REVIEW ESSAY

The New Scholarship on Family

Violence

Wini Breines and Linda Gordon

Only a few decades ago, the term"familyviolence" would have had no

meaning: child abuse, wifebeating,and incestwould have been under-
stood but not recognized as serious social problems. Child abuse was
"discovered" as a social problem in this countryin the 1870s when a
combination of factors-immigration, urbanization, changing social
roles for children, and a growing social reform movement in part
sparked by a women's rebellion-led to a refirmulationof parent-child
relations.In the late nineteenthcentury,feministsalso forced the rec-
ognition of wife beating as reprehensibleactivity,but it never gained
recognitionas a widespread social problem. The anti-childabuse cam-
paign continued in the Progressiveperiod, but lostvisibility
again in the
mid-twentieth century.
t'he 1960s and 1970s have seen a passionate awakeningof concern
withboth child and wifeabuse, and most recentlywithincest.Before we
can look at contemporaryresearchon these problems,however,it is im-
We are indebted to Inany people for help in findingoutrway thr(oughthis literature
and this new "field." R. Emertsonand Russell Dobash, Dair 1I. (illespie, Ann ,Leffler,
Car(olvni Newberger,Elizabeth Ileck, Herb Schreier,and Barrie Tho(rne took the time to
rea(d through an early draft of this and \rote out detailed c)lioments. Barrie 'Iho-rne's
work, in particular,on an unwield(l manuscril)tis greatlyappreciated. We have gained
insightsfrom discussionof these and related issues withAnne Flitcraft,
Anne Kenney,Jan
Lambertz,Paul ()'Keefe, Elizabeth Pleck,Paul Roserkrantz,and Evan Stark.Manyof these
colleagues made ciiticisimstlhatwe werTeunable to meet, 1rdid not agree with,and in the
end we mustaccept responsibility forwhatappears here. ILindaGordonlis also indelbtedto
NIMH Research (rant #R()--MH33264 for making part of this studypossible.
[SigA\:jolurnalof Womenin CultureiandSocietyI98)3. vol. 8, lo. 3]
? 1983 Iy The University of (Chicago. All ights reserved. ()00 -)7-10/(83/080-000()()8$01.()()

490
This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 491

portantto clarifythedistinctionbetweenthewidespread practiceof these

types of violence, and their labeling as a social problem. Such labeling
need not be related to any change in the actual incidence of violence.
Several factorscontributed,we suspect,to the concern about these
abuses that has emerged since about 1955. First, the last twenty-five
years marked a criticalperiod in the much-described"crisis"of the fam-
ily,of whichfamilyviolence was seen as a symptom.Second, the general
tendency toward permissive and child-centered parenthood in the
post-WorldWar II period mayhave made violentcoercion in the family
less tolerable than in the past. Third, the women's movement has
brought problems previouslyconsidered personal to public attention,
and it has particularlyrenewed critical scrutinyof the family,which
feministsidentifiedas an importantstructureof domination.Fourth,the
1960s and 1970s also brought a culture of self-exposure--commercial,
personal, and artistic;a decline in an older etiquette of modestyand
privacy about personal life; and a new acceptabilityof a confessional
mode. This particularaspect of "modernity,"a long-rangetrend which
reached an extremein the 1970s, contributedto the opening up of areas
of lifeonce hidden. At the same time,it is importantto add, urban and
housing developmentsof the twentiethcenturybrought an increase in
privacywhich made it possible for familyviolence to be hidden, among
the richand poor alike. Fifth,the 1960s also representeda low pointin the
influenceof religiouslybased moralismin the United States,and a high
point in environmentalistsocial thought; child abusers, for example,
could be considered victimsthemselves,in need of help rather than
punishment,and theirexposure thus encouraged.
Unfortunately,verylittlehistoryhas yet been writtenabout family
violence and mostof the workreviewedin thisessay is writtenwithouta
good knowledge of its own historicalcontext.1Nevertheless,the rapid
spread of concern in the past decades should be at least mentioned:after
the 1962 publicationof a keyarticleon the "batteredchild syndrome,"2
there was an explosion of attentionto child abuse. By 1967 all the states
had passed laws making reportsof child abuse mandatory.In 1974 the
federal governmentestablished a center for the identificationand pre-
ventionof child abuse. Meanwhile, the women's movementforced the
expenditure of federal and state monies on batteredwomen's shelters,
rape crisiscenters,and research on wife beating.

1. Among the rare pieces of historyon thistopic are Elizabeth Pleck,"Wife-beatingin

Nineteenth-CenturyAmerica," Victimology 4, no. 1 (Fall 1979): 62-74; Nancy Tomes, "A
'Torrent of Abuse': Crimes of Violence between Working-ClassMen and Women in Lon-
don, 1840-1875,"JournalofSocial History11, no. 3 (Spring 1978): 328-45; and Jan Lam-
bertz,"Violence against Women in 19th-CenturyEngland" (B.A. thesis,Harvard Univer-
sity,1980).
2. C. Henry Kempe et al., "The Battered Child Syndrome,"Journalof theAmerican
MedicalAssociation181 (1962): 17-24.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
492 Breinesand Gordon Review:FamilyViolence

The quantityof materialcovered in this reviewmay suggestto the

nonspecialist reader how quickly the field of familyviolence has at-
tractedresearchers,particularlysociologists.Those workingin the held,
however,will findmuch thatis omittedhere. For reasons of coherence
and lengthwe have concentratedon scholarlyworkin the United States,
that is, on original research; we have omittedstrategicand service-ori-
ented writing.Also omittedare studies of siblingviolence and violence
against the elderly-new and importantareas of study-and studies on
the largestformof mistreatment of children,child neglect,because most
of the researchunder the familyviolence rubricdoes so (withsome im-
portantexceptionsthatwe willnote below). This lastomissionseems to us
politicallysignificant:research concentrateson individuals maltreating
childrenand not on the much largerincidenceof childrenvictimizedby
poverty,inadequate child care, inadequate educational institutions, and
so on, problems for which the social and economic organizationof our
society is most clearly to blame. Furthermore,child neglect, when
studied at all, is defined in terms that reflecta class bias: these terms
emphasize the symptomsof povertyand downplay formsof emotional
neglect that are likely to be at least as common in more prosperous
families.
We reviewhere workson child abuse, woman battering,and incest,
in that order. We have adopted the rubric "familyviolence" only re-
luctantlybecause aggregating these problems tends to obscure dif-
ferences that are essential and, indeed, may obscure the heart of the
problems. We will point to these differencesand argue that, before
synthesisis possible,all formsof familyviolence require furtherseparate
analysis.We also argue that most scholarshipon familyviolence has a
weak understandingof family,of gender, and of power.
In thisconnectionwe make argumentsrelevantto all three typesof
familyviolence discussed in thisessay. First,all violence mustbe seen in
the contextof wider power relations;violence is not necessarilydeviant
or fundamentallydifferentfrom other means of exerting power over
another person. Thus, violence cannot be accuratelyviewed as a set of
isolatedeventsbut mustbe placed in an entiresocial context.Second, the
social contexts of familyviolence have gender and generational in-
equalities at their heart. There are patterns to violence between in-
timates which only an analysis of gender and of its centralityto the
familycan illuminate. Using these insightsleads us to the third argu-
ment,a critiqueof the fborm of empiricismthatdominates thisresearch
and writing.We are committedto the empirical,to work that reports
actual experience and evidence as exactly as possible. Such empirical
work can be theoreticalas well. Empiricist research, by contrast,relies
exclusively on measurable, instrumentalized data and on observableevi-
dence, asking only questions answerable by yes or no, or withnumbers.
Family violence involves questions of social and cultural meaning that

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 493

are not alwaysquantifiablein thisway. Intimateand familyrelationships

are filledwithcontradiction,withlongingsand expectationsso ambigu-
ous and ambivalent that they may not be conscious. Research on this
subject, therefore,needs qualitative and theoreticalanalysis as well as
empirical observation.
We writefroma feministperspective;thatis to say,we see the family
and familyviolence, like all other historicalphenomena, as produced
within a gendered society in which male power dominates. Feminist
analyses of the family,however, vary greatly.All flow from a concern
withwomen's rightsand freedom,but theycome to verydifferentcon-
clusions and have differentsuggestionsforchange. While fullyaware of
the limitationson what researcherssuch as ourselvescan offerto prac-
titionersand activists,we offerour criticismand interpretationin the
belief that the theoreticaland political debate is just beginning in ear-
nest.

I. ChildAbuse
The last two decades have produced a vast literatureon child abuse
by researchers and cliniciansin differentfields.We look at the central
issues primarilythrough a historicalgrouping of this scholarship into
threegeneral areas: psychological,societal,and sociologicalexplanations
of the problem.
In the late nineteenthcentury,concernwithchildabuse was defined
by child protectionagencies, run firstby volunteercharityworkersand
later by professionalsocial workers.In contrastthe most recent history
has been shaped primarilyby physiciansand to a lesser extentby legal
personnel.3 In 1962, pediatrician Henry Kempe identifiedwhat he
called the batteredchild syndrome-broken bones or other evidence of
physical trauma sometimes revealed by X-rays. These circumstances
generated, not surprisingly,a narrow definitionof child abuse that fo-
cused on specificphysicalinjuriesand symptomsthatcould be diagnosed
primarilyby physicians.In fact,a general definitionof abuse and neglect
is an extremelycomplex and criticalissue, whichwillbe discussed below;
it is enough to point out here thatan initiallynarrowdefinitionhas left
its mark on child abuse scholarshipof the last two decades.

Psychological
Explanations
Beginning in the early 1960s, pediatricians and psychiatristslike
Kempe, Brandt Steele, and Carl Pollack blamed the batteredchild syn-
3. See Eli Newbergerand Richard Bourne, "The Medicalizationand Legalization of
Child Abuse," in CriticalPerspectives
on ChildAbuse,ed. Richard Bourne and Eli Newberger
(Lexington, Mass.: Lexington Books, 1979), pp. 139-56.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
494 Breinesand Gordon Review: Family Violence

drome on the psychologicalproblems of the parents.4Abusive parents

were labeled pathological,deviant,neurotic,or psychotic.Such charac-
teristics,however, did not distinguishthem from a large sector of the
population who are not abusers, a serious problemwiththisapproach.5
A varietyof more specificpsychologicalexplanationssoon emerged.
One, mostsuccinctly summarizedbyHenryKempe and RuthKempe, was
that child abusers are immature individuals who expect the child to
behave like a grown-up and interpretthe child's inabilityto control
himselfas willful,hostile,and/orexcessivelydemanding.6A variationof
this theme is the concept of "role reversal,"7which postulatesthe par-
ents' desire for love and approval from the child, as though the child
were the parent.
The great majorityof child abuse experts believes that there is a
"cycle of abuse," in which violence is repeated fromone generation to
the next-the repetitionof a distortedrelationshipthat deprives chil-
dren of the consistentnurturingneeded for their full development.8
The psychological and sociological schools explain the problem dif-
ferently.9 The latteruse a social learningmodel, to whichwe willreturn
later.The psychoanalytically inclinedpsychologistsexplain it in termsof
the unconscious; Ruth Kempe and Henry Kempe, forexample, suggest
thatindividualsrepeat behavioror experiences fromtheirfirsttwoyears
of life,when theirawareness of such experiences could only be nonver-
bal. Parents then reproduce these preverbal abusive nmemories when

4. Kempe et al. (n. 2 above); Brandt Steele and Carl Pollack, "A PsychiatricStudy of
ParentsWho Abuse Infantsand Small Children,"in The BatteredChild,ed. Ray Helfer and
C. Henry Kempe (Chicago: Universityof Chicago Press, 1968), pp. 103-47. Also see
Leontine Young, Wednesday's Child(New York: McGraw-Hill Book Co., 1964).
5. See Richard (;elles, "(hild Abuse and Psychopathology:A Sociological Critique
and Reformulation,"in FamilyViolence,ed. Richard (elles (Beverly Hills, Calif.: Sage
Publications,1979), pp. 27-41, forthiscritiqueof much of the psychologicalliterature.See
also Edward Zigler,"ControllingChild Abuse in America: An EffortDoolmedto Failure,"
in ChildAbuseand Violence,ed. David Gil (New York: AMS Press, 1979), pp. 37-48.
6. Ruth S. Kempe and( C. Henry Kempe, ChildAbuse (Cambridge, Mass.: Harvard
UniversityPress, 1978), pp. 18-20. Also see Blair and RitaJustice,TheAbusingFamily(New
York: Human Sciences Press, 1976), p. 40; and Steele and Pollack (n. 4 above).
7. See, e.g., M. G. Morrisand R. W. Gould, "Role Reversal:A (oncept in Dealing with
the Neglected Battered-Child Synldrome,"in The NeglectedBattered-Child Syndrome: Role
Reversalin Parents(New York: Child Welfare League of America, 1973).
8. Kempe and Kempe (n. 6 above), p. 12.
9. Examples fromithe psychological literature include John Spinetta and David
Rigler,"T he Child-abusingParent: A PsychologicalReview,"and otherarticlesin Traumatic
Abuseand theNeglect(f Childrenat Home, ed. Gertrude Williamsand John Money (Balti-
more: Johns Hopkins UniversityPress, 1980); Steele and Pollack (n. 4 above); Kempe and
Kempe (n. 6 above), pp. 12 ff. Examples from the sociological school include Sidney
Wassermtan,"h'le Abused Parent of the Abused Child," in Violencein theFamily,ed.
Suzanne Steinmetzand Murrav Straus (New York: Harper & Row Publishers,1974), pp.
222-29; and Murray Straus, Richard Gelles, and Suzanne Steinmetz,BehindClosedDoors:
Violencein theAmericanFamily((arden City,N.Y.: Anchor Books, 1980).

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 495

interactingwiththeirown children,eitherbecause theyare now able to

be the strong,aggressiveparenttheyalwayswanted to be, or because the
fear and sense of rejectiontheyfeltas childrenare triggeredwhen they
perceive theirown children rejectingthem.10The Kempes suggestthat
we are constrainedas parentsby the nature of our own experiences,the
implicationbeing that,barringintervention,the abusive cycle will con-
tinue unbroken.1' In other words,the parent is not able to postpone his
or her childlikeneeds in order to functionas a parent. Related to these
theories is Carolyn Newberger's hypothesisthat the parents' level of
awareness about the process of child developmentand about theirown
parental responsibilitiesbears importantrelationto child abuse and ne-
glect.12
Gender is one of the major blind spots in thisliterature.It is rarely
acknowledgedthatwomen,not genderlessparents,desire babies, or that
women do the parenting.13One study of young women who refused
contraceptionand abortion because of a strongdesire to bear a child,
but who then abused their infants,indicates that they recreated in
their marriages the isolation and miseryof their parental homes and
were drawn to motherhood as a solution to their unhappiness. The
researcher,Dr. Ray Helfer,suggeststhatlike theirparentsbeforethem,
these motherswanted the baby to keep them company or to "role re-
verse and . . . to parent the parents."'4
Only one commentator,Gertrude Williams, points out the pro-
foundlyand traditionallyfemale issues ignored by Helfer's study. She
suggests that sexism and pronatalism have taught these women that
motherhood is women's only fulfillingactivity.In Williams's opinion
they would probably not have become child abusers had the culture
offered them viable alternativesto marriage and motherhood.15Her
commentssuggestthe necessityof a feministperspectiveon child abuse,
a perspectivethatmakes gender centralto an understandingof all forms
of intimateviolence. There are many gender-relatedissues, including
the fact that women are primaryparents, the lack of interestingand

10. Kempe and Kempe (n. 6 above), pp. 13-14. See also Selma Fraiberg,EveryChild's
Birthright:In Defenseqf Mothering (New York: Bantam Books, 1978), pp. 31-35.
11. See Richard (;elles, "Violence in the Family: A Review of Research in the Sev-
enties,"JournalqfM,arnrage and theFamily42, no. 4 (November 1980): 873-85, esp. 879, for
a discussion of caveats about intergenerationalviolence findings.
12. Carolyn Moore Newbergerand Eli Newberger,"The Etiologyof Child Abuse," in
ChildAbuse:A MedicalReference, ed. Norman F. Ellerstein(New York: John Wiley& Sons,
1981), pp. 11-20.
13. See Carolyn Kott Washburne,"Some Thoughts on a FeministAnalysisof Child
Abuse and Neglect" (paper presented at the National Conference for Family Violence
Researchers,Durham, New Hampshire,July21-24, 1981).
14. Ray E. Helfer, ChildAbuseand Neglect:The DiagnosticProcessTreatment Programs
(Washington,D.C.: U.S. Departmentof Health, Education, and Welfare, 1975).
15. (ertrude Williams, "Toward the Eradication of Child Abuse and Neglect at
Home," in Williamsand Money, ecls. (n. 9 above), pp. 588-605, esp. p. 597.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
496 Breinesand Gordon Review:FamilyViolence

well-paid work options for women, the inadequacy of contraceptive

methodsand thus the number of unwanted pregnancies,and the effect
of a pronatalistculture. Thus far,however,nonfeministclinicianshave
ignored the gender of both parent and child, and few feministshave
published commentarieson child abuse.

MaternalBonding

Psychological findings about the intergenerationalcontinuityof

abuse have dovetailed with theories of maternal bonding, creating a
loose congruencebetweenthe two. Bonding or attachmenttheory,based
on mammalian mothers and babies, postulates a critical period, im-
mediatelyafter birth,in which certain patternsof nesting,grooming,
and care behavior are manifestedby the mother.Even briefor partial
separations during this period may distortfeeding and care of the in-
fant,and, in the theorists'words, resultin deviant maternalbehavior in
mammals and humans. 6 According to John Money and Andrea
Needleman, defective mother-infantpair bonding, sometimes called
maternal deprivation,may lead to schizophrenia,child battering,and
abuse dwarfism(growthretardationin children).17The reasoning then
goes thatidentification of personalityand familyrelationshipspredictive
of child battering-a screening that may begin during pregnancy,at
delivery,or in the periodjust afterbirth-enables professionalinterven-
tion before the firstbattering.l1
In a criticalreviewof maternalbonding, WilliamRay Arneyargues
that such generalizations about human mother-childattachmentare
sloppily taken from primate studies and are scientifically unfounded.
There is no proof of a criticalperiod in the human mother-infant re-
lationship. Furthermore, as Arney points out, there is an antifeminist
intentbehind the currentrevivalof interestin the centralityof the fam-
ily; widespread acceptance of maternalbonding promoteswoman's role
as primarycaretaker,her bond with the infantas her most important
relationship.19
Ironically,while gender is ignored in the child abuse literature,it is
at the same time assumed that women are the primary parents and
consequentlythat their"bonds" withthe infantare critical.Attachment

16. Marshall Klaus and John Kennell, "Mothers Separated fromtheir Nerwb-orn In-
fants,"in Williamsand Money, eds. (n. 9 above), pp. 208-27.
17. John Money and Andrea Needleman, "Impaired Mother-InfantPair Bonding in
the Syndrome of Abuse Dwarfism: Possible Prenatal, Perinatal, and Neonatal Antece-
dents," in Williamsand Money, eds. (n. 9 above), pp. 228-39, esp. p. 236.
18. Charles SchwarzbeckIII, "Identificationof Infantsat Risk for Child Abuse: Ob-
servationsand Intfetencesin the Examinationof the Mother-InfantDyad," in Williamsand
Money, eds. (n. 9 above), pp. 240-46, esp. p. 240.
19. WilliamrRay Arney,"Maternal-InfantBonding: The Politicsof Falling in love
withYour Child," FeministStudies6, no. 3 (Fall 1980): 547-70.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 497

theoriesmake an issue of the mother'sand not the father'spotentialfor

abuse. The underlying implicationis that mothers are ultimatelyre-
sponsible for the reproduction of abuse through an intergenerational
cycle of poor bonding and parenting. Fathers-and the full array of
social relationsin and around families-are ignored as unimportant.
With theories of maternal bonding as their rationale, Henry and
Ruth Kempe propose widespread use-with the parents'permission-of
an "earlywarningsystem"which,theybelieve,would diminishabuse and
neglect. It is worth noting again that while they often use the term
"parent," they predict difficultyprimarilyon the basis of the mother's
behavior and situation.Their studies show that theycan predictwhich
parentsare likelyto abuse theirchildrenbyclose observationof parental
behavior during threeperiods: beforethe birthof the child,in the labor
and deliveryrooms, and during the firstsix weeks of the infant'slife.
The signalsof potentialtroublemaybe a mother'sconcern about weight
gain during pregnancy,parents who come fromabusive backgrounds,
mothershostileor passive towardthe baby in the deliveryroom, mothers
who fail to look in the baby's eyes or hold it face-to-faceor find it too
demanding at feeding time. The Kempes's screening enables them to
predict successful or unsuccessful parenting in 76.5 percent of their
cases.
A large part of their concern and screening revolves around the
mother'sunambivalentacceptance and love for her child, her nonsep-
aration fromit,and her "acceptance" of motherhoodimmediatelyupon
birth.20They ignore the possibilitythat a mother's relationshipto her
infantmay be influencedby social factors,and theywould deny that a
mother's need for time away from a baby could be a sign of mental
health and autonomy. Certainlythere is no evidence that minimizing
mother-childseparation will preventchild abuse.21
Another assertionof the centralityof the mother-childdyad comes
from a differentquarter altogether.FeministAlice Rossi, who several
years ago, in "A Biosocial Perspectiveon Parenting,"indicated an inter-
est in the evolutionarybiological basis of the mother-infant bond, stated
recently that we can have littleconfidence that anyone but the primary
parent, the mother, can adequately care for a child. Her doubts about
substitutingany alternativeto the biological mother for child care are
based on infant-developmentresearchon special cues and signals given
by infantsto their mothersto which motherslearn to respond.22What

20. For a critiqueof these ideas, see Arney,pp. 564-65.

21. The hospitaltreatmentof prematureinfantshas come in forspecial consideration
because premature infants,who are separated fromtheir mothersfor a prolonged time,
seem to be overrepresentedas abuse and neglectvictims.For a refutationof thisargument,
see Milt Kotelchuck, "Nonorganic Failure to Thrive: The Status of Interactional and
EnvironmentalEtiologicTheories," in Advancesin BehavioralPediatrics1 (1980): 29-51.
22. Alice S. Rossi's comment in JudithLorber, Rose Laub Coser, Alice S. Rossi, and

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
498 Breinesand Gordon Reviezw:
FamilyViolence

we have here is an argumentlike Rossi'searlierone: the maternal-infant

bond is the keydevelopmentalfactorforthe infant,men are unlikelyto
be optimalparents,and itis doubtfulwhetherinstitutionalized child care
can provide for good infantdevelopment. It is mothers and mothers
alone who are motivatedand able to nurture their infants,and who,
implicitly,can harm them."3
In "The Fantasy of the Perfect Mother" Nancy Chodorow and
Susan Contrattorecentlyoffereda criticalperspectiveon thisevaluation
of mothering.They suggest that feministsand antifeminists alike have
created a fantasyof mothersas "larger than life,omnipotent,all power-
ful, or all powerless," and in so doing reifymothers,denying them
subjectivity.The mother-childdyad is conceptuallyisolated and its spe-
cialness exaggerated.24 Ironically, the idealization of mothers also
creates an image of an omnipotentchild whose needs and demands of
maternalcare are taken,bydefinition,as realistic.The child'sexperience
of distressor frustrationis understood as entirelythe mother's fault.
Mothers,withoutlivesof theirown, are defined as people who do or do
not live up to theirchild's expectationsand needs, or to society'sviewof
the child's needs.
Critiquessuch as Chodorow and Contratto'sand Arney'sare begin-
ning to expose problematicassumptionsin maternal-bondingliterature,
assumptionsthat have in turn influencedchild abuse scholarship.First,
the viewof the mother-childunit as isolated fromother influencesleads
to a tendencyin child abuse commentaryto blame mothersexclusively.
Second, proposed screening for child abuse sometimesadopts the un-
proven assumption that there is a criticalperiod of bonding between
mother and infantsoon afterbirth.There is no clear evidence that a
child who has manycaretakersbeforethe age of fiveand is lateradopted
is unable to formattachments,or thatchildrenformattachmentsonlyto
mothers.Many children who have lived in fosterhomes or institutions
are able to formdeep and lastingattachments.25 Third, assumptionsthat
optimal motherhood should be full-timeand exclusive have not been

Nancy C(hodorow,"(n 7he Reproductionl ofthe Methein A lethological

g I)ebate," Signs:
Journalof Inomenin Cultureand Society6, no. 3 (Spring 198)0):482-514, esp. 498. See Alice
Rossi,"A Biosocial Perspectiveon Parenting,"Daeldals 106, no. 2 (Spring 1977): 1-31; and
forcritiquesof Rossi: Wini Breines, Margaret(erullo, and JudithStacey,"Social Biology,
Family Studies, and AntifeministBacklash," FeministStudies4, no. 1 (February 1978):
43-67; Nancy Chodorow, "(Considerationson a Biosocial Perspectiveon Parenting,"Berke-
leyJournal(ofSociology22 (1977-78): 179-97; Arnev (n. 19) ab)oe).
23. See Chodoroow'sreply foi-a refutationof Rossi's argument,especially to the
theory that women's motheringstenmsfi-omtheiribiology and that infantsneed their
biologicalmother("Reply by NancyCho(drowss," in l.orber et al. [n. 22 above], pp. 504-8).
24. Nancy (Chodorowand Susan C(ontratto,"The Fantasyof the PerfectMother,"in
Rethinking theFamily:SomeFeminist edl.Barrie I horne withMarilynt
Question.s, Yalom (New
York: Iongmnan,1982), pp. 54-75, esp. p. 67.
25. Michael Rutter,"Maternal Deprivation, 1972-1978: New Findings, New Con-
cepts, New Appr-oaches,"Annal, 'cadeinmy'ledliine
of 8, no. 3 (July 1979): 312-23.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 499

proven. The concepts of maternal deprivation and maternal bonding

reinforcedeep-seated sexist beliefs in woman's destinyas mother and
her completeresponsibility forthe successesand failuresof her children.
Finally,the proposed screeningfor"high-risk"parents (mothers)is
potentiallydangerous.26 Procedures that allow professionalsand gov-
ernment officials,including the police, to brand individuals, families,
and entire ethnic and racial groups as "problems" must be questioned.
We do not live in a societywhichcan be relied upon to use thisinforma-
tion humanelyor helpfully;ratherit may be used to punish and track,
and to push pronatalist and mother-blamingpolicy. For example, if
therapydoes not work for high-riskparents,will sterilizationbe used?
True, manyof the indicatorsemployed in screeningforchild abuse are
probablyuseful. In particularcases theymayindicatepotentialproblems
and maybe successfulin alertingand helping individuals.However, in a
society that cannot use such informationobjectively,screening is of
questionable use for the establishmentof public policy.

SocietalExplanations
The limitationsof an exclusivelypsychological understandingof
child abuse and neglect expose the staticaspect of its categories: chil-
dren, parents,abuse, neglect. We learn littleabout who the parents or
childrenare, whatkind of societyand culturetheylive in,whattheirlives
are like, what abuse or neglect is, and who is definingit. The contextis
missing. The systematicintroduction of sociological factors into an
understanding of child abuse and neglect came in reaction to this
abstractthinking.The firstsuch effort,David Gil's 1970 analysisof the
causes of child abuse and neglect,ViolenceagainstChildren,based on data
from national studies done in 1965, 1966, and 1967, was in some re-
spectsa long pendulum swingaway fromany psychologicalor individual
factorsand towardthe social context.Gil argued thatchild abuse cannot
be eliminated unless our inegalitarian, competitive, irrational, and
hierarchicalsocietyis transformedinto an egalitarian,cooperative, ra-
tional, and humane one. Schools, courts, day-care centers, welfare
policies that do not provide for children's optimal development,social
conditions such as poverty,racism, substandard medical care, and the
absence of meaningfuljobs are, for Gil, the prime causes of child abuse
and neglect. He reminds us of the elementarysociological observation
that the socioeconomic statusof the familyinto which the child is born
has enormous consequences forits development.Thus he definesabuse
and neglect as "any act of commission or omission by individuals,in-

26. See Eli Newbergeret al., "PediatricSocial Illness: Toward an EtiologicClassifica-

tion,"Pediatrics60 (1977): 178-85, for an analysisthat questions the accuracy of the pre-
dictivepowers of principlediscriminationfeatures.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
500 Breinesand Gordon Review:FamilyViolence

stitutions,or society... whichdeprive[s] childrenof equal rights,liber-

ties,and/orinterfere[s]withtheiroptimal development."27
Finally, Gil apportions some of the blame to the culturally
sanctioneduse of forcein child rearing,whichpermitsparentsto relieve
theirfrustrationby physicalattackson theirchildren. Since the stresses
are greaterin poorer homes, so is the likelihoodof assault; according to
Gil, it is not thatmiddle-classpeople do not believe in using forceor are
less violenttoward theirchildren,but they have less experience of the
stressthattriggersviolence. Gil suggeststhatthe use of forcemayin fact
be functionalin preparing children for inegalitarianand competitive
adult roles. Again, however, gender is absent as a significantcategory
fromhis analysisof child abuse and neglect.

SociologicalExplanations
Today the leading American sociological work on child abuse-
leading in quantityand in recognition-represents enough of a break
with David Gil's work to require separate categorization. The three
best-knownsociologistsin this field are Murray Straus (its founder),
Richard Gelles, and Suzanne Steinmetz.28
Straus,Gelles, and Steinmetzhave emphasized thatchild abuse and
wifebeatingare differentmanifestationsof the general problemof fam-
ilyviolence. Their work underscoresthe similaritiesin patterns,causes,
and remedies of child abuse and other kinds of familyviolence.29They
emphasize the need to depsychologizeand demystify familyviolenceand
to remove it from the deviance category.30Four major themes charac-

27. David C. Gil, "UnravelingChild Abuse," in Gil, ed. (n. 5 above), pp. 3-17, esp. p.
16. See also David Gil, ViolenceagainstChildren(Cambridge, Mass.: Harvard University
Press, 1970).
28. Most of theirwork,and thatof similarresearchers(oftenStraus's students),is in
essays and articles. The exceptions are Richard (elles, The ViolentHome (Beverly Hills,
Calif.: Sage Publications,1972); Suzanne Steinmetz,The Cycleof Violence:Assertive, Aggres-
sive, and AbusiveFamilyInteraction(New York: Praeger Publishers, 1977); and a more
popular book writtenbyall three,BehindClosedDoors(n. 9 above). Manyof theirarticlesare
also in anthologies.Of these, we will refermostoften to: Gelles, ed., FamilyViolence(n. 5
above); Steinmetzand Straus, eds., Violencein theFamily(n. 9 above); Gil, ed., ChildAbuse
and Violence(n. 5 above); Bourne and Newberger,eds., CriticalPerspectives (n. 3 above);
Williamsand Money, eds., TraumaticAbuse(n. 9 above); and Maurice Green, ed., Violence
and theFamily(Boulder, Colo.: WestviewPress for the American Associationfor the Ad-
vancementof Science, 1980).
29. As we have said, we are criticalof thisaggregation,but we mustnote here the ways
in which it affectsthe organizationof our remarks.We discuss child abuse, wifebeating,
and incestseparately,but we are reviewingsociological worksthat generalize about all of
them. As a result,certain themes of the Gelles, Straus, and Steinmetzschool will be dis-
cussed both here and, as theybear on wife beating,in the next section.
30. Richard Gelles, "The Social Constructionof Child Abuse," in Gelles, ed. (n. 5
above), pp. 43-53.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 501

terize their work: (1) rejectionof uniquely psychologicalexplanations;

(2) reliance on empiricistand almost exclusivelyquantitativeevidence;
(3) challenge of the notion thathome and familyare havens of tranquil-
ity;and (4) condemnation of physicalpunishmentin child rearing.
Gelles was the first,in 1973, to denounce the psychologicalinter-
pretationsof child abuse.31 He argued that such interpretationspostu-
lated mental stateswhich could not be proven; that these mental prob-
lems were in any case less importantthan social, economic, and demo-
graphic factors; and that the research of the psychologistswas un-
scientific.Gelles and Straus sensiblysuggest, as does Gil, that the life
situations and life chances of low-income people and minoritiescan
underminefamilyrelationshipsand thus make parentsunable to protect
their children.32For example, their data indicate that unemployment
(particularlythe husband's), social isolation (often a resultof poverty),
and unwanted pregnancies create stressfulsituationsfor parents, and
thus create conditions leading to possible abuse and neglect. The pre-
vention of unwanted children-through availabilityof abortion, birth
control,and sex education-is thereforecriticalin theirpoint of view.
In other words,instead of confiningthemselvesto the studyof how
the parent may behave withthe infantor consideringonly macro-social
factors(as does Gil), Gelles and Straus focus on the intersectionof these
structuralconsiderationswithpersonal ones. They thus move away from
psychologicalor social abstractionsand toward a more mediated, com-
plex descriptionof actual experience. In so doing, theyhave contributed
greatlynot only to the understanding of child abuse but also to the
development of policies toward its alleviation.
The second characteristicof this sociological research is its reliance
on empiricist,quantitativemethodologies.Straus,Gelles, and Steinmetz
won federal funding for and then conducted the firstnational survey
using a representativesample of the population, on a problem hereto-
fore considered too sensitive for this kind of research. They found,
replicatingthe experience of the women's movement,thatif people felt
safe, theywere willing,even eager, to talk about familyviolence. All the
variablesconsidered-whether stressfactors,typeof violence,or relative
power of the marital partners-were coded in quantifiablecategories.
31. Richard Celles, "Child Abuse as Psychopathology:A Sociological Critique and
Reformulation," mericalnJourna(l 43, no. 4 (July1973): 611-21, reprinted
ofOrthop,schiiatrV
in Gelles, ed. (n. 5 above), pp. 27-41.
32. In general,thereseems to be agreementthatthe incidenceof abuse and neglectis
higher among low-incornepeople, although the data is skewed because lower-income
people are reported more than are middle-class people. A few writerssuggest that
abusive parents come fromall walksof life. See, e.g., Keinpe and Kempe (n. 6 above), p.
10. But most of the cases and the data discussed in the literatureare about low-income
familieswho come to the attentionof professionals.See also Leroy H. Pelton,"Child Abuse
and Neglect: The Mythof Classlessness,"in TheSocial Context ofChildAbuseand Neglect,ed.
Leroy H. Pelton (New York: Human Sciences Press, 1981), pp. 23-38.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
502 Breinesand Gordon Review:FamilyViolence

The strengthsof this approach are considerable: it makes propositions

testableand findingscomparable. It also has significantlimitations,as is
apparent, for example, in their discussion of stress and frustration.
Many factorsare correlated with abuse and neglect,but a commitment
solely to the quantifiable leaves only that: correlations.For example,
isolation is correlated with familyviolence. Does isolation lead to vio-
lence, or violence to isolation, or are violence and isolation both
symptomsof otherproblems?This question is not considered,much less
answered.
Straus, Gelles, and Steinmetz eliminate psychologicalor cultural
factors,then,whichare not easily susceptibleto quantificationor which
can be quantifiedonly at the expense of experientialambiguity.They
eliminate,forexample, the dynamicsand contradictionsof intimatere-
lationships.As a result their theoryis abstractand does not illuminate
actual lifeas experienced by actual people. Indeed, theirthoroughgoing
empiricismappears to reflecta distrustof what can be learned from
qualitativeformsof insight.
In the thirdand fourththemescharacterizingthisapproach to child
abuse literature,the sociologistshave opened powerfulcritiquesof other
aspects of our society-the home, forexample, so oftenromanticizedas
a peaceful haven,whilein factso oftena "cradle of violence"-and of the
"legitimate"and sociallyacceptable use of violence in child rearingand
in the societyat large. Until recent feministscholarshipdemystifiedit,
the familywas accepted by most scholars as a harmonious and private
institution,perhaps in conflictwiththe larger societyas a unit,but itself
the site of love and intimacy.Straus, Gelles, and Steinmetzhave docu-
menteda verydifferentunderstandingof the family,as a place in which
violenceand hate are felt,expressed, and learned as consistentlyas love.
They suggest that the familyis, in fact,violence-proneto some degree
because it is the institutionin which intimacyis enacted, an intimacy
which can be supportiveor destructive.33They amass impressivedata
about theenormous incidenceof violencewhichtakesplace in families.34
They pointout thatthereare more murdervictimswho are membersof
the murderer'sfamilythan any other categoryof murderer-victim re-
lationship, and one is more likelyto observe, commit, or be a victim of
violence withinone's familythan in any other setting.35They especially
emphasize that thereis a license to hitin the family,and oftennowhere

33. See Gelles's "Introduction,"in Gelles, ed. (n. 5 above), for a sunmmar- of this
thesis.
34. Suzanne Steinmetzand MurrayStraus, "General Introduction:Social MyNth and
Social Systernin the Study of Intra-FamilyViolence," in Steinmetzand Straus, eds. (n. 9
above), pp. 3-25, esp. p. 3.
35. Gelles, ed. (n. 5 above), pp. 188 ff.'Ihis evidence on imurlder
coies mainlyfroin
the work of Marvin E. Wolfgang. See particularlyhis Patter-ns in CriminalHomicide(Nesw
York: John Wiley& Sons, 1958).

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 503

else, seen especially graphicallyin wifebeating,but also in child abuse.

The familyis a "trainingground" for violence,according to Gelles and
Straus,and aggressionin the familyan accepted and normalresponse to
frustrationand anger.36
Straus and Gelles do not, however,offerany analysisof the family
thatmakes the reason for so much violence comprehensible.Thus, they
go no furtherthan to point to the facts.The factsare shocking,but what
are we to make of them? Gelles, for example, lists-a listis the charac-
teristicanalyticform of this school-eleven factorsthat make it likely
thatviolencewillbreak out in the family,among themthe lengthof time
familymembers spend together,the intensityof commitmentand in-
volvementin familyrelationships,itsprivacy,the age and sex differences
of its members,ascribed roles, involuntarymembership,proneness to
stress.37But thereis no analysisor theoryof power differentials between
husbands and wives, parents and children, sistersand brothers,older
and younger siblings; or of the differingsocialization and roles of
women and men. These differentialsare, in fact,the basis of much of
familylife and thus familyviolence. Again, dynamic relationshipsare
neglected, a consequence of their unwillingnessto strayfrom the un-
quantifiable.
Straus, Gelles, and Steinmetzdo not place a familywithits power
structureinto the contextof a societywithitspower structures-gender,
generational,class, racial. They see the familyahistorically.Even the
culture of violence they decry is not analyzed in terms of its history,
relation to male supremacy,and class relations. Many of the specific
stress factors they point to-isolation, alcoholism, unemployment,
intimacy-must be understood historically,in relationto the decline of
community,to competitionand individualism,to the changes in wom-
en's status,forexamples. The failureof these sociologiststo incorporate
more feministsocial theoryand ethnographiesof the familycuts them
off fromuseful sources of insight.38
A central analytic link missing from their study of the familyis
gender. While theyconsideritsignificantin theirworkon wifebeating,it
is only brieflymentioned in their work on child abuse.39 Even at first

36. Richard Gelles and MurrayStraus,"FamilyExperience and Public Support of the

Death Penalty,"in Gil, ed. (n. 5 above), pp. 538-57.
37. GCelles, "Introduction"(n. 33 above).
38. See, e.g., Carol Stack,All Our Kin (New York: Harper & Row Publishers,1974);
Michael Young and Peter-Willmott,Familyand Kinshipin East London(Baltimore: Penguin
Books, 1957); Lillian Rubin, WorldsoJPain (New York: Basic Books, 1976). For theoretical
work, see Michelle Zimbalist Rosalclo and Louise Lamphere, eds., Woman,Culture,and
Society(Stanford,Calif.: Stanford UniversityPress, 1974); Rayna Reiter,ed., Towardan
Anthropology of Women(New York: MonthlyReview Press, 1975); Nancy Chodorow, The
Reproduction of Mothering: and theSociologyofGender(Berkeley: Universityof
Psychoanalysis
California Press, 1978).
39. Althoughtheymentiongender, e.g., in the followingplaces, thereis in factnoth-

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
504 Breinesand Gordon Review:FamilyViolence

glance, two thingsabout child abuse should make the relevance of gen-
der divisionsand sexism clear: first,that women are almost always the
primary parent; second-the same fact observed from another
perspective-that child abuse is the onlyformof familyviolencein which
women are assailants as often as men. It seems clear that something
about the gendered natureof parentingand otherhome responsibilities
can help explain abuse by mothers.
The structureof parenting,and hence of child abuse and neglect,is
gendered. Individual and unshared parentingmeans thatwomen have
fewer opportunitiesfor other work and fulfillment.Inadequate child-
care facilitiesaffectwomen mainly,and poor women particularly,an
inadequacy directlyrelated to the fact that women do more parenting
than men. At the same time,men historicallyhave more authorityand
power in families,anotherside to the sexual divisionof labor. A mother's
uniquelyfemale feelingof responsibility and of guiltforthe "success" or
"failure"of her child maylead to abuse and neglect.Even the conditions
cited by the Straus school as leading to stress,such as isolation and
poverty, are related to the sex/gender system: they more directly
characterize and impinge upon women, especially those who head
households alone.
When these writersdo discuss gender, theyfocus on the disastrous
consequences of unemploymentand povertyof the father,the tradi-
tional head of the household. They point to his potentialresortto physi-
cal forceas a substituteforother missingresourcessuch as status,power,
and money.40Male socialization, which encourages aggression as an
element of masculinity,is also identifiedas part of the source of child
abuse. (By contrast,femalesocialization,as we have pointed out above, is
not usually considered in relationto child abuse.) Indeed, although no
expert has pointed it out, it is impressivethat,althoughwomen do most
of the child care, theycommitonly about 50 percent of the abuse and
neglect. This figure means that 50 percent of those who abuse and
neglectare men, who have on average littleresponsibilityforand expo-
sure to children.A significant question flowsfromthis:whydo men who
do so littlechild care physicallyabuse children so much?
The fourththeme in thisworkis itsgeneral abhorrenceof violence.
Straus, Gelles, and Gil show the naivete of assuming that one can have
peaceful familiesin a violent society.Straus has been particularlyout-
spoken in insistingthatthe legitimation,even glorification, of violencein
the media, the popular culture,and in politicalideology contributesto
individual,personal violence. He protestsagainst violence on television,

ing slubstantialabout theirconmments: Steinmietzand Straus,"(eneral Introduction"(n. 34

above), pp. 12, 13, 20; (elles, ed. (n. 5 above), pp. 18, 33-34.
40. See William J. Goode, "Force and Violence in the Family," in Steinmetz and(
Straus, eds. (n. 9 above), pp. 25-43.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 505

but objects withequal force to governmentalviolence-the death pen-

alty, police brutality,the arms race, for example-and calls for gun
control.41However stronglyone might agree, Straus's arguments on
these points are not rigorous, nor are they supported by convincing
evidence. Indeed it is characteristicof the Straus school that once its
practitionersdepart from quantitativemethods, they seem to assume
that no other kind of argument requires proof. For example, the evi-
dence about the effectsof televisionviolenceon behavior is weak at best.
Nor have there been controlled studies, to our knowledge,comparing
familyviolence in societieswithand withoutlarge standingarmies and
police forces.
A subthemeof theirdisavowal of violence is the shared opinion of
Gelles, Straus, Steinmetz,and Gil that there is a connection, perhaps
even an unbroken continuity,between physical punishmentand child
abuse. The conflationof "force," "physical attack,""assault," or "cor-
poral punishment"with abuse is characteristicof their work. They be-
lieve these are degrees of one process. When corporal punishmentin
childrearingis sanctioned,injuryand abuse are inevitable,theyargue.
Gelles suggeststhatthe use of physicalpunishmentteaches childrenthat
it is acceptable "(1) to hit people you love, (2) forpowerfulpeople to hit
less powerfulpeople, (3) to use hittingto achieve some end or goal, and
(4) to hit as an end in itself."42Throughout their discussion the term
violence is unfortunatelyused interchangeablywith physical punish-
ment,a judgment problematicfor its failure to consider culturallydif-
ferentvalues and standards of discipline.

Valuesand Definitions
The distinctionbetweenphysicalpunishmentand abuse bringsus to
the complex topic, mentionedearlier, of how to define child abuse and
neglect.Jeanne Giovannoni and Rosina Becerra's excellentbook, Defin-
ing ChildAbuse,offersone approach.43The authors show how hard it is
to arriveat a definitionof child abuse thatcoversall the areas in whichit
is needed-personal consciousness,legal codes, social work,and policy-
making-and also the high stakes involved in such definition:a child's
life, a parent's custody of her child. Surveyingthe attitudesof several
occupational and social groups in Los Angeles (lawyers,pediatricians,
social workers,police, and lay people analyzed byclass and ethnicity),the
authors asked respondents to rank the seriousness of many different

41. MurrayStraus,"A Sociological Perspectiveon the Preventionof Wife-Beating,"in

Social Causes of Husband-Wi/fe Violence,ed. Murray Straus and Gerald Hotaling (Min-
neapolis: Universityof Minnesota Press, 1980), pp. 211-32.
42. Gelles, ed. (n. 5 above), p. 15.
43. Jeanne Giovannoni and Rosina Becerra, DefiningChild Abuse (New York: Free
Press, 1979), p. 4.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
506 Breinesand Gordon Review:FamilyViolence

"vignettes,"each of which described a specific treatmentof a child.

These included storiesof physical,sexual, and emotional mistreatment
of children,as well as mistreatmentthrough neglect. Giovannoni and
Becerra also examined the handling of'a thousand actual cases in four
Californiacounties during 1975 and 1976.
They found significantdifferencesin what formsof child treatment
various population groups consider acceptable and reprehensible.This
findingfitsan existingbody of sociologicalliteratureon the relationship
between family values and social status, which family violence re-
searchers have considered as well.44Straus and Steinmetz have sub-
stantiatedthe main conclusion of most of this literature,that parents
trainchildrenin the home forthe workand treatmenttheycan expect as
adults.45If the parentsneed verbal and interpersonalskillsin theirwork
(i.e., are middle class) and expect their children to require the same,
there will be less physical punishment; if they use motor skills,take
orders, and do not need verbal resources,there will be more.
Accordingto Giovannoniand Becerra's study,however,lower-class
familiesdo not accept mistreatment (as distinctfrompunishment)more
than do middle-classfamilies.46'I'heir data also suggest that ethnicity,
race, and culture are as importantas class; withina given educational
ancdclass grouping,ethnicdifferencesplay a powerfulrole in defining
mistreatment. For example, while blacksand Hispanics found all forms
of child abuse to be more serious than did whites,blackswere more con-
cerned by parents' failure to provide for or to supervise children; in
contrast,Hispanics ranked sexual and drug-related incidents among
the most serious abuses. Lay people in general considered matters
pertainingto drugs and stealingby the parents (moral matters)among
the most serious incidents,while professionalsrated these lower than
physicalharm to the child. These last findingsare consonant with so-
ciological work that indicates a greater tolerance of "moral deviation"
among the more highlyeducated.
The work of Giovannoni and Becerra shows that, in the study of
child abuse, the moral purposes, cultural values, and communitystan-
dards with which most people raise their children must be taken into
consideration.This considerationcannot be adequately integratedusing
the strictlyquantitativemethods of the Straus school. In arguing for

44. See, e.g., Melxin Kohn, (.lass an1dCojfo/'rmitt

((hicago: Unilversityof Chicago Press,
1977); Mirra Komarlovsky, Blue Col(lr lMarriage(New York: Vintage Books, 1967); Rubin
(n. 38 above); Herbert Cans, The Urbahn Villager-(New YIork:Fi-ee Press, 1962); (oode (n.
40 above).
45. Murrayh A. Stralus,"Some Social Antecedentsof PhysicalPunishment:A Linkage
Theory Interpretation,"in Steinmetzand Straus,eds. (n. 9 above), pp. 159-66, esp. p. 159;
Suzanne Steinmetz,"()ccupational Envir-onment in Relation to Phxsical Punishmentand
Dogmatism,"in Steinmtetand Straus, eds. (n. 9 ab(ove),pp. 166-72, esp. p. 166.
46. (ioxvannoniand Becerra (n1.43 above), p. 189.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 507

distinctionsamong differentbehaviors which may all be called "child

abuse," Giovannoni and Becerra, furthermore,indicate the problems
entailed withutilizingan even largeraggregationlike familyviolence,an
aggregation which obscures the specificsof child abuse, wife beating,
and incest.In addition,theyare rightfullyconcerned withlabelingwhich
may emerge from assumptions unique to the class, culture, time, or
situationof the outside observer,especiallyof the professionalwho is in
a positionof potentialpower over parents. Such concerns seem equally
applicable to other formsof familyviolence.

II. WifeBeating

By contrastwithscholarlyworkon child abuse, thaton wifebeating

is newer,less developed, and less clearlydivided into schools of thought.
Studies of wifebeating break down into several cross-cuttingcategories.
There is a feministschool of thought which views the problem as a
microcosmof the societal relationsbetween the sexes; in contrastnon-
feminist(but not necessarily antifeminist)analysts tend to view the
problem (often called, not wife beating, but spouse abuse or marital
violence)47as a gender-neutralor at least as a mutual problem of both
sexes. In addition there is division between those who approach the
problemprimarilypsychologically-and these mayinclude feministsand
antifeminists;those who look primarily to social stress factors for
explanations-and these too include people with a range of attitudes
toward feminism;and those who rely on systemstheory approaches.
Most nonfeministsociologistsemphasize similaritiesbetweenchild abuse
and wifebeatingboth in theirpatternsand causes and in theirremedies.
Indeed, these scholars argue that treatingthe problems separatelymay
obscure theircommonalities.48
Whatever the logical, analytical meritsof this argument, it bears
repeating that concerns about child abuse and wife beating have dif-
ferenthistories.Child abuse has existed as a recognized social problem
for over a century.Concern with the problem grew from agitationby
professionalgroups motivatedby self-promotionas well as concern for
abused children: charityand social workers,doctors, and lawyers.By
contrast,the firstlarge-scalecampaign againstwifebeating took place in
the 1970s, part of the renaissance of feminism.(Nineteenth-century

47. Straus, Gelles, and Steinmetz(n. 9 above); Gelles, The ViolentHome (n. 28 above);
Steinmetzand Straus, eds. (n. 9 above), are examples.
48. All the books in the previous note take this perspective. In addition, several
articlesargue explicitlyforthe commonalityof typesof familyviolence,e.g., Ursula Dibble
and Murray A. Straus, "Some Social Structure Determinantsof Inconsistencybetween
Attitudesand Behavior: The Case of FamilyViolence,"JournalofMarriageand theFamily
42, no. 1 (February 1980): 71-80.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
508 Breinesand Gordon Review:FamilyViolence

feminismproduced similarcampaigns, though much smaller,in a few

European countries and a few U.S. cities.)49This differencefrom the
historyof child abuse resultsin part, of course, fromthe factthatchil-
dren are less able to build a social movement. Professionalsbegan to
contributeto the anti-wife-beating movementveryquicklyin the 1970s,
as providersof servicesand as researchers.These professionalsare not
"neutral": theytend to abhor wifebeatingjust as theyabhor child abuse.
Nevertheless,the greaterpoliticizationof the wife-beatingissue, or more
preciselyitsdirectconnectionwitha powerfulsocial movement,puts the
professionalsin a differentsituation. No matterwhat their claims to
objectivity,nor how small their topic, they are, in fact,responding in
many aspects of theirwork to the challenges of feminism,and the new
concepts and social analyses it has thrownup in the last decade. How-
ever, the narrow,intrafamily, gender-neutralperspectiveso often em-
ployed by family violence scholars obscures the actual historyof the
feminist"discovery"of violence againstwomen and discourages analysis
of the problem froma contextof societal male supremacy.

BatteredWomen
Approaches
Sociology
A burstof sociological scholarshipon wifebeating appeared in the
1970s. As with that on child abuse a decade before, it began with a
critique of the previously dominant explanations based on psycho-
pathology5? (explanations which often ended in identifyingthe prob-
lems of the victimas the cause of her battering).51Thus sociologists
began withthe assertionthat neitherwife beaters nor their victimsare
necessarilycrazier than nonviolentadults.52This critique permittedre-
searchers to move toward studyingthe familyas an institution,seeing
social and economic stressas causative factorsof violence, and making
policyrecommendationsthatcould, if implemented,be genuinelyhelp-
ful to batteredwomen.
However, the argumentthatwifebeatingcould be "normal,"thatis,

49. See Elizabeth Pleck's essay, "Feminist Responses to 'Crimes against Women,'
1868-1896," in thisissue of Signsand in her forthcomingbook. We are excluding historical
work on familyviolence fromthis articlebecause there is so littleof it.
50. For a good bibliographyof these psychologicalexplanations,see Mildred Daley
Pagelow, Woman-battering: Victims (BeverlyHills, Calif.: Sage Publica-
arndTheirExperiences
tions, 1981), pp. 19-20. To save space, we limitedourselves in this articleto sociological
work on wifebeatingand have had to eliminatesome useful psychiatricand psychological
work, such as that by Elaine Hilberman and Kit Munson, "SixtyBattered Women," Vic-
timology2, nos. 3-4 (1977): 460-70.
51. E.g.,J. E. Snell, R.J. Rosenwald,and A. Robey,"The Wifebeater'sWife,"Archives
of GeneralPsychiatry 11 (August 1964): 107-13.
52. E.g., MurrayStraus,"A Sociological Perspectiveon the Preventionand Treatment
of Wifebeating,"in BatteredIt'omen,ed. Maria Roy (New York: Van Nostrand Reinhold
Co., 1977), pp. 194-239, esp. pp. 194-95.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 509

not psychopathological,shades into a rejectionof any psychologicalin-

quiry, even into normal components of human male psychology.Fur-
thermore,although sociologistsof the Straus school claim merelyto set
aside psychologicaldimensions in order to explore social ones, in fact
they often operate on the basis of an alternativeset of psychological
assumptions-behaviorist ones. Thus Gelles wrote in one of his most
recent works,an unpublished conference paper of July 1981, "Human
interactionis guided by the pursuit of rewards and the avoidance of
punishmentand costs."53Such assumptionsprohibitthe recognitionof
contradictoryor unconscious motivationswithinindividuals.Gelles and
othersof thisschool relyon a "social learning"model forexplainingwhy
some people are more violentthan others,avoiding the insightsof ego
psychology.54
Behaviorist psychological assumptions, part of an overall anti-
psychologicalstance,are relatedto the emphasis on quantificationin this
work.The leading exceptionsare venturesby Straus into systemstheory
withits attendantflowcharts.55However, the systemstheoryapproach
shares with the quantitativework a mechanisticquality,reducing each
factorto a variable with unique and nonambiguous meaning and then
representingthe linear interactionof these variables in graphic form.56
Even the more overtlyfeministwork to have emerged fromthis school
of familyviolence sociologyis empiricist.For example, the method used
for recent examinations of the impact of the women's movement on
aspects of familyviolence involved exclusivelyquantitativeindices to
the strengthof feminism,such as NOW membershipsize, or the states'
passage of ERA laws.57
The quantitative bent suits, also, the "antiideological" stance of
much contemporarysociology (although we consider the empiricist
method itselfan ideology). For example, in a recenttheoreticaloverview
Gelles uses the term "ideology" with obscure but clearly pejorative
53. Richard Gelles, "An Exchange/SocialControl Theory of IntrafainilyViolence"
(paper presentedat FirstNational ConferenceforFamilyViolence Research,Universityof
New Hampshire, Durham, July 1981).
54. E.g., see MurrayA. Straus and Gerald T. Hotaling,"Introduction,"in Straus and
Hotaling, eds. (n. 41 above), pp. 4-5; Murray Straus, "A Sociological Perspectiveon the
Causes of Family Violence," in Green, ed. (n. 28 above), pp. 7-31; and Pagelow (n. 50
above).
55. MurrayStraus, "A General SystemsApproach to a Theory of Violence between
Family Members,"Social ScienceInformation 12, no. 3 (June 1973): 105-25.
56. Steinmetzand Straus, "General Introduction"(n. 34 above), pp. 17-18.
57. Debra S. Kalmuss and MurrayA. Straus, "Feminist,Political,and Economic De-
terminantsof Wife Abuse Services in American States" (Family Violence Research Pro-
gram, Universityof New Hampshire, Durham, March 1981, mimeographed); KerstiYllo
and MurrayA. Straus,"Patriarchyand Violence against Wives: The Impact of Structural
and Normative Factors" (Family Violence Research Program, Universityof New Hamp-
shire, Durham, July 1981, mimeographed).

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
510 Breinesand Gordon Review:FamilyViolence

meaning. Under the heading "Ideology: Sexism and Racism," Gelles

writes:

Some have argued that these data [officialstatisticson child abuse

and wife abuse] support the notion that the real cause of family
violenceis not psychologicaldisordersor social learning;rather,itis
oppressive sexism, racism,and a patriarchalsocial organizationof
capitalisticsocieties.While it is easy for liberal-mindedsocial scien-
tiststo sympathizewiththese conceptualizations,thejump fromthe
relationshipbetween income and violence to a theoryof racismor
sexismis large and not yetfullysupported bythe available empirical
evidence. The use of ideology in place of scientifically informed
theoryhas become increasinglycommon in the emotionallycharged
field of domestic violence and has partiallyinhibiteda serious sci-
entificprogram of theoryconstructionin thisarea.58

This statementmeritsclose attention.In his own constructionof

theory,Gelles is objectingto certainother methodsof theoryconstruc-
tion whichhe finds"ideological" and hence nonscientific.Politically,it is
clear thatfeministsand other radicals are the targets.Methodologically,
it would seem-from the context-that nonquantitativelyverifiablear-
gumentsare the targets.Yet itmaywellbe thattheoperational definition
of ideology here is nonquantitative,or nonempiricist.
This particularpassage is exceptional because in it the implicitanti-
ideological bias pervasivein much sociologyis so explicitlyarticulated.At
the same time Straus and his collaboratorshave on the whole been sup-
portiveof feminism.Issue by issue, mostwifeabuse scholarsagree with
feministdemands-on the ERA, on the pernicious effectsof sexistim-
agery in the mass media, on the error of blaming the victimsof wife
beating,on abortion rights.The empiricistsociologists'differenceswith
many feministsbegin with their reluctance to make a syntheticinter-
pretationof the role of male supremacy in establishingconditions for
violence-an interpretationthat,we mustassume, Gelles would dismiss
as ideological.

Is All ViolencetheSame?
We have seen above how Straus,Gelles, Steinmetz,and manyothers
have patientlyreiteratedthe foolishnessand danger of scholarlyadher-
ence to the mythof the familyas a "haven."59In insistingon thisrealism
58. (elles, "Exchange/SocialControl Theory" (n. 53 above). See also Richard (elles,
by R. Emerson and Russell
review of ViolenceagalinstWilves:A Case againstthePatriarchy
Dobash, Society(September-October 1980), pp. 87-88.
59. Unfortunatelythere is now a backlash against this deromaIlticizationof the
family-not onlv fr-omthe New Right but even among liberals and radicals. Leading

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 511

about the family,sociologistscontributedto exposing the pervasiveness

and relativelegitimacyof maritalviolence.
MurrayStraus and manyothers have also, as we have seen, argued
repeatedly that all violence should be viewed as of similar origin. In
criticizinga view of familyviolence as individual deviance, they argue
that it is normal in a violentsociety.60As Colleen McGrath argued, this
approach makes violence "an abstract,transhistoricalsocial fact.... In-
stead of sick people, we get a sick society."61The violence of both child
abuse and wifebeating is thus conceptualized as a "breakdown in social
order"62 (with no distinctionsmade between its differentmanifesta-
tions), rather than as the reflectionof a power strugglefor the mainte-
nance of a certain kind of social order. An example is Straus's Conflict
Tactics Scale (CTS), whichcategorizesviolentacts on a continuumfrom
least to most severe, treatsmale and female acts equally, and makes no
allowance for the power context withinwhich violence occurs.63The
CTS assumes thatall violentacts are comparable and can be ranked; that
violence can be ordered linearly;and, implicitly,that any pushing, hit-
ting,or throwingis worse than any amount of verbal or emotional ex-
pression, no matter what pain the latter inflicts.In fact, the context
necessaryfor evaluation of violent actions would, of course, make the
data too complex for quantification.
In the contemporarypoliticsof the campaign against wifebeating,
Straus's tendencytowardabsolutismagainst violence has had both posi-
tive and negative results. An example of what we find positive-one
among many-is a 1979 article in which Straus reprintsa section of a
dialogue between himselfand U.S. RepresentativeScheuer at hearings
of the House Committee on Science and Technology. In it Scheuer
raises the objection that"violence" includes everythingfrommayhemto
a push or a shove. Straus acknowledgesthisdefinitionand defends it: he
argues that in the House of Representativesnot even a mild push or

examples fromliberalsand radicals include ChristopherLasch, Haven in a HeartlessWorld:

The FamilyBesieged(New York: Basic Books, 1977); Jean Bethke Elshtain, "Feminists
against the Family,"TheNation(November 17, 1979), pp. 497-500; National Organization
foran American Revolution,Our FamiliesAreUp to Us!. Right-wingromanticizationof the
familyshould be well known to readers today, but for a coherent summary,see Allen
Hunter,"In the Wings: Ideology and Organizationof the New Right,"RadicalAmerica,vol.
15, nos. 1 and 2 (January-April1981).
60. E.g., Steinmetz and Straus, eds. (n. 9 above), chaps. 8-9; Straus, Gelles, and
Steinmetz(n. 9 above), pp. 101 ff.
61. Colleen McCGrath,"The Crisis of the Domestic Order," Socialist Review 43
(January-February1979): 11-30, esp. 17.
62. R. Emerson and Russell Dobash, quoted in Dorie Klein, "Can This Marriage Be
Saved? Batteringand Sheltering,"Crimeand SocialJustice12, no. 12 (Winter 1979): 19-33,
esp. 23.
63. "The ConflictTactics Scale" is in theJournalofMarriageand theFamily41 (Febru-
ary 1979): 75-88.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
512 Breinesand Gordon Review:FamilyViolence

shove would be tolerated,while in the familypeople accept more vio-

lence. This double standard is wrong, Straus argues; people have the
same rightsto immunityfromviolence everywhere.64 Insistingthuson a
single standard of civilrights has been vitalto the campaign againstwife
beating.Anycompromise in thisabsolutistpositionwould surelyweaken
the feministdemand that marriage should not call upon women to
sacrificeany of theircivil rights.
Such an absolute condemnationof violence,however,willmean that
the victimmustalso be nonviolentto appear worthyof support. With a
perspectivethatthustreatsviolenceas itselfthe problem,ratherthan the
symptomof a problem,a batteredwoman appears to deserve less sym-
pathy if she too used physicalforce,or perhaps even initiatedthe vio-
lence.
This absolutism about violence has also affectedhow research on
wife beating is done. Questions turn upon specificacts, blows, pushes,
weapons, ratherthan upon the gestaltof the conflict.The ConflictTac-
tics Scale does not measure injuries,for example; nor does it deal with
the bias possiblyresultingfromthe factthat"violent"behaviorsbya man
may be ignored as nothingbeyond the ordinary,while similarbehavior
by a woman may be so out of characterthat it is well remembered.
This weakness is well illustrated by the "findings" reported by
Steinmetzin a 1977 article,"The Battered Husband Syndrome." In it
she reportedthat"the percentageof wiveshavingused physicalviolence
often exceeds that of the husbands."65 Many scholars and anti-wife-
beating activistscriticizedSteinmetz'sview and emphasized the asym-
metryof maritalviolence.66But it has been unfortunatethat most re-
sponses to Steinmetzwere limitedto denyingthe findings,challenging
the reportingof statistics, and demanding additional contextualfactors.
They have not questioned the termsof the inquiry.This kind of limited
64. Straus,"Sociological Perspective"(n. 54 above).
65. Suzanne Steinmetz,"I he Battered Husband Syndrome,"Victimolog, 2, nos. 3-4
(1977-78): 499-509.
66. Elizabeth Pleck et al., "'The Battered Data Syndromne:A Reply to Steinmetz,"
Victinology 2, nos. 3-4 (1977-78): 680-83; Richard Gelles, "The Truth about Husband
Abuse," Ms. 7 (October 1979): 65-66, reprintedin hisFamilyViolence(n. 5 above), chap. 8;
M.J. Fields and R. M. Kirchner,"Battered Women Are Stillin Need," Victimology 3, nios.1
and 2 (1978): 216-22, among others. Criticismsincluded that the source of Steinmetz's
data didinot measure outcomes of violence; thatwomen more oftenthan men strikeout in
self-defense;that women are legallydisadvantaged; that women are oftenattackedwhen
pregnant; thatwomen are miorevulnerablebecause theyare generallyless power-ful than
men; thatSteinmetzmisrepresentedthedata she was herselfusing,reportingthattheywere
more significantthan they were; that Steinmetzdismissed and/orignored contradictory
data. But in the recentpopular book byStraus,Gelles,and Steinmetz,BehindClosedDoors(n.
9 above), in a chapterattributedto Straus,Steinmetz'sfindingson husband beatingare still
repo-ted in an uncriticaland misleading way. In our opinion, the mere reportthat hus-
bands are beaten more than wiivesis probably false and surely does not co-ntributeto
general understandingof maritalXviolence.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 513

critiqueleads to a trap,forit impliesthatwomen ought to be and usually

are nonviolent,and that this virtue is what entitlesthem to our sym-
pathy.
It is admittedlydifficultto examine the fullcontextof any violence
between intimatesbecause that veryintimacycreates privacyand com-
plexity. However, the problem becomes virtuallyinsuperable with a
methodologythatdoes not encompass the ambiguityand complexityof
heterosexual intimacy,informedas thatis by gendered and hierarchical
social relations. Once discrete acts of violence, removed from overall
power relations,become the subject of study,the "data" no longer de-
scribe reality.
Steinmetz's misleading findings on husband beating are by no
means characteristicof sociology in this field. On the contrary,sexism
and issues of gender are a major theme of studies on maritalviolence.
Most workby Gelles and Strausincludes women's social powerlessnessas
a factorin batteringrelationships.For example, in a 1976 article,Gelles,
in keeping with his and Straus's critique of the victim blaming in
psychologicallyoriented literature,found that the major contributing
factor in keeping women in violent marriages was the failure of the
"helping" agencies to offerany practicablehelp when asked.67However,
the dynamicsof male supremacyare only analyzed quantitatively,and
concepts like sexism remain discrete items in lists of variables, often
added on, as in the category"other."

WhoIs "at Risk"for WifeBeating?

Another importanttheme in sociological scholarshipon wife beat-
is
ing the role of stress. Generally,the term is used to define specific
difficultiesexternal to the familysystem,68but those using it do not
ordinarilyconsider marriage itself,or heterosexual relations,automati-
cally stressful,as feministanalysismight.Yet the sociologists'own work
occasionallysuggeststhe difficulty in distinguishingbetween stressin a
relationship and the relationship itself.69The stress research is
nonethelessimportantbecause it representsa method of searching for
mediations between the individual personalityand familytensionsand
societal problems, and because it demonstratesthat wife beating is a
social problem,notjust an individualone. Furthermore,stressresearch
tends to produce progressivepolicy implications,underliningthe need
for good social services.

67. Richard Gelles, "Abused Wives: WhlyDo They Stay,"JournalofMarriageand the

Family38, no. 3 (November 1976): 659-68.
68. E.g., Keith M. Farrington,"Stressand FamilyViolence," in Straus and Hotaling,
eds. (n. 41 above), pp. 94-114.
69. Straus has generallybeen more complex on thispoint,as he is the sociologistwho
has consistentlybeen most sensitiveto the impact of sexism on violence.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
514 Breinesand Gordon Reviezv:FamilyViolence

However, as Straus has shown, the fact that "stress" may be

hypothesized as a mediating factor between psychological and social
levels does not explain why similarstressaffectspeople differently. At
the other extreme,when stressis approached theoretically,as part of a
systems-theory model and withoutthe evaluation of concrete experi-
ence, the results verge on the tautological. For example, as Keith Far-
ringtonwrites,"(1) the familyencountersa high amount of stress,(2) it
tends to be poorlyequipped to handle stress,and (3) thereis thusa great
potentialfor frustrationwithinthe family."70
Another argument about conditions conducive to familyviolence
uses "resourcetheory,"whichcan be tracedto WilliamGoode's articleon
violence in the familypublished in 1971.71 In a marketmetaphor,the
use of violence is assumed to be a more "costly"resource which will be
avoided unless "cheaper" resourcesof power, such as prestigeor wealth,
are absentor depleted. Such a theory,iftrue,would explain whypoor or
low-statusmen are more oftenwifebeatersthan are richand high-status
men. But attemptsto testthishypothesishave produced mixed results.72
The inconclusivenessof these findingsshould remind us thatresources
are hard to define except in a manner that ignores (lifferencesin per-
sonality,culture,degree of familyconflict,and manyother factors.Nor
is the assumption defensible that violence is only a last resort,never a
firstchoice.
The "resources" concept stimulatesa tendency to view personal
characteristicsand aspects of relationshipsas commodities,as alienable
and exchangeable possessions-a highlyideological and, we think,mis-
leading view of intimaterelationships.Resource theoryalso legitimizes
the comparisonof dissimilarfactors,weighingincome againstemotional
dependency, for example.
Resource theory sometimes relies on a distinctionbetween in-
strumental,that is, purposefulviolence and expressive,nonpurposeful,
violence;73 with such a distinction resources can explain only in-
strumentalviolence. This distinctionis helpfulin showingthat violence
can be purposive,as a means of enforcingone's will. But the distinction
does not illuminate actual acts of violence, which often contain both

70. Farrington(n. 68 above), pp. 110, 113.

71. WilliamJ. (Goode,"Force and Violence in the Family,"Journalq1 Marrizageand the
Famnil 33, no. 4 (November 1971): 624-36. Resource theory is in f:actan instance of
"exchange" theory,in turn an application of the kind of rationalisticpsychologyalready
discussed. Craig M. Allen and MurrayA. Straus, "Resourc-es,Power and Husband-Wife
Violence," in Straus and Hotaling, eds. (n. 41 above), pp. 188-208, esp. p. 205n.
72. Allen and Straus predictthatmen withInore resources will be less violent(n. 71
ab()oe); Rodney Starkand James MIcE(ov III findthe reverse(see "Middle (lass Violence,"
Psychology Today4 [Novelmber1970]: 52-54).
73. Richard (elles and lMur-ray Straus, "Determinantsof Violence in the Family:
Toward a 'I'heoretical Integration,"in (Cottemporary
TheoriesabouttheFamily,e(l. Weslev R.
Burr et al. (New York: Free Press, 1978).

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 515

instrumentaland expressive-that is, nonpurposeful-meanings. Fur-

thermore,wife beating may be expressive in the individual case but
instrumentalin the collective. As Jan Lambertz has shown in an un-
published essay on wife beating in nineteenth-century industrial En-
gland, individual husbands in drunken nastymoods beat theirwiveswith
no apparent purpose; but the collective effectwas to keep a class of
women diffidentand submissivetoward their husbands.74Contempo-
rary feministshave offered similarinterpretationsof wife beating and
rape: while the individualattacksmay appear irrational,taken together,
they are an important ingredient in the continued subordination of
women; even to women not directlyvictimized,these attacksteach les-
sons.75
The sociologistsin thisfieldhave generallynot looked at the collec-
tiveculturalmeaning and communitycontrolof maritalviolence. Some-
where among the individual personality,the marital system,the stress
factors,and the resourcesof familymembers,we mustalso consider the
role of social "networks"-friendsand relatives-in prescribingand reg-
ulating the use of violence. Such networksmay serve to help control
women and disciplinechildrenby licensingphysicalattackson them,just
as the networkscan set limitson the violence that is sociallyacceptable.
Indeed, we suspect thata woman generallydefinesherselfas abused on
the basis of the standards of her communityand friends;self-esteemis
sociallyconstructed.The means by whichthese standardsare communi-
cated and accepted needs study.Categories like "instrumental"and "ex-
pressive" must necessarilybecome fuzzy in social analysis like this, as
theywould, too, in a psychologicalanalysisthat included the possibility
of the irrationaland subconscious meanings and motivations.
Another line of thought,related to resource and exchange theory,
considersstatusdifferencesbetweenmaritalpartners.A recentpaper on
this topic concludes that certain types of status inconsistencyand in-
compatibilitybetween spouses produce high risksof spouse abuse, "par-
ticularly life-threateningviolence."76 Feminists have criticized these
findings,charging that they are based on assumptionsthat favor male
74. Lambertz (n. 1 above).
75. BritishanthropologistAnn Whitehead, looking at male cultural attitudesabout
maritalconflicts,described how in a Britishpub men pressureda more egalitarian-minded
friendto maintainhis personal dominance over his wifein theircollectiveinterest("Sexual
Antagonism in Herefordshire,"in Dependenceand Exploitation in Workand Marriage,ed.
Diana ,eonard Barker and Sheila Allen [New York: Longman, Inc., 1976], pp. 169-203).
76. Carlton A. Hornung, B. Claire McCullough, and Taichi Sugimoto, "Status Re-
lationshipsin Marriage: Risk Factors in Spouse Abuse,"JournalofMarriageand theFamily
43, no. 3 (August 1981): 675-92. See also Straus, "General SystemsApproach" (n. 55
above); MurrayStraus,"Sexual Inequalityand WifeBeating,"in Straus and Hotaling,eds.
(n. 41 above); and JohnE. O'Brien, "Violence in Divorce Prone Families,"JournalofMarriage
and theFamily33, no. 4 (November 1971): 692-98.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
516 Breinesand Gordon Review:FamilyViolence

dominance.77Certainlythe choice of termslike "inconsistency"and "in-

compatibility"suggests an unexamined assumption that the effectson
marriageof unconventionalcircumstances(e.g., a wifewithhigherstatus
than her husband) are deleterious. Such findings,however, are not
necessarilyantifeminist.Some writerson thisissue have speculated that
the transitionto a more sexuallyegalitariansocietymay be stressful,for
men and women, and that men may respond to loss of dominance and
privilegewithviolence.78
Sociologistshave also examined the intergenerationaltransmittalof
a tendencytoward familyviolence, the "cycleof abuse" that we earlier
discussed. The opinion that both victimsand perpetratorsof family
violence were often exposed to similar violence in their childhoods is
common among cliniciansand researchers.79Research which seeks to
testthis transmittaloften falls prey to the "clinicalfallacy"-like gener-
alizations about criminalsbased on studies of convictedcriminals-and
often lacks controls.We do not have studies about the effectsof child-
hood violence on nonviolentadults,or on adults who are violentin their
familiesbut who have not been "caught" and labeled. Yet another criti-
cismis lackof controlforclass and othervariableswhichmaywellcoexist
withpreviousexperience of violence. Previouslyabused subjectsin these
studies are very likelyto be poor, and to be classifiedas from "multi-
problem"families,and any of these other problemsmayin factcorrelate
withcontinuingviolence more than withpreviousviolentexperience.8?

Sociologyand Social Policy

Widespread and sometimesintensepublic concern withchild abuse
and wife beating has drawn scholars into relations with social policy
activists.The pressure on researchers to demonstrate the immediate
relevance and usefulness of their findingshas tended to reward the
concrete rather than the abstract.At the same time, the pressure for
77. E.g., R. Emerson Dobash and Russell Dobash, ViolenceagainstWives:A Case against
thePatriarchy(New York: Free Press, 1979), p. 23.
78. Robert N. Whitehurst,"Violence in Husband-WifeInteraction,"in Steinmetzand
Straus, eds. (n. 9 above), pp. 75-82; Straus, "Sexual Inequality and Wife Beating" (n. 76
above); Mc(;rath (n. 61 above), pp. 11-30. For a critiquein agreementwiththis,see Klein
(n. 62 above), p. 30, n. 1.
79. E.g., Gelles, The ViolentHome (n. 28 above), pp. 169-70; Straus,Gelles, and Stein-
metz (n. 9 above), pp. 99 ff.;LarrySilverset al., "Does Violence Breed Violence? Contrib-
utions froma Studyof the Child Abuse Syndrome,"American JournalofPsychiatry 126, no. 3
(September 1969): 404-7.
80. These criticismsare cogentlypresented in Dorothy Miller and George Challas,
"Abused Children as Adult Parents: A 'Iwenty-fiveYear Longitudinal Study" (Family
Violence Research l'rogram,Universityof New Hampshire, Durham, July 1981, mimeo-
graphed). Also see Linda Jayaratne,"Child Abusers as Parentsand Children: A Review,"
Social Work,vol. 22, no. 1 (1977).

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 517

immediaterelevance may have been an influencetoward narrowingthe

focus of many studies at the expense of breadth (or depth), and to the
detrimentof theorizing.In addition, most of the fundingcomes from
agencies like the National Institutesof Health, traditionallydominated
by physiciansand scientists,which preferquantificationand the avoid-
ance of "ideology."
The combinationof narrow focus,statisticalwork,and empiricism
has produced a body of sociological research with the strengthsof
specificityand precision. In regard to wife beating, it has effectively
destroyedsome myths,particularlythose blamingsuch violence on nag-
ging wives and pathological husbands. Examined as a whole, however,
thisbody of research has given us hundreds of separate puzzle pieces.81
What is needed is a big picture, an interpretationthat is historical,
encompassingthe significantchanges in culturaland social normsabout
marriage,women's rights,and violencethathave occurred in the lasttwo
centuries; one that is deep and complex, recognizing many levels of
human experience-from the psychologicalto the social to the legal-all
of which may be operative within a single situation; one that in-
corporates the actors'own understandingof theiractions; one thatrec-
ognizes violence,withoutundue moralism,as partof a continuumof ways
in which humans relate. However, as we will see in turningto feminist
studiesof wifebeating,a recognitionof sexismin itselfprovidesno easy,
immediate synthesis.

FeministScholarship
The concern withwoman batteringhas so farproduced threemajor
scholarlybooks witha feministperspectivebased on original research:
Lenore Walker'sThe BatteredWoman,offeringa psychologicaltheoryof
the victim'spredicament;ViolenceagainstWives,by R. Emerson and Rus-
sell Dobash, sociologistsworking in Scotland; and Mildred Pagelow's
Woman-battering,a studyin the Straus-Gellesmode. In addition,one long
articleis importantenough to be discussed here: "Medicine and Patriar-
chal Violence: The Social Constructionof a 'Private' Event," by Evan
Stark,Anne Flitcraft,and William Frazier.82
In Lenore Walker'sanalysiswe encountera problem found in most
scholarship based on interviews: women are usually the only in-

81. Richard (elles discusses somieof these shortcomingsin his useful methodological
critique,"Etiologyof Violence: )vercomningFallacious Reasoning in UnderstandingFamn-
ily Violence and Child Abuse," in Gelles, ed. (n. 5 above).
82. While we do not review it here, we call attentionto the useful anthology,The
Victimizationof Women,ed. Jane Roberts Chapman and Margaret Gates (Beverly Hills,
Calif.: Sage Publications,1978), whichincludes articlesby Del Martinand Lenore Walker
on battered women as well as articleson rape, the sexual abuse of children,and sexual
harassnent.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
518 Breinesand Gordon Reviezw:
FamilyViolence

forrmants.83 Walker herself, a feminist,has been accused by other

feministsof victimblaming because she has identifieda particularpat-
tern among batteredwomen. However, we have a differentinterpreta-
tion of Walker; as she puts it, "I label [the battered woman] a victim
because society ... has socialized her into believing she had no choice
but to be such a victim.Whysome women rejectthisculturalnormwhile
othersfall preyto it is partiallyexplained by the psychologicaltheoriesI
put forthhere."84
Walker's main psychologicaltheoryis "learned helplessness."It is a
behavioristtheory based on animal experimentation:dogs who have
been subjected to random electricshocks were compared with control
dogs in theirabilityto learn how to avoid an electricshock. The control
dogs learned quickly;the conditioneddogs were slow to learn. The latter
group had learned helplessness.Walkerdraws an analogybetweenthese
conditioneddogs and batteredwomen, suggestingthatthe women have
learned throughexperience that theycannot stop the beatings,cannot
controltheirfate. Walker's theoryhas two weaknesses: first,similarre-
search has not been done on humans,and second, the theoryshares the
general problems of behavioristpsychology,in that it excludes depth-
psychologicalfactorsand relies on an oversimplifiedmodel of human
learning and personalityformation.
Still, Walker's learned helplessness,if understood metaphorically,
does illuminatesome aspects of women's victimization.Most battered
women have in facttriedto avoid beatingsand have "learned" thatthey
cannot do so; like the dogs, theyare in factin "cages" constructedof law,
poverty,dependent children,lack of child care, and so on. Furthermore,
many battered women also sufferfrom a deformationof self-esteem,
another kind of learned helplessness.Walkeremploysher theorywitha
sense of complexityand contradiction,subtlyreinterpretingactions by
wolnen that have previouslybeen seen as exclusivelyself-destructive.
She shows that mostbatteredwomen tryhard to controltheirsituations
withinthe limitsof what seems possible; forexample, manywomen who
appear to be "asking for" a beating are in facttryingto get "over with"
somethingtheysee as inevitable,in order to be relieved of anxiety.
Anger against victimblaming is a theme that runs through all
feministworkon wifebeating.The workof Stark,Flitcraft, and Frazier,
and Dobash and Dobash, to whichwe turn next,is particularlyeffective
in itsproof of women's double victimization:firstby batterers,and then

83. As Del Martinhas pointed out, researchintended to discoverwhymen batterhas

often ended by producing analyses of the victims,since the battererwill rarelyacknowl-
talkabout it,a situationthathas contributedto the
edge his violenceand exen more rar-ely
extensivevictimblamingtraditionalin thisfield(BatteredWives[San Francisco: New Glide,
1976]).
84. Lenore E. Walker, The BatteredWoman(New York: Harper & RowxPublishers,
1979), p. 14.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 519

by social controlagents who assign women the blame. It may be thatthe

victim-blamingprocess creates worse damage than the battering.How-
ever, anxietyabout victimblaming should not be allowed to hold back
serious feministscholarship and theoryabout woman battering.Some
men beat women, and some do not; some women are beaten, and some
are not. It is neither useful nor credible to argue that the differences
withineach group are purelyrandom. To argue thatsexismis the cause
may be helpful in addressing polemics against the victimblamers but
does not advance theoryor strategy.In factthe line between investigat-
ing the circumstancesand responses of victimsand blaming victimsfor
them is hard to walk. If, however,we avoid the burden of keeping this
balance, we abrogate an importantfeministresponsibilityfor under-
standingand changing the world.
The Stark,Flitcraft,and Frazier articleis about blamingthe victim,
but advances an even more radical and provocativenotion,"the social
constructionof battering."85Based on a study of how the emergency
room of a cityhospitaldealt withwomen treatedforinjuries,the article
demonstrateswaysin whichhospitalworkerseffectively ignore battering
and victims' requests for help, treating symptoms(such as injuries,
alcohol abuse, headaches), while showing marked uninterest in
causes-a practice less likely, say, if the symptoms resulted from
pneumonia.
This articleoffers,in our view,one of the fewtheoreticallysophisti-
cated analyses of woman battering.It focuses on how the "helping pro-
fessions,"in this case especially those of medicine, coerce women who
are appealing for help back into the situationsand relationshipsthat
batterthem. It shows a systemtakingwomen who were hit,and turning
them into batteredwomen.
When the authorsexamine the dynamicsof male supremacywithin
individual relationships,however,theiranalysisis simpler,less convinc-
ing, and tends toward reductionism.They argue thatthe main impetus
for batteringis the male's loss of control over women's labor power,
partlydue to women being drawn into the wage labor force.Through-
out they assume, as do many feminists,that wife batteringhas been
increasing-an unproven assumption.
The best scholarlybook on wife beating is ViolenceagainstWivesby
the sociologistsDobash and Dobash.86 They too emphasize the "social
construction"of batteringand offer a fine critique of how even well-
meaning professionalsmay implyto a beaten woman that she must ac-
cept thistreatment.The Dobashes' politicsare different,however.They
focus more on individual men's use of violence to maintaintheirpower
85. Evan Stark,Anne Flitcraft,and Williamn Firazier,"Medicine and PatriarchalVio-
lence: TIhe Social (onstruction of a 'Private'Exent,"International
Journal(ofHealth Services9,
no. 3 (1979): 461-93.
86. Dobash and( Dobash (n. 77 above).

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
520 Breinesand Gordon Review: Family Violence

and privilege than on economic motivation. One might crudely

categorize the Dobash view as radical feminist,and the Stark-Flitcraft-
Frazier view as socialistfeminist.
Both the book and the article make historical mistakes. The
Dobashes use legal and prescriptivedocuments-that is, evidence about
how people were supposed to act-as if theytell us how people did act.
The resultantstoryimplies a historyof unrelieved batteringof women
and obscures women's own strategiesin response. Both book and article
fail to explain why some women were able to avoid battering,others to
resistit personally,others to resistit throughnetworksof support-not
to mention why some men were able to avoid becoming batterers.
Dobash and Dobash assume that one cannot make such distinctions
withoutblamingthe victims;Starkand his colleagues even argue against
the possibilityof making such distinctions.87
But the meat of the Dobash book is much richerthan itstheoretical
or historicalintroduction.Withoutsensationalism,but withappropriate
anger, the Dobashes offeran interpretationof wife beating from the
point of view of its victims.Their analysis of the contributionof the
"helping professions,"the police and judiciary, to the constructionof
batteredwomen is particularlymovingbecause itis detailed, not polemi-
cal. Here theirline of thoughtcloselyresemblesthatof Stark,Flitcraft,
and Frazier; both are founded on an understandingof male domiination
as a systematicset of structures,not merelyon a recognitionof sexist
attitudesor bad behavior among individual men. But unlike Stark and
his colleagues, the Dobashes do not shrinkback fromthe personal,intra-
familyaspects of violence against women, and the responsibility of men
is not allowed to be dissipatedbyanalysesof overallsocietaldomination.
Pagelow's Woman-battering is a careful and useful compendium of
information, but it is limitedby its method and its vision.88Feministin
that it focuses on women and includes discussionsof shelters,it follows
an empiricistmethodand asks onlythequestionsthatfitthatmethod.Its
psychologicalassumptions are behaviorist,and its causal explanations
correlate listed variables. However, it does conciselyand convincingly
criticizecommon mythsand stereotypesabout battering.
Althoughit is beyond the scope of thisessay,the best as well as one
of the earliest introductionsto wife beating is stillDel Martin'sBattered
Wives.89Her command of the relevantresearch and theoriesis impres-
sive,and her book coversthe legal and social-servicesystemsand survival
tactics,as well as politicalissues.
Scholarship on wife beating, or marital violence, is new and un-
developed. The fewwvorks we have labeled "feminist"stand out because
an(dFriazicr
87. Ibid.,chaps. 1 and 2 passim;Stark,Flitcraft, (n. 85 above),pp. 468.
478.
88. Pagelow (n. 50 above).
89. Martin(n. 83 above).

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 521

theyare doing somethingwhichought to be a minimalrequirementfor

all scholarshipon thisquestion-considering violence betweenmen and
women as an aspect of overall social relationsbetween the sexes, par-
ticularlyas theyare experienced withinthe family.

III. Incest

Incest has long been labeled a crime, albeit a rare one, but not a
problemof familyviolence.Althoughwe are examiningincestwithinthe
frameworkof familyviolence,it seems advisable firstto consider briefly
the general intellectualand cultural legacy about incest we all share.
Because the incesttaboo was oftenassumed to be universaland itsviola-
tion always considered with horror and dread, incest has been sur-
rounded by extreme secrecy90and is thereforedifficultto document.
To understand incest as a type of familyviolence, we must distin-
guish between at least two differentmeanings of incestand its "taboo."
The most well-knownanthropological studies suggest that the incest
taboo in simple societies was directed toward the enforcement of
exogamy.91If we accept that definitionof its purpose, it follows that
what was tabooed was reproductivemating and the formationof new
familyunitsbetweenblood relatives.Nonreproductiveor "casual" sexual
relationswere not the taboo's major targets.Furthermore,the taboo was
concerned withmany formsof potentialincest-brother-sister,mother-
son, father-daughter,and so on-although always more stringentfor
some dyads than forothers. Most societiesare successfulat maintaining
this taboo, for even if sexual contactoccurs betweenclose relatives,law
and social pressure usually prevent the pair from becoming a socially
acceptable couple.
Through the centurythathas passed since child protectionagencies
were established,but especially in the last decade, those who cared to
inquire could know that another form of incest is in fact a relatively
frequentoccurrence: sexual relationsbetween relativeswithoutthe for-
mation of new families through marriage. The most common form,
man-girl incest, is now estimated to involve at least 1 percent of all
girls.92Incest this frequentis hardlytabooed.
90. For generations,anthropologistshave debated the purposes and functionsof this
taboo, but only rarelywere the anthropologists'theoriesrelated to contemporarysociety.
For a summaryof these debates, see Herbert Maisch, Incest(London: Andre Deutsch,
1973). See also Talcott Parsons,"The Incest Taboo in Relation to Social Structureand the
Socializationof the Child," in Social Structure (New York: Free Press, 1970),
and Personality
pp. 57-77.
91. See Claude Levi-Strauss,"The Family," in The Familyin Transition:Rethinking
Marriage,Sexuality,ChildRearingand FamilyOrganization, ed. Arlene S. Skolnickand Jerome
M. Skolnick(Boston: Little,Brown & Co., 1971), pp. 50-72.
92. David Finkelhor,SexuallyVictimized Children(New York: Free Press, 1979), p. 88.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
522 Breinesand Gordon Review:FamilyViolence

The recent wave of feminismhas been primarilyresponsible for

draggingincestout of the closet,so to speak. In the lightof day, we can
now perceive three factsthat had previouslybeen obscured: (1) When
incestoccurs as a social problem-that is, when it appears to its partici-
pants, their intimates,or controllingsocial agencies as disruptive-it is
primarilya relation between older men and young girls,less often be-
tween men and boys. Sexual relationsbetween older women and chil-
dren appear negligiblein incidence, and sexual relationsbetween chil-
dren of similarages are frequentlyviewed as nonproblematicbyall who
knowabout them.(2) Man-girlincestis by no means extremelyrare. The
1 percentestimateis a low one, and it would mean thatthree-quartersof
a millionwomen eighteen years and over have had such an experience,
withthousands more added each year. Other studiesindicatethatup to
one out of every four women in the United States is a victimof sexual
molestationby the time she reaches eighteen,and 10 percentare incest
victims.93(3) Incestuous relationshipsbetweenmen and girlsare usually
experienced bythe girlsas coerciveand assaultive-in short,as a formof
child abuse. (A minorityof writers on this subject challenge the
classificationof incest as a form of abuse,94 but today incest is being
viewed as such by many researchersand social workers.)
Once placed withinthe frameworkof familyviolence,incest,it be-
comes evident,shares characteristicswithwifebeating,since it is usually
an assault by a male against a fenale. In cases of child sexual abuse,95
approximately92 percentof the victimsare femaleand 97 percentof the
assailantsare males; incestcases followthe same pattern.96Interestingly
93. Susan Forward and Craig Buck, Betrayalof Innocence:IncestandlIts Devastation
(New York: Penguin Books, 1978), p. 3. JudithHerman and Lisa Hirschmanlestimatethe
incidence of father-daughterincestat about 2-3 percent of their psychotherapyclients,
whichaccords withthe KinseyReportdata in which 1.5 percentof women surveyedstated
that they had been molested by their fathers (Judith Herman and Lisa Hirschman,
"Father-DaughterIncest,"Signs:Jour1nal qf Womenin Culturearl Society2, no. 4 [Summer
1977]: 735-56, esp. 742). See also Kee MacFarlane, "Sexual Abuse of (hildren," in Chap-
man and Gates, eds. (n. 82 above), pp. 81-109, esp. p. 86.
94. See John Money's intrLoduction to the incestsection in Williamsand Money, cds.
(n. 9 above), pp. 411-14. They believe thatwe live in a sexuallyrepressivesocietywhichis
responsible for all sorts of perversions,including the inabilityto express affection.'I'he
family,theyhold, can undo these evils by promotingopen sexuality.For a response, see
Finkelho- (n. 92 above), pp. 12, 13. Unfortunately,these advocates of sexual freedom
lack concern for or interestin male/femalepower relations,power differencesbetween
adults and children,and coercion.
95. By sexual abuse, we mean not only intercoursebut all formsof sexual f'ondling
and genitalcontactwhen theyare coercive,or when thechild is simplynotaware enough to
give informedconsent. The legal definitionsof incestvaryby state,but generallyinclude
sexual relationsbetween relatives,and may also include sex between nonkin whose re-
such as stepfathersand daughters.
lationshipis sociallylike that of famnily,
96. Herman and Hirschman (n. 93 above), p. 736. Finkelhor'ssample showed that
sexual experiences for boys are homosexual (85 percent) and fiorgirls,heterosexual (94
percent) (SexuallyVictimized [n. 92 above], p. 75). See Judith Herman (with Lisa
Childr-en

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 523

and predictablyenough, until very recentlythe clinical literatureig-

nored this feature of incest, implyingthat, for example, mother-son
incest was as prevalent as father-daughterincest. Despite the statistics
produced by the published studies, which all agree on the over-
whelminglyman-against-girlpattern of child sexual abuse, some re-
searchershave been reluctantto admit this male-dominantpatternand
have suggested, for example, that there is a great deal of woman-
against-boysexual abuse that remains hidden because the boy victims
may not wishto reveal it. This desire to "explain away" realityexpresses,
perhaps, how threateningan honest look at child sexual abuse can be to
nonfeminists.
It is not surprisingthereforethat feministshave produced most of
the recentscholarshipon incest.As the women's movementempowered
wife-beatingand rape victimsto speak out, so also were incest victims
heard and believed by consciousness-raisinggroups and/ortheirfemale
therapists.The victims'new abilityto describetheirplightmerged witha
resurgence of attentionto child abuse and led to its consideration by
doctors and child abuse teams in hospitals, by social workers,and by
others who work with children. Only very recentlyhas the school of
sociologythatdominatesthe familyviolence fieldconsidered incest.Stu-
dents of Straus are beginning to collect data on the sexual abuse of
children,and one of them,David Finkelhor,has published a book, Sexu-
ally Victimized Children,which includes informationon incest.The other
rapidly growing genre of literatureis female, first-personpopular ac-
counts by victims,some of them best sellers.97Incest, a heretoforeun-
mentionable subject, is now part of popular consciousness, explored
oftenin Ann Landers's columns,televisiontalkshows,and movies,to say
nothingof its explosion in pornographyof all types.
In the following pages we will brieflypresent the conventional
interpretationof incestfound in the nonfeministclinicalliterature,con-
sider Finkelhor'sbook, and conclude witha presentationof twoexcellent
studies by clinicians: one byJudith Herman with Lisa Hirschman,and
the other by Florence Rush.
Untilthe recentinfluenceof feminism,Freud's workwas the source
of the conventionalscholarlyanalysisof incest.In thisclinicalliterature,
first,the emphasis was on childhood sexual desire for the parent,with
many incest accusations by children dismissed as fantasy.Second, the

Hirschman),Father-Daughter Incest(Cambridge, Mass.: Harvard UniversityPress, 1981),

pp. 12-19; Florence Rush, The BestKeptSecret:Sexual AbuseofChildren(Englewood Cliffs,
N.J.: Prentice-Hall,Inc., 1980). See also the importantinterdisciplinary
book focusingon
treatment,which will not be dealt within this review: Ann Burgess et al., Sexual Assaultof
Childrenand Adolescents (Lexington, Mass.: Lexington Books, 1978).
97. E.g., KatherineBrady,Father'sDays (New York: Dell PublishingCo., 1981); Louise
Armstrong,Kiss Daddy Goodnight(New York: Pocket Books, 1978); Sandra Butler, Con-
spiracyof Silence(San Francisco: New Glide, 1979); Forward and Buck (n. 93 above).

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
524 Breinesand Gordon Review:FamilyViolence

extraordinarilyhigh prevalence of father-daughterincestwas ignored,

and all kindsof incesttreatedas iftheirincidenceand impactwere equal
to one another. Third, the main factor advanced to explain father-
daughter incest was the "collusive mother." For example, the Kempes
say thatmanyfathersmovingtowardincestuousbehavior"are giventhe
extra push by a wifewho arranges situationsthatallow privacybetween
fatherand daughter.... Stories from mothersthat they could not be
more surprised can generallybe discounted-we have simplynot seen
an innocent mother in long-standing incest, although the mother
escapes the punishmenther husband is likelyto suffer."98 The expecta-
tionsof the motherassumed or articulatedby mostcliniciansare classic
examples of the "fantasyof the perfect mother" (and reminiscentof
maternal bonding theory),expectations which they reinforce as "ex-
perts" in the managementof our privatelives: the mother's"emotional
abandonment" of the daughter and husband leads to incest.99The
mother-blamingusuallyentailsdiagnoses of frigidity or, at least,lack of
sexual interesttowardthe husband. It is because he does not get enough
sex, in thisversion,thathe turnsto his daughter.Thus it is her wifelyas
well as motherlyfailurethatcauses father-daughterincest.
A fourth common factor in this traditional understanding of
incest-one of the legacies of Freudian psychoanalytictheory-is its
placing of responsibilityupon the child's seductiveness toward her
father.Incest victimsare oftendescribed as sexual before theirage and
responsible for temptingthe man. Fifth,to the extent the fatherwas
allocated any responsibility,he is considered immature, perhaps an
alcoholic,unable to communicate,a pedophile, and so on-certainly not
normal. At the deeper level the nonfeninist clinical and the popular
literatureunilaterallyabsolve the father. He is of little psychological
interest;he is marginal to the scenario in which mother and daughter
are at centerstage.The reader of thisliteraturecould easilyforgetthatit
is the fatherwho has coerced his child,who is the adult and aggressor in
the situation.There seems no curiosityabout him-his actionsare simply
reactions to his wife and daughter. He is not responsible for what he
does; theyare.
Before turningto the feministcritiqueof thistraditionalinterpreta-
tion, let us note a solid challenge to some of its facts in Finkelhor's
SexuallyVictimized Children.Finkelhor'swork,althoughbased on a limited
sample of college students,is an excellent studyof the sexual abuse of
children.He uses quantitativeresultsskillfully, situatingthemin current
political debates raised by the women's movement about sexual reform
98. Keinpe and Kempe (n. 6 above), p. 48. In Betrayalof Innocence(n. 93 above),
Fo-rwardand Buck statethatfather-daughter incestis the resultof "the aggressor'sattempt
to findthe tendernessand understandingthatshould issue fromhis relationshipwithhis
wife but usually does not" (p. 33).
99. Forward and Buck (n. 93 above), p. 45.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 525

and liberation.He is sensitiveto the issue of gender and avoids blaming

the victim.He suggeststhat sexual victimizationis a formof social con-
trolof women,whichbegins withgirlchildren,and may have its sources
in male supremacy.
His startlingfindingsprovide one of the major sources of data on
incest and other sexual abuse of children. Twenty-eightpercentof the
women and 23 percentof the men admitteda sexual experience witha
familymember; 9 percent of the women had found the experience
coercive and victimizing.100 Finkelhor'sstudyrepresentsthe best of the
sociological school in which he was trained: a nonspeculativecoherent
statementof intelligentlycollected factswhich are shocking in and of
themselves;an avoidance of psychologicalexplanations in favorof en-
vironmentaland institutionalfactors;and an inclusive presentationof
the issues and theirpossible causes. He is, characteristically,
unwillingto
generalize or theorize, stating that there are many different kinds of
sexual abuse, each of whichmay require a differentexplanation. He has
admirablydescribed the main featuresof incestin a book thatwillbe a
resource for many years to come.
While Finkelhorprovidesus withfactsand makes suggestionsabout
their possible significance,it is the clinical work that has approached
their cultural and political meanings. These studies are Herman and
Hirschman's 1977 article, "Father-Daughter Incest"; Herman's 1981
book of the same name (to whichHirschmanalso contributed);Florence
Rush's 1980 book, The Best Kept Secret;and Kee MacFarlane's article,
"Sexual Abuse of Children." Herman begins by stating that "female
childrenare regularlysubjectedto sexual assaultsbyadult males who are
part of their intimatesocial world. The aggressors are . . .neighbors,
familyfriends,uncles, cousins, stepfathersand fathers.This disturbing
fact,embarrassingto men in general and to fathersin particular,has
been repeatedly unearthed in the past hundred years, and just as re-
peatedly buried. The informationwas simply too threateningto be
maintained in public consciousness."101The factsthat men in general
and fathersin particularare sexual abusers, and that children do not
usually fantasizeor lie about their incestuous sexual encounters,form
the center of feministinterpretationsof incest.
In Herman's chapter "The Rule of the Father" (as in Rush's "The
Freudian Coverup"), a feministanalysis of contemporaryincest starts
withthe asymmetry of the incesttaboo in a male-dominantfamilysystem
and culture. As Freud described it, the boy child learns thathe may not
possess his motherbecause his fatherwill punish him,but thatwhen he
is an adult,he willpossess a woman of his own and possiblya daughterof
his own whom he probablywill give away to another man. A girl learns

100. Finkelhor(n. 92 above), p. 83.

101. Herman (n. 96 above), p. 7.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
526 Breinesand Gordon Review:FamilyViolence

thatshe may acquire power onlyindirectly,throughher connectionto a

man. In patriarchalsocieties the rightsof ownership and exchange of
women withinthe familyare vestedin the father;because men make the
rules, the taboo against sexual contactwiththe daughter does not carry
the same force as that which prohibitsincest with the mother. Incest
betweenmotherand son, on the other hand, is an affrontto the father's
power and is punished more severely.
Herman found that victims' families shared some important
characteristicsof which the most importantwas the estrangementof
motherand daughter. Many of the motherswere dysfunctionaldue to
mental and physicalillness of all kinds or alcoholism,were invalids,or
were absent from home. Herman does not blame the mothersfor the
incest in theirfamilies,pointingout that many were themselvesvictims
of their husbands' violence and abuse, as well as of other sorts of de-
privations.The oldest daughters often took over the mother'sduties-
housework,cooking,caring foryoungersiblings.The daughtersdid not
confide in their mothers,or had their attempted communicationig-
nored. They learned that the motherswere dependent on their hus-
bands for their own survival and could not protect anyone else; the
daughtersexperienced theirmothersas cruel,depriving,or at best indif-
ferent;and the daughters feltabandoned by their mothers.
Herman also found thatmostvictimsexpressed some warmfeelings
toward the fathers,who made them feel special. They were much more
willingto forgivetheir fathersthan their mothers,or themselves.Be-
cause a molested daughter is dependent upon her fatherfor care and
protection,perhaps the only affectionshe receives,she may not allow
herselfto feel anger at being used;'02 she may see or have no alternative
but to endure the situation.
There is, however,new evidence thatincestvictimsdo in facttryto
stop the incestby seeking outside help or strikingback. Often both are
unsuccessful.The storiesin the first-personnarrativesof girlswho tell
clergymen,teachers,or neighborsonly to be rebuffed,disbelieved,ig-
nored, or humiliatedare excruciatinglypainful and contributea sober-
ing correctiveto the idea that publicitywill put an end to incest.'03
The daughter maygain a semblanceof power in the familythrough
bargainingwiththe fatherforthingsshe wants.But her mainexperience
is shame and guilt,isolation,and an oppressive,disproportionatesense
of responsibilityfor holding the familytogether,which she accom-
plishes in part by keeping her secret.
Perhaps the most disturbinginsightHerman and Hirschman offer
102. Herman and Hirschman (n. 93 above), p. 748.
103. See Armstrong(n. 96 above) and other popular accounts. See also the un-
published paper by Linda Gordon, "The 'Normality'of Incest: PreliminaryFindingsof an
Historical Study"; and Maria Ramas, "Freud's Dora, Dora's Hysteria:The Negation of a
Woman's Rebellion,"FeministStudies6, no. 3 (Fall 1980): 472-510.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 527

is that the father-assailantfeels no contrition.The fathers"reacted to

their wives' illnesses as if they themselves were being deprived of
mothering.As the familyproviders,they feltthey had the rightto be
nurtured and served at home, if not by their wives then by their
daughters."104The incestuous fathersdisplayed an overwhelminglack
of parental and protectivefeelingstowardtheirown children.They did
not take on nurturing functions to make up for the mothers' in-
capacities, as the motherswould have done in the reverse situation.105
Not onlywere the assailantsnot sorry,theyoftendid not understandthe
destructivenessof the incest,expressed no parental or nurturantfeel-
ings forthe victim,and blamed her or theirwives.106Ironicallyoftenthe
daughter-victimreinforcedthis view, blaming the mother and herself
and exoneratingthe father.When incestcomes to light,it is remarkable
how much guilt is feltby the woman-victim.
The greaterthe paternaldominance and authoritarianismin a fam-
ily, Herman suggests,the more may violationsof the incest"taboo" be
expected. Furthermore, it is importantto note thatthe "weak" mothers
often found in incest familiesreflectan extreme male-dominantfamily
power structurethatis merelyan exaggerationof normal patterns.Male
dominance is particularlynormal in sexual relations.In our culture it is
the norm for husbands to expect sexual as well as housekeeping and
child-raisingservice from wives. One source of the incestuous fathers'
lack of guilt feelingsmay stem fromthe factthat many men are accus-
tomed to the experience of sex witha weaker and unwillingpartner.107
Herman and Hirschman argue, as do Rush, MacFarlane, and Fin-
kelhor, that father-daughterincest will "disappear only when male
supremacyis ended." And theywarn that effortsto reintegratefathers
into the world of childrenis a positivedevelopment,"only on the condi-
tion thattheylearn more appropriate parental behavior."'08A mechan-
ical inclusion of fathersas primary parents is not enough to ensure
healthy and nurturant parent-child relationships. This argument is
based on feministpsychologicalwork that analyzes the significanceof
female mothering,a resultof which is the suppression in males of the
capacity for nurturanceand affectionateidentificationwith women.109
In our opinion, the child's circleof nurturancemustideallybe widened
to include the father, and others, in order to undo the blaming of

104. Herman(n. 96 above),p. 79.

105. The natureof boys'socialization and the exclusively
intomasculinity female
parentingcommoninour society meanthatmostmendo notlearnhowtonurture others.
See Chodorow(n. 38 above),and Herman(n. 96 above),pp. 55, 56, 62, 63.
106. See Hermanand Hirschman (n. 93 above);Brady(n. 96 above);Armstrong
(n.
96 above);and Forwardand Buck(n. 93 above).
107. Gordon(n. 103 above).
108. Hermanand Hirschman (n. 93 above),p. 756.
109. Herman(n. 96 above),pp. 55, 56.

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
528 Breinesand Gordonr FamilyViolence
Reviezwt:

mothers,to recognize the social factorsthatimpinge on the family,and

to eliminate sexually abusive fathers.Yet we agree with Herman and
Hirschman in supporting measures which in the meantime will
strengthenand support mothersin the family,since daughters feel vul-
nerable to theirfatherswhen their motherscannot protectthem.
A reversalof conventionalconcerns,the centralpoint in these new
interpretationsof incest,is the responsibilityof the assailant,the father
or older male relative.Freud, as Herman and Rush show,was unable to
accept his female patients' stories of having been seduced by their
fathersand believed insteadthatthese were fantasiesbased on desire for
the father.Herman and Rush suggest the storieswere true. New data
about sexual abuse of children and of incestdocument this. Instead of
focusing on the seductive child or the collusive mother,the feminist
interpretationsidentifythe fatheras the responsibleparty,and a male-
dominant society,culture,and familysystemas the context for his ac-
tions.
Rush's The BestKeptSecret:Sexual AbuseofChildrenconsiders sexual
abuse froma cultural perspective.Although Rush is a psychiatricsocial
worker,hers is not a clinical study but a sweeping cultural and social
indictmentof a male-dominantculture which encourages the sexual
abuse of childrenand exonerates theirmale assailants.Included in it is
an importantchapteron the sexual abuse of boys.The general eroticiza-
tionlof children is part of our historicalheritage and popular culture,
according to Rush, and directlyimplicatedin children'ssexual abuse.
Thus far the most perceptiveinsightsinto incest have come from
people withpsychologicaltrainingor viewpoints.Perhaps because incest
is so deeply enmeshed in culture,psychology,and sexuality,an analysis
of environmentalfactorsalone cannot tell us as much as we need to
know. It is not accidental, moreover, that most of the recent contri-
butions are by women sensitiveto the deeper meanings of gender and
sexualityembedded in our familyand sexual relations.Their work has
contributedenormouslyto exposing the biases of earlier work,showing
itssourcesin male supremacy,and describingthedreadful harmdloneto
children.

IV. Conclusion

The recognitionand analysisof violence betweenintimatesis stillin

an earlystage. Its underdevelopmentis due partlyto the loss, througha
discontinuityof attention,of insightsfrom earlier periods of concern
wviththis problem, and partlyalso to the great social changes of the
mid-twentiethcenturythat make the familyan intenselyproblematic
social form.In any case, the existingfamilyviolence scholarshipallows
no final conclusions. Each year has br-oughtnew informationand in-

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Si-ns Spring1983 529

sightsfromscholars,activists,and serviceproviders.We hope that pro-

cess of learning can continue, withstandingpressures to save public
funds and to save the "image" of the familyby sweeping its brutalities
under the rug once again.
In our attemptsto evaluate these understandingsin midstream,as it
were, from a feministperspective,several themes make repeated ap-
pearances. All of them reinforcethe importanceof analysisof the family
as a locus of struggleas well as of support, an analysis that points to
gender and age structuringas a source of power differencesand per-
sonal tensions.This analysis,developed in (by now) scores of books and
hundreds of articles,provides a tool by which familyviolence scholars
are able to connect their topic withother, nonphysical,tensionswithin
the family,and indeed witha deeper understandingof familyand in-
timacyaltogether. Unfortunately,analysis of the familyis more often
than not neglectedbyfamilyviolenceresearchers.We suspectthisresults
from fear of the challenges feminismthrowsout to professionalschol-
arship,and fromthe academic tendencyto enforcespecializationand to
focus on extremelynarrow topics.
Individuals in our societyinheritdifferentdegrees of power de-
pending upon theirsex, class, race, and age. These differencesare acted
out in intimaterelations.The familyin particularis an institutionthat
permits and sometimes even encourages its members psychologically
and/orphysicallyto hurtone another. But whatrequiresemphasis is that
people oftenhurtone anotherin recognizablepatternswhichmirrorthe
power (or lack of it) that differentfamilymembers hold in society.A
gendered and generational understandingof child abuse, incest,and
wifebeatingrequiresthatwe analyze who is violentto whom,whatforms
the violence takes,and what itseffectsare. In thisway,patternsemerge;
violence is not random and haphazard. For example, women are more
likelyto be aggressorsin child abuse than in any other kindsof violence.
They are, however, rarely sexual abusers of children; almost all such
abusers are males. Females are the victimsin incestand wifebeating,and
their engagement in spousal violence often has differentcauses and
differenteffects from men's violence. Remembering the traditional
gender definitionsof masculinityand femininityand of motherhood
ancdfatherhood (which evolve from the sexual division of labor in the
family)and seeing themwithinthe contextof our violentculturemake it
possible to begin to understandthe causes and patternsof violencein the
family.
The abstract level of many generalizationsabout familyviolence
suggeststhe need formakingdistinctionsamong the differentbehaviors
thatgo under thatrubric.For example, the same act of child abuse-say,
spanking a child with a strap until he or she is wounded-may have
different meanings depending on whether it is performed by an

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
530 Breinesand Gordon Review:FamilyViolence

alcoholic housewife, an elderly grandfather,a teenaged baby-sitter.

Child neglect-say, leaving a five-year-oldunattended-may have dif-
ferentmeaningsdepending on whetheritis done by a singlemotherin a
housing project,a suburban housewife,a couple who are illegal aliens in
New York. To understandany act of familyviolence requires lookingat
its overall contexts and patterns, but simultaneously searching for
specificmeanings. Closely connected to this need for specificityis the
importanceof qualitative,as well as quantitative,formsof insight.This
includes looking at the participants'self-understanding.It also involves
the integrationof psychologicaland cultural analyses into behavioral
descriptionsof intimaterelationships,whether between children and
parentsor betweenadults. No act of violence is simplythe pittingof one
individual against another; each contains deep cultural and psycho-
logical meanings. At the same time, no act of violence is merely the
expression of a social problem (or a culture) such as povertyor un-
employmentor male dominance; each is also the personal act of' a
unique individual.It does not sufficeto propose, as a fewfeministshave,
that wife beating mechanicallyreflectsand expresses male dominance,
since, as we have argued, thiswould lead to the false conclusion thatall
heterosexual relationshipsare violent. How male supremacy,class and
race domination,and acute social stressinformspecificacts of violence
requires an analysis of psychic processes. Consumer capitalism,
militarism,televisionand other media sexismand violence,forexample,
implicateaggression, passivity,and masochismin individual sexual be-
havior; understandinghow this happens, and how to eliminate it, re-
quires close attentionto daily life processes.
We have by no means won the battle against victimblaming in
familyviolence literature.Indeed, we suspect that one form of victim
blaming-mother blaming,an old Americantradition-is being revived.
The incestuous father and the wife beater claim less attentionfrom
many cliniciansthan the victimand the victim'smother.Child abuse is
attributedto the failureof mother-childattachmentin the "familyvio-
lence adaptation" of the maternalbonding literature.Common to this
work is a focus on the individual relationship,out of context and in
isolation,in such a wayas to obscure the largerissues and patternsand to
magnifythe woman's psychologicalshortcomings.
The largesttaskof feministfamilyviolenceanalysisshould be, how-
ever, not only to defend women against victimblaming but to demand
the constant,relentlessapplicationof analysisby gender to all discussion
and practicerelated to these problems.The simplestinsightof thisnew
wave of feministscholarshipmay be the hardestto convey: gender must
be a categoryof analysisfor all social issues. There is particulardanger
now, when many are clingingto the potentialsupport structuresof the
family,thatthe familyas the historicallocus of sharp strugglesbetween

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Signs Spring1983 531

the two sexes and differentgenerationswillbe forgottenor denied. The

pressures on men to provide, the pressures on women to parent and
nurture, the pervasive messages of inadequacy that create anger and
self-hatredin so manyof us, the povertyof communitythatplaces such
inflatedexpectations on our intimates:these and many other factors
conspire to make us attackone another,even those we love. Our frustra-
tion and anger mightcertainlybe channeled more usefully,not at each
other,but toward changing the institutionsthatoppress us.

DepartmentofSociology
Northeastern (Breines)
University

DepartmentofHistory
ofMassachusetts-Boston
University (Gordon)

This content downloaded from 205.208.116.024 on November 20, 2017 10:42:41 AM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).