The evolution and development of cranial form in Homo sapiens

The evolution and development of cranial form in
Homo sapiens
Daniel E. Lieberman*†, Brandeis M. McBratney*, and Gail Krovitz‡
*Department of Anthropology, Harvard University, 11 Divinity Avenue, Cambridge, MA 02138; and ‡Department of Anthropology, George Washington
University, 2110 G Street NW, Washington, DC 20052
Edited by Henry C. Harpending, University of Utah, Salt Lake City, UT, and approved November 26, 2001 (received for review August 20, 2001)
Despite much data, there is no unanimity over how to de?ne Homo
correlated phenotypic novelties (15, 16). Thus, interactions at
sapiens in the fossil record. Here, we examine cranial variation among
multiple hierarchical levels of development—from individual genes
Pleistocene and recent human fossils by using a model of cranial
to structural modules (integrated suites of characters that grow as
growth to identify unique derived features (autapomorphies) that
a unit)—confound efforts to define basic independent characters.
reliably distinguish fossils attributed to ‘‘anatomically modern’’ H.
Yet such autapomorphies are predicted to exist if AMHS evolved
sapiens (AMHS) from those attributed to various taxa of ‘‘archaic’’
as a separate lineage from AH.
Homo spp. (AH) and to test hypotheses about the changes in cranial
We test here the hypothesis that AMHS is a distinct species in a
development that underlie the origin of modern human cranial form.
phylogenetic sense, recognizable on the basis of one or more
In terms of pattern, AMHS crania are uniquely characterized by two
autapomorphies, against the null hypothesis that AMHS has no
general structural autapomorphies: facial retraction and neurocranial
autapomorphies, indicating inclusion in a separate lineage. To this
globularity. Morphometric analysis of the ontogeny of these autapo-
end, we report three analyses that examine cranial variation in
morphies indicates that the developmental changes that led to
recent Homo by using a developmental model of cranial evolution.
modern human cranial form derive from a combination of shifts in
First, we use factor analysis to identify structurally important
cranial base angle, cranial fossae length and width, and facial length.
combinations of variables that covary among AMHS crania. Sec-
These morphological changes, some of which may have occurred
ond, we use ANOVA and comparisons of sample ranges to test
because of relative size increases in the temporal and possibly the
whether these structural differences discriminate reliably between
frontal lobes, occur early in ontogeny, and their effects on facial
AMHS and AH. Finally, we combine two morphometric analyses
retraction and neurocranial globularity discriminate AMHS from AH
to investigate hypotheses about the developmental shifts that
crania. The existence of these autapomorphies supports the hypoth-
influence the major structural differences between AH and AMHS
esis that AMHS is a distinct species from taxa of ‘‘archaic’’ Homo (e.g.,
cranial form. First, by comparing the pattern of three-dimensional
Homo neanderthalensis).
cranial shape in adult AH and AMHS by using landmarks that
include major loci of cranial growth, we identify cranial regions that
Paradoxically,ourownspecies,Homosapiens,isoneofthemost appeartocontributetoshapedifferencesbetweenthetaxa.Second,
poorly defined species of hominids. The recent human fossil
we test whether variables that quantify the same shape differences
record has a confusing pattern of variation, with numerous vaguely
between AH and AMHS contribute during ontogeny to the major
defined taxa (e.g., ‘‘archaic’’ H. sapiens, ‘‘modern’’ H. sapiens, Homo
cranial differences between humans and our closest extant relatives,
heidelbergensis, Homo helmei, Homo rhodesiensis), most of which
chimpanzees.
are not widely accepted. A major source of this confusion is the lack
Materials and Methods
of established unique derived features (autapomorphies) of ‘‘ana-
tomically modern’’ H. sapiens (AMHS). The most frequently used
Materials. The majority of the previously proposed diagnostic
diagnosis for AMHS is Day and Stringer’s (1), which is based solely
cranial characters of AMHS listed in Table 1 and several other
on cranial features (listed in Table 1), and which has since been
variables (see below) were measured by using external landmarks
expanded and scrutinized (2–6). However, there are at least two
from several samples: 100 recent H. sapiens (50 of each sex) from
major problems with the diagnostic features in Table 1. First, most
five craniofacially diverse populations (from Australia, China,
Egypt, Italy, and West Africa; for details, see ref. 17); 10 relatively
of the features are difficult to use as phylogenetic characters
complete Late Pleistocene fossils commonly classified as early
because they describe cranial vault globularity, and are thus not
AMHS (Cro Magnon 1, Jebel Irhoud 1, Liujiang, Minatogawa 1,
structurally or developmentally independent. A second, more fun-
Obercassel 1, Predmosti 4, Qafzeh 6, Qafzeh 9, Skhul V, and
damental problem is their failure to discriminate reliably between
Zhoukoudian 101); and nine relatively complete crania typically
‘‘archaic’’ Homo spp. (AH) and AMHS. Many recent human crania
assigned to AH (but not H. erectus) comprising five Neanderthals
fall outside the supposed range of AMHS variation for some
(Gibraltar 1, Guattari, La Chapelle aux Saints, La Ferrassie 1, and
features, and a few skulls generally attributed to AH fall within the
Shanidar 1), and four crania usually attributed to H. heidelbergensis
range of AMHS variation (7, 8). Many researchers (e.g., ref. 9) thus
or H. rhodesiensis (Bodo, Dali, Broken Hill, and Petralona). Fossil
consider H. sapiens to be a morphologically diverse species with
crania lacking the upper face were not included. All external
archaic and anatomically modern grades.
measurements were taken from casts at the American Museum of
Although H. sapiens may include anatomically modern and
Natural History (New York), with the exception of the recent
archaic variants, an increasingly popular view is that AMHS is a
human sample and Skhul V, which were taken from original
distinct species. The best support for this hypothesis comes from
specimens. Geometric morphometric comparisons of cranial dif-
genetic evidence for an African origin of extant human populations
ferences between AH and AMHS were computed from two- and
between 100,000 and 200,000 years ago, and for divergence between
three-dimensional landmarks digitized from computed tomography
humans and Neanderthals about 500,000–600,000 years ago (10–
(CT) scans of four adult fossil crania (Bodo, Broken Hill, Gibraltar
12). Testing this hypothesis by using cranial features, however, is a
challenge because of the substantial integration that occurs among
the various semi-independent units of the cranium (13, 14). Recent
This paper was submitted directly (Track II) to the PNAS of?ce.
evolutionary developmental studies show that major changes in
Abbreviations: AMHS, anatomically modern Homo sapiens; AH, archaic Homo spp.; TPS,
form associated with speciation typically result from ontogenetically
thin plate spline; EDMA, Euclidean distance matrix analysis.
early alterations in the regulation of growth, leading to multiple
†To whom reprint requests should be addressed. E-mail: danlieb@fas.harvard.edu.
1134 –1139 ? PNAS ? February 5, 2002 ? vol. 99 ? no. 3
www.pnas.org?cgi?doi?10.1073?pnas.022440799

---

Table 1. Major cranial features traditionally used to diagnose anatomically modern Homo sapiens
Diagnostic feature (from refs. 1 and 2)
Diagnostic metric (if present)
Short high vault
Basibregmatic height?glabello-occipital length ? 0.70*
Parietals long and curved in mid-saggital plane
Parietal angle (PAA) ? 138°*
Parietal arch high and wide in coronal plane
Bregma-asterion chord?biasterionic breadth ? 1.19*
Occipital bone long, narrow, not markedly projecting
Occipital angle (OCA) ? 114°*
High frontal angle
Frontal angle (FRA) ? 134°*
Weak, noncontinuous supraorbital torus divided into medial
Relative size of glabella, supraciliary ridges, and lateral trigones (ST 1–5)†
and lateral portions
Canine fossa present
Inferior perimeter of zygomatic process retracted relative to superior
(orbital) perimeter*
*From ref. 1.
†From ref. 6, pp. 344 –346.
1, and Guattari) and four fairly robust recent adult male H. sapiens
vault width relative to height (measured as euryon–euryon?
(two Australian, two Native American) from the National Museum
bregma–vertex) was substituted for bregma-asterion chord?
of Natural History (Smithsonian Institution, Washington, DC).
biasterionic breadth (from ref. 1) because it better quantifies vault
Two-dimensional landmarks were digitized from lateral radio-
curvature in the coronal plane; canine fossa depth was measured as
graphs of a longitudinal study of six male and six female recent H.
the maximum subtense between zygomaxillare and alare; supraor-
sapiens from the Denver Growth Study (details in ref. 18) and from
bital torus size?shape was quantified by using Lahr’s system of
lateral radiographs of a cross-sectional sample of Pan troglodytes
grades (ref. 6, pp. 344–346); browridge length was calculated as the
(details in ref. 18). All radiographs were compared at three
midsagittal distance from glabella to the bifrontomaxillare chord.
ontogenetic stages: stage I, 50% through the neurocranial growth
Mandibular characters such as the chin and dental size measure-
phase (?3 years in H. sapiens and 1.5 years in P. troglodytes); stage
ments were not included in the analysis (see ref. 23).
II, at the end of the neurocranial growth phase (?6 years in H.
sapiens and 3 years in P. troglodytes); and stage III, adult (based on
Analysies of Variation. ANOVA and comparison of sample ranges
third molar eruption).
were used to test the hypothesis that structural changes identified
by the factor analysis discriminate between AH and AMHS.
Factor Analysis. Factor analysis identifies combinations of variables
Because of unequal sample sizes, ANOVA significance was deter-
that account for morphometric covariation among a given sample
mined conservatively by using Scheffe´’s F (19). Variables compared
(19, 20). Although identified factors need to be further tested
include the previously proposed diagnostic cranial characters of
against a priori developmental models by using methods such as
AMHS (Table 1) and three additional variables included on the
confirmatory factor analysis (21), exploratory factor analysis is a
basis of the factor analysis results (see below) that have recently
useful initial test of the hypothesis that a few structural modifica-
been proposed as structural determinants of AMHS cranial form
tions underlie much of the taxonomically important cranial varia-
(5–7, 14, 24–27): neurocranial globularity, defined as the round-
tion in recent Homo. Factors were extracted from the AMHS as
edness of the cranial vault in the sagittal, coronal, and transverse
well as the combined AMHS and AH samples described above by
planes; facial retraction, defined as the anteroposterior position of
using principal components analysis; both the initial factor solution
the face relative to the anterior cranial base and neurocranium; and
and a varimax transformation were examined (20). Included vari-
facial prognathism, defined as the orientation of the lower face
ables quantify most of the previously proposed diagnostic charac-
relative to the upper face. Table 2 provides details of how these
ters of AMHS in Table 1: frontal angle, parietal angle, and occipital
variables were measured and standardized. Facial retraction could
angle were measured following Howells (22); vault height relative
not be estimated reliably from external measurements, and was
to length was measured as basion-vertex?nasion-opisthocranion;
measured from radiographs and?or computed tomography scans of
Table 2. Comparison of facial projection, vault globularity, and other cranial features in archaic and anatomically modern Homo
Recent H. sapiens (n ? 100)
Pleistocene H. sapiens (n ? 9)
Archaic Homo spp. (n ? 10)
Overlap†,
Variable
Mean
SD
Range
Mean
SD
Range
Mean
SD
Range
%
Vault height relative to length (VHL)
0.76
0.04
0.68–0.90
0.74
0.04
0.67–0.80
0.63*
0.03
0.60–0.67
17
Parietal angle (PAA, degrees)
135.63
5.10
120–149
135.45
3.50
131–140
142.16*
7.41
133–155
83
Vault width relative to height (VWH)
0.98
0.07
0.81–1.15
0.95
0.06
0.89–1.07
0.85*
0.04
0.80–0.91
50
Occipital angle (OCA, degrees)
123.40
9.02
107–167
118.90
7.60
108–129
105.46*
8.23
96–117
40
Frontal angle (FRA, degrees)
129.49
6.26
102–145
128.44
3.59
121–131
139.54*
4.76
132–146
75
Browridge score (from ref. 6)
1.91
0.95
1–5
3.40
1.27
2–5
5.00*
0.00
5–5
100
Glabellar projection (mm)
24.02
4.29
11.4–34.8
24.47
4.39
17.3–29.8
32.79*
6.77
22.0–41.5
44
Canine fossa depth (mm)
4.82
1.74
1.8–10.1
4.75
1.17
3.5–6.3
1.74*
1.63
0–5.1
44
Prognathism (degrees)‡
35.89
4.35
27.7–52.8
35.84
2.69
32.6–40.2
34.37
3.21
30.1–41.4
100
Facial retraction?GM§
0.40
0.04
0.32–0.48
0.46
0.04
0.42–0.49
0.56*
0.03
0.51–0.58
0
Neurocranial globularity¶
0.59
0.06
0.49–0.86
0.58
0.04
0.51–0.61
0.47*
0.03
0.43–0.49
0
*Mean signi?cantly different (P ? 0.05, Scheffe´’s F, ANOVA) from combined AMHS sample (all variables distributed normally).
†Calculated as percentage of AH crania within total range of variation in combined AMHS sample.
‡The sagittal plane angle between prosthion, the most posteroinferior point on frontal squama above glabella, and the midline average of maxillary tuberosities.
§Measured as nasion–foramen cecum, standardized by a geometric mean of four cranial (mostly facial) dimensions: endocranial volume0.33, nasion–prosthion,
bimaxillary tuberosity breadth, and maxillary tuberosity–prosthion.
ANTHROPOLOGY
¶A dimensionless index of overall neurocranial globularity measured as (euryon-euryon?basion-vertex)?nasion-opsithocranion2.
Lieberman et al.
PNAS ? February 5, 2002 ? vol. 99 ? no. 3 ? 1135

---

the comparative AMHS samples and those fossils for which nasion–
foramen cecum can be measured: Bodo, Broken Hill, Cro Magnon
I, Gibraltar I, Guattari, La Chapelle aux Saints, Obercassel I,
Petralona, and Skhul V.
Geometric Morphometrics. To examine the structural and ontoge-
netic bases of differences in facial form between AH and AMHS,
two morphometric analyses were calculated by using subsets of 17
landmarks digitized directly from computed tomography scans of
fossil hominids measured ETDIPS (www.cc.nih.gov?cip?software?
etdips?) and from radiographs of the ontogenetic samples of H.
sapiens and P. troglodytes. Landmarks used were anterior nasal
spine, basion, bregma, foramen cecum, glabella, lambda, nasion,
opisthocranion, orbitale, pituitary point, posterior maxillary (PM)
point, prosthion, sella, sphenoidale, the most inferoposterior mid-
line point on frontal squama above glabella (frontex), and the
midline points of greatest elevation between nasion and bregma
(metopion), and bregma and lambda (see refs. 22 and 28 for
landmark definitions). After Procrustes superimposition (29, 30),
Thin Plate Spline (TPS) analysis (www.usm.maine.edu?
%7Ewalker?; ref. 31) was used to visualize major differences in
projected lateral view between taxa. Euclidean distance matrix
analysis (EDMA) was also used to quantify significant differences
in three-dimensional shape, by dividing all interlandmark lengths by
a global geometric mean, and by using nonparametric bootstrap-
ping (n ? 100) to determine confidence intervals of 0.90 (? ? 0.10)
for each size-corrected linear distance (32, 33).
Fig. 1.
Untransformed factor scores of external linear measurements (see
Results
Materials and Methods) that quantify most of the proposed diagnostic cranial
characters of AMHS in Table 1. Variables are: 1, frontal angle (FRA); 2, parietal
Exploratory Factor Analysis. Orthogonal and varimax solutions of
angle (PAA); 3, occipital angle (OCA); 4, vault width relative to height (VWH);
both the AMHS and combined AMHS and AH samples yield
5, canine fossa depth (CFD); 6, vault height relative to length (VHL); and 7,
virtually identical results, indicating similar, statistically robust
browridge size?shape. Sample includes recent and fossil AMHS crania (see
patterns of covariation among the diagnostic features of AMHS
Materials and Methods). Variables outside the shaded box have factor load-
listed in Table 1. Fig. 1 summarizes the initial (untransformed)
ings greater than 0.50. Factor 1 (which accounts for 26% of variance) separates
factor solution of the AMHS sample in which factors 1–3 explain
variables that quantify neurocranial globularity; factors 2 and 3 (which to-
61% of the sample variance. Variables that contribute substantially
gether account for 35% of the variance) separate variables related to facial
to factor 1 (factor scores ? 0.50) are parietal angle, occipital angle,
retraction. Factors from combined AMHS and AH samples (not shown here)
vault height relative to length, and vault height relative to width.
show a similar pattern, but account for more sample variance.
These moderately correlated variables (mean r ? 0.40 for the
combined sample) all quantify cranial vault curvature in the
not completely separate AH and AMHS (see also refs. 7 and 8).
coronal, sagittal, and transverse planes. As noted above, neurocra-
Ranges overlap considerably for these variables, especially brow-
nial globularity has previously been proposed to be diagnostic of
ridge size?shape and facial prognathism. However, measure-
AMHS (5, 6, 17, 24). In contrast, browridge size and frontal angle
ments of facial retraction and vault globularity completely
contribute to most of the variation in factor 2, and canine fossa
discriminate between the two taxa with no overlap (Table 2).
depth explains most of the variation in factor 3. Of these patterns
Thus, as characters, neurocranial globularity and facial retrac-
of covariation, the association between browridge size and frontal
angle (factor 2) is related structurally to facial retraction, another
tion appear to represent AMHS autapomorphies.
key proposed structural autapomorphy of AMHS (24, 25). It has
been well established that primates with more retracted faces have
Geometric Morphometric Comparisons. Variations in facial retrac-
smaller, shorter browridges with steeper frontal squamae, reflect-
tion are thought to be a function of interactions between several
ing the supraorbital region’s role to integrate spatially the upper
cranial components including facial size, cranial base angle, cranial
face and the neurocranium (see ref. 14); linear measurements of
base length, and brain size (14, 25–28). Likewise, variations in
facial retraction explain ?80% of browridge length and frontal
neurocranial globularity presumably derive from multiple interac-
angle variation across primates and in ontogenetic samples of
tions between portions of the brain and the size and shape of the
humans and chimpanzees (25, 34). The structural basis for canine
cranial base (17, 28, 35). Thus, to better understand the origin of
fossa depth, which contributes most of the variation in factor 3, is
AMHS cranial form it is useful to identify more proximate variables
less clear and requires further study. Variation in this feature may
that interact during growth to generate variations in facial retrac-
be a function of maxillary arch retraction relative to the zygomatic,
tion and neurocranial globularity among recent humans. As a
but could also reflect maxillary sinus expansion into the infraorbital
preliminary effort, we first used TPS and EDMA analyses of
region.
landmarks that include major loci of cranial growth to compare the
pattern of shape differences between adult AMHS and two taxa of
Analysis of Variation. Table 2 compares ranges and degrees of
AH: Neanderthals and African archaic Homo. The results, sum-
cranial variation for a number of features to test whether the two
marized in Fig. 2, not only highlight the above described differences
structural variables identified above, neurocranial globularity
in facial retraction and neurocranial globularity, but also reveal
and facial retraction, discriminate between AH and AMHS
several important differences in facial and cranial base shape that
better than the features traditionally thought to be diagnostic of
provide clues about their structural and developmental causes. The
AMHS. Although the mean values for all features in Table 1
most obvious difference is that the AMHS face is much smaller
differ significantly (P ? 0.05) between the two samples, they do
relative to overall cranial size than in either group of AH. According
1136 ? www.pnas.org?cgi?doi?10.1073?pnas.022440799
Lieberman et al.

---

Fig. 2.
Geometric morphometric comparisons of AH and AMHS cranial form. (A and B) TPS analysis based on least-squared superimposition (see Materials and
Methods) of modern human (target) and Broken Hill (warp, in green; A), and Guattari (warp, in green; B). Landmarks used in TPS: sella, sphenoidale, PM point,
foramen cecum, anterior nasal spine, nasion, glabella, bregma, lambda, opisthocranion, the most inferoposterior midline point on frontal squama above glabella
(frontex), the midline point of greatest elevation between nasion and bregma (metopion), and the midline point of greatest elevation between bregma and
lambda (see Materials and Methods for de?nitions). Arrows indicate basicranial ?exion in warp. (C and D) EDMA of four modern humans versus Broken Hill and
Bodo (C) and Guattari and Gibraltar 1 (D). Red lines indicate scaled linear distances ?10% longer in AMHS than warp crania; blue lines indicate scaled linear
distances ?10% shorter in AMHS than warp crania; dashed lines indicate linear distances calculated by using only Broken Hill (C) or Guattari (D) from a smaller
subset of landmarks. Note that the PM point, the most anterior point on the greater wings of the sphenoid, lies off the midsagittal plane.
to the landmarks used here, facial reduction in AMHS appears to
cranial base length, and middle and anterior cranial fossae size on
be concentrated in the upper face, with 10–15% decreases in both
cranial ontogeny. In addition, there are no well-preserved fossil
supero-inferior height and antero-posterior length relative to over-
Neanderthal crania with undistorted or complete cranial bases, and
all cranial size. AMHS also have smaller midfaces than Neander-
none younger than 2.2 postnatal years, by which time most cranial
thals but not the archaic Africans because of autapomorphic
base growth (e.g., flexion) is complete (18).
midfacial prognathism in Neanderthals (2).
An alternative, preliminary way to test the effects of facial
At least three important differences in shape between the AMHS
diminution, cranial base flexion, anterior cranial base elongation,
and AH samples are also evident in the cranial base. First, anterior
and expansion of the middle and anterior cranial fossae on facial
cranial base length (e.g., from sella to foramen cecum) is ?15–20%
retraction and neurocranial globularity in H. sapiens is to compare
longer relative to overall cranial size in AMHS than in either taxon
ontogenetic samples of human and nonhuman primates to test
of AH. Second, the anterior cranial base (and with it the face) is
whether the same variables contribute to homologous structural
more flexed relative to the posterior cranial base in AMHS (as
differences. This hypothesis is supported by the analyses summa-
indicated by arrows in Fig. 2; see also refs. 25 and 28). Average
rized in Fig. 3, which compare cranial ontogeny in humans and
cranial base angle in AMHS is 134° (ref. 18), but is ?15° more
chimpanzees by using just the cranial base and facial landmarks
extended in Guattari and Broken Hill. Third, the EDMA analyses
from the analysis presented in Fig. 2. Fig. 3 shows that, in contrast
indicate that the middle cranial fossa in AMHS is ?20% wider
to humans, facial retraction decreases during Pan ontogeny. During
relative to overall cranial size, as shown by the distance between the
early postnatal ontogeny (between stages I and II), facial projection
midline of the sphenoid body and the poles of the temporal lobes
is associated with a decrease in the relative length of the anterior
(the PM points). These differences in relative cranial fossae di-
and middle cranial fossae and with an increase in relative facial
mensions suggest that the temporal (and possibly the frontal) lobes
length and height (Fig. 3A). After neural growth is complete in Pan
are proportionately larger in AMHS than AH.
(between stages II and III), the relative lengths of the cranial fossae
Although the above analyses suggest that a few variables may
continue to shorten, facial height and length continue to increase,
underlie major cranial shape differences between AH and AMHS,
and the cranial base extends, rotating the face dorsally relative to
further analyses are necessary to test whether and how growth
the neurocranium (Fig. 3B; refs. 18 and 28). In contrast, develop-
differences explain these contrasting patterns, especially in terms of
ment of relative cranial base length, relative facial size, and cranial
facial retraction and neurocranial globularity. Recent geometric
base angulation is different in H. sapiens ontogeny, as the neuro-
morphometric comparisons (36) show that Neanderthal and
cranium remains highly globular and the face stays retracted under
AMHS crania have distinctive, ontogenetically early growth pat-
the anterior cranial base. Between stages I and II (while the brain
terns that may result from shifts in basicranial and facial develop-
is still growing but cranial base flexion is complete; ref. 18), the
ment. However, the available sample of infant AH crania is too
posterior cranial fossa becomes relatively shorter as facial size
small and insufficiently complete, particularly in the basicranium, to
remains constant relative to overall cranial size (Fig. 3C). As the
ANTHROPOLOGY
test directly the effects of facial size, cranial base flexion, anterior
human face increases in relative size (mostly inferiorly) between
Lieberman et al.
PNAS ? February 5, 2002 ? vol. 99 ? no. 3 ? 1137

---

stages II and III, facial retraction decreases slightly, the anterior
cranial fossa becomes relatively shorter, and the cranial base
remains flexed rather than extending as it does in Pan (Fig. 3D).
Basicranial flexion is important because it positions most of the face
beneath the anterior cranial fossa. In conclusion, the major vari-
ables that apparently underlie differences in facial retraction and
neurocranial globularity between AH and AMHS are the same
ones that contribute to similar differences evident in human and
chimpanzee cranial ontogeny: cranial base angle, the relative length
and width of the cranial fossae, and relative facial height and length
(Fig. 3E).
Discussion
The above results indicate that most of the differences previously
identified between AH and AMHS crania relate to changes in facial
retraction and overall neurocranial globularity. These two struc-
tural modules not only explain much of the covariation among
traditional diagnostic features of AMHS (1–4), but also do a better
job of discriminating AH and AMHS crania (see refs. 6–8).
Additional crania are needed to test this hypothesis more exten-
sively, but in the diverse sample studied here, there was no overlap
in the range of variation of quantitative measures of these features.
Facial retraction and neurocranial globularity probably discrimi-
nate between AH and AMHS human crania better than Day and
Stringer’s (1) characters because of the effects of integration. Most
of the characters in Table 1 are not independent, but instead
measure aspects of neurocranial shape, facial retraction and other
features that reflect morphological integration during growth
among basic structural units of the skull (e.g., nasal and oral
pharynges, eyeballs, neural lobes, etc.). As a working hypothesis
that requires further testing, we propose that only fossil crania with
an index of neurocranial globularity greater than 0.50 and an index
of facial retraction less than 0.50 should be classified as H. sapiens.
We caution, however, that such criteria are not applicable to
artificially deformed or otherwise pathological crania such as WLH
50 (37).
Although facial retraction and neurocranial globularity appear to
be AMHS autapomorphies, they are not independent units—what
Wagner (38) terms biological characters—but are instead structural
modules that likely derive from complex interactions among more
fundamental units of the skull. Determining the proximate causes
of these autapomorphies is speculative without a more sophisti-
cated understanding of cranial morphogenesis and without enough
well-preserved infant and juvenile crania to compare directly
cranial ontogeny in AH and AMHS. However, it is reasonable to
hypothesize that the evolution of AMHS cranial form may have
been caused by changes in just a few variables that influence the
relative spatial position of the face, cranial base, and neurocranium.
The most important of these shifts are increased flexion of the
cranial base, a longer anterior cranial base, a shorter face (especially
anteroposterior length), and, possibly, increased size of the tem-
poral and?or frontal lobes relative to other parts of the skull.
Ontogenetic and interspecific studies demonstrate the effects of
these variables on cranial shape among human and nonhuman
Fig. 3.
Ontogenetic TPS and EDMA analyses of cranial growth in Pan and Homo
(see Materials and Methods for details). Outlines are selected specimens (targets
in black, warps in green). (A) P. troglodytes stage II (target), stage I (warp). (B) P.
troglodytes stage III (target), stage II (warp). (C) H. sapiens stage II (target), stage
I (warp). (D) H. sapiens stage III (target), stage II (warp). (E) Stage III H. sapiens
(target), stage III P. troglodytes (warp). The TPS analysis is based on only basicra-
nial and facial landmarks: basion, prosthion, anterior nasal spine, nasion, gla-
bella, opisthocranion, sella, pituitary point, sphenoidale, posterior maxillary
plane point, foramen cecum, orbitale, and posterior nasal spine. Superimposed
on TPS are EDMA results: red lines indicate scaled linear distances that are
signi?cantly longer in target than warp crania; blue lines indicate scaled linear
distances that are signi?cantly shorter in target than warp crania.
1138 ? www.pnas.org?cgi?doi?10.1073?pnas.022440799
Lieberman et al.

---

primates. Increased cranial base flexion relative to cranial base
separately from other lineages (43, 44). If one accepts a lineage-
length and brain size is associated with increased globularity of the
based species concept, then AMHS autapomorphies are per-
brain, hence of the braincase (24, 35, 39, 40). Moreover, because the
suasive evidence that H. sapiens is a distinct species from AH
cranial base floor is the roof of the face, cranial base flexion
taxa, including the Neanderthals (H. neanderthalensis) and fossils
influences facial orientation relative to the anterior cranial fossa
sometimes attributed to H. heidelbergensis, H. rhodesiensis, or
(reviewed in ref. 28), and anteroposterior facial length relative to
other hypodigms. Another likely noncranial autapomorphy of
anterior cranial base length affects facial projection relative to the
H. sapiens (not analyzed here) may be the chin (see ref. 23).
neurocranium (reviewed in ref. 14). In addition, temporal and
In addition, from a developmental perspective, many of the
frontal lobe sizes influence the size of the middle and anterior
variables that influence facial retraction and neurocranial globu-
cranial fossae, respectively. Expansion of either lobe thus lengthens
larity (cranial base angle and temporal and frontal lobe size) may
the anterior cranial base (see above). Finally, increases in relative
be good systematic characters because they develop early in
temporal lobe size also contribute to reorienting the face more
ontogeny, and because they likely have a low degree of pheno-
vertically underneath the anterior cranial fossa because the most
typic plasticity. In fact, several recent studies show that the major
anterior points of the middle cranial fossae (the PM points) lie on
differences in cranial growth between Neanderthals and AMHS
arise prenatally or perinatally (36). One interesting exception,
the posterior margin of the face, the PM plane, which has been
however, may be facial size, which grows more slowly during
shown to be tightly constrained (90°) relative to the orientation of
ontogeny and which is partially subject to epigenetic effects from
the axis of the orbits within humans and between primates (28, 41).
mastication (45). Variations in facial size probably contribute to
Thus, temporal lobe elongation relative to cranial size rotates the
much of the variation in browridge size and other correlates of
entire face below the anterior cranial fossa (reviewed in ref. 28).
facial retraction evident within recent H. sapiens (5, 6, 14, 17).
Futher comparative and ontogenetic analyses are needed to
We have much to learn about the complex processes of cranial
test more fully the effects of cranial base flexion, anterior cranial
growth and integration, but the above results highlight how
base length, facial length, and temporal and?or frontal lobe size
efforts to tease apart these processes have the potential to yield
on facial retraction and neurocranial globularity in Homo. In
better characters for testing systematic hypotheses, and to iden-
addition, it would be interesting to know more about the
tify possible targets of selection during speciation. It is exciting
proximate causes of these changes, and their possible adaptive
to consider that only a few small shifts in growth, probably in the
bases (if any). As noted above, increases in relative temporal and
brain and possibly in the cranial base, may be responsible for
frontal lobe size probably cause relative elongation of the
most aspects of the evolution of modern human cranial form.
anterior cranial base in AMHS, and may also underlie increased
Viewed in this light, the origin of modern human cranial form is
basicranial flexion (28). It is intriguing but still premature to
more likely a result of relatively minor morphogenetic ‘‘tinker-
speculate whether such neural differences relate to possible
ing’’ than a major shift in developmental processes.
behavioral differences between AH and AMHS (42).
Regardless of their cause, the existence of several AMHS
We thank C. Dean, J. Jernvall, P. O’Higgins, G. Manzi, D. Pilbeam, R.
autapomorphies has clear systematic implications. Although a
Potts, F. Spoor, and several anonymous reviewers for helpful comments;
K. Mowbray, G. Sawyer, I. Tattersall, and M. Morgan for access to
universally acceptable definition of the species unit is a quixotic
skeletal collections; and D. Hunt, B. Frohlich, J.-J. Hublin, R. Ma-
endeavor, both phylogenetic and evolutionary species concepts
chiarelli, M. Ponce de Leo´n, H. Seidler, F. Spoor, C. Stringer, and C.
agree that species should be monophyletic lineages, evolving
Zollikofer for access to CT scans and?or radiographs of fossils.
1. Day, M. H. & Stringer, C. B. (1982) in L’Homo erectus et la place de l’homme de
22. Howells, W. W. (1973) Cranial Variation in Man: A Study by Multivariate Analysis of
Tautavel parmi les hominides fossiles, ed. De Lumley, M.A. (Centre National de la
Patterns of Difference Among Recent Populations (Peabody Museum Bulletin, Cam-
Recherche Scientifique, Nice, France), pp. 814–846.
bridge, MA).
2. Stringer, C. B., Hublin, J. J. & Vandermeersch, B. (1984) in The Origins of Modern
23. Schwartz, J. H. & Tattersall, I. (2000) J. Hum. Evol. 38, 367–409.
Humans: A World Survey of the Fossil Evidence, eds. Smith, F. H. & Spencer, F. (Liss,
24. Weidenreich, F. (1941) Trans. Am. Philos. Soc. 31, 321–442.
New York), pp. 51–135.
25. Lieberman, D. E. (1998) Nature (London) 393, 158–162.
3. Groves, C. P. (1989) A Theory of Human and Primate Evolution (Oxford Univ. Press,
26. Spoor, F., O’Higgins, P., Dean, M. C. & Lieberman, D. E. (1999) Nature (London)
Oxford).
397, 572 (lett.).
4. Tattersall, I. (1992) J. Hum. Evol. 22, 341–349.
27. O’Higgins, P. (1999) in On Growth and Form: Spatio-temporal Pattern Formation in
5. Lieberman, D. E. (1995) Curr. Anthropol. 36, 159–197.
Biology, eds. Chaplain, M., Singh, G. D. & McLachlan, J. C. (Wiley, New York), pp.
6. Lahr, M. M. (1996) The Evolution of Modern Human Diversity: A Study of Cranial
373–393.
Variation (Cambridge Univ. Press, Cambridge).
28. Lieberman, D. E., Ross, C. F. & Ravosa, M. J. (2000) Yearbook Phys. Anthropol. 43,
7. Wolpoff, M. (1986) Anthropos (Brno) 23, 41–53.
117–169.
8. Kidder, J. H., Jantz, R. L. & Smith, F. H. (1992) in Continuity or Replacement:
29. Chapman, R. E. (1990) in Proceedings of the Michigan Morphometrics Workshop, eds.
Controversies in Homo sapiens Evolution, eds. Bra¨uer, G. & Smith, F. H. (Balkema,
Rohlf, F. J. & Bookstein, F. L. (Univ. Michigan Museum of Zoology, Ann Arbor),
Rotterdam), pp. 157–177.
pp. 251–267.
9. Wolpoff, M. H., Hawks, J., Frayer, D. W. & Hunley, K. (2001) Science 291,
30. Rohlf, F. J. (1990) in Proceedings of the Michigan Morphometrics Workshop, eds.