On A Method of Investigating The Development of Institutions Applied To Laws of Marriage

On a Method of Investigating the Development of Institutions; Applied to Laws of Marriage
and Descent
Author(s): Edward B. Tylor
Source: The Journal of the Anthropological Institute of Great Britain and Ireland, Vol. 18
(1889), pp. 245-272
Published by: Royal Anthropological Institute of Great Britain and Ireland
Stable URL: http://www.jstor.org/stable/2842423 .
Accessed: 15/06/2014 20:15
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.
Royal Anthropological Institute of Great Britain and Ireland is collaborating with JSTOR to digitize, preserve
and extend access to The Journal of the Anthropological Institute of Great Britain and Ireland.
http://www.jstor.org
This content downloaded from 195.34.78.237 on Sun, 15 Jun 2014 20:15:32 PM

All use subject to JSTOR Terms and Conditions
List of Presents. 245
From the UNIVERSITY.-Journal of the College of Science, Imperial

University,Japan. Vol. ii, Parts 2, 3.
Mittheilungenaus der MedicinischenFacultat der Kaiserlich-
JapanischenVniversitat. Band i, No. 2.
From the EDITOR.- America,nAntiquarian. Vol. x. Nos. 4, 5.
July, September,1888.
Journal of Mental Science. Nos. 110, 111.
Na,ture. Nos. 974-993.
-- Photographic Times. Nos. 353-372.
American Journal of Psychology. Vol. i. No. 1.
S- Science. Nos. 281-299.
- Timehri. No. xiii.
Revue d'Anthropologie. Tom. iii, 1888. No. 4.
Revue d'Ethnographie. Tom. vi, 1888. Nos. 1-3.
Revue Scientifique. Tom. xli; No. 26; Tom. xlii, Nos.
1-19.
Bullettino di Paletnologia Italiana. T. iv. N. 5-8.
Ethnologische Mittheilungen aus Ungarn. I Jalhrgang.
1887-88. Heft 2.
The followingpaper was read by the Author:-
On a METHOD of INVESTIGATING the DEVELOPMENT of

INSTITUTIONS; APPLIED to LAWS of MARRIAGE and DESCENT.
By EDWARD B. TYLOR,D.C.L., F.R.S.

FOR years past it has becomieevident that the great need of
anthropology is that its methodsshould be strengthenedand
systematised. The world has not been unjust to the growing
science,farfromit. Whereveranthropologists have been able
to show definiteevidence and inference,for instance,in the
developmentseries of arts in the Pitt-RiversMuseum,at Ox-
ford,not only specialistsbut the educatedworld generallyare
ready to receive the results and assimilatethem into public
opinion. Strictmethodhas, however,as yet only been intro-
duced over part of the anthropologicalfield Therehas still to
be overcomea certainnot unkindlyhesitancyon the part of
men engagedin the preciseoperationsof mathematics, physics,
chemistry, biology,to admitthat the problemsof anthropology
are amenableto scientifictreatment. It is myaim to showthat
the developmentof institutions miaybe investigatedon a basis
of tabulationand classification. For this end [ have takenup
a subjectof the utmostreal as well as theoreticalinterest,the
formation of laws of marriageand descent,as to which during
many years I have been collectingthe evidencefound among

246 E. TYLOR.-On a Methodof Investigating
theDevelopment
betweenthreeand fourhundredpeoples,rangingfrominsignifi-
cant savage hordes to great culturednations. The particular
rules have been scheduled out into tables, so as to ascertain
what may be called the " adhesions" of eaclhcustom,showing
which peoples have the same custom,and what othercustoms
accompanyit or lie apart fromit. From the recurrenceor
absence of these customsit will be our businessto infertheir
dependenceon causes actiingoverthe wholerangeof mankind.
Years since,long beforemy collectionof data approachedits
presentbulk,and could be classifiedinto the elaboratetables
now presented,I becamenaturallyanxiousto know whetherthe
labour had been thrownaway,orwhetherthis social arithmetic
would do somlething to disclose the course of social history.
The questionwas how to makethe trial. I reinembered a story
I had once heard of Horace Vernet,that a friendasked him
how he planned out his huge battle-pieces. The painter took
theinquirerinto his studioanldbegan a pictureforhim by first
touchingin a bayonetin one cornerof his canvas,thendrawing
the arm and sabre of the trooperslashingover the bayonet-
thrust,and so on fromi one overlappingfigureto the next till he
reachedthe centralgroup. It seemed to me that it would be
well to beginthusin one cornerof the field. The pointI chose
was a quaint and somewhatcomic customas to the barbaric
etiquettebetweenhusbands and theirwives' relatives,and vice
versa: theymay not look at one another,muchless speak,and
they even avoid mentioningone another'snames. Thus, in
America,John Tanner,the adopted Ojibwa,describeshis being
taken by a friendlyAssineboininto his lodge,and seeinghow
at his companion's entry the old father and motlher-in-law
coveredup theirheads in theirblanketstill theirsoni-in-law got
intothe compartment reservedfor him,where his wifebrought
him his food. So in Australia,Mr. Howitt relates how he in-
advertentlytold a native to call his mother-in-law, who was
passing at some little distance; but the blackfellowsent the
order round by a third party, saying reproachfullyto Mr.
Howitt,"You know I couldinot speak to that old woman."
Absurdas this custommay appear to Europeans,it is not the
outcomeof mere local fancy,as appears on reckoningup the
peoples practisingit in various regionsof the world, who are
foundto be about sixty-sixin number,that is, mrore than one-
sixth of the whole number of peoples catalogued,which is
roughlythree-hundred and fifty. Thus:
Avotdance.
BetweenH. and W.'s Rel. Mutual. jBetweenW. and H.'8 Rel.
45 8 13

of Institutiosn;appliedto laws of Marriageand Descent. 247
Now, on lookingout fromthe schedulesthe adhesionsof this

avoidance-custom,a relationappearsbetweenit and the customs
ofthe worldas to residenceaftermarriage. This is seen in the
followingcomputationof the peoples whose habit is for the
husbandto takeup his abode withthewife'sfamiilypermanently,
or to do so temporarilyand eventuallyto reinove withher to
his own familyor home (the reverseof this does not occur),or
forthe husbandat once to take homethe wife.
Residence.
H. to W. Removal W. to H.
65 76 141
Now, if the customsof residenceand the customsof avoid-

ance were independent,or nearlyso, we should expect to find
theircoincidencefollowingthe ordinarylaw of chance distribu-
tion. In the tribeswherethe husband permanently lives with
his wife'sfamily(sixty-five out of threehundredand fifty), we
should estimatethat ceremonialavoidance betweeinhim and
themmightappearin nine cases,whereasit actuallyappearsin
fourteencases. On the otherhand,peopleswherethe husband
at marriagetakes his wifeto his home (one hundredand forty-
one out of threehundredand fifty), would rateablycorrespond
with avoidance betweenhim and her familyin eighteencases,
whereasit actuallyappearsin niniecases only. Also,ifthethirteen
cases of avoidancebetweenthe wifeand the husband'sfamiiy
were dividedrateablyamong the different modes of residence,
two or three cases should come among the peoples where the
husbandlives withthe wife'sfamily,but thereare no such cases.
On the otherhand,fivecases shouldbe foundamong,the peoples
where the wife lives in the husband's home or family,but
actuallythereare eight. Thus thereis a well markedprepon-
deranceindicatingthat ceremonialavoidanceby the husbandof
the wife'sfamilyis in some way connectedwithhis livingwith
them; and vice versa'as to the wife and the husband'sfamily.
Hereupon,it has to be enquired whetherthe facts suggesta
reasonforthisconnexion.Such a reasonreadilypresentsitself,in-
asmuchas theceremonyof not speakingto and pretendingnotto
see some well-knownpersonclose by,is familiarenoughto our-
selves in the social ritewhichwe call " cutting." This,indeed,
with us implies aversion,and the implicationcomes out even
morestronglyin objectionto utterthe name (" we nevermen-
tion her,"as the song has it). It is different,however,in the
barbariccustomwe are considering, forherethe husbandis none
the less on friendlytermswith his wife's people because they
may not take any notice of one another. In fact,the expla-

theDevelopment
nation of this ceremonialcutting may be simpler and more
directthan in civilised Europe. As the husband has intruded
himselfamonga familywhichis not his own,and into a house
where he has no riaht,it seemsnot difficult to understandtheir
marking, the differencebetweenhim and themselvesby treating
him formallyas a stranger. So like is the workingof the
human mind in all stagesof civilisation,thatour own language
conveys in a familiar idiom the same trainiof thought; in
describingthe alreadymentionedcase of the Assiineboin marry-
ing and taking,up his abode withhis wife'sparentswho pretend
not to see him whenhe comesin, we have onlyto say thatthey
do not recogntse him,and we shall have condensedthe whole
proceedinginto a singleword. In this firstexample,it is to be
nioticedthat the argumentof a causal connexionof somekind
between two groupsof phenomenabrings into view,so far at
least as the data provesound,a scientificfact. But we pass on
to less solid groundin assigningfor this connexiona reason
which may be only analogousto the real reason,or onlyindi-
rectlycorresponding with it, or onlypartlyexpressingit, as its
correlationwith other connexionsnmay eventuallyshow. This
importantreservation, once stated,may be taken as understood
throug,h the restof the enquiry.
Let us now turn to anothercustom,not less quaint-seeming
thainthe last to the European mind. This is the practiceof
namiing, the parent fromthe child. When Moffat,the mis-
sionary,was in Africa among the Bechuaiia, he was spoken
to and of,accordingto native usage, as Ra-Mary = fatherof
Mary. On the other side of the world,among the Kasias of
India, ColonelYule mentionsthe like rule; forinstance,there
b)einga boy named Bobon,his fatherwas known as Pabobon.
In factthereare above thirtypeoplesspreadoverthe earthwho
thus name the father,and, thoughless often,the mother. They
may be called,coining a name for them,teknonymous peoples.
When beginningt to notice the wide distributionof this custom
of teknonymy, and settingmyselfto reckonits adhesions,f con-
fess to have been fairlytaken by surpriseto findit lying in
close connectiolnwith the customof the husbanid'sresidencein
the wife'sfamily,the two coincidingtwenty-twotimes,where
accidentmightfairlyhave given eleven. It provedto be still
morecloselyattachedto the practiceof ceremonialavoidanceby
the husband of the wife's relatives,occurringfourteentimes,
whereaccidentmighthave givenfour.The combinationis shown
on thediagram,fig.1, the (appproximate)numberson whichgive
the means of estimatingthe probableclosenessofcausal connec-
tion. Were the three customsso distantlyconnectedas to be
practicallyindependent, theproductofthecorresponding fractions

ofInstitutions;applied to Laws of1arriage and Descent. 249
132
S- 50 X 3 50' , multiplied
X 353 31
utpi into the threehundredand fifty
it
peopleswould show that theirconcurrencemightbe expected
to happenbetweenonce and twice in the list of peoples of the
world. In factit is found elevenitimes. Thus,we have their
commoncauLsation vouched forby the heavy odds of six to one.
Many of the firmestbeliefs of mankindrest,I fear,on a less
solid basis. In tracing,out the originof the groupof customs
in conformity withtheseconditions, it is not necessaryto invent
a hypothesis,as an account of the proceedingsot the Cree
Indians will serve as a "luminous instance" to clear up the
whole situation. Among these Indians the young husband,
comingto live with his wife's parents,must turnhis back on
Fig. 1.
\ 1q~.toa W. ,
them,not speak-ingto them (especially not to his mother-in-

law), being,thus treatedas a strangertill his firstchild is born
whereuponhe takes its name,and is called " fatherof So-and-
so," and thenceforthattaches himself to his parents-in-law
ratherthan to his own parents. That is to say, he is cere-
moniallytreatedas a strangertillhis child,beingborna member
of the family,gives him a status as fatherof a memberof the
family,whereupontheyconsistently leave offthe farceofnotre-
cognisinghim. When I broughtthisargumentto the knowledge
of Dr. G. A. Wilken,of Leyden,he pointedout to me thatin his
series of papers on " PrimitiveForms of Marriage,"'wberehe
1 Gl. A.
Wilken, "Over de primitievevormenvan het huwelijk,&c.," in
" IndischeGids," 1880, &c.

250 E. TYLOR.-On the,Development
a MelAodof Investigating
gives instancesof the namingof fathersfromchildren,he had

statedthis practiceto be an assertionof paternity.Undoubtedly
it is so on the father'spart,and its being so is quite compatible
with its being a recognitionof him by the wife'skinsfolk,the
two aspectsbelongingto one social fact.
Taking the connection between residence and ceremiionial
avoidance to be substantiatedby theirrelativeadhesions,it is
necessary to notice that there are cases where the husband,
althoughhe carriestlhewifeawayfronm the homeof herparents,
neverthelessgoes throughthe formof avoiding them. This,
ulnderthe circumstances, seems a miotiveless proceeding,but is
intelligibleas a survivalfroma timewhen he would have lived
with them. These cases belong mainlyto the Malay District
and to Australia. In the Malay Districtthe habit of residence
in the wife'sfamilyis still a notableinstitutionof the country,
thoughbeing,fastsupersededby householdingon the Arab and
Europeanmodels. In Australia,the nativecustomis described
as being thatthehusbandtakes his wifeto his own home,while
at the same time he carries out the etiquette of cuttinghis
mother-in-lawto a ludicrous extreme,with slight traces of
the avoidanceof the father-in-law.It appearedto me that on
the present explanationthis must indicate a recenthabit of
residenceon the wife'sside,and referenceshoweda law of the
Kurnai tribe of Gippsland,1that whenia native kills game,
certainparts of the meat (of a kangaru, the head, neck, and
partof the back) are the allottedshlareof thewife'sparents.As
the duty of supplyingganmeto the wife's householdwhenthe
husband lives there is one of the best-markedpoints of
matriarchallaw, I wroteto Mr.Howitt,as the leadingauthority
on Australian anthropology, suggestingthat furtherenquiry
would probablydisclose evidence hithertounnoticedas to the
maternalstageof societysubsisdingin Australia. Afterexamina-
tion made,Mr. Howittreplied:-" I ant now satisfiedthatyour
surmiiisesare quite correct,"and therewithhe sent detailsbear-
ing on the quiestion, especially an account by Mr. Aldridge,of
Maryborough, Queensland,as to the practiceof the tribesin his
neighbourhood.This I will quote, as being a stronglymarked
case ofresidenceon the wife's side. " When a man marriesa
womanfroma distantlocality,he goestohert-ribelet andideiltifies
himselfwithherpeople. This is a rule withveryfewexceptions.
Of course,I speak of themas theywerein theirwild state. He
becomnes part of and oineof the family. In the event of a war
expedition,the daughter'shusbandacts as a blood-relation, and
will fightand kill his own blood-relations if blows are struckby
1 Fison and Howitt," ]Kamilaroiand Xurnai," p. 207.

and lDescent. 251
of Institutions;appltedto Laws of Mfarriage
his wife'srelations. I have seen a fatherand son fighting

under
these circunmstainces, and the son would most certainlyhave
killed his fatherif othershad not interfered."
The relativepositionsof the two groupsof customs,residence
and avoidance,nmay now be morecompletelyshown,by the aid
of the diag,ranm,
fig.2.
Fig. 2.
RE-SIDENCE, AVOtDANCEW.
W. +o Hl
14-1) /9) _ ()
Remove H4&YW.FAM.W.H. FAM.
(65)
Here the space representingresidenceis divided into three

sections, viz.,residenceon the wife'sside; the transitionalstage
of reinoval (where the couple begin marriedlife in the wife's
house,but eventuallymnove);residenceon the husband's side.
Accordingto the previouS,arguments, the ceremonialavoidance
between the husband and the wife's familyis takenito have
arisenwithinthe periodswhen he and theylived permanently
or temporarilyin contact,and to have continuedby survival

theDevelopmitent
252 E. TYLOR.-On a MethodofInvestigating
intothe period afterthis co-residencehad ceased. Therenext

appear the small group of eight cases of mutual avoidance,at
once betweenthe husband and the wife's fami]y,and the wife
and the husband's family. These consistently are foundin the
removalstage,wherebothkindsofresidencemeet,survivinginto
the stage of residenceon the husband'sside. Avoidance be-
tween the wife and the husband'sfamilyhas the same range,
but here the conditionsproducingit belong to both stagesof
residence,and thereis no questionof survival.
From this distributionof the avoidance-customs, it appears
thatin the partsoftheworldopen to the presentinspection,the
threestagesof residencehave tended to succeedone anotherin
the upwardorderof the diagram. Residenceon the wife'sside
appears earliest,afterthis the removal stage, and latest,re-
sidenceon the husband'sside. For if it be supposedthat the
courseofsocietywas in the reversedirection, as would be repre-
sentedby turninig the diagramupside down,avoidancebetween
the husbandand thewife'sfamilywouldbe represented as arising
in the stagewhen the husbandlived away fromit, whileavoid-
ance betweenthe wite and the husband'sfamily,whichought
on this suppositionto contiriueby survival into the stage of
residenceon the wife'sside,is not foundthere. The avoidance-
customs,thoughpracticallyso triflingy, are thus signalsshowing
thedirectionof a movement, of whichwe shall morefullysee the
importance, of habitualresidencefromthe
namely,the shiftilng
wife'sfamilyto the husband's.
Let us now proceedto apply a similarmethodto the investi-
gation of the great division of society into matriarchaland
patriarchal. In the matriarchalsystem,descentin the family
or clan is reckonedfromthe mother; authorityis mainly on
her side, the nmother'sbrotherbeing habituallyguardianof the
children; successionto rankand office, and inheritanceof pro-
perty,followthe same line passing to the brotheror to the
sister's son. In the patriarchalsystem descent is fromthe
father; he has the power over wife and children; succession
and inheritanceare fromhim to his offspring.Between these
extremestageslies an intermediate or transitionalstagein which
theircharacteristics are variouslycombined. The termspatri-
archal and mnatriarchal not beingquite appropriate,I shall use
in preference forthe three stagesthe termsmaternal,maternal-
paternal,and paternal. The classification is necessarilysome-
what vague,but I think will be found to have sufficient pre-
cision for the problemof determining the directionin which
mankindhas tendedto move fiomone of the stagesto another.
In dealing,withthisproblemcertaincustomsrelatingto marriage
law will be used as indicators.

of Institutions;appliedto LawvsofMiarriageand Descent. 253
Among a large proportionof the nationsof the worldup to

the middle levels of culture, the re-marriageof widows is
arranged,and more or less enforced,but the regulationsare
framedon two distinctprinciples. On the firstprinciplethe
widowbecomesthe wifeof her husbaind'sbrother, or near kins-
man,accordingto somne recogniizedorderof precedenceof claim.
The word"levirate,"fromlevir= husband'sbrother, has become
the acceDted term for this iinstitution,but its sense must in
most cases be extendedto take in a series of kinsmen,among
whom the brother-in-law only ranks first. Unfortunately, it
has seldom been thoug,htworthwhile to ascertainthisprecise
order,which mightthrowlight on familystructure,as in an
accountdrawn up by Mr. Howitt of the practicein Australian
tribeswhere any man is eligible to succeed to the widow,if he
standsin therelationof elderoryoungerbrotherto thedeceased,
beginningwith actual brotherson the male or femaleside, ac-
cordingto the rule of descent in the tribe,and extendingto
tribalbrotherswho are in our terminology cousins,mioreor less
near. The levirate appears in its various formsamong one
hundredand twentypeoplesin inylist,or about one in threein
the world. On taking out its adhesions it seems sufficiently
accountedfor as a custom of substitution,belonging to the
periodwhen marriageis a compactnot so much betweentwo
individualsas betweentwo families, oftennmadewhenthecouple
are infantsunable to understandXit,in fact sometimesbefore
theirbirth. That the levirateformspart of this familytrans-
actionis consistentwith othercustomsmore or less associated
with it, viz.,that when a wife dies or turns out ill her family
are bound to replace her by another,a rule which sometimes
even holds forbetrothal,and that the widowis not allowed to
marryout of her husband'sfamilyunless by leave of his kins-
men,who have the choice of keepingher, or partingwith her,
usually for a price. The social distributionof the levirate is
shownin fig.3 to extendthroughall threesocial stages. It is
in the maternal-paternal stage that it comes into competition
with the second principle,uriknownin the maternalstage,in
which the father'swidows pass by inheritanceto his sons,
especiallythe eldest son taking,his stepmothers.A smiallbut
importantgroupof cases formsa bridgebetweenthe two princi-
ples oflevirateand filialsuccession,combining bothin thesaimie
nation. This combinatioiiis well shown in Africa,where on a
chief'sdeath the head wifewill pass by levirateto his brother,
while her soii, the new chief,will inlherita crowd of step-
mothers,a less onerousleracy indeed than may seem,as they
are practicallyslaves who hoe and grind corn for their owIn
living. Lookingat the distribution of these groupsof customs,
VOL. XVIII. T

254 E. TYLOR.-On a Methodof InvestigatingtheDevelopment
it is seen to be only compatiblewiththe view thatthe paternal

rule followedthe maternal,bringingwithit even while its pre-
valence was but partial,the principleof paternalwidow-inheri-
tance.
The quaint customof the couvade has now to be considered
Fig. 3.
PAT.
Isot/
LEVI RATE.
XA1AT.- PAT
\ / M~~MAT.
fromthe same pointof view. In this the father,on the birth

of his child,makes a ceremonialpretenceof being the mother,
being nursedand taken care of,and performing otherritessuch
as fastingand abstainingfromcertainkinds of foodor occupa-
tion,lest the new-bornshould sufferthereby. This customis
known in the fourquiartersof the globe. How sincerelyit is
still accepted appears in a storyof Mr. Im Thurn,who on a

of Institutions;appliedto Laws of Mar-riage
and Descent. 255
forestjourneyin BritishGuiana noticedthat one of his Indians

refusedto helpto haul the canoes,and on enquiryfoundthatthe
man's objectionwas that a child musthave been bornto him at
homeabout this time,and he must not exert himselfso as to
hurt the infant. In the Mediterraneandistrictit is not only
mentionedby ancient writers,but in Spain and France,in or
near the Basque country,it went on into modern times;
Zamacola,in 1818, mentions,as but a little time ago,that the
motherused to get up and the fathertake the child to bed.
Knowingtheteniacity of thesecustoms,I shouldnot be surprised
Fig. 4.
{
PAT.
(8)
COUVADE (20)
M AT.-PAT. h
MAT.
if tracesof couvademightbe foundin thatdistrictstill. Now

examiining of the couvade by the diagram,Fi.g.
the distribution
4, we see that this farcical proceedingdoes not appear in the
maternalstage,but arising in the miaternal-paternal, at olice
takes its strongestdevelopmentoftwentycases; in the paternal
the nurnberfallsto eightcases,leadino to theinferencethathere
it is onlykept up in dwindlingsurvival.
Lookingat this position,I inust now argue that the original
initerpretationof the couvade given by Bachofenin his great
T 2

theDeveloppment
treatise'in 1861, and supportedby Giraud-Teulon, fitssubstan-

tiallywith the facts,and is justifiedby them. He takes it to
belongto theturning-point of societywhenthetie ofparentage,
till then recognisedin maternity, was extended to take in
paternity,this being done by the fictionof representingthe
fatheras a second mother. He comparesthe couvadewith the
symbolicpretencesof birthwhichin the classical world were
performed as ritesof adoption. To his significant
examplesmay
be added the fact that amongcertaintribesthe couvadeis the
legal formby whichthe fatherrecognizesa child as his. Thus
this apparentlyabsurd customa, wlhichfor twentycenturieshas
been the laughing-stockof mankind,proves to be not merely
incidentallyan indicatorofthetendencyofsocietyfrommaternial
to paternal,but the verysign and recordof thatvast change.
The distributionof customsin figs.3 and 4 is only com-
patiblewitha tendencyof societyto pass fromthe maternalto
the paternal systems,the maternal being placed as earliest
from the absence of survivals fromother stages extending
initoit, as theyfreelydo into the paternal,which is therefore
placed as latest. The argumentis a geological one. Just as
the formsof life,and even theactual fossilsoftheCarboniferonis
formation, may be traced on into the Permian,but Permian
typesand fossilsare absenitfromr theCarboniferous strataformed
beforetheycame into existence,so herewidow-inheritance and
couvade,which,if the maternalsystembad been laterthanthe
paternal,would have lasted on into it, prove by their absence
the priorityof the maternal. Thus the presentmethodcon-
firmson an enlargedand firmbasis the inferenceas to the
antiquityof the maternalsystemarrivedat by the pioneersof
the investigation, Bachofenand McLennan,and supportedby
the later researchof a generationof able investigators-Morgani,
Lubbock,Bastian,Giraud-Teulon,Fison, Howitt,Wilken,Post,
Lippert,and othiers. By thisit is not,however,meantto imply
thatthe maternalformof familyas hereset forthrepresentsthe
priinitiveconditionof mankind,but that it is a stage through
J. J. Bachofen, "Das Mutterrecht," pp. 1-7,255; Giraud-Teulon,"Les

Originesdu Marriage,"p. 138. In my accountofthe conLvade, " Early History
of Mankind,"Chap. x, I have laid stresson the magical-sympathetic nature of
a largeclass of couvaderitesas implyinga physicalbond betweenparent and
child; thus an Abiponewould not take snufflest his sneezingmighthurthis
newbornbaby,and a Caribfathermnust abstainfromeatingsea-cowlest his infant
should getlittlerouindeyeslike it. This motive,whichis explicityor implicitly
recognisedby the savagesthemselves, certainlyformspartof the explanationof
the couvadle. It is, however,secorndarv,beingdue to the connexionconsidered
as subsistingbetweenparentand child,so thattlhesesympatheticprohibitions may
be interpreted as originally
practisedbythe motheronly,and aiterwardsadopted
by the fatheralso.

of Institutions;appliedto Laws of Marriageand -Descent.257
which the inhabitantsof a greatpartof the world now in the

paternal appear to have passed, and which still continuesin
forceoverconsiderabletractsof everypart of the globe except
Europe. It seeinsprobablethatthis maternialsystemarose out
of an earlier and less organisedand regulatedconditionof
humanlife. As to this problem,however,thoughthe present
schedulesare not devoidof information,I have not been able
Fig. 5.
H.
RESIDEE
(66), ,i U,, , .
to brinctthe generalevidenceinto shape sufficientlyto justify

my offering a theoryhere.
The analogyhas alreadycomeinto view betweenthe division
of societhaccordingto residence,and accordingto the maternal
and paternal systems. This relation,the realityof which is
evident frommere consideratioln of the difference
as to family
life which must ensue fromthe husband living,in the wife's
house or the wife living in the husband's,may be corroborated

theDevelopment
fromthe schedules. Thus the numberof coincidencesbetween

peoples where the husband lives with the wife's familyand
where the maternalsystemprevails,is naturallylarge in pro-
portion,whilethefullmaternalsystemas naturallyneverappears
among peoples whose exclusive custom is for the husband to
take his wifeto his own home. But as I have pointedout,the
-naternaland paternalsystemsare inoteach a definite institution,
but combinations in which more or less strictlythe authority,
descent,succession,inheritancefollowthe femaleor the male
side. The imperfection of my schedules makes it desirablefor
me to postponean attemptto work out numericallythe intri-
cate problernof the mutual relationsof these social rules till
more perfectdata are accessible. I have made, however,a
rough sketch illustrativeof the hypothesissuggestedby the
diagramisfigs.3 and 4, namelythat in the olie simple fact of
residencewe may seek the main determiningcause of the
several usages which combineto forma maternalor paternal
system. This sketch,fig. 5, is meant to suggest the social
movementwhichthe schedulesseem to imply. Division accord-
ing to residenceon the female or male side is taken as the
fundamentalfact,and the lines show the institutionsof female
descent,avunculai authority,&c., arisingin the stageofresidence
on the femaleside, and extendinginto the stages of removal
and residence on the male side. Within these two latter
stagesit is thatmale descent,paternalauthority, &c., arise and
extendonwardin history. This directionis indeed consistent
with what our own knowledgeof human naturewould lead us
to expect. We can well understandhow when the man lives
in his wife's familyhis powerwill count for little against the
combinedauthority of hermaternalunclesand brothers, whereas
when he takes her to his own home,be is apt to becomemaster
of the household; and we should expect the rules of descent,
succession,and inheritanceto followthe same order. Actual
record of such transitionis very rare, but at any rate one
observer,the Hon. J. W. Powell, of the Bureau of Ethnologyat
Washington,has had boththe opportunitvto see and the skill
to see whathe was seeing,withthe resultof convincinghimself
that the transitionfrommaternalto paternal societyhas in
great measuredependedon residence. I quote a passage of a
letterfromhim:-" It would seem fromsuch opportunities as
I have had to collectfactsin the fieldthat huntingand other
partiesare frequenltly organisedin such a mannerthatthe male
meimbers of a clan groupproceedtogetherin companywiththeir
wivesand children. Under suchcircumstances thecontrolof the
familynecessarilyfalls into the hands of the husbands and
fathers."This happensamongthe Pueblo Indians,a matriarchal

of lnstitutions;appliedtoLaws ofMiarriageand Descent. 259
people withfemaledescent,whose clans, in consequenceof the

scarcityof waterforirrigationin theirdesertregion,are obliged
to separatewidelyforthe cultivationoflands at a distancefroin
the centralPueblo. The result is that the controlof families
and the trainingof childrenare temporarily taken out of the
lhandsof theirown kin on the mother'sside,and withtheacqui-
sitionof cattlein thesenew homiescomesthe tendencyto settle
therepernmanently. Observationof thesefactsled Major Powell
to adoptthe hypothesisthat clanship by femaledescentpassed
in this way into clanship by male descent by the segregation
of clans forindustrialpurposes.
The next diagram,fig. 6, throwsmore light on the great
social transformation.It showsthe distribution of the practice
Fig. 6.
I III' CAPTURE
PAT
J 0
MVATh-
PAT.
21
21
6 ~~~M
AT.1

260 F. TYLOR.-On a Methodc
ofInvestigating
theDevelopment
ofmarriageby capture. When the accountsofnationalcustom

are classifiedtheyshowthat captuire(whichbelongsto overone
hundredofthe peoplesscheduled)can be moreor less accurately
divided into three kinds:-Hostile capture,when warriorsof
one tribe bring away as captives women of another tribe is
a featof arms praised in historyshortof the highestlevels of
culture. There were fierceIndians of the Pampas who held
that their god, the Great Eagle, told thenm to live by making
war on all othertribes,slaying theirmen and carrying, offtheir
womenand children. The same spiritis heardin the hopes of
Sisera's host to divide the spoil,to everyman a damselor two.
Looking at hostile capture fromthe anthropologicalpoint of
view,we have to noticethatit exists equally throughthe three
stagesof society,frommaternalto paternal. Now it obviously
conflictswith full matriarchalinstitutionsthat a man should
bringin a captive wife,for he cannot take her home to his
mother-in-law.To understandsuch a customappearingwithin
the rangeof matriarchy at all,we mustrememberthata captive
has no rights,so thatwhathappensto her does not immediately
affectthe regular customof the tribe,whichapplies to native
free women. Yet even here the tendencyof capture must
always have been to upset the maternalarrangemenits.When
capturecomesto be an accepted mode of marriagebetweenor
among tribesor clans who live at peace and habituallyinter-
marry,it is evident that such " connubial capture,"as it is
describedon the diagram,can only consist with the paternal
system,inasmuchas the husband necessarilycarries the wife
to his own home,therebysettingon foota paternalhousehold.
This is true also of the cases where the capture has become
a merelyformalceremiiony, a marriagesettled
accomiipailying
beforehand,for the very formof captureinvolves the bride-
groomcomingwith his friendsto carrythe brideto his homle.
This is the interpretation of the fact, made evident in the
diagram,that connubialand formalcapturebelong onlyto the
intermediatestagewhere paternalinstitutionsare arising,and
to the later stage wheretheyare fullyestablished. The effect
of capturein breakingup the maternalsystem,and substituting
the paternalforit,has thusto be takenintoaccountas a serious
factorin social development. There is at least one regionof
the world where the operationmay be seen going on at this
day-the Malay Islands. To quote the concise description
by Riedel of the matrimonialarrangementsof the Babar
Archipelago:-" The men followthe women,and live in their
houses. The childrenalso belong to the wife'sfamily. If a
man is rich enoughhe may marryseven wives,who all remain
in the housesof theirparents. A man who has manywives is

ofInstitutions;appliedtoLaws ofMarriageand Descent. 261
respected. The robberyof a wife fromanotherclan (negari)

is an honour,and the childrenfollowthe father,withor without
paymentof the fine attachedto the deed. Smalleror weakler
clans even demand no fine."l In the Kisar and Wetar island
groupsa like stateof thingsappears,the maternalsystembeing
the recognisedrule,but always liable to pass into the paternal
system by capture,which brings wife and childreninto the
husband'shands.
At this pointit will be convenientto examinetwo institu-
tionsof earlymarriagelaw, namely,exogamyand classificatory
relationship. The principleof exogamywas broughtpromi-
nentlyintoview fiftyyearsago, by Sir George Grey,2whenhe
describedthe native Australianrule fora man notto miarry a
womanof the same familyname or bearingthe same animal-
crest or kobong as himself; and called attentionto the coin-
cidence of this with the North American systemii of clanis
namedfromtotemanimals,a man beingboundto marryoutside
his own totemor clan. Mr. J. F. McLennan3 gave thesecus-
toms the name of exogamyor "marrying-out,"and showed
themto belong to " a mostwidelyprevailinig principleof mar-
riagelaw amongprimitiveraces." Much information has since
then come in, with the result of showing that exogamyhas
hardlyto do with the capture of wives in war between alien
nations, but ratherwith the regulationof marriageswithin
groupsof clans or tribeswho have connubium; such clans or
tribesmay be more or less at strife,but theyacknowledgeties
of kindredand are usually allied by language. It is now also
understoodthat a people may at once practiceendogamyor
" withinits borders,and exogamyor " marrying-
' marrying-in
out" of its clans with one another. The situationmay be
understoodamongthe Hindus,where a man must marryin his
caste,but withinthat caste mustnot marryin his own gotraor
clan. The effectof an exogamicrule is similarwhetherclan-
ship followsthe femaleor male line of descent. Next,as to the
principleof classificatoryrelationship,
an earlymentionof this
is by Father Lafitau,4above one hundredand fiftyyears ago,
who states that " among the Iroquois and Hurons all the
children of a cabin regardall their miother'ssistersas their
mothers,and all theirmother'sbrothersas theiruncles,and for
the same reason they give the name of fathersto all their
I Riedel, "De Sluik- en XroesharigeRassen tussehenSelebes en Papua,"
p. 351; see 415, 448.
2 Grey,"Journalsof Two Expeditionsin N.W. and W. Australia,"Vol. II,
p. 225.
J. F. McLennan," PrimitiveMarriage,"pp. 48, 130.
4 Lafitau," Mceursdes SauvagesAmeriquains,"Paris, 1724, Vol. I, p. 552.

theDevelopm7ent
father'sbrothers,and aunts to all theirfather'ssisters. All the

childrenon the side of the motherand her sisters,and of the
fatherand his brothers, regardeach othermutuially as brothers
and sisters,but as regards the children of their uncles and
aunts,thatis,oftheirmother'sbrothersand father'ssisters,they
only treat them on the footingof cousins... . In the third
generationthis changes,the great uncles and great aunts be-
comeagaingrandfathers and grandmothers ofthechildrenofthose
whom they called nephewsand nieces. This continuesalways
in the descendingline accordingto the same rule." In ourown
time,Lewis H. Morgan,livingamongthe Iroquois as an adopted
Indian, was str-Lck with this systemof relationships, so unlike
what he had been broughtup among,and which he .at first
thoughtto be a peculiarinventionofhis Iroquois. But finding,
on enquiry,thatit extendedto otherNorthAmericantribes,he
eventuallyby circulatinginterrogatories succeeded in collecting
a greatseriesof systemsofrelationship, in whichhe established
the wide prevalenceof classificatory as he called them
system-s,
fromthe relativesbeing groupedin classes.' Underthe term
classificatorysystems,Mr. Morgan inciuded not onlythoseap-
proximating to the Iroquois type,but a muchsimplerand ruder
plan prevalentin Polynesia;it is,however,convenientforme to
confinemyremarkshereto the former grouponly. This system,
as foundainongtheAmericanIndians,Mr.Morganshowedto be
closelyanalogousto that of the Dravidian nations of Southern
Hindustan. This latteris a well-knownsourceof perplexityto
a newlyappointedEnglish civilian,whomaybe told by a witness
that his fatherwas sittingin the house,but presentlythe same
witness mentionshis fatheras comingin fromthe field; the
native is sharplyreprovedbv the judge for contradicting him-
self,whereuponhe explains,it was my " littlefather,"by which
he meanshis father'syoungerbrother.
I am placing tooetherthe two institutions,exogamy and
classificatoryrelationship,inasmuch as they are really con-
nected, being in fact two sides of one institution. This was
made out eight years ago, by the Rev. LorimerFison, in the
work on the Kamilaroiand Kurnai tribes of Australia by him
and Mr. Howitt.2 This importantexplanationis still scarcely
known to anthropologists, nor indeed, have I much right to
reproach others with neglectingit, for I reviewedFison and
Howitt's book withoutdistinctlyrealisingthe bearingof this
argumenton the theoryof exogamy,whichonly came roundto
1 L. H. Morgan, "Systems of Consanguinityand Affinity of the Human
Famlily" (SmithsonianContribaitions,
1871).
2 Fison and Howitt," Kamilaroi and Xurnai," 1880, p. 76.

of Institutions;appliedto Laws of Miarr-iage
and Descent. 263
me latelyin a way which I had betternow describe,as it will

enable me to explain shortlyand plainlythewholeproblem. In
tabulatingthe nationsof the world,I founda groupof twenty-
one peoples whose customas to the marriageof firstcousins
seemedremarkable;it is that the childrenof two brothersmay
not marry,nor the childrenof two sisters,but the child of the
brothermay marrythe child of the sister. It seemedobvious
that this " cross-cousinmarriage,"as it may be called,must be
the direct result of the simplestformof exogamy,where a
populationis dividedintotwoclassesorsections, withthelaw that
a imian who belongsto Class A can onlytake a wifeof Class B.
Such a division,forinstance,is farnilarin Melanesia. Dr. R. H.
Codringtondescribesit in the.Banks Islands, wherethe natives
have two families,called veve inother,which implies that
descentfollowsthe mriother'sside,and a man mustmarrya wife
of theothermotherfromhimself,or as theysay,not on his own
side of the house but on the other. Thus,takingA, a, B, b,as
inales and femalesof the class A and B, anid bearingin mind
that the mother's childrenare of her class, but the father's
childrenof the oppositeclass to his, we have:-
Fig. 7.
Two sisters,a, a, Two brothers,A, A, Brotherand sister,A, a,

their . their their
Children, A, a, Children B, b, Children B, a,
are of are of are of
same class - tribal same class = tribal differentclass - tribal
brotherand sister brotherand sister cousins
= unmarriageable. -unmarriageable. -marriageable.
Having cometo thispoint,it seemed to me that I had seen

somethinglike it elsewhere,and on lookingback to " Kamilaroi
and Kurnai" I foundthatFison had thusworkedout the origin
of the Turanianclassificatorysystern,as Morgan calls that in-
cludino the above-mentionedsysternsof North America and
India, withothers. Fig. 8 puts conciselythe main featuresof
the argumentas to a man's kin.
Fig. 8.
His
father'sbrother'schild
or is (tribal)brother
or sister,
mother'ssister'schild J
Therefore

264 E. TYLOR.-On a M1ethod
ofInvestigating
theDevelopmnent
father'sbrotheris (tribal) father,

mother'ssisteris (tribal) mother,
His
father'ssister'schild )
or Sis tribal(cousin).
mother'sbrother'schildJ
Therefore
father'ssisteris (tribal)aunt,
mother'sbrotheris (tribal)uncle.
Though not proposingto enter fully into the deductionof
classificatoryrelationshipsin all theirvarietiesfromthe rule of
exogamy,it is niecessary to pointout thatthe formof exogamy
here contemplatedis the simplest or dual form,in which a
people is divided into two interinarrying classes. Systemsof
exogamywhich are dual in their nature,that is, consistingof
two classes or groupsof classes,standin directconnectionwith
cross-cousinmarriageand classificatory relationship. But if the
numberof exogamicdivisionsis not dual, if there are for in-
stance three clans,and a man of one clan may take a wife of
eitherof theothertwoclans,it is readilyseen thatthe argument
of fig.7 breaks dowln. Althoughat preseiltonly preparedto
deal with exogamyand classificatory relationshipin their dual
form,Ii may notice that the treatmentof the problemby the
methodof adhesionsstrengthenis the view,not wantingin other
evidence,thatthedual formi of exogamymay be consideredthe
originalform. In reckoningfromthe presentschedules the
numberofpeopleswhouse relationship namesmoreorless corre-
spondingto the classificatory systemshereconsidered,theyare
foundto be fifty-three,and the estimatednumberof thesewhich
mightcoincide accidentallywith exogamywere there no close
connexionbetweenthem,would be about twelve. But in fact
the numberofpeopleswhohave bothexogamyand classification
is thirty-three,thisstrongcoincidencebeingthe measureof the
close causal connexionsubsistingbetweenthe two institutions.
The adherenceis even strongeras to cross-cousinmarriage,of
which twenty-onecases appear in the schedules,no less than
fifteenof the peoples practisingit being also knownas exoga-
mous. Here,indeed,the relationis not one of derivation, but of
identity,the cross-cousinrule being actuallya partial formor
imperfectstatemenlt of the law of exogamyitself. Such ad-
hesionsbetweentwo or morecustomshave been alreadyrecog-
nised as provingthe existenceof causal connexion,but it has
now to be pointed out that theyserve anotherpurpose. The

of Irnstitutions; to Laws of Jliarriageand Descent. 265
aopplied
conrexion,whenproved,reactson the evidenceby whichit was

proved. When onceit has beenshownthatcross-cousin marriage
is part and parcel of exogamy,it may be arg,uedthat all the
twenty-one peoplespractisingcross-cousin marriageare to be set
downas exogamous. Now as onlyfifteen ofthemare expressly
recordedto be so, the list of exogamousnationsof the worldhas
to be increasedby six. So, classificatory beingevi-
relatiolnship
dence thatthe peoplespractisingit are or have beenexogamous,
this will add some twentymore to the list of nations among
whom furtherinvestigationwill probablydisclose recordthat
exogamicsocietyonce prevailedor still prevails. Even if no
directrecordis forthcoming, the indirectproofmay with due
cautionbe sufficient forplacingthemri in the exogamousgroup,
whichmay thus numberabove one hundredpeoplesout ofthe
threehundredand fiftyofthe world. Those whoremember the
sharpdiscussionbetweenMcLelnnanand Morganyearsago, and
the view thatthe classificatory relationships werea meresystem
of addresses,will be struckwith the way in which the contro-
versyis likelyto end. For myselfI hardly know whetherI
feelnmore glad or sorrythatmy old friendMcLennanto the day
of his deathneverknewthat Morganand he,whobelievedthem-
selves adversaries, wereall the whileallies pushingforwardthe
same doctrinefromdifferent sides.
It thus appears that the numberof nations who have the
systemof intermiarrving clans is largerthan has been known.
But even this by no means measures the full importanceof
exogamyas a factorin the constitution of society. Anthropolo-
gistshave longhad beforethemtheproblemof determining how
far clan-exogamymay have been the originof the prohibited
degreesin matrimoony so variouslydefinedin the laws of nations.
The yet largerprobleinhas been opened,how far laws of per-
missionand prohibitionof marriagemay have led nations to
definerelationshipsand give themnames,distinguishing forin-
stanceunclesfromfathers, and cousinsfrombrothers. It may,
I think,conduceto the solutionoftheseproblemsto notice two
ways in which the collationof the presenttables bears on the
meaningand originof exogamy.
Thereare conditionsofsocietyunderwhichexogamyis found
side by side withwife-capture, so thata barbaricmarriageoften
iuvolvesboth in one and the samreact, as when a Tatar and a
partyof his friends,all armed to the teeth,ride offto the tents
of a distantclan,and thence, withsimulatedor evenreal violence
carryoffa bride. But on reckoniino up the peoples amongwhonm
this combinationof captureand exogamxly is found,the number,
thoughenoughto show that theyco-existfreely,falls short of
whatwouldjustifythe inferencethattheyare cause and effect.

266 E. TYLOR.-On a IMethod theDevelopnmrnt
ofImvestiogating
Moreover,it appears thatthis co-existencebelongsespeciallyto

the paternalstage of society,and to the maternal-paternal, in
whichpaternalinifluence is partlyestablished. This is intelligible
enough fromwhat has been already said as to the effectof
capturein setting on foot paternalinstitutions, fromits very
outset,by bringingthe wifeintothe husband'shands and home.
We are thus led to a more fundamental test of the positionof
exogamy,by enquiringwhetherit existedin thatearliestknown
stage of the maternalsystemof society,wherethe husbandlives
in the wife's family. The schedules show that there are in
differentparts of the world twelve or thirteenwell-marked
exogamouspeoples whose habitof residenceis forthe husband
to join the wife'sfamily. This state of thingsseems to me to
preventour regardingexogamyas a resultof capture,it being
plain that the warriorwho has carrieda wife captive froma
hostile tribe does not take up his abode in her family. If
captureleads to any formof exogamy,this must,I think,be a
paternalform,and if it be admittedthat the maternalformis
earlier,then it follows that capture is inadinissibleas the
primarycause of exogamy.
More thantwentyyearsago,in compilinga list of nationsprac-
tisinigthis customof marryingout of the tribeor kin,I noticed
thatin any fulldiscussionof the subjectwould have to be con-
sidered the wish to bind different tribestogetherin friendship
by intermarriage.2 Compilingthe presenttables has broughtto-
gether observationsto this effect. Morgan, describinghow
the alliance of the Iroquois tribes,made up of intermarrying
clans,formeda bond of union throughoutthe nationalleague,
writes: " It was the boast of the Iroquois that the greatobject
of their confederacy was peace; to breakup the spiritof per-
petual warfare, whichhad wastedthe red race fromage to age."
Alnothergroup of North Americantribes,the Tinneh,on the
Arcticcircle,are dividedinto threecastes,theirrule beingthat,
forinstance,a Chit-sanghmay not marrya Chit-sangh. When
this does take place,the personsare ridiculedand laughed at,
the man is said to have marriedhis sister,even thoughshe mray
be fromanothertribe,and therebe not the slightestconnection
by blood between them. Hardisty,wlhogives these details,
remarks:-" One good thingproceededfromthe above arrange-
ment,it preventedwar betweentwo tribeswho were naturally
hostile."'4 The Bogos of Abyssinia are exogamous, and of
I Kasia, Garo, Menangkabauand Padang, Banks Islands, Mortlock Islands,

Delaware,Iroquois,Mandan and Minitari,Moqui, Tlinkit,Arawak.
Cbirolki,
2 " Early Historyof Mankind,"p. 286.
3 Morgan, " League of the Iroquois," p. 91.
4 " SmithsonianReport,"1866, p. 315.

of Institutions;appliedto Laws of Marriageand Descent. 267
them Munzingerreportsthat they are closely bound together

by reciprocalmarriages,"so that internalwar is almost im-
possible. Blood-quarrelsamongthe Bogos are always settled
very quickly,whilst the smallest collisionwith the adjoining
tribes leads to everlastingwars."' Du Chaillu writes of
Ashango-land, " tribes and clans intermarrywith each other
and this brings about a friendlyfeelilngamong the people.
People of the same clan cannot intermarry with each other."2
Thus,it seemsthatwhenPlutarchasks in the"Roman Questions,"
" Why do theynot marrywomen near of kin ? " he has some
reason in settingdownas one possibleanswer," Whetherfrom
theirwishingto increasefriendships by marriages,and to acquire
manykinsfolk,givingwives to othersand receiving(wives)from
them."3
On lookingat the distinction betweenendogamyand exogamy
fromthis pointofview,it will be seen that thereis a periodin
the growthof societywhen it is a politicalquestionof the first
importance. While the vast forestor prairiestill affords abun-
dant food fora scantypopulation,small hordesmay wander,or
groupsof householdsmay be set up, each little tribeor settle-
mentcut offfromthe rest,and marrying withinits own border.
But when tribesbegin to adjoin and presson one anotherand
quarrel,thenthe difference betweenmarrying-in and marrying-
out becomespatent. Endogamyis a policyof isolation,cutting
offa horde or village,even fromthe parent-stockwhence it
separated,if onlya generationor two back. Among tribesof
low culturethere is buitone means knownof keepingup per-
manent alliance, and that means is intermarriage.Exogamy,
enabling a growingtribe to keep itself compactby constant
unionsbetweenits spreadingclans,enablesit to overmatchany
number of small intermarrying groups,isolatediand helpless.
Again and again in the world's history,savage tribes must
have had plainlybeforetheir minds the simple practicalalter-
native betweenmarrying-out and being killed out. Even far
on in culture,the political value of intermarriage remains.
"'Matrimonialalliances increase friendshipmore than aught
else,"is a maxim of Mohammed. "Then will we give our
daughtersuntoyou,and we will take yourdaughtersto us, and
we will dwell with you,and we will become one people,"is a
well knownpassage of Israelitehistory.
Exoganrylies farback in the historyof man,and perhapsno
observerhas ever seen it comeinto existence,nor have the pre-
cise conditionsof its originyetbeen clearlyinferred. Even the
I Munzinger," Sittenund Recht der Bogos," p. 10.
2 Du Chaillu,"Journeyto Ashango-land,"p. 427.
3 " Plutarch,Queest.Rom.," cviii.

268 E. TYLOR.-On a MethodofInvestigating
theDevelopment
historicalrelationbetween exoganmy and the systemof classes

known as totemismis not fullycleared up; whetheras Prof.
RobertsonSmith takes it,' totemismsupplied the necessary
inachineryforworkinga law of exogamy,or whetherexogamy
itselfled to toternism.But as to the law of exogamyitself,the
evidenceshows it in operationover a greatpartof the human
race as a factorof politicalprosperity. It cannotbe claimedas
absolutelypreventingstrifeand bloodshed,indeed,it has been
remarkedof somepeoples,such as the Khonds and the Banks
Islanders,that the intermarryingt clans do neverthelessquarrel
and fight. Still by bindingtogethera whole community with
ties of kinshipand affinity, and especially by the peacemak-
ina of the womiien who hold to one clan as sisters and to
another as wives,it tends to keep down feuds and to heal
them when they arise, so as at criticalmomentsto hold to-
gethera tribewhich under endog,amnous conditionswould have
split up., Exogamy thus shows itselfas an institutionwhich
resiststhe tendencyof unculturedpopulationsto disintegrate,
cementingthem into nations capable of living together in
peace and holding togetherin war,till they reach the period
of highermilitaryand political organisation. Seen fromthis
point of view, the remarkablefact is more easilyunderstood
that exogamy,passing on fromthe maternalto the paternal
stage of society,shiftsits prohibitions fromthe femaleto the
male line of descent,now allowingmarriageswhich it treated
formerlyas ilncestuous, while prohibitingotherswhich it for-
merlyallowed withoutscruple. This transformation has been
taking place withinrecenttimes amongMalay and American
tribes,and seems to be even goingon still, it malkingno dif-
ferencepoliticallywhetherkinshipfollowsthe femaleor male
line,if only marrying-out causes the requisiteintermixture of
the clans. In this connexionit is worthwhile to noticethat
thereare a small numberof peoples in different parts of the
world,who have a rule of exogamynot dependingon kinshipat
all. For instance,Piedrahita2relatesof the Panches of Bogota,
thatthoseof one towndid inotmarryany womanthereof, as all
held themselvesbrothers, and the impedimentof kinship was
sacred to them,but such was theirignorancethat if a sister
werebornin a different townfromher brother, he was not pre-
vented frommarryingher. An anthlropologist, with the list
beforehim of the peopleswho prohibita nmani frommarryingin
his own village,mightexplainthisnot as a result of ignorance,
but as an extremecase of what may be called " local exogamy."
1 W. RobertsonSmith," KinshiparndMarriagein Early Arabia," p. 184.
2
Piedrahita, " Historia General de las Conquistas del Nuevo Reyno de
Granada," 1688,page 11.

of Institutions; appliedto Laws of Mlarriageand Descent. 239
The results here brought forwardmake no approach to

exhaustingthe possibleinferencesto be drawn fromthe tables.
These need not even be confinedto workinoout thedevelopnieilt
of customsfoundin existencesomewhereon the globe,but may
in some measurerestorethe knowledgeof formsof societynow
extinct. Interesting, however,as theseproblemsare,I am more
anxious to bring under discussionthe methodby wvhich they
are heretreated,how imperfectly I am well aware. The inter-
pretationsofferedwill have to be corrected,the tabulated
materialimprovedin quantityand quality,and the principles
it involves broughtout more justly, yet at any rate it will
reinain clear that the rules of human conduct are amenable
to classificationin compact mi-asses, so as to show by strict
numericaltreatmenttheirrelationsto one another. It is only
at this point that speculativeexplanationmust beoin,at once
guided in its course and strictlylimitedin its range by well-
markedlines of factto whichit mniust conform. The key of the
positioIn is, as thatveterananthropologist, Prof.Bastian,of tie
Berlin Museum,is neverwearyof repeating,that in statistical
investigationthe futureof anthropology lies. As soon as this
is systematicallyapplied, principlesof social developmentbe-
come visible. Eveni the diagramsof this paper may sufficeto
show that the institutionsof man are as distinctlystratified as
the earth on whichhe lives. They succeedeach otherin series
substantiallyuniform over the globe, independentof what
seem the comparativelysuperficialdifferencesof race and
language,but shaipedby similar human nature acting,through
suceessivelychangedconditionsin savage,barbaric,and civilised
life.
The treatmentof social phenomenabynumericalclassification
will,it must be added,reacton the statisticalmaterialto which
the methodis applied. It is in classifyingthe re-ordsof tribes
and nationsthat one becomesfullyaware of theirimperfect and
even fragmentarystate. The descriptions happily tend to
correctone another'serrors,but the great difficulty is blanlk
want of information.As for extinct tribes,and those whose
nativeculturehas been re-modelled, thereis nothingto be donie.
But there are still a hundredor more peoples in the world,
among whorna promptandl minute investigationwould save
some fastvanishinog memoryof their social laws and customs.
The quest mightbe followedup internationally, each civilised
iation taking in hand the barbarictribeswithin its purview.
The futurewill,doubtless,be able to take care of itself as to
mostbranchesof knowledge,but thereis certainworkwhichiif
it is to be done at all, mustbe done by the present.
VOL. XVIII. U

270 Discussion.
DIscussIoN.
The PRESIDENTfelt sure that no one would have appreciated Dr.
Tylor's memoir more justly, or would have welcomed it more
warmly,than Mr. Herbert Spencer,whose effortsto erect a science
of sociology upon an inductive basis weire well known. Mr.
Spencer, as we all remember,went to great cost,and mu-chexerted
hirnself to obtain a collection of the customs of all available
nations, savage and civilised, arranged in an uniform and orderly
maijanerfor purposes of intercomparison. The result was the
publication of an amount of material that filled four very large
folio volumes. Unfortunatelyhe had been obliged to delegate to
others the task of compilation,and the work was not carried out
as accurately as was desirable, or even as completely,notwith-
standing its bulk. Much the same may be said of another and
a differentcollection. Dr. Tylor has, on the contrary,collected a
mass of well sorted and highly considered information,by means
of a sustained and scholarly investigation,extending over many
years, and there could be little doubt that a publication of his
compact notes, supplemented it might be by the notes of other
anthropologists,would be of itself a most valuable and acceptable
work. Dr. Tylor's memoir dealt both with a method and with
conclusions; it was of the method only that he (Mr. Galton)
would now speak. It consistedin ascertainingthe degree in which
the concurrence of certain customs was exceptionally frequent.
He thought that the degree of interdependence,to which the
various degrees of exceptional frequency testified,might with ad-
vatitage be expressed in terms of a scale, in which 0 represented
perfect independence,and 1 complete concurrence. By doing so,
the values of the various concurrences would become more clear.
As an example of what he meant, he would referto a scale used in
certain psycho-physicalinquiries and discussed in Fechner's book,
where the true significanceof the various percentages of success
and failure was tabulated.
It was extremelydesirable for the sake of those who may wish
to stu-dythe evidence for Dr. Tyior's conclusions,that full infor-
mation sshouldbe given as to the degree in which the customs of
the tribes and races wlhichare compared togetherare indepeudent.
It night be, that some of the tribes had derived them from a
common source, so that they were duplicate copies of the same
original. Certainly,in such an investigation as this, each of the
observationsought, in the language of statisticians,to be carefully
" weighted." It would give a useful idea of the distributionof the
several customs and of their relative prevalence in the world, if a
mnapwere so markedby shadings and colour as to present a picture
of their geographical ranges.
Professor FLOWER remarkedupon the great value of Dr. Tylor's

paper, congratulatinigIiinaon the application of a rigid statistical
method to a research whichhad generallybeelnconducted on vague

Discussion. 271
and uncertain lines. It was, however,perfectlyobvious that the
value of such a method depended entirelyupon the units of com-
parison being of equivaleintvalue, and this seemed to him to be a
very great difficultywhen dealing with groups of mankind. He
had, however,no doubt that Dr. Tylor had taken every means in
his power to eliminate the errors which might arise from this
source.
Mr. G. BERTIN, afterremarkingthat this paper would do a great
deal to elevate anthropologv,said he thought that, if Dr. Tylor
had included in his diagrams one illustrating the primitivestate
of society in which women were the commonwives of the clan or
tribe,it would explain everything. This state of things still exists
in some parts of Tibet, and traces of it are detected in Ancient
Egypt and among the primitive Semites. Women were at first
considered like other properties,and in the communiststage they
used to belong to each and all; when propertywas divided women
were assimilated to landed properties or estates, and the children
took the name of their mother,as in feudal countriestheytook that
of their estate. This is really the origin of the so-called matri-
archate, in which the mother had, in fact,no power, but gave her
name to her child. It is only with progress and civilisation that
the position of women was raised till it tends in our modern times
to place them on equal footingwith men.
SIR G. CAMPBELL agreed with a preceding speaker that the

maternal system does not mean the rule of the female,but only
that she is used as the familyseed-bed. And he would very much
like to obtain informationon one point in the historyof marriage,
viz., who inventedor how came about the very peculiar systemof
monogamy,so prevalent among all Aryan races, and under which
a man is not only confinedto one wife,but tied to her by indis-
soluble bonds. The maternal system we understand,under which
the women of a familyare the brood mares of their own family;
the patriarchal system we understand,under which a man rules
over his wives, slave girls, and children. and exchanges the
formerfor otherswhen he thinksfit. But the sacramental mono-
gamous marriage, by which a man is tied to one wife for ever,
(among the Hindoos the wife is tied to him even beyond the grave),
that is very peculiar, and he had never seen it accounted for.
Mr. BOUVERIE-PUSEY remarked that Dr. Tylor's views on the
origin of exogamy derived confirmationfrom an old Hungarian
law, according to which the Ishmaelites (Tartirs converLedfrom
Islam) were commanded to give all their daughtersin marriage to
Hungarians, and to take none but Hungarian wives for their
sons, obviously to prevent their continuing to form a separate
nationality.
Dr. TYLORcongratulatedhimselfon having been able to place the
preseintmethod before investigatorswhose criticism was of such
u 2

272 Discussion.
importance,from their thorough appreciation of the points in

which such a method has inherentweakness. With the details as
yet in an imperfectstate, he found it difficultto bring out the
resultsexcept as a temporarystep, which is, however,on the road
to permanentsettlement. The difficulty raised by Mr. Galton that
some of the concurrencesmight result from transmissionfrom a
common source, so that a single character might be counted
several times from its mere duplicates, is a difficulty ever present
in such investigations,as for instance in the Malay region,where
groups of islands have enough differentiationin their marriage
systems to justifytheir being classed separately,though traces of
commonorigin are at the same time conspicuous. The only way
of meetingthis objection is to make separate clsssificationdepend
on well marked differences, and to do this all over the world. With
regard to Professor Flower's caution as to the units of comparison,
an answer of somewhat the same kind might be given. When a
communityor grotipof communitiesfollowsa law of marriage and
descent substantiallysimilar,this may be taken as a unit,-notwith-
standinghistoricalconnectionand the consequetnt partial correspon-
dence which may exist between it and other unit systems. If this
methodbe fairlyand equably worked over the world,the correspon-
dences broughtabout by historicalconnexiontend to set offagainst
one another,leaving the results of general human action more or
less clear.
Dr. Tylor added that he had collected much material bear-
ing on the great problem raised by Sir George Campbell, but
at present without any result sufficientlydefiniteto be brought
forward.
NOVEMBER 27TH, 1888.
FRANCISGALTON, Esq., F.R.S., Pqesident,in theChair.
The minutesof the last meetingwere read and signed.

The followingelectionswere announced:-Rev. LORIMER A.
FISON, of Flemington,Melbourne,Australia, as an honorary
member; Mrs. C. A. FRASER, of 10, Craven Hill, Hyde Park;
Department
HENRY BALFOUR, Esq., B.A., of the Anthropological
of the UniversityMuseum,Oxford; and H. HAVELOCK ELLIS,
Esq., of EarlsbrookRoad, Redhill,as ordinarymembers.
The followingpresentswere announcedand thanksvoted to
the respectivedonors:


On A Method of Investigating The Development of Institutions Applied To Laws of Marriage

Uploaded by

Document Information

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

On A Method of Investigating The Development of Institutions Applied To Laws of Marriage

Uploaded by

Copyright:

Available Formats

On a Method of Investigating the Development of Institutions; Applied to Laws of Marriage

This content downloaded from 195.34.78.237 on Sun, 15 Jun 2014 20:15:32 PM

From the UNIVERSITY.-Journal of the College of Science, Imperial

The followingpaper was read by the Author:-