AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 98:173-195 (1995) Asymmetric Vault Modification in Hopi Crania LUCI ANN P. KOHN, STEVEN R. LEIGH, AND JAMES M. CHEVERUD Program in Occupational Therapy, Washington University School of Medicine, St. Louis, Missouri 63108 IL.A.I?K.); Department of Anatomy and Neurobiology, Washington University School of Medicine, St. Louis, Missouri 63110 (J.M.C.); Department ofAnthropology, University of Illinois, Urbano, Illinois 61801 (S.R.L.) KEY WORDS Cradleboard Cranial growth, Finite element scaling, Hopi, ABSTRACT Cradleboarding was practiced by numerous prehistoric and historic populations, including the Hopi. In this group, one result of cradleboarding was bilateral or asymmetric flattening of the posterior occipital. We test whether cradleboarding had significant effects on the morphology of the cranial vault, cranial base, and face. Additionally, we examine associations between direction of flattening and asymmetric craniofacial growth. A skeletal sample of Hopi from the Old Walpi site includes both nonmodified (N = 43) and modified individuals (N = 39). Three-dimensional coordinates of 53 landmarks were obtained using a diagraph. Thirty-six landmarks were used to define nine finite elements in the cranial vault, cranial base, and face. Finite element scaling was used to compare average nonmodified individuals, with averages of bilaterally, right, and left modified individuals. The significance of variation among “treatment”groups was evaluated using a bootstrap test. Pearson product-moment correlations test the association of asymmetry with direction of modification. Hopi cradleboarding has a significant effect on growth of the cranial vault, but does not affect morphology of the cranial base or face. Bilateral flattening of the cranial vault leads to decreased length and increased width of the cranial vault. Flattening of the right or left cranial vault results in ipsilaterally decreased length and width coupled with a corresponding increased length and width on the contralateral side of the cranial vault. There is a significant correlation of size asymmetry with direction of modification in the cranial vault, but not with size or shape change in the cranial base or face. 0 1995 Wiley-Liss, Inc. Numerous prehistoric and historic populations in North America (Hrdlicka, 1935; Dennis, 1940; Dennis and Dennis, 1940; Neumann, 1942; Bennett, 1973, 1975; Droessler, 1981; Heathcote, 1986; Holliday, 1993) and the Middle-East (Ewing, 1950) practiced cradleboarding. One result of this practice was the flattening of the occipital or lambdoid region (Hrdlicka, 1935; Stewart, 1937; Dennis, 1940) with either bilateral or unilateral (flattening on either the right or left sides) effects (Hrdlicka, 1935).An analy0 1995 WILEY-LISS, INC sis of crania which were modified during development provides a n opportunity to examine the growth relationships of the cranial vault, cranial base, and face. Results of previous studies have suggested that there may be a n interrelationship of growth of these Received Apnl 20, 1993; accepted March 17, 1995. Address reprint requests to Luci Ann P. Kohn, Program in Occupational Therapy, Box 8505, 4444 Forest Park Ave., Washington University School of Medicine, St. Louis, MO 63108. 174 L.A.P. KOHN ET AL. cranial regions (Young, 1959; Pucciarelli, 1978; Babler and Persing, 1982; Babler et al., 1987; Babler, 1988, 1989; Cheverud et al., 1992; Kohn et al., 1993). In this paper we evaluate whether localized modification of the cranial vault by cradleboarding during infancy has a significant effect on the morphology of the adult cranial base and face. Intentional artificial cranial vault modification has been shown to have significant effects on the growth and morphology of the cranial base and face (Cocilovo, 1975,1978). Two types of modification are generally recognized based on cranial vault morphology: antero-posterior and annular modification (Dingwall, 1931; Neumann, 1942). Anteroposterior modification was practiced by the prehistoric people from Ancon, Peru, prehistoric individuals from Makapuan, Hawaii, and the Songish from the Pacific Northwest coast. This type of modification resulted from the application of a cradleboard or headdress to the frontal and occipital regions of the cranial vault (Boas, 1921; Dingwall, 1931; Cybulski, 1975; Schendel et al., 1980; Allison et al., 1981). Anterior-posterior growth of the cranial vault is restricted, resulting in crania which appear to be short in the anterior-posterior dimension and wide in the medial-lateral dimension (Boas, 1921; Oetteking, 1930; Dingwall, 1931; Anton, 1989; Mizoguchi, 1991; Cheverud and Midkiff, 1992; Cheverud e t al., 1992). In contrast, annual modification, as practiced by the Kwakiutl and Nootka (Pacific Northwest coast), some prehistoric individuals from Peru (reported in Anton, 1989), and the Arawe (Blackwood and Danby, 1935) is produced by circumferentially wrapping the cranial vault (Dingwall, 1931). Medial-latera1 growth of the cranial vault is restricted, resulting in crania which appear to be long in the anterior-posterior dimension and narrow in the medial-lateral dimension (Oetteking, 1930; Dingwall, 1931; Cybulski, 1975; Anton, 1989; Mizoguchi, 1991; Kohn et al., 1993).Antero-posterior and annular modification produce increases in the angle between the anterior and posterior cranial base (Oetteking, 1924; McNeill and Newton, 1965; Anton, 1989). Comparable results have been produced in experimental growth studies of animals (Young, 1959; Pucciarelli, 1978) or crania in which the sutures were surgically closed prematurely (Babler and Persing, 1982; Babler et al., 1987; Babler, 1988, 1989). Antero-posterior and annular modification of the cranial vault are produced by the multi-directional application of pressure to the cranial vault. Cradleboarding produces localized, unidirectional pressure on the posterior cranial vault. As a result of differences in pressure exerted on the cranium, cradleboarding may be expected to have a different degree of effect on the morphology of the cranial base and face than has been observed in antero-posterior or annular modification. Few studies have systematically studied the effects of cradleboarding on the cranial base and face. Heathcote (1986) examined nonmodified and modified crania within populations on Kodiak and Kagamil Islands and found no evidence for a n effect of cradleboarding on dimensions outside the area of flattening within either group. Moss (1958)compared infants with “postural flattening” to nonmodified white adults in order to assess the effect of localized occipital modification on growth of the cranial base. After the statistical adjustment of infant orbital angle to that of the nonmodified adult, Moss found no difference in cranial base angle between the modified infants and the nonmodified adults. Bjork and Bjork (1964) assessed Peruvian crania for a n association of asymmetry and artificial modification. Specifically, they report a significant correlation between side of the cranial vault which was flattened (left, bilateral, or right) and the length of the cranial base and face. The cranial base and face were shorter in association with ipsilateral flattening and they were of equal length when there was bilateral flattening of the occipital. The effect of unintentional cranial vault modification on cranial morphology is of direct relevance to biological distance studies. Numerous studies have used metric traits in a n attempt to assess the degree of biological distance between prehistoric or protohistoric Indian groups within the United States based on cranial dimensions (Corruccini, 1972; Buikstra et al., 1990; El-Najjar, 1978; Droessler, 1981; Sciulli and Schneider, 1985; ASYMMETRIC VAULT MODIFICATION Sciulli, 1990).The practice of cradleboarding was particularly widespread throughout the Southwest, though the practice differed by population or by degree within a population (Hrdlicka, 1935; Stewart, 1937; Reed, 1949). The inclusion of dimensions which are a€fected by artificial modification may produce distorted estimates of biological distance between populations (Droessler, 1981). In this paper we test whether cradleboarding, which resulted in cranial modification among Hopi, had a significant effect on morphology of the cranial base and face. I n particular, we use finite element scaling to measure the effects of deformation on the differences in landmark locations. This is accomplished by comparing averages of unmodified Hopi with Hopi for whom occipital flattening has occurred on either the left side, the right side, or bilaterally. In addition, we evaluate whether or not there is significant asymmetry in cranial morphology. MATERIALS AND METHODS Samples The sample included in this study is housed a t the Field Museum of Natural History in Chicago, Illinois. All individuals in the sample were excavated from the site of Kuchaptavela, also known a s Old Walpi or Ash Hill, by C. L. Owen in 1901. The associated ceramics and ethnohistoric accounts date the site to between 1300 and 1680 A.D. Some of the Hopi from this site practiced cradleboarding of their infants, and the sample includes both individuals with nonmodified crania and individuals with modified crania. The inclusion of both nonmodified and modified individuals from the same series provides a control for interpopulation variation in craniofacial morphology (Cheverud and Midkiff, 1992; Cheverud et al., 1992; Kohn et al., 1993). Adult crania from the Old Walpi series were scored a s “not modified” (or nonmodified, nm) if there was no visible evidence of modification, “slightly modified (sm) if there was a n indication of asymmetry present, and “modified” (m) if the lambdoid region was flattened. Intra- and interobserver reliability for scoring modification was found to be high in this series (Konigsberg et al., 175 1993). The Hopi series includes 43 adult females (14 nm, 14 sm, 15 m) and 39 adult males (9 nm, 15 sm, 15 m). The modified crania were classified a s exhibiting either bilateral flattening (15 females, 9 male), flattening on the right side (5 females, 11 males), or flattening on the left side (9 females, 10 males). Gender identity was based on museum records and visual inspection of secondary skeletal characteristics. Age was assessed by museum records, dental morphology, and state (open or closed) of the sphenoccipital synchondrosis. The lambdoid flattening observed in the Hopi was a coincidental result of the practice of keeping their infants in a cradle. Hopi cradles were made of either woven saplings or a board, and infants were bound to the cradle for 20 hours or more per day from birth until roughly 6 to 12 months (Dennis, 1940). The cradle provided a place for the infant to sleep and was thought to help the infant grow straight (Hough, 1918; Dennis, 1940). Asymmetry in lambdoid flattening (bilateral, left, or right flattening) was produced by a n infant’s preferred head position while sleeping (Hrdlicka, 1935). Sample classification Discriminant function analysis was used to evaluate the relationship between linear dimensions between landmarks on the cranial vault (landmarks listed in Table 1)and our classification system. Results of discriminant function analysis indicated that nonmodified and modified crania could be reliably discriminated ( P c 0.05). Discriminant function analysis was also used to test whether the slightly modified crania could be reliably re-classified into the nonmodified or modified groups. Since the slightly modified individuals were not included in the original classification test, this is a n unbiased test of their classification. All of the slightly modified individuals could be reliably included within either nonmodified or modified. The final sample includes 43 nonmodified individuals (23 females, 20 males), 16 bilaterally modified individuals (13 females, 3 males), 10 left modified individuals (2 females, 8 males), and 13 right modified individuals (5 females, and 8 males). 176 L.A.P. KOHN ET AL. Measurements T m L E 1. Landmarks recorded for the HOD& crania Landmarks 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 Intradentale superior Premaxilla [(R) anterior alveolar ridge between the canine and premolar] Premaxilla (L) Posterior nasal spine Nasale Zygomaxillare superior (R) Zygomaxillare superior (L) Zygomaxillare inferior (R) Zygomaxillare inferior (L) Infratemporal crest (R) Infratemporal crest (L) Vomer spine Nasion Frontomalare (R) Frontomalare (L) Pterion (R) Pterion (L) Optic foramen (R) Optic foramen (L) Bregma Bregma-nasion (point halfway along bregma-nasion arc) Bregma-Pterion [(R), point halfway along bregmapterion arcl Bregma-Pterion (L) Lambda Asterion (R) Asterion (L) Bregma-lambda (point halfway along bregmalambda arc) Bregma-asterion [(R), point halfway along bregmaasterion arcl Bregma-asterion (L) Pterion-asterion [(R), point halfway along pterionasterion arcl Pterion-asterion (L) Pterion-lambda [(R), point halfway along pterionlambda arcl Pterion-lambda (L) Lambda-asterion [(R), point halfway along lambdaasterion arcl Lambda-asterion (L) Opisthion Basion Lambda-opisthion (point halfway along lambdaopisthion arcl External auditory meatus (R) External auditory meatus (L) Temporo-sphenoid (R) Temporo-sphenoid (L) Jugular process (R) Jugular process (L) Foramen lacerum (R) Foramen lacerum (L) Anterior nasal spine Maxillary tuberosity (R) Maxillary tuberosity (L) Zygomatic arch (R) Zygomatic arch (L) Optic foramen midpoint (average of 18 and 19) Pterion-asterion midpoint (average of 30 and 31) The measurements are comparable to those presented in Kohn e t al. (1993) and Cheverud et al. (19921, and will be briefly discussed here. A diagraph was used to collect the three-dimensional coordinates of 53 landmarks (Fig. 1, Table 1)from the 82 Hopi crania, and all data was collected by one individual. The X- and Y-coordinates of the landmarks were entered using a Tektronix 2dimensional digitizer, and the Z-coordinates were entered by computer keyboard. After data entry, the cranial coordinates were reoriented such that the origin was located at anterior nasal spine. The axes were arbitrarily defined as: (1)anterior-posterior axis was oriented along the line through anterior nasal spine and lambda; (2) medial-lateral axis was oriented roughly along a line connecting the right and left external auditory meatus, perpendicular to the first axis; (3) a superior-inferior axis was oriented roughly along a line through bregma and opisthion, perpendicular to the two previously defined axes. This definition of arbitrary axes was useful for screening data for outliers, provided a useful manner of referring to directions, and did not affect further analyses or their interpretation. Finite element scaling was used to measure the morphological difference between the nonmodified crania and each type of asymmetrically modified crania (bilateral, left, and right modified). Finite element scaling measures the difference between two forms, a reference (initial) and a target (subsequent) form (Lewis et al., 1980; Cheverud et al., 1983; Cheverud and Richtsmeier 1986; Richtsmeier and Cheverud, 1986; Richtsmeier e t al., 1992; Moss e t al., 1985; Moss, 1988; Lozanoff and Diewert, 1986; Bookstein, 1978, 1983, 1984, 1987). The finite element scaling analyses were performed using SCAL3D (Hammer and Bachrach, 19861, and graphical representations of the deformations are provided by FIESCA (Morris, 1989). Alternative morphometric methods are presented in Lele (1991), Lele and Richtsmeier (1991, 1992), Bookstein (19911, Goodall (19911, Kendall(1984), Mardia and Dryden (1989a,b), and reviewed by 177 ASYMMETRIC VAULT MODIFICATION A 31 24 Fig. 1. The landmarks listed in Table 1are identified by number. The nine finite elements listed in Table 2 representing the right and left anterior and posterior cranial vault, the upper and lower face, and the cranial base are depicted by connecting the appropriate landmarks with lines. A: View of the cranium from the bottom; B: Lateral view of the cranium. Note that landmarks which are not included in finite elements were not analyzed. Only landmarks and elements from the left side of the cranium are represented in the lateral view since compliments on the right side would appear superimposed in this figure. Superimposition occurs with lateral and midline landmarks 43 and 44, 45 and 46,30,31, and 53,and 16 and 17, since the cranial base element shown in this lateral view includes landmarks from both the right and left sides. Richtsmeier and coworkers (19921, Rohlf and Marcus (19931, and Bookstein (1991). Lateral and basal views of the average nonmodified and average modified (bilateral, right, and left) Hopi are illustrated in Figures 3, 5 and 7 as superimposed after generalized procrustes rotation (Goodall and Bose, 1987).Generalized procrustes rotation minimizes the squared differences between homologous landmarks in the superimposed forms. This allows the direct visual comparison of the average forms and may be useful to the reader in interpretation of the finite element scaling results. Finite element scaling is a geometric analysis of deformation, and is based on landmark Cartesian coordinates. Finite elements are formed on each cranium by connecting nodes or landmarks with straight lines, and individual anatomic regions are depicted by 178 L.A.P. KOHN ET AL. FORM = ASIZE + ASHAPE P1 A SIZE ONLY P1 PI ASHAPE ONLY P1 Fig. 2. The form strain tensor can be graphically represented by an ellipse. The ellipse represents the local size and shape change of a region surrounding a landmark when comparing a reference to a target form. The tensor describes the degree to which a standard circle changes size (increase or decrease) and is deformed into an ellipse in the deformation of the reference to a target form. P1 and P2 represent the principal axes of deformation in two dimensions. The second figure represents a transformation of size change only, with size change along all anatomical axes increasing the same amount. The third figure represents a transformation of shape only, with no change in area despite the increase along some anatomical axes and decrease along other anatomical axes. a finite element. In three-dimensions the finite elements may be defined by four (tetrahedral), five (pyramidal), six (wedgeshaped), or eight (hexahedral) landmarks. In this analysis, thirty-six landmarks are used to define nine hexahedral finite elements, delineating the upper and lower face and vault on both the right and left sides, and the cranial base (Fig. 1,Table 2). The finite element scaling analysis of the difference between two forms produces a form (or Lagrangian) strain tensor (Cheverud and Richtsmeier, 1986; Richtsmeier and Cheverud, 1986)for each landmark included in the analysis. The form strain tensor is a symmetric matrix which provides a measure of the difference between the reference and target forms in the X, Y, and Z dimensions. A spectral decomposition of the form strain tensor for each node yields the eigenvalues or principal values (ei),and eigenvectors or principal vectors (Pi)of the matrix. The principal val- ASYMMETRIC VAULT MODIFICATION TABLE 2. Finite elements used in the analysis of the effects o f annular on Houi crania' Finite element Landmarks Lower face (R) Lower face (L) Upper face (R) Upper face (L) Cranial base Posterior vault (R) Posterior vault (L) Anterior vault (R) Anterior vault (L) 1, 2, 4, 5, 6, 10, 12, 48 1, 3, 4, 5, 7, 11, 12, 49 5, 6, 10, 12, 13, 14, 16, 52 5, 7, 11, 12, 13, 15, 17, 52 16, 17, 30, 31, 43, 44, 45, 46 24, 25, 28, 30, 36, 38, 43, 53 24, 26, 29, 31, 36, 38, 44, 53 16, 20, 22, 24, 25, 27, 28, 53 17, 20, 23, 24, 26, 27, 29. 53 'Landmarks are defined in Table 1 ues describe the magnitude of the most positive, intermediate, and least positive change necessary to deform the initial into the target form. The principal vectors describe the direction along which this form change occurs, and are described with reference to the initial form and along the axes described previously. Form change can be decomposed into size change (either increase or decrease) and shape change (Fig. 21, and these can be calculated from the principal values. In two dimensions, if one considers a circle of standard size at each landmark in the reference form, size change measures the degree to which the circle increases or decreases in area when the reference form is deformed into the target form. Size change, s, is defined a s the average of the principal values once they are transformed to additive scale by the equation: li = ln[(l + 2ei)liz].The antilog of li is the proportional change in length along the ith anatomical dimension. Shape change, t, is a measure of the degree to which the standard circle is deformed into a n ellipse during the deformation of the reference into the target form. Shape change is the standard deviation of the principal values once they have been transformed to linear scale. Global size change is the local size change averaged over all landmarks, and global shape change is the standard deviation of local size change measured a t each landmark (Cheverud and Richtsmeier, 1986; Richtsmeier and Cheverud, 1986). Statistical tests Multivariate analysis of variance of chords (linear distances between cranial vault landmarks) and arcs (surface distance between cranial vault landmarks) was used to test whether unintentional cranial vault modifi- 179 cation showed a significant effect of sex, modification, and a n interaction of sex and modification. The lack of a significant interaction between sex and modification indicates the absence of differences in the response of male and female cranial vaults to the effects of cradleboarding. Consequently, data from males and females can be pooled to explore the skeletal effects of unintentional modification. For each type of modification (left, bilateral, and right) we test the hypothesis that there is no significant difference between the average nonmodified and average modified Hopi. That is, we test whether the nonmodified and modified individuals could be drawn from the same nonmodified sample based on measures of form differences supplied by finite element scaling. For each test, a weighted average of nonmodified individuals was calculated using the sex ratio observed in the modified group. Within each modification type, the weighted nonmodified average was deformed into the modified average. A parametric test of significance is not possible due to the sample size. Therefore, the significance of the observed differences between nonmodified and modified averages is tested by a nonparametric bootstrap procedure suggested by Lele and Richtsmeier (1991) and used in Cheverud et al. (1992) and Kohn et al. (1993). To perform this test, two bootstrap samples, one nonmodified and one modified, are generated by random sampling with replacement of the nonmodified sample. The random sampling is stratified by sex, such that males and females are drawn from nonmodified sample in the sex ratio observed in the modified sample. The bootstrap nonmodified average is deformed into the bootstrap modified average. Five hundred bootstrap sample pairs were generated for these significance tests. The statistical significance of the observed difference between nonmodified and modified averages is measured by the proportion of the bootstrap differences which exceed those actually observed between the groups. This procedure is used to test local size and shape change at each landmark, global size and shape change, and volume differences between nonmodified and modified individuals. The evaluation of local and global size change and volume differences are two- 180 L.A.P. KOHN ET AL. A Y L X B 2 L X Fig. 3. Superimposition of an average nonmodified Hopi (solid line) and average bilaterally modified Hopi (dotted line) after Procrustes registration on all landmarks. Unconnected landmarks were not included in the present analysis. A: Basal view of skull; B: Right side of skull, seen as though viewed from the left through a transparent left side of the skull; C: Left side of skull. ASYMMETRIC VAULT MODIFICATION 181 A Y L X B Z L, C Fig. 4. Finite element scaling results of the deformation of the average nonmodified Hopi into the average bilaterally modified Hopi drawn on the average nonmodified Hopi. The ellipses represent the magnitudes and directions of difference between the average nonmodified and bilaterally modified Hopi. Ellipses are exagger- ated by a factor of 2.5 to aid in interpretation. Unconnected landmarks were not included in the present analysis. A: Basal view of skull; B: Right side of skull, seen as though viewed from the left through a transparent left side of the skull; C: Left side of skull. 182 L.A.P. KOHN ET AL TABLE 3. Results of finite element scaling analyses of nonmodified into bilaterally modified Hopi crania' Landmark Cranial vault Bregma Bregma-lambda Lambda Opisthion-lambda Opisthion Pterion-asterion midpt. Pterion (R) Pterion (L) Bregma-pterion (R) Bregma-pterion (L) Bregma-asterion (R) Bregma-asterion (L) Pterion-asterion (R) Pterion-asterion (L) Asterion (R) Asterion (L) Cranial base Jugular process (R) Jugular process (L) Foramen lacerum (R) Foramen lacerum (L) Optic foramen midpt. Vomer spine Infratemporal crest (R) Infratemporal crest (Ll Face Fronto-malare (R) Fronto-malare (L) Zygomaxillare superior (R) Zygomaxillare superior (L) Nasion Nasale Infradentale superior Posterior nasal spine Maxillary tuber (R) Maxillary tuber (L) Premaxilla-maxilla (R) Premaxilla-maxilla (L) Global Element volume Lower face (R) Lower face (L) Upper face (RI Upper face (L) Cranial base Posterior cranial vault (RI Posterior cranial vault (L) Anterior cranial vault (R) Anterior cranial vault (L) Size (Prob) Shape (Prob) PV1 (DIR) 0.57 (0.282) -0.66 (0.598) -3.73 (0.999)" -4.77 (0.962)'x'k -2.06 (0.958)** -0.97 (0.900) 1.91 (0.024)" 1.70 (0.048)"' 3.64 (0.004)* 4.47 (0.001)" 2.17 (0.048)* 2.65 (0.054) 4.59 (0.006)* 3.62 (0.008)' -1.03 (0.800) 0.82 (0.254) 5.25 (0.0011' 5.46 (0.0141" 7.02 (0.001)" 7.39 (0.008)' 6.95 (0.001)' 7.51 (0.001)' 3.59 (0.001)' 3.14 (0.056)"* 4.18 (0.022)* 3.20 (0.1141 9.82 (O.OO1)'L 6.88 (0.028)* 7.96 (0.002)' 5.46 (0.042)* 6.85 (0.001)" 6.76 (0.002)" -4.82 (0.998)* -0.65 (0.632) 2.69 (0.188) 1.46 (0.340) 2.65 (0.044)"" 2.19 (0.034)** 3.13 (0.074) 0.07 (0.5101 -3.70 (0.986)* 1.06 (0.352) -1.40 (0.828) -1.98 (0.7401 1.08 (0.264) 0.18 (0.470) -1.40 (0.880) 3.38 (0.082) -2.60 (0.874) 3.80 (0.086) -0.64 (0.678) -2.99 (0.964)"" 0.40 (0.274) PV2 (DIR) PV3 (DIR) 4.86 (M-L) 5.45 (AS-PI) 2.04 (M-L) 1.56 (M-L) 4.54 (AI-PS) 6.08 (M-L) 6.24 (M-L) 5.37 (M-L) 10.05 (AS-PI1 9.47 (AS-PI1 17.75 (A-PI 12.47 (A-P) 17.03 (A-P) 11.14 (M-L) 6.09 (A-P) 6.84 (M-L) 4.07 (S-I) 0.80 (M-L) 0.44 (AI-PS) -0.71 (AI-PSI 1.21 (M-L) 2.89 (S-I) 2.75 (S-I) 2.53 (S-I) 2.76 (AI-PSI 3.94 (AI-PSI -0.17 (S-I) 2.62 (S-I) 4.09 (M-L) 4.26 (A-P) 1.60 (AMI-PLS) 4.98 (AI-PSI -6.39 (A-P) -7.35 (AI-PSI - 11.92 (AS-PI) -13.06 (AS-PI) -10.45 (AS-PI) - 10.24 (A-P) -2.74 (A-P) -2.41 (A-P) -0.93 (M-L) 0.94 (M-L) -7.83 (M-L) -5.41 (M-L) -4.58 (S-I) -3.19 (S-I) -9.40 (ALI-PMS) -7.98 (AS-PI) 6.20 (0.0081" 4.39 (0.210) 9.51 (0.340) 7.56 (0.482) 4.81 (0.058)'* 2.58 (0.250) 3.94 (0.460) 3.32 (0.566) 1.72 (MI-LS) 5.45 (MI-LS) 15.32 (AI-PSI 11.22 (M-L) 9.57 (M-L) 6.01 (RI-0s) 9.43 (M-L) 2.85 (M-L) -2.91 (A-P) -1.67 (A-P) 4.65 (M-L) 2.62 (S-I) 2.25 (S-I) 1.40 (A-P) 1.18 (S-I) 2.08 (S-I) - 11.57 6.11 (0.096)'" 4.54 (0.4261 3.75 (0.352) 3.24 (0.668) 2.53 (0.652) 1.55 (0.7001 2.07 (0.386) 7.34 (0.421) 3.90 (0.406) 6.95 (0.182) 4.24 (0.086)** 4.93 (0.016)" 1.89 (S-I) 5.28 (AMS-PLI) 2.08 (S-I) 0.64 (S-I) 4.77 (S-I) 1.73 (S-I) 0.64 (AS-PI) 14.80 (RI-OS) 1.98 (AMI-PLS) 14.79 (M-L) 4.17 (S-I) 2.72 (A-P) -0.74 (A-P) - 10.84 (M-L) 3.59 (AMI-PLS) -5.03 (AM-PL) 0.39 (A-P) -6.20 (M-L) -0.04 (M-L) -6.13 (A-P) - 1.35 (M-LI 0.04 (A-P) 0.80 (A-P) -1.91 (M-L) -0.59 (AI-PSI -4.06 (M-L) 2.10 (A-P) -4.65 (RS-01) -2.11 (ALS-PMI) -7.04 (M-L) 2.38 (S-I) -3.74 (A-P) 0.32 (A-P) -5.87 (M-L) -2.20 (S-I1 -8.54 (M-L) -5.13 -8.84 -7.64 -2.91 -0.49 -0.40 -4.40 (MS-LI) (MS-LI) (AS-PI) (A-P) (A-P) (RS-01) (A-P) (A-P) 2.64 (0.012)' 0.0006 (0.462) 0.0090 (0.390) 0.0336 (0.132) 0.0359 (0.120) 0.0321 (0.162) -0.0619 (0.978)* -0.0311 (0.850) 0.0011 (0.458) 0.0254 (0.142) 'Reported results include size and t h e proportion of bootstrap samples exhibiting a greater size difference than that observed In the samples; shape and the proportion of bootstrap samples exhibiting a greater shape difference than t h a t observed in the samples; the principal values (PVi) and their associate anatomical directions (transverse directions for midsagittal landmarks are indicated a s 0 for left and R for right); and the proportional volume differences between nonmodified and bilaterally modified Hopi for each finite element, and the proportion of bootstrap samples exhibiting greater volume differences. Tests of local and global size and element volume are two-tailed tests. Landmarks or elements with proportions greater than 0.975 are significantly smaller than nonmodified, and proportions smaller than 0.025 are significantly larger than nonmodified. Tests of local and global shape change are one-tailed tests, and proportmns greater than 0.05 are significantly different from nonmodified *Statistical significance a t P = 0.05 level. **Statistical significance at P = 0.10 level. sided tests, with probabilities less than 0.025 and greater than 0.975 significant at the 5% level. Local and global shape changes are one-sided tests, and probabilities less than 0.05 indicate significance at the 5% level of significance. Direction of local shape differences are uninterpretable in the absence of significant local shape change (Bookstein, ASYMMETRIC VAULT MODIFICATION 183 A B 2 L x Fig. 5. Superimposition of an average nonmodified Hopi (solid line) and average right modified Hopi (dotted line) after Procrustes registration on all landmarks. Unconnected landmarks were not included in the present analysis. A Basal view of skull; B: Right side of skull, seen as though viewed from the left through a transparent left side of the skull; C: Left side of skull. 184 L.A.P. KOHN ET AL. A Fig. 6. Finite element scaling results ofthe deformation of the average nonmodified Hopi into the average right modified Hopi drawn on the average nonmodified Hopi. The ellipses represent the magnitudes and directions of difference between the average nonmodified and right modified Hopi. Ellipses are exaggerated by a factor - of 2.5 to aid in interpretation. Unconnected landmarks were not included in the present analysis. A: Basal view of skull; B: Right side of skull, seen as though viewed from the left through a transparent left side of the skull; C: Left side of skull. 185 ASYMMETRIC VAULT MODIFICATION TABLE 4. Results Landmark Cranial vault Bregma Bregma-lambda Lambda Opisthion-lambda Opisthion Pterion-asterion midpt. Pterion (R) Pterion (L) Bregma-pterion (R) Bregma-pterion (L) Bregma-asterion (R) Bregma-asterion (L) Rerion-asterion (R) Pterion-asterion (L) Asterion (R) Asterion (L) Cranial base Jugular process tR) Jugular process (L) Foramen lacerum (R) Foramen lacerum (L) Optic foramen midpt. Vomer spine Infratemporal crest (R) Infratemporal crest (L) Face Fronto-malare (R) Fronto-malare (L) Zygomaxillare superior (R) Zygomaxillare superior (L) Nasion Nasale Infradentale superior Posterior nasal spine Maxillary tuber. (R) Maxillary tuber. (L) Premaxilla-maxilla (R) Premaxilla-maxilla (L) Global Element volume Lower face (R) Lower face (L) Upper face (R) Upper face (L) Cranial base Posterior cranial vault (R) Posterior cranial vault (L) Anterior cranial vault (R) Anterior cranial vault (Ll of finite element scaling analyses of nonmodified into right modified Hopi crania' Size (Prob) Shape (Prob) PV1 (DIR) -0.32 (0.600) -3.32 (0.912) -1.38 (0.826) 0.76 (0.418) 0.28 (0.430) 0.33 (0.400) 2.11 (0.052) 1.83 (0.084) 1.82 (0.138) -0.67 (0.642) -5.44 (1.000)* 2.32 (0.068) 3.52 (0.008)" 0.46 (0.388) -3.88 (0.998)* 1.88 (0.086) 2.60 (0.118) 3.93 (0.232) 3.21 (0.010)* 1.85 (0.8601 2.53 (0.510) 3.71 (0.012)* 3.91 (0.001)* 3.59 (0.142) 6.86 (0.002)* 4.54 (0.066)** 10.00 (0.001)* 7.14 (0.010)* 8.59 (0.001)" 2.29 (0.770) 6.11 (0.001)" 5.89 (0.004)" 2.23 (S-I) 1.57 (AOI-PRS) 2.41 (S-I) 3.48 (S-I) 3.12 (AI-PS) 3.51 (M-L) 6.95 (A-P) 6.48 (M-L) 10.52 (A-P) 5.55 (S-I) 5.59 (A-PI 12.61 (MSILI) 14.80 (ALS-PMI) 3.18 (M-L) 4.02 (A-P) 10.46 (M-L) 0.77 (M-L) -3.76 (A-P) -3.30 (AOS-PRI) -7.47 (M-L) -0.98 (M-L) - 5.20 (A-P) -0.28 (A-P) -0.79 (M-L) 0.90 (M-L) -2.98 (AS-PI) 2.56 (S-I) -4.66 (A-P) 2.88 (M-L) -2.90 (S-I) 2.14 (S-I) -2.62 (A-P) 3.13 (S-I) -6.64 (M-L) - 1.49 (A-P) -5.45 (M-L) -2.76 IS-I) -15.65 (M-L) 1.99 (A-P) -5.87 (MI-LS) 5.72 (AI-PS) -7.23 (AMS-PLI) 0.82 (A-P) -2.44 (S-I) -9.94 (M-L) -4.23 (S-I) 0.98 (A-P) -4.61 (S-I) 0.92 (0.372) 0.83 (0.328) 0.68 (0.426) 4.09 (0.136) -0.24 (0.618) 0.91 (0.280) 3.71 (0.066) 0.23 (0.474) 7.82 (0.010)* 4.45 (0.332) 4.61 (0.858) 12.93 (0.194) 4.09 (0.148) 3.61 (0.054)** 6.24 (0.088)*" 2.55 (0.836) 10.60 (AL-PM) 7.20 (AI-PSI 7.19 (S-I) 26.72 (A-PI 5.30 (M-L) 4.19 (M-L) 12.94 (M-L) 3.06 (AM-PL) 2.54 (AM-PL) -0.24 (AS-PI) -0.19 (A-P) 0.52 (S-I) -0.84 (A-PJ 2.94 (S-I) 3.50 (S-I) 0.90 (S-I) -8.52 -3.85 -4.29 -8.78 -4.69 -3.99 -3.62 -3.07 (S-I) (M-L) (M-L) (M-L) (S-I) (A-P) (A-P) (AL-PM) 0.51 (M-L) 5.24 (AI-PS) -2.59 (A-P) -0.56 (A-P) -1.39 (M-L) -0.98 (M-L) 0.67 (AS-PI) 1.41 (A-P) -4.11 -7.27 -6.82 -4.74 -3.15 -3.81 -2.23 -2.26 -3.51 -5.86 -3.73 -4.82 (S-I) (M-L) (M-L) (M-L) (A-P) (A-P) -0.19 (0.522) 1.55 (0.442) -2.12 (0.888) -0.19 (0.548) 0.17 (0.448) -0.41 (0.660) 1.35 (0.194) 5.88 (0.012)* 0.59 (0.356) 0.70 (0.420) 0.88 (0.322) -0.48 (0.614) 0.53 (0.306) 3.32 (A-P) 3.10 (0.780) 8.13 (AS-PI) 6.73 (0.612) 4.50 (0.292) 3.78 (S-I) 4.07 (0.660) 5.22 (S-I) 3.62 (0.478) 5.45 IS-I) 3.17 (0.132) 3.88 (S-I) 3.28 (0.050)** 6.00 (AI-PSI 9.09 (0.184) 22.46 6-11 3.70 (0.560) 5.72 (MI-LS) 5.32 (0.498) 7.18 (M-L) 3.96 (0.222) 6.16 (S-I) 3.99 (0.248) 4.88 (A-P) PV2 (DIR) 0.00 (A-P) 1.66 (A-P) 0.70 (A-P) -1.03 (S-I) PV3 (DIR) (M-L) (M-L) (MS-LI) (S-I) (M-L) (M-L) 2.14 (0.452) 0.0471 (0.104) 0.0241 (0.324) 0.0028 (0.512) 0.0130 (0.426) 0.0820 (0.036)** -0,0565 (0.926) 0.0345 (0.196) -0.0467 (0.882) 0.0019 (0.498) 'Reported results include size and the proportion of bootstrap samples exhibiting a greater size difference than that observed in t h e samples; shape and the proportion of bootstrap samples exhibiting a greater shape difference than that observed in t h e samples; the principal values (PVi) and their associate anatomical directions (transverse directions for midsapttal landmarks are indicated as 0 for left and R for right); and the proportional volume differences between nonmodified and right modified Hopi for each finite element, and the proportion of bootstrap samples exhibiting greater volume differences, Tests of local and global size and element volume are two-tailed tests. Landmarks or elements with proportions greater than 0.975 are significantly smaller than nonmodified, and proportions smaller t h a n 0.025 are significantly larger than nonmodified. Tests of local and global shape change are one-tailed tests, and proportions greater than 0.05 are significantly different from nonmodified. *Statistical significance a t P = 0.05 level. **Statistical significance at P = 0.10 level. 1984; Cheverud et al., 1992), since lack of significance indicates that there are no statistical size differences among anatomical dimensions. Following the methods of Bjork and Bjork (1964) we also evaluate whether there is significant asymmetry in size and shape change. That is, we test for a significant as- 186 L.A.P. KOHN ET AL A B C 2 I Fig. 7. Superimposition of an average nonmodified Hopi (solid line) and average left modified Hopi (dotted line) after Procrustes registration on all landmarks. Unconnected landmarks were not included in the present analysis. A: Basal view of skull; B: Right side of skull, seen as though viewed from the left through a transparent left side of the skull; C: Left side of skull. ASYMMETRIC VAULT MODIFICATION 187 A Y L X B Fig. 8. Finite element scaling results ofthe deformation of the average nonmodified Hopi into the average left modified Hopi drawn on the average nonmodified Hopi. The ellipses represent the magnitudes and directions of difference between the average nonmodified and left modified Hopi. Ellipses are exaggerated by a factor of 2.5 to aid in interpretation. Unconnected landmarks were not included in the present analysis. A Basal view of skull; B: Right side of skull, seen as though viewed from the left through a transparent left side of the skull; C: Left side of skull. 188 L.A.P. KOHN ET AL. TABLE 5. Results o f finite element scaling analvses o f nonmodified into left modified HoDi crania' Landmark Cranial vault Bregma Bregma-lambda Lambda Opisthion-lambda Opisthion Pterion-asterion midpt. Pterion (R) Pterion (L) Bregma-pterion (R) Bregma-pterion (L) Bregma-asterion (R) Bregma-asterion (L) Pterion-asterion (R) Pterion-asterion (L) Asterion (R) Asterion (L) Cranial base Jugular process (R) Jugular process (L) Foramen lacerum (R) Foramen lacerum (L) Optic foramen midpt. Vomer spine Infratemporal crest (R) Infratemporal crest (L) Face Fronto-malare (R) Fronto-malare (L) Zygomaxillare superior (R) Zygomaxillare superior (L) Nasion Nasale Infradentale superior Posterior nasal spine Maxillary tuber. (R) Maxillary tuber. (L) Premaxilla-maxilla (R) Premaxilla-maxilla (L) Global Element volume Lower face (R) Lower face (L) Upper face (R) Upper face (L) Cranial base Posterior cranial vault (R) Posterior cranial vault (L) Anterior cranial vault (R) Anterior cranial vault (L) Size (Prob) Shape (Prob) PV1 (DIR) 0.05 (0.536) -1.55 (0.756) - 1.64 (0.840) 0.60 (0.416) 1.66 (0.188) 0.39 (0.398) 3.00 (0.028)"" 2.94 (0.026)** -2.08 (0.832) 1.33 (0.282) -0.33 (0.588) -2.63 (0.972)"" 0.15 (0.470) 1.66 (0.204) 0.51 (0.386) -2.05 (0.918) 3.18 (0.076)"" 3.16 (S-I) 3.06 (0.582) 2.82 (S-I) 3.69 (0.006)" 2.42 (S-I) 3.11 (0.488) 4.48 (AI-PS) 5.88 (0.024)" 10.16 (S-I) 5.39 (0.001)" 5.99 (S-I) 2.32 (0.200) 5.21 (S-I) 2.28 (0.724) 5.57 (S-I) 6.36 (0.002)" 6.34 (S-I) 4.23 (AL-PM) 2.04 (0.802) 5.79 (0.016)" 7.91 (MS-LI) 6.35 (0.052)"" 2.60 (AI-PSI 5.90 (0.052)"" 6.82 (S-I) 4.46 (0.336) 6.52 (AL-PM) 4.69 (0.014)" 7.29 (MS-LI) 4.18 (0.086)** 2.67 (A-P) 4.17 (0.058) -1.84 (0.734) 7.51 (0.048)** 3.33 (0.226) 4.13 (0.004)" 0.22 (0.454) 0.84 (0.412) -0.57 (0.600) 4.88 (0.304) 5.80 (0.220) 7.49 (0.722) 4.17 (0.928) 5.70 (0.036)" 2.09 (0.478) 4.33 (0.418) 3.63 (0.606) 1.13 (0.274) 0.45 (0.562) 0.36 (0.434) 0.23 (0.492) -0.58 (0.594) 0.23 (0.434) -2.08 (0.896) - 1.67 (0.790) 2.00 (0.234) - 1.99 (0.648) -3.20 (0.946) - 1.64 (0.780) 0.36 (0.370) 11.55 (ALI-PMS) 3.71 (AI-PS) 21.28 (A-P) 9.46 (AL-PM) 13.09 (A-P) 2.09 (M-L) 5.84 (S-I) 3.09 (S-I1 10.38 (0.048)" 13.67 (A-P) 10.66 (0.542) 16.02 (S-I) 5.29 (0.158) 4.82 (MS-LI) 5.39 (0.484) 7.87 6-1) 2.83 (0.790) 3.45 (A-P) 2.47 (0.424) 3.22 (M-L) 3.14 (0.126) 1.08 (S-I) 8.48 (0.228) 9.49 (M-L) 7.44 (0.086)** 13.73 (MS-LI) 3.03 (0.922) 1.76 (M-L) 6.77 (0.010)* 3.08 (A-P) 3.29 (0.536) 2.85 (S-I) 2.26 (0.474) PV2 (DIR) 1.44 (M-L) -3.11 (M-L) -0.73 (M-L) 0.82 (M-L) 0.80 (M-L) 2.59 (M-L) 4.48 (M-L) 3.82 (M-L) -2.58 (A-P) 0.77 (AM-PL) -1.65 (A-P) 1.16 (AS-PI) 1.86 (M-L) 3.39 (S-I) -0.83 (A-P) -1.20 (MS-LI) 3.78 (AMS-PLI) 0.80 (M-L) 5.07 (M-L) 2.92 (S-I) 3.02 (M-L) 1.34 (S-I) 1.97 (M-L) 0.86 (M-L) 4.28 (S-I) -0.38 (AM-PL) 3.71 (MI-LS) -0.85 (M-L) -1.78 (S-I) 0.47 (S-I) -0.94 (A-P) -1.57 (A-P) -0.61 (A-P) -1.97 (A-P) 0.04 (S-I) -2.51 (A-P) PV3 (DIR) -4.16 (A-P) -4.02 (A-P) -6.13 (A-P) -3.21 (AS-PI) 4.82 (A-P) -6.55 (A-P) -0.26 (A-P) -0.15 (A-P) -8.71 (M-L) -0.82 (S-I) -6.23 (MI-LS) -10.34 (M-L) -7.20 (A-P) -4.24 (AM-PL) -4.26 (MI-LS) -6.99 (MI-LS) - -1.50 (MI-LS) -8.97 (AS-PI) -0.03 (S-I) 1.48 (AM-PL) -2.13 (S-I) -2.63 (A-P) -4.69 (A-P) -5.25 (A-P) - -11.30 (M-L) 10.80 (AL-PM) -6.61 (A-P) -5.44 (A-P) -3.15 (M-L) -2.83 (A-P) -5.97 (M-L) -10.74 (S-I) -5.22 (MI-LS) -5.38 (S-I) - 11.16 (M-L) -4.87 (M-L) - -0.0140 (0.6640) -0.0419 (0.7660) 0.0592 (0.0500)** 0.0538 (0.1480) 0.1101 (0.0180)" 0.0109 (0.4140) -0.0361 (0.7640) -0.0248 (0.7420) -0,0310 (0.8240) 1 Reported results include size and the proportion of bootstrap samples exhibiting a greater size difference than that observed in th e samples; shape and the proportion of bootstrap samples exhibiting a greater shape difference than t h a t observed in the samples; the principal values (PVi) and their associate anatomical directions (transverse directions for midsagittal landmarks are indicated as 0 for left and R for right); and the proportional volume differences between nonmodified and left modified Hopi for each finite element, and the proportion of bootstrap samples exhibiting greater volume differences. Tests of local and global size and element volume are two-tailed tests. Landmarks or elements with proportions greater than 0.975 are significantly smaller than nonmodified and proportions smaller than 0.025 are significantly larger than nonmodified Tests of local and global shape change are one-tailed tests, and proportions greater than 0.05 are significantly different from nonmodified. *Statistical significance a t P = 0.05 level **Statistical significance a t P = 0.10 level. sociation between the amount of size or shape change and the size of modification. The average nonmodified male Hopi and the average nonmodified female Hopi were esti- mated. Finite element scaling was used to estimate the difference between the average nonmodified male and each modified male by deforming the average nonmodified male ASYMMETRIC VAULT MODIFICATION Hopi into each modified male Hopi. The size and shape change for each deformation was calculated. Size asymmetry in males is the bilateral difference (right minus left) in size change for the deformation of the nonmodified average male into each modified male. The comparable procedure was repeated for females. Shape asymmetry is the bilateral difference in shape change from the same deformations. A Pearson product-moment correlation tests the association between magnitude of asymmetry and the side of modification. Side of modification was coded as 1 for left-sided occipital flattening, 0 for bilateral occipital flattening, and - 1 for right-sided occipital flattening. A positive correlation of size (or shape) asymmetry with side of modification indicates that individuals exhibit greater size (or shape) change on the side that is not modified. 189 and right bregma-asterion). The bilaterally modified cranial vault appears to be shorter (anterior-posterior) and wider (medial-lateral). The flattening of the lambdoid region is evidenced by a n anterior-superior to posterior-inferior decrease ( 11-14%) a t lambda, opisthion, and opisthion-lambda. A parallel increase ( 5 ~ 1 7 %occurs ) a t the right and left asterion, bregma-lambda, and bregmapterion. An anterior-posterior decrease (210%) is seen a t landmarks located along the midsagittal plane (bregma, pterion-asterion midpoint) and the anterior-lateral cranial vault (right and left pterion). A marked widening (2-11%) of the cranial vault is present a t the midsagittal landmarks as well a s bilaterally at pterion, and pterion-asterion. There is a n anterior-posterior increase (417%) a t posterior-lateral landmarks (right and left bregma-asterion, pterion-asterion), and the cranial vault is wider at pterionasterion. Little additional change in the craRESULTS nial vault occurs along the superior-inferior Bilateral modification axes, with approximate 2-5% superior-infeThe average bilaterally modified Hopi fe- rior increase in the anterior and lateral cramale was 3.7% smaller in total volume than nial vault. the average nonmodified Hopi female. The Extensive form change in the cranial vault total volume of the average bilaterally modi- is not accompanied by marked changes in fied male did not differ from the average morphology of the cranial base with bilateral nonmodified Hopi male. Prior to finite ele- modification. Compared to the nonmodified ment scaling analysis, the modified Hopi fe- average Hopi, the average bilaterally modimales were rescaled to ensure that differ- fied Hopi has a smaller right jugular process. ences observed between nonmodified and Size increases at optic foramen midpoint and bilaterally modified crania were not due to vomer spine approach significance. A signifitheir differences in overall size. cant shape change at the right jugular proThe finite element scaling results measur- cess is due to a 12% decrease along a n axis ing the differences between the average non- from medial-superior to lateral-inferior and modified and average bilaterally modified a 3% anterior-posterior decrease at this Hopi are presented in Table 3 and Figure 4. landmark. Shape change a t the optic foraThere is no significant global size difference men midpoint approaches significance with between the nonmodified and bilaterally 2-9% increases along the medial-lateral and modified Hopi, but there is a significant superior-inferior axes and a 3% decrease global shape difference. The volume of the along the anterior-posterior axis. posterior cranial vault is smaller in the bilatSignificant effect of bilateral cranial vault erally modified Hopi, with a statistically sig- modification on the Hopi face are rare: two nificant difference on the right posterior cra- facial landmarks exhibit significant or nial vault. nearly significant size or shape changes. Bilateral modification has a significant ef- There is a general trend toward a higher fect on the size and shape of the Hopi cranial (superior-inferior), narrower (medial-latvault. The posterior cranial vault (lambda, eral) face in the modified Hopi, however, few opisthion, opisthion-lambda) is smaller and of the landmarks exhibit significant shape the lateral cranial vault is larger (bilaterally change. The region surrounding the right at pterion, bregma-pterion, pterion-asterion, fronto-malare is significantly smaller, and 190 L.A.P. KOHN ET AL. the reduced size at the left premaxilla-maxilla approaches significance. The left premaxilla-maxilla is longer (anterior-posterior), narrower (medial-lateral),and shorter (superior-inferior) in the modified Hopi. The bilaterally modified Hopi right fronto-malare is narrower (medial-lateral) and higher (superior-inferior) than the landmark in the nonmodified Hopi. eral), and higher (superior-inferior) a t lambda. Landmarks on the right lateral and posterior cranial vault, including asterion, bregma-asterion, and bregma-pterion are longer (anterior-posterior) and narrower (medial-lateral) and shorter (superior-inferior) than in the nonmodified cranial vault. The only significant shape change on the left side of the cranial vault is a t left asterion, which exhibits a increased width (10%) and Asymmetrical modification: right side decreased height (5%). The total volume of the average right modThe modification of the right cranial vault ified Hopi is smaller than the average non- results in few statistically significant effects modified Hopi. The average modified male on the morphology of the cranial base. There is 2.2% smaller and the average modified are no significant localized size differences female is 3% smaller than the average non- a t any of the landmarks. There is a signifimodified male and female, respectively. The cant shape difference only a t the right jugumodified individuals are re-scaled prior to lar process, which is longer (10%) along a n analysis to ensure that observed differences axis from anterior-lateral to posterior-mein form are not due to differences in volume. dial, longer (2%)along a n axis from anteriorFinite element scaling results measuring medial to posterior-lateral,and shorter (8%) the difference between the average nonmodi- superior-inferior. The decrease (3-4%) in fied Hopi and the average right modified length, increase (4-12%) in width, and inHopi are presented in Table 4 and Figure 6. crease (3-4%) in height observed a t vomer Neither global size nor shape differences are spine and right infratemporal crest apobserved in comparing average nonmodified proaches significance. Despite the lack of with average right modified Hopi. The vol- significance of shape change, considerable ume of the right anterior and posterior cra- asymmetry in magnitude of shape change nial vault is smaller than nonmodified, but estimates are observable between the right this difference is not significant. The volume and left sides. Since the sample size of right of the cranial base is smaller than nonmodi- modified individuals is small, the estimates fied, and this difference approaches signif- of shape change may be highly influenced icance. by a few extreme individuals. Modification of the right side of the cranial Modification of the right side of the cranial vault has a significant effect on cranial vault vault also results in few statistical signifimorphology. The region surrounding right cant effects on facial morphology. There is a bregma-asterion and right asterion are sig- significant size increase at the posterior nanificantly smaller in the right modified Hopi, sal spine. Shape change at intradentale suwhile the right pterion-asterion is signifi- perior approaches significance with a n incantly larger in the right modified Hopi. crease (6%) along a n axis from anteriorThere are a number of significant localized inferior to posterior-superior and a decrease shape changes associated with modification (2%) in width. of the right side of the cranial vault. The Asymmetrical modification: left side flattening of the right posterior cranial vault is associated with a lengthening and widenThere was no difference in total volume ing of the right posterior-lateral and right between the average nonmodified female anterior-lateral cranial vault. Although not and the average left modified female. The significant, the flattening of the right occipi- average left modified male was 2% smaller tal appears to be associated with a n increase than the average nonmodified male, and the in width (medial-lateral) and a n increase in coordinates of the left modified males were height (superior-inferior) on the left side of transformed to correct this volume differthe cranial vault. The cranial vault is shorter ence prior to finite element scaling analysis. (anterior-posterior), narrower (medial-lat- This ensures that difference between aver- 191 ASYMMETRIC VAULT MODIFICATION TABLE 6. Pearson product-moment correlation of asymmetry i n size and shape change with direction of modification' Landmarks 2, 3 6, 7 10, 11 14, 15 16, 17 22, 23 25, 26 28, 29 30,31 43,44 45,46 48,49 Size (Prob) Shape (Prob) -0.052 (0.776) 0.139 (0.442) -0.194 (0.280) 0.149 (0.408) -0.149 (0.386) -0.353 (0.035)* 0.565 (0.001)* 0.567 (0.001)* -0.505 (0.002)* 0.216 (0.207) 0.102 (0.554) 0.065 (0.722) -0.043 (0.815) -0.221 (0.216) -0.098 (0.587) 0.277 (0.119) 0.100 (0.560) -0.070 (0.686) -0.016 (0.925) -0.181 (0.291) 0.232 (0.174) 0.013 (0.939) -0.096 (0.579) 0.321 (0.073)** 1 Asymmetry was measured as right minus left, and direction of modification was 1 for left modified individuals, 0 for bilaterally modified individuals, and - 1 for right modified individuals. *Statistical significance a t P = 0.05 level. **Statistical significance at P = 0.10 level. age nonmodified and average left modified individuals is not due to differences in volume. The sample size of the left modified individuals (N = 10) is smaller than that for the bilaterally (N = 16) or right (N = 13) modified individuals, complicating detection of significant differences between groups. However, the presence of trends comparable t o those observed above can be assessed. The finite element scaling results measuring the difference between the average nonmodified and average left modified Hopi is presented in Table 5 and Figure 8. There is no global size or shape difference between the average nonmodified and average left modified Hopi. The volume of the cranial base is significantly larger in the average left modified Hopi, and the larger volume of the right upper face in the average modified Hopi approaches significance. The pattern of modification observed in the left modified Hopi generally follows the patterns observed for the bilaterally and right modified individuals in that there are no significant size differences between the average forms. It should be noted that the size increase at the right and left pterion and the size decrease a t left bregmaasterion approach significance. Modification of the left cranial vault results in anteriorposterior decreases at midsagittal and left posterior-lateral landmarks of the cranial vault. The midsagittal and left posterior cra- nial vault decreases in width while the left anterior cranial vault increases in width. There is a general trend toward an increased superior-inferior height throughout the left side of the cranial vault. Several significant shape changes are observable on the right side of the cranial vault. Left modification of the cranial vault results in an anteriorposterior decrease and a medial-lateral and superior-inferior increase on the right side of the cranial vault. As was observed with right cranial vault modification, significant effects of left modification on the morphology of the cranial base are rare. There is a significant size increase at the optic foramen midpoint, and the size increase at the right foramen lacerum approaches significance. The only significant difference in shape between the average nonmodified and left modified Hopi is at the optic foramen midpoint, which is longer (12%),wider (3%), and shorter (3%) in the modified individuals. As was observed with right cranial vault modification, large and asymmetric shape change are observed in the cranial base, especially at the jugular processes and foramen lacerum. The small sample representing left modified cranial vault modification may cause extreme individuals to excessively influence the average for modified individuals. The face also exhibits little significant effect of left modification of the cranial vault. There are no significant localized size differences between the average nonmodified and left modified Hopi. The right premaxillamaxilla is longer (3%) and wider (11%)in the left modified Hopi. Right fronto-malare is longer (13%),narrower (12%),and higher (4%) in the left modified Hopi. The shape change a t the right maxillary tuberosity approaches significance. The right maxillary tuberosity is longer (13%) along an axis from medial-superior to lateral-inferior and smaller (5%)along an axis from medial-inferior to lateral-superior. Asymmetry A Pearson product-moment correlation measures the association of direction of modification and magnitude of asymmetry (the difference between the right and left sides) in size and shape change (Table 6). Apositive 192 L.A.P. KOHN ET AL. correlation indicates that the amount of size or shape change is greater on the right than on the left when the left side is flattened (or change on the left is greater on the left than on the right when the right side is flattened). There is a significant negative correlation of direction of modification with size asymmetry a t bregma-pterion (r = -0.35, P = 0.04) and pterion-asterion (r = -0.51, P = 0.002), and a significant positive correlation of direction and asymmetry for asterion (r = 0.57, P < 0.001) and bregma-asterion (r = 0.57, P < 0.001).Thus both a size reduction a t asterion and bregma-asterion on the side of modification, and a size increase a t bregma-pterion and pterion-asterion on the side that is modified are implied. There is no significant asymmetry for size change at landmarks in the cranial base or face. DISCUSSION These results indicate that unintentional cranial vault modification by cradleboarding effects cranial vault morphology in a predictable manner. Specifically, localized restriction of cranial vault growth by the cradleboard results in compensatory growth through much of the cranial vault. The effect of cradleboarding is generally limited to the cranial vault, however, and this form of cranial vault modification has little effect on the morphology of the cranial base and face. These results support Moss’s (1958) findings that posterior cranial vault flattening did not significantly affect the morphology of the cranial base. Cranial vault flattening in this sample of Hopi appears to have a more extensive effect on the morphology of the cranial vault than was observed by Heathcote (1986) in samples from Kodiak and Kagamil Islands. Asymmetric localized size change in the cranial vault is associated with cradleboarding. Asymmetric flattening of the posterior cranial vault is associated with decrease a t asterion and bregma-asterion, and lengthening a t bregma-pterion and pterionasterion on the modified side of the cranial vault. However, cradleboarding does not produce localized size change asymmetry in the cranial base or face. There is little significant asymmetry in localized shape change. These results contrast with those of Bjork and Bjork (1964) who analyzed crania from Peruvian coastal and highland samples. They observed that asymmetry of the length of the cranial base and maxilla showed a significant positive correlation with direction of modification of the cranial vault. That is, the lengths of the cranial base and maxilla were shorter on the modified side of the cranium. The effects of cradleboarding on cranial development can be contrasted to those of intentional cranial vault modification by headdresses or head wraps. In the case of the Hopi, cradleboarding results in pressure being applied to a limited area of the cranial vault, and the morphological effects of cradleboarding are largely limited to the cranial vault. In contrast, antero-posterior modification and annular modification both result from pressure applied to the cranial vault in multiple directions (Dingwall, 1931).Both forms of intentional cranial vault modification have significant effects on the growth and morphology of the cranial base and face (Cheverud et al., 1992; Kohn et al., 1993). Antero-posterior modification of the cranial vault results in a n anterior-posterior decrease and a medial-lateral increase in growth a t landmarks in the cranial vault, cranial base, and face (Cheverud e t al., 1992). Intentional antero-posterior modification (as practiced by prehistoric people from Ancon) and bilateral modification by cradleboarding modify the posterior cranial vault in the same directions. Annular modification of the cranial vault produces a n anterior-posterior increase and medial-lateral decrease in growth a t landmarks in the cranial vault, cranial base, and face. The localized effects of cradleboarding on the growth of the cranial vault, cranial base, and face are similar to those observed in scaphocephaly (Kohn et al., 1994). Scaphocephaly, or premature closure of the sagittal suture, also produces a localized restriction on growth of the cranial vault. Restriction of medial-lateral growth at the sagittal suture significantly affects the shape of the cranial vault; however, the morphology of the cranial base and face are within the normal range for the sample. It appears that there is a n interrelationship between the growth of the cranial vault, ASYMMETRIC VAULT MODIFICATION cranial base, and face. However, localized modification of the craniofacial complex does not always affect the growth of other regions of the craniofacial complex. Localized or unidirectional disturbances in direction of cranial vault growth, such as those resulting from cradleboarding or scaphocephaly, fail t o produce growth changes in the cranial base and face. In contrast, general and multidirectional disturbances in direction of cranial vault growth significantly influence the growth of the cranial base and the face. Similar conclusions arise from the experimental vault modification (Pucciarelli, 1978) and premature closure of cranial sutures (Babler and Persing, 1982; Babler et al., 1987; Babler, 1988, 1989), and from craniosynostoses (Moss, 1959; Kreiborg, 1981, 1986; Kreiborg and Pruzansky, 1981; David et al., 1990, 1982; Richtsmeier, 1985, 1987, 1988;Richtsmeier et al., 1991). These results indicate that anthropological studies of population variability can readily include samples from populations that practiced cradleboarding. Since cradleboarding does not effect the morphology of the cranial base and face, inclusion of dimensions from these regions should not introduce bias in analyses of genetic differences between populations. Thus, biological distance studies of prehistoric or protohistoric groups should not be distorted by the inclusion of these dimensions. ACKNOWLEDGMENTS We thank Dr. Glenn Cole and the Field Museum of Natural History for access to the skeletal collection. We also thank Dr. Joan T. Richtsmeier for making MGPA and FIESCA available to us. Dr. Lyle W. Konigsberg was very helpful in the early stages of this project. Nyuta Yamisita and Dr. Jim Midkiff were helpful in collecting the dta, and Susan Jacobs was helpful in the data analysis. Three anonymous reviewers provided helpful comments on this manuscript. This research was supported by NSF grant BNS8910998. LITERATURE CITED Allison M, Gerszten E, Munizaga J, Santoro C, and Focacci G (1981) La practica de la deformacion craneana 193 entre 10s pueblos Andinos Precolombinos. Chungara 7t238-260. Anton SC (1989) Intentional cranial vault deformation and induced changes of the cranial base and face. Am. J . Phys. Anthropol. 79r253-268. Babler WJ (1988) Effects of multiple suture closure on craniofacial growth in rabbits. In KWL Vig and AR Burdi (eds.):Craniofacial Morphogenesis and Dysmorphogenesis. Ann Arbor: Center for Human Growth and Development, pp. 73-90. Babler WJ (1989) Relationship of altered cranial suture growth to the cranial base and midface. In Ja Persing, MT Edgerton, and JA Jane (eds.): Scientific Foundations and Surgical Treatment of Craniosynostosis. Baltimore: Williams and Wilkins, pp. 87-95. Babler WJ, and Persing JA (1982) Experimental alteration of cranial suture growth: Effects on the neurocranium, basicranium, and midface. In AD Dixon and BG Sarnat (eds.): Factors and Mechanisms Influencing Bone Growth. New York: Alan R. Liss, Inc., pp. 333-345. Babler WJ, JA Persing, Nagorsky WJ, and Jane JA (1987) Restricted growth of the frontonasal suture: Alterations in craniofacial growth in rabbits. Am. J. Anat. 178t90-98. Bennett KA (1973) The Indians of Point of Pines, Arizona: A Comparative Study of Their Physical Characteristics. Anthropological Papers of the University of Arizona, No. 23. Tuscon: The University of Arizona Press. Bennett KA (1975) Skeletal Remains From Mesa Verde National Park, Colorado. Publications in Archaeology 7F, Wetherill Mesa Studies. Washington, DC: National Park Service. Bjork A, and Bjork L (1964) Artificial deformation and cranio-facial asymmetry in ancient Peruvians. J. Dent. Res. 43t353-362. Blackwood B, and Danby PM (1955)A study of artificial cranial deformation in New Britain. J. Roy. Anthropol. Inst. 85:173-192. Boas F (1921)Ethnology of the Kwakiutl: Based on data collected by George Hunt. 35th Ann. Rep. Bur. Am. Ethnol. 39-794. Bookstein FL (1978) The Measurement of Biological Shape and Shape Change. Lecture Notes in Biomathematics, No. 24. New York: Springer-Verlag. Bookstein FL (1983)The geometry ofcraniofacial growth invariants. Am. J. Orthod. 83t221-234. Bookstein FL (1984) A statistical method for biological shape change. J . Theor. Biol. 107:475-520. Bookstein FL (1987) Describing a craniofacial anomaly: Finite elements and the biometrics of landmark locations. Am. J . Phys. Anthropol. 74t495-510. Bookstein FL (1991) Morphometric Tools for Landmark Data. Geometry and Biology. New York: Cambridge University Press. Buikstra JE, Frankenberg SR, and Konigsberg LW (1990) Skeletal biological distance studies in American physical anthropology: Recent trends. Am. J . Phys. Anthropol. 821-7. Cheverud JM, and Midkiff J E (1992) Effects of frontooccipital reshaping on mandibular form. Am. J. Phys. Anthropol. 87:167-171. Cheverud JM, and Richtsmeier J T (1986)Finite-element 194 L.A.P. KOHN ET AL. scaling applied to sexual dimorphism in rhesus macaque (Macaca mulatta) facial growth. Syst. Zool. 35t381-399. Cheverud JM, Lewis JL, Bachrach W, and Lew WP (1983) The measurement of form and variation in form: An application of three-dimensional quantitative morphology by finite element scaling methods. Am. J. Phys. Anthropol. 62:151-165. Cheverud JM, Kohn LAP, Konigsberg LW, and Leigh SR (1992) Effects of fronto-occipital artificial vault modification on the cranial base and face. Am. J . Phys. Anthropol. 88~323-345. Cocilovo JA (1975) Estudio de dos factores que influencian la morfologia craneana en una coleccion andina: El sex0 y la deformacion artificial. Revista del Instituto de Anthropologia. Universidad Nacional de Tucuman 2t197-212. Cocilovo JA (1978) Estudio de dos factores que influyen en la morfologia craneana en una coleccion patagonica: El sex0 y la deformacion artificial. Arquivos de Anatomia e Anthropologia. Instituto de Anthropologia Professor Souza Marques 3:113-141. Corruccini RS (1972) The biological relationships of some prehistoric and historic Pueblo populations. Am. J. Phys. Anthropol. 37t373-388. Cybulski JS(1975) Skeletal variabilityin British Columbia coastal populations: A descriptive and comparative assessment of cranial morphology. Nat. Mus. Can., Nat. Mus. Man, Mercury Ser.,Archaeol. Survey Can., Pap. 30. David DJ, Poswillo D, and Simpson D (1982)The Craniosynostoses. New York Springer-Verlag. David DJ, Hemmy DC, and Cooter RD (1990) Craniofacia1 Deformities. New York: Springer-Verlag. Dennis W (1940)The Hopi Child. NewYork: D. AppletonCentury Co. Dennis W and Dennis MG (1940) Cradles and cradling practices of the Pueblo Indians. Am. Anthropol. 42r107-115. Dingwall J E (1931) Artificial Cranial Deformation: A Contribution to the Study of Ethnic Mutilations. London: John Bale Sons and Danielson. Droessler J (1981) Craniometry and Biological Distance: Biocultural Continuity and Change a t the Late-Woodland-Mississippian Interface. Evanston: Center for American Archaeology. El-Najjar MY (1978) Southwestern physical anthropology: Do the cultural and biological parameters correspond? Am. J.Phys. Anthropol. 48t151-158. Ewing JF (1950) Hyberbrachycephaly as influenced by cultural conditioning. Pap. Peabody Mus. Am. Archaeol. Ethnol. 3:l-99. Goodall, C (1991) Procrustes methods in the statistical analysis of shape. J . R. Statist. SOC.Ser. B 53t285-339. Goodall C., and Bose A (1987) Models and procrustes methods for the analysis of shape differences. Proc. 19th Symp. Interface Cornput. Sci. Statist. 86-92. Hamner RS, and Bachrach WE (1986) SCALE. Chicago: Northwestern University Rehabilitation Engineering Program. Heathcote GM (1986) Exploratory human craniometry of recent Eskaleutian regional groups from the western arctic and subarctic of North America. BAR International Series 301. Holliday DY (1993) Occipital lesions: A possible cost of cradleboards. Am. J . Phy. Anthropol. 90:283-290. Hough W (1918) The Hopi Indian Collection in the United States National Museum. Washington, DC: Government Printing Office. Hrdlicka A (1935) The Pueblos. Am. J. Phys. Anthropol. 20t235-460. Kendall DG (1984) Shape-manifolds, Procrustan matrices and complex projective spaces. Bull. Lond. Math. SOC.16:81-121. Kohn LAP, Leigh SR, Jacobs SC, and Cheverud JM (1993) Effects of annular cranial vault modification on the cranial base and face. Am. J. Phys. Anthropol. 90:147-168. Kohn LA, Vannier MW, Marsh JL, and Cheverud JM (1994) Effect of premature sagittal suture closure on craniofacial morphology in a prehistoric Hopi male. Cleft Palate-Craniofac. J. 31:385-396. Konigsberg LW, Kohn LAP, and Cheverud JM (1993) Cranial deformation and nonmetric trait variation. Am. J. Phys. Anthropol. 90:35-48. Kreiborg S (1981) Craniofacial growth in plagiocephaly and Crouzon syndrome. Scand. J. Plast. Reconstr. 15t187-197. Kreiborg S (1986) Postnatal growth and development of the craniofacial complex in premature craniosynostosis. In MM Cohen (ed.):Craniosynostoses: Diagnosis, Evaluation and Management. New York: Raven Press, pp. 157-189. Kreiborg S, and Pruzansky S(1981) Craniofacial growth in premature craniofacial synostosis. Scand J . Reconstr. Surg. 15:171-186. Lele S (1991) Some comments on coordinate free and scale invariance methods in morphometrics. Am. J . Phys. Anthropol. 85r407-418. Lele S,and Richtsmeier J T (1991) Euclidean distance matrix analysis: A coordinate-free approach for comparing biological shapes using landmark data. Am. J. Phys. Anthropol. 86:415-429. Lele S,and Richtsmeier J T (1992)On comparing biological shapes: Detection of influencial landmarks. Am. J. Phys. Anthropol. 87:49-66. Lewis JL, Lew WB, and Zimmerman J L (1980)A nonhomogeneous anthropometric scaling method based on finite element principles. J . Biomech. 13t815-824. Lozanoff S,and Diewert V (1986)Measuring histological form change with finite element methods: An application using diazo-0x0-norleucine (DON) treated rats. Am. J . Anat. 177:187-201. Mardia KV, and Dryden AJ (1989a)The statistical analysis of shape data. Biometrika 76271-282. Mardia KV, and Dryden AJ (1989b) Shape distributions for landmark data. Adv. Appl. Probability 21t742-755. McNeill RW, and Newton GN (1965) Cranial base morphology in association with cranial vault deformation. Am. J . Phys. Anthropol. 23:241-254. Mizoguchi Y (1991) Covariations in craniofacial measurements caused by artificial deformations of the cranial vault. Bull. Nat. Sci. Mus. Tokyo, Ser. D, 17:31-50. Morris GR (1989) FIESCA: Graphic and Analytical Software for Finite Element Scaling Analysis in Biological Research. Baltimore, MD: The Johns Hopkins University, Department of Civil Engineering. Moss ML (1958) The pathogenesis of artificial cranial deformation. Am. J . Phys. Anthropol. 16t269-286. Moss ML (1959) The pathogenesis of premature cranial synostosis in man. Acta Anat. 37t351-370. Moss ML (1988) Finite element method comparison of ASYMMETRIC VAULT MODIFICATION murine mandible form differences. J. Craniofac. Genet. Dev. Biol. 8t3-20. Moss ML, Skalak R, Patel H, Sen K, Moss-Salentijn L, Shinozuka M, and Vilmann H (1985) Finite element method modeling of craniofacial growth. Am. J . Orthod. 87t453-472. Neumann GK (1942)Types of artificial cranial deformation in the Eastern United States. Am. Antiq. 7:306-3 10. Oetteking B (1924) Declination of the pars basilaris in normal and artificially deformed skulls: A study based on skulls of the Chumash of San Miguel Island, California, and those ofthe Chinook. Indian Notes Monogr. 7. New York: Heye Foundation. Oetteking B (1930) Craniology of the North Pacific Coast. Mem. AMNH 15:l-391. Pucciarelli HM (1978) The influence of experimental deformation on craniofacial development in rats. Am. J . Phys. Anthropol. 48:455-461. Reed EK 91949) The significance of skull deformation in the southwest. El Palacio 56~106-119. Richtsmeier J T (1985)A Study ofNorma1 and Pathological Craniofacial Morphology and Growth Using Finite Element Scaling Methods. Ph.D. Thesis. Evanston, IL: Northwestern University. Richtsmeier J T (1987) Comparative study of normal, Crouzon, and Apert craniofacial morphology using finite element scaling analysis. Am. J. Phys. Anthropol. 74t473-493. 195 Richtsmeier J T (1988) Craniofacial growth in Apert syndrome as measured by finite element scaling analysis. Acta Anat. 133~50-56. Richtsmeier JT, and Cheverud JM (1986)Finite element scaling analysis of human craniofacial growth. J. Craniofac. Genet. Dev. Biol. 6~289-323. Richtsmeier JT, Grausz HM, Morris GR, Marsh JL, and Vannier MW (1991) Growth of the cranial base in craniosynostosis. Cleft Palate-Craniofac. J. 28.55-67. Richtsmeier JT, Cheverud JM, and Lele S (1992) Advances in anthropological morphometrics. Ann. Rev. Anthropol. 21:283-305. Rohlf FJ, and Marcus LF (1993) A revolution in morphometrics. Trends Ecol. Evolution 8:129-132. Schendel SA, Walker G, and Kamisugi A (1980) Hawaiian craniofacial morphometrics: Average Makapuan skull, artificial cranial deformation, and the “rocker” mandible. Am. J. Phys. Anthropol. 52~491-500. Sciulli PW (1990)Cranial metric and discrete trait variation and biological differentiation in the terminal Late Archaic of Ohio: The Duff Site cemetery. Am. J . Phys. Anthropol. 8219-29. Sciulli PW, and Schneider KN (1985) Cranial variation in the terminal Late Archaic of Ohio. Am. J. Phys. Anthropol. 66~429-443. Stewart TD (1937) Different types of cranial deformity in the Pueblo area. Am. Anthropol. 39~169-171. Young RW (1959) The influence of cranial contents on postnatal growth of the skull in the rat. Am. J . Anat. 105t383-415.