(WW
.45 f‘ {If . 9 {4
Fin 1 0.787. is

£13233? iii

ABSTRACT

MARRIAGE AND ADOPTION IN CHINESE SOCIETY
WITH SPECIAL REFERENCE TO CUSTOMARY LAW

By

James Pierce McGough

Anthropological field research into the customary laws of
family and household organization in rural Taiwan brought to light
a runnber of "irregular" forms of marriage and adoption--forms not
easily categorized by traditional descriptions of Chinese society.
Uxorilocal marriage, in which the newly married couple moves to live
with the woman's family instead of the man's family is an example,
as is the institution sometimes referred to as the "adopted daughter-
in-law," in which a young girl is taken into a family and is raised
to marry one of the sons of the family.

My attempt to account for these and other, related forms led
in two main directions. One was descriptive: I collected data from
field research and historical and archival sources referring to
these customary forms as they existed in Taiwan and in other parts of
China as well, in order to gauge the range of variation in forms of
marriage and adOption in traditional Chinese society generally and

to determine whether Taiwanese society deviated from the pan-Chinese

norm.

James Pierce McGough

The second direction of investigation was theoretical:
Because serious problems arose in formulating definitions of "mar-
riage" and "adoption'I that were applicable cross-culturally, I
attempted an approach which did not assume these categories as a
basis for analysis.

I found that there was great variety of forms of both mar-
riage and adoption in traditional Chinese society, that this variety
seems to have existed in all geographical regions, not peculiarly
in Taiwan, and that it seems to have been inversely related to socio-
economic class. "Irregular" forms were more likely to be found among
urban and rural poor than among the elite. This has led to a cer-
tain bias in many accounts of traditional Chinese family and kinship
organization: an emphasis on elite norms at the expense of the more
varied reality in lower classes.

My theoretical conclusion is that it is a mistake to choose
"marriage" and "adaption" as units of study, at least for the Chinese
case. I found it much more productive to focus instead upon the
household as a social, legal, and economic unit, and then to view
marriage and adoption as more or less vaguely defined areas on a
spectrum of social forms, all of which could be utilized to recruit
persons for the household or to establish connections with other
households. When this approach is taken, it becomes evident that a
large number of these recruitment forms can be treated as variations
on a basic structure and, further, that the particular variations
can be accounted for by reference to demographic, economic, and

socio-political factors.

MARRIAGE AND ADOPTION IN CHINESE SOCIETY
WITH SPECIAL REFERENCE TO CUSTOMARY LAW

By

James Pierce McGough

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Anthropology

1976

“wins”... :3...
.._
_

 

©Copyright by

JAMES PIERCE MCGOUGH
1976

TABLE OF CONTENTS

INTRODUCTION ......................... 1

Chapter
I. THE DEFINITION OF MARRIAGE IN ANTHROPOLOGICAL THEORY . 1
II. GENERAL CHARACTERISTICS OF MARRIAGE AND FAMILY IN

TRADITIONAL CHINESE SOCIETY . ._ ........... 15

The Chinese Marriage Ceremony ............ l5

The Role of Marriage in Chinese Society ....... l7

Social and Legal Aspects of Marriage ........ 20
Conventional Analyses of the Traditional Family

System ...................... 24

Rights and Duties Within the Domestic Unit ..... 35

III. IRREGULAR MARRIAGE FORMS ............... 52

Uxorilocal Marriage ................. 52

Polyandry ...................... 84

Analytic Model for Uxorilocal and Polyandrous

Marriages ..................... 89

Double Generation Marriage ............. 94

The Levirate, Sorarate, and Cousin Marriage ..... -97

Consensual Unions .................. 107

Spirit Marriage ................... 112

Purchase and Pawn Marriage ............. 120

Homosexual Unions .................. l35

IV. FORMS OF ADOPTION ................... l62

Kuo-fang Adoption in Chinese Society ........ 165

The Adopted Daughter-in-Law" ............ l72

Adoption in Taiwan ................. 186

The "Adopted Daughter-in-Law" in Taiwan ....... l93

V. MARRIAGE AND ADOPTION IN A TAIWANESE VILLAGE ..... 204

A Brief Description of T'ien-sung-p'i ........ 204

Irregular Forms of Marriage and Adaption ...... 218

ii

Chapter Page

Case Studies .................... 222
Case I: Phoa* ................... 224
Case II: P'an ................... 227 I
Case III: M. Nu; T. Go .............. 232
Case IV: M. Yeh, T. Iap .............. 238
Case V: M. Ch'iu, T. Khu ............. 241
Case VI: M. Liao, T. Liau ............. 244
Case VII: Ch‘iu, T. Khu .............. 253
VI. ANALYSIS AND CONCLUSIONS ............... 266
APPENDICES .......................... 313
A. SELECTED CONTRACTS .................. 314
B. CASE STUDIES ..................... 351
C. GLOSSARY ....................... 371
BIBLIOGRAPHY . . . . . . . .................. 389

iii

INTRODUCTION

I conducted anthrOpological field research into customary
law in T'ien-sung-p'i Village, I-lan Country, Taiwan, from August
l970 to August 1972. One of the interesting things to emerge from
this research was the variety and prevalence of "irregular" forms
of marriage, especially uxorilocal marriage, in which residence
after the wedding is with the wife's family, and simpua_("little
daughter-in-law"), or adopted daughter-in-law marriage, in which a
female child is ad0pted into the family and raised to marry one of
the sons.

This dissertation utilizes material from field research and
'a variety of other sources to examine the nature and role of these
and other "irregular" forms of marriage and ad0ption in Chinese
society. My research began with the contemporary Chinese society of
ruralTaiwan. Some have suggested that such irregular marriage pat-
terns were and are particularly Taiwanese and were due in part to
Taiwan's history, particularly its fifty years under Japanese rule.
However, investigations into data about actual, rather than ideal,
marriage patterns throughout traditional China showed impressive
historical depth and geographical spread for such irregular forms.
The description and comparison of irregular marriage forms which I

include have benefited much from a compilation of customary laws made

early in this century, Min Shang Shih Hsi-kuan Tiao-ch'a Pao-kao Lu
(abbreviated in the text MSS).

The data from Taiwan and from mainland China led me to con-
clude that for the purposes of my analysis it was best to abandon
"marriage" and "adoption" as significant and relevant categories and
to focus instead on a spectrum of social forms that blended into one
another along a number of dimensions which emerged in the course of
the research. Uxorilocal marriage blended into regular marriage,
regular marriage blended into simpug_marriage, gimpua_marriage blended
into adoption, and so on.

The problem was to determine whether there was some basic
pattern common to all these social transactional forms and to try to
identify the demographic, economic, political, and technological cir-
cumstances in which particular forms might be found. It proved
easiest and most productive to treat family or household units as
corporate interest groups, and to focus the analysis on patterns of
group recruitment, particularly in the context of the concept of
"self-exploitation" developed in A. V. Chayanov's theory of peasant
economy.

I would like to thank the Midwest Universities Consortium
for International Activities and the Center of Asian Studies, Michi-
gan State University, for financial support of this study. The
Institute of Ethnology, Academia Sinica, Taiwan, kindly accorded me
Visiting Researcher status for 197l-l972, and I would particularly
like to thank its Director, Li Yih-yuan, for his aid. Mr. Wang

Ji-ming's skillful help made the completion of the research possible,

and the unflagging hospitality of our village friends and neighbors
made it productive and enjoyable.

Dr. Bernard Gallin, Dr. Robert McKinley, Dr. Kenneth David,
and Dr. Charles Morrison, all of the Department of Anthropology,
Michigan State University, made careful and valuable comments on the
draft of the dissertation, for which I thank them.

Mr. Tien Tsung-yao, of Michigan State University, and
Mr. Alfred Pan, of the University of Hong Kong, both helped in the
translation of difficult texts, and Mr. Austin Shu, of the Michigan
State University Library, was very helpful in settling some biblio-
graphic questions. Mr. Tien also helped in checking some of my
translations.

My heartfelt thanks to David and Marymae Klein for their
careful and instructive editorial help, and to my patient and hard-
working typist, Susan Cooley.

Finally, I thank my wife, Helen, for her crucial intellec-
tual and emotional aid and support throughout all phases of work on
this dissertation.

The following abbreviations are used in the text to identify
languages or dialects used: M. (Mandarin or kggzyfl), T. (Taiwanese),
C. (Cantonese),.J.(Japanese). The romanization of Taiwanese words
fbllows Douglas (1873) and Barclay (1923), except that an asterisk

is used instead of a superscript "n" to indicate nasalization.

CHAPTER I

THE DEFINITION OF MARRIAGE IN
ANTHROPOLOGICAL THEORY

In attempting to deal with the overlapping categories of
marriage and adoption in rural Taiwanese society and Chinese society
generally, a natural response is to hold that if it is hard, and in
some cases impossible, to distinguish between the two, then in some
sense they are the same thing. From this point of view it is impor-
tant to look at definitions of marriage, to see if there are strong
reasons to reject this argument.

Although I am not concerned here with cultural universals
as such, it is in this context that a number of general definitions
of marriage have been put forward. Although Edmund Leach says that
“marriage is a 'bundle of rights'; hence all universal definitions
of marriage are vain“ (1961: 105), Hard Goodenough disagrees, and
goes on to offer one.

I would myself define marriage, then, as a transaction

and resulting contract in which a person (male, female, cor-
porate or individual, in person or by proxy) establishes a
continuing claim to the right of sexual access to a woman--
this right having priority over rights of sexual access others
currently have or may subsequently acquire in relation to her
(except in a similar transaction) until the contract resulting
from the transaction is terminated--and in which the woman
involved is eligible to bear children (l970: 12-13).
Goodenough's formulation and the arguments behind it are

interesting for other reasons than just arriving at a universal

4

definition of marriage. In attempting to arrive at universals, he
is forced to return to basics in his characterization, to the essen-
tials, independent of particular cultural forms.

The result is that the definition of marriage is broadened
to include institutionalized unions that are not usually thought
of when "marriage" is discussed, but are known from the ethnographic
record: marriages in which a woman may marry another woman, "ghost
marriages," and so on (Goodenough 1970: 13).

Both of these latter irregular forms of marriage were found
in Chinese society ("ghost marriage" is still prevalent in some
areas, but I know of no evidence that woman-woman marriage is still
practiced), so we are on the right track. Such a definition does
not deal with T. simpua, or "little daughter-in-law," marriage, however.
In this form of marriage, which was widespread in Chinese society,

a young girl was adopted into a family and raised until puberty,
when she was married to a son of the famiTy.(Thisinstitution will
be described in greater detail later.)

There are good arguments for considering her to have become
married when the adoption took place. Her status with respect to
I such things as mourning grades, surname use, kinship terminology,
and aspects of customary law all seem to indicate this. Yet
obviously no right to sexual access, or eligibility to bear children,
was immediately created.

Goodenough's arguments for the inclusion of these last two
clauses are that

In all human societies, it is necessary to have some means
of determining when a woman is eligible for sexual relations

and when she is eligible to bear children. The timing of

these eligibilities does not always coincide. In all societies,
there is a need to know when a man is eligible to engage in
sexual relations, too, and, if relevant, when he is eligible to
beget children. And in all societies, there is a need to deter-
mine who has sexual privileges with whom and on what occasions.
Everywhere, finally, criteria of some kind are needed by which
to determine where a child belongs, in what groups he has
membership rights, and what adults are responsible for his
maintenance, socialization, education, and for protecting his
interests and enforcing his rights (l970: 8).

It is further argued that in different societies these
requirements are met by various kinds and numbers of transactions,
and that to achieve a generally applicable definition of marriage
it is necessary to find a universal concern that relates in some
way to sexual reproduction and that requires a transaction between
a man and a woman. "Sexual access is the only human concern I can
find that meets this requirement" (Goodenough 1970: ll).

This is an outcome, he says, of several human characteris-
tics. One is the tendency to form more or less continuous, affect-
laden relationships, sexual and otherwise. A second is the "uni-
versal tendency" for males to be combative and competitive when it
comes to sexual access to females. Finally "for a variety of
reasons that are yet unclear," men and women who grow up together as
siblings "tend not to establish sexual liaisons with one another"
(1970: 11). This is the so-called "Hestermarck hypothesis."

A consequence of these universal human tendencies is that
men and women develOp continuing social relationships with two
categories of persons: those with whom they are intimately
associated in childhood as members of the same domestic unit,
and those with whom they establish sexual liaisons as young
adults. There is little overlap of these categories. . . .
Because rights and obligations other than those pertaining to

sexual access can be vested in the sibling relationship, a
definition of marriage that includes other than considerations

of sex and reproduction cannot be applied universally--as
long, that is, as we assume that marriage involves a transac-
tion that links two persons in a manner they were not linked
before (1970: 12).

Arthur Wolf has written three articles attempting to show
that §jm933_marriage in Chinese society is a source of evidence
which supports the "Nestermarck hypothesis" mentioned above, that
men and women who grow up as siblings tend not to establish sexual
liaisons with one another (Wolf 1966, 1968, 1970; Goodenough
explicitly refers to the first of these articles). I do not agree
with Wolf's arguments, however.

Wolf argues that such marriages are disliked by those for
whom they are arranged, and are less successful than are other mar-
riages. He measures success by such things as divorce rate, number
of visits to prostitutes by the male partners, and numbers of chil-
dren. Although some of these may be particularly culture-bound
definitions of success in marriage, even if it is true that such
marriages are comparatively unsuccessful, it has not been demon-
strated to my satisfaction that this has anything to do with the
fact of partners having been raised together, in and of itself.

Informants in T'ien-sung-p'i who had themselves been involved
in or with such unions tended to blame any distaste felt on the part
of the male (the female was not regarded as in any position to
complain) on the fact that the intended wife occupied the lowest
position in the family social structure. From the time they were
little she was the one who did all of the dirty, distasteful jobs.

She was a drudge, and was not dressed as well as the others in the

family. As a female she was inferior to the males in the household;

as adopted she was inferior to the daughters of the family. Since
her ties with her natal family had been cut, she could not appeal to
them (or threaten to appeal to them) to bolster her own position in
the ad0ptive family and her position was inferior to that of a

woman who had been married in the regular way. The latter could
always return to her natal family if she got fed up, and the affinal
family had a greater or lesser investment in her in the form of the
brideprice. Neither of these was true for the simpua,

It is not surprising that the simpug_was sometimes indis-
tinguishable from the adopted daughter, or from the T. cha-bo-kan,
or female slave/servant. The picture was not so dark in every
instance, of course, but it was a general tendency to regard her as
occupying a distinctly inferior position. Hence, for instance, it
seems that adopted daughters tended more often to become prostitutes
than did others (until fairly recently, when adoption itself began
to decline in popularity).

For these reasons I think that the institution of gimpgg_
adoption and marriage is not evidence for a postulated "universal
tendency" for those raised together not to form sexual liaisions
with each other. Certainly if there is such a tendency it seems not
to come from being raised together, but rather from certain socio-
logical characteristics of the situation, and it is by no means a.
strong tendency. I have not done a careful count of such marriages
in the village of T'ien-sung-p'i to compare with other types of
marriage, but my intuitive feel is that they are no less "success-

ful" (in terms of divorce rate or number of children) than other

marriages.

I do not think that "number of visits to a prostitute"

in an adequate measure of "success" in this case, since in Taiwanese

society frequenting prostitutes may be as much a manifestation of

conspicuous consumption as of marital distress.

This means that Goodenough's reliance on sexual access as a

cornerstone of his definition of marriage is not warranted generally,

and certainly should not be used in this particular case. Edmund

Leach has made essentially the same point:

The institutions commonly classed as marriage are concerned
with the allocation of a number of distinguishable classes of

rights.

I 63 'I'l morn-n)

I.
J.

In particular a marriage may serve:
To establish the legal father of a woman's children.
To establish the legal mother of a man's children.
To give the husband a monopoly in the wife's sexuality.
To give the wife a monopoly in the husband's sexuality.
To give the husband partial or monopolistic rights to
the wife's domestic and other labour services.
To give the wife partial or mon0polistic rights to the
husband's labour services.
To give the husband partial or total rights over property
belonging or potentially accruing to the wife.
To give the wife partial or total rights over property
belonging (n: potentially accruing to the husband.
To establish a joint fund of pr0perty--a partnership--
for the benefit of the children of the marriage.
To establish a socially significant "relationship of
affinity" between the husband and his wife's brothers.

One might perhaps considerably extend this list, but the
point that I would make is that in no single society can mar-
riage serve to establish all these types of right simultan-
eously; nor is there any one of these rights which is invari-
ably established by marriage in every known society. We need
to recognize then that the institutions commonly described as
marriage do not all have the same legal and social concomitants
(Leach 1961: lO8-lO9).

Leach adds, with respect to item "C" above, that "I use the

term 'monOpoly' advisedly. I consider this right C is to be regarded

as a monopoly control over the disposal of the wife's sexuality

10

rather than an exclusive right to the use thereof" (Leach l96l:
107).

If marriage never serves simultaneously to create all the
above rights, and no one of them always serves to define marriage
itself, then it seems clear that we must look at these and similar
rights together in any particular society, to see just which of them
are involved in particular cases.

It should come as no surprise that we may find more than one
kind of "marriage" in a particular society. If there is no single
right which can serve to define marriage as against all other social
institutions, then we can expect a more fluid situation,in which
various institutions may take on various combinations of Leach's
rights listed above. Ad0ption, in Chinese society, does establish
a good number of those rights. It might be argued that this serves
to show that the list of rights, the characterization of marriage,
is inadequate, then, because it cannot serve to distinguish such
clearly different things as adoption and marriage. As we have seen,
though, and as I will show in greater detail later, we are in this
way getting closer to the empirical situation, in which it j§_often
hard to distinguish the two.

Rodney Needham argues that the marriage institution

. . . is defined in any particular instance by what we

divisively call the "other institutions" of the society. It
is not only jural institutions, either, that we have to take
into account, but moral and mystical ideas as well, and these
in an unpredictable and uncontrollable variety. The compari-
son of marriage in different societies needs therefore to be
contextual, and ultimately ''total" in a Maussian sense, if we

are to be sure that we understand what we are trying to com-
pare (1971: 6).

ll

P. G. Riviere carries this argument even further, leaving
jural institutions far behind. In an argument in which he seems to
confuse "functions" (in the sense of mathematical functions or
interrelations between social and/or cultural variables) with "func-
tionalism" [the explanation of the existence and functioning of
social and/or cultural phenomena in terms of their contribution to
the maintenance of some presumed system], he says that "functions"
are part of the conscious model; rationalizations of an underlying
structure.

. . . marriage must be studied as one of the possible
relationships between men and women, and . . . this relation-
ship is only meaningful in comparison with other relationships
in the same structure. . . . Marriage is . . . one of the
socially approved and recognized relationships between the
conceptual roles of male and female. This relationship will
reflect an aspect of the particular society's conventional
ideas about the two categories and it will be possible to define
it by opposing it to other possible male/female relationships
which exist in society. . . . In taking marriage as a category
of study we have concentrated on a part-relationship rather
than a whole one. What I am trying to do is to get marriage
viewed as part of the total male/female relationship (Riviere
1971: 66).

I am generally in agreement with Riviere's proposed approach,
and I have found many of his arguments interesting and his sugges-
tions useful. What I have found less persuasive, and less useful
in actual analysis, is his emphasis on marriage as "an aspect of

the particular society's conventional ideas" about the conceptual

 

roles of male and female and their relationships (emphasis supplied),
to the seeming exclusion of economic, legal, and broadly sociologi-
cal perspectives.

Abner Cohen, writing of the writers whom he calls "thought

structuralists," says that

12

It is when they lose direct reference to social interac-
tion that they become one-sided and stray from the main stream
of social anthropology. Most of them are fully aware of this
danger and almost invariably begin their different disserta-
tions with a declaration of faith in "social structuralism"
and a promise to bring their analysis of thought structure to
bear upon the dynamic intricacies of social organization. But,
as the exposition proceeds, the promised analysis is put off
until the end, when it becomes largely inconsequential.

This is in no way an indication of analytical weakness,
but is rather a matter of orientation and interest. The prob-
lems that this approach poses are not sociological problems,
but principally deal with the relations between symbolic cate-
gories. . . . These are of course very important problems for
social anthropology, but only if they are systematically analysed
within the context of power relationships (Two Dimensional Man
1974: 44-45).

Elsewhere, Cohen says of these "power relationships" that
they consist of two of the four broad institutional spheres (the
economic and the political) into which most anthropologists
(implicitly or explicitly) have divided social structure. The
other two spheres, kinship and religion, he categorizes as having
the common denominator of symbolism (1974: 22).

Cohen is interested in develOping an integrated anthropologi-
cal study of "modern" industrial society which will focus on the
dialectical relationship between power relationships and symbolic
relationships. Furthermore, he argues that what will be of basic
interest in such a study will be relationships between individuals
and groups, interest groups of all sizes and sorts, groups which
may be organized formally, along Weberian lines, or informally,
"making use of the kinship, friendship, ritual, ceremonial and
other forms of symbolic patterns and activities that are implicit
in what is known as 'style of life'" (1974: 68). Furthermore, "the

study of the structure of the informally organized interest group

13

is a key to the development of an anthropology of complex society"
(1974: 124).

I have found it very productive to approach the family
(which must, after all, be an important factor in any discussion
of marriage and adoption) as one of a number of "interest groups,
protecting or developing power for their members through informal
organisational mechanisms" (Cohen 1974: 120). The family is and
was a particular kind of interest group in an on-going, ever-changing
politico-economic field. In the Chinese case, certainly for any
time period of interest to us, this politico-economic field does and
did include the existence of a formally organized state polity.

This is a very important factor because "the develOpment of
interest groups, and the nature of the relationships between them
and the state, depends to some extent on the structure of the
state. . . . The anthropologist must deliberatelyformulate his
problems in such a way as to make reference to the state a neces-
sary part of his analysis" (Cohen 1974: 128-29).

The analysis of marriage and ad0ption, and related social
transactions, must be carried out in such a way that eXplicit
reference is made to the state and to the general politico-economic
setting. Marriage, no less than adoption, was traditionally a
transaction between interest groups, between families, and not just
individuals. Thus it is necessary to focus on power variables: on
the legal and economic characteristics of such transactions.
"Failure to take proper account of the transactional nature of

social relationships--especial1y in their jural aspects, where

14

rights and duties are involved--may well have contributed to
anthropology's conceptual problems in defining marriage, family,

and kinship" (Goodenough 1970: 22).

CHAPTER II

GENERAL CHARACTERISTICS OF MARRIAGE AND
FAMILY IN TRADITIONAL CHINESE SOCIETY

The Chinese Marriage Ceremony
Marriage, in the classical Chinese formulation, had six
stages, the so-called M. 112:113 These were "asking the name"
(M. wen ming), "betrothal" (M. ting-hun), "exchange of goods"
(M. na ts'ai), "exchange of wealth" (M. ng_pj), "setting the date"

(M. ch'ingrch'i), and "welcoming the bride" (M. ch'in ying). These

 

were not always followed, but furnished a general pattern, a guide-
line, for the wedding ceremony.

According to Kataoka Iwao, observance of the "six rites"
was a class-related phenomenon during the Japanese era in Taiwan.
This seems to have been true for the rest of China as well. The
upper classes attempted to follow the traditional six rites, while
the lower classes' marriage forms were, he says, very disorganized,
including even marriage by sale and temporary marriage of one's own
wife to another man (Kataoka 1921: 20). This is described in Chapter
III under the term "pawn marriage."

This corresponds roughly to the distinction made in Taiwan
between "major" and "minor" marriage (T. toa chhoa and sio chhoa,

res pectivel y). The former consists of "regular" marriages of the

P’i'r‘ilocal type, in which an adult woman marries into, and moves

15

16

into, the husband's family. The second category, of minor marriages,
includes various forms of uxorilocal marriage and adoption-marriage
(Kataoka 1921: 20).

In major marriage the stage of discussing the potential
match includes the exchange of the horosc0pes of the two individ-
uals. Inspection and comparison of the horoscopes, often done by
a specialist, can offer an avenue for honorable withdrawal from the
negotiations should there be the need for one. At the start of the
negotiations it is usually the man's family that is approached by a
go-between who has been engaged by the woman's family.

"Asking the name" is sometimes also called "discussing the
marriage" (M. j_hun). According to a traditional saying, there are
five items of importance in the discussion of a potential marriage
match. These were the reputation of the family, the matter of
money, talent and ability, appearance, and health (Wu 1970: 125).

In the next stage, "betrothal," there is an exchange of
gifts, usually including sweet cakes (T. lg_piaf,"ritual cakes")
and other articles, often symbolic of fertility and the birth of
many sons. These are sent to the woman's family by the man's family,
and then they are divided up and redistributed to the relatives as
announcement of the impending marriage. The relatives who receive
such cakes then send gifts that are intended to make up the dowry.

The next stage is the further exchange of gifts, and includes
the old stages or rites of exchange of goods and exchange of money.

‘4 procession accompanied by the go-between conveys further presents

and the marriage contract (T. hun-su) from the man's house to the

17

woman's. The presents again include cakes and other symbolic gifts,
but this time betrothal money as well. Some of the gifts are kept,
but some are used to make up return gifts, sent back to the man's
family. Often, also, a portion of these goods is kept back by the
woman's family and later used to inflate the dowry, so as to be in a
better position when negotiating the betrothal gifts they are to
receive from the man's family. Sometimes, because of the trouble
and expense involved in this exchange of gifts, they are all con-
verted into cash. This is sometimes called T. ta*-poa* ("convert
to cash"), and it is said that if the two sides do not have close
contacts, live far apart, or if they are quite different in socio-
economic status, then in this way though one side loses face, both
profit (Wu 1970: 128).

"Asking the date," selecting an auspicious date and time for

the welcoming of the bride, is colloquially called T. sang jit-thau-a

 

("sending the horoscope"). The groom's family sends, through the
go-between, a paper (jjt-thau-a) on which is written the time for
the agreement of the bride's family. When the time arrives, the
groom or the go-between (customs differ) goes to the bride's home

and accompanies her back to the groom's home.

The Role of Marriage_in Chinese Society
As Arthur Wolf puts it, "three things happen in a marriage
of the patrilocal type [that is, virilocal type]. The bride leaves
her natal home and relinquishes membership in her family of orien-
tat:ion; she steps over the threshold of the groom's home and becomes

a~'lnember of his household; and she is presented to the groom's

18

ancestors and thereby acquires the status of wife" (Wolf 1966:

883). This last is colloquially called T. chhut-thia* ("to come

 

out to the hall"), and refers to the emergence of the bride from
her chamber on the third day after marriage to worship the gods and
ancestral tablets, and to be introduced to her husband's relations.
Inviting her to come out was supposed to be done by a little girl,

and was called T. chhia* chhut-tia* ("invitation to come out to the

 

hall"). The bride paid homage as well to the parents-in-law, and
then afterwards repaired to the kitchen to cook a chicken and begin
her service to the family.

Marriage, or marriage-like social transactions, was of course
/"'"' * L. /——J __fi_____,,_ _ .ra—‘wv -7~. _._._._

 

Mgr-*—

 

 

of great importancefljn_tr§gitiopal Chinese society. Major marriage
,_—~L_,/’* 2-“.

was at the center of the development of one's status with respect
to family and kinship. It was, of course, more a matter of concern
for the family than the individual. This was in part because of
the religious connotation that marriage carried. It was incumbent
upon the younger generation to provide offspring to continue the
sacrifices to the ancestors. These offspring had to be males, not
females, as only the former were thought to be capable of offering
effective sacrifices.

This latter distinction is important because it is related
to the fact that (as I argue) all social transactions and relation-
ships in Chinese society were traditionally, and still are to a
large degree, between parties of unequal standing. One of the

basic and most important sources of this inequality is the social

distinction of sex roles. Women, of course, played an absolutely

19

indispensable role in the social system, but it was a role which
was distinctly different from that of the males, and distinctly
inferior. The rights and duties of the children in a family dif-
fered sharply depending on sex, and the rights and duties of husband
and wife were likewise differentiated.

The major marriage system, the officially approved one, was
in theory monogamous, but was in fact often polygynous. That is,
even under the Ch'ing code it was a crime to have more than one

principal wife (M. ch'i or chenggch'i), but secondary wives, or

 

concubines (M. ghljgh) were permitted. Ostensibly these were taken
only when the principal wife proved barren, but this rule was not
much followed, and in practice a man could take as many concubines
as he wanted or could afford. The concubine had a definite and
legally protected position in the household and family, but one
which was distinctly inferior to that of the principal wife. Her
children by her "husband" were not considered illegitimate, but
their rights in the family estate were inferior to those of the
children of the principal wife.

If we approach marriage as a kind of social transaction,
then we must pay attention to the kinds of rights, duties, and
other items exchanged in the transaction. Robin Fox has pointed out
that there is a "universal tendency" to male dominance over females
and children (Fox 1967). This gives men an interest in the con-
trol of the sexuality and reproductive powers of women and of the '

labor that both women and children can provide.

20

Proprietary rights in the sexuality and labor of a woman
and her children may be vested in the husband and father or in
the brother and maternal uncle. The allocation of these rights
also relates to the distinction between the conjugal and con-
sanguine forms of the family. But like other forms of property,
they can be subject to inheritance, sale, gift, loan, rental--
whatever kinds of transactions and transfers of title a society's
rules of property allow. . . . The widespread practice of vari-
ous forms of adoption and fosterage indicate how common are
transactions in the rights of parenthood (Goodenough 1970: 22).
Major marriage, in Chinese society, was accomplished through
a transaction between two corporate groups, families, which resulted
in a contract (usually written) in which the man's family acquired
a continuing claim to the right of control over the sexuality of

the woman.

Social and Legal Aspects of Marriage

The effect of marriage in the traditional social and legal
system was important in two main areas. These were the socio-legal
status of the individual and property law. In socio-legal status,
the most important aspect was the establishment of the husband-wife
relationship (Cf. Wolf 1966: 883). This, of course, does not
consist solely in the relationship between two individuals, but
rather in the creation of a constellation of family and kinship
relationships between the two families concerned.

As far as the husband-wife relationship itself, the two were
considered to form a single unit upon marriage, but this did not
inply that there existed a relationship of equality between the two.
The woman took her basic social position from that of the husband,
and had no independent existence; she was clearly submissive in her

reVlationship to her husband (Tai 1966: 233; Taiwan Min-Shih Hsi-Kuan

21

Tiao-ch'a Pao-Kao [hereafter IMSJ: 85). In law, the relationship

 

between husband and wife was in many ways comparable to that between
genealogically senior and junior. In terms of mourning duties, that
of the wife to the husband was the heaviest degree, of coarse sack-

cloth and three years (M. chan-ts'ui). That of the husband to the

 

wife was one degree less, coarse hempen fabric for one year
(M. tzu-ts'ui).

The existence of this clearly unequal relationship did not
mean, of course, that there were no mutual obligations; though a
hierarchical relationship, it was not completely one-sided. Mutual
obligations, followed and enforced in varying degress, were those
of mutual support, cohabitation, and maintenance of sexual fidelity.
Mutual support was not defined in the Ch'ing Code, but according to
customary law, the husband was bound to support the wife and the
concubine as well; if he was unable to do so, the wife then had a
duty to support him (IMS: 850). As we will see later, there were
at times and in certain places socially recognized forms of "simul-
taneous" or "serial" polyandry that arose from this duty.

The duty of cohabitation was emphasized in the code, and of
course a runaway wife was subject to punishment (INS; 85). In
customary law this duty of cohabitation was often applied to the
husband as well, and the absence of the husband, or letters from
him, constituted, after a certain number of years, grounds for divorce
and permission to remarry.

As for sexual fidelity, this of course was perhaps the area

0f greatest inequality between husband and wife. Under the

22

traditional law, extramarital relations were considered adultery,
but (unsurprisingly) a wife's adultery was considered more serious
than that of the husband. Moreover, while a husband could take a
concubine, both in the formal and customary law, the wife had no
right to form any kind of sexual relationship with any other man.
We will see, however, that the husband, exercising his right to the
monopoly over the wife's sexuality, could enter her into a poly-
androus relationship with another man. This usually took place in
situations in which the man could not support the family.

The husband had the right to issue orders to the wife, and
she had the duty of obeying, but this right was not without limit.
The husband's ill treatment or punishment of the wife was not
actionable in court, no matter how arbitrary or unreasonable it
might have been, unless it went as far as serious injury or death.
On the other hand, in the event of injury or death caused to the
husband by the wife, the punishment usually meted out for such
offenses was increased because the wife had additionally violated
the socially and legally defined relationship between herself and
the husband. This was also true when a junior or inferior assaulted
a senior, or superior.

The husband was forbidden by the Code to mortgage or sell
lfis wife or concubine, but in fact this often happened, usually in
cases of poverty (INS; 33; also see cases described below).

By virtue of the marriage, the wife took on the husband's
socio-legal position to a certain degree. That is, the husband's

relatives, both consanguine and affinal, became the wife's relatives

23

as well. But while the wife's mourning grade with respect to her
parents-in-law was the same as her husband's her duties with respect
to his other ascendants were one degree lower than his (INS: 86).
Although the wife in some respects took on the status of a member
of the husband's kin group, she was in an ambiguous situation. She
maintained her relations to her own kin group, and she maintained
her own surname, putting the husband's surname before it. In other
words,she was not entirely absorbed by the husband's kin group. On
the other hand, her position in her natal family was not unaltered.
Her mourning duties to her natal family members were all reduced by
one degree, as were theirs to her (IMS: 86).

In this context of the constellation of family and kinship
relationships created between the two families by marriage, the
subordinate position of the woman is again demonstrated. A rela-
tionship was created between the husband and other kinsmen of the
wife's family, but the scope of these relationships was not the same
as that between the wife and the other kinsmen of the husband's
family; it was limited to the circle of mourning relatives (INS; 86).

As far as property rights were considered, the husband and
wife were thought, by some, to have comnon pr0perty in the sense
that she had no personal property except for items of personal use
that came with her dowry (clothing, utensils, etc.); other items
of the dowry, including money, land, houses, etc., were all con-
sidered to be the common property of the wife and the husband, and

hence part of the comnon family estate (INS: 89).

24

Conventional Analyses of the Traditional
Family‘System

 

It seems appropriate here to give a brief description and
analysis of the customary laws of the traditional family system,
its composition and structure, and the rights and duties incumbent
upon the various members by virtue of their status as members.

The family in Ch'ing China could be defined, from a customary law
point of view, as "a coresident organization of kinsmen with the
intention of living a common life" (IMS: 219). A family had a
family head, and the rest of the family members were family depen-
dents. The family, or M. ghjg, was called a household, M. fig,

in terms of public law.

The head of the family held his position as such by virtue of
being a male elder, but there were also cases of those of proven
ability perfbrming the role regardless of generational or age stand-
ing. If there were no male to fulfill the role, then a female elder
could act as head. The family head represented the family exter-
nally, in the realm of public law. Internally, in private law, he
was the leader of the family. Not only in matters of socioeconomic
status, but also in matters of pr0perty, the family members were
bound by his decisions.

Since the family was seen as a coresident organization of
kinsmen with the intention of living a common life, commensality and
common property were the rule, and religiously the family was
centered around ancestor worship. Thus, according to Tai Yen-hui

(7963: 2), the comnon life of the family was expressed through the

25

commonality of residence (M. 93g), pr0perty (M. tsiai), and fire
(M. hgg), the latter symbolizing the cooking stove and sacrificial
fires. Thus one comes across such terms as smoke (M. ygg), resi-
dence (M. Egg, Egg), cooking fire (M. tglgag), pr0perty (M. tsigj),
household utensils (M. chia-huo), sacrificial vessel (M. 3139),

and so on, all prefixed with "common" (M. kung, t'ung, chung) to

 

 

indicate a still undivided family, and prefixed with "separate,"
"divided," or "different" (M. fgg, hsj, j, pieh) to indicate a
divided one.

The constitutent members of the family sometimes included
members other than kinsmen, such as slaves, servants, or hired
laborers. Tai Yen-hui (1963: 3) has found it convenient to divide
the family members into basic family members (M. chi-pen ch'eng:,
yfign) and secondary members (M. fu-th ch'eng-yUan), the latter
including such persons as slaves, servants, etc. (In Taiwan of the

Ch'ing Dynasty, the T. cha-bo-kan, indentured servant, was such a

 

"secondary" family member.) Those coresident persons of freeborn
status (M. liang-jen, I'virtuous pe0ple," not slaves or servants)
were regarded as the basic members, and in ordinary usage this is
what was meant by the term "family members."

The term "family member" usually meant a coresident kinsman,
but coresident family members of different kin lines (such as
different-surname adopted children, or uxorilocal husbands), though
they had different rights vis-a-vis the family pr0perty, were still

regarded as basic family members (Tai 1963: 3).

26

This indicates a certain problem in defining the family and
family members. INS, following Tai, says of the Ch'ing family
structure that

"Family members," . . . refers to coresident persons with
mourning obligations. Coresidence is an essential factor in
the definition of family member, and a family was usually
made up of same-surname consanguineal relatives with mourn-
ing duties. But those of different surnames, relatives with-
out mourning duties, or those non-patrilineal relatives with
mourning duties (as, for example, a son-in-law), all also can
be regarded as family dependents (IMS: 182).

Historically Chinese law and custom has been male-centered
and patriarchal in its bias. Kinsmen were divided into agnatic

consanguines (M. nei-ch'in, "internal kin," or tsung ch'in,

 

descendants of the same male ancestor), consanguinal kinsmen con-
nected through females (M. wai-ch'in, "external kin," including
those related through the mother, father's sister, daughter, and

so on), and ”wife's relatives," thatisg those created through one's
marriage (in pre-Ming times this last category was not distinguished
from the previous one, relatives through females in general. INS;
9). Although law and custom basically gave more weight and impor-
tance to the agnatic kin, there was still considerable importance
given to the "maternal relatives" (M. wai-ch'in), especially in the

criminal law.

According to some analyses of the traditional social system,
there were two main types of kinship relationships: those arising
from a "natural state of affairs" (TMS: 9), such as marraige or
birth, and those created through legal fiction, such as adoption,
etc. Apart from this there was also the creation of a kind of

kinship relationship through a moral duty of support, as between

27

the concubine's son and his father's primary wife, or between a
child and his/her step-parent (TMS: 10).

Birth was regarded as the only true source of "natural con-
sanguine kinship" (TMS: 10). The parent-child relationship was the
basis of the kinship system, and such relationships as grandparent-
grandchild, uncle-nephew, and so on were regarded as extensions of
the parent-child relationship.

In general there were at least two different kinds of
parent-child relationships: that born of a socially and legally
approved marriage, and that not so born. Each of these could again
be divided into at least two types. Of the legitimate relationships,
there were children born of the primary wife, M. tj;tzg, and children
born of the secondary wife, or concubine, M. shu-tzu. Of the
illegitimate relationships, there was that created when a child was
born of fornication or adultery (M. chien-sheng) and that created
when a child was born of a domestic slave (M. pei-sheng).

In most circumstances a child had a kinship relationship
with his mother, his father, and with their own consanguineal kin,
but there were exceptions to this rule. There were two different
situations for those children born out of wedlock (M. chien-sheng).
In the one case, when the children were born of a legally ineffec-
tive marriage, then they followed the father's line. If, on the
other hand, the father did not recognize them, and did not raise
them, then they followed the mother's line. When the child was
assigned to one side, then all kinship relations with the other were

severed. Thus the common saying "Follow the man, sever relations

28

with the woman. Follow the woman, sever relations with the man"

(M. kuei nan, tuan nU. kuei nU, tuan nan). In the other case,

 

 

children born of a domestic servant, they also were regarded as
illegitimate, born out of wedlock, and their status was lower than
that of one born in wedlock, but if the father recognized the child,
then in law this created a parent-child relationship, but no kin-
ship relationship was formed with the mother's relatives (TMS: 12).
Half-siblings, children of the same mother but different

fathers, were not considered to have mutual mourning relations in

 

the customary law, though according to the Chu-tzu Chia Li

specified a mourning grade of M. hsiao-kung with a mourning period

 

of five months (TMS: 12). Thus,apart from a prohibition on inter-
marriage, they were not seen in customary law as having any kinship
relations. During the Japanese period in Taiwan, if they lived
together, then they were regarded as relatives. If not, then as
strangers. According to Toshio Ikeda (1944: 163), a saying was
current in Taiwan during the Japanese period that "those of the
same father but different mothers are true brothers; those of the
same mother but different fathers are 'dog brothers.'"

Logically, those consanguineal relationships that can in
principle be created only through birth can be terminated only
through death; only a natural function can terminate a relationship
that is the product of a natural function.

Traditionally, however, there were at least two exceptions
to this rule. The first of these concerns the relationship between

a child and its mother after divorce or remarriage. A woman who

29

has been divorced by her husband, or who has remarried after his
death, has extinguished her rights and duties with respect to him
and his kin group, but her relationships to her children by him were
not, under the traditional law, thereby severed. They were cur-
tailed, however, and limited to the persons involved, and did not
extend to other relatives, except insofar as an affinal relationship
was created between a son's wife and his mother, whether diVorced
or remarried. This is because the son takes on a single identity
with his wife (they form one unit). As for a woman who has married
an uxorilocal husband into the first husband's family, it was held
in the Japanese period in Taiwan that she maintains her mother-child
relationship with her children by the former husband until she
follows the second husband out of the family (M. ch'u-she, "leave
the house"), whereupon the relationship is extinguished (TMS: 13).
The second exception to the rule that only death terminates
aconsanguineal relationship is that of alienation through sale
(M. mai-tuan). A child who is "sold" (mai-tuan, "sold and cut off")
thereby loses any kinship relationship with his/her natal family.
If, however, the child is bought back ("redeemed," M. shu-hui),
then the original kinship status is restored. In Taiwan of the
Japanese era, concubines and female slaves were usually purchased
(M. mai-tuan), and among adoptions, the M. ming-ling-tzu was usually
purchased. Once purchased, the kinship relation (yf such a child
to his natal family was completely severed (TMS: l4).
Speaking of rights in land and buildings, Santaro

Okamatsu says:

30 L,

There are two kinds of alienation of immovable right in
Formosa as in other countries; in one of them, the person
entitled transfers all of his right to another (translative
acquisition); and in the other, the person entitled creates
a new right based on his own right; or, in other words, the
possessor of a right constitutes a new right which restricts
his own right (constitutive acquisition) (1901: 30).

As for translative acquisition, I'This term applies to cases
where a person entitled transfers his right to another person, that
is, when the subject of the right is changed. In the legal use of
the term in Formosa, if the term M. maj_(buying) or M. maj_(selling)
is used, it refers always to this kind ofalienation"(l901: 31).
Included are not only the rights enumerated above, but also rights
in persons. The "purchase" of children and women in Taiwan and
other parts of China during the Ch'ing Dynasty was one example of
this translative acquisition. It was an indication that the trans-
action was one of translative acquisition and not one of constitutive
acquisition. The transfer was of an absolute and irrevocable nature
(unless and until, of course, it was superseded by another such
transfer), and basically did not differ in form and function from one
involving, say, land. Thus there was the common saying "After you'Ve
sold a child, you can't even call out his name. After you've sold a
field, you can't even have access to its frontage" (M. mai tzu,
wu chiao ming, mai t'ien,vntgyu t'ien-t'ou hsing, INS; 32).

Fictive consanguineal relationships are those that exist
between persons who do not, in fact, have any biological relation-
ship, but are recognized in law as having an equivalent relation-

ship. In the customary legal system of the Ch'ing Dynasty, there

were very many different types of such fictive kin relationships.

31

For instance, there was the relationship between an adopted,
instituted heir, and his adoptive parents, created through the
institution of an heir through ad0ption (M. chi-szu), those rela-
tionships created between parent and child through the various other
fonms of adoption, and a host of relationships created through moral

duty (M. en-i ming-fen), such as that between step-parents and

 

step-children, primary wife and concubine's children, concubine

and primary wife's children, a concubine and the children of another
concubine, foster mother and foster children (M. tz'u-mu), a wet-nurse
and the children she has nursed, ad0ptive mother and ad0pted child,
and so on.

In the traditional socio-legal system, the patrilineal line
of descent was paramount, and if an adult male were without off-
spring, he could take an agnatically related kinsman of the pr0per
generation as his heir. This was called "instituting an heir"

(M. chi-tszu, or liz§52)- The one so instituted was called

M. gfli;£§zy_or szu-tzu, and was considered to have a relationship
with the senior party that was the same as a biological one, and

it was not limited to the two parties to the transaction themselves,
but extended to the other kin as well, as though they were really
parent and child. The relationship was an artificial one, though,
and so could be terminated by disinheritance or the return of the
child to his natal home (INS; 15).

In the traditional legal system (but not necessarily the
customary system), adoption without the goal of continuing the line

could be divided into adoptions within the kin group, or ad0ptions

32

of different—surname children. This kind of adoption was held to
create only an adoption relationship, not one of fictive kinship,
and, moreover, a relationship that did not extend beyond the imme-
diate parties to the adoptibn: no relationship was created with the
relatives of either party.

In Taiwan, at least, and one suspects that this was true of
China generally, there was some variance between the traditional
law and the customs actually followed. Ad0ption was generally
divided into two kinds: M. ku0-fang and M. ming-ling3tzu adoption.
The former, "transfer to another branch," in principle meant an
ad0ption within a kinship line, but it also included adoption within
a surname, since colloquially people of the same surname were recog-
nized as of the same line. In most cases a kuo-fang adoption did
not extinguish relationships with the child's natal family (TMS: 33).

Ming-ling-tzu ad0ption indicates an adoption in which a

 

child of another surname is ad0pted, and thereupon the child's
relations with the natal family were ended. Thus the ming-liggrtzu
were for the most part "sold" (M. mai-tuan). Adoption of abandoned
children is included within this category.

In Taiwanese custom, the relationships between the adoptive
parents and adopted children, whether it was kuo-fang or ming-lingftzu
adoption, were the same as those between biological parents and
children. The explanation for this discrepancy with the Ch'ing Code
that is offered in IM§_is that

In Taiwan, because the objective conditions at the time

of colonization were not suitable, the system of lineage organ-
ization slowly fell into disuse, and so the prohibition on

33

inter-surname adoption was not strictly followed, until it
reached the point that the status of an adopted child was
not at all changed by the fact of its being an inter-surname
adoption (TMS: 34, n. 7).

Between a step-father and the children of his wife by a
former husband, there existed in the Ch'ing law a quasi-father-
child relationship, one that was restricted to the two parties them-
selves, and did not extend to the relatives of either of them. This
was one of the relations created through moral obligation, and it
was dissolved if the step-father and the mother of the children
divorced, or if she remarried. During the Japanese period in Taiwan
this generally remained so, except that a condition for the exis-
tence of the relationship came to be that the step-father be
actually coresident with the mother and children. If he moved to
live elsewhere, then the relationship was ended.

Between step-mother and step-children a quasi-parent-child
relationship was created, one that was not restricted to the parties
themselves, but extended to the relatives of the respective parties,
just as if it were a biological mother-child relationship. This
also was a relationship created through moral obligation (M. mjgg;_
fgg), and it ended through divorce, remarriage, or death (TMS: 34).

There existed a quasi-parent-child relationship between the
primary wife in a family and the children of a concubine, a rela-
tionship not limited to the parties themselves, but extending to
the relatives of each. This relationship was created through moral
obligation (M. ming-fen , and was ended if she divorced, died, or
remarried. This remained so under the Japanese, but in practice

if the primary wife married in an uxorilocal husband after the death

34

of the first husband, this was not regarded as the same as regular
marriage, and her kinship relations continued as before.

Between the concubine and the children of either the primary
wife or another concubine there also existed a quasi-parent-child
relationship. It too was based on moral obligation, and was ended
by divorce or remarriage. In Taiwan of the Japanese period, the
concubine was called M. shg;mg_whether or not she herself had chil-
dren. The mourning duties that the other children in the family
had toward her were determined by the father (INS; 36).

The husband-wife relationship was the basic source of the
kinship relationship. The wife, through marriage, obtained a status
as a member of the husband's kin group. This in a sense meant that
the husband-wife relationship was basically incestuous, a suggestion
that is borne out by the fact that sexual relations and marriage

between a man and his brother's widow were strictly prohibited in
(:eertain segments of society (though nonetheless a practice in some
other segments, as we will see; cf. Riviere 1971: 73, n. 13).

In the traditional law, the wife had kinship relationships
With the husband's kin, that is, the husband's consanguines and
their wives. The husband was also thought to have a kinship rela—
t"‘i0lr‘tship to the wife's kin, but because of the patrilineal bias, the
"91 ationship here was narrower. and weaker than in the case of the
WT ‘Fe. The husband's relationships extended only to the wife's
consanguines, and not to their wives (T_MS_: 38).

The status of the concubine was definitely lower than that

(TV 'the primary wife, but still she was in law admitted the statusof

35

secondary wife. Not only was she therefore regarded as a relative
of the husband, but she was also a quasi-member of his kin group

(TMS: 39).

Rights and Duties Within the Domestic Unit

 

There are several different kinds of rights which must be

differentiated, particularly parental rights (M. ch'in-ch'Uan),

 

elders' rights (M. tsun-chang ch'Uan), and family head rights

 

(M. chia-chang ch'Uan). It is important to separate these in

 

analysis, though the situation is confused since in many, if not

most, cases they tended to all repose in the same individual.

Moreover, it is argued that "in Chinese traditional law, except for

elders' rights, it seems that there are no other kinship rights
Specifically basedupon the parent-child relationship" (TMS: 178).
111at is, parental rights are simply a subset of the more general
e1 ders' rights, and are not in nature different from them.

' Within one family there was often more than one elder; the
(zéitegory of "elder" was a relative one. For instance, the parents
<=<3lmtainly had elders' rights (or "parental rights") over their own
<3"I‘ildren, but if there were also grandparents living in the family,

t:'1(3l1 the parents were expected to follow their orders, and thus
1”Nair rights over their own children were to a certain degree
r‘estricted. Among the elders in general, there were distinctions
between lineal elders and collateral elders, and a division between

“‘31 e and female elders. Great-grandparents, grandparents, and

Parents of course qualified as elders, but so did uncles and aunts, ,

01der brothers, and older sisters. The rights of elders were not

\ .
\
l

36

absolute, but varied, with respect to the particular relationship
between senior and junior. The status of elder was further extended
to those senior "relatives through moral obligation" such as adop-
tive parents, step-parents, etc. But in law at least, the concu-'
bine was not seen as an elder vis-a-vis any of the family head's
children except for her own (INS; 158).
The order of precedence among the various relatives who
qualified as elders was not specified in the Ch'ing Code (INS: 180).
Thus it seems that the situation might have arisen in which more
than one elder, in acting within the limits of their rights, might
find themselves in conflict over the same junior relative. But in
practice there was an order of precedence that was based on genea-
logical distance and generation. The father first exercised parental
(elder's) rights. In his absence the grandparents did so, and in
their absence, great-grandparents. If there were no lineal elders,
then the closest collateral elders took over the rights, subject to
1:!1e approval of the eldest among them (just as the father's rights
were in theory subject to the grandparents' approval).

Under the Ch'ing system, all junior persons coresident in a
family, whether or not adult, whether or not they themselves were
paY‘ents, were subject to the elder's rights. These elder's rights
‘3S311'Id not, however, be exercised over juniors living apart, and with
SeDarate property (cf. Chuang 1972).

The most succinct and most accessible source of information

0“ the rights of the family head and the elders, from the point of

37

of view of customary law,is Tai Yen-hui's article on the Taiwanese
family system and the family estate during the Ch'ing Dynasty
(Tai 1963).

The family head was the leader of the family; he represented
the family in public law and was responsible to the government for
the reporting of census information, for payment of taxes, for the
fulfillment of pao-chia duties, etc. The post of family head was
in principle filled by the eldest of the family, with males
taking precedence over females. If a family had a capable male
member, then a woman could not in principle serve as head.

What must be noted (as was pointed out above) is that in the
traditional family law the elder had great authority over those

.younger than himself. This was the so-called elders' rights, and
‘they were differentiated according to whether the elder was a lineal

(elder or a collateral elder. In the latter case, the elders' rights

then were differentiated on the basis of genealogical distance.

'T11e elders' rights over the juniors, moreover, did not depend on
(:C)residence. In the case in which all were coresident, the family
head had rights as such over the family members, and since he
(Usually) was also the elder of the family, he had elder's rights
Over them as well. Thus one can, in analysis at least, differen-
tiate between his rights as family head and his rights as elder.
:I" ‘this more strict use of the term, the rights and powers of the
family head over the family members derive from the fact that the
family, as an organization, needs some sort of chain of command,

Ssomeone in authority to make decisions, decisions affecting the

38

family as a concrete, coresident group. This is why the family
head's rights and powers extend only to those actually coresident
(in some sense of the term) in the family. The elders' rights,
however, are based on the deference and obedience expected on the
part of juniors in their relationships to the elders.) It is based
more in the ethical and moral system, and thus is not, in principle,
limited by coresidence. But this distinction between different
kinds of rights, between rights based in slightly different socio-
political realms, is complicated by the fact that in most cases the
family head and the elder was the same person, in which case in
actual practice it was not possible to separate the different kinds
of rights. Internally, domestically, the authority system was
dualistic, and so the strength of the rights of the family head
depended on his particular genealogical connection with the family
members. When the family head was also a lineal elder, then his two
ssc>rts of rights coalesced, and he had great, if not absolute, powers
c>\/er the family members. There were other situations, however, in
vvriich the dualistic nature of the rights became evident. If for
W"Ié'lizever reason someone junior became family head, then his powers
'3 n general were limited by the elders' rights held by senior members 1‘
C’T= ‘the family. Also, when the family head happened to be a col-
1 ateral elder, his rights were more limited than those of a lineal
e‘Ider.
The content of the family head rights must be divided into

tho se having to do with the general management of the family life

39

and those having to do with the management and disposal of the
family estate. We treat the first division here.

A family is a group of family members living a common,
coresident life. The family head had the power to direct affairs hav-
ing to do with the family property, expenditures, and daily life, and
the family members had the duty of obeying his directions.

1 His most important function, as family head, was, in the
traditional view, to manage the allocation of duties, ensuring that
the men farmed, or did other kinds of external business, and that
the women busied themselves with domestic tasks. If quarrels arose
within the family, then he was to arbitrate (either by his personal
decision or through expressing the consensus of the important family
.members). If a family member violated a state law or a family regu-
lation, then he had the responsibility of punishing the violation
according to the law. But, as has been pointed out, if the family
head was not the eldest, then he had to abide by the decisions
()1: his senior, and could not criticize or admonish him. In this
7‘635pect, his powers as family head were limited. The head had the
F3<3vver, however, of admonishing and punishing other family members
(Junior to him), and if such punishment did not go so far as serious
'i'1;iler, he could not be taken to court. Of course, the family head
a1 80 had the responsibility of supporting the family members, and
"EiC1 to arrange for the marriages of those in the family. It is most
'Fuilt1ting, however (in Tai's view), to view his powers of disposal

0“: family members through marriage, indenture. and ad0ption.

40

as being based in his rights as elder and not as family head. If a

family member has another elder who is genealogically closer, then
that elder's rights take precedence in this matter (but then one

must deal with the fact that the family head had the duty of arrang-

ing the marriage of a slave).
In Chinese history the social-ethical system has long given

precedence to those senior in either generation or age (with genera-

tion taking precedence over relative age). In fact, one could

simply treat generation as a more elaborated and more indirect
expression of age, that is, birth order of the founders of particu-

lar branches of a kin line. The relationship between older and

younger (whether phrased in terms of strict age or in terms of gen-

eration) is essentially one of authority/submission. The status of

(elder is rooted in the kinship system and so is independent of the
particular geographical location of the individuals, but of course
<:(>residence is an important condition for the actual practice of

1:fiose rights.
The important point here is that this authority system,

baSed as it is ultimately upon age, is inherently a relativistic
(37162- Structurally, at least, there is always another elder more

SeI’I'ior to each elder, and another person more junior than each

4i‘1l1‘ior. Thus though parents exercise elders' rights over their own
(:rli'ldren, these rights are subject to the rights that their parents

‘1EI\IEE over them, and so on. It is for this reason that Tai concludes

(Tai 1963: 6) that China has never had, in the history of its legal

41

system, a system of "pure parental rights," no system that approaches
the modern-law system of rights of guardianship.

The rights of the elder are quite general and abstract in
nature. They are based in the general socio-political and moral
system, on the obedience and deference that the juniors are supposed
to display toward them. They are, moreover, based on the relative
hierarchy of age and generation, and so one must pay attention to the
paticular situation in each case, to just what kind of elder the
senior person is, and the particular relationship with the family
members, before one can determine what the content of the rights is
likely to be. In general, however, the types of rights enjoyed by
the elder are as follows:

1. General authority.

The elder, as has been noted, has a general authority of
<:cmnmnd over his juniors, and this includes the authority to admonish
<3!“ punish. The elder could not cause serious injury or death to
the junior without reason, but it was no abuse of his rights to beat

the junior as an admonition. But if this led to death, it was
treated as a crime.

2. The power of expulsion.

The expulsion of children from the family is not mentioned

‘i'T ‘the Ch'ing Code, but was a common practice among the pe0ple

(Ta-i 1963: 7). Many cases of parents expelling their own, or adopted,
‘ZTli'ldren can be found for Ch'ing Taiwan. Usually, at the time of
exDulsion, some way of public announcement was found; sometimes this

was to file a case with the magistrate, a case called a "burial

42

case" (M. mai-ti an). After the expulsion, all family and kinship‘
relations were completely severed, though they might be restored
later if the source of friction was removed.

3. 'The power to dispose of juniors in marriage.

According to the Ch'ing Code, marriages should be managed
by the grandparents or the parents. If neither the grandparents
nor the parents were alive, then marriages should be managed by
some other relative. If a woman's husband died, and her daughter
was to be married, then she (the mother) could manage the marriage
(Tai 1963: 7). The absolute right to dispose in marriage, however,
was limited to grandparents, parents, and M. chi-ch'in elders
(that is, elders toward whom the person concerned had a mourning
obligation of at least one year). In Taiwan under the Ch'ing
Ilynasty, the marriage was usually managed by the father or the

household head; in their absence, then by some other male elder,
such as grandparent or uncle, etc. A woman, even if she was the
mother, could not act as the manager of the marriage. This, however,
was just a formal requirement. In practice, of course, a woman had
the main say in the marriage of her daughter.

4. The right to adopt out junior family members.

At the time of ad0ption, the person being adopted out was
L‘sually not an adult, and so the ad0ption contract was executed by
‘1'1 (elder, whether grandparent, father, uncle, or older brother. If
'tjlfié adopter was dead (as, for example, in the institution of post-
hLm‘lous heirs), then his wife, parents, or brothers executed the

acklption contract. Ad0ption sometimes had the character of sale and

43

purchase, that is, a transaction involving “translative acquisi-
tion." This was in principle prohibited by the Ch'ing Code.
According to the commentary to the Code, however, such customs were
permitted in cases of extreme hardship. The Code prohibited the
hiring out of children as indentured servants, but once again the
Commentary to the Code said that "among poor people today it is a
very common occurrence to hire out one's children to serve others,
and this does not fall within the prohibition" (Tai 1963: 7).

As far as this custom was concerned in Taiwan, it was very
common to mortgage (M. tien-mai) or sell into indentured service
one's children, whether or not the child agreed, whether or not the
family was very poor, and whether the child was to be made a wife,
concubine, slave, or adopted child (Tai 1963: 7).

There has been much argument and discussion over the nature
of the traditional Chinese family, and given the pervasiveness of
the ideological bias of the ruling class in the traditional society,
it is well to try to maintain a critical attitude toward accepted
theories and ideas.

Basically, however, it seems that the family was "the

basic social unit" (though I will not enter into a discussion here

I
.‘

of exactly what is meant by this phrase). It has been pointed out
that Chinese families were not, on the average, large, extended
families, as was popularly believed by some in the West and China;i
Attention cannot be paid to size of family alone, of course, but 1
also must be paid to its structure and its social functions (in

the sense of its relationships with the rest of the social,

44

political, economic, and ecological environment). It seems safe
to say that a typical family in traditional China was made up of
three generations. Whether or not this was any different from a
contemporary American or European family is another question. As
the typical family was a peasant one, it tended toward economic
self-sufficiency (without suggesting that families were ever
"mainly" self-sufficient; this was clearly not the case). Although
three-generation families seem to have been quite common in Ch'ing
Taiwan, what information we have from the period, in the various
gazeteers, shows that the households averaged between four and six
persons (INS; 322).

It is interesting to note the sources of socioeconomic
support for the family system that are outlined in INS:

1. China has long been an agrarian country, and common,
cooperative labor is profitable in agriculture. The common ,
saying "have many sons and you'll get lots of money"(M. ting tuo,
hsieh ts'ai) indicates that only through coresident c00peration
can one bring prosperity to the family.

2. Because China was so large, the government was usually
unable to ensure peace and order in local areas, and because
self-protection was necessary, there naturally develOped large
families and other self-protection organizations.

3. To maintain a family's power, influence, and capital,
it is necessary to follow a large-family system.

4. Economy of scale. The ancient saying was “living
together, many expenses are saved; divided, each person has
expenses" (ChU chU tse pai fei chieh sheng, hsi chU tse ko

 

u fei .

5. In order to support normal human relations and develop
morality, the government promoted coresidency and common prop-
erty among families; honors were awarded to virtuous families,
and sometimes they were made exempt from tax payment. The
Code also prohibited brothers from living separately and
dividing the property as long as the parents or grandparents
were alive (INS: 323).

\
x _, _._-/“‘J

45

The same source notes, however, that because of various
factors, some of the supports for the traditional-ideal large-
family system became built-in threats:

1. The maintenance of the family relied upon the knowl-
edge and ability of the head, and if he were not competent,
then the family would lose its order. Even a competent head,
one with character, couldn't accomplish everything.

2. The maintenance of coresidence and common pr0perty
required that family members renounce selfishness and uphold
the common good. But in fact, there was usually division

because of selfish family members and unfair collateral
elders.

3. In the daily life of a large family, the women take
turns at the domestic chores. If there is any inequality in
work allocation, it is easy for gossip to get started, every-
one harbors suspicions with respect to the domestic management
of other branches, and it is easy for this to cause sisters-
in-law to argue and wives and concubines to quarrel (INS: 323).
A very important factor in the analysis of the domestiC‘;
system is the nature and control of the family estate or family :
property. I

The term "family property," of course, means "property
belonging to a family," but there are many interpretations of this
phrase. INS outlines and compares four major theories as to the
nature of the traditional family estate, and adopts the view thatfhr
the estate was not the sole property of the head, but rather was
the joint possession of the head and the dependents (those members i
of the family other than the head). This was so whether the head
was a lineal or collateral elder (INS; 325-31).

All of the terms used in the Ch'ing Code for division of

family estate (M. fen-hsi,gfen-i ts'ai-ch'an, fen-hsi chia-ts'ai

t'ien-ch'an) clearly express the idea of division of conmon property,

 

not that of inheritance. That is, it is clear that the estate was

46

a joint possession of the senior and junior family members, whether
lineal or collateral relatives. Although the possession was termed
"common possession" (M. kung-yu), the share coming to an individual
was unfixed, and changeable in nature. It could not be freely
disposed of before the actual division, and could be subsumed as
one kind of right that the individual possessed by virtue of his
status as a family member.

After the formal division, the original family estate was
divided into several new family estates, and each portion of the
original became the common property of the members of that new
famny.

A further reason to regard the estate as the common property
of the family members is that the c0parceners to the estate were
limited to the family dependents. The share due a c0parcener in
joint possession exists in the whole of each article that goes to
make up the estate. The articles themselves do not function as
shares. The share of each c0parcener is not fixed, and this is
different from ordinary joint ownership (the so-called joint owner-

ship in severalty, M. fenepieh kung-yu, in which the shares are

 

fixed). Thus if the number of c0parceners increases (through birth
or adaption) or decreases (through adoption out, uxorilocal mar-
riage, or death), then the shares will accordingly change.

The c0parceners have the right of increase (jus accrescendi),
and if only one c0parcener is left, the right of survivorship; the
estate becomes the property of that one person. The coparcener's

share is implicit, and is limited to persons with the status of

47

family dependent. Thus a c0parcener could not, before division,

at will dispose of any of the pr0perty of the estate, and could not
turn over his portion to any other person. The act of alienation ;
of the estate (through sale, mortgage, or gift) required the agree-
ment of the entire group of family members (but the lineal family
head's powers of alienation, as noted above, were not thus limited).
The debts of the family were the debts of each in the group, and
after division the basic c0parceners took on these debts. The
basic c0parceners had no right to request of the lineal elders that
partition take place, but they did with respect to the collateral
elders.

An important point here is that the family estate in the
traditional system was passed on from generation to generation
through division and not through inheritance. That is, since the
family estate is best seen as having been held in common between
the head of the family and the family dependents (other family
members), the latter had continuing rights (if implicit) in that
estate both before and after the death of the head. Thus division
of the estate before his death posed no logical or legal problems in
and of itself. This is connected with the fact that (as Tai Yen-hui E
argues) there were, in the traditional family system, no "pure" L
parental rights. The rights exercised by parents over children
were a form of elders' rights in general, and were limited and
restricted in turn by the elders' rights which the grandparents

held over the parents. Both of these reflect the continuity and

48

corporateness of the family as a unit, and a de-emphasis of the
place of any particular individual within the unit.

The result is a kind of flexibility and relativity within
the structure. The share of a c0parcener in the family estate is
not fixed, but is relative and changing. The change in one's share,
moreover, is dependent upon the others in the family. The rights
of one person over another in this situation are never absolute,
but are qualified, at least in principle, by those held by someone
higher in the hierarchy.

This sort of relativity makes for a good deal of flexi-
bility in size and complexity of structure. The basic blueprint
works for higher-order structures as well, such as lineages. This,
I think, is one of the sources of the problem that social scientists
have had in defining family and lineage and differentiating between
the two. The fact is that there I§_no clear and definite dividing
line between them. The Chinese themselves seem to give recognition
to this terminologically. The term M. gfljg, which is usually
translated "family," can also be used for much larger and higher-
order kin groupings. The distinction between this term and M. Egg
(lineage) is much more a Western (or Western-trained) social
scientist's distinction than it is a Chinese distinction. Simi-
larly a M. Iggg_(branch) can range in size and complexity from an
individual (a single, unmarried brother within a "family" can be
termed a Iggg) to a large grouping which in isolation might itself
be seen as a lineage. The term Iggg_itself refers literally to a

room within a family dwelling, and thus is simply a segment within

49

whatever is the larger unit. All of these terms are relative, and
changing in meaning except structurally; the specific meaning emerges
in context, in relation to what other units are involved.

This is also true of the term M. Ng_(household), but the
context of its use is rather different. Traditionally the distinc-
tion between "family" (M. ghjg) and "household" (M. Ng) was simply
that the household was the external aspect of a family, the family
as seen in terms of public law. In private law the same thing was
termed a family. In the Ch'ing Code it was specifically stated that
in the census records, when computing families, "household" should
be written (INS; 220). It may be objected that this is inadequate
because a household may contain persons who are not family members.
I think that this may be another problem in the translation of
cultures. We translate M. ENIg_as "family," then turn around and
expect the gNIg_to have the same characteristics as "family,"
including some kind of presumed biological coherence.

Many definitions of family in Chinese society do, in fact,
begin with the assumption that the members are "kinsmen" or rela-
tives with mourning duties to one another. Earlier in this chapter,
however, it was pointed out that it was necessary to divide the
family members into "basic" and "secondary" members, and that a
number of "unrelated" persons, including even household servants
and slaves, had to be admitted to some sort of family member status.

The distinction between the "family" and the "household"
became much less clear in Taiwan because of the influence of the

Japanese Civil Code. During most of its rule, the Japanese

50

Colonial government attempted to maintain what it could (or what
seemed useful) in the customary legal system. Inevitably, however,
court interpretations and administrative fiat had an effect on the
customary system. This was accelerated with the policy decision
that the Taiwanese legal system would eventually be brought within
the ambit of the Civil Code in effect in Japan proper.

The Japanese CivilCode,cfiicourse,differed from Taiwanese
Chinese customary law. The Japanese, for instance, assumed that the
Taiwanese system of family estate division was really the same as
the Japanese inheritance system, and treated it accordingly. This,
of course,then had an effect on the Taiwanese system of division and
on the family estate in general.

Another point of difference was that in the Japanese sys-
tem, the "family" was an abstract entity, and had come to be admin-l
istratively defined. Each person had to be registered as a member
(either a head or a dependent member) of a family in order to have
any legal status as a person. Each family had to have a head
(J. NgsNg), and the position of family/household head was passed on
in succession.

When the Japanese put essentially the same system into
effect in Taiwan, the Taiwanese family system was changed. The new
household head rights were of course placed in the hands of the
family head, and this had a great effect on the system of elders'
rights in the traditional Chinese system.

The difference that is usually emphasized between "family"

and "household" is that the former implies some sort of "kinship"

51

relationship, while the latter does not necessarily do so. Since
family members, in the data I present, were seldom actually
restricted to those with kinship relationships, I have not found
the distinction between the terms very useful, and will use them
more or less interchangeably. Any future discussion of Taiwanese
family structure and organization must pay more attention than has
been the rule to the effect of Japanese colonial policies. This is
also important in the wider problem of relating Taiwanese society to
Chinese society generally.

The major or "regular" form of marriage mentioned above
was only one of the ways a family could bring in new members. There
was also a surprising variety of alternative marriage forms and

adoption forms. The next chapters will examine these in detail.

CHAPTER III

IRREGULAR MARRIAGE FORMS

Uxorilocal Marriage

 

If one approaches it from the point of view of dialectical
logic, itis not surprising that a society which extols the virtues
of patrilineality, patriarchy, and virilocal/patrilocal residence
after marriage should also develop customs of uxorilocal marriage
and at least a limited form of "matrilineality."

Uxorilocal marriage has long been a feature of Chinese
society. It is unclear for just how long, but there is a clear
reference to it as an institutionalized form in the Shih-Chi,
written in the Han dynasty. In the section of that work devoted
to the biography of Hua Chi is the passage, "Ch'un-yU K'un was an
uxorilocal son-in-law to the state of Ch'i."

In the Han Shu, also written in the Han dynasty, in the
biography of Chia I is a passage which says, "In the state of Ch'in,
if a family is rich and the sons vigorous, then they divide the
estate and leave. If the family is poor, and the sons vigorous,
then they marry uxorilocally and leave." This passage occurs in a
memorial presented by Chia I to the Emperor in 174 B.C. to illus-
trate what depths the state of Ch'in had reached because the Lord
of Shang had abandoned propriety and benevolence.

Ch'U T'ung-tsu (1972: 252, n. 8), who translates this passage

somewhat differently, includes an interesting discussion of the

52

53

meaning of the word M. gNgi, which is common in terms for uxori-
local husbands. One explanation is that it means "an excrescence,"
something unnecessary. But it is also identified with M. QNIN,

"to pawn"; a man who is too poor to manage a betrothal gift would,
in effect, offer himself to the bride's family as a pledge, or in
bride service.

Another point of view rejects the latter interpretation,
and refers again to the Han Shu,where it is said that people pawned
their children (M. chui tzu) in bad years to get food and clothing.
It is further pointed out that there was a custom in areas south
of the Huai River of people selling their children (M. mai tzu) to
work for others as slaves. These children were called M. chui tzu,
"pawned children"; if they were not redeemed within three years, they
became permanent slaves. Thus, it is argued, the chui tzu was a
person who was “pawned" out for a fixed period--a bond servant.

A male who was pawned out in this fashion, and not redeemed, might
have then married the daughter of his master, in which case he
would be called a M. chui th, "bond son-in-law."

,Yet another scholar thinks that a chui th was simply a
male slave, one who was given a female slave as his wife, not his
master's daughter. "Whatever the explanation, the chui hsfi had an
inferior status and were treated unfavorably by the government.
Under the Ch'in and Han they were sent together with fugitives and
merchants to guard the frontiers. According to Kung Yu, chui hsfi
were also prevented from entering officialdom during Emperor Wen's

time" (Ch'u 1972: 253, n.).

54

Two other common terms for an uxorilocal husband both employ
the word M. g_h_a_o_, "toinvite" (T. gig). These are M. chao-hsll
and M. chao-fu, "invited husband." The former indicates a son-in-
law who is brought in to marry a daughter of a family, while the
second indicates a man who is brought into a family to be the second
husband of a daughter-in-law whose first husband (the son of the
family) had died. These distinctions are not always observed. The
Rev. Carstairs Douglas records that in the late 18005 in Fukien the
word T. chin-tsoe meant "marrying and going to live in the house
of the bride's father for a longer or shorter time (gen. short),
but retaining one's own surname and having full power over one's
own household, thus differing from the manner of marriage called
'ghig'" (Douglas: 1873). Barclay, in his supplement to the work of
Douglas, then says of the practice referred to by the word T. EDIE.
that "in some cases the bridegroom retains his own name" (Barclay:
1923). It seems, then, that the distinction between these two
words also is not always observed, and in fact I have never come
across any other evidence of a consistent difference of meaning
between M. gNgI_and M. gNgg_(T. t§9g_and gNIg, respectively).
Uxorilocal marriage is not only an old form, it is also
widespread. The Min Shang Shih Hsi-kuan Tiao-ch'a Pao-kao Lu
(Record of Investigations Into Civil and Commercial Customs, here-
after abbreviated N§§) records various types of such marriage from
areas in Chi-lin, Feng-t'ien (the present Liao-ning), Shan-hsi,
Ho-pei, Shan-tung, Ho-nan, Chiang-su, An-hui, Chiang-hsi, Che-chiang,

Fu-chien, Hu-pei, Shen-hsi, Kan-su, and Je-ho. A few quotes

55

from this source will give an idea of the general form and some of
its variations.
In various parts of Hu-pei,

The uxorilocal husband is treated as a son. Some follow

the wife's surname, some don't. The marrying-in ceremony is
. the same as in ordinary marriage (literally M. wan-hun, "com-

pletion of the marriage"). In Ma-ch'eng he does not follow the
woman's surname, and there is no fixed type of ceremony. In
Hsing-shan, if there is no son of the family, then the uxori-
local husband must change his surname, a go-between must
officiate, the relatives come for a feast, and then the son-in-
law is sent into the bedroom for the ritual drink of wine.

In Chu-hsi there are three types of uxorilocal marriage:
that in which the husband obtains property but no sons
(M. yu ch'an usu tzu), that in which he obtains both property
and sons (yggch'an wu tzu), and that in which he gets neither
pr0perty nor sons (wu ch'an wu tzu). In the first case the
husband continues the family line and inherits the family pr0p-
erty. He must change his name to that of his wife's family.
The second type occurs when the woman's family loves her very
much, and this sentiment is extended to the son-in-law. He
later receives a portion of the property, and need not change
his name. In the last case, the husband usually changes his
name, but there are cases in which he does not. As for the
marriage ceremony, it is the same as in ordinary marriage.

In Wu-feng all uxorilocal sons-in-law are treated as
sons, and continue the line. There is no distinction like the
one between the three types of marriage mentioned above; all
of them must change to the woman's surname. When the man
enters the woman's family it is after a go-between has arranged
it between the two families. The woman's family holds a feast
to give witness, and invites relatives and neighbors. A con-
tract of mutual agreement is written up. The contents include
the name of the one marrying uxorilocally, his father's name,
the name of the bride, and a guarantee that the children will
become permanent descendants of the family, or a specification
of exactly which children in the future will follow the son-
in-law's line. The son-in-law then changes his surname to
that of the wife's family. It is further recorded in the con-
tract, if the son-in-law (to be) is still a child, that he
will be raised (M. t'ung-yang) by the woman's family until the
couple is of age, when the marriage will be completed. If the
two are already adult, then an auspicious day is selected for
the wedding. The marriage expenses are all borne by the woman's
family, but there are also cases in which the man's family pre-
pares items of clothing and jewelry for the woman's family, or
the man's family may even send along a dowry and trousseau
(N§§; 1630-31).

 

 

56

Again, for Hu-pei, we find that

In Chu-shan this kind of uxorilocal marriage is divided
into three types based upon rights in property and in chil-
dren: those with rights in both, those with rights in neither,
and those with rights in property but none in children. In
the first form, the husband need not change his name. In the
last form, he is regarded as a son, and so must change his
surname. In the middle form, when he has rights in neither
the family estate nor the children he and his wife produce,
he is made a permanent dependent in the woman's family, col-
loquially called M. shan -men, and he also must change his
surname. This is a completely uxorilocal marriage; he does
not set up another household.

In Ching-shan he must change his name. This usually hap-
pens when there are no sons, and when the woman's parents die,
most of the time the uxorilocal husband then returns to his own
kin line. The ceremony of entering the household in uxorilocal
marriage is the same as that in regular marriage. In Ch'ien-
chiang, whether or not the uxorilocal husband changes his name
is decided in the agreement worked out by the go-between. This
is usually decided by the circumstances of the case. Generally,
if it is a case of a marriage in which the uxorilocal husband
gains rights in the family estate, but not in the sons, he
changes his surname. He rarely does so if he has rights in
both the estate and the sons. In the event that he has no
rights in either, he must pay a certain amount of money to the
woman's family for her trousseau. Sometimes he then changes
his name, sometimes he does not.

In Ku-ch'eng he not only must change his surname, he must also
go through the ceremonies of worship, lighting candles, and
drinking the nuptial cup.

In T'ung-shan no matter which of the three forms of uxo-
rilocal marriage [abbreviated M. yu yu, yu wu, wu wu] are
involved, the son-in-law does not change his name. His sons,
however, must follow their mother's surname, and continue that
line. The marriage must be witnessed by the go-between and
the relatives, a contract must be drawn up, there must be an
exchange of gifts, and the ceremony of presentation to the
ancestors (M. miao-chien) must be observed.

In Pa-tung the husband need not change his name. If he
is to continue his wife's kin line, however, either he does or
his first son does. The second son would then follow the
father's surname. The marriage ceremony is more or less the
same as that in ordinary marriage (NSS; 1644).

 

In Hupei, as elsewhere, there also existed a subvariant of
uxorilocal marriage called "inviting a son-in-law for support (of

his parents-in-law) in old age," M. chao-th yang-lao:

57

In Wu-feng there is a custom of uxorilocal marriage in
which the husband obtains rights in neither the estate nor
the sons. If one has only one daughter, for instance, and
if one wishes to rely on her care and support for the rest
of one's life, one can only invite in an uxorilocal husband.

He does not change his name, and does not continue his wife's
family line. The contract stipulates only that during the

life of the parents of the woman, they will be supported by the
son-in-law, and that in death the expenses of their funerals
will be borne by him. There is no fixed time limit set for

his leaving to set up another household (this, leaving with the
wife and children, was called M. ch'u-she).

In Chu-hsi, Ma-ch'eng, and Hsing-shan, all of these "uxori-
local sons-in-law for support in old age" have the time limit
set in the contract. It is usually said to be after the woman's
parents "are 100 years old" that the son-in-law can leave their
household. This is the so-called "half son, half son-in-law"
(M. pgn tzu pan th) (NSS; 1632).

 

In Chu-shan, Ching-shan, T'ung-shan, and Ch'ien-chiang
Hsiens, the time limit before the son-in-law can leave
(M. ch'u-she) is usually clearly set in the contract. It may
be ten years, or eight years, or when the children have grown
up, or when the wife's parents are dead and buried. This is
the "half son, half son-in-law."

In Chu-shan, however, when the agreed-upon period has been
completed, then if the son-in-law wants to leave, he must pay
over a predetermined amount to the wife's family. If there is
a difference of Opinion, the son-in-law may at times leave
before the stipulated time is up. In Chung-hsiang Hsien
whether or not the son-in-law changes his name, or whether or
not he supports the wife's parents in their old age are both
determined by the particular contract drawn up. In Ku-ch'eng
Hsien the contract does not set a specific time for the depar-
ture of the son-in-law. In Pa-tung Hsien, there is no contract
for this kind of marriage (NSS; 1645). '

The discussion so far has been mainly concerned with the
uxorilocal son-in-law, M. chui-hsfi or M. chao-hsu. There is also
the related, but somewhat different form mentioned earlier, the
M. chao-fu, the second husband of a widow, who marries uxorilocally,
typically into the household of the family of the first husband.
This form of marriage was found in many areas. In Chilin Province
he was called M. tso-t'ang chao-fu (MSS: 1311-12). In P'u-t'ien

Hsien Fukien, it is recorded that sometimes the uxorilocal husband

58

was even of the same surname as the first husband, which caused

much trouble. "Sons born of a different-surname uxorilocal father
cannot compete, with respect to inheritance, with those of the
original kin group. It is very hard to decide, however, whether the
sons of a same-surname uxorilocal husband should or should not be
regarded as the same as those born to different-surname husbands"
(NSS: 1969-70).

In Shen-hsi such a husband was called a M. Shang-men han.
He changed his name to that of the wife's former husband's family,
and in many other respects as well took on the former husband's
social identity; he enjoyed whatever rights to property the first
husband did (NSS; 1748).

It is recorded that in Shun-ch'ang Hsien, Fukien, .

If a woman's husband dies and she has no children, and the
family is poor, she will usually remarry, and move out, with
the permission of the head of the household. When they have
a little property, though, or if she has children, then it is
much more likely that she will marry in an uxorilocal husband,
or even marry out, taking the children and pr0perty with her.
These sort of customs are found mostly among the middle and
lower classes (NSS; 1579).

In Shan Hsien, Ho-nan, if a woman's first husband died, she
was able to take in a second, to live uxorilocally, and the rela-
tives of her first husband could not interfere. The second husband
changed his surname to that of the first husband, and any children
subsequently born to them also took that surname. The first hus-
band's property went to the sons by him, while the property of the

second husband, at his death, was shared equally by all of the sons

of both fathers (NSS: 1386).

59

The uxorilocal husband in Shan—tung, Te-p'ing Hsien, who
changed his surname to that of his wife was called, interestingly,
an "adapted son," M. i-tzu. His sons, however, were not called

"adopted grandsons,“ M. i-sun (MSS: 1398). A woman, in parts of

 

Shan-tung, who had no brothers or other relatives, could inherit
from her father. If her parents were both dead, and there was no
one else to arrange a marriage for her, then she could herself marry
in an uxorilocal husband, who would take her surname. This was

called M. tso-shan chao-fu. When this sort of thing happened, and

 

the uxorilocal husband inherited the estate, or if a girl had been
adopted in, and a husband married in uxorilocally for her, and they
inherited, the household was called, in Shan-tung, a "woman's

household," M. nU-hu (MSS: 1399—1400).

 

In some parts of Hu-pei (as well as elsewhere) it was stipu-
lated in the customary law that only sons could not marry uxori-
locally. In some other parts of Hu-pei (as well as elsewhere),
however, only sons were permitted to do so (N§§; 1633, 1635, 1946,
1951).

Chin-chiang Hsien, Fukien, is recorded as coming close to
fulfilling all of the logical possibilities:

In Chin-chiang Hsien, when one has no sons, then apart
from instituting a child of the same surname as an heir, or
adopting a child of another surname (M. ming-ling) to be
instituted as heir, there are other customs. If one has no
sons, but only daughters, one can invite in an uxorilocal hus-
band for a daughter, and even institute this son-in-law as
heir. Or, if one has neither sons nor daughters, one can
adopt a daughter, and then invite in an uxorilocal husband
for her, and institute him as heir. Again, even if one has
a son, who already married, but then died, then one can invite
in another husband, to live uxorilocally, and make him heir.

60

All of the above uxorilocal husbands abandon their surnames
and take on the family's surname. Later, whatever property
there is is inherited by the uxorilocal husband (NSS; 1573-74).

The report on uxorilocal marriage from ChU-yung Hsien
Chiang-su advances some basically sociological explanations for its
existence:

In ChU-yung problems relating to legal status (M. shen-fen)
and succession are exceedingly complex. Families without sons
usually invite in uxorilocal husbands for a daughter, to play
the role of son. When he marries in he changes his surname,
and signs a document as evidence, a document called an "usori-
local contract" (M. chui-shu), which is actually no different
from an adoption contract (M. chi-shu). If the person being
succeeded has sons or nephews, he will give them a certain por-
tion of the pr0perty to avoid disputes. If not, then he must
pay a certain amount to the clan, and then the son-in-law can
enter the genealogy and the clan, and enjoy all the rights of
an heir.

One reason for this kind of custom is that because of war
males become rare and the land is visited by devastation.
Another reason is that the parents love their daughter so
much, and this love is extended to the son-in-law. This lat-
ter is the most common circumstance. After it has been prac-
ticed for a long time, people come to regard it as natural.

In families without much property it is not much cause of
litigation. In cases of families with some property, or in
cases in which the son-in-law is for some reason unsatisfactory,
then lawsuits certainly do arise. In recent years litigation
has reached 80% to 90% of the cases. If one strives to main-
tain the relationship, then one finds oneself acting counter
to the law. If one is strict in following the rules, then
there are too many restrictions of the freedom of others. It
would seem better to have this matter regulated by the Draft
Civil Code (NSS: 1465-66).

pt
, V

Thus in parts of Chiang-su uxorilocal sons-in-law could be I
entered into the lineage and the genealogy. This was not true in
other areas. In Min-Ch'ing Hsien, Fukien, for example, it was made;
explicit in the customary law (according to the Min Shang Shih col-
lection at any rate) that an uxorilocal husband could not be entered

into the genealogy after death. Sons born to the union, however,

61

could change their surnames and be entered (NSS; 1601). In parts of
Hu-nan, "sometimes the [uxorilocal] son-in-law is entered into the
woman's family's genealogy, and it is recognized as a pr0perly
listed name. Sometimes this is not so" (NSS; 1690).

In some areas (Chekiang, for instance; N55: 1530-31) a
terminological distinction was made between husbands or sons-in-
law invited to live uxorilocally to manage the family estate

(M. tso-ch'an chao-fu) and those invited in to support the children

 

of the dead first husband (M. chao-fuyang-tzu).

Feng-t'ien (the present Liao-ning) had a custom which
approached the "temporary marriage" discussed elsewhere. This was
called M. tg;Ngg, "companionate marriage":

In Sui-Chung Hsien there is the custom of ta-huo, which
means to marry a husband in to live uxorilocally for support.
If a woman's husband has died or has been gone for several
years without any trace, and the family circumstances are
poor, so she cannot support herself, then she can establish
this relationship with another man. Their relationship is
the same as in ordinary remarriage, but a contract must be
written up. In the contract it must clearly state that this
is because of poverty, and that the husband will assume a cer-
tain number of debts. If the first husband is to return,
then the money will be returned to the second husband, who
must then return the wife (and any children she may have taken
with her). This kind of custom is found only too often in
both urban and rural areas, each case for its own reasons
(the relatives of the first husband, for instance, may be
desirous of the money, or may oust her from the estate). In
lawsuits the contract becomes very important. In the minds
of rural people, this is a perfectly proper custom (NSS; 1308). e

Not much has been written, in either English or Chinese,
about the various forms of uxorildcal marriage. One relatively

easily available article is by Liu Hsing-t'ang titled "Two forms

of uxorilocal marriage," published in The Eastern Miscellagy (Tung-

fang Tsa-chih) vol. 33 #15, August 1936.

 

62

In his article, Liu mentions that it is often difficult to
find material on uxorilocal marriage:

We might look for help on this from the local gazeteers.

They, however, publicized only the exalted and refined cus-
toms; this sort of "indecent" affair was forbidden by them.
According to the principle of "hiding the evil and publiciz-
ing the good," such things were generally left undiscussed.

It is in this context that I would like to make known

Mr. Ch'en Sheng-shao's work, the Wen-su Lu (A Record of Cus-
toms). Mr. Ch'en has, with a fearless spirit, filly and in
detail, exposed for Fukien Province the dark areas of feudal-
ism, the power of clan organization, the economic life, and
various other kinds of customs. This is an important archival
source for sociology. economics, social history, and ethnol-
ogy (Liu 1936: 109).

After an introduction like that, and after reading the
excerpts included by Liu, I of course wanted very much to find a
copy of Ch'en's work. This desire was only increased when I dis-
covered that sections of it were reprinted in the Komalan T'ingechih,
the gazeteer published in the Ch'ing Dynasty for the I'lan area
in Taiwan, where I did research. The sections that appear there
include information on social problems in Taiwan, inter-ethnic
disputes, the Aborigines, the process of development of the land,
banditry, industry, and so on.

I was very disappointed, then, to be unable to find a copy
anywhere, or even a reference to the work. When I did finally find
a reference to it, it was in an article by Cheng Hsi-fu published in
Taiwan Feng-wu (Taiwan Folkways), February 1970, in which Cheng
laments that the work is now lost.

The Wen-su Lu, by Ch'en Sheng-shao, written in the Tao-

kuang era II821-1851], is an important historical source for
Taiwan, compiled by a well-travelled official during the mid-

Ch'ing. Though it was written not long ago, very few peeple
know anything about it; Lien Ya-t'ang does not include it in

63

the Taiwan T'ung-shih [General History of Taiwan], and none

of his contemporaries seem to have mentioned it. It is really
a pity that it is this difficult to preserve historical
sources. Fortunately, the Komalan T'ing-chih has preserved
six excerpts from it (one of which also appears in the Tan-
shui T'ing;chih), so that we still have some impression of it
as a work—(Cheng 1970: 74).

 

 

 

 

The excerpts which appear in Liu's article, however, include

material not found in the Komalan T'ing-chih. It seems, then, that

 

Liu, writing in 1936, had access to the original, so the work may
not be completely lost, and a copy may surface some day.

Aside from their historical interest, the excerpts quoted
by Liu are important for the information they give on uxorilocal
marriage in both mainland China and Taiwan.

In Chao-an (Fukien) when a middle-class (M. chun -hu)
family marries in a wife, the bride-price is around 100 taels
(M. chin); among lower-class families (M. hsia-hu) it is 56
or 60 taels. Other gifts are not furnished. The dowry when
marrying out a daughter is similar, so marriage is quite dif-
ficult. Marrying out a daughter is quite difficult, so few
daughters are raised. Marrying in a wife is quite difficult,
so bachelors are many. It is pr0per that a man have descendants
to continue the line, so an attractive woman may invite in an
uxorilocal husband to produce children to do so. The old cus-
toms are based on this (Liu 1936: 109).

If someone of the lower class is old and without sons, and
has a daughter of marriageable age, then they invite in an
uxorilocal husband for her in order to get the most out of
their remaining years. In Chao-an people buy girls in order
to get an uxorilocal husband, and widows also invite in hus-
bands, in order to continue the line, to maintain the (first)
husband's grave, manage the estate, and husband the resources.

The sons are not discriminated against on the basis of
which husband was their father, and the clansmen do not dis-
approve on the grounds~of having disrupted the kin group. In
this situation it is agreed, in contract, that the first-born
son shall follow the wife's family and subsequent sons the
husband's line. Sometimes all follow the wife's and none the
husband's. There are some husbands who, enamoured of the wife,
and covetous of her pr0perty, completely forget their origin.

If he steals the wife's pr0perty and absconds, or does not
attend to the support of her parents, then this is cause for

64

litigation. If a husband marries uxorilocally, his son may ,
inherit (in the wife's line), or a son may be sold in adoption
and so enter that ancestral temple. Though this is confound-
ing surnames and lines, its actual occurrence in Chao-an is
very common. Such is the popularity of uxorilocal marriage

(Liu 1936: 109).

Ch'en is quoted as saying that in Lu-kang T'ing (Taiwan),
where he was an official,

When a widow invites an uxorilocal husband, or an uxori-
local husband is brought in for an ad0pted daughter, it is
called M. chui ([T. tsoe or chehl just as on the mainland.

In Taiwan wives call the husbands M. lao-kun [T. lo-kon ],
and the weakest is the M. chui lao-kung (i. tsoe lo-kogg .

He makes a virtue of obedience, is submissive, and plays the
role of wife. Moreover, there are prostitutes who, when there
is no alternative, take in uxorilocal husbands and then talk
of themselves as married women, while in fact he is no more
than an Opium-den flunky. The conjugal intimacies are no more
than vain imaginings. Once a difference of opinion arises, he
is expelled. Thus the uxorilocal husband in this situation is
like a wife, and actually even more like a female slave

(M. gel) (Liu 1936: 110).

 

Finally, he says of the Aborigines in Lu-kang that "they
do not employ a go-between, the males usually marry uxorilocally, and
after marriage they rarely indulge in immoral liaisons. Taking a
wife is called 'taking the hand' [M. ch'ien-shou, T. khan-chhiu],
and divorcing her is called 'dropping the hand' [M. fang-shou,
T. pang:chhig]" (Liu 1936: 108).

 

This last section is also found in the Komalan T'ing-chih
(chapter 5b, on the "wild" Aborigines, the appendix), but the rest
of the quotes, as far as I know, are found nowhere else than in
Liu's article. This reference to the Aborigine word for marriage
may be one of the sources for the claim (made, for instance, by
Lien Heng in his Taiwan YU-tien [A Dictionary of Taiwanese Phrases])

that the Taiwanese word for wife, khan-chhiu, is traceable to

 

65

non-Chinese, Aborigine, marriage customs. Douglas, however, in
his dictionary of Amoy vernacular, lists this word as meaning
"wife." He is generally careful to point out "local" words, and

makes no comment in this case. Furthermore, khan-chhiu is used

 

only as a noun, never as a verb, and pang-chhiu, "divorce," is never

 

used (Wu 1953: 136). I suspect that it is an old Hokkien word
unconnected with Taiwanese Aborigine customs.

Liu also provides us with some more recent information on
uxorilocal marriage in mainland China. He cites an investigation
of marriage customs in Hu-pei which was reported in the April 20,
1936, edition of Hsin-wen Pao.

Families with daughters but no sons may invite in an uxori-
local husband. He follows the wife's surname and is borrowed
(M. chieh) to continue the wife's family line. But there are
also families with both sons and daughters which still take
in an uxorilocal husband. This is probably because the mother
greatly loves the daughter and cannot bear to be parted from
her. Only in this case would there be such an irrational cus-
tom. There are also those without either sons or daughters,
who buy or adopt a girl and then invite in an uxorilocal hus-
band for her. There are even cases of families of which the _
sons-in-law, for many generations, were all uxorilocal, creat-
ing a large and flourishing family. There is more than one form,
and they are practiced by eminent families as well as poor.
These customs flourish in the Chiang-ling, Kung-an, and Pao-
k'ang areas (Liu 1936: 107-108).

 

Among widows in Pao-k'ang Hsien there is the custom called
M. tso-t'ang chao-fu, "remaining in the family, bringing in an
uxorilocal husbandi“ The uxorilocal husband is called a
M. p'ei-erh, "compensator."

In T'ung-ch'eng Hsien wives may take an uxorilocal hus-
band, after the death of the first husband, to move in, manage
the domestic affairs, and support the children. This is called
M. Shang-men. The same custom exists in Chung-hsiang Hsien,
but the name is not the same. There it is called M. chao-fu
yang-tzu, "inviting an uxorilocal husband to support the chi -
dren.‘

Those husbands who marry uxorilocally have a ready made
wife, ready-made children, and ready-made pr0perty. Nothing

66

could be more convenient. But one condition is that he must
follow his wife's surname. Some call him M. nan-ch'ieh,
meaning "male wife," but he is really a "female husband."
Usually those who take advantage of this convenience are
extremely poor men, but there are also those who endure this
humiliation for only a short time, and wait until after they
have entered the wife's family and then suddenly, with a
resurgence of pride, they change back to their own surname
(Liu 1936: 108-109).

 

Liu claims that in the Han-chung area of Shen-hsi, the
ancestors of one family usually include several different surnames,
and that this is also true of Szu-ch'uan. That one surname-group
should maintain sacrifices to ancestors of several different sur-
names, and that several different surnames should unite in one
blood line, he argues, is a natural outcome of one or the other form
of uxorilocal marriage. In the Nan-yang area of Ho-nan there is
said to be a tradition that this form of marriage was practiced in
the Ming Dynasty. Village names composed of double surnames have”:
resulted from one son's following the mother's former husband's sur--
name, and another son's following her second husband's surname. To I
the present day, says Liu, the two surnames are still regarded as
of the same blood, and do not intermarry (Liu 1936: 110). This
phenomenon involving surnames is not rare. For an example in con-
temporary Taiwan, see the discussion of the Liaos of T'ien-sung-p'i
in Chapter V.

In Taiwan, the customs surrounding the institution of
uxorilocal marriage were not much different from those on the main-
land of China. Uxorilocal marriage is perhaps best divided into
the two forms mentioned earlier--that involving the marriage of a

daughter of a family, and that involving a daughter-in-law of a

67

family. I will begin with the first form, called M. chui-fu

(T. tsoe-hu) in the Nationalist Code, but also sometimes referred
to as M. chao-hsu (T. chio-sai), or, more rarely, the institution
is called M. chiu—hun (T. chiu-hun).

If it is seen from the point of view of the inviting family,
three main goals emerge. These (not mutually exclusive) goals are
to provide heirs for the family, to support the older generation or
young children in the family, or to provide someone to manage the
family estate.

According to the legalistic interpretation of the IgINgg.
Min-Shih Hsi-kuan Tiao-ch'a Pao-kao (TMS), China has long had an
"iron rule" that there is no intermarriage between those of the same
surname, and those of different surnames don't ad0pt (the tradi-
tional phrase is M. t'ung-hsing pu hun, i-hsing puyang), so it is
impermissible in both II_("rites," "pr0priety," "morality") and law
to consider the uxorilocal husband as having been adopted. Thus
the children of the uxorilocal husband must be seen as children of
the wife's family, not his, for the goal of providing heirs to be
realized (INS; 110).

This may be fine logic, given the premises, but it is poor
description of social and legal realities. The Chung-kuo Min
Shang Shih Hsi-kuan Tiao-ch'a Pao-kao Lu (MSS) has many reports
from many different areas of customs permitting same-surname intermar-
riage. There are also many reports, in that compilation as well
as elsewhere, not only on the abstract permissibility, but the wide-

spread practice of inter-surname adoption.(In fact, this type has

68

probably always been as p0pular, if not more so, in Taiwan, than
intra-surname ad0ption, even if we are considering only male chil-
dren.) Thus it should be no surprise that in many areas the uxori-
local son-in-law Ng§_explicitly considered as adopted. In parts of
Shan-tung, as noted above, an uxorilocal husband who changed his
surname to that of his wife was called an "ad0pted son" (M. Iztgg).
In parts of Chiang-su uxorilocal husbands played the role of

son, and their contracts were no different from adoption con-
tracts. In some areas an uxorilocal husband could be entered into
the family genealogy. This last is true of Taiwan as well. The

Sungrshih Tsu-p'u, a genealogy published in Taichung in 1971 (and

 

which includes inhabitants of T' ien-sung-p'i) lists uxorilocal husbands
quite often. There was also a case in the village where I did
research of a man who was simultaneously married uxorilocally into,
and adopted into, a different-surname family.

Another goal on the part of the inviting family, as noted,
is to provide support or property management. A high court ruling
in the thirty-ninth year of Meiji (1906) said that

The M. chao-fu (T. chio-hu) institution can be divided
into two aspects, marrying out uxorilocally and marrying in
uxorilocally. Either way it is marriage with a daughter of
a family to provide that family with pr0perty management or
heirs. Thus it must imply a submissive relationship on the
part of the uxorilocal husband to the family head, and the
marrying-in husband cannot set himself up as an independent
family head with all the public and private powers that that
implies. This is an obvious fact in the customs. Thus if the
man feels that he and his family have been slandered against
to the point that he cannot endure living there any longer,
and uses this as a reason to go against the household head's
wishes, and demands that the wife be handed over so that they
can live apart from the household head, this is entirely
counter to the goal of such a marriage (INS; 110, 111, 247).

69

It should be noted that the M. chao-fu referred to in the
passage above refers to both the chao-fu pr0per (the uxorilocal
husband married to a daughter-in-law of the family) and the
M. chao-th (an uxorilocal husband married to a daughter of the
family). The Japanese seem not to have made a clear distinction
between these two in their legal rulings, and the two are lumped
together in the Nationalist Code as M. chui-fu.

On the part of the man's family, it is usually said that the
motivation for marrying a son uxorilocally into another family is
economic; the family is poor and lacks the money and presents
needed to marry in a wife in the regular way. There is another
consideration. It sometimes happens, it is claimed, that two fami-
lies are on very intimate terms, and wish to intermarry. Because
the girl's family "loves her too much," or if for some other reason
they are reluctant to part with her, they may have the son of the
other family marry in and live uxorilocally (IN§; 111).

There are two basic types of uxorilocal son-in-law marriage,
that in which the son-in-law moves permanently into his father-in-
law's family and that in which it is for a limited period. This
latter type can then be divided into two further categories, one
in which the period is indefinitely limited, and one in which there
is a definite limit. The former, for example, might be until the
children are grown (the sons, that is), or the woman's parents are
dead and buried. The second sort of limitation is set in terms of

a certain number of years, typically from three to ten.

70

After the agreed-upon period, however it is defined, the man
is then able to leave the family if he wishes, taking his wife and
children with him. Often he also receives a kind of "severance
pay," an amount agreed upon in the contract that he is to receive

when he fulfills it. The present mayor (Ts'un-Chang) of T'ien-shan

 

Ts'un (one of the administrative divisions of T'ien-sung-p'i)

had married uxorilocally into the Liao family in the 19305. He
signed an eight-year contract, but stayed with his wife's family
for ten years. He received three yen per month, and at the end of
the time period, got a lump sum of 300 yen. He laughs now, and
says that because of inflation, by the time he got the money, it
was hardly enough to buy a bottle of wine.

According to the mayor (and in addition to his personal
experience in this, his official position requires that he write up
marriage contracts among others, and help mediate marriage disputes),
if an uxorilocal husband moves out of the wife's family before the
expiration of the contract, he must pay a sum of money in compen-
sation, and then the marriage becomes an "ordinary" marriage. If
divorce occurs, then there is no compensation, but then there must
be an agreement about the fate of the children.

This form of marriage occurs only, it is said by informants,
if the wife's family is much wealthier than the man's. Education
is also important. If a man has a high school education, he could
marry into a wealthier family than if he had only an elementary

education.

71

In this kind of marriage, the husband gives the wife's
family money to buy things for the marriage ceremony, but then they
return it to him when the contract has expired. The upshot, though,
of all the exchanges is still that the marriage costs "about one-
fifth" what a "standard" one costs--NT$10,000 as Opposed to NT$50,000
for an ordinary marriage.

In the case of the Ts'un-Chang, he paid about the same as

 

he would have in an ordinary marriage because his family was very
poor, and his wife's family was of much higher status. At that
time, in addition to other business interests, they owned one-half
share in a bus line that connected T'ien-sung-p'i with Lo-tung.
Thus he was able to ally himself with one of the most powerful
socioeconomic groups in the town. He got on well with his father-
in-law, and was prevailed on to stay an extra two years beyond

his contract time, in part because his father-in-law had lost a
good deal of money on his business deals. It is not uncommon that
a husband stay on longer than was specified in the contract;
usually he gets more money.

In wealthier families this kind of marriage is treated as a
very serious matter, with many conditions set out in the contract.
In poorer families, there are fewer conditions, the husband is not
so restricted, and the whole thing is freer.

The Ts'un-changjs decision to marry uxorilocally has paid off.

 

He has a respected standing in the town, and official office. He
has "many" chia of land, he owns a building in Taipei that he

rents, and he Operates a drygoods store. He has two sons with

72

college educations; one is a businessman in Tokyo, and the other is
trained in pharmacy, and is resident in the United States part of
the time. (Both, however, are still officially registered in T'ien-
sung-p'i.) His wife's older sister also had an uxorilocal husband.
The latter married into the Liao family in 1928 and stayed for
fourteen years. Finally dividing from them in 1942, and setting

up a new household, they too are quite well-off. They are land-
owners, not tenants, they own a drygoods sh0p, two of their
daughters are graduates of the Lanyang Girls' School (said to be
the best in I-lan), and a third daughter is a college graduate,
teaching at a local middle school. In neither case of uxorilocal
marriage did any of the children take the surname Liao.

As far as the form of uxorilocal marriage is concerned,
because of its nature, there is usually no M. ch'in-yjng_(T. chhin-
geng), the part of the ceremony in which the bridegroom goes to the
bride's house to escort her to his house. Some other elements of
the usual ceremony may be changed, but one important requirement is ii
the contract.

Often the uxorilocal husband is in a rather weak position
and can be easily browbeaten by the wife's family, so in the
Yuan, Ming, and Ch'ing laws, it was ruled that a marriage
contract must be written up, called a M. chao-hun tzu [T. chio-
hun jil, clearly specifying how long he will remain in the
wife 5 family, when he will be permitted to leave with her and
their children, what rights he will obtain, what duties he
will take up with respect to the wife's family, and which kin-
ship lines the future children (particularly the males) will
follow (IN§; 111, 112).

This was true under the Japanese as well; in 1917 it was

ruled that such a contract was a legal requirement of such a

73

marriage, and it was further specified that in the absence of such

a contract the relationship was to be regarded as simple cohabi-
tation, and that to have it recognized as a valid marriage, a con-
tract had to be written up specifying the assignment of children and
whether it was a permanent or fixed-time relationship (INS; 112,
134n, 256). The simple act of registering the marriage with the
household registry did not legally establish the relationship.

The husband, through his agreement to the contract, had,
either permanently or for a specified length of time, to live with
the wife in her household. He became a family dependent in that
household, but his relationship was ambiguous. He was supposed to
take on only an affinal role in the wife's family, not the role of a
blood relation of that family or a member of that family's kin
group (INS; 112).

As we have seen, however, the situation could be much more
complicated than that. The uxorilocal husband could, at times, take
on a "quasi-blood relationship" within the wife's family. In any
case, as a dependent member of that family, he was subject to the
wife's family's elders' rights and/or household head rights. In
addition to these general duties of his, to obey those holding the
above rights over him, he had also to produce heirs and meet the
conditions set out in the contract, including that of support.

We have noted that depending on the type of relationship
entered into, and depending on the particular clauses in the con-
tract and the rights and duties thereby transferred, the husband

may or may not change his surname to that of the wife. According

74

to the compilation of Taiwanese customary law, a court ruling in
1910 held that the uxorilocal husband maintained, in any case, his
"blood" relationship to his natal family and his membership in his
original kin group. Thus he kept his own surname (INS; 112). This
may have been true in the law books, but clearly went against the
customary arrangements; though it has perhaps always been most com-
mon that the husband keep his own surname, it was not uncommon for
him to add the wife's surname to his own, or, though this was the
least common of the three, to change it completely (see Tai 1966:
146).

While the husband did not lose his original kin ties or
nominal membership in the kinship group to which he was entitled by
birth, certain other rights and duties were curtailed. During the
time that he was resident in his wife's family, for instance, all
rights and duties which arose from the family-dependent relationship
(such as the duty of following the orders of one's own family elders
or the family head) were ended, since they were replaced by the same
rights in the hands of the elders and the head of the wife's family.

Another sort of curtailment, of course, depending on the
clauses in the marriage contract, was in rights Unand over children.
As was noted earlier, during the Japanese period it was something
of a legal requirement, to make the union valid, that there exist a
contract with clauses specifying the assignment of children to sur-
name and kin line. This kind of clause was almost always found in

such contracts from other periods as well, because it was naturally

75

an important factor in the relationship and a source of potential
trouble.

Those children assigned to the husband in an uxorilocal
marriage, who were to follow his surname and give sacrifices in his
kin line, had, in the customary law, legal relations to their parents
and other kinsmen which were essentially the same as those in "regu-
lar" marriage.

The legal status of those assigned to the mother and her kin
line, or those destined to joint assignment to both the father's and
mother's lines, was a bit more confused. In Tai Yen-hui's Opinion,
the kinship relationships of such children arose from their birth,
and not from any sort of implied adOption (INS; 113). This is, in
a sense, paradoxical. The same natural process (birth) produces two
different kinds Of children with two different sets of rights and
duties, but the distinction between the two kinds, the allocation of
individuals to each group, is, of course, a cultural process.
Logically, this is connected with considerations about the status of
the uxorilocal husband with respect to the wife's family.

To the degree that the marrying-in husband is viewed as an
affinal relative in his wife's family, the custom of allocating one i
or more sons to that family to take that surname and sacrifice to
that kinship line must be seen as matrilinealdescent.(This is the
custom Often referred to as M. huan-sun [T. hoan-sunl; customarily
the first son is assigned to the inviting family, though this is
modifiable in the contract.) The logical alternative to this

(matrilineality, as noted elsewhere, is usually attributed to

76

barbarians) is to have the husband change his surname, adOpt him,
and then there is no problem with the mode of descent; it is simple ’
patrilineal descent. What this does, of course, is to shift the_
logical problem elsewhere in the system. Now a "son" is married to
a daughter, and one is uncomfortably close to the incest taboo.

I have not seen uxorilocal marriage discussed in precisely
these terms, but this is what is meant when it is argued that the
uxorilocal husband cannot be viewed as having been adopted because
those (If the same surname don't marry, and those of different sur-
names don't adopt. If the husband is viewed as adopted, then both I
parts of the rule are violated simultaneously.

However problematical the precise status of the uxorilocal
husband may be in terms of Chinese kinship theory, when it comes
to inheritance and other relations to property, it is not surprising
that a number of fairly widely accepted and fairly clear rules have
evolved.

Traditionally the main arrangement (and this seems to have
been true for Taiwan as well as the rest of China) was more or less
as follows. The uxorilocal son-in-law had the right of management
and disposal of his own personal property. This property was, in a
sense, held in common with the children following his own surname
and line. After his death (or at division if this occurred before
his death) the property went to those children. If there were no
such children, then it went to the children following the mother's

line (see Tai 1966: 146; INS; 113).

77

It was noted earlier that the uxorilocal husband maintained
"nominal" rights to membership in his natal kin group. As long as
he was resident with his wife's family, he was considered a family
dependent of that family, and not Of his natal family. Thus, during ,
this time he had no rights to the family estate of his natal family.
If, however, he returned to his family and kin line before property
division, he could participate in that diVision.

’As far as the original estate of the wife's family is con-
cerned, the uxorilocal son-in-law had no rights Of possession,
though he might have rights and duties of management of that estate.
He usually had the right, however, to share equally with the family 7'
whatever increase in the estate occurred through his management; often
this was included as a clause in the contract. In the Ming and
Ch'ing Codes, however, it was held that a permanent uxorilocal son-
in-law had the right to divide the estate equally with an adopted
heir, another instance where the uxorilocal husband seems to be
equated for some purposes with an adapted son (INS; 114). At times,
depending on the situation, a fixed-term uxorilocal son-in-law might
also be given a portion of the family estate, called currently in
Taiwan, according to the compilation Of Taiwanese customary law,

1. tam long-a pun, "carrying Off a basket of money," or T. t3g_

 

long-a pun, "measuring out a basket of money" (INS; 135n). Most

 

of the above shares were, of course, subject to his having fulfilled
the uxorilocal marriage contract satisfactorily. If the uxorilocal
son-in-law has been adopted, that is made an heir, then he can, of

course, inherit in the wife's family estate.

78

When it comes to the termination of such a marriage, the
Taiwan Min-Shih Hsi-kuan TiaO-ch'a Pao-kao points out that the
principles are the same as in ordinary marriage, but the positions
of husband and wife are exactly reversed, and the uxorilocal son-in-
law is not uncommonly driven out unilaterally by the inviting
family (INS; 114). The man's departure from his father-in-law's
household, taking his wife and some or all of his children with him,
is called T. chhut-sia (M. ch'u-she). This may happen because the
period in the contract has expired, because of mutual agreement, or
he may unilaterally leave, or be expelled. There are two main
points where this differs from the dissolution of regular marriage.
After the death of the husband, his wife still maintains her affinal
links with his family, even though she is not resident there. If she -
remarries, then they are extinguished. The death of the wife, however,
extinguishes his affinal ties with his wife's family, even though
he may, through mutual agreement, continue to reside with them.
This seems to be another indication of the relative inferiority of
the uxorilocal son-in-law's position in the relationship (INS; 115).

As far as the current legal status Of this form of marriage

is concerned, I quote from INS;

The Civil Code recognizes uxorilocal marriage as well as
regular marriage, and this is done to remain close to custom.
According to article #1002 the uxorilocal husband shall live
with the wife, which is the meaning of the Old term chiu-hun.
Also, according to article #1059, "The children of the uxori-
local husband shall take the mother's surname; but if there
are other provisions made in the contract, they shall apply."
This is a change from the old viewpoint of passing on the sur-
name and continuing the line. According to article #1000,

"The uxorilocal husband shall prefix his wife's surname to

his own. If provisions are made otherwise in the contract,
however, they shall not be limited by this."

79

This last provision does not follow the custom, so one
might ask why it has been included. It was desired to give
some recognition to sexual equality, and to express the fact
that in uxorilocal marriage the wife has transformed her
status. The relationship between uxorilocal husband and wife
is a marital one. The relationship of their children to the
husband's or the wife's family line is equally a blood rela-
tionship in each case. Because the Civil Code is based on
ideas of bilateral descent and inheritance, whether or not the
children are family dependents does not distinguish between
them, nor does a difference of surname. The recognition of
the freedom of the parties to the contract to specify in the
contract the allocation of surnames to the children and the
assignment to the ancestral lines of the mother or the father
has not been put to the question in law, but according to
article #1059, second part, the children of an uxorilocal hus-
band in principle follow the mother's surname, but if provi-
sions are made to the contrary in the contract, they shall be
followed. Also, Article #1060, part 2, says that the children
shall reside with their mother. As for the contract itself,
and its division of the children with respect to surnames and
property rights and inheritance, this can be viewed as an act
of division by grant or testament (INS; 116).

We now come to the second major type of uxorilocal marriage
common in Chinese society. This, as has been noted, is the uxori-
local husband (as opposed to the uxorilocal son-in-law); a man who
marries into a family, but to a daughter-in-law (as second husband)
instead of a daughter. There were several types of this form of
marriage. First, families lacking sons or resident sons-in-law
may invite in an uxorilocal husband to manage the family estate.
This is the so-called T. chiO-hu che-soa*. (This is one reading of
the last character, and it is the same, except for tone, as the
reading T. gggf for "mountain"; in some areas of the mainland, at
least, this type of uxorilocal husband is called M. tso-shan chao-fu
["uxorilocal husband sitting on the mountain"] [T. che-soa* chio-hu].)

At the same time, the term M. tso-shan ("sitting on the mountain")

80

itself also has the meaning "manager, overseer," and is Opposed to
M. ch'u-hai ("the one who goes to sea," "supercargo").

Second, if a couple was old, and had no sons to support them
in their Old age (and one tends to forget, in an age of Social
Security, just how serious a situation this could be), they might
marry in a husband, after their son had died, to their daughter-in-
law. This was, of course, the form called T. chiO-hu iong-lo
(M. chao-fu yang—lao,_ "inviting in an uxorilocal husband to support
the older generation").

Third, if a son dies leaving a wife and small children, his
parents might invite in another husband for the daughter-in-law,
to support and raise the children. This is called T. chio-hu iong-tsu

(M. chao-fu yang-tzu, "inviting in an uxorilocal husband to support

 

the children"). The children by the previous husband will call him
"step-father" (T. Ngzflg; M. ENI:IN; or T. au-chek,NL hou-shu), and
he will reciprocate with, for example, T. Ng;t§g_(M. chi-tzu) for
his step-son. In principle, they should all reside together and the
mourning grade of the children to him should be that for a step-
father, but in Taiwanese custom they do not do so if they are already
adult, and can independently manage the property themselves. If her
children are independent and the mother still invites in an uxori-
local husband, "then this is considered an immoral act (M. paj;tg)
on the part of the mother“ (INS; 117).

Fourth, if a family's only son has died, leaving a widow,
then a husband may be invited in specifically to provide heirs. This

is called T. chio-hu si*-tsu (M. chao-fu sheng3tzu).

 

81

Whatever the goals on the part of the inviting family, the
goals on the man's part are said to be much the same as for uxori-
local son-in-law marriage--1ack of money for the brideprice. One
of the ways that this form of marriage differs, in fact, from ordi-
nary marriage is that the expected amount Of brideprice is compara-
tively small, at times nonexistent. Customarily, the man does not
provide anything more than the cost of a mosquito net or other such
bedroom furnishings (things brought by the woman in ordinary mar-
riage). He may sometimes assume part of the cost of a feast for
the family (INS; 119).

This kind of uxorilocal marriage differs from the first type
considered, apart from the fact that the wife is a daughter—in-law
instead of a daughter, in that rarely is it a limited-term rela-
tionship, in the sense of limitation to a definite number of years.
The arrangements often specify that after the mother-in-law and
father-in-law are dead, the husband and wife can then leave the
household, or that after the birth Of sons they can leave.

As in the other form of uxorilocal marriage, provision must
be made in the contract for the allocation ofcfifildren.(Provision
is gINgyg made, says INS; 118.) In this form, however, the situa-
tion is somewhat more complicated, in part by the likelihood that
there are also present children by the first husband.

If the purpose of the marriage was not to provide heirs for
the inviting family, then the children, in customary law, would
follow the uxorilocal husband's surname and kinship line. What

the provisions are in the contract, though, is very important. A

82

court ruling in 1913 held that if in the contract it was stipulated
that the children should follow the kinship line of the inviting
family, but somehow the opposite was entered into the household
registry, this was insufficient to consider that they did not follow
the inviting family; the terms of the contract took precedence over
the household records (INS; 122).

A common principle for the division Of children, as with the
other form of uxorilocal marriage, was to have the first son follow
the inviting family's line. Again following the reasoning with
respect to the other form of uxorilocal marriage, they are considered,-
legally, to be participating in matrilineal succession (complicated
by the fact that the mother's relationship to her former husband's
family is a bit curious). The sons who follow his surname and
family line pose no legal or logical problems; they are participat-
ing in simple patrilineal succession.

As for the children of the woman by her first husband,
according to a ruling of 1921 the uxorilocal husband had absolutely
no kinship rights over them at all (INS; 122). A husband in this
type of marriage is also in a more subordinate position in the
inviting household than in the first type of uxorilocal marriage.

He becomes a family dependent and must obey the elders Of the family
or the family head. He cannot independently manage the affairs of
the family, and an important characteristic was his inability to _
live other than where his wife did. According to a court ruling of i
1908, moreover, he was barred from becoming head of the household

into which he married (INS; 119-20).

83

The status of this kind of uxorilocal husband differed from t
the first, though, in that there was no question of his being
regarded as adopted into the family. He Obtained no membership
status in the inviting family's kinship group, nor did he lose
membership in his own (according to a 1910 ruling; TMS: 120).
Although he gains no rights to the new family's estate, he does, as
long as he resides with his wife, lose his rights as c0parcener in
his own family's estate. He does not usually take the inviting
family's surname, and is considered to have only an affinal rela-
tionship to the wife's natal family, not a blood relationship (or
even "quasi-blood relationship”). In this respect this form of
uxorilocal marriage is closer to ordinary marriage than is the first
form. With respect to the inviting family, that is the family of
the wife's first husband, the second, uxorilocal, husband is held to
have no legal kinship relationship at all. While he is resident
there, he is subject, of course, to the elders' rights and the
household head's rights of that family, but after the death, for
instance, of the woman who is the center of this relationship, the
uxorilocal husband is held to have no longer any standing as a
person with kinship rights in that family (according to a 1921 court
ruling; TMS: 120).

As far as property rights are concerned, this uxorilocal
husband's rights were much the same as those of the other type. He
had rights of management and disposal over his own property, and

at his death the property was inherited by the sons following his

84

line, not the sons following the inviting family's line. If he had
no sons, though, the property went to the inviting family.

He had few rights in the prOperty of that family, but he
did not lose his position as a c0parcener in his natal family, if
and when he returned. What rights he did have in the estate of the
inviting family were subject to the limitation of the contract and
its fulfillment by him. These rights were generally limited to
helping in the management of the estate and the domestic affairs of
the family. He had no rights of inheritance nor had he any inde-
pendent rights to management or disposal of the estate (INS;

121, 122).

Polyandry

At this stage, the discussion leads naturally into a des-
cription of a number of other "irregular" forms of uxorilocal mar-
riage. I include these not for their curiosity value but because,
though they are and were statistically rare compared to "ordinary":
uxorilocal marriage, they were widespread geographically and are V
important to fill in the complete spectrum of variation on some
basic themes.

We have seen that one form of uxorilocal marriage took place
when a woman's first husband died and for a variety of reasons she
preferred to (or, more likely, had no choice but to) remain in her
husband's family and bring in another husband uxorilocally. We
have also seen that this was an institutionalized and accepted

practice in some areas even when the husband was not known to be dead

85

but had been missing for some time (see "companionate marriage"

above, and Hui-an Hsien in Fukien referred to in NSS; 1587 where

 

this was called M. shih tsung chao-fu, "inviting an uxorilocal hus-

 

band when the first husband is missing"). The next logical step is
for this form of marriage to be practiced before the husband is
dead--polyandry. In Fukien Province, according to the Chung-kuo

Min Shang Shih Hsi-kuan TaiO-ch'a Pao-kao Lu (Report on the Inves-

 

tigations Into Customary Law in China, MSS),

In Ku-t'ien Hsien wives are quite ordinarily able, with
the husband's permission, to marry in another husband uxori-
locally for a period of perhaps ten years, turning the bride-
price over to the first husband for his use. This is called
"hanging up the curtain" M. kua-chan . Those children born dur-
ing the ten-year period will Be raised by the uxorilocal, second,
husband, but sometimes, if the first husband lacks heirs, it
is agreed that the first son born during the ten-year period
will follow the second husband, but that the rest will follow
the first husband. They refer to each other as "intertwined
brothers" M. chiao-chia hsiung-ti, and if they are equally well-
off, and their mother dies, they will fight over who is to carry
out the sacrifices and burial, paying no attention to popular
opinion. This custom is practised in Ta-tung Hsiang of Ku-t'ien
Hsien, but it is not yet known to have led to litigation. When,
because there are many children, it's difficult to provide the
daily necessities, and the husband finds it difficult to sup-
port even himself, and so another, uxorilocal, husband is
brought in, he is called a "helper" (M. an -t'ui). The general
circumstances are similar to those in the first form, "hanging
up the curtain," but the period Of uxorilocal residence is only
two or three years (NSS; 1585-86).

 

In Fu-an Hsien, in the area of Hsi-pei Hsiang, there is
still an evil custom called T. ak chhe [The first character
.seems to be a dialect character, re ated, I assume, to the
Taiwanese dialect character pronounced T. 2952 which means "to
hire, rent," and which can be used in reference to persons],
"rented wife." Usually it is because party A has no wife, and
does not have the wherewithal to marry one in, and party 8 has
a wife but is unable to support her. The rental price is no
more than ten taels (M. chin), and the period of time varies
from perhaps three to perhaps ten years. The children born
within that period follow A. If she is to die within the time

86

period, then both A and 8 together arrange for the funeral.
When the time period is up, and if the first husband can redeem
her, then it must be permitted. If not, then she remains the
wife of A (NSS; 1602-1603).
In Chekiang and many other areas, this sort of marriage was
called, appropriately enough, "inviting a husband to support a
husband" (M. chao-fu yang-fu); it was described as "frequent among
the lower classes,” and it was specified that if the first husband
had no sons, sons by the second husband could inherit from the first
(NSS; 1530). The custom was also found in Hsia-p'u Hsien, Fukien
(NSS; 1604), and in P'ing-nan Hsien, where the husband was called
"helper" (M. pang-t'ui); the time period was said to be usually
three to five years, and the sons born to the second husband were
at times able to inherit from the first husband. This custom, it
was claimed, had formerly been more prevalent, but had declined in
recent years (NSS; 1606).
In Chu-shan, Ching-shan, and Ch'ien-chiang Hsiens, Hupei
(these areas seem to have been particularly well studied), it was
also called M. chao-fu yang-fu; in Chu-shan Hsien he had to change
his surname to that of the first husband, while in the other two
Hsiens, he did not (NSS; 1653). In Yuan Hsien, Hupei, where it was
described as "quite rare," the second husband also did not change
his name (NSS; 1667).
In Ch'eng-ku, Tz'u-yang, Mien, and other Hsiens of Shensi,
There are women, already with husbands, who invite in
another husband to come and live with them, in order to support
the family. This is called M. chaO-fu yaggffu. This sort Of

custom is frequent in the Hsiens south Of the Han River. In
seeking reasons for this, one finds that in most cases it is

out

87

because the first husband is disabled and cannot make a living.
The couple, after consultation, are forced to do this. There
are also those, however, who are not disabled, but through the
profligacy Of youth have expended the family property, and so
endure humiliation, swallow their pride, and allow the wife to
marry in another husband (NSS; 1702).

In other areas of Shensi,

If a woman is promiscuous, or if her husband becomes
paralyzed, or deaf and dumb with senility, or disabled through
aberration cn- some other sort of illness, and cannot make a
living for his family, so that they become poor, and exposed to
the cold and hunger, and there is no other way out, then after
discussing it and freely agreeing to do so, they can ask a go-
between to seek out a man to marry in and live uxorilocally,
in order to support the first husband. A contract is written
up, called an uxorilocal marriage contract, which clearly stipu-
lates that they cannot mistreat the first husband.

There are also cases of this happening when the first hus-
band is not at all sick or disabled, but has just, through
youthful profligacy, gambling, and inattention to his proper
business, completely dissipated the family estate, so there is no
alternative but to endure the humiliation, swallow their ride,
and allow the wife to marry in another husband (MSS: 1718)

This sort Of polyandrous marriage also was practiced through-
most of Kansu:

Kansu has a special kind of evil custom called M. chao-fu

an -fu. For instance, if A has married in B as his wife, and
if later he becomes debilitated from age, or he is struck down
with a serious illness, or he becomes poor to the point that he
cannot any longer support himself, then the wife, 8, can, with
A's permission, marry a man C into the household to live uxori-
locally and support the first husband. Everything in the house-
hold is his to use. If later, sons and daughters are born,
which father they should follow is decided through oral agree-
ment by the two sides. The relatives of A cannot interfere in
this. This form of marriage is definitely not considered shame-
ful by society (NSS; 1769).

A commentary is appended to the above, listing the names

of those making the reports on which the notice is based, and the

areas within Kansu reported on. It then continues to the effect

that “according to the investigations of this Committee, this evil

cusumiis found practically everywhere in Kansu, and if it is to be

88

forbidden by law, this will have the unfortunate effect of imme-
diately cutting off the sources of support for large numbers of
people."

I do not know of any descriptions of this form of marriage
from Taiwan. How prevalent it was in the past is now impossible to
tell. It was no doubt rarely encountered, but it did occur. The
collection of legal documents and commentaries titled (in Chinese)

the Taiwan Szu-fa Jen-Shih P'ien contains at least one contract for a

 

M. chao-fu yang:fu marriage:

The executor of this contract of chao-fu yang-fu marriage,
Wang Yun-fa, married, as wife, the daughter of Li San named
Hsiu-liang, twenty years of age this year; they have been mar-
ried four years. Hsiu-liang cares every day for her parents-
in-law, is very obedient and filial, and in her regulation of
the family, one seldom hears any recriminations; having married'
a wife like this, one can really set one's mind at rest.

In recent years, however, the executor Yun-fa has become
ill and disabled, is physically less than a man, the family
has become impoverished, cannot meet its expenses, and are unable
to borrow money anywhere.

Though they are poor, they are still alive. Thinking that
"of the three unfilial acts (leaving no descendants is the
worst)," and that youth will not return, husband and wife have
discussed the situation day and night, and have decided that
indeed there is no other way. Wishing to preserve virtue, and
thinking that a family Of several persons will hardly go without
food, they have decided that only with a chaO-fu yang-fu will
they be complete. Therefore they have asked in a gO-between,
to discuss the matter of the marriage. The go-between has found
Wu Chin-wen, the oldest son of Wu Chiu, to marry in uxorilocally, .
to form man and wife. They have agreed that they do not want
any bride-price, but that each month he must provide twenty
dollars to cover his expenses.

Should sons later be born, and grandsons in turn, then
whether there are many or few, they shall all sacrifice to the
two family lines.

 

89

The two sides are willing and in agreement, and neither
will have later regrets. Fearing that there is no evidence of
the spoken word, one copy of this chaO-fu yang-fu contract is
drawn up as evidence.

Scrivener Cheng Ju-shui
GO-between Mrs. Ch'en nee HsU
Witnesses Wang Chin-fa
Wang T'ien-fu
Executor of the chao-fu yang-fu contract Wang Yun-fa

T'ung-chih 9 (1869), third month (Taiwan Szu-fa IV: 571).

 

Analytic Model for Uxorilocal and
Polyandrous Marriages

One of our informants, a T. tgj;§g_(M. tai-shu) or "scrivener,"
one who makes a business of writing up contracts, other official
documents, and may also become a mediator and informal legal counsel,{
said that uxorilocal marriages are generally more likely to end in
divorce than regular marriages. Usually this is said to be because
the woman's parents are likely to be critical and fault-finding with
respect to the uxorilocal husband. He said that for the woman's side,
her parents want to ensure descendants and/or support in their Old
age. The motives on the man's side are primarily economic. He can,
by marrying uxorilocally, set up a household and family without most
of the usual expense. Basically, the scrivener argues, there is no
great difference between uxorilocal husbands who take their wife's
surname and those who do not. The decision is an individual one,
made by the parties involved in each case.

Thescrivenerhsarguments about the instability of uxorilocal?
marriage should sound familiar to those interested generally in :

kinship and social organization. Buchler and Selby say that:

9O

. . . Our concern must be directed at the transference of rights
between groups linked by the conjugal dyad. It seems that cor-
relates to . . . anomalous structural features are to be found
in the degree to which rights in uxorem (over a woman ggg_wife),
and rights in genetricem (over a woman ggg bearer of children,
or over the progeny of a union) are transferred from the family
of orientation of the woman to the man or his group with whom
she coresides (1968: 28).

 

All of this was writtenirithe context of a discussion of the
Nayar case, which is in many ways different from the case of uxori-
local marriage in Chinese society. They go on, however, to make a
more general point, and to suggest a cultural universal. Whereas
Gluckman correlated divorce rates with degree of patriliny (Gluckman
1950: 166), Fallers proposed a correlation with the degree to which
women were absorbed into the man's lineage:

Where a woman, either through the complete transfer of her’,
child-bearing properties or by other means, is socially absorbedf
into her husband's lineage, patriliny tends to stabilize mar-
riages; where a wife is not so absorbed and thus remains a
member Of the lineage into which she was born, patriliny tends
to divide marriages by dividing the loyalties of spouses
(Fallers 1957: 121).

Buchler and Selby point out that, following Levi-Strauss, it
does not logically matter whether we regard the exchange from the
point of view of the male or female.

From the female point of view we would state the conclusion

as follows: the degree to which the female or her group retains
rights in genetricem and in uxorem will be positively correlated

with the degree of instability in the marriage (Buchler and
Selby 1968: 29).

 

Thus, the many indications of a tendency to relative insta-
bility in such marriages are not surprising.
Buchler and Selby's work has some other points of interest

in this connection. Reporting on Selby's research in the Oaxaca

91

Valley, they consider postnuptial residence, and discuss a number of
points which should sound familiar to the reader of these pages.

Although a statistical summary of residential forms dis-
plays an amazing disparity, the jural pattern of postnuptial
residence in this village is virilocal. To an actor (and his
family) postnuptial residence (which will permit access to the
household resources by the coresident) poses problems in strategy,
which are seen to be resolved appropriately only in the context
of a number of situational factors. For a man there are two
alternative patterns of postnuptial residence, virilocal and
uxorilocal. Virilocal residence is the preferred form, not in the
statistical sense that it is the ”normal" form of residence
(which it is), but in the sense that positive value attaches to
it. Associated with residential choice are a bundle of activi-
ties . . . undertaken by both families whereby the hand of the
potential spouse is "sought"; arrangements are made for the
wedding; and financial commitments are made by both sides in
deciding how much of a contribution will be made by each side to
the expenses of the wedding. . . . By arranging to seek the hand
of the groom, the family of a potential bride can arrange an
uxorilocal marriage.

As mentioned earlier, the jural or normative rule is viri-
locality, to which positive value is attached. Positive value
is also attached (by actors) to the resources of the household,
and each actor (actor in the sense of negotiating family)
attempts to maximize the values of the transaction. A male gains
an initial "profit" by arranging a virilocal marriage. The com-
parative "loss" on the part of the bride is made up by the fact
that she marries into a household with greater resources than
her household "of orientation." Thus jurally we have a pattern
of what the anthropologist would call hypergamy, and the native
actor an equivalent exchange.

Consider the Opposite situation--that of uxorilocal mar-
riage. Uxorilocal marriage is not evaluated neutrally, but a
negative value is placed upon it (from the point of view of the
male actor). Thus a proportionately greater degree of incentive
must be injected into the transaction'hiorder to make the exchange
between the two families a fair or equal exchange. Thus in the
case of uxorilocal marriage the differential will be prOpor-
tionately greater than in the case of virilocal marriage (Buchler
and Selby 1968: 49-50).

Almost all of this could be in reference to uxorilocal mar-
riage as found in Chinese society. It is gratifying, if not entirely

surprising, to see such an interesting example of general sociological

92

principles in operation in very different and geographically very
separate societies.

They go on to suggest that the "incentive" mentioned above
can be thought of in much the same way as "relative deprivation";
"economic differential sufficient to provide adequate incentive (for
uxorilocal marriage in this example, but by inference to virilocal
marriage as well) will vary directly with the economic status of the
'lower' party in the exchange" (1968: 50). Data are presented to
show that of eighty-four households surveyed, the family that sought
the hand of the spouse was always richer than the other family in
the transaction. Though I do not have at my disposal all of the
economic information necessary to make the same argument about the
general form of uxorilocal marriage in Chinese society, I strongly
suspect that similar results would be found. This would be an
interesting topic for further research.

Buchler and Selby argue that this kind of decision-oriented
approach provides a bridge between jural accounts on one hand and
statistical summaries on the other, but it is then necessary to

. . have recourse to principles that underlie behavior,
as formulated by the native- -speakers, and implicit in their
behavior. This requires an ethnoscientific definition of the
behavioral parameters, an approach that requires of the infor-
mants that they indicate the behavioral segments they regard as
critical to ordering their behavior as well as the situational
determinants of decisions in the context of the behavioral seg-
ment. For example, it appears that residence per se is not an
appropriate frame of reference for analysis. It is subsumable
under the more general domain of analysis of status differen-
tiation and exchange. The critical observed behavior is not
the physical translation of a member of one family to the beha-
vioral environment of another, but rather the negotiation of

sexual, physical, and economic rights involved in the process
of the pedimiento (1968: 52).

 

93

"Pedimiento" refers "to that set of activities undertaken by

both families whereby the hand of the potential spouse is sought"

(1968: 49).

The underlying principles of exchange in this model of

uxorilocal marriage in the Chinese case (following Buchler and Selby

1968: 52) are as follows:

Each party to the exchange contributes what each thinks are ;

approximately equal values, defined in the following ways:

1.

A positive value is attached to postnuptial residence

(that which determines what household the actors are in,

and what position they have in the household, and hence

their rights and duties) that follows the pattern in the

dominant ideology (viri-patrilocal in this case).

A positive value will be attached to the resources

under the control of the households involved.

A positive value is attached to allocation of children

(particularly, but not exclusively, sons) to a certain

surname and to sacrifice to the ancestors of a certain

line, which follows the pattern in the dominant ideology

(patrilineal in this case).

The value of the resources attached to each actor will

be calculated in two ways.

a. The differential evaluation of resources (economic
and other).

b. The relative position of each actor in terms of

locally relevant socioeconomic divisions.

94

Knowledge, then, of the socioeconomic variables and the
relevant cultural principles underlying such exchanges should make it
possible to "generate," or predict, what line of action will be taken
in particular situations.

This model, as is noted, is one of decisions made in
exchanges involving postnuptial residence. We have seen, however, :2
that in Buchler and Selby's view, and my own, what is really important:
is not the residence itself, but status differentiation and exchange; ;
the exchange of sexual, physical, and economic rights over and in I
persons. From this it should be clear that this model, with suitable
modifications, should also work to generate or predict other kinds of
marriage and, I will argue, the various kinds of adOption as well.

In a sense this is "common sense." It should be no sur-
prise that such things as marriage and adOption, actions with social
and economic consequences, should also have social and economic con- ,5
straints. What is important, I suggest, is that there is a relatively
small number of principles which are interrelated in particular ways,
principles which can, in specifiable circumstances, generate particu-
lar, observed, social forms. In this way a kind of unity can be
postulated for what are otherwise rather disparate and seemingly
unrelated forms, and thus understanding, in at least one sense of

the term, has been advanced.

Double Generation Marriage_
When a man died leaving a number of children and little to
support them, we have seen that one way of dealing with the situation

was through uxorilocal marriage. A second husband was married in for

95

the widow to support her and her children. In certain circumstances
she might marry out and take her children with her. This was, in
fact, a common enough practice that it was associated with particular
local terminologies in a variety of geographic areas.
When a woman took her children with her to another marriage
it was quite common to differentiate between those of her children
who continued their former kinship connections (M. huo-tai, literally
"bring alive," or "bring and maintain their former status"), and
those who did not (M. szu-tai, literally "bring dead," or "bring and
extinguish their former status"). In Shan-hsi it is recorded that:
When a widow brings along her former husband's children
when she remarries, they are colloquially divided into M. huO-tai
and M. szu-tai. NO matter how long she has been widowed, those
children who do not change their surnames, and for whom it is
clearly stipulated in the marriage contract that they will
later return to their original kinship line, are called
M. huo-tai. The M. szu-tai are those children whose surnames
are changed to that of the second husband. Their given names
remain the same, and they inherit equally with the step-father's
own children. If he has no such children, then they can succeed
him (NSS; 1436-37; see also pp. 1447, 1450, 1454, 1723).

In another area of Shan-hsi it is said that step-children in

such situations called one another M. ke-shan hsiung-ti chieh-mei,

 

literally "siblings separated by mountains" (NSS; 1447).
In parts Of Chekiang this was called M. sui mu chuan-chia,
"following the mother in remarriage":

. . When a widow has no children or has children but the
family is poor and it is difficult to support them, she is per-
mitted to remarry. As for her children, if her former husband
has parents and brothers, she can leave the children with them.
If not, then she may take them with her. If there are people
in the first husband's family who can support and raise them,
but the mother and children cannot endure living with them,

. . in thatcase she may remarry and take the children with her.
Furthermore, for the children there may be a specified time

96

limit, after which they return to their original kinsmen, or
they may stay permanently, as integral parts of the family
(NSS; 1567-68).
In parts Of Fukien the former category of children, or at
least those who kept the former husband's surname after moving into

the step-father's household, were called M. yg;t'ung tzu, a term

 

that I cannot satisfactorily explain.

In a society in which parental arrangement of marriages was
taken for granted, and such practices as the adopted daughter-in-law
were so common, it is not surprising that a common variant on the
pattern under discussion was one of I'double generation marriage";
that is, the marriage of a man and a woman, and, at the same time,'
the marriage to each other of their children by their respective
former spouses.

In Shan-hsi and some other areas, this was called "brother-
sister marriage":

When a woman remarries, and takes her children with her,

she can then marry one of her daughters to a son of the second

husband. This is called M. hsiung-mei hun "brother-sister
marriage" (NSS; 1436).

 

When a widow remarries and takes one of her own daughters 1
along as wife of a son of the second husband, this is for cur-.
rent support, and future marriage. This is quite common among{
the people. It is necessary, however, that both sides agree, =
that the services of a go-between be secured, that gifts and ,
marriage documents be exchanged, and that a regular betrothal /
be carried out. When the girl has reached marriageable age, I
and the marriage has been completed, then the two families will
interact with each other just as do ordinary affines. The woman
who has remarried, however, must not again go to her former
husband's family, so as to symbolize their limited relationship
to her (NSS; 1450-51).

97

In parts of Chiang-hsi, a betrothal contract was added to
the marriage contract (of the parents) in such cases, or a separate
marriage contract was drawn up (NSS; 1515).

In Fu-ting Hsien, Fukien,

. . . When a woman's husband dies she commonly takes her
children with her when she remarries. Sometimes her son might
marry a daughter of her second husband by his former wife, or
her daughter might marry a son of the second husband by a for-
mer wife, so that the parents and the children, respectively,
marry each other. Colloquially this is called M. t'ao-hua chia-
.gNg, "grafting a peach blossom onto bamboo" (NSS; 1603).

This last account is most interesting, as this is the phrase

that is used in such cases in Taiwan (T. tho-hoe ngpeh-tek; see, for
instance, Wu 1969: 142-43). The peach blossom is a common symbol for

the female, and the bamboo here is representative of the male.

The Levirate, Sororate, and Cousin Marriage_

Thus far, a number of alternative solutions have been con-
sidered for the problem of what to do when a woman with young children
is widowed. A logically obvious solution which has not yet been
discussed is that the dead husband's position in the marriage be occu-
pied by his brother. This institution is commonly known as the
levirate, and has many advantages. Because the sociological situa-
tion does not change in terms of the relationship between the two
affinally linked families, no money or new presents need change
hands. The brother of the original husband gets a "free" wife, and
neither familyhas the problem of what to do with the woman. Exactly
whose right it was to marry her Off again, however, and so receive
the brideprice and marriage presents, was in any case variable and

subject to dispute. If there were children, the levirate could be

98

an ideal solution. The children would be kept in the first husband's
family, and his kin line would be in no danger of dying out. The
popularity and acceptability of M. kuo-fang adoption, adOption within
the patrilineal kingroup, meant that it was quite natural for the
second husband, the original husband's brother, to care for and sup-
port his dead brother's children and have them share in the division
of his estate upon his own death.

The problem with the levirate was that it was seen, at
least in the value system of the elite classes, and in the written
classics, as a rather alarming sort of incest. It was not the worst
fonm, to be sure, but it was, nevertheless, abhorrent. One writer
comments on slave girls that they might find themselves being pursued
by two brothers. The idea of two brothers having sexual intercourse
with the same woman, even if serially, was disgusting, as, in greater
degree, was the idea of father and son in the same situation.

Functionally, of course, such a prohibition would have
helped maintain harmony within a household in which more than one
married brother lived. Such a function was obviously more important
in gentry, elite, circles where such a family size was possible,
than among peasants and other poor peOple, who had little chance to
(and perhaps, in some cases, little desire to) develop and maintain
such a coresident kin group.

Structurally, one can also see this prohibition as an index
Of the degree to which a woman was seen to be incorporated into the

kin group of her husband. The act of marriage quite clearly reduced

l1£2r~ kinship ties to her natal family, and just as clearly established

99

her within the husband's group. In a sense, then, she is incorporated
into the kin group to such a degree that her sexual availability to
the males of that kin group becomes ambiguous.

Such a prohibition is not, however, common in other societies
with patrilineally organized corporate kin groupings. Robin Fox
says:

It is interesting that the exogamic ban in China extended

to a widow of an agnate. Thus, if a man died his lineage had
to marry his widow off to another lineage (if she was allowed
to remarry). Usually, in patrilineal societies the Opposite
has been the case; once the lineage has Obtained a woman it
hangs onto her. Very Often, for example, she is married to her
dead husband's brother; a custom known as the levirate. China
however was an exception to this, and I do not know the reason
why this should be (Fox 1967: 117).

I suggest that it is not really a big problem to explain the
absence of the levirate. It was, in fact, not absent. Fox says,
in another context, of China that "in such a complex society there
was room for flexibility" (1967: 116). That is certainly true,
especially given the nature and functioning of the class system.

From the point of view of the society as a whole, China was able to
have her cake and eat it too; there was a prohibition on the levirate
and there was the levirate too. There was far more variation and
heterogeneity in China than is usually depicted. In the elite value
system it was quite easy to set up a rule which prohibited the prac-
tice of the levirate, and get it cannonized in the classics. If such
a prohibition "helped" anyone at all, it was those in the elite

class with large corporate kin groups to maintain. It could then

also be used as a measure of elite status. Any family which, for

lOO

economic or other reasons, found it advantageous to practice the
levirate defined itself as nonelite by so doing.

The lower classes were generally left to go ahead and follow
whatever "disgusting" or "evil" customs they wished, so long as the
social, political, and economic status quo was not threatened. In
fact, analytically, it was in the interests Of the elite to make sure
that there would always be a clearly defined lower class, one defined
so as to be clearly different from their own.

Surely one of the reasons that there was a ban on marriage
to widows of agnates in China, while other societies with patri-
lineally organized kin groups tend to follow the levirate, is because
China had an elite class and a long and prestigious literary tradition
which was shaped by, and helped support, that elite class. In other
societies, once a lineage obtains a woman it holds onto her because
to do so makes good socioeconomic sense; it is a rational use of
resources. To do otherwise is expensive, in a number of senses of
that term. One can argue, then, that the ban on the levirate in
China is a form of conspicuous consumption. So, also, are some other
. customs_which are connected with this in various ways, like the ban on
the remarriage of widows to anyone, footbinding of women, and so on.
These things are expensive, but they were also necessary for a mini-
mal definition of membership in the elite class. This argument sug-
gests that many, if not all, of the social forms forbidden by the
elite and in the classics should be found actually to be the practice

in nonelite classes. One of the things I am attempting to demonstrate

101

is that this is in fact so. The levirate did exist, as we shall see,
but seemingly was not prevalent among the elite.

The Min Shagg_$hih Hsi-kuan TiaO-ch'a PaO-kao Lu says, for
parts of Shan-hsi, that

If the Older brother dies, and his younger brother has no
wife, then the younger brother can take his sister-in-law as
wife. If the younger brother dies and the older brother has
no wife, then he also can take his sister-in-law as wife.

The compiler's comment is as follows:

This marrying the wife or concubine of a relative is not
permitted in the law. Moreover, in regard to right human
relationships and public morals, this is truly an evil cus-
tom (NSS; 1418).

In another part of Shan-hsi,

When an older brother dies, leaving a widow, the younger
brother can, with the services of a go-between, marry his
widowed sister-in-law. This is colloquially called M. chieh-
th hun "continuing marriage" (NSS; 1434-35).

In yet another part of the same province,

When a family is poor, and among several brothers one who
has married dies, then one of the brothers who has no wife can,
after discussing it with the woman's family, marry her as his
wife. This is called M. t'e-pieh th-ch'in ("special remar-
riage").

The compiler's comments:
This custom was reported by the Hsien Magistrate Chang
Liu-hsing, and is the same as that in . . . [other named
Hsiens]. Marrying the wife or concubine Of a relative is
clearly prohibited in the current Code. These customs in these
Hsiens cannot be regarded as having a beneficial effect on pub-
lic morality (NSS; 1455).
The same custom was found in parts of An-hui, and is reported
with the comment that "no custom is worse than this" (NSS; 1477-78).
In parts of Chiang-hsi the levirate was called M. sheng-fang, lit-

erally "elevating to another branch," when it was the younger brother

102

who died, and his older brother married the widow of the former.
"But this kind of evil custom is found only among poor families"
(NSS; 1505).

In Kan Hsien of Chiang-hsi Province

The evil custom of the levirate, which is found among the
peOple, is most flourishing in Kan Hsien. For instance, if
among several brothers only one has a wife, and he dies, then
his wife is then re-married, to his older or younger brother.
If the brother then dies, she can be again remarried in this
fashion. A certain previous Hsien Magistrate strictly forbade
this custom, and it has decreased somewhat, but in remote vil-
lages it is still quite common. Also, some are forced into
this by the excessive cost of brideprice and wedding presents
(NSS; 1510).

In Lin—hai Hsien, Chekiang, the levirate was said, again,
to be found among people of poor families, and was called M. shu-sao

chieh-mien, "joining brother-in-law and sister-in-law" (NSS; 1548).

 

For another part of Chekiang a similar account is found,
and the custom is compared, implicitly, to uxorilocal marriage:

The custom of the "estate-managing uxorilocal husband"
(M. tso-ch'an chao-fu) is found in various parts of Chekiang.
In P'ing-hu Hsien it is found as well, and the second husband
is called M. tien-fu, "supplanting husband" or M. chuan-ch'e chu
"axle." Moreover, there is also the custom of a brother-in-law
marrying a sister-in-law. For instance if a woman's husband
dies, and her brother-in-law is adult, and the family is in dif-
ficult circumstances, then either with the woman's parents-in-
law making the arrangements, or the two persons themselves
agreeing, the two may become man and wife. This custom, however,
is found only among the lower classes of society (NSS; 1557).

 

 

In yet another part of Chekiang (where the custom is called

M. chuan-ch'in, "transferring the relationship"), it is said that it

 

may be practiced because the family is poor and cannot afford another
marriage, or it may because the two peOple have already formed an

illicit relationship (NSS; 1559-60).

103

It is claimed that in areas of Hu-pei (where the levirate is
called M. lun-hun, "regular, or natural, marriage," which must be a
euphemism for M. lun-hun, "rotating marriage") no marriage document
is used, and no go-between, and the marriage may be the result of the
free agreement of the two parties, or the woman may be forced to
make Obeisances to the ancestral tablets, thus concluding the mar-
riage (NSS; 1615; see also 1668).

In a number of places in Hu-pei thelevirate was called

M. chuan-fagg, "transferral to another branch," and in Wu-feng Hsien

 

it is said that it was not permitted that a man marry his own brother's
widow, but he could marry the widow of some other member of the same
'generation within his lineage (NSS; 1629-30; see also 1653).

There are also reports from various areas in Hu-nan (NSS; .
1679; 1681), Shen-hsi (1744, 1745, 1756), Kan-su (MSS: 1775, 1801), :
Chih-li, and so on. It is curious, though, that I have yet to find
references to its practice in Taiwan. I do not know why this should7
be; I would expect it to be practiced, and given the zeal with which
many writers have pointed out "evil" and “low" customs in Taiwanese
society, I would expect it to be recorded if it had been observed.

A discussion of the levirate quite naturally brings to mind 7
its logical companion, the sororate. This was evidently much less
common than the levirate was, which should occasion no great sur-
prise. If it was in the interests Of a patrilineally organized
corporate group to "hang onto" thewomen it had married in, because
it was expensive to do otherwise, it was also sensible for the same

group to marry off its own women in a regular way, so as to obtain

104

the bride-price and extend its affinal ties. In the sororate neither
of these goals is likely to be furthered. Still, the sororate evi-
dently did exist.
In Yu-yfi Hsien, Shan-hsi,
If A marries the oldest daughter of B, and then she becomes
ill and dies, and B has a second daughter, he may marry her to A
as his second wife. This happens in those families which have
good and close relationships (NSS; 1448).
This was called M. th-ch'in, "continuing the relationship."
In an area embracing five Hsien in Hunan,
. . If A has two sons, and B has only one daughter, and
[the families of] A and B have long had affinal ties, or they
are particularly good friends, and B' 5 daughter is of an age
with A's Oldest son, then they may marry. If A's oldest son
later dies, and A does not wish to break off the good relation-
ship, then he can ask that B's daughter be married to his (A's)
second son.
Also, if A and 8 each has several sons and daughters, and
B marries a daughter to a son, of the right age, of A (A's
other sons having already been betrothed or married), and later
8' 5 daughter dies, A can ask to have 8' 5 second daughter Himar-
riage to his widowed son. Both of these are attempts to con-
tinue the relationship in the face of change. In these various

Hsien, they are colloquially called M. huan-ch'in, "changing
relations" (NSS; 1681).

There is an exceedingly brief account from a Hsien in
Shensi (where it is also called H. th-ch'in, "continuing the rela-
tionship“), and one from the present Hupei (the term for it is the
same), wherein it is said to be no different from ordinary remar-
riage (NSS; 1293).

I think that it is quite significant that out of perhaps
1,500 reports, in over 500 pages, on kinship, marriage, and inheri-
tance, the Chung3kuo Min Shang Shih Hsi-kuan Tiao-ch'a Pao-kao Lu
has only a few notices on the sororate. Although they are very few,

it seems significant that the element that receives emphasis is that

105

of the existence and maintenance Of good relationships between the
two families, something that is not emphasized in the very many more
reports on the levirate. The sororate, unsurprisingly, seems not

to be a natural outcome Of structural features of Chinese kinship
organization, but is there to be used when other, particular, cir-
cumstances outweigh its disadvantages. Wolfram Eberhard (1968: 181)'
gives a number of reports of the institution in various areas and
time periods, but on the brief data he presents, they seem not to be
institutionalized patterns, but rather, isolated and idiosyncratic
cases. As with the levirate, I know of no reports on its practice in
Taiwan.

The levirate and the sororate are particular examples of
preferential marriage patterns. A natural question arises about
the occurrence of such patterns in Chinese society.

Throughout Hu-pei, but particularly in Chu-shan Hsien we
find

. . the custom of two families intermarrying. . . . As for
instance, when a girl of family A is married into family 8, and
family 8 also has a girl, or a cousin, who marries into family
A. There are even cases of such intermarriage when the genera-
tions are not matched, as when family A marries in a girl of
family 8, and then a paternal aunt, or some such relative, of
A is given to the brother of the girl of family 8 as wife.

In the commentary it is pointed out that though the original
report was for Chu-shan Hsien, this custom was found throughout the
area. It is also pointed out that except for intermarriages in which
the generations are not matched, this custom was not contrary to the

current Code (NSS; 1609).

There is a companion report about cousin marriage:

106

In Chu-shan Hsien matrilateral cross-cousin marriage is
most common, while matrilateral parallel cousin marriage is
also not uncommon (NSS; 1609).

Patrilateral cross-cousin marriage, however, is clearly prohibited.

For four Hsien in Hu-pei, it is said that

The children of a mother's brother can marry the children
of a father's sister, and the children of two sisters can
intermarry. In the customs of Chu-hsi and Yun Hsiens, except
for the children of two sisters, who can marry, only the daughter
of a mother's brother is permitted to marry the son of a father's
sister. This is colloquially called M. chih-nU sui ku "the
niece following the aunt (patrilateral),“7and it is not permitted
for the daughter of a father's sister to marry the son of a
mother's brother. This latter is called M. ku-jou huan-hsiagg
"flesh and bones returning home" (NSS; 1624).

In five other Hsien of Hupei it is said that all of the
above forms were permitted, but in one additional Hsien (Ku-ch'eng),
as in Chu-hsi and Yun Hsien, as noted above, the daughter of a
father's sister was not permitted to marry the son of a mother's
brother. The colloquial saying was M. ku-jou huan hsiang, chia pai
jen wang, "If the flesh and bones return home thus, the families will
be destroyed and the people lost" (NSS; 1650-51).

In Luang-p'ing of Jehol (present Ho-pei), very near Peking,
it was said,

If a father's sister's son took as wife his mother's
brother's daughter, or if the OffSpring of two sisters marries,
in both cases this was called M. ch'in-shagg chia ch'in "adding
relationship upon relationship" (NSS; 1808).

The cousin-marriage forms mentioned above all preserve
surname exogamy, but this itself was not inviolable. Intra-surname
marriage is specifically mentioned as practiced in the Hopei-
Tientsin region (and most of the provinces north of the Yangtze

River)(NS_S_: 1292), various parts of Shansi (1430, 1449), parts of

107

Chiang-su (1478, 1482), Kuang-tse Hsien in Fukien (1585), all of
Hu-pei (1608; see also 1618, 1639), Ch'an-an Hsien of Shensi (1739),
and all of Kansu (1768). In most, if not all,<rfthese accounts, it
was specified that such marriages were permitted and customary

as long as the persons were not actually of the same kin line, or

were not too close genealogically.

Consensual Unions

 

In addition to forms labelled "marriage," there are also a
number of marriage-like forms to consider. Because of their nature,
it should come as no surprise that there is little information about
them, historical or otherwise.
Mention has been made of the relationship known as M. ta-huo
("companionate marriage"),as found in the area now known as Liao-ning.
This was a form of uxorilocal marriage which differed from the usual
forms in that it was potentially temporary. If the woman's first
husband were to return after being missing for some time, she would
return to him. Various forms of wife rental and pawning were also
characterized by their temporary nature.
Another form of consensual union was called M. pan-lac ("com-
panion in Old age"). From Szu Hsien in Anhwei Province,
Old men without wives take in an old woman as a companion.
She is not a wife and not a concubine, and there is no marriage
document or any other conditions. This is called "companion
in old age." After the man's death, these women "companions in
old age" are constantly embroiled in lawsuits with the sons and
grandsons of the man (NSS; 1485).

This was brought to the attention of and recorded by the

authorities when one of the lawsuits came before the court in 1915.

108

Hsfi K'O reports a somewhat different custom from the Shanghai
area (or at least from the Wu dialect area) in his Ch'ing Pai-lei Ch'ao
(H50 1920: XVI, 43); the relationship called M. cha pling(pin[_
pijnlp'eng)-t'ou, which is translated by Mathew's Chinese-English

 

Dictionary_as "to form an illicit connection; to live as man and wife

 

without regular marriage." Mathews says that 9N3; usually read "ya,"
means in Shanghai dialect "to join, to crowd around."
L. C. Arlington points out of the word M. p'ing that

The modern meaning of "illicit intercourse with maid-
servants" seems to be derived from the custom in vogue during
the Han dynasty forbidding sexual intercourse during periods of
fasting, for which the penalty was a fine of four ounces of
silver. . . -(1922: 56).

This penalty was called M. p'ing.
Whatever the origins Of the term, it is more widespread than
just the Wu dialect area. HsU's account is as follows:

M. cha p'ing-t'ou means a man and woman who form an irregu-
lar union, who act as man and wife, and who, moreover, live, eat,
and drink together. It also applies to those who only form a
liaison, but do not live together. As for this term M. g'ing-
t'ou, it is used reciprocally by the man and the woman. After
such a relationship has been established, and recognized by each
side, it may also be recognized by a third party. For instance
one can say "He (she) is my ijng3t'ou," or "He (she) is her
(his) pfing-t'ou." This term "p'ing-t'ou" particularly pleases
those Of a literary turn of mind. In the Peking dialect we call
it wai-chia, "house on the side" (this term, however, implies
that there is a fixed house or apartment, while the term p'ing-
t'ou does not necessarily imply that; this is the only, slight,
difference in meaning).

M. che p'ing-t'ou is the term that is used when, either
because of some actual clash of opinion the two persons decide
to part, or their interests become Opposed, and they go their
separate ways. This is just as in business, when there is a
split in the stock of a joint-stock company (M. che-ku). After
such a split the two persons look upon each other as strangers
(H50 1920: XVI, 43).

 

 

 

109

With admirable sociological acumen, HsU establishes that this
is an institutionalized form, recognized by society; the relationship
has a clear beginning and a clear end, and it links two persons in
socially recognized, named roles. It is also clear that it is an
irregular form, clearly distinguished from marriage. It is a pity
that we have no further information on the legal status of the parties,
or of children born to the union.

Ch'en Ku-yuan divides marriage into formal and actual. The
former are those with the form of marriage but not the substance
(such as spirit marriage, or ghost marriage, in which one or more of
the partners is dead) and the latter are those with the substance of
marriage without the formal aspects. As an illustration Of this last

type, he discusses the custom of M. piing-t'ou, which he also calls

M. p'ing-tu:

As for this 'in -tu, today we use this term for what is
in effect a marriage, But the ShuO-wen cites the Han Code to
the effect that "The people of Ch'i call fornication with
someone's wife or female slave "p'ing," and the Kuan - un says
"If one has intercourse with a woman during fasting then the
penalty is a fine of four ounces of silver." Both of these
are different in meaning from the current usage. Today's
meaning of the term p'ing is perhaps equivalent to the old
M. eh-ho, "illicit un1on," etc. . . . Those examples from the
past w ich seem most like the modern p'ing-tu, and excepting
those which are obviously just illicit cohabitation, are not
fornication (M. chien), and yet are without benefit of a mar-
riage ceremony; in this they are no different from p'ing-
tu. . . .
—_' Generally speaking, for the last several thousand years
only the standard betrothal and marriage have been recognized,
and p'ing-tu, not having included the proper ceremonies, was
treated in the Ij_and law both as simple fornication. If we
look at the various historical legal codes, we can see that in
some the intercourse of man and woman without the benefit of the
rites was punishable by death. In others fornication with an
unmarried woman could be punished with one and one half years'
banishment, and in some cohabitation without the ceremonies of

 

110

g the II_was called fornication or mutual enticement, and had its
specified punishment. Today p'ing-tu is simply regarded as a
non-legal marriage, but it is still regarded as in effect a
marriage. Thus, though there are none of the legal protections
and guarantees of a legal marriage, yet the persons involved in
the union are not criminally culpable.

Customs, however, are not united in the terminology used
for this relationship. Cohabitation without marriage is cer-
tainly called p'ing. For a married person to be involved with
someone outside the marriage in a criminal relationship is also
called M. cha p'ing-t'ou. On the other hand, in some customary
usages those who are united without the proper ceremonies may,
in the eyes of society, be regarded as being formally and prop-
erly married, and moreover the term 'in is not then used.

For this reason some scholars use the atin term concubinatus

to translate piing-t'ou; this is not entirely the same as con-
cubine3 the English term used to translate M. ch'ieh (Ch'en—r937:
112-14 .

 

vIt will be noticed that the quote given by Arlington about
the origin of the term pIIgg having to do with an old prohibition on
intercourse during fasting is the same as that given by Ch'en.
Arlington, however, attributes it to the Han Code, while Ch'en
attributes it to the Kuang-yun (an Old rhyming dictionary, but con-
siderably later in origin than the Han). Given the similarity of
the passages, it is possible that the texts have been corrupted.

In a footnote Ch'en explains that what he has in mind when
he talks about some peOple without the proper ceremonies being none-
theless regarded by society as well and prOperly married is the cus-
tom, as reported in the Min Shang Shih Hsi-kuan Tiao-ch'a Pao-kao Lu
(Record of the Investigation Into Chines Civil and Commercial Cus-
tggE), of abbreviated "marriage ceremonies."

In P'ing-ch'uan and Sheng-hua Hsien, fOr’example, (in what
is now Ho-pei, northeast of Peking)

Poor peOple usually get married on the night of the last

day Of the year (M. ch'u-hsi); they need not select an auSpi-
cious day, the man's family need not send wedding presents,

 

[QT

13'

(4‘

“,3

111
and the woman's family need not reciprocate with a trousseau.
This is in order to save on expenses (NSS; 1806).
Ch'en also says that it is an "old custom" that a widow on
remarriage need not go through any open marriage ceremony (Ch'en
1936: 120n).

The report about poor peOple getting married at the new
year so as to save on expenses and trouble is very interesting
because that is what used to be the custom in Taiwan for gimpgg_
marriage. I thought that this might have something to do with the
fact that at this time the kitchen god, or the god of the hearth,
who has an influence over the fortunes of the family, and reports
family affairs to heaven, is in fact thought to be absent, in heaven
making his report. If there was any guilt or nervousness about such a
relationship this would be a natural time to "get it over with," with
the smallest amount of fuss possible and hopefully without the knowl-
edge Of the spiritual powers. I have never heard this argument
voiced, however, and usually the selection of the date, as is
reported from Hopei, is explained as an effort to save on money and
trouble (see, for example, Ikeda 1944: 202).

If this was a general custom among the poor in China, however,
it lends no special significance to simpgg_marriage in Taiwan. From
the evidence presented so far, however, it is impossible to tell. The
Pelaort for the two Hsiens in Hu-pei does not specify what kinds of
mal—riage were involved in these end-Of-the-year ceremonies. Another
gerperal report for the area does say, however, that the custom of

adopting daughters-in-law was found "among poor families" and that

"gr
1Ii

112

when the girl was over fifteen years of age an auspicious day would
be selected, "in the regular way," for them to complete the mar-
riage. This was said to be intended to save on expenses (NSS;
1805).

Information is available about some other such relationships,
but one becomes increasingly suspicious of their degree of institu-
tionalization. Yao Ling-hsi, for example, discusses the I'dew hus-
band and wife" (M. lu-shui fu-fu):

A temporary union is called a "dew marriage" (M. lu-shui
yin-yfian). It is not a regular marriage because it is as tem-
porary as the dew, which evaporates with the light of day. It
means to conduct a temporary illicit union (Yao 1940: 147).

The information on another few such relationships is even

more tenuous. HsU mentions, for the Cantonese dialect area, the
C. kai_ga-poh, M. ch'i chia-pfo:

This is the "p'ing-wife" (M. 'in -fu) of a man, or a
prostitute he especially likes. A C. Eai-ga-louh, M. ch'i

chia-lao is the "p'ing-husband" (M. p'ing-fu) of a woman

(Hsu 19 : XVI, 43,3).

Such "common-law marriages" are not unknown in Taiwan.

 

Barclay lists T. tgg_(reading pronunciation 39;); "together; said of

man and woman living together without proper marriage."

Spirit Marriagg_

Spirit marriage or "ghost marriage" is fairly well known
from reports based on research in Taiwan (see, for example, Jordan
1971). It was also found in mainland China in various Hsien in
HOpei and in Ho-nan.

. If a son dies before marriage, usually a girl of

another surname is selected who died young, before marriage-
able age, and a ghost marriage is conducted. This is colloquially

113

called M. ch'u kuei-ch'i (literally, "marrying a ghost wife"). ‘
It is also calTéd M. p'ei-ku (literally, "marrying the bones"). 7
After the marriage they are buried together (NSS; 1379). g

 

In P'ing-hu Hsien (Chekiang) in upper, middle, and lower
class families, when a son dies before marriage, usually an
appropriate, matching family is selected, with a deceased

‘ daughter of the right age, and the two are betrothed. Then her
spirit tablet is brought into the house. The ceremony is like
that for a living person. This is called M. min -hun (literally,
"hell marriage"). If this is not done, the soul w11 have nowhere
to belong, and cannot enter the ancestral temple for sacri-
fices, and moreover descendants cannot be instituted. Once the
ghost marriage has taken place, then the necessary qualifica-
tions have been fulfilled for having an heir, and one can then
have a descendant instituted (NSS; 1557-58).

 

In parts Of Shantung,

When a male has died unmarried, his parents may ask a go-
between to find an unmarried, deceased girl to marry him. Her
coffin is then buried with him in his grave. This is called
M. ch'eng yin-ch'in [literally "forming a yig_marriage," and
where "yig? is part of the familiar yin-yang dichotomy in
Chinese thought.] (M§§} 1392).

In areas of Shanhsi,

All those males who have died young and unmarried can be
married to deceased, but unbetrothed, girls. In this kind of
affair one still obtains the services of a go-between, who dis-.
cusses the arrangements with the two sides. If they then agree,
an auspicious day is selected for the marriage. The form of
the ceremony and procedure is no different from that in living
marriage, just a little simpler. When the ceremony is finished,
the two sides then interact as though they were kin. This is
colloquially called M. piei-szu hun ("marrying the dead")

(NSS; 1423).

In other parts of Shansi.

For all those males who died married, but their wife then
remarries, or for those who die unmarried, and they were twelve
years Old or more, their fathers or brothers will marry in an
unmarried, deceased girl, and bury them together. This is col-
loquially called an "old wife" (M. lao-ch'i) or "ghost wife"
(M.kuei-ch'i) (N_s_s_: 1408-1409). "T‘—

For some parts of Shensi it is reported that

Among the people if a girl of marriageable age dies, then
whether or not she has been betrothed, she may be first buried.

uni

iar

u
U

‘ E

Auk

i’ulh

\ IIJ

114

Then a go-between will be asked to discuss a union with a
deceased, unmarried male. When the two are married, then they
are buried together in the same grave, and the girl is also
dressed in new clothes as though she were being married out.
Thus this is called "marrying a ghost wife" (M. ch'U kuei-ch'i)
(NSS; 1702).

 

This form of marriage is what Ch'en Ku-yuan calls an "empty
union" (M. th-ho). He gives evidence of its having existed in pre-
Han days, and the Northern Wei, T'ang, Sung, Yuan, and Ming dynas-
ties. In the Ch'ing

. it was still flourishing; in Shansi, for example,
all those males or females who die young after being betrothed
will go through a spirit marriage ceremony. If the girl dies,
then she will be buried in the man's tomb; if the man dies,
and the girl is re-betrothed, then another, deceased, girl will
be sought to marry him, an auspicious day will be selected, and
the two buried together.

Today in the Ho- -pei, Shan- hsi, Shan- -tung, and Che- -chiang

regions this custom still exists, and is called "ng_marriage“
(M. ch' eng yin- -ch' in or M. yin-p 'ei) (Ch' en 1937: 112).

It is interesting that all spirit marriages that I have
heard of in Taiwan are initiated by the spirit of the deceased
female. Jordan (1971: 183) bears this out, and goes on to say that,(
"culturally and structurally there is every reason not to be sur- /
prised that it is the girls who become ghosts, and that they are
pacified and laid to rest by providing them with a husband and line
of descendants who undertake to worship them." It is surprising,
though, that most of the reports from the mainland of China give as
much or more weight to the deceased male as initiator of such a
relationship.

It is also surprising that from the reports that I have seen
so far, it was typically two ghosts who were married to each other,

and the remains buried together, whereas in Taiwan (in my experience)

115

it is much more typical that the spirit desires to marry a living
person.

Jordan mentions that in what he calls the "Bao-an variant
Type II," the ghost specifically demanded to marry her sister's hus-
band, and caused trouble for her natal family (1971: 184). I found
this to be the most common situation in T'ien-sung-p'i, though often
it was not exactly specified that it was absolutely necessary that
this man be the one.

Jordan finds it puzzling that the sister's husband be
singled out, and offers the hypothesis that this is because the
"supply" of deceased, unwed girls in a village may get "used up," and
by having recourse to the deceased sister of a wife, a new source of
supply can betapped. (This is because villages in southern Taiwan
tend to be exogamous, and there may be more such girls in another
village where this ceremony is not so popular.)

There is another possible explanation. In my experience
the spirit medium tends to stick to "prescriptions" which can be
carried out by, and involve, the person who has taken the initiative
to approach the medium. This only makes sense. Functionally, the
involvement of the afflicted or affected person should be psycho-
logically more effective. The typical situation is that illness
or misfortune has been visited upon a person or family, and finally
it is felt that something has to be done about it. If the medium
advises a course of action that involves the affected person, then

at least it is felt that he or she is "doing something."

116

Secondly, strangers are not going to want to get involved
with this sort of affair, simply because there is no particular reason
to put up with the trouble and inconvenience, and the possible future
complications. Jordan cites Lou Tzu-k'uang's (Lou szy-kuang in
Jordan's romanization system) account of a practice in the south of
Taiwan of inducing a stranger to participate in such a marriage by
putting money in a red envelope on the road, hiding, and waiting
until a male of the appropriate age picks it up. Then the perpetra-
tors rush out, addressing the stranger with the apprOpriate kinship
term, and try to force him to acquiesce to their demands that he play
along and marry the spirit (Lou 1968: 26-29; Jordan 1971: 185n). The
point here is that the man has to be tricked into the situation,
has to be persuaded, and often is paid a certain amount of money (the
red packet). Otherwise he would have no reason to get involved.

A typical case that is taken to a medium is one in which
one's children are ill, or have been often ill. Jordan (1971: 182)
says that the motivating problem is typically illness unresponsive
to normal medical treatment. Such illness does, I'm sure, act as
such an impetus, but more often it is repeated or extraordinary ill-
ness. It has been pointed out (Helen L. McGough, personal communi-
cation) that often in such cases there seems to be, in effect, a
distinction made between types of causation. It is recognized that
there are "rational," "Objective" causes of illness, and recourse is
made to Chinese or Western medicine to deal with them. This,

though, is in a sense treating only the symptoms. There is still the

117

kind of questions posed so well by Evans-Pritchard; the questions
"Why me?","Why now?","Why ill so Often?" and so on.

Typically a child will be ill for a while, or several times,
and the mother will take him or her to a doctor (Western or Chinese).
After the illness is over (or before, if it is a long one), she will
take the child to a medium to find out just when the illness occurred.
It does seem to be true that, as pointed out by Jordan and others
(Li Yih-yuan, personal communication), unmarried females have an
anomalous position in the social structure during life, and in death
this is extended into the spirit world. Thus, such a woman is a
"natural" for explanations involving malevolent, or at least uneasy,
spirits.

So it should be no surprise that when faced with a sick
child, the medium should say (and typically this is the way it is
phrased) "There isla (deceased) woman of the (child's) mother's genera-
tion," or "There is a woman of the father's generation who died
young," and so on. In one case which I observed, an old woman with a].
sick son was told that it was her daughter who died young who was
causing the illness; the exact relative chosen depends on the age and.
social characteristics of the client, but is usually one over whom
the client has some influence or control. It would hardly make
sense for the medium to attribute the illness to someone or something
over which the client has no influence or control, and in effect say
that it is an incurable affliction.

In the case of ill children (and in "0' experience, this was

the major motive for consulting a medium), then, it is very likely

ia

118

that the medium will single out either a sister of the child's
mother, or of its father. In most such cases, in my experience, it
is the mother who takes the child to the medium, and the father sel-
dom goes along. Thus it also makes sense for the medium to specific-
ally single out the mother's sister, thus involving the mother more
directly. By this logic it also makes sense to specify that the dead
girl wishes to marry the father Of the ill child. If the father is
not willing to go through with such a ceremony to possibly (and many
peOple are rather dubious about this whole thing, but want to "cover
all bets") help his own child, then who could be expected to do so?

I don't think, then, that it is necessary to have recourse
to some sort of village exogamy in spirit marriage (Jordan 1971:
186-87) to account for a preference for the sister's husband in such
marriages; it is a natural tendency, an outcome of the sociological
situation. Furthermore, I question the idea of such dead girls
being "used up." In my own experience I have never seen a medium at
a loss for words when told by a client that "I have no sisters and
never did." The medium may say that the client's mother had been
pregnant and had a spontaneous abortion without knowing it, and the
fetus had been a female. The medium is always absolutely certain,
and the client doubtful, but in many cases persuadable. Carrying
this argument even further, it seems that in some areas the medium
may even argue that the increased and increasing use of birth control
pills means that many spirits that never even had the chance to be
born are floating around, causing trouble for living peOple (Ch'en

Chungmin, personal communication).

119

Wu Ying-t'ao has an interesting description of such mar-
riages as practiced in Taiwan, and which he calls by another, common

name, M. ch'il shen-chu, "marrying the (woman's) soul tablet":

 

This is spirit marriage (M. min -hun, T. ben -

Ngg). In the past this custom used to flourish in aiwan.
Now, because it is common for various Buddhist temples to take
in and make offerings to soul tablets, it is less often seen.
Its characteristics are as follows:

A woman who dies before betrothal is called an "orphan
girl" (T. kO°-liang), and her family members will arrange a
ghost marriage for her. Since it is the common belief that the
spirit of the dead can still desire descendants (T. tho-su)
to offer sacrifices to them in order to prevent their becoming
evil spirits, a gO-between is engaged to find a poor man and
give him a certain amount of money to compensate him for cere-
monial expenses, and have him marry the soul tablet. Most poor
people could find a use for this money to find another woman
and to help meet the expenses of marrying her. The man's fam-
ily, however, also has the duty of ever afterwards continuing
the sacrifices, and of regarding the (woman of the) soul tablet
as the principal wife, and any other as second wife.

In the ceremony of marrying the soul tablet, the go-between
first acts for the man, selecting an auspicious day and send-
ing gifts to Offer to the soul tablet. On the day of the wed-
ding, the tablet is then put in a black sedan-chair, and only
after it has been brought into the house, and put on the family
altar, can the red sedan-chair arrive with the new bride. Some-_
time later again an auspicious day is selected, and then the
soul tablet is entered into the man's family's main ancestral
tablet (T. toa-pai) to receive homage from the later genera-
tions.

Another form occurs when a woman dies after betrothal, and
the man's family wishes to have him marry her spirit. If they
do not it is colloquially believed that the soul of the dead
woman may become an evil spirit dangerous to the second wife.
Therefore, on the day before, or on the same day as, the wedding,
the man marries the soul tablet. The form of the ceremony is
the same as the form described above.

There is also the case in which both the male and female
die after betrothal; when this happens the man's family may
marry in the girl's soul tablet, and then sacrifices are made
to them both together. A certain amount of money is then put
in front of the incense brazier to symbolize that later a M.
kuO-fan adOption will be carried out from among the families
of the brothers of the dead man to provide them with descendants.

If the male dies first, and the woman decides to remain a
chaste widow, as in the Old days, they she will don wedding
clothes on the day of his burial, and will ride in a sedan-chair

for

120

to the man's family's house, where they will welcome her as

in the regular ceremony. Afterwards, ‘the bride should for-
mally ask after the groom's health. His family members claim
that he has already recovered. Then she changes into mourning
clothes, and remains in the house, sending Off her husband with
wailing. On the other hand, if she does not intend to remain a
widow, and wants to remarry, then she does not follow the above
marriage ceremony, but only attends the funeral, just as anyone
else would (Wu 1969: 143-44).

Purchase and Pawn Marriage

 

The Chung-kuo Ch'uan-kuo Feng-su Chih (Record Of Customs

 

All China) records that in ShaO-wu Hsien in Fu-chien

 

If the husband dies, then the wife remarries. This is
regarded as quite customary. It even goes so far that a wife
has one husband, and then remarries to another, and then
remarries to yet a third. There is the custom of the M. tso-
ch'an chao-fu, "prOperty-managing uxorilocal husband," who

 

 

people call a M. pai pi-t'ou "white nose." A husband can sell
his wife to someone Else, and the wife can redeem her own self.
This sort of evil custom ought to be prohibited, for the good
of the peOple (Hu 1936: II, 5:74).

 

This last is an institution similar in some ways to that of

polyandry described elsewhere, but is basically quite different, and

is referred to by a variety of names, M. tien-ch'i, M. tsu-ch'i,

and

so on, all meaning "mortgaged wife," "pawned wife," or “rented

wife."

A report from Ching-ning Hsien, Chekiang, says that:

In the lower classes there have been cases of mortgaging
(M. tien) one's wife to someone else. The two parties discuss
and set the amount of money and length of time of the mortgage,
and a contract is drawn up, called a "mortgage marriage con-
tract" (M. tien-hun shu). The mortgaged wife may move to the
family of the one whoibought the mortgage, or may remain in
the family of the one who sold it; there is no fixed pattern.
The children born during the specified time period belong to
the one who bought the mortgage. As for the goal of this
form, it is financial gain for the one mortgaging his wife. For
the holder of the mortgage, there are two reasons for this:

 

121

1. Because he has not yet married, and can't afford to
marry in a wife, he can accept another's wife for a tem-
porary period, or
2. Because he has a wife, but she has had no children,
he takes in another, mortgaged, wife to provide heirs.
Sometimes men have been known to take two or more such
wives at the same time. This kind of custom is also practiced
in the upper and middle classes. The sale of a wife occurs when,
due to poverty and inability to support her, she is sold to
another as wife or concubine. The seller executes a contract,
called]? "contract of marriage by sale," M. mai-hun shu (MSS:
560-6 .

 

In Yung-K'ang Hsien, Chekiang,

. the custom is quite open in nature; there is a go-
between as witness, and a contract. In the contract is clearly
written the time limit and the price of the mortgage. Sons
born during the mortgaging period follow the line Of the one who
holds the mortgage. When the time limit is reached, then she
returns to her original husband and his relatives. The mortgage
price is not returned (thus this should perhaps be called
"rental marriage"). Sometimes, if they don't want to return
her, the original husband can be approached to see how much it
would cost for him to sever his relations with her (NSS; 1558).

The same custom is reported for a number of other areas of
Chekiang (NSS; 1529-30), and there is a relatively extended analysis
for Feng-hua Hsien:

1. Aspects of the custom and their changes:

In Feng-hua it is customary that when a woman's husband dies
leaving children, and there is no property, so it is hard to make
a living, they she can mortgage (M. tien) herself out to someone
else as wife or concubine. Those who are old or poor, and lack
descendants, and find it hard to marry a wife will take a mort-
gage on such a widow, and take her as wife or concubine. The
time limit is typically eight or ten years (there are also
indefinite time-periods, as agreed to and set in the contract by
the two parties). The children born during the period are recog-
nized as the sons and daughters of he who holds the mortgage.

When the time-period is up, and the contract no longer holds,

most of these children remain in the household of that husband.
The amount paid to the woman by the holder of the mortgage is

also clearly entered into the contract. Each year a certain
amount of silver is paid, and a certain amount of grain; a rather
small amount sufficient for an allowance, but not for the complete
requirements of the whole family. The variations on this are dif-
ficult to discover.

122

2. 'The areas where it is found and its legal effect:

This custom of mortgaged wives is found in various [named]
areas of Chekiang, but in those places the motivation for it is
generally poverty and difficulty in making a living. In Feng-
hua, however, about 90 percent of the cases are because the first
husband died, leaving children (if there are no children, then
the wife just remarries in the ordinary way [literally "com-
pletely," M. wan-ch'anJ). The effects of this transaction are
such that all of the children born during the time-period of the
contract have the same rights when it comes to property division.
If the holder of the mortgage has no other children, then they
have the rights of those with the status of legitimate sons of
the primary wife (M. ti—tzu), they enter into the line of kin-
ship succession, they can be entered into the genealogy, and
there is no question but that they have full rights of inheri-
tance to the estate; no one in the kin group can dispute this.

3. The relationship between this custom and the current
code:

According to the current code, in the section on the rights
to the estate held by those Of junior generations, it is held
that in division of property, one cannot simply give each son an
equal share depending on the number of sons. Illegitimate sons
(M. chien-shegg) must be given one-half the share that legitimate
sons receive. If there are no other children, then they can be
treated as heirs and given full shares. According to the Draft
Legal Code, "illegitimate" sons are limited to those born Of
fornication only (M. yin chien erh sheng), and those born of
legally invalid marriage are not included. When distinguishing
between illegitimate (M. chien-sheng) and natural (M. szu-sheng)
children, one must look at the time-periods involved. The con-
tract in this kind of transaction, whether the woman is termed
"wife" (M. ch'i), or "concubine" (M. ch'ieh), or dependent
(M. t'ou-k'ao), or even when there is no specific term used,
still these are not permanent unions, and the wife does not
move to the husband's household. The holder of the mortgage
also does not usually live with the woman, and only provides an
allowance, and does not take responsibility for the entire
expenses of the family. In effect, then, this is nothing but a
union which has been provided with a contract, and in law abso-
lutely cannot be regarded as a regular marriage. In this kind
of marriage the children that are born are of course no differ-
ent from natural children (M. szu-shen , and so have the same
rights in property division as do egitimate children born to the
primary wife (M. ti-tzu). If the holder of the mortgage has no
other children, and so treats these as his children, then other
possible heirs cannot object. Although this custom has long been
prevalent, it is in conflict with the Code, and its result is the
confusion of blood lines (this kind of mortgaged wife has no moral
standing to speak of, and whether or not the children born to her
are really the blood children of the holder of the mortgage is
truly a big problem). . . .

 

 

 

 

123

This report also has an appended item comparing this with
the institution of polyandry (M. chao-fu yang-fu) mentioned else-

where:

In (various named) Hsiens, there is the custom of inviting
in an uxorilocal husband to support the first husband. This is
to mortgage a wife, through poverty or sickness (but the wife
does not leave the first husband's family, but invites in the
second husband to live with them). Its important points are as

follows:
1. The uxorilocal husband must have obtained the permis-

sion of the first husband, and seldom is the money so
obtained used to cure illness.

2. The children born during the fixed time-period (usually

three years to five or ten) follow the second husband.

3. The second husband, during the contract period, does

not live with the first (that is, does not live in the
same room (M. pu t'ung-fang chU-chu. MSS: 1543-45).

So far the discussion has been limited to the so-called
M. tien-ch'i, which has been variously translated "mortgage" or
"pawn." The legal relationships created by entering into a tigg
contract are not, of course, exactly those implied by the English
terms used as glosses.

In Taiwan (where, as we shall see, the custom of transfer-
ring tIeg-rights over persons also existed), there were two main
forms of pledge or pawn, T. tjgg_and T. Iggj; T. tjgg_is often V
translated as "pledge" or "pawn," while tNgI_is more often trans-
lated as "mortgage" or "hypothecation."

A major consideration in all of these forms was,of course,”)
that they allowed persons to transfer rights in the reproductive ‘
Tummers of women separately from other rights, and furthermore did
not: imply that the original rights-holder irrevocably lost those
rig'hts. The argument that Okamatsu applies to rights in land must

ADDFIy to some degree to rights in persons as well:

124

In countries where the feudal system prevails and the laws
are undeveloped, the main occupation of the peOple is agricul-
ture, and, apart from the land, goods constituting capital do
not amount to much, and consequently if the right to land is
entirely lost, one has difficulty in Obtaining a livelihood.
Even if money is obtained for the land it cannot be used as
capital in the absence of commerce in which to employ it. These
are the reasons, it is supposed, why, in disposing of the right
to lands in a country like Formosa, a partial right was always
reserved and made a source of constant revenue (Okamatsu 1901:
28 .

Finally, I would like to give some attention to the rela-
tionship created through T. 2S5. This meant giving (in the case of
land) the right of tenancy over to someone else; letting the land.
The owner of the land obtained the right to a certain consideration,
a rental. The “real" rights to the land (T. giap-tsu) remained with
the lessor.

I have elsewhere discussed the use of the T. pSN_relation-
ship with respect to rights over persons as it was practiced in
mainland China. It was also used in that way in Taiwan, where, for

example, the phrase T. pgk-sin ho-lang_means "to hire out (as one's

 

son) to be like a slave" (Barclay 1923). This type of relationship,
however, was in both mainland China and Taiwan subject to a good
deal of terminological variation. For parts of Che-chiang, for
instance, we learn that

In middle and lower classes, it has been the custom of
men with wives of several years' standing, and with children,
but when later forced by poverty and inabilit to support
them, to write up a contract, renting (M. t§g_ the wife to
someone else. This is discussed and arranged through a go-
between, the time-period is set at perhaps five or ten years,
and the price at so many foreign dollars. This is all written
up clearly in the contract. If she is to take her children
along to be raised, the time-period for that is also mentioned,
and other conditions, such as one absolving the other party of

125

responsibility if the children die young, etc. When the period
is up, then the contract is abolished and the wife must return
to the original husband. Sometimes she is redeemed early (MSS:
1555-56). '___

Although this is called "rental" (M. Egg), on the informa-
tion available there would seem little to choose between such
"rental" and the M. tigg, or "pledge" mentioned above. Okamatsu
(1901: 147) has said that in regard to land and other immovables there
is considerable terminological inconsistency in the use of M. tigg_
and M. 3:31; From what little I have read in regard to rights in
persons, there seems to be little consistency in the use of M. t§g_
and M. tjgg; but these two do seem to be differentiated from M. tIgj_
on the basis of whether or not the reproductive powers (as distin-
guished from sexual access per se) are transferred.

I asked a sophisticated informant in T'ien-sung-p'i about
this rental or pawning of persons in Taiwan and the local area. His
response was that such contracts now mainly have to do with girls
going to work as prostitutes, and are entered into by their parents.

A prostitute who signs a two-year contract gets (or rather her parents.
get) about NT$20,000 if she is a virgin, and NT$17,000 if not. As
one might expect with such sums Of money involved, there is a medical
check to ascertain the state of her virginity. The I'black society"
guarantees and enforces the contract, as it is of course illegal in
the eyes of the Nationalist Code. If either side violates its pro-
visions, the other side can appeal to the society for ajudication.
If the girl runs away, they will find and return her. In T'ien-

sung-p'i while we were there, it was said that there were about

fifteen girls from the village working somewhere else (mainly in

126

Taipei) as prostitutes. Usually about 30 percent of them at any
particular time had gone out to work as prostitutes on their own
initiative. The rest had been "pawned" or "mortgaged" by their
parents. Usually this happens when they are fifteen or sixteen
years old. A broker comes to the town, asking families if they want
to pawn out their daughters. Understandably, given the legal situa-
tion, it was a bit hard to get information on this broker. I did
find out, however, that while he was sometimes a local person, often
a girl from the area who went to work in such a way would Often tell
her M. Tao-gan, "boss," about other prospects in T'ien-sung-p'i, and
he would come asking.

Most of the girls who are hired out in this way go to
Taipei. Some used to go to Taichung, in the 19605, because a woman
from the town had opened a house of prostitution there. There is
both legal and illegal (but winked at officially) prostitution in
Lotung, the closest city of any size, and only about one-half hour
by bus from T'ien-sung-p'i, but that was considered too close. Some
of the local girls were working there, but only the older, more experi-
enced ones, who were described as no longer shy and embarrassed about
their profession.

In the contracts of this sort that I have seen (all his-
torical) for Taiwan, the term used is M. 3:31, Though I have not
investigated this carefully, it seems that there is an indication for
M. ti§g_to be used when a married woman (or a widow) is pawned to a
man for a certain amount of time, in which time he obtains her ser-

Vices as cook laborer, etc., and obtains rights in her reproductive

127

powers; the children born to her become his. In the M. tigj_con-
tract, the reproductive powers are not so transferred.

Taiwan, of course, was not the only part of China where this
sort of arrangement was possible. A report from Chiangsu says that

Though peOple are not articles of commerce, yet poor fami- i
lies still treat them as pawnable goods, and send betrothed, ’
but yet unmarried, girls to the husband's family to be raised.
Sometimes they are pawned out (M. ch'u-1i k'un-ch'i, literally
"sent out in a k'un contract," where k‘un means "to tie up,"

"to bind") to houses of prostitution for perhaps three or five
years. A formal contract is quite casually written up, a con-
tract called a M. k'un, which also clearly noted that the death
of the girl would be considered an act of heaven (M. shen szu
t'ing t'ien). The conditions of her return and the length of
the time period were also stipulated. The price might be 100
uan, 4O yuan at first, then 30 yuan the second and third years
iNSS; 1460-61).

 

 

 

This introduces yet another term for the various contracts
for transferring persons, or rights over persons, but also forms a
bridge with adoption and T. SINpgg_marriage. A report for another
area in Chiang-su uses, in addition to this term M. NINE, the term
M. yg; which also means "to pawn," "to mortgage":

M. k'un-chi means to take one's daughter, or one's wife,
and send her into a house of prostitution to work for a speci-
fied period of time as security for a loan of money, setting up ‘
a M. k'un-shen contract. The time period specified can be in
terms of long, middle, and short years; the first are of six-
teen months, the second of fourteen, and the third twelve. If
she becomes ill, she must make up the days of work that she
misses, and if she gets pregnant, she must work an additional
three months after the baby is born (NSS; 1483).

This sort of pawning was not necessarily connected with
prostitution. In Hsia-pu Hsien, Fu-chien,

If poor people have too many daughters or too many adopted
daughters-in-law (M. miao-hsi), then when they are perhaps six
or seven years of age, they can write up a contract pawning
them to someone else to act as M. gel; "a servant, female
slave". 111 the contract is clearly specified the period of
time, which may be eight or ten years. The price may be four

128

or five yuan to perhaps ten-odd yuan. If it is for an indefi-
nite period, then she can be redeemed through repaying the
original sum. If she should die through illness, then no ques-
tions are asked. Should she die throu h mistreatment, however,
then her master must pay compensation iNSS; 1604).

In the Ch'ing Dynasty there were popular sayings comparing
women from the Pescadores to cattle from Taiwan, as both were thought
to be forced to do extremely hard and heavy work. In part this was
because women from the Pescadores were thought to be often sold or
pawned as slaves. In 1852 there was great famine in the late winter
in the Pescadores, and several score of women were taken to be sold
in Taiwan as slaves. Several representatives of the gentry there
felt that this was a shameful affair, and so banded together to redeem
them. They then supported them until the harvest was finished, gave
them some travel expenses, and returned them to their families.

In 1840 a proclamation concerning such slave girls said,
in part,

. . . There are rich and poor families, noble and mean
persons, but the most harmful practice in Taiwan is that of
taking slave girls! The girls of good families have, at home,
mother and father, brothers and sisters. When they marry out
they have the enjoyment of mother-in-law, father-in-law,
husband-wife relationships, sisters-in-law and children. As
for female slaves, however, their parents, through poverty,
are forced against their own wishes to sell them to act as lowly
servants to get the several tens of taels they can for them to
keep themselves alive. But are not these girls also people
just like us?

According to the M. II_a girl should be betrothed at fifteen
and married at twenty, and clearly these times should not be
missed. If I had a daughter, I should certainly marry her at the
proper time, and as long as a suitable husband was still not
found, I would be nervous and worried, and leave no stone unturned
in trying to find one. So how could one delay this year after
year, until she is old-aged and still unmarried?

In Taiwan it is customary that such slave-girls grow to
adulthood and still do not marry. They may be kept in the house-
hold, or sold and resold to other people, to spend their lives as

129

servants even unto death. They may be kept to marriageable age
and then be given to some old man as his plaything, to work like
an animal, and end up reduced to a pitiful state, a toothless
old hag. They may find themselves pursued by both father and
son, each unaware of the other's interest, or two brothers,
bewitched, may‘ each try to get her into bed. Or such a girl
may find herself whipped and injured by some jealous wife, or
the target of scheming concubines, or she may find herself con-
stantly pregnant, giving birth to bastard after bastard, whose
status and ancestry are unclear. Or she may find herself in a
house of prostitution, or may be sold to another province.

She may have no one to rely on in life, and no one to care for
her ghost after death. This is in direct opposition to heavenly
virtue, and the cause of many scandals in families. When con-
sciences are lost, what has happened to the way of heaven?
Grievances build up, and silent anger that can't be aired.

The anger of heaven is what causes the frequent disasters,
large and small, which afflict Taiwan. It is not unlikely that
it is anger at this kind of evil. . . .

This not marrying off slave girls is clearly punishable
under law, and the Taiwanese custom Of confining them, apart
from the sensibilities and the law, also goes against harmony
in heaven, and harms customs on earth. Though it has been pro-
hibited by previous officials, it has still not been changed.

A number of the gentry representatives have requested that
I again prohibit it; if it is not clearly regulated, then it
will be difficult to change evil customs. Thus after careful
consideration, I have decided that: Henceforth all, whether
gentry or common peOple, who have slave girls who have reached
the age of twenty-three years, must find a husband for them;
they may not delay beyond the age Of twenty-five. If they delay
longer than age twenty-five, then the slave girl shall be returned
to her parents, brothers, or other relatives to be married if
they apply to the government office to ask for her. The price
paid shall not be returned, and moreover the head of the guilty
family shall be punished with eighty strokes. After her marriage.
the girl shall follow the directions of her husband, and shall
not be detained by the family head. If the latter has not
received back the price that he paid, then the girl and her hus-
band may stay to serve for a number of years, but it shall be
their decision to do so (Taiwan Szu-fa: 155-57).

 

Just as this practice was not restricted to Taiwan, so it
was not restricted to the Ch'ing Dynasty. Feng Ho-fa says of Shensi
in the 1930s that

. Because of lack of food grains, and because all other

possessions were already sold, persons were bought and sold, and
there were even human markets (M. jen-shih). In Wu-king and

130

Fu-feng Hsien there were at least three of these, centers in
trading in women, where the price was dependent on age and
beauty. Attractive girls of around twenty brought eight yuan
according to the January 5, 1930, edition of the Shen Pao.
For the most beautiful, the highest price was no more than ten
yuan. The majority were around eight yuan, the next most common
price was three or four yuan. K)
In Sui-yUan human markets were set up. When the local “
government stopped it, they were disguised as marriage broker-
ages (M. shou hsi-t'ieh). One could buy a marriage contract
(M. hsi-t'iehiifor four foreign dollars, and then could take the
girl whose name was on it. One could then tell the government
that it was a marriage (M. ch'u-ch'i) or taking a concubine
(M. na-ch'ieh). The government then set up a 10 percent tax on
this, and in a short time earned as much as 10,000 yuan. . . .
In other provinces selling women would bring about 100 yuan.
. Women from Shensi were sold to Shansi to be wives and
concubines--which was better than being sold as a prostitute to
Chih-1i (Feng 1933: 142).

 

This association with marriage brokerage was not fortuitous;
an edict issued in the latter half of the eighteenth century made
reference to the same sort of thing in Fu-chien Province:

From the time that this office prohibited the custom among
the Fukien people of confining female slaves and not marrying
them to husbands, it has been gradually abolished, but they are
still not automatically married, but are given to an evil go-
between (M. chien-mei) who contracts to sell them, and they are
then taken to an illegal brokerage (M. mei-kuan), and then pro-
spective buyers are invited to come to inspect them. Some of
these buyers are shameless fellows with no consciences; they may
falsely claim that they have lost a wife, and intend to remarry,
or they may claim to want to buy a slave, or marry a concubine,
and in their deception demand a trial of her cooking ability,
whereupon they will throw a big party, with much drinking and
riotous behavior, even to the point of forcing the girl to sell
her favors to the guests, from which these fellows take the
profit. This sort of thing may go on for months and years, with
no date set for the marriage (Taiwan Szu-fa: 154-55).

Women and children could be sold as well as rented or pawned
out to others; it would seem that the difference was between trans-
lative acquisition and constitutive acquisition. While the latter
was “the creation of a certain right out Of a right enjoyed by one

Person and the transfer of the created right to another person,"

131

the former was applied "to cases where a person entitled transfers
his right to another person, that is, when the subject of the right
is changed" (Okamatsu 1901: 31). Thus the important difference
between the two has to do with the degree to which the person or
actor with the original rights to and in the thing or person retains
some of those rights after the transaction; in constitutive acquisi-
tion, some of the rights are transferred, but some of them remain
with the original rights-holder. In translative acquisition, all
of the rights are transferred, so the original rights-holder no
longer has any.
A report for An—huei says that
Poor peOple who lack the means to support their young
children may sell (M. chia-mai) them to one or another Buddhist
temple to be monks. These rural peOple, forced by poverty,
mistakenly believe that monkhood is just another profession,
and so willingly send their children to the monastery, and
receive money in return which can help support them, or in the
hope that he may in the future succeed to the position of head
of the monastery (NSS; 1488).
In another part of An-huei, it is said that it was customary
for people, if they lacked sons, to buy (M. chia-mai) a boy of a
different surname to inherit and continue the kin line. In the
contract for such a purchase it was clearly stated that he would
never return to his original kin line (NSS; 1493). In both of these
cases the result for the child's original family was much the same,
in that the child was completely lost to them, both as a person and
as a way to produce the next generation.

This was also said to be common in Chin-chiang Hsien of Fu-

chien, ‘where- people, whether or not they had their own sons, liked

132

to buy (M. mai-yang) boys of other surnames, change the surname to
their own, treating them as their own children. Any family property
was then divided equally among the sons, whether they were purchased
or born to the family. This was not objected to, it is said, even
though it confused kin lines (NSS; 1573). In another part of Fukien
this was said also to be the case, except that when the sons were
entered into the genealogy the purchased ones were differentiated
from the ones born to the family by the use of different colored
ink--blue for the former, red for the latter--when drawing lines to
indicate M. chao-mu,"generational standing'(NSS; 1598).

In parts of Hu-pei poor peOple sold their daughters to
others to be their wives, and it was noted that this was through
the pressure of poverty and was something that the law “could not
successfully prohibit" (NSS; 1668).

In Hu-nan, it is said of selling one's wife that

This kind of custom is followed because worthless people

squander their money, and find it difficult to stay alive, so
much so that they sell their wives. As for the wives, they are
often willing to do this because the husband is unreliable, and
they can't endure the suffering to which they are subjected.
They must, however, gain the permission of the husband, and in
the case of those who have not reliably obtained such permis-
sion, a go-between must be used. If the wife's permission is
obtained, and her relatives' permission, and the purchaser's
elders' permission as well, then she is taken in as wife, with
a palm-print as evidence (on the contract). The purchaser must
also provide a certain number of presents and a feast.

There then follows a sample contract:

The executor of this contract to sell a first wife is

Mr. so-and-so. Today, because of incompatibility with my first
wife, so-and-so, I have asked the go-between, so-and-so, to
help in selling my wife to Mr. so-and-so, and in settling the

amount in cash and presents and to act as witness. I wish to
make it clear that after selling her, we will have no further

133

contact whatsoever, and the purchaser can establish a husband-
wife relationship with her, and no one shall interfere. Fear-
ing that there is no proof of the spoken word, this is written
up as evidence (NSS; 1683).

This sort of sale of a wife at times became difficult to
distinguish from "regular" marriage; in Nan-an Hsien, Fu-chien,

Marriage, in other than gentry families, can still be
concluded through purchase. The price of a wife may even be
as much as several hundred yuan. If it is no more than a few
yuan, then this is actually betrothal money. It is called
"buying a wife," though, and not "marrying a wife" (M. p'in-
ch'i). This is the colloquial in Nan-an. What is most
strange, though, is that whether she is bought or married,
after the woman is brought into the house, her husband can
freely resell her, and her natal family cannot interfere. The
money that is so received goes to his family, and the woman's
natal family does not share in it (NSS; 1583).

. -'-., A“"\,
\\ -

In parts of Shensi,

Poor people raise daughters to marriageable age, and then
simply treat them as valuable commodities. Those who would wish
to marry Off a daughter ask a go-between to arrange it and to
discuss the price. Ordinarily this is around twenty-four
strings of copper cash or twenty taels of silver. After gifts
are exchanged, then a marriage card is sent, and the price is
discussed; this may be as much as 100 taels or more. Col-
loquially this is called "marriage by purchase" (M. mai-mai
hun-yin), but afterwards the families of the man and of the
woman in this kind of marriage may not recognize each other as
relatives (NSS; 1740).

Soon we find that we are discussing brideprice in general,
and there is little or no question of marriage by purchase. In
Chahar:

. . When poor peOple marry out daughters, they always
demand many presents, so that there is an element of purchase-
marriage. At the time of marrying out a daughter they repeat-
edly demand clothing, jewelry, money and presents. These are
all written in the marriage document. If family A pays out a
lesser number of presents, family 8 can step in and offer more;
there are constant lawsuits over this (NSS; 1817).

134

In Hong Kong, at the present time, this is still practiced

as an institution. According to the South China Morning Post of

 

August 10, 1975:

. . Immigration Department sources say that as many as
950 overseas Chinese girls were married in Hong Kong last
year.

Despite the fact that both Vietnamese and Khmer girls
are a little less available because of the political situation
in those countries, 500 girls from Southeast Asia have been
married here in the first six months of the year--and in most
cases the weddings were arranged by intermediaries who are
making a small fortune out of the business.

As with everything else, inflation has reared its ugly
head. The price of a transaction has gone up from HK$8,000
to $10,000--but men of Chiu Chow and Fukienese descent are
apparently quite prepared to part with this sort of sum pro-
vided they get the right type of girl.

The "wrong" type at the moment would seem to be the
Cantonese girl. "They spend too much time playing mahjong
and wasting money. We want women who are frugal and will
look after us in our old age" said one Chiu Chow bridegroom who
appeared to be at least twenty years older than his bride.

Complications, of course, do arise, because some of the
girls arriving--particularly those from Bangkok--believe the
streets of Hongkong are paved with gold. Many are village
girls from humble backgrounds, whose parents sell them to
brokers for something in the region of $2,000. The Hongkong
and Bangkok connections then split the rest of the $8,000
(out of the $10,000) paid by the prospective bridegroom.

The bride price fluctuates, but not by very much. . . .

This sort of brokerage occurs in Taiwan as well, but the
most notorious form of such transactions, as mentioned earlier, is
connected with prostitution, not marriage. I have mentioned that
in T'ien-sung-p'i I was told that about 60 percent of the girls
who became prostitutes did so through a broker, and through a con-
tract entered into by their parents. Little reliable information is
available on general figures for Taiwan, but what is available seems
to suggest that this figure (60 percent) is about right. Although

it is complicated by different categories of prostitutes, it would

135

seem that around 50 percent have entered into some sort of contractual

arrangement (Hsieh 1971: 456-60).

Homosexual Unions

 

To end this section on irregular marriage and marriage-like
forms, I would like to discuss homosexual unions. To some the term
"homosexual" implies unions in which there is physical, homosexual,
activity. I do not mean to imply this, but simply mean by the term
that these are unions with a greater or lesser degree of institu-
tionalization and permanency, in which the partners are of the same
sex. As we shall see, this often did lead to a physical homosexuality,
but this is not a sociologically necessary part Of any definition or
characterization of these unions.

In the discussion so far, attention has been given to hetero-
sexual unions only. The great variety of these, however, should
lead one to suspect that marriage-like unions might not be restricted
to heterosexual pairs, and once one starts looking for such arrange-
ments, a certain amount of information on them does come to light.

Of all the forms mentioned so far, however, these were probably the
most rare. This, however, is only a relative evaluation; the num-
bers of women, in particular, involved in such relationships may have
been quite substantial indeed, even if the practice of such rela-
tionships seems to have been rather limited geographically. .

The first of these forms to be considered is that involving:
females, and it seems to have been restricted to the P'an-yfi

(C. Pun-yu) and Shun-te (C. Sun-tak) areas of Kuangtung province

136

(these areas are quite near Canton city). It seems at first glance.
to be fairly recent, and to be a response to particular techno-
environmental and economic conditions. There actually were a number
of different forms, interrelated in various ways. i

The data are limited at present. I suspect that a thorough
search of written sources would produce a good deal more, however,
and there is ongoing research in Hong Kong among women who had par-
tiCipated in these and similar forms by Marjorie Topley and Andrea
Sankar.

Following Fei Hsiao-t'ung's argument, the introduction of
Western manufactured goods upset the rural economy, particularly
that sector involved with rural handicrafts. At about the same time
foreign capital, and Chinese capital, was being used to set up cer-
tain kinds of factories, particularly in the treaty-port areas, the
large urban centers on the coast of China. In the P'an-yu and
Shun-te areas, the important local economic activity was silk pro-
duction. The develOpment of a degree of mechanization of the silk
industry, and its centralization in more capitalized factories, meant
the creation Of wage labor, and thus jobs. The existence Of jobs at
which women and young girls could make enough by their own labor to
support themselves meant that they had a much greater degree of
freedom and independence, and they,of course,took steps to increase
and protect this new-found freedom. One of the results was the
development of an "anti-marriage" movement, as marriage and the
family were seen by some of these women as one of the most, if not

Iflg most, oppressive of social institutions. It is interesting,

- e * 'fi\¢"‘

137

if not surprising, that alternatives to male-female marriage and family?
organization were still often phrased in terms of family organization 5
and marriage. The girls involved in these alternate forms had to ,
work with a given repertoire of concepts and values. To some degree
this movement was phrased in indigenous terms. It seems that it

had its beginning well before the May 4th movement in China, which

did much to spread I'modern" and Western ideas about individuality,
freedom to marry, the small, conjugal family, and so on.

Most of the data that I present here are from the Chung-hua
Ch'uan-kuo Feng-su Chih (Record of Customs for All China). This work
was published in 1936, but the work of collecting the data seems
to have been done a number of years earlier. The picture that the
author, Hu Pu-an, gives is one of a rural countryside in a good deal
of economic trouble, but one that is near enough to the big cities
and ports connected with overseas trade that alternative means of
employment are available, means that have particular effects on local
social organization.

One measure of widespread poverty and great socioeconomic
class divisions in Kuangtung is the "slave-girl" system (C. mui-tsai), :
which flourished up through the nineteenth century.

In Kuang-tung there are middle-aged women who specialize

in trading in women and girls. They put up their savings and
purchase attractive girls from poor families, dress them gaily
and make them up, and then send them to various ports in
Southeast Asia, where they are sold to overseas Chinese to be
concubines or slaves. The profit is not inconsiderable. The ,
girls who are thus sold are called C. mui- -tsai ("little sister")

Men who are sold this way are called C~ jyu- -tsai ("piglet")
(Hu 1936: II, 7,10).

138

This mui-tsai system flourished in Hong Kong as well, where
it attracted the attention of the League of Nations and caused a
good deal of embarrassment for the British government in inter-
national circles. Various measures were taken to eradicate the prac-
tice, over the Objections of some of the Chinese community in Hong
Kong, objections which attempted to show that the C. mui-tsai was
not really a slave, but just a household servant, and almost an
adOpted daughter. It seems fairly Obvious that it was the Kuang-tung

equivalent to the Taiwanese cha-bo-kan system, but that it was perhaps

 

open ix) a greater degree of abuse because of its capitalization and
develOpment as a business.
Chen Han-seng says that

The rubber manufacturers of Malaya are by no means the
only people who appreciate the rich reservoir of cheap labour
Offered by Kwangtung. The cheapness of that labour, especially
of women, is, indeed,appalling. From such villages as Shun-te,
Pan-yu, Chung-shan, and Sin-hwei, nearly twenty thousand peasant
girls have been taken to Hong Kong and are now working as
housemaids with the status of a mui-tsai. Other streams of
peasant girls, bought and sold, move toward the cities of the
province itself. In Swatow, the market price of a housemaid
used to be about 100 yuan . . . , and that of a concubine about
three times more; but at the present time, with the increased
inability of the peasants to support their daughters, the
average price may be said to be from 30 to 40 percent lower.
There was a time when military officers of the lower ranks made
lavish purchases of peasant girls. Now the purchasing power of
these circles, too, and of other urban classes is diminished;
and a far greater number Of peasant girls are being pawned
rather than sold outright. The price obtained by pawning is,
of course, much lower than Obtained by sale; but since the peasant
father is often not in a position to redeem the pledge, pawning
often simply represents a disguised form of selling at much less
than the customary price (Chen 1936: 112).

139

Kuang-tung in the 19305, according to Chen, had more women
than men employed as day-laborers, a certain sign of rural economic

trouble:

One of the surest proofs that a condition exists in which
agricultural production is discouraged is an abnormal prOpor-
tion of female labor in the total labor power applied to the
land--and this even in regions where field work by women is
customary. As the more mobile part of the laboring population,
it is the men who first come out from under an intolerable
burden, either by engaging in some other form of labor if
such can be found or by migrating. When rural employers
increase the prOportion of women laborers, this is a sign that
male labor has become scarcer or less docile or sufficiently
less efficient to make the employment of women preferable. An
increased employment of women may also reflect a desperate
general economic situation in which a peasant woman who nor-
mally works for her own family is forced to seek employment
for wages--even though the wages offered may not suffice to
support even her alone (Chen 1936: 103-104).

Female labor was less well-paid than that of men: in ten
representative villages of Pan-yu the wages for females were on the
average only three-fifths those of the males (Ch'en 1936: 104).

This was due to a number of factors, but one of them was that female
labor, in agricultural work which demanded great physical exertion,
was simply less productive.

In work which was not so strenuous, the labor of women was
not less productive than that of the males, and since the price of
female labor was depressed, employers could reap an unearned profit,
in a sense, by hiring females exclusively. This seems to have been?/
true in silk production; the work did not demand great labor, but i
did demand long hours of careful and tedious work. Thus it became
common to hire only women.

There is some evidence that the economic situation had much

deteriorated by the 19305, compared to twenty years or so earlier,

140

the approximate time of the collection of the data contained in the
Record of Customs of All China, if for no other reason than the
world economic depression. Even so, things were not rosy for most
peasants in the earlier days.

The argument that the depressed economic situation was in /‘=
large part responsible for the concentration of women laborers in
the silk industry is given support by the fact that during the Ming
Dynasty, before the impact of Western manufactured goods on the
domestic handcraft industry, the silk workers in Chaingsu and Anhwei
were all men (see Chung-kuo Tz'u:pen Chu-i Meng-ya Wen-t'i T'ao-lun
Chi [Collected Discussions on the Problem of the Sprouts of Capital-
ism in Ching] Peking 1957: I, 12).

This is an indication that the specialized reliance on women
laborers in the silk industry in Kuang-tung may well have been a
recent phenomenon (Ch'ing and later), and may well have been, as
Ch'en says, an indication of basic rural economic problems, and the
exodus of male laborers. This is important because, as we will see
later, the concentration of women laborers in the silk industry led
to some decidedly irregular social forms.

In a section on marriage customs in Pan-yu, Hu says that
rural women usually waited three to five years after the formal
wedding until the next ceremony, the welcoming of the bride by the
groom, took place, and that it was another three to five years before
she moved to live virilocally. This, he says, was because it was an
Old custom, but also because of poverty. Marrying in a woman took

several hundred taels of silver. The money usually was Obtained

141

from a money-lender or from a credit association. After the expenses
of the wedding itself, it took three to five years to repay the debt
incurred. The same was true of the next ceremony, the "welcoming";
only after the debt incurred then was repaid could the husband begin
to support a wife and children.

In rural areas it has become quite customary that the newly
married women return to their natal homes, as is customary, but
then do not return to the husband's home. The bride may stay
for a whole month, and then return to the husband's home for a
few days and then later she is permitted to return to his home
at major festivals. . . . Some are unwilling to do even this,
and hide to avoid it. . . . Some run away and hide permanently,
and do not return to their husbands, nor do they return to the
natal home. . . . There are also those who set up concubines for
the husband to make him agreeable to their not returning (Hu
1936: 11, 7,27).

It seems that because economic circumstances (and there were
no doubt a number of other factors at work), marriage and post-
marital residence customs changed enough, and in the right direction,
those women who were for various reasons upset with their lot
were able to avoid what they saw as the trap of marriage. Hu records
that the common pattern was for young girls, after marriage, to spend
a few days in residence at the husband's home, and then return to
live in the natal home except for the New Year's and other holidays.
In the customs of Kuang-tung a woman was supposedly not permitted to
pass such holidays in her natal home (Hu 1936: 11, 7,32). Her
parents-in-law would often use subterfuges to get her to return to
them, and she would counter with her own. The saying "The mother-in-
law is always ill, the daughter-in-law is always a bridesmaid"

expresses this. The husband's family tried to get the woman to

return on the pretext of the mother-in-law's illness, but the

142

daughter-in-law would leave again very soon on the pretext that she
had to participate in the wedding of a relative (Hu 1936: II, 7,28).

Though I wonder at the demographic validity of the exact
figures, Hu claims that in middle-class families marriages were
typically at age seventeen or eighteen, so the age of actually taking
up virilocal residence was twenty or even thirty. In poor families
this was delayed even further, so often the age of forty is reached
before a son is born, and so rich families usually were one or two
generations deeper than were poor families (Hu 1936: II, 7,28-29).

Anticipating, in a sense, what Chen later said about male
laborers, Hu says that

In the rural areas, men may marry, and a month or so later

find it advantageous to go Off looking for work abroad. The
husband is not upset at this, nor does the wife complain. They
are husband and wife in name only, and not in reality. Thus
they take this lightly. Several years later, if he returns,
only then will his wife move in to live with him. Then the men
are always upset about the age or ugliness of the wife, and so
buy a concubine. This is an aberration of family life, and a
shameful blot on customs (Hu 1936: II, 7,29).

The same author goes on to praise the rural women of Pan-yu,
saying that they take chastity very seriously, and that there were
many virtuous, exemplary, and chaste women there. Not as many as
one or two out of a hundred, he says, would remarry when widowed.
Those who did were immediately marked Off for ever. If a woman were
violated, she would think nothing of death. This was said to have
happened in the Hsien-feng period (1851-1862), when many women
committed suicide when raped by bandits.

In the really glorious cases Of such chastity, the women

just drowned themselves, and no one knew what had happened.
This is very regrettable. One still sees today those who would

143

rather die than be disgraced, and among them some have become
too radical in this. Those who are not willing to live with
their husbands, and so take poison, to commit suicide, or run
away because of a quarrel, or because the husband has used up
the savings to buy a concubine, they view the cohabitation of
man and wife as a great shame, and life, in comparison, they
treat very lightly. This is the special nature of rural women
(Hu 1936: II, 7,30).

In the difficult economic circumstances, a disproportionate
share of the burden fell on the women and young girls, giving an
added incentive to a search for some sort of alternative life-style
and source of support, something that most women would no doubt have
welcomed in the best of times.

As we have seen, one way out was the C. mui-tsai system.
Another, one of the few alternatives in the traditional economic and
political system of China, was prostitution. He mentions the "blind
singing girls of Kuang-tung":

The so-called blind singing girls of Kuang-tung are attrac-
tive girls who have been purchased by procurers and procuresses
who follow this kind of business. They teach them to sing and
dance and how to entertain customers. Then they wait until they
are fourteen or fifteen years old, and cruelly pluck out both
eyes, and give them false eyes and dark glasses to conceal them.
Thereupon the girls go out to solicit for customers to enter-
tain, and make a business of "selling songs"; they cannot dis-
tinguish between old, young, handsome, or ugly customers, and
don't entertain thoughts of running away, but instead obey every
command of the proprietor. This is really reprehensible (Hu
1936: II, 7,10).

This is of interest not only in this context, but also because it is
an indication that the custom Of buying/adopting young girls and
raising them to be prostitutes was not, as has been claimed by some,

confined to Taiwan Province. (See Chapter IV, on adoption.)

144

fl

Another alternative for women and young girls was to become’:
a servant in a wealthier family; an "amah," or C. ma-je:

You see them here and there in [certain named areas of the
city of Canton], where they are more common than are ordinary
women. Some let it be known that they are married, but are
forced by poverty into this occupation. Others are women from
rural villages who were married out, but refused to go to the
husband's household. These latter have set up their own organ-
ization, and their united strength is very great. Some of them
take their monthly income and save it in a credit association.
They may also raise a young female slave as an investment.
These are called C. ma-je mui ("amah's mui-tsai"). There are
many dark deeds in this sort of situation, and abundant evil
practices (Hu 1936: II, 7,11).

Another alternative strategy, one which has attracted the
most attention (but was not necessarily, for that, the most common),
was made possible by silk production:

In P'an-yu the land is fertile, and the people mainly make
their living from silk production. Whether rich or poor, the
daughters of the families there all know how to pick mulberry
leaves and how to reel off silk from cocoons. In one day they
can earn as much as eight or nine M. chiao, but at least three
or four chiao. The standard of living in rural areas is of
course not as high as in the cities, and this amount is enough
for self-sufficiency, and even for a bit of surplus. Such
girls, then, have a means of support, and regard marriage as
the most shameful of human affairs. Thus they form pacts not to
marry. If they are forced to marry by their parents, they do
not move to reside virilocally. These latter, after marriage,
do not have intercourse with the husband, and the next day return
to their natal family, to be companions to their sisters and
female cousins. This means that they are lost to the husband's
family (Hu 1936: II, 7,30).

It is in this context that we find the transactional social .
forms that are similar to, or are, marriage. The nonmarrying pacts
were in the form of groups, or associations, of girls. They all
entered into a kind of sworn sisterhood, called a M. chin-lan hui,

"golden orchid society," or colloquially M. k'ua—hsiang chih,

145

C. kua-seung chi, "a recognition of mutual admiration," or M. shih

 

 

p'eng-yu, C. sihk pang-yau, "workmates (society)'I (Hu 1936: II,

 

7,34).
In P'an-yu, the group of sworn sisters would attempt to
enforce the nonmarrying agreement if the parents of one of them
forced her to marry. They would at times abduct her and hide her
when the date of the wedding approached, or they might start an
uproar to embarrass the girl's family. If they were unsuccessful in
this, and the girl was/forced to marry, then their second line of
defense was to keep her from moving to reside virilocally.
Thus on the day Of the wedding, they take her and strip her
clothing off, and using long strips of cloth, bind up her body,

. and then thoroughly stitch it with thread, as a protection.
Then on the next day, when she returns to her natal family,
they investigate to see if she is still "sealed up" with the
thread. If it is not still in its original form, then they all
beat her, and there is nothing that the parents can do about it
(Hu 1936: II, 7,31).

In Shun-te there was a refinement; when she was bound up
"like a corpse about to be put in the coffin," they make her eat
"white melon" (M. pai-kua) and other medicines, to retard the pro-
duction of urine, so she will not have to urinate (Hu 1936: II,
7,35).

The obvious goal of all this is to try to interfere with
the consummation Of the marriage. The girl is also expected, by her }
sworn sisters, at least, to resist her husband's advances, even to
the point of using force.

Hu says that there are a number of motivations for girls to,

act this way. One kind of girl saw marriage to a man to be a great

shame, or did not want to marry because of the sufferings of

146

childbirth and raising children, or wanted to avoid the drudgery of
housework.. Another motivation had to do with the desire for a freer
life, and to indulge in promiscuity (II, 7,31).

There are also those who . . . are only using the pretext
of not marrying to advance their own profligacy and promiscuity.
The great degree of hidden licentiousness is, I am afraid,
breaking out into the open. . . (Hu 1936: II, 7,32).

In another section Hu complains about the so-called "free

woman" (M. tzu-yu n0):

In recent years the clamoring for women's liberation has
grown louder day by day. In Kuang-tung there already is the
"free woman." After becoming adult, she does not submit to the
fetters of the family, and the family also no longer wishes to “
restrict her. Those who are already married are permitted to i
be free and active. They are colloquially called "free women."i
These women go so far as to actually become promiscuous. To ‘
spend the night with the highest class of them costs no more
than ten yuan; second class, only five yuan. They intend by
these means to be able to get a good husband, and the two may
even reach an understanding that they will obtain the various
necessities for a wedding in this way. This behavior is not
considered shameful by the girl, nor is it by the families
either. Society treats it very lightly and would just as soon
forget it. This is very strange (Hu 1936: II, 7,11).

Hu says that this sort of behavior even led to the "abso-
lutely immoral practice" of abortion, a practice that, "though it has
been discussed as a support for freedom in love, yet must not be
adopted" (Hu 1936: II, 7,32).

Just what eventually happened to the majority of the non-
virilocal wives is unclear. Some, as has been noted, went to work as
servants, or amahs, in the large cities like Canton and Hong Kong.
Some must have become prostitutes, and some no doubt simply remained
in their natal homes for the remainder of their lives. A certain

number of them eventually did take up virilocal residence.

147

As for those non-virilocal wives, although they reside
with their parents, yet because continuing the kin line is
regarded as the greatest duty of the individual in our society,
they dare not flaunt their selfishness, and Openly go against
the usages of a kinship-based society. Thus those who are rich
pay a sum of money over to the husband to enable him to buy a
concubine. . . . Only then will she herself be free.

If she does not do this, then each year at festival times,
when she must visit her husband, and stay with him for a night
. . . and man and woman are together in the night, one cannot
avoid something happening. Moreover, it may be a long time
after the wedding, and far from the control of her sworn sisters.
In the quiet of the night nature will take its course. Even
steel, if it's been smelted a hundred times, may become soft and
pliable. Certainly this is very probable.

Thus the non-virilocal wife and her husband, through these
several acts of intercourse, may develOp a physical relationship,
and a pregnancy may occur. If she gets pregnant, then she cannot
but move virilocally, in the normal way.

Thereupon, the so-called "non-marrying chaste girl" success-
ively becomes married, and after marriage a non-virilocal wife,
and finally a pregnant, virilocal, housewife (Hu 1936: 11, 7,32).

A certain proportion of them, however, ended up living together
as couples, with emotional ties like those between husband and wife
(Hu 1936: II, 7,30). Hu says that he is not sure when this custom .
started, but that some say that it started among girl workers in the i
silk factories. Two girls would establish a sworn bond, and unite as
a pair. The other women would imitate this, and soon it had even
spread to the daughters of influential families. This, says Hu,
went too far, and became something of a custom.

In the drawing up of the contract, both sides must agree,
much as in the legal form. If both parties are interested,
then one will prepare . . . gifts Of respect, to symbolize her
sentiments. If the other side accepts them, this constitutes
consent. If not then it is refusal. When the contract is writ-
ten up and put into effect, if they have any savings, they invite
friends for a night's feast, and the friends all congratulate
them. After this they are inseparable, day and night, and happier
than any married couple. After the contract has been established,
if there is any difference of Opinion, and one of them wishes to
go back on the agreement, then they must submit to the judgment

148

of one of their sisterhood leaders (M. niang-tzu chUn, "amazon").
Usually the punishment is some kind of beating or di§grace.
Thus it has become a kind of customary law.

As far as their cohabitation is concerned, though they
cannot completely live as do man and wife, yet they can actually
have the pleasure that men and women do, either, it is said,
through using massage and caresses, or through using mechanical
aids. This is not a refined account, and it is difficult for
one of the elite (M. chin-shen hsien-sheng) to speak of it.
These women even select descendants, instituting a daughter to
succeed to their estate, and later their adopted daughter will
also enter into one of these compacts, as though it were a
daughter-in-law, and it is all as if there were a blood rela-
tionship. This is really strange (Hu 1936: II, 7,34).

 

Hu was not the only one to notice these customs, though he
seems to have recorded more specific information than anyone else
whose accounts I have seen. Another such account is reproduced in
Ch'en Tung-yUan, Chung-kuo Fu-nU Sheng-huo Shih (A History of Women's
Life in China):

Chang Hsin-t'ai, in his YDeh-yu Hsiao-chih (A Short Record
of Travels in Kuang-tung). says "The women of Kuang-tung all
call sworn sisterhoods 'golden orchid societies' (M. chin-1an
Ngj). After a woman has been married out, she returns, as is
customary, to her natal home, and then does not return again
to the husband's family. Sometimes, in fact, those who have
not completed the marriage must wait until her sworn sisters
are all married out, and only then can she return to her hus-
band's family. If they try too hard to persuade her to do so,
then all of the sworn sisters will band together and forbid
it." He also says "In the last ten-odd years, the custom has
taken yet another turn; two of these sworn sisters live together
in a close relationship, and moreover in their cohabitation, one
of the women acts the role of the husband. This custom began in
Shun-te, then later spread its influence into the P'an-yfi and
Sha-chiao areas, and its followers became even more widespread,
so that Canton itself could not avoid it. It is also called a
"sworn agreement" (M. pai-hsiang chih), and after entering the
agreement, they enter into a close union, whose pleasures exceed
even those of married life. They even spend their entire lives
this way, without marrying." ‘

To refrain from marrying because of lesbianism is really
unnatural, and very harmful to the women's health, but after
the great changes in the circumstances of earning a livelihood,
women not getting married at the proper time, and falling into
lesbianism, has become much more widespread, and is truly a big
problem (Ch'en 1933: 300).

 

 

149

Institutionalized male homosexual unions were evidently much
less widespread, though at the same time these are also precisely
the kinds of union that one would expect to be least talked about
and written about. This indeed is a general problem for anyone try-
ing to investigate many aspects of social life in traditional Chinese
society. Only a small proportion of the people were literate, and
the use and contents of literary works were heavily influenced by
socio-political factors. It is no accident that, as we have seen
elsewhere in this study, the standard histories and other works tend
very much to ignore nonstandard social forms and processes, and have
very little to say, and little of even that is reliable, about non-
elite aspects of social life. It is often hard, for example, to find
in these works material on peasant life that is specific enough to
be genuinely useful.

This has been acknowledged by historians as a great pity,
but not something about which anything much could be done. Recently,r?
however, attention has been paid by some historians to the use Of such]
things as short stories and plays as sources of historical material
for the study of the "other side" of traditional Chinese society.

There is another special significance to using short

stories and plays in the study of the social and intellectual
history of China. Because they were a form of pOpular litera- .
ture, then though they were not taken seriously by the literati’y
of the past, still they were very widespread among the masses, ’
and their influence was very great. Also, the audience that

they were aimed at was not limited to the privileged ruling
class, but also included peddlers, clerks, and, in fact, all }
classes. Thus the thoughts and sentiments these works reflected
were very broad and their influence went far.

At the same time, though the study of history in China is

more than 2,000 years old, yet the historians and compilers of
the past were under the management of government authorities,

150

and so not only were most of their materials from official

archives, their viewpoints were always those of the officials.

If we wish to enlarge and deepen our understanding of past his-

tory, if we wish to obtain new knowledge of past social phe-

nomena, and if we want new facts about the lives of peOple of

the various social classes, then we must go beyond the limited

and biased histories and archives of the past, and boldly search

for new material and new research methods. The use of material

from pOpular literature to study history must be one of these

new attempts (Lo 1975: 56).

One author who will no doubt hold an important place in the 3

study of the non-elite aspects of the late Ming and early Ch'ing is
Li YU. Li lived from 1611 to 1679, and so his life straddled the
two above-named dynasties. He was a playwright and literary critic,
and led an interesting and rather unorthodox life, well described by
LO (1975). From writing plays it was a short step to writing short
stories; Li in fact held that the two were really no different, and
titled one of his collections of stories The Silent Play_(Wu-shegg
Ngi). Li is perhaps best known in the West as possibly being the
author of a mildly (by current Western standards) erotic novel called
The Prayermat of Flesh (Jou P'u-t'uan). This novel was translated
into German by Franz Kuhn and then retranslated into English, with the

title The Before-Midnight Scholar, and used to be available in

 

paperback in supermarket book-stands in the United States.

For our purposes an important aspect of his literary philos-
ophy is that he advocated a sort of "literature for the masses." He
divided literature into standard and popular. In his own words,

Stories and dramas are not comparable to essays. The
latter are written for those who are literate, and of course
it is quite all right if they are profound. Plays are written
for both the literate and the illiterate to see, including
illiterate women and children. Thus they should be simple, and
not profound. If we use the form of the essay, and still intend

151

that the literate, the illiterate, even women and children, all
appreciate it, then even the classics passed down from the ancient
sages will be simple, and not profound, like the short stories of
today (Lo 1975: 63).

A natural outcome of this philosophy is that Li held that
popular literature should appeal to the emotions and should be real-
istic. In his own words:

The way of the ancient kings was rooted in human nature.
Those who write plays and stories must restrict themselves to
what is immediately observable through their own eyes and ears.
They must not have recourse to things and events outside their
own experience. Whether in poetry, plays, ancient or modern
literature, it is all the same. Those who illuminate the work-
ings of human nature will never be forgotten (Lo 1975: 63).

People say that the affairs of Old are all true, and those
related of today are all false. I do not agree. If a writer Of
stories and plays is not factual, it's usually just because he's
being allegorical. If you want to motivate people to be filial,
then you use as an example a well-known filial son. But you
need not record everything about him, just (appropriate) actions.
You can then judiciously choose from among the rest to add (where
appropriate) (LO 1975: 64).

The ancient poets never failed to make literary allusions,
never failed to include the names of people, and never failed
to write with fixed phrases. There is another way, however, to
use these allusions, names of people, and fixed phrases. If one
refrains from using remote allusions, from using the names of
Obscure persons, and takes phrases from the speech of the
streets, then it may happen that in one's writing there will be
dialogue and phraseology that are familiar to the ear and tongue
and, though it be literature, it is still no different from what
is said in the streets (Lo 1975: 64). «

I am citing Lo and Li himself in such detail to establish
Li's credentials as an obserVer, and to show that he in particular is
a good source of the sort of information discussed by LO. Li is
still looked upon by many Chinese scholars of a traditional and
literary turn of mind as rough, crude, and not quite acceptable.
His style, though, is, to my eyes at least, vigorous and direct.

He writes in the vernacular, or M. pai-hua, and often uses simplified

152

and nonstandard characters, and many colloquial expressions. This
is evidence of a sincere effort to follow his own philOSOphy of
literature, but makes life hard for the non-native speaker who
wishes to read and translate his work.

In his article on Li, Lo Hsfieh-lun lists the sorts of tOpics
that Li uses in his various writings. His collection of short stories

titled Shih-erh Lou, for example, treats mainly problems of marriage

 

and the relations between the sexes; seven stories out Of twelve have
this as a topic. The important characters in the twelve stories are:
gentry (two stories), merchants (two stories), "upper-class students"
(five stories), "ordinary people" (M. pai-hsing: three stories).

The collection titled The Silent Play_treats of a great
variety of social problems. The important characters in these stories
are: "lower-class scholars" (five stories),ordinary peOple (M. p_a_i_-_
.Ngigg: seven stories), merchants (three stories), and officials (one
story) (Lo 1975: 77).

It is one of the stories in this collection, a story whose
main character is a scholar, that is of most interest to us. The
story is called "Nan meng-mu Chiao-ho San-ch'ien," which means
something like "Male solicituous and conscientious mother"; it is a
reference to the mother of Mencius, who is a model of motherly
virtue. It is story number six in the edition edited by Helmut
Martin and published by the Chin-hsueh Bookstore, Taipei, 1969.

The story, briefly, is of a young, brilliant, and handsome
scholar in P'u-t'ien Hsien, Hsing-hua Prefecture, Fukien, in the

late Ming,named Wei Chi-fang, who was a homosexual. When he was

153

quite young he managed to pass the first examinations, but was so
handsome, and so given over to seeking pleasure among his friends,
that he had little time to prepare for the M. chu-gen examinations.
When he got a little older, however, a moustache began to develop on
his upper lip, and slowly a number of such inconvenient, and unat-
tractive, things began to develop, and he found himself with more
time for study. In the examinations he was successful enough to
enter the academy, and became a "subsidizedscholarm (M. ling-shan
hsiu-ts'ai).

Later he realized that "of the three unfilial acts, to
have no descendants is the worst," 50, though he was completely
indifferent to women, he married a woman named Shih, and eventually
had a son by her, though they did not live together most of the time.
The wife then died of post-partum infection, and so he entrusted the
child to a wet-nurse to raise. Because he already had a son, there
was no need to remarry another woman, and instead, he decided to get
an attractive young homosexual as a "second wife“ (M. hsfi-hsfian, "to
replace the string on one's lute; to remarry"). He looked around for
a long time, and after a number of trials and tribulations, settled
on a boy named Yu Jui-lang, son of a poor rice-merchant.

After a number of false starts, the two became lovers, and
Wei paid a bride-price (500 taels) to Jui-lang's father, and the two
of them then took up residence together as man and "wife."

When Jui-lang approached puberty, and began to feel the
stirrings of sexual desire, he started to masturbate while Wei was

having anal intercourse with him. Wei became alarmed at this,

154

saying, in effect, that a young man has only so much yggg_in him,
and if it is thus spent, he will age and die prematurely. What
alarmed him even more was, as he said to Jui-lang, that the boy,
though happy with Wei, would soon start to see women in a different
light, which is only normal, and that because this is so they were
doomed to part.

This sort of talk upset the boy greatly, as he had come
genuinely to love Wei, and was happy living with him. When Wei was
away on a business trip, Jui-lang determined to remove this threat to
their relationship, and so took a knife and totally emasculated him-
self. Wei returned just in time to find him in a pool of blood, but
still alive. He rushed the boy to a doctor, who was able finally to
save him. It is noted, however, that the scar that remained looked
very much like the external genitalia of a woman, and that the boy
Jui-lang took to dressing like a woman, and to staying home to
attend to domestic affairs.

There was a brief "run-in" with the local magistrate, who
accused Wei of doing this to the boy on purpose, violating one of the
important Confucian commands that one should not harm the body,
which is a gift of one's parents, and making it impossible for
Jui-lang to have descendants. This, however, was cleared up, and
the two returned home, whereupon they retrieved Wei's son from the
nursemaid, and, essentially, I'lived happily ever after."

This is, of course, fiction, and is presented neither by me
nor by Li as a true story. The setting, however, the concepts,

ideas, the descriptions of customary forms, and the names for these

155

sorts of relationships, are useful data, especially in a situation
in which there is very little else to use. i

The general concepts, for instance, that are used to explain
sex and its workings, and the nature of homosexuality, are very
interesting and important, especially when it is remembered that
these stories are intended to be popular literature, accessible,
acceptable, and understandable to the literate and nonliterate alike.
The terms in which, the concepts with which, these things are ana-
lyzed and discussed must therefore have been common in the society
generally, among the elite and nonelite alike.

I will present some excerpts in free translation.

The poem says:

It's unknown where or when the "southern custom" started.

That the role of a woman should be given to a man,

That the back should be turned and buttocks presented,

That the male role should be vigorously played, in the absence
of the female,

50 that one of the pair is left frustrated, out in the cold;
What enjoyment can be derived from this?

This is unbearably dirty and sordid; yet still ,there is another
side to it. This poem . . . is particularly intended to Offer
an admonition to homosexuals. It isn't known when homosexuality
began, or who started it, but it has been handed down to the
present, and indeed, has become a challenge to the natural
male-female relationship. How can this not be regarded as
strange?

How do we know that the male-female relationship is natural?
Simply by noting that the male's body has a projection, while
the female's body has an indentation. How was this kind of
structure created? The male and female embody the forms given
by heaven and earth; that which is in abundance supplements that
which is lacking. When this supplementation reaches just the
right point, then pleasure automatically results. How could
this sort of process be forced? After intercourse the male's
semen and the female's blood congeal together to form the fetus.
After a full ten months a boy or girl is born. How could such a
result be by chance? It is simply one of the conditions of
acting in accordance with the intercourse of ng_and yang,
modeled on the cover and support of ch'ien and k'un. It resembles
the moulding of creation; it is natural and so can be taken for

156

granted. Though this is all better left unmentioned, it is no
threat to 11, Though just dalliance, it is beneficial to
propriety.

As for homosexuality, in form there is no distinction
between surfeit and lack. In emotions there is no a5pect of
mutual pleasure and enjoyment. In the result, there is no
childbirth function. I don't know where it gets its interest.
Since its origin it has caused suffering to people, and has
brought no benefit. Of what use is it? . . . .

It might be that an older widower, poor and unable to marry
another wife, would use this as an outlet for sexual desire. A
young and handsome boy might, because of poverty and hunger,
use this as a means to make a living. These cases are under-
standable. But nowadays there are actually well-off youths who
also like it. This is much less understandable.

This practice is fashionable everywhere, but it is nowhere
more flourishing than in Fukien. Even Chien-ning and Shao-wu
north, prefecture by prefecture, and county by county, it is more
and more common.. Not only peOple follow this practice. Non-
sentient beings like shrubs and trees, influenced by the prac-
tice, do as well. Deep in the mountains there is a kind of
banyan tree, also called a Nan-feng tree (M. nan-fen , the
"southern custom," means homosexuality). In each case there
is a small tree which grows in front of the banyan tree. The
banyan tree eventually leans its trunk forward, making improper
advances towards the small tree. After a while, the advances
are successful, and a branch tightly wraps around the trunk
of the small tree. It also slowly enters into the banyan's
embrace, and the two trees become one, so that it is impossible
to separate them, even with axe or saw. Thus it is called the
Nan-feng tree (Li 1969: 349-54).

Li mentions in the beginning that there is another side to
all this, and has his hero, Wei Chi-fang, make the argument. After
describing Wei's intelligence and beauty, he goes on:

There was just one thing. The women looked at him with
burning interest, while he looked at them with cold indiffer-
ence. Why was this? Just because his nature was ruled by the
South, and was the enemy of the North. He would Often say to
others that there were seven things about women that he detested.
When asked about these, he would clearly list them.

"With makeup and powder they falsify reality. That's one.

Binding the feet and piercing the ears, they alter their

physical nature. That's two.

Their breasts;hxtout, are appendages like drooping tumours.

That's three.

One can't get away from them; they're like clinging vines.

That's four.

157

One is tied down by women and children, and has no freedom.

That's five.

After menstruation the bed is wet and clothes dirtied.

That's six.

In childbirth they get stretched without limit. That's seven.

How can this be compared with the attractions of a handsome
man? For better or for worse there is absolutely no artificiality.
From head to toe he is completely natural. I can go anywhere with
him, and we don't arouse suspicion or run into difficulties.
How nice it is, forming a clean 'husband and wife' pair (M. t§g_
i-tui chieh-Ch'ing fu-ch'i)."

His listener, agreeing with everything else, fears that in
that word "clean" (M. chieh-ch'igg) he has gone too far.

The response is that those who do not like it, think that
it is unclean. For those who do like it, in the contemplation
and in the experience of it, there is a different feeling entirely.
This argument is easily made for those who have experienced it,
but it is difficult to explain to the uninitiated.

The listener, not wishing to get into a heated argument,
just let it go at that (Li 1969: 357-59).

 

 

Jui-lang and Wei met at a Matsu festival that was held
annually, at which the homosexuals gathered, to look over the year's
new "crop" of young boys; they even wrote up a catalogue of the
names, addresses, and descriptions of likely candidates, and selected
the most attractive. Jui-lang of course came out well in this con-
test. His father, however, had been reluctant to let him go to the
festival, but finally did so on the understanding that he should be
accompanied by some neighbor boys, and that he should not go off to
any secluded place to chat with any man.

Why was this? One must understand that with Fukien homo-
sexuals, as with women in general, first marriages are dis-
tinguished from later ones. For a virgin, men are prepared to
pay a much higher brideprice (M. p'in), and to carry out a
complete and prOper wedding ceremony. If, then, one is not
strict in controlling young boys, and they are seduced by some-
one, then they are called "fallen" Uh.pai-1iu ts'an-hua).

They do not then become completely worthless, but a buyer

(M. shu-chu) will generally be interested only in a casual and
temporary relationship, and will not enter into a permanent
marriage relationship (M. ch'Ueh-p'iflg). and he will not select
the boy to come to live with him. Thus Jui-lang's father could

158

not relax his guard; he intended to raise Jui-lang for future
sale.

Moreover, since the publication of the "catalogue of young
boys" (M. mei-t'ung k'ao-an), everyone knew that Jui-lang was
selected as the winner; alT of his schoolmates, except for the
really poor ones, couldn't help but fantasize about him, and the
others, who were fairly well-off, all drooled over him. Friends
had long since made this all known to Jui-lang's father, and he
said to his closest friends, "It was very unlucky that the boy
was born in this evil place; I don't suppose that one can avoid
the influences of customs. I will just have to swallow my
pride, and will have to put up with this barbaric custom of
marrying for money. I am burdened with a debt of three hundred
taels, funeral expenses of one hundred taels, and I need one
hundred to prepare for my own funeral expenses. If anyone can
afford a brideprice of five hundred taels, tell him to come
along. If he can't afford that, tell him to forget it."

From this time on, he increased his control over Jui-lang,
not only forbidding him to leave the house, but also forbidding
anyone from coming to see him.

In Fukien it happens that homosexuals pay a brideprice,
but usually it's no more than symbolic; several tens of taels at
most, but as little as several taels. It is a symbol of entreaty
(M. hsiang-ch'iu). How could one be willing to pay five hundred
taelsT‘brideprice for a boy? Everyone looked on this as being
real greed, and they all talked about it. Those without many
means said "Does he think that boy's ass is inlaid with gold
and plated with silver? Why doesn't he just marry a good and
virtuous wife and forget this boy? He's of no use." Those of
means said "The boy is not yet sixteen years old. It's better
to wait a few years until he gets even poorer; the price will
naturally go down then."

So Jui-lang's wedding date was delayed a while, during
which time he could only think of Wei, but could not see him
(Li 1969: 374-77).

 

 

Eventually, however, Wei was able to sell some land and

houses, and so come up with the money. He paid it over to Jui-lang's

father, who was then able to settle his debt, meet the funeral

expenses, and still put away an amount for his old age and own

funeral expenses. This was all said to be illustrative of the old

saying "with a son all things are complete" (M. yg_tzu, wan shih tsu).

Wei then selected an auspicious day, and on that day Jui-

1ang was taken into his household (M. kuo-men). The form of the

159

wedding-night ceremonies was said to be nothing out of the ordinary
for contemporaneous marriages, and three wedding verses (M. ch'e-

chang shih) are quoted as illustration; they are full of double-

 

entendre, and more-or-less veiled references to the postures to be
used that night by the couple, and so on (Li 1969: 384-86).

The term sometimes used by Li in this story to mean homo-
sexuals is M. lung-yang. Van Gulik, in Sexual Life in Ancient
.Shigg, has the following to say about the term:

Some profligate princes . . . kept young boys as catamites
(M. lOan-t'ung), or had homosexual relations with adult men.
Sources of the Han period and later aver that a number of min-
isters designated as pi, "favorites," had homosexual relations
with the ruler. However, the term pI_has the general meaning
of "a man--or woman--who obtained the master's favor by fawn-
ing upon him and by encouraging him in his vices"; whether, in
the case of a man this implied a homosexual relation is diffi-
cult to ascertain, since the ancient texts are so terse as to
leave room for widely varying interpretations. An exception
is the case of a certain Lung-yang-chUn, who in the fourth
century B.C. served as minister of the prince of Wei; a passage
in the Chan-kuo-ts'e, a source of the third century B.C.,
proves that he had indeed homosexual relations with his master.
Lung-yang-chUn has become so notorious in Chinese history that
lun - an is used as a common literary term for male homosexu-
ality (Van Gulik 1961: 28).

 

 

What is of most importance is the degree of institutionali-
zation and permanence that these relationships had. Homosexuality in
and of itself is of no sociological interest; the social relation-
ships that it creates or affects are. Leaving aside all the data on
such things as the role of male actors (who were traditionally
expected to be homosexual) and male prostitution (see Wang 1934),
there is information that supports Li's claim that there were insti-
tutionalized, marriage-like, homosexual unions in China. The evi-

dence is fragmentary, but significant. It might be argued that some

160

of the later references were actually based on a reading of Li's
work. Many of the later references, however, use additional terms
not used by Li:
The term "little brothers" (M. hsiung-ti erh) is the same
as the Fukienese M. ch'i-ti ("bond younger brother"). . .

Homosexuals raise them like younger brothers. . . . Fukienese
homosexuals consider themselves as brothers (Yao 1940: 154).

 

This term M. ENII_shows up in many different contexts (see
Chapter IV culadoption); the term M. ch'i-ti ("bond younger brother“)
is common in Cantonese (C. kai-dai) and is used as invective, meaning
"homosexual," in Hong Kong today. HsU K'o says that in Cantonese it
means a male prostitute (H50 1920: XVI, 43,2).

It seems fairly clear U) me, then, that there were in late
Ming and Ch'ing China institutionalized relationships between males,
in Fukien at least, and that these relationships were often expressed
in terms of marriage, and were carried out in some of the social
forms connected with marriage. It is notable that in this story of
Li YU's, it is taken for granted that there is a clear status and
role difference between the partners in these unions. Intercourse is
assumed always to be anal, and in the physical act one party is con-
sistently active and one passive. Sociologically, as well, one of
the couple clearly plays the male role, and one the female. The
discussion of the nature of homosexuality is carried out in terms
that suggest that it is seen as a particular interplay of maleness
and femaleness, not something different from either of these two.
This brings to mind such general philOSOphical concepts as the one

discussed by Van Gulik in these words:

161

I believe that at an early date already the Chinese had
realized . . . that every man has a more or less strongly pro-
nounced feminine element in him, and every woman in her a more
or lesser develOped masculine element. And that it was because
of their recognizing this psychological truth that the Chinese
when choosing images that would express as accurately as possible
the male and female nature, preferred more complicated and waver-
ing symbols to the simplistic, definite Older ones such as heaven
and earth, sun and moon, etc. Hence they chose the "lesser Yang"
Green Dragon and the "lesser Yin" White Tiger, rather than "Yang"
Red Bird and the "Yin" Dark Warrior (1961: 42).

CHAPTER IV

FORMS OF ADOPTION

A means of recruitment to the family, in addition to mar-
riage, that was evidently always widespread and common in Chinese
society, and most certainly was in Taiwanese society, is that of
adoption. There is some problem in terminology here. In English
there is the problem of whether the forms under discussion should be
labeled "adoption" or "fosterage." "Adoption" usually refers to
transactions in which rights over a child are completely and irre-
vocably (except, in principle, in the case that there is another
adoption transaction) transferred from one person to another.
"Fosterage" tends to refer to those cases in which the transaction
transfers only part of the rights, or cases in which all the rights
over the child are transferred, but are done so only temporarily.

In some cases, as will be seen, it is clear that a particu-
lar form is "adoption," or is "fosterage." In many other cases,
however, it is not at all clear which term should be used, and
instead of getting involved in a terminological dispute with myself,
I have generally just used the term "adoption" in its most general
sense, specifying the nature of the transferral of rights where
appropriate. In any case, as a study of land and other rights in

traditional Chinese society will show, the distinction between

162

163

long-term "temporary" rights and permanent rights is sometimes a
very fine one.

There is also the problem of the comparability of the
English "adOption" and the various Chinese terms that are used.
The most common one, M. yagg, means, according to its entry in

Mathews' Chinese-English Dictionary, "To nourish; to rear; to bring

 

up; to care for. To bear children. To support; to maintain.

Oxygen" (all these when read in the third tone). Read in the fourth,
falling, tone, it means "To support, as scholars, superiors, etc.

To care for, as one's parents." Again read in the third tone, it
also has the (rare) meaning, "distressed," and finally in the third
tone it can be used for M. yggg_(different character), meaning "to
itch."

The meaning of the term M. yggg_is much wider in scope than
is the English "adopt,” even if we restrict ourselves to the usages
involving children and other persons: yggg_in many cases is best “I
translated as the English "support" or "to raise." Iggg_is freely i
and commonly used in this sense to refer to one's own (biological) ”I
children, a usage which 'hs not common to the English "adOpt." Thus
the phrase M. yang tzu, written in characters in isolation, can mean
either "to bear or bring up a son,” or "an adOpted son" (Mathews).

In many cases, then, when we automatically translate
M. yggg_as "adopt" or "adapted," we may be reading more into the
situation than is warranted. When we say that a child is "adOpted"

in the United States, certain information is communicated about

the social and legal status of that child in the adoptive family.

164

It is a big mistake to read anything like the same set of rights, or
the same status, into the situation in Chinese society when a child
enters into a yggg_relationship. In some cases the rights, duties,
and even the conventional views of the nature of the change in the
child's "substance" are quite similar in the two situations, and in
many cases they are not at all similar.

"Supporting" or "raising" a child did entitle that child
to claim certain rights in the household, but these were quite vari- [f
able. The English "adoption" implies, or connotes, something like
"making one's own," as in the phrase "I have adopted the structural-
ist approach in this analysis." An "adopted child," to English-
speakers, is in principle the same (whether or not in fact) as one's

biological child. This is what I mean by "change of substance." _

Though a child, in Chinese society, had certain rights by
virtue of having been."supported" or "raised," this might or might {
not have involved a change in substance. In some cases the rights
of a family servant might in fact approach those of some of these
children, though in the case of the servant there was no question of
transformation of substance.

This is part of the reason why Chinese adoption sometimes
seems so strange to Western eyes. We have certain expectations about
"adopted" children, and are then shocked, for example, to see
"adopted" daughters sold into prostitution. This is not an apologia:

there were indeed many cruel and inhumane practices associated with

certain forms of "adOption." To understand what happened, and

165

happens, we need to be aware that our adoption is not necessarily
their adoption.

I will begin this section, as I have other such sections,
with a survey of some of the forms of adoption found in Chinese
society generally, before going on to discuss Taiwan as a particular
case. Since my main interest is customary law in this context, I
will again rely mainly on the compilation of customary law made in the
first decades of this century, the Chung-kuo Min ShaggIShih Hsi-kuan
Tiao-ch'a Pao-kao Lu (abbreviated NSS).

 

Kuo-Fang Adgption in Chinese Society
The first kind of adoption I want to discuss is the form
most approved by the elite and canonized in the written legal codes:
M. kuo-fang adOption, or "transferring (a son) to another branch (of f
the family, lineage, etc.)." In parts of Honan this was called ‘
M. ta-fan ("wielding the funeral banner"):

When a parent dies, it is usually his son who carries the
funeral banner at the head of the procession. Thus, if someone
dies without a son, and without having designated an heir,
usually one of his nephews will carry the flag to the cemetery.

If later there are legal proceedings, this is taken as evidence
of having entered into the line of succession (NSS; 1382).
Thoughthis is, in effect, a posthumous M. kuo-fang adoption,
it raises no particular problems in customary law. In Anhwei there
was a further elaboration of this form. If one branch of a family
died out and there were two or more remaining branches, the remaining

branches would compete to nominate one of their sons to act as heir

to the discontinued branch. In a compromise situation, each branch

166

would nominate one of its sons to participate in a joint succession

(called M. kuan-chi. MSS: 1497).

 

As the saying, "those of the same surname do not marry, If
those of different surnames do not adOpt" indicates, adoption, I
according to the elite norm, should only have been M. kuo-fang
adoption. In Hu-nan:

If one without sons institutes an heir, then of course he
must select the closest relative, within the agnatic kin group,
of the proper M. chao-mu generation classification. In the
customs of I-Yang Hsien, if someone wishes to institute as heir
a more distant relative who is particularly virtuous, or is
particularly liked by him, then he must turn part of his
estate over to whatever closer relatives there are. This is
called M. kuo-fang_li, the kuo-fang present, and is to secure
the concurrence of the branch being passed over. The abuse
may arise, however, that the closer relatives insist on an
absolute right to this "present," and this is easily and often
a source of contention. This custom is not limited to I-yang,
but is also found in Ning-hsiang and Hsiang-t'an (NSS; 1693-94).

 

This sort of M. kuo-fang adoption is well known and unexcep-
tionable, so I will not give any more illustrations of it. In its
typical form a nephew is "adOpted" to his uncle. He then becomes
the heir of that uncle, and when he marries and has children, the
children are deemed to be continuing the uncle's kinship line.

There are illustrations of this sort of adoption in the Liao lineage
from T'ien-sung-p'i and the Ch'iu lineage. Though M. kuo-fang adop-
tions are usually of males, there is no logical or legal problem in

adopting females in the same way.

There are, however, a number of variations on this theme.
The "adopter," for example, need not be alive. As there is a
"spirit marriage" in which one of the partners is deceased, so there?

also is "spirit adoption." This makes sense both religiously and i

l

167

economically. The kinship line of the dead man certainly needs
someone to continue it if his spirit is to have sacrifices and so 1
not become a dangerous force, and there may be an estate of some
sort which can be kept within the larger kin group if a related
heir is instituted.

It would seem that such spirit adOption was and is very
corrmon, but it does not "get into print" as often as some other forms
because often it is for religious succession as a goal, and if there
is no estate over which to contend, such an adoption would have little
or no effect outside the family religious Observances.

In parts of Shensi,

One who died with no heirs, and without designating an

heir, had no one to act as the master of the funeral ceremony.
The kinsmen would discuss this, and then select someone of the
proper M. chao-mu generation classification to succeed him and
to "take over the brazier." At the time of the burial, he
smashes the brazier, and accompanies the coffin to the ceme-
tery. The kinsmen then all recognize him as the successor to
the dead man. A11.prOperty left by the dead man reverts to his
possession, and no third party can contest this. This is
called "taking over the brazier" or "taking possession of the
family's share" (NSS; 1722).

We have seen that several nephews, or other relatives, can
be adopted simultaneously as M. kuo-fang sons to jointly succeed to
a branch of a family, and it was conversely possible for two or more
branches, or M. fang, to be jointly continued by a single male.

The simplest of these forms was that which consisted of
one male continuing two lines, most typically that of his father

and that of his uncle. This was referred to by a number of names,

most commonly M. shuang-t'iao or M. ch'eng-t'iao, the former meaning

 

168

"to continue two ancestral halls, or lines," the latter "to succeed
to (another) ancestral hall, or line."

It is in this context that we find the one exception to the
rule that Chinese society, as far as the jural rules were concerned,
was monogamous. The wife was always carefully distinguished from
the concubine (or, to put it another way, the primary wife was
always distinguished from the secondary wife). To marry more than
one primary wife was to commit bigamy, and my impression is that
such bigamy was no more common in the traditional society than in,
say, the contemporary United States.

There was an exception, however. If a man died before having
fathered a son, it was permitted that his nephew, or some other
agnatic relative of the proper generation, marry two ines (primary
ines) if he was, through M. kuo-fang adoption, instituted as heir to
his uncle. His children by one wife would then follow his father's
kin line, and his children by the other would follow the dead uncle's
kin line.

Strictly speaking, this was not M. kuo-fang adoption, as
the latter involved the total transfer of a child from one branch
of a family or lineage to another branch, while the M. shuang-t'iao
son simultaneously continued both lines. The distinction is not,
however, of great moment here.

This concept of "double inheritance and succession" could

also be generalized. In parts of Shensi:

169

If there are nine brothers or agnatically related cousins,
but among these nine eight lack descendants, then though there
is only one person, with only one son, and if the eight others
wish to choose this one child to succeed them, and there are
not enough males of the right generation in the kin group to
go around, or if there are, but their parents do not wish to
part with them, etc., that one child can be instituted as heir
to the nine kin lines. This is called "one branch having a son,
nine will not die out" (M. i men yu tzu, chiu men pu chfieh.

NSS; 1722-23).

 

The same thing is reported for another part of Shensi, where the
number is ten instead of nine (NSS; 1749).

Thus far the discussion has concerned adoption within the
agnatic kin group. It was also possible to go beyond the agnatic
kin group proper to find an heir:

In various Hsiens of Chiang-hsi, it is customary among the
masses that those without sons or grandsons, who wish to insti-
tute an heir, can do so by inviting (M. chao) a sister's son
(M. wai-sheng) to continue the kinship line of his mother's
brother. He obtains, moreover, the estate left by his uncle,
and changes his surname to that of the uncle. The kinsmen can-
not interfere in this. Since it is a blood relationship, it
is, for those who are particular about such things, different
from the case involving a child of a different surname (sic;

M. i-hsing chih tzu). This kind of extraordinary arrangement
is therefore recognized by the members of the kin group (NSS;
1499-1500).

 

In Feng-hsiang Hsien, Shensi,

One without sons can, with the agreement of his relatives,
have his sister's son succeed him. After he has done so, how-
ever, if a son is then born to the man, then the instituted
heir cannot enjoy equal rights in the division of the estate
with the son (NSS; 1763).

Something similar is reported for all of Kansu; "If the

household is dying out, then the son Of one's own daughter can
inherit the rights in the estate" (NSS; 1768). This last is very

similar to the custom Often called "borrowing the child and returning

170

the grandchild." In various Hsien in the Pao-ting area of the
prsent Hopei,

The system Of "borrowin the son and returning the grand-
son" (M. chieh tzu, huan sun? is followed when a person has
already undertaken to continue another line, and then the origi-
nal line finds itself lacking heirs; if that other line has
sons, one of them is returned to the original line to continue
it (E13; 1299).

This was a very widespread custom, and was well represented (as we
shall see) in Taiwan. Sometimes it went by different names, and
had somewhat different forms. In Chilin (Kirin) generally, it was
quite common:

If we look into the custom of designating a grandson as
heir when a son is lacking, we find that it is basically not
permitted in law. In the whole Province of Kirin, however,
it is very common, and the relatives regard it as quite natural,
and raise no objections to it (NSS; 1312).

In Hei-lung-chiang it was said that this was permitted only
in the case where one had a son, but he died; then his son, one's
grandson, could become heir. If one had never had a son, and this
was done through institution, then it had no legal effect unless a
majority of the kinsmen agreed to it (NSS; 1323, 1338, 1352). In
other areas such a proviso did not apply (NSS; 1348). These
arrangements caused trouble because they were thought to confuse the
distinction of generations within a kin grouping. Since such a
distinction was very important for various aspects Of kinship organi-
zation, it was quite serious.

It has been noted above that females as well as males could
be adopted within a kin group (M. kuo-fang , and in the section on

uxorilocal marriage it is pointed out that it was very common to

171

adoptadaughter with the idea of then inviting in an uxorilocal
husband for her. In Kan-su, we find the logical combination of
these:

. . . Those who are aged and without sons may invite in an
uxorilocal husband for a daughter, and make him heir. If there
is no daughter, and there is no one among the kinsmen who can
succeed to the line, then one can adopt in a niece to continue
the line . . . and invite in an uxorilocal husband for her
(NSS; 1783-84).

All of these different forms of adoption have concerned
males and females related to the adOpter in one way or another.
Except for the cases involving females, this was generally the
jurally standard form of adOption in the traditional society.

The adoption, however, of unrelated, different-surname,
males and females was very common. In HsUan-en Hsien of Hupei,

. The customs of succession treat kinship with great
importance, and give little weight (to the rights of) those
of different surnames. If one has no sons, then one must first
go to one's closest kin to find a male to adopt (M. kuo-chi).
If they have no sons that can do so, then one must go to more
distant kin to find one. If they also have none, only then can
one institute as heir a male of a different surname. When one
adOpts such a child, one must first get the permission of the
kinsmen, and only then will he have rights in the estate. If
the agreement of the kinsmen has not been obtained, then though
he has been (formally) instituted as heir, still he will have
no rights in the estate (NSS; 1672-73).

We find, however, that there is great variation in this, and
the customary arrangements in some cases scarcely sound like des-
criptions of Chinese society at all; in various places in Hunan.

Succession is not limited to those of the same agnatic kin
group (M. t'ung-tsung); one's sisters' sons, one's daughters,
and the sons of oné's wife's sisters or brothers, though they
are of different surnames from one's self, all can succeed,
become heirs, and be entered into the genealogy. Except for
the closest of them the relatives cannot object. This is,
just because of their blood relationship with the husband or

 

172
the wife, to bring other surnames, and thus confusion, to the
kin groups; it is forbidden by the current code (NSS: 1680).
There are other reports of different-surname adoption from

parts of Shensi (NSS; 1706, 1721, 1724, 1729, 1748, 1761, 1764, 1765),
Kansu (1768, 1777-78), and what is now HOpei (1805, 1809, 1810, 1816,
1819, 1821). I think it safe to say that it was common throughout
China, and perhaps more widespread than the more “standard" M. Egg-

fang variety.

The "Adopted Daughter-in-Law"

The T. gimggg, or M. t'ung-yang-hsi, is a woman who is
"adOpted" into a family with the idea in mind of later marrying her
to one's son (who in turn may also be adapted). This was a very
widespread custom in Chinese society, and was found in many geo-
graphical areas. It is sometimes said to be a peculiarity, or to be
peculiarly well develOped, in Taiwan (see Taiwan Szu-fa II, 188),
but the Chung-kuo Min Shang Shih Hsi-kuan Tiao-ch'a PaO-kao Lu
says that about 80 to 90 percent of households in Shun-ch'ang Hsien
of Fukien province had a simpua (MSS: 1578), and goes on to report it

for the following areas:

Chih-1i, various hsien p. 1293
HO-nan
YU-nan I-tao p. 1377
Sung Hsien p. 1385
Shan-tung
Li-ch'eng p. 1390
Tung-a "
Te-p'ing "
Shan-hsi
Wu-hsiang Hsien p. 1437
Kao-p'ing Hsien p. 1453

173

An-huei
Wu-wei Hsien
Ho-fei Hsien
T'ai-hu Hsien
Kuang-te Hsien
Ch'iu-p'u Hsien
Chiang-hsi, various Hsien
Nan-ch'ang Hsien
Hsin-chien Hsien
Che-chiang, various Hsien
T'ai-shun
Chu-yfian Hsien
Fu-chien
Chang-p'u Hsien
Lung-hsi Hsien
Cheng-ho Hsien
P'u-t'ien Hsien
Shun-ch'ang Hsien
Ch'ang-lo Hsien
Ch'ung-an Hsien
Hu'pei
Han-yang Hsien
Chu-ch'i Hsien
Ma-ch'eng Hsien
YDan Hsien
Ku-ch'eng Hsien
Chu-shan Hsien
T'ung-shan Hsien
Ch'ien-chiang Hsien
Pa-tung Hsien
Kuang-ch'i Hsien
Hu-nan
Lin-feng Hsien
Shensi
Ch'ang-an Hsien
Fu-feng Hsien
Ch'i-shan Hsien
Hsing-p'ing Hsien
Wu-kung Hsien
Tso-shui Hsien
Lo-nan Hsien
Ho-yang Hsien
Kan-su, whole province
Je-ho, whole province

p.
p.

'U U'U'U UUU'U'U 'U

'U U'UUU

1475-76

1480-81
1491
1495
1501-4
1506

1525
1560
1568

1571, 1583

1576

. 1578, 1579
. 1591
. 1505

1619
" , 1628

. 1639

" , 1642

. 1712

1772, 1775
1805

A more complete and detailed survey could be prepared if

there were any reason to do so, but the simple point that I want to

make with this list is that this custom is far from being peculiar

174

to Taiwan, but was found throughout China. Some of the accounts of
this custom recorded in the Min-Shang Shih (MSS) are Of considerable
interest.

In various parts of the present Ho-pei the adOpting family
did not call her "daughter-in-law," but M. ya-t'ou ("slave girl")
because when she was very young, they didn't know whether or not she
would actually become a daughter-in-law. After she actually married
a son of the family then they called her M. t'ung-yang-hsi (MSS:
1293). In rural areas of HO-nan, among poor people, daughters could
be sent in adoption to other families after a go-between had arranged
the exchange of marriage contracts between the two families. Later,
when the two had reached adulthood, the two families decided on an
auspicious time to carry out a regular marriage ceremony, with
invited guests, etc. (NSS; 1377).

In parts of Ho-nan it was noted that it was important to
have a written contract for such transactions, in case disputes later
developed (NSS; 1384). In areas of Shan-tung, it was recorded, poor
families, to save on wedding expenses, sent their young daughters to
the husband's family to be raised to adulthood, then married

(M. ch'eng-nun, or "complete the marriage"). At this time she would

 

sometimes return to the natal home, and an auspicious day would be
selected for her to be formally taken to the man's family (M. ngg;_
gNIQ). This, it was noted, was called M. t'ung-yang-hsi, "daughter-
in-law raised from youth," but another name was M. t'uan-yUan hsi-fu,

"incorporated daughter-in-law" (NSS; 1390-91).

175

Poor families in certain parts of Shan-hsi, with an eye to
the gifts and money to be obtained, would betroth daughters to other
families, but because of the young age of the girls, they didn't
formally take her to the man's family (M. ying-ch'fi), but they would
exchange the marriage contracts, and then send her to the man's house
to be raised. When she was of age, then they would perform another

wedding ceremony. This was called M. t'ung-yang_hsi, "daughter-in-

 

law raised from youth," and was often the cause of lawsuits over the
gifts and money, or some other reason (NSS; 1453).

In Chiang-su the M. t'ung-yang hsi was called M. miao-hsi,

 

"infant daughter-in-law." The marriage relationship was created by
the exchange between the two sides of the marriage contract, and if
either side should later regret its action, it was difficult to get
out of the agreement (NSS; 1459-60).
In An-huei, all that was needed for adopting a girl as future
daughter-in-law was to have the girl's parents draft a horoscope
(M. pa-tzu, "eight [horoscope] characters") for the girl and send
it to the adopting family. There was no regular marriage contract,
whether or not the adopting family later married her to their son
or an adopted son (NSS; 1476). The report for Ta-hu Hsien in An-huei
says:
There is a definite time for marriage for all men and
women. Though the Code prohibited "prenatal marriages" (mar-
riages by "pointing at the belly" and "cutting the garment"),
in Ta-hu it is very common for those without sons to adopt and
raise a daughter-in-law. It is said that if it is difficult
to give birth to a son, then adopting a daughter-in-law will
make it easier. Thus she is called M. wang-lang_hsi, or

M. ten hsi (a daughter-in-law waiting for a husband). If later
a son 15 orn and married to the girl, then no marriage contracts

176

and no go-between are used, and the two sides seldom dispute.
Sometimes the son may be born late. If the girl is already
eight or nine years old before he is born, then it is impos-
sible to avoid a great age discrepaney. If she reaches mar-
riageable age and a son has still not been born, then her
status can be changed to that of an adopted girl (M. yi-nU),
and she can be married out. Her natal family, however, will
then want to arrange the marriage, and this causes many law-
suits (NSS; 1480-81).

For Kuang-te Hsien (also in An-huei) it is recorded that the

M. t'ung-yagg hsi was sometimes adopted to the man's family after she

 

was old enough to do manual or household work, and sometimes,
because of poverty, she was sent when only a few months old. In
neither case were marriage documents or a go-between used, and
"therefore disputes and regrets are seldom seen" (NSS; 1491).

There are a number of interesting and revealing accounts
from Chiang-hsi, free translations of which follow:

Those who adopt daughters-in-law to avoid expenses are
about 50 to 60 percent. Apart from the M. t'ung-yang hsi there
is also the M. hua-teng n0 (this is a daughter-in-law adopted to
wait for the birth of a son of the family to be her husband,
and the term is based on the saying "transplanting a flower, and
waiting for a son," M. ch'a to hua-erh tai erh shegg; sometimes
called M. hua-tun n0, where M. tgg_is another character for
M. teng, "waiffli. This kind of adOpted girl is also called
M. wang-lang hsi. This kind of custom is found in every area,
but is particularly flourishing in Chiang-hsi. If after adOpt-
ing her there is no son or instituted heir to marry her to,
then the adopting family may select a husband to marry her out
to. If the husband is not suitable, or if they propose to marry
her to their adopted son, then her natal family will certainly
intervene, and the majority of the wedding gifts will go to
them. . . . This form is rarer than the M. t'un - an hsi, but
still amounts to 30 percent or 40 percent; "regular" marriages,
apart from those among wealthy families, account for no more
than 10 or 20 percent (NSS; 1501-1502).

 

 

 

 

In Chiang-hsi there are two types of marriage, minor and
major, so the associated marriage documents differ. The minor
marriages are M. t'ung-yang hsi marriages and M. hua-teng nu
(the latter is an adopted-in daughter-in-law waiting for the
birth of a son to the family to be her family). Once the

177

gO-between has achieved an agreement then the man's family
adopts the girl, so in the marriage contract (M. keng'shu)
there is written only the girl's horoscOpic characters. In
major marriage, the two sides should discuss the time of mar-
riage and the girl's family should send her "eight characters"
to the man's family. This is colloquially called the "draft
eight characters." If the marriage negotiations are unsuccess-
ful, this must be returned. If successful, then it is evidence
of agreement to marry. When it comes time for the formal wel-
coming of the bride (M. ying-ch'U), then a few days before, the
man's family writes his "eight characters" on the side of the
marriage contract and sends it via the go-between to the girl's
family. Her family then writes her "eight characters" on it.
Only then is it completed. This is traditional, but there is
no fixed form. . . .

On investigation in Chiang-hsi, it is found that the mar-
riage gifts in regular, major, marriage are very expensive, so
apart from among the very wealthy families, major marriages
account for no more than 10 percent or 20 percent, and minor
marriages 80 percent to 90 percent. Also, in major marriages
conflicts and disputes over the incompleteness of the marriage
contract are common. As for remarriage out, the marriage con-
tract given to the man's family in the first marriage is used
as evidence. The form and contents or its articles are no
different from those in a contract of sale (NSS; 1502-1503).

The pairing of ng_and yang is the constant principle of
heaven, and the prOper timing Of marriage was important in the
old rites, but in Chiang-hsi, because of poverty and the diffi-
culty of making a living, there is the custom of taking a woman
in preparation for her later becoming a daughter-in-law. This
is called M. hua-tun n0, where M. tgg_means to graft or trans-
plant branches. This kind Of custom is a very common cause of
lawsuits; it is an evil, male—chauvinistic custom. Poor families
with a daughter cannot take on the duty of supporting and rear-
ing her to adulthood, and after a few days or few months of
suckling, look for a woman with milk to take the girl as a
M. hua-tun nu. The girl's family first prepares a marriage
contract, called the "eight characters," and asks a go-between
to transmit it to the adopting family. They then send a few
eggs and some wine. This is all there is to it. . . .

The motive for adopting such a girl may be because one's
child has died young, or because one's own daughter has already
been adopted to someone else. The goals are several. One is the
hope that she can be married to a son of the family when they
are old enough; the two sides agree, and simple gifts are sent
to her natal family as betrothal gifts, and the years of support-
ing her are treated as equivalent to the expense and trouble of
a regular marriage. A second goal is "planting jade"; to protect
the continuity of a line. This happens when no son is born, she
remains single, grows Old, and there is not much of an estate.

 

 

178

Luckily one still can adopt an unmarried son of a relative as son.
Usually her family's agreement is sought, and in accordance with
rural custom a simple marriage ceremony is carried out, and the
two can then support each other. This second situation arises
when there are too many corrupt practices in the family or because
the son was born too late and the age discrepancy between the two
is too great, or one's heir has died and the family is poor, and
there is no relative with a son to be instituted as heir to
continue the line. If the girl is not willing to spend the

rest of her life thus, then one can re-marry her out to try and
profit from the wedding presents, hoping for lots of money, and
selecting a son-in-law without regard for his quality of char-
acter. The goal of sending a dowry is to get wealth in return

to support one in old age.

It turns out that the girl's parents or other kinsmen may
struggle hard for the right to select her husband, and this often
results in lawsuits. Her parents will say "the ceremony is not
complete; the marriage contract is not clear. Though you sup-
ported her, she is still our daughter, and you must attend to
our right to select her husband." The adOptive family will say
"We worked hard and so have been compensated for our poor fate
in having no son. How dare you interfere in our affairs? We
cannot allow you to determine who she marries." In this situa-
tion, with respect to Ij_she is still an unmarried daughter.
According to sentiment, though, how can hard work not be recom-
pensed? . . . This Often comes up in lawsuits, and custom has
decided that the natal family should present twenty-four or
thirty-two foreign dollars to the adoptive family as "milk
money," and then the right to manage her marriage reverts to
them. In the Chiang-hsi area there is little female infanti-
cide, but many lawsuits over M. hua-tun n0. A serious problem
is not easily solved. This is the woman supporting the man's
family (M. lang-men nU t'un); problems apart from that of the
adopted daughter-ihilaw must also be attended to (NSS; 1503-1505).

 

 

A commentary to this goes on to say that according to the
custom in Chiang-hsi, the cost of gifts for a major marriage is over
100 gNIg, so those who can afford it, and so have such a marriage,
amount to only 10 percent to 20 percent, and that one shouldn't be
surprised, then, when ordinary people, to save on such expenses,
adOpted girls as daughters-in-law even before they had sons, and
this became customary.

For another area of Chiang-hsi, it is recorded that the

services of a go-between may be used, or the two families may exchange

179

marriage contracts, but in either case the betrothal takes effect
after she has been adopted into the man's family (M. kuo-men, Npass
through the door"), which ceremony consists only of worshipping the
ancestors (NSS; 1506).

Che-chiang had a custom called "grafting a flower"
(M. ch'a hua) that was different from the one mentioned above for
Chiang-hsi. If a family had adapted in a girl to be later married
to a son, then the kinship relation was held to already exist before
they actually married, and she was regarded as the "wife-to-be" of

the son (M. wei-hun ch'i). If, during the period after she was

 

adopted in, and before the marriage to the son, the clan genealogy
happened to be revised and updated, then she would be entered as

the son's wife. If her intended husband were to die before they
could be married, then with her parents' consent, she could be mar-
ried to his brother instead. Ihi the custom, "grafting a flower,"

a new marriage contract was not needed, the other clan members had
nothing to say about it, and when the genealogy was later re-edited,
it would have to be acknowledged.

A lengthy commentary to this report claims that this custom
was found in lower classes (M. chung-teng i-hsie chih she-hui),
because the parents were afraid that in the future their children
might not have the means to arrange a marriage, and so, planning
ahead, adopteda girl to be daughter-in-law. The attitude of the
adOptive parents toward the girl was usually the same as if she had
been their own child, and her attitude toward them was also the same

as if they had been her own parents; the relationships were very

180

affectionate and close. If the intended son happened to die, neither
the girl nor the parents would want her to marry out of the family,
so if their second son was not too different in age from the girl,
they might marry her to him. At the same time, if the relations
between the affinal families are good, then it will not only be hard
for daughter-in-law and mother-in-law to part, but it will be diffi-
cult to break off relations between the two families. Moreover, if
they marry her out to someone else, the comnentator argues, and so the
son has to find some other wife, the economic consequences for a poor
family could be very great. "There is no one who does not take for
granted that clan relatives should show pity and compassion, and so
accordingly this kind of precedent grew up; though on the surface it

smacks of an incestuous relationship (M. tu-lun chih hsien), in fact

 

there is no confusion of blood lines. Thus one can see it from one
point of view as the annulment of the old marriage contract, and from
the other point, the establishment of a new contract" (NSS; 1524-25).
For another area of Che-chiang, the custom of adapting in a
daughter-in-law is attributed to the lower classes specifically, and
then is also said to be due to an attitude which values male children
but not female. Families which either can't support their daughters,
or just don't care to keep them, would adopt them out to others,
even though they would get only a few ygan from doing so. The
commentary to the account blames excessive parsimony (NSS; 1525-26).
It is also related for Che-chiang that this custom finds its
origin in the inability of the natal family to support the girl,

and the desire to save on wedding expenses. The girls that are so

181

adopted, it is said, were Often around ten years old, in which case
the boy's family would pay out ten or twenty yuan as a gift. If the
girls were only a few months Old, then the girl's family would have

to "sweeten the pot" (M. chin-t'ieh) with three or four yuan

 

(NSS; 1568).

The section of the Min Shang Shih Hsi-kuan Tiao-ch'a Pao-
kgg_I§_on customs of Fukien Province is,of course,particular1y impor-
tant because the majority of the Taiwanese came from there, bringing
many practices with them. The report for Chang-p'u, Lung-hsi, and
Cheng-ho Hsiens says:

Families in Chang-p'u without sons will adopt in a girl
(M. miao-hsi) to later marry a son, an instituted heir, or an
adopted son of that family. If there is later no son for mar-
riage, then they must invite in a son-in-law (M. chao-chui) for
her, to carry on the ancestral worship. This is true to the
point that it is very rare to see a family with such a girl who
has not been so married.

Long-hsi even has the custom of buying boys to act as
heirs, and peOple go to the foundling hospital (M. yU-ying
t'ang) to adOpt girls as miao-hsi so that they can raise them up
and then invite in a son-in-iaw, to act as son. To invite in
another husband, because of the death of the first, is also not
prohibited.

In Cheng-ho, poor families, apart from their usual practice
of marriage by purchase (M. mai—mai hun-yin), also like to adopt
a young girl as M. miao-hsi, and raise her up to become wife to
the son. A gO-between is not used, nor is there any marriage
contract. After the marriage is concluded (M. ch'eng-hun), an
affinal relation between families is recognized as being created,
though in some cases the affinal relation is created by the
adoption (NSS; 1571-72).

The account for P'u-t'ien is particularly detailed:

In P'u-t'ien 70 to 80 percent of the lawsuits arise from
adopting daughters, adOpting daughters-in-law, or inviting in
sons-in-law. Among the people, whether or not one has sons or
daughters, most like to buy (M. chia-mai) a young girl of the same
or different surname as adopted daughter (M. yang-n0) or adopted
daughter-in-law (M. yang-hsi), and then wait until they are of
marriageable age. If it is a case of an adopted daughter, then

182

the adOptive parents will select a man of the same or different
surname (depending on whether they have a son of their own or
not) to invite in as an uxorilocal son-in-law, or as an insti-
tuted heir (M. chi-tzu; if the adoptive parents already have a
son, then they will invite in a different-surnamed man as uxori-
local son-in-law. If not, then they will invite in a same-
surnamed man, who can at the same time act as son). If it is

a case of an adopted daughter-in-law, then she is of course
married to the son specified in the contract drawn up at the
time of the adOption. If there was no contract written up, but
just a verbal agreement, and at the time of marriage [completion
of marriage, M. ch'eng-hun], the specified son is already dead,
then they can at their pleasure marry her to another son. As for
the case in which the son, or an uxorilocal husband, dies after
the marriage has been completed, then they can invite in another
husband for her.

Because of this, questions about the right to arrange the
marriage (M. chu-hun) are constantly arising. The right to
arrange the marriage of an adOpted daughter of course resides
with the adoptive parents. Whether the right to arrange a
remarriage when the first uxorilocal husband dies and the woman
wishes to invite another in resides with the dead first husband's
family and kin, or still is in the hands of the adoptive family,
is settled in the customs of P'u-t'ien by saying that the mother's
family (M. mu chia) arranges the remarriage, and the husband's
family hasn't a leg to stand on. As for a widowed adopted
daughter-in-law, the right to arrange an uxorilocal remarriage
for her also remains with her adOptive parents and their kinsmen.
For instance if A has a widowed adopted daughter-in-law, then he
has the right to invite in a uxorilocal husband for her. If this
husband also dies, then he has also the right to invite in yet
another. Any children born to the uxorilocal husband cannot be
disposed of in marriage by the husband or his family. If they
attempt to arrange such a marriage, and try to get the wedding
presents, then the mother's family must, according to custom,
intervene and 5tOp them. In P'u-t'ien these have already become
crucial matters, and so this special account is given here.

The husband invited in to live uxorilocally, no matter
whether he is an only son or not, must, if his elder kinsmen have
given their permission and allowed him to marry thus, submit to
the direction and supervision of the woman's family, has the duty
of permanently residing with and supporting that family, and has
no right to interfere with the management of the family or its
estate. Nowadays this custom is found not only in rural areas,
but also in the towns and cities. . . (NSS; 1576-77).

 

In Sun-ch'ang, it is said that 80 percent to 90 percent of

the families practice this custom, called locally M. yUan-fang,

183

"completing the branch? (NSS; 1578). It is specified, for the same
area, that this custom is found mainly among the lower classes
(NSS; 1579).

For Ch'ang-p'u it is said that there were three kinds of women
who could invite in uxorilocal husbands. One was a "virgin" (a
young unmarried girl), another an adOpted daughter-in-law, and the
third a widow. It was most common for a widow to do so, and less
common for an adOpted daughter-in-law. It was more common for this
latter than for an unmarried girl, however. In uxorilocal marriage,
the report continues, it is usually specified in the contract whether
the man simply becomes a son-in-law or becomes an instituted heir and
at the same time supports the parents. In the case of a widow marry-
ing in'a husband, it was usually just to be a husband. In the case
of an adopted daughter-in-law, then even a lineal relative of the
mother could be married in. This was sometimes true of widows as
well (NSS; 1582).

In Ch'ang-lo poor families adopted young girls with, it is
said, the goal of saving on wedding expenses. The term used here was
M. miao-hsi, "infant daughter-in-law." The two sides needed only
agree to the match, and then some meat, noodles, and a few gifts
were sufficient as betrothal money. Later, when they were Of age, a
lucky day was selected for the consummation of the marriage (m. Ng;_
gNig). In very poor families the last day of the year was selected
for this (NSS; 1591).

In Hsia-p'u there was a custom of pawning six- or seven-year-

old girls to others to work as servants (M. pei-nU). A contract was

184

written up which clearly indicated the time period, perhaps eight or
ten years. The price was not fixed, but might have been four or
five to ten yggfl, When the period had expired, then she could be
redeemed for the original price. If she died during this period,
then if it was because of illness, no more was said. If it was
because of maltreatment, though, the pawner had to be compensated
(NSS; 1604).

A notice for Hu-pei (Han-yang, Chu-ch'i, Ma-ch'eng, Yuan,
We-feng, and Hsing-shan Hsiens) describes the ceremony used in adopt-
ing a daughter-in-law:

The ceremony for adopting a daughter-in-law in Han-yang
Hsien is as follows. If the girl is under ten years of age,
then she just worships the ancestors alone. If she is above
ten, but still not of marriageable age, then she and the husband-
to-be do so together. Later, the time of consummating the union
(M. yUan-fang) is not fixed, but is at the latest at twenty
years of age. The ceremony consists of sending the girl back
to her natal home, and then at the appointed time carrying her in
a sedan chair to the man's home. Then in the evening she is
taken before the ancestral tablets, and the two of them worship.
Finally, she is taken into the bedroom for the ceremony of drink-
ing the nuptial cup (M. ho-chin).

In Wu-feng Hsien those who become adOpted daughters-in-
law are for the most part daughters of poor families. The
families send their daughters to the man's family to grow to
adulthood, and then consummate the marriage. Before the time
for the consummation has been set, it is necessary to send a go-
between to the girl's family to settle this. After it has been
agreed to, then the man's family must prepare articles such as
tea, fruit, wine, meat, and so on. A card is also written up and
sent to the girl's family, notifying them of the time that has
been selected. The man's family then arranges for sedan carriers
and horses to go to the girl's family to escort her. Her escorts
may be one or two persons of each sex, called M. Shang-ch'in.

The girl's family also sends along some bedclothes, chests, and
other kinds of articles used by women as her trousseau.

In YUan Hsien, there is no set ceremony for adopting a
daughter-in-law. The age of consummating the marriage (M. yUan-
fang) is no later than eighteen or twenty years. The girl's fam-
ily prepares a trousseau and the man's family prepares musicians,

185

wine, and a feast for the wedding ceremony. This is a proper
ceremony. Very poor families, who cannot afford to do this,
need not.

In Hsing-shan Hsien when a family adopts in a daughter-in-
law, they must employ a go-between, throw a feast for the rela-
tives, have the pair worship the ancestors, and drink the nuptial
cup, but they need not go through the ceremony of welcoming the
bride (M. yjng-ch'in). Later, at the consummation (M. yUan-fang)
they once again throw a feast for the relatives, worship the
ancestors, but there is no set time for this ceremony.

In Chu-ch'i Hsien, the ceremony for adopting such a girl is
no different from that of marriage between two adults. There is
no set time for the later ceremony of consummation (M. yUan-fang).

In Ma-ch'eng Hsien there is no set form for the ceremony of
adOption. The ceremony of consummation is not set at a particu-
lar time, but depends on when they reach adulthood (NSS; 1620-21).

 

Much the same is reported for other areas of Hu-pei, except
that it is noted that in Ku-ch'eng Hsien, when the girl is adopted
in, she is just treated as an adOpted daughter, and there is no cere-
mony until she has come of age; and then the consummation ceremony is
carried out. This ceremony, though, was for the most part quite
simple. In Chu-shan Hsien, however, sometimes a regular ceremony of
welcoming a bride and taking her into the family (M. yjng-chieh
kuO-men) was carried out when she was adopted. Again, however, in
Pa-tung Hsien the ceremony was exceedingly simple, and really amounted
to little more than the girl and her husband paying a visit to her
natal home one month or three days after the consummation. This was
called M. hui-men, "going home." In Ch'ien-chiang Hsien, all that was
necessary was for the family to invite friends and relatives to a
feast, and then announce that "this is our son's wife." Later, when
they were of age, the marriage was consummated. In most of these
cases, however, it is specified that they do worship the ancestors

(often expressed as M. miaO-chien, “introduced to the ancestral

 

tablets").

186

Thus, in sum, the M. t'ung-yang-hsi (or M. miao-hsi,

 

M. wang-lang-hsi, M. hua-tun-hsi, etc.) was, and is, a young girl who

 

 

has been Uadoptedf into a family with the intent (usually, but not
always) of later marrying her to a son (or an instituted heir, or an
adopted son, or an uxorilocal husband). The motives for entering
into such a relationship are various. From the point of view of the
man's family, they save expenses, gain a source of labor, and make sure
that she becomes early and well adjusted to the family. From the
point of view of the girl's family, it is because they cannot support
her, or they want to get a good price for her.

Tai Yen-hui argues that legally the t'ung-yang-hsi should be
recognized as a "coresident betrothed daughter-in-law-to-be"

(M. t'ung-chfi ting-hun hsi. Tai 1966: 248). Her status was some-

 

times changed, however, into that Of an adopted daughter (M. yang-n0,
or M. j;g§) because the intended husband had died, or the two were
not willing to marry. In these cases, then, the institution has

been explained as adOption with marriage as the goal, and attain-
ment of the goal then taken as grounds for the dissolution of the
adoption relationship. If it has for one reason or another been
decided that the goal will not be attained, then the effect of the

adoption continues to exist (INS: 125).

Adoption in Taiwan

 

There is much more material on adoption in Taiwan than there
is for the mainland of China. This is in part because of the records
kept by the Japanese Colonial Government, but also because, of

course, it has long been Open, as the mainland has not, to field

187

research of various kinds. I will turn my attention to the various
forms of adoption in Taiwanese Chinese society.

As we have seen, the goal of adoption in China was usually
said to be the continuation of the worship of the ancestors. It
was necessary to make sure that the sacrifices were not cut off, and
the sacrifices are the duty of the sons. Even if they may not actually
carry them out themselves, but rather let their wives do it, it is done
in their name. There were, of course, other reasons for the practice--
economic, social/political, charitable, and so on. It was still the
religious goal that was emphasized in the traditional society.

The collection of Taiwanese customary law titled Taiwan Min-

shih Hsi-kuan Tiao-ch'a Pao-kao (abbreviated INS) says that the prac-

 

tice was generally for the same motives in Taiwan as the rest of
China, and says that it was very prevalent in the Ch'ing Dynasty.
During the first period of opening up an area, there is
a great need for large family and kin groups, whether or not
they were related, and so many persons adOpted children to
increase their labor power. The motive of continuing the kin-
ship 1ine was not, of course, entirely drOpped in these cases
(INS: 151).

The same source then goes on to discuss another motive for
adoption: that it will promote the birth of a child of one's own.
Adopted daughters were often given names like T. chio-ti, "inviting
a younger brother," or T. chiO-ti (different characters from the
first), "inviting regulation, direction." The second character of
the second name, T. ti, has the same pronunciation (including tone)
as T. t1, "brother." Other such names are T. oat-ti, and T. iO-ti.
The two characters T. ggt, and T. 19 are said to be vulgar characters

for T. iong, "to rear, to adopt, to give birth to." In all cases the

188

meaning is the same; the expression of hope that the adOpted child,
so named, will bring a little brother to the family. In not all
cases is the girl with a name like these adopted; one may be given to
one's own biological daughter as well if sons are lacking in the
family.

Adoption for these reasons continued during the Japanese
period, but an added factor was the custom, as mentioned in the
section on the T. simpua, of adopting girls who were in fact house-
hold slaves or servants.

In the present day,

According to the investigation conducted for this report,

Taiwanese adoption, whether among Hakka or Min-nan speakers,

is mainly for the purpose of continuing the kin line and con-
tinuing the sacrifices to the ancestors. Apart from this,
there are also those who adopt to have an heir for the family
property, for support in old age, to help in earning a liveli-
hood, for superstitious reasons, etc., and not only because one
has no children of one's own; often those with their own, or
wit? other adopted children also adopt more children (INS;

151 .

The four main types of adoption proper in Taiwan that I
will discuss are what I will call M. kuO-fang adOption (T. koe-pang,
"to transfer to another branch"; I will use Kuo-yu [Mandarin] pro-
nunciations for these terms, as they are not peculiar to Taiwan),

M. ming-ling adoption (T. beng-leng), posthumous adoption ("spirit
adoption," or institution of an heir), and the adoption Of girls (the
first forms referring almost exclusively to male children).

There is a certain amount of confusion in the use Of these
terms, and the distinctions between them seem to have blurred a good

deal during the Japanese and the Nationalist periods. A M. kuo-fang

adoption is, strictly speaking, the adoption of a child from within

189

one's own kin group--a nephew in a classificatory sense. Colloquially
however, in Taiwan and the mainland, M. kuo-fang seems usually to

have been Opposed to the adOption of a different-surname child.

This makes terminologically ambiguous the adoption of a child of

one's own surname but not of one's own kin group. Generally, but

not always, the term M. kuO-fang has then come to mean same-surname
adoption, on the implicit theory that those of the same surname are
ultimately, if distantly, related. The adoption of a child of another
surname then was often glossed as M. ming-ling adoption. This term

is interesting and, perhaps, revealing. It is based on a passage in

the Shih-chigg, the Book of Odes (sometimes the Classic of Poetry):

 

 

"When the ming-ling caterpillar has offspring, they are entrusted to
the solitary wasp." The M. ming-ling is a sort of caterpillar, the
larvae of which are sometimes imprisoned by the M. £19213, solitary
wasp, to provide food for its young.

A further complication in all of this is that in Taiwanese
customary law, at least, adopted children were divided into two
categories based on the degree of the completeness of the transferral
of the rights over them. Those who were entirely and irrevocably
transferred, cutting off all ties with their natal family, were
called by various terms all signifying outright sale, or translative
acquisition. Since M. kuo-fang adoption, presumed to be between
relatives, does not, in principle, completely sever such relation-
ships, M. ming-ling adoption, meaning adoption of (or the purchase
of) a child of another surname, came to have the meaning of "trans-

lative acquisition," or "irrevocable transfer."

190

As if this were not enough, it seems that in Taiwan the

M. ming-ling-tzu (the child adopted in M. ming-ling adoption) some-

 

times had the status of a male T. gimpgg, being first adopted and
then later married to a woman within the family. One form involves
the marriage of a M. ming-ling-tzu to the widowed daughter-in-law
of the family (a form which involves elements of M. ming-ling adOp-
tion, "male T. gjmggg_ad0ption andnmrriage"),uxorilocal marriage
(the second type, involving the remarriage of the woman), and the
levirate (INS; 153, 272 nl77).

During the Japanese period, the term M. ming-ling continued
to be used, and indeed is commonly found in the household registry
material of the times. The Japanese government, however, regarded
the custom of the translative acquisition of rights in children,
phrased in terms of buying and selling, as against public morals, and so
it was not permitted in law, and the legal status of the M. Nigg;_
ling-tzu and the M. kuo-fagg;tzu became technically the same (INS;
153).

Posthumous adoption has to be divided into that intended to
have religious effect only and that intended to have legal and eco-
nomic effect. Traditionally it was quite possible for a widow, or
other relative, to establish an heir (M. Ijzggg) after the death
of a man, though usually the deceased had to have been of a reason-
able age; it was not done for infants. During the Japanese time as '
well, this was possible. The police who were in charge of the
household registration did permit the registration of posthumous

adoption. After the Japanese period,

191

. . The establishment of heirs (M. li-szu) quickly became

quite rare, and moreover it is not recognized in the current
~law. But the so-called "designated heir"(M. chih-ting_chi-ch'eng)
that is to bequeath to an adOpted child, still preserves a part
of the meaning of M. li-szu. After the death of the husband,
the wife can of course aaopt, and if both are dead, custom still
permits family members to institute an heir for them, though
this is quite rare. Moreover, in this situation the person
instituted is just entered into the ancestral tablets as a

M. kuo-fang-tzu. There are only mourning and ritual duties with
respect to the deceased, and no valid kinship relationship in
law has been created between the two. To institute an heir

after death is in Taiwanese custom called T. chia to- an , "to
continue a branch which has (otherwise) died out" (INS; 155-56).

The adoption of girls is also a very complicated affair.
It shades Off into marriage in one direction (the T. simpua) and

into slavery in another (the T. cha-bo-kan). The T. simpua will

 

be treated separately, and here I will confine myself to other forms
of the adoption of daughters.

The Taiwan Min-shih Hsi-kuan Tiao-ch'a Pao-kao (TMS) says
of the adaption of daughters Hithe Ch'ing that:

The flourishing custom of adopting daughters in Taiwan,
apart from the so-called T. chio-ti (an adOption intended to
stimulate the birth of a child), was because one could take
advantage of their labor. If a family had no children, they
could adopt a daughter; if they already had a daughter, they
could adopt another, and even if they already had an adopted
daughter, they could adopt yet another one. If there was no
male among the relatives who could be adopted as M. kuo-fang,

a girl could be, and could be made to inherit.

Taiwan also had the system of the adOpted daughter-in-law,
colloquially called T. simpua. The adOpted daughter and the
adOpted daughter-in-law were different as far as the motivation
for the adoption was concerned. The latter was a young girl of
another surname adOpted with the intention of later making her
one's daughter-in-law. It did not matter whether or not it was
yet determined to whom she should be married; she added the
adoptive family's surname to her own just the same as if she had
already married in, and a "quasi-affinal relationship" was
created between her and the relatives of the adoptive family.

The adopted daughter was different from this. The motiva-
tion for adoption was not the same as the above, she took the

192

adoptive family's surname as her own, and a kinship relation
with the relatives of the adoptive family was created, just as
if she were born to that family.

However, in the northern part of Taiwan, the people usually

mixed up the two terminologically, calling both T. Simpgg_
(INS; 154).

I found this latter to be true. In T'ien-sung-p'i, where
I did field research, the term T. gimggg_tended to be a general term
perhaps best just rendered as "adOpted daughter." Certainly in a
situation in which one could adopt in a "daughter-in-law“ even though
one had no son yet to marry her to, there was Often little differ-
ence between the two.

There were a number of conditions of various kinds affecting
adOption, conditions actually followed to a greater or lesser degree.
Traditionally, for example, the adopter was supposed to be male.
Though a widow could institute a posthumous heir for her late hus-
band, this was done in his name, and she was not acting for herself.
This was true in Taiwan until the mid 19205, when single women were
allowed to adapt (INS; 156). Under the current law a married person
must jointly adOpt with the spouse, but a single adult of either sex
may adopt.

The adopting person did not traditionally have to be of a
specific age, but the age differential between the one adopting
and the one adOpted had to be approximately that between parent and
child. In the Nationalist Code, under article 1073, the age dif-
ference must be at least twenty years.

In the Ch'ing Code the lack of sons was a condition for

adOption; "in fact this was not followed by the people" (INS; 157),

193

and it was not even a formal condition during the Japanese period.
Neither could the adopted child be an only child, according to the
Ch'ing Code, but this also was not actually enforced (INS; 158).

Customari 1y, money was and is paid to the natal family in most
cases of adOption. We have seen that this was sometimes regarded
as a price for the child, in outright purchase. At times, however,
it was phrased as money in compensation for the raising of the child,
and was called "milk money" (T. leng-po--gin). The amount was not
fixed, but was and is dependent on the status of the families,
their relationship, and so on.

Formally, the child sometimes paid homage to the ancestors
of the family into which he or she was being adopted, but this was
not considered a necessary condition of the adoption. In the present
"this custom still exists, but among the lower classes there is no
ceremony whatsoever; they just hand over the child and exchange the
'milk money' or the price, and that is all" (INS; 161).

A contract of adoption was usually written up, but this was
and is not a necessary condition (INS; 161), and neither was the

participation of a gO-between.

The "Adopted Daughter-in-Law" in Taiwan
The category of adOpted-daughter-in-law is being treated
separately from other forms of adopted daughters since it is most
relevant to the argument being presented on the distinction between
marriage and adOption in Taiwanese and Chinese society.
In Taiwan an M. t'ung-yang-hsi was called a T. gimpgg, but

the two terms do not strictly correspond to one another. The

194

M. t'unggyang hsi was a female adOpted with the intention of marry-

 

ing her to one's son. An adOption without this intention was a
"simple" adoption. A source of confusion is that in Taiwan, and
especially in Northern Taiwan, both kinds of adopted girls were often
colloquially called T. gjmgga, When one also realizes that it was
neither uncommon nor improper for a family to adOpt in a daughter-
in-law even though there was no son for her to marry and then to
wait for the birth, adoption, or marriage in of a son or son-in-
law, the situation becomes even more complex. If a girl is termed
a T. gimpgg, then, she could be an adOpted daughter, with a position
in the family essentially similar to that of a biological daughter;
she could be an adopted daughter whom the family intends to marry to
a specific, living, son (in which case it is obvious that she is
better termed a "daughter-in-law," the marriage relationship having
already been established, if not completed); she could be an adOpted
daughter whom the family intends to marry to a son who will be
adopted into the family; she could be intended for a husband who
will be married in to live uxorilocally; she could be intended for a
son it is hoped will be born later; she could be intended to be held
"in reserve," to be dealt with in one of the above ways depending on
future events; she could actually be little more than a slave, work-
ing as a household servant; or she could be destined to be sold or
pawned into prostitution.

In Taiwanese customary law, the T. gjmggg_is divided into

two main types:

195

1. Those adOpted with the intention of marrying her to a
specific son of the family. This type is known as
T. u-thau-tui, "paired."

2. Those adopted without any definite spouse in mind.

This is known as T. bO-thau-tui, "unpaired" (Ikeda

 

1944: 195; INS; 126).

In the customary law an important condition was that the girl
be sent to live with the adopting family. There were no restric-
tions on age, but in practice the age difference between the adopt-
ers and the adOpted usually corresponded to that between parents and
children. Most of those adopted were twelve or thirteen years of
age, but they also could be two or three years, or even two or three
months. If they were as Old as seventeen or eighteen, then they
were simply married in (INS; 127). One important condition, a con-
dition that can in some cases help to distinguish the different
types of T. gimpgg, is that if it is intended to marry her to a son
of the family, then the rule of surname exogamy means that she must
be of a different surname than is the family. If she is of the same
surname, then she is a "real" adopted daughter, and cannot marry a
son of the family.

Traditionally there was no fixed form to such an adoption,
apart from writing up a contract, but sometimes marriage documents
were exchanged. Usually an auspicious day was selected, though, and
friends and relatives invited to a feast which marked the public
proclamation of the establishment of the relationship. One source

says that the amount of the "brideprice" (M. p'in-chin) depended

196

on the beauty of the girl, but that when she later grew up and mar-
ried, the adOpting family needed pay no more (INS; 127). Usually
she took the adaptive family's surname and added it in front of her
natal surname, just as if she were already married. This was dif-
ferent from the case with an adopted daughter (T. iong-lu); the
latter replaced her natal surname with that of the adoptive family,
and her kinship relationships with relatives of the adopting family
were the same as those of the family's own children. Thus the legal
status of the two types of adOpted girls was completely different,
but the T. sjmgga, "adopted daughter-in-law," could be changed into
a T. iong-lu, “adOpted daughter" (INS; 128). The reverse change was
obviously not as easy to make: if her status in the family were
already that of (adopted) daughter-in-law, she was defined as eli-
gible to marry within the family and hence there was no logical
problem to her marrying outside the family, just as a widow might
remarry out.

But if her status were that of (adopted) daughter, she was
then defined as eligible only to marry outside the family, and it
would be very difficult to change that status, to permit marriage
within the family. Thus though there was much confusion between the
two kinds of "adoption," and one could not always tell what any
particular case represented, the two forms were not interchangeable;
a T. gimpgg_might "really" be a T. iong-lu, but a T. iong-lu could
not "really" be a T. §jmggg_(cf. INS; 129 n91).

There are a number of reasons for arguing that a T. gimpgg_

should be regarded as "married" when she is "adopted." She added,

197

as has been noted, the new surname to her original one, and her
mourning grades with respect to the members of the family were the
same as those of a "regular" wife. If she had illicit relations with
someone other than her intended, then whether this should be regarded
as adultery or as fornication was not consistently ruled by the
courts at the time (INS; 128). If after the death of the husband-
tO-be (and before the marriage), however, she remained a widow in

the adOptive family, she could inherit the management of the hus-
band's prOperty, and could institute a posthumous heir. Further-
more, sexual relations between the T. sjmpgg_and her husband/fiance
were not regarded as illicit, but were rather punished under the
provisions for disobedience (Hoang 1915: 77n; INS; 128). After
entering the new family, the T. gimpggfs kinship relationships in

the family were the same as those of a wife, her relations with her
natal family were the same as if she had married out in a regular
way, and the kinship relationships between members of her natal
family and the members of her new family were the same as if she had
married in an ordinary way. Furthermore, she used a wife's kinship
terminology with her affinal relatives.

Thus in Taiwanese customary law the kinship relations created
by T. §Igpgg_ad0ption were affinal relations, and not consanguineal
or "quasi-consanguineal," and this was reaffirmed by a court decision
of 1922 (INS; 129). Just as did a "regular" wife, the T. gimggg_
entered the man's family as a dependent, and so had to obey the
family elders. Relations with her natal family, except for the nat-

ural relations of parent-child, were terminated (INS; 129).

198

The M. t'ung-yang-hsi relationship is for the most part

 

similar to that of regular marriage. Termination may occur through
death, abandonment, or mutual agreement. If the relationship was
dissolved before marriage to the son of the family (the latter was
called M. yuan-fang, M. ho-fang, M. Shang-fa, M. Shang-t'ou, etc.,

and in Taiwanese sang-tsoe-tui), the girl was not supposed then to

 

marry anyone of the intended husband's line (INS; 130).

The report on Taiwanese customary law, however, says that
there were some special grounds for dissolution of this relationship
not shared with regular marriage. If the husband-to-be married
someone else, this was grounds for dissolution. If a T. bO-thau-tui,
(unpaired; with no fixed husband-to-be),§_i_rmnad reached mar-
riageable age, but the adoptive family delayed an unreasonable length
of time in the selection of a husband for her, she could request
dissolution. Legally, it was determined by court decree that a
T. gimpgg_was a girl who had been adopted by contract with the
intention of marrying her to a specific son of the family or one to
be selected later, and who moves in with the adoptive family. The
son so selected had something of a duty to marry her, but he could
not be forced to do so by law. If he refused, then the main goal
of the adoption was lost, and if he married someone else, then this
was grounds for dissolution of the marriage/adoption (INS; 130).

During the Japanese period in Taiwan, there were no great
differences in this custom, and it was not declared contrary to good
morals and public order. Japanese law did, however, prohibit the

custom of selling slaves (T. tsa-bo-kan), and in response, peOple

199

got around the prohibition by calling the slaves adopted daughters or
daughters-in-law. This terminology also worked as a screen to hide
the practice of adopting girls for the purpose of selling them into
prostitution.

Later, in the Nationalist period, the custom was not per-
mited to continue unchecked, and, in fact, it is rather rare now.
Article 972 of the Code bases marriage and the marriage contract on
the principle of free agreement of male and female. The socio-
political forces that spelled the end Of arranged marriages worked
with particular effectiveness on this, an arranged marriage par
excellence. Such T. gjmggg_adoptions as did take place after the
implementation of the Nationalist Code (the custom did not,of course,
end all at once) were disguised as "regular" adoptions. In many
cases, of course, e5pecia11y with the many social and economic
changes that have taken place in the last twenty-five years, the
goal of marrying the girl to one's son is not carried through. Thus
in the household records, for instance, many household members
listed simply as "adopted daughters" were really intended to be
T. gimpgg, If this plan was indeed given up, then there is no way
to distinguish them from "real" adopted daughters. If the parents
do manage to be successful in their desires, and the two do marry,
then it is usually easy to determine this from the household registry,
as the adoption is first dissolved, and then the marriage is com-
pleted.

One of the reasons I was; given for the decline of the

custom in the village in which I lived and did research was that an

200

adopted girl, on the least pretext, could go to the Women's Asso-
ciation and file accusations of maltreatment, accusations which,

I was told, would be very difficult to disprove. The Women's
Association tends to side with the girl in such cases, and the
trouble and expense that can arise from this, and the fact that the
girl can "just run away if she feels like it, and then get the
Association to back her up" means that adopting a girl is seen as
more trouble than it is worth.

One of the reasons for the vigilance of the Women's Asso-
ciation is that it is fact that in the past "adopted daughters" and
"adopted daughters-in-law" were often associated with prostitution.
Barclay (1923) lists as the meaning of T. iong-lu "an adOpted
daughter; term for concubine or inmate of house of ill-fame in rela-
tion to owner." This is somewhat similar to the custom in Chinan,
Shan-tung, of calling rural prostitutes "little daughter-in-law
(M. hsi-fu-erh. Hu 1936: II, 2,12).

Wu, in Taiwan Min-su, says that the custom in Taiwan was

 

limited by surname exogamy, so the parties to the "adoption" trans-
action were always of different surnames, and that the age of adop-
tion of the girl was from birth to around two or three years

(Wu 1969: 106). According to statistics taken from the household
registration for each hsien, there were on January 31, 1960, 93,550
unmarried adopted girls (M. yang-n0; including T. gimpgg) still
living with their adoptive parents (though this figure also includes
those who had actually already become prostitutes). The vast

majority of them (74,066) had been adopted under the age of five

201

(Liu 1969: 60-63). Though this includes two different types of
"adopted girls," it is evidence that there was a tendency to adopt
such girls at a rather young age.

The goal, according to Wu, and often heard, was to save on
future marriage expenses and bride-price. He also says, and this
is heard less often, that at the same time raising such a girl from
an early age could increase the closeness of relationships, adding
the affinal relationship to the parent-child relationship, and so
helping smooth family affairs. Thus this method of adoption/marriage
"kills two birds with one stone" (Wu 1969: 106). Counter to Wolf's
arguments for the incestuous tinge to such marriages, and their
general unsuccessfulness, there is the proverb T. tang-tng leng:_
thau kau tang-Sim, "those who suckle at the same nipple are more of
the same mind," indicating that a son and a T. gimpgg_who share the
same woman's milk will be able to have an even closer relationship.
Furthermore, there is the belief that after adopting a T. gjmggg_
it will be easy to get grandsons. This is called T. teh-chhi*
(Wu 1969: 106), "to plant a slip, or seedling," and seems to be
' similar to the custom of T. chiO-ti, T. chio sio-ti, or T. chio-ti-ti,
whereby an adopted child is thought to "call," or "invite" another,
biological,Ch11d. (Adopting a child can also be called T. teh-chhi*;
see Barclay 1923.)

The method of adoption, according to Wu and various infor-
mants, was variable, and could include the inspection Of the girl's
"eight characters" (her "fate"), a "betrothal" (T. sang-tia*, "to

send earnest-money"), which was to hang some money tied around her

202

neck with a red thread (called T. koa* am-chi*), sending some money,
and some presents. This is all not necessary, however, and in fact
sometimes not only is it eliminated, but the girl's natal family may
even send along some additional clothing for the girl (Wu 1969:
106).

If there is indeed a son in the family for her to marry
(there wasn't always), and if she indeed marries him (she didn't
always), then the marriage is called "marriage within the household“

(T. ho- lai hun-in, a term that no doubt arose during the Japanese

 

period, after the system of household registration was put into
effect). The "marriage," or "consummation of the marriage," was

itself called T. thui tsoe-tui, "to push together into a heap,"

 

or T. tsoe taO-lang, "to become adult," or "to become married."

 

The first term is interesting because it is translated by Douglas
as also meaning "to begin to cohabit, as betrothed persons with the
regular ceremonies" (Douglas 1873: 531; see Wu 1969: 107).

The circumstances in which she found herself after she was
taken into the adoptive family varied a good deal, all the way from
those similar to the circumstances of a biological child of the
family to those of a female slave, or T. tsa-bo'-kan. This depended
in part on the fortunes of family subsequent to the transaction.

If all things seemed to be going well, then she was judged to have

a “good fate," or "good luck" (T. ka chiah-ui ho, a term that,
according to Barclay [1923: 20], was used for a bride), and so would
meet with the family's approval. If, however, the family fortunes

took a turn for the worse, she might well be judged as having "bad

203

luck" (T. ghai* ka-chiah-ui), and in such circumstances could easily

 

become the focus of the family's resentment. Generally, however,
she seems often to have had a difficult life, a circumstance summed

up in the tenn"simpua'sfate" (T. sim-pu-a Cheng, T. sim-pu-a the,

 

 

sim-pu-a tsai, etc., terms with decidedly negative connotations,

 

Wu 1969: 107; Wu seems to have taken much of this material from

Toshio Ikeda, A Survey of Taiwanese Family Life, 1944, though he does

 

not cite this work. See particularly pp. 194-203 in Ikeda).

That there may at times have been more positive aspects of
her position in the new family is indicated by the saying, "Rear
a daughter and she becomes someone else's; rear a T. gimpga, and
she is your own9~(T. chhi tsa-bo' kia* pat lang-e, chhi sim-pu-a

tsoe khien ka [Wu 1969: 107], or sim-pu-a ka-ti e, tsa-bo‘-kia* pat-

 

lang-e [Ikeda 1944: 267]). These sayings also illustrate another
point, to be returned to later--that a T. simpua was in many ways
more "valuable," or "useful," to a family than their own, biological,

daughter might be.

CHAPTER V

MARRIAGE AND ADOPTION IN A TAIWANESE VILLAGE

In my discussion of marriage, adOption, and related forms
so far, I have drawn on material from various parts of China, with
particular emphasis on Chinese society as found in Taiwan. My
interest in this tOpic began, however, with field research in one
particular village in one particular area of rural Taiwan. I have
tried to establish in previous chapters the various forms of mar-
riage and adoption that existed in traditional Chinese society, in
order to give some indication of their variety, historical depth,
and geographical spread.

In the present chapter I want to discuss more specific data,
data from that one specific village where I did research. I will
examine patterns of adoption and marriage as they occurred through
the Japanese period and up to the present. TO do so I will give a
brief description of the research site--its setting, history, and
population. I will then select a number of case studies, to be
summarized in this chapter and presented in more detail in

Appendix.B.

A Brief Description of T'ien-sung-p'i
The first research on which this dissertation is partly

based was carried out in T'ien-sung-p'i, I-lan Hsien, from August

204

205

1971 to August 1972. The village of T'ien-sung-p'i is a large one
by rural Taiwanese standards, and lies at the head of the triangular
plain in northeastern Taiwan that occupies much of I-lan Hsien or
county. It contained, at the end of June 1971, about 575 households,
and over 3,500 persons. Administratively, T'ien-sung-p'i has no
independent existence. This was the name of the settlement in the
Japanese era and before, but after the Japanese left Taiwan, the
Nationalists divided it into two administrative villages, thgg,
Sociologically, however, it remains one settlement, one unit, and
there are no significant differences between the two administrative
units; the division was carried out simply because of a general
policy that thgg_should be of approximately similar sizes. The
division resulted in T'ien-shan Ts'un, the southern part of the vil-
lage, with a population in June 1971 of 1,617, in 263 households,
and T'ien-fu Ts'un, in the northern part, with 1,967 persons, and
312 households (the figures are from the San-hsing Ngjggg_or township
government, of which T'ien-sung-p'i is a part).

The main economic activity of the population is wet rice
agriculture, and the most important secondary economic activities
are also connected with agriculture; these all have to do with
products grown in the adjacent mountain areas, primarily fruit.
Located at the apex of the I-lan (or Lan-yang) Plain, it is quite
near the foothills in the southwestern part of the county. It is
surrounded by rice fields, and in fact a good prOportion of the

population (but by no means the majority) actually resides in small

206

hamlets or isolated residences scattered throughout the fields
around the demographic center of the village.

T'ien-sung-p'i is the last stOp on the bus line from Lo-tung,
a commercial center on the I-lan rail line, which connects with Taipei.
This means that the village is something of a market center for the
upriver areas; though there are roads which lead farther up than
T'ien-sung-p'i, they are quite primitive and unpaved, and are easily
cut off by heavy rain and slides. The convenience of the bus ride on
to San-hsing, the township seat about three kilometers away, or
LO-tung, however, also acted to undercut the village as a trading
center. After coming down from high in the mountains, few found
anything of interest in T'ien-sung-p'i, and continued on to San-hsing
or to Lo-tung.

In the 19205 the Japanese built a narrow-gauge railway from
Lo-tung which passes through T'ien-sung-p'i and continues far up
into the mountains. In those days it was used to bring down camphor
and lumber. The trade in camphor has died down, but lumber still
comes down to LO-tung in large quantities, both by this railway
and by large lumber trucks.

After the Nationalists received Taiwan from the Japanese at
the end of World War II, some of the mountain areas above the I-lan
Plain were opened up to fruit orchards, developed as part of the
government's plan to provide jobs for retired servicemen.

All of these things--the need for mountain road expansion

and maintenance, the lumber industry, and fruit orchards of various

207

kinds--mean that there are a good number of laborer's jobs available
relatively nearby.

In the village of T'ien-sung-p'i younger men, whose educa-
tion took place after the Japanese period (thus equipping them to
speak Nggzyy, or Mandarin), tend to go to Taipei when looking for
work. Older men, whose education, if there was any at all, took
place during the Japanese period (and who thus have difficulty in
speaking or understanding Nggzyyg,tend to go into the mountains to
work.

It was said that at Li-shan, a center of fruit-growing in the
mountains, almost all of the work crews contained at least one
T'ien-sung-p'i man. The village is close enough to the mountains
that a man can go to work there without cutting himself off entirely
from friends and home. (This is a more important consideration for
the older men, who are more bound by family responsibilities.)

There is usually some other fellow villager working nearby, and it
is not difficult to return to the village every month or so, or on
short notice, either by the railway, bus, or on one of the trucks
which are constantly going back and forth between the mountains
and Lo-tung (where there are some of the largest lumber mills in
East Asia).

Thus a substantial number of the households in T'ien-sung-p'i
are households which combine agricultural work with mountain labor-
ing jobs for the men (and to a lesser degree for the women, as cooks

and washers). This enables them to remain resident in the village,

208

so its population is no doubt a good deal greater than it would have
been otherwise.

This is particularly true in recent years, as the profit-
ability Of wet rice agriculture has declined. Such mountain
products as a variety of citrus fruits, tea, ginger, Liu-ch'iu
pine, and many more are grown either on the foothills or higher in
the mountains, and have been taking on increasing economic impor-
tance. Some latecomers to the area, who were denied access to low-
lying land suitable for wet-rice agriculture, were forced to turn
to the mountains, specializing in other crOps. Initially at a
disadvantage vis-a-vis the wet rice cultivators, some of them are
very well-Off now, as other forms of agriculture become more
profitable than growing rice.

The village is also a collection and distribution center for
the surrounding area on the plain, and for a number of settlements
further up the river in the mountains. Rice is brought into the
village to be milled, and there is a citrus-fruit packing station
which is used each year when the fruit is harvested to collect,
grade, and pack fruit from the surrounding orchards.

T'ien-sung-p'i is something of a small market center, and is
large enough to be called a "small town" (see Ch'en Cheng-hsiang
(1959-1961). It has a market population of over 2,500 and
has a "shOp index" of over 50 percent. (The “shop index" is the
percentage of 5hOps along 100 meters of the main street.) Not much
fresh produce is traded locally; most people raise most of

what they eat, shopping only for small quantities of certain

209

vegetables, for small quantities of pork at one or the other of the
two butcher shops, or for such things as bean curd or for fish,
which was occasionally brought in from outside on a small truck.
One of the reasons that it has not specialized further as a col-
lection and distribution center for agricultural produce and manu-
factured goods is that only about three kilometers away (only about
five or ten minutes on the bus) is the town of San-hsing, which has
a public market for foodstuffs and many more specialized stores.
Even Lo-tung, one of the largest towns on the I-lan Plain, is only
about one-half hour away by frequent bus.

The stores in the village are almost all operated as addi-
tions to farming or some other economic activity; few of them are
the sole source of income for the families concerned. Most of them
are simply retail outlets, selling such things as canned goods,
candy, cigarettes, eggs, soy sauce (though many preferred to make
their own), wine, mock money for ritual purposes, salt, batteries,
light-bulbs, and so on. While I was resident there, the only
product produced entirely in the village was a sort of large knife
or machete which was fashioned by the single blacksmith (one of the
families said to be of non-Chinese, Aborigine, descent, but reluctant
to be reminded of the fact), and is used by those working in the
mountains. When I made a brief return visit in 1974, however, a new
industry had been started with capital from Taipei. The Ngiagg, or
township, government had built a structure which had been intended
to be a public market for the village. After it was Opened, few

showed any interest in taking up a stall there (which was not

210

surprising; all of the stores were also residences for the owners'

families and so, in a sense, were rent free), and so it was rented

out to a businessman who is using it to house a small factory which
processes the wood from a kind of large fern tree, cutting it into

slabs which can be used to grow orchids; the product is exported to
Korea and Japan. This is the first new source of wage labor to be

developed in the village for a long time; about twenty persons are

employed, either part or full time.

The only other major source of wage labor in the village is
a hydroelectric power station, which employs perhaps thirty of the
villagers for guard duty, maintenance work, and so on. The tech-
nicians who operate the power station itself are for the most part
from outside the village, live together in a compound nearby, and
are not really part of the social life of the village.

The original population of the Taiwanese Chinese has always
been divided, roughly, into three groups: the Hakkas (mostly
originally from Kuangtung Province), those (originally) from Chang-
chou, and those (originally) from Ch'uan-chou. (Both of these latter
are areas not very far apart from each other in southern Fukien
Province.) The history of Taiwan has been characterized by dis-
putes and struggles among these three groups, in varying combinations
(and this apart from their united struggle against the mountain
Aborigines). I-lan Hsien has been known, from its early days, as an
area of particular concentration of Chang-Chou speakers. Those from

Chang-chou and those from Ch'uan-chou both speak dialects Of Hokkien

211

or Min-nan; though the languages are mutually intelligible to their
respective speakers, there are clear differences in pronunciation.

Ethnically, however, the village of T'ien-sung-p'i is
atypical Of I-lan, and is in many ways more representative of the
ethnic groupings as they are found in Taiwan as a whole. The dis-
tinction between Chang-chou speakers and Ch'uan-chou speakers has
been on the decline since, I would speculate, the beginning Of the
Japanese period. Linguistically, for instance, a Taiwanese language
is developing which is not clearly either Chang-Chou or Ch'uan-chou.
This is happening in I-lan generally, as in the greater part of the
rest of Taiwan, but is perhaps more advanced in T'ien-sung-p'i than
in the rest of the hsien.

A comparable but separate development has been the assimi-
lation of other ethnic groups to this Chang-chou/Ch'uan-chou com-
posite Taiwanese majority. Thus it is clear in T'ien-sung-p'i, for
instance, that there is emerging a sort of "pan-Taiwanese" linguis-
tic and ethnic standard. This makes a precise delineation of con-
temporary ethnic divisions difficult.

The earliest settlers in the T'ien-sung-p'i area were evi-
dently Sinicized Aborigines, quickly followed by (mainly) Chang-Chou
Han Chinese, from other parts of I-lan. Then there seems to have
been an influx of Ch'uan-chou Speakers, and Hakkas, the latter
mostly from the Hsin-chu and T'aO-yuan areas on the other side of
the island. The precise relative percentages of the descendants of
these groups in the present-day village of T'ien-sung-p'i would be

difficult to arrive at. My own research, and investigation into

212

the household records of the Japanese and Nationalist periods, lead
me to confirm what I was told by the inhabitants themselves--
that about 30 percent of the pOpulation are (or were) Hakka. About
10 percent are or were Sinicized Aborigines, and the remainder are
composite Chang-chou/Ch'uan-chou, with the predominant element
being Chang-chou.

The figures cannot be much more precise than that since the
categories used are of very uncertain validity. A casual visitor
could spend much time in the village without suspecting that the
inhabitants were anything but Chang-chou/Ch'uan-chou Han Chinese.
Only if one is sensitive to the ethnic implications of certain sur-
names (both Hakka and Aborigine), or if one asks the outright ques-
tion of ethnic make-up, would the heterogeneity of the population
become clear.

I knew only one person, for example, who could speak any
of the Aborigine language, and he could not remember more than a
few score words. By the same token very seldom did I hear Hakka
being Spoken, and when I started asking about this, I was told that
a substantial and increasing number of young "Hakka" peOple do not
know how to speak the language. This is important, because it is
one of the local definitions Of ethnic-group affiliation. When I
tried to get more precise information on the numbers of Hakka, for
example, it quickly became Obvious that the criterion of definition
was the ability to speak Hakka.

Not only is there this kind of confusion, there has been a

good deal of intermarriage as well. This is hard to quantify, but

213

in the Japanese-era records, for example, where a sort of ethnic
identification of individuals is made (I say "sort of" because the
identification is not really ethnic, but place of origin of one's
ancestors for the Han Chinese; Kuangtung origin generally means
Hakka, but some Hakka did move to Taiwan from Fukien), there is no
lack of Han inter-group marriages or Han-Aborigine marriages. One
would expect, also, that social, political, and economic changes
since the end of the Japanese period would only have increased the
numbers of such marriages.

I was told by a local writer of marriage and other con-
tracts that there is some feeling that Hakka, Holo (Hokkien-speaking
Han Chinese), and Sinicized Aborigines should all marry within their
own groups, but that this feeling was not strong. A stronger rule
applied only to Han Chinese-Sinicized Aborigine marriages--that
males from the former group may marry into the latter group
but it is less common and less acceptable for the females to do so.

The picture, then, is one of a village which is ethnically
diverse, but the ethnic divisions do not at present account for
much. The peOple are mainly engaged in agriculture, but a large
proportion of households, perhaps a majority, are involved in some
sort of supplementary work for wages.

The unusual size (yf'Tien-sung-p'i and its ethnic hetero-
geneity are a result of both its history and geographic location.

The history of conflict with Taiwanese Aborigines in the area,
as well as some other factors, is of special importance. As late as

1910, Aborigines attacked a nearby police post, killing over ten

214

policemen, and the history of the settlement largely reflects the
continued presence of hostile Aborigine groups.

The original settlement in the area of T'ien-sung-p'i was
a small hamlet near what is now the power station. Before 1920,
when the power station was built, there was an irrigation pond
(M. ij, T. pi) from which the village gets its name. The literal
translation of T'ien-sung-p'i is "irrigation pond sent from heaven,"
but it is also said that "T'ien-sung" is really the name of one of
the early landlords of the area and thus the name of the village
should not be translated literally.

The original settlement was walled for protection from the
Aborigines by intertwined wood, bamboo, and thorn plants. The area
of the original hamlet is still called T. Sia*-te (M. ch'eng-ti),
literally "the walled place." The original compactness of the
settlement can be explained in terms of the defense it offered the
settlers from Aborigine attacks. (North Taiwan is more generally
characterized instead by dispersed settlement patterns. See Ch'en:
1959-61.) When the threat from Aborigines had abated by about 1920,
scattered, dispersed settlements began to appear on the Plain
around the village, but the foundations had already been laid for
the nucleated settlement that is seen today.

Land was a valuable commodity in develOping Taiwan. The
only thing which kept the T'ien-sung-p'i area from being Opened up
and settled far earlier than it was, was the danger from the
Aborigines. This began to change when Liu Ming-ch'uan became

Governor, toward the end of Ch'ing rule Of Taiwan. Under his

215

leadership, develOpment plans were laid out for the pacification and
opening up of the frontier areas. The Japanese also, not much
later, set up develOpment bureaus, the goals of which were to pacify
and control the Aborigines, to open up new lands, and generally to
promote and secure production in agriculture and such things as
exploitation of forest products and mining.

The Japanese also set up, again following the lead of the
previous Ch'ing government, a system of guard lines to contain the
mountain Aborigines and protect the lowland settlers. Around the
turn of the century around the fringes of the I-lan Plain (and
including T'ien-sung-p'i) were 140 guards stationed for the protec-
tion of villages along the borders with the Aborigine areas. There
were also ninety guards whose job it was to protect the areas of
camphor-wood production in the mountains (Wen 1957: 699). Some of
these guards are included in the earliest household registration
records for Tien-sung-p'i, and I interviewed one, a Sinicized
Aborigine, who was still alive in 1972.

All of this meant the possibility for further develOpment
Of the village of T'ien-sung-p'i. It became the center for some
of the administrative connections that the Japanese Colonial govern-
ment had with the aborigines, and its economic position was no doubt
improved when it was made one of only two areas in I-lan where trade
with the Aborigines was permitted. Japanese government policies
worked to improve security, which meant that there were substantial
areas of land opened up for settlement--land the Japanese were

interested in developing.

216

The government household records at my disposal show that
the earliest pOpulation in the village was for the most part Sini-
cized Aborigine. Later Holo families from other parts of I-lan
began to move in. Later, as security improved and there was increased
land available, groups of peOple and families began to move in from
further afield. Three Hakka lineages, the Liaos, Ch'ius, and Sungs,
all came in from the other side of the island, from the Hsin-chu/
T'ao-yuan area, around the turn of the century.

Evidently the "private-enterprise," or entrepreneur, system
of land development was in use, and the first of the Hakka groups
were attracted to the area by the prospect of acquiring agricultural
land. They supplied the manpower and management, and in turn got a
portion of the land (though it is still unclear whether they got
this land "free and clear," or rather acquired rights of "permanent
tenancy"). Just what combination of events and factors led to the
arrival of the three Hakka groups from a relatively small area on
the other side of the island at about the same time, I do not yet
know. In many ways more is known about I-lan in the Ch'ing

Dynasty (through the Komalan T'ing-Chih and other writings) than

 

about I-lan in the Japanese period, particularly its earlier years.
Thus T'ien-sung-p'i is known as the area in I-lan with the

greatest concentration of Hakkas (Liao Han-ch'en, ILHC II, 3,19),

and the San-hsing Hsiang area, the larger administrative unit in

which the village is located, is known as having the greatest con-

centration of Sinicized Aborigines (Mao I-po, LIHC II, 2,50-53;

YU 1951; Juan 1968).

217

Households in T'ien-sung-p'i range in size from one person

to twenty-two persons, with the distribution as follows:

 

 

Size of Households Number of Households
l 44
2 34
3 41
4 58
5 77
6 82
7 81
8 71
9 48

10 27
11 13
12 10
13 8
l4 8
15 3
l6 5
17 4
l8 1
22 l

The average size is six persons per household, and the median is the
same. ("Household'I here is simply that defined by the household
registration office; this usually corresponded to sociological

units which were actually economically independent to some degree,
but the degree, in particular cases, was quite variable.) Conjugal
families accounted for 55 percent of families, stem for 31 percent,
joint families for 8 percent, and others, those which did not easily

fit into the other categories, accounted for 6 percent of the

218

population. The same structural patterns can be expresed simply in
terms of generation depth; 10.2 percent of the families are of only
one generation, 54.4 percent are two generations in depth, 32.4 per-
cent three generations, and 3.1 percent were four generations in
depth.

A good index of the heterogeneity of the population of
T'ien-sung-p'i is the number of surnames represented. There are a
total of seventy-one surnames among all the heads of households, a
sobering number to an anthrOpologist used to reading about single-
surname villages. Twenty-two of these surnames, however, are rep-
resented by one household only. Eight of the surnames account for
over one-half of all the households. These are Huang (T. Ng),
with fifty-one; Sung (T. Song), Lin (T. Lim), and Ch'iu (T. Khu),
with forty-six each; Ch'en (T. Tan), with thirty-five; Li (T. Li),
with thirty-four; Liao (T. Liau), with thirty-three; and Chang

(T. Tiu*), with twenty-seven.

Irregular Forms of Marriage and Adogtjon
The precise numbers and percentages are not of crucial
importance, and given the large size of the village population
difficult to compute, but I do have some rough figures for the
incidence of some irregular forms of marriage and adOption. These
figures are taken from the current household registration data for
1972, and do not include any information from the Japanese-period

records.

219

Individuals Households % of Households

 

 

 

Uxorilocal marriage 90 83 13
Suspected uxorilocal

marriage 45 32 5
Simpua 50 48

Adopted daughters 206 180 29
Suspected adopted

daughters 74 71 10

The "suspected" uxorilocal husbands and adopted daughters
above are those cases in which there is reason to believe that one
or the other of the relationships holds, but the evidence is insuf-
ficient to conclusively prove it. In the case of a su5pected
uxorilocal husband, for example, the suspicion may be based on the
fact that the children are listed as taking the mother's surname.
In such cases, when it was not explicitly stated that he was an
uxorilocal husband, I have simply categorized it as "suspected."

It is possible in most of these cases to prove the su5pi-
cion one way or the other by having recourse to the Japanese-era
records. In almost every case in which I have done this, my sus-
picions have been borne out. Thus I think that it is safe to say
that well over 15 percent of the households have participated in
one way or another in uxorilocal marriage within the last thirty or
forty years.

Since it is Often difficult to distinguish T. gimpgg_from
adopted daughters proper, and because the "suspected" cases, as in
"suspected" uxorilocal marriage, almost always do turn out to be

cases of adoption of one form or another, we can say that perhaps

220

45 percent or more of the households have participated in one way
or another in one or more forms of the adoption of young girls over
the last forty or fifty years.

The qualification "in one way or another" is inserted
because a household was counted when one of the household members
was identified in the records as having been, say, adopted. Thus a
household may not have actually adopted a daughter itself, but the
wife of the head of the household may have been married in from her
adoptive family; since she had been an adopted daughter, that house-
hold was counted. Such cases were relatively few, though, and the
great majority of the above cases involve households directly
involved in the transaction.

Obviously this means that a particular adOption may be
counted twice if both of the families which were parties to the
transaction were from T'ien-sung-p'i. A large number of such adop-
tions or uxorilocal marriages, however, were contracted with families
outside the village. Still, what is of most interest is the fre-
quency with which households engage in such transactions (whether
it is, for example, to bring a girl into the household in adOption
or to get "rid" of one that way).

The information contained in the current household regis-
tration is about the living members of the household only (with
the minor exception of those who have died so recently that their
entries have not yet been removed). Since few of the persons
counted in the cases above would be Older than about sixty years of

age, and any earlier household members who might have been involved

221

in such transactions can be assumed to be deceased and their records
removed from the current registry, this means that their marriages
would most likely have taken place between thirty and forty years
ago, and any adoptions would have taken place earlier than that,

say forty or fifty years ago. I am using these figures as a rough
measure of the time-span being considered. In fact, it is an
outside limit for the time-span: in practice, little information
from the Japanese period is carried over into the Nationalist
household records other than the names and birthdates of persons.
Often there is no information about the circumstances of marriages,
for instance, in the current records if they took place during the
Japanese era, and this is so even if both of the parties are still
alive. The effective time-span covered by the current records (for
the sorts Of information that we are interested in here, at least)
is in most cases no more than about twenty-five years. Since the
death or change of the household head and a number of other factors
may lead to the retirement of the old forms, and the transferral of
some (but not all) of the information onto new forms for the current
records, this means that in a substantial percentage of the cases
the records cover much less than twenty-five years.

For these reasons, I suspect that there is substantial
underreporting (If these and other transactions--underreporting in
the sense that over thelast twenty-five years or so there must have
been many more adoptions and more uxorilocal marriages than the

above figures suggest. Just how many more is impossible to say.

222

Case Studies

 

To illustrate the kinds of social transactions which caught
my eye in T'ien-sung-p'i, and to give more detailed illustration of
the social and economic background of the village in general and the
transactions in particular, I have chosen a number of case studies.
The information contained in these case studies comes from inter-
viewing peOple in the village and from the household registration
data of the Japanese and Nationalist periods.

I attempted to c0py all of the available records for the
entire village, and came very close to doing so; all that is missing
is a small amount Of information on "temporary residents" in the
Japanese period. Considering the size of the village, and the
length Of time covered, it should be no surprise that there are
many pages of records. Some of the material I was able to photo-
c0py, using the machine belonging to the San-hsing Hsiang Kung-so,
the township government office, but all of the Japanese-era material
had to be hand-COpied.

I originally wanted to go through all of the household data
in order to plot marriages and adoptions of various types over time.
Because of the way the records were organized, however, this proved
very difficult in practice. When the head of a household died, or
some other change took place which could not easily be incorporated
into the current form, a whole new form was written up for that
household, and the old one removed and filed by year of removal.
(The earliest year of removal in our records was 1910.) Thus the

body of the registration material consists of the current, running,

223

record, and the yearly removals since 1910. (Some of these are miss-
ing for the early years.)

Thus, the set of records removed in 1927, for example, might,
depending on the history of the particular family, contain infor-
mation dating from the 18405 and 18505, or it might date from 1926
only. More likely, the information will span those dates. To
total all of the gimpgg adoptions, say, that took place in 1927,
one would have to go through all of the material (including the
current records) except, of course, those forms removed prior to
1927. While this kind of reconstruction of data is perfectly
possible in principle, it would be extremely tedious and time-
consuming in practice.

I have therefore chosen an alternative method (Hi using the
material for case studies. I have selected seven surname groups
and then have followed each and every member, past and present,
tracing each marriage and adOption. A5 I have mentioned, some of
the records are missing. In some instances it simply wasn't known
what had happened to the removals for particular years, and in
others it was known that typhoon damage was responsible. In general
the coverage is very good, though, and in only a few cases (noted
in the genealogies)were the data on individuals incomplete.

The seven groups that I have selected are surnamed P'an
(T. Phoa*; two groups), Wu (T. Go), Yeh (T. Iap), Ch'iu (T. Khu;
two groups), and Liao (T. Liau). The selection of these groups is
by no means random. I have tried to select from different socio-

economic categories and ethnic groupings. I have not, however,

224

selected any of the groups to illustrate any particular marriage or
adoption form and I can think of no particular reason why these
groups should not be considered representative of the village as a

whole.

§E§£_l} Phoa*

The founders of the group selected as case one were Phoa*
Lim Iam-bun and Phoa* A-hoan, both Sinicized Aborigines.(I will use
Taiwanese Romanization here, as the Aborigine given names are
Taiwanese transliterations of the original, and thus represent
Taiwanese, not gug:yg, sounds.) All of the members of this family
were Sinicized Aborigines, except for some of the members who were
married or adopted in. I will indicate those members who were not
Aborigine.

The main occupation listed for this household was day-
1aboring. They moved into the village from T'ou-wei, the present
T'ou-ch'eng Chen in the extreme northeast of the I-lan Plain, in
1908. Phoa* A-hoan, the male ancestor of the family, was born in
1871. He married Lim Iam-bun, also fromT'ou-wei, in 1892, when
he was twenty-one and she fifteen. Their marriage was evidently a
"regular" one. Her father was surnamed Lim, and her mother surnamed
Phoa*. There is no particular reason, however, to suspect any sort
of preferential marriage system; Phoa* is the most common surname
among Sinicized Aborigines. It was "sold" to the Aborigines by the
Chinese in the Ch'ing Dynasty on the basis that when the character

is analyzed into parts, it yielded M. shui ("water"), M. mi_("rice"),

225

and M. tljgg_( pady field). It is no accident that it also yields
M. jag, "barbarian, Aborigine" (Wen 1957: 563).

Phoa* and Lim then had four children--three sons and a
daughter. The oldest son was Phoa* Bu-li, who was born in 1898,
six years after the parents' marriage. (Onewould strongly suspect
that there were other, unrecorded, children who died in infancy:
the infant mortality rate was evidently quite high in the Ch'ing,
and the rate among Sinicized Aborigines has always been higher than
among the Han Chinese.) In 1906 a one-year-old girl, Phoa* Tan
A-Liang,was adOpted in from a different place in T'ou-wei. As her
name indicates, she was surnamed Tan originally, and then took the
surname Phoa* as a prefix. This indicates that she most probably
was intended to be a T. simpua, (Thisis one case in which the
records are insufficient, and I have been unable to find what hap-
pened to her.)

In 1910 a second son was born, and died soon after birth.
The next year a third son also died soon after birth. Finally the
last child, a daughter, A-eng, was born in 1912. She evidently
became the main hope of the family for the production of descendants;
in 1930 she married in Khu Tiong-chin as her uxorilocal husband.
Khu (M. Ch'iu) was from T'ien-sung-p'i (see the "Ch'iu B" case),
and was not an Aborigine, but a Hakka. He did not change his surname,
or add his wife's to his own, but their two children both took the
surname Phoa*. Both of these were girls, and one died at infancy,
in 1934. In the same year A-eng and Khu divorced, and he left the

household. Two years later an illegitimate son was born to A-eng, a

226

son who took the surname Phoa*. Then in 1938 she married in a
second uxorilocal husband (thus an "uxorilocal husband" as opposed
to Khu, who was an "uxorilocal son-in-law“). This man was named
Liau Boa*-hong, another Han Chinese (though itis not Clear whether
Hakka or Holo) from outside the I-lan area. He did not change his
name, and all of the children (up to 1945 there were two boys and
two girls) took his surname.

Meanwhile, the son of the family, Phoa* Bu-li, had adapted
in a daughter (her status was M. yang-n0, "adopted daughter") from
Chuang-wei (in I-lan) in 1931. Her name was A-iu, and her ances-
tral place of origin was listed as Fukien. She took the surname
Phoa* when she was adopted (at age twelve), and her illegitimate
child, It-long (a Japanese name; Ichiro), who was born in 1939,
also took the surname Phoa*.

Phoa* Bu-li married out uxorilocally to Phoa* O-be, thus
violating surname exogamy, but divorced her in 1934. Finally he
married out uxorilocally again in 1938 to I-lan city, a woman named
Te* (M. Cheng).

At some time in the 19405 the family moved out of the village,
which is one of the reasons that there are gaps in the records.

In this case, a family followed over four generations, there
is only one "regular" marriage, but four uxorilocal marriages, two
in and two out. There were two daughters adopted in, one evidently
a m, the other an "ordinary" adopted daughter, and there were

two illegitimate children.

227

Case II: P'an

This group of relatives is also surnamed T. Phoa* (M. P'an),
but they say that they are not Aborigines, but completely Han
Chinese, so I will use the Mandarin romanization "P'an." The
data that I have, as far as they go, bear this out.

The founders of the local branch of the group were two
brothers, P'an T'ien—mu and P'an Shih-ying, who arrived in the
village together in 1924 from Yuan—shan (in I-lan). The older
brother (1871-1938) was the head of the household, which included
his younger brother and the children of both of them. The older
brother's wife, however, died before they moved to T'ien-sung-p'i,
and so there is little information on her. The second brother's
wife, Yu Fen, was born in 1880, and was adOpted into the P'an family
when she was three years old. She was married to the second son
of the family in 1897, when she was seventeen and he twenty-four.

The older brother and his wife had four children of their
own, three males and a female. The senior male of this generation,
P'an A-ts'ai, married a woman surnamed Yu, but she died before they
moved to the village, so information is lacking on her status.

The wife of Han-ch'uan, the second son, was a simpua, She was
adopted in from Yuan-shan in 1906, at the age of one year. They
married in 1924, when she was nineteen and he twenty-one. He lived
only six more years, though, dying in 1930. Three years later an
uxorilocal husband was married in for her, a man named Kao Niang-huo.
Ten years after that, in 1943, they divided from the estate and

moved elsewhere in T'ien45ung-p'i.

228

The T. simpua, Lin A-yeh, had three children by the P'an
son, and another son later. This later son, P'an Hsi-shui, was born
eleven months after Nan-ch'uan's death, and so was classed as illegi-
mate. In later records, however, his father is listed as Nan-ch'uan.
In 1935 a two-year-old T. simpua, named Ch'en Hung-ts'ao was adopted
in from A-li-shih (in I-lan, between the village of T'ien-sung-p'i
and Lo-tung), and she and Hsi-shui were married in 1952 when she
was nineteen and he twenty-one. This couple has had four children,
the first born in 1954.

The uxorilocal husband, Kao Niang-huo, adOpted in a girl
surnamed Li in 1955, and she changed her surname to Kao. She then
married out (within the village) in 1962, when she was twenty.

The third brother of this generation of siblings, the brother
(to P'an A-ts'ai and P'an Wan-ch'uan, was named Chin-ch'ien. He
lived from 1908 to 1960. He and his wife, who was married in nor-
mally, had a large family. Though two children, both sons, died
when young, they still had eight remaining children. In 1950, when
they had three living children, two of them male, they adopted in a
T. simpua_named Lan Hsiu-mei, who was then one year of age. In
1966, when she was seventeen and her intended, the third son, Tsai-
sheng, was twenty-seven, they were married. They have had three
children so far, born in 1966, 1969, and 1971.

This is an interesting case, as the age difference between
the two (ten years) is greater than usual. This means that when
she was adopted in, he was already ten years old. There cannot be

muchtyf an argument in this case for sibling socialization. They

229

were married November 10 (the date in the household records), and
their first child was born December 11, 1966, just a month later.

To follow up the rest of the family, we must go back to the
two brothers who arrived in T'ien-sung-p'i in 1924. The younger
brother's wife, as has been noted, was a T. simpua, They had two
‘children, a son and a daughter. The son, Nan-tzu, was born in 1905.
He and his two sons and their wives and children were no! neighbors
and friends when I lived in T'ien-sung-p'i. He died in 1974, about
a year and a half after I left the village.

His wife also was a T. simpua, She had been adopted in in
1920 (when she was thirteen years old), before they moved to T'ien-
sung-p'i. The two of them married five years later. They had two
sons, the first born in 1930, five years after their marriage. His
name is Jui-mao. The second son, Ch'uan-ch'in, divided from the
rest of the family in 1971, but remained in the house; he and his
family ate separately after the split, and Operate in most ways as
an independent economic unit. The grandfather, as he was called
(T. A-kong), was still listed in the records as the head of the
household that contains the older son, and he took turns eating
with the families of the two sons.

One day in 1972 my assistant and I accompanied the wife
(Liu Mei-yun) of the younger brother, her two children, and another
neighbor to a female spirit medium in a hamlet not far from T'ien-
sung-p‘i. The older girl, Wen-1i, had been ill for several days and
had a fever. In a "trance“ (this was one of the few times that I

saw a medium who appeared to be "faking" the whole performance;

230

usually one can be easily persuaded that the medium is actually in
an altered mental and physical state for perhaps thirty seconds or a
minute), the medium first said that there was something wrong with
the child's fate, and that a ceremony called "settling the horoscope"
(T. an peh-Ji) was needed. She then said that a girl of the father's
generation, who had died early, was causing trouble. The girl's
mother said that she knew nothing about it. The spirit medium
became impatient, and said, "A female. A female of the father's
generation. You go home and ask him about it. There are already
two reasons why the child is sick, and furthermore this is the
child's bad year. She had better watch out for water--all kinds of
water, hot water, ditch water, and so on--and traffic. If she walks
on the street or crosses the road, she'd better be careful."

She then prepared four written charms (T. by) and a packet
of "fairy pills" (T. sian-tan). One charm was to be burned before

the gods of the bed (T. chhng-kong and chhng-bo), one was to be

 

burned in front of the house for the wandering spirits, one was to
be burned and mixed with water to bathe the child, and the last
was to be burned in front of the ancestors' tablets while the
female of the father's generation was prayed to to protect the
child. The pills were to be put in water for her to drink.

It was then the second child's turn. The medium said that
the child had a cold, and then asked whether or not the mother had
taken her to a fortune teller. The mother said no. The medium
laughed, and said, "Well! She has a simpuafs fate (T. sim-pu-a mia*)

and if you don't give her away to someone else, then at least you

231

should give her in 'pro-forma' adoption (T. khe chhui-bi)." The

 

mother said nothing, and the medium repeated this several times.
Then she said, "There is a woman of the child's mother's genera-
tion. . . ." The mother, seemingly embarrassed, admitted that yes,
there was an older sister. "Right!" said the medium. "That sister
of yours wants to share your husband. You had better choose a day,
and then at midnight welcome your sister to your home. Do you
understand?" The mother was silent. The medium then continued,
saying that the child also had fright, a common folk illness. She
prepared four charms similar to those for the first child, except
that one of them was to be taken and burned at the mother's home in
I-lan City while a "ghost marriage" was concluded between the mother's
husband and her dead sister.

There were two further clients at this session, and both of
them were advised by the medium that part, at least, of their troubles
was due to female relatives. In one case it was a woman's daughter
who had died young (the woman was past middle-age), and in the other
some unnamed woman of a man's father's generation (the client was an
adult male).

Later, when we had returned to the village, this suggestion
of the ghost marriage was put to the father of the children, P'an
Ch'uan-ch'in. He simply snorted and refused to have anything to
do with such an affair. This caused an argument between him and
his wife, which was joined by his sister-in-law. As far as I know,

the marriage never did take place.

232

This case is interesting in a number of ways. There are
twelve marriages about which enough is known to assign them to par-
ticular types. Of these twelve, five are T. simpua marriages, and
one is uxorilocal. There is furthermore an adopted daughter, and
at least two daughters of the last generation who have been adapted
out.

The T. simpua_wife of P'an Han-tzu, Lai Ying, was adapted
in when thirteen years old. This seems a case near the borderline
between adoption and marriage. In the household records for the
Japanese era, at least, there is evidence for girls being married
as young as fourteen. It seems obvious that just a few years made
the difference between her being brought into the household as a
T. sjmpg§_and as an "ordinary" wife.

One of the other cases of simpua marriage, that between P'an
Tsai-sheng and Lan Hsiu-mei, lies toward the other extreme. Since
he was ten years old when she was born, he was almost as close, in
age, to being a parent as he was to being a sibling; it was thought
that the age difference between the adOpter and the child being

adopted should be about twenty years.

Case III: M. Wu; T. Go

This family has a very complicated history, which involves
many surnames. It is perhaps best to start with a woman named
Li Wei, who was born in 1874 and died in 1925. She married three
times at least, and possibly a fourth. She may have earlier married

a man surnamed Lin, but in the earliest records she was then married

233

to Wu Lao-hsia. When he died or they divorced (I have not been
able to determine this), she married, in 1914, a man named Ch'en
Shan-chih. Two years before their marriage, he had recognized as
his child an illegitimate daughter of hers who was born in 1911.
Long before this, in 1893, she (the woman Li Wei) had given birth to
a son, Wu Chin-sheng, fathered by Wu Lao-hsia. It is to this son
that she and her daughter, Ch'en A-mei, went in 1918, after her
husband Ch'en Shan-chih had died in 1917.

Then in 1919 Li remarried again, this time to Huang A-tsung.
At the same time he adapted, as simpua, Li's daughter; when Li
married in, bringing her daughter, it was what is called in gugzyg
M. szu-tai; the daughter's connections with her natal kinship line
were severed, and she became a Huang. This family then was composed
of Huang, Li, the daughter of the latter, now surnamed Huang, and
an adopted son, Huang A-ch'eng. He was the son of a Sinicized
Aborigine named Ch'en, who was married to another Aborigine, sur-
named HsU, and had first been adopted to another HsU (who may well
have been a relative of the mother). In l9l3,when he was two years

old, he was adopted (M. ming-lingftzu) to Huang A-tsung. In 1931,

 

when he was twenty, he married out uxorilocally to the adopted
daughter of another man surnamed Huang, in Wu-chieh (in I-lan).
His occupation was listed as manual laborer.

Meanwhile, Li's daughter, who had been adOpted in by Huang,
was married in 1927 to an uxorilocal husband named Chang Lien—mao.

They had two children, and then, with the children, moved out, back

234

to Chang's original home in Erh-wan-wu (in San-hsing). This is what
is called M. ch'u-she, "leaving the (uXorilocal) household."

To return to Li Wei's first (or second) marriage, to Wu
Lao-hsia, she had by him a son named Wu Chin-sheng (she and her
illegitimate daughter had resided with the son for a while after
the death of Li's second husband). Chin-sheng was born in 1893,
and in 1914, when he was twenty-one, he married uxorilocally into a
family surnamed Tai.

This Tai family had, in a sense, started out surnamed Huang.
Huang T'ien-szu had been the head of the household until his death
in 1905. He was married to a woman with the name Tai Lin A-lou.
There is no conclusive evidence, but it seems probable that Huang
was an uxorilocal second husband to Lin A-lou, who had first been
married to a man surnamed Tai. 'At any rate they had a daughter in
1880, Huang Chiu-mu. In 1901, when she was twenty-one, Tai A-shih
was simultaneously adopted in and married uxorilocallyto Chiu-mu.
She did nine years later, and in 1912 he married a second time, to
Lin YU. (This marriage was an ordinary one.) Tai had three children
by his first wife, but two years before their marriage he had
adopted in a T. simpua; she was intended to marry a son of his later,
and so kept her original surname--Liao A-yU. Subsequently, however,
plans changed, and she took the surname Tai and her status was
changed from that of T. simpua_to that of adOpted daughter. Then,
as adopted daughter, she married in an uxorilocal husband, Wu
Chin-sheng, in 1914. Four years later, in 1918, they left the Tai

family (M. ch'u-she).

235

Wu Chin-sheng died only seven years later, in 1921. By
that time, however, he and Liao (now Tai) A-yfi had produced three
sons. After his death, she adopted in three daughters and married
in an uxorilocal husband for one of them. The first such daughter
was adopted in 1922, when she was not yet one year old, from YUeh-mei
(the present San-hsing). Ten years later, for reasons unknown to
me, her adoption relationship was ended, and she was returned to
her natal home.

The second daughter was adopted in in 1930, from T'ien-
sung-p'i, when she was two years old. Her surname had been Shen,
but was changed to Wu, so her status was that of adopted daughter,
not adopted daughter-in-law. Thirteen years later, in 1943, she
was married to an uxorilocal husband, Chu Wan-chin, a Sinicized
Aborigine from Yiieh-mei. (The rest of the family is Holo.) Wan-Chin's
brother, Chu A-huo, by a strange coincidence, is mentioned in an
article by Juan Ch'ang-jui on the Sinicized Kavalan Aborigines in
the I-lan area (Juan 1968: 6). The fact that A-huo's affinal con-
nections extend from T'ien-sung-p'i to Wu-chieh (in I-lan) is used
as part of an argument by Juan that the Sinicized Aborigines have a
wider area of marriage ties than do the Han Chinese.

Chu Wan-chin remained in the Wu household for four years,
after which he and the Wu's adopted daughter moved out with their
daughter, who had been born in the same year. The daughter took the

Chu surname, not Wu.

236

In 1946 a ten-year-old girl surnamed Liao was adopted in
(changing her surname to Wu). She stayed only five years, however,
and then married out in 1951.

The oldest son of Wu Chin-sheng is Wu A-shu. He was one of
our neighbors when we lived in T'ien-sung-p'i. His mother, Tai
A-yfi, is still alive at age seventy-seven, and in good health.

A-shu married a woman surnamed Ch'iu, by whom he had two sons. She
died in 1948, and shortly thereafter he married a woman named
Chuang A-hsiang, by whom he had one daughter, born in 1956.

The household as presently constituted includes four gen-
erations: Tai-A-yU, her son Wu A-shu, his son and his wife, and four
of their children. A-shu's second wife, A-hsiang, lives in Taipei
and works as a cook. She maintains economic ties with the family,
however, and returns once or twice each year to the village.

The family is quite poor. It owns five fen of rice land,
which it farms itself, but its main source of income is
manual labor. Both A—shu and his son work, usually in the mountains
on construction jobs. They would not be in such a difficult situa-
tion if A-shu's second wife had not gone bankrupt trying to run a
drygoods store. They had to go into debt to pay the creditors, and
still have five or six years to go before the debt is paid off.

The daughter of A-shu and Chuang A—hsiang became pregnant
by her lover in 1972. They have not even had the M. ting-hun
(betrothal) because A-hsiang disapproves of him. This is regarded
as unfortunate, but it is said that they can just go and live together

as man and wife anyway.

237

A-shu's second son, Shu-sung, was married not long before
his accidental death while working in the mountains in 1967. (SUCh
deaths are not uncommon, and have dissuaded some from taking such
jobs.) His wife gave birth to a son, but after a dispute with
A-hsiang, she was driven out of the household, taking the child with
her. The brideprice (NT$38,000) was not yet all paid to her family,
so the son was used as a pawn in the bargaining. The wife's family
said that only if the brideprice were fully paid would the child be
surnamed Wu, and follow that kin line. A-hsiang refused to allow
the money to be paid, and so the child took the mother's surname
and not the surname Wu.

Shu-sung thus had a son but no descendants, and so in 1972
one of his older brother's sons, Wu Chang-thn, was posthumously
adopted to his uncle, to continue that line and the sacrifices.

The contract of that adoption was written up by the village Taoist
priest, Yeh Liang-ch'uan (see the Yeh case) on red cloth, and it
was presented to Shu-sung's soul tablet, along with incense and
sacrifices. A copy and translation of the document can be found in
Appendix B.

A few words should be said about the Chuang family, from
which came Wu A-shu's second wife, Chuang A-hsiang. Her father,
Chuang A-k'o, had at one time been a temple keeper, and was evi-
dently a builder of sorts; around 1920 he built a number of houses
on the main street of the village. He died in 1952, and ownership
of the houses passed to his oldest daughter, since his only son had

died two years before.

238

By this time A-hsiang, the second daughter, had already
married Wu A-shu (the second marriage for both of them), and he
bought one of the houses from his sister-in-law (at a good price,
it was said, since they were related). She kept one of the houses,
the one next door, and so she and her husband were our landlords,

and the Wu family our neighbors.

Case IV: M. Yeh, T. Iap

This family is that of the village Taoist priest (T. sgj;_
kggg) and his son, an apprentice Taoist priest. They used to be
resident in Wu-chieh (in I-lan), and moved to T'ien-sung-p'i in
1923. The head of the household then was Yeh A—li. He was born
in 1868, and had become head of the household when he was nine years
old, when his father died. He married in 1886, when he was eighteen
and his wife fifteen. They had one son, then adOpted two girls in.
One was adopted in as a T. simpua.(keeping her natal surname) at
the age of seven, in 1900. Nine years later, at the age of sixteen,
she married the Yeh son, who was then twenty.

The second girl was brought in as an adopted daughter, and
so changed her original surname (Chiang) to Yeh. She was adOpted at
age two in 1906. She later married out, to a man surnamed Lin, in
the village of T'ien-sung-p'i.

The son, Yeh Ch'eng-t'ung, and his T. simpua_wife Wei
A-chfi, then had a large number of children--two sons and at least
six daughters. Two of the daughters are listed in the household
records as first daughter, and two are listed as fifth daughter; I

have not been able to clear up this confusion in birth order.

239

One of the daughters married in an uxorilocal husband, and
one of the sons (the second) married out uxorilocally. Two daugh-
ters were married out in the regular way, and another was adopted
out as a simpua, Her adoptive relationship was ended, though, when
she returned to the family, and then married out in the ordinary
way. One daughter remained unmarried, and one was adopted when she
was fifteen (a very advanced age for adOption).

The chronologically oldest daughter was married to an uxori-
local husband, Lin Ch'ing, in 1925, when she was fifteen and he
eighteen. They remained in her natal home for eleven years, and
also had a large family; though three of their children died in
infancy, seven survived. The oldest son married out uxorilocally
in 1954 (when he was twenty-seven), to an adopted daughter of a man
in Keelung, the port on the northeast coast. One daughter married
out when she was seventeen to a man who lived in Taipei. The
third daughter was adopted to the man who later became the mayor
of one of the administrative villages that make up T'ien-sung-p'i,
and later married out to Lo-tung. The fourth daughter was adopted
out as a simpua_to Ch'iu Wang-chih (see the Ch'iu case) when she
was nine years old. In 1958, however, when she was twenty-one,
the relationship was dissolved, and she returned to her natal home.
The rest of the children followed the parents when they moved to
Taipei in 1957. None of the children took the surname Yeh.

The oldest son, Yeh Liang-ch'uan, born in 1919, became
T'ien-sung-p'i's Taoist priest. In 1943 he married a woman who had

been surnamed LU, but had been adopted out, and so changed her name

240

to Chuang; when she married him she took his name, and so changed
again, to Yeh. They then had six children, four of whom survived,
and adapted one girl. This adoption is puzzling; she was eighteen
years old when adopted, and stayed with the Yeh family for only
four years, marrying out to Tung-shan (in I-lan).

Yeh Liang-ch'uan, the Taoist priest, is the head of the
household. Also resident (officially) is his brother Yeh K-un-
sung and his family (see further on in this case). Resident
unofficially is the head's "comnon-law" wife,Ch'iu P'an Chin-kuei.
This woman is the widow of one of the Ch'iu surname group members.
Her husband died in 1961, leaving five children ranging in age from
new-born to eleven. Just when her relationship with Yeh Liang-ch'uan
started, [am not sure. A daughter born to her in 1963, and listed as
illegitimate, is widely accepted, however, as Yeh's child. In
1970-71 she, her father-in-law, and all of her children were officially
coresident with Yeh. Then they set up a new household. Though she
is officially registered there, she in fact lives with Yeh. Though
their relationship is not formally marriage, it is treated by others
in the village as equivalent to marriage in most respects. When
she fights with Yeh and the fights become common knowledge (a: very
common occurrence in a rural village, of course), she is sometimes
berated by her father-in-law and others of the Ch'iu group for
complaining. Though Yeh has not married her, they say, he is
better than many husbands; he provides for her, he does not beat
her, and he does not waste money on drinking and such things. She

retorts that he is always away nights (the implication seemingly

241

being that he is seeing other women), and their counter-argument is
that that is just part of the job of being a Taoist priest; one
has to make night "house-calls."

Yeh Liang-ch'uan's son is following in his footsteps, and
is training to be a Taoist priest. In 1970 he married a woman from
within the village--eighteen-year-old Lin Mei-ling. She was orig-
inally surnamed Chang, but had been adopted out to a Lin family.

The younger brother of Yeh Liang-ch'uan, Yeh K'un-sung,
married out uxorilocally to a woman surnamed Weng in T'ao-yuan,
on the other side of the island. Two years later, they divorced,
and he returned to this household. In 1952 he remarried, and has
had four children by his second wife (this marriage is virilocal).

He is a laborer, and works in the mountains (forestry).

§g§g_y; M. Ch'iu, T. Khu

This family should probably be included in the other Ch'iu
case, but since I cannot prove without a doubt that they are related,
I will treat them separately.

The family moved from Wu-chieh in 1924. The head of the
family at that time was Ch'iu Ch'i-yuan, who lived from 1861 to
1933. He and his descendants are listed as having Fukien as their
place of ancestral origin, but they are southern Fukien Hakka
speakers.

Ch'i-yuan married Huang Huan in 1885, when he was twenty-
four and she sixteen; it was an ordinary virilocal marriage. They
had three sons, born in 1888, 1901, and 1909, and adopted in three

girls as T. simpuas for the sons. The first of these T. simpuas,

242

Lu Wen, was identified as a Holo, of Fukien origin, but her name
(Wen, which has many variant characters, often meaning "lips,"

"to kiss") is usually a Sinicized Aborigine name. She was adopted
in from Erh-chieh (in I-lan) in 1901 when she was ten years old,
and five years later she married the oldest son of the family.

The second T. simpua_was Yang A-ts'ung. She was originally
surnamed Lin, and evidently was earlier adopted to someone named
Yang (no further information on this). She was adopted in by the
Ch'iu family in 1922 when she was fourteen. Two years later, when
she was sixteen, she married Ch'iu Chung-sung, the second son.
They then had ten children--seven boys and three girls. The
oldest daughter married out, and the second one was adapted out in
1936 when she was five years old. The oldest son married in 1954,
when he was twenty-seven, to a girl from T'ien-sung-p'i, and the
second son married in 1960, when he also was twenty-seven. When
his wife married in, however, she brought with her two daughters,
one born in 1956 and one born in 1958. He recognized the younger
one at the same time as the marriage, and her name was changed to
Ch'iu. He did not, however, recognize the first daughter, and she
retained the surname Yeh. In 1960 Ch'iu Chung-sung and all of his
family moved out to Su-ao Chen.

A third son (Ch'iu Chung-chin; he is listed as the sixth
son, but this seems to be an error--there is no evidence of any
other sons) married out uxorilocally to P'an A-ying (see the P'an
case), later divorcing her and moving elsewhere. A T. simpug_

was adopted in, evidently for him, in 1912 when she was two years

243

cflcL (This is Lin A-fen.) This marriage, however, did not work out;
in 1930 she had an illegitimate daughter, in 1931 her status was
changed from that of T. §iflgg§_to adOpted daughter, and one month
after that she was married out to Taipei.

The oldest son, meanwhile, Ch'iu Chung-chien, and his simpug_
wife Lu Wen, had three of their own children, and adopted in a
fourth. Their oldest son was Ch'iu A-pai, who was born in 1913.

In 1923 a simpua, Lin A-wei, aged four, was adOpted in for him from
Wu-chieh. In 1938, when she was nineteen and he twenty-one, they
were married. This couple then produced five children of their own,
and adOpted in a T. simpug_for the oldest son. This girl was adopted
in from T'ien-sung-p'i in 1948, when she was three years old, taking
the surname Ch'iu. In 1960, when she was fifteen, the adoption
relationship was dissolved, and she reverted to her original sur-
name, Liao. At the same time she married the oldest son of

the family. They have so far produced six children; the sixth and
last was born after I left T'ien-sung-p'i, in 1973.

This household now includes the T. simpug_wife (Lin A-wei)
of Ch'iu A-pai (he died in 1958), her oldest son, Ch'iu Shun-t'ien,
and his simpua wife, Ch'iu Yfi-lan (and their children), and Shun-
t'ien's brother, Ch'iu T'ien-lai. The mother, A-wei, is the head
of the household. The family owns about 400 tangerine trees in
the foothills of the mountains, and so is fairly well-off (solidly
in the middle class of a three-class system). Ch'iu Shun-t'ien is
an employee of the Taiwan Electric Company, T'ien-sung-p'i Hydro-

electric Station. When I visited T'ien-sung-p'i last, in 1974, he

244

was away in Taipei for close to a year's advanced technical training
for the Electric Company. He is the graduate of a Keelung technical
school.

This family is more wealthy than most, is better educated
than most, dresses better than most, and presented the image of
being more "modern," 50phisticated, and upwardly mobile than most,
all characteristics which contradict most of the commonly held
stereotypes about those who indulge in T. simpug_ad0ption and marriage.
This after three generations of successive T. simpug_marriages.

Of Ch'iu A-pai's other children, the second child, a
daughter, was married out in 1964 and the second daughter was
adopted out to a Lin family in Wu-chieh in 1961, when she was twelve
years old.

Out of eleven marriages contracted by members of this family
for which we have sufficient evidence for classification, four of
them were simpua_marriages, and one was uxorilocal. There was one

unconsummated simpua marriage, and two girls were adopted out.

Case VI: M. Liao, T. Liau

This and the preceding case are different from the previous
ones in that they are really a large number of cases interrelated
by kinship; the overall unit is a lineage.

I selected the Liao group because they are very important in
the history of the village of T'ien-sung-p'i as well as in its
present. It is also a large, prolific group, which has been in the
village for close to seventy-five years, and so there is a good deal

of information on the members.

245

The ancestors of the Liao lineage in its local version were
Liao Cheng-ch'ing and his brother Liao Cheng-fa. They arrived in
the village from the other side of the island around the turn of
the century. This was around the time that the Japanese Colonial
government began to gain greater control over the Mountain Aborigi-
nes, and was able to increase security in the area. This Opened
up a good deal of land. A man named Ch'en Shan-chu from I-lan
City got control of the land, and recruited a number of Sinicized
Aborigines to open it up and begin to develop it. Later, I was
told by some informants, he sold much of the land that was thus
prepared for agriculture to the Hakkas, particularly the Ch'ius
and the Liaos.

The Liaos themselves, on the other hand, say that they were
poor and landless tenants in those days, and the ancestors came to
the I-lan area searching for land of their own. They had first
gone to P'u-li, in the center of the island, but the older brother
got sick on the bananas there. They took this as an ill omen, and
moved on.

They heard that there was good land available in the I-lan
area, and came to look around. The Liaos joke today that the
brothers decided to settle in the village through a misunderstanding.
The two had arrived shortly after the New Year, and noted piles of
rice still standing in front of the houses. Their conclusion was
that it must be a rich land to have so much rice left so late in
the year. The joke, of course, is that the I-lan Plain is known as

one of the wettest areas in Taiwan, and getting the rice dry before

246

it is damaged by mildew or fungus attacks is still a big problem
for farmers today, one that has an influence on overall production
statistics.

However it happened, the Liaos managed to obtain large
areas of land near what was then the irrigation pond. In the early
days they maintained an ancestral temple there, but later, as the
lineage grew and its organization weakened, one of the branches
gained control over the land on which the temple stood, and con-
verted it into an ordinary house.

Apart from this temple, they did not own any land or other
property in common. They were poor, they say, when they arrived,
and what land they had was divided among the sons, and there was
none in common. The lineage seems to have suffered its greatest
segmentation in the eighteenth generation; the temple stood on about
five fen of land. Today two members of the nineteenth generation
own the land.

Around 1910 the Liaos built the first rice-mill in the
village. It was part of a general store, selling rice, meat, wine,
and articles of daily use. The mill was at first man-powered, then
powered by a water-wheel, and last by electricity; in the 19205
the Japanese finished the hydroelectric station. Thus the Liaos
were "in on the ground floor," and were in a position to profit from
the boost given to business when the many laborers and others moved
in to build the power station and the narrow-gauge railway system.

From those early days, the Liaos have been one of the dominant

groups economically. They may have been poor when they arrived, but

247

they quickly became one of the richer groups in the area. They are
always owner-cultivators, never tenants, they have extensive busi-
ness interests in the village and elsewhere, and their sons and
daughters have a consistently higher level of education than do
most of the others in the village. The dominant branch still oper-
ates the rice mill, they live in the biggest and one of the newest
houses in the village, and they have recently purchased a new truck,
their second, for hauling rice and fertilizer, costing NT$30,000.
There are, generally, two different kinds of Liaos. One
kind, the type of which this lineage belongs, is called "split" or

"portmanteau" Liao (M. shuang-Liao, T. siang-Liau). The other kind

 

is called "single," or "simple" Liao (M. tan-Liao, T. sien-Liau).
The former become Chang (T. Tiu*) at death; that name is inscribed
on the tomb, and sacrifices are offered to the ancestors in the name
of Chang. The second kind, the tan-Liaos, are Liao in death as well
as in life. The former group, the "split Liaos," cannot marry those
surnamed Chang, Chien (T. Kan), or, of course, other Liaos. Their
relationship with the Chien surname is different from that with
Chang; there are no Liaos who become Chien at death. It was admitted
that "times have changed" and that there was little that one could
do if such forbidden marriages were actually proposed. Still, it
is considered wrong.

The explanation for all of this, according to the Liaos
themselves, traces back to events on the mainland of China several
hundred years ago. They were originally surnamed Chang, but one of

their ancestors, forgotten now, married uxorilocally into the Liao

248

surname group, and they took that surname. They remain Chang,
however, "in the bones," and so identify themselves as Chang to
the gods when they T. pai-pai ("worship") and have Chang inscribed
on the ancestral tablets. This was called "eating their surname"

(T. chidh i-e si*); taking on the surname of a group in order to

 

gain protection or other benefits from that group.

All of the Liaos in this lineage are "split" Liaos; there
are other, "single" Liaos in the area, though. In P'o-pu-wu, the
settlement a kilometer or so the east of the village, there are
"chicken-head," "single," Liaos, whose male members cannot eat the
heads of chickens, but must give them to the women of the group to
eat.

For whatever reason, the Liaos of this lineage (there are
other, unrelated Liaos living in the village) seem not to have
attempted to convert their economic standing into one of political
power or influence. They are in no sense community leaders, and are
the focus of a good deal of what class conflict might be said to
exist in T'ien-sung-p'i. They interact very little in the social
life of the village. They seldom participate in T. pai-pais,
religious festivals, and in fact Liao Ta-chu, the head of the
economically dominant branch in the nineteenth generation, the
group which now runs the rice-mill, is a nominal Catholic.

There are only about thirty Catholics in the village. They
are clearly from the lower socioeconomic class in the village, and
are seen as socially marginal in local terms. At times, in fact,

they are seen as troublesome because of their refusal to participate

249

in various religious observances, a refusal which is seen as pos-
sibly having adverse effects on the village as a whole.

Why the Liaos have chosen to alienate themselves from the
social life of the village is not clear. Liao Ta-chu's Catholicism
is nominal in the sense that he is not a regularly practicing and
active Catholic, but seems rather to bring up this identification to
define himself in Opposition to the majority of the other villagers.
This feeling is reciprocated.

We were told by the present mayor (M. ts'un-chang) of
T'ien-shan Ts'un, who is Liao's brother's son-in-law, that Liao Huo,
the father of Liao Ta-chu and his brothers, and who ran, with his
brothers, the store and rice mill in the earlier generation, was
dishonest and cheated the people by falsifying their accounts with
him. He was so notOrious at this, and so hated, that when he died.
in 1965, no one was willing to act as pallbearer. This caused
great scandal and embarrassment, and the Liaos finally approached
the mayor and appealed for his intercession. He was able to get a
sufficient number by arguing that this was all making the village
lose face, and by trading on his own personal standing, asking
peOple to do it as a personal favor to him, not the Liaos.

It may be that the Liaos are upwardly mobile enough that
they now identify with classes in the cities, outside T'ien-sung-p'i,
where they do have economic interests. There are indications that
they do see themselves as in some sense superior to most of their
fellow villagers, and they may remain there simply because of their

holdings in land, the rice mill, and the store.

250

However one might define rural social classes, and draw
the lines between them, it is clear that the Liao lineage as a group,
and indeed most of the individual families, are in the "rich" cate-
gory. Those who are not are clearly in the middle category of a
three-class system, and usually in the upper part of that. None
are "poor." It is thus interesting, for comparative purposes, to
examine the types of marriage and adOption practiced by the Liaos
during their residency in T'ien-sung-p'i. A summary of this infor-
mation, taken from the appendedgenealogy, is as follows: Of a total
of seventy-five marriages, twenty-three were marriages of women out,
and the remaining fifty-two resulted in the couple's residence, for
a longer or shorter period of time, in the Liao group. Of the
fifty-two, thirty-two were "regular" marriages, fifteen were
T. simpug_marriages, and five were uxorilocal marriages. Thus mar-
riages into the group can be distinguished as "regular" and "irregu-
lar" types; there were thirty-two and twenty, respectively. If,
however, we widen the scope of interest to "persons recruited through
a social transaction involving rights, duties, control over repro-
ductive powers, etc.," the picture changes a bit.

Fifteen children were adopted into the group (not counting
those adoptions which resulted in marriage; those have been classed
as "T. simpua_marriages" and not "adoptions"). If these are then
added to the "irregular" category, the result is that the "irregu-
lar" forms are about equivalent in number of the "regular" forms.

In fact, Of the adOpted children, ten were female. As has been

noted in the sections on adoption and T. simpua marriage, it is very

251

hard in many cases to distinguish "true" adopted daughters from

T. simpua_adopted girls. (hisome cases the adopters themselves are
not sure what will eventuate.) Of the ten adOpted girls, at least
three are explicitly mentioned in the records as having been
intended as T. simpua, but later having been changed in status to
"adopted daughter" to be married to someone else.

There were also adoptions out--fourteen, all females. Of
the fourteen, four were specified to be adoptions out as T. simpua,
and Of course a good prOportion of the rest must also have become
T. simpua, Taking marriages and adoptions together, one can get an
idea of the expansion of the Liao lineage; thirty-seven persons left
the group, either through adOption or marriage, all of them females.
Sixty-seven persons, however, entered the group, ten of them males
(five in uxorilocal marriage and five adopted) and fifty-seven Of
them females (thirty-two in regular marriage, fifteen in T. simpua
marriage, and ten in adOption).

The Liao group is too large to permit a systematic descrip-
tion here, but a few particularly interesting aspects can be pointed
out. On page one, for instance, of the genealogy, the case of the
T. simpua_Wu Shuang is a good illustration of the degree to which
such an adopted girl is incorporated into the adopting group. She
was adopted into the Liao family in 1900, when she was one year old.
(The adopter, Liao Cheng-fa, was one of the two brothers who
established the T'ien-sung-p'i Liao lineage.) In 1918, when she
was nineteen years Old, her status was changed from that of T. simpua_

to "adopted daughter," and the next year she was married out, in

252

regular fashion. Then in 1925 she divorced that husband, and
returned, to marry a second time in 1926. It is significant here
that after her divorce, she returned to the Liao family and not her
natal family. Thus her status in the Liao family was essentially
different from the affinal relation created by marriage. A woman
who was married into the family would never lose her ties with her
own natal family, which means that the husband's family's rights
over her and her reproductive powers were curtailed to some degree.
The transference of rights in T. simpua adoption, however, was much
more complete. It sometimes happened that such an adOption was
dissolved, and the girl returned to her natal family, but it was
much more common for her, if she did not marry a son of the family,
to be transformed from a "daughter-in-law" to a "daughter,"

so that the family still retained rights in her and her sexuality
(the right to arrange and conclude a marriage for her).

There is also a very complicated set of M. kuo-fang adOp-
tions. The older of the founding brothers, Liao Cheng-Ch'ing, had
three sons. The youngest of these, Liao Ch'ing, did not have any
sons until late in life (when he had two), and so he adOpted in two
sons of his oldest brother, Liao Shan (who had six sons). These
adoptions, in various records, were termed both M. kuo-fang and

M. ming-ling-tzu. One of these M. kuO-fang adOpted sons, Liao

 

Ta-wan, then married and had three sons Of his own. One of these
he then adopted back to his original branch, to one of hj§_brothers,
Liao Shan's second son, Liao Ta-ang. The latter had married twice,

but still had no sons, so he adopted a different-surname child

253

(termed a M. ming-lingytzu, and who had been the illegitimate child

 

a woman surnamed Yu). Two years later, at age three, the boy died.
The M. kuo-fang adopted son "saved the game," however, producing
two sons. This is called "borrowing the son and returning the

grandson" (M. chieh tzu, huan sun), though this phrase is more often

 

used in non-kuO-fang adoptions (see Chapter IV).

Case VII: Ch'iu, T. Khu

This case, like the Liao case, is really a series Of
cases connected by kinship. The Ch'ius arrived in T'ien-sung-p'i
perhaps ten years after the Liaos did. The records state that they
arrived in 1915, but informants indicate that the actual date was
somewhat earlier, and 1915 is the date of the official transferral
of the household registration. The head Of the lineage then was
Ch'iu A-ch'uan, also known as Ch'iu Ch'fian-liu. The former version
is from the household records, and the latter from a genealogy
loaned to me by the Ch'iu group. There are a number of such name
differences. Unfortunately I was not aware of this when I was
still in T'ien-sung-p'i. Only when I left the field and
was able to finish the index of my household registry data, and
so could compile all of the information on the Ch'ius, did the
discrepancies appear. There are a number Of possible explanations
for this situation.

One is that genealogies, as ritual, retrospective documents,
seem to be more organized and complete with respect to generational

names than are the names actually used. Either this is because

254

these are "ritual" names, used in complete form only for genealogi-
cal and other uses, or there is in fact a good deal Of falsifica-
tion of names after the deaths of individuals, to bring them into
line with the generational pattern. This is linked to the differ-
ence between the formal, full, name and its informal version with
the T. A prefix.

Another explanation, of course, and one that was Obviously
Operative to some degree, is marginal literacy and "misspellings."
Thus there may be two versions of certain names, but versions the
pronunciation of which are identical. Thus there are two versions
of Ch'iu Yu-sheng (Appendix B, p. 365). A reading pronuncia-
tion Of "to have" in Taiwanese, 13, is identical to the pronuncia-

tion of the other character used in the other version, T. 1g,
"friend."

A third explanation is that some of the members of the
lineage are not actually related at all, and that there has been
large-scale falsification of the genealogy to attempt to reflect or
affect contemporaneous socio-political realities.

Since I have not been able to interview the people involved,
or look further into historical sources, I cannot conclusively solve
this problem. On the basis of the evidence at hand, though, I tend
to support the first and second explanations in combination; I have
no reason to believe that any "creative" use has been made of the

genealogy in the sense of altering it to incorporate unrelated

groups or individuals.

255

The Ch'iu lineage differs from the Liao lineage in that
they are considerably less well-off than the Liaos are. Though the
constituent families are seldom "poor," they are what might be
called "lower-middle class" inaasocial-class system. Where the
Liaos are landowners, the Ch'ius are tenants; where the Liaos are
businessmen, the Ch'ius are laborers.

Out of 116 marriages for which there is enough information
to identify the type of marriage, twenty-eight were marriages of
women out to another family or surname group in the regular manner,
and forty were marriages in the regular manner. Twelve were uxori-
local marriages, eight in and four out of the surname group.
Thirty-six were T. simpua_marriages. Thus sixty-eight of the mar-
riages were "regular," and forty-eight "irregular."

Once again, however, as with the Liao case, it is instruc-
tive to analyze the adOptions as well. Eighteen children were
adopted irL (When it is specified that the status of a girl is
T. simpua, this is counted in the category "T. simpua.ad0ptions,")
Sixteen Of them were female, and two male. Seventeen children were
adOpted out, fifteen females and two males. Forty-five girls were
adopted in as T. simpua, Since thirty-six Of them were then actually
married to a son of the family, this leaves nine who were converted
into "adopted daughters" and married out, or were simply sent back
to their natal homes or adopted out.

Thus the sixty-eight "regular" transactions involving the
exchange of rights over persons must be weighed against ninety-two

"irregular" transactions (marriages and adoptions).

256

Once again, as in the Liao case, I cannot go into a detailed
account of marriages and adoptions for the entire lineage, but I
can isolate out a number of interesting cases.

Page 366 Of Appendix B, for example, illustrates how one
can create a family when one lacks sons. Yu A-"ban" (the last
character in the name is a Taiwanese dialect character meaning "last
child") was married to Hsfieh A-chi, but lacked sons or daughters.
When he and his family moved into T'ien-sung-p'i in 1951, he already
had adopted in a daughter, Yu YU-lan, who had originally been sur-
named Ch'en, and had then married in an uxorilocal husband, Chien
Chao-tsan, for her. Two years before they moved into the village,
he had also adopted in another daughter, this time a T'ien-sung-p'i
girl surnamed Ch'en (Ch'en YUeh-O). In T-ien-sung-p'i, in 1953,
he then married in another uxorilocal husband for the second adopted
daughter,(HuACh'in-chang, from the village. While the family was in
the village, the first adOpted daughter died, and in 1953 her
widowed uxorilocal husband married a woman who had been a married
T} simpua_in the Ch'iu lineage, but had divorced her husband the
year before. They then remained in the village when Yu A-"ban," his
wife, the second adOpted daughter, and her uxorilocal husband all
left for LO-tung in 1953.

There is another example of the same sort Of thing in which
one Of the members of the Ch'iu lineage married out uxorilocally.
Lin A-fu, who lived from 1898 to 1965, was married to Liu A-shou,
born in 1903 and still alive. Having no sons, they adOpted in two

daughters. In 1951 Ch'iu Mu-ch'eng married in uxorilocally, and is

257

still there. His wife was originally surnamed Cheng, and changed
that to Lin when she was adopted in. Then when Ch'iu married in,

he prefixed Lin to his own surname. Six children have been born to
them, four surviving.

The other adopted daughter, Lin Feng A-yfi, had an illegiti-
mate daughter in 1959, named Lin Mei-hui. Since it was an illegiti-
mate birth, and the father Officially unknown, the child took the
surname Lin. Because her father was an uxorilocal husband, two of
the other adOpted daughter's children also took the Lin surname.

It is interesting that whether surnamed Ch'iu or Lin, and whether
male or female, the children were given the same generation names--
Lien- for the males, and Mei- for the females.

In the Ch'iu lineage also occurs one of the few instances
of concubinage that 12 found in the history of T'ien-sung-p'i.

In 1918 (the Official date of change of registration) Ch'iu Yu-sheng,
the ancestor of what I have labeled branches ten, eleven, and

twelve, moved into the village from T'aO-yUan with two of his three
sons (Ch'iu Chieh-hsing and Ch'iu Shih-shui). The first son brought
his second wife, and the third son his concubine. She had been
married into the family in 1902, when she was twenty-four. This

was in the same month as the marriage of Ch'iu Shih-shui's primary
wife. The latter, however, moved back to her natal family in 1905.
There is no record of divorce.

The primary wife had no children, so Ch'iu Shih-shui's line
has been perpetuated by the secondary wife, or concubine's, sons.

They produced six children--four daughters and two sons. The oldest

258

daughter (as the children of a secondary wife, they have no birth
order in the earliest records, just the word M. shu-tzu, "child of a
concubine") was married out, and the remaining three daughters were
all adopted out within the first few years of birth. Two T. simpuas,
however, were then adopted in for the two brothers. The T. simpua
intended for the older son did not marry him, however, and she was
married out after her status was changed to that of "adopted
daughter."

The proportions Of nonstandard marriage forms and adoptions
become quite high in some branches of the Ch'iu family. Ch'iu A-tung,
the ancestor of what I have numbered branch twenty-three, had a
T. simpua_wife, and they produced four sons. His oldest son was also
married to a T. simpua, and they had two sons, each of whom was
married to a T. simpua_wife. The Oldest of these sons had one son
who survived, and he_ was married to a T. simpua,

Ch'iu A-tung's second son did not marry at all, so he first
adopted in a daughter, who later married out, and then adOpted in a
T. simpua_and a M. kuo-fang son from his brother in 1928, and married
the two to each other in 1936.

The third son of Ch'iu A-tung married in the ordinary,
"standard" fashion, and had four sons. One of them has disappeared
from view, but the other three all married T. simpua_wives (one Of
the sons is the one adopted to A-tung's second son, above); the

Oldest of these then married hj§_first son to a T. simpua.

259

The fourth son of Ch'iu A-tung married a T. simpug_as well,
and married both of his sons to T. simpuas. That is twelve T. simpua_
marriages, and there were only fifteen "regular" marriages.

The chart of the following page is a summary of marriage
and adOption forms in the seven cases. The figures generally bear
out my intuitive feeling about the relative proportions of marriage
and adoption forms, but they do force some modifications in that
original guess. If we total regular marriages in and out, uxori-
local marriages in and out, and T. simpua_marriages, we get 249
“marriages" represented in the case materials. Of these 249 mar-
riages, thirty, or 12 percent, are uxorilocal.

If we then take the remaining 219 marriages (having removed
the thirty uxorilocal marriages), we find that sixty-four, or almost
25 percent, were T. simpua_marriages.. This figure of 219, however,
includes sixty marriages out, to other families and surname groups;
"T. simpua_marriages" here,of course,include only marriages into the
group (or marriages within the group; in either phrasing, the couple
remains in the surname group or family after marriage). A more
proper comparison, then, would be with regular marriages in. If
we compute this, we find that of 159 marriages of women into the
groups in the cases, 40 percent were T. simpug_marriages.

A puzzling aspect of these figures is that there are a total
Of ninety-five marriages in, and sixty marriages out. One would
expect, of course, a much more balanced set of figures. Here we
are talking about group recruitment, and at first I thought that

these figures were thrown out of balance by the large and expanding

260

 

 

o_ m o a P o o o meawpaaea.mmmwnmmm
om op N NF _ o N a use meewpaoaa emcee
oe m_ o N_ a o _ N a? meawaaoaa cacao
NN N, m ma m _ m P e? anVaaaaa.mmman
an m_ a mm m o m P memaeeeae.mmmamw
m o P a _ o N o N So .83th Zoo—Cox:
_N m o m a N _ _ E 882:2 82:2:
om NN m NN N o F m use meaaweaas ea_=mem
mm Nm m oe N _ N a =_ mamaeaaae ea_=mem
_aaae ee_auH> 3w.euu> =W.;UHHH> azuHHH aao_a H ea.a NH ee>n>H amaze

 

ammo xn mega» cowpaon< ucm mmmwggmz mo gmnsaz

cowpaou< new mmmwggmz

261

lineage, which was "using up" women. This imbalance, however, is
not reversed in any of the cases studied; large or small, each group
(with one exception, Chi'iu Case IV), took in more women than it
sent out.

The figures Of ninety-five marriages in and sixty marriages
out.refer only to regular marriages, and it must be suggested
that this imbalance is being made up for elsewhere in the system.

To get an overall view, then, we must include T. simpug_marriages
as well. This, then, implies that we must widen our definition, so
as to include adoptions generally, since adopted daughters are
usually potentially, at least, T. simpua_adopted daughters, and
hence "potential spouses."

These computations do not, however, clear up the confusion;
if we total all individuals entering the groups studied, whether
through "marriage" or "adOption," we get a figure of 234. Those
leaving the group in marriage or adoption, 114. The figures are
even more imbalanced, and it is clear that a major source Of the
imbalance is adoption; 118 individuals were adopted in, the majority
of them explicitly intended as simpuas (seventy-eight), but only
thirty-six adopted out. Thus there is a net gain of eighty-two
individuals, a total which goes a long way toward explaining the
imbalance of figures above. What remains to be explained, of course,
is the imbalance in adOptions.

The explanation for this imbalance, I think, lies in the
historical and environmental situation. An imbalance in the recruit-

ment and 1055 figures is puzzling only under the assumption of a

262

static population; only then must one look somewhere else in the
local system for "compensating factors." In a dynamic and expand-
ing population, however, there is no need to (and indeed, it is a
mistake to) look for homeostatic mechanisms; there is no homeo-
stasis.

The village Of T'ien-sung-p'i is a young one, even by
Taiwanese standards (when compared to areas of the mainland, for
example). All of its development must have taken place within the
last 150 years, and most Of it within the present century. The
information contained in the case studies effectively brackets the
period of greatest expansion Of the area Of arable land and popu-
lation. If homeostatic mechanisms are to be looked for, it is in
terms of a much larger system than the local one in the village
itself.

What seems to have been the case is that the influx of per-
sons into the village was made possible by the local development of
land, and the improvement in security, which took place beginning
not long after the Japanese took over Taiwan. The influx of persons,
as sources of labor, of course, in turn made that development pos-
sible. Here my unit Of reference has shifted from the groups appear-
ing in the case studies to the village population as a whole. Some
of the adoptions and marriages in the above cases,of course,took
place within the village as a unit, but the majority of them were
extra-village transactions. The preponderance of movement in over
movement out in the case materials corresponded to an actual pre-

ponderance of movement of peOple into the village.

263

People from areas of greater population density, and greater
pressure on the land, moved into the expanding and developing fron-
tier areas. This is exactly what I was told by the Liaos about the
motivations of their ancestors Liao Cheng-Ch'ing and Liao Cheng-fa;
that they had come to the I-lan area looking for land because their
home area in Hsinchu had no land available. Those areas gave up
the people that T'ien-sung-p'i gained.

A larger proportion of the people so gained were recruited
through adOption of females. Given the cultural values which gave
females much less importance than males, and a system of customary
law which Offered many alternative forms of transaction of rights
over persons, females could be much more flexibly "used" than males
in this situation. A female was cheaper to raise than was a male,
since her intrinsic value was seen as lower, and she could very early
begin to contribute labor to the household. Even today, in my
experience, the females of families, whether adopted or not, begin
to work much earlier for the family than do male children. Girls
of nine or ten can and do contribute much valuable labor in looking
after younger siblings, preparing food, cleaning up the house, and
so on, usually while their Older brothers are out playing.

Thus one can get eight or ten years of increasingly produc-
tive labor out of a girl before she is of an age to marry. If she
is own's own daughter, then she will marry out, and be lost to the
family (unless, of course, an uxorilocal husband is married in for
her; as is shown in the section on uxorilocal marriage, however, one

must "sweeten the pot" to get such a man). One will, of course,

264

get the brideprice in return, but the alternatives available in the
"disposal" of a daughter are not as many as in the case of an
adopted daughter. One can get just as much labor out of her, and
more; it is commonly felt that since such a girl is not one's own,
it is easier to work her hard, and spend less on her. When she is
of marriageable age, one is not restricted to arranging a marriage
out or uxorilocal marriage in, but might also marry her to one's son
(or adopted son). An added bonus, which has been mentioned by
Arthur Wolf, is that she has been socialized in the family, and so
can be expected to be less disruptive and more docile than a "stranger"
who is married in (Gallin 1966: 163-66). This phrasing is in terms of
socialization, but one can also put it in terms of customary law; such
an adapted daughter has been more fully incorporated into the family.
There has been a more total transferral of rights, and thus the
adopting family has greater powers over her than it would over a
woman who had married in. Uku~natal family still has some rights over
her, and she can return to them in extreme situations.)

A Taiwanese folk-song puts it well:

Bring a basket of rice to feed the chicken;

If you raise a chicken,it:can crow the time.

If you raise a dog, it can bark at ghosts.

If you raise a pig, you can butcher it.

If you raise a sister'sson, it's all expense, and no return.

If yggmrare a daughter-in-law, she'll support the husband's

'If you raisea adaughter, she becomes someone else's.
(Li 1936: 240).

We can see from the figures that in fact the alternative Of
marrying such a girl to one's son was the most popular; of 118 girls

adopted in, sixty-four, or 54 percent, did marry this way. The

265

remainder were dealt with in a variety Of ways, but figures on the
relative numbers are simply unavailable. Some were sold or pawned
as prostitutes, some remained in the adOptive households as ser-

vants, or T. cha-bo'-kan, some were re-adOpted out, some were married

 

out, and some were simply returned to their natal families.

The only serious drawback to such T. simpug_marriage was
that it restricted affinal ties. When the adopting family acquired
relatively complete rights over the girl, it at the same time, and
to the same degree, reduced its future powers to activate affinal
ties for social or economic reasons. Though the personal ties between
such a girl and her natal family will not have completely dis-
appeared, they will have weakened considerably after ten or twenty-
years' residence in the adOptive family, and as the natal family
has few rights over the girl after her adoption, they also have cor-
respondingly few duties toward her, and by extension, the family to

which she belongs.

CHAPTER VI

ANALYSIS AND CONCLUSIONS

A basic argument of this dissertation is that "marriage" and
"adoption" are inapprOpriate categories for the analysis of social'
transactions in traditional Chinese society, and that attention
should instead be paid to a spectrum of social forms, a spectrum
which contains a wide variety of types of transactions. What is
common to all of these types Of transactions is that they are forms
of recruitment; recruitment to family and household groups. Any
further analysis, then, requires that we look into the nature and
role Of these family and household groups.

I have found it useful in this analysis to view the family t»
ai.§.§lflésgfninformal interest group, one which makes use of various
kinship, ritual, and other formsof symbolic patterns and activities
torpgotector develop power (in the broadest sense) for its mem:
Q§£§a This view requires that the general political and economic

context must be taken into account. To quote Abner Cohen, "The

anthrOpologist must deliberately formulate his problems in such a

 

way as to make reference to the state a necessary part of his
analysis" (1974: 128-29).

When we come to see the family as a corporate interest group,
we must, particularly for an agrarian society like that of traditional
China, consider the features of the family farm that are said to

266

267

characterize peasant societies. In this context I have found the
theories of A. V. Chayanov to be most fruitful. One Of Chayanov's
central ideas, seen by some as thg_central concept for analyzing
family economics and the key to his arguments (Thorner 1966: xv),

was that of the labor-consumer balance between the satisfaction of
family needs and the drudgery (or irksomeness) Of labor. His argu-
ment was that capitalist economic theory could not be indiscriminately
applied to any society whatever, that it really worked only in

capitalist economic systems. Furthermore,

 

The economic theory of modern capitalist society is a
complicated system of economic categories inseparably connected
with one another--price, capital, wages, interest, rent, which
determine one another and are functionally interdependent. If
one brick drops out of this system, the whole building collapses.
In the absence Of any one of these economic categories, all the
others lose their specific character and conceptual content,
and cannot even be defined quantitatively (Chayanov 1966: 4).

This seems obvious if one takes, for instance, the category
of price. Chayanov argues, however, that it is no less true of the
category Of wages, the one Of most interest to us here because it
bears on the nature of the family farm labor force.

Even if out Of all the possible economic systems lacking
this category we choose one in which exchange and credit (and
thus the categories Of price and capital) are present (for
example, the system of peasant and artisan family labor units
held together economically by monetary and exchange processes),
we shall still find that the structure of such an economy lies
outside the conceptual systems of an economics adapted to cap-
italist society.

On the family labor farm, the family, equipped with means
of production, uses its labor power to cultivate the soil and
receives as the result of a year's work a certain amount of
goods. A single glance at the inner structure of the family
labor unit is enough to realize that it is impossible without the
category of wages to impose on this structure net profit, rent,
and interest on capital as real economic categories in the capi-
talist meaning of the word.

268

Indeed, the peasant or artisan running his own business
without paid labor receives as the result of a year's work an
amount of produce which, after being exchanged on the market,
forms the gross product of his economic unit. From this gross
product we must deduct a sum for material expenditure required
during the course Of the year; we are then left with the increase
in value of material goods which the family has acquired by its
work during the year, or, to put it differently, their labor
product. This family labor product is the only possible category
of income for a peasant or artisan labor family unit, for there
is no way of decomposing it analytically or objectively. Since
there is no social phenomenon of wages, the social phenomenon of
net profit is also absent. Thus, it is impossible to apply the
capitalist profit calculation. . . .

The amount of the labor product is mainly determined by the ;2
size and composition of the working familya.the number of its 6
members capable of work, then by the productivity of the labor
unit, and--this is especially important--by the degree of labor
effort-5the degree of self- exploitation through which the work-
ing members effect a certain quantity of labor units in the
course of the year.

Thorough empirical studies of the peasant farms in Russia
and other countries have enabled us to substantiate the follow-
ing thesis: the degree of self-exploitation is determined by a
peculiar equilibrium between family demand satisfaction and the
drudgery of labor itself (Chayanov 1966: 5-6).

This latter point is elaborated as follows: each increment
of the family labor product must be seen from two standpoints:
first, from the importance it has with respect to the satisfaction
of family needs, and second from the viewpoint of the drudgery that
isainvolved. With each additional increment, the importance of that
movement for the satisfaction of the needs of the family decreases.
The drudgery, however, will increase continuously and at an accererat-
ing rate. There must be a point, then, where these two curves,
falling importance and increasing drudgery, intersect. This point Of
equilibrium is Chayanov's labor-consumer balance. Before it is
reached, as long, that is, as the unsatisfied needs are perceived
as more important than the irksomeness of the work involved, there is

motivation for more and more labor output by the family members.

269

When the equilibrium point is reached, however, the anticipated
benefit of the increments no longer outweighs that irksomeness, and
further labor is pointless.

Chayanov holds that his work, and that of a number Of
others, has shown that

. . . This moment Of equilibrium is very changeable. It
is reached as follows: on the one hand, through the actual
specific conditions Of the unit's production, its market situa-
tion, and through the unit's location in relation to markets
(these determine the degree of drudgerY); on the other hand, by
family size and composition and the urgency of its demands,
which determine the consumption evaluation. Thus, for example,
each increase in labor productivity results in gain of the same
quantity of products with less labor. This allows the economic
unit to increase its output and to satisfy family demands in full.
On the other hand, the significance Of each ruble of gross
income for consumption is increased in a household burdened with
members incapable of work. This makes for increased self-
exploitation Of family labor power, so that the family's standard
of living, threatened by increased demand, can be kept up in some
way.

Starting with the nature of the basic consideration des-
cribed above, the family labor farm has to make use of the market
situation and natural conditions in a way that enables it to
provide an internal equilibrium for the family, together with
the highest possible standard of well-being. This is achieved by
introducing into the farm's organizational plan such labor invest-
ment as promises the highest possible labor payment per labor
unit (1966: 6-7).

Both Basile Kerblay (1971: 159) and Daniel Thorner (1966: xxi)
have pointed out that Chayanov's theory

. . worked better for thinly populated countries than for
densely pOpulated ones. It also worked better in countries where
the agrarian structure had been shaken up (as in Russia after
the emancipation of the serfs in 1861) than in countries with a
more rigid agrarian structure. Where the peasants could not
readily buy or take in more land, his theory would have to be
seriously modified.

Since Chayanov did not work out these modifications, he did
not elaborate a full-blown theory of peasant family farming for
any country other than Russia. Nonetheless, he indicated that
he thought one single universal theory of the peasant family
farm at the micro level could be devised. In his view, the

270
Russian case, which he developed so fully, was only an illus-
tration of this larger theory (Thorner 1966: xxi).

As pointed out in a footnote to this passage, however,
Chayanov himself recognized this fact and made a few suggestions as
to the proper direction of analysis. He pointed out that "the peasant
farm tends, by means of renting land, to bring the area it is
exploiting agriculturally into Optimal relationship with family size"
(1966: 111). He then acknowledges "that in cases where the land
regime is not very flexible the relationship between land and family
is regulated by a change in the amount of labor hired or hired out"
(1966: 112).

He further argues that a peasant family without sufficient
land and means of production can put the surplus labor force that is
thereby created into another form of economic activity, such as
crafts and trading. The sc0pe of these activities, however, may be
limited, and income low.

In this case, it is sometimes advantageous for the peasant
farm to violate the optimal combination of production elements
for its activity and to force its labor intensity far beyond
the Optimal limits. Inevitably losing on unit labor payment,
it nevertheless considerably expands the gross income of its
agricultural undertaking and reaches a basic equilibrium between
the drudgery of labor and consumption--within the limits of
agricultural activity, of course--at a level Of well-being lower
than would occur given a farm Optimal in size and proportions.

This forcing up Of labor intensity, buying increased annual
agricultural income at the price of reducing labor unit payment,
is achieved either by an intensification of work methods or by
using more labor-intensive crOps and jobs (Chayanov 1966: 113).

Chayanov's mode of analysis is, furthermore, useful because it can
help to re-orient the anthropologist's examination of peasant families

in rural, agrarian societies.

271

Peasant societies are sometimes described, and indeed defined,
as being composed of "relatively" autonomous units, such as family
households and villages. That is, households, and villages, are often
thought of as being "reasonably" self-sufficient. Some authors tend
to take this literally, and assume that such peasant households are
in reality quite autonomous, little involved with affairs and events
of larger order than their day-tO-day and season-tO-season agricul-
tural pursuits. Others seem to take it as a sort of "ideal type," a
type they acknowledge didn't "really" exist, but one which nonethe-
less furnishes the model for analysis. Even when it is made explicit
that peasants usually exist in larger-scale social, economic, and
political units such as states, there is still a tendency to assume
that their contacts with representatives Of those units are limited
and unidirectional, and that they are contacts in which the peasant
is passive.

Chayanov's model of peasant economy does not restrict us to
such a view. He recognizes two varieties of family economy. One was
the "natural economy," which resembles the picture of the autonomous,.
self-sufficient units described above. The other he called the
"commodity economy." It differed from the first mainly in that it
was influenced more by commercialization. It differed, for example,
in that it possessed the economic categories of commodity price,
interest on capital in the form of rentier's income, differential
rent, and land price. The presence of these economic categories is
sufficient to distinguish commodity economy from natural economy, but

they are both still systems of family economy, and in Chayanov's

272

argument are to be distinguished from other types of economy. The
important factor is not how "isolated" or self-sufficient families
are, but what sort Of labor force they employ--one made up of the
family members, variously defined, or of a hired labor force.

These considerations bring us close to the conditions that
obtained in the traditional Chinese economy, and even closer to
those of Ch'ing Taiwan. This "forcing up of labor intensity" should
also ring a bell for those familiar with Clifford Geertz's work on
sawah, or wet-rice cultivation (Geertz 1968).

I want to turn now to a consideration of the implications
of Chayanov's approach for analyzing Chinese peasant family structure
in general, and adoption and marriage in particular.

In the Chinese case, I think that there is no question but
that rural economy has long been a "commodity economy," a subdivision
of Chayanov's family economy. This is true, Of course, in varying
degrees for different parts of China and for different periods of
Chinese history. It seems to be particularly so for southeastern
China, the area of greatest interest to us, and particularly so for
the Ming Dynasty and later. E. S. Rawski has shown in some detail
how in parts of sixteenth-century Fukien and eighteenth-century Hunan
provinces

. . . and in other areas, commercial develOpments had an

impact which extended beyond urban artisan and mercantile worlds,
into the rural household economy. This effect is illustrated in
Fukien, where the expansion Of foreign trade in the sixteenth
century brought identifiable changes to both the urban and rural
economy. . . . Because the historical materials on Chinese com-
merce are scanty and provide little basis for evaluating changes

in either the quantity or quality of trade, study of Chinese
agriculture can provide a useful measuring stick against which

273

the largely descriptive evidence of markets can be judged. In
comparison with commerce, materials on farming are rich, and

it is possible to examine develOpments in agriculture in some
detail. One paradoxical conclusion of this study is that sig-
nificant changes in commerce can be best studied indirectly, by
tracing their effect on agriculture. The interaction of trad-
ing and agricultural activity provides a focus of interest. . .
(Rawski 1972: 2).

This close, inextricable involvement of commerce in agricul-
ture must also have characterized the history of the Chinese settlement Of
Taiwan. In fact, such involvement must have been in many ways even
further-reaching than on the mainland. The very beginnings of effec-
tive Chinese settlement of the island were tied up with the course
of international trade within East Asia and the international trade
between East Asia and EurOpe. The very early (beginning with the
Dutch occupation) and heavy involvement with the cultivation of sugar-
cane, as a cash-crop for export is evidence of the effect of this on
the development of agriculture in Taiwan. Another index of this
effect is the diversity of the types of rice grown in Taiwan during

Ch'ing rule. Lien Heng, in his Taiwan T'ung-shih, lists forty

 

two different strains of rice that were available (fifty-six strains
if subvarieties are counted) in Taiwan, a greater variety than was
available on the mainland (according to Chou 1957: 25; Rawski, how-
ever, says that "Ming gazeteers for Fukien list more than 150 dif-
ferent seeds, more than two thirds of which are to be found in only
one part of the province" [1972: 40]). The range of coinage in use in
Taiwan toward the end of Ch'ing rule also reflects the extent of
monetization of the economy and involvement with world trade. The
foreign silver coinage included Mexican, Hong Kong, Spanish, American

trade, Annamese trade, Japanese, Japanese trade, Peruvian, Indian,

274

Portuguese, and Siamese dollars (though the last four were compara-
tively rare; Chou 1957: 112).

The issue,then,is not the extent Of isolation of individual
households, their self-sufficiency, or market orientation. The

Shih Ch'a Lu records that in Taiwan "the inhabitants truly know how to

 

pursue profit. They go Off in boats or shouldering their goods,
trading without stop. This is also why fresh grain is hoarded."

The Komalan T'ingrchih, which quotes this passage, says that in
(Komalan, now I-lan, it was the same, and that less well-Off families
sold their rice and other crops while they were still on the stalk,
accepting a lower price to do so. Even so, it is claimed, they
managed to end up with a surplus (5A; 364). The same gazeteer
records that double crOpping was prevalent, and that the land was
quite rich, so that only when twg_good harvests were achieved in a
year was it said to be a good year. There was a surplus, then, over
and above what was needed in Komalan, and this surplus was sent to
the granaries at Chen-hai and Cha-p'u in Chekinag, Shanghai, Chang-
chou, and Ch'Uan-chou (Komalan T'ing-chih: 363).

The crucial issue for the applicability of Chayanov's model
is the extent of use Of hired agricultural labor. On this point,
Yhi-min Ho says that for the period 1901-1960 in Taiwan, "farm hired
labor accounted only for a small share in the total agricultural
employment, and its share remained extremely stable in the range from
3.3 to 4.3 percent of the total agricultural employment" (1966: 42).
He arrives as these figures by comparing the agricultural population

as recorded in the household registration records with that in the

275

census enumerations. The numbers of agriculturally employed consis-
tently differ in these two sources, and HO argues that this is
because the agricultural hired laborers were not included in the
category Of agricultural pOpulation as used in the household records,
which included only three subgroupings--owner-cultivators, part-
owners, and tenants. If the household figures are compared with the
census figures for the census years 1905, 1915, and 1920, the ratios
of farm hired to total population amount to 4.27, 3.34, and 4.67
percent, respectively. Official data on hired agricultural labor
are then available for the period 1947 to 1960, and the ratio for
1947 is found to be 3.40 percent. This gives strong support to his
agrument about the overall numbers of hired laborers throughout the
Japanese period. If we accept these figures, it seems to be that
there is no reason to think that conditions in this respect were
significantly different during the Ch'ing period.

Some comparative figures from other countries will show that
this is quite a low percentage of hired farm workers; the figures
were 35 percent for England in 1911, 36 percent for Germany in 1920,
25 percent for the United States in 1920, and 32 percent for France
in 1911 (Sun 1938: 69).

It seems, then, that in China hired farm labor was never as
important as: it was and is in other countries, that this generali-
zation is even more true Of the wet-rice areas of the Yangtse Valley
and further south (Sun 1938: 69), and that Taiwan has until recently '
been characterized by quite low rates of hired farm labor use. This

means that Taiwanese and most other Chinese farms on the mainland

276

were peasant family farms in Chayanov's sense, that they practiced a
"domestic mode of production," and thus that his concept of "self-
exploitation" is applicable.

Chayanov's view is strikingly echoed by Feng HO-fa in his
massive compilation of materials on Chinese peasant economy published
in 1933:

The Chinese family system causes a situation in which the

relatives of many of the peasants help them in the management
of the fields, and receive in return food, lodging, and a cer-
tain amount of cash. If one has more relatives, then one has
more help. Thus it is not easy to compute the costs of culti-
vation. The labor of the women and children of the family can
also reduce expenditure on hired labor. The amount paid for
such help is then used within the family and kin group, so
just who receives the benefit is hard to determine. But if we
want to know, for comparative purposes, the amount expended
for the management Of each plot, we must know the value of the
unpaid labor, and this must be added to those expenses. Thus
this kind of help, computed at the current labor wage, and
including food and lodging, is reckoned at ten yuan per year
for a male, two thirds of that for a female, and one third of
that for a child (Feng 1933: 846-47).

These latter figures are clearly very arbitrary and rule-Of-
thumb, and illustrate the difficulty that Chayanov has discussed of
using capitalist economic categories in what he argues is an essen-
tially noncapitalist economic system.

Thus, in the traditional system, there were large numbers
of "interest groups,"'h1Abner Cohen's use of the phrase, that were
involved as units in agriculture. It is in this context, among
others, that we must view such transactions as adoption and marriage,
since they function in part as methods of group recruitment. "Self-
exploitation" means not only that the head of the family worked, and
often worked harder than if he had been hired by someone else, but

also that the other members of the family worked as well.

277

Ho points out the extent of family labor on farms in Taiwan:

The rate of participation of the agricultural labor force is
calculated for the six census years, 1905, 1915, 1920, 1930,
1940, and 1956 based on the census's agricultural employment
data. TO define labor force in the conventional Western sense,
i.e., pOpulation over 14 years of age and under 65, is not
appropriate in Taiwan. The rate of participation so obtained is
persistently over 100 percent. The agricultural labor force
defined in the Western conventional manner underestimates the
actual size of the labor force. It is not uncommon in rural
Taiwan, as well as in other underdeveloped countries, for chil-
dren under 14 and adults over 65 to be actively engaged in
agricultural production. Therefore, the portion of agricultural
population with age over 12 is defined here as agricultural labor
force. Admittedly, this definition tends to overestimate by
including peOple over 65 in the labor force, but, in the mean-
time, it excludes those child laborers under the age of 12, which
makes the size Of the labor force estimated here smaller than the
actual size of the labor force in agriculture. The agricultural
labor force so defined and the series so obtained can be regarded
as reasonably satisfactory. . . .

Women laborers also occupy a large share in the total agri-
cultural labor force and contribute importantly to agricultural
production; they should not be excluded from the category of
agricultural labor force (HO 1966: 40-41).

While I am not arguing that the need for agricultural labor
is the single "cause" of the various forms of adOption and marriage
discussed here, the character of the agricultural economy is crucial
for the understanding of those forms. Lin Heng-tao, in an article
on the Ch'ing development of Taiwan, discusses some of the customs
and usages in contemporary Taiwan that can be traced back to the
pioneer era. The prohibition during the Yung-cheng era (1723-1736)
upon the families of immigrants accompanying them to Taiwan, he
argues, led to the custom in Taiwan of adOption of children of dif-
ferent surnames. These adoptions, which ran counter to the Ch'ing
Code (which restricted adOptions for the most part to same-surname
adoptions), were, he says, for the purpose Of ensuring against

poverty and for providing care in Old age. The practice was then

278

continued for a long enough time that it became customary, and even
families with many of their own children would adopt (Lin 1968: 59).

I would not agree that such adoptions were continued through
inertia, as it were, (n: with the implication that they were a
peculiar development of conditions in Taiwan. There is much evidence,
as I have shown, that such adoptions and other "irregular" forms were
widespread wherever there were Chinese populations. Lin has, how-
ever, put his finger on a very important aspect of the practice.
PeOple without them could make use of a variety of transactional forms
to create families or "pseudo-families." Moreover, it was neces-
sary to do so in many cases for survival because there were few
opportunities or places for single individuals in the traditional
politico-economic structure.

The phenomenon of self-exploitation noted by Chayanov meant
that the peasant family farm was agriculturally quite productive.
It is no accident that the family, and the cultural values that help
to support it, were endorsed and promulgated by the Imperial state.
Agricultural productivity meant increased revenues for the state,
since the majority of such revenues came from land taxes and other
imposts connected with agriculture and the peasantry. At the same
time that productivity was increased (not everywhere and in all situ-
ations, but generally) because of self-exploitation internal to the
family, socio~political stability was increased, because the state
was in effect operating a system Of indirect "colonial" rule,
co-Opting and delegating authority to the head of the family. The

head then found himself working for his own personal good (and, given

279

the sociO-political situation, for the good of the other members of
{the family) by serving the good of the state. The legal, and to a
great degree the customary, powers Of the head vis-a-vis the other
family members were quite far-reaching. At the same time, however,
he was also hostage to the state for the good behavior Of the other
members. Thus, to the degree that he pursued his own interests and
the family's interests, he was also benefiting the state by ensur-
ing revenues, through productivity, and ensuring social order by
controlling, with his great powers, the other members of the family.

We have noted that in Chayanov's model of self-exploitation,
a family faced with insufficient land and insufficient availability
Of alternatives to agricultural labor may force its labor intensity
far beyond "Optimal" limits. By so doing, it loses on unit labor
payment, but increases the gross income of its agricultural activi-
ties. That is, the overall income for the family is increased, but
at the cost of greatly increased drudgery of labor, an increase that
might not be borne by a hired laborer. This forcing-up of labor
intensity was done in the Chinese case, as Chayanov suggests for the
general case, "either by an intensification Of work methods or by
using more labor-intensive crOps and jobs" (1966: 113).

It is in the context of the insufficient availability of
alternatives to agricultural labor mentioned above that another argu-
ment becomes interesting: that the elite in the traditional Chinese
bipolar class structure monOpolized, or perhaps a better word is

co-Opted, the economic alternatives to agricultural work.

280

It is clear . . . that successful mobility is measured by
power, and that power in turn is conditional upon holding politi-
cal status, either as a ranking or retired bureaucrat in govern-
ment or as a titled candidate. Power is given by association
with government, and wealth follows power. Inlaid chopsticks
and porcelain cups come after the fact. Wealth has neither
prestige nor security withOut power. Upward mobility is achieved
only when the motivated social climber comes to stand among the
rulers or among those eligible to rule. Arrival is marked by
holding one or more examination degrees, won or purchased from
the state or conferred by the state. "Only the scholar stands
high."

Elites, therefore, are first of all power persons. Power,
in the name of empire, enables them to draw income from land,
service, business, or Office; to find security of property; and
to enjoy prestige and well-being Of person. Rank, not wealth,
is the social measure of elite status. Rank is formalized in
terms of academic degrees and Official titles, for which paper
diplomas and certificates (displayed in the ancestral temple)
and different costumes (worn on public occasions) are the visible
evidence. . . .

Strictly speaking, the elite are not merely one socio-
economic class within Chinese civilization. They are, in their
own view, the very carriers of civilization; the peasants are
just so many energy slaves. . . .

Gentry families necessarily live off the peasant landscape
in their own localities. But the power to do so is gained on
a translocal basis. Each family centers in a tissue of influen-
tial connections. Reaching out from home base, the approach to
power is personal. . . (Stover 1974: 118-19).

Thus wealth could be safeguarded from the exactions of offi-
cials only through political connections and power, and from theft,
loss, or destruction only through putting it in land, a form of
investment with quite a low rate of return, but vfith high security.
The goal of government was to ensure domestic tranquility and pro-
tect sources of state revenue and the sources of state and elite
power. In such a situation family and kinship ties could come to be
just as important to a family group as the provision of laborers for

the agricultural enterprise.

281

Abner Cohen discusses how ethnicity is "used" by Hausa com-

munities in Yoruba towns in connection with long-distance trade

between the savannah and the forest belt:

Under the pre-industrial conditions prevailing in Nigeria,
long-distance trade is attended by a number of technical prob-
lems which can be effectively overcome when men from one ethnic
group, speaking the same language and observing the same code
Of conduct, control all Or most Of the stages of the trade in
specific commodities. Such an ethnic control, or monopoly, can
usually be achieved only in the course of continual and bitter
rivalry with competitors from other tribes. In the process,
the monopolising ethnic community is forced to organize for
political action in order to deal effectively with increasing
external pressure, to co-ordinate the co-operation of its mem-
bers in the common cause, and to mobilise the support of commu-
nities from the same ethnic stock in neighboring towns (1974:
95 .

This is a very interesting passage for a number of reasons.

First of all, it brings to mind the trading firms, or "guilds" of
trading firms, that monopolized the trade between Taiwan and the

China mainland during the Ch'ing period, guilds made up of persons

Of Ch'Uan-chou origin. When one thinks Of the endemic inter-ethnic
rivalry and fighting that took place in Taiwan at the time, and of _
Lin Heng-tao's argument that in Taiwan, and especially in the north;

of Taiwan, villages composed of persons from the same areas of the

mainland took the structural place of the single-surname, clan-

organized villages of Fukien and Kuangtung, Cohen's discussion is

quite apt.

Second, and more to the point, the passage brings to mind
the situation with respect to banking in traditional China, particu-
larly during the last half of the nineteenth century. King points

out that indigenous Chinese banking (the so-called "native banks")

1

was to a certain degree monopolized by Shan-hsi "ethnic" groups, and '

282

that the managers Of the branch Offices were adopted into the families
of the head managers (King 1965: 94). Under the pre-industrial con-
ditions prevailing in China, there was little in the way of politico-
legal mechanisms for the support and protection of banking and other
financial affairs, so that such activities had to be carried out in

a context of other personal ties, in a context of "multiplex" rela-
tionships.

The question Of the nature of these personal ties and how
they are able to serve such functions must be considered here. Since
many such ties were founded in the context of family and kinship,
attention is drawn to Meyer Fortes' ideas about the nature of kin-
ship, ideas that have sometimes been expressed in such terms as the
"morality" Of kinship, the "irreducibility" Of kinship, "prescriptive
altruism," and "amity" (see Bloch 1973: 75-89; Pitt Rivers 1973;
89-105). Fortes has held that kinship is different from all other
social relationships in that it is characterized by a moral force,
such that individuals are motivated by the common good, and not by
their own self-interest. There are,of course,opposing views among
anthropologists, some Of whom choose to ignore or minimize "morality"
as a motivating force (Bloch 1973: 75).

For Fortes, the essence of kinship morality consisted in
sharing without reckoning, a relationship Marshall Sahlins character-
izes as "generalized reciprocity" (Sahlins 1965: 147). The impli-
cation seems to be that balance is not sought in the short term
because the relationship is presumed to endure. Another way to look

at it, of course, is that the goal, the motive, in the relationship is

283

morality, not reciprocity. An imbalanced relationship is tolerated
because of considerations of morality. Fortes adOpts this view and
singles out kinship relationships as basically different from other
social ties because of their moral content—-a moral content which
cannot be reduced to economic or political reward (Fortes 1969:
238).

. . In following Fortes and stressing the significance
Of "moral" relationships and their irreducibility to perceived
economic or political reward. . . I am not in such complete
agreement with him that there are not certain qualifications
which are necessary to make here. First, for Fortes the moral
character of kinship is unique and marks Off kinship relations
from all others. By contrast I am quite prepared to believe
that in some societies other types of relationships seem to
have equal if not more "moral" commitment than kinship. When
I stress the significance of the morality of kinship for politi-
cal and economic organization it is because kinship is, very
often, the best type of long—term moral relationship. I would
not prejudge the issue whether it is the moral relationship par_
excellence although I am prepared to accept that this might
indeed be so. The second point which should be stressed is
that even if kinship is the most "moral" social relationship
there are many kinds of kinship,some implying shorter term com-
mitment, some longer term commitment. Fortes' own distinction
for the Tallensi between descent ties and ties of affinity is
an example. Finally, the third qualification follows from the!”
other two. If the effect Of morality is the existence of long ,
term commitments then there is no sharp break between kinship
and other commitments but rather we should regard kinship as
the end of a continuum consisting of commitments of different
terms (Bloch 1973: 76-77; emphasis supplied).

 

Whatever the precise "nature" of kinship, or that of the
"morality" mentioned above, it seems clear that the Chinese family
was and is characterized by relationships of generalized reciprocity,
reinforced, it should always be remembered, by hierarchical rela-
tionships Of an authoritarian nature, the latter in turn reinforced,
in traditional times, by the state political order. This is one of

the reasons that families and households were able to function as

'9.
1
. ‘_,

284

domestic units of production; mechanisms of authority and compliance
were built into the structure and were supported and sanctioned by
the wider social and political realm.

Recruitment to the family and/or household was accomplished
in part through the wide variety of social transactional forms dis-
cussed earlier in this dissertation. The question now arises, what
emerges from an analysis and comparison of these various forms of
marriage and adoption?

Before turning to a consideration of specific forms, we must
once again discuss some of the logical and conceptual problems
inherent in the situation. Rodney Needham has discussed, in the
context of kinship, marriage, and other topics of concern to social
anthropologists, the implications of some Of the arguments of
Wittgenstein, particularly those outlined in the Blue Book (1958).

Needham points out, and this has been implicit in most of
the general arguments in this dissertation, that anthropological
research has been misguided in even the supposedly basic topics of
kinship and marriage, by certain conceptual failings (Needham
1971: 2).

The most important failing here is an inapprOpriate con-
ception of the nature and definition of a "class." This is that a
class, in the logical sense, has been always assumed to have defi-
nite and specific criteria of membership. This is, in a sense, a
very mechanical and atomistic view of logic and reality, and might
best be seen as an inheritance from the Judan-Christian tradition

in the West, and not as an eternal given truth. Reality is seen

285

in this view as divided up into neatly discrete categories, the
items within each category being "the same."

Wittgenstein argues that this is a mistaken, or misguided,
view. Whereas we tend to think of classes being defined or char-
acterized by at least one specific attribute which controls member-
ship in that class and represents the "essence" of the class,
Wittgenstein argues that the members of a class may in fact be
related to each other in many different ways; that there may ngt_
be any specific attribute common to all members.

Thus, usually unconsciously,

We are inclined to think that there must be something

in common to all games, say, and that this common property is
the justification for applying the general term "game" to the
various games; whereas games form a family the members of which
have family likenesses. Some of them have the same nose, others
the same eyebrows and others again the same way of walking; and
these likenesses overlap (Wittgenstein 1958: 17; cited in
Needham 1971: 30 and Needham 1974: 69).

If we use the metaphor of family resemblance, then classes
become much less discrete and relationships more complex and criss-
crossing, both within and between classes.

There need not be, accordingly, any one thing that the
phenomena of "kinship" have in common, or that the phenomena
of "marriage," "descent," "relationship terminology," or
"incest" respectively have in common. Under each general term
the instances may be classed together, instead, as in spinning
a thread we twist fibre on fibre: "And the strength of the
thread does not reside in the fact that some one fibre runs
through its whole length, but in the overlapping of many
fibres" (Needham 1971: 30; 1974: 69; Wittgenstein 1953: sec. 67).

As Needham points out, we "are inclined by our conceptual
habits to assume that there must be something in common to all the

phenomena that we class together as 'kinship' and the rest, and it

286

is this unanalysed premise that leads us into theoretical diffi-
culty" (1971: 31; 1974: 69).

What must be done, in specific cases, is to investigate
the possibility of commonalities among items habitually classed
together, not to assume that they exist. One should then be prepared
for the situation in which there are no specific common features at
all, or that those which do exist are so general as to make proposi-
tions about the class itself difficult.

This problem is a very general one, common to many of the
sciences, but perhaps an occupational hazard to anthropologists
because Of the nature of their discipline. It amounts to faulty
abstraction and the faulty handling of abstractions. What Wittgen-
stein attacks as a problem is similar in many respects to the
general problem of cross-cultural comparison and research, and is
also similar to lines of thought in rather different disciplines
5uch as, for example, "general semantics," from Alfred
Korzybski to S. I. Hayakawa.

We abstract from reality, forming general classes along
lines Of interest to us for various reasons. We then label these
classes with general names, and then proceed to think in terms of,
use, and react. an the names labelling the general classes, instead
of the reality behind them. This is,of course,necessary for thought,
is thought. We fail to notice that the reality has changed, or the .
classification, in some unconscious way, has changed, and so there.

is no longer a good fit between the abstraction and reality. We

287

go on assuming, however, that there is still such a fit. This leads
us into difficulties.

When we react to the label of a class, and not to its members
themselves, and the label is one which has been translated or trans-
ferred from another culture or subculture, we have problems of cross-
cultural comparison, cultural relativity, and ethnocentrism. When
we react to the label of a class, react to the words themselves
instead of their referents, then semantic and logical problems can
arise such as the ones studied by the "general semanticists." In
the examples given by Wittgenstein, labels have been given to classes,
and it has been assumed that the classes are of a certain kind, with
certain definitions and logical characteristics. The problem arises
in that this has been done erroneously; the classes simply lack those
characteristics.

This kind of logical problem exists always and everywhere;
it is the anthrOpologist, often attempting to describe and explain
one culture in terms and concepts derived from another culture, who
meets it head on, in its most visible and unavoidable form. This
dissertation explores a minor example of the general problem--that
Of the description, analysis, and definition of "marriage," "adop-
tion," and related institutions in Chinese society. Error and con-
fusion can come about in at least two ways: the faulty equation Of
labels Of classes (equating the English term "marriage“ with one or
another Chinese-labelled class), and a faulty understanding Of the

nature of the classes (the assumption, for instance, that there

288

exists at least one common attribute defining membership in the class,
when this is not, in fact, the case).
Thus, with respect to the second error, Needham says that

What is common to descent systems may be only that they
variously exploit certain logical possibilities in a sexually
defined transmission Of rights from one generation to the next.
What is common to the myriad forms of marriage may be only the
contractual union of sexual statuses. What is common to rela-
tionship terminologies is the exploitation of certain formal
possibilities in the classification of sexual statuses. What
is common to incest prohibitions is merely the fact of pro-
hibition. There may thus be something in common, under each
general term, but not necessarily a definite set of character-
istic, specific, or essential features. The common feature Of
prohibition, for example, does not entail that the incest regu-
lations of different societies shall in any specific respect be
at all comparable. . . .

This does not mean, though, that comparative studies are
unfeasible or should be put at a discount. On the contrary,
if we give up the reificatory typologies that are usual in
social anthropology we shall actually be in a better position
to compare, for we shall at least see the social facts in a
less distorted way. More positively still, we can carry out
comparisons by reference to logical features, and also by the
formulation of more suitable abstractions than have been cus-
tomary (1971: 31; 1974: 70).

Needham's suggested commonality for all forms of marriage,
"the contractual union of sexual statuses," gives us a good start
on the job of dealing with the myriad forms OfChinesenmrriage.
We have noted that such classes may share a family resemblance only,
which is why his suggested commonality is at such a high level of
generality. This also implies that some of the characteristics of
the "family members" may be found to apply to those outside the
"family circle" proper; we should not be alarmed at crossing class
boundaries.

If we try to characterize marriage in traditional Chinese

society as a contractual union of sexual statuses, we run into the

289

problem of adOption. It has been pointed out how marriage blends
into adOption, through the institution of T. simpug_adoption, or the
"adOpted daughter-in-law." This would cause no problems to the
above characterization as long as the adopted daughter-in-law was
then clearly distinguishable from other adopted children; these are
just variant forms Of contractual union of sexual statuses.

These two forms of adoption are not always distinguishable,
however, and we have to say that marriage blends, step by step, into
"regular" adoption of females. This is true for males as well;
there was a continuum leading from "regular" adoption of young boys
through "adopted sons-in-law" (IMS; 272 nl77, 153) to uxorilocal
marriage.

One can argue, then, that "regular" marriage is structurally
identical to "regular" adoption. In the former case the union is
of different sexual statuses, whereas in the latter case the union
is of different generational statuses. It should be implicitly
understood in this formulation that the contracting parties are the
heads of households or families (males, in the Chinese case), and
not the individuals themselves. General formulations or definitions
of marriage which tend to assume that the parties to the contract
are the marrying couple themselves are inadequate cross-culturally,
and certainly would not fit the characteristics of the traditional
Chinese marriage system.

The postulation of the structural similarity between mar-
riage and adoption is exciting because it offers a more satisfying

solution to the "confusion" of the categories of marriage and

290

adoption; they are seen to be the same structural form, in which a
change of the characteristics of the elements of the structure leads
to an alteration in the "surface," or observed forms.

The objection can be anticipated that this structural simi-
larity is at such a level of generality that almost apy_social
transaction involving two (or more) persons can be subsumed under it.

There are several answers to this Objection. First, the
arguments cited from Needham and Wittgenstein above show that we
cannot expect much more than very general, nonspecific, character-
izations of even more delimited classes such as "marriage." If we
Operate as though we can be more precise at this level, then we are
deluding ourselves, and are headed for logical and conceptual
trouble.

Second, of course, the postulation Of this structural simi-
larity is justified by the empirical fact that marriage and adoption
are Often difficult, even in specific cases, to distinguish from
each other. If two phenomena are very difficult to distinguish, then
they are in some sense the same.

Third, there is every reason, in terms of the indigenous
Chinese systems of logic and world view, to make an equation between
relative sex statuses and relative generational statuses. The
statuses themselves are, of course, defined and characterized dif-
ferently, but the relations between the statuses in each sphere (the
relations between the statuses of male and female, and the rela-
tions between the statuses of generational elder and junior) are

structurally the same.

291

This line of reasoning brings us squarely into the realm
of symbolic variables mentioned by Cohen, and a brief excursion must
be made into Chinese traditional systems of thought, to see how the
society conceptualized the nature of the categories with which we
are dealing and the nature of their relationships.

This is a sticky area for the anthropologist because China
is a literate society with a long tradition of religious, philo-
SOphical, and sociO-political writing. The relationship of these
writings to Chinese society and culture is, of course, dialectical:
they both affect and reflect sociO-cultural reality. It places a
great burden on the anthrOpologist interested in Chinese society
that there is a vast amount of such writing, that it is obviously
so important to a thorough understanding of Chinese society as a
whole and of any particular part, and that it is written in a vari-
ety of classical Chinese styles. The ability to read such material,
to understand it, and to put it in proper sociO-cultural and politi-
cal perspective is a specialized study in and of itself.

The anthropologist must rely upon specialists from other
disciplines to tell him what he wants to know about these writings.
Often, however, these specialists do not know, or do not even want
to know, about the sorts of things Of interest to anthrOpology or
sociology.

If the anthropologist does not use what he can from the
written tradition, he can be accused Of naiveté, of having treated
China as a vast, nonliterate, "primitive" society with nothing but

local oral traditions. If he does use such material,he runs the risk

292

of being accused of a kind of elitism--of relying on material written
by the elite for the elite, material with little relevance, if any,
for the life and thought of the vast majority, the peasants.

The varieties of marriage and adoption, and the character-
istics Of each, seem remarkably constant and consistent over vast
geographical areas in China. Some of the more exotic forms, such as
those involving homosexual, or same-sex, "marriage,“ are known only
from circumscribed geographical areas, but the other forms seem to
be found just about everywhere. It has also been seen that some of
these forms have been known for 2,000 years or more.

All of these varieties Of marriage and adoption have grown
up and have develOped in a certain ideological environment; one
which seems to have Spread to wherever there were Chinese popula-
tions. This ideological environment is what anthrOpologists have
called the "world view"; it contains ideas about the cosmology
generally, and man's place in it. It also contains certain char-
acteristic modes of thought and analysis. It is these latter, in
particular, that seem to be useful in explaining phenomena in the
elite world and peasant world alike.

I will explore some of these modes of thought as analytic
tools; I do not claim that this is the way people actually think,
in each and every case (it is difficult, in any case, to specify
"what goes on in peOple's minds"), just that it is a useful and
productive mode of analysis. In this, I will Often have to rely
upon the analyses Of other specialists, and will use a number of

the "elitist" writings referred to above.

293

The traditional modes Of analysis and exposition exhibited
what has been called a "coordinate framework" and a kind of dialec-
tical method. The dialectic, of course, was most frequently expressed
in terms of M. yjn_and M. ynng, In most instances these are defined
as the "female principle" and the "male principle," respectively.

A kind of dialectical dualism, however, seems to be the important
thing in the dichotomy; references to "male" and "female" are only
one manifestation of a structural relationship which can be phrased
in any of a large number Of ways.

_jn_stands for, or is, the moon, earth, night, water, cold,
dampness, darkness, the hidden aspect of things. Ynng_partakes of
the sun, heaven, day, fire, heat, dryness, light, the open and visible
aspect of things, and so on. There are a vast number of these cate-
gories which are Opposed as yin_and ynng, categories which bear
little or no relationship to either "male" or "female."

This is clearly a case in which Wittgenstein's concept of
"family resemblance" is applicable. There is no single attribute
which can define a human family resemblance in most cases; it is
rather a topological arrangement of certain of the physical features
of the person, a certain relationship of the features to one another.
In a similar way there is no single attribute which can serve to
identify, or define, M. yjn_or M. ynng; rather it is the relationship
between the two which is of importance.

an_and ynng_are complementary. They imply the existence

of each other, they define each other; neither has any meaning apart

294

from that given it, in a very Hegelian way, by its "opposite." It
is the relational structure which has meaning.

In this kind of dialectical relationship the two are oppo-
sites and are complementary. They are never in the balanced and
complete confrontation we are used to in the West in the fight
between light and dark, goodness and evil. One is always the
ascendant, although since each contains the seeds of the other,
there is always the implication, at least, that this is in the pro-
cess Of changing. This is well summed up in the common symbol of
yin, ynng, and their interaction, a symbol which is perhaps most
familiar to Americans as the logo for one Of the major railways.

[in tends to flow downwards, to contract. It is soft,
passive, and submissive. inng_tends to expand, to flow upwards, and
outwards. It is hard, active, and controlling. As heaven, ynng_
fertilizes the earth, which is yin, It is the interaction of yin
and ynng_which produces all tfifings. In some texts the sexual
interpretation is made more explicit; it said that it is the inter-
course of heaven and earth which produces the myriad things and
beings of the universe, and that in sexual intercourse man and woman
re-enact, or more accurately, participate in, the universal inter-
action of yin_and ynng,

This persuasive, dialectical, dualism is obviously of impor-
tance in understanding such things as male/female relationships,
marriage, and human interactions in general. As an analytic device,
for explanation and exposition, it becomes more powerful when tied

up with the coordinate framework mentioned earlier.

295

Chinese cosmology might well have been authored by a strict
functionalist anthropologist with a decided bent toward animism, or
animatism. All things were seen as interconnected in a vast system
that is usually called, in more religious or philOSOphical writings,
the M. inn, usually (and inadequately) translated "the way." This
interconnectedness Of aspects of the universe is raised to the level
of explicit discussionauuibecomes, as a concept, the object of specu-
lation itself in these elitist philosophical and religious writings.
It seems to be there, however, in peasant world-views as well, though
perhaps in less explicit, and certainly in less abstracted, form.

The qualities of things in this coordinate framework are to
be found from their positions in the structure of the system, while
at the same time their position in the structure is "because of"
their qualities. The causality we are used to in the Judaeo-
Christian world (discrete, mechanical, uni-directional and perhaps
to be sought ultimately in the idea of a single creator-god) does
not have a clear role here.

A change in one area, or part, of the great system will
bring about, or be correlated with (since causality is unclear),
change in another part Of the system.

Joseph Needham says

A number of modern students-~H. Wilhelm, Eberhard, Jablonski,

and above all Granet--have named the kind of thinking with which
we have here to do, "coordinative thinking" or "associative
thinking." This intuitive-associative system has its own
causality and logic. It is not either superstition or primi-
tive superstition, but a characteristic thought-form of its

own. H. Wilhelm contrasts it with the "subordinative" thinking

characteristic of European science, which laid such emphasis on
external causation. In coordinative thinking, conceptions are

296

not subsumed under one another, but placed side by side in a
pattern, and things influence one another not by acts of
mechanical causation, but by a kind of "inductance". .

The key word in Chinese thought is Order and above all Pat-
tern. . . .The symbolic correlations or correspondences all
formed part of one colossal pattern. Things behaved in par-
ticular ways not necessarily because of prior actions or impul-
sions of other things, but because their position in the ever-
moving cyclical universe was such that they were endowed with
intrinsic natures which made that behavior inevitable for them.
If they did not behave in those particular ways they would lose
their relational positions in the whole (which made them what
they were), and turn into something other than themselves.

They were thus parts in existential dependence upon the whole
world-organism. And they reacted upon one another not so much
by mechanical impulsion or causation as by a kind of mysterious
resonance (Needham 1956: II, 280).

 

Thus the dynastic histories and other works had sections
recording strange, unusual events Of various kinds as indications
and portents of trouble or change in the system. The Sou-shen Chi
(Record Of Investigations Into the Supernatural) has several such
entries referring specifically to male/female role relations:

In the latter part of the seventh year of the Chien- an
reign period of the Emperor Hsien Ti [of the Later Han;
202 A. D. ], in Sui [an ancient district in Szuchuan], a male
changed into a female. Chou chun submitted a memorial, saying
that "in the time of Ai Ti [of the Han; reigned6 B. C. to A. D. 1]
there was also this kind Of change. There will be a change in
affairs of state in the future. " In the 25th year [A. D. 200]
Shan Yang Kung was enfeoffed (Sou- shen Chi VI, 11a).

In the time of A1 Ti's Chien-p'ing reign period [6 B.C.
to 2 B.C.], in Kiangsi, a man changed into a woman, married
out as wife, and gave birth to a son. In Ch' ang- -an Ch' en Feng
became general, and then lost all his descendants. Viewing
this as the appearance of mutual production [which implies
mutual destruction in the coordinate five-"element" system],
some said that he who had become a wife and had given birth to
a son would have his line eradicated in the next generation.
Thus later Ai Ti died, P' ing- -Ti was lost, and Wang Mang took
this as an Opportunity to plot [for power] (Sou- shen Chi VI:
7b

In the time of Hui Ti [Of the Western Tsin; reigned
A. D. 290- 307], in Lo-yang, there was a person with the physical

297

characteristics of both male and female, who moreover could act
the part of either sex, and what is more, was extremely licen-
tious. The wars and disorders of the empire arise from the
monstrous shapes which are created by the chaotic and improper
[mingling of] male and female essences (Sou-shen Chi VII: 3a).

The general system which is the universe has as its motivat-
ing force the interaction of M. yin_and M. ynng, It has been noted
that the interaction of M. yin and M. ynng_was sometimes phrased in
terms of sexual intercourse. It is no surprise, then, to find that
sexual intercourse, conceptions, and birth are all explained (in
these literary works at least) in terms of M. yin, M. ynng, and
M. nnii_(the "breath" or essence of things), and the five elements, or
processes.

The Fu-jen Liangefang(Tried and True Prescriptions for
Wnngn), by Ch'en Tzu-ming, of the Sung, says that

The intercourse Of man and woman is the intermingling of
their two essences. If yin_blood arrives first, and later the
yang semen attacks, the blood Opens and envelops; the semen
enters to form bones, and a male form is created. If the yang
semen enters first, and the yin blood then admixed, the semen
will Open, and envelop the blood. The blood enters to form the
base, and a female is formed. The yang essence collects on the
ventral aspect. Thus males are heavier ventrally, and the corpses
Of drowning victims will always lie prone in the water. The
yin_essence collects dorsally, so women are heavier dorsally,
and drowning victims will die supine in the water. When animals
die Of drowning, they also will be supine or prone in this way.

If the yin_and yang arrive together, then the form is
neither male nor female. If the semen and blood separate, then
Siamese twins result. Indications for judging the nature Of the
fetus are as follows: If the husband is young, the wife older,
and a daughter is born, she will be thin, haggard. If the
mother is strong and the father declining, and a son is born,
then he will be weak. The Old prescriptions first noted this
(KCTSCC v. 56: 640).

The Wan-ch'uan Kuang-ssu Chi'yao (The Complete Essential Guide

for Increasing Progeny) has a similar injunction:

 

298

The male's semen abounds in thinking of his wife. The
woman's blood abounds in pregnancy. Male and female unite in
order to multiply descendants and to continue the Obligations
of morality. The consequences are no light matter, so one
cannot but speak of the method of seeking a son. In this the
male is ruled by semen and the female by blood. If the yang
semen is copious and not exhausted, and if the yin blood flows
in a timely manner and is not prolonged, if in and yang are
freely intermingled, and semen and blood combine, then a fetus
will form, and children will multiply. ,

If the yang is weak, and cannot descend to meet the in,
if the yin_is deficient and cannot rise to follow the an ,
then yin and yang are Opposed, and semen and blood wil eviate
from their proper paths. This is why there are no children.

The ignorant in the past did not know this, and Opposed
natural law, mistakenly promulgating the art of seeking the
inner breath, extending the production-destruction relationships
from the five-elements theory, seeking nourishment in pills,
making falseness equal to truth, and trying to make the works
of man to surpass those of nature. Though they had children,
many did not give birth, and many could not rear their children
(KCTSCC v. 56: 642).

These last mutterings about the "inner breath,“ the five-
elements theory, and pills are a reference to a kind of indigenous
tantrism advocated by some Taoist writers. These sexual techniques
were only one kind of a number of different sets of techniques for
attempting to achieve material immortality. The search for this
immortality was constant among certain strains of Taoism and led,
according to Joseph Needham, to a considerable amount of scientific
or protO-scientific thinking, observation, and experimentation.

The sexual techniques

Owing to Confucian and Buddhist antagonism . . . have

remained much the most recondite, yet they have considerable
physiological interest. It was quite natural, in view of the
general acceptance of the iinrYang theories, to think of human
sexual relations against a cosmic background, and indeed as
having intimate connections with the mechanism of the whole
universe. The Taoists considered that sex, far from being an
obstacle to the attainment Of hsien-ship (the "immortality"

referred to above), could be made to aid it in important ways.
Techniques practiced in private were called "the method of

299

nourishing the life by means Of the Yin and the Yang" (iin

Yang yang_sheng chih tao), and their DB's—i'c aim was to conserve

as much as possible of the seminal essence (ching) and the divine
element (shen), especially by "causing the ching to return"

(huan ching). At the same time, the two great forces, as
incarnated in separate human individuals, were to act as indis-
pensable nourishment the one for the other. . . (Needham 1956:
146-47 .

 

 

These techniques, which viewed sex as a sort Of alchemy,
could not be sharply distinguished from more ordinary sexual tech-
niques "of the lay bedchamber" (Needham 1956: 147).

A number of books advocating and describing sexual prac-
tices have survived Confucian and Buddhist censorship but are
quite rare. Thus it might be thought that they and their ideas had
little currency among the majority of the people. Still, Needham
says that

POpular versions of some of these still circulate (or did

so until recently) in the lending-libraries Of pedlars in China,
and others are also passed privately from hand to hand. I
always remember the reply given to me by one of the deepest
students of Taoism at Chheng-tu when I asked him how many
people followed these precepts: "Probably more than half the
ladies and gentlemen of Szechuan" (Needham 1956: l47n).

Chinese society, if I can anthropomorphize this way, has
Opted for the M. yang, or male, principle as the dominant one for
the ordering of social life. This could hardly be otherwise, since
the M. yang principle is the active, controlling one, although some
Taoists have at times advocated a kind of nonaction, passive, M. yin_
approach to social organization and government.

A social order based on, or associated with, M. yang, however,

necessarily implies one based on M. yi_n. This logical fact has been

recognized by Chinese thinkers, and such a social order, matrilineal

300

and matriarchal, has been attributed to their own proto-human ances-
tors of the mythological past, to animals, or to foreigners. How-
ever, one can argue that analytically such an order does exist as
the M. yin, or hidden, aspect of Chinese society.

Although M. yang is the active, hard, controlling element,
and M. yin_the passive, soft, submissive one, this does not mean
that the former is necessarily the more "powerful" of the two. In
some theories about pregnancy there are indications that a fetus
is "basically" or "naturally" female, and that the male's M. yang-
essence, or semen, must be strong to overcome this basic tendency if
a male child is to result; certainly conception is usually phrased
in terms of some sort of contest between the M. yang and M. yin_
forces, the outcome Of which determines the sex of the child.

In some of the Taoist techniques mentioned above, a male
should have intercourse, without ejaculation, with nine or eleven
different women (Odd numbers being M. yang) before having intercourse,
with ejaculation, with his wife. In this way his M. yang essence
would be strengthened, and the child would be sure to be male. As
Needham puts it:

The purpose of the Taoist techniques was to increase the
amount of life-giving M. ching ("essence") as much as possible
by sexual stimulus, but at the same time to avoid as far as
possible the loss Of it. Moreover, if the Yang force in man
were continually fed by the iin_force, it would not only conduce
to his health and longevity, but its intense maleness would
ensue that when emission did take place, the sex of the result-
ing child would be male (Needham 1956: 149).

Some.of the manuals of sex techniques mentioned earlier

advocate a kind of "sexual vampirism"--methods of strengthening

oneself at the expense of women. (This critical view was that of

301

R. H. van Gulik. He later adOpted a more positive view of Taoist
sex practices in general, but for this argument, I think that
his earlier judgment must remain [Needham 1956: l46n].)

In the manual called YU Fang Pi ChUeh (Secret Instructions

Concerning the Jade Chamber), it is explicitly advised that no one

 

practicing these methods should let women know anything at all about
them. Not only would the arts be of no use in the nourishment of

M. ynng, but harm and illness could result for him because this would
be giving the woman a weapon that she could use against him. One
woman, Hsi Wang Mu, a sort of personification of M. yin, is mentioned
as skilled in these arts. She is described as practicing her sexual
vampirism on young boys.

The nourishment Of M. ynng_on the part of the male continues,.
in a sense, after death. The soul, as might be expected, is also
analyzable in terms of M. yin_and M. ynng, The theories of the soul
are complex and varied, but usually they agree in dividing the soul
essence into two aspects, a M. ynng_aspect, the M. nngn, and the
M. yin aspect, the M. knni, At death the M, §n§n_goes upwards, the
natural M. ynng_direction, to heaven, and the M. kngi_goes downwards,
the natural M. yin direction, to the underworld. The M. gngn.becomes
a benign spirit, which requires continuing nourishment from this
world, and the M. 5291 becomes an unfriendly spirit, which can be
placated by those in this world. The picture is much more complex
than this in part because a M. nnnn_can become a M. kngi_if there is

no one to sacrifice to it, and, as has been pointed out by Chiao

302

Chien, "Chinese thinking on the supernatural is not categorical but
rather dialectical" (Chiao 1975: 393).

The argument that I am making is that it would be possible,
and might be very valuable, to analyze Chinese society using the
tools given in the traditional modes of thought, dividing the society
into M. ynng_and M. yin aspects, related in the characteristic
M. yin/ynng_dialectic.

The patrilineal, Confucian, "standard" aspects of the
society can be seen as M. ynng, but a M. yin_aspect is also present.
This seems to accord well with Marjorie Wolf's findings in Women and

the Family in Rural Taiwan (1972) that a look at patrilineal

 

Taiwanese society from the point of view of the women reveals the
importance of a "hidden" social structure of linked, woman-centered
uterine families. She says that these uterine families are in
conflict with, but also mesh at certain points with, the patrilineal
social structure. She further states that this "woman's subculture,"
as she calls it, is below the level of consciousness, and that it has
no ideology, but "appears almost as a response to the traditional
family organized in terms Of a male ideology."

One can argue, however, that there in, in a sense, an
ideology, and that the uterine family arises as a response to the‘
male ideology but also as a logical requirement of the dialectic.
The uterine family is part of the M. yin aspect of social organiza-
tion, as are many Of the other types of forms mentioned here, and
the M. yin/ynng_dualism provides the structure and outline of its

ideology. Both aspects Of the social structure are necessary for

303

the working of the entire system, of which they are complementary
parts.

Using this kind of analysis, one can say that the kinds of
contractual unions discussed here are, at the most general level,
transactions between two statuses which stand in a certain relation-
ship to each other. This relationship is never one of equality, but
is 9"? pfcomplementary inequality: a M. yin/ynng_relationship.

At this level Of generality, the statuses and associated
roles are not male/female, but superiority/inferiority or, perhaps
better in some situations, activity/passivity. One actor is always
clearly in the M. ynng_role, and one quite clearly in the M-.Xifl
role. In marriage proper, this is manifested as male/female. In
adoption proper, as senior/junior. 1m1uxorilocal marriage the struc-
ture remains the same; it is just that it is the biologically male
actor who takes the M. yin role, and the biologically female actor
who takes the M. ynng_role. At the level of ideology this is one
of the reasons that the union is negatively valued by the male: the
structure of the union is disharmonious with respect to the structure
of the society generally. In same-sex unions, though there is very
little information on them, it seems that again there is always a
clear differentiation Of roles: one partner always plays the M. ynng_
role and one the M. yin.

These sorts Of transactions involve the transferral Of
certain rights and duties over the persons concerned. These are
expressed ideologically as a kind of transformation of "substance"

or "nature." This "substance" or "nature" can be identified with

304

the M. ynng aspect of the individual, or the individual's soul. The
soul tablet of a woman, for example, remains on her husband's family's
altar after her death, and not on her natal family's altar. This

is true in most forms of adoption as well. In the section on
biographies Of meritorious women of the Taiwan T'ung-shih (The

General History of Taiwan), this is well described in the case of

 

an "unmarried" T. simpua whose husband-to-be had died:

Shih Chin-niang was from [a certain named place] in Chang-
hua. When she was 14 years Old, she was made a T. simpua of
Wu Mao-shui. She was very docile and agreeablein nature, and
was filial to her parents-in-law, who loved her. They waited
until she was of marriageable age, and then divined for a lucky
day, whereupon Wu Mao-shui (her husband-to-be) suddenly died.
She was then just 16. She was grief-stricken, and did not want
to live. Her parents-in-law exhorted her, compelling her to
attend the funeral rites. She wailed day and night, so pitiably
that it brought tears to the eyes of those who heard. One day
she returned to her natal home, and her mother, thinking of
her daughter's youth, wanted her to remarry. Chin-niang nobly
said "In life I am a person of the Wu family; in death I will
be a spirit (M. kuei) of the Wu family. How could I remarry?",
whereupon she returned, without taking leave Of her mother
(1918?: 1128).

This transformation of substance was not always complete,
symbolizing the less than complete transferral of rights from one
family to another, and it could be treated l'creatively" to justify,
or make possible, certain lines of action.

I have described in Chapter V how the Liao lineage in T'ien-
sung-p'i was what is called "split Liao"; they had originally been
surnamed Chang, but an ancestor had married uxorilocally into the
Liao surname group, whereupon they took that surname for most pur-
poses in this world, retaining Chang for their dealing with the

other world (or worlds). This means that they are not supposed to

305

marry into the Chang surname; this would be, in effect, a viola-
tion of surname exogamy.

I knew, however, that one of "0' neighbors, surnamed Chang,
had married a Liao woman. The latter is Liao YUeh-O; she married a
man named Chang Cheng-ming. When I asked informants from the Liao
surname group about this, they said that this was different because
"her bones aren't Liao, only her skin is,“ and so the prohibition
did not apply to her. This was because her father, originally sur-
named Wang, had married uxorilocally into the Liao family (simul-
taneously being adOpted as a Liao as well). Thus, uxorilocal mar-
riage, adoption, change of surname, and the fact that all of his
children then became Liaos did nothing to change his basic "Wang-ness"
(if that became important or useful, of course). He and his chil-
dren have Liao skins but Wang bones.

This interpretation is reinforced by the description of one
kind of M. ting-hun, or "betrothal" ceremony given me by another
informant, not a Liao. In this ceremony, an open, or "public" cere-
mony, the go-between comes to the bride's home with six men (though
children can be counted among the six), three chickens, two pieces
of pork, three M. 92:2:31 (a fish like a trout), the betrothal money,
and so on.

The chickens are boned. This is explained as having the
meaning, in Taiwanese, bah beh hong chiah, kut bo ai hong khe ("We
give the flesh to eat, but not the bones to gnaw.") Douglas (1923)
has T. bah hO-lang chiah, kut m'chia* no-lang khoe, "Let people eat

306

the flesh but not gnaw on the bones, to allow our friends to be
somewhat injured or reviled, but will not let it go too far" (gin).

In terms of rights over the bride, this can be interpreted
as meaning something like "Now she is part Of your family and you
can treat her as you see fit, but if it goes too far, and she is
treated too badly, it will reflect discredit on our family, and we
will have to do something about it."

In terms of the ideological system, it can also be interpreted '
(and these are not mutually exclusive interpretations) as a symbolic
statement that in this transaction the "substance" of the bride is
in part transformed to that of the groom's family, but not all of
it; she remains in part still identified with her natal family.

To return to the analytic mode, the general form of such
contractual unions is one of a joining together in a particular kind
of relationship two different, if complementary, statuses. The
general characterization of this union is conveniently expressed in
terms of M. yin_and M. ynng.(lhis is my analysis, and I am not sug-
gesting thereby that this is the conscious categorization in the minds
of any particular peOple.) In the case of adoption proper, the
structure is manifested in terms of generation differences (and not
just age). In the case of marriage prOper, it is manifested in
terms of male/female differences.

In both cases, in adOption and in marriage, there is a trans-
formation or transferral of, in the ideological system, "substance,"
which I have identified with some sort of M. ynng_essence. In terms

of rights and duties, this is expressed as a transferral of rights

307

in the monopoly control, the disposal of, sexuality and reproduc-
tive powers; these things are transferred in marriage and in adOp-
tion.

In this analysis T. §inpnn_ad0ption poses no problems; a
female has been transferred physically from one household to another,
and at the same time control over her sexuality has been transferred
as well. From the point Of view of the head of a household, in the
case of either adoption prOper or marriage prOper (of one Of his
sons), he has brought into the household and the family a female Of
junior status, over whose sexuality and reproductive powers he has
control. He can choose to transfer these rights to another family,
marrying her out, or adopting her out, in which case, if she is
below a certain age, it is a case of "regular adoption." On the
other hand, he may wish to himself exercise these rights (through
his son), in which case it is LmadOption/marriage. If the
female is already of marriageable age when she is brought into the
family, then it is marriage proper.

The vast majority of the transactional forms, including all
of the common ones, can then be "generated" by a small number of
binary variables. These are sex, age relative to the culturally
defined age Of marriage, and whether the disposal of sexual and
reproductive powers is internal (consume) or external (exchange)

to the family. This can be shown in a simple diagram:

308

Exchange (regular adoption)
Male-——{:

ADOPTION Consume (adopted son-in-law)

Younger than
marriage age -Exchange (regular adoption)
Female-{—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Person to be Consume (T. simpua and
recruited T- cha-bo-kan)
MA . o
MarrigégeaEge alefi fiJJxor1local Marr1age
or Older .- _ .
emale t1r1local Marr1age

Rights other than just those pertaining to control of, or
the disposal Of, sexual activity and reproductive powers are trans-
ferred in these transactions, but they are not necessary variables
in the generation of the above forms.

Not all of the forms are accounted for, nor are all of the
subvarieties of the ones in the diagram accounted for; to do so in
the latter case would simply make the diagram, which is intended
simply as a heuristic device, too complicated. In the former case,
we must keep in mind Wittgenstein's "family resemblance." Some of
the types of "marriage" and "adOption" that have been considered
simply do not fit in with the above variables; their resemblance to
the "ordinary" forms lies elsewhere.

The types of marriage listed in the diagram, for instance,
are all implicitly arranged marriages; in a transaction between
two parties, rights over a third party are being transferred. In
the little information we have on same-sex "marriages," it is so
only for the case of males (and this is very shaky information); in

the female case it seems always to have been a union entered into by

309

two individuals. Even so, in that case, as we have seen, one

partner seems to have taken the dominant, male role, and the other
the female, submissive role (I do not mean physically, in intercourse,
but socially, in terms Of status and role), so the structural pattern
is one of virilocal marriage.

If we return to the particular transactional forms listed
in the diagram, we can see that the various types of uxorilocal
marriage are not included. In principle further variables can be
entered here to account for these varieties. These, however, become
more and more specific to particular cases and have no applicability
to the other contractual unions in the diagram.

For instance, the difference between the uxorilocal son-
in-law and the uxorilocal husband is dependent on whether the family
has a daughter or a daughter-in-law to marry. If it is a daughter-
in-law, then the presence or absence of children, their ages, the
economic standing Of the family, the amount of property in the
family estate, and so on will combine and interact in order to pro-
duce the "property-managing" uxorilocal husband, the uxorilocal
husband for support of the children, the uxorilocal husband for
production of children, or even the polyandrous uxorilocal husband
("uxorilocal husband for support of the [first] husband").

The conclusion that adoption and marriage are at one level
the same is not just an analytic suggestion. Arthur Wolf relates
a curious incident in his dissertation (Wolf 1964: 78), one which
in: presents without comment. The incident concerned a prostitute

who adOpted a baby girl. After a prostitute has been in the business

310

for a number of years, her chances of marriage are rather slim, and
she starts thinking of alternative plans for the future. At this
point, if it's evident that sheis never going to get out of her
calling, and if she has no children of her own, there is a good
chance that she will adopt a baby girl to follow in her footsteps.
It is often said that in the traditional Chinese society, a woman
could have descendants only through a man. The custom of spirit
marriage shows that this was in part so. When it was difficult to
find any man, however, even that rule could be ignored, and a
single prostitute, or, as we have seen, women united in homosexual
marriage, could adopt children to provide descendants for them-
selves.

In Wolf's incident, the adopting mother paid a price for the;-
child, a price called a brideprice. She dressed the child in red,
the traditional color of marriage, when she took it home, and three
days later she brought the child back to its natal home for a visit,
to I'play the role Of guest." This latter is exactly what a new bride
does three days after the wedding. The expression "play the role Of
guest" indicates that after marriage she is a guest in her own natal
home, the same as any other stranger. Her substance has been altered
through marriage.

The adoptive mother in this case was saying in symbols what
I have argued in this dissertation. She was both adopting and mar-l
rying the child. It may have been that she chose to use the symbolsfl
Of marriage instead Of the simpler procedure of "ordinary" adoption

because she intended to "use" the reproductive powers of

311

the child herself; the child was being adopted in as a T. ginpnn,
would stay in the woman's household to provide descendants and support
in old age, and would not be married out.

For Chinese society, the implications of this approach are
various." First, more heterogeneity and variability must be expected
than has usually been the case. In both China and the West, ideas and
expectations about the traditional family structure and the forms
and incidences of marriage and adoption were heavily influenced and
biased by the Chinese elite ideology. More explicit reference must
be made to the role of socioeconomic class differences, and this in
turn means an increased emphasis on the political economy Of the tra-
ditional society.

Abner Cohen has suggested that the anthrOpologist should
deliberately formulate problems so that they make explicit reference
to the state. Perhaps those interested in various forms Of social
organization should also deliberately formulate their problems so
as to make explicit reference to irregular and nonstandard forms.
Information on such irregularities is sometimes hard to find in the
case of Chinese society, but, as I have tried to show in this dis-
sertation, a careful search of sometimes unorthodox sources may be
surprisingly productive.

This leads to another generalization: that anthropologists
and others interested in the analysis and description of Chinese
society should make more use of the vast amount Of written material

that has been produced by the Chinese themselves. It is Obvious that

312

an understanding of the Chinese peasantry is crucial to an under-
standing of modern, if not all of, Chinese history. The anthro-
pologist, armed with the basic sinological tools, an orientation
toward long-term participant-observation fieldwork, and a compara-
tive perspective of peasant societies the world over, stands in a
good position to make important contributions in this area.

As far as marriage and adoption are concerned, it is a
mistake to focus upon isolated forms as unitary objects of study.
What should be the Object of study is the spectrum of such forms,
their relations one to the other, and the relationships Of all of

them to the politico-economic setting.

APPENDICES

313

APPENDIX A

SELECTED CONTRACTS

314

APPENDIX A

SELECTED CONTRACTS

The following is a free translation of the contract of "spirit kuo-
fang adOption" of Wu Chang-hsDn.

A Contract Of Kuo-fang AdOption.

The executor of this contract, and his wife Ch'en Su-lien, are the
parents Of a third son, named Wu Chang-thn, now two years of age,
born Min-kuo 60 [1971], eighth lunar month, 19th day,_n§n hour [7-9
p.m.]. Since the executor's younger brother, Wu Shu-sung died young,
without having raised sons or daughters to continue the line of
sacrifices, Wu Shu-huang and his wife agree, out of the sentiment
proper between brothers, to adOpt their son Wu Chang-thn to the other
branch of the family. His entire person will be transferred to the
late Wu Shu-sung so that the smoke of incense and sacrifices will be
plentiful and will continue forever. There will be no later regrets,
and the contract will not be violated, but it is feared that there is
no evidence of the spoken word, so on this day, the 15th day of the
second lunar month of the 6lst year of the Republic, one c0py of

this contract is drawn up, with the Offering of incense, and it is
presented before the brazier as evidence. The late uncle Wu Shu-sung
shall protect and care for the kuo-fang child Wu Chang-hsOn, and will
ensure that he will grow to adulthood, marry, find a good position,

have many sons, and have a happy and wealthy life.

315

316

Concluded this day Jen-tzu, 6lst year of the Republic, 2nd lunar
month, 15th day.
A TRUE DOCUMENT

Executor Wu Shu-huang (his seal)
Witness Ch'en Su-lien (her seal)
Guarantor Wu A-shu (his seal)
Scrivener Yeh Liang-ch'uan (his seal)

317

Uxorilocal Marriage Contract #1

The executor of this contract, Ma Chi-te, because his younger
brother, Ma Ts'ung-te,died of illness, leaving a wife surnamed
Chang, who has no one to look after her and her children, has,
through a go-between,invited Ho Chiang-lien to move in as uxorilocal
husband, in order to facilitate their agricultural livelihood and
provide food and clothing. Ho Chiang-lien shall not be permitted
to be lazy or profligate. Should anyone in the Ma household raise
any troubles about this marriage, Ma Chi-te guarantees that none Of
this shall affect the position of HO Chiang-lien.

Fearing that there is no evidence of the spoken word, this

marriage contract is written up as proof.

Go-between Hsieh Ch'eng-yfian

Drawn up for the executor Ma Chi-te on the 25th day of the 9th
month, 1916.

Shensi Province (MSS: 1703-1704).

318

Uxorilocal Marriage Contract #1

3C

:4

mh+flméfm3gpp
Eafifififififlfiffififififi
>

4.5.. i
'r—IZS
any»
1‘"qu
fi”
..,
31'

‘1 {Iv—L: /:.':'_"4';;L
3&1 : '1 ".
E.
iii-l

\-
F?
a:
a-B
fi

_g+<~,
5133-”;

‘1») ._ F

at“

‘5'. o
$.?
E:

h!

' "34 ifiiélnfiiz.
3%

n it

319

Uxorilocal Marriage Contract #2, Part 1

 

The executor of this contract, Li Shih-hsing, is now approach-
ing thirty years of age, but has not yet married because his family
is poor. Now the Oldest son of Liu K'ai-en has died, leaving a wife
surnamed Wu. She finds it difficult to endure the loneliness, and
is unwilling to marry out, so a go-between has been asked to approach
the Liu family to arrange an uxorilocal marriage, to the woman Wu,
creating everlasting marriage ties.

Liu has agreed and has selected an auspicious day for the
uxorilocal marriage (M. Shang-men), and has met and discussed this
with the kinsmen.

From this time on, Li will serve and support the older gen-
eration, and will see to their funeral arrangements. He will also
maintain harmony with the younger generation, work diligently in
the fields, labor hard, and not shirk the difficulties even of night
work. Should, in the future, new land and property be acquired, it
will be divided into three shares, and shared equally, with no dis-
putes. Li shall not idle away his time, or waste money on prosti-
tues, or go into debt. If he does not obey, then he shall be turned
from the family, alone. Should the man and wife both become alien-
ated and both wish to leave the parents-in-law [of the woman] then
they [the man and woman] shall provide a string of 100 cash as a
present to the Liu family. The land shall all remain with the Liu
family. As long as the woman Wu maintains her resolve to remain and

work to raise the descendants, the Liu family shall not take away

320

the land and cut Off their sustenance. Li shall not later malign
the Liu family, nor shall he form liaisons elsewhere.

Husband and wife; Older brother and younger; united in a
harmonious group, how beautiful!

After this uxorilocal marriage, both parties shall be content,
and never shall there be regrets or recriminations. Fearing that
there is no evidence of the spoken word, this contract is written
up as eternal proof. May their offspring multiply over the gen-

erations.

GO-betweens Wang Yung-chih
Mrs. Li, nee Liu

Older Brother Li Shih-1in
Elders Liu K'ai-mou

Liu Sheng-lan

Liu Teng-ch'eng

Scrivener Wu Shou-shan

Written up this 6th day, 10th month, 1919, for Li Shih-hsing.

-eaaeme4mnam

321

Uxorilocal Marriage Contract #2, Part 1

TI -4* QE ‘
m ~——a~n a
..tis*aa
I" 3. NR 1‘ 1b u’
%.:§ 5 matazfi
5&5}?

LIL-3.
I

aamwnm+msanem>ne

EST

0’:

L:

emmmaee

. z:
$21815
..‘ '11! A 11'
>3} t'i- 9.511 if: 13?. a?
aaenan
‘ififw-i ‘35
iii "‘1 it: 76 iii. if:
It It 4‘ if.“- " 1'3
121’. 1111;— £11 3'43 182.11
1% 40% " 161 RR 1"]
11% iii in G? B in}
.satt
l' 7 'r t‘ 15
am 5a 1.11 If ’5
ESL id DE ifi: Q?

322

Uxorilocal Marriage Contract #2, Part 2

 

The executor of this contract of uxorilocal marriage to a
daughter-in-law to continue the kinship line, Liu Ka'i-en, because
his Oldest son Sai-lan died leaving a wife surnamed Wu, but no off-
spring, and because she finds it difficult to endure the loneli-
ness of the many years to come, without a husband and family, and
after discussion, an uxorilocal husband is being married in.

If the second son,Yu-lan, or the third son, Chang-lan, have
extra sons, one will be instituted as heir to the line of the Oldest
son; how could it be permitted that his line be extinguished?

The various kinsmen are agreeable to presenting one portion
Of land, at Ta-wang Ch'ang-kou, local name Schoolhouse Hill, including
one house, one pig sty, and a vegetable garden, along with the rental
contracts, to Wu. Also included are three persimmon trees, and at
Hua-chia, land (sufficient to require one inn two inn of seed rice
to plant), two parcels of land in all. This property is to be used
in the future as capital by Wu to support herself and her children,
to support those of the Older generation, and pay for their burial
expenses, and to raise up the heirs in the younger generation to
continue the Oldest son's line and the sacrifices thereto.

She cannot abandon the parents-in-law, and cannot later
change her mind about this. As long as she uses the land and property
given to her to act filially, and works hard to support the heirs,
to care for the elderly and arrange for their funerals, the land shall

not be taken from her, and her income will not thus be cut off.

323

Fearing that there is no evidence of the spoken word, this
contract of succession to the kin line is given to Wu to act as
proof of her possession of the landlord rights.

This contract is to be kept under the-protection of the kins-

man Wang Yung-ch'ang.

Witnesses Liu K'ai-mao
Liu Sheng-lan
Wang Yung-chih
Mrs. Li née Liu
Li Shih-1i

Scrivener Wu Ch'i-hsiu

Written up this 6th day, 10th month, 1919 for Liu K'ai-en and wife
in their presence.

Shen-hsi Province (NSS; 1718-20).

mmgfimw§m+8>@m%%
2

fl

:4? 9:11

cut 5::-

"a

we”
wfifi
“$¢%

324

Uxorilocal Marriage Contract #2, Part 2

. ...—_..-
I

1

51-: Q!-
'JESUXB
2*}.

at»
E. \
_,

‘e‘d'
7J6
>46
[75.1
3§N§$
.F
i

a
:1.—
U
u.
., s:-
a%hfiN%§fi%E£$
E3

rawm

an _
'E% a

' In]
fiiflj: ‘ =
E an it: 11. m . {,1
43415 2 :m
4%: {N 'NB 2%. ;g_
21? 21s ‘1 1L1! '

.Emgfifi

" H'Eaiiifii
. -: ‘1‘; A _: ' .
4% in: wt ”F 33 iL‘a .2?
all 41 5} 11a _.

fi
2?
H
HE
«as
I”
.1—
E$l
E

m\
Q1" L51 -
a ' “"53"“ a'

1.

new
am
ee
as
$2
‘dg-A

H§EH

fi
‘1‘, \ I
Em ,
E

h

{1

”iv

mamas
a

“‘4:
00 hala-
5’

w

I.

l
J
\ Q;P
E
m.

W3§

figfilafifikr
AL
.31

T; 5?: ’é. HQ 31
' I“. it}: Ii‘clt [)5 1 ‘t
' 9h, 51.; .1: ‘ :13
i: 11L}. :Ji: in; 1;).
98 181 IE; L1 ten
35"fi‘%zé
£2 :ié £2. £3 i§£
)Ufllfiifi %
"330 m a
3,1 5‘:- 1’4. 711-: at
33?” ’L'n if??? 'fl'l. 4‘)!
fi'Oifi

ea
2%
msg:

e:

N

§fi§$3Tfi

fibfi?

51

”a“
U-
I. ' ;

'..
h c

55

35‘
,‘_h
any

25%

325

Uxorilocal Marriage Contract #3

 

The executors of this contract are Wu Shih-tou and Chang
Chin-hsien. The wife, Li Yao-niang, age 36, of the grandson of Wu
has not yet given birth to any children, but she has adopted a young
girl, Wang Chin-liang, aged 16, who is now of marriageable age, but
does not yet have a mate. Wu and Li have discussed this, and now
agree to invite in an uxorilocal husband for the girl Wang, to support
the Older generation and provide heirs.

At the same time, Chang Ts'ung-liu, aged 25, nephew of Chang
Chin-hsien, also desired to be married, and so the two families have
discussed marriage plans through a mediator. They have agreed, and
have selected an auspicious day for the two to be married.

If the couple later give birth to sons or daughers, it is
stipulated that they will divide the surnames Wu and Chang between
them evenly. If there is only one son, then he shall worship both
lines.

Li shall be supported through the end of her life, and through
good and bad each side shall fulfill its duty. It is stipulated
that when a grandson reaches 16, the Wu family shall be permitted to
call him back to be reunited with the ancestral tablets. This has
all been clearly stipulated by the go-between, and both sides are
willing.

May their children be numerous and luck with them. Because
proof is desired, both sides jointly draw up this document, two

copies, and each will keep one copy as evidence.

326

It is again clearly stipulated that Li will be supported by
Chin-liang the rest Of her life, that each will carry out his or her
duty, and not change in intention, and that at age 16, a grandson
shall be allowed to return to the Wu family. If they live together
for a long time and amass great wealth, then it shall be divided

equally between the Wu and the Chang brothers.

T'ung-chih 11 (1872), 12th month.

Scrivener Chang Nai-shih
Go—between Cheng San-yuan
Witness Wu Shih-wei
Arranger Mrs. Li
Executors Wu Shih-tou
Chang Chin-hsien

Taiwan, Taiwan Szu-fa, pp. 412-13.

 

327

Uxorilocal Marriage Contract #3

 

nnmmédiAEifiskm 3mm

.K . EBMEERZEE .. +1.55 Emmkmfi .. em. finance... Imififi

. EX [mm : Tnxma mmtwamAnmm,

Enema: timwvumw ofifiw MAmnfi mpmaTEAnfi
EfiE @EEW Efififlé fimfifi

.EEEEarEHEEE. sz OEREEa tee

owimfifia .. 023w an"???

“MERE «RR, f:+flfl.kfimu

km 9 vmmrgulonfim

'J'EJ‘E

aE m amtm..

Mfl6w+7mfimalmza 9 fitmfi¢%3_Tmtmnmfibr o Lflurmnfifilt a :mfimm: :mwmfiawurv
rum ..m%.mw.. Wflfimc. .. Hfimm .Acmmi. mining» .. wakefiiflua o

mmm. mfiflmz FEaEE; fimflw
.Efimmm BEEOEkAW
”Mlfimwfim.%aumnn B

REA
EEA
flfiA
ifitA

mnamem¢x

m $Mfl¥a¢mwfi+§mmmuw
“WWL%M1LH115\EXDNMWAI%91

mwfi
mam-
Eiw
fifim
Emfi
Emfi

328

Uxorilocal marriage contract #4

 

Just as the flowers bloom, and the moon is full, it is natural
to find a mate.

Now two surnames are uniting in marriage, and establishing an
uxorilocal marriage. The two sides are in accord on the following
clauses:

1. That the uxorilocal husband shall not take the wife's

surname, but shall retain his own.

2. That he shall provide and care for his wife's father,
and shall heed the father-in-law's instructions.

3. That of the children born to the two, apart from the
second son and the first daughter, who shall follow the
mother's surname, all shall follow the father's surname.

May all the five generations be prosperous, and may they live
long lives. This document is written up in two copies, for each to
have proof.

Husband xxx
Wife xxx
Agreeing to the contract xxx

Witnesses xxx
xxx

Taiwan, sample form from Tai-shu Su-chuang Ta-ch'fian (The Scrivener's
Complete Handbook of Legal Forms) 1970:Vll3.

329

Uxori10ca1 Marriage Contract #4

 

KHH@&6X&.#Wfihmfi% EgfifimE..
.xfiiooomm#k.mfi¥&o
.\OOO£¥EQsfiafia W$§§..
Tania}. 3a fitfigfifififi. figmifiw .
iwxé Hafiz. #fifizwfiahg. .
A. i
x xx
fifiA.
g. a» .a

wgaa. a. .3 .a.

00000

00000

09000:

\

@0000

330

Uxorilocal Marriage Contract #5

 

The executors of this contract are xxx (hereafter referred to as
party A) and xxx (hereafter referred to as party B).

The two parties agree to the following conditions in uxori-

local marriage:

l. Party 8 willingly enters into uxorilocal marriage with
party A.

2. Neither party shall take the other's surname.

3. Party B will take as his own place of residence, the
place of residence of party A.

4. The children born after this marriage, except for the
second son, who shall take the surname of party B and
will continue that kin line, will all take the surname
of party A, and continue that line.

5. Party B has the duties of supporting the parents of party
A, and of maintaining the family's economy.

6. This agreement shall have effect as soon as it is signed.

Executors of the contract:
Party A, the invitor xxx
Pary B, uxorilocal husband xxx

.Agreeing to the contract; the father and mother of the invitor:
Father xxx

Mother xxx

Witnesses xxx
xxx

Year of the Republic, XX, X month, X day.
Taiwan, from HsU l97l: 78-79.

3N

Uxorilocal Marriage Contract #5

 

vmm

nmr:

UAOOOAJTmfiLfiV OOOAuTmVWEVokWAfmw

.szfi.fiu Afi. Wfifimo
z~$3§fiflflxflflo
Esafiflmfifimfimmo
m,m&$fi¥t.ma+iafia mncmumn.mmwmfiu.wmmwmm.

n.5waflmmfixozawxm. wasznnn.

fifiEfifiEflfi$fiéflo

.nmmex mfimrx 0
awxax o

0
0

m A m fl,A w m 30 0
0
0

W

m . A 0

w fi E m xwfit xfl . x8

c>0.00<0<2
El@|@@ @Ij

W
m

332

Uxorilocal Marriage Contract #6

 

The executors of this uxorilocal marriage contract, Mrs. Ch'u
of the Hung family, and Chang Tsai-sheng, venture to hold that a man
who has grown to adulthood shall marry and a woman who has reached
maturity should have a mate. Hung, nee Ch'u, has a daughter, l8
years old, named Lien-liang, who has not yet married. Thus through a
go-between, they have invited Chin-lan, brother of Tsai-sheng, to
marry in uxorilocally, to go through the ceremonies and become man
and wife. On this day, in consort with the go-between, a price of
124 large "dragon dollars" has been fixed. On this day 24 dollars
shall be paid over to Hung, and the remainder, l00 dollars, shall
be deferred until Lien-liang's mother, Mrs. Lin, returns to Taiwan,
when the account shall be settled, and no other arguments shall be
entertained.

If Mrs. Lin decides that she does not wish to receive the
money, but would rather be supported and cared for by her daughter,
then this too shall be permitted. Thus she shall receive either the
money or support, but not both, as one should reasonably excpet.

May their children and grandchildren be as numerous as grasshoppers,
and may their fame become widespread.

This is willingly agreed to by both sides, and neither will
have regrets or recriminations.

Fearing that there is no evidence of the spoken word, two
c0pies of this contract are drawn up, each side to keep as proof.

Hung, nee Sh'u, has this day' received, through the go-

between, the full sum of 24 "dragon Buddha dollars."

333

It is again stipulated, that should Lien-liang's mother wish
to reside with her daughter, and be supported by her, she shall be
cared for, clothed, and fed in life, until her death, and that then
Chin-lan shall take full responsibility for her funeral expenses, and
her soul tablet shall go to the Hung family. This is clearly stipu-
lated.

It is again stipulated that should Lien-liang's mother not
return to Taiwan, after many years, it is humbly requested that it
would be fitting that the child born to Chin-Tan and Lien-liang be
equally divided between the two sides, and that if there should be
only one son, he shall take on the sacrifices of both family lines.
Moreover, it is agreed, with the go-between, that a grandson should
then be returned [literally "married back"] to the Hung family, in
order to continue their posterity, and it is repeated that there shall
be no second thoughts about this later.

It is again stipulated that after Chin-Ian has married into
the family, he shall pay over six silver dollars per month to cover
the daily expenses of himself and his wife.

It is further stipulated that the Hung family's buildings,
property, and household goods all belong to them, and have nothing
to do with the Chang family.

28th year, 8th month, of Kuang-hsu (1901)

Scrivener Kao Jui-ch'ing
Go-between Mrs. Hsieh née Lin
Witnesses Uncle, Chang Shih-miao
Uncle, Hung Wang-sheng
Executors Chang Tsai-sheng
Mrs. Hung, nee Ch'u

Taiwan, Taiwan Szu-fa: III, 4l5-l7.

 

334

Uxorilocal Marriage Contract #6

 

A? Awfio
Eflfi. .Law :A BA :5

’mmgmo

fifluw. :mmvigwfiu” www.37\ wand/W ’ExH/éH’fifi—HJ. dim/mumfimflxl!

AAAAAH.AAA.

A A
A A

AAAAA

Aflz+tAfiAAH a

fiifi

Wflmq afiAan+A

HIE BE). i153 F55

*y#%
H
fi

WAAJA NA} AA AAA A.AW%AAA Afita 1wz-¢o
AAAAA§A AAA+A A ATA.AAAA.AAAwAg.rAmnmm
.AzyéAAA MAAA.#AA Zalyzmmm.m¢g agkfiyAq
AOAA aarwthmAAAAAmr.ArAamAA.Afla: AT
Wfiyfifiwflfi Km A3:%£#RTE&EQ.T%W#¥WQT Mfifim
owiflflxAflf %:fi #.fi, ”3:.Wz :fiMfim MMH+AXAoA
Azmfim Av Km n%A@.HnPfiéz?lm€5.$Afimfiflo
BathWA¥Hfi$WK+¢MAKAHAHTEA AG
AAA vAAAAAwfi€m A AAAAA AA AA:5 AAAAA
fififmwrW/m .r.1/..n...$fi¢~m :mmmwmmo .
. AAA: miBfiAfLAKHW Rm AAAAAzL.AA$fifi;AW
my? Afimxw%.m$ HAT(.WAEAEA, AAEH AMfifi.mw

AAAAAMAAA

Awfio

ilflffi$
Afixg"
Awt .

35 E374

Eéaé

335

Contract for Pawning a Girl

 

The executor, Ch'en Chao née Chang, of Tainan . . . has a
daughter of her own, the second, aged 14. Because of lack of money
she now wishes to pawn this girl, and so, through a go-between, has
agreed with Huang Chien-Chih, of Tainan, on a price of 30 silver
dollars of 7.3 weight, without interest. He, Huang, shall be per-
mitted to teach the girl how to sing, and how to dress and make her-
self up, to be a sing-song girl or prostitute. Huang shall pay
Ch'en five tales per month out of the girl's profits.

The girl shall be pawned for a period of four years; within
that period she cannot be redeemed, but after four years Ch'en shall
be permitted to redeem the contract for 30 silver dollars, and the
girl shall be restored to her parents.

Should anything untoward happen, this shall be considered
an act of heaven, and no affair of the pawn-holder.

The two sides willingly enter into this agreement, and
neither shall have any regrets or recriminations. Fearing that there
is no proof of the spoken word, and in the desire for proof, a pawn
contract is written up to serve as permanent evidence.

On this day received: the sum of 30 dollars, 7.3 weight.
Twenty-eighth year of Kuang-hsu (l902), 2nd month and 13th day of

the lunar calendar.

Middleman Lin Lien-hua

Witness The girl, Ch'en Mei
Executor Ch'en Chao (née Chang)
Scrivener Neng Chin-yun

Taiwan, Taiwan Szu-fa: l80-81.

 

336

 

Contract for Pawning a Girl

EEEEEEA.J EEMEMEEEE+EEEEE EE EEEWE
.E+EE° OEEZEEE mmEENEEE.E¢mmmEEEE WE+_E
Emawta EEEE EEE,L EEEEEEE,EEEEE muioEflE
.,EELQEHEEEEEOLEEEEWEEE;EEEWEEEE.EWEEE
EEEEEEEEEE EEIHEEEOEEEEEE.EEEE.EEEEEO

EEzw EE.Eer.nEEH.EEJE,EyEEEE Em. HEEEO

Ea. LtaézE Em EEO

EE 2+AEEmzfi+EB . .EEA EEE
ERA EEE

1mm fins... immiA fimflwfi
EEEA EEE

337

Contract for Pawning an Adopted Daughter

 

The executor, Shih Lin Shun, of Tainan T'ing . . . has an
adopted daughter named Chiu-chih, l2 years old. Herfhther-in-law
is ill, and she has no money to pay a doctor to cure him,so she has
asked a middleman to help her pawn out the girl. Huang Tsa-bo, of
Ma-tsu Kang, has agreed to loan her 60 big "dragon dollars," and has
paid out 30 dollars in advance. The remainder, 30 dollars, is to be
put in escrow with the pawn-holder, to be retained until Mrs. Shih's
death, when it shall be used to cover the funeral expenses. The 30
dollars has been received in full, through the middleman; the girl
is turned over to the pawn-holder for his management. If she does
not obey his orders, he is permitted to chastise her.

The girl shall be pawned out from the fourth lunar month of
the twenty-eighth year of Kuang-hsfi [1902] to the fourth month of
the thirty-sixth year of Kuang-th [the reign title "Kuang-hsfi" did
not last that long, but the year indicated was l9l0]. After this
period, interest of five fgg_per dollar per month shall be paid.

The usual fee of three dollars per month for teaching her to sing
need not be paid.

When the capital, 60 dollars, is returned, then the pawn-
holder shall not dare to make any trouble over it, and shall allow
the girl to be restored to Mrs. Shih. If it is desired to redeem
her before the expiration of the contract, then if the money, the
interest, and the presents given by Huang are all given back,

she shall be permitted to be redeemed.

338

If something untoward should happen, then both sides shall
consider it an act of the creator, and no affair of the pawn-
holder.

Both sides enter willingly into the agreement, and neither
shall have any regrets or recriminations. Fearing that there is no
proof of the spoken word, and desiring such proof, this contract is
drawn up, and sent with the money to the pawn-holder for later
evidence.

It is stipulated that: The girl can be redeemed during the
time-period, and she cannot be sold to work as a prostitute for
anyone else.

It is stipulated that: If she is able to work as a wine-
house girl, her mother shall be paid three dollars per month; if she
is able to work as a prostitute, the mother shall be given three

more dollars per month, six dollars in all.

Middleman Hung Chu

Witness Shih Tan

Executor The mother, Lin Shun
Scrivener Wu Mei-ching

Twenty-eighth year of Kuang-hsu (1902), fourth month, twentieth day
of the lunar month.

Taiwan, Taiwan Szu-fa: 18l-83.

 

339

Contract for Pawning an Adopted Daughter

 

1%mfi$A. fifkmflfi Wt mmfimd+kfifiwmfiafi.aiii
Lg m%.¢§ufiro4mairmMfimg.mmréia.L§mv mm

+1.1...“ fawn”...
._ W2. 31K ... ...Ju a». x ts ... ... .UWRhtrTixs
mk7/PU1/nAMFhamW/Hufi snmlfiww ......lflm ”MMJAAHkBH; ’Fmi “MM "I’LL/”*5.” o. ....7MP?.W1PL .1....L.n.....~m..w

it? fifia mkmm.%. whiz? #firakifimfiwa;tt§¢fl
$fi.$r;gxeal mmiv-or ammo wniinmne; wz+
nannes.mnancmWhfinfiraomha w an raw amp
owammnnrq canto,L.jt:lmammuogamexamu.m
memkmea.mumnw.fiemmm n.r.zm mazmfim mm
imfommzm ..4nam ug:t.¢mna.emmmm¢aa.mxfi

ieeazem.
L~%fi. .L..mmm.
L.fifi.. ¥fit %#
nglx Mmfl MWMM o

mam xwmwngfle.mao
flamngrwea;%maefi fiancee.

Ed

m...
ffl

eex & x
.fiix mamm

.. . . meeae amaame
%%:+Amfi%mfiz+a ‘ fieA fimm

340
Contract for Pawning a Simpua

The executor of this contract is Ts'ai Niu-kang, nephew of
Ts'ai Tsung, whose ancestors came from Chang-Chou, Chang-p'u.

The grandfather moved to Komalan, and had no business or
profession, and Ts'ai wishes, with the permission of his uncle, to
pawn the uncle's simpua, aged l3 years, born in the eighth month,
13th day, at the sheg_hour [which is the time-period 3-5 p.m.],
because they lack money.

He wishes to pawn the simpua to the brothers Hu Lien and
Hu Chu, to be an actress. The time-period shall be six years. The
three parties agree to a price of three silver dollars per year.
The money has this day been paid over, and the simpua shall now
follow the orders of the Hu brothers.

No hidden conditions shall be added; should complications
arise, they shall be the responsibility of Ts'ai, and not of the
pawn-holders. Fearing that there is no proof of the spoken word,
one copy of this contract is drawn up to serve as evidence.

It is stipulated: That in the intercalary eighth month Ts'ai

was paid four foreign dollars.

First year of T'ung-chih (1862), eighth month, twenty-eighth day.

Scrivener Lin I
Middleman

Witness Huang Li
Executor Ts'ai Niu-kang

Taiwan, Taiwan Szu-fa: l9l.

 

341

Contract for Pawning a Simpua

 

I} 7?? ..hP
- a o. ... lsh.) II 440') .‘C‘ .4 \.\ \
31.. afla.L.fl....~x urJfimwur... Liz. ...E..m.+: :wfi... tfiffl.

sun.“ 3
5.» Or a I .. w.

mfifir£flflflflm xz+.ag¢fiz $.1LLHWL
Q£¥H fiflLffi.a:Wa §$W%Eikfifio .WB.a

.mnxlnmxh .umnom WLM$13fl+fiTfi$ng¢kié W...
:flfii. “mm in: 6.30
WLLzmnm; LLL LL Liao
LLLLAE2+AB..

.LLLLL LLLL LL. LLAL LLLLL .LLLL
.91..

... 3....ka ...w

L #LL
w LLL
.A

LLL

342

Contract of Re-sale of‘a Female Slave

 

The executor of this contract, the Ch'en family of Mao-kang-
wei . . . openly bought one female slave from the Yen family, aged nine
years. Because of incompatibility, they wish to re-sell this female
slave, and through a go-between found the Kuo family . . . to buy her.
The three sides have agreed on a price of 60 large silver dollars of
6.8 weight. The amount was handed over this day, and the female
slave is hereby turned over to the buyer to be their slave, and to
be renamed by them. If there is any incompatibility, they are
allowed to resell her. If anything untoward should happen, it shall
be considered an act of heaven, and no affair of the buyer.

Both parties are in willing agreement, and neither shall
have any regrets. Fearing that there is no proof of the spoken word,
and desirous of such proof, one copy of this contract, and two copies
of palm-prints, three c0pies in all, are prepared as evidence.

Again it is stipulated that the complete sum of 60 large

silver dollars, 6.8 weight, has been received.

Fifteenth year of Kuang-hsu (l889), third month.

Go-between Mrs. Chang
Executor The Ch'en family
Scrivener Tseng Yun-chih

Taiwan, Taiwan Szu-fa: l70-7l.

 

343

Contract of Re-sale of a Female Slave

 

LLLLLA.LLLML L .LMfl LELLLLD LLLLOLT
LL.LLLLLLL ELMLLLJLLLmLLLLF.ELLLLLLALL
# LEOLLL aflfiiLLfizjfi hmafi :LLLLL. .ELLLLL;%
LLLLLLLLO LLLLL. LLL. LLLLLLOL LLL.LLLL.LL

.. . v . ....MM
fifl.4$fi3.éflfififi$#LgfiL$,L.%#fifi Hflfi o
LLLLLLL ALLL .LLLL. .LL. . LLA L Li
mwfiw+43m+sfiau nu . LLL;+%W.$1A mw ,am .
RnwA .QEE

344

Contract of Sale of a Girl to Be a Slave

 

The executor of this contract, Wu Pen, of Tainan, Feng-shan
T'ing . . . has a second daughter, named Ch'ien, aged five years,
born on the fifth day of the second month, at the mg9_hour [5-7 a.m.].

Now, because the family is poor, and it is difficult to get
daily meals, and the debtors are pressing, they wish to sell her,
through a go-between, to be female slave to Hsfieh Wang of Tainan
City. . . . The three sides have agreed on a price, and it has been
paid this day, through the go-between, and the girl is turned over
to the buyer for his use. If there is any incompatibility, he can
re-sell her.

Should someone kidnap her, the executor shall file a complaint
with the court; if he does not, then he shall pay back the money
that he has received, in compensation.

If something untoward should happen, it shall be considered
predestined. Moreover, this is the executor's own daughter, and
there will be no trouble about frauduently dealing with stolen
property.

Both sides willingly agree, and neither shall have any
regrets or recriminations. Fearing that there is no proof of the
spoken word, and wishing to have such proof, one c0py of this con-

tract is written up to serve as evidence.

345

It is stipulated: That the executor this day received,

through the go-between, 38 large silver dollars, of 7.3 weight.

Twenty-eighth year of Kuang-hsfi (1902), eleventh month.

Go-between Ts'ai Nien
Witness Wife, née Lin
Executor Wu Pen
Scrivener Ts'ai Ts'ai

Taiwan, Taiwan Szu-fa: l7l-72.

346

Contract of Sale of a Girl to Be a Slave

 

yfiawuaeA.ammmmxfiememem+iafiee.aaaaezt.
;%:%.zw.¢ia,:nmfiaww¢oemgflfiw.gang.m$afiw

ll. . at... J... 7 a ...... ..FPI. Sun-...! vIWri
..E . “mktvrfifiawfimaflmn?E3fi$fii+®fiPWfififiw =:m.az ,

mammuwofit4waen$meem;wxam.mxnfimgofiamAww
.mmmga;%exkfie.fimfixim.ammmm.x¥fiizasnaam
mm.nexmoeefiflézt;.mmmmMAwmemownzmae.afia
maamm.uefim.emam.ayufimaaeflm.ememo
nam¢ueemeameemxa.mo.
.%%:+A$+_H a ._ : .emA fiAma
_ fiEA afina
mfltmameA a a
RaA fl 2

347

Contract for Sellingga Wife

 

The executor of this contract, Yang P'ien, of Hsieh-lung Pao,
Fan-a-liao Village, bought, by Open transaction, the wife of the
brother of Huang Ch'uan-hsien of Hs'ueh-chia Pao, Chung-chiu Village,
and brought her in as wife. After entering the household, she re-
fused intimate relations with him. He could not endure the circum-
stances so he has asked a go-between to introduce someone else, and
has contracted to sell the woman, Tseng Tuan, to Yang Pin, of Pai-
t'ou Village.

The three parties have agreed on a price of 68 dollars; the
money has been paid, and the woman has been turned over, to be a
concubine. She shall not obstruct the transaction, and shall not
make any trouble.

It is guaranteed that this woman, Tuan, was bought by P'ien
in an Open transaction, and has no connection with others which
could cause hidden obstructions; if there are any hidden circum-
stances, P'ien will take the consequences, and the other party shall
not be liable.

This is agreed to by both parties, and neither shall go back
on their promises. There is no proof of the spoken word, and as proof
is desired, this document is drawn up in one copy, and a prologue, one

cOpy, two copies in all, to furnish evidence.

Thirtieth year of Kuang-hsfi (1904), second month.

Go—between Yang Sheng
Witness Mrs. Hsieh
Executor Yang P'ien
Scrivener Yang Hsin-sheng

Taiwan, Taiwan Szu-fa: 406.

348

Contract for Selling a Wife

 

nsugrAmmfit fiZk.aHMflmmmfimmmemfia%znfim
see.Arm2.mfixg. # m:.mmmmm§.ammm.meazmm
#Aflfl$fimmuwflméfifi.zr.mmymmxmAkmF;%mwafiwfi
flAmEMHui¥3 £7043 Ammm.nxfl$a$WLTOQMfl
§;#fiT21 mu m+.ifiamim¢&;mfimmmnpknmmih x
faizmomtz um.¢mm ummm.emam.na. ems ¢_m a:

aeflm. u:m.wugm
.eanfxorfitWfixm sz&.m .
mma+¢+zfi 3 @mA m &
A.. . . .mRA w a

E

k

teamwxsmnfi
“kaxmwra

349
Contract for Pawning a Girl

The executor of this contract, Cheng Szu-yung, of I-jen Li,
Chung-chou village, because of family poverty, wishes to pawn his
daughter, Yen, to Cheng Tan-chi to be a slave for 20 silver dollars.

It is guaranteed that this Yen is the daughter of Szu-yung,
and that she is l4 years old. The relatives were approached first,
but were not willing to receive her, and so she is pawned to Cheng
Tan-chi for 20 silver dollars, for a period of six years. She shall
be returned to Yung upon full repayment of the money, and Tan-chi
shall not be obstructive.

Fearing that there is no proof of the spoken word, this
document, in one copy, is drawn up to serve as evidence.

On this day received the full sum of 20 dollars of 7.3
weight.

Twenty-seventh year of Kuang-hsfi (l90l), first month.

Go-between Sin Te-chi
Executor Cheng Szu-yung
Scrivener Lin Ch'eng-chi

Taiwan, Taiwan Szu-fa: 357.

 

350

Contract for Pawning a Girl

 

£.ZW

amt;$Amtmmwfimmtfiwfi EZfim &.%mm§tflz&.m
kmenflfiwm .fih mmsxmowwarflnzietE.mfi+mmok
AEWERE Twanmmm -mmtsm m:m MWEEE .wm mfinm

L4

fifi fljwh ou%m3.$%fim HfiEmo
EBBEfifitEfiIEfiWE EMo

%E2+t$mfl a EMA EEE
axmflfic+w+xn .Lfimrr
.REA Emfl

APPENDIX B

CASE STUDIES

351

APPENDIX B

CASE STUDIES

Poa* Case I (1)

‘7‘] .......... 1333. 6
W Bu-l'n

mo

mag .
pm. (khan-bolt
.‘39‘5 (simpua)
Phoz'Tan A-liang
IQ" @

pm‘ A-hoan ‘A‘;~ A- 51m
'77" m3 Khu A-bai '

Phoa' Lim
Inna-bun an;

Khu W3 A‘s“
1211-
KhuM-chin

no

    

Phoa' A—e

 

 

..‘3‘1‘ Afinor'flocal);

Liau Boathong M39

cm

 

352

maAGMQtimatQ

 

Mm A-‘m

Phoa* It-Mns

nu Phoa‘ Unit-gig"
.q ( o deca‘
Agezuam) mfi Phoa* A-hoe

Phoa* A-chhlu (meg‘fim)

%L\au A-b5k

Liau Sek'ho

Liau Siu-un

fl® L‘au SW.- Ch“;

353

Pan Case II (1)

"“5 95.. A4633

P'an ‘IW chi}.

l'hn Wan-ch'an

rim Lm A-yeh (“MP“)

 

K30 Niang- Mo (um-3 had)

7.3» Chm-Ch'ien

lt1| L
4 \Q!‘ PanT mu-” '4“ "a“ Tsen9-h"‘a
Ch'en Ya
Pin (hug-ch'i“

 

 

 

 

 

 

“13 f! P . .
an ”"Imfi Pfin Wan-ha
Lin A-hou
—- P‘an D's “as (simpua)
231°. y. Fen "m "a“ “m
Igl'

P'an A-§s‘ai

 

“IO!

”30 A I’lan Wan-ch'an

 
 
 

 

h--- .‘39‘:

o P'an Lin A-yeh
(Sim pua)

non

A K30 Nians-hvo

(uxorilocal)

P'an Chin-Ch'iu

 

ms 'in Tseng-Ma

 

‘A® P'an Hsienshao

 
 

 
  

    
  
 
 

  
   

354

Pan Case II (2)

fig?“ Fons-me}
mu I -
I!“ "n Chan ha
we,“ Him-thin
(mrr'ned cut)

I «6 ' .
MP.“ Wu sung
‘|_41]_®P'an “sin-{eng

(adefledow)
.flfiQan Su-iien
431% r'an Su-mei

(3:10pr on?)
_A"‘“ Pig“ Ming-hung
man Su-y’deh
.L‘DLAr'u Ming-M

UQSo

"so g
n

.Jq—S‘LAP'an K'un-hsians

£E.@r3n Hui-wen
——-A““" P‘an Khan-me}

“5" P3“ K'un-+3
m1

r5" Shea-«MM

  
 

r'an Chen Nuan-wu

 

 

P'an Sta-Ch'ing

P'an “sh-Chu

 
 
 
    

n31 . P'an “sin-131

\qu . . .
“' fan “SI-Shut

 
 

'3‘}. P'an Ch’en Hung-i53-
(simpua)

EEL® Kao Ch‘eng-‘i n

'—"'—® Phn Ch’ing-hs'lang
, i‘ "an Fu- \I'I
W01 0 Chang Su-fien

I131 ' ’ . '
Wat’an Chan he"

P‘an 191-Sheng

an She-13M
.326.) P'an 5MP "“
[‘62 ® P‘an Shu'ch'i.“

   
  
 

 

  
 

‘“ o P'an Pooh“?
A P'an 96““"3

   

lan “SN-me"
(simpua)

M P’an Wan-fien
flA P'an I'hs'umg

«a
mg

m@ I’lan HsRu-yfin
fill-A f'an Fu-yung

.. . m Kuan-ms
NBA P‘an Kuan—chieh
.flA P'an K'uan-ckeh

 

 

Pan Shun-+e

 

 

3% P'an Ch’ing- {ans

'5» Wan-flu

 

 

P'an Lai Yin
“‘1 (“my“) 3

Jflfi®f5n A-hao

 

 

 

355

Pan Case II (3)

#0 ('3! “En-shah
435.6993“ wen-huh

mes - . '
-—Z:> P n Kuan-““30
“3° | P'an Ink-mac a

M“ H Li Poo-kuei Imcaf'an Wen-h“

 

 

 

. W" 2. P'an K'un-shen [fly—G)?“ Wen-‘3

 

 

HS! 1 Liao Ina-1mm 7'3“ Wen-me“.

356

wu Case III (1)

Huang Tien-szu . .

  
  

"“303 Chin-mu

 

0" Lin A-lou IIS‘I 1‘9! 13': A-sh‘u k

(W; uxorilocai)

..——{“‘°3 Xm I-cweng

ms Tai A-stu

 

1817

 

Lin YA?

II‘H- . a
W0 Lm Ch en
”15 Huang Li Wei

 

Wu Lao -mia

{$91- Ch'en Shau- ch'uh

\tfli

P-O

  

Liao A-y'd.
(stmpua)

 

Wu Chin-she
(urorkloca t)

A Ch'en 5* me} (Elks;-
4W§e5 reusing“
“Op-I'M; man-fied)

 

 

,' :12" Huang Afisung

= " (hang lion-mu
(WI-Elna!)

 

HI! - ' - _
---i in A—che (mus
|i‘:- a; margin!
0 uxorilocafiy)

357

Wu Case III (2)

 

 

q

ales: 2. 23012.32.

3133 52-229 :3 gnaw...

 

 

9.3-3 5.6%

 

5.5-955 $36!}:

€23 $1..
57929 :3 €54

 

2:75 a? 3...

 

 

 

95: 36 53%
93:: -23 :3 /.\1~.. m! ‘

2....
.2...

 

 

35-... :3
9.3.5.5 339%“!
man..£.< 93:5 :2

   

 

I!

C3... verse: .2:
«.... gang .5... 3.3: s3 mmhn
“8.3 5.25.: m...
3103 3.3 as .iOfifi
A3833?» 5.5.53 :5 ...muan
As. 31.3 =§.< =3 3.1
«£75. «5.5 23. v ....» .< “a...
me.
I... 0.72.: :3 ES 95 3...... .3...
.- u. - s mfismuqfivnsbw my! 0.30 :Efiv 5 .Su 3
099.:me 3} .5 ,

358

Yeh Case IV (I)

  

 

 

  

(uxorilocal)
MG” we.
(nuuvuuioufi>
flflA‘leh vans-ch'an
WIS“ ‘Ieh (Heng-tuna Yeh ”1“
Yeh A"; A 1 ad:fl:egrt,
fig Yceh We} A-ch'd mu ”3"“ ”0
sfim ua ‘ -
Yeh usu chm-ho P > “(5)“ A ""3

 

MIN .
»--- v h A-fi
<9 as... :2; m I...

(mm-Red out)
pm K'un-k"!

ried out)
flEEAYeh Khrsung

 

359

Yeh Case IV (2)

W LIn flan—5m

MA 1.3.. A-ch‘i (married out, uxorilocauy)

£6)” kuei-mam (married out)

,LQL“.@LM Kue's-hua (ado'find conga: L‘ao

Yeh Luau-flying
“no 'qgg 2 Lin T'uns-KM (a.k.3. LI“ Shih-k'un)

Lin Ch'ing H01

(uxordocaD “3‘! @Lin ‘m (3493- LinSu-‘hu; adopted 00* It {\mpua,
la‘kr returned heme)

5%‘4‘ TWS‘CM": (34:3. Lin ShKh-dnaug)

flAun Chih-chien
”4% Lin Keel-Ian
HM

,% Lin A-mu
\14!

 

P--- ----0----

360

Yeh Case IV (3)

Yeh Ch'u-
M‘n- IQ b;

 

‘ MIN ‘Ie H M- M'umg
moo ‘Ieh Pi-dn'dm
n52. Lm Mei-“n3 (3W daughter)
, I '.
‘ as q Yeh She the»
Mn

‘Ieh Cheng-ch'u

£31691“ Sun-hsi:

 

 

 

I ‘ch Linus «chum

“5 ‘A Yeh Cheng - kue"

\n;

 

Mung (norm 00-0)

m.®‘leh A-

Cw; P50 hiookuea “1216)“... Swhua (”\egm'nfle)

 

 

Ch'iu. A-chu
no; : ' ? ’iu (hag-suns
0““ WI "ion-mg"
(33mm)
‘ u H
Wang A‘ym
(moraumm

 

 

  

dtvorcoa)

1 Yeh K’un-suns

"%€h’iu Ts'u lag-Man

12% Ch's'u Ik'uen-ch’m

film Ch't'u Rafi-fins
H55
——@““ Ch'iu yea-m:

L‘EIA Ch'iu \Is'oeno drum,

 

Ch'iu \k’uen- duh:

m”%ven Cheng-m3

 

HI!

Yeh Tu kin-Chih

“‘5‘va $94953":
AIS—(91¢ Su-chen

 

[M‘leh Cheng-“n

Ch’iu Ch'ing

-piao
I!“
H

liao Lian, ‘ 61

H”

361

Ch'iu Case V (1)

Ch?“ C h'i-yuan

Ch'iu Owns-sag

Fu__®chh A Mia)
(fllflied 00")

‘ ma (SCUM Slum-fie
W31: “sin-Ian
(“0912‘ out)

PAM? Ch'iu Ch‘in,-ch'in

MANN (Wig-huan,

"3' Cfiiufl'a'ng-‘ien

%€€iu A-dv'ina

““3 Ch'iu (Nah-ha

 

 

 

Ch'iu "an, “um

ch‘u Yang A-
Hang
(‘in 910)

 

1%Ch'iu Vin-cans

 

I!“

Ch'iu WW shu

C .
hiu Chung chum}? 95 Ch'iu A-pii

 

  

 

 

 

  

--- chia Lu Wen y «m cw}; A-kue’t
"‘5 (s‘mP‘i) ——® 030”“ M)
0m gm '.° l' A-w '
Mfih‘lgmgw on? ' $25.31) ea
WOVI|OG\N;
see Phrase)
.31. Was Lin A-{m H15 Ch'iu A'h‘fii'fi
(fizfiedow)

\qao

Ch'iu shun-Wen
(mega; mate)

362

Ch'iu Case V (2)

| Ch'iu Shanda [HS—SACK“. Nerf:
a 'in Va. Ya-m Mum Mei-ma

(mag Hunk
magnum! e5

 

 

 
 
 
   

 

I Ch'iu Sta-136‘

Yeh Ch‘un (ineg'fi'llm‘f)

LVL'G Ch‘! u Ch'un- Inc's
"“3_@ Ch’iu Tai-h
a. - - Liao 7343“

‘ .0
0:" A 933 (s MP“) "“5 Ch'iu “such-Ma
In '
MN! ‘ Ch‘a Shun-hm

‘ “NO ACh‘u'u Shmd’icn

 

 

 

(mama 0“) nan .. . _
Ch'iu Lin Ame; ~ Chm than «ten
““"P‘W ....‘gmh Ch'iu Shun-ch'ih
IQ'.‘ .%Chh‘u ';“,'k'ans
Wt Ch'iu Shun-I33
(adored on

 

»““ mum-m

363

Liao Case VI (1)

Lin Shun

Hcii Fa.
(simpua)

 

LiuTing

“sick Tsu

 

it“ u... Clans Ch‘ing

  

He'd. Ch’a“

 

 

Lib, Chit 'm ‘ chi-
LiuChien ma ("“9“) m: L” "‘5

 

It“ “an, Ch‘ing
th’m Hei

Line “0

 

Chu Mi

 

Li Lin
(simpua)

 

Li» A“

 

H65 Liao Mao

no: Yeh ‘Iu-mci

HIE-@Liao ”Ch‘iu"

(married 0M)

1% Liu “Bani mi
(named on?)

3511 Wu Shea (simpua;

adapted 005M";

married out; divorced;
returned; married on)

 

364

Liao Case VI (2)

 

Liao Sim-
mu - “65

win um . ,
(slay-l) “A Liao‘Iu-iu

Liao Shu- - 9 - --°’-
(Koo-fig 1'- m)

A Liao Ta- “in;

 

ti

\m ch'u mei pE-A‘Z'th

 

 

 

  
    
 

u 9 I ‘
nu Chan ‘EOEI‘I‘Q;A;;I;M- -f¢

”@qu “flu-1""!
Liu 9‘? mfgLiu Sim - - Q - - -6 -

.--—-..-e-us-ee

-..-.. -¢-____-‘.u
4-- -- 9- ---’---

 

‘5‘

-Q...
..-
q.
...a

Li» L33 “0“ I. W
l—A ' Ibii he .'

"'6 3.; Lil. TO-WJI-‘- '(‘“’ A “a . . ."'

b Lia his!» 1 ung- °

chia km?“ a
“M u) l—ALiu Iain-m
‘ Liao Ta-mu

m Li» Ch‘iu-hsien:
1 1| LiuTa- k‘un

w. Ch‘iu-ch'u!“
" (Sin-pa?)
we... “on

(MON 00*)
‘-'Ou” A chm

(3“,‘9‘ ow.)

 

 

365

Ch'iu Case VII (1)

 

 

 

 

 

 

 

 

 

 

 

BRANCH
r—A Ch‘iu-nu |
”'A Ch’in. - pie 4
'h—ACh’atn .731 5
Chin- km: 6
Chin-Mia” 1
.. Mung-Shag 3
Want - using q
r-A‘R“ ohm'h‘i“ Ath’i‘u szhens (Heb. hung ‘0
' j” A Ming-ho u
Shih-Chu; H.
flung-Lung t3
Ch'iu “Ii-Kai Ni 4!» H
A__H w_A;..I..Wmngrim H
r—A man, . k’ai u.
—A Niang- +s'ai n
"AW-chin I:
ALAQ‘W "Oi-she; —-Allian5 453 H
—A "Bang-‘90. 19
. . ‘-A A- ch'ing n
—£‘£m M" fi’fi A-ch'm n
"A A’ ku 13
and“... (Emu-nu a #93» A- hm: "A A‘ Y“ 1"
—-A A-mei. 25'
, h-A A-ch’un 1‘
"Am—Hi“ hue“ 5 Ch» - ch'oh 1"

 

366

Ch'iu Case VII (2)

«...»...osx :35 ® 3::

2......03. :a..:u@ 3'.

 

 

 

 

 

932.103
9.29.3.3 :5 4.. «any.
nonmasimgx BOEVA ”v hm... 5‘033 _
0.. sum) :> 0 puma...
9.5...3. :35 3... «vote»?! ....u.< .13. a»
. «1.3.320 «358.....3» .3.
, 5.5.8.. 3.5 :35 . :3-< s»
O O
7:303. 3259:. I «.33 ....ousu
‘ 93.35 52.5 man...
3.5393: . «Etovauown. ........... .... i i - .2. .
aria. safe .3! ..s....< 99.5 «flflauw cw...

rut .

.3.
......< 5.6 3»

 

mac.

5.6.3 ..>

367

Ch'iu Case VII (3)

L‘EAch'm Lien-ii

.MAch'iu Lien-dim

‘Lfl-ACW‘“ Lien-fan
W31

‘ NS? 4.3 Ch'iu Lien-hung
IQSC

‘j§‘|_ Lin Cheng A-yiin MGLM Moi-ha!

(34 6
Lin A-fia 0?“ )

was-ms “"3 sin mu Mu-ch‘e --A"“ L' Lien-shu.
(uxorilocal) .3 m
Lin Lia 4‘3“.“

M03 - "
gym Lin Mia-mi
(adopted) (altegiiimate)

 

 

368
Ch'iu Case VII (4)

5. 2.12.3
Asia...» 3.50434

+2. 92.5on 0 .
grip-s: ......3 5c. A0325
...-3.. 53¢:

nasaeau i I In.

57¢ :3 3.5 r...
«9.3-5.3 .....5 2.: ,

 

 

 

at... 1.3.3
:5... .....5 v.5 €593.38...
A as... .. ..oE...a..... «:33
S_.< .3 3.5 a .. T’s ..os.-. ..
185.2.» mafiimsi 3.5 .2... “39:3 .0... 95 x .0878... 9.8.:
... 00 .15.: . u I u .
«Limbs... 3136 i i 5.... «51. ...... 5.. 9.05...» .... "HM
«3.5.23» «5% ......G can... ....vsiocoas ......
3:... ......a . .

 

rah «$.80?! A0553 gflwv

. £0.51
at... 3.5.3.5
“39.93:... ......3 10...

«5.3.5.5 ......G
3...... 35$”

   

 

 

369

Ch'iu Case VI

 

 

 

 

 

 

I (5)

  

‘h'juTai A-chiu
“Imp“; divorced)

 

r—fichhflnnafig
gap cu’m Huan W i
(simpua), Y a,

99-, cn‘au A—‘en (adopted

in; married out)
5|}; Ch‘iu ‘I dun
M" (I‘M-fangadofl‘uon)
“1" “(in A'yu
(sompu)

("‘9 3:223:23; A“
~ "w Ch'iu A-jih

' ‘ Ch’iu “flea-{u
use @qu Lin A-ch'un

" 41 NJ
‘ ‘ lb ‘33 Inna.) (“0P
__|3 Ch'iu Hsieh A'yu
(sh-pun)

 

 

Ch'iu A-‘hmg
"n Ch’iu I'mci J—
.. - W"
Eh at Lin)fluan __
“up“
mq- «m :31; Ch’iu. Hs‘é can

Ch’iu A-ch'un

 

._A"" “‘5“ “ck-Wins

 

 

Ch'iu (Men Tu.
(“my“)

 

mo arm Huan, A-yiian
I” ’ ( Simpua)
MN 1. - -
"‘1 C. II Shh-chm
' 1| Ch’i; A- Villa
J!!! (“’3‘ L“. I
(simpua)

H_M_!_® Ch’iu A-hao
(married om)

 

fifi®cvca cm. an. (aorta! m,

r“ Wudgmuriu out)

1 'in Ch'en Shut-bu
M08

370

Ch'iu Case VII (6)

P931 Ch’m A- ch'i
(3409 {ed In;
mama! om)

mLGCWiu mama
(mart-Bed ou‘l')

 

cm" I -
‘1 «'3‘: a" m

 

 

j“(‘33 Ach'm
(simpua;
divorced)

W“ Ath‘iulu'n-p'ing

Ch'iu Wu Pi-d'm
n :4 (56mm)

 

 

“‘16) Ch‘iu ‘Ifi-ju

 

 

(k‘un k3“!
UQ|O

Ch'iuCh'on A-suu
mo (sir-P03)

Ch’iu A-M‘ao
“'33:! (adapted in;
man-fied on?)

‘ l .. .,
'figflu m TS). 180

..‘LSEACh’iu Chan-m3

"‘22.: Ch’iu. Chm-hsien

EQG Ch‘iu Wick - 0

 

 

 

LNAucau Te-l'icn

MGCRM. A-mi
(matrhd out)

IQ“ 3. .
H‘H Gimme“ h“!

. mo Que". Wag-Nuns

‘ W“ Ch’iu Ch’ing-i

%Ch’iu A-shan
m

‘ ”SA 03m Chin-ching

 

‘ I'm ‘3' CR6 «no.
h'A‘caigfieagw)

APPENDIX C

GLOSSARY

371

333323;3333;333:3323-13-13-1-1—4-14—4

APPENDIX C

GLOSSARY

. a PET
. an peh-j1‘ \£/\‘3'

bah beh hong chiah, kut bo a1' hong khefi] kgfifififijjg
bah ho Iang chiah, kut m-chia* ho Iang khoe Yhé’ékfifim
beng-Ieng jififié‘}

b0 thau-tui sim-pu-a £4 3 %fifi4§

cha éL,

cha-bo-kan’g‘fiyfij

cha p'ing-t'ouflfifi‘fi

. ch'a huagfdj 3?,

ch' a to hua- erh tai erh Sheng fifififiififii’f

. Chang 5%

Chang Hs1‘n-t' a1’ 32$?»

. Ch' ang- -an Hsien Nigé:

Chang Chou :

U

Chang- p' u Hsien I? Sfifiix
chao 37,

\

Chao-an 5?; ‘3?

Ch'ao-Chou 2%‘N
chao-chui 3.7% a

. chao-fu 3m :7";
. chao-fu Sheng-tzu flv‘E/i;

. chao-fu yang-fu 2}"? fi ék

chao-fu yang-lac if; {& i,

. chao-fu yang-tzu ”{3} i ii}-

chao-hun tzu {lai/365‘:

. chao- hsu 33:5?-

chao- hs'u yang- -1ao 2T7? ifi-ét

chao- -mu H.772

. ch'e-chang shih fiflifi

372

7133333334433

33333333233333.3333-1—1—1-1—1

373

. che-ku fiaufl
. che p'ing-t'ou gldg‘ia
. che-soa* chio-hu 55, LN 32; ‘fi

ch'eng-nun 69%;}?

. Ch'eng-ku Hsienfisbrfllgi

. Ch'eng-ti £441 its

. ch'eng-t'iao $4136

. ch'eng yin-ch'in fifiifi)

chhi cha-bo-kia* pat-lang-e, ’23 éJI-mg'lxfi?)
chh1' sim-pu-a tsoe knian ka fififijma
. chhia* chhut-tia* 3% m/E‘
. chhng-bo ff-Q‘

chhng-kong fix’fi

chhut-sia ”lg/g?

. chhut-thia* 716/33;

ch'i fir,

ch'i a

ch'i-ch'in 33%

ch'i chia-lac 523225

ch'i chia p'o {7% 31%

. chi-fu {'E‘ 52 .

. chi-pen ch'eng yiian EZPEQ-i
. chi-shu 1% $-

. chi-szu {1&3}?

. ch'i-ti 52%

. chi-tzu 14% 3-

. chia

chia-chi if;

. chia-Chang ch'iian am

. chia-fu ix

. chia-huo 537' . k

.Ch1'aI $2

#333343333333333333.3333333333433333

374

. chia-mai 4% i
. chia-tsu $536.

. Chiang~11°ng Hsien 3,1 Riga

chiao

. chiao-chia hsiung-ti '3? #{M

chiap to-pang {Kg {291%

. chieh-hsi] hun fifijfi ‘
. chieh tzu, huan sun 4% 3 £31;-

chien

*
. ch'ien $2,

Ch'ien-chiang Hsien 3315591951
chien-mei fig
cm'en-m'ng 33, "

. chien-Sheng :5;

chih

. Chih-ch'in é %%
. chin-nu sui ku fig im

. Chih-ting chi-ch'eng :jgy‘agggk

chin

Chin-chiang Hsien giifi

.chin-ch'in iiifia

. ch'in-ch'ijan 21.1%

. chin-Ian hu1' fifié

. ch'in-Shang chia ch'in fiurjiofl
. chin-shen hsien-Sheng fifl‘?éfi’i

. chin-t'ieh Z? E‘i

chin-tsoe {l 2%

. ch'in ying 2W4
. Ch'ing ch'i £3? ‘53
. Ching-m’ng Hsien 56% M

Ching-shan Hsien ill—3%

. ch10 4272

3 3 3.3 3 3 3 -l -1 -'| —| -I -l -1 -i -l

375

chio-hu che-soa* Q7; fiat/g.
chio-hu fang-10 flféi
ch1’o-hu iong-tsu f2; fiégg-
ch1'o-hu si*—tsu 32 ii}

chio-hun j1’ 212332}?

. ch1‘o sio-ti fl/J\%

. ch1o-t1' 3%

. Chic-ti 3?; if;

. chio ti-t1‘ flfifi

. Ch'iu fl?

. chiu-hun flag

. Cho-shui H51 3% 3K i’é
. chu {i

. ch'a fl;

. ch'u- ch'i ¥§

. chi] ch'u tse pa1' fei chieh shengflw‘fiagfle‘

hsi chu tse ko yu fei MK) I a-fi a

M. ch u-hai t‘. if}

M. ch'u-hsi ‘ffi. 9

M. Chu-hsi Hsien 795' Lifi

M. chu-hun if?

M. chii-jen fiE/L

M. ch'u kuei-ch'i $924? ,
M. ch'u-11 k'un-ch'i 3.117%?
M. Chu-shan Hsien 7‘? L“ ail

M. ch'u-she $9.5

M. ch'i] shen-chu @fi‘fi'i};

376

. Chfi-yung Hsien affii

chuan-ch'e chu fl i %

. chuan-ch'in fl $4

. Ch'Uan Chou fl .
chuan-fang

. chuan-hun fit?
ch'Ueh-p'ing ’ fir’
chui
chui-fu 7";

chiu-hun \ fig—
chui hsu & {g

. chui-shu %
. chui tzu §'

chung ?
Chung-hsiang Hsien fifi‘ifi

. chung-hu q! f"

chung-tang i-hsia chih [she-hui CF %- I‘AT iéié
en-i ming-fen 55‘ 3.54713

. fan %

fang %
fen é}
fen-hsi 45‘

fen-hsi chia-ts'ai t'ien-ch'an fifirfiifigfi
fen-1' ts'ai-ch'an @% 533$,
fen-pieh kung-yu fifi] 3E}?

. Feng-hsiang Hsien “ fl ’
Feng-hua Hsien 42,
at: 11

Feng-t'ien

Fu-an Hsien fig! ‘f E41

Fu-feng Hsien

. fu-hs'ti ch'eng y'tian Eaégé

Fu-ting Hsien

. Go
. Han-chung ‘ CF

377

ho-ch1’n 1: ~25,

ho-fang {3*
ho-1a1’hun-1’n fWfiW
hoan-sun fiafi

hs1'

hsi-fu-erh fis’fifig‘,
hsi-ma ((119 fig

Hsi-pei Hsiang Ejje. égf
hsi-t'ieh $4";

hsia-hu “F f"

Hsia-p'u Hsien fififi
hsiang 635’

hsiang-ch'iu 38 it
hsiao-kung Axm

hsien

Hsin-chu Hsien ifr’f‘rfi

Hs1’ n-wen Pao 5% iii a
Hsing-hua Fu $42 $1

. Hsing-shan Hsien ’5? wig};
. hsiung-mei hun ififi

. hsiung-ti erh 2,354,?)
. hsu-ch'in kg: 39.1

. hsu-ho fl?

. hs'd-hsiian *3 4g“

. Hs'dan-en Hsien ‘ififi

th—T

.hu/é'

Hua c1115)? 4%

. Hua-Hen Hsien £2 "gt/5%..

hua-teng nil 17; if #
hua-tun nii 2‘2 fix‘éfi
huan-ch'in :5; $ng
huan-sun i 22%,
Huang f

(K

. Hui-an Hsien g» jag

hui-men E] 333

3

378

. ,T. hun it;

hun-su fig $-

huo ‘l;

huo- tai 3% $
1%

1'- hsing chih tzu gfiig
1' hun gfifi

1- Ian Hsien Egg

I- 1an Shih WEE

1' men yu tzu, chiu men pu chiieh - lagjj‘fa'QiLfi 15:2.
gm

jen-shih A115

. jit-thau-a E1 iii/H-
. Jou P'u-t'uan BE] 2.5 I?)
. jyu-tsai $4}

ka ch1’ah-u1’ h9 75¢ #Jifi}

kai-dai ‘

ka1’ ga-louh fiat

ka1‘ ga-poh ék‘élsg

Kan Hsien gag

ke- hu 71$)“

ke- shan hsiung- -t1‘ chieh-mei Rial“ mm
ke- tsu éfi 3—

keng shu R %

khan- chhiu £1 f'
khe chhui- b1“ ’5'}?! 7K7B

Khu if
k1’u-k1’ong 7kg
ko- -11’ang 3&1?»

. koshu f1
. Ku-ch'eng Hsien fififi

. ku-jou huan-hsiang ‘fi'fiflfi

. ku-jou huan-hsiang, chia pai jen wangfifiifi‘szfikt

379

. Ko-ma-Ian “516%1 é;
. koa*-am chi* .
. koe-pang {if

\

Ku-t'ien Hsien ‘3; 9 fl
kua-chang 1i} ‘1"
k'ua-hsiang chih ”35le 9‘32

kuan-chi

Kuang-tse Hsien 5‘7... 3,? fl
Kuang-yun

kuei

kuei nan, tuan nii, kuei nil tuan nanflgfltgfiifl%
kuei-ch'i % a

k'un

k'un w )W

k'un chi 7p]

k'un-shen «1Q

kung 9'”:

Kung-an 1T3;
kung-yu #Efi
kuo-ch1’ flag
kuo-fang fill
kuo-fang 11' 1?! fig
kuo-lo 2X31
kuo-men E) F?

. Lan-yang Pg

lang-men n13 t'un 3’ PH: E
lac-pan *4 R

lac-ch'i 3%; g

le p1'a* fig? ‘
1eng-p0°-gin ’31. fi fig,

3

. L1’au 7
Lim

380

,1. L1 §‘

11
H-chiao 914;; fl;
11-szu 3‘1: 56%

. L1 Yu 7&3}?!

Hang-jen gal.

Liao

Lin M
Lin-hai Hsien E169 ifimk

. L'in Han-Sheng fl fié

Iing-shan hsiu-tsaifififij’
liu-H /\

Io-kong fi/‘k

Lo-tung @1-

1u-shu1 fM-fu 3.7K 7t éfi

Luan-p' ing Hsien 1%?

man-t'ung j'fii

lun-bun fig fink?
lung-yang fig. 5

. Ma-ch'eng % #5..
. ma-je fififl.
. ma-je mui fififi'

mai {i .

.maié

. mai-hun shu éiéé-

. mai-mai hun-yin‘gfigfl

mai-ti an i‘i/fi
mai-tuan

. mai tzu a}
. mai tzu, wu chiao ming, ma1‘ t'ienf§fifltz

wu yu t'ien-t:ou hang ‘iflfifiiflfifif
. mai-yang $1!
. mei-kuan fl?

381

. mei-t'ung k'ao-an 32% 1&3}.

.mi ix

. miao-chien @%

. miao-hsi ”a

. Mien. Hsien 5.133%

. Min-Ch'ing Hsien 7Q ififi

ming-fen in?)

ming-hun ii?
ming-h‘ng-tzu fifé}

. mu-chia Q3:

mui-tsa1 W}
na-ch'ieh 6*] i

na pi 11W ‘33?

na t'sai m 5*
Nan-an Hsien \g‘ififi.
nan-ch'ieh %E.
nan-feng ¥R

__ \
Nan meng mu chiao ho san ch'ien Qififi/‘zg
. N9 ’1;

niang-tzu chiin 45‘ 3f.
nfi-hu if
o'-a-h1' 543 \37

. Oat-ti 1% 1;

Pa-li-sha-nan w \ °12§W§

Pa-tung Hsien E 1311

pa-tzu / \ ‘?

pai-hsiang Chih fi 14%!

. pai-hsing E {It

. pai-hua {3 ii

pai-kua {g/K\
pai-liu ts'an-hua Mfifé

. pa'i-pai fi—fi

pai pi-t'ou égfi
4%

pa1-te

pék $218K”

33333333333333.371-1713333332333-13333-1—1

382

. pak chhe 71:31: 2

pak-sin ho-l ang aififik

. P'an *

. pan-1ao ’fiéifia

. pan-tzu pan-hsi] ¥3¥ffia
. P'an-yfi [gig]

. pang chhiu fiké

. pang-t'ui Qflfl

. pao-chia fig?

. Pao-k'ang 4%}4

. Pao-ting 43:53

. pe1‘

. p'ei-er‘h ag’f’i)

. p'ei-ku $24 %

. pei-n'd flit

. pei-sheng-fifi

. p'ei-szu hun @5925»

MW ka-chiah-ui 5 {112-$1,311

. Pho-po-o Mi

Phoa* 9%

pig}

p'14

. p1'eh 5%

. p'in

. p"1'n-ch'1’ £2 a
. p'in-chin Egg
. p'ing fl

. ‘P'ing-ch'uan Hsien ifigafl

p'ing-fu 33'- 1E

. p'ing-fu jg fl

P'ing-hu Hsien Sf-Efi M
P'ing-nan Hsien flfifi
/ .—

p'ing-t'ou fi’f

. p'ing-tu jg"
. P'o-pu-wu Wig

383

. Pu-t'ien Hsien 5&7 57%.

pu t'ung-fang chu-chu1fi%g’&

. Pun-yu %%

sai-kong ET ’2:
San-hsing 53 IE.

San-tiao :25. fl

. San-tiao L1'ng 2 212%

sang-jit-thau-afi g EX}
sang-t1a* Ki?
sang-tsoe-tui i’fiiifi
Sha-chiao z‘x/b’éfi

. Shan Hsien 35%ng

Shang-ch'in J:

Shang-fa j: a
Shang-men I F‘i
Shang-men han J”: F5 ifi
Shang-t'ou j; E; ‘
Shao-wu Hsien 77.2? flag:

shen 9f?
shen-fen j) 43

Sheng-fang 4}

. Sheng-hua Hsien FIE/{ufii

Sheng szu t'1’ng t'ien iflfii

. 511111 a

Shih p'eng-yu éaflfli
sh1’h tsung chao-fu \fifli
shou hsi-t'ieh‘ a '71:

\\I /-
Shuo-wen 39.; 3k
4* £1:

. Shou-chu <2 ..

. Shih-erh Lou +2 712%;

shu-hui 3% FE]

shu-mu flfi‘

shu-sao jEjen-mien fififi$
shu-tzu 3

Shuang-Liao 43,3]?—

384

shuang-t'iao 4%. 7‘36

shui 7K

Shun-ch'ang Hsien ”(i Erfi
Shun-te ME 41%.

sian-tan «(11.19“

Sia*-te @590

siang-Liau ”Li 7

sien-Liau :3;

simpua fiéfifif

simpua Cheng fifi§4§fi

simpua ka-kie, tsa-bo'-k1’a* pét-langefififl}fi5fi9
simpua m1'a* 33333444?- § Xmfl/xfib
simpua the fiafiIfl—fi
simpua-chai fig§4§fi

s1'o chhoa IL

sio-tso {gig

siOk-sin

Song yr):

Su-ao 3315?

Sui-chung Hsien A *fiL

su1’ mu chuan-chia fiiig'fifig
Sun-tak ME 6‘?

Sung a

Szu Hsien BPS] a?!

szu-Sheng fii

szu-tai 591$

szu-tzu fig é

ta-fan 1"] \T'%

. ta-huo 31$ %

ta-kung fi :75
ta*—poa* if fl

. Ta-tung Hsiang Kiéflf

t'ai q

. T'ai-chung Hsien {E2 €74;

T'ai-pei Hsien <2," 15%

33332333333323—1714443-433-1-4-1ZZ-IZZ—I—I—IZ

385

. tai-shu “{‘E%'

tai-su /l‘t%
tam Tong-a pun 2?. fi43$

tan

,1.
tan-liao $5,?-

. Tan-shui fiy- 2K

tang-'gng Teng-thau kha tang-Sim fifinggi fifildn

. tao

t'ao-hua chia-chu jig/TE, 5&fo

. T'ao-yUan Hsien fié@£¥§\

tau
tau long-a pun lkfigfiz:
t'e-pieh th-hun fifi” %] iii?

. Te- -'p 1'ng Hsien :fis‘z’ai

teh- chhi*
teng- -hs1' é

thai Hf;

. Thau- u1‘ gfi D
. tho- hoe ;k1'ap- -tek fix“; ‘KT'T

tho- -su ..j' fifi
thui tsoe tu1’ é’fi/ka/é

ti- tzu "fi‘% ;
t'ien ff]
tien ;&

.tiench'1’~$h%%

t1en-fu

. T' 1'en- -fu i2

. tien- hun shu Q;-
. tien- -ma1' 3%?

. T' ien- s-han {Lb

.T'1’en- --sungp'1' Kit?

t1ng
ting-Chan men-fen jg/foa/ZS

ting-Chih hui-p'en TE flffifi

. t1ng-hun flig-

386

. t1‘ng p'en Ta @-

. ting tuo, hsieh ts'ai T 21333”
. T1'u* 65V

. to' E31

toa chhoa 7k 3.}
toa-pai {#2

. toa- p1'a* k 3%

toa- tso 7kg

. T' ou- -'ch eng E2155;

t'ou—k'ao 3%” a?

.T'ou-wei jg]
.ts'ai 51'

tso- ch' an chao- fu ’15 __&d7l+\
tso 1' -tu1‘ chieh— ch' 1'ng fu- ch' 1M‘\§i 1??? if

. tso- shan chao- fu ’32 114 314—1:
. tso- t' ang chao- fu 2“: 3%‘7’

tsoe

tsoe 10- kong 1% ik A
tsoe toa- Tang KKK/K
tsu fl

. tsu 7555

tsu-ch'i fl?

ts'uan (9'

ts' un $\
ts' un- chang )H’fi

tsun- -chang ch' 'uan im

tsung ch' 1n ;m;

tu- Tun ch1'h hsien igfiLfi

. t'uan-yiian hsi-fu I? [a £41?

tun 71’
t' ung F]

. t' ung- -chu ting- -hun hs1' FUSE-{m
. t' ung- -hs1'ng pu hun, 1'- hs1ng pu yang fii’fq~i§%l%~

. T' ung- -shan Hsien LU
. t'ung-tsung E1337:

3333333333333333333233ZZZZZZZ-IZZZZZ

387

. t'ung-yang-hsi if %:‘kg~

t'ung-yang hsi] E rgz‘éii'
tz'u-mu 238’

. Tz'u-yang Hsien %?%%

tzu-yu nil @é‘gt

u thau-tui simpua 7g Efi‘fififizfi.
wai-chia 7p31

wai-sheng fl &

wan-ch'uan"7%_,

wan-hun ‘7’":

1%
wang-Tang hs1' fiéwfi
wei Chi-fang fié~¥

wei-hun ch'i iii???
wen-ming 351%
Wu 1%

. Wu ch'an wu tzu Qé?§'

Nu-feng Hsien If %~ %
wu-fu i AF:
wu-fu ch'in i Rafi

. Wu-kung Hsien ink 175%

ya-t'ou \J’ E3

ya-t'ung tzu/g 7kg 3.

. yang ?%

yang -

\

. yang

. yang-hsi ’ifi.
. yang-nu 3&3}:

yang tzu 1%;3—

yeh-ho 2} A;

. Yeh

. yen Via

388

. yin T’é

. yin fl

. yin chien erh sheng ET] 2?: a? é.
. yin-p'ei F; @E. 1

. ying-chieh kuo-men fléikflfi F?

ying-ch'in H $34
ying-ch'Li K21, g

. yu ch'an wu tzu fiéfifi $-

yu ch'an yu tzu/fi/ifi }
yii-ch'in fig?)

Yu Jui-Tang 7C1 {*1 BF

yu tzu, wan shih tsu Z3333§1
yu wu £193—

W W 7% IE

Yu-yii Hsien fi fig;

yii-ying t'ang fig}?

y'fian-fang [a

Y'dan Hsien a? £41 \

yiieh yu hsiao chih @{ifi/‘kéfi‘
Yung-k'ang Hsien figfikfii

BIBLIOGRAPHY

389

BIBLIOGRAPHY

ARLINGTON, L. C.

1922 Sinological notes. The New China Review 4,1.(Feb.).
BARCLAY
1923 Supplement to dictionary of the vernacular or spoken

language of Amoy. Shanghai.

BLOCH, MAURICE
1973 The long term and the short term: The economic and
political significance of the morality of kinship.
The character of kinship, ed. Jack Goody. London.

BUCHLER, IRA, AND HENRY SELBY

1968 Kinship and social organization. New York.
CHANG HSIN-T'AI fii'mfi .
l899 Yiieh-yu hsiao-chi. gfi/RZYQ

CH'EN CHENG-HSIANG fii T? \ . a,
1959-61 Taiwan ti-chih. ’3 2,5 fizw

CHEN HAN-SENG

1936 Agrarian problems in southernmost China. Shanghai.
CH'EN KU-YUAN Fifi 2%,
1937 Chung-kuo huh-yin shih. Shanghai. Rep. 1966.

Commercial Press. wgfififii
CH'EN TUNG-YUAN B’ifigi

1933 'Chung-kuo fu-nu sheng-huo shih. Shanghai.
Wfififliiéi
CH'EN TZU-MING Fifi ‘3
Sung Fu-jen ligng-fang. fik 365 , in KCTSCC.

CHENG HSI-FU Eéi'ikl
1970 éhien sheng-shao y'u Wen-su lu. a ifl’éifia‘i’ifiifi

\
Taiwan Folkways (Taiwan Feng-wu €35, ) 12.1:
74-77.

CHIEN, CHIAO

1975 Review of gods, ghosts, and ancestors by D. K. Jordan
American Anthropologist 77,2: 393.

390

391

J—
CHOU HSIEN-NEN a ‘3}. . .
1957 @1154“ Taiwan ching-chi shihl’é/fté‘zfifizfii

CHUANG YING-CHANG
1972 The adaptation of family to modernization in rural
Taiwan: A case study. Bulletin of the Institute of
Ethnology. Academia Sinica 34: 85-98. In Chinese.

CHUNG-KUO JEN-MIN TA1%§QEP CHUNQ-KUO LI-SHIH \
CHIAO YEN-SHIH 1:11 5 &7K~>;cv@)7,g

1957 Chung-kuo Tz'u-pen chu-i meng-ya‘fien-t'i t'ao-lun chi.

2 vol. Peking.¢@§$§fififl‘agiffi

M35
1930 Chung-kuo min-shang-shih hsi-kuan tiao-ch'a pao-kao lu.
Record of investigati ns i to civ'l n ommer ial
customs). Nanking.‘¥&&a $¥qéaé£fifivfli
CH'U T'UNG-TSU
1972 Han social structure. Seattle, University of Washing-
ton Press.
COHEN, ABNER
1974- ” Two dimensional man. London.

DOUGLAS, CARSTAIRS
1873 Chinese-English dictionary of the vernacular or
spoken language of Amoy. London.

EBERHARD, WOLFRAM

1968 The local cultures of South and East China. Leiden.
FALLERS, L. A.
1957 Some determinants of marriage stability in Busoga:
A reformulation of GluCkman's hypothesis . Africa 27:
106-21.
FENG HO-FA .15 352 fl
1933 Chung-kuo nung-tsiun ching-chi tzu-liao. Shanghai.
mamaeflfifi
FOX, ROBIN
1967 Kinship and marriage. Harmondsworth.
GALLIN, BERNARD .
1966 Hsin-hsing, Taiwan; a Chinese village 1n change.
Berkeley.

GEERTZ, CLIFFORD
1963 Agricultural involution. Berkeley.

392

GLUCKMAN, MAX
1950 Kinship and marriage among the Lozi of northern

Rhodesia and the Zulu of Natal. African systems of
kinship and marriage, ed. A. R. Brown and Daryll
Forde, pp. 166- 206.

GOODENOUGH, WARD

1970 Description and comparison in cultural anthropology.
Chicago.
GOODY, JACK, ed.
1973 The character of kinship. London.
HO YHI-MIN
1966 Agricultural development in Taiwan 1903-1960.
Nashville.

HOANG, PIERRE

1915 Le mariage Chinois au point de vue legal. Chang—hai
(Shanghai).
HSIEH K'ANG at);
1971 Tziwana ch' ang- -chi wen- -t' i. Pei-t'ou Taiwan.
bgkfiffifi
Hsii K'O (£13)
1920 Ch' ing- pai Iei- ch' ao. Shanghai 1966 Taiwan reprint,

Commercial Press Taipei. t figfifi

HsiJ T'IEN-LIN 4% {gr}

1971 Tai-shu chuan-chuan ‘pu ch'iu- -jen. Kaohsiung:
Kuang-ming Press. jté a? 913;sz /\_
HU P'U-AN *‘7
1936 gaingxfigajgh' uan- -kuo f ng- su chih&; Shanghai. 1968
reprint, Taipei. dag 4% (55:9,

I-1an Hsien-Chih PE: fig %‘t£~
1969 (abbreviate ILHC)

IKEDA, TOSHIO

1944 A survey of Taiwanese family life. In Japanese.
JORDAN, D. K.
1971 Two forms of spirit marriage from Taiwan. Bijdragen

tot de taa1--1and--en volkenkunde van Nederlandseh-
Indie 127,1: 181-89.

JUANICH'ANG-JUI 141 h H h 1
968 an t -c u an- ua t §va a 1

E1.“ 11 Nen- hsien 2’th fléfib’gg “5%é' RAW

chin

KCTSCC see Ku- t' u- -shu chi- ch' eng.

393

I
KATAOKA, 1on F: 155] - H .
1921 Taiwan fuzoku shi. Taipei. ’aaéfiL/ixgiu

KERBLAY, BASILE
1971 Chayanov and the theory of peasantry as a specific
type of economy. Peasants and peasant societies,
ed. Teodor Shanin.

KING, FRANK H. H.

1965 Money and monetary policy in China 1845-1895.
\
Komalan t'ing-chih flgigfiflafi
Ch'ing

Ku-chin t'u-shu chi-ch'eng KQ’IE %駢u (abbreviated KCTCC).
LEACH, EDMUND

1961 Rethinking anthropology. London.
LI HSIEN-CHANG 45% fl \. L.
1936 Taiwan min-chien wen-hsiieh {5- ZQE—‘FgX-‘ég

(Taiwanese folk literature). 1970 reprint, Taipei..

u /
LI YU 7534!. -
1969 Nu-sheng hsi.@:f?[§§. Taipei reprint, Helmut Martin,

ed.
LIEN HENG 3876;: Taipei. $75381?»

1920 Taiwan t'ung-shih.

1963 Taiwan yU-tien. Taiwan wen-hsien ts'ung-k'an #161.
Reprint. fa‘fig §%
LIN HENG-TAO . ‘ . -
1968 Ch'ing-tai Taiwan k'ai-t'o chien snu.>§i€€@53%

‘ Chun - Nen- Fu-hsing Yfieh-k'an 2,3.
I3”: wgdmi&fiii
LIU CH'ING-PO

1969 Yang-ngérchgung-hun, t'ung-chien yen-chiu.
*a

LIU HSING-T'ANG (fiifila .
1936 Chui-hun-chih te liang-ko hsing-shigggflm

Tang-fang Tsa-chih 33,15: 107-11%
L0 HSUEH-LUN m ‘gxfi

1975 Li yU hsi-chfi,\ s' o h Q f - 'n t zu-ggiang
yii shih-tai.§;§.% cafliéséé é‘) 33:7

Ta-1u Tsa-chih 50,2: 56-87.

394

LOU TZU-K'UANG ij-E

1968 un-su chih. Taipei. M £5.»
MARTIN, HELMUT, ed.
1969 Nu-sheng hsi. Taipei.

MARTIN, RICHARD, trans.
1965 The before-midnight scholar. London.

MATHEWS, R. H.
1966 Chinese-English dictionary. Rev. ed. Cambridge,
Harvard University Press.

NEEDHAM, JOSEPH

1956 Science and civilization in China. Vol. II.
Oxford.
NEEDHAM, RODNEY, ed.
1971 Rethinking kinship and marriage. London.
1974 Remarks and inventions. New York.

OKAMATSU, SANTARO
1901? Provisional report on investigations of laws and
customs in the island of Formosa. Kobe.

PITT-RIVERS, JULIAN
1973 The kith and the kin. The character of kinship, ed.
Jack Goody.

RANSKI, E. S.
1972 Agricultural change and the peasant economy of South
China. Cambridge.

RIVIERE, P. G.
1971 Marriage: A reassessment. Rethinking kinship and
marriage, ed. Rodney Needham, pp. 57-74.

Sou-shen chi {Kg 8"“? 131.1

Kan Pao, compiler. 7)" gm
STOVER, LEON

1974 The cultural ecology of Chinese civilization.
New York.

SUN HSIAO-TSUN
1938 The problem of farm management in contemporary China,
part 3. Quarterly Review of the Sun Yat-sen Institute
3,2. Reprinted in translation in Agrarian China.
Institute of Pacific Relations, Chicago, 1939.

395

Sung-shih tsu-p ' u 9E R751}? $4

1971 Taichung.
SUZUKI, SEIICHIRO £919? ’8?
1934 Taiwan kyukan kankon sésai to nenju gyoji. Taipei.

tabég“aflgfifiijg%g§;ffi.za£FJ¢zfi g;

TAI YEN-HUI a we it?
1963 TC 'ingg -tai T Twan cih chia-chih chi chia-ch'an.
ggkc Taiwan Men-hsien 14,3:
' 9°~az- 3;

1966 Chung-kuo fa-chih snih.€i'@iiifil§t Taipei.

Tai-shu su-chuang ta-ch'uan .
1970 (The scrivner's complete handbook of legal forms).
Taipei, Ta- -fang Ch'u-pan She.

Taiwan Szu-fa e “364/.32.
Abbreviation for Taiwan szu- -fa jen- -shih p' ien.
{f’gfiiflgffi Taipei, Bank of Taiwan.

THORNER, DANIEL

1966 . Chayanov's concept of peasant economy. A. V. Chayanov:

- The theory of peasant economy, ed. Thorner, Kerblay,
and Smith. Homewood, 111., for Am. Econ. Assn.

THORNER, DANIEL, KERBLAY, AND SMITH, eds.
1966 A. V. Chayanov: The theory of peasant economy.
Homewood, 111., for Am. Econ. Assn.

TMS
1969 .Taiwan min-shin si- kuan ao- kao (abbre-
viated TMS).’€—1‘S’@‘1td%fic%% Taipei.
VAN GULIK, ROBERT
1961 Sexual life in ancient China. Leiden.
Nan- ch' uan kuang- -ssu chi-yao gzé/iffifig
In KCTSCC.

HANG SHU-NU 33. a fix
1934 Chung-kuo ch'ang-chi shinci‘flihaafii‘l Shanghai.

WEN CHI 33%

1957 Taiwan fan-Cheng Chih/é‘zgfilifsqaipei.
NITTGENSTEIN, LUDWIG
1953 Philosophical investigations, trans. G.E.M. Anscombe.
Oxford.

1958 The blue and brown books. Oxford.

WOLF, A. P.
1964

1966

1968

1970

WOLF, MARJORIE
1972

WU YING-T'AO
1970

YAO LING-HSI
1940

16 CHIN-CH'UAN
1951

396

Marriage and adoption in a Hokkien village. Unpub-
lished Ph.D. dissertation, Cornell University.

Childhood association, sexual attraction, and the
incest taboo: A Chinese case. American Anthropolo-
gist, new (2nd) series 68,4: 883-98.

Ad0pt a daughter-in-law, marry a sister: A Chinese
solution to the problem of the incest taboo. American
Anthropologist, new (2nd) series 70,5: 864-74.

Childhood association and sexual attraction: A
further test of the Westermarck hypothesis. American
AnthrOpologist 72,3: 503-15.

Women and the family in rural Taiwan. Stanford,
California.

%1 wgmin- tsu. é‘ifififihipei.

{9%ng wei-yen. 31% 515146; a

An analysis of the population of P'ing-p'u in San-
hsing Hsiang, I-lan Hsien.
Taiwan Wen-hsien 2.3-4: 1-5.

Y'u-fang pi-chlieh 331% $15 fl

 

HICHIGQN TATE UN V.

1111 1111

W
01643

RRR

111“

5
1111111111111in
31 2 31

 

9