Volume29, Number2, AprilI988 1 349

earlyAtlantic settlementin Scania. Acta Archaeologica Lundensia,series4, 12.
MARKOVIC-MARJANOVIC,
J. 1978. "Geology and
in Vlasac: A Mesolithicsettlementin
stratigraphy,"
the Iron Gates, vol. 2. Edited by D. Srejovicand Z.
Letica,pp. 11-27. Beograd:SerbianAcademyof Sciences.
NEMESKERI,
J., AND L. SZATHMARY.
1978. "Anthropology,"in Vlasac: A Mesolithicsettlementin
the Iron Gates, vol. 2. Edited by D. Srejovicand Z.
Letica,pp. 69-I85. Beograd:SerbianAcademyof Sciences.
ROZOY,
J-G. 1978. Les dernierschasseurs.Bulletinde
la Societe ArcheologiqueChampenoise,special number.
SREJOVIC,
D., AND Z. LETICA.
1978. "Archeology,"in
Vlasac: A Mesolithic settlementin the Iron Gates,
vol. i. Edited by D. Srejovicand Z. Letica, pp. 70170. Beograd:SerbianAcademyof Sciences.
TRINGHAM,
R. 1971. Hunters,fishers,
and farmersof
easternEurope.London: Hutchinson.
VAN DER LEEDEN,
F. 1975. Waterresourcesof the
world.PortWashington,N.Y.: WaterInformation
Center.
S. 1975. A note on the anthropological
ZIVANOVIC,
characteristicsof the Padina population.Zeitschrift&
ffirMorphologieund Anthropologie66:I61-75.
Z. K. I983. Prehistoricalskeletal remains
ZOFFMANN,
fromLepenskiVir (IronGate, Yugoslavia).Homo

illustrations and measurements (Gillispie 1974:540).2

To providespecimens forhis anatomylectures,he collected skulls representingnumerouspopulations (Stanton i960:27-28).
As the collection grew,Mortonbegan
to use it as the focus of his ethnologicalresearch.By
I849 the MortonCollection ofHuman Skulls contained
over8oo human craniafromthroughouttheworld(Morton i849a:vi).
With Blumenbach,Morton believed that therewere
fivemajorraces, each characterizedin partby the shape
of the head (Stantoni960:4-I,
29). He consideredcranial capacityan especially good indicatorof race (StanTo determinecranial capacity he filled
ton i960:3i).
crania with siftedmustardseed and then emptied the
seed into a homemade volumetric cylinder (Stanton
i960:29-32).
In I839, he used 256 of these values to
calculate a mean cranial capacity for each of the five
races. The table listingthese means, reproducedhereas
table i, was publishedalong with manyofMorton'sraw
data in his Crania Americana (i839:260).
The thesis of
this workwas thatnative Americanswere one race distinctfromEskimos and Mongolians. Crania Americana
implied, thoughit did not state, that each race had an
independentorigin,3and since Morton arguedthat the
races were as immutableas species thereaderwas leftto
conclude that the racial differencesin cranial capacity
detailed in this table were as old as humanity.
As Morton continued his ethnologicalresearch and
enlargedhis collection,he improvedthe accuracyofhis
measuringtechniqueby substitutinglead shot formus34:129-48.
tard seed and in I849 used 623 of these "shot values"
to constructan even more elaborate table of cranial
capacities. This I849 table, reproducedhere as table 2,
and all of the data used in its constructionwere published in Morton's (I849a) Catalogue of Skulls (pp. viiviii). In this work,Mortonproposedthat the fiveestablished races would be more aptlydescribedas "groups"
and divided these "groups" into "families" in turndiJOHN S. MICHAEL
vided into "races." Late in his career,he began to apply
7446 OverhillRd., Melrose Park,Pa. 19126 U.S.A.
the results of his research to enhancingthe scientific
IO VI 87
understandingof species. From his I849 table he concluded that each race was characterizedby "a collective
Samuel George Morton,M.D. (1799-i85i), was one of identityof physical traits"and said thathe favoredthe
the giantsof the Americanscientificcommunityof his independentoriginofraces (p. ix). By I 85I he was openly
time. In addition to being a practicingphysicianand a declaringthathuman races werein factspecies; physical
professorof anatomy,Mortonwas active in geologyand formwas forhim the ultimatecriterionfordetermining
ethnology.Because of his expertisein paleontology,he species, and his ethnologicalresearchshowed thateach
was regularlycalled on to examine and describenewly race had a distinctform(Stanton i960:140-41).
Moruncoveredfossils.He was also lauded forhis innovative ton's definitionof species contradictedthe widely acapproachto ethnologicalresearch,especiallyhis use of ceptednotionofspecificinfertility.
Mixed-racehumans,
unlike hybrid animals, are indeed fertile,and many
I. ? I988 by The Wenner-Gren
FoundationforAnthropological scholars used this to
support the specific unity of
Research. All rights reserved OOII-3204/88/2902-0007$I.oo. This

A New Look at Morton's
CraniologicalResearch'

paperis based on researchreported
in "An AnalysisofSamuelG.
Morton's Catalogue of the Skulls ofMan and the InferiorAnimals,

ThirdEdition,Basedon a Remeasurement
ofa RandomSampleof
theMortonCollectionofHumanCrania,"whichwas presented
to
the MacalesterCollege Honors Programin the Departmentof
Geologyon May i, I986. I am gratefulto JanetMonge,Gerald
Webers,JoeV. Michael,and Tom and BonnieMichaelforsupport
and encouragement.

2. Gillispierefersto Mortonas "a founder
ofinvertebrate
paleontologyin the United States" and describeshis firstethnological
publicationas "a landmarkin anthropology."
3. StantonfeelsthatMorton,a man of social standing,
mayhave
shied away froma directdiscussionof racial originsin orderto
avoidanypotentialforscientificor religiouscontroversy
(PP.3233).

350

1 CURRENT

ANTHROPOLOGY

I

TABLE

Mean Cranial Capacities (in.3) ofRaces accordingto
Morton(1839)
Race
Caucasian
Mongolian
Malay

American

Ethiopian
SOURCE:

Sample
Size

Mean

52
Io
i8

87
83
8i

I47

29

82

78

Largest
IO9

93
89
IOO

94

TABLE

2

Mean Cranial Capacities (in.3) ofRaces accordingto
Morton(I849a,b)
Sample
Largest Smallest Mean
Size

Smallest

Race and Family

75
69
64
6o

Modem Caucasian Group
Teutonic Family
Germans
English
Anglo-Americans
Pelasgic Family
1
Persians
Armenians
J
Circassians
Celtic Family
Native Irish
Indostanic Family
Bengalees, &c.
Semitic Family
Arabs
Nilotic Family
Egyptians

65

Morton (i839:260).

humankind.Accordingto Morton,however,specificinwas a mythand hybridswere actually as comfertility
species of animals as they were among
among
mon
"species" of humans (StantonI960:140-41).
Gould (1978, 198I) has criticizedMorton's ethnological resesarch,especially his I839 and I849 tables, for
conformingto the conventionalAmerican racist tenet
that the "Caucasian" race is superiorin intelligenceto
"Negroes." Morton,he claims, "regardedcranial capacHe
ityas an overallindicatorofintelligence"(1978:503).
numbers
unequal
contain
samples
Morton's
that
shows
of male and female crania,an importantfactorbecause
cranialcapacityis generallyless in women thanin men.
He also notes that Morton'ssamples are not uniformin
size and should not have been comparedand that this
sample-sizeinequalityis compoundedin the I849 table
when Morton averages the means of "races" and presents the resultingvalues as the means of "families" or
"groups." In addition, Gould findsthe I839 and I849
tables inconsistentwith theirrespectivedata because of
miscalculations and omissions. Finally,he argues that
Morton's I839 "seed data" are inconsistentwith his
He goes on to suggestthat
I849 "shot data" (pp. 505-9).
these errorsindicate that Morton unconsciously doctoredhis results in terms of his a prioriconvictionof
Caucasian racial superiority.He tracesthis bias to Morand claims thattheerrorss erve
ton's culturalupbringing
to skew his results to conformto this bias. He even
speculates that Morton may have systematicallymismeasuredcrania recordedin his I839 "seed data" in accordance with this bias. In the final analysis, Gould
views this as a prime example of how a prioriconvictions can influencethe outcome of ostensiblyobjective
research(pp. 504-9).
To assess Gould's claims, I remeasuredthe cranial
capacities of 2oi specimens fromthe Morton Collection4and comparedthevalues withMorton's.I also used
the I849 data to recalculatemean cranial capacities for
populationssimilarto thosepresentedin Morton's i 849
table. The goals of this work were to determine(i)
4. The Morton Collection of Human Crania is currentlystored in
the University Museum, the University of Pennsylvania, and was
remeasured duringJanuaryof I986.

Ancient Caucasian Group
Pelasgic Family
Graeco-Egyptians
Nilotic Family
Egyptians
Mongolian Group
Chinese Family
Malay Group
Malayan Family
Polynesian Family
American Group
Toltecan Family
Peruvians
Mexicans
Barbarous Tribes
Iroquois

Lenape

Cherokee
Shoshone, &c.

i8
5
7

II4
I05

Negro Group
Native AfricanFamily
American-Bom Negroes
Hottentot Family
AlforanFamily
Australians
SOURCE: Morton (I849a:viii;

90

97

9I

82

96 92
go J

IO

94

75

84

6

97

78

87

9I

67

8o

3

98

84

89

I7

96

66

8o

i8

97

74

88

55

96

68

8o

6

9I

70

82

20

97
84

82

68

86
83

I55
22

IOI

58
67

75
79

32

3

I

70

92

85

79

i
i6x

0

104

70

8
84

0

I2

3

99
89
83

65
73
68

83 18
82 J 3
75

8

83

63

75

62

i849b:222).

whetherMorton accuratelymeasured cranial capacity
usingshot,(2) whetherMorton'sI849 table is incompatible with his I849 data, (3) whetherthereis evidence
that Morton unconsciously doctoredhis I849 table in
terms of a racial bias, and (4) whetherMorton's I849
table is scientificallysound.
To avoid confusion,all of Morton's measurementsof
cranial capacity are called "measurements" and my
measurements of identical crania "remeasurements."
Likewise, Morton's data were "calculated" in i849 and
"recalculated" forthis study.His termsforthe populations in his I849 table are enclosed in quotations (e.g.,

Volume 29, Number 2, April I988
TABLE

| 35'

3

Mean Cranial Capacities (in.3)fromMorton'sI849 Measurementsand I986
Remeasurements

Group and Subgroup ("Race" or "Family")

Sample
Size

American
Mexican
Peruvian
Various tribes ("Barbarous Tribes")

9I
IO
54
27

Caucasian
Ancient Pelasgic ("Graeco-Egyptians")
Anglo-American
Celtic
Egyptian
English
Fellah
German
Indostanic
Pelasgic ("Persians, Armenians, Circassians")
Semitic

52a

Malayan
Malayan
Polynesian

II

Mongolian
Mongolian ("Chinese")
Negro
AfricanNegro ("Native AfricanFamily")
American Negro ("American-BornNegroes")
Australian
Hottentot
Mixed

Measurement
Mean

Remeasurement
Mean

78

76
88
88
8i

92

go
82
84
95
83
78
86
98
73
89
78
88
84

84
9i
8o
73
84
96
73
88
77
84

8
3

85
86
84

83
84

4
4

87
87

83
83

40

I

82
84
8o
73
75

79
8i
78
7I
74

3

8o

77

3
7
I

4
3
6
Io

II

3
3

28

7
4

82

82

aIncludes one Caucasian group cranium of unknown subgroup.

uses molded acrylicballs in place of shot.7Remeasured
capacities were then enteredinto a computerfile along
with the capacities recordedin Morton's I849 catalog.
Mean cranial capacitieswere determinedforgroupsand
subgroups and compared with those presented for
"groups,""families,"and "races." Each craniumwas remeasured threetimes, the largestvariationbeing only
i.8% by volume.8Individualcranial capacities thus determinedwere consistentwith Morton's9but on the average32.48 cm3(roughly2 in.3)lower.Means calculated
forgroupsand subgroupsfromremeasuredvalues were
also consistentwith but lower than means calculated
fromMorton'sdata (table 3). Because the I849 table does
5. This catalog,unlike Morton's,lists the crania accordingto not representall the measuredcraniaMortonlisted,it is
"race."Mygroup-subgroup
classification
was createdsolelyforthe impossibleto determineexactlywhich craniahe used in
analysisoftheMortonCollectionandhas no biologicalorsociolog- his calculations. Nonetheless, mean cranial capacity
ical significance.
Althoughbasedon Morton'sandMeigs'ssystems was recalculatedforgroupsand subgroupsusing all the
ofclassification,
itis presented
notas a correction
ofthesesystems crania listed,and 75% of the "family"or "race" means
but as a rigorously
definedalternative.
6. I originally
remeasuredI 2 randomly
selectedcraniafromeachof reportedby Morton in I849 were found to be within
"group," "Egyptianrace"); my alternativesare not. All
of the crania listed in Morton's i849 catalog were assignedto rigorouslydefinedgroupsand subgroupsbased,
respectively,on Morton's "groups" and on his "families" and "races" (all loosely defined).Some ofthe crania
measured by Morton do not fit clearlyinto any of the
"races" in his I849 table but were assignedto subgroups
accordingto Meigs's (I857) catalog of the Morton Collection.5
Of the craniameasuredbyMorton,2oi wererandomly
selected forremeasurement.6Cranial capacity was remeasuredby a techniqueadaptedfromOlivier(i969) that

threepopulations:(i) Black Africansand Americans,(2) Native
Americans,and (3) Teutonic Europeansand Americans.I then 7. Moldedacrylic"no-hole"balls are availablefromGreenePlastchose to enlargethis studyto include all populationslisted as ics Corp.,Hope Valley,R.I.
"races"in Morton'sI849 tableandremeasuredI65 craniafromall 8. Twenty-nine
craniahad a coefficient
of variationovero.99%,
"races" at random.The sampleof2oi craniatherefore
containsa whilefivewereundero.io%.
slightly
largersubsampleoftheabovethreepopulationsrelativeto 9. Over95% ofMorton'smeasurements
werewithin4 in.3ofthe
mostsubsamplesremeasured.
measurements;
fewerthan7% weresmaller.

352

| CURRENT

ANTHROPOLOGY

TABLE

4

Mean Cranial Capacities (in.3)fromMorton'sI849 Table and I986 Recalculation
Morton's

I849 Table

Group and Subgroup ("Race" or "Family")
American

Mexican
Peruvian
Varioustribes("BarbarousTribes")

Caucasian
Ancient Pelasgic ("Graeco-Egyptians")

Anglo-American
Celtic
Egyptian

English

Fellah
German
Indostanic
Pelasgic ("Persians, Armenians, Circassians")
Semitic

Malayan
Malayan

Polynesian

Mongolian
Mongolian ("Chinese")
Negro
AfricanNegro ("Native AfricanFamily")

AmericanNegro("American-born
Negroes")
Australian
Hottentot

Sample
Size
338
22
I55
i6I
I7I

i8
7
6
55
5
I7

i8
32
I0

3

23
20

3

6
6
85
62
I2

8
3

Recalculation

Mean

Sample
Size

79
79
75

335
27
I52

84
--a

88

90

87
8o
96
8o
90

8o
84
89

I57

I85
I7

Io
6
55
5
i8

20

33
I0

8

Mean
8o
82

75
84
83
87
85
88
8i
96
79
87

8o
84
85

27
23

85
85
83

82

8
8

85

a

93

82
84

85
86
83
82

83
82

75
75

4

67
I0

I0

3

85

8I

75
75

aNot calculated by Morton.

in.3ofthe subgroupmeans recalculatedfromthe same it was impossible to determinewhich three,I included
them all in the Semitic subgroup(Meigs i857:50)."
data (table 4).
There is no indicationthat Morton's miscalculations
Many of the means presentedin Morton's I849 table
are inconsistentwith his data, probablybecause ofmis- or omissions had any substantial effecton his overall
calculations and omissions. As table 4 shows, for ex- results.Whenall ofMorton'sI 849 data areused to deterample, the Celtic and Egyptianmeans are roundedoff mine mean cranial capacities forsubgroups,the resultincorrectly.'0Again, he lists 670 crania measured for ing means are quite similar to those presentedin his
cranial capacity and his table (p. vii) shows only 623 I849 table.'2 The only errorofMorton's I849 table that
measurements. While he intended to omit "idiots," may indicate bias is his unfaircomparisonof samples.
"mixed races," and childrenfromhis table (p. ix), this His samples are unequal in size and sexual distribution,
does not account forall the omissions.Accordingto his and Gould (1978:505-6) has convincinglyargued that
criteria,no Fellah, Indostanic,Malayan, or Polynesian Mortonhad some knowledgethat sample size could afsubgroup crania should have been omitted from the fectmeans. It is possible thatMortonunconsciouslyadtable,and yet some are omittedfromeach ofthese sam- justed his resultsby limitingthe size of some samples,
ples. Table 4 appears to indicate furtherproceduraler- but because his other errorsdo not indicate bias it aprors,but these may be an artifactof the group-subgroup pearsequally possible thattheinequalityofsample sizes
system.For example,Mortonlists the capacities of five was a resultof his ignoranceof statistics.
crania describedas "Semitic" (Meigs I857:34-35), two
Unlike Morton's calculations, his specific cranial
"Baramka, or Baramacide Arabs," and one "Bedouin,"
but ofthese eightcraniahe tabulatesonlythree.Because i i. Likewise,the Mongoliansubgroupcontaineda "Laplander"
2

thatmaynothavebeenusedbyMortonin the"Mongolianfamily"
io. Althoughsubgroupsdo not alwaysmatchMorton's"races"in mean.
do. It is quiteprob- i2. Andthisis so eventhoughMortonincorrectly
determined
the
samplesize,theCelticandEgyptiansubgroups
ablethatthedataused to recalculatethesetwomeansareidentical meansfor"groups"and "families"byaveraging
themeansoftheir
to thoseused byMorton.
constituent
"races.'"

Volume 29, Number 2, April I988

|

353

not Gould's onlyerror.He presents
This is, furthermore,
a recalculation of the I849 table entitled "Corrected
values forMorton's final tabulation" (p. 508) showing
means of 86 in.3for"Native Americanpeoples," 85 in.3
for"Mongolian peoples," "Modern Caucasian peoples,"
and "Malay peoples," 84 in.3 for "Ancient Caucasian
peoples," and 83 in.3 for"Africanpeoples." In fact,this
table is not a correction;it is based on data and terminologynot used byMorton.Gould uses data fromI839 that
Morton does not use in final tabulation,and his "peoples" do not contain all the samples in Morton's
"groups." His "Africanpeoples," for example, do not
include Morton's "Australians" or "Hottentots." He
justifiesdroppingthe "Australians" fromhis "African
peoples" by citing modem anthropologicalevidence
showingno relationshipbetweennativeAustraliansand
black Africans(p. 508). Yet he does not combine "Modem Caucasian peoples" and "Ancient Caucasian peoples" as modem anthropologistsusually do.'5 He omits
the "Hottentot" subsample fromhis "Africanpeoples"
because Hottentots"are very small in stature,and all
three crania are female."'6 Elsewhere,however,he includes a subsample offive"English" crania,all male, in
his "Modern Caucasian peoples" mean (p. 508). Also,

measurementscontain few errors.To be sure, theyare
consistentlylargerthan my remeasurements,but current anthropometric
c onventionaccepts that measurements of cranial capacity will varywith the technique
(Olivieri969:134). Gould's speculationthatMortonsystematicallymismeasuredcrania using seed is based on
his statisticalanalysis of Morton's I839 and I849 data
and the assumption that Morton held "Caucasians"
superiorto "Americans" and these in turn superiorto
"Negroes."'13 He admits that his own resultsmay have
been due to the increasein sample size afterI 839, buthe
presentsotherevidencesuggestingthatMortonadjusted
his values to position "Negroes" below "Americans."
Gould finds that Morton's I839 data for "American"
craniaproducea mean capacityof 8o in.3while the I839
table gives an "American" mean of 82 in.,3 above the
"Ethiopian" mean of 78 in.3 Because he findsno evidence that Mortonknew of this error,Gould concludes
that it was unconscious (pp. 505-6). I have, however,
uncoveredevidence suggestingthat Mortonwas aware
of it. The libraryof the Academy ofNatural Sciences of
Philadelphia holds a copy of Crania Americana that
contains a dedication handwrittenby Morton. In this
copy thereis also a zero penned in over the "2" of the
"82 cubic inches" recordedas the "American" mean in
question.'4 Gould considersthis errora result of Morton's desire to adjust the "American" mean above the
"Ethiopian" but below the "Caucasian." If there had
been a similar errorreducing the "American" mean,
however,it could also have been interpreted
to indicate
bias toward"Caucasians." This hypotheticalbias is subfromthat spelled out by Gould but just as
tly different
reasonable. If conflictingevidence can supporttwo almost identical interpretations,
it cannot be said clearly
to indicate anything.Gould's statisticalanalysis would
supporthis suspicion of systematicmismeasurement
onlyifMortonhad the bias he attributesto him. Since I
have foundno indicationof thatbias, and giventhe accuracy of Morton's shot data, it seems unlikely that
Mortonsystematicallymismeasuredcrania in I839.
Gould makes no mention of any errorsthat do not
appear to favor Morton's assumed racial bias. I have
foundsuch an error:the i849 table and my recalculations give the "Malay group" and Malayan groupmean
as 85 in.3 even thoughthe recalculatedsample is larger
thanMorton's; simplyaveragingthe recalculatedmeans
forthe Malayan and Polynesian subgroupsproduces a
value i in.3 lower than Morton's. This errorcannot be
attributedto Morton's bias, and it is all the more
noteworthybecause Gould misses it in determininga
"Malay" mean of 85 in.3 using Morton's data (p. 508).

AlthoughGould is mistaken in many of his assumptionsabout Mortonand his work,he is correctin asserting thatthese tables are scientificallyunsound.He fails,
however,to mentionthe overridingreason forrejecting
them,namely,Morton's acceptance of the existenceof
race. Most anthropologistsfeel that there is too little
evidence to conclude with certaintywhetherthe concept ofrace is a biologicalrealityor simplyan artifactof
classification(Weiss and Maruyama 1976:47). If race
dose not really exist, then Morton's samples are
meaningless, and this criticism overshadows Gould's

I3. "I have reanalyzedMorton'sdata and I findthattheyare a
patchwork
ofassumption
and finagling,
controlled,
probably,
byhis
conventionala prioriranking(his folkson top,slaveson thebottom)"(Gould I978:504).
I4. The dedicationreads,"Presented
to the Academyof Natural
SciencesofPhiladelphiabytheAuthor.December24, I839." The
inkusedforthecorrection
matchestheinkusedforthededication,
and the handwriting
ofthe dedicationmatchesthatfoundin the
library'scollectionofMorton'spersonalletters.

I 5. I haveyetto findanyanthropological
workotherthanMorton's
orMeigs'sthatseparatesCaucasiansintoancientandmodempopulations.
i 6. Gould did this even afterhe criticizedMortonforhaving
dropped"Hindu" crania fromthe "Caucasian" sample because
theywerenotablysmallerthantherestofthesample.
I7. I have chosento followStanton'sinterpretation
becauseit is
directlysupported
byhistoricalreferences.

Crania iI9I,

and 1248, "Lap"Germanof Frankfort,"

lander,"are omittedfromthe "Corrected"table (P. 508).
I conclude that
Contraryto Gould's interpretation,
Morton
Morton'sresearchwas conductedwithintegrity.
was one of the firstscholars to attemptthe study of
human diversitythroughobjective measurements,and
it is not surprisingthathe made mistakes.Althoughhe
cannotbe excused forhis errorsor his unfaircomparison
ofmeans, he should be givencreditforhavingtakenthe
risk of experimentingwith a new and innovativetechnique. He was attemptingto understandracial variation
and not, as Gould claims, tryingto prove Caucasian racial or intellectual superiority.The science historian
William Stantonconcludes that "Mortonhimselfnever
equated cranial capacity with intelligence" (Stanton
i96o:35 |.17

354

CURRENT

ANTHROPOLOGY

criticismsof the size and sexual distributionsin Morton's samples. Gould nowhere questions the scientific
authenticityof racial classificationand in fact,by presentinga "Corrected"table,lends supportto the notion
of racial c4assificationas a biological reality.
Morton'stables containmiscalculationsand omissions
of data, but his I849 data are reasonablyaccurate and
thereis no clear evidence that he doctoredthese tables
forany reason. His tables are neverthelessscientifically
unsound because his so-called samples were never adequately defined.His failureto define"race" makes his
workstatisticallymeaningless.I do not arguethatostensibly objectiveresearchmay not sometimesbe affected
by the unconscious desires of the researcher;I simply
suggestthattheworkofMortonis not a clearexampleof
this phenomenon.His workis flawed,and the scientific
communitywithin which it took place was racially
biased, but a connection between Morton's errorsand
this conventionalracismis simplynot supportedby the
evidence at hand.

Half a WifeIs BetterThan None:
A PracticalApproachto
NonadelphicPolyandry'
M. E. STEPHENS

DepartmentofAnthropology,Universityof Calgary,
Calgary,Alta., Canada T2N iN4. i8 VI 87

Nonadelphic polyandry,definedfor this paper in culturalratherthan biologicalterms,2is a sexual union between a woman and two or more unrelatedmen such
that the children born to the woman are considered
legitimateoffspring.
Implied in this definitionis a temporal aspect; a series of monogamous unions does not
qualify as polyandry.Also implied is an agreement
amongthe partiesinvolvedin the union; cuckoldrydoes
not qualify.
The early literature on polyandryamong human
groups stressed its rarity. Murdock (I949:25) flatly
ReferencesCited
statedthat"the polyandrousfamilyoccursso rarelythat
it may be regardedas an ethnological curiosity"and
C. C. Editor.I974. Dictionaryofscience
GILLISPIE,
mentionedonly the inhabitantsof the Marquesas and
biography.New York: Scribner.
the Toda of India as usual practitioners.Nonadelphic
cranial
S. J. I978. Morton'srankingofraces by
GOULD,
has been found only among the Nayar (see
polyandry
200:503-9.
capacity.Science
Gough i968) and the Marquesans as a normalpractice;
1 I98I.
The mismeasureofman. New York:
sporadic polyandryis found occasionally among other
Norton.
groups,
usually relatedto infanticide(MurdockI949).
from
MEIGS,
J. A. I857. Catalogue ofhuman crania
The
recentliteraturehas studiedthe phenomemore
of
Sciences
Natural
of
the
Academy
of
collection
the
Philadelphia. Philadelphia:Merrihewand Thompson non froma sociobiologicalperspective,examiningpolyandry as a reproductivestrategy.Wrangham's (i980:
Printers.
263) definitionof strategy,"a set of decisions which
G. I839. Crania Americana. PhilaSAMUEL
MORTON,
produce
behaviour patterns with a particular result:
delphia: J.Dobson.
the
ultimate probable outcome is to increase
namely,
. I849a. 3d edition.Catalogue ofskulls ofman
the
access to a given resourcespecificto the
behaver's
and
Merrihew
and the inferioranimals. Philadelphia:
strategy,"is used here.As Mulder (i987:6) has reminded
Thompson Printers.
us, "The evolutionaryargumentis essentiallythatdeci. I849b. Observationson the size ofthe brainin
sions, choices and customarypatternsofbehaviourhave
of
the variousraces and familiesofman. Proceedings
the
ultimate effectof increasingreproductivesuccess,
Philadelphia
Natural
Sciences,
theAcademy of
whetherthis is a conscious goal or not."
4:22I-24.
Most studies of polyandryassume that the evolutionSpringfield:
G. I969. Practical anthropology.
OLIVIER,
benefitsto females are obvious. In cultural terms,
ary
Thomas.
polyandry
may be describedas benefittingfemales by
spots.
The
leopard's
I960.
WILLIAM.
STANTON,
makingthe labourand resourcesoftwo men available to
Chicago: Universityof Chicago Press.
These are benefitsin termsof reproducI976. Archeology, theiroffspring.
K. M., AND T. MARAYAMA.
WEISS,
tive
the
emotional costs,as shown in the intersuccess;
anpopulationgenetics,and studies of human racial
view
with
a
shared
wifein the i982 AustralianfilmFirst
Anthropology
Physical
of
Journal
American
cestry.
Contact,may be high.
44:3I-49.
Benefitto males has receivedmorethoroughexamination. Followingthe lead of Trivers(I972), notions such
i. ? I988 by The Wenner-Gren
FoundationforAnthropological
Research.All rightsreservedOOII-3204/88/2902-0006$I.00.
ofpolyandry
is morespecific.In biolog2. The biologicaldefinition
ical terms"cooperativepolyandry"
is the union of two or more
males with a singlefemaleduringa singlebreedingseason and
theircooperationin raisingthe subsequentoffspring
(Faaborgand
Pattersoni 981).