London School of Economics

London School of Economics
The Incest Taboo in Relation to Social Structure and the Socialization of the Child
Author(s): Talcott Parsons
Source: The British Journal of Sociology, Vol. 5, No. 2 (Jun., 1954), pp. 101-117
Published by: Wiley on behalf of London School of Economics and Political Science
Stable URL: http://www.jstor.org/stable/587649
Accessed: 29-11-2015 12:45 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
Wiley, London School of Economics and Political Science and London School of Economics are collaborating with
JSTOR to digitize, preserve and extend access to The British Journal of Sociology.
http://www.jstor.org
This content downloaded from 130.64.11.153 on Sun, 29 Nov 2015 12:45:53 UTC
All use subject to JSTOR Terms and Conditions
The Incest Taboo in Relation to Social
Structure and the Socialization of the
Child [I]
TALCOTT PARSONS
t FTER SOMETHINGlike a generation in which the attention of

^ anthropologistsand sociologistshas been focusedon the phenomena
s Swhich differentiateone society from anotherand the differentstruc-
tures within the same society from each other,in recentyears there has been
a revival of interestin the problemof what featuresare commonto human
societies everywhereand what are the forces operatingto maintain these
commonfeatures. One reasonfor my presentinterestin the incest taboo is
that it is one of the most notable of these commonfeatures. With all the
variabilityof its ineidenceoutside the nuclearfamily, there is the common
eore of the prohibitionof marriageand in general of sexual relationships
between membersof a nuclearfamily except of coursethe conjugalcouple
whose marriageestablishesit.
In the older discussionsthe prevailingtendencywas to attempt to find
a specific" cause" of the taboo, thus instinctiveaversionor Westermarck's
contentionthat aversionwas acquiredthroughbeing broughtup in the same
household. As our empiricalinformationand theoretical resourceshave
accumulated,however, it seems less and less likely that this is the most
fruitfulapproach. On the contraryanythingso generalas the incest taboo
seems likely to be a resultantof a constellationof differentfactorswhichare
deeply involvedin the foundationsof humansocieties. Analysisin termsof
the balanceof forcesin the social system ratherthan of one or two specific
" factors" seems much more promising. Furthermore,it seems highly
probablethat a combinationof sociologicaland psychologicalconsiderations
is involved; that a theory which attempts to by-pass either of these fields
will find itself in difficulties.
The elementof constancyclearlyfocusesin the nuclearfamily. Perhaps
the most recentauthoritativesurveyis that of Murdock,[2] and we have his
authoritythat no society is knownwhereincest betweenmother-son,father-
daughterorfull brother-sister is permittedexceptthe fewcasesof brother-sister
IOI
I02 THE INCEST TABOO
marriagein royal families,but neverfor the bulk of the people. Thereare a
fewcasesof marriagepermittedbetweenhalf-brother andhalf-sister,andsimilar
cases of closeness,but only a few. I shall thereforetake the nuclearfamily
as my point of departureand attempt to review a few highlightsof it as
a sub-systemof the society. But the nuclearfamily is, in my opinion,only
the focus of the structuralproblem,not the whole of it. I shall therefore
next attempt to link with the relevant considerationsabout the family,
a series of problemsabout its place in and articulationwith the rest of the
society. Then, given this widersetting of social structure,I will attempt to
analyse some of the relevantproblemsof psychologicalmechanismin terms
of the characteristicsand signiScanceof eroticismin personalrelationships
and in the personalityitself.
I. THE STRUCTURE
AND FUNCTIONS
OF THENUCLEAR
FAMILY
The universalityof some orderof incest taboo is of course directlycon-
nected with the fact that the nuclearfamily is also universalto all known
human societies. The minimalcriteriaof the nuclearfamily are, I suggest,
first that there should be a solidaryrelationshipbetween mother and child
lasting over a period of years and transcendingphysical care in its signifi-
cance. Secondly,in her motherhoodof this child the woman should have
a specialrelationshipto a manoutsideherowndescentgroupwhois sociologically
the " father" of the child}and that this relationshipis the focusof the " legiti-
macy" of the child, of his referentialstatus in the largerkinshipsystem. [3]
The commonsense of social sciencehas tendedto see in the universality
and constancyof structureof the nuclearfamily a simple reflectionof its
biological function and composition; sexual reproduction,the generation
differenceand the differentiationby sex in the biologicalsense While I in
no way questionthe importanceof this biologicalaspectand am in agreement
with the view that the human family is an " extension" of a subhuman
preculturalentity, on the human-culturallevels there is, I am sure}another
aspect of the problemof constancy. The two biologicalbases of differentia-
tion, sex and generation,may be regarded,that is, as "-pointsof reference"
of a type of socialorganizationthe sociologicalsignificanceof whichis general
in the structureof small groups.
Evidence from the experimentallaboratorystudy of small groups [4]
hasshownfirstthat smallgroupswith no priorinstitutionalizeddifferentiation
of status, differentiatespontaneouslyon a hierarchicaldimension,whichI may
call " power" in the sense of relativeinfluenceon the outcomeof processes
in the system. This is the case when this differentiationis measuredby
any one of a variety of possiblemeasures,both fromthe point of view of the
observerand that of participantsin group process. We may say there is
a differentiationbetween " leaders" and " followers".
Secondly, there appears a differentiationwhich cuts across this one,
with referenceto qualitativetype of functionin the group. The first broad
TALCOTT PARSONS Io3
qualitativetype of differentiationwhich appearsin this sense is what Bales
and I have called that between primarily" instrumental"function in the
group and primarily" expressive" function. An instrumentalfunction is
one primarilyconcernedwith the relations of the group to the situation
external to it, includingadaptationto the conditionsof that situation and
establishmentof satisfactorygoal-relationsfor the system vis-i-vis the situa-
tion. Expressivefunction on the other hand is concernedprimarilywith
tho harmonyor solidarityof the group,the relationsinternallyof the members
to each other and their " emotional" states of tension or lack of it in their
roles in the group.
Levelof differentiation is of coursea functionof size of the group. By the
time we reacha membershipof four there can be a typical four-rolepattern,
differentiatedhierarchicallyinto leadershipand followershiproles, and quali-
tatively into more instrumentaland more expressiveroles. I would like to
suggestthat it is fruitfulto treat the nuclearfamily as a specialcase of this
basicallyfour-rolepattern,with generationas the mainaxis of superior-inferior
or leader-followerdifferentiation,sex the axis of instrumental-expressive
differentiation. Obviouslythe helplessnessof the child, particularlyin the
first years, is the main basis of the former. The universalfact that women
are more intimately concernedwith early child care than are men (with
lactation playing a very fundamentalpart) is the primaryreason why the
femininerole, in the family as well as outside, tends to be moreexpressive
in this sense than the masculine.[5]
My first point is thus that the nuclearfamily has certaincharacteristics
commonto small groupsin general. The effectivenessof its performanceof
functionas a familyis, I think, dependenton its havingthese characteristics.
The primaryfunctionsI have in mind aLrea certain significancefor main-
taining the emotionalbalances of all membersof the family includingthe
adults, and its paramountrole as an agency for the socializationof children.
The general charactensticsI have in mind are three. The first is that it
should be a small group,especiallyin its higherstatus-echelon. Given age-
specificdeath rates as well as birth rates presumablyin no society does the
effectivenuclearfamily averagemorethan about seven members,and gener-
ally fewer. The secondcharacteristicis that the main structuraldifferentia-
tion of the family as a group should be along these two axes, namely that
of poweror hierarchyand the instrumental-expressive distinction. 'Thethird
is that boththe latter should be representedin the " leadership" structure
and that there shouldbe a strong " coalition" betweenthem. [6] The fact
that the two parents are of opposite sex and that marriage,though with
variations, always constitutes an important structuralbond of solidarsty
transcendingthe parentalfunctions,in a broadway insuresthis. It should
be clear from the above that sex role differentiationin its more generalized
sensewhichimpingeson manycontextsotherthan the structureof the nuclear
family itself is portantly involved in this structuralcomplex.
But this does not mean that just any kind of small group which met
Io4 TlIE INCEST TABOO
these speeifieationseould performthe funetionsof the family. It elearlyhas
to be a group whieh has relatively long duration- a eonsiderablespan of
years. But it is not indefiniteduration. One of its most importantchar-
aeteristiesis that the family is a self-liquidatinggroup. On attainmentof
maturity and marriagethe ehild eeases in the full sense to be a memberof
his familyof orientation; insteadhe helpsin the establishmentof a new one.
The implieationsof this basie faet will be brieflydiseussedin the next seetion.
Seeondly,it must be a group whieh permits and requiresa high level
of diffuseaffeetiveinvolvementfor its members; thoughthis of eoursevaries
with the differentroles,beinghighestfor the youngehild. Clearlyno evanes-
eent experimentalgroupeould performthe funetionsof a family. The fact
that with few exeeptionsthe nuelearfamily is the main unit of resideneeis
of critieal importaneein this eonneetion.
Finally, third, I suggestthat it is essentialto the familythat morethan
in any other groupingin soeieties, overt erotie attraetionand gratifieation
shouldbe given an institu-tionalized plaee in its strueture. But when we say
this is institutionalizedwe mean that erotieismis not only permittedbut
earefullyregulated; and the ineesttaboois merelya very prominentnegative
aspeet of this more generalregulation.
This aspeetwill be morefully diseussedin the thirdseetionof the paper.
But at this point it does seem worthwhile to summarizethe familiarfeatures
of the erotie organizationof the family. First genital erotieismis both per-
mitted to and expeetedof the maritalpair. Only in eertainspeeialreligious
groups is its justifieationeven in theory eonfinedto the direet proereative
funetion; it is itself a bond and a very importantsymbol of the solidarity
of the marriagepair as responsiblefor a family. But at the same time-
and this faet aeeentuatesthis meaning the maritaleouplehave a monopoly
of the right to genital erotieismwithin the nuelearfamily, though of eourse
not neeessarilyoutside.
Secondly,pre-genitalerotieismis positively institutionalized,always in
the early mother-ehildrelation,and probablyusuallyto some extent in that
of father and ehild. But clearly it is generallyfar more importantin the
ease of mother and ehild.
Third, with probably few exeeptions, overt erotie expressionexeept
possibly autoerotieismin some eases, is tabooed as between post-oedipal
ehildrenand both parents,and in the relationsof the ehildrento eaeh other,
except where an older sibling plays a partly parentalrole to a small ehild.
Finally,no homosexualityis permittedat all withinthe nuclearfamilyunless
we wish to eall the attraction between mother and pre-oedipaldaughter
homosexual. In view of what we knoxvon psychologieallevels of the erotie
potentialsof human beings this structureis elearly not one of unrestricted
permissiveness,but of a systematiccombinationof eontrolledexpressionand
regulatoryprohibition. Nloreover,in view of the wide variety of humaneus-
toms in so many respects,its relativeuniformityis impressiveand deserves
to be counted as one of the most importantuniversalsof human society.
TALCOTT PARSONS Io5
It would be rash to suggest that the socializationof childrencould not
be carriedout except in a groupof the specilAcbiologicalcompositionof the
family, or even without this specific set of erotic relationships. I think,
honvever,that it is fairly safe to contendthat the primarysocializingagency
must be a small group with broadly the sociologicalcharacteristicsI have
suggested,and that even the erotic factor could not vary extremelywidely.
For example,it couldnot be completelysuppressed,by havingall fertilization
occurby artificialinsemination,and by a carefulpolicy of avoidingarousing
any eroticintereston the partof children,or at the otherextremeby removing
all restrictionson fulfilmentof any and all erotic impulsesas and when they
might be aroused.
II. THEFAMILY
ANDTHEWIDERSOCIAL
STRUCTURE
One of the car(linaluniformitiesof social structurewhich is most inti-
mately connectedwith the incest taboo is the fact that nuclearfamiliesare
never found as independenttotal " societies" on a human cultural level.
There is never simply extra-socialbiologicalmating outside the family, but
the nuclearfamilyis alwaysa unit within a societywhichcontainsa plurality
of other families,and other types of units; " solidarity" extends over these
areas and the other groupings,and even where they are kinshipgroupings,
sociologicallythey have characteristicsvery different from those of the
nuclearfamily.
Undoubtedlyone of the main characteristicsof the more " primitive"
societieslies in the fact that a far largerproportionof the total socialstructure
is organizedabout kinship than is the case with the more " advanced"
societies. Indeed there are some where it is difficult to speak of any
" statuses" or groupswhicharenot in someimportantrespectkinshipstatuses
and groups. But two main things need to be said. First, though always
includingnuclear families, the kinship system always also includes groups
whichdifferfundamentallyfromnuclearfamilies. Secondly,it can, following
Leach, [7] probablybe said that a kinship system cannot be a completely
" closed" system in that features of it always have to be analysed with
referenceto economic,politicaland otherconsiderations whichare not peculiar
to kinshiprelations,which do not disappearin social structureswhich have
entirely cast loose from a kinship base.
Whetherthe groupingswhichtranscendthe nuclearfamilyare organized
about kinshipor not, relativeto the familythey have in general- with a few
exceptions like friendships-certain characteristicsin common. They are
groups in which the personalemotionalinterests of the individualare not
so closelyboundup as in the family; wherethe accentis moreon impersonal
functionsof the group. A good kinshipcase of this type is the lineage as
a corporateentity w-ith referenceto its political functions. The case of
ora,anizations composedprimarilyof occupationalroles in modernsociety is
one where kinship is not prominent. Broadly one may say that in such
Io6 THEINCEST TABOO
casesthe role or the organizationis charaGterized by primacyof functional

responsibilityon a social system level, and by relatively severe control of
spontaneity- by what I have elsewhere[8] called" affectiveneutral-
affective main-
ity". These are the structuresin which the main functionsof direct economic
tenanceand goal-attainmentin the society are performed; viz.etc.
provision,politicalstabilization and defence, religious expression,
Wherethe main basis of compositionof such groupingsrests in kinship,
marriage has direct functionalsignificanceas a mechanismwhich establishes
important directties of interpenetration of memberships betweenthe different
Under such circumstances marriage
elementsin the structural network. difficult
be merelya " personalaffair " of the parties to it. Where it is
cannot
tohave solidaryrelationshipswhich do not involve kinship the intermarriage
betweengroups can establish a pattern of such solidaritiescross-cutting
thosebased (lirectlyor primarilyon relationshipsby descent. has
As Fortune was one of the first to emphasize,and Levi Strauss
farther,[9] in this kindof situation it is not so much the prohibition
developed obligation
ofincest in its negative aspect which is importantas the positive out.
toperformfunctionsfor the subunit and the larger society by marrying
to contribute to the formation
Incestis a withdrawalfrom this obligation political
andmaintenanceof suprafamilialbonds on which majoreconomic,
atldreligiousfunctionsof the society are dependent.
the
Where extended kinship groupingshave a critical importancein
this kind which 1lnderlie the patterns
socialstructure,it is considerationsof the
of extension of the incest taboo beyond the nuclearfamily. Broadly which, first,
principlesseem to be that intermarriage is forbidden within units
the social
areorganizedprimarilyas kinshipUIlitS,second,have functionsin family
systemwhichtranscendthe personalinterests of the members of small
and,
groups,which thereforeinvolstea more impersonalset of disciplines
daily interaction with reference
third,groupswithinwhich,as kinshipgroups, and the
to these interestsis relativelyclose. The lineage and its segments of such
male local successiongroup which Leach discusses, are prototypes
breaks
groups. Illustratingthe last criterionit is typical that exogamyoften within their
downwithinthe most extenslvelineagegroups but is maintained
lower-ordersegments.LIO] way it
Recent work on kinshipseems to indicatethat in a very rough is the
this respect. At one end
is possibleto constructa series of types in
cross-cousin
so-called Karieratype which is characterizedby symmetrlcalbut is very
marriage. This forms a very " tight " form of organization,
social ties which can be established
limited in the rangeof differentkinds of under
through it. It makes for a rigid social structure,though probably
certain conditionsa relatively stable one.
cross-
Levi-Straussis probably right that the asymmetricaltype of
on marriage urith the mother's brother's
cousinmarriagenvhichrestsprimarily and more
daughterconstitutes an importantstep towardsa wider ranging type and with
flexibleset of arrangements as compared both with the Kariera
TALCOTT PARSON S Io7
marriageto the father'ssister'sdaughter. It is interestingto note that this
is connectedwith the asymmetryof the structureof the nuclearfamily itself
as that was discussedabove. If the masculinerole is more instrumental
than the femininein the senses I have discussed,then the men shouldhave
more direct and important anchoragesin the extended kinship groupings
than the women. Then for a woman who has marrtedout of her descent
group,the strongestsourceof supportwouldnot be her sisterbut her brother.
This is first becausethe sister may well have marriedeither into ego's own
post-maritalgroupor into anothercontrolledlargelyby her husband'sagnatic
kin and second becausein the descent group the men have more controlin
extrafamilialaffairsthan the women.
The father's brother,on the other hand, is in a status directly similar
to that of the fatherand not complementaryto him, while the father'ssister
belongsto this same agnaticgroup. Put a little differentlyan alliancewith
the mother'sbrotheris the stablestkind of alliancewith a distinctlydifferent
group and at the same time bolsters the structureof the nuclearfamily in
such a way as to redressthe balanceresultingfrom its internalasymmetry
by giving the mother external support through a channel independentof
her husband.
Levi-Straussthereforeseemsto be rightin sayingthat asymmetricalcross-
cousin marriagethroughthe mother'sbrother'sdaughterrelationshipopens
up a widercirclewhichis both stablerand moreextensivethan any alternative
where the kin involved are so close. Leach, [II] however, has made an
importantadditionalcontributionby showingthat on such a basis the kinship
system cannot be closedthroughmarriage-exchange relationsalone, but that
there are several alternativeways in which such a system can work out.
Whichof them will developwill dependon the economicand politicalrelations
of the exchangingkinshipunits, and hence on the natureand values of the
" considerations" which enter into the marriagearrangementsother than
the exchangeof spouses as such.
But all this is compatiblein a broadway with Levi-Strauss'view that
this makes women, though in somewhatdifferentways also men, a kind of
symbolic " counters" in a processof exchange. PerhapsI may state it in
somewhatdifferentterms by saying that the woman or man, in marrying
outside his own descent group, is performinga role-obligationin a social
groupor collectivitywhichtranscendshis own family of orientation,and one
to which to some degree his family is subordinated; it is a superordinate
unit in the social structure. He is no more free to marrywhomhe chooses
in such a situationthan is an industrialworkerfree within the organization
to performany job-taskhe choosesregardlessof how it fits irstothe plan for
how the total processis to be organized.
It is in this sense that incest wouldbe sociallyregressivein the sense in
which Levi-Straussanalysesthe problem. It would,in an area of the higher
integrative structuresof the society, constitute giving membershipin the
lower-levelstructurepriority over that in the higher. It is only on the
Io8 THE INCEST TABOO
impossibleassumptionthat families should constitute independentsocieties
and not be segmentalunits of higher-levelorganizations,that incest
as a
regularpracticewould be socially possible.
These considerationsgive us the basis for a furthergeneralizationcon-
cerningthe differencebetweenextendedexogamoussystemsand those
in modernsocieties. So far as the higherlevel functionsof the found
society are
performedby collectivitiesthe compositionof whichis determinedin kinship
terms,therewill be a tendencyto extendthe incesttabooto such
collectivities.
So far, however,as social function, economic,politicaland religious,
to be organizedin groups not put together out of kin, the whole comes
issue of
exogamy with referenceto them will cease to be signiEcant.
There is, however, complete continuity between these two types
cases so far as certainaspects of the social functionsof the incest taboo of
are
concerned. We may say that there are two primary interconnectedbut
independentaspectsof this function. In the firstplace,it is sociallyimportant
that the nuclearfamily shouldnot be self-perpetuating and hencethat adults
shouldhave a personalitystructurewhich motivatesthem to found new
and
independentnuclearfamilies. Erotic attractionto personsof opposite sex
butoutsidethe nuclearfamilyis clearlya mechanismwhichaids in this.
But,
secondly,it is essential that persons should be capable of assumingroles
whichcontributeto functionswhich no nuclearfamily is able to
whichinvolvethe assumptionof non-familialroles. Onlyif such perform,
rolescan be adequatelystaffed can a society function. I suggest non-familial
that the
criticalrolesin this class are rolesin whicheroticinterestsmust be
altogether
subordinated to other interests.
I thus see the " problem" of the incest taboo in the following
sofar as social structureis concernGd.It seemsto be clearthat setting
humanper-
sonalitiesare universallysocializedin nuclearfamilies,whichare smallgroups
ofthe specialtype sketchedabove. Includedin their special
is the role of erotic attraction between their members. Thecharacteristics
incest taboo
operates to " propel" the individualout of the nuclearfamily,not in one but
intwo senses. He is propelledinto a new nuclearfamily formed
by his
marriage.Here the erotic componentof his personalityis positively made
useof. But also he is propelledinto non-familialroles, which of
courseare
differentiatedby sex and other status-characteristics,but in some sense such
rolesmust be assumedby all adults. This correspondsto the fact that
known every
society consistsin a pluralityof nuclearfamiliesthe durationof which
islimited to one generationand also the fact that these familiesare
relatively always
low-levelunits in a social structurethe higher-levelunits of which
havedifferentfunctionsin the society, functionswhich cannot be
byfamily groups. [I2 It is in this setting that I wish to discuss performed
theproblemsof the psychologicalcharacteristicsof eroticismand its some of
thedevelopmentof personality. place in
TALCOTT PARSON S IO9
III. THEPSYCHOLOGICAL
CHARACTERISTICS
ANDFUNCTIONS
OFEROTICISM
After all, the most distinctivefeatureof the incest taboo is the regulation
of eroticrelationships,within the family and in relationto the establishment
of new families. The considerationsabout social structure which I have
advancedthereforeneed to be supplementedby a discussionof the nature
of eroticismand its functionsin the developmentof personalityand in the
personalityof the adult. I shall here put forwarda view which has three
main emphases. First, eroticismwill be held to play a very importantpart,
probablyan indispensableone, in the socializationof the child, in taking a
raw organismand making a " person" out of it. Second, however, the
awakeningof eroticinterestsnot only performsfunctions,it createsproblems.
There are importantpsychologicalreasonswhy erotic needs seem to be par-
ticularlydifficultto control. Makinguse of this instrumentof socialization
thereforeconstitutes a kind of " pact with the devil ". Once present the
questionof what is to be done with this forceis a seriousproblem. Finally,
third, the view of eroticismI take here will dissociateit considerablyfrom
what is ordinarilymeantby the " sex instinct" or the instinctof reproduction.
Thoughthe interestin genitaleroticismof the post-adolescentis undoubtedly
genuinelypart of the erotic complex,and a very importantpart, it is only
part, and the complex is far broaderthan such an instinct in two senses.
On the one handits childhoodor pregenitalaspectsare of fundamentalimport-
ancefor ourproblemandpresumablyhave nothingto do with the reproductive
function. Secondly,though there undoubtedlymust be a basis on constitu-
tional predisposition,the aspects of eroticismwhich are importantfor our
purposes involve a very large component which is learned rather than
" instinctive" in the usual sense.
I shall rely heavily on Freudfor my views of the eroticcomplex,though
I think Freudcan be supplementedby some considerationsderivedfrom the
sociologicalstudy of the processof socialization. But after all one of the
greatestof Freud'sdiscoverieswas the fundamentalimportance of the eroticism
of childhood-the fact of its existencewas not discoveredby Freud, but as
so oftenin the historyof sciencewell-knownfacts excitedlittle interestbecause
nobody knew how to assess their importance. Furthermore,Freud clearly
saw the importanceof the processesof learningin the developmentof erotic
interests. I may recallhis famousstatementthat " the infant is polymorph
perverse". This I interpretto mean that any normalchild has the potenti-
ality of developingany of the well-knowntypes of erotic orientation,homo-
sexuality, autoeroticismand the perversionsas well as what we think of as
normalheterosexuality. This can only meanthat the latter is in considerable
measurethe productof the processof socialization,not simplythe expression
of an instinct. [I3]
What, then, are the most importantcharacteristicsof eroticism? Erotic
interestis, I think,the interestin securinga particulartype of organicpleasure,
whichis in one aspectorganicallyspecificin a way comparableto the pleasure
IIO THE INCEST TABOO
of hunger-gratification or warmth. But this is only one aspect of it. What
is most importantabout eroticismis, I think, its dual character,the com-
binationof this organicspecificity,the possibilityof intensepleasurethrough
the strmulationof specificparts of the body, with a dsffusespreadinginto
a generalsensation of well-besng. From stimulationof an erogenouszone
then, it is not a very big step to leartiingthat almostany type of bodilycontact
with the agentcan cometo be felt as a sourceof pleasure. I maytake a specific
examplefromearly childhood. Being fed by the motheris a sourcenot only
of hunger-gratification but very early, accordingto psychoanalyticmews,
of oral-eroticgratificationas well. But from stimulationof this oral-erotic
interest there is generalizationto pleasurablesensation from any physical
contact with the sourceof the origonaloral gratification; hence being held
and fondledby the mother,is a sourceof pleasureand a focusof an incipient
system of expectations.
Put in-psychologicalterms, erotic gratificationis a peculiarlysensitive
source of conditioningin the " classical" Pavlovian sense. From desinng
the specificstimulation,the child comesto desirediffusenon-specificcontact
with the objectwhichhas servedas agentof the onginalgratification. Erotic-
ism is thus a major,in the earlierstages probablythe majormechanismfor
the " generalizationof cathexis' by which a diffuseattachmentto an object
comes to be built up. [I4]
The great importanceof diffuseattachmentin this sense to the process
of learninghas come to be well-recogrtized.So long as a socializingagent
is only a source of specific segmentalgratifications,the omission of such
gratificationwill cause the child very rapidlyto lose interestin the object.
But the processby whichthe deeperkind of learning[I5] iS possibleinvolves
the buildingup of need-systemsand then their fmstrationas a preliminary
to the learningof new goals and needs. [I6] lUheessentialpoint is that the
socializingagent shouldbe in a positionto fmstratethe child-really senously
-without losing control of him.
Anotheraspect of the point is that it is by this orderof generalization
of cathexis that the child is made sensitiveto the-aJtstr4des of the socializing
agent, say the mother. This sensitinty to attitude is possibleonly through
transcendingthe specificity of interest in organic gratificationsas such.
What mattersto the child is whetherand how muchhe feels that his mother
" cares". The very fact that erotic gratificationis not essentialto any of
the basic physiolopcalneeds of the indindual organismmakes it a suitable
vehicle for this generalization.
A furthercharactensticof eroticismseemsto be importantin the general
situation,it is what underliesmy referenceabove to its arousalconstituting
in a sensea " pact wlth the delril". Erotic need, that is, seemsto have some
of the characteristicsof addiction. The erotic interestsof childhoodcannot
be allowed to be dominantin later phases of delrelopment,and in normal
developmentare not. But the evidenceis that by and large they are not,
as the psychologlstsput it, successfully" extinguished", but ratherhave to
TALCOTT PARSONS
III
be repressed. From this it comesthat the psychoanalysisof any " normal"
adult will bring to light " infantile" erotic patterns which are still there,
though they have not been allowedovert gratificationfor many years. The
evidenceis very clearthat normaland pathologicaldifferin this respectonly
in degree,not in terms of presenceor absence.[I7] If this generalview is
correctthen the mechanismsfor handlingsuchpermanentlyrepressedmaterial
must be of great importancein the normaladult personality.
Let us look at the matterin moreof a sociologicalperspective. A social-
izing agent at any given maiorstage of the processplays a dual role, in two
systemsof socialinteraction. On the one handhe-or shc participateswith
the child at the level which is appropriateto the beginningof the phase in
question,as in the caseof the mother-childlove attachmentof the immediately
pre-oedipalperiod. On the other hand she the mother in this case --also
participatesin the full four-rolefamilysystem. In disturbingthe equilibrium
of the formerinteractionsystem she acts as an agent of the latter. This act
of disturbanceconstitutesfrustrationto the child and producesamongother
things anxiety and aggression. If, on the otherhand, there were no positive
motivationin his involvementin the relationshipother than what he is now
denied expression,the attachmentwould simply break up and no progress
could be made, since he is not yet motivatedto assumehis rew role in the
new and higher level interactionsystem.
But the specificpart of the erotic attachmentis a focus of preciselythe
elementof " dependency" at the relevantlevel whichhas to be overcome
if the newlevel is to be attained. Underthe conditionspostulated,however,
the diffuseaspect of the erotic attachmentcan survivethe frustrationof the
focal specificdesire,and it can thus becomea main lever by whichthe thild
is positively motivated to learn a new role which, it must be remembered,
involves learning new goals, not merely new instrumentalmeans for the
attainmentof given goals.
Thusthe child'seroticattachmentto the motheris the " rope" by which
she pullshim up froma lowerto a higherlevel in the hardclimbof " growing
up ". But becausethe pointsof attachmentof this " rope" remainsensitive,
interestin them is not extinguished,there is a permanentchannelback into
the still operatisreinfantile motis-ationalsystem. Serious disturbancesof
ttie equilibrium*f the personalityean always re-openthese channels This
is what is esrclinarily
meantby " regression" and early eroticpatternsals-ays
play a promlnentpart in regressivetendencies.[X8]
Thereseerrlto be threestages at whichthe rnc)ther is the priJnarXT
object
of eroticattractlotlvf the child; these are what Frelldidentifiedas the Oral;
tne Arzalan(l the Phallic phases. They correspondto tllree relatively clis-
ccntinueus ti steps" in the yvrocessof learning new levels of personallty
organizatlon; ew goals, and capacities for independentand responsible
perfornance Each oIle leaves a residuumof tEseerotic structureswhich
have been essential in order to make the step, but which if allowed to
remainactive wouldinterferewith the subsequentsteps. Thus thereis in all
II2 THE INCEST TABOO
personalities,grantingmy hypothesisof addiction,a channelthrougherotic
associations,right down into the lowest and most primitive strata of the
Id the most regressiveparts of the personalitysystem. These can be re-
activatedat any time. Theconnectionof this situationwiththe problemof the
probablepsychologicalsignificanceof incest seems to be clear.
Fromthis point of view the problemof incest fits into the largercontext
of the structuringof eroticmotivationin the personality,over time and with
referenceto choice of objects. The context includesthe problemof homo-
sexuality and of the status of the perversions. The goal of socialization-
with many variationsbut in its broad pattern universal is to establishat
least the primacy,if not the completemonopolyover other possibilities,of
normalgenital erotic attractionwhich includeschoice of object outside the
family, and stability of orientationto objects.
Only mother-sonincest is as such directlyinvolved in the constellation
I have sketched. Here the regressiveimplicationsseem very clear. This
agrees with psychoanalyticopinion that such incest, where it does appear
in our society, is always deeply pathological,on both sides but particularly
that of the son.
The case of the daughter vis-a-vis her father is somewhat different.
But when she is forcedto abandonher primaryattachmentto her mother,
it should be clear that the next availablealternativeis the father. This is
furthermade " plausible" by the fact that she is taught that it is normal
for a femaleto have a primaryattachmentto a masculineobject,but in this
case erotic developmentof the attachmentis blocked. This clearly has to
do on one level with the internalequilibriumof the nuclearfamilyas a system.
The erotic attachmentof the parentsto each other is a primaryfocus and
symbolof their coalitionas the leadershipelementof the familyas a system.
To allow the child who has just been forcedout of an erotic attachmentto
the motherto substituteone with the fatherwouldimmediatelyweakenthis
coalitionas a sourceof generalizedpressureto grow up for childrenof both
sexes.
But there is a broader" functional" aspect of the problem. If it is
exceedinglyimportantthat the boy should find a feminineobject outside
the family, this is obviouslyonly possiblein a generalizedway if girls also
typically do so. Furthermore,in orderto performher functionsas a social-
izing agent, as mother,it is extremelyimportantthat a woman'sregressive
need systemsshouldnot be uncontrolled. Indeedit is probablymoreimport-
ant than in the case of the man, becauseas a motherthe womanis going to
have to enter into much strongereroticreciprocitieswith her young children
than is her husband,and at the sarnetime she is in due coursegoingto have
to act as the agent of their frustrationin these respects. If she is not able
to control her own regressiveneeds, then the mother-childsystem is likely
to get " stuck " on one of the early levels and be unable to take the next
step. Indeed such phenomellaare prominentin the pathology of family
relationsill relation to the genesis of mental disorders. Thus the " over-
TALCOTT PARSONS II3
protective" mother, instead of, at the propertime, refusingto reciprocate

her child'sdependencyneeds, positivelyencouragesthem and therebymakes
it more difficultfor him to grow up.
I;inally,there is the case of the prohibitionof brother-sisterincest. It
seems to me that in the first instance this relates to the symmetryof the
nuclearfamily. Oncethe oedipalcrisishas beenpassed,the mostsymmetrical
arrangementis that svhichreservesa monopolyof erotic relations within
the family to the marriedcouple. But in a broadercontext functionallythe
more importantthing is at the relevant time to achieve complete though
temporary repressionof erotic needs for both sexes. Fulfilmentof this
recluirementwould be blocked by permissivenessfor brother-sistererotic
relations.
For childhooderoticismregardlessof the sex of the child the original
object is the mother. Oncethis attachmentto the motherhas ceasedto be
useful to the developmentof the personalityit tends, I have noted, to be
repressedaltogether. This meansthat not only is the originalobject denied,
but those " next in line", that is all other membersof the originalnuclear
family, are tabooed. This in turn, it seems, is an aspect of what I referred
to above as the process of self-liquidationof each particular nuclear
family.
What lTreudcalled the periodof " latency", i.e. from the point of view
of overt eroticism,thus seems to be the period in which the individualis
above all learningto performextrafamilialroles. Childhooderotic attach-
menthas playeda part in layingthe necessaryfoundationsfor these processes,
but beyonda certainpoint it becomesa hindrance. Just a wordmay be said
about the first of these steps whichseems to have a bearingon the problem
of brother-sisterattachment.
One of the primaryfeatures of the oedipal transitionin the course of
whichthe last phaseof childhooderoticismis normallyrepressedis the assump-
tion of sex role, or the first majorstep in that process. Thoughthe points
of referencefor the differentiationare unmistakablybiologicallygiven, there
is strongreasonto believe that the role, includingthe psychologicalcategor-
izationof the self, mustbe learnedto a muchgreaterextent than has ordinarily
been appreciated. It seemsto be significantthat just at this periodchildren
begin to be much more independentof their familiesand to associatepar-
ticularlywith other children. There will be many variationsas a function
of the structureof extended kinship grouy)sand the nature of residential
communities,but it seemsto be broadlytrue that thereis a generaltendency
to segregationof the sexes at this period. The phenomenonso familiarin
Westernsociety of the one-sex peer groupseems to have a nearly universal
counterpartto some extent elsewhere. The turning of primary interests
into the channelof relationsto friendsof the same sex and nearlythe same
age seems to have a dual significance. On the one hand it reinforcesthe
individuals self-categorization by sex by creatinga solidaritytranscending
the family between personsof the same sex. On the other hand, for the
I
first time the individualbecomesa memberof a groupwhichboth transcends
the family and in whichhe is not in the stronglyinstitutionalizedpositionof
being a memberof the inferiorgenerationclass. It is the first major step
towarddefininghimselfas clearlyindependent of the authorityand help of the
parentalgeneration.
Adolescencecomes only after a considerableperiodof this latency-level
peer group activity. Along with the fact that the emerginggenital erotic
interest of adolescenceand after involves symmetricalattractionto persons
of oppositesex, it is of the first importancethat now for the first time erotic
attractionis experiencedwith an object which is broadlyan equal, instead
of a generation-superior.On both counts there must be a considerable
reorganizationof the erotic complexin the personalityand its relationsto
the other componentsbefore mature erotic attachmentsbecome possible.
It is a psychiatriccommonplacethat much of the pathology of marriage
relationshipsand of the erotic interestsof adults otherwise,has to do with
inadequatesolution of these two problems,namely how to form a stable
attachmentto a single personof oppositesex and how to treat the partner
as fundamentallyan equal,neitherto be dependenton him or her in a childish
sense nor, by a mechanismwhichincludesreaction-formation to dependency,
to take the parentalrole and have a compulsiveneed to dominate.
When all this has taken place the circle is closed by the individual's
maxTiageand parenthood. He has had his erotic ties within the nuclear
family of orientationbroken. But he has also built up the noneroticcom-
ponentsof his personalitystructurewith the doubleconsequencesof building
a relativelysecuredam againsthis still-presentregressiveneeds,and building
a positive set of motivationalcapacitiesfor the performanceof the non-
familialroleswithoutwhichno societycouldoperate. Onlywhenthis process
has reacheda certain stage are the gates to erotic gratificationre-opened,
but this time in a greatly restructuredway and carefullycontrolled.
Finally,it mustnot be overlookedthat the eroticmotivationalcomponent
of the adult personalityis used not only to motivatethe mantal attachment,
but also constructivelyas itself an instrumentof the socializationof the next
generation. For it is clear that eroticismis fundamentallya phenomenon
of social relationships. Strongeroticmotivationis built up in the child only
becausethe mother,and to a lesserdegreethe father,enjoys reciprocalerotic
relationswith the child. But as in the case of the genital eroticismof mar-
riage, this must be controlledby strong ego and super-egostructuresin the
personality,lest the parent be unable to renouncehis own need when the
time comes.
I expressedagreementabove with the view of Fortuneand Levi-Strauss
that on the social level incest must be regardedas a regressivephenomenon,
a withdrawalfrom the functions and responsibilitieson the performance
and fulfilmentof which the transfamilialstructuresof a society rest. The
review of the role of eroticismin the developmentof the personality,which
TALCOTT PARSON S II5
I have just presented,showsa strikingparallel. Incestuouswishesconstitute

the very prototypeof regressionforthe matureperson,the pathto the reactiva-
tion of the pnmitive layers of his personalitystructure. But surely this is
more than merely a parallel. There is the most intimate causal inter-
dependence. Societiesoperateonly in and throughthe behaviourof persons,
and personalitieson the humansocio-culturallevel are only possibleas par-
ticipantsin systems of sociallyinteractivebehaviour,as these are relatedto
the needs of human organisms.
I have argued that erotic gratificationis an indispensableinstrument
of the socializationof the humanchild,of makinga personalityand a member
of societyof him. But equally,unrestrictederoticgratificationstandsdirectly
in the way, both of the maturationof the personality,and of the operation
of the society. Indispensableto certain processesof learning,it becomes
probablythe most seriousimpedimentto furtheressentialstages of maturity.
The incest taboo is a universalof human societies. I suggest that this is
becauseit constitutesa main focusof the regglationof the eroticfactor. The
institutionalizationof the familyprovidesthe organizedsettingfor the positive
utilizationof the erotic factor,both in socializationand in strengtheningthe
motivation to the assumption of familial responsibility. But the taboo in
its negative aspectis a mechanism whichpreventsthis positiveuse fromgetting
" out of hand", which ensures the self-liquidationof the particularfamily
and the productionof personalities it which are capableof fulfillingthe
by
functionsof transfamilialroles.
Admittedly,as far as originsareconcerned,this is verylargelya functional
argumentand does not solve the problemsof how incest taboos came into
being. It does, I think, serve to illuminatethe manifoldways in which the
incest taboo is involved in the functioningof any going society and gives
a basis for predictionof the probableconsequencesof variousformsof inter-
ferencewith it or modificationof it. It placesthe problemin the context of
analysisof the socialsystemin such a way as also to showthe interdependence
of social systems with the processesof the personality. Once this level of
analysis has been worked out the problem of ongins assumes a lesser
significance,but also can be approachedwith better hope of success.
There is one final importantpoint. At the beginningof this paper I
referredto the earliertendencyto attempt to find a specificexplanationof
the incest taboo, and expressedmy own belief that an analysisof the inter-
dependeneeof a numberof factorsin a systemwas muchmorepromising. A
commoncounterpartof this specificfactorview,is the demandthat an explana-
tion in some one simple formula adequatelyexplain all the variationsof
incidenceof the taboo. It seemsto me clearthat, on the basisof the analysis
I have presented,this is an illegitimateand unnecessaryrequirement. I have
emphasizedthat there is a solid commoncore of incidence,namelycentering
on the nuclearfamily. But we know that even this is brokenthroughunder
very exceptionalcircumstances,namely the brother-sistermarriageof a few
royal families. This case is not an embarrassmentfor the kind of theory
I have presented. Iior if the taboo is held to be the resultantof a balance
of forces,then it is alwayspossiblethat the balanceshouldbe alteredso as
to relax it under certain circumstances. As Fortune [I9] correctlypoints
out a better test case wouldbe the full legitimationof morganaticmarriages
in royal families i.e. as taking the place of politicallysigniIScantalliances.
Essentiallythe same holds where it is a questionof variationsof incidence
outsidethe nuclearfamily. Only a sufficientlyfull analysisof the conditions
of stability of the partic?lar socialsystem in questioncan furnishan adequate
answerto the questionof why this ratherthan a differentpattern is found
in a particularcase. But such variations,and the elementsof contingency
involved in them, (lo not alter the importanceof the massivefundamental
facts that no humansociety is knownwithoutan incest taboo, and in no case
does the taboo fail, for a society as a whole, to includeall the relationships
within the nuclearfamily. It is to the understandingof these massivefacts
that this analysis has been primarilydirected.
NOTES
I. This paper was presented at a meeting of the Association of Social Anthropologists,

University College, London, January IO, I954.
2. Soczal Structure,I949, chap. IO.
3. It will be notecl that I deliberately assume the incest taboo as part of the
constitution of the family itself.
4. See R. 11. Bales, " The Equilibrium Problem in Small Groups ", in Parsons,
Bales and Shils, TVorkingPapers irl the Theoryof Actiorl}I953.
5. The best (locumentation of this generalization available so far is I think a paper
by M. Zelditch, Jr., " Role Differentiation in the Nuclear Family ", to appear as chap. IIt
of Parsons, Bales, Zelditch ancl Olcls, Family, Socialization and Interaction Process (to be
publishecl during I954 by The Free Press). Zelditch studied a sample of fifty-five societies
and found first an overwhelming preponderance of relative instrumentalisrn in the father
role, second no cases where the available eviclence was unequivocal that the mother
role in the nuclear family is moreinstrumental than that of the father. The greatest
difficulties for this thesis occur in the cases of matrilineal kinship systems where the
mother's brother takes over some of the functions of tlle father in other systems. The
weight of Zelditch's evidence, however, suggests that even in these cases the relatizoe
differentiation on this axis holds, though the span of it is greatly narrowed.
The im?ortance of these four roles for family structure is, I think, emphasized by
kinship terminology. I believe it is true that, with all the variation of kinship ter-
minology, there is no knosvn system where these four roles, namely mother, father,
brother, sister and, conversely, self, spouse, son, daughter, are not cliscriminated from
each other. Of course frequently incumbents of these roles are classifiecl together with
other kin, as father with his brothers. But there is no known system which fails to
discriminate the four carclinal roles in the nuclear family from each other. This is to
say that generation and sex within the family are universally made bases of cliscrimina-
tion. There is no other set of roles in kinship systems of which this is true.
6. The connection between the leaclership eoalition of the small group and the
erotically bound marriage partners was first stated by Bales, " The Equilibrium Problem
in Small Groups ") in Parsons, Bales all(l Shils, op. cit., chap. I V.
7. E. R. Leach, The Strmst?wral
Igplications of lUlatriltneal Cross-Coustn Marriage,
Royal Anthropological Society, Lon(lon, I 95 I .
8. Cf. Parsons, l he Social System, chap. 11.
TALCOTT PARSONS II7
9. R. F. Fortune, " Ineest ", in Encyclopediaof the Social Sciences,edited by Selig-

man and Johnson, and Claude Levi-Strauss, Les structureselementairesde la parentd.
IO. For a reeent survey, ef. Murdoek, op. cit., ehap. X.
I I. Op. Cit.
I2. There are good reasons for believing that there is an intimate eonneetion between
the oGrereomingof the exeessive autonomy of the nuelear family and the possibility of
a eultural level of soeial development. In the first plaee sueh a group is apparently
too small to support an independent language with its minimum of extensity of general-
ization and eommunieative range. It is also probable that it is too " ingrown ", eultur-
ally rather than biologieally. One of the important eonsequenees of the ineest taboo
is to enforee the mixing of family eultures (on the distinetiveness of the eultures C1 par-
tieular households, see J. M. Roberts, Three Navaho Households,Peabody Museum
Monographs, Cambridge, Mass.). There is an analogy here to the biologieal funetions
of sexual repr( duetion. If, therefore, I may hazard an extremely tentative hypothesis
about soeio-eultural origins it would be that the earliest societyhad to be a multifamily
unit whieh enforeed an ineest taboo.
I3. The best general referenee for this aspeet of Freud's work is his TAlree
Contributions
to Sexual Theosy.
I4. Freud's views on this problem are most fully developed in the late paper,
Hemmung,Symptomund Angst: English title: The Problemof Anxiety.
I5. Meaning the internalization of cultural values cf. Parsons, The Social System,
chap. ATI.
I6. This involves what Olds calls the " law of motive growth ". See James Olds,
PsychologicalPapers it the Theory of Action, chaps. I and II, I954.
I7. Eroticism in this respect seems to be a member of a larger class of strong affective
interests. Thus the work of Solomon and Wynne on conditioned anxiety in dogs has
shown that a sufficiently acute anxiety is almost impossible to extinguish.
I8. This sociological aspect of the socialization process is much more fully analysed
in Parsons, Bales, et al., Family Socializationand Interactio Process, The Free Press,
I954, especially chaps. IV and Vl.
I 9. Op. Clt.

London School of Economics

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

London School of Economics

Uploaded by

Copyright:

Available Formats

London School of Economics

t FTER SOMETHINGlike a generation in which the attention of

casesthe role or the organizationis charaGterized by primacyof functional

protective" mother, instead of, at the propertime, refusingto reciprocate

I have just presented,showsa strikingparallel. Incestuouswishesconstitute

I. This paper was presented at a meeting of the Association of Social Anthropologists,

9. R. F. Fortune, " Ineest ", in Encyclopediaof the Social Sciences,edited by Selig-

You might also like