA ‘to clap’
dap (MC) 蹋 tà ‘to trample/tread’ > H *dæD ‘to step on/tread’

5b. HM *æ (54, 149)

H *æ : HM *æ : M *æ
1 2 3 4 5 6 7 8 9 10 11 HM
18. molar tooth – pa2 pua1 – – – – – – – ba1 *pæ
19. mountain pass – – – – – e7 – ɛ7 e1 ɛ7 a1 *ʔæ(C)

Loan in set 5b:

ngwæX (MC) 瓦 wǎ ‘tile’ > HM *ŋʷæX

5c. HM *æj (169)

H *æ : HM *æj : M *aj
1 2 3 4 5 6 7 8 9 10 11 HM
20. dry/thirsty qha1 – ɴqhua1 ŋka1b ɴʔqheiA ŋkhei1 – gaːi1 gaːi1’ – gɔi1 *NKhæj
 Proto Hmong-Mien 123

5d. HM *æw (114, 152, 171)

H *æ : HM *æw : M *au
1 2 3 4 5 6 7 8 9 10 11 HM/M
21. liquid/soup – ca3 kua5 ka3a tɕiB – – – – kla3 – *klæwX
*N-
22. to play – tsa4 – – – – – dzjaːu6 dzaːu6 dza6 –
dzæwH
*N-
23. wind tɕen5 ci5 cua5 – tɕiC tʃi5 tɕi5 ʥau5 ȡaːu5 hja5 dzau5
cæwH
24. to measure (rice) gaːu1 gaːu1' ɬa1 dzau1 *hrauA
25. young man bjaːu2 baːu2 – – *mbjauA
26. to repay – klaːu3 klau3 kau3 *qlauB

Loans in set 5d:33

*lˤaw (OC) > daw (MC) 桃 táo ‘peach’ > HM *ɢlæw
*[g]aw (OC) > haw (MC) 號 háo ‘to shout/cry out’ > HM *Gæw ‘to sing/cry out’
*[kh]ˤawʔ (OC) > khawX (MC) 槀 gǎo ‘dry/withered’ > HM *KhæwX
*mə.[ts]ˤawʔ (OC) > tsawX (MC) 澡 zǎo ‘to wash (hands)’ > HM *ntsæwX
tsrhæwX (MC) 炒 chǎo ‘to fry’ > M *chauB

5e. HM *æp (86, 144)

H *æ : HM *æp : M *æp
1 2 3 4 5 6 7 8 9 10 11 HM
27. to cut ken7 ɴqɑ7 – – – tʃei7 – kap7 kjap7 klan7 kɛp7 *klæp
28. fan zen8 mjɑ7 ntsua8 mpja8 mpziD mpli8 mi8 bjap8 bjaːp8 bjɛn8 dzjɛp8 *mbljæp

5f. M *æt (164)

1 2 3 4 5 6 7 8 9 10 11 M
29. to twist/rub ɕat7 saːt7’ saːt7’ sjɛt7 *ɕætD

Discussion
There was a merger of Hmong-Mien *au (3d) and *æw (5c) into Mienic *au. ‘To measure
rice’, ‘young man’, ‘to repay’, and ‘to stir fry’ have been placed here, in rime 5c, given a
slightly better fit with the reflexes of Hmong-Mien ‘wind’. But it should be remembered
that these words could be taken back to either *au or *æw in Proto Hmong-Mien.
5g. Hmong-Mien *-æk (72) contains only Chinese loans:

百 bǎi ‘hundred’: OC *pˤrak > MC pæk > HM *pæk

拍 pāi ‘to strike’: OC *mə-pʰˤrak > MC phæk > (HM *mphæk > ) M *bhækD ‘to clap’
白 bái ‘white’: OC *bˤrak > MC bæk > M *bækD

Another Chinese loanword falls into this set:

客 kè ‘guest’: OC *kʰˤrak > MC khæk > HM *Khæk

33
Another Chinese loan which might fit here would be the irregular Mienic word for ‘claw/talon’ (M
*ʔɲauB); Downer (1973) noted the resemblance to Amoy /jiaù/.
124 Chapter 2

Notes
1. See also Shimen /l̥ a1/, Qingyan /l̥ o1/, and Gaopo /l̥ hɑ1/ ‘bridge’ (all West Hmongic).
Compare Tibeto-Burman *(s-)ley ‘bridge, ladder’ (Matisoff 2003), which is a closer
semantic match than Chinese 梯 ‘wooden steps, stairs’ (OC *l̥ [ə]j > MC thej > Man. tī).
8. This word is probably the same as Hmong-Mien *mphi̯ ɛk ‘chaff/husk’ (1.5/1); note the
Pa-Hng (7) form /m̥ e5/ which provides a link.
9. Compare Tibeto-Burman *pʷak ‘pig’ (Matisoff 2003).
11. Tan Trinh Pa-Hng (Niederer 1997) has been substituted for Baiyun Pa-Hng (7).
16. In White Hmong, ‘lead metal’.
17. This word would have ended in either -p or -t (Hmong-Mien tone 7 words ending in -k
merged with tone 5 in Hmongic).
18. In White Hmong, ‘jaw, chin’.
19. A final stop consonant is tentatively reconstructed on the basis of three reflexes with
tone 7.
21. In Biao Min (10), this means ‘dew’.
22. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
23. Jiwei (2) form taken from Office of Miao-Yao Research Lexicon (1987).
25. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
20 & 28. These words may involve expressive phonology.
29. Hainan Mun (Shintani and Yang 1990) has been substituted for Lanjin Mun (9). This
word is probably the same as Hmongic *S-phu̯ aA ‘twist/rub’ (1.2/15).
 Proto Hmong-Mien 125

6
6a. H *ɔ
1 2 3 4 5 6 7 H
1. many/much nɛ5 – ntau5 ntɔ5a nʔtuC ntei5 – *ntɔC
2. shoes – ɕɔ5 khau5 kɔ5b khuC ʃu5 ŋkʉ5 *khjɔC
3. to sieve – ɕɔ5 tʂhau5 – – – – *tshjɔC
4. bracelet – pɔ6 pau6 po6 vuC – po6 *bɔC
5. ant – – ɳtʂau8 mpʐɔ8 mpjuD – – *mbrɔD

Loans in set 6a:

*[d]rak ‘to put on clothes; be drjak zhuó ‘to wear,
> 著 > H *drɔC ‘to hit target’
attached’ (OC) (MC) touch, etc.’
trjak zhuó ‘to wear, H *trɔC ‘to put on/wear
*trak ‘to place’ (OC) > 著 >
(MC) touch, etc.’ (shoes)’

6b. HM *ɔi (131)

H *ɔ : HM *ɔi : M *əi
1 2 3 4 5 6 7 8 9 10 11 HM
6. pangolin ɣo6 ʐɔ6 ʐau6 ʐo6 wjoŋA – ɦje6 gai6 gjai6 – – *rɔiH

6c. HM *əuk (241)

H *ɔ : HM *əuk : M *əuk
1 2 3 4 5 6 7 8 9 10 11 M
7. straw sandals ɕu7 tθu7’ ȶu7 – *səukD
8. to dissolve ju8 ju6 – – *jəukD
9. sword ʥu8 ȡu6 ȡu8 dziu8 *ɲɟəukD

Loans in set 6c:

*tʕək (OC) > tok (MC) 得 dé ‘to get/gain’ > HM *təuk
*truk (OC) > trjuwk (MC) 竹 zhú ‘bamboo’ > HM *ɲcəuk ‘bamboo strip’
*[dz]ʕawk (OC) > dzak (MC) 鑿 záo ‘chisel’ > HM *dzəuk

Discussion
This may have been a diphthong in Proto Hmongic: *ɔw.

Notes
1. There are Mun forms for ‘many/much’ with a final velar nasal (/duŋ5/ or /duːŋ5/) which
correspond in all but the coda. It is not clear how this word should be reconstructed, so
provisionally only the Hmongic forms are given above.
2. In Jiongnai (6) and Pa-Hng (7), ‘straw sandals’. Perhaps from Chinese 屨 ‘sandals,
straw shoes’ (OC *kro-s > MC kjuH > Man. jù), although the Chinese word is not
aspirated. Compare Mienic *sjəukD ‘straw sandals’, #7 this set.
4. Green Mong has been substituted for White Hmong (3).
5. Compare Mon-Khmer *srmuuc (Shorto #873), Proto Tai *mu̯ it D2S, Malay /semut/.
126 Chapter 2

7
7a. H *o
1 2 3 4 5 6 7 H
1. to boil (TR) tho1 – – to1b thuA – – *thoA
2. thunder ho1 so1 sɔ1 so1b suA xu1 – *S-phoA
3. female/woman – – pɔ2 – vuA – – *boA
4. step ȶə2 ʈo2 ʈɔ2 ʈo2 ʐuA tʃa2 tɕa6 *droA
5. to cover self – – vɔ3 wo3a ʔwuB – va3 *ʔwoB
6. to look after ɣə3 – ʐɔ3 ʐo3a ʔwjuB – – *ʔroB
7. to play (flute) – – tʂhɔ3 pʐo3b – – – *phroB
8. to sharpen/grind xhə3 ho3 hɔ3 ho3b huB xu3 ho3 *hoB
9. thread fhə3 so3 sɔ3 so3b suB – – *S-phoB
10. turbid – ɳʈo3 ɳʈɔ5 ɳʈo3a ɳʔtʂuB ɳtʃu4 ȵɔ̃3 *ŋkroB/*ŋgroB
11. cow/ox – ʑu4 – – ʑuB – – *joB
12. thorn pə4 to4 pɔ4 pɯ4 vuB – – *-boB
13. CLF-mouthfuls lo5 – lɔ5 lo5a – lu5 – *ʔloC
14. to remember – – ɲcɔ5 ȵʨo5a ȵʔʨuC – – *ɲcoC
15. to rest ʨhə5 ɕo5 ʂɔ5 so5b suC θjeu5 – *sjoC
16. span, finger l̥ o5 – dɔ7 l̥ o5a qlouC – – *qroC
17. to wring – pʐo5 – pʐo5a pjuC – pja5 *proC
18. to cover mə6 – mpɔ6 mpu6 mpuC – pho5 *mboC
19. to track – ɳɑŋ6 ɳʈɔ6 – – – – *ndroC
20. armpit ɕə5 ʨo5 tʂɔ7 so7a tsuC – ʨa5 *tsjoC/D
21. muddy – – ŋkɔ7 ŋko7a – – ŋa7 *ŋkoD
22. to grind
– ʐo8 ʐɔ8 ʐo8 wjuD – – *roD
(grain)

Loans in set 7a:

gjuwX (MC) 臼 jiù ‘mortar/treadmill’ > H *ɟoB
tsuX (MC) 祖 zǔ ‘ancestor’ > H *tsjoB ‘tiger’
*lr̥ u (OC) > trhjuw (MC) 抽 chōu ‘to pull out’ > H *throC
*[r]ˤo-s ‘to leak’(OC) > luwH (MC) 漏 lòu ‘to drip’ > H *ŋgroC
*(mə-)tsʰˤuʔ-s (OC) > tshawH (MC) 造 zào ‘to arrive at’ > H *dzoC ‘to reach/arrive at’

7b. HM *o (225)
H *o : HM *o : M *o
1 2 3 4 5 6 7 8 9 10 11 HM/M
23. big ȴ̥hə1 – hl̥ ɔ1 ɭo1b l̥ uA l̥ jeu1’ ȴo1 lo1 lu1’ ɬu1 lɔu1 *hljo
24. husband – po3 – – – – – – – bu3 – *N-poX
25. to soak – – ta2 tɔu2 *doA

Loans in set 7b:

*[ts]ˤuk-s (OC) > tsawH (MC) 竈 zào ‘stove’ > HM *N-tsoH
*(mə-)pˤuk-s (OC) > pawH (MC) 報 bào ‘to announce/report’ > M *ʔboC (< *mpoH) ‘to tell’
 Proto Hmong-Mien 127

7c. HM *u̯ o (206, 225)

H *o : HM *u̯ o : M *u
1 2 3 4 5 6 7 8 9 10 11 HM
26. deep to1 to1 tɔ1 to1a tuA ðu1 to1 do1 du1 du1 – *N-tu̯ o
27. old qo5 qɔ5 qu1 – – ku5 ko5 ko5’ ku5 ku5 ku5 *qu̯ oH

7d. HM *əw (206)

H *o : HM *əw : M *əu
1 2 3 4 5 6 7 8 9 10 11 HM
28. bamboo l̥ ho3 l̥ ho3 hl̥ ɔ3 – – – – l̥ au3 lɔu3 ɬau3 lau3 *hləwX
29. to chop to3 – ntɔ3 nto3a nʔtuB ntu3 no3 – – dau3 dau3 *ntəwX
30. to come back lo4 lo4 lɔ4 lu4 luB lɔu4 lo4 lau4 lau4 lau4 lau4 *ləwX

Loan in set 7d: ʔæwX (MC) 拗 ǎo ‘to break’ > H *ʔloB, M *ʔnəuB (< *ʔəwX)34

7e. HM *i̯ ou (134)

H *o : HM *i̯ ou : M *i̯ ou
1 2 3 4 5 6 7 8 9 10 11 HM
31. cooking
– ȵʨho5 ɲchɔ5 ȵʨo5b ȵʔʨhuB – – ɕeu5 sɔu5’ – – *ɲchi̯ ouH
smoke

Loans in set 7e:

*Nə.[ts]ˤuʔ (OC) > tsawX (MC) 早 zǎo ‘early’ > HM *ntsi̯ ouX
syoX (MC) 暑 shǔ ‘warm’ > HM *sji̯ ouX

7f. HM *əp (106)

H *o : HM *əp : M *əp
1 2 3 4 5 6 7 8 9 10 11 HM/M
32. to bite tə8 to8 tɔ8 to8 ðuD – ta8 tap8 tap8 than8 – *dəp
33. centipede ɕap7 tθap7’ sən7 tsap7 *səpD
34. lightning flash gap7 – – dzjɛp8 *ljəpD

Loans in set 7f:

*kˤ<r>ep (OC) > kɛp (MC) 夾 jiā ‘to pinch’ > M *ʔɟəpD ‘to pick up’
*(kə.)rəp (OC) > lip (MC) 笠 lì ‘bamboo hat’ > M *rəpD

34
This word appears with different “cover” consonants in Hmongic (ʔl- ) and Mienic (ʔn- ). It is not clear if
this was borrowed once as *ʔəwX, or separately into the two branches, from Chinese sources that had
already added a cover consonant, as H *ʔloB and M *ʔnəuB; they have been entered separately in the
onsets section and in the index.
128 Chapter 2

7g. HM *ət (222, 108)

H *o : HM *ət : M *ət
1 2 3 4 5 6 7 8 9 10 11 HM/M
35. to weave to7 nto7 ntɔ7 – – ntu7 na7 dat7 dat7 dan7 dat7 *ntət
36. to laugh ȶə7 ʈo7 ʈɔ7 ʂo7a tʂuD tsu7 ʨa7 kjat7 kjet7 klan7 tut7 *krət
37. near θat7 tθat7 – – *sətD
38. thick – kat8 – – *gətD

Loan in set 7g:

kjwot (MC) 蕨 jué ‘fern’ > M *kʷjətD

7h. HM *u̯ ət (108)

H *o : HM *u̯ ət : M *u̯ ət
1 2 3 4 5 6 7 8 9 10 11 HM
39. to see poŋ8 – pɔ8 po8 mpuD pu8 pa8 pwat8 fat8 phi8 – *bu̯ ət

Loan in set 7h:

phjut (MC) 拂 fú ‘to brush’ > M *phu̯ ətD ‘to sweep’

7i. HM *ək (167)

H *o : HM *ək : M *ək
1 2 3 4 5 6 7 8 9 10 11 M
40. crossbow n̥ ak5 na7’ – – *hnəkD
41. to swallow na7 na7 na7 – *ʔnəkD

Loan in set 7i:

*kə.rək (OC) > lik (MC) 力 lì ‘strength’ > HM *-rək, M *khləkD

Discussion
Pa-Hng (12) shows a rime difference that reflects whether or not the Hmong-Mien rimes of
this set (as reflected in the Mienic forms) were open or closed. Open Hmong-Mien rimes
yield Pa-Hng /o/ (see rimes 7b-d above), whereas closed Hmong-Mien rimes yield Pa-Hng
/a/ (see rimes 7f-h above). The reflection of the open/closed Hmong-Mien contrast as a
vowel quality difference in Pa Hng can also be seen in rime 4. Similarly, Jiwei (Qo Xiong)
(2) shows vowel quality differences reflecting elements of the rime not preserved
elsewhere in Hmongic in rimes 4, 12/17, and 13. This suggests that Pa-Hng and Jiwei may
have split off the main Hmongic branch at an early stage (see Strecker 1987a,b on the
conservatism of “Na-e”, or Pa-Hng).
The Mienic reflexes of rimes 7b and 7c are almost identical, and those of 7c are clearly to
be reconstructed as Mienic *u, since they merge with 3f and 16c (including the common
development of Mienic *u to Liangzi Mun /-o/). The key difference between the two sets is
in Zao Min (23):

7b HM *o > M *o > Zao Min /-ɔu/

7c HM *u̯ o > M *u > Zao Min /-u/
 Proto Hmong-Mien 129

There is no evidence from the Hmongic side to place the words in rime 7g (*-ək) here.
However, the fact that *-əp and *-ət belong to this rime allow us to provisionally include
them here as well.
The two words in Wang and Mao rime set 105 belong here as well, but they are not listed
above, since they are both Chinese loanwords. They have another rime—*ap—that merged
with the others above into Hmongic *o (rather than into Hmongic *a, as we might expect).
They are Hmong-Mien *ʔap ‘duck’ from Chinese 鴨 (MC ʔæp > Man. yā), and Hmong-
Mien *dap ‘to put on/wear shoes’, possibly from Chinese 踏 ‘step on’ (MC thop > Man
tà).

Notes
2. Perhaps the same as Hmong-Mien *mpuə ‘thunder’ (1.4/16); see discussion under onset
1.2.
3. Perhaps borrowed from Chinese 婦 ‘wife, married woman’ (MC bjuwX > Man. fù),
although the tones do not correspond.
4. The Jiwei (2) form is from Office of Miao-Yao Research Lexicon 1987.
7. Compare the similar Mienic *phlu̯ əmB ‘to play (flute)’ (1.32/29).
16. This is the span of thumb to middle finger in White Hmong.
19. White Hmong nrog /ɳʈɔ6/ means ‘to accompany; with’ (‘to track’ > ‘to follow’ > ‘to
accompany’ > ‘with’).
21. Hainan Mun /glɔɔ6/ ‘muddy’ looks similar (Shintani and Yang 1990), and also fits with
the forms under Mienic *ŋglokD ‘turbid’ (5.36/13).
23. In White Hmong, hlob /hl̥ ɔ1/ is ‘to grow; elder; great in volume, proud’. The simple
word for ‘big’ is loj /lɔ2/, which appears to have the same root. The tonal and onset
difference may be attributed to a prefix that devoiced the initial of ‘to grow’: *hlj- yielded
upper-register tone 1, while *lj- yielded lower-register tone 2 (see chapter 4).
24. Compare Chinese 父 ‘father’ (OC *[b](r)aʔ > MC bjuX > Man. fù). See also Hmong-
Mien ‘father/male’ *pjaX (1.16/4).
27. Wang and Mao (1995:327–328) list ‘old’ in the sense of ‘aged’, and ‘old’ as the
opposite of ‘new’ separately because of the presence of two forms in Ho Ne (She), with
those meanings (kɤ5 and kɔ5, respectively). Since the forms are otherwise identical, they
are presented together here. This is probably from Chinese 故 ‘old (not new)’
(OC *kˤa(ʔ)-s > MC kuH > Man. gù), especially since in White Hmong it appears before
the noun it modifies, as in Chinese, rather than in the native position after the noun.
30. In White Hmong, ‘to come back to one’s home’; compare HM *dɑj (2.3/15) ‘to come
(back to a place other than one’s home)’.
32. Compare Malayo-Polynesian *ketep ‘to bite’.
33. There are similar forms in both Tai-Kadai (Lao /khep7/) and Mon-Khmer (Khmer
/kʔæp/).
34. Hmongic *ljeD ‘lightning flash’ (2.42.1/10) would have ended in either -p or -t
(Hmong-Mien tone 8 words ending in -k merged with tone 6 in Hmongic), so these two
words are clearly related. Compare to Tibeto-Burman *(s-)lyap ‘flash; lightning’ (Matisoff
2003).
38. Also Jiangdi Mien /ʥat8/ and Xiangnan Mien /jə8/. Compare Proto Malayo-Polynesian
*buket ‘thick, viscous’ from Proto Austronesian root */-keC/; Proto Malayo-Polynesian
*/-ket/ ‘adhesive, sticky’ (ACD).
130 Chapter 2

40. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). This is
undoubtedly the same word as Hmongic *hnænB ‘crossbow’ (2.8/19). Compare Mon-
Khmer *snaʔ (Shorto #97), Chinese 弩 (OC *[n]ˤaʔ > MC nuX > Man. nǔ), and Proto Tai
*hnaa B1.
 Proto Hmong-Mien 131

8
8a. H *ʉ
1 2 3 4 5 6 7 H
1. to push earth
phɛ1 – phua1 pou1b – phau1 phɪ1 *phʉA
aside
2. marrow/brains l̥ hɛ1 l̥ hə1 hl̥ ɯ1 lou1b l̥ uA – – *hlʉA
3. to rot la2 – lɯ2 lou2 lenA – – *lʉA
4. 2SG (you) – – kɔ2 – – – – *gʉA
5. to pack/wrap qhɛ3 – qhɯ3 hou3b qhuB khau3 – *qhʉB
6. to flow – – ntɯ4 – qluB ljɔu4 – *ɴɢlʉB
7. cuckoo – qu4 – hu4 – – – *ɢʉB
8. elephant shɛ5 – ntshɯ5 – nʔtshuC – – *ntshʉC
9. lung zɛ6 mʐə5 ɳtʂɯ7 mpʐou5a mʔpjuC – mɔ5 *mprʉC
10. to stroke l̥ hɛ5 phʐə5 phlɯ7 plou5b phluC – – *phlʉC
11. which tei6 – tɯ6 tu6 ðuB – ti1 *dʉC
12. shoulder – pə6 pɯ6 – vuC – – *bʉC

Loans in set 8a:

luH (MC) 露 lù ‘dew’ > H *lʉC
bjuX (MC) 腐 fǔ ‘curd’ > H *hʉB ‘bean curd’

8b. HM *u̯ ei (39, 170, 242)

H *ʉ : HM *u̯ ei : M *u̯ ei
1 2 3 4 5 6 7 8 9 10 11 HM/M
13. 2SG (you) moŋ2 mɯ2 – – – mɐŋ2 mɯ6 mwei 2 mei2 məi2 mui2 *mu̯ ei
14. tail tɛ3 tə3 tɯ5 tou3a tuB ðau3 tɛ3 twei3 tei3 dwai3 dui3 *tu̯ eiX
15. younger brother – kɯ3 kɯ3 kou3a kuB – – – – ki3 kui3 *ku̯ eiX
16. malodorous – ʨə5 tʂɯ5 sou5a tsuC tʃau5 ʨɔ5 θwei5’ tθei5 tswəi5 ti5 *tsju̯ eiH
17. to sleep/ lie
pi5 pə5 pɯ5 pu5a puC pau5 pɛ5 pwei5’ fei5 – bui5 *pu̯ eiH
down
4 4 4 4
18. bee mwei ŋwei mi mui *mu̯ eiB
19. to shed
blwei6 blei6 – – *mblu̯ eiC
leaves/drop

Loans in set 8b:

*[k]ʷˤraj (OC) > kwæ (MC) 蝸 wō ‘snail’ > H *ɢʉA
*[k]ʷˤraj (OC) > kwæ (MC) 蝸 wō ‘snail’ > M *klu̯ eiA/B
*[dz]ˤo[j]ʔ (OC) > dzwaX (MC) 坐 zuò ‘to sit’ > M *dzu̯ eiB
lwojH (MC) 累 lèi ‘tired’ > M *lu̯ eiC ‘lazy’

8c. HM *ui (226)

H *ʉ : HM *ui : M *ui
1 2 3 4 5 6 7 8 9 10 11 HM
20. nose zɛ6 mʐə6 ɳtʂɯ6 mpʐu6 mpjuC mpjau6 mjɔ6 pa6 – bli6 – *mbruiH
132 Chapter 2

Discussion
It is clear that the medial -w- in the Mienic forms of rime 8b point to a rounded element in
the rime, not secondary rounding of the initial consonant, since it occurs consistently in
this rime.

Notes
2. According to Sagart (1999:67), this is the same as Chinese 髓 ‘marrow’ (OC *s-lojʔ >
MC sjweX > Man. suǐ).
4. Also attested in Western Hmongic Xianjin /kau2/, Shimen /ʨy6/ and Bunu /kau2/.
5. Possibly from Chinese 裹 ‘to wrap’ (OC *[k]ˤo[j]ʔ > MC kwaX > Man. guǒ).
8. Compare Lolo-Burmese *ts(h)aŋ ‘elephant’ > Written Burmese chaŋ.
10. Compare Mienic *phlunA (1.32/27) and Chinese 撫 ‘to caress’ (OC *[pʰ](r)aʔ > MC
phjuX > Man. fǔ).
13. Similar to the Austronesian 2PL root *-mu (ACD); the 2PL root was extended to the
2SG in Austronesian. Western Hmongic has a different root in set 8a above: *gʉA ‘you’
(5.3/8).
15. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Quanzhou Biao Min
(10).
20. Compare Chinese 鼻 ‘nose’ (OC *m-[b]i[t]-s > MC bjijH > Man. bí). For the
reconstruction of this rime, see Jiangdi Mien /bjuːi6/, Xiangnan Mien /bu8/, and Changping
Mien /blui6/ (Wang and Mao 1995).
 Proto Hmong-Mien 133

9
9a. H *uw
1 2 3 4 5 6 7 H
1. gourd – tɔ1 tau1 tu1a toA – – *tuwA
2. to butt (of bulls) ȶu5 – ɳʈau5 ɳʈu5a ɳʔtʂoC – – *ntruwC
3. husked rice – ntsɔ5 – – – ɳtʃei5 ȵɯ5 *ntsuwC
4. male – – lau5 ɭu3a ʔloB – – *ʔljuwB
5. centipede khu7 – – – – khjɔ5 – *khluwD
6. bamboo hat – ku7 kau7 ku7a koD – kɯ7 *kuwD
7. to beat
– – ntau7 ntu7a – – ta7 *ntuwD
(someone)
8. crooked – ŋkhu7 ŋkhau7 ŋko7b ŋʔkhaD – ŋɔ1 *ŋkhuwD
9. dark tsə7 pʐu7 tʂau7 pʐu7a pjoD pjɔu7 – *pruwD
10. houttuynia
ȶu7 ʈu7 ʈau7 ʈu7a tʂoD – – *truwD
cordata
11. back (of body) kə8 – qau8 hu8 ʁoD kɔ8 kɯ8 *ɢuwD

Loans in set 9a:

syuw (MC) 收 shōu ‘to receive/gather in’ > H *sjuwA
ʔjiwH (MC) 幼 yòu ‘young’ > H *ʔjuwC ‘small/young’
*l̥ ˤik (OC) > thet (MC) 鐵 tiě ‘iron’ > H *hluwC

9b. HM *uj (243, 260)

H *uw : HM *uj : M *uj
1 2 3 4 5 6 7 8 9 10 11 HM/M
12. sour ɕhu1 ɕɔ1 – su1b soA θjɔ1 sɯ1 ɕui1 tθui1’ swəi1 si1 *suj
13. mouth/beak ȵu2 ȵɔ2 ɲcau2 ȵʨu2 ȵʨoA ntjɔ2 ȵʑɯ2 dzuːi2 – ʨi2 dzi2 *ɲɟuj
14. loud/sound bui1 buːi1’ bəi1 bai1 *ʔbujA
15. bat (animal) bui1 buːi1’ bəi1 bai1 *ʔbujA
16. clothes gui1 gui1 lwəi1 – *ʔrujA
17. thigh ʨui2 ȶuːi2 – si2 *djujA
18. steep ʨui4 ȶui4 – ki4 *gjujB
19. thread ɕui5 tθuːi5’ sui5 si5 *sujC

Loan in set 9b:

dwoj (MC) ‘violent wind; collapse’ 頹 tuí > M *dujA ‘to fall (of rain)’

9c. HM *up (271)

H *uw : HM *up : M *op
1 2 3 4 5 6 7 8 9 10 11 HM/M
20. grasshopper – – – – – – ka-jɔ7 tɕop7 – klin7 – *klup
21. to bury plop7 plɔp7 – – *plopD
134 Chapter 2

Loans in set 9c:

*qʰˤ[ə]p (OC) > xop (MC) 欱 hē ‘drink/smoke’ > HM *hup
*[t]ˤ[ə]p (OC) > top (MC) 荅 dá ‘small bean’ > HM *dup ‘bean’35

9d. HM *ut (239)

H *uw : HM *ut : M *ut
1 2 3 4 5 6 7 8 9 10 11 HM
22. glutinous 8 8 8 8 D 8 8 8 8
nə nu mplau mplu mplo – – but blut blun bit *mblut
/sticky

Loan in set 9d:

*l̥ ˤot (OC) > thwat (MC) 脫 tuō ‘to peel off/escape’ > M *ʔdutD

9e. HM *uk (204, 203)

H *uw : HM *uk : M *ok
1 2 3 4 5 6 7 8 9 10 11 HM
23. six ȶu5 ʈɔ5 ʈau5 ʂu5a tʂoC tʃɔ5 tɕɯ5 kwo7 kjɔ7 klɔ7 tɔu7 *kruk
24. to shrink xhu7 – – – – – – ɕo7 – – – *hjuk

9f. HM *u̯ ɛp (163)

H *uw : HM *u̯ ɛp : M *ɛp
1 2 3 4 5 6 7 8 9 10 11 M
25. to drip djep7 – djan7 dɛp7 *ʔdjɛpD

Loans in set 9f:

*[g][i]p (OC) > dzyip (MC) 十 shí ‘ten’ > HM *gju̯ ɛp36
dzjep (MC) 捷 jié ‘quick’ > HM *cu̯ ɛp

9g. HM *u̯ ɛt (70)

H *uw : HM *u̯ ɛt : M *ɛt
1 2 3 4 5 6 7 8 9 10 11 HM
26. itch(y) 5 5 7 7b D 7 7 7 7’ 7 7
ʨhu ɕi khau ko kho ʃɔ ŋɯ ɕet set hin kɛt *khju̯ ɛt
/scratch(y)

9h. HM *u̯ ɛk (224)

H *uw : HM *u̯ ɛk : M *ɛk
1 2 3 4 5 6 7 8 9 10 11 HM
27. to enter pə6 pɔ5 – – – pɔ6 pɯ6 pja8 – phja8 pjɛ8 *bju̯ ɛk

Loan in set 9h: *l̥ ˤik (OC) > thet (MC) 鐵 tiě ‘iron’ > M *hrɛkD

35
See Sagart 1999:187.
36
This could also have been borrowed from Tibeto-Burman, as were the numerals from ‘four’ through
‘nine’ (see chapter 5, section 2). Benedict (1987a:14) reconstructs *g(j)iap for Hmong-Mien.
 Proto Hmong-Mien 135

Discussion
Hmong-Mien *ɴɢu̯ at ‘smooth/slippery’ from Chinese 滑 ‘slippery’ (OC *Nə-gˤrut > MC
hwɛt > Man. huá) also belongs to this group.

Notes
2. This is probably from Chinese 鬥 ‘fight, wrangle’ (MC tuwH > Man. dòu).
3. ‘Husked rice’ corresponds in only two of its four forms (Jiongnai (6) and Ho Ne /tsi5/ do
not correspond), and may thus prove not to belong in this set.
4. In White Hmong, lau /lau5/ is male of birds, while laug /lau6/is male of sheep, goat, cow.
5. Compare the Central Tai and Southwest Tai forms (Downer 1978): Lao /(khii3-) khep7/,
Thai /takhaap7/, which also appears in Kam as /khep7/. See also Proto Mon-Khmer *kʔaip
> Khmer /kʔæp/ (Shorto #1226).
6–11. These words must have ended in either *-p or *-t, since Hmong-Mien *-k would
have yielded Hmongic forms with tone 5.
8. Compare Chinese 曲 ‘bend, crooked’ (MC khjowk > Man. qū), Tibeto-Burman *kuk
(Matisoff 2003), and Malayo-Polynesian *buŋkuk (ACD).
10. Also known as 魚腥草 (Man. yú xīng cǎo), this is a plant used primarily for medicinal
purposes (see chapter 7).
11. In White Hmong, this refers to the upper back. Tan Trinh Pa-Hng (Niederer 1997) has
been substituted for Baiyun Pa-Hng (7).
12. This may come from Chinese 酸 ‘sour’ (OC *[s]ˤo[r] > MC swan > Man. suān).
13. Perhaps from Chinese 嘴 ‘beak’ (OC *[ts]ojʔ > MC tsjweX > Man. zuǐ).
20. Jiangdi Mien (Wang and Mao 15) has been substitued for Luoxiang Mien (8).
21. Hainan Mun (Shintani and Yang 1990) has been substituted for Lanjin Mun (9).
22. In Hmongic, this is used for ‘sticky’ rice, in Mienic it means both ‘glutinous’ and
‘snot’. Compare Proto Malayo-Polynesian *belit/*bulit ‘viscous, sticky’ (ACD) and
Chinese 秫 ‘glutinous millet’ (OC *m.lut > MC zywit > Man. shú) (Sagart 1999:79).
23. A borrowing from Tibeto-Burman *k-ruk (see chapter 5 and Benedict 1987a). The
Hmong-Mien reconstruction must differ from the Mienic reconstruction, since the Mienic
reflexes exactly match those of ‘few, lack’ (rime 13), indicating a merger of two rimes in
Mienic.
24. See also Xianjin Miao /xau7/, Jiangdi Mien /su7/, and Changping Mien /ðut7/. The final
-t in Changping Mien and the Hmongic tones suggest that this word ended in -t, not -k. But
the word is otherwise similar to Chinese 縮 ‘to shrink’ (OC *[s]ruk > MC srjuwk > Man.
suō), so a -k is tentatively reconstructed here.
25. The correspondence of Mienic tone 8 to Hmongic tone 6 is regular when the Hmong-
Mien word ended in -k.
136 Chapter 2

10
10a. H *e
1 2 3 4 5 6 7 H
1. finger ring – – mphlai1 – mʔphleA – – *mphleA
2. to tie/cross qhei1 – qhai1 he1b qheA khai1 – *qheA
3. twin pi1 – ntsai1 mpje1a mʔpleA – – *mpljeA
4. to slant/slanting – qa2 qai2 – ʁeA – – *ɢeA
5. bowl – – tai4 te4 – – ta4 *deB
6. half – – – pje7b phjeC – – *phjeC
7. quick xhi5 ʂɑŋ5 ʂai5 – seC v̥ wei5 – *hreC
8. to scold/curse – – – he5a qeC – – *qeC
9. to conceal ɣi7 ʐɑ7 ʐai7 ʐe7a ʔwjeD vwei7 ɛ7 *ʔreD
10. daughter/girl phi7 mphɑ7 ntshai7 mpje7b mʔphjeD phai7 phe7 *mphjeD
11. to ladle out hei7 – hai7 ɦe7b heD – fɛ7 *heD
12. to peel/shave – – chai7 ʨe7b ʨheD – – *cheD
13. to pick
– tɑ7 tai7 te7a teD – næ̃ 6 *teD
up/clamp
14. to suck – – ntsai7 – – – – *ntseD
15. chin/lower jaw ɕi1 ʨɑ8 tʂai8 se8 zeD – – *dzjeD
16. lightning flash ȴi8 ȴɑ8 lai8 ɭe8 leD – – *ljeD
17. to seek shelter – – ɳʈai8 ɳʈe8 ɳtʂeD – – *ŋgreD

Loan in set 10a:

*N-kˤ<r>ep (OC) > hɛp (MC) 狹 xiá ‘narrow’ > H *ɴɢeD

10b. HM *e (73)
H *e : HM *e : M *e
1 2 3 4 5 6 7 8 9 10 11 HM/M
18. hungry – ɕi1 tʂhai1 se1b tsheA – sa1 ɕa1 sa1’ – – *tshje
19. fear/be afraid ɕhi1 ȵʨha5 ɳtʂhai5 ntse5b nʔtsheC ɳtʃei1 ȵ̥ɛ5 ʥa5 ȡa5’ ȡa5 dzjɛ5 *ntshjeH
20. to fool/humor
– – – – – kle1 – – kja4 kljɔu4 kɛi4 *glje(X)
(children)
21. that wa3 va3 wə3 vɛi5 *ʔweB

10c. HM *ej (37)

H *e : HM *ej : M *ej
1 2 3 4 5 6 7 8 9 10 11 HM/M
22. meat ŋi2 ɲa2 ɴqai2 ŋke2 ɴqeA ŋkai2 ŋɛ2 – – ʥi2 hɔi2 *NGej
23. animal fat/oil m̥ ei1 mei1’ m̥ əi1 mi1 *hmejA
24. sore/boil/blister ʨwei2 sei2 – – *kʷjejA
 Proto Hmong-Mien 137

Loans in set 10c:

dej (MC) 蹄 tí ‘hoof’ > HM *dej
*[g](r)aj (OC) > gje (MC) 騎 qí ‘to ride’ > HM *ɟej
*[k]ʷˤraj (OC) > kwæ (MC) 蝸 wō ‘snail’ > M *kʷlejA
tsyeX (MC) 紙 zhǐ ‘paper’ > M *tjejB
dzyeX (MC) 是 shì ‘to be’ > M *djejB
drijH (MC) 緻 zhì ‘beautiful’ > M *dzejC
dijH (MC) 地 dì ‘non-irrigated land’ > M *dejC

10d. HM *ep (48, 68, 87, 91, 124, 220)

H *e : HM *ep : M *ep
1 2 3 4 5 6 7 8 9 10 11 HM
25. bear ȴ̥i7 – dai7 l̥ e7a – – – kjep7 kjaːp7 kljɛ7 – *qrep
26. to slice l̥ hei7 l̥ hɑ7 hl̥ ai7 le7b – l̥ ai7 – le7 – – – *hlep
27. to lick ʑi8 ʑɑ8 jai8 ʑe8 – ŋklai8 jɛ8 – – – dzjɛp8 *-jep

Loans in set 10d:

*[ts]ap (OC) > tsjep (MC) 接 jiē ‘to receive/accept’ > HM *tsep ‘to receive/borrow’
*mə-[tsʰ]<r>op (OC) > tsrhɛp (MC) 插 chā ‘to insert’ > HM *threp
*mə-tsˤr[a]p (OC) > tsrɛp (MC) 眨 zhǎ ‘to wink’ > HM *ntsjep
hɛp (MC) 狹 xiá ‘narrow’ > M *hepD

10e. HM *et (51)

H *e : HM *et : M *et
1 2 3 4 5 6 7 8 9 10 11 HM
28. tongue ȵi8 mjɑ8 mplai6 mple8 mpleD mpli8 mje8 bjet8 bjet8 blin8 bɛt8 *mblet

Loans in set 10e:

*[tsʰ]i[t] ‘varnish’ (OC) > tshit (MC) 漆 qī ‘lacquer’ > HM *thjet
*[k]ʷʰˤet ‘break; defective’ (OC) > khwet (MC) 缺 quē ‘to have a gap’ > HM *NKʷet
*qʷˤet (OC) > ʔwet (MC) 挖 juē ‘to dig’ > M *ʔwetD

Discussion
An important loanword which falls into this rime set, but into none of the subsets above, is
Hmong-Mien *Kəi ‘chicken’ (onset 5/6.1) from Chinese 雞 ‘chicken’ (OC *kˤe > MC kej
> Man. jī). This is an areal word: see also Proto Tai *kəi (Li 1977).

Notes
4. In White Hmong (3), also ‘to lean’.
9. Perhaps Chinese 諱 ‘taboo; to conceal’ (MC xjwɨjH > Man. huì). See correspondence to
HM *ʔræi ‘vegetable’ (2.55/12).
14. This word is found in other West Hmongic languages as well.
16. Clearly related to Mienic *ljəpD (2.42.1/7), since the onsets are identical and the
Hmongic word would have ended in either -p or -t (Hmong-Mien tone 8 words ending in
-k merged with tone 6 in Hmongic). Compare to Tibeto-Burman *(s-)lyap ‘flash;
lightning’ (Matisoff 2003).
138 Chapter 2

17. In White Hmong (3), ‘to hide’.

20. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10). The meaning of this word is given as hǒng as in hǒng xiǎo há ‘to handle children’ in
Wang and Mao 1995.
23. Compare Austronesian *SimaR ‘grease/oil/fat’ (ACD).
24. Since this word has a labiovelar initial, it is probably a loan from an unknown Chinese
source.
25. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10).
26. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
Compare Tibeto-Burman *(s-)lep ‘to slice’(Matisoff 2003). Given the tonal reflexes in
Hmongic, only -p and -t are possible coda consonants in this word. The striking similarity
to Tibeto-Burman *(s-)lep makes -p seem more likely.
28. Probably from Chinese 舌 ‘tongue’ (OC *m.lat > MC zyet > Man. shé); see parallel
with Hmong-Mien *mblut ‘glutinous/sticky’ (1.36/9) from Chinese 秫 ‘glutinous millet’
(OC *m.lut > MC zywit > Man. shú) (both discussed in Sagart 1999:79).
 Proto Hmong-Mien 139

11
11a. H *u̯ ɛ
1 2 3 4 5 6 7 H
1. to grasp mɛ1 me1 mua1 muɑ1a – – – *ʔmu̯ ɛA

11b. HM *ɛj (55, 93)

H *u̯ ɛ : HM *ɛj : M *ai
1 2 3 4 5 6 7 8 9 10 11 HM/M
2. to have mɛ2 me2 mua2 – maA mɔ2 mɪ2̃ maːi2 naːi2 ma2 – *n-mɛj
3. to squeeze ȵai5' ȵai5 – – *hɲaiC

Loans in set 11b:

*m.rˤajʔ (OC) > mɛɨX (MC) 買 mǎi ‘to buy’ > HM *mɛjX
*m.rˤajʔ-s (OC) > mɛɨH (MC) 賣 mài ‘to sell’ > HM *mɛjH

11c. HM *u̯ ɛj (75)

H *u̯ ɛ : HM *u̯ ɛj : M *u̯ ɛi
1 2 3 4 5 6 7 8 9 10 11 HM
4. soft mɛ6 ne6 mua6 noŋ6 mjaC mɔ6 mɪ6̃ mwei6 ŋwei6 mi6 mai6 *mlu̯ ɛjH
5. eye mɛ6 me6 mua6 moŋ6 maC mɔ6 mɪ6̃ mwei6 ŋwei6 mi6 mai2 *mu̯ ɛjH

11d. M *u̯ ai (61)

1 2 3 4 5 6 7 8 9 10 11 M
6. to ask naːi6 naːi6 na6 nɔi6 *nu̯ aiC

11e. M *u̯ ɔi (188)

1 2 3 4 5 6 7 8 9 10 11 M
7. sun/day n̥ ɔi1 nɔːi1’ n̥ wai1 nai1 *hnu̯ ɔiA

Loans in set 11e:

khoj (MC) 開 kāi ‘to open’ > M *khu̯ ɔiA
twojH (MC) 碓 duì ‘tilt-hammer/pestle’ > M *tu̯ ɔiC
ʔojH (MC) 愛 ài ‘to love’ > M *ʔu̯ ɔiC

Discussion
Hmong-Mien rime 11b shares with rimes 12/17b-c and 15b, d the characteristic that the
final high front vocalic element triggered the introduction of a back rounded on-glide in
the development of Hmongic. For discussion, see chapter 2, section 1. Thus all words in
rime 11 have an intermediate Proto Hmongic form in -u̯ ɛ, whether or not a back rounded
on-glide is reconstructed at the Hmong-Mien level on the basis of its appearance in Mienic.
140 Chapter 2

Rime sets 11d and 11e are included here because their Hmongic counterparts appear under
the similar nasal rime 22 below: ‘to ask’ (*nɛŋC) and ‘sun/day’ (*hnɛŋA). The members of
each pair are clearly related. A similar problem exists with four out of the five Hmongic
words in sets 11a-c: these words have nasal initials, and show a nasal coda in Zongdi (4),
as well as Gaopo, Fengxiang, and Bunu (not given here). It may prove to be best to
separate Hmongic and Mienic in these cases as well, but since the Hmongic forms do not
fit into any of the rimes 18–30 exactly, they appear here, and the final nasals are assumed
to be secondary.

Notes
1. In Zongdi, ‘grasp’ > ‘take’ has narrowed to mean ‘to take a wife’.
1 & 2. If Hmongic ‘to grasp’ had Mienic cognates, we would have four words dealing with
the possession and transfer of objects involving the same root at the Hmong-Mien level:
*ʔ-mɛj ‘grasp’, *n-mɛj ‘have’, *mɛj-X ‘buy’, *mɛj-H ‘sell’ (see chapter 4). The question
of a possible genetic relationship between Hmong-Mien and Chinese is still open, but the
deep and pervasive appearance of this basic root in both families is intriguing.
4. Medial *-l- reconstructed on the basis of Gaopo /mləŋ6/ (Wang and Mao 6). Compare
Proto Malayo-Polynesian *ma-lumu ‘soft, tender, gentle’.
5. Possibly from Chinese 目 ‘eye’ (OC *[m][u]k > MC mjuwk > Man. mù).
 Proto Hmong-Mien 141

12/1737
12/17a. H *u̯ ei
1 2 3 4 5 6 7 H
1. belly ʨhu1 ʨhi1 chau1 ʨɔ1b ʨhuA the1 – *chu̯ eiA
2. to pour ȴu1 – lau1 ɭo1a ʔluA – – *ʔlju̯ eiA
3. to repay pə2 pi2 pau2 pɔ2 vuA – – *bu̯ eiA
4. to burn (TR) ȶu3 – ʈau3 ʂɔ3a tʂuB – – *kru̯ eiB
5. flour – – – plɔ5a plouB – – *plu̯ eiB
6. sinew ɕhu3 ɕi3 – sɔ3b suB – si3 *sju̯ eiB
7. old lu4 – lau4 lo4 lauB – – *lu̯ eiB
8. river deer – – – ŋko4 ŋkuB – – *ŋgu̯ eiB
9. axe to5 tɔ5 tau7 tɔ5a tuC ðei5 ti5 *tu̯ eiC
10. to boil (TR) hu5 hɔ3 hau5 – – – – *hu̯ eiC
11. to write xho5 ʂei5 ʂau5 ʐɔ5b – – – *hru̯ eiC
12. to fall (over) ku6 qɔ6 qau6 ho6 ʁuC – – *ɢu̯ eiC
13. pus pu6 pɔ6 pau6 po6 vuB pei6 pɪ6 *bu̯ eiC
14. thin (person) su5 ntsei5 ntsau6 ntso6 – ɳtʃe6 nʑi6 *ndzu̯ eiC
15. young man ȵo6 – ɳʈau6 ɳʈo6 ɳtʂuC ɳtʃi6 – *ndru̯ eiC
16. ice ȴ̥u7 – dau7 – – – – *qru̯ eiD

Loans in set 12/17a:

*tˤek (OC) > tek (MC) 蹢 dí ‘hoof’ > H *tru̯ eiC ‘hoof/claw/nail’
*(mə-)t(ˤ)rek (OC) > trɛk (MC) 摘 zhāi ‘to pluck’ > H *ndru̯ eiC ‘to strip leaves’

12/17b. HM *ei (39, 170, 208)

H *u̯ ei : HM *ei : M *ei
1 2 3 4 5 6 7 8 9 10 11 HM/M
17. cloth – ntei1 ntau1 ntɔ1a nʔtuA nte1 ni1 di1 – di1 dɛi1 *ntei
18. four l̥ u1 pʐei1 plau1 plɔ1a plouA ple1 pi1 pje1 pjei1 pləi1 pɛi5 *plei
19. hair ȴ̥u1 pi1 plau1 pɭɔ1a plouA ple1 pi1 pje1 pjei1 pli1 pɛi1 *pljei
20. to know pu1 – pau1 pɔ1a – pe1 pɪ1 pei1 pei1 pəi1 bɛi1 *pei
21. shell/covering l̥ hu7 – phlau1 – phlauC – – – – – fɔu1 *phlei
22. head fhu3 pʐei3 hau3 wɔ5b huB – pɦi3’ pje3 pjei3 pli3 pɛi3 *S-phreiX
23. head louse ɕhu3 ȵʨhi3 ɳtʂhau3 ntsɔ3b nʔtshuB – ʨhi3 ɕei3 tθei3 – dzɛi3 *ntshjeiX
24. son-in-law – – vau3 wɔ3a ʔwuB ve3 vi3 vei3 – – vɛi3 *ʔweiX
25. soybean pu3 – pau3 – – – – – bei3 bi3 mɛi3 *N-peiX
26. to dream pu5 mpei5 mpau5 mpɔ5a mʔpuC mpe5 mi5 bei5 bei5 bəi5 bɛi5 *mpeiH
27. taste/try flavor m̥ hu5 – – – – – – – – m̥ i5 – *hmeiH
28. porcupine ɕen6 ʨi6 tʂau6 so6 zuC ʃa6 ʨi6 dei6 dei6 – hɛi6 *N-dzjeiH
29. child ʨwei3 sei3 hwjəi7 – *kʷjeiB

Loans in set 12/17b:

dzjeH (MC) 漬 zì ‘to soak’ > HM *dzeiH
mejX (MC) 米 mǐ ‘husked rice’ > M *hmeiB

37
This category merges two of the rime categories in Wang 1994, 12 and 17. These two categories differ
only in the Jiwei (Qo Xiong) reflex, for which an explanation is proposed in section 2.1.3.1.
142 Chapter 2

12/17c. HM *æi (132)

H *u̯ ei : HM *æi : M *æi
1 2 3 4 5 6 7 8 9 10 11 HM/M
30. vegetable ɣu1 ʐei1 ʐau1 ʐɔ1a ʔwjuA ji1 ɦji1 gai1 gjai1 lai1 ɛi1 *ʔræi
31. shy ȵai3 ȵai1 – – *ʔɲæiA/B
32. thin kje6 kjai6 klai6 kɛi6 *glæiC

Loans in set 12/17c:

drij (MC) 遲 chí ‘late’ > M *djæiA
sejH (MC) 細 xì ‘fine, small’ > M *sæiC ‘small/fine’

12/17d. HM *u̯ əi (39, 207)

H *u̯ ei : HM *u̯ əi : M *u̯ əi
1 2 3 4 5 6 7 8 9 10 11 HM
33. ashes ɕhu3 ɕi3 tʂhau3 sɔ3b tshuB θe3 ɕi3 ɕwai3 saːi3 swai3 sɔi3 *tshju̯ əiX
34. knee ʨu6 ʨɔ6 cau6 ʨo6 ʑuC – ʨi6 ʨwai6 – ȶwai6 – *ɟu̯ əiH

Discussion
Hmong-Mien rimes 12/17b-c share with rimes 11b and 15b, d the characteristic that the
final high front vocalic element triggered the introduction of a back rounded on-glide in
the development of Hmongic. For discussion, see section 2.1.3. Thus all words in rime
12/17 have an intermediate Proto Hmongic form in *-u̯ ei, whether or not a back rounded
on-glide is reconstructed at the Hmong-Mien level on the basis of its appearance in Mienic.
This set appears as two smaller sets in Wang 1994: 12 and 17. However, the two Hmongic
sets differ only in one language: Jiwei (Qo Xiong) (2), which shows both /ei~i/ and /ɔ/ in
this rime. As with two other rimes (4 and 13), conservative Jiwei may reflect an old
Hmong-Mien rime contrast here. The Jiwei vowel is /ei~i/ if the Hmong-Mien rime does
not have a rounded on-glide (rimes 12/17c-d above), and is /ɔ/ if the Hmong-Mien rime
has a rounded on-glide (rime 12/17e above). The Jiwei form /ʐei1/ for ‘vegetable’ (12/17d)
supports this. The one exception is ‘ashes’, where /ɕɔ3/ is expected, but /ɕi3/occurs instead.
Mienic counterparts have not been found for the words in rime 12/17a, but if we can find
them, ‘axe’, ‘pus’ and ‘fall over’ should show rounded on-glides. Conversely, Hmongic
counterparts have not been found for ‘shed leaves’ in rime 12/17b, but if they can be
found, Jiwei should show the /ɔ/ vowel.
A Chinese loanword that belongs to this set is HM *mpu̯ æiH ‘to boil (INTR)’ from Chinese
沸 ‘to boil (INTR)’ (OC *-p[u][t]-s > MC pjɨjH (MC) > Man. fèi).

Notes
1. In White Hmong, cognate /chau1/ means ‘to crawl on one’s belly’ (compare the verbal
use of ‘belly’ in English: ‘to belly up to the table’).
5. Compare Proto Malayo-Polynesian *palu ‘sago flour’ (ACD).
7. The individual forms look very much like Chinese 老 ‘old’ (MC lawX > Man. lǎo), but
at the Hmongic level the rime does not correspond.
11. Compare Chinese 鏤 ‘to incise’ (OC *[r]ˤo-s > MC luwH > Man. lòu).
18. From Tibeto-Burman *b-ləy ‘four’ (Matisoff 2003). See chapter 5.
 Proto Hmong-Mien 143

20. Possibly from Chinese 別 ‘to separate, distinguish’ (OC *N-pret > MC bjet > Man.
bié).
21. The majority of other West Hmongic cognates have tone 1. Yanghao (1) from Office of
Miao-Yao Research Lexicon 1987. The word in both Yanghao (1) and Fuyuan (5) means
‘soft membrane inside shell’. Compare Proto Tai *pl–k ‘shell, bark’.
22. Extended semantically in White Hmong to ‘base, root, origin’ as well as ‘summit’
(tone 3) and ‘leader, headman’ (tone 5).
23. Tan Trinh Pa-Hng (Niederer 1997) has been substituted for Baiyun Pa-Hng (7).
24. Changping Mien (Wang and Mao 18) has been substituted for Luoxiang Mien (8).
Compare Tibeto-Burman *krwəy ‘son-in-law’ (Matisoff 2003). This connection is
strengthened by the parallel ‘sister-in-law/daughter-in-law’ *ʔɲam (4.7/24) which closely
resembles Tibeto-Burman *nam ‘daughter-in-law’ (Benedict 1987a).
26. Jinxiu Mien (8) from L-Thongkum 1993:205.
27. Possibly related to Chinese 味 ‘taste (n.)’ (OC *m[ə][t]-s > MC mjɨjH > Man. wèi).
28. Tan Trinh Pa-Hng (Niederer 1997) has been substituted for Baiyun Pa-Hng (7).
29. Compare Pucheng (Min) /kiãi3/ ‘child’ (Norman 1988:243).
30. Perhaps Chinese 卉 ‘herbs, vegetation’ (MC xjwɨjH > Man. huì). See correspondence
to ‘to conceal’ (2.55/10).
31. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
32. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10). A word meaning ‘thin (person)’ with matching rime and tone exists in Hmongic, but
its initial is *ndz- (3.6).
33. Chinese 灰 ‘ashes’ (MC xwoj > Man. huī) has a very different onset and a different
tone, but the rime corresponds.
144 Chapter 2

13
13a. H *ow
1 2 3 4 5 6 7 H
1. plum – – khaɯ1 kə1b khoA – – *khowA
2. soot ʨhu1 – ŋkhaɯ1 ŋkə1b ŋʔkhoA – – *ŋkhowA
3. white l̥ u1 qwə1 daɯ1 l̥ ə1a qloA klau1 kwɔ1 *qlowA
4. able su2 – tsaɯ2 sə2 – – – *dzowA
5. heart ȴ̥u3 – plaɯ3 pɭə3b ploB – – *pljowB
6. paper tu3 ntə3 ntaɯ3 ntə3a nʔtoB ntau3 nɔ3 *ntowB
7. face – – – pɯ4 – – – *bowB
8. foot tu5 tə5 taɯ5 tə5a toC – – *towC
9. salty ȴ̥u5 – daɯ5 l̥ ə5a – – – *qrowC
10. to swallow – – – ŋkə5a – ŋkau5 ŋɛ5 *ɴqowC
11. walnut – – – l̥ ə5a qloC – – *qlowC
12. to gnaw ki7 ku7 kaɯ7 kə7a koD – – *kowD
13. to ignite/light tu7 – taɯ7 te7a toD – – *towD
14. to open (door) pu7 pu7 – – poD – pɔ7 *powD
15. to peck/dig ʨu7 ȵʨu7 ɲcaɯ7 ȵʨə7a ȵʔʨoD – ʨɔ1 *ɲcowD
16. pheasant – – tʂaɯ7 sə7a – – – *tsjowD
17. to pick up ʨhə1 – khaɯ7 kɯ7b khoD – – *khowD
18. to shut (door) shu7 – – ʂə7 – – – *sowD
19. bamboo – – ʈaɯ8 ʈə8 – – – *drowD
20. mountain pass – – daɯ8 l̥ ə8 – – – *ɢlowD
21. pair ȵu8 ŋoŋ8 ŋkaɯ8 ŋkə8 ŋkoD ɳtʃau8 ŋɤ8 *ŋgjowD
22. to stumble – – – – – – – *ndzjowD
23. whip – – mplaɯ8 mplə8 mpʔloD – – *mblowD

Loans in set 13a:

tsjuwX (MC) 酒 jiǔ ‘liquor’ > H *cowB
*[r]awʔ (OC) > ljewX (MC) 燎 liǎo ‘to singe’ > H *hljowB ‘to burn (mountain)’

13b. HM *ou (263)

H *ow : HM *ou : M *ou
1 2 3 4 5 6 7 8 9 10 11 HM/M
24. to stand/get up ɕhu3 ɕə3 ʂaɯ3 sə3b soB θjau3 sɤ3 ɕou3 – səu3 fu3 *sjouX
25. fire tu4 tə4 taɯ4 tɯ4 ðoB tau4 tɤ4 tou4 tɔu4 təu4 tu4 *douX
26. explode tu6 tə6 taɯ6 tɯ6 ðoC – tɤ6 – tɔ6 – – *douH
27. bellows lou2 lɔu2 ləu2 lu2 *louA
28. hold w/2 hands pou2 pɔu2 – – *bouA
29. to forget – ȵɔu3 n̥ au3 nau3 *hɲouB
30. grave/tomb θou3 tθɔu3 tsəu3 – *tsouB
 Proto Hmong-Mien 145

Loans in set 13b:

tsyhoH (MC) 處 chù ‘bed/place’ > HM *chouH
drjoH (MC) 箸 zhù ‘chopsticks’ > HM *drouH
hjuH (MC) 芋 yù ‘taro’ > HM *wouH
pjuX (MC) 斧 fǔ ‘axe’ > M *pouB
tsyoX (MC) 煮 zhǔ ‘to boil’ > M *tjouB
kjuH (MC) 句 jù ‘CLF-sentences’ > M *kjouC

13c. HM *op (237)

H *ow : HM *op : M *op
1 2 3 4 5 6 7 8 9 10 11 HM
31. skin tu3 tə3 taɯ5 tə3a toB – – dup7 dup7 din7 dip7 *N-top

13d. HM *ot (256)

H *ow : HM *ot : M *ot
1 2 3 4 5 6 7 8 9 10 11 HM
32. to stop up/plug – ntshu7 ɳtʂaɯ7 ntsəu5a nʔtsuD ɳtʃau7 sɔ7 tsot7 – – – *ntsjot

Loans in set 13d:

*[s][r][o]t (OC) > srwæt (MC) 刷 shuā ‘to brush’ > M *sotD ‘to wipe’
khwot (MC) 堀 kū ‘hole’ > M *khotD

13e. HM *ok (186, 203)

H *ow : HM *ok : M *ok
1 2 3 4 5 6 7 8 9 10 11 HM/M
33. few/lack ɕu6 – tʂaɯ6 – zoC ʃau6 ʨɤ6 θo8 tθɔ6 tshɔ8 hu8 *dzjok
34. finger do7 dɔ7 dɔ7 – *ʔdokD
35. turbid glo8 glɔ6 – – *ŋglokD

13f. HM *əut (255)

H *ow : HM *əut : M *ut
1 2 3 4 5 6 7 8 9 10 11 HM
36. navel tu7 ntu7 ntaɯ7 ntə7a nʔtoD ntu7 – dut7 duːt7’ din7 – *ntəut

Discussion
The variant reflexes in Jiwei (Qo Xiong) (2) and Pa-Hng (12) are in complementary
distribution. In Jiwei the rime is /u/ in tone 7 and /ə/ elsewhere, which follows the
generalization made in the introduction to the reconstruction (2.1.3) that traces of coda
consonants can be detected in vowel distinctions in this language. In Pa-Hng the
distribution of rimes is conditioned by the register of the initial consonant: it is /ɤ/ after
low register tones 4, 6, 8 and /ɔ/ after high register tones 1, 3, 7.
146 Chapter 2

Notes
7. Also Shimen /bey4I/, Qingyan /pau4/, Gaopo /pə4/, Fengxiang /pɔ4/ (Wang and Mao
1995). Xuyong /bleu4/ with medial -l- may also be cognate.
13, 25 & 26. The first of these may stand in a morphological relationship to the other two:
*tow ‘to ignite/light’, *douX ‘fire’, *douH ‘to explode’ (see chapter 4).
15. There is a similar word in Mienic: Jiangdi Mien (Wang and Mao 15) /dzo7/ and
Xiangnan Mien (Wang and Mao 16) /dzəu7/ ‘to peck’; however, the rimes do not
correspond. The word is listed twice in Wang and Mao, once as ‘to peck’ and once as ‘to
dig’ because in Yanghao (1) ‘to peck’ has tone 5 and ‘to dig’ has tone 7. The forms in all
other locations with these meanings are identical, which suggests the differentiation is
limited to Yanghao. It is also used in White Hmong to mean ‘to hoe/clean the ground’.
Compare Chinese 撅 ‘to dig’ (OC *[g]ot > MC gjwot > Man. jué).
20. This may be the same as Chinese 峽 ‘gorge between two mountains’ (OC *N-kˤ<r>ep
> MC hɛp > Man. xiá; the same word as 狹 ‘narrow’), but the Chinese word does not have
a rounded vowel.
22. This word is only represented by two West Hmongic languages in Wang 1994:
Xianjin /ɳtʂeu8/ and Shimen /ɳdʐɦey8II/, but even on the basis of such scant evidence seems
to fit into rime 13.
27. Possibly from Chinese 蘆 ‘reed’ (MC lu > Man. lú), since Hmong and Mien bellows
are made from a cylinder with a pump inside (Lemoine 1972:130–131).
28. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
29. Compare Hmongic *hnuŋB ‘to forget’ (2.8/27).
32. Compare Cantonese /tsat/, Hainan Mun /tjat7/ ‘to plug (a hole)’. Jiangdi Mien (Wang
and Mao 15) has been substituted for Luoxiang Mien (8).
34 & 35. Given the merger of two different Hmong-Mien rimes to Mienic *-ok, these
words could equally well have been placed in rime 9e.
34. Compare Hmongic *ntaB ‘finger’ (2.4/4).
35. Compare Hmongic *ŋkroB/*ŋgroB ‘turbid’ (5.49/7).
 Proto Hmong-Mien 147

14
HM *u̯ i (20)
H *u̯ u : HM *u̯ i : M *u̯ i
1 2 3 4 5 6 7 8 9 10 11 HM
1. two o1 ɯ1 ɔ1 ɔ1a uA u1 va1 vi1 i1 wəi1 vi5 *ʔu̯ i

Discussion
A correspondence “set” containing only one word is not of great significance to the
reconstruction. ‘Two’ is a common word which is phonologically slight. There is no
supporting external evidence for this reconstruction from Chinese loans that pattern the
same way. Nonetheless, this reconstruction is faithful both to the Hmongic and Mienic
reflexes and to the pattern of preservation of the initial part of the rime in Hmongic, and
seems reasonable.
148 Chapter 2

15
15a. H *u̯ a
1 2 3 4 5 6 7 H
1. otter ɕha1 ȵʨhɑ1 ɳtʂhua1 ntsa1b nʔtshaAθja1 – *ntshju̯ aA
2. spinning wheel – – tʂhua1 – tshaA – – *tshju̯ aA
3. to twist/rub fha1 – sua1 sa1b saA va1’ – *S-phu̯ aA
4. to open out/undo tha3 nthɑ3 nthua3 – nʔthaB – – *nthu̯ aB
5. to marry (a
qha5 – – ha5a kaC – – *qu̯ aC
woman)
6. to tear ȴ̥i5 – dua5 ɭ̥a5a qlaC – kwɛ3 *qru̯ aC
7. to throw on
pa5 – mpua5 mpa5a – – ma5 *mpu̯ aC
(clothes)
8. to be used to ȴa6 ȴɑ6 – ɭəa6 laC – – *lju̯ aC
9. duck ka6 – – – ʁwaC – – *ɢu̯ aC
10. to meet ʨa6 – cua6 ʨəa6 ʑaC – – *ɟu̯ aC
11. rat – – tʂua6 pʐəa6 vjaB – – *bru̯ aC

Loans in set 15a:

bjuX (MC) 腐 fǔ ‘rotten’ > H *bu̯ aB ‘bad/spoiled’
dzjak (MC) 嚼 jué ‘to chew’ > H *cu̯ aC 38

15b. HM *əj (94, 151)

H *u̯ a : HM *əj : M *əi
1 2 3 4 5 6 7 8 9 10 11 HM/M
12.excrement qa3 qɑ3 qua3 ha3a qaB ka3 ka3 dai3 dai3 kai3 kai3 *N-KəjX
13. to do ɛ5 – ua5 aŋ5a aC – ɪ5̃ – ai5 wəi5 ai5 *ʔəjH
14. to kill ta5 tɑ5 tua5 ta5a – ta5 ta5 tai5’ tai5 tai5 dai5 *təjH
15. to die ta6 tɑ6 tua6 təa6 ðaC ta6 te6 tai6 tai6 tai6 tai6 *dəjH
16. bamboo shoot
za6 mʐɑ6 ɳtʂua6 mpʐəa6 mpjaC mpja6 mje6 bje6 bjai6 blai6 bɛi6 *mbrəjH
/orphan
17. diligent ŋa6 ŋɑ6 ɴqua6 ŋkəa6 ɴqaC ŋka6 ŋe6 – – gwai6 – *NGəjH
18. to fly dai5 dai5’ dai5 – *ʔdəiC

Loans in set 15b:

*[r][ə][j] (OC) > lij (MC) 梨 lí ‘pear’ > HM *rəj
tsij (MC) 粢 zī ‘sacrificial grain’ > M *tjəiA ‘millet’
dzyiX (MC) 柿 shì ‘persimmon’ > M *djəiB

15c. HM *u̯ əj (222)

H *u̯ a : HM *u̯ əj : M *u̯ əj
1 2 3 4 5 6 7 8 9 10 11 HM/M
19. sweet potato na2 – – – – – – du2 dɔːi2 dwai2 dai2 *ndu̯ əj

38
If this is indeed the source of the word in Hmongic, an explanation is needed for the register difference.
See also Chinese 噍 ‘to chew’ (OC *[dz]ewk-s > MC dzjewH > Man. jiào).
 Proto Hmong-Mien 149

15d. HM *aj (74, 112, 150, 151)

H *u̯ a : HM *aj : M *ai
1 2 3 4 5 6 7 8 9 10 11 HM/M
20. medicine ʨa1 ŋkɑ1 – ka1a kaA ðja1 ȵa1 – gjaːi1 hja1 tsɛi1 *N-kjaj
21. to come ta2 – tua2 ta2 ðaA – te2 taːi2 taːi2 ta2 tɛi2 *daj
22. waist l̥ a3 qwɑ3 dua3 l̥ a3a qlaB kla3 la3 klaːi3 klaːi3 kla3 lai3 *qlajX
23. to crow qa5 qɑ5 qua5 ha5a qaC ka5 ŋa5 kaːi5’ kaːi5 – kɔi5 *KajH
24. sharp ɣa6 ʐɑ6 ʐua6 – wjaC ŋkja6 – gai6 gjai6 lai6 hɛi6 *-rajH
25. river – – da2 dɔi2 *ndaiA
26. which laːi5 – ha2 – *[hl]aiC

Loans in set 15d:

*pʰˤaj-s (OC) > phaH (MC) 破 pò ‘to break’ > HM *phajH ‘to cut open’
*[k]ʷˤaj-s (OC) > kwaH (MC) 過 guò ‘to pass/surpass’ > HM *KʷajH ‘to cross (river)’
*[dz]ˤaj (OC) > dza (MC) 鹺 cuó ‘salty’ > M *ndzaiA

Discussion
Hmong-Mien rimes 15b, d share with rimes 11b and 12/17b-c the characteristic that the
final high front vocalic element triggered the introduction of a back rounded on-glide in
the development of Hmongic. For discussion, see chapter 2, section 1. Thus all words in
rime 15 have an intermediate Proto Hmongic form in *u̯ a, whether or not a back rounded
on-glide is reconstructed at the Hmong-Mien level on the basis of its appearance in Mienic.

Notes
2. The meaning of this word in White Hmong is ‘machine’. It is probably from Chinese 車
‘cart, chariot, thing with wheel’ (MC tsyhæ > Man. chē).
3. This word is probably the same as Mienic *ɕætD ‘to twist/rub’ (4.13/5).
6. Tan Trinh Pa-Hng (Niederer 1997) has been substituted for Baiyun Pa-Hng (7).
7. In White Hmong, ‘to wrap around the body’. Possibly from Chinese 被 ‘to cover self’
(OC *m-p(r)ajʔ > MC bjeX > Man. bèi).
9. Also Fengxiang (West Hmongic) /qa6/ (Wang and Mao 9). Expressive phonology.
14. Yanghao (1) from Office of Miao-Yao Research Lexicon 1987.
14 & 15. These words are important in the reconstruction of morphology in Hmong-Mien,
as they bear a morphological relationship to each other: see chapter 4. They are also to be
related in some fashion to proto-Austronesian *ma-aCay ‘to die’ and *pa-aCay ‘to kill’
(ACD).
16. This root also means ‘orphan, widow(er)’ in all Hmongic languages except Jiongnai,
Ho Ne and Pa-Hng. It may be that ‘bamboo shoot’ is basic and ‘orphan, widow(er)’ is a
derived meaning since the distribution of ‘bamboo shoot’ is wider. However, given Proto
Tai *br[a] ‘orphan’, perhaps ‘orphan/widow(er)’ is the basic meaning.
21. To come to a place other than one’s home.
23. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10). Possibly from Chinese 歌 ‘to sing; a song’ (OC *[k]ˤaj > MC ka > Man. gē), although
this word is not used for the song of animals, and the tones do not correspond.
150 Chapter 2

16
16a. H *u
1 2 3 4 5 6 7 H
1. to blow tsho1 phʐo1 tʂhua1 pʐa1b phjaA phu1 phjo1 *phruA
2. to trap/ensnare – – cua1 ʨæ1a ʨaA – – *cuA
3. fern xhə1 ʂo1 ʂua1 ʐa1b saA ŋkjeu1’ – *-hruA
4. on top of – – – sa1b saA – – *sjuA
5. sound/noise xhə1 ʂo1 ʂua1 ʐa1b saA ŋkjeu1’ ɕo1 *-hruA
6. tree fhu3 – – wa1b – – – *S-phruA
7. fog/cloud ho1 ho1 hua1 ɦa1b haA – ho1 *huA
8. green/blue zo2 mʐo1 ɳtʂua1 mpʐa1a mʔpjaA – mjo1 *mpruA
9. firm/durable – – ʈua2 ʈa2 – – – *druA
10. to sprinkle
– ȵʨho3 ɲchua3 ȵʨha3b ȵʔʨhaB – – *ɲchuB
/spill
11. rabbit – – lua3 ɭa3a ʔlaB – – *ʔljuB
12. sorghum – – ɳtʂua7 ntsa7a – – – *ntsjuC
13. to weed/hoe – ntho5 nthua5 – – – – *nthuC
14. ragged/shabby nei6 – ntua6 – qlaC – – *ɴɢluC

Loans in set 16a:

ʔu (MC) 烏 wū ‘crow’ > H *ʔuA
*[k](r)u (OC) > kjuw (MC) 鳩 jiū ‘turtledove’ > H *ɴquA ‘pigeon’39
ʔjiewH (MC) 要 yào ‘important/to want’ > H *ʔjuB ‘to ask for/want’

16b. HM *u (111, 225)

H *u : HM *u : M *u
1 2 3 4 5 6 7 8 9 10 11 HM/M
15. drum ȵə4 ɳɑŋ4 ɳʈua4 ɳʈəa4 ɳtʂaB ɳtʃu4 mjo4 dʑo4 ȡu4 du4 – *ndruX
16. to launder sho5 ntsho5 ntshua5 ntsa5b nʔtshaC ɳtʃu5 ȵo5 do5 du5’ dzu5 du5 *ntshuH
17. ramie/hemp no6 no6 ntua6 ntəa6 ntaC ntu6 mjo6 do6 du6 du6 – *nduH
18. tortoise – – – – – – – to6 tu6 – – *duH
19. return home mu4 mu4 mu4 mu4 *muB

Loans in set 16b:

*tˤuʔ (OC) > tawX (MC) 搗 dǎo ‘to husk/pound’ > HM *tuX
*[dz]ˤu (OC) > dzaw (MC) 槽 cáo ‘trough’ > M *dzuA

16c. HM *uə (191)

H *u : HM *uə : M *uə
1 2 3 4 5 6 7 8 9 10 11 HM
20. nine ʨə2 ʨo2 cua2 ʨa2 ʑaA tʃu2 ko2 du2 du2 ju2 ku2 *N-ɟuə
21. cattle pen ŋə2 – ŋkua2 ŋka2 ŋkaA – tɕõ2 – gu2 glu2 dzu2 *ŋgluə
22. thunder – – – – – – mo1 bu1 bu5 – bjau1 *mpuə

39
This word also could have been borrowed from Tibeto-Burman *k(r)əw ‘dove’ (Matisoff 2003).
 Proto Hmong-Mien 151

Loans in set 16c:

*[g](r)aʔ (OC) > gjoX (MC) 粔 jù ‘rice cake’ > HM *ŋkjuəX
*[b]ək-s (OC) > bjuwH (MC) 伏 fú ‘to hatch’ > HM *buəH ‘to hatch/embrace’

16d. HM *ow (172)

H *u : HM *ow : M *ow
1 2 3 4 5 6 7 8 9 10 11 M
23. shuttle gou4 glɔu4 – – *ɴɢlowB

Loans in set 16d:

*[g](r)aw (OC) > gjew (MC) 橋 qiáo ‘bridge’ > HM *ɟow
mjuH (MC) 霧 wù ‘fog’ (Cantonese /mou6/) > M *mowC

Discussion
There was a merger of Hmong-Mien *uw (3f), *u̯ o (7c), and *u (16c) to Mienic *u.
‘Tortoise’, ‘to return home’, and ‘trough’ have been placed here, in rime 16c. But it should
be remembered that these words are not necessarily to be taken back to *u in Proto
Hmong-Mien.

Notes
1. There are Chinese words with ph- meaning ‘to blow’, but this may be independent use
of expressive phonology in both language groups.
6. The Yanghao (1) form is a classifier for vegetation (Office of Miao-Yao Research
Lexicon 1987), as is West Hmongic Shimen /faɯ1/ (Wang and Mao 4).
8. ‘Green/blue’ and Mienic *ʔbɔuB (1.4/3) ‘grey’ may be cognate: Mienic ‘grey’ in Hainan
Mun also means ‘blue’ (Shintani and Yang 1990).
10. Also ‘to pour out’.
12. Upon inspection of more cognates, it becomes clear that tone 7 in these two forms for
‘sorghum’ are tone sandhi forms derived from tone 5.
14. Reflexes in Xianjin (/ɳʈua6/) and Qingyan (/ɳʈo6/) have retroflex initials, so this word
might better be reconstructed as *ɴɢruC. It is possibly from a prefixed form of Chinese 陋
‘humble, mean’ (OC *[r]ˤo-s > MC luwH > Man. lòu).
18. Reconstructed to Proto Hmong-Mien on the basis of its appearance in Mienic and
Qibainong Bunu /tu6/ (Wang and Mao 10).
20. From Tibeto-Burman *gəw ‘nine’ (Matisoff 2003). See chapter 5.
21. Biao Min (10) is from Solnit 1982. This may be the same as Chinese 圉 ‘prison, pen’
(OC *m-q(ʰ)(r)aʔ > MC ngjoX > Man. yǔ).
22. This may be related to Hmongic *S-phoA ‘thunder’ (1.2/7). See discussion under onset
1.2.
23. This is probably related to the Hmongic word for ‘shuttle’, *ɴɢəŋB (5-6.6/21).
152 Chapter 2

18
18a. H *in
1 2 3 4 5 6 7 H
1. to dry in sun – – ʐia1 ʐæin1a – – – *ʔrinA
2. monkey lei1 – lia1 læin1a – lai1 – *ʔlinA
3. red – – lia1 læin1a ʔlenA – – *ʔlinA
4. skirt tei1 tɛ1 tia1 tæin1a tenA ðai1 tĩ1 *tinA
5. to weave/plait hei1 hɛ1 hia1 ɦæin1b henA xan1 – *hinA
6. alive – – cia2 ʨein2 ʑiA – – *ɟinA
7. front surface – – ntia2 – ɴqlenA – – *ɴɢlinA
8. level – – tia2 tæin2 – – – *dinA
9. crop of bird pi3 – tsia5 pjein5a puB pai3 pẽ1 *pjinB
10. ginger khi3 – qhia3 hæin3b qhwjenB khai3 – *qhinB
11. grape qei3 qɛ3 – hæin3a qenB – ji3 *qinB
12. cooking
vi4 wɛ4 jia4 ʑein4 wenB van4 vɪ4̃ *wjinB
pot/pan
13. cool (water) sei4 tsɛ4 tsia4 sein4 – tʃai4 ʨɦɪ4̃ *dzinB
14. persimmon mi4 mɛ4 – – menB – – *minB
15. to drive in
– ȵʨɛ5 ɳtʂia5 – nʔtsenC – ni5 *ntsjinC
(nail)
16. mussel/clam l̥ iə6 pɹɛ4 plia6 – – – – *blinC
17. to be wedged
– ɳʈhɑ7 – ɳʈei5b – – – *ntrinC
in
18. female mi8 – nia8 mi8 menC mai8 mɛ7̃ *minD

Loan in set 18a:

sying (MC) 升 shēng ‘liter measure’ > H *sjinA ‘measure for grain’

18b. HM *im (4, 5)

H *in : HM *im : M *im
1 2 3 4 5 6 7 8 9 10 11 HM/M
19. bitter i1 ɛ1 ia1 æin1a – an1 jɪ1̃ im1 im1 in1 jɛm1 *ʔim
20. thorn jim3 ȡim3 – – *ʔɟimB

Loans in set 18b:

ljem (MC) 鎌 lián ‘sickle’ > HM *ljim
sim (MC) 心 xīn ‘heart’ > M *simA
tsyim (MC) 箴 zhēn ‘needle’ > M *simA

18c. HM *in (7)

H *in : HM *in : M *in
1 2 3 4 5 6 7 8 9 10 11 HM
21. tooth m̥ hi3 ɕɛ3 hn̥ ia3 mi3b m̥ jenB m̥ ai3 mɪ3̃ – – mjen3 – *hmjinX
 Proto Hmong-Mien 153

Loans in set 18c:

*-[b]ˤe[n]ʔ (OC) > benX (MC) 辮 biàn ‘a braid’ > HM *mbjinX ‘to braid/a braid’
pjien (MC) 鞭 biān ‘whip’ (Cant. /pin1/) > M *pinA

18d. HM *iŋ (8)

H *in : HM *iŋ : M *iŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
22. field ȴi2 – lia2 ɭæin2 lenA leŋ2 lĩ2 giŋ2 giːŋ2 ljɛ2 ljaŋ2 *ljiŋ
23. monkey biŋ1 biːŋ1 – bjaŋ1 *ʔbiŋA
24. crop (of
ʨiŋ3 ȡiŋ3 – – *ciŋB
bird)/gizzard

Loans in set 18d:

syeng (MC) 聲 shēng ‘sound’ > M *siŋA
sengX (MC) 醒 xǐng ‘to sober up’ > M *siŋB
Cant. /pɛːŋ5/ ‘to hide onself’ > M *piŋC

18e. HM *i̯ əm (135)

H *in : HM *i̯ əm : M *i̯ əm
1 2 3 4 5 6 7 8 9 10 11 HM/M
25. to sniff at m̥ hi5 – hni̥ a5 mi5b m̥ jenC m̥ ai5 mi5 hom5 hɔːm5’ m̥ ja5 – *hmji̯ əmH
26. to cover
hom6 hɔːm3 – hum3 *hi̯ əmB/C
(with tile)

18f. HM *i̯ ən (119, 137)

H *in : HM *i̯ ən : M *i̯ ən
1 2 3 4 5 6 7 8 9 10 11 HM
27. liver xhi1 ʂɛ1 ʂia1 ʐæin1b senA ŋkheŋ1’ hɪ1̃ gan1 gjen1’ ɬan1 – *-hri̯ ən

Loan in set 18f:

ngin (MC) 銀 yín ‘silver’ > HM *ɲʷi̯ ən

18g. HM *i̯ əŋ (83, 159)

H *in : HM *i̯ əŋ : M *i̯ əŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
28. high/ tall xhi1 ʂɛ1 ʂia1 ʐæin1b senA ŋkheŋ1’ hɪ1̃ gaŋ1 gjaŋ1’ ɬaŋ1 dzɔŋ1 *-hri̯ əŋ
29. new xhi1 ɕɛ1 tʂhia1 sæin1b senA ŋkheŋ1’ sɪ1̃ ɕaŋ2 saŋ2 saŋ1 sjaŋ1 *tshji̯ əŋ
30. plum – – – – – – le3 gaŋ3 – ɬaŋ3 ȵɔŋ3 *hli̯ əŋX
31. shallow – – ŋkla3 – – – – glaŋ3 gjaŋ3 – lɛŋ3 *ŋkli̯ əŋX
32. smooth
– mjɛ6 mpla6 mplein6 mplenC – wɪ6̃ – bjaŋ6 – – *mbli̯ əŋH
/glossy
33. POSSESSIVE ȵaŋ1 ȵaŋ1 nin2 – *ʔɲi̯ əŋA

Loan in set 18g:

*N-s-r̥ eŋ (OC) > tshjeng (MC) 清 qīng ‘clear’ > HM *ntshji̯ əŋ
154 Chapter 2

Discussion
Note the close similarity in the reflexes of rime 18 (*-in) and 20 (*-en). The final Hmongic
*-n disappeared in most languages: it survives only in West Hmongic (as in Zongdi (4) and
Fuyuan (5) above), and in Pa-Hng as nasalization on the vowel.

Notes
2. The Mienic word for ‘monkey’ (*ʔbiŋA 1.4/18) shares the same tone and rime with this
word, but begins with a different initial, suggesting the possibility of an ancient disyllabic
source. Compare Proto Tai *liŋ ‘monkey’ (Li 1977).
2 & 3. These may be the same word, if the monkey was referred to as ‘the red one’. How
the connection between ‘monkey’ and ‘red’ may relate to the connection suggested above
between the Hmongic and Mienic words for ‘monkey’ is unclear.
7. There are also Hmongic forms that match this word in meaning, rime, and tone but have
an *mbl- initial: Gaopo /mplɛ2/, Zongdi /mplæin2/, and Fengxiang /mplen2/ (Wang 1994).
16. Yanghao (1) and Jiwei (2) from Office of Miao-Yao Research Lexicon 1987.
18. ‘Female’ probably did not originally have a nasal coda: expected nasal codas in Gaopo
(7) above, as well as in Qingyan and Qibainon Bunu (West Hmongic) do not appear, and
the tone category 8 suggests a final -p or -t. But based on the overall correspondence, it is
placed here for the time being. In White Hmong, ‘mother’.
21. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10). This word for ‘tooth’ could also fall under rime 18f and be reconstructed *hmji̯ ənX.
22. Compare the almost identical Chinese 田 ‘field; to hunt’ (OC *lˤiŋ > MC den > Man.
tián) and Tibeto-Burman *liŋ ‘field’ (Matisoff 2003). Both Haudricourt and Strecker
(1991) and Sagart (1999:183–184) entertain the possibility that this was a loan from
Hmong-Mien to Chinese.
23. The Hmongic word for ‘monkey’ (*ʔlinA 2.40/18) shares the same tone and rime with
this word, but begins with a different initial, suggesting the possibility of an ancient
disyllabic source.
24. Downer (1973) gives Cantonese /khan/ ‘gizzard’ as the source of Mien /kien1/,
although this is not obviously the same word.
25. Both ‘to sniff at’ and ‘to taste/try’ (Hmong-Mien *hmeiH, 1.8/12) mean to smell or
taste intentionally. Different words exist for ‘to smell’ and ‘to taste’ unintentionally.
Compare Tibeto-Burman *s-nam ‘to smell’ (‘directive, causative, intensive’ s- prefix; see
Matisoff 2003:99 ff.).
26. These forms for ‘to cover’, as well as Mienic *komB (5.1/21), appear to be borrowings
from Tai-Kadai (see entries in section 2.2). The Mienic reflexes in both tone 3 and tone 6
suggest an initial voicing alternation as well as a tone alternation, but the tones may simply
be “loan-tones” (see chapter 3), chosen to reflect the tone of the donor language(s) as
closely as possible.
29. This may be from a variant of Chinese 清 ‘clear’, given its close resemblance to pre-
nasalized *ntshji̯ əŋ in this set (3.20/18); see chapter 4.
30. This may be from Chinese 李 ‘plum’ (MC liX > Man. lǐ).
31. Green Mong has been substituted for White Hmong (3).
32. In White Hmong, ‘brilliant’.
 Proto Hmong-Mien 155

19
19a. H *æn
1 2 3 4 5 6 7 H
1. bamboo pipes ki2 – qen2 hæin2 ʁwjaŋA – – *ɢænA
2. person nɛ2 ne2 nɛŋ2 – naA nai2 – *nænA
3. crossbow n̥ hen3 – hn̥ ɛŋ3 nein3b n̥ aB – n̥ ẽ3 *hnænB
4. horse ma4 – nɛŋ4 mein4 maB me4 mɪ4̃ *mjænB
5. lazy ŋi4 ɲe4 ŋkɛŋ4 ȵʨein4 ŋkaŋB – ŋɛ4 *ŋglænB

19b. HM *æn (42, 175, 247)

H *æn : HM *æn : M *æn
1 2 3 4 5 6 7 8 9 10 11 HM/M
6. footprint/track m̥ ha3 – nen3 mein3a m̥ aB – – m̥ wan3 ŋwaːn3’ m̥ jɛn3 – *hmjænX
7. tiger ʥan2 gjaːn2 jɛn2 kjɛn2 *ŋgljænA
8. person/Mien mwan2 mun2 min2 min2 *mjænA
9. spirit/ghost m̥ wan3 ŋwaːn3 m̥ jɛn3 mjɛn3 *hmjænB

19c. HM *an (99, 118)

H *æn : HM *an : M *an
1 2 3 4 5 6 7 8 9 10 11 HM/M
10. cogongrass qɛ – ɴqɛŋ1
1
ŋkæin1a – ŋkan1 ŋɛ gaːn1
1
gaːn1 gwan1 gɔn1 *NKan
11. CLF-people lɛ2 le1 lɛŋ2 læ2 – laŋ1 læ̃ 2 laːn2 laːn2 – na1 *lan
12. back (of body) tan4 tan4 – – *danB
13. to snore gaŋ2 daːn2 – – *ŋgjanA
14. to disperse daːn5 daːn5’ dzan5 – *dzhanC
15. pullet/hen kaːn5 kjaːn5 – – *klanC

Loans in set 19c:

*[tʰ]ˤa[n]-s (OC) > thanH (MC) 炭 tàn ‘charcoal’ > HM *thanH
kjɨn (MC) 筋 jīn ‘sinew’ > M *kʷjanA

Notes
2 & 8. Wang and Mao link these Hmongic forms in n- that mean only ‘person’ to the
Mienic forms in m- that mean both ‘person’ and serve as autonyms for the different Mienic
groups. This Mienic word *mjænA (1.24/19), which probably comes from Chinese 民
‘people’ (OC *mi[ŋ] > MC *mjin > Man. mín), is entered separately here. Pa-Hng /mjɛ2/
patterns with Mienic, as it does in a number of other instances (see ‘chaff/husk’ and
‘thunder’).
3. This is undoubtedly the same word as Mienic *hnəkD ‘crossbow’ (2.8/7). Compare
Mon-Khmer *snaʔ (Shorto #97), Chinese 弩 (OC *[n]ˤaʔ > MC nuX > Man. nǔ), and Proto
Tai *hnaa B1.
4. Although Hmongic ‘horse’ is ultimately related to Chinese 馬 (OC *mˤraʔ > MC mæX
> Man. mǎ), these forms suggest a more direct connection to some Tibeto-Burman
language (TB *mraŋ, Matisoff 2003). Furthermore, the rime matches item 6 in this set,
156 Chapter 2

‘footprint/track’, which is also a Tibeto-Burman borrowing. Mienic words for ‘horse’ are
all borrowings from Chinese.
5. Compare Chinese 懶 ‘lazy’ (OC *[r]ˤanʔ > MC lanX > Man. lǎn).
6. Probably from a derivative of Tibeto-Burman *s-naŋ ‘to follow’ (Matisoff 2003).
7. Compare Mon-Khmer *klaʔ ‘tiger’ (Shorto #197).
8. See also Mien dialects Jiangdi (Wang and Mao 15) /mjen2/ and Xiangnan (Wang and
Mao 16) /mjəŋ2/. Pa-Hng /mjɛ2/ patterns with Mienic here as it does in a number of other
instances (see ‘chaff/husk’ and ‘thunder’). The probable source of the Mienic autonym,
which takes the variant forms ‘Mien’, ‘Mun’, and ‘Min’, is Chinese 民 ‘people’ (OC
*mi[ŋ] > MC mjin > Man. mín). Wang and Mao link these Mienic forms in m- that mean
both ‘person’ and ‘group’ to the Hmongic *nænA ‘person’ (2.9/19) since the rimes and
tones correspond. The Hmongic word for ‘person’ is entered separately here for the time
being.
8 & 9. The sporadic -w- is secondary: -j- dropped, then -w- developed between a labial and
/a/.
11. Compare Chinese 郎 ‘young man’ (Man. láng); in Min a homophonous word means
‘person’.
12. This word is reconstructed with an -a- on the basis of the long vowel in Jiangdi Mien
/taːn4/ (Wang and Mao 15).
13. Compare Chinese 鼾 ‘to snore’ (MC xan > Man. hān), which matches in rime, but not
in onset.
 Proto Hmong-Mien 157

20
20a. H *en (24)
1 2 3 4 5 6 7 H
1. forehead ȵ̥haŋ1 ɕɛ1 – ȵi1b – – nɪ1̃ *hɲenA
2. cooked/ripe ɕhaŋ3 ɕɛ3 ʂia3 sæin3b senB θeŋ3 sɪ3̃ *sjenB
3. shallow ȵi4 – – ȵi4 – – ȵɦɪ4 *ɲenB

Loan in set 20a:

nyemX (MC) 姌 rǎn ‘slender’ > H *ɲenB ‘thin’

20b. HM *em (59)

H *en : HM *em : M *em
1 2 3 4 5 6 7 8 9 10 11 HM/M
4. to steal ȵaŋ6 ȵɛ6 ȵia6 ȵi6 ȵenC ȵiŋ6 ȵɪ6̃ nim6 neːm6 – ȵɛm6 *ɲemH
5. petal ɬim5 leːm5 – ɬiŋ5 *hlemC

20c. M *i̯ em (40)

1 2 3 4 5 6 7 H HM 8 9 10 11 M
6. mountain 2 2 2
kem kiːm kleŋ – *gli̯ emA
(range)
7. to scold/curse heːm5 – – – *hi̯ emC

Loan in set 20d:

*N-kˤem ‘to doubt’ (OC) > hem (MC) 嫌 xián ‘to dislike’ > M *ɣi̯ emA ‘to dislike’

20d. HM *eŋ (44)

H *en : HM *eŋ : M *eŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
8. 1st day of 1 1b A 1
– – sia sæin sen – – – – – hɛŋ *seŋ
month
9. to lift – – – – – – ʨõ1 ʨɛŋ1 ȶeːŋ1 kjɛ1 kaŋ1 *kjeŋ
10. to choke – – – – – ɳtʃei5 ni5 ʥɛːŋ5 – ʥɛ5 jaŋ5 *ntsjeŋH
11. pair – geːŋ6 lɛ6 dzaŋ6 *reŋC

Loan in set 20e:

ngængH (MC) 硬 yìng ‘hard/solid’ > M *ŋjeŋC

20e. HM *æm (79)

H *en : HM *æm : M *æm
1 2 3 4 5 6 7 8 9 10 11 HM
12. to weep ȵaŋ1 ȵɛ3 ȵia3 ȵi3a ʔȵenB ȵiŋ3 – ȵem3 ŋjim3 ȵan3 ȵɛm3 *ʔɲæmX
158 Chapter 2

Discussion
Note the close similarity in the reflexes of rime 18 (*in) and 20 (*-en). Notable is the
absence of Hmong-Mien *en in this rime set, which has been reconstructed for rime 25
instead (Hmong-Mien *en > H *ɛŋ, M *ei). It may be that Hmong-Mien *en was
inherently unstable.
Most of the words in this rime set have palatal onsets; this may eventually make it possible
to combine it with another rime set. The distribution remains unclear, however.
Wang and Mao rime sets 23 and 24 also belong here. Wang and Mao rime 23 consists of
one word: Hmong-Mien *kʷjan ‘catty’ from Chinese 斤 (MC kjɨn > Man. jīn ). Hmong-
Mien *an goes to Hmongic *en in this word rather than to the expected Hmongic *æn
(rime 19 above). Wang and Mao rime 24, in addition to ‘forehead’, ‘cooked’, and ‘thin’
above, includes Chinese ‘thousand’ and ‘money’, which can be reconstructed for Hmong-
Mien with *i̯ en:

tshen (MC) 千 qiān ‘thousand’ > HM *tshi̯ en

dzjen (MC) 錢 qián ‘money’ > HM *dzi̯ en

Notes
3. ‘Shallow’ could as easily have been reconstructed *ɲinB and placed in rime 18a above. It
does not fit into either set precisely. Wenjie Pa-Hng (Wang and Mao 12) has been
substituted for Baiyun Pa-Hng (7).
5. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
6. Sunjiang Biao Min (Wang and Mao 22) has been substituted for Dongshan Biao Min
(10).
7. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). This may
belong with Hmongic *qeC (5-6.1/10) ‘to scold/curse’, as represented in Wang and Mao
1995 (327).
10. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
 Proto Hmong-Mien 159

21
21a. H *əŋ
1 2 3 4 5 6 7 H
1. to be at/live ȵaŋ1 ȵi1 ȵɔ1 ȵɔ1a ʔȵoŋA ȵɐŋ1 ȵõ1 *ʔɲəŋA
2. breast/chest – – ɳʈɔ 1 ɳʈoŋ1a – – – *ntrəŋA
3. animal fat/oil ȶaŋ2 – ʈɔ2 ʂoŋ2 ʐoŋA – – *grəŋA
4. boat/ship ȵaŋ2 ŋɑŋ2 ŋkɔ2 ŋkoŋ2 ŋkoŋA ɳtʃoŋ2 ŋɯ2 *ŋgjəŋA
5. to sink taŋ2 – tɔ6 toŋ6 toŋA – tæ̃ 4 *dəŋA
6. to low/bellow – – ɴqɔ3 ŋkoŋ3a ɴʔqoŋB – – *ɴqəŋB
7. to return ȶaŋ3 – ʈɔ3 ʈoŋ3a tʂoŋB – – *trəŋB
8. to roll down ȴ̥aŋ3 cɑŋ3 dɔ3 – qloŋB – – *qləŋB
9. to bury ȴaŋ4 ȴɑŋ4 lɔ4 ɭoŋ4 loŋB – læ̃ 4 *ljəŋB
10. Allium – – dɔ4 l̥ oŋ4 ʁloŋB – – *ɢləŋB
11. shuttle ŋaŋ4 nɑŋ4 ɴqɔ4 ŋkoŋ4 ɴqoŋB ŋwaŋ4 – *ɴɢəŋB
12. to swallow ŋaŋ4 ŋu4 – – ɴqoŋB – – *ɴɢəŋB
13. to wait taŋ4 tɑŋ4 tɔ4 toŋ4 ðoŋB – te4 *dəŋB
14. to let go ɕaŋ5 ʨɑŋ5 tʂɔ5 – – – ʨõ5 *tsjəŋC
15. to wipe ɕhaŋ5 ɕɑŋ5 ʂɔ5 soŋ5b soŋC – – *sjəŋC
16. to be – – jɔ6 ʑoŋ6 ʑoŋC – ȵɯ6 *jəŋC
17. interior ȵaŋ6 ɳɑŋ6 ɳʈɔ6 – – – – *ndrəŋC

Loan in set 21a:

*[k]ʰˤoŋʔ (OC) > khuwngX (MC) 孔 kǒng ‘hole’ > H *qhəŋB

21b. M *əm (229)

1 2 3 4 5 6 7 8 9 10 11 M
18. to be at/live jem1 jam1 jan1 – *ʔjəmA
19. bracelet ʨem2 sam2 – – *gjəmA

Loan in set 21b:

*[r][ə]m (OC) > lim (MC) 淋 lín ‘to pour’ > M *rəmA ‘to water’

21c. HM *ən (157, 176, 194)

H *əŋ : HM *ən : M *ən
1 2 3 4 5 6 7 8 9 10 11 HM/M
3 3 3 3a B 3 3
20. to shoot paŋ pɑŋ pɔ poŋ poŋ poŋ pɔ̃ pwan3 fan3 – bun3 *pənX
21. fine (flour) moŋ4 mɑŋ4 mɔ4 – – – mæ̃ 4 mwən6 – mun6 mun6 *mənX/H
22. oneself kan2 kan2 kan4 – *gənA

Loan in set 21c:

ʔwanX (MC) 碗 wǎn ‘bowl’ > M *ʔwənB

21d. HM *əŋ (140)

H *əŋ : HM *əŋ : M *əŋ
1 2 3 4 5 6 7 8 9 10 11 HM
23. night m̥ haŋ5 m̥ hɑŋ5 hm̥ ɔ5 mɔ5b m̥ oŋC – mæ̃ 4 m̥ aŋ3 – m̥ ɔŋ5 mɔŋ5 *hməŋH
160 Chapter 2

Loans in set 21d:

*(mə.)rəʔ (OC) > liX (MC) 里 lǐ ‘village’ > HM *rəŋX
*N-[dz]eŋ-s (OC) > dzjengH (MC) 淨 jìng ‘clean’ > M *ndzəŋC

21e. HM *əum (193)

H *əŋ : HM *əum : M *om
1 2 3 4 5 6 7 8 9 10 11 HM/M
24. to hear n̥ haŋ3 n̥ hɑŋ3 hn̥ ɔ3 nɔ3b n̥ uB n̥ ɐŋ3 n̥ ɔ3̃ nom3 num3 n̥ ən3 – *hnəumX
25. to swell aŋ5 ɑŋ5 ɔ5 oŋ5a oŋC oŋ5 ʔõ5 om5’ ɔm5 ən5 – *ʔəumH
26. to cover kom3 kɔm3 – – *komB

21f. HM *əun (214)

H *əŋ : HM *əun : M *un
1 2 3 4 5 6 7 8 9 10 11 HM
27. tender ɣaŋ5 ʐɑŋ5 – – – ɳtʃoŋ5 – gun5’ – lun5 in5 *ʔrəunH
28. to be fat ȶaŋ6 ʈɑŋ6 ʈɔ6 ʂoŋ6 ʐoŋC ʃoŋ6 ʨõ6 kun6 kun6 klin6 tin6 *grəunH

21g. HM *əuŋ (218)

H *əŋ : HM *əuŋ : M *uŋ
1 2 3 4 5 6 7 8 9 10 11 M
29. deaf – duŋ1 dwə1 dɔŋ1 *ʔduŋA
30. to bark ʨuŋ5’ ȶuŋ5 ȡwə5 – *cuŋC

Loans in set 21g:

*mˤoŋʔ (OC) > muwngX (MC) 蠓 měng ‘midge’ > HM *məuŋX ‘fly (N.)’
kjang (MC) 薑 jiāng ‘ginger’ > M *khjuŋA
pjangH (MC) 放 fàng ‘to let go’ > M *puŋC

21h. HM *əaŋ (101)40

H *əŋ : HM *əaŋ : M *aŋ
1 2 3 4 5 6 7 8 9 10 11 HM
31. cooked rice – – hn̥ ɔ5 ɳɔ5b – – – naːŋ5 naŋ5’ n̥ aŋ5 nɔŋ5 *hnrəaŋH

Discussion
Set 21b, made up of Mienic words entirely, is included here because it is assumed that
Hmongic ‘to be at/live’ with a palatal nasal initial in set 21a is related to Mienic ‘be at/live’
with a palatal glide initial. Sets 21c and 21d contain words with labial initials that
developed a rounded glide in Mienic before -əN (see also rime 28).
There is only one loanword to exemplify each of two more rimes belonging to set 21:

40
All other words in Wang and Mao rime set 101 are attested in Mienic only, and can be placed in rime 24.
‘Rope’ has Hmongic cognates meaning ‘sash/cord/belt’, and the expanded set also belongs in rime 24
rather than here.
 Proto Hmong-Mien 161

21i. HM *əan (194): HM *phəan ‘CLF-quilts’ from Chinese 片 ‘half; partial; one-sided’
(MC phenH > Man. piàn).
21j. HM *u̯ en (194): HM *chu̯ en ‘to thread’ from Chinese 穿 ‘to bore through’ (MC
tsyhwen > Man. chuān).

Notes
1. Mienic *ʔjəmA (4.10/21) ‘to be at/live’ is probably related to this word.
3 & 28. Although only attested in Hmongic, ‘animal fat/oil’ clearly involves the same root
as ‘to be fat’ under rime 21f. It appears that the widespread ‘to be fat’ derived from the
now more restricted ‘animal fat/oil’ via C-tone derivation (see chapter 4).
4. Compare Central Malayo-Polynesian *waŋka[ŋ] ‘canoe’ (ACD).
9. This may be from Chinese 斂 ‘to dress a corpse’ (Man. liàn), although the tones do not
correspond.
11. Mienic forms for ‘shuttle’ are probably related to this set (Luoxiang Mien /gou4/,
Liangzi Mun /dlou4/, Lanjin Mun /glɔu4/). But there is no final nasal in Mienic, and this
rime correspondence is not attested elsewhere; the initials are also problematic.
18. Hmongic *ʔɲəŋA (4.7/21) ‘to be at/live’ probably fits with the Mienic *ʔjəmA, but the
absence of a nasal initial in Mienic is hard to explain (Wang and Mao explain it as function
word reduction).
19. Also Changping Mien /kjəm2/ (Wang and Mao 1995). Compare Lungchou (Tai-Kadai)
/kiim2/ (Li 1977).
20. Compare Malayo-Polynesian *panaq ‘to shoot’ and Mon-Khmer *paɲʔ ‘to shoot’
(Shorto #905). In White Hmong, ‘to throw’.
20 & 21. The labial initial gave rise to a medial -w- in Mienic. The -w- persists in
Luoxiang (8), and rounded the vowel in Biao Min and Zao Min (10 & 11).
22. This is technically Hmong-Mien, since this word also appears in Ho Ne (Hmongic):
/kan4/. This form is irregular, however, and looks like a loanword from some Mienic
language.
23. Chinese 晚 ‘evening’ (Man. wǎn) > Mien /muən1/ seems to be different. More likely
connections are to Tibeto-Burman *s-muːŋ ‘dark’ (Matisoff 2003) or Mon-Khmer
*m[h][ɯə]h ‘evening, night’ (Shorto #264) or *maŋ ‘night, evening’ (Shorto #638).
26. This word for ‘to cover’, as well as Mienic *hi̯ əmB/C ‘to cover (with tile)’ (7.13/18),
appear to be recent borrowings from Tai-Kadai. Li (1977) reconstructs a number of similar
words with this general meaning: *xr[um]B1 ‘to cover, veil; *ɣ[um]B2 ‘to cover, protect’;
*homC1 ‘to cover up’. See also the Mienic forms under ‘to cover’ with initial g- in L-
Thongkum’s reconstruction: Mien /gom3/, Mun /gəm3/, etc. (1993:204).
31. Possibly from Chinese 饟 ‘bring food to workers in the field’ > ‘food brought to
workers in the field’ (OC *n̥ aŋ(ʔ)-s > MC syangH > Man. xiǎng).
162 Chapter 2

22
22a. H *ɛŋ
1 2 3 4 5 6 7 H
1. black l̥ ɛ1 qwe1 du1 l̥ oŋ1a qlaŋA klaŋ1 – *qrɛŋA
2. clay pot/earthen
– – hu1 ɦoŋ1b haŋA – – *hɛŋA
jar
3. CLF-
lɛ1 le1 lu1 noŋ1a ʔlaŋA xoŋ1 lɔ̃1 *ʔlɛŋA
bowls/houses
4. to feed pɛ1 – pu1 – – – – *pɛŋA
5. star qɛ1 qe1 qu1 hoŋ1a qaŋA ka1 kɔ̃1 *qɛŋA
6. sun/day n̥ hɛ1 n̥ he1 hn̥ u1 noŋ1b n̥ aA n̥ ɔ1’ n̥̥ ẽ1 *hnɛŋA
7. tree of heaven – – ju2 ʑoŋ2 ʑaŋA – – *jɛŋA
8. 1SG (I/me) – – ku3 koŋ3a kaŋB – – *kɛŋB
9. artemisia xhi3 – – ʐoŋ3b saŋB – – *hrɛŋB
10. husked
shɛ3 se3 tshu3 soŋ3b tshaŋB – – *tshɛŋB
rice/millet
11. short lɛ3 le3 lu3 loŋ3a ʔlaŋB laŋ3 lɔ̃3 *ʔlɛŋB
12. 1SG (I/me) vi4 we4 – – – va4 vɔ̃4 *wɛŋB
13. CLF-horses tɛ4 – tu4 toŋ4 ðaŋB – – *dɛŋB
14. midday meal – – ʂu5 soŋ5b saŋC – – *sjɛŋC
15. to ask nɛ6 ne6 nu6 noŋ6 naC nɔ6 nɪ6̃ *nɛŋC

Loans in set 22a:

ljang (MC) 量 liáng ‘to weigh/measure’ > H *ljɛŋA ‘to measure (rice)’
ʔjang (MC) 秧 yāng ‘seedling’ > H *ʔjɛŋA
yangH (MC) 養 yàng ‘to keep, support’ > H *jɛŋC ‘to raise (chickens)’

22b. HM *ɛm (174)

H *ɛŋ : HM *ɛm : M *əm
1 2 3 4 5 6 7 8 9 10 11 HM/M
16. body louse tɛ3 te3 tu3 toŋ3a taŋB taŋ3 nɔ̃3 tam3 tam3 dan3 dam3 *tɛmX
17. to ladle dam3 – – da3 *ʔdəmB

Loans in set 22b:

*(pə.)r[ə]mʔ (OC) > limX (MC) 廩 lǐn ‘barn/granary’ > HM *rɛmX
nemH (MC) 念 Cant. /nɐm3/ ‘to think of’ > M *hnəmB~*hləmB

22c. HM *u̯ ɛn (100)

H *ɛŋ : HM *u̯ ɛn : M *u̯ ɛn
1 2 3 4 5 6 7 8 9 10 11 HM
18. son/boy tɛ1 te1 tu1 toŋ1a taŋA – tɔ̃1 – tɔːn1 twan1 dan1 *tu̯ ɛn
19. wet – nte1 ntu1 ntoŋ1a nʔtaŋA ntaŋ1 tɔ4̃ – dɔːn1 – – *ntu̯ ɛn

Loan in set 22c:

ljen (MC) 聯 Cant. /lyːn2/ ‘to sew’ > M *lju̯ ɛnA
 Proto Hmong-Mien 163

22d. HM *ɛŋ (102, 104)

H *ɛŋ : HM *ɛŋ : M *əŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
20. sky/heaven vɛ2 – – – – – vɔ̃2 – – – vaŋ2 *wɛŋ
21. to snap tɛ5 te5 tu5 toŋ5a taŋC taŋ5 tɔ5̃ – – taŋ5 daŋ5 *tɛŋH
22. to put on
taŋ2 taŋ2 – – *dəŋA
/wear(jewelry)
23. to resemble n̥ aŋ2 naŋ3’ – – *hnəŋB
24. short naŋ3 niŋ3 naŋ3 naŋ3 *ʔnəŋB

Loans in set 22d:

*s-təŋ-s (OC) > tsingH (MC) 甑 zèng ‘rice steamer’ > HM *tsjɛŋH
tongH (MC) 凳 dèng ‘stool/bench’ > M *təŋC

22e. HM *u̯ ɛŋ (252)

H *ɛŋ : HM *u̯ ɛŋ : M *u̯ ɛŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
25. full pɛ3 pe3 pu3 poŋ3a paŋB paŋ3 pɔ̃3 pwəŋ3 pɔŋ3 pɔŋ3 baŋ3 *pu̯ ɛŋX
26. cold (day) ʨwəŋ3 sɔŋ3 ȶɔŋ3 kuŋ3 *kju̯ ɛŋB
27. quilt ɕwəŋ5 tθɔŋ5’ sɔŋ5 suŋ5 *su̯ ɛŋC

Loans in set 22e:

tsyowngH (MC) 種 zhòng ‘to grow’ > HM *n-tju̯ ɛŋH
phjowngX (MC) 奉 fèng ‘to present with both hands’ > M *phu̯ ɛŋB ‘to carry in 2 hands’
bjuwX (MC) 婦 fù ‘woman/wife’ > M *mbu̯ ɛŋB ‘daughter-in-law/bride’

22f. HM *ɛɔm (100, 173, 245)

H *ɛŋ : HM *ɛɔm : M *ɛɔm
1 2 3 4 5 6 7 8 9 10 11 HM/M
28. hot (water) khi1 ce1 ku1 koŋ1a kaŋA khjaŋ1 kɔ1̃ kɔm1 kjɔːm1 klœn1 tsam1 *k(h)lɛɔm
29. CLF-bowls
nɔm1 nɔ1 nɔ1 na1 *ʔnɛɔmA
/houses

22g. HM *ɛɔŋ (215)

H *ɛŋ : HM *ɛɔŋ : M *ɛɔŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
30. horn ki1 ce1 ku1 koŋ1a kaŋA kjaŋ1 kɔ̃1 kɔŋ1 kjɔːŋ1 klɔ1 kɔu1 *klɛɔŋ
31. to crawl ȵɔŋ1 ȵɔːŋ1 – – *ʔɲɛɔŋA

22h. HM *ɛuŋ (268)

H *ɛŋ : HM *ɛuŋ : M *ɛuŋ
1 2 3 4 5 6 7 8 9 10 11 HM
32. sky/heaven – – – – ɴqwaŋA ŋkuŋ2 ŋwɔ2 guŋ2 guːŋ2 – – *ɴɢɛuŋ
33. sky/heaven – – ntu2 ntoŋ2 – – – – – – – *ndɛuŋ
164 Chapter 2

Discussion
On the Hmongic side, this rime is similar to oral rime 11. The similarity points to a
common origin for Hmongic *nɛŋC (2.9/22) and Mienic *nu̯ aiC (2.9/11) ‘to ask’, and for
Hmongic *hnɛŋC (2.8/22) and Mienic *hnu̯ ɔiA (2.8/11) ‘sun/day’.

Notes
1. This is no doubt related to Mienic *qri̯ ɛkD ‘black’ (6.46/1).
3 & 29. These two classifiers are related: see discussion in chapter 6, section 6.2.1.
4. In Yanghao (1), ‘to allow’; in White Hmong, ‘to give’. Yanghao (1), Xuyong /po1/ ‘to
feed’, Shimen /pu1/ ‘to feed’ from Office of Miao-Yao Research Lexicon 1987.
7. Ailanthus altissima.
8. This is the Hmong-Mien first person singular pronoun with the widest distribution; see
chapter 3. Although it appears to be cognate to Dongshan Biao Min /kəu3/ ‘I/me’ (Wang
and Mao 21) and Sunjiang Biao Min /kɔ3/ ‘I/me’ (Wang and Mao 22), it is not possible to
place these forms into any existing Hmong-Mien rime set. The Proto Hmong-Mien form
must have had an initial *k- and a back rounded vowel, however, since the *k- did not
retract to q- in Hmongic; see chapter 2, section 1.2.
11 & 24. These two words for ‘short’ are related: see discussion in chapter 6, section 6.2.1.
17. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). Possibly
from Chinese 斟 ‘to ladle out’ (OC *t.[q][ə]m > MC tsyim > Man. zhēn).
20, 32, & 33. These three forms for ‘sky/heaven’ share the same rime, and have similar
onsets. Since no language shows more than one of the three, it appears they belong
together in a very irregular set.
23. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
28. Jiwei (2) from Office of Miao-Yao Research Lexicon (1987). Sunjiang Biao Min
(Wang and Mao 22) has been substituted for Dongshan Biao Min (10). Aspiration has been
tentatively reconstructed for Hmong-Mien because of the aspirated initials in Hmongic.
30. The forms for Zongdi (4) and Fuyuan (5) come from Wang 1994. These forms were
entered inconsistently in Wang and Mao 1995. Although the tones do not correspond, this
is probably the widespread word for ‘horn’, Chinese 角 (OC *k.rˤok > MC kæwk > Man.
jiǎo).
31. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
32. Tan Trinh Pa-Hng (Niederer 1997) has been substituted for Baiyun Pa-Hng (7).
33. Representing Mienic, see Changping Mien /ðuŋ2/ (Wang and Mao 18).
 Proto Hmong-Mien 165

23
23a. HM *euŋ (153)
H *ɨŋ : HM *euŋ : M *eu
1 2 3 4 5 6 7 8 9 10 11 HM
1. intestines – ɕe3 ȵ̥u3 ȵoŋ3b ȵ̥enB ȵ̥ɔ3 ŋ̥ŋ3 ȵ̥̥eu3 ȵɔu3 – – *hɲeuŋX
2. mother’s
nen4 ne4 nu4 noŋ4 ȵiB nɔ4 nɯ4 nau4 nau4 nau4 nuŋ4 *neuŋX
brother

23b. HM *iəŋ (269)

H *ɨŋ : HM *iəŋ : M *iəŋ
1 2 3 4 5 6 7 8 9 10 11 HM
3. flea m̥ hen1 – hm̥ ɔ1 mɔŋ1b m̥ enA mɔ1 n̩ 1 mɛŋ1 muŋ1’ m̥ au1 mɔŋ1 *hmiəŋ

Discussion
None of the words in set 23 is limited to Hmongic, thus none directly represents the
intermediate *ɨŋ. The final nasal of ‘water buffalo/cow’ was innovated (compare Proto Tai
*ŋwue ‘ox/cow’ and Proto Tibeto-Burman *ŋwa ‘cattle’) at the Hmong-Mien level
(Downer 1973).
Two other words in this set are Chinese, and each represents a unique pattern; both have a
front vowel followed by a back vowel which merged to the intermediate Hmongic rime
*ɨŋ.

23c. Wang and Mao rime 196: HM *ŋiuŋ ‘water buffalo/cow’ from Chinese 牛 ‘ox/cow’
(OC *[ŋ]ʷə > MC ngjuw > Man. niú)
23d. Wang and Mao rime 265: HM *kjeəm ‘gold’ from Chinese 金 ‘metal’ (OC
*[k](r)[ə]m > MC kim > Man. jīn)
166 Chapter 2

24
24a. H *aŋ
1 2 3 4 5 6 7 H
1. attic – – ntha1 ntua1b nʔthenA – – *nthaŋA
2. to float – nten1 nta1 – – – – *ntaŋA
3. spirit/ghost ȴ̥aŋ1 qwei1 da1 l̥ ua1a qlenA kli1 kwɛ1 *qraŋA
4. stomach – – pla1 – – – – *plaŋA
5. winnowing fan vaŋ1 wei1 va1 wua1a ʔwenA ven1 vɛ1 *ʔwaŋA
6. armspan ȴ̥aŋ2 ci2 da2 l̥ ua2 ʁleiA – – *ɢraŋA
7. to groan/ moan zaŋ2 mʐei2 ɳtʂa2 mpʐua2 mpenA mpwen2 – *mbraŋA
8. long sword – – nta2 ntua2 ntenA – – *ndaŋA
9. perilla (sū má) n̥ haŋ3 n̥ hen3 hn̥ a3 – n̥ enB – – *hnaŋB
10. mouse/rat naŋ4 nei4 na4 – – nen4 nĩ4 *naŋB
11. to breed/
ɕhaŋ5 – ɕa5 – – – – *ɕaŋC
propagate
12. to fly ʑaŋ5 ʑi5 ja5 ʑua5a ʔʑenC i5 je5 *ʔjaŋC
13. grave/tomb saŋ5 ntsei5 ntsa5 ntsua5a – – ȵe5 *ntsaŋC
14. wild dog m̥ haŋ5 – hm̥ a5 ma5b – – – *hmaŋC
15. maggot kaŋ5 – ka7 kua7a ʨenC – kɛ1-lõ7 *klaŋC

Loans in set 24a:

*N-kʷˤaŋ (OC) > hwang (MC) 黃 huáng ‘yellow’ > H *ɢʷaŋA ‘bright/light’
*N-kʷˤaŋ (OC) > hwang (MC) 黃 huáng ‘yellow’ > H *ɢʷaŋA ‘yellow’
*l̥ ˤaŋ (OC) > thang (MC) 湯 tāng ‘hot liquid’ > H *hlaŋA ‘to scald’41
trjang (MC) 張 zhāng ‘CLF-tools’ > H *traŋA
yang (MC) 煬 yáng ‘to melt’ > H *jaŋA ‘to dissolve/melt’
dzam (MC) 慚 cán ‘ashamed’ > H *dzaŋA
*[ɢ]ʷan (OC) > hjwon (MC) 園 yuán ‘garden’ > H *waŋA
nyem (MC) 髯 rán ‘whiskers’ > H *ɲaŋC ‘beard’

24b. HM *am (78, 97, 154)

H *aŋ : HM *am : M *am
1 2 3 4 5 6 7 8 9 10 11 HM/M
16. sister-in-law ȵaŋ1 ȵi1 ȵa1 ȵa1a ʔȵenA ȵi1 – – ȵaːm1 ȵan1 ȵɛm1 *ʔɲam
17. rib taŋ3 – ta3 tua3a tenB – – – taːm3 – – *tamX
18. blood ɕhaŋ3 ȵʨhi3 ɳtʂha3 ntsua3b nʔtshenB θi3 ȵ̥e3 ɕam3 saːm3 san3 dzjɛm3 *ntshjamX
19. half (of day) taŋ4 – ta4 – – – – daːm2 daːm2 – – *N-dam(X)
20. to spill ɕaːm5 jaːm6 ja6 – *(ʔ)jamC

Loans in set 24b:

*mə-tˤam (OC) > tam (MC) 擔 dān ‘to carry on shoulder’ > HM *ntam
*[k]ˤam (OC) > kam (MC) 甘 gān ‘sweet’ > HM *Kam
*Cə.[r]ˤam (OC) > lam (MC) 藍 lán ‘indigo’ > HM *ŋglam
Early Min /tsjäm/ ‘bland’42 > M *tsamB

41
The meaning of the Hmongic word is the same as a related Chinese word with the C tone now written 燙
‘to scald’ (Man. tàng).
42
Jíyùn dictionary: Norman 1988:232.
 Proto Hmong-Mien 167

24c. HM *aŋ (101, 120, 158)

H *aŋ : HM *aŋ : M *aŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
21. fragrant – – – – – nten1 nɛ1 daːŋ1 daːŋ1 daŋ1 dɔŋ1 *ntaŋ
22. snake naŋ1 nei1 na1 na1a ʔnenA nen1 nẽ1 naːŋ1 naːŋ1 naŋ1 nɔŋ1 *ʔnaŋ
23. sash/cord/rope l̥ haŋ1 – hl̥ a1 ɭua1b – – – l̥ aːŋ1 laːŋ1 ɬaŋ1 – *hljaŋ
24. neck – – da1 – – – – klaːŋ1 klaːŋ1 klaŋ1 kuŋ1 *qlaŋ
25. back (vs.front) qaŋ1 – qa1 hua1a – – – daːŋ1 daːŋ1 – kɔŋ1 *N-Kaŋ
26. to put on/wear naŋ4 n̥ hei3 hn̥ a3 na3b – nen4 n̥ ẽ3 – – – nɔŋ3 *(h)naŋX
27. eagle/hawk l̥ aŋ3 qwei3 da3 l̥ ua3a – klen3 kwɤ3 klaːŋ3 klaːŋ3 klaŋ3 kjaŋ3 *qlaŋX
28. firewood θaːŋ2 tθaːŋ2 tsaŋ2 hɔŋ2 *dzaŋA
29. boat/ship daːŋ3 daːŋ3 dzaŋ3 dɔŋ3 *ʔdzaŋB
30. in front daːŋ6 daːŋ6 – – *ndaŋC
31. word/ character – daŋ6 dzaŋ6 – *ndzaŋC

Loans in set 24c:

*[tsʰ]raŋ (OC) > tsrhjang (MC) 瘡 chuāng ‘sore (n.)’ > HM *tshaŋ ‘sore/boil/blister’
*maŋ-s (OC) > mjangH (MC) 望 wàng ‘to look at’ > HM *maŋH
*lraŋ (OC) > drjang (MC) 腸 cháng ‘intestines’ > M *ɢljaŋA
ʔjang (MC) 秧 yāng ‘seedling’ > M *ʔjaŋA

24d. HM *i̯ aŋ (121)

H *aŋ : HM *i̯ aŋ : M *i̯ əŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
32. flower paŋ2 pei2 pa2 pua2 venA pen2 pɛ2 pwaŋ2 faŋ2 pjaŋ2 pjaŋ2 *bi̯ aŋ
33. male (human) ʨaŋ6 – – ʨua6 – ɳtʃi6 – ʨaŋ2 – – kjaŋ2 *ɟi̯ aŋH
34. winnowing fan ɕaŋ1 saŋ1’ ȶaŋ1 – *ɕi̯ əŋA
35. to remember ʨjaŋ5’ saŋ5 ȶaŋ5 kɛŋ5 *kji̯ əŋC

Loans in set 24d:

*kʷˤaŋʔ (OC) > kwangX (MC) 廣 guǎng ‘wide’ > HM *Kʷi̯ aŋX
*N-kʷˤaŋ (OC) > hwang (MC) 黃 huáng ‘yellow’ > M *ʔgʷi̯ əŋA ‘bright’
*[ɢ]ʷˤaŋ (OC) > hwang (MC) 皇 huáng ‘yellow and white spotted’ > M *wi̯ əŋA ‘yellow’

24e. HM *æŋ (63)

H *aŋ : HM *æŋ : M *eŋ
1 2 3 4 5 6 7 8 9 10 11 HM
36. insect/worm kaŋ1 ci1 ka1 kua1a ʨenA kjen1 kɛ1 kɛŋ1 kjeːŋ1 klɛ1 tsaŋ1 *klæŋ
37. frog qaŋ3 – qa3 hua3a qenB ken3 kɤ7 – ȶeːŋ3 kwa3 – *KæŋX

24f. HM *i̯ æŋ (139)

H *aŋ : HM *i̯ æŋ : M *i̯ eŋ
1 2 3 4 5 6 7 8 9 10 11 HM
38. beeswax ʨaŋ1 ʨi1 cia1 ʨua1a ʨenA – – – – ʥjɛ1 – *N-ci̯ æŋ
168 Chapter 2

24g. HM *ɔn (45, 121)

H *aŋ : HM *ɔn : M *ɔn
1 2 3 4 5 6 7 8 9 10 11 HM
39. grain head/bag n̥ haŋ1 n̥ hei1 hn̥ a1 na1b n̥ enA n̥ en1’ n̥ ẽ1 – nɔ1 nan1 – *hnɔn

Notes
6. In White Hmong, this measure is for a double arms’ length (fingertip to fingertip).
9. Perilla is a medicinal herb in the mint family (also known as “shiso”).
10. This is probably the same as Mienic *nauB ‘mouse/rat’ (2.9/3).
12. Also ‘to run’ in Pa-Hng (Niederer 1997:120).
13. In Yanghao (1), ‘to store a body before burial’ (Wang 1994).
14. This word refers variously to ‘wolf’, ‘fox’, ‘jackal’, etc. Compare Proto Tai *hma1
‘dog’ (Li 1977).
15 & 34. Hmong-Mien ‘insect’ and Hmongic ‘maggot’ have the same root. The medial -l-
may be evident in the Pa-Hng (7) disyllabic form. ‘Maggot’ seems to have been derived
from ‘insect’ by C-tone derivation (see chapter 4). Compare Austronesian */qulej/
‘maggot’ and the following note by Blust: “Most problematic is the evidence from
Puyuma, Western Bukidnon Manobo and Tagabili that PAN */qulej/ meant not only
‘maggot’, but also ‘insect’” (ACD). Forms with tone 7 are secondary tones derived by tone
sandhi.
16. Also ‘daughter-in-law; bride’. Compare Tibeto-Burman *s-nam ‘daughter-in-law’
(Benedict 1987a:18).
18. Compare Mon-Khmer *jhaam ‘blood’ (Shorto #1430).
20. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). This
word has reflexes with both tone 5 and tone 6; it could equally well be reconstructed
*jamC.
21. Compare Kam /taaŋ1/ and Mak /taaŋ1/ ‘fragrant’ (Downer 1978).
23. The meaning ‘sash/cord/belt’ is associated with this word in Hmongic; the meaning
‘rope’ is associated with this word in Mienic. Hmongic forms from Wang 1994. Possibly
from Chinese 繩 ‘string, cord’ (OC *m.ləŋ > MC zying > Man. shéng).
24. Compare Chinese 頸 ‘neck’ (OC *keŋʔ > MC kjiengX > Man. jǐng). Wang and Mao
(1995) give two different forms for Zao Min ‘neck’, and note that both are irregular in
rime: /kan1/ (341) and /kuŋ1/ (500). It is not clear if these are variants, or if one form was
entered in error. /kuŋ1/ is arbitrarily entered here.
25. See also Mien (Wang and Mao 16) /gaŋ1/.
26. Since tone 3 and tone 4 are equally well-attested, the initial h- is placed in parentheses.
27. Compare Mon-Khmer *klaŋ ‘hawk, large raptor’ (Shorto #714), Tibeto-Burman
*(g)laŋ ‘eagle’ (Matisoff 2003).
28. Downer (1973:21) says this is a loanword from Chinese 柴 ‘firewood’ (MC dzrɛ >
Man. chái), one of a number that show a relationship between final -i or -u in Chinese and
final -ŋ in Mien.
31. This is probably from Chinese 字 ‘word, character’ (OC *Cə-[dz]ə-s > MC dziH >
Man. zì), despite the difference in rime. Downer (1973:21) cites this as one of a number of
examples where Middle Chinese -i or -u corresponds to Mienic -ŋ.
 Proto Hmong-Mien 169

32. The -w- in Luoxiang Mien (8) is secondary: -j- dropped, then -w- developed between a
labial and /a/. Compare Austronesian *buŋa ‘blossom of fructifying tree or bush’ (ACD),
Written Burmese /pwɑŋ1/ ‘blossom’, and Laqua (Tai-Kadai) /puŋ/ (Benedict 1975:295).
33. In White Hmong, the cognate may be /ca6/ in the phrase ob cag /ʔɔ1 ca6/ ‘others’.
34. Probably from Chinese 扇 ‘fan’ (MC syenH > Man. shàn).
37. Compare Proto Austronesian *keŋkeŋ ‘frog; hollow sound’ (ACD).
39. Chinese 囊 ‘bag’ (OC *nˤaŋ > MC nang > Man. náng) may have been borrowed from
Hmong-Mien, since it shows only the derivative meaning (see chapter 7). Yet compare
Tibeto-Burman *s-nam ‘ear of grain’ (Matisoff 2003).
170 Chapter 2

25
25a. HM *ein (26, 36, 170)
H *ein : HM *ein : M *ei
1 2 3 4 5 6 7 8 9 10 11 HM
1. vine – ɕi1 hm̥ a1 ma1b m̥ oŋA m̥ e1’ m̥ a1 m̥ ei1 mei1’ m̥ əi1 mɛi1 *hmein
2. this noŋ3 nei3 nɔ5 na3a ʔnoŋB ne3 n̩ 3 nei3 nei3 na3 na3 *ʔneinX
3. heavy ȵ̥hoŋ3 hei3 hȵ̥a3 ȵa3b ȵ̥oŋB ȵ̥e3 ȵ̥a3 ȵ̥i3 ni3 n̥ i3 nɛi3 *hnjeinX

Discussion
All three Hmong-Mien words in set 25 begin with a nasal consonant. This could suggest
the nasal in the coda of some Hmongic languages is secondary, given the absence of word-
final nasal consonants in Mienic for these three words. However, three facts argue against
this analysis: (1) It is just as common in this family for nasal-initial words to lose final
nasalization through dissimilation (see ‘to sniff’, for example 1.23/18); (2) If the Mienic
forms reflect the protolanguage rime and these words originally had syllable-final -i, they
should have yielded a rime in Proto Hmongic with a rounded onset (see section 2.1.3); and
(3) The Chinese word for ‘creeping plant’ 蔓 (OC *ma[n]-s > MC mjonH > Man. wàn),
which may be the same as this word for ‘vine’, ends in an -n. Therefore, the analysis
chosen here is that *-n in Mienic dropped in a heavy syllable after a nasal onset (*Nein >
Nei).
 Proto Hmong-Mien 171

26
All of the words in this set are Chinese. They are placed together because they merged into
one rime in Hmongic, here reconstructed as *i̯ aŋ.

26a. Wang and Mao rime 161, Wang 1994 rime 26:
H *ɲci̯ aŋA ‘straight’ from Chinese 正 (MC tsyengH > Man. zhèng)
H *ji̯ aŋA ‘poplar tree’ from Chinese 楊 (MC yang > Man. yáng)
H *wi̯ aŋC ‘ten thousand’ from Chinese 萬 (MC mjonH > Man. wàn)

26b. Wang and Mao rime 65:

HM *dzi̯ ɔŋH ‘craftsman’ from Chinese 匠 (MC dzjangH > Man. jiàng)
M *ɣji̯ ɔŋB ‘elephant’ from Chinese 象 (MC zjangH > Man. xiàng)

26c. Wang and Mao rime 161:

H *li̯ aŋB and M *lji̯ oŋB ‘tael (40 grams)’ from Chinese 兩 ‘tael’ (MC ljangX >
Man. liǎng)
172 Chapter 2

27
27a. H *uŋ
1 2 3 4 5 6 7 H
1. to eat naŋ2 noŋ2 nɔ2 nɔ2 noŋA nɐŋ2 nɔ2 *nuŋA
2. to forget n̥ hoŋ1 noŋ3 hn̥ ɔ3 na3a n̥ oŋA n̥ ɐŋ3 nõ3 *hnuŋB
3. wheat maŋ4 – mɔ6 məŋ6 muC – – *muŋC

27b. M *um (136)

1 2 3 4 5 6 7 8 9 10 11 M
4. to winnow ȵum5 ȵum5’ n̥ ən5 num5 *hɲumC

Loan in set 27b:

nyemX (MC) 染 rǎn ‘to dye (cloth)’ > M *ɲumC 43

27c. HM *un (195, 231)

H *uŋ : HM *un : M *un
1 2 3 4 5 6 7 8 9 10 11 HM/M
1 1 1 1a A 1 1
5. illness/pain moŋ moŋ mɔ məŋ ʔmoŋ mɐŋ mæ̃ mun1 mun1 mən1 man1 *ʔmun
6. round – – – – – – ka1˗lɔ̃4 klun2 klun2 klin2 – *ɢlun
7. to stroke phjun1 – phlən1 – *phlunA
8. to water dzun2 – – – *ɲɟunA

Loans in set 27c:

gjun (MC) 裙 qún ‘skirt’ > M *gjunA
hjwon (MC) 園 yuán ‘garden’ > M *hwunA

27d. HM *u̯ ɔm (98)

H *uŋ : HM *u̯ ɔm : M *əm
1 2 3 4 5 6 7 8 9 10 11 HM
9. cold – noŋ5 nɔ5 nɔ5a – nɐŋ5 ȵɔ̃5 nam5’ nam5 nan5 nam5 *ʔnu̯ ɔmH

27e. M *ən (62)

1 2 3 4 5 6 7 8 9 10 11 M
10. to eat ȵen6 ŋjin6 ȵin6 ȵan6 *ɲənC

Loan in set 27e:

kjinX (MC) 緊 jǐn ‘tight’ > M *kjənB

Notes
1 & 10. Rime 27e is provisionally placed here on the possibility that the Hmongic and
Mienic words for ‘to eat’ are the same.

43
This may be a borrowing through a neighboring language. Compare Lao /njɔm/, Mak /njum/, and
Vietnamese nhuộm.
 Proto Hmong-Mien 173

2. Compare Mienic *hɲouB ‘to forget’ (4.8/13).

6. Compare Chinese 輪 ‘wheel; round’ (MC lwin > Man. lún); see chapter 4 on the
significance of Pa-Hng disyllabic form for interpreting Hmong-Mien clusters. Yanghao
/l̥ en2/ and Xuyong /khun2/ ‘round’ may be related, but the rimes do not correspond (forms
from Office of Miao-Yao Research Lexicon 1987).
7. Mien (8) from Office of Miao-Yao Research Lexicon 1987 and Biao Min (10) from
Solnit 1982. Compare Hmongic *phlʉC ‘to stroke’ (1.32/8) and Chinese 撫 ‘to caress’ (OC
*[pʰ](r)aʔ > MC phjuX > Man. fǔ).
8. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
9. In Mienic, the meaning is ‘cool (water)’.
174 Chapter 2

28
28a. H *oŋ
1 2 3 4 5 6 7 H
1. ditch koŋ1 – – kaŋ1a koŋA kjɐŋ1 – *koŋA
2. steep shoŋ1 ntshoŋ1 ntsha1 ntsaŋ1b nʔtshoŋA ɳtʃɐŋ3 – *ntshoŋA
3. to lead – ʨoŋ1 ca1 ʨaŋ1a ʨoŋA – – *coŋA
4. trough ȴ̥oŋ1 coŋ1 da1 l̥ aŋ1a qloŋA kjɐŋ1 ljõ1 *qroŋA
5. root ʨoŋ2 ʨoŋ2 ca6 ʨaŋ6 ʑoŋA – kõ2 *ɟoŋA
6. bellows thoŋ3 thɑŋ3 – thaŋ3b thoŋB – tho3 *thoŋB
7. CLF-rooms ʨhoŋ3 – cha3 ʨaŋ3b khoŋB – – *choŋB
8. neck qoŋ3 qoŋ3 – haŋ3 – kɐŋ3 – *qoŋB
9. plain (level
– – ɳʈa4 – ɳtʂoŋB – – *ndroŋB
ground)
10. rain noŋ6 noŋ6 na6 naŋ6 noŋC nɐŋ6 mõ6 *m-noŋC
11. to crawl ȵoŋ6 – ŋka6 ȵʨaŋ6 – – – *ŋgjoŋC
12. owl l̥ ioŋ5 – pla7 – – – – *ploŋD

Loans in set 28a:

duwng (MC) 筒 tóng ‘tube’ > H *droŋA ‘flute/tube’
luwng (MC) 聾 lóng ‘deaf’ > H *loŋA
pjuwngH (MC) 風 fèng ‘to blow’ > H *poŋC ‘air’

28b. HM *oŋ (178, 216)

H *oŋ : HM *oŋ : M *oŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
13. to stretch ɕhoŋ1 – ɕa1 – – – ɕõ1 ɕoŋ1 – – – *ɕoŋ
14. to ask for noŋ6 nɔŋ6 nɔŋ6 – *[n/l]oŋC

Loans in set 28b:

*truŋ (OC) > trjuwng (MC) 中 zhōng ‘center/middle’ > HM *ntroŋ
*mə-roŋ (OC) > ljowng (MC) 龍 lóng ‘dragon’ > HM *-roŋ
*mə-loŋ-s (OC) > yowngH (MC) 用 yòng ‘to use’ > M *[n/l]oŋC

28c. HM *uŋ (179, 180, 249, 250)

H *oŋ : HM *uŋ : M *uŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
15. earth-worm ʨoŋ1 coŋ1 ca1 ʨaŋ1a ʨoŋA jɐŋ1 jõ1 duŋ1 – hwjə1 dziŋ1 *N-cuŋ
16. bone shoŋ3 soŋ3 tsha5 saŋ3b tshoŋB θɐŋ3 sõ3 θuŋ3 tθuŋ3 swə3 hiŋ3 *tshuŋX
17. seven ɕoŋ6 ʨoŋ6 ɕa5 ʨaŋ6 zoŋC ʃɐŋ6 tɕɤ6 (sje6) – – – *djuŋH
18. root duŋ2 duːŋ2 – – *ndzuŋA
19. pig tuŋ4 tuŋ4 twə4 tiŋ4 *duŋB
20. rain bluŋ6 buŋ6 blə6 biŋ6 *mbluŋC
 Proto Hmong-Mien 175

Loans in set 28c:

yang (MC) 羊 yáng ‘sheep/goat’ > HM *juŋ
suwngH (MC) 送 sòng ‘to send/deliver’ > HM *suŋH
tsyang (MC) 獐 zhāng ‘fallow-deer’ > M *cuŋA ‘deer’
trjang (MC) 張 zhāng ‘CLF-tools’ > M *truŋA
yang (MC) 煬 yáng ‘to melt’ > M *juŋA ‘to dissolve/melt’
ljowng (MC) 龍 lóng ‘dragon’ > M *kljuŋA
sræwng (MC) 雙 shuāng ‘pair’ > M *suŋA
yangX (MC) 養 yǎng ‘to nourish, raise’ > M *juŋB

Discussion
It is interesting that Hmong-Mien *uŋ patterns differently from Hmong-Mien *um and
*un: *uŋ merged with what is reconstructed here as Hmongic *oŋ, whereas *um and *un
merged with what is reconstructed here as Hmongic *uŋ (rime 27 above).

Notes
1. Compare Biao Min /khlɛ1/ ‘ditch, ravine, mountain stream’ (Solnit 1982). Zao Min (11)
/ku1/ appears to have been borrowed from Chinese 溝 ‘ditch, irrigation canal’ (MC kuw >
Man. gōu).
4. Wenjie Pa-Hng (Wang and Mao 12) has been substituted for Baiyun Pa-Hng (7).
5 & 16. These two words for ‘root’ may be related.
6. This word bears a slight resemblance to Chinese 橐 ‘sack, bellows’ (MC thak > Man.
tuó).
7. Possibly from Chinese (房) 間 ‘room’ (MC kɛŋ > Man. (fáng)jiān), although the rime
does not correspond.
8. Compare Chinese 頸 ‘neck’ (OC *keŋʔ > MC kjiengX > Man. jǐng).
10 & 20. Although the initials do not match, Hmongic and Mienic ‘rain’ may well belong
together, since they both belong to this rime set.
12. Yanghao (1) from Office of Miao-Yao Research Lexicon 1987.
14. This word alternates l- and n- onsets, like the Chinese loanword ‘to use’ in this rime
set. It is not possible to reconstruct an onset.
17. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8). The
initial seems to tie the Mien form to the Hmongic set, but the rime obviously does not
agree. Mienic *ŋjiC (5.24/1) may be related, but the correspondence is difficult. Hmong-
Mien ‘seven’ is from Tibeto-Burman *s-nis ‘seven’ according to Benedict (1987a:13); see
chapter 5.
19. Perhaps from Chinese 豚 ‘young pig’ (OC *[d]ˤu[n] > MC dwon > Man. tún), although
the tones and final nasals do not correspond.
20. Compare Khmer /phliəŋ/ ‘rain, to rain’.
176 Chapter 2

29
29a. H *ɔŋ
1 2 3 4 5 6 7 H
1. crops qə1 – qɔŋ1 haŋ1a – – – *qɔŋA
2. needle ʨu1 ʨu1 kɔŋ1 kaŋ1a koŋA tʃwaŋ1 kɔ1̃ *kjɔŋA
3. puttees ȶhu1 ɳʈhu1 ɳʈhɔŋ1 ɳʈaŋ1b ɳʔtʂhoŋA ɳtʃwaŋ1’ ȵɔ
̥ ̃
1
*nthrɔŋA
4. thin – ɕu1 ʂɔŋ1 saŋ1b soŋA θjɔŋ1 – *sjɔŋA
5. thrush ʨu1 ʨu1 cɔŋ1 ʨaŋ1a ʨoŋA tʃwaŋ1 ʨɔ1̃ *cɔŋA
6. crazy – – – mpjaŋ2 mpzoŋA – – *mbljɔŋA
7. leaf nə2 nu2 mplɔŋ2 mplaŋ2 mploŋA mplɔŋ2 mjɔ2̃ *mblɔŋA
8. table/bench – – ʈɔŋ2 ʈaŋ2 ʐoŋA – tɕe2 *drɔŋA
9. forest ɣu3 ʐu3 ʐɔŋ3 – ʔwjoŋB ŋkjɔŋ3 jɔ3̃ *-ʔrɔŋB
10. mosquito ʑu3 – jɔŋ3 ʑaŋ3a ʔʑoŋB – jõ3 *ʔjɔŋB
11. to winnow – pʐu3 tʂɔŋ3 pʐaŋ3b pjoŋB pjɔŋ3 – *prɔŋB
12. to finish ʨu4 ʨu4 – ʨəŋ4 – – – *ɟɔŋB
13. cloud/haze en5 – ɔŋ5 aŋ5a – – – *ʔɔŋC
14. year – ʨu5 ɕɔŋ5 ʨaŋ5a – – ʨɔ5̃ *ɕɔŋC
15. thick (liquid) ȵu6 – ŋkɔŋ6 ŋkəŋ6 ŋkoŋC – ŋa8 *ŋgjɔŋC

29b. M *ɔm (192)

1 2 3 4 5 6 7 8 9 10 11 M
16. leaf nɔm2 nɔːm2 nan2 num2 *nɔmA
17. asleep gɔm2 – – – *ŋgjɔmA

Loan in set 29b:

*-[g]ˤ[ə]m (OC) > hom (MC) 含 hán ‘to hold in mouth’ > M *ʔgjɔmA

29c. HM *ɔp (219, 254)

H *ɔŋ : HM *ɔp : M *ɔp
1 2 3 4 5 6 7 8 9 10 11 HM
18. to cough ŋo4 – hn̥ ɔŋ7 naŋ7 n̥ oD – – n̥ op7 – n̥ ən7 – *hnɔp

Loan in set 29c:

kop (MC) 鴿 gē ‘pigeon’ > M *kɔpD

29d. HM *ɔŋ (142)

H *ɔŋ : HM *ɔŋ : M *ɔŋ
1 2 3 4 5 6 7 8 9 10 11 HM
19. good ɣu5 ʐu5 ʐɔŋ5 ʐaŋ5a ʔwjoŋC ŋwaŋ5 jɔ̃5 gwəŋ5’ gɔŋ5 lɔŋ5 dzɔŋ5 *-ʔrɔŋH
20. to put on/
tə5 ntu5 ntɔŋ5 ntaŋ5a nʔtoŋC ntwaŋ5 no5 doŋ5’ dɔŋ5 dɔŋ5 dɔŋ5 *ntɔŋH
wear (cap)
 Proto Hmong-Mien 177

Loans in set 29d:

*-lˤoŋ (OC) > duwng (MC) 銅 tóng ‘copper’ > HM *dɔŋ44
thuwngX (MC) 桶 tǒng ‘bucket’ > HM *thɔŋ(X)
tsuwng (MC) 騣 zōng ‘mane’ > M *ʔdzɔŋA 45
nowng (MC) 膿 nóng ‘pus’ > M *nɔŋC

29e. HM *u̯ ɔŋ (85)

H *ɔŋ : HM *u̯ ɔŋ : M *u̯ əŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
21. to listen – – mlɔŋ6 mʐəŋ6 – – mæ̃ 5 moŋ5 muŋ5’ m̥ ɔŋ5 maŋ5 *(S-)mru̯ ɔŋH
22. carrying pole moŋ5 muŋ4 mɔŋ4 – *mu̯ əŋB

29f. HM *ɔk (205)

H *ɔŋ : HM *ɔk : M *ɔk
1 2 3 4 5 6 7 8 9 10 11 HM/M
23. bird nə6 nu6 nɔŋ6 nəŋ6 noC nwaŋ6 mõ6 nu8 nɔ6 n̥ ɔ8 nɔu8 *m-nɔk
24. bamboo bowl – – – – *khɔkD

Loan in set 29f:

tsræwk (MC) 捉 zhuō ‘to take’ > M *tjɔkD

29g. HM *u̯ əm (212, 246)

H *ɔŋ : HM *u̯ əm : M *u̯ əm
1 2 3 4 5 6 7 8 9 10 11 HM/M
1 1 1a A 1
25. water ə u – aŋ oŋ waŋ ʔɔ̃ wəm1
1
wɔm1 ən1 m̩ 3 *ʔu̯ əm
26. pillow – ȵʨɯ5 ɲcɔŋ5 ȵʨaŋ5a ȵʔʨoŋC ntjoŋ5 ȵɔ5 dʑum6 ȡam5 – – *ɲcu̯ əmH
27. to play (flute) pwəm3 pjɔm3 – – *phlu̯ əmB

29h. HM *u̯ əŋ (102)

H *ɔŋ : HM *u̯ əŋ : M *əŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
28. tree tə5 ntu5 ntɔŋ5 – nʔtoŋC ntwaŋ5 – djaŋ5’ gjaŋ5 djaŋ5 djaŋ5 *ntju̯ əŋH
29. year ȵ̥hu5 – – – – ȵ̥waŋ5 – ȵaŋ5 ȵaŋ5’ ȵ̥aŋ5 ȵaŋ5 *hɲu̯ əŋH
30. to walk jaŋ2 jaŋ2 ȵaŋ2 dzaŋ2 *jəŋA
31. wife of m.’s
mwaŋ2 ŋwaŋ2 mjaŋ2 mjaŋ2 *mjəŋA
brother

Loan in set 29h:

tsyhingH (MC) 秤 chèng ‘balance’ > HM *nthju̯ əŋH

44
Laurent Sagart points out that in the Hmongic language Bunu, ‘copper’ has the initial l- of Old Chinese
(/loŋ2/, see Mao, Meng, and Zheng 1982:192), which suggests that this word was borrowed more than
once.
45
Chinese ‘mane’ was borrowed independently by Hmongic: *tsʉŋA (rime 30a).
178 Chapter 2

29i. HM *u̯ em (60)

H *ɔŋ : HM *u̯ em : M *em
1 2 3 4 5 6 7 8 9 10 11 HM
32. seed ȵ̥hu1 ɳu1 nɔŋ1 ɳaŋ1a n̥ oA nwaŋ1 ȵõ1 ȵem1 ŋjim1’ n̥ jɛn1 num1 *hnru̯ em
33. raw/unripe ȵu4 ȵu4 ȵɔŋ4 – ȵoB ȵwaŋ4 ȵõ4 – ŋjim4 – – *ɲu̯ emX

29j. HM *u̯ eŋ (46)

H *ɔŋ : HM *u̯ eŋ : M *eŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
34. door/gate ȶu2 ʈu2 ʈɔŋ2 ʂaŋ2 ʐoŋA ʃwaŋ2 tɕɔ2̃ kɛŋ2 keːŋ2 klɛ2 – *gru̯ eŋ
35. green/blue mɛŋ1 meːŋ1 mɛ1 mɛŋ1 *ʔmeŋA
36. rock bɛŋ5’ – bɛ5 bɛŋ5 *ʔbeŋC

Loan in set 29j:

*breŋ (OC) > bjæng (MC) 平 píng ‘even’ > M *beŋA ‘level’

Discussion
There is only one loanword to exemplify another set 29 rime:
29k. HM *u̯ ən (248): HM *mphu̯ ənH ‘to scatter/sprinkle’ from Chinese 噴 ‘to spatter,
sprinkle’ (OC *pʰˤur-s > MC phwonH > Man. pèn).
A final rime fits best in this set as well, although this is a tentative placement, since these
Chinese loanwords to Mienic have no Hmongic counterparts:
29l. M *u̯ am (211): M *su̯ amA ‘beard’ from Chinese 髟 ‘long hair’ (MC sræm > Man.
shān) and M *ju̯ amA ‘armspan’ from Hakka /njam/ (Downer 1973:19).

Notes
2. Compare Chinese 箴 ‘needle’ (OC *t.q[ə]m > MC tsyim > Man. zhēn).
3. Tan Trinh Pa-Hng (Niederer 1997) has been substituted for Baiyun Pa-Hng (7).
7 & 16. Although the initials do not match, these two sets for ‘leaf’ may well belong
together, since they both belong to this rime and have the same tone. The two words in
rime 29b are thus included here, even though they appear to be attested in Mienic only.
Compare Proto Malayo-Polynesian *buluŋ ‘medicinal herbs’ which has come to be used as
‘leaf’ in many Austronesian languages (ACD).
9. Jiongnai (6) from Mao and Li 2001.
14 & 29. These may be related to Chinese 年 ‘year’ (OC *[n]ˤi[n] > MC nen > Man. nián),
but as suggested by Mortensen (2002), a Tibeto-Burman source (TB *s-niŋ ‘year’,
Matisoff 2003) better explains the final velar nasal.
17. Also Hainan Mun /gjoːm2/.
18. Jiangdi Mien (Wang and Mao 15) has been substituted for Luoxiang Mien (8).
21. Biao Min (10) from Office of Miao-Yao Research Lexicon 1987.
23. The final -k of ‘bird’ is reflected in a velar nasal in Hmongic (the shift to a final nasal
conditioned by the nasal initial). Compare Proto Malayo-Polynesian *manuk
‘bird/chicken’ (ACD) and Proto Tai *n-lok ‘bird’ (Li 1977).
24. Not attested in these four Mienic languages, but Jiangdi Mien and Xiangnan Mien
/khɔ7/ correspond.
 Proto Hmong-Mien 179

26. Perhaps from a C-tone variant of Chinese 枕 ‘pillow’ (OC *[t.q][ə]mʔ > MC tsyimX >
Man. zhěn).
27. Changping Mien (Wang and Mao 18) /phləm3/. Compare the similar Hmongic *phroB
‘to play (flute)’ (1.47/7).
28. Perhaps the same as Chinese 樹 ‘tree’ (OC *m-toʔ-s > MC dzyuH > Man. shù), but
compare Mon-Khmer *t2ʔɔɔŋ ‘tree’ (Shorto #491).
31. The -w- in Luoxiang Mien (8) and Lanjin Mun (9) is secondary: -j- dropped, then -w-
developed between a labial and /a/.
32. Downer (1973) noted the similarity between this word and Cantonese /ŋan/ ‘seed’.
32 & 33. The rimes of ‘seed’ and ‘raw/unripe’ differ strikingly in Jiangdi Mien and
Xiangnan Mien, not represented here (Jiangdi /ȵiːm1/ and /ȵom4/; Xiangnan /ȵiŋ1/ and
/ȵəŋ4/). Without a better alternative, the two words are put together here.
35. Hmongic *mpruA (1.49/16) ‘green/blue’ may be related.
180 Chapter 2

30
30a. H *ʉŋ
1 2 3 4 5 6 7 H
1. to fall – – pɔŋ1 paŋ1a poŋA – – *pʉŋA
2. Hmong m̥ hu1 ɕoŋ1 hm̥ ɔŋ1 maŋ1b m̥ joA – ŋ̥ŋ1 *hmʉŋA

Loans in set 30a:

*[ɢ]ʷˤaŋ (OC) > hwang (MC) 蝗 huáng ‘locust’ > H *gʉŋA ‘grasshopper’
tsuwng (MC) 騣 zōng ‘mane’ > H *tsʉŋA 46

30b. HM *ɨŋ (9)

H *ʉŋ : HM *ʉŋ : M *ɨŋ
1 2 3 4 5 6 7 8 9 10 11 HM/M
3. to go moŋ4 moŋ4 mu4 məŋ4 muB ŋ̰ŋ4 nɪ2̃ miŋ2 niŋ2 – mi2 *n-mʉŋ(X)
4. to lift niŋ5 niŋ5’ – – *hnɨŋC

Notes
1. Possibly from Chinese 崩 ‘to fall, collapse (of mountain)’ (OC *Cə.pˤəŋ > MC pong >
Man. bēng).
2. ‘Mien’ and its variants is probably from Chinese 民 ‘people’ (OC *mi[n] > MC mjin >
Man. mín).
3. Note that the tones do not correspond across the Hmongic-Mienic divide (see chapter 2,
section 1.4).
4. Despite the tone mismatch, this is probably the same word as Chinese 拎 ‘to carry, lift’
(Cant. /nɪŋ1/, Man. līng).

46
Chinese ‘mane’ was borrowed independently by Mienic: *ʔdzɔŋA (rime 29d).
 Proto Hmong-Mien 181

Additional Mienic sets

It is not possible to place the following three sets of loanwords into one of the rimes above,
since there is no Hmongic evidence to show the direction of the merger, nor do they
pattern exactly like any other sets in Mienic.

A. M *iu (3)
Loans in set A:
kjuwX (MC) 韭 jiǔ ‘Allium’ > M *ɟiuB
tsjuwX (MC) 酒 jiǔ ‘liquor’ > M *tiuB
mjewH (MC) 廟 miào ‘temple’ > M *miuC

B. M *u̯ ok (258)
Loans in set B:
tsyuwk (MC) 粥 zhōu ‘porridge’ > M *tju̯ okD
dzyuwk (MC) 熟 shú ‘cooked/ripe’ > M *dju̯ okD

C. M *ien (6)
Loans in set C:
mjienH (MC) 面 miàn ‘face’ > M *hmienA
khen (MC) 牽 qiān ‘to lead’ > M *khienA
3 Tonogenesis and tone development

3.1 Introduction
The earliest in-depth work on the Hmong-Mien language family involved the
reconstruction of the tones by Kun Chang in a series of articles published over several
decades (Chang 1947, 1953, 1966, 1972). This important work is summarized in Niederer
1998 (216–250). For the most part, Hmong-Mien tonal categories correspond to those of
Chinese, Vietnamese, and Tai, the languages of the ‘Sinosphere’47—Chinese and those
languages in close contact with and profoundly influenced by Chinese—and the similarity
of their tonal systems makes it appear that these languages all developed tones in the same
way. We therefore assume that tonogenesis in Hmong-Mien followed a series of
developments first elucidated by Haudricourt (1954b) to explain tonogenesis in
Vietnamese. By this account, tonal contrasts first arose upon the loss of syllable-final
laryngeal consonants *-ʔ and *-h (< *-s), yielding a three-way tonal contrast:

A (*ø) B (*ʔ) C (*h)

Haudricourt (see also Matisoff 1973) has proposed that these three tones were level, rising,
and falling respectively.48 Syllables with final voiceless stops *-p, *-t, *-k developed their
own tonal characteristics at a later date, yielding a fourth major tone category, D.
These original tones subsequently doubled to eight following the merger of syllable-
initial voiced and voiceless obstruents (preglottalized and voiceless sonorants patterned
with the voiceless initials), as represented below:49

1 (voiceless initials) A1 (1) B1 (3) C1 (5) D1 (7)

2 (voiced initials) A2 (2) B2 (4) C2 (6) D2 (8)

It is generally accepted that the ‘upper register’ tones (A1, B1, C1, and D1 from syllables
with voiceless initials) were initially higher in pitch, while the ‘lower register’ tones (A2,
B2, C2, and D2 from syllables with voiced initials) were initially lower in pitch. The
physiological connection between initial consonant voicing and the development of pitch
contrasts has been widely discussed, although there have been problems with finding a

47
“Sinosphere” is a term coined by James A. Matisoff (Matisoff 1990), as is the invaluable term
“tonogenesis” itself (Matisoff 1970).
48
On the other hand, Benedict believed that the original contrast was between a falling tone (A) and a rising
tone (B), while C was a level sandhi tone, added later (see, for example, Benedict 1997).
49
The most common way of referring to these historical categories in the literature on Asian tone
languages, and in the present work, is by the numbers 1–8, as indicated.
183
184 Chapter 3

direct causal link between the two: see for example Hombert 1978, Hombert, Ohala and
Ewan 1979, Maddieson 1984a and section 3.3 below.
In a few languages of the family the obstruent merger and the tone split did not both
occur: Fuyuan (West Hmongic) has only four ‘unsplit’ tones, while Shimen (West
Hmongic) has split the tones on the basis of initial consonant voicing, but did not lose its
voicing contrasts. Subsequent tonal splits are local developments, and did not necessarily
affect all tone categories. Most common among these is a secondary tone split based on
aspiration, as in the West Hmongic language Zongdi and the Mun of Hainan Island (for
more examples, see Niederer 1998:246–247). Prenasalization conditioned another type of
secondary tone split in North Hmongic (Niederer 1998:247–248). Almost all languages
exhibit tone mergers of two or more of the original eight categories: for example, the
reflexes of categories B2 and D1 have merged in White Hmong.
For those who are unused to thinking in terms of historical tone categories (A1, A2, B1,
B2, etc.) as opposed to phonetic tone values (high level, low rising, etc.), it is useful to
think of them this way: all the words in a particular tone category have a common
historical origin in terms of initial and final consonantism (A1 < *voiceless initial, open
syllable or syllable with a nasal coda). This ensures that when the original consonantism is
transphonologized into tone, all of the words belonging to each original category as
defined by syllable type will continue to pattern together tonally. Although phonetic
studies have shown that the newly emergent tones will have certain properties due directly
to the type of consonant lost, once tones are created, they morph quite quickly into other
things: originally high tones may lower, low tones may raise, tones may merge, contours
may simplify, etc. Therefore, across languages in a family, words that belong to a
particular tone category may have quite different phonetic realizations. For example,
Hmong-Mien words in the A1 category have a variety of different phonetic values: they
may be mid rising, high level, low rising, mid falling, or mid level. This cross-linguistic
variability is true of every tone category. The categories themselves, on the other hand, are
remarkably stable: in all Hmong-Mien languages, the members of the group of cognates
which includes ‘to give’, ‘deep’, ‘three’, ‘thatch grass’ and ‘snake’ will all have the same
tone in each language of the family (the A1 reflex), regardless of the phonetic value of that
tone in any particular language. The stability of tonal categories explains why, of all
aspects of Hmong-Mien historical phonology, tone was the first to be reconstructed.
Although the accepted account of Asian tonogenesis sketched above explains the
general structure of the Hmong-Mien tone system, many difficult and interesting topics
with regard to tonogenesis remain to be addressed. Two such topics make up the substance
of this chapter. In section 3.2, the question of the timing of tonogenesis in Chinese and
Hmong-Mien is raised through an examination of Chinese loanwords in Hmong-Mien.
Since these loanwords show perfect tone category correspondence across the two families,
it is argued that the loans were made at a time when neither donor nor borrower was tonal.
In section 3.3, a detailed, family-internal account of the development of tones in Hmong-
Mien is presented. On the basis of tone reflex patterns and the characterization of certain
modern-day tones by non-modal phonation, it is proposed that two overlapping layers of
phonation type contrasts preceded, and led to, the development of full tonal contrasts in
Hmong-Mien languages.
 Tonogenesis and tone development 185

3.2 Relative timing of Chinese and Hmong-Mien tonogenesis50

The classic account of the segmental origin of tonal contrasts works well for Hmong-Mien:
there is good indirect evidence for the presence of final laryngeals that led to the earliest
tonal contrasts (see section 3.3 below), and ample comparative evidence for the subsequent
tone split. There is a problem, however, in understanding how this Sinospheric system of
tonal contrasts might have been propagated across unrelated language groups. It has been
claimed, explicitly by Paul Benedict, but implicitly by others, that this whole system of
tonal contrasts (4 tones times 2 registers) was borrowed from Chinese by Hmong-Mien,
Tai and Vietnamese, all of which were originally atonal. Benedict writes ‘…Vietnamese,
under direct Chinese domination lost the…initial syllables of MK [Mon-Khmer] while
directly borrowing the tonal system…’ (Benedict 1997:4, emphasis added).
The perfect tone category correspondences in Chinese loanwords from the Early Middle
Chinese period is also somehow attributed to the fact that Hmong-Mien borrowed its four-
by-two tonal system from Chinese. For example, Lin Ying (1972:56), in an article on
Chinese loans in Hmong-Mien, writes “If we compare these loans with Qièyùn rhyme
tables, we find that the tones have been borrowed by Miao primarily on the basis of the
píng, shǎng, qù, and rù tone categories”51 (emphasis added). Following are examples of
Middle Chinese loanwords in Hmong-Mien which have the same tone category in both the
donor and the borrower:

(1) MC PHM Tone category in both

金 jīn ‘metal’ kim *kjeəm ‘gold’ A1
秧 yāng ‘seedling’ ʔjang A
*ʔjɛŋ (PH) A1
千 qiān ‘thousand’ tshen *tshi̯en A1
銅 tóng ‘copper’ duwng *dɔŋ A2
羊 yáng ‘sheep/goat’ yang *juŋ A2
銀 yín ‘silver’ ngin *ɲʷi̯ən A2
桶 tǒng ‘bucket’ thuwngX *thɔŋ(X) B1
瓦 wǎ ‘tile’ ngwæX *ŋʷæX B2
甑 zèng ‘rice steamer’ tsingH *tsjɛŋH C1
炭 tàn ‘charcoal’ thanH *thanH C1
竈 zào ‘stove’ tsawH *N-tsoH C1
箸 zhù ‘chopsticks’ drjoH *drouH C2
漆 qī ‘lacquer’ tshit *thjet D1
百 bǎi ‘hundred’ pæk *pæk D1
十 shí ‘ten’ dzyip *gju̯ɛp D252

At first, this seems reasonable for Hmong-Mien, because not only is there internal
evidence of the development of the Chinese type of tone system within Hmong-Mien, the
languages in this family show the imprint of Chinese contact at every level of structure
(see chapter 6, section 6.2, and Downer 1973, Ratliff 1999b, 2000, 2001a, 2009a,b). But
how do speakers hear and borrow tone categories? How do speakers hear and borrow

50
Earlier versions of this section were given at the 28th Annual Meeting of the Berkeley Linguistics Society
Meeting (2002) and at the Ninth Spring Workshop on Theory and Method in Linguistic Reconstruction in
Ann Arbor, MI (2002). It was first published as Ratliff 2005.
51
These are the traditional Chinese names for the four historical categories here labelled A, B, C, and D.
52
This might also be a loan from Tibeto-Burman (see chapter 5, section 2).
186 Chapter 3

whole systems? There is nothing in the speech signal that would enable them to do this.
And how does a borrowed word that has a particular niche within the donor language
system, the historical antecedents of which have been lost because the rise of tones
depends on their being lost, embed that borrowing in a perfectly analogous place within a
whole system that mirrors that of the donor language?
In order to support the idea that the borrowing of tone categories happened in the distant
past, one would ideally like to point to a modern-day contact situation which has recently
yielded, or is in the process of yielding, a pattern similar to this one: identically structured
systems, and borrowed words which occupy analogous places in the systems of both donor
and borrower language. What follows is a review of some recent situations where words
have been borrowed under each logically possible combination of tonal and atonal donor
and borrower languages to show that in three out of the four possible contact situations
between donor and borrower—atonal donor and tonal borrower, tonal donor and tonal
borrower, tonal donor and atonal borrower (the dominant theory)—all the known cases
have yielded patterns quite different from this one. The best conclusion is that, by
elimination, and with the knowledge of how easily prosodic systems have diffused through
Southeast Asia, it was the fourth contact situation—atonal donor and atonal borrower—
which held at the time of these borrowings. According to Sagart, Chinese tones developed
after Old Chinese but before Early Middle Chinese, so somewhere between 500 BCE and
500 CE (1999:101). On the basis of good segmental correspondences, the Chinese
borrowings above can be dated to only slightly before Early Middle Chinese, which we
know was a tonal language, or to the first five hundred years of the Common Era.53
Tonogenesis was ready to happen at this point, but it had not happened yet. It is not clear
that the innovation began with Chinese and then spread to the other languages of the area;
it is just as likely that the languages of the Sinosphere all developed tone together, at
roughly the same time.

3.2.1 Tone assignment in different contact situations

The four possible Chinese/Hmong-Mien contact situations with respect to tone at the time
of borrowing are outlined below. A discussion of the tonal outcomes each type of contact
situation has yielded in known cases will be presented and will be compared to the pattern
of Chinese/Hmong-Mien correspondences that require an explanation.54

Donor atonal, Borrower tonal

For the case where the donor is atonal and the borrower is tonal, Matisoff (2001:321–322)
reports that two variations of one basic strategy are employed in Southeast Asia: one or
two common tones are selected as ‘loan-tones’, or a rare tone is used as the loan-tone
which instantly identifies the word as a borrowing (such as English borrowings in Lahu
and Cantonese). In no reported case are all native tonal contrasts realized in words
borrowed from an atonal language.

53
Pulleyblank (1978), however, suggests that the development of tones was rather late, and that certain
rimes with final -s persisted down to the 6th century CE.
54
Matisoff (2001) gives a very useful catalog of contact situations involving borrowing of words between
tonal and atonal languages similar to this one. He also includes consideration of the relative prestige of
donor and borrower.
 Tonogenesis and tone development 187

The ‘common loan-tone’ strategy often involves an attempt to reflect stress or

intonation contour in the donor language through tone. Court’s 1975 study of Malay
borrowings in Satun Thai indicates that the pitch correlates of stress and intonation are re-
interpreted within a tonal context as tones. He reports that the high rising-falling tone of
Satun Thai is used to represent the falling intonation on phrase-final open syllables in
Malay loanwords. The high level tone is assigned to other syllables which do not occur at a
point which corresponds to the Malay intonation peak. Gandour (1979) found a similar
stress-to-tone mapping in a majority of English polysyllabic loanwords in Thai. Although
the details are considerably more complex, in general the stressed syllable of a polysyllabic
English word is assigned a high tone.
Although Hmong speakers in the United States do not now transliterate nativized
English borrowings into Hmong orthography which makes tone assignment easy to see
(they simply use the English spelling), variation in early nativized loans show employment
of the same basic strategies: an attempt to mirror either English intonation or stress through
tone assignment, or the use of a ‘loan tone’:

(2) ‘America’
ɑmɛlikɑ̀ Mid-Mid-Mid-Low (an attempt to represent English intonation)
ɑ̀mɛ̀lǐkɑ̀ Low-Low-Rising-Low (an attempt to represent English stress)55
ɑ̀mɛ̀lı̀kɑ̀ Low-Low-Low-Low (loan-tone assignment)

If Hmong-Mien had been tonal and Chinese had not been tonal at the time the loans listed
above were made, presumably all pre-Early Middle Chinese loanwords into common
Hmong-Mien would have been realized with one or two tones: the then-current ‘loan-
tones’, either a common tone (or two), or a rare tone. The tonal categories for all the
loanwords of this stratum would then have belonged to a restricted subset of the eight
across which native words are distributed, and would have been the same for all Hmong-
Mien languages—which is clearly not the case.

Donor and Borrower both tonal

The ‘rare loan-tone’ strategy may also be used in the case of borrowings between tone
languages (Matisoff 2001:321). Bunu, for example, a Hmong-Mien language of Guangxi
Province in China, assigns sandhi tones to Chinese and Zhuang borrowings, tones which
are restricted in use in the native portion of the vocabulary. Apparently speakers feel that
these minor tones are appropriate for marking all ‘special purpose’ words—they are also
used to mark baby-names such as ‘little rat’, ‘little frog’, etc. (Mao and Chou 1972, Meng
1983).
However, when both parties are tonal it is more common for borrowers to look for a
match between the tone of the loanword in the donor language and a tone in their language.
Since tone categories (as identified by sets of words that all have the same tone) are
remarkably stable over time, but tone values are remarkably changeable, it is easy to see
that this surface mapping strategy would not yield a neat historical pattern like the one we
are attempting to explain.
For example, recent Chinese loanwords in Hmong-Mien languages are borrowed by
matching the tone value of the word in the local variety of Chinese with the closest tone

55
Although for native speakers, the stress falls on the second syllable, not the third syllable.
188 Chapter 3

value in the borrowing language. ‘Basically, modern loans use Miao initials, finals, and
tones to reflect southwestern Mandarin phonology.’ (Ying 1972:64). In the varieties of
local Chinese spoken in the area of three major Hmongic languages—Yanghao (Eastern
Hmongic), Jiwei (Northern Hmongic), and Xuyong (Western Hmongic)—there is one
rising tone, a reflex of category C1. The tone it will have within each borrowing Hmongic
language is determined by phonetics: the rising tone of each Hmongic language is chosen,
even though in Yanghao the rising tone is a reflex of category B1, in Jiwei the rising tone
is a reflex of category A1, and in Xuyong the rising tone is a reflex of category D2:

(3)
丈 zhàng ‘3 1/3 meters’
Yanghao tsɑŋ B1 [35], Jiwei ʈɑŋ A1 [35], Xuyong tʂɑŋ D2 [13]
上 shàng ‘to start (class)’
Yanghao sɑŋ B1 [35], Jiwei ʂɑŋ A1 [35], Xuyong ʂɑŋ D2 [13]
(Office of Miao-Yao Research Lexicon 1987)

L-Thongkum (1997) presents an interesting case study of language change in progress

among younger speakers of Mien, a tone language, in a village near Chiangmai. Standard
Thai, also a tone language, is used as the medium of instruction, and both Northern Thai
and Standard Thai are spoken widely. In this dialect of Mien, speakers are in the process of
reducing their number of tones from six to five, bringing Mien in line with Standard Thai,
and are also adjusting the phonetic values of the tones themselves to be more in accord
with the phonetic values of Thai tones. However, tone category correspondences do not
result as a by-product of this contact-induced change:

(4)
‘evolving’ Mien Thai Mien Thai
33 33 A1 A1a, A2
31 51 A2 B1, C2, D2L
45 45 B1 B2
34 15 B2, C1 A1b
21 21 C2 C1, D1
(L-Thongkum 1997:158)

Finally, Chamberlain (1972) has also given a brief account of how tones are assigned to
Lao borrowings in five minority Tai languages of Northeast Thailand. All of these
languages, both donor and borrowers, are tonal. The purpose of his paper was to support
the proposal that speakers borrow tone shapes, but not patterns of splits and mergers, by
showing that tone borrowing in these five cases were governed by judgments of phonetic
similarity. Although tone category membership correspondences concern us here, whereas
Chamberlain was interested in patterns of tone category coalescence and splitting, both are
footprints of earlier historical events, and his findings are consistent with the claim made
here—that the process of borrowing does not cause borrowers to reproduce the history of
the donor language within their own language.
Therefore if both Chinese and Hmong-Mien had been tone languages at the time of the
loans, then Middle Chinese loanwords in Hmong-Mien would have the phonetic values of
reflexes of various Hmong-Mien tone categories just as they do in modern Chinese loans.
This would reflect an effort to match the tone in the donor language to the closest tone in
 Tonogenesis and tone development 189

the borrowing language. A borrowed word would not yield reflexes of only one tone
category across the board, moreover that one tone category which provides a perfect match
to the donor language in terms of its place in an identical system.

Donor tonal, Borrower atonal

This is the traditional view of the state of things at the time of early contact, especially
under the Austro-Tai hypothesis which holds that Hmong-Mien was originally atonal like
its Austronesian relatives (Benedict 1975). But in the modern day, we only have evidence
for three basic types of outcomes from contact of this sort. First, words may be borrowed
without tones because the borrowers cannot hear tone, especially if contact is minimal to
non-existent and the number of borrowings is small, as is the case for the handful of
Chinese borrowings in English:

(5)
麻將 má jiàng > ‘mahjong’
烏龍 wū lóng > ‘oolong (tea)’
山東 Shāndōng > ‘shantung (silk)’

Second, in a situation of greater contact, we may get the odd situation that has been
reported by Filbeck (1972) for one dialect of Mal (called T’in at a higher node in the tree),
an atonal Mon-Khmer language of Southeast Asia spoken in the northeast part of Thailand.
Mal has acquired tones in two ways. First, Northern Thai numerals have been borrowed
along with their tones. Although this may have arisen as a recitation effect, the numerals
are pronounced with their Northern Thai tones even in running speech. At the same time,
other Thai words have been borrowed with a rising pitch profile—presumably because of
its salience, because it does not correspond to the pitch contours of the borrowings as
pronounced in Thai:

(6)
Thai Mal
yâak ɲǎak ‘difficult’
lâak lǎak ‘to drag’
khɛ̀ɛk khɛ̌ɛk ‘guest’
klaaŋ kǎaŋ ‘middle’
(Filbeck 1972:115)

This rising contour also marks words borrowed from other atonal Mon-Khmer languages,
and, inexplicably, a couple of native words well.
The third situation, one of intense and prolonged contact, is probably closest to what
proponents of the theory that Hmong-Mien acquired tones from Chinese must assume for
early Chinese/Hmong-Mien contact, given the strong influence Chinese has had upon
Hmong-Mien languages over the centuries. There is one good modern day case where, in
the face of widespread multilingualism, the prosodic model of a dominant tonal language
(or in this case, languages) caused an atonal language to develop a complex system of
tones. This is the case of the development of tones in the originally atonal Austronesian
language Tsat of Hainan Island as described by Thurgood in a series of publications (see
Thurgood 1999 for a full account and further references). Tones developed in Tsat under
190 Chapter 3

the influence of two tone languages: the local (Min) varieties of Chinese, and the Tai-
Kadai language Li. The tonal system that developed in Tsat is similar in nature to the Thai-
influenced Mien described by L-Thongkum: the number of tones and the tone values are
almost the same as those of these two neighboring languages which the Tsat speakers
know well (Thurgood 1999:231), a case of surface convergence:

(7)
Tsat Chinese (Tanchou) Li (Tongshi) Li (Yuanmen)
high level 55 55 55 55
falling 42 -- 43 42
mid level 33 22 33 44
rising 24 35 13 13
low level 11 11 11 11

But the key difference between this situation and the early Sinospheric contact situation
that we are trying to understand is that Tsat developed tones in its own way, in a two-by-
three system, where the initials started the tonogenetic process and the later split was
conditioned by the finals. Thus there can be no corresponding categories in Tsat for
Chinese loanwords to map onto—a four-by-two system cannot map directly onto a two-by-
three system. Chinese borrowings in Tsat are pronounced with whatever tone is
phonetically most similar (Thurgood, p.c. 2002). Although given the assumptions about
tone spread in Southeast Asia this should be the model we are looking for, the kind of
cross-linguistic category correspondence pattern characteristic of the Sinosphere is not
found here. As Thurgood (1999:231) writes, “it is not being argued that the Tsat tonal
system is borrowed from one of these languages”.

Donor and Borrower both atonal

Given our inability to find a present-day (or recent past) model that could account for the
correspondence of tones in the Sinosphere under these three contact situations, we must
turn to the last possible scenario. If neither Hmong-Mien nor Chinese had tones at the time
these early loans above were made, we can indeed imagine how the striking cross-family
correspondence of tone categories in loanwords could have arisen. Hmong-Mien could
have borrowed the Chinese words with the (perhaps already decomposing) segmental
material which eventually gave rise to tones intact. Then if both developed tones in the
exactly same way, out of the laryngeal features of word-final consonants as tonogenesis
swept across the area—started by an unknown trigger language, not necessarily by
Chinese—then we would get these regular correspondences. This is the most likely
explanation of what happened.
The following table contains a summary of the four types of contact described above:
 Tonogenesis and tone development 191

(8)
1. Donor atonal, Borrower tonal
2. Donor and Borrower both tonal
 assignment to a common tone
 phonetic mapping:
(interpretation of stress/intonation as
Modern Chinese > HM
tone):
Lao > Tai dialects
English > Hmong, Malay > Thai,
 surface convergence: Thai > Mien
English > Thai
 assignment to a rare tone:
 assignment to a rare tone:
Chinese, Zhuang > Bunu
English > Cantonese, Lahu
3. Donor tonal, Borrower atonal 4. Donor and Borrower both atonal
 words borrowed without tones:  the languages both stay atonal
Chinese > English  the languages both become tonal
 individual words borrowed with tones:
N. Thai > T’in (numerals)
 assignment to a distinctive pitch profile:
N. Thai > T’in (other borrowings)
 systems of independent origin, surface
convergence: Chinese, Li > Tsat

3.2.2 Sinospheric languages developed tones together

One striking piece of evidence in support of the hypothesis that Chinese itself was atonal
when it lent Hmong-Mien the words in the first set above is the fact that an even older
stratum of Chinese loans shows regular tonal correspondences between the two families.
Most scholars believe that Old Chinese was toneless (Mei 1970, Baxter 1992, Sagart
1999). How else, then, can the correspondences in this oldest stratum of loanwords be
explained other than to say that tones developed in the two languages in a parallel fashion
after the words were borrowed? And how else can the identical pattern in the later stratum
of loanwords be explained other than to say that tones developed in the two languages
families after these words were borrowed as well?

(9) MC OC PHM Tone category in both

廩 lǐn ‘barn/granary’ limX *(pə.)r[ə]mʔ *rɛmX B2
鐵 tiě ‘iron’ thet *l ̥ˤik D
*hrɛk (PM) D1
力 lì ‘strength’ lik *kə.rək *-rək D2

Although Chinese contact is probably to be credited with making Hmong-Mien and

other languages of the south monosyllabic word and thus ‘tone prone’ (Matisoff 1973:89),
it is not clear that Chinese acquired tones before other languages in the area. Since the later
merger of initials and compensatory tone split are understood as happening in a wave
across the whole area, it is plausible that the initial wave of tonogenesis was an areal
phenomenon as well. If we accept the proposal that Hmong-Mien borrowed its earliest
Chinese loanwords before tones developed, it will no longer be necessary to answer the ill-
conceived question of how these languages might have borrowed a tone system and
historical tone categories. Furthermore, to the extent that these early loans can be dated on
the basis of the segmental correspondences, so, too, can the onset of tonogenesis be dated
in both Chinese and the languages that have been so profoundly influenced by Chinese.
192 Chapter 3

Future discussions of tonogenesis in Chinese should no longer rest exclusively on internal

evidence, but should also make use of the external loanword evidence in attempting to date
the onset of this important typological change.

3.3 Phonation-to-tone development56

In the secondary tone split conditioned by the voicing of initial consonants, pitch contrasts
were not realized immediately, as in the classic scenario sketched in the introduction, but
rather through an intermediate stage of contrasting phonation type. An originally voiced
initial consonant left its trace as a lower register tone through the intermediate stage of
breathiness on the vowel: ba > pa̤ > pà.57 The extensive cross-linguistic support and
phonetic justification for this stage of development is laid out clearly in Thurgood 2002
and will not be repeated here. In this section, facts about patterns of pitch level, remnant
phonation contrasts, vowel quality, and complex contour tones across the modern-day
Hmong-Mien languages are presented, facts which can only be explained by positing an
even earlier phonation contrast that arose upon loss of final laryngeal distinctions in the
syllable coda. The idea that the loss of final glottal stop and final -h should give rise to
creakiness and breathiness is not controversial (Thurgood 2002:341–342). But in order to
account for observed patterns in the modern-day languages, we must assume that there was
a period when these two sets of phonation contrasts co-existed—the more recent arising
from the breakdown of consonant contrasts at the front of the syllable, and the older arising
from the loss of consonant contrasts at the back of the syllable. In other words, the new
phonation contrast that arose upon loss of the initial consonant voicing contrast was
‘layered’ upon a pre-existing phonation contrast that arose upon loss of final laryngeal
distinctions. The three traditional tonal categories (A, B, C) were thus initially phonation
categories—modal, creaky, and breathy voice respectively—and these ancient ‘tones’ were
still primarily differentiated by phonation when the voiceless and voiced initial obstruents
merged, yielding a ‘tone’ split (A1/A2, etc.) which was originally the introduction of a
second layer of phonation contrasts.
The historical development would have thus passed through four stages. The first stage
would have preceded the development of either phonation or tone:

A B C
Stage I CV(N) CV(N)ʔ CV(N)h

The second stage would have involved the loss of final laryngeal consonants and the
consequent development of phonation contrasts:

A B C
Stage II CV(N) CV̰ (N) CV̤ (N)

56
An earlier version of the material in this section was given at the 1999 Meeting of the Linguistics Society
of America in Los Angeles (Ratliff 1999a). I am indebted to Haiyong Liu for his help in the early stages
of research on this topic.
57
Breathy phonation is often represented as a voiced aspirated initial in the Chinese literature (pɦa). The
breathiness may in fact have been more prominent at the beginning of the syllable in the early stages of
the tone split.
 Tonogenesis and tone development 193

The third stage would correspond to a time when these phonation contrasts had not become
fully tonal but when the ‘tone’ split (first, and in part to this day, a phonation split),
triggered by the merger of initial voiced and voiceless obstruents, was beginning. Thus,
one layer of phonation contrasts—modal vs. breathy—came to overlap another layer of
phonation contrasts—modal vs. creaky vs. breathy:

A B C
Stage III CV(N) CV̰ (N) CV̤ (N)
CɦV(N) CɦV̰ (N) CɦV̤ (N)

Finally, the fourth stage represents the present day, a time in which most of the Hmong-
Mien languages have tones characterized by modal voice and pitch contrasts alone. Note,
however, that across the family breathy voice is occasionally found in reflexes of category
A2, is still common in reflexes of category B2, and is even more common than modal
voice in reflexes of category C2.

A B C
Stage IV CV(N) CV(N) CV(N)
CV(N) ~ CV̤ (N) CV(N) ~ CV̤ (N) ~ CV̰ (N) CV̤ (N) ~ CV(N)

The evidence for this hypothesis consists of four wholly independent facts about
Hmong-Mien, to be discussed in sections 3.3.1 through 3.3.4 below, which can only be
explained by positing a phonation contrast that arose upon loss of the coda consonants, and
that still existed at the time of the tone split.

3.3.1 Distribution of low-pitch tones in modern languages

The first pattern which suggests the historical scenario described above is that relative
pitch height values as a reflection of ancient initial consonant voicing are more or less
preserved in categories B and C across the family, whereas in category A they are not.
That is, reflexes of the “lower register” tones B2 and C2 are actually lower in pitch than
their upper register counterparts B1 and C1. On the other hand, in category A it is
impossible to predict, for any given Hmong-Mien language dialect, whether the A2 reflex
will be lower or higher in pitch than its A1 counterpart. A survey was made of 70 Hmong-
Mien “tongues” (some different languages, some different dialects of the same language
which differ tonally): 22 languages/dialects of the Wang and Mao 1995 reconstruction
(omitting Fuyuan, which did not undergo a tone split), 43 languages/dialects from Chang
1972 (omitting dialects represented in Wang and Mao 1995 and dialects that did not
undergo a tone split), Green Mong (Andruski and Ratliff 2000), two dialects of Pa-Hng
(Niederer 1997), Houei Sai Mun (Strecker 1990), and Biao Min (Solnit 1985). In 81% of
the 70 dialects (57 locations), B2 is lower than B1; and in 89% of the 70 dialects (62
locations), C2 is lower than C1. This demonstrates the expected depressor effect of the
initial voiced consonant. On the other hand, in only 50% of the 70 dialects (exactly 35
194 Chapter 3

locations) is A2 lower in pitch than A1.58 Many B2 and C2 tones are also breathy (C2
slightly more often than B2), which may help explain why the pitch of these tones are
comparatively low.
But the question arises: why is this not also true of category A2, since ancient voiced
initials in words of this category must also have yielded breathiness on the following
vowel, presumably leading to a depressed pitch? The difference between categories B and
C on the one hand, and category A on the other can be explained if we assume that the
phonation type of syllables differed at the time of the tone split. Then the breathiness from
the original voiced initial, supported by the non-modal quality of the rest of the syllable,
whether breathy or creaky, could have kept the pitch of B2 and C2 depressed. In A2,
however, initial-induced breathiness on a modal voice syllable was not as easily sustained.

3.3.2 Residual phonation contrasts in modern languages

The manifestation of breathy voice in tone categories derived from original voiced initials
in modern languages differs according to tone category: (1) it is rare in category A2, and
when present weak: for example, Yaoli Bunu (Wang and Mao 1995:35); (2) it is common
in category C2, and when present robust: for example, Yanghao Hmu (Wang and Mao
1995:32) and Green Mong (Andruski and Ratliff 2000); (3) B2 words across the family
show an intermediate status between these two in both frequency and robustness. The
relative rarity of breathiness in tone A2 can be attributed to the dominance of unmarked
modal phonation in the A syllable. The common occurrence of breathiness in tone C2 can
easily be attributed to breathiness from the back of the syllable meeting breathiness from
the front of the syllable. Tone B2, on the other hand, presents a mixed picture, with two
marked phonation types meeting mid-syllable. Across the family, words with tone B2 are
less often breathy than C2 words, but more often breathy than A2 words, which were
originally characterized by modal voice. An implicational universal thus holds with respect
to the manifestation of breathy voice in syllables with originally voiced initials: if A2 is
breathy in a particular dialect, then so are B2 and C2; if B2 is breathy in a particular
dialect, then so is C2.59
In addition, there appears to be retention of category B creakiness in modern varieties of
Mun. The best evidence for this comes from Houei Sai Mun, where categories B1 and B2
(53ʔ) are glottalized at the end: “Tone B appears to have had a short period of
glottalization at the very end of the syllable. […] Tone B corresponds to Middle Chinese
shang sheng or rising tone, which has been reconstructed as ending in a glottal stop (see
especially Mei 1970)” (Strecker 1990:29). However, tone C2 in this dialect is also
glottalized, and moreover is glottalized throughout the syllable. Strecker tentatively

58
An analysis of pitch by tone register across tone categories in Niederer 1998 (239–243) did not yield such
dramatic results, but it is interesting that for the five languages of her sample that have level tones for
both upper and lower register in category A, the lower register is always higher than the upper register.
This may be accidental, however.
59
Tone mergers can obscure this pattern. Iu Mien of Chiang Rai shows breathiness only in category B2, but
C2 and D2 have merged into a 11 tone (Downer 1961, Purnell, forthcoming) which obscures the original
distribution of phonation types. Similarly, Biao Min shows breathiness only in category A2, but B2, C2,
and D2 have merged into a 42ʔ tone (Solnit 1985:182). Since A2 is breathy, it is likely that both B2 and
C2 were breathy originally, and the merged tone terminated by a glottal stop reflects the detritus of final
stop consonants. Shimen (A-Hmao) is a true exception, where breathiness is found in A2, B2, and D2, but
not in C2.
 Tonogenesis and tone development 195

attributes C2 glottalization to a final glottal stop in category C as well, but that could
hardly be true—B and C must have corresponded to different syllable types. More likely,
since C2 and D2 (2̰2̰) have merged, the tenseness is due to properties of the D2 syllable,
which ended in a final stop consonant (D1 is also glottalized).
In Lanjin Mun (Wang and Mao 1995:37), category C2 words are 31 and tense;
however, since this tone falls to the bottom of the pitch range, this may be attributed to the
natural breakup of creakiness that accompanies falling below one’s normal pitch range
(“vocal fry”). But the category B2 tone is not a low tone, and it is “especially tense” (4̰4̰).60

3.3.3 Breathy voice and vowel quality splits

Additional support for this hypothesis from vowel quality evidence comes from only one
language to date. In Zongdi, a Mashan dialect of West Hmongic, there is regular vowel
quality raising, a lax-glottis effect (Matisoff 1973:76), in categories B2 and C2 but not in
A2, which is characterized by weaker breathiness cross-linguistically. Zongdi tones B2 and
C2 are no longer characterized by breathiness; the vowel quality contrast is all that remains
of the earlier phonation type contrast.

(10)
Tones B2 and C2 Other tones
Rime 361 e æ
Rimes 5, 15, 16 əa a
Rimes 6, 12, 17 o ɔ
Rime 7 u o
Rime 8 u ou
Rime 13 ɯ ə
Rimes 18, 19 ein æin
Rimes 29, 30 əŋ aŋ
(Wang 1994:4–5; Wang and Mao 1995:34)

3.3.4 Complex contour tones as phonation ‘gear switches’

Complex contour tones occur almost exclusively in reflexes of category B.62 This can be
understood as a ‘gear-switching’ adjustment required in the mid-syllable transition from
modal to creaky (B1) and from breathy to creaky (B2). These modern-day B-tone words
with complex contours are no longer characterized by creaky voice; the complex contour is
all that remains of the earlier phonation type contrast.
Examples of complex contour tones from Wang and Mao 1995 (WM), Niederer 1998
(N), and Chang 1972 (C) appear below:

60
Niederer (1998:249–250) reports that Iu Mien as spoken in Thailand has creaky voice in category C2, and
only in category C2. However, Downer (1961:532) describes tone C2 as ‘very low, falling slightly, with
some weak glottal closure (‘creakiness’)’. Given the fall from ‘very low’, the creakiness here can also be
attributed to vocal fry.
61
Rime numbers refer to the categories in this book (and in Wang 1994).
62
Exceptions include the following: A1 Jiongnai (Wang and Mao 1995 #13) 343, A1/C1/D1 Kim Di Mun
(Niederer 1998 #37) 354, A1 Houei Sai Mun (Strecker 1990) 534, D1 Houei Sai Mun (Strecker 1990)
4232, C2 Xinzhai (Niederer 1998 #24) 343.
196 Chapter 3

(11)
B1 complex tones (< modal-creaky) B2 complex tones (< breathy-creaky)
Zongdi Mang (WM #7) 232 Hsiao-miao-chai Miao (C #15) 213
Shuijungping Mang (N #23) 232 Tu-an Miao (C #51) 231
Biao Muen (N #32) 354 Qibainong Bunu (WM #10) 232, 454
Thailand/Laos Mien (N #33) 452 Meizhu Bunu (N #25) 231
Liangzi Mun (WM #19) 545 Dapingjiang Mien (N #30) 231
Lanjin Mun (WM #20) 434 Thailand/Lao Mien(N #33) 231
Hainan Mun (N #35) 354 Jiangdi Mien (WM #15) 231
Hwei-kang-pa Yao (C #67) 453 Xiangnan Mien (WM #16) 232
Luoxiang Mien (WM #17) 213
Changping Mien (WM #18) 121
Hsing-an Yao (C #63) 231
Hwei-kang-pa Yao (C #67) 231

The recent birth of a complex 213 tone in White Hmong—called the -d tone, after its
written form—gives an illustration of how complex contours develop when a transition
between two phonation types must be made. A small number of words with this tone arise
from a phrase consisting of a locative with a 21ʔ contour and a demonstrative with a 33
tone (Ratliff 1992a:112–120, after Downer 1967:597–598).63 The first syllable of the
phrase is glottalized or creaky due to its being a reflex of category D2, a syllable with an
original stop consonant in the coda, and the second syllable has modal phonation. There is
no glottalization or creakiness in the resulting complex contour.

(12)
tɔ21ʔ nɔ33  tɔ213
there DEM over there

3.3.5 Summary
Below is a representation of Hmong-Mien tonogenesis under the ‘layered phonation’
hypothesis:

HM tone HM initial Ancient voice quality Modern voice quality

A modal voiceless A1 modal-modal > modal
voiced A2 breathy-modal > modal (rare: breathy)
B creaky voiceless B1 modal-creaky > modal
modal or breathy
voiced B2 breathy-creaky >
(rare: tense)
C breathy voiceless C1 modal-breathy > modal
voiced C2 breathy-breathy > breathy or modal

As has been demonstrated above, evidence for this hypothesis comes from a striking
asymmetry in the pitch heights of category A2 words versus B2 and C2 words in a sample

63
Other words with a demonstrative (or vocative) force have acquired the 213 tone by analogy with this set
of words.
 Tonogenesis and tone development 197

of 70 Hmong-Mien dialects. In most dialects, B2 and C2 are lower than B1 and C1,
demonstrating the expected depressor effect of the original voiced initial, whereas in only
35/70 dialects is A2 lower in pitch than A1. It is clear that breathiness from the original
voiced initial, supported by the non-modal quality of the rest of the syllable, has kept the
pitch of B2 and C2 depressed. In A2, however, the effect of initial-induced breathiness on
a modal voice syllable was comparatively transitory. Predictably, in A2 syllables
breathiness is rare and when present weak (Yaoli), in C2 syllables it is common and when
present robust (Yanghao). B2 and C2 words show regular vowel quality raising in the
Zongdi dialect, a lax-glottis effect. Some Mun dialects have been characterized as still
having tense or creaky voice qualities in words of category B2 (Houei Sai Mun, Lanjin
Mun). Finally, double contour tones (rising-falling or falling-rising) are common only in
the B category across dialects, reflecting an old phonation gear switch to creaky mid-
syllable.
Layered phonation contrasts are typical of Mon-Khmer languages, a few of which have
only recently started to develop tones (see especially the Pearic language Chong—Diffloth
1989, L-Thongkum 1991), but have never been demonstrated for Chinese, Tai-Kadai, or
Hmong-Mien, the three families of Asia which are today fully tonal and which share
similar modern tonal systems and tonal histories. The reconstruction of layered phonation
systems for one of these families may thus have relevance for all three.
4 Morphology

4.1 Introduction
Other than compounding, there is little internal word structure in present-day Hmong-Mien
languages, and there is only fragmentary evidence for complex words at the Proto Hmong-
Mien level. However, given the fact that languages rarely remain constant with regard to
morphological type throughout the course of their history, one expects to find, and can
find, evidence of both new morphological developments and relics of old morphological
contrasts in these “isolating” languages.
Among the morphological processes that can be found in all languages of the
Sinosphere are both single compounding of two morphemes and double compounding of
four morphemes (of the ABAC or ABCB type, where repeated elements are either
identical or closely related semantically). Reduplication is also widespread. Within
Hmongic, reflexes of the reciprocal prefix *sjɨA-D (2.28/2) in some languages harmonize
with the verb root in tone and/or rime (Wang 1994:43, Ratliff 1992a:40–42).64 In White
Hmong, compounds phonologically linked by tone sandhi can signal subtle meaning
differences (Ratliff 1992a:58–62). Vowel ablaut in classifiers reflecting characteristics of
the associated noun has been reported for the Hmongic language A-Hmao (Wang 1972),
and meaningful patterns in consonant, vowel, and tone alternations in expressives
(ideophones) have been reported for White Hmong (Ratliff 1986b, 1992a:136–163). Such
morpheme-internal alternations as these are undoubtedly to be found in other languages of
the family; further language-specific work on these topics is eagerly anticipated.
This chapter, however, given the historical focus of the book, will be devoted to four
cases of what appear to be traces of ancient morphology. The first two—nominal prefixes
(section 4.2) and prenasalization (section 4.3)—are still quite robust in present-day
languages, even though their original functions are difficult to recover. The second two
(section 4.4)—initial voicing contrasts suggesting grammatical prefixes and derivation by
tone change—are only preserved in a few pairs of words in present-day languages. Under
the standard theory that differences in major tone category arose from final laryngeal
contrasts (see chapter 3), derivation by tone change might be pushed back even further to
the reconstruction of old suffixes.

64
The Hmongic reciprocal was probably borrowed from Chinese 廝 (MC sje ‘mutually’ > Man. sī ‘with
each other’); see Sagart 1999:70 on its use as a reciprocal in Chinese.

198
 Morphology 199

4.2 Nominal prefixes65

Nouns in Hmong-Mien are frequently characterized by the presence of initial minor
syllables; more occur in some languages and fewer in others. Some of these initial
syllables are clearly prefixes today, while others look like they might have been prefixes in
the past. Reconstruction of these syllables is especially difficult since they function more
like initial syllable word formatives than meaningful prefixes in the usual sense, and there
is little consistency in the association of a particular prefix with a particular word or
semantic class of words from language to language. This section considers the extent to
which these initial syllables/prefixes can be reconstructed.

4.2.1 Hmong-Mien nominal prefixes today

Prefix form and function

In form, Hmong-Mien nominal prefixes are open syllables with single consonant onsets (or
zero onsets) and monophthong nuclei. The consonants and vowels which appear in these
prefixes represent only a subset of the full set of contrastive phonemes in any language.
They are unstressed, and in some languages they may harmonize with the nouns to which
they are attached (Chen 1993).
The function of nominal prefixes in Hmong-Mien is weakly classifying. Within
individual languages it is usually possible to group nouns into classes like “human”,
“animal”, and “tool”, etc., and to show that each class more commonly takes one prefix
than another, although in some languages this system has weakened to the point where
only one prefix remains as a marker of noun class membership. The function of these
prefixes immediately raises the question of the status of this system in comparison with the
status of the numeral classifier system, since the two systems both serve the function of
noun classification. In chapter 6, evidence is presented to show that the prefix system is the
native system of noun classification in the family, and the classifier system has been
borrowed from Chinese.

Age of the prefix system and distribution of prefixes within the family
In addition to evidence from phonological and semantic degradation, Chen (1993) has
given two further arguments from modern-day usage in support of the claim that prefixes
are quite old in the family: (1) prefixes are used more often among older speakers than
younger speakers, and (2) prefixed forms occur more often in frozen expressions as
opposed to innovative combinations of words. Furthermore, the distribution of prefixes
within the family points to the same conclusion (see below). Finally, evidence from
independent lines of research suggesting a Proto Hmong-Mien disyllabic template
indirectly supports a claim of the antiquity of these prefixes (see section 4.2.4 below).
Within the Hmong-Mien family, there is a correlation between the intensity of contact
with Chinese and the health of the prefix system. It is well-known that in both culture and
language, the Mienic side of the family shows more Chinese contact effects than the
Hmongic side of the family. For example, unlike the Hmong, Mien people are Taoist, and
use the Chinese language in rituals and songs and Chinese characters in religious writings

65
This section is adapted from Ratliff 2006.
200 Chapter 4

(Lemoine 1982, Purnell 1991, 1998). They use Chinese numerals alongside native
numerals (Purnell, forthcoming). It is also common to find a split between Hmongic and
Mienic cognate sets, where the Hmongic side has retained the native word, and the Mienic
side has replaced the native word with a Chinese borrowing (see chapter 7, section 1).
In line with this difference in intensity of contact with Chinese, the Hmongic languages
have retained more native nominal prefixes than the Mienic languages. Impressionistically,
it also appears that the Hmongic languages that display the most archaic features generally,
like Pa-Hng, are also richer in prefixes (see section 4.2.2 below). On the Mienic side,
Mien, Mun, and Biao Min show fewer prefixes than the most conservative Mienic
language, Zao Min (northwest Guangdong province), which differs enough from the others
to constitute a separate sub-branch. Mao, Meng, and Zheng (1982) note that the regular use
of the prefix /ʔa-/ on nouns in Zao Min contrasts with languages of the other two sub-
branches wherein only a few nouns optionally bear prefixes. In an early field report, the
Zao Min language was described as “isolated” and the “one least influenced by the
neighboring Chinese dialects” (Wong 1939:425). There thus appears to be a correlation
between isolation and prefix retention.
Yet despite the apparent antiquity of the prefix system, in some languages of the family
nominal prefixes are abundantly evident in the present day, and even appear in ethnic
autonyms, such as A-Hmao, Qo-Xiong, Pa-Hng, Pu-nu, etc. This is in part due to the fact
that class nouns provide an on-going source of new prefixes.

4.2.2 Cross-linguistic prefix variation

Nominal prefixes derive from class nouns as the first member of noun-noun compounds.
This process is independent and on-going in different languages of the family. For
example, cognates of White Hmong kab /ka1/ in the Eastern and Western branches of
Hmongic mean ‘insect’ and are also used as class nouns in insect name compounds (kab
laug sab /ka1 lau6 ʂa1/ ‘spider’, kab tuas /ka1 tua4/7/ ‘caterpillar’, kab nqos vias /ka1 ɴqɔ4/7
via4/7/ ‘locust’, etc.) but are beginning to generalize, now serving as the first element in
compounds denoting shrimp, crabs, and moles. And tub /tu1/ ‘son/boy’ in White Hmong is
now used as a class noun in compounds that refer to male humans, whether young or not:
for example, tub-rog /tu1 rɔ6/ ‘soldier’ (literally,‘boy-war’).
An older, but still recoverable example of the same historical process can be seen in the
common White Hmong class noun or prefix pob /pɔ1/, which serves as the first element in
compounds that denote lumpy things, such as stones, knots, tree stumps, and earlobes. Pob
is a loanword from Chinese 包 (Man. bāo) ‘to wrap; a bundle, lump’, and was probably
originally borrowed as a noun. Although this word can be used as a classifier as well,
which is not typical of prefixes, (ib pob mov /i1 pɔ1 mɔ3/ ‘one CLF-clumps rice’), it is more
typically used as a prefix attached to a noun which requires yet another word as its
classifier (ib lub pob-zeb /i1 lu1 pɔ1 ʐe1/ ‘one CLF-bulky things clump-stone’). See chapter
6, section 6.2.4 on pob in connection with a discussion of language contact and the two
Hmong-Mien nominal classification systems.
From the variation that is evident in the mapping of particular prefixes to particular
nouns across the family, it would seem that what the languages share, and all that can be
reconstructed as a likely characteristic of the protolanguage, is a “prefixing predilection,”
rather than the combination of particular morphemes. As can be seen from a glance at the
 Morphology 201

forms for ‘tail’, ‘stone’ and ‘nose’ given below, individual cognate nouns across the family
may take a variety of different prefixes.

(1) ‘tail’ (Hmong-Mien *tu̯eiX (2.1/8) > White Hmong ko-tw /ko5 tɯ5/)
Yanghao (Hmu) qɑ1-tɑ3
Jiwei (Qo Xiong) pi -tɤ3
3

Meizhu (Bunu) ku1-tau3

Baiyun (Pa-Hng) tɛ2-tɛ3

‘stone’ (Hmong-Mien *-ʔrəu (2.55/3) > White Hmong pob-zeb /pɔ1 ʐe1/)
Yanghao (Hmu) po3-ɣi1
Jiwei (Qo Xiong) qo1-ʐɯ1
Shimen (A-Hmao) a1-və1
Meizhu (Bunu) fa3-ɣe1
Baiyun (Pa-Hng) ʔã[33]-jo1

‘nose’ (Hmong-Mien *mbruiH (1.51/8) > White Hmong qhov-ntswg /qhɔ3 ɳtʃɯ6/)
Yanghao (Hmu) po3-nɛ6
Jiwei (Qo Xiong) pɑ3/7-mpɤ6
Xuyong (Hmong) tsʅ5-ɳdʐu6
Shimen (A-Hmao) a1-mpy6
Meizhu (Bunu) pi3-ntsau6

Even at very low levels the variation is no less remarkable, as is seen in the following
cognate noun pairs from two dialects of the Hmongic language Pa-Hng. These sets were
chosen to demonstrate variation; several other nouns have the same prefix in both dialects
(all forms are from Niederer 1997).

(2) Baiyun (Guangxi, China) Tan Trinh (North Vietnam)

thigh tɤ[33]-pe1 ko[31ʔ]-pe1
tooth ʔæ̃ [35]-mɪ3̃ tɤ[33]-mɪ3̃
finger ka[35]-wa3 tɤ[33]-tɔ̃[33]-ŋgwa3
pig n̩[21ʔ]-me5 ta[33]-mbe5
fish n̩[21ʔ]-mjo4 ta[33]-mbjo4
hawk ʔn̩[21ʔ]-kwɤ3 ta[33]-kwɤ3
insect n̩[21ʔ]-kɛ1 ta[33]-kɤ1
bedbug n̩[21ʔ]-pɪ5 ta[33]-pɪ5
ant ka[21ʔ]-m̥ je3 ta[33]-ka[31ʔ]-phe3
tree tɛ[33]-pja1 ʔa[33]-pa1
flower ʔæ̃ [35]-pɛ2 ʨo[31ʔ]-pɤ2
chisel tɛ[33]-ʨɪ4̃ pa[31ʔ]-ʨɪ4̃
balance tɛ[33]-ȵ̥̥ɔ5̃ pa[31ʔ]-ȵ̥̥ɔ5
crossbow ka[21ʔ]-n̥ẽ3 pa[33]-n̥ĩ3

Two observations are in order. First, in the Tan Trinh words for ‘finger’ and ‘ant’ there are
both “inner” and “outer” prefixes, which gives witness to the active nature of the
prefixation process. Second, certain prefixes can be associated with certain semantic
202 Chapter 4

classes in each of these dialects, but as soon as one rises to the language level, it is difficult
to assign classifying functions to individual prefixes. Obviously, the difficulty in assigning
a general meaning to these prefixes at the Proto Hmongic and Proto Hmong-Mien level is
accordingly much greater.
A partial explanation for the considerable cross-linguistic variation exemplified here is
the fact of a certain amount of prefix variation within each particular language and dialect,
more for some nouns and less for others.66 The generalization of one prefix over the others
would then follow as an independent development in each language, yielding the confusing
picture presented above. For example, in White Hmong, the noun ntsej ‘ear’ can occur
with the following prefixes, depending on which aspect of the ear is in focus: pob-ntsej
(lump-ear) ‘the outer ear’, qhov-ntsej (hole-ear) ‘the ear canal’, taub-ntsej (squash-ear) ‘the
earlobe’.
We may be able to reconstruct aspects of certain individual prefixes. Strecker (1989)
has reconstructed a Proto Hmongic *qouA prefix, and it is true that the most frequently
encountered prefix has either a voiceless velar or uvular onset (or a glottal stop onset) and
the A tone. But it is difficult to associate a core meaning with this prefix, since it is found
on nouns that refer to animates, plants, inanimates, and abstractions alike across the
Hmongic family. And it is clearly not possible to reconstruct particular prefix-noun
combinations. What does seem possible is the reconstruction of a prefix-root template for
the class of nouns as a whole.

4.2.3 Prefix pre-emption

In work on the reconstruction, it was not unusual to find one or more forms in a set of
cognates which corresponded perfectly in terms of rime and tone, but which had aberrant
initials. In the case of nouns, this is most easily explained as a process of “prefix pre-
emption” (Matisoff 1997), wherein the vowel of the prefix collapsed and the initial of the
prefix displaced the initial consonant of the root. Given the commonplace variation in
prefixes evident in the data above, however, it would be foolish to reconstruct one
particular prefix at the level of the protolanguage. Examples include the following three
words, each of which has only one form with an odd initial:

(3) ‘bone’ (Hmong-Mien *tshuŋX (3.2/28) > White Hmong txha /tsha5/)
Yanghao (Hmu) shoŋ3 Jinxiu (Mien) buŋ3
Jiwei (Qo Xiong) soŋ3 Liangzi (Mun) tθuŋ3
Xianjin (Hmong) tshaŋ Sunjiang (Biao Min) sjɔŋ3
5

Fuyuan (Hmjo) tshoŋB

Although Peiros (1998:158) proposes a Mon-Khmer connection for this word, neither
lookalikes in that family nor a straightforward story of phonological development within
Hmong-Mien can account for the labial initial for ‘bone’ in Mien. It seems to be a good
example of prefix pre-emption, especially given the widespread use of prefixes with body-
part terms.

66
I am grateful to David Mortensen for reminding me of this fact.
 Morphology 203

(4) ‘mane’ (Hmongic *tsʉŋA 3.1/30 > White Hmong txoob /tsɔŋ1/;
Mienic *ʔdzɔŋA 3.4/29)
Yanghao (Hmu) soŋ1 Jiangdi (Mien) dzoŋ1
Jiwei (Qo Xiong) tsoŋ Liangzi (Mun)
3
dɔŋ1
Xianjin (Hmong) tsoŋ1 Sunjiang (Biao Min) tsɔŋ1
Fuyuan (Hmjo) pjoŋA

This is clearly a loanword from Chinese 騣 ‘mane’ (MC tsuwng > Man. zōng). Given the
differences in rime correspondence, it seems to have been borrowed independently into the
two branches of the family. There is no evidence of a labial initial in either branch or in the
Chinese source, however: the Fuyuan form with a labial initial thus shows evidence of
prefix pre-emption.

(5.) ‘eagle/hawk’ (Hmong-Mien *qlaŋX (6.31/24) > White Hmong dav /ʔda3/)
Yanghao (Hmu) l ̥aŋ3 Jiangdi (Mien) ʨaːŋ3
Jiwei (Qo Xiong) qwei Liangzi (Mun)
3
tlaŋ3
Xianjin (Hmong) tl ̥aŋ3 Dongshan (Biao Min) klaŋ3
Gaopo (Hmong) ploŋ3

The word for ‘eagle/hawk’ is not unique to Hmong-Mien; it is a well-known word shared
by almost all of the languages of Southeast Asia, and everywhere the word has a dorsal-
liquid cluster. The Gaopo form with a labial initial thus shows evidence of prefix pre-
emption.
The complications of one particularly puzzling initial correspondence may also be
attributed to the same process. The word for ‘hand’ is the sole member of its initial
correspondence “set” in the reconstructions of Wang (1994) and Wang and Mao (1995).
These words are uncontroversially cognate because the rime and tone correspondences are
perfectly regular. The forms from the nine Hmongic languages he lists, with the addition of
Pa-Hng (from Niederer 1997) and Ho Ne (or She, from Wang and Mao 1995) are given
below (the cognate Mienic forms are not listed here because they all begin with /p-/):

(6) ‘hand’ (Hmong-Mien *-bɔuX (1.3/3) > White Hmong tes /te4/)
Yanghao (Hmu) pi4 Zhongdi (Mang) ʂe4
Jiwei (Qo Xiong) tɯ 4
Fuyuan (Hmjo) weiB
Xianjin (Hmong) te4 Fengxiang (Hmong) pi4
Shimen (A-Hmao) di 4
Baiyun (Pa-Hng) tɛ-pɯ4
Qingyan (Hmong) tsa4 Duozhu (Ho Ne) khwa4
Gaopo (Hmong) kæ 4

On the basis of the Mienic evidence, we can be confident that the original word for ‘hand’
had a labial initial (in this case, because of the low-register tone, a *b-). The Pa-Hng form
shows that the absorption of different prefixes by various languages of the Hmongic family
is most likely to blame for the difficulty of reconstructing an initial for this word which
reflects the phonetic values of the initials in all the daughter languages (Wang first
reconstructed *bʑ- and later *bdz- for Proto Hmongic, and *bwdz- for Proto Hmong-Mien,
an approach which involves ignoring some initials and simply combining others).
Significantly, body-part words are among those nouns in Hmong-Mien which appear with
204 Chapter 4

a prefix most consistently (Chen 1993). But in White Hmong while the isolation form for
‘foot’, ko-taw /ko5-taɯ5/, does have a prefix, the isolation form for ‘hand’, tes /te4/, does
not—not because it never did, but because it absorbed its prefix.
Although ‘hand’ is unique in the complexity of its initial correspondence, ‘thorn’
(Hmongic *-boB 1.3/7) is similar to ‘hand’ in that the set is basically labial, but shows a t-
initial in Jiwei and a k- initial in Ho Ne:

(7) hand thorn

Yanghao (Hmu) pi4 pə4
Jiwei (Qo Xiong) tɯ4 to4
Xianjin (Hmong) te4 po4
Shimen (A-Hmao) di4 bo4I
Qingyan (Hmong) tsa4 pau4
Gaopo (Hmong) kæ4 pə4
Zhongdi (Mang) ʂe4 pɯ4
Fuyuan (Hmjo) weiB vuB
Fengxiang (Hmong) pi4 --
Baiyun (Pa-Hng) tɛ-pɯ4 --
Duozhu (Ho Ne) khwa4 khɯ4

Prefix pre-emption can explain how (if not why) these irregular correspondences
developed: sporadically, different prefixes were absorbed in different languages. When
prefix pre-emption has occurred, we should not expect the resulting set of initials for one
etymon to be duplicated exactly in the set for any other etymon.

4.2.4 Nominal prefixes and Proto Hmong-Mien disyllabism

Despite the arguments for the antiquity of the prefix-noun construction, there may still
remain a question about how old the prefix-noun structure actually is in the Hmong-Mien
family, and whether or not it is appropriate to reconstruct a nominal prefix position for the
protolanguage. It could be argued that this is a fairly shallow development, the result of a
new turn toward compounding, supported by the fact that prefixes do not appear
consistently in all of the languages of the family. But most hypotheses about the
reconstruction of a particular construction are informed by a body of evidence that
supports a general picture of the typological cast of the language at an earlier stage.
Independent lines of evidence all suggest that Proto Hmong-Mien was not as isolating as
most Hmong-Mien languages are today. The reconstruction of prefix variation within
disyllabic nouns in the family is part of a larger picture of early disyllabism across
syntactic categories that is supported by research on a number of different fronts. One
pattern that suggests disyllabism is discussed in section 4.4 below: the reconstruction of
old prefixes based on tonal doublets. Two other patterns are briefly addressed here: (1) the
correspondence of words with stop-liquid clusters to Pa-Hng disyllabic words, and (2) the
resemblance of certain words with unusual initial correspondences to disyllabic words in
other language families.
 Morphology 205

The interpretation of stop-liquid consonant clusters

First, as mentioned in the introduction to the reconstruction (chapter 2, section 2.1.2.3) the
reconstruction of stop–liquid clusters is difficult in that there appear to be more
correspondences than there are protolanguage stop–liquid cluster candidates (Ratliff 1995,
Solnit 1996, Neiderer 1999). Solnit presents a good methodological discussion of the
different logically possible solutions for the problematic dorsal/liquid cluster
correspondences in Hmong-Mien: they include increasing the number of liquids,
increasing the number of dorsals, or increasing the word-shape inventory to include
disyllabics. The solution chosen here (and in Ratliff 1995, 1999b) is that there were both
“tight” and “loose” consonant clusters in the protolanguage, that is, structures both of the
shape CCV and C-CV, the second not being clearly distinguishable from a disyllabic
structure. This accounts for the sporadic appearance of the first C in the “loose” cluster
correspondences, and fits in well with the areal “syllable-and-a-half” template best
exemplified by Mon-Khmer and Tibeto-Burman languages. Sagart has proposed a similar
picture in his discussion of Old Chinese word-families, where he distinguishes bare roots
both from roots with “fused” prefixes and from roots with “loosely attached” prefixes: “I
assume that each prefix ... existed in these two shapes, fused and loosely attached, and that
the three types of forms existed side by side in Old Chinese, perhaps as stylistic or social
variants” (Sagart 1999:15).
In Hmong-Mien, even words which seem to belong to the “tight cluster” (CCV)
category can show signs of loosening, which suggests the same type of variation described
by Sagart for Chinese may be at work in this family. For example, Niederer (2004:140) has
presented the following three disyllabic forms from Pa-Hng, shown here in contrast to
monosyllabic forms in other languages of the family:

(8) ‘dog’ (Hmong-Mien *qluwX (6.31/3) > White Hmong dev /ʔde3/)
Jiwei (Qo Xiong) qwɯ3 Baiyun (Pa-Hng) ta1-ljɔ̃7
Xianjin (Hmong) tl ̥e3 Tan Trinh (Pa-Hng) ka/ta1-ljɔ̃7
Fuyuan (Hmjo) qlei B

Liangzi (Mun) klu3

‘maggot’ (Hmongic *klaŋC (5.31/24) > White Hmong kas /ka7/)

Yanghao (Hmu) kaŋ5 Baiyun (Pa-Hng) kɛ1-lõ7
Shimen (A-Hmao) kaɯ7 Tan Trinh (Pa-Hng) kɤ1-lõ7
(see cognate /klaŋ1/ ‘insect’ in Dongshan Biao Min)

‘round’ (Hmong-Mien *ɢlun 6.33/27)67

Yanghao (Hmu) l ̥en2 Baiyun (Pa-Hng) ka1-lɔ̃4
Xuyong (Hmong) khun Tan Trinh (Pa-Hng) ka1-lɔ̃4
2

Dongshan (Biao Min) klin2

About these words Niederer (2004:140–141) writes, “The above cited examples…are not
complete enough to allow us to speak of regular correspondences. Yet they remain highly
suggestive, urging us to look beyond monosyllabism, as the Pa-Hng forms point to ancient
disyllabics in many cases”. It should be noted that whether the discussion of nominal
67
In the case of ‘round’, given the possibility that the root is a loanword from Chinese 輪 ‘wheel; round’
(MC lwin > Man. lún), it is hard to know if Pa-Hng shows a loosened cluster or a prefix.
206 Chapter 4

prefixes should be linked to the issue of consonant clusters, be they loose or tight, is not
clear. Since all syntactic categories are represented in these correspondence sets, consonant
cluster variation may be a reflection of something different, and more general.

Connections to external disyllabics

Finally, there is evidence for early disyllabism in the nature of particular correspondences
which can best be understood by comparing these words to structurally and semantically
similar words in other language families of the area. For example, there is the well-known
example of the Hmong-Mien word for ‘bird’ *m-nɔk (2.9/29), which bears an uncanny
resemblance to Proto Malayo-Polynesian *manuk (ACD). A ‘loose’ cluster is suggested
for Proto Hmong-Mien by the distribution of the initials: the initial in cognates from all
languages is a coronal nasal—the root nasal—with the exception of a labial in Pa-Hng
/mo6/ (see a discussion of this word in the context of an exchange on Pa-Hng (or Na-e)
archaisms and the place of this language in the family in Benedict 1986 and Strecker
1987a,b).
Another example of a connection between an unusual Hmong-Mien correspondence and
a word from another family is ‘centipede’—*khluwD (5.32/9) in Hmongic and *səpD
(2.13/7) in Mienic. The sum of these two reconstructions resembles Mon-Khmer *k-əl-ʔeːp
‘centipede’ (Diffloth 1994).
Of course, it is not clear whether these borrowed (or possibly inherited) disyllabic
words were ever disyllabic within Hmong-Mien. In the case of ‘centipede’, at first glance
it looks like Hmongic borrowed the first syllable while Mienic borrowed the second
syllable. But tone 7 in Hmongic, attested by the Yanghao (Hmu) form /khu7/, regularly
derives from a syllable with a final -p or -t. The tone correspondence with Mienic, where
final -p is preserved, suggests that the whole disyllabic word was borrowed by Proto
Hmong-Mien. The survival of elements from both syllables in both of these cases, and in a
number of others, leaves open the possibility that the words were originally disyllabic in
Hmong-Mien and reduced to monosyllabic forms at a later date.
In conclusion, it seems quite reasonable to think that classifying prefixes served the
function of noun classification in the protolanguage much as they do now in certain
conservative languages of the family. Cross-linguistic prefix variation, however, makes it
impossible to reconstruct particular prefixes and associate them with particular nouns. The
variation we see in the modern languages was probably characteristic of the protolanguage
as well. It seems most prudent to reconstruct a classifying prefix position for nouns,
without specifying its contents, against an evolving picture of ancient disyllabism.

4.3 Prenasalization
Prenasalization, a hallmark feature of the phonological inventories of most Hmongic
languages, suggests an element fused to the onset, and thus either (1) a relic of an ancient
disyllabic word (see discussion above), or (2) a remnant of old morphology. However,
with regard to the first possibility, although it is reasonable to think a prenasalized
obstruent might have been formed as a result of the collapse of the first light syllable of a
disyllabic word (with the nasal either in the onset or, more likely, in the coda of the first
syllable), there is simply no evidence to show that this occurred. Disyllabic forms with a
nasal in the onset of the first syllable have been reconstructed here only in the cases where
the first nasal replaces the nasal in the onset of the second syllable: *m-nɔk ‘bird’ (2.9/29),
 Morphology 207

*n-mɛj ‘to have’ (1.9/11), *n-mʉŋ(X) ‘to go’ (1.9/30). And with regard to the second
possibility, prenasalization plays a morphological role only in a few Chinese loanword
pairs in Hmong-Mien. In the native component of the lexicon, no family-wide contrasts
pointing to an ancient morphological role for prenasalization have yet been found.
The Chinese loanwords that seem to show a morphological role for this nasal element
are nonetheless quite interesting. In the two pairs below, the semantic relationship between
the members of each pair is quite different, and the two do not seem to signal the presence
of the same morpheme:
(9) 下 HM *GaX ‘low/short’ (5-6.3/4) vs.
H *ɴɢaB (<*NGaX) ‘to descend’ (5-6.6/4); M *ɣaC ‘to descend’ (5.13/4)
清 HM *tshji ̯əŋ ‘new’ (3.17/18) vs. HM *ntshji ̯əŋ ‘clear’ (3.20/18)
Downer (1973:14–16) also discusses two pairs of Mien words, both pairs of Chinese
loanwords, in which he sees a morphological role for prenasalization reflected in the
voiced initial of the second member of each pair:
(10) 拆 /tshɛʔ7/ ‘to pull down, pull apart’ vs. /dzɛʔ7/ ‘to be cracked’
開 /khɔi1/ ‘to open (TR)’ ( < M *khu̯ɔiA, 5.2/11) vs. /gɔi1/ ‘to open (INTR)’
He writes “Although Miao cognates for these pairs of words do not apparently exist, it
seems likely that we must suppose a prefix (pre-nasalization) in these cases, with PY
[Proto Yao] *khɔ̄i, *tshɛ́k as active verbs, *nkhɔ̄i, *ntshɛ́k as intransitive verbs; and that
since these are found in words of Chinese origin, we must assume that this prefixation was
still an active process at the time of borrowing” (16). He thus attributes the prenasalized
form to a native derivation process within Hmong-Mien, although as we will see below,
there is no independent evidence to support the reconstruction of a native Hmong-Mien
nasal prefix that is preserved as prenasalization. It seems more likely that this
morphological contrast was borrowed from Chinese along with the words themselves.

Elsewhere, prenasalization seems to be phonological. For example, in the White Hmong

pair
(11) kaum [kau8] ‘ten’ vs. nees nkaum [nɛŋ4/7 ŋkau8] ‘twenty’
the nasal element in nkaum is a close juncture carry-over from the nasal coda of the
preceding word nees. A similar effect can be seen in the White Hmong xeeb-ntxwv
‘descendants’ borrowed from Chinese 孫子 Man. sūnzi ‘grandson’ (the second morpheme
is the Chinese nominalizer). Additional evidence comes from loanword phonology: for
example, White Hmong speakers use prenasalized stops when borrowing words from
languages with voiced stops (thus White Hmong npias ‘beer’ from French bière via Lao
/bìa/). In some languages, it may thus be appropriate to analyze prenasalized obstruents as
simply fully voiced obstruents, or lenis as opposed to fortis obstruents.
The correspondences involving prenasalization in words that can be reconstructed to
Proto Hmong-Mien fall into one of two main types. The most common pattern is for
prenasalized stop and affricate initials on the Hmongic side to correspond to voiced stop
and affricate initials on the Mienic side. For example:
208 Chapter 4

Yanghao Jiwei WHmong Mien Hmong-Mien

cogongrass qɛ1 -- ɴqɛŋ1 gaːn1 *NKan (5-6.4/19)
salt ɕi3 ntɕɯ3 ɳtʂe3 dau3 *ntsjəuX (3.19/3)
fish zɛ4 mʐɯ4 ɳtʂe4 bjau4 *mbrəuX (1.51/3)
to dream pu5 mpei5 mpau5 bei5 *mpeiH (1.4/12)
to fear/be afraid ɕhi1 ȵʨha5 ɳtʂhai5 ʥa5 *ntshjeH (3.20/10)
name pi5 mpu5 mpe5 bu5 *mpɔuH (1.4/3)
put on/wear (cap) tə5 ntu5 ntɔŋ5 doŋ5 *ntɔŋH (2.4/29)

In these sets, representative of many more, no semantically related pairs with and without
prenasalization have been found, nor is there any semantic coherence to the prenasalized
words. It is also clear from the short list of examples above that words with prenasalization
do not fall into any particular word class.
In a second minor pattern, voiceless stop and affricate initials on the Hmongic side
correspond to voiced stop and affricate initials on the Mienic side. For example:

Yanghao Jiwei WHmong Mien Hmong-Mien

deep to1 to1 tɔ1 do1 *N-tu̯o (2.1/7)
earth ta1 tɯ1 te1 dau1 *N-təu (2.1/3)
excrement qa3 qɑ3 qua3 dai3 *N-KəjX (5-6.1/15)
half (of day) taŋ4 -- ta4 daːm2 *N-dam(X) (2.3/24)
porcupine ɕen6 ʨi6 tʂau6 dei6 *N-dzjeiH (3.18/12)
wing ta7 tei7 ti7 daːt7 *N-tat (2.1/4)

No trace of ancient morphology can be read from this pattern either: again, there are no
related pairs with and without this loosely-adjoined voicing element, here taken to be a
nasal, nor does any semantic or grammatical feature link the words with pre-nasal
elements.

4.4 Traces of ancient affixes68

Apart from the nominal prefixes, traces of old affixal morphology in Hmong-Mien are not
realized as affixes in the modern day, nor are they usually realized in onset consonants.
Instead, they are deeply embedded, and are realized as tones on the nucleus of the word.

T
free morpheme  prefix  (C)C(C)V(C)  suffix  free morpheme

As grammatical elements coalesce, they first cause the outer layer to morph, the
consonants of the coda and onset. Since properties of these consonants play a role in
tonogenesis and tone split (see chapter 3), these consonant feature contrasts may transfer to
the nucleus as tonal contrasts during the process of tone development. The oldest
morphological contrasts are therefore to be discovered in the first instance by identifying
pairs of words that share both core meaning and segmental phonology, but differ in tone.

68
This section contains some material first presented in Ratliff 1986a and 1986c.
 Morphology 209

By undoing the effects of tonogenesis, one can theoretically take the tonal contrasts back to
their segmental origins.
In practice, it is very difficult to reconstruct affixes in this way. To be confident that any
particular Proto Hmong-Mien prefix (or suffix) has been identified, three requirements
need to be met: (1) a number of word pairs must exemplify the relationship, (2) a
consistent form/meaning relationship must hold across these pairs, and (3) evidence from
both branches of the family must support the reconstruction. Given these requirements, it is
not yet possible to identify a single bound morpheme of the protolanguage with confidence
(other than the nominal prefixes discussed in section 4.3 that have been continually
refreshed). However, tantalizing traces of early morphology exist. They will be described
here, in the hope that more work will yield the kind of evidence—especially in quantity—
that will make the reconstruction of one or two ancient bound morphemes possible.

4.4.1 Odd/even tonal pairs within the same tone category

The tones in the doublets presented in this section can be traced back to a single historical
tone that split in two upon the neutralization of an original voicing contrast in onset
obstruents into an upper register tone from a voiceless onset, and a lower register tone
from a voiced onset (see chapter 3). Working backward, it is most reasonable to assume
that this voicing contrast in the initial consonant was caused by either the voicing or de-
voicing influence of an ancient prefix.

Odd/even tone doublets with different meanings

The first two pairs suggest the existence of either a causative voiceless prefix or a stative
voiced prefix—or both:
(12) *təjH ‘to kill’ (2.1/15) vs. *dəjH ‘to die’ (2.3/15)
This doublet shows a striking similarity to the Austronesian pair *pa-aCay ‘to kill’ and
*ma-aCay ‘to die’(ACD). In fact, the correspondence between the two reconstructed forms
is so strong that it requires a historical explanation, either contact or common inheritance;
it cannot be due to chance.69
(13) *ʔ-mɛj (>*ʔmu̯ɛA) ‘to grasp’ (1.7/11) vs. *n-mɛj ‘to have’ (1.9/11)
The relationship between ‘to grasp’, with a voiceless initial, and ‘to have’ with a voiced
initial, parallels that of ‘to kill’ and ‘to die’ above. A voiceless causative prefix could
account for the former, and a voiced stative prefix for the latter. Unlike other words
reconstructed with *m- (onset 1.9), ‘to have’ shows an initial /n-/ in Mun, leading to the
reconstruction of a pre-nasal element in this word.70 These two words appear to share a
root with *mɛj-X ‘to buy’ (1.9/11) and *mɛj-H ‘to sell’ (1.9/11); the four constitute a set
of verbs of possession. ‘To buy’ and ‘to sell’ would have had suffixes which gave rise to
tones B and C respectively, or ‘to sell’ would have been derived from ‘to buy’ by C-tone
derivation (see section 4.4.2 below). Since these words are the familiar words for ‘to buy’

69
With respect to ‘to kill’ and ‘to die‘ the connection between Austronesian and Hmong-Mien is stronger
than that between Austronesian and Tai-Kadai: Proto Tai has *tɑːi ‘to die’ (with a voiceless rather than a
voiced initial), and no match for ‘to kill’, whereas Hmong-Mien has both members of the pair, with traces
of both prefixes.
70
‘To go’ patterns the same way, and is also reconstructed with a pre-initial nasal: *n-mʉŋ(X) (1.9/30).
210 Chapter 4

and ‘to sell’ in Chinese, it has always been reasonably assumed that these were loanwords
from Chinese. But the relationship of these words to ‘to take’ and ‘to have’ in Hmong-
Mien complicates the picture of a straightforward loan of the single pair ‘to buy’ and ‘to
sell’ from Chinese. If one accepts that the four words are built on the same root, we will
need more evidence to show which of the following is correct: (1) Hmong-Mien borrowed
all four from Chinese (and two were subsequently lost in Chinese), (2) Chinese borrowed
only two of the four from Hmong-Mien, or (3) ‘to buy’ and ‘to sell’ are in both language
families as a consequence of common inheritance.
The third pair of words illustrates a part-whole relationship: the only example of its type
to be found so far:
(14) *cæwB ‘body/trunk’ (4.1/3) vs. *ɟæwB ‘leg/branch’ (4.3/3)
Of course, on the basis of one example, it is not possible to draw any conclusions about the
nature of the prefix that caused the voicing change in one or the other member of this pair.
Doublets that involved a voiceless/voiced contrast also existed in Old Chinese (Norman
1988:85). Given the high number of Chinese loanwords in Hmong-Mien, it is therefore not
surprising to see two loanword doublets in Hmongic that illustrate the same relationship:
(15) 炙 *ciC ‘to bake/toast’ (4.1/1) vs. *ɟiC ‘to burn/be alight’ (4.3/1)

(16) 著 *trɔC ‘to put on/wear (shoes)’ (2.46/6) vs. *drɔC ‘to hit target’ (2.48/6)
The second loanword doublet appears odd until one looks at the range of meanings
represented by the Chinese character 著 and their sources: (1) Old Chinese *trak > Middle
Chinese trjak ‘to place, put, apply; wear’ > Man. zhuó, and (2) Old Chinese *[d]rak >
Middle Chinese drjak ‘to touch, be attached’ > Man. zhuó. The phonology provides a good
match as well: tone C in Hmongic derives either from tone C in the protolanguage, or from
a syllable ending in final *-k.
Finally, doublets of the same kind have been found within White Hmong, both
involving splits of the A-tone. The first is a native word and the second is borrowed from
Chinese:
(17) hlob /hl ̥ɔ1/ ‘to become big (to grow)’
vs. loj /lɔ2/ ‘to be big’ (HM *hljo ‘big’, 2.41.1/7)
(18) 正 ncab /nca1/ ‘to straighten’
vs. ncaj /nca2/ ‘to be straight’ (H *ɲci̯ aŋA ‘straight’, 4.4/26)
Both roots are attested widely with upper-register tones; see the reconstructions given
above for the meanings ‘big’ and ‘straight’. The words with lower-register tones in White
Hmong now have these basic meanings, and the words with upper-register tones now have
an active (achievement or causative) sense. It is difficult to know what to make of pairs
that appear in only one language: one could say that internal reconstruction points to the
same analysis, although one would ideally like to see support from related languages.71 It is
not clear whether these pairs preserve an old contrast lost elsewhere, or represent a local
innovation.

71
Xianjin Hmong, a close relative of White Hmong, also has /ȵʨaŋ2/ ‘straight’ with tone 2, but there is no
evidence that Xianjin has the causative verb in tone 1 as well.
 Morphology 211

Odd/even tone doublets with the same meaning

Kun Chang (1972:563) noted that a few words display both upper and lower register tone
reflexes, which he also found suggestive of ancient prefixes that either devoiced one
member or the pair or voiced the other. In the cases he presents, the original meaning
contrasts represented by the reconstructed voicing contrasts can only be tentatively
proposed (he does not attempt to do this) since the meanings of the words in all dialects are
now the same.72 But given the pairs ‘kill’/’die’ and ‘to grasp’/’to have’ above, two of the
four pairs could have involved the same hypothetical voiceless active/causative (or voiced
stative) prefix. For what is given here as *(h)naŋX ‘to put on/wear (clothes)’ (2.8/24), with
the “h” in parentheses to indicate that there are both tone 3 and tone 4 reflexes of the same
original tone category (see chapter 3), we could reconstruct the following contrast:
(19) *hnaŋX ‘to dress’ vs. *naŋX ‘to wear’ < *s-naŋX ‘to dress’ vs. *naŋX ‘to wear’
Similarly, Chang notes that certain East Hmongic dialects have lower register tones for
what is here reconstructed as *pʉŋA ‘to fall’ (1.1/30). Just as for ‘to dress’ (high register,
causative) vs. ‘to wear’ (low register, stative), we might instead reconstruct the following
pair of words:
(20) *pʉŋA ‘to fell’ vs. *bʉŋA ‘to fall’ < *s-bʉŋA ‘to fell’ vs. *bʉŋA ‘to fall’
In the case of ‘to put on/wear,’ tones 3 and 4 are fairly equally represented in modern
dialects. In the case of ‘to fall’, however, far more dialects have tone 1, the reflex of the
voiceless initial, which runs counter to expectations under this hypothesis. Since the
meaning of the word is clearly intransitive today, if this were right we would expect ‘to
fall’ to have more reflexes with tone 2, from the voiced initial.
Another possible pair can be extracted from tone 5 and tone 6 reflexes of Hmong-Mien
*(S-)mru̯ɔŋH (1.54/29) ‘to listen’:
(21) M *s-mru̯ əŋC ‘to listen’ vs. H *mrɔŋC ‘to hear’
It should be emphasized that the reconstruction of meanings in examples 8–10 is highly
speculative, and is based solely on the upper-register/lower-register tone split in these
words, coupled with an analogical extension of the semantic contrast between members of
pairs with an attested difference in meaning.

4.4.2 Derivation by tone change

Another pattern of tonal contrasts evidenced by some intriguing pairs of words involves
the C-tone (from syllables ending in a final -h < *-s) and either the A-tone or the B-tone.
C-tone derivation has been described as a morphological process for Chinese (Downer
1959) and for Tai (Li 1970, Manomaivibool 1980); given the position of Hmong-Mien in
the Sinosphere, it is reasonable to think these pairs of words reflect the same process. It is
not at all clear, however, whether the C-tone word was derived from a word of another
tone through a true tonal derivation process, or whether this took place earlier, before the
syllable-final consonants that gave rise to the tones disappeared, in which case these pairs
of words would reflect an old *-s suffix.
72
The two words not discussed here are nouns: ‘field’ (White Hmong loo ‘crops’) and ‘tongue’. Although a
prefix might account for the difference between the tone 5 and 6 reflexes listed for ‘field’, it is not easy to
imagine what it could have been; the single tone 7 form for the tone 8 word ‘tongue’ Chang reports could
simply be an irregularity.
212 Chapter 4

Again, the evidence for this morphological process is slight, and the function of the
derivation process is not clear. Three pairs of words suggest that verbs were derived from
nouns in this manner (the second member of each pair belongs to the C-tone category,
marked here by the capital letter H; see chapter 3):
(22) Hmongic *grəŋA (< HM *grəun) ‘animal fat/oil’ (5.48/21) vs. *grəunH ‘to be fat’
(5.48/21)

(23) *pji̯ əuX ‘fruit’ (1.16/1) vs. Hmongic *pjiC (< HM *pji̯ əuH) ‘to bear fruit’ (1.16/1)

(24) *douX ‘fire’ (2.3/13) vs. *douH ‘to explode’ (2.3/13)73

Note that in the case of ‘animal fat/oil’ > ‘to be fat’, the derivative ‘to be fat’ is now more
widespread: the noun is attested only in Hmongic languages. In the case of ‘fruit’ > ‘to
bear fruit’ the situation is the opposite: the noun is widespread, and the derivative ‘to bear
fruit’ is only attested in Hmongic.
A fourth pair suggests that a noun with a narrowed sense could be derived from another
noun in the same way (the derivative is attested only in Hmongic):
(25) *klæŋ ‘insect/worm’ (5.31/24) vs. Hmongic *klaŋC (< HM *klæŋH) ‘maggot’
(5.31/24)
A final pair involves verbs that can be reconstructed to Proto Hmongic:
(26) Hmongic *ʔɲəŋA ‘to be at/live (4.7/21) vs. Hmongic *jəŋC ‘to be’ (4.12/21)
This pair is not an obvious one, because in Hmongic ‘to be at/live’ has a voiceless palatal
nasal onset and ‘to be’ has a voiced palatal glide onset. However, there is variation
between palatal nasal and palatal glide onsets within both sets of forms (see not only the
related M *ʔjəmA ‘to be at/live (4.10/21), but also the White Hmong triplet nyob ‘to be
at/live’, nyog ‘to be right/suitable’, yog ‘to be/yes’). Additionally, the semantic relationship
between these two words is not clear: although both words can be subsumed under our
notion of ‘to be’, the C-tone word is not an obvious semantic derivative of the A-tone
word.

73
Another word may belong to the family represented by the third pair: Hmongic *towD ‘to ignite/light’
(2.1/13), from a hypothesized Hmong-Mien form *touC (where “C” represents some stop consonant).
The causative prefix discussed in section 4.4 could have devoiced the initial *d- of the root. The tone,
however, indicates that the word ended in either a *-p or a *-t, which would have to be explained.
5 Numerals, personal pronouns, and
demonstratives

5.1 Introduction
The reconstruction of closed classes of words—such as numerals, pronouns, and
demonstratives—is of interest to historical linguists because these words are usually more
resistant to borrowing, and can hence give better quality evidence about distant
relationships. To the extent that particular words in these classes are borrowed, however,
they may also contribute to a more refined understanding of what should count as ‘core
vocabulary’, if such a universal set of concepts can be identified. In Hmong-Mien, the
class of numerals consists primarily of borrowed words, whereas the pronoun and
demonstrative classes consist primarily of native words.
In addition to admitting borrowed elements, these three ‘closed’ classes in Hmong-Mien
are slightly ‘open’ to one other, being conceptually and thus historically related: for
example, in several Hmongic languages the numeral ‘two’ is built into the dual pronoun
(‘you-two’ > 2DU), in some West Hmongic languages the 2SG pronoun is the base of the
‘near-hearer’ demonstrative (‘you’ > ‘that near you’), and in White Hmong a
demonstrative is the base of the 3PL pronoun (‘there’ > ‘those over there’).

5.2 Numerals
The following are the oldest reconstructable forms of the numerals 1-10, 100, and 1,000 in
Hmong-Mien. Source language (or language group) and proto-forms are given for
borrowed words. The Chinese forms are from on-going work by Baxter and Sagart (2009)
and the Tibeto-Burman forms are from Matisoff 2003 ([j] has been substituted for [y] in
this source to allow for easier comparison).

HM Source form Source Onset/Rime

one *ʔɨ *ʔi[t] Old Chinese (一) 7.1/2
two *ʔu̯i (none) (none) 7.1/14
three *pjɔu (none) (none) 1.16/3
four *plei *-ləy Tibeto-Burman 1.31/12
five *pra *-ŋja Tibeto-Burman 1.46/4
six *kruk *k-ruk Tibeto-Burman 5.46/9
seven *ŋjiC (M) *ni Tibeto-Burman 5.24/1
*djuŋH (HM) 2.18/28
eight *jat *-rjat Tibeto-Burman 4.12/4
nine *N-ɟuə *gəw Tibeto-Burman 4.3/16
213
214 Chapter 5

ten *gju̯ɛp *g(j)ip Tibeto-Burman 5.18/9

*[g][i]p or Old Chinese (十)
hundred *pæk pæk Middle Chinese (百) 1.1/5
thousand *tshi ̯en tshen Middle Chinese (千) 3.2/20

Native numerals
It has been known for a long time that most lower numerals in Hmong-Mien were
borrowed from Tibeto-Burman (Downer 1971; Benedict 1987a; Dempsey 1995; Peiros
1998; Mortensen 2002), while the higher numerals in Hmong-Mien were borrowed from
Chinese. The native core of the system is therefore very small, consisting of only ‘two’ and
‘three’, and perhaps ‘one’.

Two: *ʔu̯i. This is a stable word in Hmong-Mien, although it presents a unique set of
rime reflexes which are presented as a ‘correspondence’ both here and in Wang 1994
(rime 14), for want of a place to put it. Initial /v-/ in Pa Hng and some Mienic
languages appears to be a fortition of *ʔu̯- .

Three: *pjɔu. On the Hmongic side of the family ‘three’ is identical to the 1PL pronoun
in many languages (see section 5.3), leading one to hypothesize that these words are
cognate, which is especially sensible in a language with dual pronouns (‘I’, ‘we-two’,
‘we-three or more’). Furthermore, the meaning of the Mienic 1PL pronoun is not
strictly ‘we’, but may mean more generally ‘group’, and as such can enter into second-
and third-person plural compounds (e.g., Mien /mwei2-bwo1/ ‘thou-group’, which is
certainly not ‘thou-we’). But the Hmongic reflexes suggest a medial *-j- in ‘three’, and
the voiced initials in Mien and Mun suggest a pre-nasal element in Mienic in ‘we’—
‘three’ is reconstructed here as *pjɔu and ‘we’ as *N-pɔu.74 Although these words may
ultimately be related, it is not possible to demonstrate the relationship at this time.75

Tibeto-Burman numerals > Hmong-Mien

As evident from the table above, ‘four’ (with a pre-initial *p- ), ‘six’, Mienic ‘seven’,
‘eight’, and ‘ten’ are very similar to the reconstructed Tibeto-Burman forms, and pose no
particular difficulty.

Nine: *N-ɟuə. ‘Nine’ is somewhat difficult, though: Benedict (1987a:14) reconstructs

Hmong-Mien *gj[ou]A, which he derives from Tibeto-Burman *d-gəw via a ‘Donor-
Miao-Yao’ Tibeto-Burman form *-g(j)u. This seems reasonable, but there are two
problems related to the initial /d-/ in Mien and Mun forms for ‘nine’: its place of
articulation (explained if this is a case of prefix pre-emption, see chapter 4, section
2.3), and the fact that Mienic /d-/ is voiced but not matched by prenasalization in
Hmongic, which suggests a pre-initial *N- (see chapter 2, section 1). However, there is
no direct evidence that the pre-initial that causes exceptional voicing in Mienic is a
nasal: a voiced stop would also work. This detail aside, the Hmong-Mien word is
certainly the same as the Tibeto-Burman word and its Chinese cognate 九 (OC
*(tə.)[k](w)uʔ > MC kjuwX > Man. jiǔ).

74
Another problem is that the Mun of Liangzi Township, Hekou Yao A. C., Yunnan also shows a higher
vowel in ‘we’ /bu1/ than in ‘three’ /po1/ (Wang and Mao 1995:113).
75
Hmong-Mien ‘three’ also shows a resemblance to the Mon-Khmer word for ‘three’: *piʔ (Shorto #98).
 Numerals, personal pronouns, demonstratives 215

The two other numerals in the run from ‘four’ to ‘ten’—‘five’ and the Hmongic forms for
‘seven’—are more difficult to explain as Tibeto-Burman borrowings. However, given a
certain degree of resemblance, Benedict (1987a:13-14) reasonably sought to link these two
to Tibeto-Burman as well, given the likelihood that the Hmong-Mien people borrowed
these numerals as a set. His explanations of how the Hmong-Mien words may be linked to
Tibeto-Burman through his hypothetical Tibeto-Burman ‘Donor-Miao-Yao’ language are
presented below. Although the larger picture is plausible, the details of these explanations
are less persuasive.

Five: *pra. Benedict (1987a:13) writes that Tibeto-Burman *r-ŋa (Matisoff 2003 *ŋa)
alternated with *b-ŋa. In the ‘Donor-Miao-Yao’ Tibeto-Burman language, instead of
replacing the *r-, the labial was added to it. The *-r- then pre-empted the root-initial
nasal (as it has done in attested cases), giving *b-ra. The prefix then devoiced to *pra.
Mortensen (2002:4–5) notes that despite the problems with the initial consonants, the
rime correspondence is the same as that between Tibeto-Burman *s-la and Hmong-
Mien *hlaH ‘moon/month’ (see chapter 7).

Seven: Hmong-Mien *djuŋH; Mienic *ŋjiC. Benedict (1987a:13-14) writes that the
source of both Mienic ‘seven’ and Hmongic ‘seven’ was Tibeto-Burman *snis (the
reconstructed root *ni from Matisoff 2003 is given in the table above). To get from
Tibeto-Burman *snis to his ‘Donor Miao-Yao’ form *zńis, Benedict appeals to a
voicing rule and an areal tendency to palatalization. But to get from this intermediary
form *zńis, through his reconstructed Hmong-Mien form *zń[ia]C to something like
the Proto Hmong-Mien *djuŋH of the present reconstruction is quite difficult.
Benedict posits a secondary nasalization of the final, and then loss of the initial nasal—
but we expect Hmong-Mien initial nasals to persist in Hmongic. Furthermore, his rime
reconstruction cannot account for the rounded vowels in Hmongic: e.g. Mhu /ɕoŋ6/, Qo
Xiong /tɕoŋ6/, Ho Ne /tshuŋ6/, etc. (rime set 28). One other word shows this kind of
irregular correspondence: Hmong-Mien *hɲu̯əŋH ‘year’ (4.8/29) and Hmongic *ɕɔŋC
‘year’ (4.13/29, White Hmong xyoo), both probably from Tibeto-Burman *s-niŋ ‘year’
(Mortensen 2002). The most likely explanation for these difficult correspondences is
that different forms of the same word were borrowed separately in the two branches of
the family, perhaps from two different immediate sources; we should therefore not
attempt to reconstruct a single form for Hmong-Mien.

Chinese numerals > Hmong-Mien

One: *ʔɨ. Although Mien has native words for ‘two’ and ‘three’, it has clearly
borrowed Chinese 一 ‘one’: the form is /jet8/, compare Old Chinese *ʔi[t] > Middle
Chinese ʔjit > Mandarin yī. The tone 1 forms elsewhere in the family, however, were
taken to be native Hmong-Mien by Benedict (1975, 1987a), or rather, native Hmong-
Mien as a sub-family of Austro-Tai.
‘One’ is a special numeral, in that it is often shows connections to word classes and
concepts beyond the numeral set. Matisoff (1997:17) writes ‘Its high frequency
encourages morphophonemic irregularity, and idiosyncratic fusions with other
morphemes. (Compare the multiple English allofams [members of the same word
family] which all descend somehow from PIE *oino-: one, an, once, only, alone, anon,
onion, eleven…).’ It is not necessarily the case that natural semantic developments
should lead to the reinterpretation of the old word for ‘one’ and the need to borrow a
new word to take its place (this has not happened in English), but it is possible to
imagine that this might have happened at an older stage in Hmong-Mien as it later
216 Chapter 5

happened in Mien, and that for this purpose bilingual speakers might have easily
adopted the Chinese word well known to them. ‘One’ in Hmong-Mien also shows
phonological irregularity across the family, unlike the phonologically more stable ‘two’
and ‘three’,76 and the nature of the irregularity also supports the hypothesis that Chinese
and Hmong-Mien share this word77. The conservative languages Pa-Hng and Qo
Xiong (North Hmongic), along with Jiongnai, have forms with tone B instead of the
tone A which appears more widely. Recall from chapter 3 that tone B derives from
syllables with a final glottal stop in the protolanguage. Given that Pa-Hng and Qo
Xiong occasionally conserve features from the back of the syllable, including distinct
vowel reflexes for Hmong-Mien closed syllables (see chapter 2, section 1), these B-
tone forms may reflect the final *-t of Chinese. If the final glottal stop of tone B
syllables derived from an original voiceless stop, we would expect—if any Hmongic
languages showed tone B reflexes for a Chinese loanword with a final stop—that they
would be precisely these languages.

Ten: *gju̯ɛp. The source of Hmong-Mien ‘ten’ is either Tibeto-Burman *g(j)ip or Old
Chinese *[g][i]p (> MC dzyip > Man. shí). Situated as it is between the lower
numerals ‘four’ through ‘nine’, which are borrowed from Tibeto-Burman, and the
higher numerals, which are from borrowed from Chinese, it is hard to determine which
of the two is the source.

Hundred: *pæk. The source of the Chinese borrowing for ‘hundred’ was Middle
Chinese (c. 500 CE) rather than Old Chinese (c. 1500 BCE), given Baxter and Sagart’s
reconstruction of a medial *-r- for Old Chinese. There is no trace of a medial -r- among
the Hmong-Mien reflexes of this borrowed word, and the Hmong-Mien and Middle
Chinese rimes correspond well.

Thousand: *tshi ̯en. A Middle Chinese source tshen (> Man. qiān) poses no difficulty.

In addition to these individual Chinese borrowings which serve as primary numerals in

Hmong-Mien, many languages retain a secondary set of numerals ‘two’ through ‘nine’
borrowed from Chinese for use in compound forms (either the ‘10 + 1’-type combinations
for eleven through nineteen or the ‘1 x 10 (+ 1)’-type combinations for twenty and above).
In Mien, two full sets of numerals, the native-plus-Tibeto-Burman set and the Chinese set,
are used side-by-side in a systematic way (see Purnell 1968, forthcoming), whereas in
White Hmong, all that is left of the parallel Chinese system is the word for ‘two’ which
appears in the combinations involving ‘twenty’: ‘ten’ is /kau8/ and ‘twenty’ is /nɛŋ4-
ŋkau8/78 where /nɛŋ4/ derives from Chinese 二 (MC nyijH > Man. èr).79 Of particular

76
The low back vowel reflex in Qo Xiong /ɑ3/ is not a problem: it is a regular development from the
intermediate Proto Hmongic form *ʔɨ in that language.
77
The possibility also exists that this word—along with other basic vocabulary such as ‘tongue’ and ‘to
drink/smoke’—gives evidence of a very deep genetic relationship between Sino-Tibetan and Hmong-
Mien (see chapter 6, section 6.2.2).
78
See chapter 4, section 3 for a discussion of the sources of prenasalization, which in the case of /ŋkau8/ is
clearly phonological and not morphological.
79
The fact that /nɛŋ4/ is a borrowed Chinese word for ‘two’ is not evident from this form alone: the Hmong
word has a final nasal, and the tones do not correspond. But this form corresponds to other ‘second set’
words for ‘two’ across the family, and these words in turn are part of larger sets that are clearly Chinese
in origin.
 Numerals, personal pronouns, demonstratives 217

interest is the Hmongic language Jiongnai, which uses both sets of numerals, and also
exhibits very interesting allomorphy of numerals in compounds (Mao and Li 2001:299).

5.3 Personal pronouns80

Recent discussions of Southeast Asian pronoun systems (for example, Compton 1994)
have characterized these systems as relatively ‘open’ compared to pronoun systems in
other parts of the world, interacting with kinship terminology in complex ways and
admitting borrowings from other languages. Different registers may call for different
pronoun options for each person/number combination. Hmong-Mien pronoun systems do
not seem to be characterized by this kind of fluidity, perhaps as a reflection of the more
egalitarian nature of Hmong-Mien society. The relative stability of personal pronouns in
Hmong-Mien makes all of the singular personal pronouns, as well as the first person plural
pronoun, eligible for core vocabulary status and useful in reconstruction. The second and
third person plural forms and the dual forms typically extend the singular forms in
different ways in different languages, although there is a good candidate for an ancient
Hmong-Mien second person plural pronoun as well.
The following table shows only the most widely attested (and presumably the oldest)
root for each person/number combination. Three of the pronouns—1SG, 2SG and 2PL—
show a striking resemblance to Tai-Kadai and Austronesian pronouns, a fact noted by
Benedict (1975:202 ff.) in connection with his Austro-Tai hypothesis, which places
Hmong-Mien, Tai-Kadai, and Austronesian in one super-family.
Competing roots are noted in the text that follows (the numbers indicate the onset and
rime correspondence sets to which each reconstructed pronoun belongs):

Singular Plural
First *kɛŋB (5.1/22) < *ku- *N-pɔu (1.1/3)
Second *mu̯ei (1.9/8) *mi ̯əu (1.9/1)
Third *ni ̯æn(X) (2.9/1) --

First person singular: *kɛŋB. Three different roots for ‘I/me’ are found in Hmong-
Mien, but Hmongic *kɛŋB (White Hmong kuv) is the oldest of the roots for ‘I/me’,
since it appears on both sides of the family (in West Hmongic and in Biao Min /kəu3/,
/kɔ3/). Since it is not possible to place these forms into any existing Hmong-Mien rime
set, no Proto Hmong-Mien form is reconstructed. However, at the earliest stage it must
have had an initial *k- and a back rounded vowel, since *k- in words before other
vowels retracted to q- in Hmongic, and the *k- in this word did not retract (see chapter
2, section 1.2). A Hmong-Mien *ku-X deduced in this fashion resembles the Tai-Kadai
1SG pronoun as reconstructed for Proto Tai *k[əu]A (Li 1977) and Proto Kra *kuA

80
This section is based on a paper delivered at the 11th Annual Meeting of the Southeast Asian Linguistics
Society, Bangkok (Ratliff 2001b). The sources of specific language forms in this section and the next are
Office of Miao-Yao Research Lexicon 1987 (Yanghao, Jiwei, Xuyong, Shimen, Bunu, Mien, Biao Min);
Niederer 2001a (Chongpaoyao=Peitungnuo); Niederer 2001b (Pa-Hng, Nau-Klau); Edmondson 2001
(Na-Meo); Wang, ed. 1985 (Yanghao, Jiwei, Xianjin=Dananshan), Heimbach 1979 (White Hmong),
Lyman 1974 (Green Mong), Mao, Meng, and Zheng 1982 (Mien, Bunu, Lakkja), Mao and Meng 1986
(She=Ho-Ne); Shintani and Yang 1990 (Hainan Mun), Solnit 1982 (Biao Min), and Ostapirat 2000
(Gelao). Lakkja and Gelao are Tai-Kadai languages.
218 Chapter 5

(Ostapirat 2000). It also resembles Proto Malayo-Polynesian *-ku ‘my’ (ACD).81

Interestingly, a second Proto Kra root for ‘I/me’ based on Gelao forms is *ʔeA
(Ostapirat 2000), which bears a resemblance to another Hmong-Mien 1SG root, Proto
Mienic *ʔjaA (4.10/4).

First person plural: *N-pɔu. See the section above on the numeral ‘three’ *pjɔu (White
Hmong peb), which is similar to the first person plural pronoun *N-pɔu (White Hmong
peb). The original meaning of this pronoun was something like ‘group’—a meaning
that may provide a bridge to ‘three’, since it is used in combination with singular
pronouns to create plurals in a number of languages (see the sections on 2PL and 3PL
below). The function of the pre-initial voicing element is unknown. It is reconstructed
to account for Mien and Mun reflexes with the voiced initial /b-/ (see chapter 2, section
1.2.1 on pre-initials). The Biao Min 1PL pronoun /ȶa3/ is different, and resembles
Lakkja /ta2/ and Gelao /ta1/.

Second person singular: *mu̯ei. Reflexes of Hmong-Mien *mu̯ei appear in all branches
except West Hmongic, and it therefore appears to be the original 2SG pronoun. It bears
a resemblance to the second syllable of Proto Austronesian *Simu ‘second person
pronoun’ > Proto Malayo-Polynesian *imu (ACD). Some West Hmongic languages
use another root, *gʉA (5.3/8): Xianjin /kau2/, White Hmong koj /kɔ2/, Shimen /ʨy6/,
Bunu /kau2/, Nau-klau /kɯ2/.

Second person plural: *mi ̯əu. The main root is *mi ̯əu, which appears in Hainan Mun
as well as in Hmongic (White Hmong nej). The similarity between the Hmong-Mien
reconstructed 2SG and 2PL forms (*mu̯ei and *mi ̯əu) recalls the use of *-mu in AN as
a second person pronoun base, independent of number: the words that derive from AN
*Simu are variously 2SG and 2PL while AN *kamu is 2PL (ACD). It may be that the two
second person Hmong-Mien roots reflect one, number-neutral second person root in
Hmong-Mien as well. Shimen and Mien use the 2SG pronoun with a suffix for ‘group’
to make a 2PL pronoun: Shimen /mi6-tsao2/ and Mien /muei2(~mei2)-buo1/.

Third person singular: *ni ̯æn(X). Reflexes of Hmong-Mien *ni ̯æn(X) (White Hmong
nws) appear in all languages except Jiwei (North Hmongic), which has /wu3/. This is
somewhat surprising, since cross-linguistically 3SG is not a particularly stable
pronoun (neither the Swadesh 100-word list nor the Matisoff 200-word list
(1978:283–296) includes 3SG, but both include 1SG and 2SG). There are reflexes of
*ni ̯æn(X) in both tone 2 (in both Hmongic and Mienic) and tone 4 (widespread, but in
Hmongic only), so the original tone is unclear. Wang and Mao (1995) propose that
tone 2 is basic since it occurs in Hmongic languages that do not have tone sandhi, but
2 > 4 is not a typical tone change (see chapter 2, section 1.4), so 4 cannot be
explained as a sandhi tone in other languages.

Third person plural. It is not possible to reconstruct a 3PL pronoun for Hmong-Mien,
nor is it possible to reconstruct one for either Hmongic or Mienic separately. The
following is the most widely shared root:

81
Another Hmongic 1SG pronoun *wɛŋB (1.12/22) is in the same rime set as *kɛŋB (5.1/22). The two appear
in complementary distribution across Hmongic.
 Numerals, personal pronouns, demonstratives 219

Yanghao (East Hmongic) mɛ2

Bunu (West Hmongic) mu2
Nau-klau (West Hmongic) mbe2
Chang-pao-yao (West Hmongic) mu2

It is somewhat confusing that in these languages the 2SG, the 2PL, and these 3PL forms
all have an m- initial and the second tone. But they are not the same—compare
Yanghao /moŋ2/ ‘you (sg.)’ , /maŋ2/ ‘you (pl.)’, /mɛ2/ ‘they’.82
It is more common, however, to see the 3PL expressed by a phrase which includes the
3SG, for example:

Yanghao (East Hmongic) nen2-to4 ‘s/he-those’

Xianjin (West Hmongic) ȵi4-mpau8 ‘s/he-many’
Pa-Hng (Hmongic) nɯ4-mɯ4 ‘s/he’-?’
Ho-ne (Hmongic) nuŋ4-ne2 ‘s/he-?’
Mien nen2-bwo1 ‘s/he-group’

In addition, some languages employ opaque phrases: Jiwei /ʨi3-mji2/, Na-Meo /mi1-
ʔa1/, /na1-se1/ ‘others’. Other languages adapt words with related meanings for 3PL
pronouns: Green Hmong puab /pua1/ < ‘group’, White Hmong lawv /laɯ3/ < lawm
/laɯ8/ ‘(those) over there’ (Ratliff 1992a:123-125).

Dual pronouns
Most dual pronouns involve the numeral ‘two’ itself. It is therefore not possible to
reconstruct dual pronouns as separate items, although derivatives of ‘two’ and phrases
involving ‘two’ in reference to pairs of people have no doubt been used in the family for a
long time.
If there is a single form for the first person dual, it is either identical to the word for
‘two’, or it is a phonologically reduced form of ‘two’:83

Yanghao (East Hmongic) ʔo1 ‘two’, ‘we-two’

Xuyong (West Hmongic) ʔɑo1 ‘two’ > ʔɑ1 ‘we-two’
Bunu (West Hmongic) ʔau1 ‘two’ > ʔa1 ‘we-two’

The first person dual thus seems to be the most basic dual pronoun. If a phrase is used, it is
on the order of Xuyong ‘we-two-CLFperson’, Mien ‘we/group-two’.
Second and third person duals often have the segments of the corresponding plurals
(occasionally reduced) with the tone of the numeral ‘two’ (tone 1), suggesting that these
are blends, wherein each word of an original phrase (‘you-two’, ‘they-two’) has
contributed an element.

82
However, there are instances of second and third person plural mergers in West Hmongic: Xuyong
/ne~me4/, Pei-tung-nuo /ma1/. In these two cases, there is not enough information to say which category
has been generalized, since the forms do not correspond exactly to the dominant root in either category.
83
White Hmong wb /ʔɯ1/ ‘we-two’ does not seem to be synchronically related to ob /ʔɔ1/ ‘two’. But there
is variation in the pronunciation of the 1DU in the closely related Green Hmong according to Lyman
(1979:29): ʔɯ1 ~ ʔi1 ~ ʔe1.
220 Chapter 5

Yanghao (East Hmongic) maŋ1 ‘you-two’ maŋ2 ‘you (pl.)’

White Hmong (West Hmongic) ne1 ‘you-two’ ne2 ‘you (pl.)’
Bunu (West Hmongic) mi1 ‘you-two’ mi2 ‘you (pl.)’
Bunu (West Hmongic) mu1 ‘they-two’ mu2 ‘they’
Nau-klau (Hmongic) ma1 ‘you-two’ ma2 ‘you (pl.)’

It is also quite common for these source phrases to persist intact.

Inclusives/Exclusives
An inclusive/exclusive first person plural contrast has been recorded for the widely
separated languages Na-Meo (East Hmongic) and the Mienic languages Mien and Mun:

inclusive exclusive
Na-Meo (East Hmongic) mi -pai
1
nɔŋ3-pai
Mien (Mienic) buə1 jiə1-buə1
Hainan Mun (Mienic) ʔban1 ʔbuu1 (= ‘group’)

Given the genetic and geographical separation between Na-Meo and these two Mienic
languages, the inclusive/exclusive contrast may be old in the family, now lost in most
places, or perhaps not consistently elicited.
Shintani and Yang (1990) also record an odd third person plural inclusive/exclusive
contrast for Hainan Mun:

inclusive exclusive
ta2-ʔbuu1 ‘?-group’ nan2-ʔbuu1 ‘s/he-group’

No information on how these are used is provided; they might work as a third
person/fourth person contrast, a proximate ‘they’ versus a distal ‘they’, but this is just
speculation.

5.4 Demonstratives84

Person-based systems
The only demonstrative that can be confidently reconstructed for Hmong-Mien is the
proximal demonstrative ‘this’: *ʔneinX (2.7/25, White Hmong no). For the distal,
Hmongic *ʔɨB ‘that/the former’ (7.1/2, White Hmong i) can be reconstructed, as well as the
possibly related Mienic *ʔweB ‘that’ (1.10/10). Another Hmongic demonstrative,
represented by White Hmong ub ‘far’, indicates even greater distance from the speaker. It
cannot be reconstructed in full, but a reconstructed form would have tone 1 and a glottal
initial. The meanings and uses of these forms do not correspond exactly from language to
language, so it is difficult to reconstruct a Hmong-Mien demonstrative system and their

84
This section is based in part on Ratliff 1997.
 Numerals, personal pronouns, demonstratives 221

place in it. These forms may have broader deictic rather than simple demonstrative
functions.
Across the Hmong-Mien family, one finds not only expected contrasts in
demonstratives based on distance from speaker (‘this’/‘that-near’/’that-far’), but also a
sporadically attested (or only a sporadically recorded) contrast based on proximity to the
hearer (‘this-near-me’/‘that-near-you’/‘that-near-neither’/‘that-far’), hereafter called the
first-, second- and third-person demonstratives and the ‘far’ demonstrative respectively.
Like the dual pronouns described above, it appears that the second person demonstrative is
created anew in each language. Although it may be based on the 2SG pronoun, it may be
created out of other lexical material as well.

this-1st that-2nd that-3rd that-far

West Hmongic
White Hmong nɔ5 kɔ5 ndaɯ8b u1
(kɔ2 ‘you’)
Green Hmong nɔ3 kɔ3 ndaɯ8b u5
(kɔ2 ‘you’)
Bunu nau3 kau2 ‘that-middle’ uŋ1
(kau2 ‘you’)
East Hmongic
Yanghao noŋ3 nen3 moŋ2 ɛ1
(moŋ2 ‘you’) (nen2 ‘he’)
Mienic
Mien nai3 nai6 ‘that-near you’ uə3
(məj2 ‘you’)

Person-based demonstrative systems in Hmong-Mien

In White Hmong, Green Hmong, and Bunu, the second-person demonstrative is taken
directly from the 2SG pronoun. In White and Green Hmong it has taken on the tone of the
first-person demonstrative to signal that it belongs to the demonstrative form class.
In Yanghao, there has been an interesting shuffling of forms and meanings and a tonal
convergence. The 2SG pronoun serves as the basis of the neutral third-person
demonstrative, whereas the 3SG pronoun serves as the basis of the second-person
demonstrative. And while the 2SG pronoun and the third-person demonstrative are
identical, the second-person demonstrative has shifted to tone 3 to establish a form class
with the first-person demonstrative, as in White Hmong and Green Hmong.
Mien uses a different source for its second-person demonstrative: it is not cognate with
the Mien 2SG pronoun—it is a tonally altered form of the first-person demonstrative (Court
1985, Mao, Meng and Zheng 1982).
Although it is also not possible to reconstruct the form of a second-person
demonstrative for Hmong-Mien, the new generation of forms that serve this purpose in
widely separated languages of the family suggests that the contrast itself is old. On the
other hand, the evidence does not warrant the reconstruction of other contrasts attested in
222 Chapter 5

the demonstrative systems of individual Hmong-Mien languages to higher levels: Gerner

(2009) has recorded demonstrative contrasts in individual Hmongic languages based on
altitude (high, equal, low) and orientation (front, back), and Ratliff (1992a:104-112) has
provided etymologies for West Hmongic demonstratives based on nouns for geographical
features (‘mountain’, ‘plain’)85.

85
The geographical demonstratives discussed by Ratliff and the altitude demonstratives discussed by
Gerner overlap.
6 Language contact

6.1 Introduction
For the entire 2,500-year period of their reconstructed history and for long before, the
Hmong-Mien languages have been spoken in an area of rich agricultural resources which
could support large numbers of people and thus a high degree of linguistic diversity
(Bellwood 1996, 2001, 2005). From a time that antedates the reconstruction, Hmong-Mien
speakers have been in contact with speakers from each of the other language families of
southern China and Southeast Asia: Sino-Tibetan, Tai-Kadai, Austroasiatic, and
Austronesian. The ultimate challenge is to find persuasive evidence that will demonstrate
that Hmong-Mien is genetically related to one or more of these great families more closely
than it is to the others. However, since the purpose of this book is to provide a new
reconstruction of Proto Hmong-Mien and a discussion of certain family-internal historical
topics, arguments for the superiority of one proposal of distant relationship over another
will not be presented here.
It is relevant to note, however, within the context of a chapter on language contact, that
a genetic relationship has been proposed between Hmong-Mien and every one of its
neighbors at one time or another. Hmong-Mien (Miao-Yao) has been represented as part of
the Sino-Tibetan family by most Chinese scholars (for example Wang 1986, Pan 2006).
For those who favor a “greater Sino-Tibetan” family, a connection between Hmong-Mien
and Tai-Kadai is established through their common lexical and typological ties to Chinese.
Forrest (1973 [1948]:93-103), Haudricourt (1966), and Peiros (1998:155-160) raise the
possibility of a genetic link to Mon-Khmer (or the higher family Austroasiatic) on the basis
of a few shared high-quality lexical items. Benedict (1975) places Hmong-Mien in a
family with Tai-Kadai and Austronesian as part of his Austro-Tai hypothesis. Although no
one has proposed a close genetic relationship between Hmong-Mien and Tibeto-Burman,
those who believe that Hmong-Mien is related to Sino-Tibetan—either directly as a
subgroup, or adjoined to Sino-Tibetan as part of Sino-Tibetan-Austronesian (Sagart
2005)—would have to embrace the notion of a very distant genetic relationship between
these two families as well.
The lexical links between Hmong-Mien and languages from these various families have
been persuasive enough to make different eminent scholars propose quite different family
constellations. One might suppose this points to the existence of an ancient and shadowy
“Proto East Asian” language family, as was proposed by Starosta (2005). Although such a
construct is not unreasonable, it is very difficult to defend, given limited evidence. The
nearer goal would be to identify one other language family as the best candidate for either
a coordinate or superordinate relationship to Hmong-Mien within the time frame (c.
10,000–12,000 years) to which our methodology limits us. Since that work has not been
completed, lexical look-alikes shared by Hmong-Mien and Tai-Kadai, or Austroasiatic, or

223
224 Chapter 6

Austronesian are presented in the reconstruction as native Hmong-Mien; no position is

taken with respect to the question of a contact versus an inheritance account. The special
case of Chinese, by far the most important contact language in the Hmong-Mien-speaking
area, will be addressed in section 6.2. The last section of this chapter, section 6.3, will
summarize the best lexical connections found to date between Hmong-Mien and the other
language families of the area.

6.2 Chinese

6.2.1 Phonology
The phonological features of Hmong-Mien that are most likely due to contact with Chinese
include both segmental features and prosodic features. The consonant features that appear
frequently in Chinese loanwords include aspiration, frication, affrication, and palatalization
(see the introduction to the reconstruction, chapter 2, section 1.2, on aspiration and
frication). In the tables of reconstructed onsets in chapter 2, the correspondence sets that
are composed solely of Chinese borrowings appear in parentheses. These onsets can be
grouped as follows:

1) all of the labialized initials: ɲʷ, ŋʷ, kʷj, gʷj, qʷ

2) all of the velar nasals: ŋ, ŋʷ, ŋj
3) many onsets with a palatal glide: mpj, thj, ndj, kʷj, gʷj, ŋj, xj, ɣj, klj, ɢlj
4) some fricatives: ɣ, xj, ɣj, ɦ

Of these groups, only the labialized initials and velar nasals are represented exclusively by
Chinese loanwords; native words have also been reconstructed with medial -j- and with
fricative onsets.
Proto Hmong-Mien rime correspondence sets composed solely of Chinese borrowings
appear in italics in the tables of reconstructed rimes in chapter 2. These rimes are all closed
rimes (-i̯ ɛp, -u̯ at, -æk, -u̯ am, -eəm, -u̯ ən, -əan, -iəŋ, -iuŋ), but this does not appear to be
significant, since closed rimes are common in native words as well. The only Proto
Hmongic rime composed solely of Chinese borrowings is *-i̯ aŋ (rime set 26).
Prosodic features that can be attributed to Chinese contact include vowel length (chapter
2, section 2.1.3.2), tone, and tone split occasioned by the merger of voiced and voiceless
obstruent initials (chapter 3, section 3.1). It is possible that these prosodic features,
although characteristic of Chinese, did not originate with Chinese: vowel length and tone,
as prosodic features, are especially diffusible (Matisoff 2001, L-Thongkum 1997), and
could have originated anywhere in the area, subsequently spreading to neighboring
languages (chapter 3, section 3.2.2). They therefore may represent substratum effects upon
Chinese rather than superstratum effects of Chinese on other languages of the area; it is
impossible to tell. However, once adopted by Chinese, the spread of these features would
have been greatly facilitated.
 Language contact 225

Alternations of n- and l- in Chinese loanwords

In addition to the large-scale effects of Chinese on Hmong-Mien phonology sketched
above, the following group of words with both n- and l- onsets in Hmong-Mien (onset
211.1 in Wang and Mao 1995) presents an interesting minor pattern:

H *ʔloB (2.40/7) ‘to break’ 拗 ‘to break’ (MC ʔæwX > Man. ǎo)
M *ʔnəuB (2.7/7) ‘to break’

H *ʔlɛŋA (2.40/22) ‘CLF-bowls/houses’ 庵 ‘thatched hut’ (MC ʔom > Man. ān)
M *ʔnɛɔmA (2.7/22) ‘CLF-bowls/houses’

H *ʔlɛŋB (2.40/22) ‘short’

M *ʔnəŋB (2.7/22) ‘short’

On the basis of ‘to break’, we may hypothesize that these three words use different
strategies to cover an onset-less word: Hmongic covered with an l- , and Mienic covered
with an n- . Equally likely is that the donor dialects of Chinese had developed these initial
consonants before the loans were made, and that Hmongic borrowed the words from an
“l-covering” Chinese dialect, whereas Mienic borrowed the same words from an
“n-covering” dialect. The proposed Chinese source for the all-purpose classifier (here
‘CLF-bowls/houses’, see section 6.2.4 below) is quite speculative, and no Chinese source
has yet been found for ‘short’.
There are also words in Mienic only that show an alternation of initial n- and l-. In the
first word below, alternation of an original *n- exists in the donor language, which is
apparently Cantonese:

念 ‘to think of/love’

M *hnəmB~hləmB (2.8/22) ‘to think of’
(MC nemH > Cant. /nɐm3/ (~lɐm3/)

In the next word, ‘to use’, Old Chinese *l- > y- is reflected variously as /n-/ or /l-/. This
word shows a different correspondence pattern from those above, and no Mienic onset can
be reconstructed (‘to ask for’ patterns in the same way, and may also be a loanword,
although no source has been identified).86

用 ‘to use’
M *[n/l]oŋC (rime 28) ‘to use’
(OC *mə-loŋ-s > MC yowngH > Man. yòng)
M *[n/l]oŋC (rime 28) ‘to ask for’

86
There is also low-level variation between /n-/ and /l-/ in the Chiang Rai Mien reflexes of a third word,
Hmong-Mien *-ʔrɔŋH ‘good’ (2.55/29): /loŋ5/ ~ /noŋ5/ (Purnell, p.c. 2008). This onset correspondence
does not show /n-/ ~ /l-/ variation across the Mienic sub-family as the loanwords discussed here, but
suggests that these two sounds tend to alternate.
226 Chapter 6

6.2.2 Lexicon
Out of the 829 cognate sets in Wang and Mao 1995, approximately 35% represent words
of Chinese origin.87 Most of these words, given their close similarity to Middle Chinese or
modern-day Chinese forms, and their cultural nature, are clearly borrowings.88
There are also certain cultural words in Proto Hmong-Mien for which Old Chinese
provides a better match than Middle Chinese, such as ‘iron’, ‘village’, and ‘barn/granary’.
Then there are a number of basic words—one body part term (‘tongue’) and a few verbs
(including ‘to descend; low/short’, ‘to be far’, ‘to drink/smoke’, ‘to sing/cry out’, ‘to cut
open’)—that are ancient in both Hmong-Mien and Chinese. This last group of words is of
special interest, in that these words do not fit into the usual semantic fields for loanwords,
such as commerce, technology, agriculture, animal husbandry, and society. The correlation
of Proto Hmong-Mien reconstructions with both Middle Chinese and Old Chinese forms
and the presence in the Sino-Hmong-Mien vocabulary of both cultural and basic words
suggest three possible historical accounts of the facts: (1) the contact situation was so
prolonged and intimate that basic words as well as cultural words were borrowed (see
Thomason 2001:63–65 on the borrowability of everything); (2) words in the Middle
Chinese set are borrowings and words in the Old Chinese set are shared retentions from a
common ancestor; or (3) words in the Middle Chinese set are borrowings and the Old
Chinese set represents a mix of borrowed words and shared retentions. Although an
important working assumption of this reconstruction is that the Chinese element in the
Hmong-Mien lexicon overwhelmingly reflects language contact (see chapter 1, section
1.3), this does not exclude the possibility that a small core of true cognates exists. Much
more work on this question remains to be done, and will need to be undertaken with the
collaboration of specialists in the history of Chinese.

6.2.3 Syntax
The morphosyntax and syntax of Hmong-Mien languages have been heavily influenced by
Chinese. Although the details may differ, all of the gross properties of the Chinese type are
apparent in Hmong-Mien languages: in morphology, Hmong-Mien languages are
characterized by the lack of inflectional morphology and the heavy use of compounds and
reduplication. In syntax, Hmong-Mien languages show SVO word order, paratactic rather
than embedded structures (serial verb, serial noun, and serial sentence constructions), use
of both a polar question particle and the A-not-A construction, Wh-question words in situ,

87
This set of 829 words already excludes the most obvious Chinese loanwords (Wang and Mao 1995:19). A
recent study of White Hmong (Ratliff 2009a, b), in which only 15% of 1290 entries are identified as
Chinese loanwords, might be seen to challenge this picture of Chinese dominance. This is a misleading
figure, however. In White Hmong, many meanings are represented by a compound or a phrase rather than
by a single word. Given the guidelines of the larger project of which this study was a part, if a compound
or phrase were a mix of Chinese and native words (or all Chinese words assembled into a new phrase), it
was not counted as a loan.
88
Recent loans tend to be too consistent in their segmental make-up across languages, but not consistent
enough in tone (for example, from Chinese 上 ‘to start’ (Man. shàng) > Yanghao sɑŋ3 [35], Jiwei ʂɑŋ1
[35], Xuyong ʂɑŋ8 [13]). It is clear that speakers are bilingual, and know these words and their Chinese
pronunciations. The native tone assigned to each word is simply the tone in each borrowing language that
comes closest to the Chinese tone phonetically, so that the tones will be of different historical categories,
and will not correspond cross-linguistically (see chapter 3, section 3.2.1).
 Language contact 227

adversative passives, and aspect markers rather than tense markers. Final discourse
particles signal speaker attitude.
In some Hmong-Mien languages, Chinese loanwords figure prominently in the set of
grammatical function words. Thus in some cases we can say that a particular construction
in the grammar of Chinese has been transferred to Hmong-Mien on the back of the word
that signals it. The following examples come from White Hmong of Southeast Asia, but
can be assumed to be typical of other languages in the family:

tau [tau5] < HM *təuk (2.1/6) 得 (OC *tˤək > MC tok > Man. dé)
verb: ‘to get/gain’ verb: ‘to obtain, get’
auxiliary: past (preverbal)
modal: ability, possibility (postverbal)

sib [ʃi1] reciprocal prefix < H *sjɨA-D (2.28/2) 廝 Man. sī ‘with each other’

lawm [laɯ8] perfective marker 了 Man. le perfective marker89

tabsis [ta1 ʃi4/7] ‘but’ 但是 Man. dànshì ‘but’

vim(chij) [vi8 chi2] ‘because’ 為此 Man. wèicǐ ‘for this reason’

6.2.4 The classifier construction

The case of the classifier construction can be examined in some detail in this connection,
as an example of grammatical change brought about through contact with Chinese. The
fact that it is the only syntactic construction to receive close attention here is not to suggest
that it is the only example of grammatical accommodation that could be discussed;
interesting work remains to be done with, for example, the effect of Chinese on word
order, verbal aspect, and clause linking in various Hmong-Mien languages.
The Hmong-Mien languages typically possess two systems for noun classification, both
well-represented in languages of Southeast Asia: numeral classifiers and classifying
prefixes. It is not unusual to find cases where representatives of each system occur side-by-
side, each apparently classifying the noun in the same way, as in the White Hmong noun
phrase: ib lub pob-zeb, roughly ‘one CLF-bulky things clump-stone’. This redundancy of
function in the two grammatical systems is a result of long-term language contact and
multilingualism. Evidence will be presented below to show that the classifier construction,
although a hallmark feature of all modern Hmong-Mien languages,90 has been borrowed
from Chinese, while the system of classifying prefixes, now only marginally functional in
modern languages, is the native one. Given the fact that the White Hmong classifier lub in
the example above appears across the family—Hmongic *ʔlɛŋA (2.40/22) and cognate
Mienic *ʔnɛɔmA (2.7/22) ‘CLF-bowls/houses’—it seems probable that the protolanguage
had both a productive prefix system and classifiers in a few phrases which formed the
stimulus for the development of a full-fledged classifier system. However, even this oldest
classifier may have been borrowed from Chinese (see section 6.2.1 above), along with
most of the other most common Hmong-Mien classifiers, as will be shown below.

89
This connection was first reported in Fuller 1986.
90
Nichols (1992:132–133) places Hmong-Mien in the middle of a numeral classifier “hotbed”.
228 Chapter 6

Numeral classifiers and classifying prefixes not only have the same general classifying
function, but also pick out overlapping semantic classes: for example, for the
human/animal/animate class (different languages will sub-divide this class differently)
there are both classifiers and prefixes, for inanimates there are both classifiers and prefixes,
and classifiers based on shape are used alongside shape-based body-part prefixes in the
body part class (for example, White Hmong ib tus ko-tw ‘one CLF-short lengths handle-
tail’, or simply ‘one tail’).91
There are, however, important structural differences between the two types of
classifying elements. The core structure of the noun phrase in all Hmong-Mien languages
is the same, with the exception of the demonstrative, which is phrase-final in Hmongic
languages and phrase-initial in Mienic languages:

[[NUM + CLF]QP + [(prefix)-NOUN + ADJ]NP]NP

The hierarchical structure of the noun phrase in Hmongic languages, in which the
quantifier and the classifier form one constituent (QP = quantifier phrase) and the noun and
its modifiers form another (NP = noun phrase) is reflected in those languages that have
tone sandhi relationships between certain elements of the noun phrase but not others: the
numeral may change the tone of the classifier; the prefix may change the tone of the noun.
But the classifier never changes the tone of the prefix; these elements seem not to “see”
each other. Similarly, prefixes may display phonological variation in the initial and/or may
change to harmonize (in vowel or tone) with the head noun, whereas classifiers do not
show this kind of harmony with the head noun.92
Further distinctions include the following (see also Chen 1993): (1) Classifiers (and
class nouns in [class noun-noun] compounds) are semantically and phonologically more
robust than prefixes. It is often easy to identify the source of a classifier for this reason
(White Hmong thoob ‘bucket’ > ib thoob dej ‘one bucketful water’), whereas prefixes are
semantically and phonologically lightweight. (2) Classifiers are obligatory with quantified
or definite nouns; prefixes are often optional. In languages where prefixes are optional, the
use of a (true) prefix appears to be governed by stylistic or prosodic needs in given
contexts. (3) Classifiers can be used pronominally to stand in for the full noun phrase;
prefixes cannot be used pronominally.

Two paths of grammaticalization

The classifier construction was borrowed from Chinese, probably at first along with the
Chinese measure words which would be necessary for both sides to use in trade situations.
That this construction was identified with its Chinese origins can be shown by the high
number of Chinese words that serve as classifiers in Hmong-Mien (some of which will be
presented below). Discourse considerations may also have favored borrowing: for the
North Arawak language Resígaro, which has borrowed many classifiers from the unrelated
neighboring language Bora, Aikhenvald (2000:387) notes that “[t]he extensive borrowing

91
Ko is a White Hmong prefix that is also used for feet (ko-taws) and bamboo sprouts (ko-kaus). The noun
ko meaning ‘handle’ (e.g., ko taus ‘axe handle’) appears to be related.
92
A seeming exception to this generalization is the West Hmongic language A-Hmao of Shimenkan
township, Weining Yi, Hui and Miao Autonomous County, Guizhou Province, which has classifiers that
can modify vowel and tone to signal whether the head noun is “grand and imposing”, ordinary, or
diminutive (Wang 1972). This appears to be an independent development.
 Language contact 229

of classifiers…can be explained by the important role classifiers play in discourse: once

the referent is established it is referred to with a classifier, so that classifiers appear to be
more frequent in discourse than nouns themselves.” Since the establishment of the
classifier structure within Hmong-Mien, however, the first path of grammaticalization
from noun to classifier has been direct, and has pulled in native words as well.
A second path of grammaticalization derives the prefixes from class nouns as the first
member of class noun-noun compounds. This process is independent and on-going. For
example, cognates of White Hmong kab ‘insect’ in the Eastern and Western branches of
Hmongic are used as class nouns in insect name compounds, but are also beginning to
branch out to serve as class nouns in compounds denoting shrimp, crabs and moles
(‘insect’ > ‘little animal’). Similarly, the White Hmong tub ‘son/boy’ is used as a class
noun in compounds that refer to male humans, whether young or not (such as tub-rog
‘soldier’, but literally ‘boy-war’). The fact that these words are phonologically lightweight
as well as appropriately general in meaning have conspired to facilitate this change.
But notably, in HmM, there is no evidence that there is a

noun  class noun  classifier

continuum, as described by Delancey (1986) for Tai and as claimed for Hmong, Thai and
Vietnamese by Bisang (1999) as the usual “Southeast Asian” origin of classifiers. If there
were evidence that class nouns developed into classifiers rather than into prefixes, then we
could say that classifiers developed internally, and there would be no need to appeal to a
contact explanation. But not only is there no evidence for this change, there is evidence in
White Hmong for change going in the opposite direction, from classifiers to class nouns, in
an “anti-grammaticalization” process (a similar development was described for Chinese by
Loke (1997) as “re-grammaticalization”).
This classifier > class noun development in White Hmong is a limited phenomenon that
occurs when the head noun is “underspecified” semantically and the group classifier cov
‘CLF-groups’ is used (Ratliff 1991):

ib phau ntawv cov phau-ntawv

1 CLF-piles paper CLF-groups book
‘book’ ‘books’

ib lo lus cov lo-lus

1 CLF-mouthfuls language CLF-groups word
‘word’ ‘words’

Ntawv ‘paper’ cannot interpreted as ‘book’ without the classifier phau for ‘piles or stacks’.
So when it is important to emphasize that many books are being referred to, since there can
only be one classifier per noun phrase, the ‘group’ classifier displaces the ‘piles or stacks’
classifier. The disambiguating classifier then attaches to the underspecified noun as a class
noun. The same shift can be seen in the phrase for ‘words’ above: the classifier lo for
‘mouthfuls’ is displaced when cov is added, and attaches to the underspecified noun lus
‘language’.
230 Chapter 6

Three more examples of White Hmong classifiers that may also function as class nouns
appear below. These examples do not involve underspecified nouns, so it is not possible to
trace a pathway of development:

ib zaj kwv txhiaj ob zaj zaj-tshoob

1 CLF-lines song 2 CLF-lines wedding song
‘a song’ ‘2 lines of a wedding song’

ib txoj kev txoj-hmoov

1 CLF-lengths road fortune
‘a road’ ‘fortune’

ib pob mov ob lub pob-zeb

1 CLF-clumps rice 2 CLF-bowls/houses stone
‘a ball of rice’ ‘2 stones’

All three of these classifiers are of Chinese origin: two (zaj and txoj) were borrowed as
classifiers (see below) and then shifted “to the right” to become class nouns, whereas the
third (pob) was probably borrowed as a noun from Chinese 包 (Man. bāo) ‘to wrap; a
bundle, lump’ and since has come to be used in three ways—as an independent noun, as a
classifier, and as a class noun. It is also now on its way to becoming a prefix. One might
say that the classifier function of pob derived from the class noun function, but since it is
the only hyper-functional noun of its type, and it is a Chinese borrowing, it hardly supports
a language-internal account of the grammaticalization of class noun to classifier.
To summarize, in White Hmong some nouns get used as classifiers (xib ‘arrow’, thoob
‘bucket’, nplooj ‘leaf’ ), and some nouns get used as class nouns in compounds (kab
‘insect’ and tub ‘son/boy’), and some class nouns degrade to prefixes (qhov ‘hole’, a
nominalizer for ‘thing’, and pob ‘clump’ for any round or bulky thing), but nouns do not
usually become classifiers via class nouns, which argues for the existence of two separate
systems of classification, and two separate processes of grammaticalization, rather than
one:

noun class noun

 ↗  but *class noun  classifier

classifier prefix

In addition to this analysis of two functionally-similar constructions in one language that

only rarely have anything to do with one another, four types of indirect evidence drawn
both from unrelated languages and from languages across the family provide further
support for the idea that the classifier construction in Hmong-Mien languages was
borrowed from Chinese: (1) the “borrowability” of classifier systems cross-linguistically;
(2) the etymology of the most widespread classifiers in Hmong-Mien; (3) the relative
chronology of the two systems; and (4) the geographical and genetic distribution of the two
systems within the Hmong-Mien family and the correlation of these facts with degree of
contact with Chinese.
 Language contact 231

The “borrowability” of classifier systems cross-linguistically

Numeral classifiers are borrowable, especially between analytic languages in which they
are independent words:

The more lexico-syntactic the noun categorization is, the easier it is to diffuse. That is, the
most frequently diffused new patterns involve syntactic constructions with classifiers which
have a clear lexical source rather than with more grammaticalized noun class agreement
systems. Possessed classifier constructions of the type generic noun plus possessed noun
were borrowed from Carib languages into a number of North Arawak languages, e.g.
Palikur, Island Carib, and Bahwana … In the South Asian linguistic area numeral classifiers
have been shown to have spread from the Indo-Aryan languages of the region to Dravidian
languages … Indirect areal diffusion may result in the partial restructuring of classifier
systems. This involves introducing new classifier types into a system which already has
classifiers. (Aikhenvald 2000:383–384)

A well-known instance in Asia is the borrowing of Chinese classifiers by Japanese. Also,

system overlay in general is not unusual, even when it does not seem to serve a clear
purpose: “…worrying about where one grammar ends and the next grammar begins is a
totally meaningless and futile pursuit. For ‘the new grammar is constantly being created on
top of the willing and yielding ruins of the old’” (Matisoff 1991:447).

The etymology of individual classifiers

In addition to the most common Hmong-Mien classifier discussed above (Hmongic *ʔlɛŋA
(2.40/22) and cognate Mienic *ʔnɛɔmA (2.7/22) ‘CLF-bowls/houses’) which may be
derived from a Chinese noun meaning ‘hut’, many of the oldest and most basic classifiers
and some of the newest abstract classifiers are clearly Chinese in origin.
Several of the most common classifiers are loanwords which demonstrate regular
correspondences and are found across the family (White Hmong classifiers are given both
in italicized orthographic form and in phonetic transcription):

1. CLF-tools rab /ʈa1/ < H *traŋA 張 OC traŋ > MC trjang > zhāng
M *truŋA ‘CLF-flat things’ (< ‘spread’); first used
as CLF for ‘bow’, then ‘zither’ in Han
period c. 200 BCE

2. CLF-quilts phob /phɔ1/ < HM *phəan 片 MC phenH > piàn ‘one-sided’

3. CLF-lines 行 MC hang > háng ‘CLF-rows’

(writing, speech)
Yanghao /ɣoŋ1/, Jiwei /ʐoŋ2/, White Hmong zaj /ʐa2/, Xuyong /ʐɑŋ2/, Bunu /haŋ4/, Mien
/hɔːŋ2/

4. CLF-things 件 MC gjenX > jiàn ‘CLF-items’

Yanghao /ʨin /, Jiwei /cɛ /, Bunu /cin /, Biao Min /iɛn6/
6 6 6

Another common classifier that is used both for animates (animals and humans) and
short objects classified by length seems likely to have had a Chinese origin, given the
232 Chapter 6

semantic similarity of the objects classified in both families by morphemes of similar

shape. However, the word ends in a nasal in Hmongic, and has an open syllable in
Chinese:

5. CLF-horses tus /tu4/ < H *dɛŋB 頭 MC duw > tóu ‘CLF-animals’

also CLF-short cf. use of the noun tóu in compounds
lengths (e.g. referring to pieces of chalk, pencil stubs,
chalk) cigarette butts
2 2
(Mien /tau /, Mun /tau / are recent loans from tóu)

There are also many transparent (recent) Chinese loans, some of which are quite abstract,
as evidenced by these White Hmong classifiers:

6. CLF-long things txoj /tsɔ2/ 條 Man. tiáo

‘CLF-long things’ (< ‘twig’)
(independent non-corresponding
borrowings occur on both sides of family)

7. CLF-kinds/sorts yam /ja8/ 樣 Man. yàng

‘appearance, pattern; kind, type’

8. CLF-kinds/sorts hom /hɔ8/ 號 Man. hào

‘name, mark; order, size, number’

Relative chronology
The analysis of a newer, borrowed nominal classifier system layered upon a native
nominal classification system of prefixes gains support from independent evidence that
prefixes are old in the family. The prefix is the characteristic affix type in the Southeast
Asian linguistic area. Not only do prefixes predominate in the languages of the Tibeto-
Burman and Mon-Khmer families, they are now being reconstructed for the most analytic
languages at the core of the Sinosphere — Chinese and Hmong-Mien. We are able to
reconstruct a small bit of Hmong-Mien morphology, as explained in chapter 4; and with
the exception of a C-tone derivation process (section 4.4.2), all of the evidence suggests
that Proto Hmong-Mien bound morphemes were prefixes. First, there are the nominal
prefixes under discussion here (section 4.2), which from the earliest level have left
confusing correspondence patterns wherein the consonants from different prefixes have
displaced root-initial prefixes in different languages (section 4.2.3). Modern-day complex
clusters also suggest the existence of disyllabic—possibly prefixed—words, given their
survival in Pa-Hng as disyllabic words (section 4.2.4). Finally, there are pairs of words that
suggest that ancient prefixes in the (pre-)protolanguage signaled a causative/stative
distinction in verbs (section 4.4.1). Chen (1993) has supplied additional evidence of the
antiquity of prefixes in the family: prefixes are used among older speakers more frequently
than younger speakers, and prefixed forms are preserved in set phrases but are less
frequently innovated.
Classifiers, on the other hand, do not show evidence of great age. The only classifier
that can be reconstructed to Proto Hmong-Mien is the classifier for houses and bowls—
 Language contact 233

‘bulky things’—discussed above. This particular classifier may be native, but a Chinese
source has also been proposed. Furthermore, the reconstructions of the Hmong-Mien
classifiers that are obvious Chinese loanwords correspond to Middle Chinese rather than
Old Chinese. This is not surprising, given that classifiers have only developed within the
historical period in Chinese (see Erbaugh 1986 and Loke 1997). Their early use in the
Shang and Zhou dynasties (1400 BCE–221 BCE) was limited to objects of value only;
widespread use of classifiers was not established before the Han dynasty (206 BCE–220
CE). Sagart (1999:107) has hypothesized that the loss of an Old Chinese *k- prefix used to
distinguish count and mass nouns may have provided an additional stimulus for the
development of the classifier system within Chinese. There is a possibility that all
languages in the area, under stimulus from either Tai or Chinese, developed classifiers at
roughly the same time.

Geographical and genetic distribution of prefixes

As explained in chapter 4, section 4.2.1, there is a correlation between the degree of
contact with Chinese and the preservation of classifying prefixes. In general, Hmongic
languages have been more isolated from Chinese than Mienic languages, and Hmongic
languages preserve more prefixes than Mienic languages. Within Mienic, the Mien/Mun
and Biao Min languages show fewer prefixes than the most conservative Mienic language,
Zao Min, which differs enough from the others to constitute a separate sub-branch. Zao
Min regularly marks nouns with the prefix /ʔa-/ (Mao, Meng and Zheng 1982), whereas
only a few nouns may optionally be prefixed in other Mienic languages. In an early field
report, the Zao Min language was described as “isolated” and the “one least influenced by
the neighboring Chinese dialects” (Wong 1939:425).

6.3 Lexical connections to other language families

As explained at the beginning of this chapter, although Chinese is the primary contact
language, the Hmong-Mien languages show lexical connections to other languages of the
area as well. In this final section, lists of the strongest lexical evidence connecting Hmong-
Mien to the other language families will be presented, along with brief discussion. These
lists could be extended indefinitely, as the comparanda become more and more
speculative. The intention here is to exclude all but the best of these comparanda. For these
words, an historical account of either contact or common inheritance is required—chance
resemblance seems unlikely, given close similarities on all measures.

6.3.1 Tibeto-Burman
A number of Hmong-Mien words belonging to both the basic and cultural components of
the lexicon appear to be loanwords from Tibeto-Burman. The most commonly accepted
view is that these loans were made at a time when the centers of gravity of these two
families were not as widely separated as they are today. Today, the Hmong-Mien center is
to the east—along a north-south line from western Hunan province south through eastern
Guizhou and the western half of the Guangxi Zhuang Autonomous Region—and in terms
of numbers of languages and speakers, the Tibeto-Burman center is to the west—in
western China, Tibet, Burma, Bangladesh, Nepal, and Bhutan (Matisoff 2003:3-4). Studies
234 Chapter 6

on Tibeto-Burman loanwords in Hmong-Mien include Benedict 1987a, Ratliff 2001a, and

Mortensen 2002. Benedict observed that the most important Tibeto-Burman loanwords fall
into well-defined semantic sets, which suggests that they were borrowed as such. He
writes, “These loans … reflect an entirely different ‘borrowing milieu’, so to speak, than
the MY/Chinese loans: much less ‘diffuse’, with loans sparse and rigidly confined to
specific categories” (Benedict 1987a:20).
The words in the list below show a high number of voiceless sonorants in Hmong-Mien
corresponding to a prefixed *s- in Tibeto-Burman, but this is probably an accidental
consequence of the fact that the author once focused on this particular correspondence
(Ratliff 2001a). Tibeto-Burman reconstructions are from Matisoff 2003.

Numerals
The numerals ‘four’, ‘five’, ‘six’, ‘seven’, ‘eight’, ‘nine’—and possibly ‘ten’—were
borrowed as a set from some Tibeto-Burman source; see chapter 5, section 5.2.

The heavens
Hmong-Mien Tibeto-Burman
1. sun/day H *hnɛŋA (2.8/22) *s-nəy
M *hnu̯ ɔiA (2.8/11)
2. moon/month HM *hlaH (2.41/4) *s-la
3. night (also ‘dark’) H *hməŋH (1.8/21) *s-muːŋ ‘dark’

Affinial kinship terms

4. sister-in-law HM *ʔɲam (4.7/24) *nam ‘daughter-in-law’
(also daughter-in-law)
5. son-in-law HM *ʔweiX (1.10/12) *krwəy

Other
6. to slice HM *hlep (2.41/10) *s-lep
7. footprint/track HM *hmjænX (1.23/19) *s-naŋ ‘to follow’93
8. horse H *mjænB (1.24/19) *mraŋ

6.3.2 Tai-Kadai
The difference between the effect of Tibeto-Burman on Hmong-Mien and the effect of Tai-
Kadai on Hmong-Mien is just the opposite from what we might expect. Today, speakers of
Tai-Kadai languages are the closest neighbors of speakers of Hmong-Mien languages
throughout the area: in Guizhou province (Zhuang, Kam, Sui), in the Guangxi-Zhuang
Autonomous Region (Zhuang, Kam), in Southeast Asia (Lao, Thai), on Hainan Island
(Hlai). Nonetheless, most words that these two families share are recent loanwords. Very
few basic old words are shared by the two families, and of this small group, half are
common to Austronesian as well. The difference in borrowing patterns as they pertain to
these two families (Tibeto-Burman and Tai-Kadai in relation to Hmong-Mien) may shed
light on how the linguistic map of 2,500 years ago differed from that of the present day.

93
Within Tibeto-Burman, see also Jingpho *mənāŋ ‘companion’ (Matisoff 2003:304), and Lai neʔ-hnaŋ
‘footprint’ (Kenneth Bik, p.c. 1999).
 Language contact 235

Studies on the lexical connections between Hmong-Mien and Tai-Kadai include Benedict
1975 (who places Hmong-Mien with Tai-Kadai and Austronesian in his “Austro-Tai”
family) and Downer 1978 (on Tai–Mienic lexical connections). Listed below are only a
few of the oldest and best links between the two families. No words that show an exclusive
link between Mienic and Tai-Kadai languages are included (for example, ‘to cover’, ‘to
bury’, ‘young man’), under the assumption that they are loanwords. Similarly, words
shared by Chinese, Hmong-Mien, and Tai-Kadai are assumed to be Chinese loanwords (‘to
split’, ‘chicken’) and are not included here.
Proto Tai reconstructions are from Li 1977, Proto Kam-Sui reconstructions are from
Thurgood 1988, Proto Kra reconstructions are from Ostapirat 2000, and Austronesian
(AN) and Malayo-Polynesian (MP) reconstructions are from Blust’s Austronesian
Comparative Dictionary (ACD), both in this list and the lists to follow.

Hmong-Mien/Tai-Kadai
Hmong-Mien Tai-Kadai
1. fish HM *mbrəuX (1.51/3) Tai *plaA1
Kam-Sui *mprai3
Kra *p-laA
2. orphan94 HM *mbrəjH (1.51/15) Tai *braB2
Kra *bɣuŋB2
3. wild dog H *hmaŋC (1.8/24) Tai *hmaA1 ‘dog’
Kam-Sui *k-hma1 ‘dog’
Kra *x-maA ‘dog’
4. monkey95 H *ʔlinA (2.40/18) Tai *liŋA2

Hmong-Mien/Tai-Kadai/Austronesian
Hmong-Mien Tai-Kadai Austronesian
5. to die HM *dəjH (2.3/15) Tai *tɑːiA1 AN *ma-aCay
6. bird HM *m-nɔk (2.9/29) Tai *nlokD2S MP *manuk
Kam-Sui *mluk8
Kra *ɳokD
7. 1SG H *kɛŋB (5.1/22) Tai *ki̯ əuA MP *-ku ‘my’
Kra *kuA
8. 2SG HM *mu̯ ei (1.9/8) Tai *m[aï][ŋ]A AN *-mu96
Kra *məA/B

6.3.3 Mon-Khmer
A few important Hmong-Mien words have Mon-Khmer counterparts with similar form and
meaning. As mentioned above, these resemblances struck early researchers as being
94
In Thai and Lao ‘orphan’ bears a /kam-/ prefix, which may be related to the prenasalization in Hmong-
Mien. This word also means ‘bamboo shoot’ in Hmong-Mien, but does not have this meaning in Tai-
Kadai.
95
Compare Mienic *ʔbiŋA ‘monkey’ (1.4/18), which shares the same rime and tone. Forms in Tai-Kadai do
not help explain the different onsets. The b- may reflect the first syllable of an old disyllabic form (also
reflected in the *ʔ- > upper register tone in Hmongic).
96
Austronesian *-mu is a 2nd person suffix with unspecified number (ACD). See section 6.3.4 on Hmong-
Mien/Austronesian links in which Tai-Kadai does not participate.
236 Chapter 6

especially significant, and a genetic relationship between the two families has been
proposed on the basis of the quality of these matches. However, until quite recently a full
reconstruction of Mon-Khmer has not been available,97 making it difficult to demonstrate
either a connection to Proto Mon-Khmer itself (previous studies have cited either sub-
group reconstructions or words from individual languages) or regular correspondences
between Mon-Khmer and Hmong-Mien. Sidwell’s recent publication of the late H. L.
Shorto’s Proto Mon-Khmer reconstruction (Shorto 2006) has improved this situation
greatly.
The following are among the best of the resemblances found to date. The first list
contains unique Hmong-Mien/Mon-Khmer comparanda, whereas the second list contains
words which are in Austronesian as well, suggesting an “enriched” picture of Austric
(Austroasiatic, of which Mon-Khmer is a subfamily, and Austronesian). The Mon-Khmer
reconstructions are from Shorto 2006; the numbers refer to his entries.

Hmong-Mien/Mon-Khmer
Hmong-Mien Mon-Khmer
1. water *ʔu̯ əm (7.1/29) *ʔ[o]m (#1298)
2. blood 98
*ntshjamX (3.20/24) *jhaam (#1430A)
3. to weep99 *ʔɲæmX (4.7/20) *yaam (#1381)
4. tree *ntju̯ əŋH (2.19/29) *t2ʔɔɔŋ (#491)

Hmong-Mien/Mon-Khmer/Austronesian
Hmong-Mien Mon-Khmer Austronesian
5. full *pu̯ ɛŋX (1.1/22) *[d]pu[ə]ŋ (#907B) MP *penuq
6. to shoot *pənX (1.1/21) *paɲʔ (#905) MP *panaq
7. tail100 *tu̯ eiX (2.1/8) *[k]ɗuut (#1017B) MP *buntut
8. to dream *mpeiH (1.4/12) *mp[ɔ]ʔ (#105) MP *hi(m)pi

Interestingly, there are also some Hmong-Mien insect words that resemble their Mon-
Khmer counterparts. These Mon-Khmer insect words appear in a number of other
languages of the area as well. One wonders about the nature of the contact situation that
would lead to the borrowing of the names of such humble creatures (both Mon-Khmer and
“other” forms from Shorto 2006):

97
Peiros (1998) did reconstruct certain Mon-Khmer forms represented in the list below, but he indicated the
tentative nature of these reconstructions by calling them “pre-reconstructions”.
98
See Ferlus 2009 for discussion of this word in both families and alternative Mon-Khmer reconstructions.
He writes that the similarity is strengthened by the correspondence between final -X (> tone B, see
chapter 3) in Hmong-Mien and final glottalization in Proto Viet-Muong.
99
Like Hmong-Mien, several Mon-Khmer languages show a palatal nasal onset in this word ‘to weep’:
Chrau *ɲiːm, Sre *ɲim, Bahnar *ɲəm, Jeh *ɲaːm. Shorto believed this was from the reduplicated form
*yaamyaam, where -my- > ɲ-.
100
Both the Mon-Khmer and Austronesian words refer to the tail (or rump) of a chicken (Khmer /kɔntùːt/
‘rump of fowl’). The Hmong-Mien word is a more general word for ‘tail’.
 Language contact 237

Hmong-Mien Mon-Khmer Other

H *mbrɔD *suuc ‘to sting’(#873B) > Khmer Tai mót ‘ant’,
5. ant
(1.51/6) srəmaoc ‘ant’ Malay semut ‘ant’
M *səpD
6. centipede *[k]ʔiip (#1226G) > Lawa saʔaip Lao khep7
(2.13/7)
HM *klup *kntuəp (#1256aA) > Khmer
7. grasshopper
(5.31/9) kɔndoːp
HM *ntshjeiX *ciiʔ (#39) ‘louse’ > Lawa səiʔ
8. head louse
(3.20/12) ‘louse’, Kuy ɲcɛː ‘head louse’

6.3.4 Austronesian
The inclusion of Austronesian forms in the sections on Tai-Kadai and Mon-Khmer contact
above makes it clear that the Austronesian had contact with the languages of the mainland
before it became the dominant language family of insular Southeast Asia. The question of
the mainland origins of the Austronesians has fostered much recent work over the last
fifteen years, not only by linguists, but also by geneticists and archaeologists (for example,
see the papers from the Conference on Asia-Mainland/Austronesian Connections
(Honolulu 1993) published in Oceanic Linguistics 33.2 and Sagart, et al. 2005).
In addition to the words listed above (‘bird’, ‘full’, ‘to shoot’, etc.), there are some
unique lexical connections between Hmong-Mien and Austronesian; the following list
contains some of the best of these. Three of the words below—‘to die’, 1SG, and 2SG—are
repeated from the Hmong-Mien/Tai-Kadai/Austronesian list in section 6.3.2 above: in
Hmong-Mien (but not in Tai-Kadai) the match of ‘to die’ is extended to the
morphologically related ‘to kill’ (see chapter 4, section 4.4.1)—and the match of the 1SG
and 2SG pronouns is extended to the 2PL pronoun (see chapter 5, section 5.3).

Hmong-Mien/Austronesian
Hmong-Mien Austronesian
1. to die HM *dəjH (2.3/15) AN *ma-aCay
2. to kill HM *təjH (2.1/15) AN *pa-aCay
3. 1SG H *kɛŋB (5.1/22) MP *-ku
4. 2SG HM *mu̯ ei (1.9/8) AN *-mu ‘2nd person’
5. 2PL HM *mi̯ əu (1.9/1) AN *-mu ‘2nd person’
6. soft HM *mlu̯ ɛjH (1.39/11) MP *ma-lumu
7. bite HM *dəp (2.3/7) MP *ketep
8. insect/worm/ HM *klæŋ (5.31/24) AN *qulej ‘maggot’
maggot H *klaŋC (5.31/24)
9. body louse HM *tɛmX (2.1/22) MP *tumah ‘clothes louse’
7 The ancient Hmong-Mien world

7.1 Introduction
The approach to the study of the ancient Hmong-Mien world taken in this short chapter is a
preliminary and rather superficial catalog of non-basic vocabulary appearing in Wang and
Mao 1995 that appears to be both old and native, with the goal of sketching a few features
of the environment and of daily life at the time the proto-language was spoken. Each one
of these criteria needs to be further explained, however, and the numerous problems with
attempting such a sketch need to be acknowledged.
First, for this project, non-basic vocabulary is of greatest interest, because this is the
vocabulary that reveals with some specificity the surroundings, practices, and beliefs of a
group of people. Basic words like ‘water’ serve best when the goal is to elucidate distant
genetic relationships, since by definition words with basic meanings are presumed to exist
in all languages, therefore ensuring that comparanda can be found. A word like ‘puttees’
(leg wrappings), on the other hand, may not find a match in a candidate sister language
group given different habits of dress, but the fact that it is old in Hmong-Mien sheds
considerable light on one particular cultural detail of the ancient world. Some basic words
are included here, however, if they are members of a culturally significant set (numbers,
color terms, etc.).
The second criteria, that the word be old, is satisfied in the first instance if the word
appears on both sides of the family. This is an uncontroversial way of establishing the
oldest layer of the lexicon, especially when there are no potentially confusing issues like
diglossia in standard and local varieties of the same language, as is common in larger
language groups spoken by people in organized nation-states, or other types of known
intra-family borrowing.
More controversially, however, Hmongic words are treated here as old if their semantic
counterparts in Mienic are obvious Chinese borrowings. This seems to be justified by the
frequency with which the same meaning is represented by a native word in Hmongic and a
Chinese word in Mienic. For example:

Hmongic Mienic
face *bowB *hmiənA < 面 miàn
heart *prowB *simA < 心 xīn
pus *bu̯eiC *nɔŋC < 膿 nóng
fog/cloud *huA *mowC < 霧 wù (Cant. /mou6/)
white *qlowA *bækD < 白 bái
red *ʔlinA *sekD < 赤 chì
axe *tu̯eiC *pouB < 斧 fǔ

238
 The ancient Hmong-Mien world 239

bowl *deB *ʔwənB < 碗 wǎn

trough *qroŋA *dzuA < 槽 cáo
thick *tæA *ɦɔuB < 厚 hòu

Although other explanations are possible, the fact that so many native words in
Hmongic have synonyms in Mienic which are relatively recent Chinese loans leaves open
the possibility that these Hmongic words may be more ancient than they appear, having
been replaced by Chinese words in Mienic after the separation of the two groups. There is
other evidence of relatively greater contact between the Chinese and speakers of Mienic
languages: for example, unlike speakers of Hmongic languages, the Mien use Chinese in
rituals, songs, and religious writings (Lemoine 1982, Purnell 1991, 1998). The Mien also
use Chinese numerals alongside native numerals (Purnell 1968, forthcoming).
The third criterion—that the word be native—is both difficult to justify and difficult to
determine. First, in terms of sketching the cultural context of the ancient world, one could
say with good reason that this criterion is not very important: for example, we know that
the Chinese words ‘gold’, ‘silver’, ‘copper’ and ‘iron’ were in the Hmong-Mien
protolanguage over two thousand years ago, and that these metals were thus an important
part of the ancient Hmong-Mien world. But words with clear connections to Chinese have
not been included here, because it is of interest to know what terms for the environment,
cultural traits, and artifacts were particular to the Hmong-Mien group. If Chinese and
Hmong-Mien are not genetically related, a question which remains open (see discussion in
chapter 6), these isolated words may be more likely to suggest aspects of the ‘pre-
protolanguage’ world, the world before contact with Chinese. Identifying Chinese
borrowings is occasionally difficult, however, and sinologists will no doubt be able to
identify a Chinese source for some of the presumably native Hmong-Mien words listed
below. Furthermore, some words with Chinese connections (such as ‘rice/grain head’) are
deliberately included here because there is evidence to suggest that Chinese borrowed
them from Hmong-Mien rather than the other way around. Finally, since the genetic
relationships among the other families of the region have not yet been worked out, it is
always possible that a word common to both Hmong-Mien and another family of the
region (such as ‘wild dog’, shared by Hmong-Mien and Tai-Kadai) reflects a genetic
relationship (and thus native status) rather than a borrowing relationship. Pending further
research, such words are included here.
This chapter is intended to provide an outline of semantic fields which merit further
focused study. More detailed work on each one would ideally include the oldest layer of
Chinese borrowings in Hmong-Mien, and would add many more words to those few which
continue to be used in comparative studies.

7.2 Homeland101
Since the publication of folk tales which tell of an ancient migration from a cold land in the
north where animal furs were worn and the sun shone for only half the year (see especially
Savina 1924), curiosity about the origins of the Hmong-Mien people has been steadily
expressed by both the people themselves and by outsiders interested in this group.
However, no linguistic evidence has been discovered to date to suggest that the Hmong-

101
Sections 7.2, 7.3, and 7.4 are updated versions of material first presented in Ratliff 2004.
240 Chapter 7

Mien people have ever lived in any place other than southern China. As Sagart (1995:341)
writes:

Converging evidence from history, linguistics and human genetics suggests that the rice-
growing speakers of Proto-Hmong-Mien should be counted among the distant descendants
of the early domesticators of rice in the Yangzi basin, rather than of the early millet farmers
in the Huang He basin … Chinese historical records mentioning them as the Man reliably
place them in the mid-Yangzi basin in Han times; the rice-growing tradition of the Hmong-
Mien people, evident in the reconstructed vocabulary … goes in the same direction; finally,
recent work by human geneticists has brought evidence for two great centers of development
of populations in present-day China: one of these centers is the Huang He basin, and the
other, the Yangzi basin. Their data show Hmong-Mien speakers to relate massively to the
second center.

In support of this picture of the homeland of Proto Hmong-Mien speakers, there are
ancient words for three animals and three plants explicitly linked to the provinces south of
the Yangzi River. These independent evocations of the Yangzi represent a striking
convergence, and add to the evidence that places the Hmong-Mien people in southern
China at the time when Proto Hmong-Mien was spoken, which is as far back as the
linguistic evidence can take us.102
The three animals are the pangolin (manis pentadactyla or manis javanica), the ‘painted
eyebrow’ thrush, huàméi niǎo 畫眉鳥 (garrulax canorus canorus), and the river deer
(hydropotes inermis). One is reconstructed to Proto Hmong-Mien: *rɔiH ‘pangolin’,
(2.57/6, White Hmong zaug). The other two are reconstructed to Proto Hmongic: *cɔŋA
‘painted eyebrow thrush’ (4.1/29, White Hmong coob), and *ŋgu̯ eiB ‘river deer’ (5.6/12).103
In China, the pangolin is found in the southern Yangzi River basin (Wildlife of China: 69).
The dictionary Cí hǎi 辭海 records that the painted eyebrow thrush is a non-migratory
bird, common in the south Yangzi River basin, and that the river deer is found in the
middle and southern Yangzi River valley, as well as on reed beaches and grasslands on the
southeast coast.
The three plants are cogongrass, or thatch grass (imperata cylindrica), Allium, also
known as the Chinese onion or rakkyo (allium chinense), and the medicinal plant
houttuynia cordata, in Chinese yú xīng cǎo 魚腥草. One is reconstructed to Proto Hmong-
Mien: *NKan ‘cogongrass’ (5-6.4/19, White Hmong nqeeb). The other two are
reconstructed to Proto Hmongic: *ɢləŋB ‘Allium’ (6.33/21, White Hmong dos ‘small
onion, leek’), and *truwD ‘houttuynia cordata’ (2.46/9, White Hmong raus in tshuaj kab
raus). The Zhōngguó Cǎodì Zīyuán Tújì 中國 草地 資源 圖記 [Atlas of Grassland
Resources of China] (1996) shows that the tropical cogongrass used for thatching roofs
gradually disappears as one moves north from the 33rd to the 34th latitude, which places it
in southern China. Allium has both medicinal purposes and may be eaten as a vegetable. It
is cultivated in Guangxi, Hunan, Guizhou, Sichuan, and Guangdong (Cí hǎi 辭 海). Yú
xīng cǎo is a plant with medicinal uses: the leaves and roots may also be eaten (Cí hǎi 辭
海; Anderson 1993:213). It grows in provinces south of the Yangzi River.

102
Of course, if the Hmong people were from a northern region before the period recoverable by linguistic
research, terms for northern animals and plants would have either been lost or reassigned to southern
species, so this does not disprove a theory of an ancient migration from the north.
103
This word does not exist in White Hmong (it should have the form nkaus), but compare kauv ‘barking
deer’.
 The ancient Hmong-Mien world 241

7.3 Hunting
All of the old words related to hunting either appear to be native or bear a resemblance to
words in other Southeast Asian languages rather than exclusively to Chinese. This
suggests, unsurprisingly, that hunting was once important to the now primarily agricultural
Hmong-Mien people.104
Hmong-Mien ‘crossbow’ (Hmongic *hnænB 2.8/19, White Hmong hneev; Mienic
*hnəkD 2.8/7) is a famous pan-Southeast Asian root: cf. Mon-Khmer *snaʔ ‘crossbow’
(Shorto #97); Proto Tai *hnaa B1 ‘crossbow’ (Li 1977); and Chinese 弩 ‘crossbow’ (OC
*[n]ˤaʔ > MC nuX > Man. nǔ). Hmong-Mien *pənX ‘to shoot’ (1.1/21, White Hmong pov
‘to throw’) is also a widely distributed root: cf. Mon-Khmer *paɲʔ ‘to shoot’ (Shorto
#905), and Malayo-Polynesian *panah ‘to shoot; a firearm’ (Dempwolff 1938).
Some languages retain the original meaning of the Hmongic word *ndroC ‘to track’
(2.51/7, White Hmong nrog), but in Southeast Asian dialects of Hmong a semantic
development from ‘to track’ > ‘to follow’ > ‘to accompany’ > ‘with’ has occurred. A noun
*hmjænX ‘footprint/track’ (1.23/19, White Hmong (h)neev) can also be reconstructed to
Proto Hmong-Mien.
The word *NGej ‘meat/wild game’ is reconstructed to Proto Hmong-Mien and appears
to be native (5-6.6/10, White Hmong nqaij). Although it is impossible to know which
animals were hunted, Proto Hmong-Mien words for wild animals include *qrep ‘bear’
(6.46/10, White Hmong dais) and ‘monkey’ (Hmongic *ʔlinA 2.40/18, White Hmong liab;
Mienic *ʔbiŋA 1.4/18), while Proto Hmongic wild animals include *pljɨD ‘wildcat’
(1.31.1/2, White Hmong plis) and *hmaŋC ‘wild dog’ (1.8/24, White Hmong hma).

7.4 Agriculture
Rice must figure prominently in any picture of the ancient Hmong-Mien world. Unlike the
terms for a number of non-rice crops (corn, soybeans, buckwheat, various types of potatoes
and beans), Hmong-Mien terms for rice are less likely to show similarities to forms with
similar meanings in neighboring families. Quite remarkable is the poor correlation with
rice terms in Tai-Kadai or Austronesian, the languages of major rice growing peoples
(Hartmann 1998, Blench 2005). Although some words associated with wet-rice
cultivation—such as ‘seedling’ and ‘sickle’—have been borrowed from Chinese, the key
terms ‘rice plant/paddy’, ‘husked rice’, and ‘cooked rice’ do not show strong evidence of
having been borrowed.105 Furthermore, a deep influence of Hmong-Mien upon Chinese
may be evident in the word ‘rice/grain head’. This pattern is understandable if the
prehistoric millet-growing Chinese learned how to grow rice from their southern Hmong-
Mien neighbors (Te-Tzu Chang 1983:70).

104
To this set could be added a list of old words that might be related to “gathering”, such as ‘mushroom’,
‘flower’, ‘fruit’, ‘grape’, ‘plum’, ‘persimmon’, ‘egg’, ‘shell’, ‘lay eggs’, ‘nest’, ‘beeswax’, ‘root’,
‘green/raw’, and words for various insects—but since words with these meanings are common to many
cultures, they are not especially revealing.
105
Baxter (quoted in Haudricourt and Strecker 1991:339) and Sagart (1995:337) have both entertained a
possible connection between Chinese 稻 ‘rice; paddy’ (OC *[l]ˤuʔ > MC dawX > Man. dào) and Hmong-
Mien *mbləu ‘rice plant’. Although more skeptical now, they still see it as a possibility (p.c., 2008). They
also see a possible connection between Hmong-Mien *hnrəaŋH ‘cooked rice’ and Chinese 饟 ‘bring food
to workers in the field’ > ‘food brought to workers in the field’ (OC *n̥ aŋ(ʔ)-s > MC syangH > Man.
xiǎng).
242 Chapter 7

The three words for which no convincing outside connections have been found are
Hmong-Mien *mbləu ‘rice plant/paddy’ (1.36/3, White Hmong nplej), Hmong-Mien
*hnrəaŋH ‘cooked rice’ (2.53/21, White Hmong hno), and Hmongic *ntsuwC ‘husked rice’
(3.4/9)106. A third native word which may have originally been related to rice agriculture,
but which now has a more general meaning, is Hmongic *ʔrinA ‘to dry (rice) in sun’
(2.55/18, White Hmong ziab). The word whose presence in Chinese may be attributable to
contact with southern rice-growers is Hmong-Mien *hnɔn ‘grain head’ > ‘bag/pocket’
(2.8/24, White Hmong hnab ‘head of grain; bag/pocket’). The word is widespread in
Hmong-Mien with the meaning ‘grain head’, but appears with its metaphorically extended
meaning ‘bag/pocket’ only in Hmongic. Its counterpart in Chinese 囊 ‘sack/bag’ (OC
*nˤaŋ > MC nang > Man. náng) has only the derivative meaning. The contact situation was
more complex than this, however: Tibeto-Burman also has *s-nam ‘ear of grain’ (Matisoff
2003), which seems to be the same word.

7.5 Reckoning and measuring

As discussed in chapter 5, very few native numerals or quantity words can be reconstructed
for Hmong-Mien: the core members are *ʔu̯ i ‘two’ (7.1/14, White Hmong ob), *pjɔu
‘three’ (1.16/3, White Hmong peb), and Hmongic *ntɔC ‘many/much’ (2.4/6, White
Hmong ntau). Both ‘one’ and the higher numerals are Chinese, and Hmong-Mien
borrowed the numerals ‘four’ through ‘nine’ from some Tibeto-Burman language (‘ten’
was certainly a borrowing, although it is difficult to determine whether the source was
Chinese or Tibeto-Burman). Although it is possible that these borrowings replaced native
words for the numerals ‘four’ through ‘ten’, it is certainly possible to imagine a world in
which ‘two’, ‘three’, and ‘many/much’ would suffice for the needs of daily life.
Old terms of measure include Hmongic *ɢraŋA ‘span, (double) arm’ (6.48/24, White
Hmong daj) and *qroC ‘span, (thumb-to-middle) finger’ (6.46/7, White Hmong dos). Using
the body as a reference in measuring is characteristic of many languages of Southeast Asia:
see, for example, L-Thongkum 2002 (415-418) for a long list of measurement terms of this
type which can be reconstructed for proto-Katuic with very precise semantics.

7.6 Material culture

Words for house and parts of the house and grounds are old in Hmong-Mien: Hmong-Mien
*prəuX ‘house’ (1.46/3, White Hmong tsev), *gru̯ eŋ ‘door/gate’ (5.48/29, White Hmong
rooj); Hmongic *nthaŋA ‘attic’ (2.5/24, White Hmong nthab), *qroŋA ‘trough’ (6.46/28,
White Hmong dab).
The Hmong-Mien word for ‘fan’, *mbljæp (1.36.1/5, White Hmong ntxuam) does not
look like Chinese; it is provisionally considered to be an old native word. The Hmong-
Mien word for ‘bamboo strip’ *ɲcəuk (4.4/6, White Hmong ncau), used for tying things
together, is also quite old, however, this may be a loanword from Chinese 竹 ‘bamboo’
(OC *truk > MC trjuwk > Man. zhú). Four tools can also be reconstructed to Proto
Hmongic: *mblowD ‘whip’ (1.36/13, White Hmong nplawm), *tu̯ eiC ‘axe’ (2.1/12, White

106
Not attested in White Hmong. Another word for ‘husked rice’ is Hmongic *tshɛŋB (3.2/22, White Hmong
txhuv). This probably meant ‘millet’ originally, since the word has that meaning in Xianjin and Shimen
(West Hmongic).
 The ancient Hmong-Mien world 243

Hmong taus), *thoŋB ‘bellows’ (2.2/28)107, and *tsæC ‘billhook (brush knife)’ (3.1/5, White
Hmong txuas).
In the domain of clothes and clothes-making, the following words can be reconstructed
to the Hmong-Mien level: *ntət ‘to weave’ (2.4/7, White Hmong ntos), *ntei ‘cloth’
(2.4/12, White Hmong ntaub) and *nduH ‘ramie/hemp’ (2.6/16, White Hmong ntuag)108.
The word ‘shuttle’ can only be reconstructed with confidence to the Hmongic level—
*NGəŋB (5-6.6/21, White Hmong nqos)—although the Mienic forms for ‘shuttle’ are
probably related (*ɴɢlowB, 6.36/16). No particular clothing items can be reconstructed to
Proto Hmong-Mien, but four may be reconstructed to Proto Hmongic, and thus following
discussion above may represent items in use at the time Proto Hmong-Mien was spoken:
*tinA ‘skirt’ (2.1/18, White Hmong tiab), *nthrɔŋA ‘puttees’ (2.50/29, White Hmong
nrhoob), *bɔC ‘bracelet’ (1.3/6, Green Mong paug), *mphleA ‘finger ring’ (1.35/10, White
Hmong nplhaib).
Two other words of material culture are semantically isolated. Further study may reveal
either that they are old borrowings, or that their original meanings were not as ‘modern’ as
they appear today: Hmong-Mien *N-kjaj ‘medicine’ (5.16/15),109 and Hmongic *ntowB
‘paper’ (2.4/13, White Hmong ntawv).

7.7 Non-material culture

The only color term that dates back to Hmong-Mien is ‘black’, and perhaps ‘green/blue’—
‘white’ and ‘red’ can only be reconstructed for Hmongic. Since the Mienic languages use
the Chinese words 白 (Man. bái) and 赤 (Man. chì) for ‘white’ and ‘red’, it is probably the
case that these three color terms are equally old, and two have simply been replaced by
loanwords in Mienic. Typological research on color term sets (Berlin and Kay 1969, Kay
and McDaniel 1978, Baxter 1983) shows that if a language has only four colors, a likely
set will be ‘black’, ‘white’, ‘red’, and ‘green/blue’110—so it is probable that the Chinese
loanword ‘yellow’—黃 (OC *N-kʷˤaŋ > MC hwang > Man. huáng) > Hmongic *ɢʷaŋA (5-
6’.3/24, White Hmong daj)—and later a separate word for ‘blue’, etc. were additions to
this basic set, rather than replacements. The word for ‘black’ is difficult to reconstruct at
the Hmong-Mien level because of rime and tone mismatches between forms in the two
sub-families, although they no doubt reflect the same root: Hmongic *qrɛŋA (6.46/22,
White Hmong dub) and Mienic *qri̯ ɛkD (6.46/1). ‘Green/blue’ presents even greater
difficulties in onset and rime correspondence, but nonetheless Wang and Mao (1995:57)
put Hmongic *mpruA (1.49/16, White Hmong ntsuab) and Mienic *ʔmeŋA (1.7/29)
together. ‘White’ is Hmongic *qlowA (6.31/13, White Hmong dawb), and ‘red’ is
Hmongic *ʔlinA (2.40/18, White Hmong liab).
The old and stable family word for ‘physical affliction’ encompasses the three notions
of pain, illness, and injury that are often represented by separate words in other languages:
Hmong-Mien *ʔmun ‘illness/pain’ (1.7/27, White Hmong mob).

107
This word does not exist in White Hmong (it should have the form thav).
108
Additionally, although not important for our purpose here, are the Hmong-Mien native words *(h)naŋX
‘to put on/wear (clothes)’ (2.8/24, White Hmong hnav) and *ntɔŋH ‘to put on/wear (cap)’ (2.4/29, White
Hmong ntoo).
109
This word does not exist in White Hmong (it should have the form kuab).
110
The other possible four-term set is white-dark (black/green/blue)-red-yellow (Baxter 1983:4–5).
244 Chapter 7

Words in the semantic field of belief can be reconstructed for both Hmongic and Mienic
separately: Hmongic *qraŋA ‘spirit/ghost’ (6.46/24, White Hmong dab), Mienic *hmjænB
‘spirit/ghost’ (1.23/19)111, and Hmongic *bljaA ‘soul’ (1.33.1/4, White Hmong plig).
‘Sun/day’ (Hmongic *hnɛŋA [2.8/22 White Hmong hnub] and Mienic *hnu̯ ɔiA [2.8/11])
and ‘moon/month’ (Hmong-Mien *hlaH [2.41/4, White Hmong hli]) are ancient, but
clearly not native: they were borrowed from Tibeto-Burman *s-nəy and *s-la as a pair
(Benedict 1987a, Mortensen 2002). Benedict reasonably hypothesized that they were
associated with sun and moon cults, and were borrowed because of the nature of the
cultural contact between the two groups. Of the words for musical instruments, only the
word ‘drum’ is native and ancient: Hmong-Mien *ndruX (2.51/16, White Hmong nruas). It
best fits into this category because of the common use of musical instruments in religious
ceremonies.

111
It is possible that two of these words are loanwords: for Hmongic ‘soul’ (< Tibeto-Burman *b-la), see
Benedict 1987a and Mortensen 2002; for Mienic ‘spirit/ghost’ (< Old Tibetan *sman or Kam-Sui
*hmaaŋ) see Ratliff 2001a.
 Index of proto forms by English

OC=Old Chinese (Baxter and Sagart 2009); MC=Middle Chinese (Baxter 2000);
Mandarin; O=onset category; R=rime category; English gloss; Reconstr=Reconstruction;
L=level: (1) Hmong-Mien, (2) Hmongic, (3) Mienic; WO=Wang and Mao 1995 onset
category; WR=Wang and Mao 1995 rime category;112 Chap=chapter and section where
word is discussed.

OC MC Man O R English Reconstr L WO WR Chap

1.12 22 1SG (I/me) *wɛŋB 2 12 215 5.3
5.3,
5.1 22 1SG (I/me) *kɛŋB 2 541 268 6.3.2,
6.3.4
4.10 4 1SG (I/me) *ʔjaA 3 442 130 5.3
5.2,
1.1 3 1PL (we/us) *N-pɔu 1 61 244
5.3
5.3,
1.9 8 2SG (you) *mu̯ei 1 21.1 242 6.3.2,
6.3.4
5.3 8 2SG (you) *gʉA 2 xx xx 5.3
5.3,
1.9 1 2PL (you-all) *mi ̯əu 1 xx xx
6.3.4
3SG 2.1.4,
2.9 1 *ni ̯æn(X) 1 213.1 81
(he/she/it) 5.3
3.3 13 able *dzowA 2 159 263
*prəm-s pjuwngH fèng
風 1.1 28 air *poŋC 2 1 178
*prəm pjuwng fēng
4.3 18 alive *ɟinA 2 xx xx
6.33 21 Allium *ɢləŋ B
2 651 138 7.2
韭 *[k](r)uʔ kjuwX jiǔ 4.3 A Allium *ɟiu B
3 435 3
5.48 21 animal fat/oil *grəŋ A
2 279 214 4.4
1.8 10 animal fat/oil *hmej A
3 8.1 37
2.1 3 answer *tau 1 205 95
1.5 5 ant *mphæ B
2 5 168 4.2
1.51 6 ant *mbrɔ D
2 78 241 6.3.3
4.5 1 ant *ɟhi ̯əu B
3 185 210
3.16 7 armpit *tsjo C/D
2 313 258

112
Words that have “xx” under the Wang and Mao 1995 onset and rime category columns (WO and WR)
come either from Wang 1994 or the author’s notes.
245
246 Indexes

OC MC Man O R English Reconstr L WO WR Chap

2.56 22 artemisia *hrɛŋB 2 581 100
慚 *[dz]ʕam dzam cán 3.3 24 ashamed *dzaŋA 2 111 158
3.17 12 ashes *tshju̯əiX 1 326 207
2.9 22 ask *nɛŋ C
2 225 61
2.9 11 ask *nu̯ai C
3 225 61
xx 28 ask for *[n/l]oŋ C
3 213.3 216 6.2.1
要 *ʔew-s ʔjiewH yào 4.10 16 ask for/want *ʔju B
2 442 191
5.21 29 asleep *ŋgjɔm A
3 474 192
2.1.2.2,
2.5 24 attic *nthaŋA 2 209 121
7.6
7.1,
2.1 12 axe *tu̯eiC 2 205 39
7.6
斧 *p(r)aʔ pjuX fǔ 1.1 13 axe *pouB 3 1 263 7.1
6.3 9 back (of body) *ɢuw D
2 615 204
2.3 19 back (of body) *danB 3 207 80
5-6.1 24 back (vs. front) *N-Kaŋ 1 613 45 2.1.2.2
腐 *[b](r)oʔ bjuX fǔ 1.3 15 bad/spoiled *bu̯a B
2 3 151
袋 *-lˤək-s dojH dài 2.3 1 bag *di C
3 207 1
炙 *tAk tsyek zhì 4.1 1 bake/toast *ciC
2 xx xx 4.4
秤 *thəŋ-s tsyhingH chèng 2.20 29 balance *nthju̯əŋH 1 376 102 4.2
2.41 7 bamboo *hləwX 1 215 206
2.48 13 bamboo *drow D
2 255 263
5-6.2 29 bamboo bowl *khɔk D
3 542 205
5.1 9 bamboo hat *kuwD 2 541 224 2.1.2.2
笠 *(kə.)rəp lip lì 2.57 7 bamboo hat *rəp D
3 605 106
5-6.3 19 bamboo pipes *ɢæn A
2 xx xx
bamboo shoot
1.51 15 *mbrəjH 1 78 94 6.3.2
/orphan
竹 *truk trjuwk zhú 4.4 6 bamboo strip *ɲcəuk 1 352 241 7.6
4.1 21 bark, to *cuŋ C
3 433 218
3.2,
廩 *(pə.)r[ə]mʔ limX lǐn 2.57 22 barn/granary *rɛmX 1 605 174
6.2.2
1.4 9 bat (animal) *ʔbujA 3 16 243
4.12 21 be *jəŋ C
2 444 214 4.4
是 *[d]eʔ dzyeX shì 2.18 10 be *djej B
3 351 37
4.7 21 be at/live *ʔɲəŋ A
2 403 229 4.4
4.10 21 be at/live *ʔjəm A
3 403 229 4.4
荅 *[t]ˤ[ə]p top dá 2.3 9 bean *dup 1 207 271 7.4
腐 *[b](r)oʔ bjuX fǔ 7.13 8 bean curd *hʉB 2 xx xx
6.46 10 bear *qrep 1 673 124 7.3
1.16 1 bear fruit *pji C
2 xx xx 4.4
 Index of proto forms by English 247

OC MC Man O R English Reconstr L WO WR Chap

髯 *nam nyem rán 4.9 24 beard *ɲaŋC 2 405 45
髟 *s[ˤ]ram sræm shān 2.13 29 beard *su̯am A
3 203 211
beat (someone) *ntuw
D
2.4 9 2 208 110
緻 *[d]ri[t]-s drijH zhì 3.3 10 beautiful *dzej C
3 174 37
處 *t.q aʔ-s h
tsyhoH chù 4.2 13 bed/place *chouH 1 350 263
1.16 1 bedbug *pji 1 25 1 4.2
1.9 8 bee *mu̯ei B
3 21.1 37
4.1 24 beeswax *N-ci ̯æŋ 1 457 139
橐 *t ˤakh
thak tuó 2.2 28 bellows *thoŋ B
2 206 178 7.6
蘆 *[r]ˤa lu lú 2.42 13 bellows *lou A
3 216 263
4.2 12 belly *chu̯ei A
2 434 208
2.41.1 7 big *hljo 1 287 225 4.4
3.1 5 billhook *tsæ C
2 157 149 7.6
2.1.2.1,
2.1.4,
2.9 29 bird *m-nɔk 1 213.2 205
4.2.4,
6.3.2
2.3 7 bite *dəp 1 207 106 6.3.4
7.1 18 bitter *ʔim 1 709 4
6.46 22 black *qrɛŋA 2 673 252 7.7
6.46 1 black *qri ̯ɛkD 3 673 110 7.7
3.18 5 bland/tasteless *dzjæ C
2 315 171
/tsjäm/ 113
3.1 24 bland/tasteless *tsam B
3 109 154
1.34 3 blind *ʔblɔu B
3 16 244
3.20 24 blood *ntshjamX 1 341 97 6.3.3
1.47 16 blow, to *phruA 2 74 191
5.21 21 boat/ship *ŋgjəŋA 2 474 214
3.4 24 boat/ship *ʔdzaŋB 3 160 101
4.1 3 body/trunk *cæwB 2 xx xx 4.4
沸 *-p[u][t]-s pjɨjH fèi 1.4 12 boil (INTR) *mpu̯æiH 1 16 39
2.2 7 boil (TR) *tho A
2 xx xx
7.13 12 boil (TR) *hu̯eiC 2 731 39
煮 *taʔ tsyoX zhǔ 2.16 13 boil (TR) *tjou B
3 349 263
3.2 28 bone *tshuŋX 1 146 180 4.2.3
假 *Cə.kˤraʔ kæX jiǎ 5-6.1 4 borrow *KaX 1 613 92
2.3 10 bowl *de B
2 207 73 7.1
碗 ʔwanX wǎn 1.10 21 bowl *ʔwən B
3 10 157 7.1
1.3 6 bracelet *bɔ C
2 3 187 7.6
5.18 21 bracelet *gjəmA 3 472 229

113
Early Min (Norman 1988:232)
248 Indexes

OC MC Man O R English Reconstr L WO WR Chap

辮 *-[b]ˤe[n]ʔ benX biàn 1.21 18 braid/a braid *mbjinX 1 54 7 4.4
拗 *qˤriwʔ ʔæwX ǎo 2.40 7 break *ʔlo B
2 211.2 206 6.2.1
拗 *qˤriwʔ ʔæwX ǎo 2.7 7 break *ʔnəu B
3 211.2 206 6.2.1
2.49 21 breast/chest *ntrəŋ A
2 xx xx
24 breed/propagate *ɕaŋ
C
4.13 2 443 121
橋 *[g](r)aw gjew qiáo 4.3 16 bridge *ɟow 1 435 172
梯 *l ̥[ə]j thej tī 2.41 5 bridge *hlæ A
2 215 168
黃 *N-kʷˤaŋ hwang huáng 5-6’.4 24 bright *ʔgʷi ̯əŋ A
3 640 121
黃 *N-kʷˤaŋ hwang huáng 5-6’.3 24 bright/light *ɢʷaŋ A
2 663 121
1.9 3 brother *mɔu B
3 21.1 77
3.2,
桶 thuwngX tǒng 2.2 29 bucket *thɔŋ(X) 1 206 142
6.2.4
蕎 gjew qiáo 4.3 3 buckwheat *ɟæu 1 435 259 7.4
1.1 3 burn *pɔu B
3 13 77
灼 *[t]ewk tsyak zhuó 2.18 1 burn *dji ̯ɛk D
3 183 148
burn
燎 *[r]awʔ ljewX liǎo 2.41.1 13 *hljowB 2 287 263
(mountain)
1.17 3 burn (TR) *phjæwB 2 50 244
5.46 12 burn (TR) *kru̯ei B
2 277 39
炙 *tAk tsyek zhì 4.3 1 burn/be alight *ɟiC
2 183 148 4.4
斂 *[r][a]m-s ljemH liàn 2.42.1 21 bury *ljəŋ B
2 288 214
1.31 9 bury *plop D
3 73 271 6.3.2
鬥 *tˤok-s tuwH dòu 2.49 9 butt (of bulls) *ntruw C
2 256 243
1.34 2 butterfly *mpleuH 1 52 35
2.1.3.1,
買 *m.rˤajʔ mɛɨX mǎi 1.9 11 buy *mɛjX 1 9.1 93
4.4
奉 *pʰ(r)oŋʔ phjowngX fèng 1.2 22 carry in 2 hands *phu̯ɛŋ
B
2 14 252
4.8 4 carry on back *hɲaC 3 403 130
carry on
擔 *mə-tˤam tam dān 2.4 24 *ntam 1 208 97
shoulder
carry on
4.1 3 *cæwC 2 xx xx
shoulder
1.9 29 carrying pole *mu̯əŋB 3 21.1 85
圉 *m-q(ʰ)(r)aʔ ngjoX yǔ 5.36 16 cattle pen *ŋgluə 1 546 191
斤 *[k]ər kjɨn jīn 5’.16 20 catty *kʷjan 1 481 23 2.1.2.2
中 *truŋ trjuwng zhōng 2.49 28 center/middle *ntroŋ 1 256 178
5.32 9 centipede *khluw D
2 578 153 4.2.4
4.2.4,
2.13 7 centipede *səpD 3 167 106
6.3.3
1.5 1 chaff/husk *mphi ̯ɛk 1 40 148 2.1.2.2
1.17 5 chaff/husk *S-phjæC 2 59 149
 Index of proto forms by English 249

OC MC Man O R English Reconstr L WO WR Chap

炭 *[tʰ]ˤa[n]-s thanH tàn 2.2 19 charcoal *thanH 1 206 99 3.2
1.32 xx cheek/face *phl- C
2 xx xx
1.10 4 cherry *ʔwa A
2 xx xx
嚼 *[dz]ewk dzjak jué 4.1 15 chew *cu̯a C
2 433 94
2.1.2.2,
雞 *kˤe kej jī 5-6.1 10 chicken *Kəi 1 613 261
6.3.2
/kiãi3/114 5’.16 12 child *kʷjeiB 3 481 39
10 chin/lower jaw *dzje
D
3.18 2 xx xx
鑿 *[dz]ˤawk dzak záo 3.3 6 chisel *dzəuk 1 111 241 4.2
3.19 20 choke *ntsjeŋH 1 436 44
2.4 7 chop *ntəwX 1 208 206
箸 *[d]rak-s drjoH zhù 2.48 13 chopsticks *drouH 1 255 263 3.2
拍 *mə-pʰˤrak phæk pāi 1.6 5 clap *mbæ A
2 6 168
拍 *mə-pʰˤrak phæk pāi 1.5 5 clap *bhæk D
3 28 72
4.7 5 claw/talon *ʔɲau B
3 403 152
clay pot
7.13 22 *hɛŋA 2 719 47
/earthen jar
3.5 2 clean *ntshɨA 2 161 35
淨 *N-[dz]eŋ-s dzjengH jìng 3.6 21 clean *ndzəŋ C
3 114 140
清 *N-s-r̥eŋ tshjeng qīng 3.20 18 clear *ntshji ̯əŋ 1 317 159 4.3
隻 *tek tsyek zhī 3.1 1 CLF-birds *tsi ̯ɛk D
3 157 148
CLF- 6.2.1,
2.40 22 *ʔlɛŋA 2 211.2 173
bowls/houses 6.2.4
CLF-bowls 6.2.1,
2.7 22 *ʔnɛɔmA 3 211.2 173
/houses 6.2.4
2.40 1 CLF-fields *ʔli ̯əuC 3 238 210
2.3 22 CLF-horses *dɛŋ B
2 207 100 6.2.4
2.40 7 CLF-mouthfuls *ʔlo C
2 214 225 6.2.4
2.42 19 CLF-people *lan 1 216 118
片 *pʰˤe[n]-s phenH piàn 1.2 21 CLF-quilts *phəan 1 2 194 6.2.4
4.2 28 CLF-rooms *choŋ B
2 xx xx
句 *[k](r)o-s kjuH jù 5.16 13 CLF-sentences *kjou C
3 397 263
張 *traŋ trjang zhāng 2.46 24 CLF-tools *traŋ A
2 265 250 6.2.4
張 *traŋ trjang zhāng 2.46 28 CLF-tools *truŋ A
3 265 250 6.2.4
4.4 3 climb *ɲcæw C
2 436 96
2.4 12 cloth *ntei 1 232 208 7.6
2.55 9 clothes *ʔruj A
3 580 260
7.1 29 cloud/haze *ʔɔŋ C
2 709 248

114
Pucheng Min (Norman 1988:243)
250 Indexes

OC MC Man O R English Reconstr L WO WR Chap

4.3,
5-6.4 19 cogongrass *NKan 1 616 99
7.2
2.7 27 cold *ʔnu̯ɔmH 1 211.1 98
5.16 22 cold (day) *kju̯ɛŋ B
3 481 252
2.57 5 comb *ræ C
2 582 149
2.3 15 come *daj 1 207 150 2.1.3.1
2.42 7 come back *ləwX 1 216 206
2.55 10 conceal *ʔreD 2 580 48
2.28 20 cooked/ripe *sjen B
2 323 24
熟 *[d]uk dzyuwk shú 2.18 B cooked/ripe *dju̯ok D
3 363 258
cooking
1.27 18 *wjinB 2 36 7
pot/pan
4.5 7 cooking smoke *ɲchi ̯ouH 1 437 134
3.3 18 cool (water) *dzinB 2 159 83
3.2,
銅 *-lˤoŋ duwng tóng 2.3 29 copper *dɔŋ 1 207 142
7.1
2.8 29 cough *hnɔp 1 212.1 254
18 cover (with tile) *hi ̯əm
B/C
7.13 3 720 135 6.3.2
1.10 7 cover self *ʔwoB 2 10 206
1.6 7 cover, to *mbo C
2 6 225
5.1 21 cover, to *kom B
3 553 193 6.3.2
7 cow/ox *jo B
4.12 2 xx xx
匠 *S.[b]aŋ-s dzjangH jiàng 3.3 26 craftsman *dzi ̯ɔŋH 1 111 65
5.21 28 crawl *ŋgjoŋ C
2 474 216
4.7 22 crawl *ʔɲɛɔŋ A
3 403 215
1.36.1 29 crazy *mbljɔŋ A
2 xx xx
冠 *[k]ˤon kwan guān 1.10 1 crest/comb *ʔwi ̯æn 1 10 81
曲 *kʰ(r)ok khjowk qū 5.5 9 crooked *ŋkhuw D
2 545 224
crop (of bird)
4.1 18 *ciŋB 3 400 8
/gizzard
1.16 18 crop of bird *pjinB 2 49 43
5-6.1 29 crops *qɔŋA 2 xx xx
過 *[k]ʷˤaj-s kwaH guò 5-6’.1 15 cross (river) *KʷajH 1 661 112
2.1.4,
2.8 19 crossbow *hnænB 2 212.1 129 4.2,
7.3
2.1.4,
2.8 7 crossbow *hnəkD 3 212.1 129
7.3
烏 ʔu wū 7.1 16 crow (bird) *ʔuA 2 xx xx
歌 *[k]ˤaj ka gē 5-6.1 15 crow, to *KajH 1 625 150
6.3 8 cuckoo *ɢʉ B
2 xx xx
瓜 *kʷˤra kwæ guā 5-6’.1 4 cucumber *Kʷa 1 661 92 2.1.2.2
 Index of proto forms by English 251

OC MC Man O R English Reconstr L WO WR Chap

5.31 5 cut *klæp 1 577 144
割 *mə-kˤat kat gē 5-6.1 4 cut *katD 3 553 88
破 *pʰˤaj-s phaH pò 1.2 15 cut open *phajH 1 2 150 6.2.2
1.46 9 dark *pruw D
2 73 110
1.20 10 daughter/girl *mphjeD 2 53 48 2.1.2.2
媳 sik sī 4.13 1 daughter/girl *ɕi ̯ɛkD 3 359 148
daughter-in-law
婦 *Cə.[b]əʔ bjuwX fù 1.6 22 *mbu̯ɛŋB 3 18 252
/bride
聾 *[r]ˤoŋ luwng lóng 2.42 28 deaf *loŋA 2 216 179
2.4 21 deaf *ʔduŋ A
3 244 218
2.1 7 deep *N-tu̯o 1 205 225 4.3
獐 *taŋ tsyang zhāng 4.1 28 deer *cuŋ A
3 397 249
4.3,
下 *[g]ˤraʔ hæX xià 5-6.6 4 descend *ɴɢaB 2 618 130
6.2.2
*mə- 4.3,
下 hæH xià 5.13 4 descend *ɣaC 3 402 130
[g]ˤraʔ-s 6.2.2
露 *[r]ˤak-s luH lù 2.42 8 dew *lʉC 2 xx xx
4.4,
2.3 15 die *dəjH 1 207 151 6.3.2,
6.3.4
抉 *qʷˤet ʔwet jué 1.10 10 dig *ʔwetD 3 10 50
5-6.6 15 diligent *NGəjH 1 630 151
嫌 *N-kˤem hem xián 5.13 20 dislike *ɣi ̯em A
3 402 40
3.5 19 disperse *dzhan C
3 161 99
D
4.12 6 dissolve *jəuk 3 444 241
煬 *laŋ yang yáng 4.12 24 dissolve/melt *jaŋ A
2 456 66
煬 *laŋ yang yáng 4.12 28 dissolve/melt *juŋ A
3 456 66
5.1 28 ditch *koŋ A
2 541 26
7.1 15 do/work *ʔəjH 1 709 151
狗 *-[k]ˤ(r)oʔ kuwX gǒu 6.31 3 dog *qluwX 1 649 115 4.2.4
5.48 29 door/gate *gru̯eŋ 1 279 46 7.6
龍 *mə-roŋ ljowng lóng 2.57 28 dragon *-roŋ 1 582 216
龍 *mə-roŋ ljowng lóng 5.31.1 28 dragon *kljuŋ A
3 601 249
4.3,
1.4 12 dream *mpeiH 1 4 208
6.3.3
欱 *qʰˤ[ə]p xop hē 7.13 9 drink/smoke *hup 1 719 271 6.2.2
漏 *[r]ˤo-s luwH lòu 5.51 7 drip *ŋgro C
2 282 206
2.19 9 drip *ʔdjɛp D
3 424 163
3.19 18 drive in (nail) *ntsjin C
2 316 5
2.51 16 drum *ndruX 1 258 111 7.7
3.1 2 dry (clothes) *tsek D
3 157 17
252 Indexes

OC MC Man O R English Reconstr L WO WR Chap

18 dry (rice) in sun *ʔrin
A
2.55 2 580 119 7.4
5-6.5 5 dry/thirsty *NKhæj 1 617 169
槀 *[k ]ˤawʔ
h
khawX gǎo 5-6.2 5 dry/withered *KhæwX 1 614 114 2.1.2.2
鴨 ʔæp yā 7.1 7 duck *ʔap 1 709 105
6.3 15 duck *ɢu̯a C
2 627 151
染 *n[a]mʔ nyemX rǎn 4.9 27 dye, to *ɲumC 3 405 136 6.3.2
6.31 24 eagle/hawk *qlaŋX 1 649 120 4.2.3
1.51 3 ear *mbræu 1 78 113
早 *Nə.[ts]ˤuʔ tsawX zǎo 3.4 7 early *ntsi ̯ouX 1 184 134
2.1 3 earth *N-təu 1 205 96 4.3
160, 250,
4.1 28 earthworm *N-cuŋ 1
433 180
2.9 27 eat *nuŋA 2 237 62 2.1.4
4.9 27 eat *ɲən C
3 237 62 2.1.4
5-6.1 3 egg *qæw C
2 685 96
5.31 3 egg *kləuC 3 685 96
茄 gja qié 4.3 4 eggplant *ɟa 1 435 130
5.2,
4.12 4 eight *jat 1 444 127 6.3.1,
7.5
2.3 4 elder brother *da 1 207 130
elder sister
媼 *ʔˤuʔ ʔawX ǎò 7.1 3 *ʔəuX 1 709 96
/wife
3.5 8 elephant *ntshʉC 2 161 242
象 *s-[d]aŋʔ zjangX xiàng 5.30 26 elephant *ɣji ̯ɔŋ B
3 195 65
伏 *[b]ək-s bjuwH fú 1.3 16 embrace *buəH 1 15 172
1.18 9 enter *bju̯ɛk 1 27 224
立 *(kə.)rəp lip lì 2.57 1 erect, to *ri ̯ɛp D
3 605 162
脫 *lˤot dwat tuō 5-6’.3 4 escape *ɢʷa D
2 663 203
2.1.2.2,
5-6.1 15 excrement *N-KəjX 1 613 151
4.3
2.3 13 explode *douH 1 207 263 4.4
目 *[m][u]k mjuwk mù 1.9 11 eye *mu̯ɛjH 1 21.1 75
1.3 13 face *bow B
2 3 263 7.1
面 *[m]e[n]-s mjienH miàn 1.8 C face *hmien A
3 32.1 6 7.1
崩 *Cə.pˤəŋ pong bēng 1.1 30 fall *pʉŋ A
2 1 9 4.4
頹 dwoj tuí 2.3 9 fall (of rain) *duj A
3 207 243
6.3 12 fall (over) *ɢu̯ei C
2 615 39
1.36.1 5 fan *mbljæp 1 78 86 7.6
迂 *q (r)a
w
ʔju yū 6’.1 3 far *qʷuw 1 661 115 6.2.2
5.48 21 fat, to be *grəunH 1 279 214 4.4
 Index of proto forms by English 253

OC MC Man O R English Reconstr L WO WR Chap

父 *[b](r)aʔ bjuX fù 1.16 4 father/male *pjaX 1 61 92
3.20 10 fear/be afraid *ntshjeH 1 341 73 4.3
1.1 22 feed *pɛŋ A
2 xx xx
1.9 18 female *min D
2 33.2 13
female
1.3 7 *boA 2 xx xx
/woman
2.56 16 fern *-hruA 2 581 191
蕨 *-kot kjwot jué 5’.16 7 fern *kʷjət D
3 481 146
3.18 13 few/lack *dzjok 1 315 186
田 *lˤiŋ den tián 2.42.1 18 field *ljiŋ 1 312 8
1.9 21 fine (flour) *mənX/H 1 21.1 176
2.4 4 finger *ntaB 2 208 203 2.1.4
2.4 13 finger *ʔdokD 3 208 203 2.1.4
2.1.2.2,
1.35 10 finger ring *mphleA 2 xx xx
7.6
4.3 29 finish *ɟɔŋB 2 399 123
2.3 13 fire *douX 1 207 263 4.4
A
3.3 24 firewood *dzaŋ 3 111 101
2.48 16 firm/durable *dru A
2 xx xx
first day of
2.13 20 *seŋ 1 167 44
month
4.2.2,
1.51 3 fish *mbrəuX 1 78 96 4.3,
6.3.2
5.2,
1.46 4 five *pra 1 73 92 6.3.1,
7.5
1.8 23 flea *hmiəŋ 1 20 269
2.4 2 flip with finger *ntik 1 208 15
2.4 24 float *ntaŋA 2 xx xx
浮 *m.b(r)u bjuw fú 1.6 1 float *mbi ̯əu A
3 30 210
1.31 12 flour *plu̯ei B
2 85 208
6.36 8 flow *ɴɢlʉ B
2 654 242
流 *[r]u ljuw liú 2.57 1 flow *ri ̯əu C
3 605 210
2.1.2.2,
1.3 24 flower *bi ̯aŋ 1 39 121
4.2
筒 *lˤoŋ duwng tóng 2.48 28 flute/tube *droŋA 2 255 178
蠓 *mˤoŋʔ muwngX měng 1.9 21 fly (N.) *məuŋX 1 9.1 218
4.10 24 fly, to *ʔjaŋ C
2 442 121
2.4 15 fly, to *ʔdəi C
3 208 151
254 Indexes

OC MC Man O R English Reconstr L WO WR Chap

霧 *m(r)o-s mjuH /mou /
6 115
1.9 16 fog *mowC 3 9.1 172
7.13 16 fog/cloud *hu A
2 719 191 7.1
fool/humor
5.33.1 10 *glje(X) 1 603 73
(children)
2.1 13 foot *towC 2 205 263
6.3.1,
1.23 19 footprint/track *hmjænX 1 32.2 247
7.3
1.33 xx forehead *bl-A 1 87 197
4.8 20 forehead *hɲen A
2 404 24
2.55 29 forest *-ʔrɔŋ B
2 580 102 2.1.2.2
2.8 27 forget *hnuŋ B
2 236 58
4.8 13 forget *hɲou B
3 236 58
5.2,
1.31 12 four *plei 1 85 208 6.3.1,
7.5
2.4 24 fragrant *ntaŋ 1 208 158 6.3.2
5-6.1 24 frog *KæŋX 1 625 63 2.1.2.2
6.36 18 front surface *ɴɢlin A
2 654 64
2.1 3 frost *tæw C
2 205 96
1.16 1 fruit *pji ̯əuX 1 49 210 4.4
攪 *kˤruʔ kæwX jiǎo 5.31 2 fry *kleu 1 541 35
炒 tsrhæwX chǎo 4.2 5 fry *chau B
3 350 171
1.1 22 full *pu̯ɛŋX 1 61 252 6.3.3
3.16 1 gall bladder *tsji ̯əu 1 313 210
園 *[ɢ]ʷan hjwon yuán 1.12 24 garden *waŋA 2 12 45
園 *[ɢ]ʷan hjwon yuán 1.11 27 garden *hwun A
3 11 231
葫 *[g]ˤa hu hú 5-6.3 4 garlic *ɢa A
2 615 92
5’.18 4 get up *gʷjaX 1 507 130
得 *tˤək tok dé 2.1 6 get/gain *təuk 1 205 241 6.2.3
5-6.2 18 ginger *qhin B
2 638 83
薑 *kaŋ kjang jiāng 5.17 21 ginger *khjuŋ A
3 167 218
秫 *m.lut zywit shú 1.36 9 glutinous/sticky *mblut 1 90.1 239
5.1 13 gnaw *kow D
2 541 273 2.1.2.2
2.1.2.1,
*n- 2.1.4,
1.9 30 go 1 9.2 9
mʉŋ(X) 4.3,
4.4
3.2,
金 *[k](r)[ə]m kim jīn 5.16 23 gold *kjeəm 1 469 265
7.1
2.55 29 good *-ʔrɔŋH 1 580 142 2.1.2.2

115
Cantonese
 Index of proto forms by English 255

OC MC Man O R English Reconstr L WO WR Chap

5.2 3 good/tasty *khuB 3 542 115
2.1 9 gourd *tuwA 2 205 260
212.1, 121,
囊 *nˤaŋ nang náng 2.8 24 grain head/bag *hnɔn 1 7.4
236 45
5-6.1 18 grape *qinB 2 613 119
1.7 11 grasp *ʔmu̯ɛA 2 7 75 4.4
1.23 4 grass *hmjaB 3 44 130 2.1.2.2
5.31 9 grasshopper *klup 1 517 271 6.3.3
蝗 *[ɢ]ʷˤaŋ hwang huáng 5.3 30 grasshopper *gʉŋA 2 543 233
3.4 24 grave/tomb *ntsaŋC 2 160 66
3.1 13 grave/tomb *tsouB 3 109 263
1.49 16 green/blue *mpruA 2 76 57 7.7
1.7 29 green/blue *ʔmeŋA 3 76 57 7.7
1.4 3 grey *ʔbɔuB 3 16 77
2.57 7 grind (grain) *ro D
2 582 222
1.51 24 groan/moan *mbraŋ A
2 78 250
種 *toŋʔ-s tsyowngH zhòng 2.16 22 grow *n-tju̯ɛŋH 1 169 252
4.3 1 grow (TR) *ɟi
B
2 435 210
客 *kʰˤrak khæk kè 5-6.2 5 guest *Khæk 1 xx xx 2.1.4
1.31.1 12 hair *pljei 1 97 208
1.17 10 half *phjeC 2 xx xx
*N-
2.3 24 half (of day) 1 207 97 4.3
dam(X)
里 *(mə.)rəʔ liX lǐ 2.42.1 2 half kilometer *ljɨX 1 288 35
2.1.2.1,
1.3 3 hand/arm *-bɔuX 1 63 77
4.2.3
硬 ngængH yìng 5.24 20 hard/solid *ŋjeŋC 3 573 44
伏 *[b]ək-s bjuwH fú 1.3 16 hatch *buəH 1 15 191
2.1.2.1,
2.1.3.1,
1.9 11 have *n-mɛj 1 9.2 93
4.3,
4.4
缺 *[k]ʷʰˤet khwet quē 5-6’.4 10 have a gap *NKʷet 1 664 91
have food with
1.36.1 5 *mbljæC 2 xx xx
rice
11,
1.47 12 head *S-phreiX 1 208
97
2.8 21 hear *hnəumX 1 212.1 193
1.31.1 13 heart *pljow B
2 97 263 7.1
心 *səm sim xīn 2.13 18 heart *sim A
3 167 5 7.1
2.23 25 heavy *hnjeinX 1 404 26
256 Indexes

OC MC Man O R English Reconstr L WO WR Chap

/pɛːŋ /
5 116
1.1 18 hide oneself *piŋC 3 1 8
2.56 18 high/tall *-hri ̯əŋ 1 581 159 2.1.3.1
著 *[d]rak drjak zhuó 2.48 6 hit target *drɔ C
2 255 187 4.4
1.8 30 Hmong *hmʉŋ A
2 8.2 9
1.4 5 hold in mouth *mpæ B
2 4 168
含 *-[g]ˤ[ə]m hom hán 5.19 29 hold in mouth *ʔgjɔm A
3 568 192
hold with 2
抱 bawX bào? 1.3 13 *bouA 3 3 263
hands
hold/grasp hand *tæ
B
2.1 5 2 xx xx
孔 *[k]ʰˤoŋʔ khuwngX kǒng 5-6.2 21 hole *qhəŋ B
2 614 218 6.2.4
堀 *k.tʰˤut khwot kū 5.2 13 hole *khot D
3 554 256
2.55 4 honey *ʔra A
2 xx xx
蹄 *[d]ˤe dej tí 2.3 10 hoof *dej 1 207 37
蹢 *tˤek tek dí 2.46 12 hoof/claw/nail *tru̯ei C
2 253 208
鉤 *mə-qˤ(r)o kuw gōu 5-6.4 3 hook *ɴqæw C
2 616 115
2.2 2 hoop *thɨC
2 206 35
角 *k.rˤok kæwk jiǎo 5.31 22 horn *klɛɔŋ 1 577 215
馬 *mˤraʔ mæX mǎ 1.24 19 horse *mjæn B
2 33.2 247 6.3.1
517, 245,
5.32 22 hot (water) *k(h)lɛɔm 1
542 100
1.46 3 house *prəuX 1 73 96 7.6
houttuynia
2.46 9 *truwD 2 253 224 7.2
cordata
2.1.4,
百 *pˤrak pæk bǎi 1.1 5 hundred *pæk 1 1 72 3.2,
5.2
3.17 10 hungry *tshje 1 314 73
父 *[b](r)aʔ bjuX fù 1.1 7 husband *N-poX 1 1 225
擣 *tˤuʔ tawX dǎo 2.1 16 husk/pound rice *tuX 1 205 111
6.46 12 ice *qru̯ei D
2 697 167
2.1 13 ignite/light *tow D
2 xx xx 4.4
1.7 27 illness/pain *ʔmun 1 7 195 7.7
2.6 24 in front *ndaŋ C
3 210.1 101
藍 *Cə.[r]ˤam lam lán 5.36 24 indigo *ŋglam 1 522 154
4.2,
4.4,
5.31 24 insect/worm *klæŋ 1 517 63
6.2.4,
6.3.4
*mə- 137,
插 tsrhɛp chā 2.47 10 insert *threp 1 87
[tsʰ]<r>op 254

116
Cantonese
 Index of proto forms by English 257

OC MC Man O R English Reconstr L WO WR Chap

2.51 21 interior *ndrəŋC 2 xx xx
4.8 23 intestines *hɲeuŋX 1 404 153
腸 *lraŋ drjang cháng 6.33.1 24 intestines *ɢljaŋ A
3 675 120
3.2,
鐵 *l ̥ˤik thet tiě 2.41 9 iron *hluwC 2 311 224 6.2.2,
7.1
3.2,
鐵 *l ̥ˤik thet tiě 2.56 9 iron *hrɛkD 3 311 224 6.2.2,
7.1
itch(y)/
5.17 9 *khju̯ɛt 1 470 70
scratch(y)
3.1 5 join *tsæD 2 xx xx
4.4,
2.1 15 kill *təjH 1 205 151
6.3.4
4.3 12 knee *ɟu̯əiH 1 447 39
別 *N-pret bjet bié 1.1 12 know *pei 1 1 39
識 *s-tək syik shí 3.1 1 know/recognize *tsi ̯ɛk
D
3 181 148
漆 *[tsʰ]i[t] tshit qī 2.17 10 lacquer *thjet 1 134 51 3.2
7.13 10 ladle, to *he D
2 719 68
斟 *t.[q][ə]m tsyim zhēn 2.4 22 ladle, to *ʔdəm B
3 208 174
land,
地 *[l]ˤej-s dijH dì 2.3 10 *dejC 3 207 37
non-irrigated
2.42 1 late *liC 2 216 81
遲 *lrəj drij chí 2.18 12 late *djæi A
3 351 132
5.46 7 laugh *krət 1 277 222
3.5 16 launder *ntshuH 1 161 111
2.6 3 lay eggs *ndəuH 1 210.1 96
懶 *[r]ˤanʔ lanX lǎn 5.36 19 lazy *ŋglæn B
2 546 247
累 *[r]ˤuj-s lwojH lèi 2.42 8 lazy *lu̯ei C
3 228 39
4.1 28 lead, to *coŋ A
2 433 216
牽 *[k]ʰˤi[n] khen qiān 5-6.2 C lead, to *khienA 3 542 6
1.36 29 leaf *mblɔŋ A
2 90.1 192 6.2.4
2.9 29 leaf *nɔm A
3 90.1 192
5.51 4 lean (on stick) *ŋgraC 2 282 92
4.3 3 leg/branch *ɟæwB 2 xx xx 4.4
3.16 21 let go *tsjəŋC 2 313 218
放 *paŋ-s pjangH fàng 1.1 21 let go *puŋC 3 1 218
2.3 18 level *dinA 2 xx xx
平 *breŋ bjæng píng 1.3 29 level *beŋ A
3 3 46
4.12 10 lick *-jep 1 444 87
5.16 20 lift *kjeŋ 1 397 44
258 Indexes

OC MC Man O R English Reconstr L WO WR Chap

拎 /nɪŋ1/117 2.8 30 lift *hnɨŋC 3 212.1 9
2.28 2 light(weight) *-sji ̯eu 1 59 35
2.42.1 10 lightning flash *lje D
2 288 91
2.42.1 7 lightning flash *ljəp D
3 498 106
5.31 5 liquid/soup *klæwX 1 xx xx
酒 *[s.q]uʔ tsjuwX jiǔ 4.1 13 liquor *cow B
2 433 263
酒 *[s.q]uʔ tsjuwX jiǔ 2.1 A liquor *tiu B
3 205 3
*(S-
1.54 29 listen 1 21.2 85 4.4
)mru̯ɔŋH
2.56 18 liver *-hri ̯ən 1 581 119 2.1.3.1
2.4 3 long *ntauX 1 208 76 6.3.4
2.42 3 long (time) *ləu 1 216 96
2.55 7 look after *ʔro B
2 xx xx
望 *maŋ-s mjangH wàng 1.9 24 look at *maŋH 1 9.1 158
1.4 9 loud/sound *ʔbuj A
3 16 243
2.1 22 louse, body *tɛmX 1 205 174 6.3.4
3.20 12 louse, head *ntshjeiX 1 317 208 6.3.3
愛 *[q]ˤəp-s ʔojH ài 7.1 11 love, to *ʔu̯ɔi C
3 709 188
5-6.4 21 low/bellow *ɴqəŋ B
2 xx xx
4.3,
下 *[g]ˤraʔ hæX xià 5-6.3 4 low/short *GaX 1 615 130
6.2.2
底 *tˤijʔ tejX dǐ 2.4 1 lower level *ʔdiB 3 232 18
1.49 8 lung *mprʉ C
2 76 242
4.2.4,
5.31 24 maggot *klaŋC 2 517 63 4.4,
6.3.4
2.40.1 9 male *ʔljuwB 2 286 260
4.3 24 male (human) *ɟi ̯aŋH 1 435 121
3.16 8 malodorous *tsju̯eiH 1 121 242
騣 tsuwng zōng 3.1 30 mane *tsʉŋ A
2 157 233 4.2.3
騣 tsuwng zōng 3.4 29 mane *ʔdzɔŋ A
3 157 233 4.2.3
2.4 6 many/much *ntɔ C
2 208 232 7.5
髓 *s-lojʔ sjweX suǐ 2.41 8 marrow/brains *hlʉ A
2 xx xx
marry
6.1 15 *qu̯aC 2 xx xx
(a woman)
量 *[r]aŋ ljang liáng 2.42.1 22 measure (rice) *ljɛŋA 2 288 252 2.1.2.2
2.56 5 measure (rice) *hrau A
3 581 171
measure for
升 *s-təŋ sying shēng 2.28 18 *sjinA 2 323 64
grain
5-6.6 10 meat *NGej 1 630 37 7.3

117
Cantonese
 Index of proto forms by English 259

OC MC Man O R English Reconstr L WO WR Chap

7.1 4 meat *ʔaB 3 709 187
5.16 15 medicine *N-kjaj 1 472 112 7.6
4.3 15 meet *ɟu̯aC 2 xx xx
2.28 22 midday meal *sjɛŋC 2 xx xx
粢 *[ts]ij tsij zī 2.16 15 millet *tjəiA 3 349 94
1.1 5 molar tooth *pæ 1 1 149
錢 *N-ts[a][n] dzjen qián 3.3 20 money *dzi ̯en 1 111 24
6.3.2,
2.40 18 monkey *ʔlinA 2 214 43 7.1,
7.3
6.3.2,
1.4 18 monkey *ʔbiŋA 3 28 8
7.3
2.1.2.1,
2.41 4 moon/month *hlaH 1 215 92 6.3.1,
7.7
mortar
臼 *[g](r)[u]ʔ gjuwX jiù 4.3 7 *ɟoB 2 435 225
/treadmill
4.10 29 mosquito *ʔjɔŋB 2 442 248
慈 *[N-ts]ə dzi cí 2.21 1 mother/aunt *ndji B
3 378 18
mother’s
2.9 23 *neuŋX 1 213.1 153
brother
1.3 3 mountain *bæwB 2 3 244
mountain
5.33 20 *gli ̯emA 3 579 40
(range)
7.1 5 mountain pass *ʔæ(C) 1 709 54
*N-
峽 hɛp xiá 6.33 13 mountain pass *ɢlowD 2 651 273
kˤ<r>ep
2.9 24 mouse/rat *naŋB 2 213.1 177
2.9 3 mouse/rat *nau B
3 213.1 177
嘴 *[ts]ojʔ tsjweX zuǐ 4.6 9 mouth/beak *ɲɟuj 1 354 243
泥 *nˤ[ə]j(ʔ) nej ní 2.8 1 mud *hni A
3 212.1 1
5.4 7 muddy *ŋko D
2 544 222
菇 *mə.kˤa ku gū 5.19 3 mushroom *ŋkjæu 1 472 259
1.33 18 mussel/clam *blin C
2 xx xx
1.4 3 name *mpɔuH 1 64 77 4.3
*N-
狹 hɛp xiá 5-6.6 10 narrow *ɴɢeD 2 618 48
kˤ<r>ep
*N-
狹 hɛp xiá 7.13 10 narrow *hepD 3 618 48
kˤ<r>ep
2.4 13 navel *ntəut 1 208 255 2.1.2.1
2.55 3 near *-ʔræw C
2 580 259
2.13 7 near *sətD
3 155 89
260 Indexes

OC MC Man O R English Reconstr L WO WR Chap

頸 *keŋʔ kjiengX jǐng? 6.31 24 neck *qlaŋ 1 649 120
頸 *keŋʔ kjiengX jǐng 6.1 28 neck *qoŋ B
2 613 216
箴 *t.q[ə]m tsyim zhēn 5.16 29 needle *kjɔŋ A
2 469 142
箴 *t.q[ə]m tsyim zhēn 2.13 18 needle *sim A
3 167 5
2.57 3 nest *-rəuX 1 605 96
清 *N-s-r̥eŋ tshjeng qīng 3.17 18 new *tshji ̯əŋ 1 314 83 4.3
2.1.2.1,
1.8 21 night *hməŋH 1 20 140 2.1.2.2,
6.3.1
5.2,
4.3 16 nine *N-ɟuə 1 447 191 6.3.1,
7.5
1.48 1 noose/trap *briB 2 xx xx
2.1.2.2,
鼻 *m-[b]i[t]-s bjijH bí 1.51 8 nose *mbruiH 1 78 226
4.2.2
故 *kˤa(ʔ)-s kuH gù 6.1 7 old *qu̯oH 1 613 206
2.42 12 old *lu̯ei B
2 xx xx
2.28 16 on top of *sju A
2 xx xx
5.2,
一 *ʔi[t] ʔjit yī 7.1 2 one *ʔɨ 1 709 35
7.5
5-6.3 21 oneself *gənA 3 543 157
開 *N-[k]ʰˤəj khoj kāi 5.2 11 open *khu̯ɔi A
3 554 188 4.3
1.1 13 open (door) *pow D
2 1 256
2.5 15 open out/undo *nthu̯aB 2 209 151
3.20 15 otter *ntshju̯aA 2 317 94
1.31 28 owl *ploŋ D
2 xx xx
裹 *[k]ˤo[j]ʔ kwaX guǒ 5-6.2 8 pack/wrap *qhʉ B
2 614 242
5.21 13 pair *ŋgjow D
2 474 186
雙 *[s]ˤroŋ sræwng shuāng 2.13 28 pair *suŋA 3 179 218
2.57 20 pair *reŋ C
3 582 44
2.57 6 pangolin *rɔiH 1 605 131 7.2
7.6,
2.4 13 paper *ntowB 2 208 263
6.2.4
紙 *[k.t]eʔ tsyeX zhǐ 2.16 10 paper *tjejB 3 349 37
桃 *lˤaw daw táo 6.33 5 peach *ɢlæw 1 651 171
梨 *[r][ə][j] lij lí 2.57 15 pear *rəj 1 582 94
撅 *[g]ot gjwot jué 4.4 13 peck/dig *ɲcow D
2 352 273
撅 *[g]ot gjwot jué 4.4 13 peck/dig *ɲcow D
2 436 273
peel off
脫 *l ̥ˤot thwat tuō 2.4 9 *ʔdutD 3 208 239
/escape
4.2 10 peel/shave *cheD 2 xx xx
 Index of proto forms by English 261

OC MC Man O R English Reconstr L WO WR Chap

2.8 24 perilla *hnaŋB 2 xx xx
1.9 18 persimmon *minB 2 33.2 7
柿 dzyiX shì 2.18 15 persimmon *djəi B
3 351 94
民 *mi[ŋ] mjin mín 2.9 19 person *næn A
2 45 247
民 *mi[ŋ] mjin mín 1.24 19 person/Mien *mjæn A
3 45 247 2.1.2.2
2.41 20 petal *hlem C
3 215 59
3.16 13 pheasant *tsjow D
2 313 273
5.2 13 pick up *khowD 2 xx xx
夾 *kˤ<r>ep kɛp jiā 4.4 7 pick up *ʔɟəp D
3 436 106
pick up
2.1 10 *teD 2 205 87
/clamp
4.2,
1.4 5 pig *mpæC 2 4 168
7.1
豚 *[d]ˤu[n] dwon tún 2.3 28 pig *duŋB 3 243 180 7.1
鳩 *[k](r)u kjuw jiū 6.4 16 pigeon *ɴqu A
2 616 191
鴿 *[k][ə]p kop gē 5-6.1 29 pigeon *kɔp D
3 541 219
4.6 3 pillar *ɲɟæu 1 438 259
枕 *[t.q][ə]mʔ tsyimX zhěn 4.4 29 pillow *ɲcu̯əmH 1 364 246
plain
2.51 28 *ndroŋB 2 xx xx
(level ground)
4.3 21 plant, to *ɟəŋC 2 xx xx
*N-
3.6 5 play 1 198 114
dzæwH
1.47 7 play (flute) *phroB 2 74 212
1.32 29 play (flute) *phlu̯əm B
3 74 212
李 *[r]əʔ liX lǐ 2.41 18 plum *hli ̯əŋX 1 239 159
5.2 13 plum *khow A
2 542 263
楊 *laŋ yang yáng 4.12 26 poplar tree *ji ̯aŋA 2 xx xx
3.18 12 porcupine *N-dzjeiH 1 318 208 4.3
粥 *t-quk tsyuwk zhōu 2.16 B porridge *tju̯ok D
3 169 258
4.7 18 POSSESSIVE *ʔɲi ̯əŋ A
3 403 159
1.46 4 pour *pra C
2 xx xx
2.40.1 12 pour *ʔlju̯ei A
2 286 39
*mə.qˤ<r>aʔ
價 kæH jià 5-6.4 4 price *ɴqaC 2 616 92 2.1.2.2
-s
抽 *lr̥u trhjuw chōu 2.47 7 pull out *throC 2 254 206
5.31 19 pullet/hen *klan C
3 577 99
1.3 12 pus *bu̯ei C
2 3 39 7.1
膿 *[n]ˤuŋ nowng nóng 2.9 29 pus *nɔŋ C
3 213.1 142 7.1
push earth
1.2 8 *phʉA 2 2 242
aside
262 Indexes

OC MC Man O R English Reconstr L WO WR Chap

put on/wear 4.3,
2.4 29 *ntɔŋH 1 208 142
(cap) 7.6
put on/wear 4.4,
2.8 24 *(h)naŋX 1 212.1 158
(clothes) 7.6
put on/wear
2.3 22 *dəŋA 3 207 102
(jewelry)
put on/wear
踏 thop tà 2.3 7 *dap 1 207 105
(shoes)
put on/wear
著 *trak trjak zhuó 2.46 6 *trɔC 2 253 232 4.4
(shoes)
2.1.2.2,
2.50 29 puttees *nthrɔŋA 2 257 123
7.6
捷 *[dz][a]p dzjep jié 4.1 9 quick *cu̯ɛp 1 191 163
2.56 10 quick *hreC 2 581 37
2.13 22 quilt *su̯ɛŋC 3 179 252
2.40.1 16 rabbit *ʔljuB 2 286 172
ragged
陋 *[r]ˤo-s luwH lòu 6.36 16 *ɴɢluC 2 654 191
/shabby
2.9 28 rain *m-noŋC 2 213.2 178 2.1.2.1
1.36 28 rain *mbluŋ C
3 90.1 180
raise
養 *[ɢ](r)aŋʔ-s yangH yàng 4.12 22 *jɛŋC 2 444 252
(chickens)
raise
養 *[ɢ](r)aŋʔ yangX yǎng 4.12 28 *juŋB 3 456 180
(chickens)
耙 *[b]ˤra bæ pá 1.3 4 rake, to *baA 3 3 149
2.6 16 ramie/hemp *nduH 1 210.2 111 7.6
1.48 15 rat *bru̯aC 2 xx xx
4.9 29 raw/unripe *ɲu̯emX 2 405 41
*(mə-) reach
造 tshawH zào 3.3 7 *dzoC 2 159 206
tsʰˤuʔ-s /arrive at
receive 109,
接 *[ts]ap tsjep jiē 3.1 10 *tsep 1 68
/borrow 157
receive
收 *s-kiw syuw shōu 2.28 9 *sjuwA 2 xx xx
/gather in
4.1,
廝 *se sje sī 2.28 2 RECIPROCAL *sjɨA-D 2 xx xx
6.2.3
7.1,
2.40 18 red *ʔlinA 2 214 43
7.7
7.1,
赤 *[t-qʰ](r)Ak tsyhek chì 2.13 2 red *sekD 3 167 17
7.7
4.4 7 remember *ɲcoC 2 436 225
5.16 24 remember *kji ̯əŋ C
3 469 121
補 *(mə)-pˤaʔ puX bǔ 1.19 4 repair/mend *mpjaX 1 40 130
12 repay *bu̯ei A
1.3 2 3 208
 Index of proto forms by English 263

OC MC Man O R English Reconstr L WO WR Chap

6.31 5 repay *qlauB 3 649 152
2.8 22 resemble *hnəŋB 3 212.1 102
2.28 7 rest, to *sjoC 2 323 225
2.46 21 return *trəŋB 2 253 218
1.9 16 return home *muB 3 9.1 225
2.1 24 rib *tamX 1 205 154
粔 *[g](r)aʔ gjoX jù 5.19 16 rice cake *ŋkjuəX 1 508 191
rice
1.36 3 *mbləu 1 90.1 96 7.4
plant/paddy
甑 *s-təŋ-s tsingH zèng 3.16 22 rice steamer *tsjɛŋH 1 313 104 3.2
饟 *n̥aŋ(ʔ)-s syangH xiǎng 2.53 21 rice, cooked *hnrəaŋH 1 260 101 7.4
3.4 9 rice, husked *ntsuw C
2 160 260 7.4
米 *[m]ˤijʔ mejX mǐ 1.8 12 rice, husked *hmeiB 3 8.1 170
rice/millet,
3.2 22 *tshɛŋB 2 158 252 7.4
husked
騎 *[g](r)aj gje qí 4.3 10 ride *ɟej 1 435 37
6.33 3 river *ɢlæw A
2 651 95 7.2
2.6 15 river *ndai A
3 210.1 74
12 river deer *ŋgu̯ei B
5.6 2 546 39 7.2
5.31 3 road/way *kləuX 1 517 96
7.13 5 rob *hæ D
2 xx xx
1.4 29 rock *ʔbeŋ C
3 4 46
6.31 21 roll down *qləŋB 2 xx xx
4.3 28 root *ɟoŋA 2 435 178
3.6 28 root *ndzuŋA 3 162 249
2.41 5 rope *hlæC 2 215 149
2.42 8 rot *lʉ A
2 216 242
rough
粗 *s.[r̥]ˤa tshu cū 3.5 4 *ntshaA 2 xx xx
/coarse
6.33 27 round *ɢlun 1 651 231 4.2.4
3.19 3 salt *ntsjəuX 1 316 96 4.3
6.46 13 salty *qrow C
2 697 263
鹺 *[dz]ˤaj dza cuó 3.6 15 salty *ndzai A
3 162 74
sandals,
2.13 6 *səukD 3 179 241
straw
sash/cord
繩 *m.ləŋ zying shéng 2.41.1 24 *hljaŋ 1 215 101
/rope
湯 *l ̥ˤaŋ thang tāng 2.41 24 scald *hlaŋA 2 215 45
scatter 5, 142,
噴 *pʰˤur-s phwonH pèn 1.5 29 *mphu̯ənH 1
/sprinkle 17 248
5.34.1 2 scissors *ʔglji ̯euB 3 532 2
5-6.1 10 scold/curse *qe C
2 613 40
264 Indexes

OC MC Man O R English Reconstr L WO WR Chap

7.13 20 scold/curse *hi ̯emC 3 613 40
1.3 7 see *bu̯ət 1 15 108
2.53 29 seed *hnru̯em 1 259 60
3.2,
秧 *ʔaŋ ʔjang yāng 4.10 22 seedling *ʔjɛŋA 2 454 160
7.4
秧 *ʔaŋ ʔjang yāng 4.10 24 seedling *ʔjaŋA 3 454 160 7.4
5.51 10 seek shelter *ŋgre D
2 282 220
2.1.3.1,
賣 *m.rˤajʔ-s mɛɨH mài 1.9 11 sell *mɛjH 1 9.1 93
4.4
send
送 *[s]ˤoŋ-s suwngH sòng 2.13 28 *suŋH 1 155 180
/deliver
5.2,
2.18 28 seven *djuŋH 1 339 180 6.3.1,
7.5
5.2,
5.24 1 seven *ŋjiC 3 573 18 6.3.1,
7.5
聯 ljen /lyːn2/118 2.42.1 22 sew *lju̯ɛnA 3 558 100
5.34 18 shallow *ŋkli ̯əŋX 1 604 83
4.9 20 shallow *ɲenB 2 405 64
2.57 15 sharp *-rajH 1 605 151
畟 *[tsʰ]rək tsrhik cè 3.19 3 sharp *ntsjæwC 2 xx xx
sharpen
7.13 7 *hoB 2 551 225
/grind
sharpen
剉 *[tsʰ]ˤoj-s tshwaH cuò 2.20 3 *dhjæuC 3 376 259
/grind
shed
1.36 8 *mblu̯eiC 3 90.3 170
leaves/drop
羊 *ɢ(r)aŋ yang yáng 4.12 28 sheep/goat *juŋ 1 456 179 3.2
shell
1.32 12 *phlei 1 86 208 2.1.2.2
/covering
屨 *kro-s kjuH jù 5.17 6 shoes *khjɔC 2 470 187
6.3.3,
1.1 21 shoot *pənX 1 13 176
7.3
2.40 22 short *ʔlɛŋB 2 211.2 104 6.2.1
2.7 22 short *ʔnəŋB 3 211.2 104 6.2.1
1.3 8 shoulder *bʉ C
2 3 242
縮 *[s]ruk srjuwk suō 4.11 9 shrink *hjuk 1 551 203
2.13 13 shut (door) *sow D
2 167 273
5-6.6 21 shuttle *ɴɢəŋB 2 618 138 7.6
6.36 16 shuttle *ɴɢlow B
3 654 172 7.6

118
Cantonese
 Index of proto forms by English 265

OC MC Man O R English Reconstr L WO WR Chap

4.7 12 shy *ʔɲæiA/B 3 403 132
2.1.2.2,
鎌 *[r]em ljem lián 2.42.1 18 sickle *ljim 1 312 5
7.4
3.17 6 sieve, to *tshjɔC 2 xx xx
3.2,
銀 *[ŋ]rə[n] ngin yín 4’.9 18 silver *ɲʷi ̯ən 1 417 137
7.1
2.28 12 sinew *sju̯eiB 2 323 208
筋 *(Cə.)[k]ə[n] kjɨn jīn 5’.16 19 sinew *kʷjan A
3 481 99
號 *[g]aw haw háo 5-6.3 5 sing/cry out *Gæw 1 615 152 6.2.2
2.3 21 sink *dəŋ A
2 207 176
4.7 24 sister-in-law *ʔɲam 1 403 97 6.3.1
坐 *[dz]ˤo[j]ʔ dzwaX zuò 3.3 8 sit *dzu̯ei B
3 123 39
2.1.4,
5.2,
5.46 9 six *kruk 1 277 204
6.3.1,
7.5
2.1 13 skin *N-top 1 205 237 2.1.2.1
2.1 18 skirt *tinA 2 205 43 7.6
裙 *[g]ur gjun qún 5.18 27 skirt *gjun A
3 411 195
1.12 22 sky/heaven *wɛŋ 1 12 104
2.6 22 sky/heaven *ndɛuŋ 1 210.1 268
6.6 22 sky/heaven *ɴɢɛuŋ 1 630 268
slant
5-6.3 10 *ɢeA 2 xx xx
/slanting
sleep
1.1 8 *pu̯eiH 1 13 242
/lie down
2.1.2.1,
2.41 10 slice *hlep 1 215 91
6.3.1
細 *[s]ˤ[i][j]-s sejH xì 2.13 12 small/fine *sæiC 3 11 132
small
幼 *[ʔ](r)iw-s ʔjiwH yòu 4.10 9 *ʔjuwC 2 442 243
/young
smooth
1.36 18 *mbli ̯əŋH 1 90.2 83
/glossy
smooth
滑 *Nə-gˤrut hwɛt huá 6.6 9 *ɴɢu̯at 1 630 183
/slippery
蝸 *[k]ʷˤraj kwæ wō 6.3 8 snail *ɢʉA 2 639 242
蝸 *[k]ʷˤraj kwæ wō 5.31 8 snail *klu̯ei A/B
3 637 242
蝸 *[k]ʷˤraj kwæ wō 5-6’.31 10 snail *kʷlej A
3 589 37
2.7 24 snake *ʔnaŋ 1 211.1 158
2.1 22 snap *tɛŋH 1 205 104
1.23 18 sniff at *hmji ̯əmH 1 32.2 135 2.1.2.1
5.21 19 snore *ŋgjanA 3 474 99
266 Indexes

OC MC Man O R English Reconstr L WO WR Chap

漬 *[dz]ek-s dzjeH zì 3.3 12 soak *dzeiH 1 111 39
3.19 3 soak *ntsjæw C
2 316 259
2.3 7 soak *do A
3 207 225
醒 *[s]ˤeŋʔ sengX xǐng 2.13 18 sober up *siŋ B
3 143 8
1.39 11 soft *mlu̯ɛjH 1 93 75 6.3.4
4.2,
2.1 22 son/boy *tu̯ɛn 1 217 100
6.2.4
1.10 12 son-in-law *ʔweiX 1 10 39 6.3.1
5.5 13 soot *ŋkhowA 2 xx xx
sore/boil
瘡 *[tsʰ]raŋ tsrhjang chuāng 3.2 24 *tshaŋ 1 158 120
/blister
sore/boil
5’.16 10 *kʷjejA 3 171 37
/blister
3.19 16 sorghum *ntsjuC 2 xx xx
1.33.1 4 soul/spirit *blja A
2 99 92 7.7
聲 *[l ̥]eŋ syeng shēng 2.13 18 sound *siŋ A
3 167 8
2.56 16 sound/noise *-hru A
2 581 191
酸 *[s]ˤo[r] swan suān 2.13 9 sour *suj 1 119 260
1.46 3 to sow *præw A
2 73 113
1.1 12 soybean *N-peiX 1 1 39
6.48 24 span, arm *ɢraŋA 2 699 121 7.5
/njam/119 4.12 29 span, arm *ju̯amA 3 444 211
6.46 7 span, finger *qroC 2 697 206 7.5
1.51 4 spicy *mbrat 1 78 107
4.10 24 spill *(ʔ)jam C
3 442 78
spinning
車 *[t.qʰ](r)A tsyhæ chē 3.17 15 *tshju̯aA 2 xx xx
wheel
6.46 24 spirit/ghost *qraŋA 2 697 45 7.7
2.1.2.2,
1.23 19 spirit/ghost *hmjænB 3 44 42
7.7
劈 *(m-)pʰˤek phek pī 1.2 2 split/chop *phek 1 50 34 6.3.2
sprinkle
4.5 16 *ɲchuB 2 437 191
/spill
4.8 11 squeeze *hɲaiC 3 403 55
squeeze
2.40 2 *ʔlɨC 2 214 35
/pinch
stand
2.28 13 *sjouX 1 323 263
/get up
5-6.1 22 star *qɛŋA 2 613 100
4.9 20 steal *ɲemH 1 405 59
3.2 3 steel *tshæwA 2 158 77

119
Hakka (Downer 1973:19)
 Index of proto forms by English 267

OC MC Man O R English Reconstr L WO WR Chap

3.5 28 steep *ntshoŋA 2 161 178
5.18 9 steep *gjujB 3 411 260
2.48 7 step *droA 2 255 206
step on
蹋 *lˤap dap tà 2.3 5 *dæD 2 xx xx
/tread
1.46 4 stick *praX 1 73 92
1.31 24 stomach *plaŋA 2 85 45
2.13 1 stomach *siA 3 167 1
2.1.2.2,
2.55 3 stone *-ʔrəu 1 580 96 4.2.2,
6.2.2
凳 tongH dèng 2.1 22 stool/bench *təŋC 3 205 102
stop up
3.19 13 *ntsjot 1 316 256
/plug
竈 *[ts]ˤuk-s tsawH zào 3.1 7 stove *N-tsoH 1 157 225 3.2
正 *teŋ-s tsyengH zhèng 4.4 26 straight *ɲci ̯aŋ A
2 436 161 4.4
2.1.4,
力 *kə.rək lik lì 2.57 7 strength *-rək 1 605 206
3.2
力 *kə.rək lik lì 5.32 7 strength *khləkD 3 578 167 2.1.4
4.13 28 stretch *ɕoŋ 1 443 216
摘 *(mə-)t(ˤ)rek trɛk zhāi 2.51 12 strip leaves *ndru̯ei C
2 258 208
撫 *[pʰ](r)aʔ phjuX fǔ 1.32 8 stroke *phlʉ C
2 xx xx
撫 *[pʰ](r)aʔ phjuX fǔ 1.32 27 stroke *phlun A
3 xx xx
3.21 13 stumble *ndzjow D
2 xx xx
3.4 10 suck *ntseD 2 xx xx
2.1.2.1,
2.8 22 sun/day *hnɛŋ A
2 224 188 6.3.1,
7.7
2.1.2.1,
2.8 11 sun/day *hnu̯ɔiA 3 224 188 6.3.1,
7.7
5-6.6 21 swallow, to *ɴɢəŋB 2 618 214
6.4 13 swallow, to *ɴqowC 2 616 263
2.7 7 swallow, to *ʔnəkD 3 211.1 129
4.2 3 sweep *chæwA 2 434 259
拂 *pʰut phjut fú 1.2 7 sweep *phu̯ət D
3 14 108
4.4 3 sweep *ʔɟæuC 3 400 259
613, 45,
甘 *[k]ˤam kam gān 5-6.1 24 sweet *Kam 1 2.1.2.2
625 154
2.6 15 sweet potato *ndu̯əj 1 222 189
7.1 21 swell *ʔəumH 1 709 193
4.6 6 sword *ɲɟəukD 3 354 241
268 Indexes

OC MC Man O R English Reconstr L WO WR Chap

2.6 24 sword, long *ndaŋA 2 210.1 120
2.6 3 table *ndɔuA 3 210.1 123
2.48 29 table/bench *drɔŋA 2 255 85
tael
兩 ljangX liǎng 2.42 26 *li ̯aŋB 2 300 161
(40 grams)
tael
兩 ljangX liǎng 2.42.1 26 *lji ̯oŋB 3 300 161 2.1.2.2
(40 grams)
4.2.2,
2.1 8 tail *tu̯eiX 1 217 242
6.3.3
捉 *[ts](ˤ)rok tsræwk zhuō 2.16 29 take *tjɔkD 3 349 205
芋 *[ɢ]ʷ(r)a-s hjuH yù 1.12 13 taro *wouH 1 12 263
味 *m[ə][t]-s mjɨjH wèi 1.8 12 taste/try *hmeiH 1 xx xx
5.18 1 tea *gji B
2 471 81
茶 dræ chá 4.3 4 tea *ɟa A
3 399 149
6.46 15 tear, to *qru̯a C
2 697 94
*(mə-)
報 pawH bào 1.4 7 tell *ʔboC 3 16 225
pˤuk-s
廟 *[m](r)aw-s mjewH miào 1.9 A temple *miuC 3 33.1 3
3.2,
4.3,
十 *[g][i]p dzyip shí 5.18 9 ten *gju̯ɛp 1 471 163 5.2,
6.3.1,
7.5
萬 *ma[n]-s mjonH wàn 1.12 26 ten thousand *wi ̯aŋC 2 xx xx
2.55 21 tender *ʔrəunH 1 580 214
10 that *ʔwe B
1.10 3 10 73 5.4
that
7.1 2 *ʔɨB 2 xx xx 5.4
/the former
2.1 5 thick *tæA 2 205 168 7.1
厚 *-[g]ˤ(r)oʔ huwX hòu 7.15 3 thick *ɦɔu B
3 720 115 7.1
5.21 29 thick (liquid) *ŋgjɔŋ C
2 474 123
thick
5-6.3 7 *gətD 3 543 108
(porridge)
1.9 4 thick/dense *makD 3 9.1 184 2.1.2.2
1.1 5 thigh *pæ A
2 1 168 4.2
2.18 9 thigh *djuj A
3 351 243
2.28 29 thin *sjɔŋ A
2 323 142
姌 *n[a]mʔ nyemX rǎn 4.9 20 thin *ɲen B
2 405 24
薄 *Cə-[b]ˤak bak bó 1.3 1 thin *bi ̯ɛk D
3 39 148
5.33 12 thin *glæi C
3 579 132
3.6 12 thin (person) *ndzu̯ei C
2 162 132
 Index of proto forms by English 269

OC MC Man O R English Reconstr L WO WR Chap

*hnəmB
念 nemH /nɐm3/120 2.8 22 think of 3 212.2 174 6.2.1
~*hləmB
渴 *Nə-qʰˤat khat kě 5-6.5 4 thirsty *NKhat 1 617 221
2.7 25 this *ʔneinX 1 211.1 36 5.4
1.3 7 thorn *-bo B
2 3 225 4.2.3
4.4 18 thorn *ʔɟim B
3 436 5
3.2,
千 *s-n̥ˤi[n] tshen qiān 3.2 20 thousand *tshi ̯en 1 134 24
5.2
1.2 7 thread *S-phoB 2 59 206
2.13 9 thread *sujC 3 179 243
穿 *tʰo[n] tsyhwen chuān 4.2 21 thread, to *chu̯en 1 446 194
5.2,
1.16 3 three *pjɔu 1 61 77 5.3,
7.5
throw on
被 *m-p(r)ajʔ bjeX bèi 1.4 15 *mpu̯aC 2 4 94
(clothes)
thrush (huà méi
4.1 29 *cɔŋA 2 433 85 7.2
niǎo)
1.4 16 thunder *mpuə 1 16 191
1.2 7 thunder *S-phoA 2 59 206
5-6.2 10 tie/cross (legs) *qheA 2 614 261
祖 *[ts]ˤaʔ tsuX zǔ 3.16 7 tiger *tsjoB 2 313 206
5.36.1 19 tiger *ŋgljænA 3 510 175
緊 *Ki[n]ʔ kjinX jǐn 5.16 27 tight *kjən B
3 469 62
2.1.2.2,
瓦 *[ŋ]ʷˤra[j]ʔ ngwæX wǎ 5’.9 5 tile *ŋʷæX 1 561 149
3.2
tilt-hammer
碓 twojH duì 2.1 11 *tu̯ɔiC 3 217 188
/pestle
3.2 5 tin *tshæD 2 xx xx
4.4,
舌 *m.lat zyet shé 1.36 10 tongue *mblet 1 90.1 51
6.2.2
1.23 18 tooth *hmjinX 1 32.2 7 4.2
2.3 16 tortoise *duH 1 207 115
2.51 7 track, to *ndro C
2 258 206 7.3
4.1 16 trap/ensnare *cu A
2 xx xx
樹 *m-toʔ-s dzyuH shù 2.19 29 tree *ntju̯əŋH 1 424 102 6.3.3
1.47 16 tree *S-phru A
2 11 191 4.2
tree of
4.12 22 *jɛŋA 2 xx xx
heaven
7.1,
6.46 28 trough *qroŋA 2 697 180
7.6

120
Cantonese
270 Indexes

OC MC Man O R English Reconstr L WO WR Chap

槽 *[dz]ˤu dzaw cáo 3.3 16 trough *dzuA 3 111 115 7.1
*ŋkro B

5.49 7 turbid 2 280 206

/ŋgroB
5.36 13 turbid *ŋglokD 3 522 203
1.35.1 10 twin *mplje A
2 xx xx
1.2 15 twist/rub *S-phu̯a A
2 59 164
4.13 5 twist/rub *ɕætD 3 59 164
5.2,
7.1 14 two *ʔu̯i 1 709 20
7.5
4.12 3 uncle *jæuX 1 444 259
2.57 4 urine *-raX 1 84 130
用 *mə-loŋ-s yowngH yòng xx 28 use *[n/l]oŋC 3 213.3 216 6.2.1
2.42.1 15 used to *lju̯a C
2 288 151
2.55 12 vegetable *ʔræi 1 580 132
vertical
直 *N-t<r>ək drik zhí 3.21 3 *ndzjæwC 2 xx xx
/upright
里 *(mə.)rəʔ liX lǐ 2.57 21 village *rəŋX 1 605 140 6.2.2
蔓 *ma[n]-s mjonH wàn 1.8 25 vine *hmein 1 8.2 170
6.31 15 waist *qlajX 1 649 150 2.1.3.1
2.3 21 wait *dəŋB 2 207 214
2.16 3 wait *tjɔuB 3 361 244
3.18 4 wake *dzja D
2 315 92
4.12 29 walk *jəŋ A
3 444 102
6.31 13 walnut *qlow C
2 xx xx
暑 *s-taʔ syoX shǔ 2.28 7 warm *sji ̯ouX 1 347 134
warm self
2.4 3 *ntauH 1 208 76
by fire
澡 *mə.[ts]ˤawʔ tsawX zǎo 3.4 5 wash (hands) *ntsæwX 1 160 114
10, 212,
7.1 29 water *ʔu̯əm 1 6.3.3
709 246
water buffalo
牛 *[ŋ]ʷə ngjuw niú 5.9 23 *ŋiuŋ 1 549 196
/cow
淋 *[r][ə]m lim lín 2.57 21 water, to *rəmA 3 605 229
4.6 27 water, to *ɲɟun A
3 354 231
2.4 7 weave *ntət 1 208 222 7.6
織 *tək tsyik zhī 3.1 1 weave *tsi ̯ɛk D
3 181 110
7.13 18 weave/plait *hin A
2 719 7
2.49 18 wedged in *ntrinC 2 xx xx
2.5 16 weed/hoe *nthu C
2 xx xx
4.7 20 weep *ʔɲæmX 1 403 79 6.3.3
2.4 22 wet *ntu̯ɛn 1 220 100
3.16 2 what *tsjɨ B
2 xx xx
 Index of proto forms by English 271

OC MC Man O R English Reconstr L WO WR Chap

1.9 27 wheat *muŋC 2 xx xx
2.3 8 which *dʉC 2 207 242
2.41 15 which *[hl]aiC 3 719 74
1.36 13 whip *mblowD 2 90.1 273 7.6
鞭 *pe[n] pjien biān 1.1 18 whip *pinA 3 1 6
7.1,
6.31 13 white *qlowA 2 649 263
7.7
7.1,
白 *bˤrak bæk bái 1.3 5 white *bækD 3 3 72
7.7
whorl at top
4.12 2 *jɨB 2 xx xx
of head
廣 *kʷˤaŋʔ kwangX guǎng 5-6’.1 24 wide *Kʷi ̯aŋX 1 661 121
wife of
1.24 29 mother’s *mjəŋA 3 45 102 2.1.2.2
brother
wife
1.10 3 *ʔwæwB 2 10 259
/elder sister
6.3.2,
1.8 24 wild dog *hmaŋC 2 xx xx
7.3
貍 *[m]ə.rə li lí 1.31.1 2 wildcat *pljɨD 2 97 15 7.3
4.1 5 wind *N-cæwH 1 457 171
2.1 4 wing *N-tat 1 205 107 4.3
眨 *mə-tsˤr[a]p tsrɛp zhǎ 3.19 10 wink *ntsjep 1 340 220
1.46 29 winnow *prɔŋ B
2 73 248
4.8 27 winnow *hɲumC 3 403 136
winnowing
1.10 24 *ʔwaŋA 2 10 121
fan
winnowing
扇 *[l ̥][a][n]-s syenH shàn 4.13 24 *ɕi ̯əŋA 3 191 121
fan
2.28 21 wipe *sjəŋC 2 323 218
刷 *[s][r][o]t srwæt shuā 2.13 13 wipe *sot D
3 179 256
1.16 4 wooden basin *pjat D
3 37 127 2.1.2.2
word
字 *Cə-[dz]ə-s dziH zì 3.6 24 *ndzaŋC 3 162 101
/character
1.46 7 wring *proC 2 73 206
鏤 *[r]ˤo-s luwH lòu 2.56 12 write *hru̯ei C
2 581 208
寫 *s-qʰAʔ sjæX xiě 5.28 4 write *xja B
3 143 130
年 *[n]ˤi[n] nen nián 4.8 29 year *hɲu̯əŋH 1 404 102 5.2
年 *[n]ˤi[n] nen nián 4.13 29 year *ɕɔŋ C
2 443 102 5.2
黃 *N-kʷˤaŋ hwang huáng 5-6’.3 24 yellow *ɢʷaŋ A
2 687 121 7.7
皇 *[ɢ]ʷˤaŋ hwang huáng 1.12 24 yellow *wi ̯əŋ A
3 687 121 7.7
2.41.1 5 young *hljæ D
2 287 168
272 Indexes

OC MC Man O R English Reconstr L WO WR Chap

2.51 12 young man *ndru̯eiC 2 258 39
1.21 5 young man *mbjauA 3 30 171 6.3.2
younger
5.1 8 *ku̯eiX 1 541 242 2.1.2.2
brother
 Index of proto forms by White Hmong

WHm=White Hmong; O1=onset category; O2=onset subcategory; Rime=rime category;

English; Recon=reconstructed form; Lev=level: Hmong-Mien (1), Hmongic (2), Mienic
(3)

WHm English Reconstr L Onset Rime Chapter

cab earthworm *N-cuŋ 1 4.1 28
cab lead, to *coŋA 2 4.1 28
cag root *ɟoŋA 2 4.3 28
caij ride *ɟej 1 4.3 10
caug knee *ɟu̯əiH 1 4.3 12
B
cawv liquor *cow 2 4.1 13
cej buckwheat *ɟæu 1 4.3 3 7.4
ces, ceg leg/branch *ɟæw B
2 4.3 3 4.4
cev body/trunk *cæw B
2 4.1 3 4.4
ci bake/toast *ciC 2 4.1 1 4.4
cig burn/be alight *ɟiC 2 4.3 1 4.4
ciab beeswax *N-ci ̯æŋ 1 4.1 24
ciaj alive *ɟinA 2 4.3 18
cog plant, to *ɟəŋ C
2 4.3 21
B
cos mortar/treadmill *ɟo 2 4.3 7
coob thrush (huà méi niǎo) *cɔŋ A
2 4.1 29 7.2
cua chew *cu̯a C
2 4.1 15
cua wind *N-cæwH 1 4.1 5
cuab trap/ensnare *cuA 2 4.1 16
cuag meet *ɟu̯aC 2 4.3 15
cuaj nine *N-ɟuə 1 4.3 16 5.2, 6.3.1, 7.5
chav CLF-rooms *choŋB 2 4.2 28
chais peel/shave *cheD 2 4.2 10
chaub belly *chu̯ei A
2 4.2 12
chaw bed/place *chouH 1 4.2 13
cheb sweep *chæw A
2 4.2 3
chob thread, to *chu̯en 1 4.2 21
choj bridge *ɟow 1 4.3 16
dab neck *qlaŋ 1 6.31 24
dab spirit/ghost *qraŋA 2 6.46 24 7.7
dab trough *qroŋA 2 6.46 28 7.1, 7.6
daj span, arm *ɢraŋA 2 6.48 24 7.5

273
274 Indexes

WHm English Reconstr L Onset Rime Chapter

daj yellow *ɢʷaŋA 2 5-6’.3 24 7.7
dav eagle/hawk *qlaŋX 1 6.31 24 4.2.3
dav wide *Kʷi ̯aŋX 1 5-6’.1 24
dais bear *qrep 1 6.46 10 7.3
daus ice *qru̯eiD 2 6.46 12
daw salty *qrowC 2 6.46 13
dawb white *qlow A
2 6.31 13 7.1, 7.7
dawm mountain pass *ɢlow D
2 6.33 13
deb far *qʷuw 1 6’.1 3 6.2.2
dej river *ɢlæw A
2 6.33 3 7.2
dev dog *qluwX 1 6.31 3 4.2.4
dib cucumber *Kʷa 1 5-6’.1 4 2.1.2.2
dim escape *ɢʷaD 2 5-6’.3 4
dos Allium *ɢləŋB 2 6.33 21 7.2
dos span, finger *qro C
2 6.46 7 7.5
dov roll down *qləŋ B
2 6.31 21
dub black *qrɛŋ A
2 6.46 22 7.7
dua tear, to *qru̯a C
2 6.46 15
duaj peach *ɢlæw 1 6.33 5
duav waist *qlajX 1 6.31 15 2.1.3.1
dhau cross (river) *KʷajH 1 5-6’.1 15
hais ladle, to *heD 2 7.13 10
hau boil (TR) *hu̯eiC 2 7.13 12
hau head *S-phreiX 1 1.47 12
haus drink/smoke *hup 1 7.13 9 6.2.2
hiab weave/plait *hin A
2 7.13 18
hov sharpen/grind *ho B
2 7.13 7
hub clay pot/earthen jar *hɛŋ A
2 7.13 22
huab fog/cloud *hu A
2 7.13 16 7.1
hlab sash/cord/rope *hljaŋ 1 2.41.1 24
hlab scald *hlaŋA 2 2.41 24
hlais slice *hlep 1 2.41 10 2.1.2.1, 6.3.1
hlau iron *hluw C
2 2.41 9 3.2, 6.2.2, 7.1
hlawv burn (mountain) *hljow B
2 2.41.1 13
2.1.2.1, 6.3.1,
hli moon/month *hlaH 1 2.41 4
7.7
hlob big *hljo 1 2.41.1 7 4.4
hlov bamboo *hləwX 1 2.41 7
hlua rope *hlæ C
2 2.41 5
hluas young *hljæ D
2 2.41.1 5
hlwb marrow/brains *hlʉ A
2 2.41 8
hma wild dog *hmaŋC 2 1.8 24 6.3.2, 7.3
 Index of proto forms by White Hmong 275

WHm English Reconstr L Onset Rime Chapter

hmab vine *hmein 1 1.8 25
2.1.2.1, 2.1.2.2,
hmo night *hməŋH 1 1.8 21
6.3.1
hmob flea *hmiəŋ 1 1.8 23
hmoob Hmong *hmʉŋ A
2 1.8 30
hnab grain head/bag *hnɔn 1 2.8 24 7.4
hnav perilla *hnaŋB 2 2.8 24
hnav put on/wear (clothes) *(h)naŋX 1 2.8 24 4.4, 7.6
hneev crossbow *hnæn B
2 2.8 19 2.1.4, 4.2, 7.3
hneev footprint/track *hmjænX 1 1.23 19 6.3.1, 7.3
hnia sniff at *hmji ̯əmH 1 1.23 18 2.1.2.1
hniav tooth *hmjinX 1 1.23 18 4.2
hno rice, cooked *hnrəaŋH 1 2.53 21 7.4
hnov forget *hnuŋ B
2 2.8 27
hnov hear *hnəumX 1 2.8 21
hnoos cough *hnɔp 1 2.8 29
2.1.2.1, 6.3.1,
hnub sun/day *hnɛŋA 2 2.8 22
7.7
hnyav heavy *hnjeinX 1 2.23 25
hnyuv intestines *hɲeuŋX 1 4.8 23
i that/the former *ʔɨB
2 7.1 2 5.4
ib crest/comb *ʔwi ̯æn 1 1.10 1
ib one *ʔɨ 1 7.1 2 5.2, 7.5
iab bitter *ʔim 1 7.1 18
4.2, 4.4, 6.2.4,
kab insect/worm *klæŋ 1 5.31 24
6.3.4
kaj bright/light *ɢʷaŋA 2 5-6’.3 24
kas maggot *klaŋ C
2 5.31 24 4.2.4, 4.4, 6.3.4
3.2, 4.3, 5.2,
kaum ten *gju̯ɛp 1 5.18 9
6.3.1, 7.5
kaus bamboo hat *kuwD 2 5.1 9 2.1.2.2
kaws gnaw *kowD 2 5.1 13 2.1.2.2
kev road/way *kləuX 1 5.31 3
-ki child *kʷjeiB 3 5’.16 12
kib fry *kleu 1 5.31 2
A
koj 2SG (you) *gʉ 2 5.3 8 5.3
A
koob needle *kjɔŋ 2 5.16 29
kooj grasshopper *gʉŋ A
2 5.3 30
kub gold *kjeəm 1 5.16 23 3.2, 7.1
kub horn *klɛɔŋ 1 5.31 22
kub hot (water) *k(h)lɛɔm 1 5.32 22
kuv 1SG (I/me) *kɛŋB 2 5.1 22 5.3, 6.3.2, 6.3.4
276 Indexes

WHm English Reconstr L Onset Rime Chapter

kua liquid/soup *klæwX 1 5.31 5
kwv younger brother *ku̯eiX 1 5.1 8 2.1.2.2
khau shoes *khjɔC 2 5.17 6
itch(y)/
khaus *khju̯ɛt 1 5.17 9
scratch(y)
khawb plum *khowA 2 5.2 13
khaws pick up *khowD 2 5.2 13
lag deaf *loŋA 2 2.42 28
las tael (40 grams) *li ̯aŋ B
2 2.42 26
laim lightning flash *lje D
2 2.42.1 10
lau male *ʔljuw B
2 2.40.1 9
laub pour *ʔlju̯ei A
2 2.40.1 12
laus old *lu̯ei B
2 2.42 12
leej CLF-people *lan 1 2.42 19
li squeeze/pinch *ʔlɨC 2 2.40 2
lig late *liC 2 2.42 1
liab monkey *ʔlinA 2 2.40 18 6.3.2, 7.1, 7.3
liab red *ʔlin A
2 2.40 18 7.1, 7.7
liag sickle *ljim 1 2.42.1 18 2.1.2.2, 7.4
liaj field *ljiŋ 1 2.42.1 18
lo CLF-mouthfuls *ʔlo C
2 2.40 7 6.2.4
los bury *ljəŋ B
2 2.42.1 21
los come back *ləwX 1 2.42 7
lov break *ʔloB 2 2.40 7 6.2.1
lub CLF-bowls/houses *ʔlɛŋA 2 2.40 22 6.2.1, 6.2.4
luj measure (rice) *ljɛŋA 2 2.42.1 22 2.1.2.2
luv short *ʔlɛŋ B
2 2.40 22 6.2.1
luav rabbit *ʔlju B
2 2.40.1 16
lwg dew *lʉ C
2 2.42 8
lwj rot *lʉ A
2 2.42 8
mob illness/pain *ʔmun 1 1.7 27 7.7
mog wheat *muŋC 2 1.9 27
mos fine (flour) *mənX/H 1 1.9 21
mos fly (N.) *məuŋX 1 1.9 21
2.1.2.1, 2.1.4,
mus go *n-mʉŋ(X) 1 1.9 30
4.3, 4.4
muab grasp *ʔmu̯ɛA 2 1.7 11 4.4
muag eye *mu̯ɛjH 1 1.9 11
muag sell *mɛjH 1 1.9 11 2.1.3.1, 4.4
muag soft *mlu̯ɛjH 1 1.39 11 6.3.4
2.1.2.1, 2.1.3.1,
muaj have *n-mɛj 1 1.9 11
4.3, 4.4
 Index of proto forms by White Hmong 277

WHm English Reconstr L Onset Rime Chapter

muas buy *mɛjX 1 1.9 11 2.1.3.1, 4.4
mloog listen *(S-)mru̯ɔŋH 1 1.54 29 4.4
nab snake *ʔnaŋ 1 2.7 24
nag rain *m-noŋC 2 2.9 28 2.1.2.1
nas mouse/rat *naŋB 2 2.9 24
nej 2PL (you-all) *mi ̯əu 1 1.9 1 5.3, 6.3.4
neej person *næn A
2 2.9 19
nees horse *mjæn B
2 1.24 19 6.3.1
niam female *min D
2 1.9 18
no cold *ʔnu̯ɔmH 1 2.7 27
no this *ʔneinX 1 2.7 25 5.4
noj eat *nuŋA 2 2.9 27 2.1.4
noob seed *hnru̯em 1 2.53 29
2.1.2.1, 2.1.4,
noog bird *m-nɔk 1 2.9 29
4.2.4, 6.3.2
nug ask *nɛŋC 2 2.9 22
nus mother’s brother *neuŋX 1 2.9 23
nws 3SG (he/she/it) *ni ̯æn(X) 1 2.9 1 2.1.4, 5.3
ncaj straight *ɲci ̯aŋ A
2 4.4 26 4.4
ncau bamboo strip *ɲcəuk 1 4.4 6 7.6
ncauj mouth/beak *ɲɟuj 1 4.6 9
ncaws peck/dig *ɲcow D
2 4.4 13
nce climb *ɲcæw C
2 4.4 3
nceb mushroom *ŋkjæu 1 5.19 3
ncej pillar *ɲɟæu 1 4.6 3
nco remember *ɲcoC 2 4.4 7
ncoo pillow *ɲcu̯əmH 1 4.4 29
ncuav rice cake *ŋkjuəX 1 5.19 16
ncho cooking smoke *ɲchi ̯ouH 1 4.5 7
nchuav sprinkle/spill *ɲchu B
2 4.5 16
nkag crawl *ŋgjoŋ C
2 5.21 28
nkaj indigo *ŋglam 1 5.36 24
nkawm pair *ŋgjowD 2 5.21 13
nkees lazy *ŋglænB 2 5.36 19
nkoj boat/ship *ŋgjəŋA 2 5.21 21
nkos muddy *ŋko D
2 5.4 7
nkoog thick (liquid) *ŋgjɔŋ C
2 5.21 29
nkuaj cattle pen *ŋgluə 1 5.36 16
nkhaus crooked *ŋkhuw D
2 5.5 9
nkhawb soot *ŋkhow A
2 5.5 13
npau boil (INTR) *mpu̯æiH 1 1.4 12
npau dream *mpeiH 1 1.4 12 4.3, 6.3.3
278 Indexes

WHm English Reconstr L Onset Rime Chapter

npe name *mpɔuH 1 1.4 3 4.3
npua pig *mpæC 2 1.4 5 4.2, 7.1
npua throw on (clothes) *mpu̯aC 2 1.4 15
npuaj clap *mbæA 2 1.6 5
npuav hold in mouth *mpæB 2 1.4 5
nphoo scatter/sprinkle *mphu̯ənH 1 1.5 29
nplag smooth/glossy *mbli ̯əŋH 1 1.36 18
nplaig tongue *mblet 1 1.36 10 4.4, 6.2.2
nplaum glutinous/sticky *mblut 1 1.36 9
nplawm whip *mblow D
2 1.36 13 7.6
nplej rice plant/paddy *mbləu 1 1.36 3 7.4
nplooj leaf *mblɔŋA 2 1.36 29 6.2.4
nplhaib finger ring *mphleA 2 1.35 10 2.1.2.2, 7.6
nqaij meat *NGej 1 5-6.6 10 7.3
nqaim narrow *ɴɢe D
2 5-6.6 10
nqe hook *ɴqæw C
2 5-6.4 3
nqe price *ɴqa C
2 5-6.4 4 2.1.2.2
nqes descend *ɴɢa B
2 5-6.6 4 4.3, 6.2.2
nqeeb cogongrass *NKan 1 5-6.4 19 4.3, 7.2
nqos shuttle *ɴɢəŋ B
2 5-6.6 21 7.6
nqos swallow, to *ɴɢəŋB 2 5-6.6 21
nqov low/bellow *ɴqəŋB 2 5-6.4 21
nquab pigeon *ɴquA 2 6.4 16
nquag diligent *NGəjH 1 5-6.6 15
nqhis thirsty *NKhat 1 5-6.5 4
nqhuab dry/thirsty *NKhæj 1 5-6.5 5
nrab center/middle *ntroŋ 1 2.49 28
nras plain (level ground) *ndroŋ B
2 2.51 28
nraim seek shelter *ŋgre D
2 5.51 10
nrau butt (of bulls) *ntruwC 2 2.49 9
nraug young man *ndru̯eiC 2 2.51 12
nrig lean (on stick) *ŋgraC 2 5.51 4
nro turbid *ŋkro /ŋgro
B B
2 5.49 7
nrob breast/chest *ntrəŋ A
2 2.49 21
nrog drip *ŋgro C
2 5.51 7
nrog interior *ndrəŋ C
2 2.51 21
nrog track, to *ndro C
2 2.51 7 7.3
nruas drum *ndruX 1 2.51 16 7.7
nrhoob puttees *nthrɔŋA 2 2.50 29 2.1.2.2, 7.6
ntab float *ntaŋA 2 2.4 24
ntaj sword, long *ndaŋA 2 2.6 24
ntais have a gap *NKʷet 1 5-6’.4 10
 Index of proto forms by White Hmong 279

WHm English Reconstr L Onset Rime Chapter

ntau many/much *ntɔC 2 2.4 6 7.5
ntaub cloth *ntei 1 2.4 12 7.6
ntaus beat (someone) *ntuwD 2 2.4 9
ntaws navel *ntəut 1 2.4 13 2.1.2.1
ntawv paper *ntowB 2 2.4 13 7.6, 6.2.4
nte warm self by fire *ntauH 1 2.4 3
nteg lay eggs *ndəuH 1 2.6 3
ntev long *ntauX 1 2.4 3 6.3.4
ntiv finger *nta B
2 2.4 4 2.1.4
ntiv flip with finger *ntik 1 2.4 2
ntiaj front surface *ɴɢlin A
2 6.36 18
ntos weave *ntət 1 2.4 7 7.6
ntov chop *ntəwX 1 2.4 7
ntoo put on/wear (cap) *ntɔŋH 1 2.4 29 4.3, 7.6
ntoo tree *ntju̯əŋH 1 2.19 29 6.3.3
ntub wet *ntu̯ɛn 1 2.4 22
ntuj sky/heaven *ndɛuŋ 1 2.6 22
ntuag ragged/shabby *ɴɢlu C
2 6.36 16
ntuag ramie/hemp *nduH 1 2.6 16 7.6
ntws flow *ɴɢlʉ B
2 6.36 8
nthab attic *nthaŋA 2 2.5 24 2.1.2.2, 7.6
nthua weed/hoe *nthuC 2 2.5 16
nthuav open out/undo *nthu̯aB 2 2.5 15
ntsaj groan/moan *mbraŋA 2 1.51 24
ntsais wink *ntsjep 1 3.19 10
ntsaum ant *mbrɔ D
2 1.51 6 6.3.3
ntsaws stop up/plug *ntsjot 1 3.19 13
ntse sharp *ntsjæw C
2 3.19 3
ntse soak *ntsjæw C
2 3.19 3
ntseg ear *mbræu 1 1.51 3
ntseg vertical/upright *ndzjæwC 2 3.21 3
ntses fish *mbrəuX 1 1.51 3 4.2.2, 4.3, 6.3.2
ntsev salt *ntsjəuX 1 3.19 3 4.3
ntsim spicy *mbrat 1 1.51 4
ntsia drive in (nail) *ntsjin C
2 3.19 18
ntsuab green/blue *mpru A
2 1.49 16 7.7
ntsuag bamboo shoot/orphan *mbrəjH 1 1.51 15 6.3.2
ntsuas sorghum *ntsju C
2 3.19 16
ntswg nose *mbruiH 1 1.51 8 2.1.2.2, 4.2.2
ntsws lung *mprʉC 2 1.49 8
ntshav blood *ntshjamX 1 3.20 24 6.3.3
ntshai fear/be afraid *ntshjeH 1 3.20 10 4.3
280 Indexes

WHm English Reconstr L Onset Rime Chapter

ntshauv louse, head *ntshjeiX 1 3.20 12 6.3.3
ntshiab clear *ntshji ̯əŋ 1 3.20 18 4.3
ntshuab otter *ntshju̯aA 2 3.20 15
ntxa grave/tomb *ntsaŋC 2 3.4 24
ntxaib twin *mpljeA 2 1.35.1 10
ntxais suck *ntse D
2 3.4 10
ntxaug thin (person) *ndzu̯ei C
2 3.6 12
ntxiv repair/mend *mpjaX 1 1.19 4
ntxias braid/a braid *mbjinX 1 1.21 18 4.4
ntxov early *ntsi ̯ouX 1 3.4 7
ntxuag have food with rice *mbljæ C
2 1.36.1 5
ntxuam fan *mbljæp 1 1.36.1 5 7.6
ntxuav wash (hands) *ntsæwX 1 3.4 5
ntxhab steep *ntshoŋA 2 3.5 28
ntxhais daughter/girl *mphje D
2 1.20 10 2.1.2.2
ntxhib rough/coarse *ntsha A
2 3.5 4
ntxhua launder *ntshuH 1 3.5 16
ntxhw elephant *ntshʉ C
2 3.5 8
nyab sister-in-law *ʔɲam 1 4.7 24 6.3.1
nyiag steal *ɲemH 1 4.9 20
nyiaj silver *ɲʷi ̯ən 1 4’.9 18 3.2, 7.1
nyias thin *ɲenB 2 4.9 20
nyiav weep *ʔɲæmX 1 4.7 20 6.3.3
nyob be at/live *ʔɲəŋ A
2 4.7 21 4.4
nyoos raw/unripe *ɲu̯emX 2 4.9 29
nyuj water buffalo/cow *ŋiuŋ 1 5.9 23
o swell *ʔəumH 1 7.1 21
ob two *ʔu̯i 1 7.1 14 5.2, 7.5
os duck *ʔap 1 7.1 7
oo cloud/haze *ʔɔŋC 2 7.1 29
pa air *poŋ C 2 1.1 28
paj flower *bi ̯aŋ 1 1.3 24 2.1.2.2, 4.2
paub know *pei 1 1.1 12
paug pus *bu̯ei C
2 1.3 12 7.1
pauj repay *bu̯ei A
2 1.3 12
pauv soybean *N-peiX 1 1.1 12
peb 1PL (we/us) *N-pɔu 1 1.1 3 5.2, 5.3
peb three *pjɔu 1 1.16 3 5.2, 5.3, 7.5
pes mountain *bæwB 2 1.3 3
poj female/woman *boA 2 1.3 7
pom see *bu̯ət 1 1.3 7
pos thorn *-boB 2 1.3 7 4.2.3
 Index of proto forms by White Hmong 281

WHm English Reconstr L Onset Rime Chapter

pov shoot *pənX 1 1.1 21 6.3.3, 7.3
poob fall *pʉŋA 2 1.1 30 4.4
pub feed *pɛŋA 2 1.1 22
puv full *pu̯ɛŋX 1 1.1 22 6.3.3
pua hundred *pæk 1 1.1 5 2.1.4, 3.2, 5.2
puab molar tooth *pæ 1 1.1 5
puab thigh *pæ A
2 1.1 5 4.2
puag embrace, hatch *buəH 1 1.3 16
puas bad/spoiled *bu̯a B
2 1.3 15
pw sleep/lie down *pu̯eiH 1 1.1 8
pwg shoulder *bʉ C
2 1.3 8
phob CLF-quilts *phəan 1 1.2 21 6.2.4
phua cut open *phajH 1 1.2 15 6.2.2
phuab push earth aside *phʉA 2 1.2 8
plab stomach *plaŋA 2 1.31 24
plas owl *ploŋ D
2 1.31 28
plaub four *plei 1 1.31 12 5.2, 6.3.1, 7.5
plaub hair *pljei 1 1.31.1 12
plawv heart *pljow B
2 1.31.1 13 7.1
plig soul/spirit *blja A
2 1.33.1 4 7.7
plis wildcat *pljɨD 2 1.31.1 2 7.3
pliag mussel/clam *blinC 2 1.33 18
pliaj forehead *bl-A 1 1.33 xx
plhaub shell/covering *phlei 1 1.32 12 2.1.2.2
plhu cheek/face *phl- C
2 1.32 xx
plhws stroke *phlʉ C
2 1.32 8
qab back (vs. front) *N-Kaŋ 1 5-6.1 24 2.1.2.2
qab sweet *Kam 1 5-6.1 24 2.1.2.2
qav frog *KæŋX 1 5-6.1 24 2.1.2.2
qaib chicken *Kəi 1 5-6.1 10 2.1.2.2, 6.3.2
qaij slant/slanting *ɢeA 2 5-6.3 10
qaug fall (over) *ɢu̯eiC 2 6.3 12
qaum back (of body) *ɢuw D
2 6.3 9
qe egg *qæw C
2 5-6.1 3
qej garlic *ɢa A
2 5-6.3 4
qes low/short *GaX 1 5-6.3 4 4.3, 6.2.2
qev borrow *KaX 1 5-6.1 4
qeej bamboo pipes *ɢæn A
2 5-6.3 19
qoob crops *qɔŋA 2 5-6.1 29
qub old *qu̯oH 1 6.1 7
qub star *qɛŋA 2 5-6.1 22
qua crow, to *KajH 1 5-6.1 15
282 Indexes

WHm English Reconstr L Onset Rime Chapter

quaj sing/cry out *Gæw 1 5-6.3 5 6.2.2
quav excrement *N-KəjX 1 5-6.1 15 2.1.2.2, 4.3
qwj snail *ɢʉA 2 6.3 8
qhaib tie/cross (legs) *qheA 2 5-6.2 10
qhiav ginger *qhinB 2 5-6.2 18
qhov hole *qhəŋ B
2 5-6.2 21 6.2.4
qhua guest *Khæk 1 5-6.2 5 2.1.4
qhuav dry/withered *KhæwX 1 5-6.2 5 2.1.2.2
B
qhwv pack/wrap *qhʉ 2 5-6.2 8
rab CLF-tools *traŋ A
2 2.46 24 6.2.4
raj flute/tube *droŋ A
2 2.48 28
rau hoof/claw/nail *tru̯eiC 2 2.46 12
rau put on/wear (shoes) *trɔC 2 2.46 6 4.4
2.1.4, 5.2,
rau six *kruk 1 5.46 9
6.3.1, 7.5
raug hit target *drɔC 2 2.48 6 4.4
rauv burn (TR) *kru̯eiB 2 5.46 12
rawg chopsticks *drouH 1 2.48 13 3.2
rawm bamboo *drow D
2 2.48 13
rog fat, to be *grəunH 1 5.48 21 4.4
roj animal fat/oil *grəŋ A
2 5.48 21 4.4
roj step *dro A
2 2.48 7
ros laugh *krət 1 5.46 7
rov return *trəŋB 2 2.46 21
rooj door/gate *gru̯eŋ 1 5.48 29 7.6
rooj table/bench *drɔŋA 2 2.48 29
ruaj firm/durable *druA 2 2.48 16
rhais insert *threp 1 2.47 10
rho pull out *thro C
2 2.47 7
sai quick *hre C
2 2.56 10
sau receive/gather in *sjuw A
2 2.28 9
sau write *hru̯ei C
2 2.56 12
sawv stand/get up *sjouX 1 2.28 13
sib light(weight) *-sji ̯eu 1 2.28 2
sib RECIPROCAL *sjɨA-D 2 2.28 2 4.1, 6.2.3
siab high/tall *-hri ̯əŋ 1 2.56 18 2.1.3.1
siab liver *-hri ̯ən 1 2.56 18 2.1.3.1
siab measure for grain *sjin A
2 2.28 18
siav cooked/ripe *sjen B
2 2.28 20
so rest, to *sjo C
2 2.28 7
so wipe *sjəŋ C
2 2.28 21
sov warm *sji ̯ouX 1 2.28 7
 Index of proto forms by White Hmong 283

WHm English Reconstr L Onset Rime Chapter

soob thin *sjɔŋA 2 2.28 29
su midday meal *sjɛŋC 2 2.28 22
suab fern *-hruA 2 2.56 16
suab sound/noise *-hruA 2 2.56 16
tas half (of day) *N-dam(X) 1 2.3 24 4.3
tav rib *tamX 1 2.1 24
tais bowl *de B
2 2.3 10 7.1
tais pick up/clamp *teD
2 2.1 10
tau get/gain *təuk 1 2.1 6 6.2.3
taub gourd *tuw A
2 2.1 9
taum bean *dup 1 2.3 9 7.4
taus axe *tu̯eiC 2 2.1 12 7.1, 7.6
taw foot *towC 2 2.1 13
tawg explode *douH 1 2.3 13 4.4
taws fire *douX 1 2.3 13 4.4
taws ignite/light *tow D
2 2.1 13 4.4
tawv skin *N-top 1 2.1 13 2.1.2.1
te frost *tæw C
2 2.1 3
teb answer *tau 1 2.1 3
teb earth *N-təu 1 2.1 3 4.3
tes hand/arm *-bɔuX 1 1.3 3 2.1.2.1, 4.2.3
tij elder brother *da 1 2.3 4
tis wing *N-tat 1 2.1 4 4.3
tiab skirt *tin A
2 2.1 18 7.6
tiaj level *din A
2 2.3 18
tob deep *N-tu̯o 1 2.1 7 4.3
tog sink *dəŋ A
2 2.3 21
tom bite *dəp 1 2.3 7 6.3.4
tos wait *dəŋ B
2 2.3 21
tooj copper *dɔŋ 1 2.3 29 3.2, 7.1
tu snap *tɛŋH 1 2.1 22
tub son/boy *tu̯ɛn 1 2.1 22 4.2, 6.2.4
tus CLF-horses *dɛŋ B
2 2.3 22 6.2.4
tuv louse, body *tɛmX 1 2.1 22 6.3.4
tua kill *təjH 1 2.1 15 4.4, 6.3.4
tuab thick *tæ A
2 2.1 5 7.1
tuag die *dəjH 1 2.3 15 4.4, 6.3.2, 6.3.4
tuaj come *daj 1 2.3 15 2.1.3.1
tuam step on/tread *dæD 2 2.3 5
tuav hold/grasp hand *tæB 2 2.1 5
tuav husk/pound rice *tuX 1 2.1 16
tw tail *tu̯eiX 1 2.1 8 4.2.2, 6.3.3
284 Indexes

WHm English Reconstr L Onset Rime Chapter

twg which *dʉC 2 2.3 8
thee charcoal *thanH 1 2.2 19 3.2
thi hoop *thɨC 2 2.2 2
thoob bucket *thɔŋ(X) 1 2.2 29 3.2, 6.2.4
tsaim chin/lower jaw *dzjeD 2 3.18 10
tsaug porcupine *N-dzjeiH 1 3.18 12 4.3
tsaus dark *pruw D
2 1.46 9
tsawg few/lack *dzjok 1 3.18 13
tsaws pheasant *tsjow D
2 3.16 13
tseb to sow *præw A
2 1.46 3
tsev house *prəuX 1 1.46 3 7.6
tsi pour *praC 2 1.46 4
-tsi what *tsjɨB 2 3.16 2
tsib five *pra 1 1.46 4 5.2, 6.3.1, 7.5
tsib gall bladder *tsji ̯əu 1 3.16 1
tsim wake *dzja D
2 3.18 4
tso let go *tsjəŋ C
2 3.16 21
tsos armpit *tsjo C/D
2 3.16 7
B
tsov tiger *tsjo 2 3.16 7
tsoov winnow *prɔŋ B
2 1.46 29
tsu rice steamer *tsjɛŋH 1 3.16 22 3.2
tsuag bland/tasteless *dzjæC 2 3.18 5
tsuag rat *bru̯aC 2 1.48 15
tsw malodorous *tsju̯eiH 1 3.16 8
tshaib hungry *tshje 1 3.17 10
tshau sieve, to *tshjɔ C
2 3.17 6
tshauv ashes *tshju̯əiX 1 3.17 12
tshiab new *tshji ̯əŋ 1 3.17 18 4.3
tshov play (flute) *phro B
2 1.47 7
tshuab blow, to *phruA 2 1.47 16
tshuab spinning wheel *tshju̯aA 2 3.17 15
txaj ashamed *dzaŋA 2 3.3 24
txais receive/borrow *tsep 1 3.1 10
txaug chisel *dzəuk 1 3.3 6 4.2
A
txawj able *dzow 2 3.3 13
txi bear fruit *pji C
2 1.16 1 4.4
txiv father/male *pjaX 1 1.16 4
txiv fruit *pji ̯əuX 1 1.16 1 4.4
txia crop of bird *pjinB 2 1.16 18
txiaj money *dzi ̯en 1 3.3 20
txias cool (water) *dzinB 2 3.3 18
txog reach/arrive at *dzo C
2 3.3 7
 Index of proto forms by White Hmong 285

WHm English Reconstr L Onset Rime Chapter

txos stove *N-tsoH 1 3.1 7 3.2
txoob mane *tsʉŋA 2 3.1 30 4.2.3
txuas billhook *tsæC 2 3.1 5 7.6
txuas join *tsæD 2 3.1 5
txha bone *tshuŋX 1 3.2 28 4.2.3
txhab sore/boil/blister *tshaŋ 1 3.2 24
txhib split/chop *phek 1 1.2 2 6.3.2
txhiab thousand *tshi ̯en 1 3.2 20 3.2, 5.2
txhuv rice/millet, husked *tshɛŋ B
2 3.2 22 7.4
txhuas tin *tshæ D
2 3.2 5
ua do/work *ʔəjH 1 7.1 15
uab crow (bird) *ʔuA 2 7.1 16
vab winnowing fan *ʔwaŋA 2 1.10 24
vaj garden *waŋA 2 1.12 24
vam ten thousand *wi ̯aŋ C
2 1.12 26
vauv son-in-law *ʔweiX 1 1.10 12 6.3.1
vawg taro *wouH 1 1.12 13
viv wife/elder sister *ʔwæw B
2 1.10 3
vov cover self *ʔwo B
2 1.10 7
vuas tile *ŋʷæX 1 5’.9 5 2.1.2.2, 3.2
xa send/deliver *suŋH 1 2.13 28
xiab first day of month *seŋ 1 2.13 20
xob thunder *S-phoA 2 1.2 7
xov thread *S-pho B
2 1.2 7
xua chaff/husk *S-phjæ C
2 1.17 5
xuab twist/rub *S-phu̯a A
2 1.2 15
xya breed/propagate *ɕaŋ C
2 4.13 24
xya seven *djuŋH 1 2.18 28 5.2, 6.3.1, 7.5
xyab stretch *ɕoŋ 1 4.13 28
xyoo year *ɕɔŋC 2 4.13 29 5.2
ya fly, to *ʔjaŋC 2 4.10 24
yaj dissolve/melt *jaŋA 2 4.12 24
yaj sheep/goat *juŋ 1 4.12 28 3.2
yaim lick *-jep 1 4.12 10
C
yau small/young *ʔjuw 2 4.10 9
yim eight *jat 1 4.12 4 5.2, 6.3.1, 7.5
yis whorl at top of head *jɨB
2 4.12 2
yias cooking pot/pan *wjin B
2 1.27 18
yog be *jəŋC 2 4.12 21 4.4
yoov mosquito *ʔjɔŋB 2 4.10 29
yub seedling *ʔjɛŋA 2 4.10 22 3.2, 7.4
yug raise (chickens) *jɛŋ C
2 4.12 22
286 Indexes

WHm English Reconstr L Onset Rime Chapter

yuj tree of heaven *jɛŋA 2 4.12 22
yuav ask for/want *ʔjuB 2 4.10 16
zaj dragon *-roŋ 1 2.57 28
zais conceal *ʔreD 2 2.55 10
zaub vegetable *ʔræi 1 2.55 12
zaug pangolin *rɔiH 1 2.57 6 7.2
ze near *-ʔræw C
2 2.55 3
2.1.2.2, 4.2.2,
zeb stone *-ʔrəu 1 2.55 3
6.2.2
zes nest *-rəuX 1 2.57 3
zib honey *ʔra A
2 2.55 4
zis urine *-raX 1 2.57 4
ziab dry (rice) in sun *ʔrin A
2 2.55 18 7.4
zog strength *-rək 1 2.57 7 2.1.4, 3.2
zom grind (grain) *roD 2 2.57 7
zos village *rəŋX 1 2.57 21 6.2.2
zov look after *ʔroB 2 7
zoo good *-ʔrɔŋH 1 2.55 29 2.1.2.2
zoov forest *-ʔrɔŋ B
2 2.55 29 2.1.2.2
zuag comb *ræ C
2 2.57 5
zuag sharp *-rajH 1 2.57 15
 References

ACD = Blust, Robert. n.d. Austronesian comparative dictionary. Manuscript.

Aikhenvald, Alexandra Y. 2000. Classifiers: a typology of noun categorization devices.
Oxford: Oxford University Press.
Aikhenvald, Alexandra Y. and R. M. W. Dixon (eds). 2001. Areal diffusion and genetic
inheritance: problems in comparative linguistics. Oxford: Oxford University Press.
Anderson, Edward F. 1993. Plants and people of the Golden Triangle: ethnobotany of the
hill tribes of Northern Thailand. Portland: Dioscorides Press.
Andruski, Jean and Martha Ratliff. 2000. Phonation types in production of phonological
tone: the case of Green Mong. Journal of the International Phonetic Association
30:37–61.
Anttila, Raimo. 1972. An introduction to historical and comparative linguistics. New
York: Macmillan.
Baxter, William H. 1983. A look at the history of Chinese color terminology. Journal of
the Chinese Language Teachers Association 18:1–25.
Baxter, William H. 1992. A handbook of old Chinese phonology. Berlin: Mouton de
Gruyter.
Baxter, William H. 2000. An etymological dictionary of common Chinese characters.
Manuscript.
Baxter, William H. and Laurent Sagart. 2009. The Baxter-Sagart reconstruction of Old
Chinese. Personal Communication.
Bellwood, Peter. 1996. The origins and spread of agriculture in the Indo-Pacific region. In
David R. Harris (ed.) The origins and spread of agriculture and pastoralism in
Eurasia, 465-498. London: University College London Press.
Bellwood, Peter. 2001. Archaeology and the historical determinants of punctuation in
language family origins. In Alexandra Aikhenvald and R. M. W. Dixon (eds), 27–43.
Bellwood, Peter. 2005. Examining the farming/language dispersal hypothesis in the East
Asian context. In Sagart, Blench, and Sanchez-Mazas (eds), 17–30.
Benedict, Paul K. 1942. Thai, Kadai, and Indonesian: a new alignment in Southeastern
Asia. American Anthropologist 44.4:576–601.
Benedict, Paul K. 1972. Sino-Tibetan: a conspectus. Cambridge: Cambridge University
Press.
Benedict, Paul K. 1975. Austro-Thai language and culture with a glossary of roots. New
Haven: Human Relations Area Files Press.
Benedict, Paul K. 1986. Miao-Yao enigma: the Na-e language. Linguistics of the Tibeto-
Burman Area 9.1:89–90.

287
288 References

Benedict, Paul K. 1987a. Early MY/TB loan relationships. Linguistics of the Tibeto-
Burman Area 10.2:12–21.
Benedict, Paul K. 1987b. Archaic Chinese initials. Manuscript.
Benedict, Paul K. 1996. Monic *clur and other Southeast Asian ‘dogs’. Mon-Khmer
Studies 26:3–6.
Benedict, Paul K. 1997. Interphyla flow in Southeast Asia. Mon-Khmer Studies 27:1–11.
Berlin, Brent and Paul Kay. 1969. Basic color terms: their universality and evolution.
Berkeley: University of California Press.
Bertrais, Yves. 1979 [1964]. Dictionnaire hmong-français. Bangkok: Sangwan Surasarang.
Bisang, Walter. 1999. Classifiers in East and Southeast Asian languages: counting and
beyond. In Jadranka Gvozdanović (ed.) Numeral types and changes worldwide, 113–
185. Berlin: Mouton de Gruyter.
Blench, Roger. 2005. From the mountains to the valleys: understanding ethnolinguistic
geography in Southeast Asia. In Sagart, Blench, and Sanchez-Mazas (eds), 31–50.
Blust, Robert. n.d. see ACD.
Chamberlain, James R. 1972. Tone borrowing in five northeastern dialects. In Jimmy G.
Harris and Richard B. Noss (eds) Tai phonetics and phonology, 43–46. Bangkok:
Central Institute of English Language, Mahidol University.
Chang, Kun. 1947. Miáoyáoyǔ shēngdiào wèntí [On the tone system of the Miao-Yao
languages]. Bulletin of the Institute of History and Philology 16:93–110.
Chang, Kun. 1953. On the tone system of the Miao-Yao languages. Language 29:374–378.
Chang, Kun. 1966. A comparative study of the Yao tone system. Language 42:303–310.
Chang, Kun. 1972. The reconstruction of Proto-Miao-Yao tones. Bulletin of the Institute of
History and Philology 44:541–628.
Chang, Kun. 1976. Proto-Miao initials. Bulletin of the Institute of History and Philology
47:155–218.
Chang, Te-Tzu. 1983. The origins and early cultures of the cereal grains and food legumes.
In David N. Keightley (ed.) The origins of Chinese civilization, 65–94. Berkeley:
University of California Press.
Chen, Qiguang. 1993. Miáoyáoyǔ qiánzhùi [Miao-Yao prefixes]. Mínzú yǔwén 1993.1:1–
9.
Cí hǎi. 辭 海 [Sea of Words], third ed. 1979. Shanghai: Shànghǎi císhū chūbǎnshè
[Shanghai Dictionary Press].
Compton, Carol. 1994. Choosing Lao pronouns for conversational purposes: negotiating
and establishing relationships verbally. In Karen L. Adams and Thomas John Hudak
(eds) Papers from the Second Annual Meeting of the Southeast Asian Linguistics
Society (1992), 135–146. Tempe, Arizona: Program for Southeast Asian Studies.
Court, Christopher. 1975. The segmental and suprasegmental representation of Malay
loanwords in Satun Thai: a description with historical remarks. In Jimmy G. Harris
and James R. Chamberlain (eds) Studies in Tai linguistics in honor of William J.
Gedney, 67–88. Bangkok: Central Institute of English Language, Office of State
Universities.
 References 289

Court, Christopher. 1985. Fundamentals of Iu Mien (Yao) grammar. PhD dissertation,

University of California, Berkeley.
Craig, Colette (ed.). 1986. Noun classes and categorization. Amsterdam and Philadelphia:
John Benjamins.
Culas, Christian and Jean Michaud. 2004. A contribution to the study of Hmong (Miao)
migrations and history. In Tapp, Michaud, Culas, and Lee (eds), 61–96.
Delancey, Scott. 1986. Toward a history of Tai classifier systems. In Craig (ed.), 437–452.
Dempsey, James M. 1995. A reconsideration of some phonological issues involved in
reconstructing Sino-Tibetan numerals. PhD dissertation, University of Washington.
Dempwolff, Otto. 1934-38. Vergleichende Lautlehre des Austronesischen Wortschatzes
(Band I-III). Berlin: Dietrich Reimer.
Diffloth, Gérard. 1989. Proto-Austroasiatic creaky voice. Mon-Khmer Studies 15:139–154.
Diffloth, Gérard. 1994. The lexical evidence for Austric, so far. Oceanic Linguistics
33:309–321.
Downer, Gordon B. 1959. Derivation by tone-change in Classical Chinese. Bulletin of the
School of Oriental and African Studies 22:258–290.
Downer, Gordon B. 1961. Phonology of the word in Highland Yao. Bulletin of the School
of Oriental and African Studies 24:531–541.
Downer, Gordon B. 1963. Chinese, Thai, and Miao-Yao. In H. L. Shorto (ed.) Linguistic
Comparison in South East Asia and the Pacific, 133–139. London: School of
Oriental and African Studies, University of London.
Downer, Gordon B. 1967. Tone-change and tone-shift in White Miao. Bulletin of the
School of Oriental and African Studies 30:589–599.
Downer, Gordon B. 1971. The further relationships of the Miao-Yao languages. Paper
presented at the 4th International Conference on Sino-Tibetan Languages and
Linguistics, Indiana University, Bloomington.
Downer, Gordon B. 1973. Strata of Chinese loanwords in the Mien dialect of Yao. Asia
Major 18.1:1–33.
Downer, Gordon B. 1978. The Tai-Yao connection. Paper circulated at the 11th
International Conference on Sino-Tibetan Languages and Linguistics, University of
Arizona, Tucson, Arizona.
Downer, Gordon B. 1979. The rimes of Proto-West-Miao. Paper presented at the 12th
International Conference on Sino-Tibetan Languages and Linguistics, Paris.
Downer, Gordon B. 1982. Problems in the reconstruction of Proto-Miao-Yao. Paper
presented at the 15th International Conference on Sino-Tibetan Languages and
Linguistics, Beijing.
Duanmu, San. 1999. The syllable in Chinese. In Harry van der Hulst and Nancy A. Ritter,
(eds) The syllable: views and facts, 477–500. Berlin: Mouton de Gruyter.
Edmondson, Jerry. 2001. Na-Meo. Manuscript.
Erbaugh, Mary S. 1986. Taking stock: the development of Chinese noun classifiers
historically and in young children. In Craig (ed.), 399–436.
Ferlus, Michel. 2009. Le mot ‘sang’ en austroasiatique. Mon-Khmer Studies 38:25–39.
Filbeck, David. 1972. Tone in a dialect of T’in. Anthropological Linguistics 14:111–118.
290 References

Forrest, R. A. D. 1973 [1948]. The Chinese language, third ed. London: Faber and Faber.
Fuller, Judith W. 1986. Chinese le and Hmong lawm. Paper presented at the 19th
International Conference on Sino-Tibetan Languages and Linguistics, The Ohio State
University, Columbus, Ohio.
Gandour, Jackson T. 1979. Tonal rules for English loanwords in Thai. In Theraphan L-
Thongkum, Vichin Panupong, Pranee Kullavanijaya, and M.R. Kalaya Tingsabadh
(eds) Studies in Tai and Mon-Khmer phonetics and phonology in honour of Eugénie
J. A. Henderson, 94–105. Bangkok: Chulalongkorn University Press.
Gerner, Matthias. 2009. Deictic features of demonstratives: a typological survey with
special reference to the Miao group. Canadian Journal of Linguistics 54.1:43-90.
Gordon, Matthew. 2002. A phonetically driven account of syllable weight. Language 78:
51–80.
Gordon, Raymond G., Jr. (ed.). 2005. Ethnologue: languages of the world, 15th edition.
Dallas: SIL International. Online version: http://www.ethnologue.com/
Hartmann, John. 1998. A linguistic geography and history of Tai meuang-fai [ditch-dike]
techno-culture. Journal of Language and Linguistics 16.2:67–100.
Haudricourt, André G. 1954a. Introduction à la phonologie historique des langues Miao-
Yao. Bulletin de l’École Française d’Extrême-Orient 44:555–574. (Reprinted in
Problémes de phonologie diachronique, by André G. Haudricourt, 1987, 183–208.
Paris: Société pour l’Étude des Langues Africaines=SELAF.)
Haudricourt, André G. 1954b. De l’origine des tons en vietnamien. Journal Asiatique
242:69–82. (Reprinted in Problémes de phonologie diachronique, by André G.
Haudricourt, 1987, 147–160. Paris: Société pour l’Étude des Langues
Africaines=SELAF.)
Haudricourt, André G. 1966. The limits and connections of Austroasiatic in the northeast.
In Norman H. Zide (ed.) Studies in comparative Austroasiatic linguistics, 44-56. The
Hague: Mouton.
Haudricourt, André G. and David Strecker. 1991. Hmong-Mien (Miao-Yao) loans in
Chinese. T’oung Pao 77.4-5:335–341.
Heimbach, Ernest E. 1979. White Hmong–English dictionary, revised ed. Linguistics
Series 4, Data Paper 75. Ithaca: Cornell University Southeast Asia Program,
Department of Asian Studies.
Hombert, Jean-Marie. 1978. Consonant types, vowel quality, and tone. In Victoria A.
Fromkin (ed.) Tone: a linguistic survey, 77–112. New York: Academic Press.
Hombert, Jean-Marie, John J. Ohala, and William E. Ewan. 1979. Phonetic explanations
for the development of tones. Language 55:37–58.
Jenks, Robert D. 1994. Insurgency and social disorder in Guizhou: the “Miao” Rebellion,
1854–1873. Honolulu: University of Hawaii Press.
Johnson, Michael. 2002. The reconstruction of labial stop + sonorant cluster in Proto-Far-
Western Hmongic. Transactions of the Philological Society 100.1:25–58.
Kay, Paul and Chad K. McDaniel. 1978. The linguistic significance of the meanings of
basic color terms. Language 54:610–646.
Kosaka, Ryuichi. 2002. On the affiliation of Miao-Yao and Kadai: can we posit the Miao-
Dai family? Mon-Khmer Studies 32:71–100.
 References 291

Kwan, Julia Chin. 1966. Phonology of a Black Miao dialect. MA thesis, University of
Washington.
L-Thongkum, Theraphan. 1991. An instrumental study of Chong registers. In J. H. C. S.
Davidson (ed.) Austroasiatic languages: essays in honour of H. L. Shorto, 141–160.
London: School of Oriental and African Studies, University of London.
L-Thongkum, Theraphan. 1993. A view on Proto-Mjuenic (Yao). Mon-Khmer Studies
22:163-230.
L-Thongkum, Theraphan. 1997. Tone change and language contact: a case study of Mien-
Yao and Thai. In Arthur S. Abramson (ed.) Southeast Asian linguistic studies in
honour of Vichin Panupong, 153–160. Bangkok: Chulalongkorn University Press.
L-Thongkum, Theraphan. 2002. The role of endangered Mon-Khmer languages of Xekong
Province, Southern Laos in the reconstruction of Proto-Katuic. In Marlys Macken
(ed.) Papers from the Tenth Annual Meeting of the Southeast Asian Linguistics
Society (2000), 407–429. Tempe, Arizona: Arizona State University: Program for
Southeast Asian Studies.
Lemoine, Jacques. 1972. Un village Hmong Vert du haut Laos. Paris: Centre National de la
Recherche Scientifique.
Lemoine, Jacques. 1982. Yao ceremonial paintings. Bangkok: White Lotus Co.
Li, Fang-Kuei. 1970. Some tonal irregularities in the Tai languages. In Roman Jakobson
and Shigeo Kawamoto (eds) Studies in General and Oriental Linguistics, 415–422.
Tokyo: TEC Corp.
Li, Fang-Kuei. 1977. A handbook of comparative Tai. Oceanic Linguistics Special
Publication 15. Honolulu: The University Press of Hawaii.
Li, Yunbing and Zongwu Mao. 1997. Bāhēngyǔ yánjiū [Studies on the Pa Hng language].
Shanghai: Yuǎndōng chūbǎnshè [Far East Press].
Loke, Kit-Ken. 1997. The grammaticalisation and regrammaticalisation of Chinese
numeral classifier morphemes. Journal of Chinese Linguistics 25.1:1–19.
Lyman, Thomas Amis. 1974. Dictionary of Mong Njua. The Hague: Mouton.
Lyman, Thomas Amis. 1979. Grammar of Mong Njua (Green Miao). Sattley, California:
The Blue Oak Press.
Maddieson, Ian. 1984a. The effects on F0 of a voicing distinction in sonorants and their
implications for a theory of tonogenesis. Journal of Phonetics 12:9–15.
Maddieson, Ian. 1984b. Patterns of sounds. Cambridge: Cambridge University Press.
Manomaivibool, Prapin. 1980. Tonal derivation in Thai. Computational Analyses of Asian
and African Languages 13:165–172.
Mao, Tsung-Wu [Mao, Zongwu] and Tsu-yao Chou. 1972. A brief description of the Yao
language. In Purnell (ed.), 239–255.
Mao, Zongwu and Yunbing Li. 2001. Jiǒngnàiyǔ yánjiū [A study of Jiongnai]. Beijing:
Zhōngyāng mínzú dàxué chūbǎnshè [Central Nationalities University Press].
Mao, Zongwu and Chaoji Meng. 1986. Shēyǔ jiǎnzhì [A sketch of the She language].
Beijing: Mínzú chūbǎnshè [Nationalities Press].
Mao, Zongwu, Chaoji Meng, and Zongze Zheng. 1982. Yáozú yǔyán jiǎnzhì [A sketch of
the languages of the Yao people]. Beijing: Mínzú chūbǎnshè [Nationalities Press].
292 References

Matisoff, James A. 1970. Glottal dissimilation and the Lahu high-rising tone: a tonogenetic
case-study. Journal of the American Oriental Society 90.1:13–44.
Matisoff, James A. 1973. Tonogenesis in Southeast Asia. In Larry M. Hyman (ed.)
Consonant types and tone. Southern California Occasional Papers in Linguistics No.
1, 71–95. Los Angeles: The Linguistics Program, University of Southern California.
Matisoff, James A. 1978. Variational semantics in Tibeto-Burman. Philadelphia: Institute
for the Study of Human Issues.
Matisoff, James A. 1990. On megalocomparison. Language 66.1:106–120.
Matisoff, James A. 1991. Areal and universal dimensions of grammatization in Lahu. In
Elizabeth Closs Traugott and Bernd Heine (eds) Approaches to grammaticalization,
Vol. 2, 383–453. Amsterdam and Philadelphia: John Benjamins.
Matisoff, James A. 1997. Sino-Tibetan numeral systems: prefixes, protoforms and
problems. Canberra: Pacific Linguistics.
Matisoff, James A. 2001. Genetic versus contact relationship: prosodic diffusibility in
South-East Asian languages. In Alexandra Y. Aikhenvald and R. M. W. Dixon (eds),
291–327.
Matisoff, James A. 2003. Handbook of Proto-Tibeto-Burman. Berkeley: University of
California Press.
Mei, Tsu-lin. 1970. Tones and prosody in Middle Chinese and the origin of the rising tone.
Harvard Journal of Asiatic Studies 20:86–110.
Meng, Chaoji. 1983. Yáozú bùnǔyǔ 1’ zhì 4’ diào de xíngchéng hé fāzhǎn [The origin and
development of tones 1’–4’ in the Bunu language of the Yao people]. Mínzú yǔwen
[Nationalities languages and literatures] 2:56–59.
Mortensen, David. 2002. A preliminary survey of Tibeto-Burman loanwords in Hmong-
Mien languages. Paper presented at the 35th Annual Conference on Sino-Tibetan
Languages and Linguistics, Arizona State University, Tempe.
Nichols, Johanna. 1992. Linguistic diversity in space and time. Chicago: University of
Chicago Press.
Niederer, Barbara. 1997. Notes comparatives sur le Pa-hng. Cahiers de Linguistique Asie
Orientale 26.1:71–130.
Niederer, Barbara. 1998. Les langues Hmong-Mjen (Miáo-Yáo): phonologie historique.
Munich: Lincom Europa.
Niederer, Barbara. 1999. Further investigation into Pa-hng dialectal diversification. Paper
presented at the 9th Annual Meeting of the Southeast Asian Linguistics Society,
University of California, Berkeley.
Niederer, Barbara. 2001a. Changpaoyao personal pronouns. Manuscript.
Niederer, Barbara. 2001b. Pa-hng and Nau-klau personal pronouns. Manuscript.
Niederer, Barbara. 2004. Pa-hng and the classification of the Hmong-Mien languages. In
Tapp, Michaud, Culas, and Lee (eds), 129–146.
Norman, Jerry. 1988. Chinese. Cambridge: Cambridge University Press.
Norman, Jerry and Tsu-lin Mei. 1976. The Austroasiatics in ancient south China: some
lexical evidence. Monumenta Serica 32:274–301.
 References 293

Office of Miao-Yao Research Lexicon = Miáoyáoyǔ fāngyán cíhuì jí [A Lexicon for Miao-
Yao Dialects]. 1987. Produced by the Office of Miao-Yao Research of the Central
Minorities Institute (= Zhōngyāng mínzú xuéyuàn miáoyáoyǔ yánjiūshì). Beijing:
Zhōngyāng mínzú xuéyuàn chūbǎnshè [Central Minorities Institute Press].
Ostapirat, Weera. 2000. Proto-Kra. Linguistics of the Tibeto-Burman Area 23.1.
Pan, Wuyun. 2006. On the genetic relationship between the Miao-Yao languages and the
Sino-Tibetan languages. Paper presented at the Workshop on Language and Genes in
East Asia/Pacific, December 12-13, Uppsala, Sweden.
Peiros, Ilia. 1998. Comparative linguistics in Southeast Asia. Canberra: Pacific Linguistics.
Posner, Rebecca. 1997. Linguistic change in French. Oxford: Oxford University Press.
Pullum, Geoffrey K. and William A. Ladusaw. 1996. Phonetic symbol guide, second ed.
Chicago: University of Chicago Press.
Pulleyblank, Edwin G. 1978. The nature of the Middle Chinese tones and their
development to Early Mandarin. Journal of Chinese Linguistics 6.2:173–203.
Pulleyblank, Edwin G. 1991. Lexicon of reconstructed pronunciation in Early Middle
Chinese, Late Middle Chinese, and Early Mandarin. Vancouver: University of
British Columbia Press.
Purnell, Herbert C., Jr. 1965. Phonology of a Yao dialect spoken in the province of
Chiengrai, Thailand. Hartford Studies in Linguistics 15.
Purnell, Herbert C., Jr. 1968. Appendix A: Numbers. In Sylvia J. Lombard (complier) Yao-
English dictionary. Linguistics Series 2, Data Paper 69, 317–320. Ithaca, New York:
Cornell University Southeast Asia Program, Department of Asian Studies.
Purnell, Herbert C., Jr. 1970. Toward a reconstruction of Proto-Miao-Yao. PhD
dissertation, Cornell University.
Purnell, Herbert C., Jr. (ed.). 1972. Miao and Yao linguistic studies: selected articles in
Chinese. Translated by Chang Yu-hung and Chu Kwo-ray. Linguistics Series 5, Data
Paper 88. Ithaca, New York: Cornell University Southeast Asia Program,
Department of Asian Studies.
Purnell, Herbert C., Jr. 1991. The metrical structure of Yiu Mienh secular songs. In
Jacques Lemoine and C. Chien (eds) The Yao of south China: recent international
studies, 369–398. Paris: Pangu, Editions de l’A.F.E.Y.
Purnell, Herbert C., Jr. 1998. Putting it all together: components of a secular song in Iu
Mien. In Shobhana Chelliah and Willem de Reuse (eds) Papers from the Fifth
Annual Meeting of the Southeast Asian Linguistics Society (1995), 277–302. Tempe,
Arizona: Arizona State University Program for Southeast Asian Studies.
Purnell, Herbert C., Jr. Manuscript. A Iu-Mien–English Dictionary with Cultural Notes.
Ratliff, Martha. 1986a. The morphological functions of tone in White Hmong. PhD
dissertation, University of Chicago.
Ratliff, Martha. 1986b. Two-word expressives in White Hmong. In Glenn L. Hendricks,
Bruce T. Downing, and Amos S. Deinard (eds) The Hmong in Transition, 219–236.
New York: Center for Migration Studies.
Ratliff, Martha. 1986c. An analysis of some tonally differentiated doublets in White
Hmong (Miao). Linguistics of the Tibeto-Burman Area 9.2:1–35.
294 References

Ratliff, Martha. 1991. Cov, the underspecified noun, and syntactic flexibility in Hmong.
Journal of the American Oriental Society 111.4:694–703.
Ratliff, Martha. 1992a. Meaningful tone: A study of tonal morphology in compounds, form
classes, and expressive phrases in White Hmong. DeKalb, Illinois: Northern Illinois
University Center for Southeast Asian Studies.
Ratliff, Martha. 1992b. Tone language type change in Africa and Asia: !Xu, Gokana and
Mpi. Diachronica 9.2:239–257.
Ratliff, Martha. 1992c. Form and function in tone languages. In Laura A. Buszard-
Welcher, et al. (eds) Special Session on the Typology of Tone Languages, Berkeley
Linguistics Society 18, 134–44. Berkeley: Berkeley Linguistics Society.
Ratliff, Martha. 1995. The reconstruction of Proto-Hmong-Mien liquids. Paper presented
at the 5th Annual Meeting of the Southeast Asian Linguistics Society, University of
Arizona, Tucson.
Ratliff, Martha. 1997. Hmong-Mien demonstratives and pattern persistence. Mon-Khmer
Studies 27:317–328.
Ratliff, Martha. 1998. Ho Ne (She) is Hmongic: one final argument. Linguistics of the
Tibeto-Burman Area 21.2:97–109.
Ratliff, Martha. 1999a. Layered systems of phonation contrast in Hmong-Mien. Paper
presented at the 73rd Annual Meeting of the Linguistic Society of America, Los
Angeles.
Ratliff, Martha. 1999b. Sino-Hmong-Mien phonology. Paper presented at the 32nd
International Conference on Sino-Tibetan Languages and Linguistics, University of
Illinois, Urbana-Champaign.
Ratliff, Martha. 2000. Numeral classifiers and classifying nominal prefixes in Hmong-
Mien: one more story of contact and its grammatical consequences. Paper presented
at the 10th Annual Meeting of the Southeast Asian Linguistics Society, University of
Wisconsin, Madison.
Ratliff, Martha. 2001a. Voiceless sonorant initials in Hmong-Mien: Sino-Tibetan
correspondences. In Graham Thurgood (ed.) Papers from the Ninth Annual Meeting
of the Southeast Asian Linguistics Society (1999), 361–375. Tempe: Arizona State
University Program for Southeast Asian Studies.
Ratliff, Martha. 2001b. A review of Hmong-Mien personal pronouns. Paper presented at
the 11th Annual Meeting of the Southeast Asian Linguistics Society, Mahidol
University, Bangkok.
Ratliff, Martha. 2002. Reconstruction arts: vowel feature condensation in Hmongic. Paper
presented at the 35th International Conference on Sino-Tibetan Languages and
Linguistics, Arizona State University, Tempe.
Ratliff, Martha. 2004. Vocabulary of environment and subsistence in the Hmong-Mien
protolanguage. In Tapp, Michaud, Culas, and Lee (eds), 147–165.
Ratliff, Martha. 2005. Timing tonogenesis: evidence from borrowing. In Patrick Chew
(ed.) Special Session on Tibeto-Burman and Southeast Asian Linguistics, Berkeley
Linguistics Society 28, 29–41. Berkeley, California: Berkeley Linguistics Society.
 References 295

Ratliff, Martha. 2009a. Loanwords in White Hmong. In Martin Haspelmath and Uri
Tadmor (eds) Loanwords in the World's Languages: A Comparative Handbook, 638-
658. Berlin and New York: De Gruyter Mouton.
Ratliff, Martha. 2009b. White Hmong vocabulary. In Martin Haspelmath and Uri Tadmor
(eds) World Loanword Database, 1,637 entries. Munich: Max Planck Digital Library
<http://www.loanwords.info/White_Hmong>.
Ratliff, Martha. 2006. Prefix variation and reconstruction. In Thomas D. Cravens (ed.)
Variation and Reconstruction, 165–178. Amsterdam and Philadelphia: John
Benjamins.
Ratliff, Martha. 2007. Contrastive vowel length in Mienic: inheritance or diffusion? In
Shoichi Iwasaki, Andrew Simpson, Paul Sidwell, and Karen Adams (eds) Papers
from the Thirteenth Annual Meeting of the Southeast Asian Linguistics Society
(2003), 225–231. Canberra: Pacific Linguistics.
Ratliff, Martha and Judith R. Holst. 2005. De-coupling basic and stable. Paper presented at
the 17th International Conference on Historical Linguistics, University of Wisconsin,
Madison.
Sagart, Laurent. 1995. Chinese ‘buy’ and ‘sell’ and the direction of borrowings between
Chinese and Hmong-Mien: a response to Haudricourt and Strecker. T’oung Pao
81.4-5:328–342.
Sagart, Laurent. 1999. The roots of Old Chinese. Amsterdam and Philadelphia: John
Benjamins.
Sagart, Laurent. 2005. Sino-Tibetan-Austronesian: an updated and improved argument. In
Sagart, Blench, and Sanchez-Mazas (eds), 161–176.
Sagart, Laurent, Roger Blench, and Alicia Sanchez-Mazas (eds). 2005. The peopling of
East Asia: putting together archaeology, linguistics and genetics. London: Routledge
Curzon.
Savina, F. M. 1924. Histoire des Miao. Paris: Société des Missions-Etrangères.
Shintani, Tadahiko L. A. and Zhao Yang. 1990. The Mun Language of Hainan Island: its
classified lexicon. Tokyo: Institute for the Study of Languages and Cultures of Asia
and Africa.
Shorto, H. L. 2006. A Mon-Khmer comparative dictionary, ed. by Paul Sidwell. Canberra:
Pacific Linguistics.
Solnit, David B. 1982. Biao Min field notes. Card file.
Solnit, David B. 1985. Introduction to the Biao Min Yao language. Cahiers de
Linguistique Asie Orientale 14.2:175–191.
Solnit, David B. 1996. Some evidence from Biao Min on the initials of Proto-Mienic (Yao)
and Proto-Hmong-Mien (Miao-Yao). Linguistics of the Tibeto-Burman Area 19.1:1–
18.
Starosta, Stanley. 2005. Proto-East Asian and the origin of dispersal of the languages of
East and Southeast Asia and the Pacific. In Sagart, Blench, and Sanchez-Mazas
(eds), 182–197.
Strecker, David. 1987a. Some comments on Benedict’s “Miao-Yao enigma: the Na-e
language”. Linguistics of the Tibeto-Burman Area 10.2:22–42.
296 References

Strecker, David. 1987b. Some comments on Benedict’s “Miao-Yao enigma”: addendum.

Linguistics of the Tibeto-Burman Area 10.2:43–53.
Strecker, David. 1989. Hmongic noun prefixes. Paper presented at the 6th Annual Meeting
of the Linguistic Society of America, New York.
Strecker, David. 1990. The tones of the Houei Sai dialect of the Mun language. Cahiers de
Linguistique Asie Orientale 19.1:5–33.
Sun, Hongkai. 1992. Language recognition and nationality. International Journal of the
Sociology of Language 97:9–22.
Taguchi, Yoshihisa. 2005. Purnell’s *T and the reconstruction of Proto-Hmong-Mien
cluster initials. Paper presented at the 38th International Conference on Sino-Tibetan
Languages and Linguistics, Xiamen University.
Taguchi, Yoshihisa. 2006. Rhotic and cluster initials in Proto-Hmong-Mien. Paper
presented at the 39th International Conference on Sino-Tibetan Languages and
Linguistics, University of Washington.
Tapp, Nicholas, Jean Michaud, Christian Culas, and Gary Yia Lee (eds). 2004.
Hmong/Miao in Asia. Chiang Mai, Thailand: Silkworm Press (distributed by
University of Washington Press).
Thomason, Sarah G. 2001. Language contact: an introduction. Washington, D.C.:
Georgetown University Press.
Thomason, Sarah G. and Terrence Kaufman. 1988. Language contact, creolization, and
genetic linguistics. Berkeley: University of California Press.
Thurgood, Graham. 1988. Notes on the reconstruction of Proto-Kam-Sui. In Jerold A.
Edmondson and David B. Solnit (eds) Comparative Kadai: linguistic studies beyond
Tai, 179–218. Dallas, Texas: SIL and University of Texas, Arlington.
Thurgood, Graham. 1999. From ancient Cham to modern dialects: two thousand years of
language contact and change. Oceanic Linguistics Special Publication No. 28.
Honolulu: University of Hawai’i Press.
Thurgood, Graham. 2002. Vietnamese and tonogenesis: revising the model and the
analysis. Diachronica 19.2:333–363.
Wang, Fushi. 1972 [1957]. The classifier in the Wei Ning dialect of the Miao language in
Kweichow. In Purnell (ed.), 111–185.
Wang, Fushi. 1979. Miáoyǔ fāngyán shēngyùnmǔ bǐjiào [The comparison of initials and
finals of Miao dialects]. Monograph presented at the 12th International Conference
on Sino-Tibetan Languages and Linguistics, Paris.
Wang, Fushi. 1986. A preliminary investigation of the genetic affiliation of the Miao-Yao
languages. Paper presented at the International Symposium on the Minority
Nationalities of China, January 27-29, Santa Barbara, California.
Wang, Fushi. 1994. Miáoyǔ gǔyīn gòunǐ [Reconstruction of the sound system of proto-
Miao]. Tokyo: Institute for the Study of Languages and Cultures of Asia and Africa.
Wang, Fushi (ed.). 1985. Miáoyǔ jiǎnzhì [A sketch of the Miao language]. Beijing: Mínzú
chūbǎnshè [Nationalities Press].
Wang, Fushi and Zongwu Mao. 1995. Miáoyáoyǔ gǔyīn gòunǐ [Reconstruction of the
sound system of proto-Miao-Yao]. Beijing: Zhōngguó shèhuì kēxué chūbǎnshè
[China Social Sciences Press].
 References 297

Wessing, Robert. 1986. The soul of ambiguity: the tiger in Southeast Asia. DeKalb:
Northern Illinois University Center for Southeast Asian Studies.
Wildlife of China. n.d. Zhōngguó yěshēng dòngwù bǎohù xiéhuì [China Wildlife
Conservation Association]. Beijing: China Forestry Publishing House.
Wong, Sik-Ling. 1939. Phonetics and phonology of the Yao language: description of the
Yau-ling dialect. Lingnan Science Journal 18.4:425–455.
Wurm, Stephen A., et al. (eds). 1988. Language atlas of China. Hong Kong: Longman.
Ying, Lin. 1972. Chinese loanwords in Miao. In Purnell (ed.), 55–81.
Zhōngguó cǎodì zīyuán tújì [Atlas of Grassland Resources of China]. 1996. Zhōngguó
kēxuéyuàn guójiā jìhuà wěiyuánhuì zìrán zīyuán zōnghé kǎochá wěiyuánhuì [The
Comprehensive Investigation Committee on Natural Resources under the National
Strategy Committee of the Chinese Academy of Sciences]. Beijing: China
Cartographic Publishing House.