C:::A.

- b
-=+
@::
210 Gayle Rubin
\,.

- .."'t. .
and political analyses are incomplete if they do not conslder Karen Sacks
women, marriage, and sexuality.
thropology and social science--such as the evolutlon of social
Engels Revisiled: Women, Ihe Organizalion
S&atit1catJon aod tbe::::zmgll'} ol the sEate must_be reworked
,1Q. include the implications ';)Tmatniate;ili O, prOdUC'Vd Private Properly
,ri@.ge, surplus extractea 01 daughters, the conver
fii6or-.ito male ...wealth, the conversion t;)f
fimale lives of mar
political power, and the WhlCh aU of .
vanedJlSpects 01 in the course
-
This paper reexamines Engels' ideas on the bases oC women's
This sort of endeavor iS7 in the final analysis, exactly what
social position relative to that of men. Engels s almost alone
Engels tried to do in his effort to weave a coherent analysis
in providing a materialist theory-one that sees women's posi
of so many of the dlverse aspects of social life. He tried to
tlon as varying trom society to society, or epoch to epoch,
relate men and women, town and country, kinship and state,
according to the prevailing economic and political relation
fonns of property, systems of land tenure, convertibility of
ships of the society. Though he made l number ot speeific
wealth, fonns of exchange, the technology of food produc
ethnographic errors," 1 think his main ideas are correct-and
tion, and fonns af trade, to name a few, into a systematic
remain the best way of explaining data gathered since he
historical account. Eventually, someone will have to write a
wrote-namely ethnographic and historical data which show
new version of The Origin o( the Family, Priuate Property,
that women's social positlon has not always been, everywhere
and the State, recognsing the mutual interdependence of
or in most respects, subordinate to that of meno
sexuality, economcs, and politics without underestimating
the full signicance ol each in human society.
This i. a somewhat revised version ol a paper oC the same lille lbat
appeared In Woman, Culture, and edited by Mlehelle ZimbaJisl
ROl&Ido and Louise Lamphere, and Is reprinted with the permission ol
the publlshers, Stanford University Presa. Copyriaht 1974 by the
Board ol Truslees ol the Leland Stanlord Junior Univenily. SpeciaJ
persona! thanks to Shelley ROlaldo and Louise Lamphere Cor facilila(,
'. inf this reprintilla. Judith K. Drown, Kathleen Gouah, Bridaet O'Lauih
IIn, Dorothy Remy, Jean Willlam., anctoon Younf Yoon aJl eontrib
uted a reat deal to thia paper throuah tJleir valuable aUlrestions and
critlcismL
I have exeluded enul:3eratlon or theae partJy for laek or .pace, but
aJso beeauae they are substantively aecondary and are more than amply
dealt wilh by othen. Two auch erron, however, are aermane lo the
diacu.ion in this papero Enaela belleved that men were aJway. the
collecton or producers ol aubaiatenee. It hu aince become clear that for
,atherin,-huntln, loeietie. the revene ia dOler to the norm (Lee and
DeVo"" 1968); and lor horticultura! loeletJes, it ia often the women'.
211



Since capitalism has dominated anJ transformed the social
orden oC most oC the worJd's peoPlerit is uselul to look to
the past, as Eogels dd, through ethn graphlc and hlstorical
reconstrucnon, both to undersLand th present state ol atlairs
and to heJp shape the !uture. at noncapltalst
WBYS 01 organizing economic and relations and how
these aClccted the relative positions ot men and women pro
vided with an answer as to .,hy women were sub
ordinate lo men in capitaJist society, what political and
economic changas "ere needed lo end inequaJity.
The Ori8n o{ te Family, Private operty, and lhe State
(1891) is more than an anaJysil ol ornen's status. It is a
contrast between nonclass and class so ieties. Set in an evolu
Lionary lrameworlc, it shows how prv te property orginated;
and how, once on the scene, it und nnined an egaJitarian
tribal order, creating Iamilies as econo ic units, inequaJity ol
property ownership, and, nally, exp oitative class societies.
IL includes a descripton ol how W'o n's social position de
clned as private property gained str ngth as an organizing
principie ot soclety, and weaves in an analysis ol why prop
erty had Lhe effect it did speclicaU , how it transCormed
women's work organizatiol.I and, rnore generaJly, the relation
ship ol property to class and sexo
The fnit part of this paper tether sorne al Engels'
key pointa on how the sexuaJ egalit anjsm ol pceclass soci
eties WIIS undermined by changes in amen 's work, and by
the growth al the lamily as ao impo t economjc unit. U is
a selective and somewhat intecpretive summary: selective in
that it locuses on Engels' ide-as about!pubJic labor, the Cam
ily, and prvate property as they lo women's status,
and excludes his discusaion oC incest, ,xogarny, and the early

T
borticullural actiYiliea whJc:h are lbe bula aub.iltenc:e (lee Judith
Brown, lbia yolume). En,ela alao beUned al tlJe domelUcaUon oC
.mm... preceded eulUyaUon ol lb. ay, u a reault 01 more
ncent reaean:h, a more commonJy I lbat cuUivation
and putorlliun developed It the SAIne te in lb. aame milieu, u
proCJUl"ely divel1ent and lome"hat interd ,pendent .daptaUo'" (Lato
timare, 1957).
stages ol human- social evolution in general"; interpretive in
that ( have used sorne oC what has been learned' since Engels
wrote as the lens through which to view his ideas. A second
section redefines sorne oC Engels' termlnology and framework
in the context ol nonclass societies. The third secton
examines, using eLhnogrllphic data, Engela' ideas about the
importance ol public labor, private property, and the lamily
lor detennining women's status. WhUe these indeed appear lo
be lactora determining wornen's position, this examination
suggests sorne modification oC Engels' idea that women are
eiter social adults or witely dapendents, Baaed on this mod
fication, namely that the 01 private property does
not dilectly lead to a lower social status lor wornen, the
Courth section suggests an altemate explanation lor Engels'
obssrvation that class societies have used the Camily to cir
cumscribe and subordnate women..
Women in Enge/s' Theory: A Reeonstruction
Engels presents a historical procesa by which women are
transformed from free and equal productive members of socio
ety to subordinate and dependent wives and wards. The
growth ol maJeowned prvate property, with the lamily as
the mstitution that appropriates and perpetuates it, is the
cause oC this transfonnation;
According to Engels, in the early stages ol society produc
tive resoueces were owned communally by the tribe or the
clan. Food had to be collected and cooked daily. Production
WIUl lor use only, that is. to meet people's subsistence needs.
There was no surplus produced COl exchange. ** The group,
For a rull disclWion or En.eU' entire "ork in the Iigbt or current
knowlede., Eleanor Leacodl:', Intro.!uc&ion to the 1972 .dition iI oC
by Importance. Gou.h, in "The Orilin or lbe Family" ( 1971
and lbll .olwn.), and in "An An&hropololilt Loob at Eneell" (1972),
providu important reexaminatiof15 oCEn,eb;' theory oC "omen.
Thou,h Engell doe, not deal with lbis situalion, people in many
Donel... and noncapitaliat ,odetie. do in r.d produce ror excbanle.
Tbe questlon oC ho" produc:tloo Cor exchane in tbocetie. diller
214 Karen Sachs .
consisting of husband, wife, and dependent children, was
neither a productive unit nor one for perlorming housework
-,-or did it own property. Since Engels saw economc func
tions (production, consumpton, and property ownership), as
defining the lamilYt and since this group was in no way an
economic unit, the family did not exist at this stage; it had
not precipitated out ol the larger household. The household,
which was the basic social and economic unit, was commu
nistic in that all food stores were held in common, and all
work was done lor the household rather than lor individual
members or couples. Women did the housework and ran
these households.
In the old communl:ltic household, whlch embraced numerous
couples and ther children, the admnlstratlon or the household,
entrusted to the wO'.nen, wu just as much a public, a socially
necessary Industry II!. the providing of lood by the meno (Engels,
1891:120)
Instead of the lamily, the context 01 men's and women's
life and labor was a larger group based on kinship or resi
dence in a common territory. This was a cornmunal property
owning group, called the gens by Engelll. Although indio
viduals of both sexes owned tools and oersonal effects, on
their death these passed 10 other members of the same SE:! in
their gens, not necessarily to their owJ1 children. Decision
making, both economic and political, involved the equal par
ticipation of all members, both men and women. Both sexes
were equal members of the group because both made crucial
contributions 10 its economic life.
Engels concluded that the absence of private property
made men 's productive work and women's household work
of equal social significance. Men and women were simply
involved in diflerent stages of the producton of the same
Crom lhat oC capitaJi.t .ocietiea iI a complex one. A load dilcuuion 01
the Cundamental dlCCerence. involved can be Cound in Manhall Sahlln.,
Stone Are Economica (l971).
Engels Reuisited 215
kinds of goods-the production of subsistence. AH produc
tion was of the same kind: production for use.
Engels focused on the public rghts of women in the enrly
stages of society: their participation in political decisin
making and (lor the lroquois) their collective right to depose
a chief. These rights carne from membership in the gens.
which in turn was based on the perfonnance of public or
social labor. He was also impressed with the high status of a
wife relative to that o!- her husband, which he attributed to
the solidarity and kinship among the wornen, who were the
core ol the household.
The material base lor women's transfonnation from equal
members ot society to subordinate wives lay in the develop
ment of valuable productive resources, initially the domesti
catin of large animals, as private property. For Engels,. the
words "prvate" and "property" had a specific meaning. Only
goods or resources with productive potential could be consid
ered property. He was aware that people held personal goods
individually. Though these were prvate, they were not prop
'erty in the sense Engels meant the word. Only things with
productive potential can be considered property. Tools (pro
ductive means) are unimportant because the skills and mater
als for their manufacture are equally available to all. In non
industrial societies, the most important types of prvate
property are domesticated animals and cultivated land. These
are productive resources.
Engles' use of the word "private" is broader than ts use
under capitalism-where there 'are almost no restrictions on
what the owner can do with property. For Engels, prvate
seems to mean property owned by an individual or by a
famUy where rights to manage it are vested in one of the
owners. It also means that these goods can be disposed of
with sorne leeway-that is, to acquire wives, clients, or service
from others. Engels saw "gaining a livelihood" as always
men's work, and the means ol production as aJways having
been owned by the user (with the stipulation that inheritance
remained in the geM). From this he reasoned that the earliest
216 KCJren SCJchs
private property, which seemed to be omestic animala, must
have been owned by men." For Engel , prvate property be
l.ime possble in history on when technological
deveJopment and natunil resources allo ed a socety to devel
op the skills needed te domestcate an mals or to nvest labor
in land so that ita productvity last for sorne appreciable
length ol time. He believed that end ng productivty led te
enduring prvate ownenhip.
Domesticated animals were lLIsimila ed nto the older pat
terns of tool ownership-that is, they were privately owned.
Yet animals were a qualitativel)' new ind 01 item: they met
subaatence needs, and they :reprod ced themselves. The
growth ol private property shattered he communal poli tica!
economy of the gens. The fou.ndatio for ita egaJitarianism
had been the collecUve owne.rship o productve property.
Now that property was privately owne (by men), the famUy
grew in importance and soon oversh owed the gens as the
key economc and decsion-making oup. Unlike the gens,
though, interna! family structure was n t egalitarian. Families
contained propertyless dependents women and children,
and sorne propertyless men).
Prvate property transfonned the elations betwsen men
and women within the hausenold o y because it also radi
cally changed the polticaJ and econom1C relaUons in larger
society. For Engels the new weaJth animaJs
meant that there was a surplus 01 oods available tor ex
change between producve units. W time, production by
men specifically for exchange expanded,
and carne to overshadow the househo 's produdion tor use.
Industrial capit.alism had now reach the stage where pro
It i.I worth notnr that Enle" DW these new itenu (domeaticated
animal.. cultivated land) .. beinl auimiJata loto an already existin,
lOCiaI contad: tIIe panern oCownin penon eCCectl. The qualitatiyely
dirterent nature oC thue new "effecta"- at tIIey could reproduce
themaelyea and their Cruiu-Ied lo Ule de. cUon oC the communal
poliUcll.l and eeonomic order Lhat had crea ed them. Engel. doea not
attribute Lbe development oC private prope to a nedy maje nature.
Engels Remsited 217
.
duction waa a1Jnost exclusively social, outside lhe household,
and for ezchange, Ieaving women's work as prvate malnte
nance for (CJmUy use.
As production for exchange eclipsed production for use, it
changed the nature of the household, the significance of
women's work within It, and consequently women's position
in society. Women worked tor their husbands and families
instead of for society as a whole. Prvate property made ita
owner the ruler of the household. Women and other prop
ertyless dependents worked te maintain and augment the
household head's property, for he was now engaged in com
petltive production and exchange with other heads of house
holds. Women's labor was a necessary but socially subordi
nate part of producng an exchangeable surplus. Women
became wards, wives, and daughters instead of adult members
o! the society. .
Families perpetuated themselves through time by the in
heritance of property. Thus changes took place in the defini
tlon of children. From new members of a societal group, they
became either prvate heirs or subordinate, dependent work
ers, This meant that women's reproductive labor, like ther
productive work, also underwent a transformation from so
cial te prvate. That is, women bore men 's heirs-to both
property and social position-whereas before they had borne
new members of a social group that included men and wom
ell' People and property became intertwined, and each be
carne part of the definition of the other.
With the further development ol technology and accumu
laton of weaJth, the property owners separated themselves
trom their subordinate ldnsmen and allied with other prop
erty owners to preserve and defend their holdings against the
claims of the nonpropertied. This marked lhe end of kinship
based' productive groups, and Ule beginning of class
and the state.
218 Karen Sacks
Engels' Theory and Nonclass Societtes
To use Engels' concepts oC social labor and production for
exchange and apply them to nonclass societies, I will have to
redefine them so that they are more in lne with the ways
these societies are organized. Engels' use of social or public
labor in nonclass societies emphasizes work tor and in the
context of one's own corporate property-owning group. But
marriage often jolns two such groups; this leneral1y means
that at least one partner is not working for and in the context
of his or her natal group, At the same time, he or she is not
necessarily doing what Engels would call dcrnestic work
work for one's own household. Therefore 1 will stretch his
concept of social labor to include any work done (singly or in
a group) for use or appropriation by someone of another
household. Sorne examples of social labor, iIIustrated in the
next section, indicate the wide range oC organizations it
covers: participation in a cooperative worz group, tributary
labor for a chief, corve, collective Ivestock raiding, etc.
Engels' discussion of production tor exchange in the con
text of noncJass societies has to be amplified somewhat.
People do not spontaneously work to produce a surplus as
Engels implies (1891:264). There has to be some power
forcing them to produce more than they use. People in all
societies give hospitality and gifts, and these a1ways put the
recipient under an obligation to mue a return. I ~ a general
way, as long as everyone has equal access to th; means of
subsistence, production is planned to include hospitality and
gifts, and these are things which everyone has or can expect
to have by his or her own effort-and can thus make an
equivalent return in goods. But when the means of subsis
tence are privately and unequally held, a recipient is often
unable to make an equivalent retum in goods. He or she may
then be expected to retum the favor with service and become
a loyal dependent or client-follower, perhaps part of a retinue
helping entorce unequal exchange.
Both situations, the return oC equivaJent goods and the
return of sen/ice tor goods, are instances of exchange. But
Engels Reuisited 219
only the second gives one of the parties the ability to harness
the labor power of others for his or her own ends. Thus the
production of goods to gain control over the services oC
others must also be included in production for exchange.
Indeed, Engels argues that the domestic and private owner
ship o! cattle brought with it increased productivity of labor
and the use of cattle to conquer or purchase labor to serve
the wealthy (ibid.:265).
While production oC goods for trade or barter between
groups exists in societies without private property, it exists in
the fabric of a political economy geared principally toward
production for use. Perhaps this fabric was rU"St rent when
prvate property allowed the use of wealth to gan fol1owers,*
and then the need to use the productive and military labor ot
one's fol1owers to create still more wealth to keep their loyal
ty. In any case, this particular kind of production for ex
change in nonclass societies goes hand in hand with prvate
property, arad with economic and political inequality.
For example, there is wealth inequality and clientage in
nonelass societies with large domesticated animals. These
animals not.only contribute to subsistence, but they are also
necessary in order for a man to marry and to have sorne
political standing. Thus, in much o! East Atrica prior to
imperialist rule, men obtained cattle from kinsmen or Crom
service to a chief or other wealthy man, to whom they then
owed loyalty in exchange !or the cattle. The production of
cattle was a kind of production for exchange, in that loyalty
and service were given for Iivestock and were used to aug
ment the wealth and power of the benefactor, whether
kinsman or noto Regardless o! overlapping rghts and obliga
tions of various people to the livestock, the cattle were pri
vate property because there was sorne choice in how they
would be allocated, and because an individual was em
powered to make that choice.
A dllcuulon o( lbe vuiety oCcondltion. under which loyal (ollowe...
would be dirabJe betonl' in a conaideriltlon o( lb. orilin o( .tratjfi
eatlon and th. ,tate, wbicb 11 beyond lb. teope o( tbi. papero
220 Karen Sacks
I
Women as Social Adults and Wiues: African Societie
Though Engels has an integrated eory, at the risk oC
sorne distortion I would lke to sep ate out two sets oC
ideas: (1) Thcse about the immediate eterminants or mate
rial bases of women's .tatus-that cial or public labor
makes men or wol'nen adult citizens n the eyes ot society
and that men 's ownership ol prvate property establishes
their dominance over women in lwniJy and society.
(2) Those about the evoJutionary women's status
became solely subordinate and domef.tic with the develop
ment of maJe private property, prodJion for exchange, and
elass society.
In this section I will discuss the im diate determinants of
women's status by using ethnograPh'iJlustrations. This has
the advantage of focusing first on the material bases of
women's poston. Even iC is . ht in a general way,
that women are worse oCf. in class th in nonclass societles,
we stiJl need to know what gives . e to this state. Using
ethnographic reconstruction allows look at sorne of the
variety in women's status in noncapit ist socetes-rnonclass
as well as cJass-and to use the co parisons 10 illuminate
Engels' ideas.
1 do not believe that Engels' evol tionary explanation ls
correct as it stands: there iS too m eh data showing thal
women are not the complete equals men in most nonclass
societies lacking private property. Th re are also rnany soci
eties, with and without classes, whe women do own and
inherit property. The final section wi use some illustrations
from class societies 10 suggest a di! erent route to Engels'
conclusions.
The following illustration is a reco+struction, mainJy from
ethnographic sourcea, ol women's in tour African
societies prior to the imposition imperialist dom
inati.ln.
f
Thi. Mction slUDmarize. a portian oC my di.uen.tion (Secks, 1971).
TI_eae 1I0cieIJe. were selected from the on Eut ....d So.,them
Arrica beca.,.e the data on women are adeduate .... d comparable. Tbu
Engels Reuisiteti ::l
The .Mbuti oC Zai"re can be characterized as a barid society
with subsistence based on communal net hunting and the
gatherng oC vegetable lood. In South Arica, the Lovedu
were principally hoe agriculturalists, while the Pondo como
bined agriculture with Iivestock. The Ganda, a class society in
Vianda, were also hoe agriculturalists.
If we place these societies on a continuwn from egalitarian
10 class socety, our rankings can be seen 10 hold in three
principal respecta. First, Mbuti and Lovedu have economies
oC production for use; the Pondo have the beginnings of pro
duction for exchange centered aroWJft"attle; and in Ganda
production for exchange is quite important. Second, in Mbuti
and Lovedu both sexes perform social labor in a use econ
omy; among the Pondo, this remains the organization of
women's labor, but the men perform social labor at least in
part in an exchange economy; and in Ganda women's work is
individual domestic production for household use, while men
work in groups, almost totally in productlon for exchange.
Third, the Mbuti band owns the productive resources; these
aru largely patrilineal fwnUy estates in Lovedu and Pondo;
and in Ganda they become leas enmeshed in family obliga
tions and are in male hands.
Among the Mbuti, Lovedu, and Pondo, women's produc
tive activities Ale social and women have an aduJt social
status. In Ganda, where women's 'productive activities are
domestic, the status of woman i.& that 01 wife and ward only
-despite the fact that women produce tbe bulk of the food.
Thia suggests that Engels is right in seeing publc or social
labor as the bass for social adulthood. A more detailed look,
however, shows that women do not have to be eharacterized
as either social aduJu or wifely wards. Rather, the data sug
leat that women can be both simultaneously, and that
women's status in a marital relationship seems to vary inde
pendently 01 their statUJI in the larler society. Engels does
seem correct in seeing the status ol wife relative 10 husband
lQany upeclr-for e:Eample. concernin, womeh in trade .... d mukelihl
rol.-tboulh &heyare important, .imply cannol be deall wi&h here.
222 Karen Sacha
as dependent on their relationships to the property01 the :
household; that is, the spouse who owns the property rules
the household.
Table 1 summarizes some ndices of women's status in socio
ety and in the lamily, and their relationship to women's
organization 01 productive activities and to property owner
ship. * EssentiaJly, Mbuti and Lovedu women are the equals
ot men, whereas Ganda women are subordinate, and Pondo
women tall somewhere in between. The rst nine variables I
see as representing social adulthood, and the lirst five 01
these involve egalitarian relationships with people outside the
household.
A look at the rst variable, mutual-aid relationships, sug
gests that social adulthood is based on perforrning collectve
social labor. Though the Mbuti have no categories of relation
ship that are specifically identifiable as mutual aid, Lovedu
men and women both have some sort of age groupings that
are mobilzed to do sorne work for the district head and
queen. ** For women, at least, those of a neighborhood age
group mhy take col1ective action against the group of a
person who has offended one of its members. Pondo women
of a neighborhood work together and cooperate in the per
formance 01 girIs' initiation ceremonies, and women of the
same household cooperate in arranging extramarital sexual
affairs. Women's collective action s recognized by men when
they collectively punish women and gills for what they deem
to be sex olfenses. Ganda men enter a number 01 mutual-aid
reIationships with nonkn, or even non-Canda; women have
no such reIationships.
Self-representation in legal proceedings indicates that a
woman is regarded as able to be wronged or to do wrong in
the eyes 01 society, as is the case among Mbuti, Lovedu, and
The variablel and lbeir catelorization, rather lban beinll detenuined
in advanee by a IOllical .cheme, emefled u relult ol comparinl the
pOIition ol women in eaeh al lheae Courloeieti.l.
... "Queen," the lide uI.d in th. ethnoluphiel, ia a milnomer. Thoe
olnee i. aelually that ol a tribal chiel, earryinl moral luthority but
\IUle If any c:oercive power.
Engels Reuisited 223
I
's Social and Domestic Status Compared with Men
in Four African
Discrimlnation against
women's participatian
Indexes of womens status Mbuti Louedu Pondo Ganda
Social
Mutual lid n.a. nene active
Self-representation none none none active
SoclaIlzlng opportunlty non. none none active
Extramaritalsex none none none active
Dvoree none nene none active
Social disposal oC wealth none none active active
Polltica! otce none none active active
Extradomestlc dispute settlement none none active acti,e
Extradomestic mediatlon
with supemalur.d none none active active
Domestic
Wlfe's inheritance of marital estate none active active active
Wlfe's autbority OVI!r
domestlc afrain none active active actlve
WU. 81 priyate reproducer
(adultery compensatlon) none active active active
Menstrual and
preenancy restrietlons none weaJc weaJc active
Note: Ownenhip ol malor produetiv. resourees: the band in Mbuti. the
lamUy In Lovedu and Pondo, and th. individual in GandL Collec:tive
aoelal produc:tion by ..omen. u allainat that by men: equal in Mbuti
and Lovedu, unequal in Pondo, and abaent In GandL
224 Karen Sachs
Pondo. A Ganda woman, by contrast, needs a male guardian
(generally a husband or lather) to br ng her case 10 court.
The guardien is held responsible lor her acts and receives
compensation lor wrong5 done to her. .
Though Mbuti, Lovedu, and Pondo en and women pert
cipate in most of the same social ties, in the latter two
societies young wives are kept busy at domestic work, which
significantly restricts their ability to e these events. But
as older wives, as ssters visiting theirl own kinsmen, and as .
diviners, women attend social events treely as do men. In
Ganda, a large portion ol the social Bftivities are patrn- or
state-oriented; lrom these, women are
Mbuti and Lovedu have a single stahdard regarding extra
marital sexual aflara. Pondo women their extramarital
allairs as right and proper, but the menisee women's aflairs as
irnmoral. A Gnda husband may kili his wile lor real or sus
I
pected adultery, but a wile has Jittl, recourse against ber
husband. Men may use the courts to deal out severe punish
ment to their wives' lovers. In Ganda restricts extra
marital sexual activity much more than do the other soci
eties. Exceptions are made lor high-nlJlking men who have
affairs with peasant women, but men land women in the re
verse situation are punished severely. i'ehudi Cohen's (1969)
pont that restricted sexual activity serves to strengthen the
marital bond at the expense ol bonds Wh ch could serve as a
basis lor rebellion in class societies seems borne out here.
While marital and social status are v*ry closely related, the
ease of divorce lor men versus women indicates the relative
importance ol marital and social statu; lor each. In Ganda a
husband can effectively end a maniage by simply ignoring his
w.ife, but a woman who wishes a divqrce must contend not
only with her husband but with her who is partial
guardian and generally acta to presE'''Ve marriage.
Being able to give and receive lod and items ol social
exchange is the 'material buis lor exer!cising poltical power.
Engels suggests that real power only with produc
tion tor exchange and private In societies without
these-that is, in societies based on Pfoduction lor use-the
Engels Reuisited 225
perfonnance ol socia} labor gives a person the right to join
with other adults ID making political decisions and settling
disputes. This is because politlcal decision-making and dis
pute setUement are responsibilities of adult members ol an
egalitarlan society. Among Mbuti and Lovedu both sexes give
and receive lood. Lovedu women give ami receive catUe and
may marry a wile with them; they become husbands in social
status. Pondo women, though they are social producen,
cannot dlnpose ol the most important exchange item: live
stock. Perhaps the explanation les in the nature of Pondo
production lor exchange. Women's agricultural work is for
use; work s geared over the long and short run to the needs
ol the households. But men's organization lor livestock raid
ing involves them in production for exchange. Over the short
run, warfare is geared more to the power need 01 a cruel,' who
keeps a loUowing by having catUe to dlstribute, than to
household needs. Over the long run a chief keeps power by
actually distributing catUe more or less widely. He owns the
cattle captured in warlare, but sooner or later he distributes
them among the warriors by virtue of their role in raiding.
These livestock are the chief's lo allocate. They are the most
important tem ol exchange (in bridewealth, loans, and
feasts], and ol establishing long-term relationships (marriage
and service). Because Pondo women do not partcpate in
production lor exchange [radlng], they cannot dispose of the
property which establishes these power relationships. Thus,
they do not hold overt political power.
Lovedu women hold poltcal olfice, enter the decision
making arenas ol the society, and predominate in officiating
in religious rituals on behall of their lineages. Ganda peasant
women are barred lrom even the minimal access to poltical
positions available to peasant meno Yet the mother and one
sister ol the ldng do hold important olfices and exercise sorne
power predicated on their relationship to the king.
A wile's position vis-a-vis her husband is based on her
ownership, or lack ol it, ol the marital estate. In Lovedu,
Pondo, and Ganda, productive resources are inherited patri
lineally. Here there is a contradiction, or opposition, between
226 Karen Sacks
the tact thnt production is organized in a social oro public
way, but that lamilies or individuals appropriate and inherit
the productive resources. A wte does not participate in the
ownership ol resources ot her marital household. On the
other hand, the Mbuti's appropriation tor use by families
seems to me qualltatvely dilferent from the appropriation
tor inheritance and exchange ot the other societies. Mbuti
resources are owned by the territorial band as a whole. Resi
dence entitles a person to use these, and there is no nher
itance. Thus, Mbuti husbands and wives have the same rela
tionship to the band resources.
Lovedu, Pondo, and Canda wives labor tor their husbands
and their husbands' patrkin, but 'do not belong to the group
that appropriates the product ol their labor. Wives provide
heirs, raise child.ren, and do the bulk ot the domestic work
under the authority ot the husband and his kin. They do not
represent the household to outsiders. By contrast, Mbuti mar
riage carries no restrictions on a woman's authority over her
work, children, or socializing. Her tertility cannot be said to
be private since her husband receives no compensation tor
her extramaritaJ sexual relationships.
Menstrual and pregnancy restrictions on women's activities
among Lovedu, Pondo, and Ganda seem to operate to sepa
rate women's reproductive tunctions trom contact with the
social production ot exchange goods; that is, trom contad
with warriors, cattle, crait, and some medical practices. In
these three societies children inherit property and continue
the tamily line. Regardless ot how women's productive activi
ties are organized, their reproductive potential is prvate, But
among the Mbuti, where children are social members rather
than private heirs, menstmation and pregnancy are not sur
rounded by any such restrictons. This contrast suggests that
menstrual and pregnancy restrictions are based on private
property, antl that they serve to symbolize a contradiction
between social production ot exchange goods and private or
lamilial appropriation. Since men are also involved in the
reproductive pracess, and contain the same contradiction.
logically they should-and actually they do-tace analagous
Enge/s Reuisited 227
restrictions. Lovedu, Pondo, and Ganda men must separate
sexual relations trom their participation in social production
lor exchange. By contrast Mbuti regard the collective hunt as
an ideal time lor sexual lialsons,
A final point rernains. Though Canda is a class society, 1
have not dealt with the diflerences between women of rulng
and peasant tamilies. There are severa! privileges accorded to
wives, sisters, and daughters ol the king. Each category ot
ruling-class women shares sorne privilege with ruling-class
men which distinguishes them lrom peasant women: fredom
tram productive labor for some wives; sexual freedom tor
sisters and daughters; political and economic power lor the
....
queen mother and sister. But none ot these women has all the
privileges of the men of their class, which seems 10 reflect the
contradictory position ot ruling-class women: they are ot a
privileged class, but ol a subordinate sexo I have not dealt
with thern in depth because their existence does not reaHy
change the generalizations made on the basis ot peasant
women. This should not be surprising il we recall Queen
Victoria and her times in England, but it should rnake one
wary ol generalzatons based on a lew women holding prom
inent positions.
Though I have separated women's position as wives frorn
their position as social beings, in reality the two are inter
related. Wilely subservience reduces the ability ol Lovedu
and Pondo women to exercise their social prerogatives. They
are held back trom social activities to the extent that they
work under the authority ol husband and his kn. Similarly,
while Pondo women may become diviners, and while most
diviners are women-allowing the women opportunities tor
travel, soclalzng, and financial reward-they may not be
initiated to practice without their husband's consent.
Things can work the other way also. It a woman is socially
regarded as an adult, this can limit the extent to which ahe
can be subordinated as a wile. Thus. while a Pondo woman's
lertility may be said to belong to her husband, and while he
may claim compensation lor her extrarnarital sexual altairs,
this is a matter between meno Women regard these aflairs as
,.. u ... u, "" ....U ..,... I
proper, and are assisted in arranging thefll by their husband's
own kinswomen. Moreover, should a wo*,an choose to end a
marriage or vlslt her own kin, there is lidUe her husband can
do to prevent it.
I have suggested that there are two
position-women as social adults, and
that these can vary somewhat indepen enUy. What deter
mines how, or whether, women are reg ded as adults is not
the same as what detennines their po itions vis-a-vis their
hwsbands. Basically, women are social adulta whe,re they
\yor collectively as pan ol a productive group larger than or
sepuate lrom t or domestlc es 15 me . he meaning and
status ol "wife,"""lhough. epen on thesture ol the lamily
in much the way Engels suggests. Where he estate is lamilial,
and the wife works lor it but does not sh in its ownership,
she is in much the same relationship to er husband and his
kin as is a worker to his boss. Where here are no private
estates, or perhaps where the family es te is jointly owned,
the domestic relationship is a more eg it.arian one (Friedl,
1967). This last point is overstated, sin e the domestic and
social spheres ol lile are not really inde mdent, On the basis
ol the American experience, lt is dilfic lt to conceive ol a
completely egalitarian domestic relation hip when only the
male partner is regarded as fully adult byond the bounds of
the household. J
Women in Class Societies: A Reinterprettion .
1f we agree that the position ol decli'ned from
Mbuti and Lovedu to Ganda, as aboye, in direct
correlation to the domestication ol wo en's work and the.
development of production for exchan and private prop
erty, it is tempting to conclude that E gels was right alter
all-that prvate property and ;>roducti<f lor exchange lead
to women's domestication and subordinr.tion. Many anthro
pologists accept something like E'ew ol the relation
ship between private property and the owth ol social in
equality and classes. While 1 suspect th. t Women in general
AI,t6Coh' ... '-c.".", ...
stand in more equal relationship to men in nonclass soceties
than in class societles, I do not think that male propero
QJIQlcnhip is the basis lor the male's supremacy. First, not all
males own (![oductive propeltr. Second, In many clan
ocieties-even m those wlEh attero 01 mate
dommance-women as we lllI men own productive property,
arnra- wlfe's ownership ol properu gives her a substatl
Q.unt ol domestic power vis-a-viii ber bllsba
nd
(ibid.). But
class societies make a sharp dchotomy between the domestic
and public spberes ol lile, and th is domestic power is not
translatable into social power or postion in the public
spbere. Moreover, in class societies tbe economic and poltical
autonomy ol a bousebold is quite restricted. Tbus, in neces
sary dealings in tbe public sector women are at an overt or
covert disadvantage. This probably militates against even
domestic equality.
It seems likely, then, tbat in class societies tbe subordinate
position of women derives not from domestic property rela
tions but from sometbing outside tbe housebold whicb denies
women adult social status, Tbe question is tben wby do male
publc power and ideais of male socia! dominance pre
'domnate in class societies? For an explanation, tbe fOl'US bas
to shift from tbe domestic to tbe societallevel.
We have seen that public or socia! labor is tbe material
basis for adult social status, lt foUows tbat a society would
have to exclude women from public labor or in some way
denigrate women's perfonnance of sucb labor in order to
deny tbem social adultbood for any lengtb of time. * Tbe
fonner seems to have been the case, at least for many pre
capitalist agrarian states oC Eurasia (Boserup, 1970). Leaving
aside lor the moment tbe apparent exception of industrial
capitalism, wbat were the circumstances tbat may bave led
class societies te exclude women from socia! production?
An euJier ....nion or lhil paper arrued lhal all precepilaHst el...
locietiea excluded women rrom public labor. Kalbleen Goulh hu
polnted out (penonal communication) thal lbia wu nol lbe cae in
precapilaliallndian ltalea. I haye lhu. modified lhis lediaR.
230 Karen Sacks
Class societies are exploitative, which means that many
people must work (or the benelit ot a few. While tithes and
taxes on domestically produced goods can serve this end,
even agrarian societies do not rely exclusively, or even
mainly, on this fonn of production. Corve for public works,
both sumptuary and productive, conscription and predatory
war, and collective agricultural or wage work tor the rulen
a1l collective fonns of social or public labor--are important
productive activities in class societies. While these may not
necessarily seem large trom the local viewpoint, they are
crucial nationally-for creating the "surpluses" by which
rolers and their states are maintained.
Though women may or may not engage in domestic agrio
culture, they seem rarely to participate in these large-scale
forms o( social production. It seems that class societies tend
to socialize the work of men and domesticate that ot women.
This creates the material and organizational foundations for
denying that women are adults and allows ruling classes to
define them as wards of meno
But why would this happen in a class society? With the
c'Jevelopment of socialized production for a ruling class,
domestic production (or subsistence becomes more preeari
ous, forcing people into greater reliance on production (or
exchange-Iaboring for the rulers in exchange for their subsis
tence (arternatively, rulen can force people to work for them
as a condition of access to subsistence resources). Ruling
~ l a s s e s tend to select men as social laborers partly because
they are more mobile, but probably more signicantly be
cause they can be more intensively exploited than women,
not having to nurse and real children. t ... \ 4- IJ L ~ ~
Alice Clarlc (1968) provides rather gruesome data from
seventeenth-century England, a period preceding and setting
the social conditions tor later industrialization. Peasants were
being torced off the land and swelling a class o( rural, landless
laborers. The idea ot wages as something paid (or a task was
not yet fuUy institutionalized; and it contlicted .with the
earlier notion that an employer was in some way obUgated to
meet the subsistence needs of the worlcer. Yet payments were
Enge/s Revisited 231
so low that a landless family had difticulty surviving. Aman
or woman without children could survive, but prevailing re
muneration did not allow for reproduction and rearing of the
next generation of laborers. Indeed, they did not reproduce
themselves. Clark shows that the laboring class grew in size
only from constant new recruits from the peasantry. Women
and children were deliberately excluded lrom wage work by
employers, who felt an obligation to, but could not or would
not, bear the burden ot supporting nonprodr,ctive depen
dents. In human tenns the results were the abandonment of
women and theirearly death, and in organizational terms a
largely male public labor force.
Once such a dichotomy is made-women in domestic work
(or family use, men in social production for exchange-there
is an organizational. basis for a sexual divide-and-rule policy.
Whether such policy is conscious or not is irrelevant. The
effect ot state legal systems and other aspects of ideology
developed mainly by ruling classes has been to convert differ
ences between men and wornen in terms oi their roles in
production into differential worth. Through their labor men
are social adults; women aredornestc wards.
Men are more directly exploited and more ofien collec
tively so-a stuation which gves them the possibility of
doing something about it, Women's field of activity and.
major responsibility s restricted to the household, which
neither produces nor owns the means of production lor more
than domestic subsistence, a level ot organization at which
little can be done to institute social change in a class society.
This situation has severaJ consequences. First, women are
relegated to the bottom of a social pecking order (a man's
home is his castle), Second, because of their isolation and
exclusion from the public sector, women can be used as a
conservative force, unconsciously upholding the status quo in
their commitment to the values surrounding maintenance of
home, (amUy, and children. Finally, the famUy is the sole
institution with responsibility for consumption and for the
maintenance o( its members and rearing ot its children, the
futre generation o( exchange workers. It is necessary labor
232 Karen Sacks
for the rulers, but women are forcedlto perfonn it
compensation.
Modem capitalism hu maintained pattern of explot
ing the private domes tic labor of wom n, but since industrial
ization women have also been involv d heavily in public or
wage labor. Meeting the labor burden that capitalism places
on the tamily remains socially w men's responsbility.
Responsibility for domestic work is o e of the material bases
for present barriers to women workil g lor money and lor
placing them in a more exploitable po ition than men in the
public labor force. As Margai'et Bens on (1969) shows, this
domestic work s not considered "re .. work because it has
only private use value and no exchang value--it is not public
labor. Women's greater exploltability] in the modern wage
labor force may derive from a preindustrial adaptation to
being excluded from public labor oro "caJly, because women
were less exploitable in a pre-wage m lieu). Only after they
had been defined as inadequate lor ublic labor were the
1
conditions right for industrial caPitalijm to discover women
as a source of cheap labor.
However, there have been precaPfaJist societies where
women have participated in social pr duction, On the one
hand, ths means that the exclusion ol women is not a neces
sary condition lOl their exploitation, or lor sexual divide
and-rule. On the other hand, the POSi!:t"n ol women involved
in such labor seems to reinforee the th is that social labor is
the material basis of social adulthood It suggests, too, that
social adulthood is not synonymous Ith sexual equality in
class :locieties.
At least some of the precapitalist of India contained
a class ol state slaves-lor the Chola and
Vijayanagar empires.. Both men and women of this elus,
which was recruited from the "exte 'or" or Untouchable
castes, served as agricultural laboren lor religious, military,
and govemment officials, as well as corved lor public
AIl the nfoonation on India has been lenerolllly supplled by
Kalhleen Oough.
Engels Reuisited 233
works. However, women were paid considerably less Ior their
labor than meno At the same time, Gough points out that in
Untouchable tenant-farming and village-service castes or
classes, where women work today for village communities in
similar relationships of production, they "have greater sexual
freedom, power of divorce, authority to speak and witness in
caste assemblies, authority over children, ability to dispose of
their own belongings, rights to indemnity for wrongs done 1;0
them, rights to have disputes settled outside the domest.ic
sphere, and representation in public rituals." In short, women
who perform social labor have a higher status yjs-a-vis men of
jheir OWD class .than do Women who labor only in the domes
tic sphere or do no labor..
In sum, I aro suggesting two patterns: (1) Intensive exploi
tation in social produetion by and for ruling classes favored
making this men's work. In tum, ruling classes capitaJized on
the situation, legitimizing the division of labor by a thorough
going system ol differential worth. In retum for the loss of
economic autonorny, they conerred upon men exclusive
social adulthood and guardianship of women. Under these
circumstances, even if women own property the state inter
yenes to limit what they can do with it publicly, and to
subordinate the household to the larger society. (2) States
incorporate women of the poorer or propertyless classes into
social production. Here there is a "second Une of dp.(ense"
against equality nstitutionaliaed through pay differentials.
While these women are sociaJ adults with respect to men of
their class, economic policies prevent actual equaJity. The
key aspect o( women's position, especially in class societies,
i.s social adulthood, and this comes from participation in
social production.
This brief examination of the bases of women's domestic
and sociJ status suggests sorne tentative conclusions about
the kinds ol economic and social changes necessaI}' for full
sexual equaJity. While property ownership seems important
for women's domestic position vis-a-vis a husband, the exer
cise of domestic power, particularly in class societies, is
234 Karen Sachs
Jimited by whether or not women have adult status 'in the
social sphere. This in tum is determined by their partic
patio n in social production. But the dichotomization of
lamily and society, which is especially strong in class soci
eties, malees women responsible for the production 01 prvate
use value and makes men responsible for the production 01
exchange values. The distinction between production lor use
and production for exchange places a heavy responsibility on
women lo maintain themselves as well as exchange workers
and to rear future exchange and maintenance workers. In this
context, wage work (or social labor) becomes an additonal
burden and in no way changes women's responsibility tor
domestic work. For full social equality, men's and women's
work must be 01 the same kind: the production 01 social use
values. For this to happen, tamily and society cannot remain
seprate economic spheres al Iife. Production, consumption,
child-rearing, and economic decision-rnaking all need to take
place in a single social sphere-something analogous to the
Iroquois gens as described by Engels, or to the production
brigades al China during the Great Leap Forward. What is
now prvate family work must become public work for
_ ~ m e ~ n to become fully social adults.
Judith K. Brown
Iroquois Women: An Ethnohistoric Note
My purpose is to investigate the relationship between the
position. 01 women and their economic role. At Ieast three
possibities are suggested in theIlterature, Robert H. Lowie
(1961:201) felt that the two were unrelated, that in deter
mining women's status, economic considerations could be
"offset and even negatlved" by historical factors. On the
other hand, Bronislaw Malinowski (1913) maintained that
the considerable economic contribution of Australian aborig
ine women confirmed their subservient position, since their
labors were extorted from thern through male "brutaliza
tion." The opposite point of view is expressed by Jenness:
1( women amone the Iroquois enjoyed more privileges and pos
sessed (leater Creedom than the women o( other tribes, this was
due ... to the Important place that acriculture held in ther eco
nomic Ii(e, and the dlstribution o( labor ... [which leftl the
enUre cultivation ot the tields and the acquisitlon o( the greater
part gt the tood supply to the women. (1932: 137)
His explanation for the high status of women among the
Iroquois stresses the extensiveness al their economic contri
bution. A similar position is taken by B. H. Quain (1961),
Thia ia a reviaed and shertened venion oCBrown, 1970. Th&llks are due
to the edlton ol Ethnahtory Cor penniuion to use portions oC lhat
papero Tb. reMarch waa made pouible by lhe lleneroul support ol the
aadellCCe lnatitute. MUda Kahne. Yin, Yin, Yuan, and Peter Bertocci
hlve made numerous helpCul eQmmenl.s on prevous venians oC the
manulcript.
235