Journal of Archaeological Science 34 (2007) 1649e1658http://www.elsevier.com/locate/jas

Artificial cranial deformation in South America: a geometricmorphometrics approximation

S. Ivan Perez*

CONICET, Division Antropologıa, Facultad de Ciencias Naturales y Museo, Universidad Nacional de La Plata,Paseo del Bosque s/n, 1900 La Plata, Argentina

Received 17 August 2006; received in revised form 16 November 2006; accepted 5 December 2006

Abstract

The bioarchaeological record of South America is characterized by the high frequency of individuals with a variety of cranial deformationsconcentrated in three areas (North-West, Central-West and South) of this subcontinent. The general purpose of this paper is to study the variationin artificial cranial deformation in several regions of Central-West and South of South America. Cranial variation related to artificial deformationof human cranial remains is analyzed by means of geometric morphometrics and multivariate statistic methods. The results of this work showthat there are no large differences in variation among states, chiefdoms and bands of hunter-gatherers. The pattern of variation observed incranial deformation among regions can be interpreted principally according to the chronological and spatial distribution of the cranial samplesanalyzed.� 2006 Elsevier Ltd. All rights reserved.

Keywords: Bioarchaeology; Cultural cranial variation; Cluster identification; Semilandmarks

1. Introduction

The practice of altering the normal growth and develop-ment of the head during the first few years of life throughapplication of external forces was a widespread culturalphenomenon in modern humans (Buikstra and Ubelaker,1994; Dembo and Imbelloni, 1938; Dingwall, 1931; Ubelaker,1984). This practice, named artificial cranial deformation, canbe intentional (to produce a specific shape), or unintentional(the by-product of other behavior) (Brothwell, 1981; Ubelaker,1984). While in the first case the deformation is generally usedas indicative of group identity (Blom, 2005a; Getszten, 1993;Munizaga, 1992; Torres-Rouff, 2002, 2003), unintentional de-formation is mainly the consequence of the use of a board orband to swaddle the infant to the cradle (Brothwell, 1981). De-spite the differences in degree of intentionality, such activities

* Tel.: þ54 221 425 9819.

E-mail address: iperez@fcnym.unlp.edu.ar

0305-4403/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jas.2006.12.003

result in a modified head shape that is associated with somecultural practices performed during infancy (Buikstra andUbelaker, 1994; Brothwell, 1981; Torres-Rouff, 2003;Ubelaker, 1984), and therefore carry some social significance(Torres-Rouff, 2003).

The bioarchaeological record of South America is charac-terized by the high frequency of individuals with a varietyof cranial deformations (Imbelloni, 1933a) concentrated inthree areas (North-West, Central-West and South) of thissubcontinent (Fig. 1). The classic works of Imbelloni andco-workers (Bormida, 1953e54; Dembo and Imbelloni,1938; Imbelloni, 1933a,b) established that the highest varia-tion in deformations occurs in the Central-West region, coin-ciding with the chiefdom and state of agriculturalists and/orpastoralists from the Andeans (see Raffino, 1988). In particu-lar, these researchers pointed out that the Tabular erect, Tabu-lar oblique and Annular types (in the classification of Demboand Imbelloni (1938) as well as their varieties (e.g. Annularerect and oblique) occur in this area (Blom, 2005a; Demboand Imbelloni, 1938; Imbelloni, 1933a,b; Munizaga, 1992;

1650 S.I. Perez / Journal of Archaeological Science 34 (2007) 1649e1658

Torres-Rouff, 2002, 2003; Weiss, 1962). In contrast, the bandsof hunter-gatherers from the Southern region (i.e. Pampa andPatagonia; see Borrero, 1999, 2001; Orquera, 1987; Politisand Madrid, 2001) have been characterized by low variabilityof cranial deformations, with all of them considered unin-tentional (Bormida, 1953e54; Imbelloni, 1924e25, 1933a).The different deformations found in this area were consideredvarieties of the Tabular erect type (Bormida, 1953e54; Demboand Imbelloni, 1938) generated by swaddling the infant to thecradle (although this idea was discussed more recently byMendonca et al., 1988e89).

The analysis of the variation in artificial cranial deforma-tion was a relevant issue of earlier anthropological work. Inthe late nineteenth and early twentieth centuries numerousstudies were carried out to construct classifications of cranialdeformation (e.g. Broca, 1878, 1879; Dembo and Imbelloni,1938; Gosse, 1855, 1861; Hrdlicka, 1912; Imbelloni, 1924e25;Neuman, 1942; Topinard, 1879). These traditional classifica-tions were produced mainly by pooling crania that presentedsimilarities in external vault morphology (Imbelloni, 1924e25, 1933b) and using a typological (Hull, 1992; Hunt,1981; Sober, 1992) approach. This procedure is subject tosignificant observer error, and it has limited capacity for

Fig. 1. Geographical distribution of cranial deformations in South America

(gray shading; modified after Imbelloni, 1933a) and samples analyzed in

this work (abbreviations as in Table 1).

the study of morphological variation. In the first half of thetwentieth century, morphometric procedures were developed(based in planes, lineal measurement and angles), alongwith new classifications based on inferring the cause of de-formation (e.g. Dembo and Imbelloni, 1938; Imbelloni,1924e25). In addition to building classifications, the earlystudies were focused in the use of cranial deformation as cul-tural trait in the analysis of ethnographic and prehistoricgroups (Bormida, 1953e54; Imbelloni, 1924e25, 1933a,b;Weiss, 1962, among others), as well as in the influence ofcranial deformation on racial classifications (Bormida,1953e54; Dembo and Imbelloni, 1938). Likewise, more re-cent studies have been mainly interested in both the effectsof cranial vault deformation on craniofacial growth and de-velopment (e.g. Anton, 1989; Anton and Weinstein, 1999;Cheverud et al., 1979, 1992; Corruccini, 1976; Friess andBaylac, 2003; Kohn et al., 1993; Konigsberg et al., 1993;Richtsmeier et al., 1984), and the use of deformation as in-dicative of group identity (e.g. Blom, 2005a; Munizaga,1992; Torres-Rouff, 2002, 2003). However, although newand more comprehensive morphometric approximationshave been applied, in particular to the study of the effectsof cranial deformation over craniofacial anatomy (e.g.Cheverud et al., 1992; Friess and Baylac, 2003; Kohnet al., 1993), the analyses in these studies have been basedon the traditional classification. In like manner, the recentinvestigations on artificial cranial deformation in the Central-West region of South America from a bioarchaeologicalperspective (e.g. Blom, 2005a,b; Munizaga, 1992; O’Brien,2003; O’Brien and Sanzetenea, 2002; Torres-Rouff, 2002,2003) has been addressed from a typological perspective,using the traditional classification (e.g. Dembo and Imbelloni,1938; Imbelloni, 1924e25; Steward, 1973) to examine broadregional differences.

The general goal of this paper is to study the variation inartificial cranial deformation in several regions of Central-Western and Southern South America (Fig. 1). This workaims specifically at: (a) analyzing vault variation generatedby cultural factors in samples that represent prehistoric popu-lations of band of hunter-gatherers and chiefdoms and state ofagriculturalists and/or pastoralists; (b) grouping individualswith similar culturally modified vault morphology withineach sample; and (c) comparing cultural variation among sam-ples. Cranial variation related to artificial deformation of hu-man cranial remains will be analyzed by means of geometricmorphometrics (Adams et al., 2004; Marcus, 1990; Rohlf,1990; Rohlf and Marcus, 1993) and multivariate statisticalmethods. The fundamental advantage of geometric morpho-metrics over traditional approaches (Marcus, 1990; Reymentet al., 1984) is in the development of powerful statisticalmethods based on models for shape variation of an entire con-figuration of points corresponding to the locations of morpho-logical landmarks (Bookstein, 1991; Rohlf, 1999, 2000) andsemilandmarks (Bookstein, 1997). Thus, it will be possibleto contrast previous studies dealing with the variation of cra-nial deformations in broad regional scale with a new, moreobjective and powerful procedure.

1651S.I. Perez / Journal of Archaeological Science 34 (2007) 1649e1658

2. Materials and methods

The samples of human remains included in this study cor-respond to adult (ca. 18 to 45 years) male and female individ-uals from several South American regions (Fig. 1, Table 1).Sex and age estimation were made on the basis of cranialand pelvic features (Buikstra and Ubelaker, 1994). The sam-ples analyzed belong to late Holocene prehistoric populationswith differences in subsistence modes, hierarchy and social or-ganization (Table 1). While the groups from Andes (B, J-S andNC) are agriculturalists and/or pastoralists with chiefdom orstate organization, the Patagonian (NR, ChR and SP), SouthCuyo (SC) and Chaco (Ch-E) groups are bands of hunter-gatherers.

For morphometric analyses the individuals were positionedin the Frankfurt plane and digital images were obtained fromthe crania in lateral (left side) norm with an Olympus CamediaC-3030 digital camera. Coordinates for two landmarks (-)and 78 semilandmarks (C) were recorded on the lateralnorm of the crania (Fig. 2). Alignment ‘‘fans’’ at equal angulardisplacements along a curve were placed using the MakeFan6software (Sheets, 2003), and the coordinates of landmarks andsemilandmarks were recorded by means of the tpsDIG 1.40software (Rohlf, 2004).

Two series of observations were performed on a sample of15 crania selected by means of random numbers from theDivision Antropologıa of the Facultad de Ciencias Natural yMuseo (UNLP) collection, with the purpose of evaluatingintraobserver reliability for recording the coordinates of land-marks and semilandmarks used in this study. These series ofobservations were separated by a 15-day interval. Analysisof intraclass correlation was performed on these coordinatesto assess the degree of concordance between series (Shroutand Fleiss, 1979; Zar, 1999).

The effects of location, scaling and orientation in landmarkconfigurations were removed using Generalized Procrustesanalysis (GPA; Bookstein, 1991; Rohlf and Slice, 1990).

Semilandmarks of the crania in lateral norm were aligned bymeans of the sliding semilandmark method proposed byBookstein (1997). This operation extends the GPA in orderto slide the semilandmark points along the outline curve untilthey match as well as possible the positions of correspondingpoints along an outline in a reference specimen (Sheets et al.,2004). This is done because the curves or contours should behomologous from subject to subject, whereas their individualpoints are not (Bookstein et al., 2002). The remaining variationamong individuals is considered as difference in shape. In thisstudy we used the Semiland6 software (Sheets, 2003) to alignthe semilandmarks along their respective curves, sliding themalong it, in order to minimize the Procrustes distance betweenthe subject and a reference (Sampson et al., 1996; Sheetset al., 2004).

Fig. 2. Landmarks (-) and semilandmarks (C) recorded from the craniofa-

cial structure.

Table 1

Samples analyzed

Samples Abbr. N\ N_ Region Years BPa

Bolivia1 B 26 36 Altiplano 500e1000

Jujuy-Salta1,2 J-S 33 35 Tilcara, La Poma and La Paya 500e1000

North Cuyo1,2 NC 18 20 Calingasta, Angualasto,

Pachimoco and Jachal

600e800

South Cuyo3 SC 9 23 Atuel river 350e800

Negro River1,2 NR 58 74 Negro river (nearby to the coast) 350e600

2000e3700b

Chubut River1 ChR 66 78 Chubut river (nearby to the coast) 500e1100

1500e2000b

South Patagonia1,2,4,5,6 SP 25 33 NW of Santa Cruz, Magallanes

and Rıo Gallegos

350e1100

Chaco-Entre Rıos1,2 Ch-E 10 17 Formosa, Chaco and Entre Rıos 200e1000

Total 245 316

1Division Antropologıa, Museo de La Plata (La Plata); 2Museo Etnografico (Buenos Aires); 3Museo Municipal de San Rafael (Mendoza); 4Instituto de la Patagonia

Austral (Punta Arenas); 5Instituto Nacional de Antropologıa y Pensamiento Latinoamericano, (Buenos Aires); 6Museo Regional Provincial ‘‘Padre Manuel Jesus

Molina’’ (Rıo Gallegos).a Approximate date established in base to radiocarbon dating and contextual information.b Some areas present two moments with different cranial deformations.

1652 S.I. Perez / Journal of Archaeolog

Shape variation within each sample was evaluated usingFoote’s (1993) disparity measurement. This is defined as mor-phological disparity D ¼

P(di

2)/(n � 1) where di representsthe distance of the specimens to the group centroid (Zelditchet al., 2004). Disparity was measured using DisparityBox6software (Sheets, 2003), which uses the Partial Procrustes dis-tance as di measure.

The method of Relative Warps was used for the comparisonof the configurations of landmarks and semilandmarks (Book-stein, 1991). Relative warps are principal components of thepartial warps plus component uniform vectors (Bookstein,1991; Rohlf, 1993), and were used to describe the major trendsin shape variation among and within samples. The partial warpscores are components along the orthogonal eigenvectors ofthe bending energy matrix and describe non-affine patternsof shape difference (Bookstein, 1989, 1991; Rohlf, 1993,1996), whereas the uniform components describe affine shapedifferences (Bookstein, 1996; Rohlf and Bookstein, 2003). Animportant aspect of this analysis is that the results of statisticalanalysis can be expressed as an intuitive deformation grid di-agram of each case with respect to the mean form or reference.The analyses were made using the Relative warps 1.40 soft-ware (Rohlf, 2004). The value of the alpha parameter was0 (zero), thus including variation at all scales.

A Kernel analysis (Baxter et al., 1997; Browman andFoster, 1963; Wand, 1994; Wand and Jones, 1995) was usedto investigate the distribution of the shapes related to cranialdeformation in the morphometric plane generated by the Rel-ative Warps analysis. This technique provides a smoother rep-resentation of the data and can be used as an informal kind ofclustering method (Baxter et al., 1997). Kernel analysis wasperformed using KernSmooth 2.22 package for R 1.9.1 (Ihakaand Gentleman, 1996).

Finally, the variation among the consensus (mean shape;see Bookstein, 1991) of the groups defined by the RelativeWarps and Kernel analyses of each sample was analyzed bymeans of Relative Warps analysis. The patterns of ordinationproduced by this analysis were compared with the geographiclocation of the samples using the Procrustes method (Digbyand Kempton, 1987; Gower, 1971; Peres-Neto and Jackson,2001). This method scales and rotates the ordinations, usinga minimum squared differences criterion, in order to find anoptimal superimposition that maximizes their fit. The sum ofthe squared residuals between configurations in their optimalsuperimposition can then be used as a measurement of associ-ation (Gower, 1971). A permutation procedure (PROTEST;10,000 permutations) was then used to assess the statisticalsignificance of the Procrustean fit (Peres-Neto and Jackson,2001). Procrustes analysis was made using vegan 1.4.4 pack-age for R 1.9.1 (Ihaka and Gentleman, 1996).

3. Results

The digitization of landmarks and semilandmarks showsexcellent consistency (Fleiss, 1981) between observationalseries (ICC ¼ 1.00).

The samples from Bolivia, Jujuy-Salta and Negro Riversamples showed the highest values of disparity when within-sample variation was estimated (Table 2). These values aretwo (Jujuy-Salta and Negro River samples) and three (Boliviasample) times higher than those for the undeformed sample(Chaco-Entre Rıos) and for the samples with deformed and un-deformed crania from Cuyo, Chubut River and South Patago-nia (this latter sample presented few cases with evidence ofdeformation, most of them coming from the north of SantaCruz province, Argentina).

The scores of the two first relative warps of the cranial vaultin lateral norm are illustrated only for male individuals of foursamples (two samples of Andean region and other two samplesof Patagonian region) in Figs. 3 and 4. The results for femaleindividuals are not presented here, but similar findings wereobtained for both sexes. Interestingly, the patterns of concen-trations of points were different among the four samples ill-ustrated (Figs. 3 and 4). Three clearly differentiated sets,grouping concentration of morphologies, were distinguishedfor the Bolivia sample using kernel density contours(Fig. 3a). The largest group (ND) corresponds to the unde-formed crania. The grid that corresponds to the first groupinterpretable as deformed crania (B1) shows intense compres-sion in both occipital (around the nuchal crest area) and frontalregions, as well as expansion in the posterior area of the pari-etal bones. In contrast, the grid corresponding to the othergroup interpretable as deformed crania (B2) presents slightcompression in both the occipital and frontal regions. Clearlydifferent results are obtained when the deformation of theJujuy-Salta sample is analyzed. In this case no clearly diffe-rentiated groups of concentration of morphologies can bedistinguished using kernel density contours (Fig. 3b). The indi-viduals that correspond to the undeformed (ND) and deformed(J-S) crania are not differentiated, presenting a continuous dis-tribution along the first relative warps. The grids correspondingto the individuals interpretable as deformed (J-S) present slightcompression in both occipital and frontal regions.

Table 2

Foote’s measurement of within-sample disparity for the regions analyzed

Samples Disparity

Bolivia \ 0.0024

Bolivia _ 0.0027

Jujuy-Salta \ 0.0020

Jujuy-Salta _ 0.0016

North Cuyo \ 0.0013

North Cuyo _ 0.0009

South Cuyo \ 0.0013

South Cuyo _ 0.0011

Negro River \ 0.0018

Negro River _ 0.0018

Chubut River \ 0.0008

Chubut River _ 0.0009

South Patagonia \ 0.0010

South Patagonia _ 0.0011

Chaco-Entre Rıos \ 0.0010

Chaco-Entre Rıos _ 0.0009

ical Science 34 (2007) 1649e1658

Fig. 3. Relative Warps and Kernel analyses performed using the Bolivia (a) and Jujuy-Salta (b) male skull samples.

1653S.I. Perez / Journal of Archaeological Science 34 (2007) 1649e1658

Fig. 4 shows the results for two Patagonian samples. Fig. 4ashows three groups of concentration of cultural morphologies(NR1, NR2 and NR3) and the undeformed group (ND) forNegro River sample. These groups are distributed more

continually that the Bolivia sample. The grid correspondingto the first group interpretable as deformed (NR1) presentscompression in both occipital (around the nuchal crest area)and frontal regions, as well as slight expansion in the posterior

Fig. 4. Relative Warps and Kernel analyses performed using the Negro River (a) and Chubut River (b) male skull samples.

1654 S.I. Perez / Journal of Archaeological Science 34 (2007) 1649e1658

area of parietal bones. The grid corresponding to the secondgroup interpretable as deformed (NR2) presents slight com-pression in both occipital and frontal regions. On the otherhand, the NR3 group presents compression in the lambdaarea. Finally, the Kernel analysis does not allow to findwell-defined groups of morphologies within the Chubut Riversamples (Fig. 4b). The deformed cases (ChR1 and ChR2) aredistributed around the undeformed morphologies (ND). Thegrid corresponding to the individuals interpretable as de-formed in the negative values of relative warp 1 present slightcompression in the frontal region (ChR1), whereas the individ-uals interpretable as deformed in the positive values of relativewarp 1 present slight compression in the lambda area (ChR2).

Fig. 5 shows the comparison of the consensus for culturallygenerated cranial morphologies for all the samples (the firsttwo relative warps comprise 88.3% of the total variationamong samples). Fig. 5 shows that the cases with slight com-pression of the frontal bone only occur in the Chubut Riversample. The samples with compression at the lambda regionare distributed in Central and Northern Patagonia (ChR2 andNR3) and Southern Cuyo (SC). On the other hand, the caseswith compression in the frontal and occipital regions occurin the Bolivia, Jujuy-Salta, Negro River and Northern Cuyosamples. The latter sample presents an intermediate

Fig. 5. Relative Warps analyses performed using the consensus of cranial

deformation groups.

morphology between the fronto-occipital compression andlambda compression morphologies. The samples from NegroRiver and Bolivia show deformations with compression inboth occipital and frontal regions, as well as an expansion inthe posterior area of the parietal bones. Finally, the patternof variation in cranial deformations among samples of laterlate Holocene is mainly related with their geographic distribu-tion (Fig. 6). The values of association between these variablesis higher for male samples (correlation-like statistic derivedfrom the Procrustes fit m12 ¼ 0.88, p ¼ 0.001; Fig. 6a) thanfor female samples (m12 ¼ 0.80, p ¼ 0.016; Fig. 6b).

4. Discussion

Artificial cranial deformation has been analyzed mainly bymeans of visual and traditional morphometrics (based on linearmeasurement and angles) techniques (e.g. Anton, 1989; Blom,2005a,b; Bormida, 1953e54; Dembo and Imbelloni, 1938;Neuman, 1942; O’Brien, 2003; O’Brien and Sanzetenea,2002; Torres-Rouff, 2002, 2003; Weiss, 1962). Traditionalmorphometrics (e.g. Anton, 1989; Anton and Weinstein,1999; Brown, 1981; Moss, 1958), as well as some earlier geo-metric morphometrics techniques (i.e. Elliptic Fourier andFinite Element analyses; e.g. Cheverud et al., 1992; Friessand Baylac, 2003; Kohn et al., 1993), have been employedby several researchers for the study of the relationship betweenartificial cranial deformation and craniofacial growth anddevelopment. In these works the morphometric informationhas been used as dependent variable and traditional classifica-tions (e.g. Dembo and Imbelloni, 1938; Steward, 1973) havebeen employed as the independent variable, without previouslyestablishing an objective independent grouping. Similarly,recent analyses of cranial deformation in the context ofbioarchaeological studies have based their assessments uponvisual determination. South American studies have mostlyused the Imbelloni classification (e.g. Blom, 2005a,b;Munizaga, 1992; O’Brien, 2003; O’Brien and Sanzetenea,2002; Torres-Rouff, 2002, 2003).

The semilandmarks-based geometric morphometric andmultivariate analyses of cranial vault in the present workhave shown the existence of different cultural morphologiesin South America. These morphologies correspond to lambdicand antero-posterior compression, as well as antero-posteriorcompression plus superior expansion (Figs. 3e5). The resultsobtained generally agreement with previous descriptions ofcranial deformations in this geographic area that were madeusing morphoscopic techniques (Bormida, 1953e54; Demboand Imbelloni, 1938; Imbelloni, 1924e25). However, somedifferences between the results of the present work and previ-ous ones can be highlighted. In the first place, although themethods used to deform the human cranium to achieve a partic-ularly desired shape may be similar within human groups andvery different among them, the final shape of crania may varydue to other factors, such as osseous remodelation and varia-tions in the degree and duration of the of the deformation prac-tice performed on each individual, among others. Likewise,due to the fact that the degree of standardization of the

1655S.I. Perez / Journal of Archaeological Science 34 (2007) 1649e1658

Fig. 6. Procrustes fit of geographic coordinates (circle) onto the position of female (a) and male (b) cranial deformations on relative warps 1 and 2.

deformation can vary among regions and human groups due tocultural causes, the ability to measure such differences is rel-evant for anthropological research. For instance, the differ-ences observed in the pattern of variation between Boliviaand Negro River regions can be due to the higher standardiza-tion of cranial deformation in the former, which could there-fore be related to social and cultural differences betweenregions. These subtle differences in the resultant cranial defor-mation can only be studied by means of quantitative methods.In the second place, the recognition of discrete groups with as-sociated cultural meaning (in the way the cranial deformationtypes were traditionally defined) is very subjective and it issubject to significant observational errors. In some areas thereare clearly differentiated cultural groups (e.g. Bolivia andNegro river regions) whereas in other areas this differentiationis not present (e.g. Jujuy-Salta and Chubut river regions). Theuse of quantitative methods is a useful approach to control ob-servational error, reducing the subjectivity of the cranial defor-mation classification. This work uses some of the possibleapproximations; particularly in some samples (e.g. Boliviaand Negro River) the criterion for establishing groups waschosen on the basis of sets of the most densely clustered pointsin the multivariate space (Baxter et al., 1997), while in othersamples groups were divided by choosing a value along rela-tive warp 1. However, the delimitation of groups with similarmorphology is not a simple task and the criteria to be appliedrequire further discussion. Finally, the cultural pattern of var-iation among regions has not been analyzed satisfactorily inpast studies. The use of quantitative methods in this work al-lows obtaining a pattern of variation in cultural characteristicsthat can be used together with several models of cultural trans-mission and evolution (Cavalli-Sforza, 2000).

The results of this work show that there are no large dif-ferences in the variations of artificial cranial deformation inCentral-West and Southern South America among states,chiefdoms and bands of hunter-gatherers. In particular, al-though the Bolivian altiplano region shows the greatest varia-tion in morphologies produced by cultural deformation, it issimilar to the one observed in Negro River region (Table 2).However, these areas can be distinguished by the fact that

the Bolivian altiplano sample shows more differentiatedgroups, whereas the Negro River sample shows a greater num-ber of less differentiated- groups (Figs. 3a, 4a and 5). On theother hand, although the samples from Chubut River andJujuy-Salta did not show a clear boundary between deformedand undeformed individuals, the latter shows greater variation(Figs. 3b and 4b; Table 2). Thus, my results do not support theview held by Imbelloni and co-workers, which emphasized theexistence of large differences in the variation occurring inchiefdoms and states with respect to bands of hunter-gatherers(Bormida, 1953e54; Imbelloni, 1924e25, 1933a). These in-vestigators pointed out that the hunter-gatherers from Patago-nia present only one culturally modified morphology (Tabularerect) with several variants (plano-lambdic, plano-frontal andpseudocircular), and that these are only unintentional deforma-tions derived form using a table or band to swaddle the infantto the cradle (Bormida, 1953e54; Imbelloni, 1924e25). Theattribution of low cultural (although not morphological) varia-tion to Patagonian groups was primarily the result of the eth-nological concepts extracted from Kulturkreislehre thinking(see Lowie, 1946).

The pattern of variation in cranial deformation observedamong regions can be interpreted according to the chronolog-ical and spatial distribution of the cranial samples analyzed.All the samples from the Andean region belong to the laterlate Holocene (ca. 500e1000 years BP), but only the individ-uals with compression in the lambda region have been dated tolater late Holocene in Patagonia and Southern Cuyo (ca. 350e1100 years BP; Beron and Baffi, 2003; Perez, 2006). The otherdeformations from Patagonia have been radiocarbon-dated toearly late Holocene (ca. 1500e3700 years BP; Beron andBaffi, 2003; Perez, 2006). Thus, the variation observed withinthe Patagonian region is to some extent related with temporaldifferences. In contrast, the variation in deformations amonglater late Holocene samples is principally related to the geo-graphic distribution of the samples. The Procrustes analysesshow that the morphometric variation among cranial deforma-tions is associated with the geographic distance separatingthem (Fig. 6). In particular, male samples display higher asso-ciation than female samples, with females ChR2, NR3 and SC

1656 S.I. Perez / Journal of Archaeological Science 34 (2007) 1649e1658

showing greater difference from NC, J-S and B2 (Fig. 6a) thanthe one found among males (Fig. 6b).

The spatial pattern found could be explained by means ofa model of isolation by distance, which has been successfullyapplied to other cultural traits (Cavalli-Sforza, 2000). Thismathematical model, originally developed to explain biologi-cal variation, has three relevant parameters in the case of cul-tural phenomena; of these, the first two are responsible for theincrease of traits differences: innovation rate (equivalent tomutation rate in biology), sampling or coping error (equivalentto genetic drift in biology), as well as the migration rate of thetraits between neighborsdwhich decreases trait differences(Cavalli-Sforza, 2000). However, this model cannot explainall the variation observed in later late Holocene, because animportant difference in cranial deformations between neigh-bors is observed in the Bolivian altiplano (Fig. 6). The pres-ence of different variety of cultural deformations in thisregion has been interpreted as an ethnic marker in this area(Blom, 2005a). An additional interruption in the geographicordination of the samples is observed in the Cuyo region(Fig. 6). The use of cranial deformation as an ethnic markerprobably takes place also in the early late Holocene of Patago-nia and nearby areas, where there is archaeological evidenceof an increase of territoriality (Madrid and Barrientos,2000). The increase of territoriality is concordant with greaterdiversity of cranial deformations in Patagonia (ChR1, NR1and NR2). This diversity contrasts with the homogeneity incranial deformation observed in the later late Holocene ofPatagonia (Fig. 5), a period marked by greater cultural ho-mogeneity in the area, probably caused by the expansion ofNorth Patagonian populations to nearby regions (Barrientosand Perez, 2004).

5. Conclusion

The research performed after the works of Imbelloni andco-workers (Bormida, 1953e54; Dembo and Imbelloni,1938; Imbelloni, 1933a) have mainly discussed the presenceor absence of the types defined by these authors in differentareas. This approximation was not particularly useful. Thecontinuous nature of the morphologies analyzed in some areas,the different expression of these deformations among regions,plus the high subjectivity involved in the determination of cra-nial deformations, make the study of cranial deformation usingmorphoscopic and typological approach way too narrow fora modern anthropological perspective. The application of geo-metric morphometrics and multivariate techniques providesa wider approach to analyze the vault variation generated bycultural factors, cluster individuals with similar culturallymodified vault morphology within each sample, and comparecultural variation among samples. In particular, the semiland-marks-based geometric morphometric and multivariate analy-ses of cranial vault used in the present work allowed to findsubtle differences in the cranial deformation between samples,as well as to recognize discrete groups of individuals with sim-ilar deformation in a lesser subjective way that the previous

studies, and then analyze quantitatively cultural patterns ofvariation among regions.

The analyzed area, the Central-West and South of SouthAmerica, was characterized by a diversity of cranial deforma-tions. The morphologies found in this area correspond tolambdic and antero-posterior compression, as well as antero-posterior compression plus superior expansion. The resultsof the present work show that there are no great differences re-garding the overall variation of artificial cranial deformationbetween states-chiefdoms and bands of hunter-gatherers, likeit was pointed out previously (Bormida, 1953e54; Demboand Imbelloni, 1938; Imbelloni, 1933a,b). However, theseareas can be distinguished by the fact that the sample fromthe Andeans state (i.e. Bolivian altiplano) show more definedgroups that the Patagonian hunter-gatherers which display anmore homogeneous dispersion on the shape space (i.e. NegroRiver). Finally, the pattern of variation in cranial deformationobserved among regions can be interpreted according to thesome chronological differences in the Patagonian samples, togeographical distribution during the later late Holocene andto ethnic separation for Bolivia and Cuyo-Patagonian bound-ary as well.

Acknowledgements

I am sincerely grateful to Valeria Bernal and Paula Gonza-lez for their comments that greatly improved the quality of themanuscript. I thank Hector M. Pucciarelli [Division Antropo-logıa. Facultad de Ciencias Naturales y Museo, La Plata (Ar-gentina)], Ines Baffi and Leandro Luna [Museo Etnografico‘‘J. B. Ambrosetti’’, Buenos Aires (Argentina)], Paula Novel-lino [Museo Municipal de San Rafael (Mendoza, Argentina)],Mateo Martinic [Instituto de la Patagonia Austral (PuntaArenas, Chile)], Rafael Goni [Instituto Nacional de Antropolo-gıa y Pensamiento Latinoamericano (Buenos Aires, Argen-tina)], and the staff at Museo Regional Provincial ‘‘PadreManuel Jesus Molina’’ (Rıo Gallegos, Argentina) for grantingaccess to the human skeletal collections under their care. I alsothank to Amelia Barreiro and Cecilia Morgan for help me withthe English version of the manuscript. Marina Perez made thedrawings for Figs. 1 and 2. Grant sponsor: Universidad Nacio-nal de La Plata. Grant sponsorship: Doctoral FellowshipArgentina Government (CONICET).

References

Adams, D.C., Rohlf, F.J., Slice, D.E., 2004. Geometric morphometrics: ten

years of progress following the ‘revolution’. Italian Journal of Zoology

71, 5e16.

Anton, S.C., 1989. Intentional cranial vault deformation and induced changes

of the cranial base and face. American Journal of Physical Anthropology

79, 253e267.

Anton, S.C., Weinstein, K.J., 1999. Artificial cranial deformation and fossil

Australians revisited. Journal of Human Evolution 36, 195e209.

Barrientos, G., Perez, S.I., 2004. La expansion y dispersion de poblaciones del

norte de Patagonia durante el Holoceno tardıo: evidencia arqueologica y

modelo explicativo. In: Civalero, T., Fernandez, P., Guraieb, G. (Eds.),

Contra viento y marea. Arqueologıa de Patagonia, Buenos Aires, pp.

179e195.

1657S.I. Perez / Journal of Archaeological Science 34 (2007) 1649e1658

Baxter, M.J., Beardah, C.C., Wright, R.V.S., 1997. Some archaeological appli-

cations of kernel density estimates. Journal of Archaeological Science 24,

347e354.

Beron, M.A., Baffi, E.I., 2003. Procesos de cambio cultural en los cazadores-

recolectores de la provincia de La Pampa, Argentina. Intersecciones en

Antropologıa 4, 29e43.

Blom, D.E., 2005a. A bioarchaeological approach to Tiwanaku group dynamics.

In: Reycraft, R. (Ed.), Us and Them: Archaeology and Ethnicity in the An-

des. Cotsen Institute of Archaeology Press, Los Angles CA, pp. 153e182.

Blom, D.E., 2005b. Embodying borders: human body modification and diversity

in Tiwanaku society. Journal of Anthropological Archaeology 24, 1e24.

Bookstein, F.L., 1989. Principal warps: Thin-plate splines and the decomposi-

tion of deformations. IEEE Transactions on Pattern Analysis and Machine

Intelligence 11, 567e585.

Bookstein, F.L., 1991. Morphometric Tools for Landmark Data: Geometry and

Biology. Cambridge University Press, Cambridge.

Bookstein, F.L., 1996. Biometrics, biomathematics and the morphometric syn-

thesis. Bulletin of Mathematical Biology 58, 313e365.

Bookstein, F.L., 1997. Landmark methods for forms without landmarks:

Localizing group differences in outline shape. Medical Image Analysis

1, 225e243.

Bookstein, F.L., Streissguth, A.P., Sampson, P.D., Connor, P.D., Barr, H.M.,

2002. Corpus callosum shape and neuropsychological deficits in adult

males with heavy fetal alcohol exposure. NeuroImage 15, 233e251.

Bormida, M., 1953e54. Los antiguos patagones. Estudio de craneologıa, Runa

6 (1e2), 55e96.

Borrero, L.A., 1999. The prehistoric exploration and colonization of Fuego-

Patagonia. Journal of World Archaeology 13 (3), 321e355.

Borrero, L.A., 2001. Ambios, continuidades, discontinuidades: discusiones

sobre arqueologıa Fuego-Patagonica. In: Berberian, E.E., Nielsen, A.E.

(Eds.), Historia Argentina Prehispanica T. 2. Brujas, Cordoba, pp. 815e838.

Buikstra, J., Ubelaker, D., 1994. Standards for Data Collection From Human

Skeletal Remains. Arkansas Archaeological Survey Research Series 44.

Broca, P., 1878. Sur des cranes et des objets dindustrie provenant des fouilles

de M. Ber a Tiahuanaco (Perou). B.S.A.P 3-1, 230e235.

Broca, P., 1879. Sur un mode peu connu de deformation toulousaine. B.S.A.P

3-2, 699e701.

Brothwell, D.R., 1981. Digging Up Bones. Cornell University Press, New

York.

Browman, A., Foster, P., 1963. Density based exploration of bivariate data.

Statistics and Computing 3, 171e177.

Brown, P., 1981. Artificial cranial deformation: a component in Pleistocene

Australian Aboriginal crania. Archaeology in Oceania 16, 156e167.

Cavalli-Sforza, L.L., 2000. Genes, Peoples, and Languages. University of Cal-

ifornia Press, London.

Cheverud, J.M., Buikstra, J.E., Twichell, E., 1979. Relationship between non-

metric skeletal traits and cranial size and shape. American Journal of

Physical Anthropology 50, 191e198.

Cheverud, J.M., Kohn, L.P., Konigsberg, L.W., Leigh, S.R., 1992. Effects of

fronto-occipital artificial cranial vault deformation on cranial base and

face. American Journal of Physical Anthropology 88, 323e345.

Corruccini, R.S., 1976. The interaction between nonmetric and metric cranial

variation. American Journal of Physical Anthropology 44, 285e294.

Dembo, A., Imbelloni, J., 1938. Deformaciones Intencionales del Craneo de

Caracter Etnico. J. Anesi, Buenos Aires.

Digby, P.G.N., Kempton, R.A., 1987. Multivariate Analysis of Ecological

Communities. Chapman and Hall, London.

Dingwall, E.J., 1931. Artificial Cranial Deformation. London.

Fleiss, J.L., 1981. Statistical methods for rates and proportions, second ed.

John Willey, New York.

Friess, M., Baylac, M., 2003. Exploring artificial cranial deformation using

Elliptic Fourier analysis of Procrustes aligned outlines. American Journal

of Physical Anthropology 122, 11e22.

Foote, M., 1993. Contributions of individual taxa to overall morphological

disparity. Paleobiology 19, 403e419.

Getszten, P.C., 1993. An investigation into the practice of cranial deformation

among the pre-Columbian peoples of northern Chile. International Journal

of Osteoarchaeology 3, 87e98.

Gosse, L.A., 1855. Essai sur les deformations artificielles du crane. Annales

dHygiene publique et de Medicine legale 2e3, 317e393.

Gosse, L.A., 1861. Questionnaire relatif aux deformations artificielles du

crane. B.S.A.P 2, 101e104.

Gower, J.C., 1971. Statistical methods of comparing different multivariate

analyses of the same data. In: Hodson, F.R., Kendall, D.G., Tautu, P.

(Eds.), Mathematics in the Archaeological and Historical Sciences. Edin-

burgh University Press, Edinburgh, pp. 138e149.

Hrdlicka, A., 1912. Artificial deformations of the human skull with special

reference to America. Actas del Congreso Internacional de Americanistas,

147e149.

Hull, D., 1992. The effects of the essentialism on taxonomy. Two thousand

years of stasis. In: Ereshefsky, M. (Ed.), The Units of Evolution.

Essays on the Nature of Species. The MIT Press, Cambridge, MA,

pp. 199e225.

Hunt, E., 1981. The Old Physical Anthropology. American Journal of Physical

Anthropology 56, 339e346.

Ihaka, R., Gentleman, R., 1996. R: a language for data analysis and graphics.

Journal of Computational and Graphics Statistics 5, 299e314.

Imbelloni, J., 1924e25. Deformaciones intencionales del craneo en Sudamer-

ica; polıgonos craneanos aberrantes. Revista del Museo de La Plata 28,

329e407.

Imbelloni, J., 1933a. America, cuartel general de las deformaciones. Actas del

XXV Congreso Internacional de Americanistas 1, 59e68.

Imbelloni, J., 1933b. Los pueblos deformadores de los Andes. La deformacion

intencional de la cabeza, como arte y como elemento diagnostico de las

culturas. Anales del Museo de Historia Natural 37, 209e254.

Kohn, L.P., Leigh, S.R., Jacobs, S.C., Cheverud, J.M., 1993. Effects of annular

cranial vault modification on cranial base and face. American Journal of

Physical Anthropology 90, 147e168.

Konigsberg, L.W., Kohn, L.A.P., Cheverud, J.M., 1993. Cranial deformation

and nonmetric trait variation. American Journal of Physical Anthropology

90, 35e48.

Lowie, R., 1946. Historia de la etnologıa. Fondo de Cultura Economica,

Mexico.

Madrid, P., Barrientos, G., 2000. La estructura del registro arqueologico del

sitio Laguna Tres Reyes 1 (Provincia de Buenos Aires): nuevos datos

para la interpretacion del poblamiento humano del Sudeste de la Region

Pampeana a inicios del Holoceno tardıo. Relaciones de la Sociedad Argen-

tina de Antropologıa 25, 179e206.

Marcus, L.F., 1990. Traditional morphometrics. In: Rohlf, F.J., Bookstein, F.L.

(Eds.), Proceedings of the Michigan Morphometrics Workshop. Special

Publication Number 2. The University of Michigan Museum of Zoology,

Ann Arbor, MI, pp. 77e122.

Mendonca, O.J., Bordach, M.A., Cignetti, A.N., 1988e89. La deformacion

craneana en las Lagunas (Neuquen), Relaciones de la Sociedad. Argentina

de Antropologıa (NS) 17 (2), 57e73.

Moss, J.L., 1958. The pathogenesis of artificial cranial deformation. American

Journal of Physical Anthropology 16, 269e286.

Munizaga, J.R., 1992. Antropologıa fısica de los Andes del Sur. In:

Meggers, B. (Ed.), Prehistoria Sudamericana. Nuevas Perspectivas. Tarax-

acum, Washington, pp. 65e75.

Neuman, G.K., 1942. Types of artificial cranial deformation in the eastern

United States. American Antiquity 7, 306e310.

O’Brien, T.G., 2003. Cranial microvariatin prehistoric south central andean

populations: an assesment of morphology in the Cochabamba collection,

Bolivia, Ph.D. work, Binghamton University, New York.

O’Brien, T.G., Sanzetenea, R., 2002. Deformacion craneana artificial: un estu-

dio de la coleccion de Ibarra Grasso de Cochabamba, Boletın del INIAN

Museo. Serie Arqueologica Boliviana 4 (26).

Orquera, L.A., 1987. Advances in the Archaeology of the Pampa and Patago-

nia. Journal of World Prehistory 1 (4), 333e413.

Perez, S.I., 2006. El poblamiento holocenico del Sudeste de la Region

Pampeana: un estudio de morfometrıa geometrica craneofacial. Doctoral

Thesis work. Universidad Nacional de La Plata.

Peres-Neto, P.R., Jackson, D.A., 2001. How well do multivariate data sets

match? The advantages of a Procrustean superimposition approach over

the Mantel test. Oecologia 129, 169e178.

1658 S.I. Perez / Journal of Archaeological Science 34 (2007) 1649e1658

Politis, G., Madrid, P., 2001. Arqueologıa pampeana: estado actual y perspec-

tivas. In: Berberian, E., Nielsen, A. (Eds.), Historia Argentina Prehispan-

ica. Editorial Brujas, Cordoba, pp. 737e814.

Raffino, R., 1988. Poblaciones Indıgenas de la Argentina. Tipografica Edito-

rial, Buenos Aires.

Reyment, R.A., Blackith, R.E., Campbell, N.A., 1984. In: Multivariate Mor-

phometrics, second ed. Academic Press, New York.

Richtsmeier, J.T., Cheverud, J.M., Buikstra, J.E., 1984. The relationships

between cranial metric and nonmetric traits in the rhesus macaques from

Cayo Santiago. American Journal of Physical Anthropology 64, 213e222.

Rohlf, F.J., 1990. Morphometrics. Annual Review of Ecology and Systematics

21, 299e316.

Rohlf, F.J., 1993. Relative warps analysis and an example of its application to

Mosquito wings. In: Marcus, L.F., Bello, E., Garcıa-Valdecasas, A. (Eds.),

Contributions to Morphometrics. Monografıas del Museo Nacional de

Ciencias Naturales, Madrid, pp. 132e159.

Rohlf, F.J., 1996. Mophometric spaces, shape components and the effects of

linear transformations. In: Marcus, L.F., Corti, M., Loy, A., Slice, D.,

Naylor, G. (Eds.), Advances in Morphometrics. Plenum Press, New

York, pp. 117e129.

Rohlf, F.J., 1999. Shape statistics: Procrustes superimpositions and tangent

spaces. Journal of Classification 16, 197e223.

Rohlf, F.J., 2000. Statistical power comparisons among alternative morphomet-

ric methods. American Journal of Physical Anthropology 111, 463e478.

Rohlf, F.J., 2004. TPS Series Software. Available at: http//life.bio.sunysb.edu/

morph/.

Rohlf, F.J., Bookstein, F.L., 2003. Computing the uniform component of shape

variation. Systematic Biology 52, 66e69.

Rohlf, F.J., Marcus, L.F., 1993. A Revolution in morphometrics. Tree 8-4,

129e132.

Rohlf, F.J., Slice, D.E., 1990. Extensions of the Procrustes Method for the

optimal superimposition of landmarks. Systematic Zoology 39, 40e59.

Sampson, P.D., Bookstein, F.L., Sheehan, F.H., Bolson, E.L., 1996. Eigen-

shape analysis of left ventricular outlines from contrast ventriculograms.

In: Marcus, L.F., Corti, M., Loy, A., Naylor, G.J.P., Slice, D.E. (Eds.),

Advances in Morphometrics. Nato ASI Series, Series A Life Science,

vol. 284. Plenum, New York, pp. 131e152.

Sheets, H.D., 2003. IMP Software Series. Available at: http://www.canisiu-

s.edu/wsheets/morphsoft.html/.

Sheets, H.D., Keonho, K., Mitchell, C.E., 2004. A combined landmark and

outline-based approach to ontogenetic shape change in the Ordovician

Trilobite Triarthrus becki. In: Eelewa, A. (Ed.), Applications of Mor-

phometrics in Paleontology and Biology. Springer, New York, pp.

67e81.

Shrout, P.E., Fleiss, J.L., 1979. Intraclass correlations: uses in assessing rater

reliability. Psychological Bulletin 2, 420e428.

Sober, E., 1992. Evolution, population thinking, and essentialism. In:

Ereshefsky, M. (Ed.), The Units of Evolution. Essays on the Nature of

Species. MIT Press, Cambridge, MA, pp. 247e278.

Steward, T.D., 1973. The People of America. Charles Scribner’s Sons, New

York.

Topinard, P., 1879. Des deformations ethniques du crane. R. d’A 2, 496e506.

Torres-Rouff, C., 2002. Cranial vault modification and ethnicity in middle

horizon San Pedro de Atacama, Chile. Current Anthropology 43,

163e171.

Torres-Rouff, C., 2003. Shaping Identity: Cranial Vault Modification in the

Pre-Columbian Andes, Doctoral Thesis work, University of California.

Ubelaker, D., 1984. Human Skeletal Remains. Taraxacum, Washington.

Wand, M.P., 1994. Fast computation of multivariate kernel estimators. Journal

of Computational and Graphical Statistics 3, 433e445.

Wand, M.P., Jones, M.C., 1995. Kernel Smoothing. Chapman and Hall,

London.

Weiss, P., 1962. Tipologıa de las deformaciones cefalicas de los antiguos

peruanos, segun la osteologıa cultural. Revista del Museo Nacional 31,

15e42.

Zar, J.H., 1999. Biostatistical Analisis. Prentice Hall.

Zelditch, M.L., Swiderski, D.L., Sheets, H.D., Fink, W.L., 2004. Geometric

Morphometric for Biologists: A Primer. Academic Press, London.