You are on page 1of 18

Middle Pleistocene human facial morphology in an evolutionary

and developmental context


Sarah E. Freidline
a, b, c, d,
*
,1
, Philipp Gunz
a
, Katerina Harvati
b, c, d
, Jean-Jacques Hublin
a
a
Max Planck Institute for Evolutionary Anthropology, Department of Human Evolution, Deutscher Platz 6, Leipzig 04103, Germany
b
Anthropology Ph.D. Program, City University of New York Graduate School, 365 Fifth Avenue, NY 10016, USA
c
New York Consortium in Evolutionary Primatology, USA
d
Paleoanthropology, Department of Early Prehistory and Quaternary Ecology, Eberhard Karls Universitt Tbingen and Senckenberg Center for Human Evolution and Paleoecology,
Rmelinstr. 23, 72070 Tbingen, Germany
a r t i c l e i n f o
Article history:
Received 24 December 2011
Accepted 8 August 2012
Available online 11 September 2012
Keywords:
Allometry
Face
Ontogeny
Homo heidelbergensis
Neanderthals
Semilandmark geometric morphometrics
a b s t r a c t
Neanderthals and modern humans exhibit distinct facial architectures. The patterning of facial
morphology of their predecessors, the Middle Pleistocene humans, is more mosaic showing a mix of
archaic and modern morphologies. Signicant changes in facial size and robusticity occurred throughout
Pleistocene human evolution, resulting in temporal trends in both facial reduction and enlargement.
However, the allometric patterning in facial morphology in archaic humans is not well understood. This
study explores temporal trends in facial morphology in order to gain a clearer understanding of the
polarity of features, and describes the allometric patterning of facial shape.
The modern human sample comprises cross-sectional growth series of four morphologically distinct
human populations. The fossil sample covers specimens from the Middle Pleistocene to the Upper
Paleolithic. We digitized landmarks and semilandmarks on surface and computed tomography scans and
analyzed the Procrustes shape coordinates. Principal component analyses were performed, and
Procrustes distances were used to identify phenetic similarities between fossil hominins. In order to
explore the inuence of size on facial features, allometric trajectories were calculated for fossil and
modern human groups, and developmental simulations were performed.
We show that facial features can be used to separate Pleistocene humans into temporal clusters. The
distinctly modern human pattern of facial morphology is already present around 170 ka. Species- and
population-specic facial features develop before two years of age, and several of the large-scale facial
differences between Neanderthals and Middle Pleistocene humans are due to scaling along a shared
allometric trajectory. These features include aspects of the frontal bone, browridge morphology, nasal
aperture size and facial prognathism. Infraorbital surface topography and orientation of the midface in
the European Middle Pleistocene hominins is intermediate between the African Middle Pleistocene and
Neanderthal condition. This could suggest that the European Middle Pleistocene hominins display
incipient Neanderthal features.
2012 Elsevier Ltd. All rights reserved.
Introduction
The Eurasian and African Middle Pleistocene hominins (MPh)
are characterized by a combination of craniofacial morphological
features, variously aligning them with Homo erectus sensu lato, H.
neanderthalensis and H. sapiens. The most complete faces include
Bodo and Kabwe (i.e., Broken Hill) from Africa, and Arago 21, Pet-
ralona, and Sima de los Huesos 5 from Europe. MPh specimens are
characterized by massive facial bones, broad upper faces, projecting
browridges, especially in the lateral region, large interorbital
breadth, and details of the nasal margin and palate (Stringer, 1974;
Arsuaga et al., 1997; Rightmire, 1998, 2008; Trinkaus, 2003, 2006;
Athreya, 2009). They also share several facial features with Nean-
derthals, such as facial prognathism, broad nasal apertures, and
a lack of concavity in the infraorbital region (Rak, 1986; Trinkaus,
1987, 2003, 2006; Arsuaga et al., 1997; Harvati et al., 2010).
Striking differences between MPhs, Neanderthals and recent
modern humans occur in their facial size and robusticity. In
particular, Bodo, Petralona and Kabwe are characterized by their
* Corresponding author.
E-mail addresses: sarah.freidline@eva.mpg.de, sfreidline@gmail.com
(S.E. Freidline), gunz@eva.mpg.de (P. Gunz), katerina.harvati@ifu.uni-tuebingen.de
(K. Harvati), hublin@eva.mpg.de (J.-J. Hublin).
1
Present address: Max Planck Institute for Evolutionary Anthropology, Depart-
ment of Human Evolution, Deutscher Platz 6, Leipzig 04103, Germany.
Contents lists available at SciVerse ScienceDirect
Journal of Human Evolution
j ournal homepage: www. el sevi er. com/ l ocat e/ j hevol
0047-2484/$ e see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jhevol.2012.08.002
Journal of Human Evolution 63 (2012) 723e740
massive facial bones compared with Neanderthals and recent
modern humans (Rightmire, 1998), and both the MPhs and Nean-
derthals have longer and more prognathic faces than recent
modern humans (Trinkaus, 2003). To what extent the facial
differences among recent modern humans, MPhs, and Neander-
thals are related to facial size is uncertain.
Allometry, as dened by Gould (1966), is the study of size and its
consequences. Conventionally, it is used to investigate the rela-
tionship between the total body size of the organism and its shape,
anatomy, and physiology among other characteristics (Gould,
1966). Julian Huxley (1932) introduced the concept of allometry
into studies of growth, evolution and function, and later S.J. Goulds
seminal work in both allometry (e.g., Gould, 1966, 1975) and het-
erochrony (e.g., Gould, 1968, 1977), the dissociation of size, shape
and age (Gould, 1977; Alberch et al., 1979), set the framework for
comparative allometric ontogenetic investigations among living
and extinct organisms. Since Gould, a substantial body of work
investigating heterochrony and growth allometry in human and
non-human primates has accumulated (e.g., Shea, 1981, 1983; Leigh
and Cheverud, 1991; Ravosa, 1991, 1998; Collard and OHiggins,
2001; Penin et al., 2002; Leigh et al., 2003; Berge and Penin,
2004; Strand Viarsdttir and Cobb, 2004; Mitteroecker et al.,
2004, 2005; Leigh, 2006; Lieberman et al., 2007; Bruner and
Ripani, 2008; Gonzalez et al., 2011), however studies on fossil
hominins are less common.
Rosas (1997, 2000) and Rosas and Bastir (2004) identied
several features on the Middle Pleistocene Sima de los Huesos (SH)
mandibles associated with an increase in size, including the pres-
ence of a retromolar space, the curvature and orientation of the
symphysis, and the position of the mental foramen with respect to
the dentition. Similarly, in their geometric morphometric study on
modern human and Neanderthal mandibular shape, Nicholson and
Harvati (2006) found that the retromolar gap was related to an
increase in mandibular size. Size-correlated shape changes in later
hominins and recent modern humans have also been observed in
the relative breadth of the nasal aperture (Rosas and Bastir, 2002;
Holton and Franciscus, 2008), nasoglabellar prole (Rosas and
Bastir, 2002), alveolar prognathism (Rosas and Bastir, 2002), and
infraorbital surface topography (Maddux and Franciscus, 2009). In
the latter study, Maddux and Franciscus (2009) identied an allo-
metric relationship between infraorbital size and surface topog-
raphy within Middle to Late Pleistocene Homo and modern
humans. They found that individuals with large infraorbital regions
tend to exhibit at, or inated, surface topographies (characteristic
of Neanderthals), while individuals with small infraorbital areas
possess depressed surface topographies (like modern humans).
They argued that depressed versus inated infraorbital shapes are
not dichotomous congurations, but fall along a continuous size
gradient. Together, these studies demonstrate that some Neander-
thal apomorphies (e.g., retromolar space and infraorbital ination)
are size-correlated.
However, a major limitation to several of the studies mentioned
above is the lack of ontogenetic data. Without a developmental
growth series, these studies can only demonstrate that there are
allometric shape changes among adults (i.e., static allometry). An
understanding of the developmental basis of morphological
features is crucial to interpreting the taxonomic importance of
characters because features that develop earlier in ontogeny are
thought to be less susceptible to epigenetic factors, or less likely to
respond to exposure to environmental effects (Lieberman et al.,
1996, 2002; Wood and Lieberman, 2001; but see; Roseman et al.,
2010).
Due to a growing body of ontogenetic studies, it is becoming
increasingly clear that taxon-specic craniofacial morphology
among hominin species developed prenatally or very early
postnatally (e.g., Ponce de Len and Zollikofer, 2001; Ackermann
and Krovitz, 2002; Lieberman et al., 2002; Strand Viarsdttir
et al., 2002; Williams et al., 2002; Krovitz, 2003; Bastir and Rosas,
2004; Mitteroecker et al., 2004; McNulty et al., 2006; Gunz et al.,
2010, 2012; Gunz, 2012). These morphometric studies largely
support descriptive research identifying unique Neanderthal
morphology at an early age (e.g., Tillier, 1989, 2011; Rak et al., 1994;
Akazawa et al., 1995). For example, at birth the ear labyrinth of
Neanderthals are different in size and shape when compared with
modern humans (Hublin et al., 1996; Ponce de Len and Zollikofer,
1999), and by at least two years of age Neanderthal features on the
occipital, such as the suprainiac fossa and a bilaterally protruding
occipital torus, have already developed (Hublin, 1980). Additionally,
studies have shown that at the time of birth, the face of a Nean-
derthal is already larger than that of a modern human (Ponce de
Len and Zollikofer, 2001; Ponce de Len et al., 2008; Zollikofer
and Ponce de Len, 2010; Gunz et al., 2010, 2011, 2012). On the
other hand, several Neanderthal facial features, such as midfacial
projection, maxillary ination, and double-arched browridges, are
not fully developed until later in childhood (Tillier, 1996).
Ponce de Len and Zollikofer (2001) showed that Neanderthals
and modern humans share a common pattern of cranial and
mandibular shape change from an early age and onward. The
general pattern of shape change comprised a projection and
downward elongation of the face and mandible combined with
a contraction of the cranial vault. They concluded that the cranio-
facial and mandibular differences between these two groups
probably results from differential activity of growth elds early in
ontogeny (Ponce de Len and Zollikofer, 2001). Similarly, in their
geometric morphometric study on facial development in great apes
and Australopithecus africanus, Ackermann and Krovitz (2002)
found that facial features arose very early in development fol-
lowed by parallel postnatal developmental patterns (with the
possible exception of the gorillas). Additionally, they found that
aspects of facial growth were more similar between A. africanus and
modern humans, relative to the great apes. Therefore, their study
suggests that our early human ancestors were already demon-
strating some human-like aspects of facial growth. The implications
of these ndings are that one can interchange hominid postnatal
growth trajectories without producing signicant differences in the
end results. This has been further demonstrated by the study of
McNulty et als. (2006) on the taxonomic afnities of the subadult
Taung fossil specimen. To evaluate the adult morphology of the
Taung child, McNulty et al. (2006) performed a series of develop-
mental simulations. They grew Taung along various hominine
developmental trajectories and compared its adult morphology
with adults of both Australopithecus and Paranthropus. While
McNulty et al. (2006) found the developmental patterns of extant
hominine species to be statistically different, the results from their
developmental simulations indicate that the postnatal develop-
mental differences between hominines have little impact on the
estimation of the adult morphology. McNulty et al. (2006)
demonstrated that the adult morphology of Taung can be reliably
estimated even through the application of an incorrect develop-
mental trajectory.
However, several geometric morphometric analyses on humans
and primates have found that both early postnatal cranial
morphology and later postnatal growth contribute to further
differentiate populations and species (Richtsmeier et al., 1993;
OHiggins and Jones, 1998; OHiggins et al., 2001; Strand
Viarsdttir et al., 2002; Bastir and Rosas, 2004; Cobb and
OHiggins, 2004; Strand Viarsdttir and Cobb, 2004; Bastir et al.,
2007). For example, Bastir et al. (2007) argued that both pre- and
postnatal ontogenetic growth are important in establishing
morphological differences in mandibular shape between
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 724
Neanderthals and modern humans. Their results showed divergent
ontogenetic shape changes between Neanderthals and modern
humans and signicantly different allometric scaling patterns.
The aim of this study is to place the MPhs in both a geographi-
cally and chronologically broad evolutionary and developmental
context to gain a clearer understanding of how archaic (i.e., MPhs
and Neanderthals) and modern human facial features are affected
by facial size and how they change through time. More specically,
the goals of this study are to: 1) explore temporal trends in facial
morphology in order to gain a clearer understanding of the polarity
of facial features, 2) describe the shape changes associated with
allometric scaling in archaic and modern human faces, and 3) test
whether differences in facial shape between archaic and modern
humans are attributable to the differential extension or truncation
of common growth allometries (i.e., ontogenetic scaling), or
whether shape variation is due to divergent growth patterns.
Temporal trends in facial morphology are explored through the
means of a principal component analysis (PCA) of Procrustes shape
variables and Procrustes distance comparisons. The second and
third objectives are addressed by comparing the angle and orien-
tation of Neanderthal and modern human ontogenetic allometric
trajectories in both shape and Procrustes form space, and by
applying developmental simulations (McNulty et al., 2006;
Neubauer et al., 2010; Gunz et al., 2010, 2012; Gunz, 2012) to grow
modern human and Neanderthal faces to the size of selected Middle
Pleistocene fossils. The predicted facial shapes are then visualized
and compared between archaic and modern human groups.
In this study, we recognize three categories of allometry:
ontogenetic, static and evolutionary (Cock, 1966; Cheverud, 1982;
Klingenberg, 1998). We use cross-sectional data on recent modern
and Pleistocene humans to explore patterns in ontogenetic allom-
etry, dened as the covariation of traits with size across different
age groups (i.e., ontogenetic stages) of a given species (Klingenberg,
1998). Static allometry reects trait covariation with size within the
same ontogenetic stage of a single species (Klingenberg, 1998). This
type of allometry is observed among the adults in our sample.
Lastly, evolutionary allometry arises from covariation of traits with
size and their phylogenetic changes and can be analyzed either
within one ontogenetic stage or across stages (Klingenberg, 1998).
In his inuential paper evaluating the relationship between these
three types of allometry, Cheverud (1982), following work by Cock
(1966), Gould (1966) and Shea (1981), demonstrated that patterns
of static adult allometry cannot be assumed to reect ontogenetic
processes (although see Klingenberg and Zimmermann, 1992;
Klingenberg, 1998). This is mainly because adult data usually
represent a very restricted (in both size and time) subset of the
ontogenetic range (Inouye and Shea, 1997).
Following Gould (1975), an allometric regression line can act as
an ontogenetic criterion of subtraction such that the points along
the line can be explained in terms of size-required changes. Devi-
ations from the line are non-allometric shape changes and the
morphological differences may indicate specic functional adap-
tations (Gould, 1975). In this study, we used ontogenetic scaling as
a criterion of subtraction because it is a powerful means for iden-
tifying shared effects of size change between groups or species, as
well as derived dissociations of ancestral allometries (Gould, 1975;
Shea, 1981, 1983; Inouye and Shea, 1997). Here we use a multivar-
iate approach to allometric scaling that differs in certain method-
ological aspects from the classical allometric studies. Departing
fromthe conventional denition and applications of allometry, e.g.,
we substitute facial size for body size because we are specically
interested in how differences in facial size inuence the expression
of facial features and overall facial shape. Surface semilandmark
geometric morphometric techniques are used to quantify facial
features that are otherwise difcult to capture, such as the
infraorbital surface topography, and developmental simulations are
applied to visualize the effects of allometry on the face.
Evolutionary models of Middle Pleistocene hominins
The two main competing phylogenetic interpretations of the
Middle Pleistocene human fossil record are: 1) most of the MPhs
comprise a single, cross-continental taxon, spanning Africa, Europe
and possibly Asia, or 2) the European and African MPhs, at least
those fromMarine Isotope Stage 11 and onward, belong to separate
clades. In the rst model, the MPhs are assigned to the taxon H.
heidelbergensis and are considered to be the last common ancestor
to both Neanderthals and modern humans. This view is supported
by the strong morphological and metric similarities between the
European and African specimens (e.g., Stringer, 1974, 1983; Arsuaga
et al., 1997; Rightmire, 1998, 2007, 2008; Mounier et al., 2009;
Harvati, 2009a).
In the second model, the European MPhs are seen as ancestral
to Neanderthals and are classied as either H. neanderthalensis
(Hublin, 1998, 2009) or as the exclusive members of H. heidelbergensis,
a chronospecies of the Neanderthal lineage (e.g., Arsuaga et al., 1997;
Manzi, 2004; but see Wolpoff et al., 1994; Rosas et al., 2006; Tattersall
and Schwartz, 2006; Bruer, 2008 for alternative interpretations of
the fossil record). In this scenario, the African MPhs are interpreted as
being ancestral to H. sapiens and often classied as H. rhodesiensis
(Hublin, 2009). As evidence for the Neanderthalization process,
proponents of this model refer to a series of features on the skeleton
that foreshadow the Neanderthal condition and that occur uniquely
in the European MPhs (Dean et al., 1998; Hublin, 1998, 2009). These
features are found on the face, occiput, temporal bone and dentition.
They include, but are not limited to, a convex and receding horizontal
infraorbital prole, anteriorly advanced and sagittally oriented face,
wide occipital torus, incipient suprainiac fossa, bilaterally protruding
occipital torus, strong juxtamastoid eminence (Dean et al., 1998;
Hublin, 1998), and derived conditions expressed in different non-
metrical traits of the dentition (Martinn-Torres et al., 2012).
However, quantitative support for this model is limited. This may in
part be due to the difculty in quantifying these complex features
especially on the face (Harvati et al., 2010).
Materials and methods
Sample
This study includes a comprehensive sample of subadult and
adult Middle to Late Pleistocene fossil hominins (Table 1) and recent
modernhumans (Table 2). The fossil sample was designedtoinclude
all available Middle to Late Pleistocene fossils that preserve rela-
tively complete faces (see section on Missing data reconstruction
below). Table 1 lists the 26 fossils, their broad geographical loca-
tion, chronology and repository.
The modern human sample (Table 2) comprises a cross-
sectional growth series from four geographically diverse human
populations spanning three continents: Africa, North America and
Europe. The individuals within each growth series range in age
from two years to adulthood. The African sample consists of an
archaeological Khoisan population from South Africa. The North
American sample is divided into two groups: a combined Native
American archaeological sample from Canyon del Muerto, Arizona,
and Grand Gulch, Utah, and an archaeological population from
Point Hope, Alaska. Lastly, the European sample consists of
a temporally more recent combined sample from Strasbourg
(France) and Greifenberg (Austria). The modern human cranial data
were obtained fromspecimens housed in the American Museum of
Natural History (AMNH, New York), Iziko South African Museum
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 725
(Cape Town), the University of Cape Town, and the Medical Faculty
of Strasbourg.
Individual specimens in the human subadult skull collection
from the Medical Faculty of Strasbourg have known ages with
a precision ranging from one month to one day (Rampont, 1994).
Age estimates for all other subadult individuals were assessed
according to dental eruption patterns following Ubelaker (1989).
Each specimen in the modern human sample was classied by
developmental stage according to dental eruption sequence. Age
group one is composed of subadults with a deciduous dentition,
lacking the eruption (as dened by the exposure of cusps) of any
permanent teeth. Age group two is dened by the eruption of the
rst molar. Age group three is dened by the eruption of the second
molar. Age group four is dened by the eruption of the third molar
(i.e., complete set of permanent dentition). Table 2 lists the number
of subadults distributed among each age group within each modern
human population, and Table 3 lists the age estimates and group-
ings for the subadult fossil specimens. Because of the difculty and
uncertainty in identifying the sex of subadult individuals (Scheuer
and Black, 2000), we made no attempt to do so, although studies
have shown that size and shape differences occur between males
and females throughout ontogeny and are most pronounced during
Table 2
Recent modern human adult and subadult specimens used in the analysis.
Population/Geographic region AG
a
1 AG 2 AG 3 AG 4 (adult) Total
Khoisan, South Africa
c,d,e
8 8 4 38 (M
b
: 14; F: 24) 58
Arizona (Canyon del Muerto), Utah (Grand Gulch), USA
e
7 6 4 52 (M: 25; F: 27) 69
Alaska (Point Hope), USA
e
6 10 4 48 (M: 26; F: 22) 68
Strasbourg, France
f
and Greifenberg, Austria
e
7 4 4 49 (M: 27; F: 22) 64
a
The abbreviation AG represents Age Group. See text (Material and Methods e Sample) for denition of each age group.
b
The specimen sex is denoted as M for males and F for females.
c
Iziko South African Museum
d
University of Cape Town
e
American Museum of Natural History
f
Medicine Faculty of Strasbourg
Table 1
Fossil specimens used in the analysis, their abbreviation (Ab.), repository and chronology.
a
Specimen Ab.
b
Repository
c
Chronology
Middle Pleistocene: Africa
Bodo Bd NME ca. 600 ka (Clark et al., 1994)
Kabwe Kb NHM 700e400 ka (Klein, 1994); late
Middle Pleistocene (Stringer, 2011)
Middle Pleistocene: Europe
Arago 21
d
Ar UM 600e350 ka (Cook et al., 1982;
Falgures et al., 2004)
Petralona Pt AUT 670eca. 250 ka (Harvati et al., 2009)
Sima de los Huesos 5
e
Sm5 UCM ca. 530 ka (Bischoff et al., 2007)
Late MiddleeLate Pleistocene: Africa
Jebel Irhoud 1 I1 MA ca. 160 ka (Smith et al., 2007)
Late MiddleeLate Pleistocene: Europe
Gibraltar 1 Gb1 NHM 71e50 to 35 ka (Klein, 1999)
Guattari Gt MNPE ca. 50 ka (Schwarcz et al., 1991)
La Chapelle-aux-Saints LCh MH 56e47 ka (Grn and Stringer, 1991)
La Ferrassie 1 LF1 MH 71e50 to 35 ka (Klein, 1999)
Pech-de-lAz I Pech MH 51e41 ka (Soressi et al., 2007)
Late MiddleeLate Pleistocene: Asia
Liujiang
e
Ljg IVPP 139e111 ka (Shen et al., 2002)
Qafzeh 6 Q6 UT 135e100 ka (Grn et al., 2005)
Qafzeh 9 Q9 UT 135e100 ka (Grn et al., 2005)
Shanidar 1
e
Sh1 IM ca. 50 ka (Trinkaus, 1983)
Shanidar 5
e
Sh5 IM ca. 50 ka (Trinkaus, 1983)
Skhul 5 Sk5 PM 135e100 ka (Grn et al., 2005)
Teshik Tash
e
T-T MSU ca. 70 ka (Movius, 1953); 57e24 ka
(Vishnyatsky, 1999)
Upper Paleolithic: Europe
Cro-Magnon 1 e MH 28e27 ka (Holt and Formicola, 2008)
Grotte des Enfants 6
e
e MAP Gravettian (Henry-Gambier, 2001)
Mladec 1 e NM ca. 31 ka (Holt and Formicola, 2008)
Predmost 3
e
e DVM (original
material missing)
Early Upper Paleolithic (Smith, 1982)
Oberkassel 1 e RL ca. 12 ka (Street, 2002)
Oberkassel 2 e RL ca. 12 ka (Street, 2002)
Upper Paleolithic: Asia
Zhoukoudian 101
d
e Original material missing ca. 13e33 (Chen et al., 1989;
Brown, 1992; Hedges et al., 1992)
Zhoukoudian 102
d
e Original material missing ca. 13e33 (Chen et al., 1989;
Brown, 1992; Hedges et al., 1992)
a
Subadult individuals are italicized.
b
Individual European and Asian Upper Paleolithic fossil material were not labeled in the gures and therefore no abbreviation is provided for these specimens.
c
The complete names of the repositories are available as a table in the SOM.
d
Casts from the Department of Human Evolution, Max Planck Institute for Evolutionary Anthropology (Leipzig, Germany).
e
Casts from the Division of Anthropology, American Museum of Natural History (New York).
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 726
and after adolescence (e.g., Strand Viarsdttir et al., 2002;
Bulygina et al., 2006). Adult individuals were sexed according to
Howells (1973) criteria, and when possible an equal number of
males and females were included.
Measurement protocol
Computed tomography (CT) and surface scans CT or surface scans
of the fossil and modern human crania were acquired. The CT scans
were made with either an industrial (BIR ACTIS 225/300) or
medical (Toshiba Aquilion) CT scanner, and the remaining speci-
mens were surface scanned with either a Minolta Vivid 910 or
a Breuckmann optoTOP-HE. The pixel size of the CT scans ranged
from 0.24 to 0.49 mm and the slice thickness was between 0.25
and 1.00 mm. The resolution of the surface scanners ranged from
w30 microns to w6 microns in the z plane. The differences in
resolution among these scanners are much smaller than the
digitizing error and differences among the specimens. For
example, although the adults and juveniles in the European
sample were scanned using two different methods (adults:
Minolta surface scanner, juveniles: medical CT scanner), they
cluster together in a principal component analysis (PCA).
Three-dimensional surface models were extracted from either
the surface or CT scans and saved as .ply polygon model le format.
Surface scans were processed using either Geomagic Studio or
OptoCat (Breuckmann) software, depending on the scanner model
used. For the CT data, three-dimensional surfaces were extracted
using Avizo (Visualization Sciences Group Inc.). The landmarks and
semilandmarks were digitized on the surface models using Land-
mark Editor (Wiley et al., 2005). If CT or surface scan data of the
original fossil material were not available, surface scans of high
quality casts (see Table 1) from the Division of Anthropology of the
AMNH (New York) or the Max Planck Institute for Evolutionary
Anthropology (Leipzig) were made.
Landmark data Landmarks and semilandmarks, dening curves
and surfaces, were digitized by one observer (S.F.). 3D coordinates
of landmarks (Table 4) and semilandmarks along curves were
digitized on all specimens and a mesh of surface semilandmarks
was digitized on one template individual (see Fig. 1) using
Landmark Editor software (Wiley et al., 2005). This template
mesh of surface semilandmarks was warped into the vicinity of
the surface of every specimen using a thin-plate spline (TPS)
interpolation according to the landmark and curve data. The
warped points were then projected onto the surfaces by picking
the closest vertices from the specimens .ply le following other
studies (e.g., Gunz, 2005; Gunz and Harvati, 2007; Neubauer
et al., 2009, 2010; Gunz et al., 2009a,b; Harvati et al., 2010;
Stanseld and Gunz, 2011; Freidline et al., 2012). This protocol
guarantees that every specimen has the same number of curve
semilandmarks and surface semilandmarks in approximately
corresponding locations. A detailed description can be found in
Gunz et al. (2005, 2009b) and Mitteroecker and Gunz (2009).
The initially equidistant semilandmarks were slid along
tangents to the curves and tangent planes to the surfaces so as to
minimize the bending energy of the TPS interpolation between
each specimen and the Procrustes consensus conguration. After
sliding, landmarks and semilandmarks were treated the same in
subsequent statistical analyses. To convert the landmark and
semilandmark coordinates to shape variables, a generalized
Procrustes analysis (GPA) was performed. GPA removes the effects
of translation and rotation in the raw coordinate data and stan-
dardizes each specimen to unit centroid size, the square root of the
sum of squared distances from each landmark to the specimens
centroid (Dryden and Mardia, 1998). All data processing and
statistical analyses were performed in Mathematica (Wolfram
Research) and R (R Development Core Team, 2010).
Missing data reconstruction As geometric morphometric methods
require all specimens to have the same number of homologous
points, some minor data reconstructionwas necessary for somefossil
specimens. First, bilateral symmetry was exploited by mirroring the
surface of the better-preservedside alongthemidsagittal planeusing
Geomagic Studio and Avizo. The following specimens were
completed by mirror-imaging: Bodo, Cro-Magnon 1, Gibraltar 1,
Guattari, Kabwe, La Ferrassie 1, Oberkassel 1, and Shanidar 1.
Second, if missing data occurred bilaterally on a specimen or
along its midline, landmarks were estimated using geometric
reconstruction via TPS interpolation (Gunz et al., 2009a). This
method has been applied in several other studies (e.g., Gunz, 2005,
2012; Gunz andHarvati, 2007; Gunz et al., 2009a,b, 2010, 2011, 2012;
Harvati et al., 2010; Bastir et al., 2011; Benazzi et al., 2011; Stanseld
and Gunz, 2011; Freidline et al., 2012; Neubauer et al., 2012) to
Table 4
Homologous landmarks used in the analysis.
a
Landmark Abbreviation Denition
Alveolare ids
Anterior nasal spine ans Thin projection of bone
on the midline at the
inferior margin of the
nasal aperture
Auriculare
b
au
Frontomalare orbitale
b
fmo
Frontomalare temporale
b
fmt
Frontotemporale
b
ft
Glabella g
Jugale
b
ju
Medial orbital margin
b
mm The point of maximal
concavity on the medial
inferior corner of the
orbital margin, where
the orbital margin and
the frontal process of
the maxilla meet
Nasion n
Nasospinale ns
Porion
b
po
Rhinion rhi
Sphenopalatine suture
b
ss Suture between palatine
and sphenoid bones
posterior and inferior to
maxillary tuberosity
following Harvati (2001)
Staphylion sta
Zygomatic process root inferior
b
zri The malar root origin
projected onto buccal
alveolar surface
Zygomatic process root superior
b
zrs The point where malar
root arises from the
maxilla (often a point
of concavity between
alveolare region and
zygomaxillare) following
McNulty (2003)
Zygomaxillare
b
zm
a
All landmarks are dened following White et al. (2012). Denitions and refer-
ences (if available) are provided for the less common landmarks
b
Paired right and left landmarks
Table 3
Ontogenetic age and age group classication of the subadult fossil specimens used in
the analysis.
a
Specimen Ontogenetic age (ca. yr.) Age group
Pech-de-lAz I 2 (Tillier, 1996) 1
Teshik-Tash 9e11 (Tillier, 1989; Williams et al., 2002) 2
Grotte des Enfants 6 13e15 (Henry-Gambier, 2001) 4
a
See text (Material and Methods e Sample) for denition of age group
classications.
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 727
reconstruct fragmentary fossil material. Arago 21 required the most
extensive reconstruction. A detailed account of this reconstruction
and the protocol can be found in Gunz et al. (2009b).
Statistical analyses
To explore the temporal variability in facial morphology among
Pleistocene humans, a PCA was performed in shape space. A PCA
reduces the dimensionality of high dimensional shape space and
provides summaries of large-scale trends within the data
(Bookstein, 1991; Rohlf, 1993). We used TPS warping of the
Procrustes mean shape along the rst two principal components
(PCs) to visualize the shape changes (see Gunz and Harvati, 2007;
Mitteroecker and Gunz, 2009). Additionally, to identify which
individuals are most phenetically similar to one another, nearest
neighbors were calculated using inter-individual Procrustes
distances. To visualize the morphological similarities shared
between nearest neighbors, each of the Middle Pleistocene homi-
nins was superimposed on its nearest neighbor.
In morphometrics, a distinction is made between shape and
form. Shape refers to the geometric properties of an object that are
independent of its overall size, position and orientation, whereas
the form of an object includes both its shape and size (Mitteroecker
and Gunz, 2009). Because this study focuses on allometry, the
inclusion of size in our statistical analyses is essential. To quantify
facial size, we used the variable centroid size, which is computed
during the Procrustes superimposition. Following Mitteroecker
et al. (2004, 2005), we performed a PCA in Procrustes form space
(also called size-shape space) on the Procrustes landmark and
semilandmark coordinates. Form space includes the geometric size
(as the natural logarithm of centroid size) of each specimen
(Mitteroecker et al., 2004) and is valuable because the relationship
between shape and size can be readily explored.
In this study, we compute the ontogenetic allometric trajectories
as multivariate regressions of the Procrustes shape variables on the
natural logarithmof centroid size. As outlined by Mitteroecker et al.
(2004, 2005), properties of these trajectories (e.g., length, shape,
divergence) can be compared between populations and species.
Coincidental ontogenetic trajectories indicate a common pattern of
shared development between the groups under investigation. We
interpret coincidental growth trajectories for the extant (recent
modern humans) and extinct groups (Neanderthals) as a shared
pattern of relative growth that also may be truncated or extended.
Parallel trajectories reveal that the morphological divergence
between groups has already occurred prior to the age of the youn-
gest specimen, and that there is a common pattern of subsequent
growth shared between groups. By contrast, divergent trajectories
indicate that development further accentuates group differences.
Ontogenetic trajectories were plotted in shape and Procrustes
form space for Neanderthals, each of the four modern human
populations, and the modern human mean. The modern human
mean trajectory was calculated by regressing Procrustes shape
coordinates on the natural logarithm of centroid size on the entire
modern human sample. Because only adult MPhs faces are included
in this study, static, rather than ontogenetic, allometric trajectories
were calculated and plotted in Procrustes shape and form space for
these hominins.
To see if modern humans and Neanderthals share a common
ontogenetic allometric trajectory during postnatal ontogeny, we
computed the angle between the Neanderthal and the mean
modern human ontogenetic allometric vectors. We then tested
whether the angle was statistically different from zero using
a permutation test (McNulty et al., 2006; Gunz, 2012). We rst
mean-centered the respective groups and then compared the
actual angle with angles obtained from regressions computed for
5000 permutations that randomly reassigned group membership,
permuting the entire sample for each iteration. Likewise, the angles
between the ontogenetic allometric trajectories of the four modern
human populations were calculated and tested to see if they were
signicantly different from zero. To assess the effects of the small
Neanderthal sample size on the computation of the ontogenetic
allometric trajectory, we bootstrapped subsamples from the recent
modern human sample, simulating the composition of the Nean-
derthal sample (i.e., one specimen from age group 1, one from age
group 2 and six adults).
Developmental simulations and visualization techniques
The aim of the series of developmental simulations was to
identify the allometric shape changes that occur when scaling an
adult modern human and Neanderthal to the size of a MPh. To do
so, we calculated the mean adult modern shape and grew it
following the modern human mean ontogenetic allometric trajec-
tory to the size of a) Kabwe, an African MPh that has a particularly
inated infraorbital surface topography for a MPh, and b) Bodo, an
African MPh that has the largest face in our sample. Shape changes
were visualized by TPS warping of the adult modern human mean
shape to the modern human mean grown to the size of Bodo and
Kabwe. The same approach was applied to Neanderthals. We war-
ped the surface of La Ferrassie 1 to the mean adult Neanderthal
shape, and then grew it following the Neanderthal ontogenetic
allometric trajectory to the size of Kabwe and Bodo.
Results
Principal component analyses and permutation tests
To explore temporal trends in facial morphology, a PCA was
performed in Procrustes shape space of only the adult individuals
(Fig. 2). The rst two principal components (PCs) represent 46.9% of
the total shape variation, and neither of the PCs correlate with size.
There is a clear separation between the modern humans, including
Figure 1. The landmark and semilandmark dataset. 671 landmarks and semiland-
marks digitized on all specimens, red: homologous landmarks; blue: curve-
semilandmarks; yellow: surface-semilandmarks. Homologous landmarks are abbre-
viated. The full names are listed in Table 4. (For interpretation of the references to
colour in this gure legend, the reader is referred to the web version of this article.)
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 728
the early modern and Upper Paleolithic humans, and the archaic
humans. Additionally, members of each temporal group cluster
together. However, there is some overlap between Middle Pleisto-
cene humans and Neanderthals. The main shape changes that occur
along PC 1 are in the curvature of the anterior portion of the frontal
bone, browridge projection, width of the nasal aperture and facial
prognathism. Middle Pleistocene specimens, such as Bodo and
Sima de los Huesos 5, and Neanderthals, such as La Chapelle-aux-
Saints, cluster at the extreme positive end of PC 1. These fossils
share a more receding anterior frontal bone, anteriorly projecting
browridge, wide nasal aperture, and prognathic mid and lower face
when compared with recent modern humans, which generally plot
at the negative end of PC 1. While several shape changes are shared
between PC 1 and PC 2, such as frontal bone curvature, facial
prognathism and infraorbital surface topography, PC 2 primarily
demonstrates changes in facial length and orientation of the
maxillary body (i.e., maxillary body facies). Shanidar 5, Petralona
and Kabwe share a superiorlyeinferiorly long face and their
maxillary body is oriented near vertically. On the opposite end of PC
2, the modern humans express a shorter face and a maxillary body
that is oriented down and slightly posteriorly.
Based on pairwise Procrustes distances (PD), Bodo is most
similar to Sima de los Huesos 5 (PD0.068), Sima de los Huesos 5 is
most similar to Arago 21 (PD 0.06), Arago 21 is most similar to La
Chapelle (PD 0.05), Petralona and Kabwe are most similar to one
another (PD 0.056), and Shanidar 5 is most similar to Petralona
(PD 0.056). To illustrate the resemblances between these speci-
mens, we superimposed each MPh onto its nearest neighbor
(Fig. 3aeo). Bodo and Sima de los Huesos 5 share similarly projec-
ting lateral browridges, wide nasal apertures and interorbital
breadths. However, Bodos glabellar region is more anteriorly pro-
jecting and its mid andlower face morphology is especially different
from Sima de los Huesos 5 (Fig. 3aec). For example, Sima de los
Huesos 5 has a more prognathic mid and lower face, a slightly more
inated infraorbital surface topography and a more parasagittally
oriented midface when compared with Bodo (Fig. 3c: Sima de los
Huesos 5 is green and Bodo is bone color). Sima de los Huesos 5 and
Arago 21 (Fig. 3def) share a similar mid and lower face morphology
and diverging upper faces (i.e., browridges). Thus they share
a similar infraorbital surface topography, and they exhibit a similar
degree of facial prognathism and parasagittal rotation. However,
Arago 21 has a more projecting browridge, especially the middle
Figure 2. PCA in shape space of adult fossil and modern humans. A convex hull is drawn for the Upper Paleolithic modern humans. The arrows indicate each fossil specimens
nearest neighbor according to inter-individual Procrustes distances. The black arrows specify if an individual is most similar to another individual outside of its temporal group. The
full names and abbreviations for the fossil specimens are listed in Table 1. The surface visualizations represent the mean shapes at the positive and negative ends of PC 1 and 2.
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 729
portion (Fig. 3f: Sima de los Huesos 5 is green). The similarities
between Arago 21 and La Chapelle (Fig. 3gei) are in midfacial
prognathism and infraorbital surface topography. La Chapelle has
a more anteriorly projecting browridge than Arago 21, except for the
lateral portions, its midface is more parasagitally oriented, and its
nasal aperture is more anteriorly projecting (Fig. 3i: La Chapelle is
red). Kabwe and Petralona (Fig. 3jel) primarily share similarities in
the lateral midface, mainly inzygomatic bone projection. Kabwe has
a more anteriorly projecting browridge, and a superiorlyeinferiorly
longer and more prognathic subnasal region. However, the
infraorbital topography of Petralona is more inated than Kabwe
(Fig. 3l: Kabwe is green). Lastly, Shanidar 5 and Petralona (Fig. 3me
o) share a similar degree of mid and lower facial prognathism,
infraorbital ination, and nasal aperture width. The main difference
between these two specimens is in the morphology of their brow-
ridge. Petralona expresses a more anteriorly projecting browridge
and upper face (Fig. 3o: Shanidar 5 is red).
The results of the PCA in Procrustes shape space of the complete
sample (i.e., subadult modern humans and Neanderthals) are
shown in Fig. 4. The rst three PCs were plotted and represent
Figure 3. Procrustes superimposition of each of the Middle Pleistocene hominins on its nearest neighbor in shape space. The Procrustes shape information for each specimen is
contained in the landmark and semilandmarks and everything in between these landmarks is interpolated. The landmarks and semilandmarks are color-coded according to the
specimens surface color (dark yellow is associated with the bone colored surfaces; dark green with the green surfaces; and dark red with the red surfaces). Thus, when two different
colored landmarks are exposed this indicates overlapping (i.e., similar) morphology. Different surface morphologies are present when only one landmark color and corresponding
surface is visible. Nearest neighbors are indicated by the arrows in Fig. 2. The full names and abbreviations for the fossil specimens are listed in Table 1. (For interpretation of the
references to colour in this gure legend, the reader is referred to the web version of this article.)
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 730
60.0% of total shape variance. PC 1 is correlated with size
(r y 0.79). The solid lines represent the ontogenetic allometric
trajectories for each of the four modern human populations, the
Neanderthal and mean modern human ontogenetic allometric
trajectories, and the MPh static allometric trajectory. The shaded
region represents the 95% single prediction interval of the modern
human regressions of shape on the natural logarithm of centroid
size. It is evident that all Neanderthal and MPh adults fall outside
this condence interval for allometric scaling. The thin lines
represent the bootstrapped estimates of the modern human allo-
metric trajectory computed from subsamples that simulate the
Neanderthal sample composition.
None of the Neanderthal or recent modern human ontogenetic
allometric trajectories coincide. Because each ontogenetic
trajectory has a unique starting point, this plot shows that pop-
ulation and species-specic facial morphology is present before
permanent teeth erupt. As in Fig. 2, clusters of temporal groups are
also apparent when including the entire sample (i.e., all subadults
included) although the MPh and Neanderthals overlap. Apart from
Qafzeh 9, the early and Upper Paleolithic modern humans fall
within the range of recent human variation.
The variance explained by ontogenetic allometry is 26.0% for
recent modern humans and 59.7% for Neanderthals. The angle
between the modern human average ontogenetic trajectory and
the Neanderthal ontogenetic trajectory in the subspace of the rst
three PCs is 28.0

. A permutation test using all dimensions of shape


space, however, reveals that the slopes of the Neanderthal and
mean modern human (MH) ontogenetic trajectories are not
Figure 4. PCA in shape space including subadult and adult fossil and modern humans. The rst three PCs are plotted. PC 1 represents 32.5% of total shape variance, PC 2 represents
17.8%, and PC 3 represents 8.8%. The lines indicate the ontogenetic allometric trajectories for Neanderthals (in red), the mean modern human trajectory (in dark gray), and the
trajectory for each modern human population (in black). The static allometric trajectory for the MPhs is in green. The trajectories were calculated by regressing shape on the natural
logarithm of centroid size. The specimen color-coding is the same as in Fig. 2. The shaded region represents the 95% single prediction interval of the modern human regressions of
shape on the natural logarithm of centroid size. The thin gray lines represent the bootstrapped estimates of the modern human allometric trajectory computed from subsamples
that simulate the Neanderthal sample composition. (For interpretation of the references to colour in this gure legend, the reader is referred to the web version of this article.)
Figure 5. PCA in form space of subadults and adult fossil and modern humans. Form space includes the log centroid size for each individual. The rst three PCs are plotted. PC 1
represents 79.0% of total form variance, PC 2 represents 5.0%, and PC 3 represents 3.3%. The lines indicate the ontogenetic allometric trajectories for Neanderthals (in red), the mean
modern human trajectory (in dark gray), and the trajectory for each modern human population (in black). The static allometric trajectory for the MPhs is in green. The trajectories
were calculated by regressing form on the natural logarithm of centroid size. The specimen color-coding is the same as in Fig. 2. The shaded region represents the 95% single
prediction interval of the modern human regressions of shape on the natural logarithm of centroid size. The thin gray lines represent the bootstrapped estimates of the modern
human allometric trajectory computed from subsamples that simulate the Neanderthal sample composition. The two circles demonstrate where the predicted shapes of a modern
human overgrown along the mean modern human ontogenetic allometric trajectory to the size of Kabwe and Bodo plot in the PCA (see text for more information). (For inter-
pretation of the references to colour in this gure legend, the reader is referred to the web version of this article.)
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 731
signicantly different from zero (p < 0.31). Therefore, we cannot
reject the null hypothesis that the mean modern human and
Neanderthal ontogenetic allometric trajectories are parallel.
To further explore the effects of size on facial morphology, we
performed a PCA in Procrustes form space (Fig. 5). The rst three
PCs represent 85.0% of total variance. The Neanderthal and modern
human mean ontogenetic trajectories do not coincide. The angle
between the two trajectories in the subspace of the rst three PCs is
8.8

. However, a permutation test cannot reject the null hypothesis


that these trajectories are parallel (p <0.35). As expected in a form-
space analysis, PC 1 is highly correlated with centroid size
(r 0.99). The variance explained by ontogenetic allometry in
recent modern humans is 80.1% and in Neanderthals it is 92.4%. As
in shape space, the shaded region in Fig. 5 represents the 95% single
prediction interval of the modern human regressions of formon the
natural logarithm of centroid size, and, as in shape space, all
Neanderthal and MPh adults fall outside this modern human
condence interval for allometric scaling. The thin lines represent
the bootstrapped estimates of the modern human allometric
trajectory computed from subsamples that simulate the Neander-
thal sample composition.
The Neanderthal infant Pech de LAz and the youngest recent
MH subadult individuals cluster at the left end of the plot (indi-
cating smaller size) and the individuals with larger faces, e.g., Bodo,
Petralona, and Kabwe fall at the opposite end. The MPhs fall at the
end of the Neanderthal ontogenetic allometric trajectory and the
orientation of the Neanderthal ontogenetic and MPh static trajec-
tories appear similar. The position of the larger MPhs, Kabwe,
Petralona and Bodo suggests that allometric scaling along
a common trajectory explains some differences in facial
morphology between them and Neanderthals. The early MH and
Upper Paleolithic humans cluster with the adult recent MH.
Table 5 lists the angles between the ontogenetic allometric
trajectories for each of the four modern human populations. The
angles were computed in the subspace of the rst three PCs to
make them easier to interpret. The permutation tests were
computed using all dimensions of shape space and Procrustes form
space, respectively. The results of the permutation test indicate that
in both shape and formspace, the trajectories of each of the modern
human populations is signicantly different from zero (i.e., not
parallel). The only exceptions are the angles between the Point
Hope and Khoisan populations (shape space: angle 4.3

, p <0.23;
form space: angle 18.5

, p < 0.07).
Developmental simulations
To visualize the allometric shape changes that occur in the face
of recent modern humans, the modern human mean shape was
calculated and grown to the size of Bodo, the largest specimen in
our sample, following the modern human mean ontogenetic allo-
metric trajectory. Fig. 6def illustrates the allometric shape changes
that occur in recent modern humans. Developmental simulations
were also performed on Neanderthals using the Neanderthal mean
facial shape and its ontogenetic allometric trajectory. These
developmental simulations are depicted in Fig. 7def. Table 6 lists
the facial features affected by allometry in both recent modern
humans and Neanderthals. Among the recent modern humans in
our sample, the Point Hope group has the largest faces according
centroid size. Therefore, to verify that this population was not
driving our results, we removed them and recalculated the mean
modern human developmental allometric trajectory. The same
allometric shape changes occurred in modern humans when the
Point Hope sample was not included. The following facial shape
changes occur as facial size increases in recent modern humans and
Neanderthals: 1) a more receding and narrower frontal bone, 2)
a more robust frontal process of the zygomatic, 3) an anterolateral
projection of the zygomatic, 4) an anteriorly projecting nasal bone
and a more prognathic nasal aperture, 5) a superoinferiorly smaller
orbital length, and 6) the subnasal region becomes superoinferiorly
elongated and more prognathic. The main differences between
recent modern humans and Neanderthals are in the shape of the
browridge and nasal aperture. In Neanderthals, the entire brow-
ridge becomes more anteriorly projecting, not only the glabellar
and lateral regions as in modern humans, and the nasal aperture
becomes wider.
Fig. 8 depicts a modern human (in bone color) grown to the size
of Kabwe (Fig. 8a, b) and Bodo (Fig. 8c, d), and superimposed on the
latter two MPhs. Both fossil specimens have a more projecting and
laterally expanded browridge, prognathic midface and inated
infraorbital region compared with a modern human grown to their
size. Additionally, in the inferior perspective Kabwes infraorbital
plate and zygomatic bone appear to be more sagittally rotated and
its nasal aperture is more anteriorly projecting. By contrast, an
overgrown modern human has an entirely different facial archi-
tecture maintaining two planes: a coronal infraorbital plate
combined with sagittally oriented lateral nasal walls, as well as
a less projecting nasal aperture. In these features, Bodo also appears
to express two midfacial planes like the modern human condition.
However, it exhibits a much more projecting nasal aperture and
a more inated infraorbital surface topography than a modern
human grown to its size.
Fig. 9 visualizes the mean adult Neanderthal facial shape (in
bone color) grown to the size of Kabwe and Bodo, and super-
imposed on them (Kabwe e a, b e and Bodo e c, d). Kabwe has
a more robust and projecting browridge and zygomatic bone when
compared with the scaled Neanderthal. Additionally, its zygomatic
bone is more anterolaterally projecting and the Neanderthal
infraorbital surface topography is more inated. Despite these
differences, however, they share a similar orientation of the
infraorbital plate and zygomatic body. The comparisons between
Bodo (Fig. 9c, d) and the scaled Neanderthal are similar to those of
Kabwe (Fig. 9a, b). The main differences are in the projection of the
browridge and the frontal process of the zygomatic. These features
are less anteriorly projecting when compared with the scaled
Neanderthal. Additionally, Bodos zygomatic body is more ante-
rolaterally projecting.
Table 5
Angles between the ontogenetic allometric trajectories of the four recent modern human populations.
a
Africa North America
b
Europe North America
c
Africa 0 23.7

(p < 0.003) 30.8

(p < 0.0005) 18.5

(p < 0.07)
North America
b
6.3

(p < 0.006) 0 19.9

(p < 0.004) 19.7

(p < 0.004)
Europe 8.4

(p < 0.0005) 5.3

(p < 0.021) 0 33.1

(p < 0.005)
North America
c
4.3

(p < 0.23) 5.9

(p < 0.004) 9.0

(p < 0.001) 0
a
The angles were calculated in the subspace of the rst three PCs in shape (upper) and form space (lower). Signicant values are denoted in bold.
b
Canyon del Muerto (Arizona) and Grand Gulch (Utah)
c
Point Hope (Alaska)
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 732
Figure 6. Ontogenetic allometric shape changes associated with recent modern humans. The mean adult modern human shape was calculated and grown to the size of Bodo
following the mean modern human ontogenetic allometric trajectory. aec: modern human mean shape in a) anterior; b) lateral; and c) inferior orientations. def: scaled modern
human mean shape in d) anterior; e) lateral; and f) inferior orientations. The main allometric shape changes in modern humans include a narrowing of the frontal bone, a more
anteriorly projecting glabellar region, an anterolateral projection of the zygomatic, a more prognathic nasal region, and changes in subnasal morphology. See text for a more detailed
description of shape changes.
Figure 7. Ontogenetic allometric shape changes associated with Neanderthals. The mean adult Neanderthal shape was calculated, warped to the surface of La Ferrassie 1 and grown
to the size of Bodo following the Neanderthal ontogenetic allometric trajectory. aec: Neanderthal mean shape in a) anterior, b) lateral, and c) inferior orientations. def: scaled
Neanderthal mean shape in d) anterior, e) lateral, and f) inferior orientations. Neanderthals shared many allometric shape changes with modern humans. The main allometric shape
differences between Neanderthals and modern humans are in the browridge and nasal aperture morphology (see text and Fig. 6 legend for a more detailed description).
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 733
Discussion
The aim of this study was threefold: 1) to explore temporal
trends in facial morphology in order to gain a clearer understanding
of the polarity of facial features, 2) to describe the allometric
patterning of facial shape in archaic and modern humans, and 3) to
test the hypothesis that differences in facial shape between archaic
and modern humans are attributable to ontogenetic scaling along
a common growth allometry.
Temporal trends and evolutionary models
Our PCA results and nearest neighbor calculations illustrate
a temporal trend in facial morphology during Pleistocene human
evolution. Although MPhs and Neanderthals overlap with one
another in Procrustes shape space (Fig. 2), each individual is most
similar to a member within its respective group. The main excep-
tion to this is Arago 21, which is most similar to La Chapelle-aux-
Saints. Additionally, our PCA results show that there is no clear
Table 6
Recent modern humans and Neanderthal facial features affected by allometry.
a
Anatomical region Recent modern humans Neanderthals
Upper face Glabellar region and lateral browridge becomes more
robust and anteriorly projecting
Entire browridge becomes more anteriorly projecting
Receding frontal bone X
b
Narrower frontal bone X
Midface Widening of the frontal process of the zygomatics X
Anterior and lateral projection of zygomatics X
Superoinferiorly taller and mediolaterally narrower
nasal aperture
Wider nasal aperture breadth
More projecting nasal bones and prognathic nasal aperture X
Smaller orbits in superioreinferior dimensions X
Lower face Superoinferiorly longer subnasal region X
Prognathic subnasal region X
a
Allometric shape changes are described as a comparison between the modern human and Neanderthal mean shape and their respective scaled versions (e.g., the glabellar
region and lateral browridge becomes more robust and anteriorly projecting with size when compared with the modern human or Neanderthal mean shape).
b
X indicates that the Neanderthal condition is the same as in the modern humans.
Figure 8. Overgrown modern human superimposed on Kabwe and Bodo. The mean adult modern human shape was calculated, grown to the size of Kabwe and to the size of Bodo
and superimposed onto them. Modern human is bone color and Kabwe (a, b) and Bodo (c, d) are in transparent purple. (For interpretation of the references to colour in this gure
legend, the reader is referred to the web version of this article.)
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 734
distinction between African and European MPhs. This is consistent
with previous morphometric studies (e.g., Stringer, 1974, 1983;
Arsuaga et al., 1997; Rightmire, 1998, 2007, 2008; Mounier et al.,
2009; Harvati, 2009a; Harvati et al., 2010). For example, our near-
est neighbor calculations support morphological and metric simi-
larities between Petralona and Kabwe (Stringer et al., 1979; Murrill,
1981; Stringer, 1983; Rightmire, 2001; Harvati, 2009a; Friess, 2010;
Harvati et al., 2010, 2011), as well as between Bodo and Sima de los
Huesos 5 (Arsuaga et al., 1997).
In the PCA in formspace, Arago 21 and Sima de los Huesos 5 plot
on one end of the MPh static allometric trajectory, and Petralona,
Kabwe and Bodo fall on the opposite end, indicating that the latter
group of hominins express larger faces. This signal may suggest
sexual dimorphism within the MPhs sample, with Arago 21 and
Sima de los Huesos 5 being females, and Bodo, Kabwe and Petra-
lona being males. Arago 21, Bodo, Kabwe, and Petralona have been
described as males in earlier studies (de Lumley and de Lumley,
1973). However, the sex assignment of Sima de los Huesos 5 is
more ambiguous. Compared with other specimens from the SH
assemblage, as well as other MPhs and Neanderthals, Arsuaga et al.
(1997) notes that the calvarial and cranial base measurements
suggest that Sima de los Huesos 5 is either a small male or a large
female. On the other hand, Rosas (1997) assigns the Sima de los
Huesos 5 mandible to a male. While we cannot rule out the
possibility that the size discrepancy between these MPhs is due to
sexual dimorphism, the variation seen in these fossils could also be
attributed to interpopulation variation, directional trends or
diachronic uctuations.
As noted previously, Arago 21 is most similar to La Chapelle,
Sima de los Huesos 5 is most similar to Arago 21, and Shanidar 5 is
most similar to Petralona. This latter point could suggest that the
European Middle Pleistocene hominins display incipient Neander-
thal features, as put forward in the accretion hypothesis. According
to our nearest neighbor comparisons and PCA visualizations in
shape space, the European Middle Pleistocene hominins share
a more parasagittally rotated midface when compared with the
African Middle Pleistocene hominins and a slightly more inated
infraorbital surface topography. Both the African and European
Middle Pleistocene hominin groups share a wide nasal aperture
and midfacial prognathism with Neanderthals. Our results indicate
these latter two features are largely inuenced by facial size and are
most likely symplesiomorphies.
Whether the African and European Middle Pleistocene speci-
mens are conspecic (i.e., H. heidelbergensis) or the European MPhs
are pre-Neanderthals that look like African H. rhodesiensis due to
symplesiomorphies cannot be clearly determined based on our
data. A more comprehensive analysis including the entire cranium
could provide us with greater insight. However, one intriguing
alternative possibility is that Arago 21, Sima de los Huesos 5, and
possibly Petralona are early members of the Neanderthal lineage
and Bodo and Kabwe form a Middle Pleistocene group that is either
part of a separate lineage leading to H. sapiens, or similar in facial
Figure 9. Overgrown Neanderthal superimposed on Kabwe and Bodo. The mean adult Neanderthal facial shape was calculated, grown to the size of Kabwe and to the size of Bodo
and superimposed onto them. Neanderthal is bone color and Kabwe (a, b) and Bodo (c, d) are in transparent blue. (For interpretation of the references to colour in this gure legend,
the reader is referred to the web version of this article.)
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 735
form to the common ancestral species of H. sapiens and
H. neanderthalensis. Under this model, H. rhodesiensis is ancestral to
H. sapiens in Africa. According to this model, substantial levels of
genetic drift are responsible for the origins of Neanderthals in
Europe and modern humans in Africa (Endicott et al., 2010). The
primary problemwith this model is the date of the SHmaterial. The
SH material clearly exhibits Neanderthal dental and mandibular
features (Arsuaga et al., 1997; Rosas, 2001; Martinn-Torres et al.,
2007, 2012) and are too derived in the Neanderthal direction for
such an early geological date of ca. 600 ka (thousands of years ago)
(Endicott et al., 2010; Stringer, 2012). If the taphonomic doubts (see
Endicott et al., 2010; Stringer, 2012) concerning the SH material are
substantiated and the previous dates of between 200 and 400 ka
(Arsuaga et al., 1997; Bischoff et al., 2003) are correct, then this
scenario becomes more plausible. Alternatively, Martinn-Torres
et al. (2012) propose that more than one hominin lineage may
have coexisted during the Middle Pleistocene in Europe. Their
comprehensive study on the SH dentition shows that the SH teeth
are more Neanderthal-like than any of the other pene-
contemporaneous Middle Pleistocene samples (e.g., Mauer or
Arago), and in some aspects even more derived than Neanderthals.
However, the main problem with this interpretation is that the
relative ages of Mauer, Arago and the Sima de los Huesos material is
highly debated (Hublin, 2009; Endicott et al., 2010).
Our results conrm previous morphometric studies (e.g.,
Howells, 1973; Stringer, 1974; Weber et al., 2006; Athreya, 2009;
Harvati, 2009b; Harvati et al., 2010; Stanseld and Gunz, 2011)
showing that modern humans, including the early modern
humans, have a facial morphology that is distinct from archaic
humans. All of our modern human groups are most similar to one
another based on nearest neighbor calculations, reecting a high
degree of homogeneity among these groups, as well as a clear
H. sapiens lineage separation fromarchaic humans at this later time
period (ca. 200 ka). The similarities in facial morphology between
the three temporal groups of modern humans (early, Upper
Paleolithic and recent modern humans) in our study indicate that
the characteristically modern human facial anatomy had already
evolved in Jebel Irhoud 1, around 170 ka. This is not the rst study to
recognize the modern-like morphology of the Jebel Irhoud 1 face
(Hublin, 1992; Harvati, 2009b; Harvati and Hublin, 2012). However,
the transitional and early modern human specimens, especially
Jebel Irhoud 1, express morphologically more primitive frontal
bones (Hublin, 1992; Athreya, 2009; Freidline et al., 2012). While
Smith (2002) termed the African fossils from the time span of ca.
250 to 160 ka (e.g., Jebel Irhoud 1e4, Florisbad, Ngaloba, Eliye
Springs, Guomde and Omo Kibish 2) as the African Transitional
Group (ATG), other researchers (Bruer, 2008; Rightmire, 2009;
Gunz et al., 2009a) have proposed that Jebel Irhoud, along with
Laetoli 18, Florisbad, Ileret, and several other Late Pleistocene
African specimens represent the beginning of the H. sapiens lineage.
Our results tentatively support this.
Allometric patterning of facial shape in recent and archaic humans
Our results are consistent with a growing body of evidence
demonstrating that taxon-specic craniofacial morphology among
hominin species developed prenatally or very early postnatally
(e.g., Ponce de Len and Zollikofer, 2001; Ackermann and Krovitz,
2002; Lieberman et al., 2002; Strand Viarsdttir et al., 2002;
Williams et al., 2002; Krovitz, 2003; Bastir and Rosas, 2004;
Mitteroecker et al., 2004; McNulty et al., 2006; Gunz et al., 2010,
2011, 2012; Gunz, 2012).
A comparison of the developmental simulations in Figs. 6 and 7
demonstrate that modern humans and Neanderthals share many
aspects of ontogenetic allometric patterning in facial shape changes
(see Table 6), but they are not identical. Shape changes associated
with an increase in size in these two groups include a receding and
narrower frontal bone, a more robust frontal process of the zygo-
matic, an anterolateral projection of the zygomatic, an anteriorly
projecting nasal bone and a more prognathic nasal aperture,
a superoinferiorly smaller orbit, and the subnasal region becomes
superoinferiorly elongated and more prognathic. Strand
Viarsdttir et al. (2002) also identied many of these features as
aspects of ontogenetic allometry shared by a geographically diverse
modern human sample including a decrease in relative orbit size,
a relative expansion of the zygomatic, and a relative reduction in
frontal breadth. The general allometric pattern of robusticity
associated with browridge size in Neanderthals and modern
humans is not overly surprising. Data from primates indicate that
circumorbital features (e.g., browridge height; Ravosa, 1991) and
neurocranial vault thickness (Gauld, 1996) scale with strong posi-
tive allometry. Larger anthropoids will have disproportionately
larger browridges simply because they are big.
We found localized allometric differences in facial morphology
between Neanderthals and modern humans; in particular, details
of the nasal aperture and browridge morphology. These results are
consistent with several geometric morphometric analyses on
humans and primates that have found that both early postnatal
cranial morphology and later postnatal growth contribute to
further differentiate populations and species (Richtsmeier et al.,
1993; OHiggins and Jones, 1998; OHiggins et al., 2001; Strand
Viarsdttir et al., 2002; Bastir and Rosas, 2004; Cobb and
OHiggins, 2004; Strand Viarsdttir and Cobb, 2004; Bastir et al.,
2007).
In a recent ontogenetic study examining the relationship
between nasal breadth, anterior palatal breadth and facial prog-
nathism, Holton and Franciscus (2008) found that variations in
nasal breadth in recent and fossil Homo is determined by basione
prosthion length rather than anterior palatal breadth dimensions.
Lower facial prognathism, a plesiomorphic retention in Neander-
thals (Trinkaus, 1987, 2003), explains their wide nasal aperture. Our
results demonstrate that both facial prognathism and nasal aper-
ture breadth are inuenced by facial size in Neanderthals. Thus, the
broad nasal apertures in the MPhs are most likely due to their large,
prognathic faces.
In our study, the Neanderthal and modern human allometric
trajectories are not coincident and the ontogenetic trajectories
diverge. However, the angular difference between the modern
human and Neanderthal allometric trajectories does not reach the
level of statistical signicance. Therefore, our results cannot reject
the null hypothesis that the Neanderthal and modern human
postnatal ontogenetic allometric trajectories are parallel. While our
results are consistent with other comparative morphometric
studies that have reported parallel ontogenetic postnatal trajecto-
ries between modern humans and Neanderthals (Ponce de Len
and Zollikofer, 2001; Williams et al., 2002; Krovitz, 2003), the
inability to reject the null hypothesis may be an artifact of our small
subadult Neanderthal sample size (see McNulty et al., 2006 for
a discussion of this issue). This is an especially strong possibility
since we nd signicant differences between the slopes in three out
of the four modern human populations. Due to the small Nean-
derthal subadult sample size, there is a lot of potential error in its
slope (as can be seen by our bootstrap simulations in Figs. 4 and 5),
and the position and orientation of the Neanderthal ontogenetic
allometric trajectory is strongly inuenced by Pech-de-lAz and
Teshik-Tash. Therefore, we caution against over-interpretation of
this particular result. Future studies with a larger sample of
Neanderthal subadults would greatly improve the accuracy of these
results. Additionally, the discrepancy between our ndings and
those of previous studies (e.g., Bastir et al., 2007) may be explained
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 736
by differences in the anatomical region analyzed (e.g., neuro-
cranium versus face) or by differing landmark data sets (see Bastir
and Rosas, 2004 for further discussion). For example, in their
geometric morphometric study on mandibular ontogeny, Bastir
et al. (2007) found modern human and Neanderthal postnatal
ontogenetic allometric trajectories to be signicantly different from
one another. They argue that this is strong evidence for divergent
postnatal ontogenetic shape changes between modern humans and
Neanderthals. Different anatomical regions have different rates and
patterns of growth. For example, longitudinal data reveal that
among modern humans adult brain size and by extension neuro-
cranial growth is nearly achieved (90e95%) by six years of age,
whereas facial growth continues into puberty, especially among
males (Bulygina et al., 2006).
The position of the MPhs in shape and formspace in conjunction
with the position and orientation of their static allometric trajec-
tory suggests that allometric scaling explains large-scale differ-
ences in facial morphology between MPhs and Neanderthals.
However, scaling cannot explain the morphological differences
between the latter two groups and recent modern humans.
Lieberman (2008) noted that in many ways MPhs are scaled-up
versions of H. erectus with larger brains and faces, and that the
evolutionary transformation from a MPh to a modern human
involves more complex changes in the underlying cranial archi-
tecture. In a morphometric study on cranial development and
integration, Lieberman et al. (2002) demonstrated that the main
architectural craniofacial differences between archaic and modern
humans can be attributed to differences in the length of the middle
cranial fossa and cranial base angles, and that these features explain
facial retraction in modern humans.
The results of our developmental simulations (Fig. 6 and 7)
suggest that the following features that are present in the largest
MPh faces (Bodo, Petralona and Kabwe) can be explained by
ontogenetic allometric scaling along a shared archaic human
trajectory (i.e., MPh and Neanderthal trajectory): narrow and
receding frontal bone; wider, more projecting and entirely more
robust browridge, especially the lateral portion; robust frontal
process of the zygomatic; wide nasal aperture; and overall facial
prognathism. There are several facial features that are commonly
shared between MPhs, especially the European MPhs, and Nean-
derthals that cannot entirely be explained by allometry including
parasagittally oriented zygomatic bones and infraorbital plate, and
at or inated infraorbital surface topography. According to the
results of our Neanderthal developmental simulations in Fig. 9,
these features are less pronounced in the African MPhs when
compared with a Neanderthal face, even when size differences are
accounted for. Therefore, the degree of their expression can be used
to differentiate between Neanderthals and MPhs.
In a recent geometric morphometric study evaluating the
accretion hypothesis, Harvati et al. (2010) proposed that a convex
and receding infraorbital prole may be the primitive condition for
Neanderthals. They found that the infraorbital prole of African and
European Middle Pleistocene hominins are nearly indistinguish-
able and similar, but not as inated or parasagittally rotated, to the
Neanderthal mean conguration. These authors reached a similar
conclusion regarding the projection and position of glabella. MPhs,
Neanderthals and Jebel Irhoud 1 showan anteroinferior placement
of glabella, a condition different from H. erectus and recent modern
humans (Harvati et al., 2010). Their results indicate that this
morphology is most pronounced in the African MPhs (Harvati et al.,
2010). Our study and others (e.g., Rosas and Bastir, 2002; Spoor
et al., 2007) demonstrate that there is an allometric relationship
between glabellar projection and size. Thus, the large facial sizes of
Bodo and Kabwe most likely explain why the African MPhs show
the most anteriorly projecting glabellae.
In a recent study testing the effects of evolutionary and static
allometry on the midface, Maddux and Franciscus (2009)
concluded that differences in infraorbital surface topography
across the genus Homo can be explained by allometric scaling.
However, we show that a modern human grown to the size of Bodo
clearly displays a less inated infraorbital surface topography than
Bodo, and a Neanderthal grown to the size of Bodo expresses
greater ination in this region. This is most likely due to the
fundamental differences in their craniofacial architecture. Accord-
ing to our results, features like infraorbital surface depression or
ination can be used to distinguish between modern and archaic
populations. One key difference between our studies is that
Maddux and Franciscus (2009) scaled against infraorbital region
size rather than facial size. This may partly explain the discrep-
ancies between our studies. Future work testing facial integration
in MPhs and Neanderthals would provide greater insight to how
this feature covaries with the surrounding facial morphology and
its taxonomic value.
Conclusion
In this paper, we set out to explore temporal trends in facial
morphology among Middle to Late Pleistocene humans to gain
a clearer understanding of the polarity of facial features. We
identify patterns of allometric shape changes in facial morphology
in these hominins, and apply developmental simulations to
examine how size affects facial features.
Pleistocene humans can be divided into temporal groups
according to their facial morphology, however, the boundary
between the MPhs and Neanderthals is less obvious. Additionally,
our nearest neighbor calculations show that the African and
European MPhs share an overall similar facial morphology,
although Arago 21, Sima de los Huesos 5 and Petralona exhibit
several facial features that are intermediate between the African
Middle Pleistocene and Neanderthal condition. These features
include an intermediate ination of the infraorbital surface
topography and a parasagittally rotated midface. In line with the
predictions of the accretion hypothesis, this could suggest that the
European Middle Pleistocene hominins display incipient Neander-
thal facial features.
Our ndings contribute to a growing body of research that
highlights the important role of allometry in hominin craniofacial
development and evolution. The results of our growth studies
complement previous research showing that the distinctive
modern human and Neanderthal facial morphology developed
early in ontogeny, by at least two years of age. While both the
position of the Neanderthal and recent modern human postnatal
ontogenetic trajectories and developmental simulations suggests
that they share a similar pattern of ontogenetic allometric scaling,
the slopes of three out of four modern human ontogenetic allo-
metric trajectories analyzed in our study are signicantly different
from one another. This indicates that in modern humans, both
prenatal and postnatal ontogeny contributes to further differentiate
populations. By extension, this is most likely true for Neanderthals
as well. The lack of statistical signicance between the Neanderthal
and modern human ontogenetic allometric trajectories is probably
due to the restricted subadult and adult Neanderthal sample sizes.
Further studies with a larger Neanderthal sample size are needed to
verify this.
The overall position of the MPhs in shape and form space in
conjunction with the direction and orientation of their static allo-
metric trajectory suggests that ontogenetic allometric scaling
explains large-scale differences in facial morphology between
MPhs and Neanderthals. Features present in the MPhs that are due
to allometric scaling include a narrow and receding frontal bone;
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 737
a wider, more projecting and entirely more robust browridge,
especially the lateral portion; a robust frontal process of the
zygomatic; a wide nasal aperture; and overall facial prognathism.
Whereas, the parasagittally oriented zygomatic bones and
infraorbital plate and at or inated infraorbital surface topography
present in MPh faces cannot be explained by allometric scaling.
Furthermore, ontogenetic allometric scaling does not explain
differences between archaic and modern humans. Jebel Irhoud 1
undoubtedly expresses a modern human pattern of facial
morphology, reecting a clear modern human lineage separation at
this later time period.
Acknowledgments
We thank all of the curators who have so generously allowed us
to scan the fossil and modern human material from the following
institutions: American Museum of Natural History (New York),
Aristotle University of Thessaloniki, Hrvatski Prirodoslovni Muzej
(Zagreb), Institut de Palontologie Humaine (Paris), Muse Arch-
ologique (Rabat), Muse de LHomme (Paris), Museo Nazionale
Preistorico Etnograco L. Pigorini (Rome), National Museum of
Ethiopia (Addis Ababa), Natural History Museum (London), Natur-
historisches Museum (Vienna), Peabody Museum at Harvard
University (Cambridge), Rheinisches Landesmuseum (Bonn), Iziko
South African Museum (Cape Town), and the University of Cape
Town. We also thank Pr. J.-L. Kahn from the Department of
Anatomy, Medicine Faculty, Strasbourg, for access to the osteolog-
ical collection and Pr. F. Veillon from the Department of Radiology,
Hautepierre Hospital, Strasbourg, for access to the medical scan-
ners. The authors show their deepest gratitude to Dr. Begun,
the Associate Editor, and the anonymous reviewers who have
improved the quality of this paper. We would also like to thank Dr.
Fred Grine for allowing us to use a collection of CT scans in our
modern human comparative sample, and Drs. Eric Delson and Will
Harcourt-Smith for their comments on earlier versions of this
manuscript. This work was supported by the Marie Curie Actions
grant MRTN-CT-2005-019564 EVAN, the Max Planck Society, the
National Science Foundation (0333415, 0513660 and 0851756), the
L.S.B. Leakey Foundation, and Sigma Xi. This is NYCEP Morpho-
metrics contribution 60.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jhevol.2012.08.002.
References
Ackermann, R.R., Krovitz, G.E., 2002. Common patterns of facial ontogeny in the
hominid lineage. Anat. Rec. 269, 142e147.
Akazawa, T., Muhesen, S., Dodo, Y., Kondo, O., Mizoguchi, Y., 1995. Neanderthal
infant burial. Nature 377, 585e586.
Alberch, P., Gould, S.J., Oster, G.F., Wake, D.B., 1979. Size and shape in ontogeny and
phylogeny. Paleobiol. 5, 296e317.
Arsuaga, J.L., Martinez, I., Gracia, A., Lorenzo, C., 1997. The Sima de los Huesos crania
(Sierra de Atapuerca, Spain). A comparative study. J. Hum. Evol. 33, 219e281.
Athreya, S., 2009. A comparative study of frontal bone morphology among Pleis-
tocene hominin fossil groups. J. Hum. Evol. 57, 786e804.
Bastir, M., OHiggins, P., Rosas, A., 2007. Facial ontogeny in Neanderthals and
modern humans. Proc. R. Soc. B 274, 1125e1132.
Bastir, M., Rosas, A., 2004. Facial heights: evolutionary relevance of postnatal
ontogeny for facial orientation and skull morphology in humans and chim-
panzees. J. Hum. Evol. 47, 359e381.
Bastir, M., Rosas, A., Gunz, P., Pea-Melian, A., Manzi, G., Harvati, K., Kruszynski, R.,
Stringer, C., Hublin, J.J., 2011. Evolution of the base of the brain in highly
encephalized human species. Nat. Commun. 2, 588. http://dx.doi.org/10.1038/
ncomms1593.
Benazzi, S., Douka, K., Fornai, C., Bauer, C.C., Kullmer, O., Svoboda, J., Pap, I.,
Mallegni, F., Bayle, P., Coquerelle, M., Condemi, S., Ronchitelli, A., Harvati, K.,
Weber, G.W., 2011. Early dispersal of modern humans in Europe and implica-
tions for Neanderthal behaviour. Nature 479, 525e528.
Berge, C., Penin, X., 2004. Ontogenetic allometry, heterochrony, and interspecic
differences in the skull of African apes, using tridimensional Procrustes analysis.
Am. J. Phys. Anthropol. 124, 124e138.
Bischoff, J.L., Shamp, D.D., Aramburu, A., Arsuaga, J.L., Carbonell, E., Bermdez de
Castro, J.M., 2003. The Sima de los Huesos hominids date to beyond U/Th
equilibrium (> 350 kyr) and perhaps to 400-500 kyr: new radiometric dates.
J. Archaeol. Sci. 30, 275e280.
Bischoff, J.L., Williams, R.W., Rosenbauer, R.J., Aramburu, A., Arsuaga, J.L., Garcia, N.,
Cuenca-Bescos, G., 2007. High-resolution U-series dates from the Sima de los
Huesos hominids yields 600/- kyrs: implications for the evolution of the early
Neanderthal lineage. J. Archaeol. Sci. 34, 763e770.
Bookstein, F.L., 1991. Morphometric Tools for Landmark Data: Geometry and
Biology. Cambridge University Press, Cambridge.
Bruer, G., 2008. The origin of modern anatomy: by speciation or intraspecic
evolution? Evol. Anthropol. 17, 22e37.
Brown, P., 1992. Recent human evolution in East Asia and Australasia. Phil. Trans. R.
Soc. Lond. B 337, 235e242.
Bruner, E., Ripani, M., 2008. A quantitative and descriptive approach to morpho-
logical variation of the endocranial base in modern humans. Am. J. Phys.
Anthropol. 137, 30e40.
Bulygina, E., Mitteroecker, P., Aiello, L., 2006. Ontogeny of facial dimorphism and
patterns of individual development within one human population. Am. J. Phys.
Anthropol. 131, 432e443.
Chen, T., Hedges, R.E.M., Yuan, Z., 1989. Accelerator radiocarbon dating for the upper
cave of Zhoukoudian. Acta Anthropol. Sinica 8, 216e221.
Cheverud, J.M., 1982. Relationships among ontogenetic, static, and evolutionary
allometry. Am. J. Phys. Anthropol. 59, 139e149.
Clark, J.D., De Heinzelin, J., Schick, K.D., Hart, W.K., White, T.D., Wolde Gabriel, G.,
Walter, R.C., Suwa, G., Asfaw, B., Vrba, E., 1994. African Homo erectus: old
radiometric ages and young Oldowan assemblages in the Middle Awash Valley,
Ethiopia. Science 264, 1907.
Cobb, S.N., OHiggins, P., 2004. Hominins do not share a common postnatal facial
ontogenetic shape trajectory. J. Exp. Zool. B 302, 302e321.
Cock, A.G., 1966. Genetical aspects of metrical growth and form in animals. Q. Rev.
Biol. 41, 131e190.
Collard, M., OHiggins, P., 2001. Ontogeny and homoplasy in the papionin monkey
face. Evol. Dev. 3, 322e331.
Cook, J., Stringer, C.B., Currant, A.P., Schwarcz, H.P., Wintle, A.G., 1982. A review of
the chronology of the European Middle Pleistocene hominid record. Am. J. Phys.
Anthropol. 25, 19e65.
Dean, D., Hublin, J.J., Holloway, R., Ziegler, R., 1998. Onthe phylogenetic positionof the
pre-Neandertal specimen from Reilingen, Germany. J. Hum. Evol. 34, 485e508.
Dryden, I.L., Mardia, K.V., 1998. Statistical Shape Analysis. John Wiley & Sons,
Chichester.
Endicott, P., Ho, S.Y., Stringer, C., 2010. Using genetic evidence to evaluate four
palaeoanthropological hypotheses for the timing of Neanderthal and modern
human origins. J. Hum. Evol. 59, 87e95.
Falgures, C., Yokoyama, Y., Shen, G., Bischoff, J.L., Ku, T.L., de Lumley, H., 2004. New
U-series dates at the Caune de lArago, France. J. Archaeol. Sci. 31, 941e952.
Freidline, S.E., Gunz, P., Jankovic, I., Harvati, K., Hublin, J.J., 2012. A comprehensive
morphometric analysis of the frontal and zygomatic bone of the Zuttiyeh fossil
from Israel. J. Hum. Evol. 62, 225e241.
Friess, M., 2010. Calvarial shape variation among Middle Pleistocene hominins: an
application of surface scanning in paleoanthropology. C.R. Palevol. 9, 435e443.
Gauld, S.C., 1996. Allometric patterns of cranial bone thickness in fossil hominids.
Am. J. Phys. Anthropol. 100, 411e426.
Gonzalez, P.N., Perez, S.I., Bernal, V., 2011. Ontogenetic allometry and cranial shape
diversication among human populations from South America. Anat. Rec. 294,
1864e1874.
Gould, S.J., 1966. Allometryandsize inontogenyandphylogeny. Biol. Rev. 41, 587e640.
Gould, S.J., 1968. Ontogeny and the explanation of form: an allometric analysis.
Paleontol. Soc. Mem. 2, 81e98.
Gould, S.J., 1975. Allometry in primates, with emphasis on scaling and the evolution
of the brain. Contrib. Primatol. 5, 244e292.
Gould, S.J., 1977. Ontogeny and Phylogeny. Harvard University Press, Cambridge.
Grn, G.R., Stringer, C.B., 1991. ESR dating and the evolution of modern humans.
Archaeometry 33, 153e199.
Grn, G.R., Stringer, C.B., McDermott, F., Nathan, R., Porat, N., Robertson, S., Taylor, L.,
Mortimer, G., Eggins, S., McCulloch, M., 2005. U-series and ESR analyses of
bones and teeth relating to the human burials from Skhul. J. Hum. Evol. 49,
316e334.
Gunz, P., 2005. Statistical and geometric reconstruction of hominid crania: recon-
structing australopithecine ontogeny. Ph.D. dissertation, University of Vienna.
Gunz, P., 2012. Evolutionary relationships among robust and gracile australopiths:
an evo-devo perspective. Evol. Biol., 1e16.
Gunz, P., Bookstein, F.L., Mitteroecker, P., Stadlmayr, A., Seidler, H., Weber, G.W.,
2009a. Early modern human diversity suggests subdivided population structure
and a complex out-of-Africa scenario. Proc. Natl. Acad. Sci. 106, 6094e6098.
Gunz, P., Harvati, K., 2007. The Neanderthal chignon: variation, integration, and
homology. J. Hum. Evol. 52, 262e274.
Gunz, P., Mitteroecker, P., Bookstein, F., 2005. Semilandmarks in three dimensions.
In: Slice, D.E. (Ed.), Modern Morphometrics in Physical Anthropology. Plenum
Publishers, New York, pp. 73e98.
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 738
Gunz, P., Mitteroecker, P., Neubauer, S., Weber, G.W., Bookstein, F.L., 2009b. Princi-
ples for the virtual reconstruction of hominin crania. J. Hum. Evol. 57, 48e62.
Gunz, P., Neubauer, S., Golovanova, L., Doronichev, V., Maureille, B., Hublin, J.J., 2012.
A uniquely modern human pattern of endocranial development. Insights from
a new cranial reconstruction of the Neandertal newborn from Mezmaiskaya.
J. Hum. Evol. 62, 300e313.
Gunz, P., Neubauer, S., Maureille, B., Hublin, J.J., 2010. Brain development after
birth differs between Neanderthals and modern humans. Curr. Biol. 20,
R921eR922.
Gunz, P., Neubauer, S., Maureille, B., Hublin, J.J., 2011. Virtual reconstruction of
the Le Moustier 2 newborn skull. Implications for Neandertal ontogeny. Paleo
22, 155e172.
Harvati, K., 2001. The Neanderthal problem: 3-D geometric morphometric models
of cranial shape variation within and among species. Ph.D. Dissertation, City
University of New York.
Harvati, K., 2009a. Petralona: link between Africa and Europe? In: Schepartz, L.,
Bourbou, C., Fox, S. (Eds.), New Directions in the Skeletal Biology of Greece
Wiener Laboratory Series. ASCSA, Athens, pp. 31e47.
Harvati, K., 2009b. Into Eurasia: geometric morphometric analysis of the upper cave
(Zhoukoudian) specimens. J. Hum. Evol. 57, 751e762.
Harvati, K., Hublin, J.J., 2012. Morphological continuity of the face in the late Middle
and Late Pleistocene hominins from northwestern Africa: a 3D geometric
morphometric analysis. In: Hublin, J.J., McPherron, S.P. (Eds.), Modern Origins:
a North African Perspective. Springer, New York, pp. 179e188.
Harvati, K., Hublin, J.J., Gunz, P., 2010. Evolution of middle-late Pleistocene human
cranio-facial form: a 3-D approach. J. Hum. Evol. 59, 445e464.
Harvati, K., Panagopoulou, E., Runnels, C., 2009. The paleoanthropology of Greece.
Evol. Anthropol. 18, 131e143.
Harvati, K., Stringer, C., Grn, R., Aubert, M., Allsworth-Jones, P., Folorunso, C.A.,
2011. The Later Stone Age calvaria from Iwo Eleru, Nigeria: morphology and
chronology. Plos One 6, e24024.
Hedges, R.E.M., Housley, R.A., Bronk, C.R., Van Klinken, G.J., 1992. Radiocarbon
dates from the Oxford AMS system: archaeometry datelist 14. Archaeometry
34, 141e159.
Henry-Gambier, D., 2001. La Spulture des Enfants de Grimaldi (Baouss-Rouss):
Anthropologie et Palethnologie des Populations de la Fin du Palolitique
Suprieur. Comite des Travaux Historiques et Scientiques, Paris.
Holt, B.M., Formicola, V., 2008. Hunters of the Ice Age: the biology of upper
Paleolithic people. Am. J. Phys. Anthropol. (S47), 70e99.
Holton, N.E., Franciscus, R.G., 2008. The paradox of a wide nasal aperture in cold-
adapted Neandertals: a causal assessment. J. Hum. Evol. 55, 942e951.
Howells, W.W., 1973. Cranial Variation in Man: a Study by Multivariate Analysis of
Patterns of Differences among Recent Human Populations. Papers of the Pea-
body Museum of Archaeology and Ethnology. Harvard, Cambridge.
Hublin, J.-J., 1980. La Chaise Suard, Engis 2 et La Quina H8: developpement de la
morphologie occipitale externe chez lenfant pr-nandertalien et nan-
dertalien. C.R. Acad. Sci. Paris 291, 669e672.
Hublin, J.-J., 1992. Recent human evolution in northwestern Africa. Phil. Trans. R.
Soc. Lond. B 337, 185e191.
Hublin, J.-J., 1998. Climate changes, paleogeography, and the evolution of the
Neandertals. In: Akazawa, T., Bar-Yosef, O. (Eds.), Neandertals and Modern
Humans in Western Asia. Plenum Press, New York, pp. 295e310.
Hublin, J.-J., 2009. Out of Africa: modern human origins special feature: the origin of
Neandertals. Proc. Natl. Acad. Sci. 106, 16022e16027.
Hublin, J.-J., Spoor, F., Braun, M., Zonneveld, F., Condemi, S., 1996. A late Neanderthal
associated with Upper Palaeolithic artefacts. Nature 381, 224e226.
Huxley, J.S., 1932. Problems of Relative Growth. Methuen, London.
Inouye, S.E., Shea, B.T., 1997. Whats your angle? Size correction and bar-glenoid
orientation in Lucy (A.L. 288-1). Int. J. Primatol. 18, 629e650.
Klein, R.G., 1994. Southern Africa before the Iron Age. In: Corruccini, R.S. (Ed.),
Integrative Paths to the Past: Paleoanthropological Advances in Honor of F.
Clark Howell. Prentice Hall, Engelwood Cliffs, pp. 471e519.
Klein, R.G., 1999. The Human Career. The University of Chicago Press, Chicago.
Klingenberg, C.P., 1998. Heterochrony and allometry: the analysis of evolutionary
change in ontogeny. Biol. Rev. 73, 79e123.
Klingenberg, C.P., Zimmermann, M., 1992. Static, ontogenetic, and evolutionary
allometry: a multivariate comparison in nine species of water striders. Am. Nat.
140, 601e620.
Krovitz, G.E., 2003. Shape and growth differences between Neandertals and modern
humans: grounds for a species-level distinction. In: Thompson, J.L., Krovtiz, G.E.,
Nelson, A.J. (Eds.), Patterns of Growth and Development in the Genus Homo.
Cambridge University Press, Cambridge, pp. 320e342.
Leigh, S.R., 2006. Cranial ontogeny of Papio baboons (Papio hamadryas). Am. J. Phys.
Anthropol. 130, 71e84.
Leigh, S.R., Cheverud, J.M., 1991. Sexual dimorphism in the baboon facial skeleton.
Am. J. Phys. Anthropol. 84, 193e208.
Leigh, S.R., Shah, N.F., Buchanan, L.S., 2003. Ontogeny and phylogeny in papionin
primates. J. Hum. Evol. 45, 285e316.
Lieberman, D.E., 2008. Speculations about the selective basis for modern human
craniofacial form. Evol. Anthropol. 17, 55.
Lieberman, D.E., Carlo, J., Ponce de Leon, M., Zollikofer, C.P., 2007. A geometric
morphometric analysis of heterochrony in the cranium of chimpanzees and
bonobos. J. Hum. Evol. 52, 647e662.
Lieberman, D.E., McBratney, B.M., Krovitz, G., 2002. The evolution and development
of cranial form in Homo sapiens. Proc. Natl. Acad. Sci. 99, 1134e1139.
Lieberman, D.E., Wood, B.A., Pilbeam, D.R., 1996. Homoplasy and early Homo: an
analysis of the evolutionary relationships of H. habilis sensu stricto and
H. rudolfensis. J. Hum. Evol. 30, 97e120.
de Lumley, H., de Lumley, M.-A., 1973. Pre-Neanderthal remains from Arago cave in
southwestern France. Yearb. Phys. Anthropol. 17, 162e168.
Maddux, S.D., Franciscus, R.G., 2009. Allometric scaling of infraorbital surface
topography in Homo. J. Hum. Evol. 56, 161e174.
Manzi, G., 2004. Human evolution at the Matuyama-Brunhes boundary. Evol.
Anthropol. 13, 11e24.
Martinn-Torres, M., Bermudez de Castro, J.M., Gomez-Robles, A., Arsuaga, J.L.,
Carbonell, E., Lordkipanidze, D., Manzi, G., Margvelashvili, A., 2007. Dental
evidence on the hominin dispersals during the Pleistocene. Proc. Natl. Acad. Sci.
104, 13279e13282.
Martinn-Torres, M., Bermdez de Castro, J.M., Gmez-Robles, A., Prado- Simn, L.,
Arsuaga, J.L., 2012. Morphological description and comparison of the dental
remains from Atapuerca-Sima de los Huesos site (Spain). J. Hum. Evol. 62, 7e58.
McNulty, K.P., 2003. Geometric morphometric analyses of extant and fossil homi-
noid craniofacial morphology. Ph.D. Dissertation, City University of New York.
McNulty, K.P., Frost, S.R., Strait, D.S., 2006. Examining afnities of the Taung child by
developmental simulation. J. Hum. Evol. 51, 274e296.
Mitteroecker, P., Gunz, P., 2009. Advances in geometric morphometrics. Evol. Biol.
36, 235e247.
Mitteroecker, P., Gunz, P., Bernhard, M., Schaefer, K., Bookstein, F.L., 2004.
Comparison of cranial ontogenetic trajectories among great apes and humans.
J. Hum. Evol. 46, 679e697.
Mitteroecker, P., Gunz, P., Bookstein, F.L., 2005. Heterochrony and geometric
morphometrics: a comparison of cranial growth in Pan paniscus versus Pan
troglodytes. Evol. Dev. 7, 244e258.
Mounier, A., Marchal, F., Condemi, S., 2009. Is Homo heidelbergensis a distinct
species? New insight on the Mauer mandible. J. Hum. Evol. 56, 219e246.
Movius, H.L., 1953. The Mousterian cave of Teshik-Tash, southeastern Uzbekistan,
Central Asia. Bull. Am. Sch. Prehist. Res. 17, 11e71.
Murrill, R.I., 1981. Petralona Man: a New Description and Comparative Study, with
New Information on Rhodesian Man. Charles Thomas, Springeld.
Neubauer, S., Gunz, P., Hublin, J.J., 2009. The pattern of endocranial ontogenetic
shape changes in humans. J. Anat. 215, 240e255.
Neubauer, S., Gunz, P., Hublin, J.J., 2010. Endocranial shape changes during growth
in chimpanzees and humans: a morphometric analysis of unique and shared
aspects. J. Hum. Evol. 59, 555e566.
Neubauer, S., Gunz, P., Weber, G.W., Hublin, J.J., 2012. Endocranial volume of Aus-
tralopithecus africanus: new CT-based estimates and the effects of missing data
and small sample size. J. Hum. Evol. 62, 498e510.
Nicholson, E., Harvati, K., 2006. Quantitative analysis of human mandibular shape using
three-dimensional geometric morphometrics. Am. J. Phys. Anthropol. 131, 368e383.
OHiggins, P., Chadeld, P., Jones, N., 2001. Facial growth and the ontogeny of
morphological variation within and between the primates Cebus apella and
Cercocebus torquatus. J. Zool. 254, 337e357.
OHiggins, P., Jones, N., 1998. Facial growth in Cercocebus torquatus: an application of
three-dimensional geometric morphometric techniques to the study of
morphological variation. J. Anat. 193, 251e272.
Penin, X., Berge, C., Baylac, M., 2002. Ontogenetic study of the skull in modern
humans and the common chimpanzees: neotenic hypothesis reconsidered with
a tridimensional Procrustes analysis. Am. J. Phys. Anthropol. 118, 50e62.
Ponce de Len, M.S., Golovanova, L., Doronichev, V., Romanova, G., Akazawa, T.,
Kondo, O., Ishida, H., Zollikofer, C.P., 2008. Neanderthal brain size at birth
provides insights into the evolution of human life history. Proc. Natl. Acad. Sci.
105, 13764e13768.
Ponce de Len, M.S., Zollikofer, C.P., 1999. New evidence from Le Moustier 1:
computer-assisted reconstruction and morphometry of the skull. Anat. Rec.
254, 474e489.
Ponce de Len, M.S., Zollikofer, C.P., 2001. Neanderthal cranial ontogeny and its
implications for late hominid diversity. Nature 412, 534e538.
R Development Core Team, 2010. R: a Language and Environment for Statistical
Computing. R Foundation for Statistical Computing, Vienna.
Rak, Y., 1986. The Neanderthal: a new look at an old face. J. Hum. Evol. 15, 151e164.
Rak, Y., Kimbel, W.H., Hovers, E., 1994. A Neandertal infant from Amud cave, Israel.
J. Hum. Evol. 26, 313e324.
Rampont, M., 1994. Les squelettes, os et dents de foetus, nouveaux-n et enfants du
muse anatomique de Strasbourg. Aspect historiques et catalogues. Ph.D.
Dissertation, Universit Louis Pasteur.
Ravosa, M.J., 1991. Ontogenetic perspective on mechanical and nonmechanical
models of primate circumorbital morphology. Am. J. Phys. Anthropol. 85, 95e112.
Ravosa, M.J., 1998. Cranial allometry and geographic variation in slow lorises
(Nycticebus). Am. J. Primatol. 45, 225e243.
Richtsmeier, J.T., Corner, B.D., Grausz, H.M., Cheverud, J.M., Danahey, S.E., 1993. The
role of postnatal growth pattern in the production of facial morphology. Syst.
Biol. 42, 307.
Rightmire, G.P., 1998. Human evolution in the Middle Pleistocene: the role of Homo
heidelbergensis. Evol. Anthropol. 6, 218e227.
Rightmire, G.P., 2001. Patterns of hominid evolution and dispersal in the Middle
Pleistocene. Quatern. Int. 75, 77e84.
Rightmire, G.P., 2007. Later Middle Pleistocene Homo. In: Henke, W., Tattersall, I.
(Eds.), Handbook of Paleoanthropology. Springer, Heidelberg, pp. 1695e1715.
Rightmire, G.P., 2008. Homo in the middle Pleistocene: Hypodigms, variation, and
species recognition. Evol. Anthropol. 17, 8e21.
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 739
Rightmire, G.P., 2009. Out of Africa: modern human origins special feature: middle
and later Pleistocene hominins in Africa and southwest Asia. Proc. Natl. Acad.
Sci., 16046e16050.
Rohlf, F.J., 1993. Relative warp analysis and an example of its application to
mosquito wings. In: Marcus, L.F., Bello, E., Garca-Valdecasas, A. (Eds.), Contri-
butions to Morphometrics. Museo Nacional de Ciencies Naturales, Madrid,
pp. 131e159.
Rosas, A., 1997. A gradient of size and shape for the Atapuerca sample and Middle
Pleistocene hominid variability. J. Hum. Evol. 33, 319e331.
Rosas, A., 2000. Ontogenetic approach to variation in middle Pleistocene hominids.
Evidence from the Atapuerca-SH mandibles. Hum. Evol. 15, 83e98.
Rosas, A., 2001. Occurrence of Neanderthal features in mandibles from the
Atapuerca-SH site. Am. J. Phys. Anthropol. 114, 74e91.
Rosas, A., Bastir, M., 2002. Thin-plate spline analysis of allometry and sexual dimor-
phism in the human craniofacial complex. Am. J. Phys. Anthropol. 117, 236e245.
Rosas, A., Bastir, M., 2004. Geometric morphometric analysis of allometric variation
in the mandibular morphology of the hominids of Atapuerca, Sima de los
Huesos site. Anat. Rec. A 278, 551e560.
Rosas, A., Martnez-Maza, C., Bastir, M., Garca-Tabernero, A., Lalueza-Fox, C.,
Huguet, R., Ortiz, J.E., Juli, R., Soler, V., de Torres, T., Martnez, E., Caaveras, J.C.,
Snchez-Moral, S., Cuezva, S., Lario, J., Santamara, D., de la Rasilla, M., Fortea, J.,
2006. Paleobiology and comparative morphology of a late Neandertal sample
from El Sidron, Asturias, Spain. Proc. Natl. Acad. Sci. U S A 103, 19266e19271.
Roseman, C.C., Willmore, K.E., Rogers, J., Hildebolt, C., Sadler, B.E., Richtsmeier, J.T.,
Cheverud, J.M., 2010. Genetic and environmental contributions to variation in
baboon cranial morphology. Am. J. Phys. Anthropol. 143, 1e12.
Scheuer, L., Black, S.M., 2000. Developmental Juvenile Osteology. Elsevier Academic
Press, San Diego.
Schwarcz, H.P., Bietti, A., Buhay, W.M., Stiner, M.C., Grn, R., Segre, A., 1991. On the
reexamination of Grotta Guattari: uranium-series and electron-spin-resonance
dates. Curr. Anthropol. 32, 313e316.
Shea, B.T., 1981. Relative growth of the limbs and trunk in the African apes. Am. J.
Phys. Anthropol. 56, 179e201.
Shea, B.T., 1983. Allometry and heterochrony in the African apes. Am. J. Phys.
Anthropol. 62, 275e289.
Shen, G., Wang, W., Wang, Q., Zhao, J., Collerson, K., Zhou, C., Tobias, P.V., 2002. U-
Series dating of Liujiang hominid site in Guangxi, southern China. J. Hum. Evol.
43, 817e829.
Smith, F.H., 1982. Upper Pleistocene hominid evolution in south-central Europe:
a review of the evidence and analysis of trends. Curr. Anthropol. 23, 667e703.
Smith, F.H., 2002. Migrations, radiations and continuity: patterns of middle and late
Pleistocene humans. In: Hartwig, W.C. (Ed.), The Primate Fossil Record. Cam-
bridge University Press, Cambridge, pp. 437e456.
Smith, T.M., Tafforeau, P., Reid, D.J., Grn, R., Eggins, S., Boutakiout, M., Hublin, J.-J.,
2007. Earliest evidence of modern human life history in North African early
Homo sapiens. Proc. Natl. Acad. Sci. 104, 6128e6133.
Soressi, M., Jones, H.L., Rink, W.J., Maureille, B., Tillier, A.M., 2007. The Pech-de-lAz
I Neandertal child: ESR, uranium-series, and AMS 14C dating of its MTA type B
context. J. Hum. Evol. 52, 455e466.
Spoor, F., Leakey, M.G., Gathogo, P.N., Brown, F.H., Anton, S.C., McDougall, I.,
Kiarie, C., Manthi, F.K., Leakey, L.N., 2007. Implications of new early Homo fossils
from Ileret, east of Lake Turkana, Kenya. Nature 448, 688e691.
Stanseld, E., Gunz, P., 2011. Skhodnya, Khvalynsk, Satanay, and Podkumok calvar-
iae: possible Upper Paleolithic hominins from European Russia. J. Hum. Evol. 60,
129e144.
Strand Viarsdttir, U., Cobb, S., 2004. Inter- and intra-specic variation in the
ontogeny of the hominoid facial skeleton: testing assumptions of ontogenetic
variability. Ann. Anat. 186, 423e428.
Strand Viarsdttir, U., OHiggins, P., Stringer, C., 2002. A geometric morphometric
study of regional differences in the ontogeny of the modern human facial
skeleton. J. Anat. 201, 211e229.
Street, M., 2002. Ein Wiedersehen mit dem Hund von Bonn-Oberkassel. Bonn. Zool.
Beitr. 50, 269e290.
Stringer, C.B., 1974. Population relationships of later Pleistocene hominids:
a multivariate study of available crania. J. Archaeol. Sci. 1, 317e342.
Stringer, C.B., 1983. Some further notes on the morphology and dating of the Pet-
ralona hominid. J. Hum. Evol. 12, 731e742.
Stringer, C.B., 2011. The chronological and evolutionary position of the Broken Hill
cranium. Am. J. Phys. Anthropol. (S52), 287.
Stringer, C.B., 2012. The status of Homo heidelbergensis (Schoetensack 1908). Evol.
Anthropol. 21, 101e107.
Stringer, C.B., Howell, F.C., Melentis, J.K., 1979. The signicance of the fossil hominid
skull from Petralona, Greece. J. Archaeol. Sci. 6, 235e253.
Tattersall, I., Schwartz, J.H., 2006. The distinctiveness and systematic context of
Homo neanderthalensis. In: Harvati, K., Harrison, T. (Eds.), Neanderthals Revis-
ited: New Approaches and Perspectives. Springer, Dordrecht, pp. 9e22.
Tillier, A.-M., 1989. The evolution of modern humans: evidence from young
Mousterian individuals. In: Mellars, P., Stringer, C. (Eds.), The Human Revolu-
tion: Behavioural and Biological Perspectives on the Origins of Modern
Humans. Princeton University Press, Princeton, pp. 286e297.
Tillier, A.M., 1996. The Pech de lAz and Roc de Marsal children (Middle Paleolithic,
France): skeletal evidence for variation in Neanderthal ontogeny. Hum. Evol. 11,
113e119.
Tillier, A.M., 2011. Facts and ideas in Paleolithic growth studies (paleoauxology). In:
Condemi, S., Weniger, G.C. (Eds.), Continuity and Discontinuity in the Peopling
of Europe. Springer, New York, pp. 139e153.
Trinkaus, E., 1983. The Shanidar Neandertals. Academic Press, New York.
Trinkaus, E., 1987. The Neandertal face: evolutionary and functional perspectives on
a recent hominid face. J. Hum. Evol. 16, 429e443.
Trinkaus, E., 2003. Neandertal faces were not long; modern human faces are short.
Proc. Natl. Acad. Sci. 100, 8142e8145.
Trinkaus, E., 2006. Modern human versus Neandertal evolutionary distinctiveness.
Curr. Anthropol. 47, 597e620.
Ubelaker, D., 1989. Human Skeletal Remains. Taraxacum Press, Washington.
Vishnyatsky, L., 1999. The Paleolithic of Central Asia. J. World Prehist. 13, 69e122.
Weber, G.W., Gunz, P., Mitteroecker, P., Stadlmayr, A., Bookstein, F.L., Seidler, H.,
2006. External geometry of Mladec neurocrania compared with anatomically
modern humans and Neandertals. In: Teschler-Nicola, M. (Ed.), Early Modern
Humans at the Moravian Gate: Mladec Caves and Their Remains. Springer,
Wien, pp. 453e471.
White, T.D., Black, M.T., Folkens, P.A., 2012. Human Osteology. Academic Press,
Burlington.
Wiley, D.F., Amenta, N., Alcantara, D.A., Deboshmita, G., Kil, Y.J., Delson, E., Harcourt-
Smith, W., Rohlf, F.J., St John, K., Hamann, B., 2005. Evolutionary morphing. In:
Proceedings of IEEE Visualization 2005.
Williams, F.L., Godfrey, L.R., Sutherland, M.R., 2002. Heterochrony and the evolution
of Neandertal and modern human craniofacial form. In: Minugh-Purvis, N.,
McNamara, K.J. (Eds.), Human Evolution through Developmental Change. The
John Hopkins University Press, Baltimore, pp. 405e441.
Wolpoff, M.H., Thorne, A.G., Jelinek, J., Yinyun, Z., 1994. The case for sinking Homo
erectus. 100 years of Pithecanthropus is enough! Cour. Forsch.-Inst. Senckenberg
171, 341e361.
Wood, B., Lieberman, D.E., 2001. Craniodental variation in Paranthropus boisei:
a developmental and functional perspective. Am. J. Phys. Anthropol. 116, 13e25.
Zollikofer, C.P., Ponce de Len, M.S., 2010. The evolution of hominin ontogenies.
Semin. Cell. Dev. Biol. 21, 441e452.
S.E. Freidline et al. / Journal of Human Evolution 63 (2012) 723e740 740