The Ancient Inhabitants of Jebel Moya Redux: Measures of Population Affinity Based on Dental Morphology

Abstract

This paper reexamines some of the methods and craniometric findings in the classic volume The Ancient Inhabitants of Jebel Moya (Sudan) (1955) by Mukherjee, Rao & Trevor, in light of recent archaeological data and relative to a new dental morphological study. Archaeological evidence characterises these inhabitants as having been heavily influenced by outside sources; yet they managed to establish and maintain their own distinctive culture as seen in the site features and surviving artefact collections. The dental study, modelled after the original craniometric-based investigation and using the same or similar comparative samples, detected complementary indications of outside biological influence. In the study, up to 36 dental traits were recorded in a total of 19 African samples. The most influential traits in driving inter-sample variation were then identified, and phenetic affinities were calculated using the Mahalanobis D2 statistic for non-metric traits. If phenetic similarity provides an estimate of genetic relatedness, these affinities, like the original craniometric findings, suggest that the Jebel Moyans exhibited a mosaic of features that are reminiscent of, yet distinct from, both sub-Saharan and North African peoples. Together, these different lines of evidence correspond to portray the Jebel Moya populace as a uniform, although distinct, biocultural amalgam. Copyright © 2006 John Wiley & Sons, Ltd.

Full text

The Ancient Inhabitants of Jebel
Moya Redux: Measures of
Population Affinity Based on
Dental Morphology
J. D. IRISH
a
*AND L. KONIGSBERG
b
a
Department of Anthropology, University of Alaska Fairbanks, Fairbanks,
AK 99775-7720, USA
b
Department of Anthropology, University of Tennessee, Knoxville, TN 37996-0230, USA
ABSTRACT This paper reexamines some of the methods and craniometric findings in the classic volume
The Ancient Inhabitants of Jebel Moya (Sudan) (1955) by Mukherjee, Rao & Trevor, in light of
recent archaeological data and relative to a new dental morphological study. Archaeological
evidence characterises these inhabitants as having been heavily influenced by outside
sources; yet they managed to establish and maintain their own distinctive culture as seen
in the site features and surviving artefact collections. The dental study, modelled after the
original craniometric-based investigation and using the same or similar comparative samples,
detected complementary indications of outside biological influence. In the study, up to 36
dental traits were recorded in a total of 19 African samples. The most influential traits in driving
inter-sample variation were then identified, and phenetic affinities were calculated using the
Mahalanobis D
2
statistic for non-metric traits. If phenetic similarity provides an estimate of
genetic relatedness, these affinities, like the original craniometric findings, suggest that the
Jebel Moyans exhibited a mosaic of features that are reminiscent of, yet distinct from,
both sub-Saharan and North African peoples. Together, these different lines of evidence
correspond to portray the Jebel Moya populace as a uniform, although distinct, biocultural
amalgam. Copyright ß2006 John Wiley & Sons, Ltd.
Key words: dental anthropology; trait variation; phenetic affinity; Sudan; Africa
Introduction
Fifty years ago R. Mukherjee, C.R. Rao and J.C.
Trevor (1955) wrote The Ancient Inhabitants of Jebel
Moya (Sudan). The volume served as the official
report on skeletal remains recovered by the
Wellcome Expedition between 1911–1914 at
the site—which was initially thought to date
to 1000–400 BC (Addison, 1949). Originally
entrusted to such luminaries as Sir Arthur Keith
and G.M. Morant by the Trustees of the Estate of
Sir Henry Wellcome, the long overdue report was
finally delegated to the aforementioned authors
after a series of setbacks and delays, including two
world wars. Unfortunately, the 40-year hiatus
between excavation and analysis proved to be
catastrophic for the remains, which were poorly
housed and moved repeatedly after shipment to
England (Mukherjee et al., 1955; Addison, 1956).
As a result, of more than 3000 skeletons originally
excavated, only 98 crania, 139 mandibles and
a handful of post-cranial elements survived to
allow study. Nevertheless, the three authors were
initially optimistic that a serviceable report could
still be produced, as field cards for 2903 skeletons
International Journal of Osteoarchaeology
Int. J. Osteoarchaeol. (in press)
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/oa.868
* Correspondence to: Department of Anthropology, PO Box
757720, University of Alaska Fairbanks, Fairbanks, Alaska, AK
99775-7720, USA.
e-mail: ffjdi@uaf.edu
Copyright #2006 John Wiley & Sons, Ltd. Received 11 October 2005
Revised 2 February 2006
Accepted 20 February 2006

had been saved; these cards contained osteolo-
gical observations and measurements made by
Expedition physical anthropologists in the field.
The outlook turned negative, however, when
over three-quarters of the cards were found to
contain erroneous or useless data (Mukherjee
et al., 1955; Gerharz, 1994); thus, the mood of the
authors at the time of writing can be summarised
by the statement ‘ ...what at present survives of it
[ Jebel Moya skeletal sample] represents hardly a
tithe of the original, and anthropological science
has been denied a unique opportunity to deal
with a substantial body of data, for which the
field measurements are, alas, only too imperfect’
(Mukherjee et al., 1955: 31).
Despite these and other deficiencies, the report
was notable for its groundbreaking use of advanced
statistical analyses to understand the Jebel Moya
population’s make-up and affiliations. Although
retaining terms and some methodology equated
with racial typology of early physical anthropol-
ogy (see Keita, 1990, 1992, 1996), Mukherjee and
associates applied the now-common Mahalanobis
D
2
distance to craniometric data for the first
time; the result was a measure of group divergence
between Jebel Moya and 19 other African samples
(Mukherjee et al., 1955). This fresh approach,
directed away from typology and towards the
concept of population affinity, would not other-
wise become a focus of physical anthropologists
until the 1970s and beyond (Berry & Berry, 1972,
1973; Greene, 1972; Howells, 1989; Konigsberg,
1990; Keita, 1990, 1992; Irish, 1993, 1998a,b,c,d,
2005; Brace et al., 1993; Johnson & Lovell, 1994;
Prowse & Lovell, 1996; Hemphill, 1998; Roseman
& Weaver, 2003; Pietrusewsky, 2004; among many
others).
Recent reappraisals of the long-neglected
archaeological collections (e.g. Caneva, 1991;
Gerharz, 1994) revealed that, although Jebel Moya
‘has been a centre of controversy’ (Adams, 1977:
718) and its ‘reputation ...suffered’ (Caneva, 1991:
263), important information can still be gleaned
for both descriptive and comparative purposes.
This positive outlook, together with new dating
(Gerharz, 1994) prompted the current first author
to find and study the site’s skeletal remains
while conducting other research; the intent was
to see whether the old sample might yet be able
to provide some new insights about the site
inhabitants. Thus, using the now-classic Mukher-
jee et al. (1955) report as a model, the present study
will contrast the remains with the same or similar
African comparative samples originally used.
However, in place of D
2
distances based on
craniometrics, highly heritable dental variants are
compared among samples using an analogous
distance statistic for non-metric traits. The
objectives are to: (1) compare results between
studies; and (2) further explore the biological
‘place’ of Jebel Moyans in African prehistory— a
subject that has received scant attention over the
last half century. In the process, several peopling
scenarios based on inter-region cultural parallels
will be addressed.
The site
Jebel Moya is located 250 km south-southeast
of Khartoum, Sudan, in the southern part of
the Gezira plain between the White and Blue
Niles (Figure 1) (Martin, 1921; Addison, 1949;
Mukherjee et al., 1955). According to Addison
(1949), author of the official archaeological site
report, ‘Jebel Moya’ refers to a massif that actually
Figure 1. The location of Jebel Moya relative to some
other landmarks mentioned in the text.
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
J. D. Irish and L. Konigsberg

comprises a compact group of granite hills
interconnected by ridges and valleys; however,
the name later came to be associated with the
archaeological site situated within the north-
easternmost of these valleys.
The site is roughly 104,000 m
2
in area, of
which about a fifth was excavated during the
Wellcome Expedition’s four field seasons (Addi-
son, 1949; Mukherjee et al., 1955). Because a
permanent water source was present (Williams
et al., 1982), numerous habitation and grave sites
were established to accommodate the nomadic
pastoralists who occupied the valley (Addison,
1949; Gerharz, 1994). An enormous quantity of
cultural material, much of which evidenced extra-
regional influence in manufacture and design, was
recovered from both types of sites; artefacts
included thousands of lipstuds, beads and other
ornaments, hundreds of stone tools, some
imported objects (see below), and ‘several tons
of potsherds’ (Addison, 1956: 4; see also Martin,
1921). Like the skeletal remains, only a fraction of
these items survive today to allow study (e.g.
Caneva, 1991). Regarding graves, nearly 2800
were cleared by excavators; some contained
multiple inhumations, whereas others were empty
or consisted of animal burials. The number of
human skeletons recovered, many of which were
fragmentary, was 3137. Both sexes and all age
groups were represented (Addison, 1949;
Mukherjee et al., 1955). Addison (1949: 40)
noted that the dead appeared to ‘have been
disposed of with scant ceremony’. Almost half of
the individuals were buried without grave
offerings. Most others received only a few
offerings—often ornaments that may have been
worn in life (Addison, 1949). Moreover, there was
little evidence of a standard mortuary practice;
tomb types differed in appearance, body position
varied widely, and graves ‘were oriented to every
point of the compass’ (Addison, 1949: 40; 1956:
4; also Gerharz, 1994).
As mentioned, the site was originally thought
to date to 1000–400 BC, corresponding to much
of the Napatan period of Upper Nubia, a region
located along the Nile north of Khartoum.
Addison (1949) based this dating on the presence
of Napatan objects within some burials and,
primarily, on four distinct strata observed during
the course of excavation. He later modified his
interpretations (Addison, 1956), placing the site
occupation between 500 BC–AD 400. The new
timing corresponds to the Meroitic period;
objects from this later Nubian culture were also
found in a number of burials (Hofmann, 1967;
Gerharz, 1994). More recently, Gerharz (1994)
found that both sets of dates are incorrect.
Problems with stratigraphic information (Caneva,
1991; Gerharz, 1994) led to misinterpretations by
Addison. The new dating is thought to comprise
three main temporal phases between ca. 5000–
100 BC. Phase I (ca. 5000–3000 BC) was
identified by the presence of diagnostic dotted
wavy line pottery; however, the original settle-
ment horizon was said to have been destroyed by
the later inhabitants (Gerharz, 1994). Surviving
site features, including all graves, date to Phase II
(ca. 3000–800 BC) and Phase III (800–100 BC).
By comparing their horizontal distribution,
Gerharz (1994) assigned individual graves to
phases. When uncertain, the date could still be
determined if Nubian and certain other grave
goods that originated in Phase III were present.
These determinations are corroborated by 3rd
millennium BC radiocarbon dates from basal
layers of the site (Clark & Stemler, 1975; Clark,
1984; Haaland, 1987).
Materials and methods
The report by Mukherjee et al. (1955) provided a
relatively complete study of the remnant Jebel
Moya sample, including inventory, ageing and
sexing of the remains. They also attempted a
‘racial’ description and comparison of the crania
(i.e. Negroid, non-Negroid) by stratum, using a
number of obscure and more common non-metric
characteristics. However, the primary analysis of
interest involved the affinity study between Jebel
Moya and 19 comparative samples using seven
cranial measurements (e.g. maximum length,
breadth, height, etc.) in the males only; refer
to the report for a full list and descriptions. Use of
additional measures was deemed impossible due
to the large number of calculations required for
the Mahalanobis statistic in pre-computer 1955.
Comparative samples came from Egypt, Nubia,
Ethiopia, Kenya, and several west African locales;
they are listed on the left side of Table 1 in the
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
Ancient Inhabitants of Jebel Moya

Table 1. Jebel Moya and comparative samples used in the craniometric and dental studies
Craniometric samples from Mukherjee et al. (1955) Same, equivalent, or similar samples used in present dental study
a
Sample Origin Period Sample Origin Cultural affiliation Date No.
Jebel Moya
(JEM)
b
Sudan Prehistoric Jebel Moya (JEM)
c
Jebel Moya Jebel Moya Complex c. 3000–100 BC 58
Egyptian ‘E’
(EGE)
Egypt 26–30th Dynasty Egyptian ‘E’ (EGE) Giza 26–30th Dynasty 664–332 BC 62
Naqada (NAQ) Egypt Predynastic Naqada (NAQ) Naqada Predynastic Egyptian c. 4000–3200 BC 65
Badarian (BAD) Egypt Predynastic Badarian (BAD) Badari Predynastic Egyptian c. 4400–4000 BC 40
Sedment (SED) Egypt 9th Dynasty Saqqara (SAQ) Saqqara 4th Dynasty Egyptian 2613–2494 BC 41
Egypt ‘Negroes’
(EGN)
Egypt Dynastic El Hesa (HES) El Hesa Roman Period Egyptian AD 200–400 72
Kerma (KER) Nubia 12–13th Egypt
Dynasty
Kerma (KER) Kerma Kerma Classique Nubian 1750–1500 BC 63
A-Group (AGR) Nubia <4th Egypt
Dynasty
A-Group (AGR) Faras to Gamai A-Group Nubian 3000 BC 52
B-Group (BGR) Nubia 4–6th Egypt
Dynasty
NA
d
C-Group (CGR) Nubia 7–16th Egypt
Dynasty
C-Group (CGR) Faras to Gamai C-Group Nubian 2000–1600 BC 62
D-Group (DGR) Nubia 18th Egypt
Dynasty
þ
D-Group (DGR)
e
Faras to Gamai Pharonic Nubian 1650–1350 BC 38
Meroitic (MER) Nubia 1st–2nd cents AD Meroitic (MER) Semna; Faras/Gamai Meroitic Nubian 100 BC–AD 350 94
X-Group (XGR) Nubia 3rd–6th cents AD X-Group (XGR) Semna; Faras/Gamai X-Group Nubian AD 350–550 63
Taita (TAI) Kenya 20th cent. AD Kenya (KEN) Kenya, Tanzania Kikuyu, Swahili,
Chaga, Pare
19–20th cent. AD 114
Cameroons
(CAM)
Cameroon 19th cent. AD Nigeria–Cameroon
(NIC)
Nigeria, Cameroon Efik, Ibibio, Boki, Anyang 19th cent. AD 57
Tigrean (TIG) Ethiopia 19th cent. AD Ethiopia (ETH) Ethiopia, Eritrea Tigre, Amhara,
Danakil, etc.
19–20th cent. AD 40
Ashanti (ASH) Ghana 19th cent. AD Ghana (GHA) Ghana Ashanti, Fanti 19th cent. AD 47
Tetela (TET) Congo,
Gabon
19th–20th
cents AD
Congo (CNG) Congo, Gabon Teke, Kongo, Binga, etc. 19–20th cent. AD 52
Fernand Vaz
(FRV)
Gabon 19th cent. AD Gabon (GAB) Gabon Fang, Nkomi, Lumbo, etc. 19–20th cent. AD 39
Ibo (IBO) Nigeria Recent Togo–Dahomey
(TOD)
Togo, Benin Ewe, Fon 19th cent. AD 25
a
See text for details of sample comparability between studies.
b
Three-letter abbreviations of craniometric study samples used in Figures 4 and 5.
c
Three-letter abbreviations of dental study samples used in Tables 2 and 4, and Figure 3.
d
The ‘B-Group’ is no longer recognised as a distinct Nubian cultural period, so there is no equivalent sample used in the dental study.
e
‘D-Group’ is referred to today as Pharonic Nubians.
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
J. D. Irish and L. Konigsberg

same order as described in Mukherjee et al.
(1955). Considering the limited availability of
such data at the time, these samples represent a
decent cross-section of North and northern sub-
Saharan African cultures.
The dental samples
The present investigation emulates the cranio-
metric study in several ways, most notably by
comparing the same or similar samples. Ideally,
dental data would have been collected in all of the
samples originally used by Mukherjee et al.
(1955). Unfortunately, many are now dispersed,
lost, possess an insufficient number of recordable
dentitions, or are otherwise unavailable. An
overview of the samples and their comparability
to those in the craniometric study is presented
below and in Table 1. In-depth information is
available in Irish (1993, 1997, 1998a,b,c,d, 2005).
The Jebel Moya dental sample (abbreviated
JEM in tables and subsequent figures) consists of
58 specimens with the most complete upper and
lower dentitions of the 98 crania. The compara-
tive Egyptian ‘E’ (EGE), Naqada (NAQ) and
Badarian (BAD) samples are exactly the same as
those used by Mukherjee and associates. Saqqara
(SAQ) replaces the original Sedment sample;
although the former is several hundred years
more recent, both samples are from the same area
in Lower Egypt. Likewise, El Hesa (HES) replaces
the Egyptian ‘Negroes’ sample; El Hesa is younger
but, again, both were recovered near Aswan in
extreme southern Upper Egypt. Kerma (KER) is
the same early Nubian sample used in the original
study. The A-Group (AGR), C-Group (CGR), D-
Group (DGR), Meroitic (MER) and X-Group
(XGR) samples are not those used in the first
study, but derive from the same Nubian
populations, so may be considered equivalent.
It should be noted that although the term ‘D-
Group’ is retained to facilitate comparability to
the original study, this culture is most commonly
called Pharonic Nubians today (Nielsen, 1970;
Adams, 1977; Calcagno, 1986). Moreover, the
‘B-Group’ (again see Table 1) is no longer
considered a distinct Nubian cultural period;
thus, there is no corresponding dental sample in
this one case. Kenya (KEN) consists of Bantu-
speakers from Kenya and Tanzania who are
closely related biologically and linguistically to
the original Taita (Kitson, 1931). Nigeria–
Cameroon (NIC), Ethiopia (ETH), Ghana
(GHA), Congo (CNG), and the Gabon (GAB)
samples are also closely related to, or come from,
the same population as Cameroons, Tigrean,
Ashanti, Tetela and Fernand Vaz, respectively.
Only Togo–Dahomey (TOD) is not directly
equivalent to the original Ibo sample from
Nigeria, although both derive from spatially
proximate west African locations. In total, 1084
dentitions in these 19 samples were analysed for
the present study.
Dental trait recording
The study is concerned with morphological
variation of the permanent dentition. Thirty-six
non-metric traits employed in previous African
affinity studies (Irish, 1993, 1994, 1997,
1998a,b,c,d, 2000, 2005, 2006; Irish & Guatelli-
Steinberg, 2003) were recorded in the 19 total
samples (see list in Table 2). The rationale for
selecting them has been previously detailed
(Irish, 1993, 1998d, 2005); of critical importance,
however, is the high genetic component reported
for many of these traits (Scott, 1973; Larsen,
1997; Scott & Turner, 1997), which makes them
ideal for biodistance analyses (Larsen, 1997).
Excluding midline diastema, each trait is part of
the Arizona State University Dental Anthropol-
ogy System (ASUDAS) (Turner et al., 1991). The
ASUDAS procedures have proven reliable in
many studies (e.g. Scott, 1973, 1980; Turner,
1985a, 1987, 1990, 1992; Sakuma & Ogata, 1987;
Haeussler et al., 1988; Turner & Markowitz, 1990;
Irish & Turner, 1990; Irish, 1993, 1994, 1997,
1998a,b,c,d, 2000, 2005; Jackes et al., 2001).
Using 23 rank-scale reference plaques to stan-
dardise scoring (Turner et al., 1991), bilateral
traits are recorded in both antimeres and the side
with highest expression counted (Turner & Scott,
1977). This approach assumes scoring for the
maximum genetic potential (Turner, 1985a). Due
to minimum trait sexual dimorphism (Scott, 1973,
1980; Smith & Shegev, 1988; Bermudez de
Castro, 1989; Turner et al., 1991; Hanihara,
1992; Irish, 1993), it is also standard procedure to
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
Ancient Inhabitants of Jebel Moya

pool the sexes (Irish, 1997). The ASUDAS is
described fully in Turner et al. (1991) and Scott &
Turner (1997).
Quantitative analyses
Once recorded, frequencies of occurrence for the
36 traits were calculated for each sample, after
having first been dichotomised, as necessary, into
categories of present or absent based on their
appraised morphological thresholds (Haeussler
et al., 1988). For the most part, dichotomisation
was done according to standard procedure
(Turner, 1985b, 1987; Irish, 1993). However, a
small subset of traits was considered present at
higher ranks (e.g. LM1 cusp 7 is considered
present at ASU grades 2–4 compared to standard
1–4) to differentiate better between North
African and sub-Saharan samples (for additional
discussion see Irish, 1993). Trait dichotomisation
is required (Sjøvold, 1977) before submitting the
rank-scale data to most multivariate statistics (see
below). The resulting frequency table contains
information that is useful for identifying and
qualitatively comparing the sample trait variation.
A more definitive way to assess this variation is
to use a distance statistic, which provides a
quantitative estimate of inter-sample biological
divergence based on similarities among traits.
Rightmire (1999: 2) related that ‘ ...it is all but
certain that these phenotypic [dental] patterns
reflect underlying genetic variation’; thus, it is
assumed that phenetic similarity approximates or
is an estimate of genetic relatedness (Scott et al.,
1983). Of many distance statistics previously
used, the authors and others (e.g. Berry & Berry,
1972; Sjøvold, 1973, 1977; Greene, 1982; Scott
et al., 1983; Turner, 1984, 1985a; Konigsberg,
1990; Turner & Markowitz, 1990; Lukacs &
Hemphill, 1991; Ishida & Dodo, 1997; Irish,
1997, 1998a,b,c,d, 2000, 2005; Donlon, 2000;
Jackes et al., 2001) have often employed two: the
modified Mahalanobis D
2
statistic for non-metric
traits derived by Konigsberg (1990), and the
mean measure of divergence (MMD) which
incorporates the Freeman and Tukey angular
transformation to correct for low (0.05) or high
(0.95) frequencies and small sample sizes
(n10) (Berry & Berry, 1967; Sjøvold, 1973,
1977; Green & Suchey, 1976).
To emulate Mukherjee et al. (1955) most
closely, the modified Mahalanobis statistic was
used to compare Jebel Moya with the compara-
tive samples. It extends the original Mahalanobis
generalised distance by utilising a tetrachoric
correlation matrix with the dichotomised dental
data. Correlations are calculated within each
sample and pooled using sample size for each trait
pair to find the weighted average correlation. As
such, this statistic is effective in correcting for the
small sample sizes that characterise many
archaeological skeletal collections. Furthermore,
it provides an advantage over other methods
(incl. MMD) by adjusting for phenotypic
correlations between traits; this adjustment
avoids any undue weight on groups of charac-
teristics that co-occur. Additional methodologi-
cal details and the formulae are listed in
Konigsberg (1990), Konigsberg et al. (1993),
Ishida & Dodo (1997) and Bedrick et al. (2000).
Prior to applying the D
2
statistic, several
problematic dental traits were deleted, including
those with many small sample sizes (e.g. <10
cases) and shared high (i.e. fixed at or close to
100%) or low frequencies (0%). Correspondence
analysis (CA), using the SPSS Procedure Corre-
spondence, was then used to quantify which
remaining traits vary most among the samples.
This technique has been used in many prior
anthropological studies (Greenacre & Degos,
1977; Schneider, 1986; Sciulli, 1990; Gerharz,
1994; Kitagawa et al., 1995; Coppa et al., 1998;
Irish, 2005, 2006). A variant of principal
components analysis (PCA), CA factors non-
metric data comprising columns and rows of a
contingency table and displays them in reduced
space to illustrate association. Among other
output, a biplot is produced that combines
column and row points to identify which traits
are most influential relative to the samples: see
Irish (2005) for a complete description of using
CA with ASU dental data. Publications by
Greenacre & Degos (1977), Clausen (1988),
Benze
´cri (1992) and Phillips (1995) provide
methodological details.
An additional step, employed here for the first
time, was to quantify the amount of intra-sample
trait variability affecting the Jebel Moya dentitions.
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
J. D. Irish and L. Konigsberg

Evidence of cultural diversity in site features and
artefacts prompted Addison (1949) and other
researchers to suggest that the Jebel Moyans may
have been biologically diverse as well. Among
other approaches, Mukherjee et al. (1955) tested
for intra-sample variability by calculating cranio-
metric standard deviations and coefficients of
variation to contrast with those in a ‘racially
homogeneous’ comparative sample. Until now, no
statistics were available to allow a corresponding
comparison for ASUDAS rank-scale/ordinal data.
However, in the following it is shown how the
multiple threshold model that characterises ASU-
DAS trait variation (e.g. Scott, 1973) can be used
to find relative variability in such data.
To provide some context for this method, a
normal distribution is assumed in the multiple
threshold model. As such, when examining one
sample it is possible to estimate the threshold values
by reading the standard normal cumulative back-
ward from probabilities to quantiles. For example,
with a trait that comprises five ordered states (i.e.
grades of 0, 1, 2, 3, 4) of expression, the four
thresholds can be located on this distribution for a
largepooledsamplewheresomegrandmeanand
standard deviation are assumed. To find the relative
mean and relative standard deviation for any
subsample (i.e. relative to grand mean and standard
deviation), the method of maximum likelihood
(Eliason, 1993) can then be used to estimate these
two parameters while keeping the threshold values
fixed for the total sample. This approach is only
relevant with traits scored on a rank scale, not those
few initially recorded at a binomial/dichotomous
level (see below). With binary traits there are two
parameters to be estimated—the mean and
standard deviation — but only one ‘piece’ of
information: the trait frequency.
Calculations of rank-scale sample means and
standard deviations were undertaken for each
selected trait using a program written by the
second author in R (http://cran.r-project.org/).
This program (available upon request from the
first author) uses areas under a normal curve
between thresholds based on cumulative fre-
quencies to fit such distributions using maximum
likelihood. To illustrate, when LM1 cusp 7, a trait
considered here to comprise five states, was
compared among samples (refer to Table 2), the
numbers of specimens exhibiting each grade were
initially listed by sample in 19 rows. These
frequencies were summed down each column to
yield five grand sample totals which were then
cumulatively summed across this row. Next, the
first four of these cumulative sums were divided
by the final sum to yield cumulative relative
frequencies. A normal distribution with a mean of
5.0 and a standard deviation of 1.0 was then used
to determine quantiles. These two values are
arbitrary; alternative selections work just as well
and would not impact on the relative variability
comparisons (Thomas, pers. comm., 2005). In the
ensuing step, the four quantiles are assumed to
come from a normal distribution with mean and
standard deviation values unknown; maximum
likelihood was used to estimate these values.
Finally, the 19 estimated standard deviations were
compared to describe relative variability, and
identify the sample(s) exhibiting most variability
for LM1 cusp 7 expression. The process was
repeated with additional traits to help quantify
whether Jebel Moya is dentally diverse relative to
the remaining samples.
Lastly, once the edited list of dental traits has
been submitted to the Mahalanobis statistic, a
matrix of D
2
values among samples is produced.
However, a more intuitive manner in which to
interpret the results is to present the patterning of
inter-sample affinities visually using multi-dimen-
sional scaling (MDS). Procedure Alscal in SPSS
12.0 was used here. MDS provides a spatial
representation of 1 to n dimensions consisting of
a geometric configuration of points (the dental
samples) (Kruskal & Wish, 1978). Therefore,
plotting of samples into groups indicates degrees
of relationship. In the present case interval-level
MDS was used, as the large number of traits
causes the matrix of distance values to approxi-
mate continuous data.
Results
Dental trait frequencies
Table 2 lists the 36 dental traits for the 19
samples. Samples are listed in the same order as
that on the right side of Table 1. The percentage
of individuals exhibiting each trait and the total
number of individuals scored are presented.
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
Ancient Inhabitants of Jebel Moya

Table 2. Dental trait percentages (%) and number of individuals scored (n) for Jebel Moya and the 18 comparative samples
a
Trait
b
JEM EGE NAQ BAD SAQ HES KER AGR CGR DGR MER XGR KEN NIC ETH GHA CNG GAB TOD
Winging UI1 % 9.1 4.3 6.0 5.6 2.8 6.4 5.4 2.6 16.3 5.6 12.8 5.7 4.0 3.5 0.0 0.0 9.1 3.5 10.0
(þ¼ASU 1) n 33 47 50 36 36 63 56 39 49 54 39 35 99 29 36 23 22 29 20
Labial Curvature
UI1
% 39.3 17.7 12.5 50.0 9.1 65.0 38.5 43.5 57.9 46.2 24.4 38.9 37.5 60.0 36.4 53.9 40.0 75.0 64.3
(þ¼ASU 2–4) n 28 17 8 20 11 20 13 23 19 13 41 18 16 5 11 13 5 8 14
Palatine Torus % 0.0 0.0 0.0 5.6 0.0 0.0 1.8 0.0 0.0 5.2 10.7 10.2 0.9 0.0 10.8 2.2 3.9 8.1 0.0
(þ¼ASU 2–3) n 31 47 50 36 39 68 55 24 42 58 84 49 108 55 37 46 52 37 25
Shoveling UI1 % 0.0 15.4 14.3 25.0 0.0 29.4 22.2 15.8 10.5 15.4 38.9 25.0 7.1 42.9 28.6 41.7 0.0 12.5 54.6
(þ¼ASU 2–6) n 25 13 7 16 7 17 9 19 19 13 36 12 14 7 7 12 5 8 11
Double Shoveling
UI1
% 0.0 0.0 0.0 0.0 0.0 15.8 0.0 0.0 0.0 0.0 4.6 6.7 6.7 0.0 0.0 0.0 0.0 0.0 0.0
(þ¼ASU 2–6) n 27 16 7 16 8 19 7 21 17 11 44 15 15 7 7 13 5 8 11
Interruption
Groove UI2
% 0.0 4.2 9.1 10.0 33.3 50.0 9.1 4.4 45.0 27.8 36.2 40.0 11.5 25.0 23.1 7.1 12.5 0.0 6.3
(þ¼ASU þ) n 35 24 11 20 9 22 11 23 20 18 47 15 26 8 13 14 8 6 16
Tuberculum
Dentale UI2
% 5.9 25.0 27.3 36.4 66.7 61.9 8.3 13.0 35.0 23.5 40.5 42.9 36.0 25.0 41.7 20.0 42.9 66.7 37.5
(þ¼ASU 2–6) n 34 24 11 22 6 21 12 23 20 17 42 14 25 8 12 15 7 6 16
Bushman
Canine UC
% 10.0 6.3 0.0 0.0 0.0 6.5 16.7 11.5 0.0 0.0 19.6 10.0 13.0 22.2 5.6 7.7 8.7 6.3 35.3
(þ¼ASU 1–3) n 40 32 22 22 10 31 18 26 26 19 51 20 46 27 18 26 23 16 17
Distal Acc.
Ridge UC
% 11.5 7.1 15.0 12.5 0.0 13.6 18.2 33.3 12.5 18.2 31.0 21.4 36.6 57.9 27.3 62.5 64.7 58.3 61.5
(þ¼ASU 2–5) n 26 28 20 16 6 22 11 18 16 11 42 14 41 19 11 16 17 12 13
Hypocone UM2 % 73.3 84.2 90.9 86.7 95.7 75.4 91.7 73.7 76.1 81.4 78.5 85.7 78.0 90.9 66.7 83.8 94.6 86.2 100
(þ¼ASU 3–5) n 45 38 44 30 23 57 48 38 46 43 79 35 91 44 33 37 37 29 21
Cusp 5 UM1 % 5.3 5.7 17.5 10.0 0.0 8.7 24.1 12.0 40.0 15.8 10.9 21.2 14.8 25.8 7.4 38.1 18.2 18.8 30.8
(þ¼ASU 2–5) n 38 35 40 20 9 46 29 25 25 19 64 33 81 31 27 21 11 16 13
Carabelli’s
Trait UM1
% 18.2 72.7 68.4 64.7 100 47.6 51.6 73.1 75.0 54.6 58.6 60.7 55.2 38.5 60.7 59.4 61.1 69.2 58.8
(þ¼ASU 2–7) n 33 33 38 17 16 42 31 26 24 22 58 28 87 39 28 32 18 26 17
Parastyle UM3 % 0.0 0.0 0.0 0.0 0.0 2.7 5.4 0.0 2.5 2.6 0.0 4.0 2.7 0.0 0.0 0.0 3.0 4.0 4.6
(þ¼ASU 1–5) n 38 26 28 23 15 37 37 40 40 38 58 25 73 40 24 31 33 25 22
Enamel Extension
UM1
% 2.6 6.4 15.2 6.5 0.0 3.3 4.0 5.0 2.3 4.3 13.5 6.7 1.1 20.0 5.6 22.0 2.6 0.0 27.3
(þ¼ASU 1–3) n 38 47 46 31 18 61 50 40 43 47 89 45 93 50 36 41 38 30 22
Root Number UP1 % 77.3 62.5 76.1 70.6 89.7 63.8 80.4 72.4 83.0 71.1 53.9 70.7 68.6 64.7 53.6 57.9 62.8 58.8 65.0
(þ¼ASU 2þ) n 22 32 46 17 29 47 51 29 47 38 78 41 102 51 28 38 43 34 20
Root Number UM2 % 96.3 72.7 73.5 80.0 82.6 62.2 90.2 77.8 86.7 86.7 81.7 90.3 90.6 76.7 85.0 91.7 76.9 84.0 95.0
(þ¼ASU 3þ) n 27 22 34 15 23 37 41 27 30 30 60 31 85 43 20 36 39 25 20
Peg-Reduced UI2 % 0.0 1.8 0.0 10.3 6.1 0.0 1.6 4.8 0.0 6.7 1.9 4.8 0.0 6.9 2.6 0.0 15.4 11.1 5.3
(þ¼ASU P or R) n 49 57 60 39 33 24 63 42 56 60 54 42 30 29 38 37 13 9 19
Odontome P1–P2 % 0.0 0.0 0.0 0.0 0.0 1.7 0.0 0.0 5.3 0.0 0.0 0.0 1.2 0.0 0.0 0.0 0.0 3.6 0.0
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
J. D. Irish and L. Konigsberg

(þ¼ASU þ) n 50 42 49 31 12 59 41 36 38 33 82 31 85 41 27 34 34 28 21
Congenital
Absence UM3
% 2.0 15.4 3.7 8.6 20.0 30.8 16.7 4.4 7.1 6.8 7.3 17.8 3.0 1.9 12.8 2.4 0.0 0.0 4.2
(þ¼ASU ) n 51 52 54 35 35 65 60 45 56 59 82 45 100 52 39 42 47 35 24
Midline Diastema
UI1
% 2.9 0.0 0.0 5.3 0.0 3.0 3.3 5.9 0.0 1.7 8.7 3.6 15.3 9.1 10.3 2.8 17.4 16.7 10.5
(þ0.5 mm) n 34 52 52 38 33 66 60 34 52 58 23 28 72 44 39 36 23 30 19
Lingual Cusp LP2 % 48.7 61.9 95.7 79.2 66.7 61.5 86.4 80.0 72.4 58.8 86.0 100 40.0 69.0 37.5 86.2 66.7 72.7 83.3
(þ¼ASU 2–9) n 39 21 23 24 12 52 22 35 29 17 50 18 15 29 8 29 15 11 18
Anterior Fovea LM1 % 64.0 17.4 18.8 9.1 14.3 32.0 43.8 20.0 100 28.6 40.0 57.9 69.2 85.0 57.1 85.7 71.4 61.5 91.7
(þ¼ASU 2–4) n 25 23 16 11 7 25 16 15 16 7 35 19 13 20 7 21 7 13 12
Mandibular Torus % 0.0 0.0 1.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
(þ¼ASU 2–3) n 45 51 58 40 37 72 60 43 50 55 81 52 21 37 12 39 36 23 23
Groove Pattern LM2 % 48.9 29.6 45.8 35.3 22.7 18.2 41.3 63.4 50.0 39.4 10.5 29.0 83.3 58.3 66.7 39.4 75.0 68.4 55.0
(þ¼ASU Y) n 45 44 48 34 22 66 46 41 46 33 76 38 18 36 12 33 28 19 20
Rocker Jaw % 2.1 13.7 24.1 10.3 24.3 20.0 5.3 5.9 27.3 10.9 22.0 13.2 14.3 5.4 0.0 7.9 0.0 4.4 8.3
(þ¼ASU 1–2) n 47 51 54 39 37 70 57 34 44 46 82 53 21 37 12 38 35 23 24
Cusp Number LM1 % 6.5 2.3 7.9 0.0 0.0 1.9 0.0 3.1 5.7 0.0 6.9 5.7 5.6 37.0 8.3 46.4 11.8 31.6 20.0
(þ¼ASU 6þ) n 46 43 38 24 10 53 28 32 35 17 72 35 18 27 12 28 17 19 15
Cusp Number LM2 % 50.0 25.0 27.8 19.1 25.0 31.0 41.2 32.1 56.3 25.8 33.3 33.3 52.9 82.9 37.5 81.3 81.0 66.7 73.7
(þ¼ASU 5þ) n 38 36 36 21 12 58 34 28 32 31 75 33 17 35 8 32 21 18 19
Deflecting Wrinkle
LM1
% 18.9 5.7 15.2 9.5 0.0 4.6 11.1 12.0 36.4 30.0 7.0 11.1 33.3 24.0 18.2 40.0 33.3 42.9 40.0
(þ¼ASU 2–3) n 37 35 33 21 8 44 27 25 22 10 57 27 12 25 11 20 6 7 10
C1-C2 Crest LM1 % 0.0 2.9 3.0 9.5 0.0 6.5 0.0 0.0 0.0 0.0 4.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
(þ¼ASU þ) n 41 34 33 21 5 46 27 26 26 12 61 28 12 27 12 26 7 13 14
Protostylid LM1 % 2.4 11.4 22.2 12.5 7.1 16.1 11.5 2.9 32.3 21.4 49.3 27.5 35.3 15.2 0.0 25.8 31.3 23.5 16.7
(þ¼ASU 1–6) n 41 35 36 24 14 56 26 34 31 14 69 40 17 33 12 31 16 17 18
Cusp 7 LM1 % 19.6 4.3 10.9 13.3 0.0 6.7 17.1 7.1 11.6 0.0 3.5 14.6 11.1 31.3 16.7 26.5 34.8 5.3 28.6
(þ¼ASU 2–4) n 51 47 46 30 20 60 35 42 43 29 85 48 18 32 12 34 23 19 21
Tome’s Root LP1 % 14.3 0.0 10.7 5.6 6.7 11.1 25.0 15.4 19.6 9.5 6.0 2.9 25.0 13.9 0.0 28.6 27.6 15.8 50.0
(þ¼ASU 3–5) n 21 47 56 18 30 45 52 26 46 42 50 35 20 36 5 28 29 19 16
Root Number LC % 0.0 1.9 5.1 4.2 6.1 0.0 1.9 2.6 8.2 4.8 1.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
(þ¼ASU 2þ) n 24 52 59 24 33 31 52 38 49 42 65 43 18 33 6 21 30 20 10
Root Number LM1 % 0.0 0.0 5.1 0.0 0.0 2.2 2.0 2.7 2.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.2 0.0 0.0
(þ¼ASU 3þ) n 26 34 39 19 26 46 49 37 39 31 45 29 18 37 11 37 31 21 23
Root Number LM2 % 100 82.4 86.1 75.0 86.7 86.7 94.0 84.6 91.2 84.4 89.6 92.0 100 94.3 88.9 94.7 100 100 95.7
(þ¼ASU 2þ) n 20 34 36 16 30 45 50 26 34 32 48 25 17 35 9 38 27 20 23
Torsomolar Angle
LM3
% 2.4 2.9 2.2 5.9 0.0 17.0 15.7 4.9 4.8 2.6 16.7 16.7 12.5 13.5 0.0 20.6 7.7 5.9 13.6
(þ¼ASU þ) n 42 35 46 34 23 47 51 41 42 39 60 30 16 37 11 34 26 17 22
a
JEM ¼Jebel Moya, EGE ¼Egyptian ‘E’, NAQ ¼Naqada, BAD ¼Badari, SAQ ¼Saqqarah, HES ¼El Hesa, KER ¼Kerma, AGR ¼A-Group, CGR ¼C-Group,
DGR ¼D-Group, MER ¼Meroitic, XGR ¼X-Group, KEN ¼Kenya, NIC ¼Nigeria–Cameroon, ETH ¼Ethiopia, GHA ¼Ghana, CNG ¼Congo, GAB ¼Gabon,
TOD ¼Togo–Dahomey.
b
ASU rank-scale trait breakpoints from Irish (1993, 1997, 1998a,b) and Scott & Turner (1997).
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
Ancient Inhabitants of Jebel Moya

ASUDAS presence/absence dichotomies are
listed under each trait name. As can be seen,
several traits suffer from small sample sizes,
particularly those in anterior teeth that are
commonly missing post-mortem. Such data
should be guardedly interpreted because they
may not be representative of the populations from
which they derive.
Overall, Jebel Moya’s trait percentages appear
largely intermediate in occurrence relative to
those of the remaining samples, although it
does exhibit exceptionally low incidences of
UI2 tuberculum dentale and, particularly, UM1
Carabelli’s. Moreover, in moving down the
columns it appears that none of the other samples
closely parallel the Jebel Moyans for more than
just a handful of trait occurrences.
Trait editing
Before applying the Mahalanobis statistic, six traits
exhibiting four or more samples with <10 cases
were dropped from the analysis (UI1 labial
curvature, UI1 shoveling, UI1 double shoveling,
UI2 interruption groove, UI2 tuberculum dentale,
LM1 anterior fovea). Another nine traits character-
ised by shared low or high expression were also
deleted (palatine torus, UM3 parastyle, peg-reduced
UI2, premolar odontome, mandibular torus, LM1
C1-C2 crest, LC root number, LM1 root number,
LM2 root number). This editing yielded 21 highly
variable dental traits with adequate sample sizes for
submission to CA. Correspondence analysis tabular
output (omitted for brevity but available from the
first author) and the biplot (Figure 2) then identified
the most influential of these traits in driving inter-
sample variation.
The biplot’s enumerated white diamonds
denote the 21 dental traits. The 19 samples are
represented by black dots; although not labelled
to facilitate plot legibility, they can be seen to
form a Y-pattern. Sub-Saharan samples comprise
the left and North Africans the right half of the
‘Y’. This distribution is discussed further below.
Over 57% of the variance explained by the CA
model is illustrated in Figure 2. Most of it (42.5%)
is distributed along the first dimension or x-axis.
Thus, for example, traits 15 (6-cusped LM1) and 3
(UC distal accessory ridge) distinguish the
sub-Saharan samples, whereas14(rockerjaw)and
10 (UM3 congenital absence) occur most frequently
in North Africans. Because only 15% of the
variation occurs on the second dimension
(y-axis), trait/sample associations are less clear cut.
Still, high frequencies of traits 7 (UM1 enamel
extension) and 21 (LM3 torso-molar angle) are
foundinthetopmostsamples(fromlefttoright,
GHA, MER and HES), while 13 (LM2 Y-groove)
characterises the bottom two (ETH and JEM),
among others. Beyond this, a drop to <10% of
the total variance on a third dimension indicates
that sufficient information is provided in the two-
dimensional biplot to identify the most influential
traits. They include: UC Bushman canine, UC distal
accessory ridge, UM1 cusp 5, UM1 enamel
extension, UM3 congenital absence, UI1 midline
diastema, LM2 groove pattern, rocker jaw, LM1
cusp number, LM1 deflecting wrinkle, LM1
protostylid, LM1 cusp 7, LP1 Tome’s root, and
Figure 2. Two-dimensional correspondence analysis
biplot illustrating relationships among the 19 samples
and 21 dental traits. Samples are depicted by unlabelled
black dots and traits with enumerated white diamonds.
Numbers correspond to the following traits: (1) UI1 wing-
ing, (2) UC Bushman canine, (3) UC distal accessory
ridge, (4) UM2 hypocone, (5) UM1 cusp 5, (6) UM1
Carabelli’s trait, (7) UM1 enamel extension, (8) UP1 root
number, (9) UM2 root number, (10) UM3 congenital
absence, (11) UI1 midline diastema, (12) LP2 lingual
cusp, (13) LM2 groove pattern, (14) rocker jaw, (15)
LM1 cusp number, (16) LM2 cusp number, (17) LM1
deflecting wrinkle, (18) LM1 protostylid, (19) LM1 cusp
7, (20) LP1 tome’s root, and (21) LM3 torsomolar angle.
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
J. D. Irish and L. Konigsberg

LM3 torsomolar angle. Standard PCA of trait
percentagesyieldsananalogouslist.
Intra-sample variability
Of these 14 traits, UM3 congenital absence, UI1
midline diastema, LM2 groove pattern and LM3
torsomolar angle are initially recorded as dichot-
omous data. Thus, rank-scale standard deviations
could only be determined for the ten remaining
traits among samples. Jebel Moya was not found
to exhibit the greatest variability for any of these
traits; in fact, several of its standard deviations are
lower than those in the majority of comparative
samples—including several that were long
assumed to be temporally and spatially homo-
geneous based on site reports and other
information (e.g. X-Group, A-Group, Kenya, El
Hesa, etc.).
The amount of tabular information required to
present comparisons of 10 ASUDAS trait
standard deviations in the 19 samples is con-
siderable. Thus, only a representative example,
i.e. results for the LM1 cusp 7 comparison, is
presented here. The remaining nine tables may be
requested from the first author. In Table 3 the
numbers of individuals scored in each sample are
listed, as are the means—which vary relative to
the arbitrary mean of 5.0; the mean gives an
indication of the number of individuals receiving
each of five LM1 cusp 7 grades across the rows,
but is otherwise not used to compare variability
among samples. Lastly, the corresponding stan-
dard deviations and their ranks, which indicate
the most to least variable samples, are presented.
It can be seen that Jebel Moya has a lower
standard deviation for this particular trait than 13
other samples (Table 3).
Dental affinities
The distance matrix generated from the suite of
14 traits is presented in Table 4. Values for Jebel
Moya support the view implied by the qualitative
comparison of frequencies; the range of 1.53–
3.62 is roughly intermediate to those of the
comparative samples. Likewise, its mean D
2
value
of 2.59 (i.e. the sum of 18 pairwise comparisons
divided by total) is less than that in seven,
although greater than that in 11 samples; this
mean suggests there is some degree of phenetic
distinction. Lastly, looking at individual D
2
values, Jebel Moya is most akin to the sample
from Ethiopia and least like Meroitic Nubians.
The output from two-dimensional MDS of
distances among the 19 dental samples is seen in
Figure 3. North African comparative samples are
represented by black squares; those of sub-
Saharan origin are identified by white triangles.
Their overall distribution is similar to that in
Figure 2. Two-dimensional MDS of a 21-trait
MMD comparison (not shown), undertaken prior
to write-up, is also equivalent. That is, Jebel Moya
is intermediate to both groups on the x-axis, and
distinct along the y-axis. Such methodological
concordance implies that the dental affinities are
real, and not an artefact of the statistic or
illustrative method used.
Although there is not, of course, a direct 1:1
relationship between MDS diagram (Figure 3)
and distance matrix, the former does provide a
good representation; Kruskal’s stress formula 1
value is 0.174 and the r
2
is 0.841. For the purposes
of methodological comparison, two-dimensional
MDS of the original craniometric-derived D
2
Table 3. LM1 cusp 7 rank-scale variation
a
among the
samples
Sample Trait N Mean SD SD Rank
Jebel Moya 51 5.423 0.832 14
Egyptian ‘E’ 47 3.933 1.308 6
Naqada 46 4.478 1.289 7
Badarian 30 4.563 1.406 5
Saqqara 21 5.720 0.111 19

El Hesa 60 4.429 1.070 11
Kerma 35 4.742 1.558 4
A-Group 42 3.465 1.670 3
C-Group 43 4.529 1.163 10
D-Group 29 5.699 0.113 18
Meroitic 85 5.056 0.677 16
X-Group 48 4.184 1.781 1

Kenya 18 4.720 1.227 9
Nigeria–Cameroon 32 5.327 1.238 8
Ethiopia 12 4.203 1.762 2
Ghana 34 6.020 0.316 17
Congo 23 5.786 0.771 15
Gabon 19 4.940 0.945 12
Togo–Dahomey 21 5.617 0.938 13
a
See text for description of the method.

Sample exhibiting lowest trait variation.

Sample exhibiting greatest trait variation.
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
Ancient Inhabitants of Jebel Moya

values among 20 samples from Mukherjee et al.
(1955) is presented in Figure 4. It too displays
a separation between the sub-Saharan and
North African samples, with an intermediate,
yet distinct position for Jebel Moya. The
corresponding stress and r
2
values are 0.164
and 0.892. For both MDS figures, minimal
improvement in these measures of fit did not
warrant the use of more-difficult-to-interpret
three-dimensional graphing.
Figure 3. Two-dimensional MDS of 14-trait Mahalano-
bis D
2
dental distances between Jebel Moya and 18
comparative samples from the present study. North
Africans are depicted with black squares and sub-
Saharan Africans with white triangles. The three-letter
sample abbreviations are defined in the text and Table 1.
Table 4. Mahalanobis distance matrix for 14 non-metric dental traits among the 19 samples
Sample jem ege naq bad saq hes ker agr cgr dgr mer xgr ken nic eth gha cng gab tod
Jebel Moya 0
Egyptian ‘E’ 2.65 0
Naqada 2.63 1.78 0
Badarian 2.09 1.47 1.60 0
Saqqara 2.64 1.20 1.84 1.71 0
El Hesa 2.97 1.80 2.37 1.58 1.59 0
Kerma 2.13 2.37 2.53 1.69 2.38 1.70 0
A-Group 1.61 2.34 2.28 1.75 2.32 2.50 1.70 0
C-Group 2.92 2.35 1.55 2.07 2.15 2.44 2.42 2.59 0
D-Group 2.60 1.82 1.53 1.70 1.72 2.24 2.32 2.01 1.38 0
Meroitic 3.62 2.51 2.59 2.38 2.58 2.04 2.73 3.03 3.04 2.63 0
X-Group 2.82 1.64 2.06 1.45 2.03 1.44 1.77 2.33 2.24 2.09 1.73 0
Kenya 2.64 3.28 2.93 2.59 3.21 3.00 2.38 2.13 2.56 2.49 3.11 2.67 0
Nigeria–Cameroon 2.56 3.33 2.80 2.70 3.46 3.33 2.64 2.25 3.31 3.13 2.86 2.49 2.59 0
Ethiopia 1.53 2.44 2.51 1.68 2.48 2.54 1.89 1.27 2.81 2.37 3.23 2.26 2.24 2.16 0
Ghana 3.10 3.52 2.59 2.85 3.47 3.27 2.87 2.80 2.91 2.96 2.82 2.59 2.95 1.43 2.76 0
Congo 2.58 3.97 3.46 2.96 4.03 3.71 2.76 2.50 3.50 3.33 3.62 3.17 1.93 2.06 2.26 2.57 0
Gabon 2.78 3.55 3.06 2.92 3.39 3.55 3.07 2.35 2.96 2.58 3.24 2.97 1.81 2.14 2.31 2.37 1.93 0
Togo–Dahomey 2.73 3.70 3.04 2.94 3.61 3.35 2.42 2.29 3.32 3.19 3.10 2.89 2.59 1.45 2.39 1.77 2.22 2.53 0
Figure 4. Two-dimensional MDS of Mahalanobis D
2
craniometric distances between Jebel Moya and 19
comparative samples: derived from distance matrix in
Mukherjee et al. (1955). North Africans are depicted with
black squares and sub-Saharan Africans with white
triangles. The three-letter sample abbreviations are listed
in Table 1.
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
J. D. Irish and L. Konigsberg

Discussion
Inter-sample variation
The most obvious pattern among dental samples
(Figure 3) is, as noted, the divergence between
sub-Saharan and North Africans. It emulates
those derived from numerous other biocultural
data (among others, Murdock, 1959; Greenberg,
1966; Hiernaux, 1975; Mourant, 1983; Nurse
et al., 1985; Roychoudhury & Nei, 1988; Howells,
1989). And, not surprisingly, the present results
are equivalent to those in prior dental affinity
studies (Irish, 1993, 1997, 1998a,b,c); as before,
sub-Saharan samples, especially those from west
Africa, exhibit complex mass-additive traits,
including: UC Bushman canine (i.e. 2 in Figure 2),
UC distal accessory ridge (3), UI1 midline diastema
(11), 6-cusped LM1 (15), LM1 cusp 7 (19), and LP1
Tome’s root (20). Conversely, North Africans have
been shown to possess simple mass-reduced
morphology (Irish, 1998a, b), with low frequencies
of the above, and a high incidence of UM3 absence
(10) and rocker jaw (14). The present trait
combinations correspond with the previously
defined Sub-Saharan African- and North African
Dental Trait Complexes, respectively (Irish, 1997,
1998a,b).
This divergence also carries over to the
craniometric data, as evidenced by the MDS of
D
2
values (Figure 4) that are provided in
Mukherjee et al. (1955). Their Tigrean sample
from Ethiopia is associated with North Africans
cranially, and intra-regional sample affinities do
differ, but the north–south dichotomy is other-
wise maintained. By comparison, dental-based
affinities (Figure 3) actually correspond to
geographical provenance more closely than
those based on craniometrics. As can be seen,
the west (GHA, TOD, NIC), central (CNG,
GAB), and east (KEN, ETH) sub-Saharan dental
samples form separate subgroups. Egyptians
(BAD, NAQ, HES, EGE, SAQ) also cluster
together, although the Nubian samples are
somewhat dispersed.
Before proceeding, an interesting side note is
worth mentioning. Although not specified, the
original two-dimensional graph of craniometric
affinities published in Mukherjee et al. (1955, their
Figure 5.1: 85), and recreated here in Figure 5,
clearly involves translation to geometric dis-
tances, like with MDS. In fact, there is a marked
concordance between their graph and the current
Figure 4. For the most part, there is only some
minor shifting of points (e.g. Egyptian ‘Negro’ is
moved toward the sub-Saharan samples, Ibo is
closer to Jebel Moya, etc.), which is under-
standable in their conversion of the D
2
inter-
sample matrix values to distances on a hand-made
graph. However, there is also one major
difference; Mukherjee and associates placed their
Badarian Egyptian sample within the sub-Saharan
cluster, while puzzling over this unexpected
affinity (Mukherjee et al., 1955: 86). Inspection of
the original D
2
matrix (their Table 5.6: 84) does,
in reality, indicate a Badarian affiliation to North
Africans, not sub-Saharan samples. It is therefore
likely that an error was made in construction of
their original figure when converting inter-sample
distances to x- and y-coordinates. A similar
plotting inaccuracy would have taken place in
Figure 4 if the Badarian (BAD) sample had
erroneously received a negative rather than
positive x-coordinate.
Figure 5. Reproduction of the D
2
craniometric distance
graph from Mukherjee et al. (1955, Figure 5.1: 85): rotated
908clockwise and otherwise modified relative to
the original to facilitate comparison with Figure 4. Note
the erroneous position of the Badarian sample
within the cluster of sub-Saharan Africans (see text for
details).
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
Ancient Inhabitants of Jebel Moya

The place of Jebel Moya
Beyond a north–south divergence, the most
obvious similarity between Figures 3 and 4 is
Jebel Moya’s relative location. As noted, it appears
intermediate to, yet distinct from, sub-Saharan and
North Africans in dental trait expression. Similar
observations were reported regarding cranial
variation (Mukherjee et al., 1955). The main
difference is that Jebel Moya, based on individual
D
2
values, shows the closest dental affinity to
North Africans, whereas cranially it is more akin to
sub-Saharan samples. This contrast may involve,
among others, differential heritability in dental
versus cranial trait expression, a cline in selective
forces favouring sub-Saharan cranial features with
decreasing latitude (e.g. Hiernaux, 1975) and/or,
simply, the several differences in sample compo-
sition between studies. In any event, craniodental
features characterising both large comparative
groupings appear to be manifest in the Jebel
Moyans. Therefore, if the present samples are
representative of their respective populations, and
phenetic approximates genetic relatedness, the
reason for this overall affinity may be that Jebel
Moyans were: (1) an admixed people, comprising
genetic elements from populations living in various
regions surrounding central Sudan; and/or (2) a
heterogeneous populace consisting of actual
individuals from these regions.
As noted, population heterogeneity was first
intimated by Addison (1949). Similarities in
pottery between regions suggested influence from
northern Sudan, while some of the disparate
mortuary customs reflect those from farther south.
Yet, he believed that the original Jebel Moya
settlers came from the west. Although mentioning
a stone tool resemblance with far west Africa, the
place of origin was said to be the outlying desert
west of the White Nile. He then posited that
‘physical characteristics of the original settlers
would soon become modified by interbreeding
with local stocks’ (Addison, 1949: 260).
The possibility of a western origin was later
reprised. Caneva (1991) reported pottery sim-
ilarities between the Central Sahara, near Tibesti,
Borkou and Ennedi in Chad, and Jebel Moya from
the end of the 5th millennium BC. Connah
(1981) also noted that bone tools and ceramic
figurines at Daima, Nigeria (1st millennium BC to
2nd millennium AD) are reminiscent of those at
Jebel Moya. Yet the idea of a northern Sudan link
was also sustained. Pottery motifs, vessel forms,
lip-plugs, and stone tools of the Butana Industry
(ca. 3rd millennium BC), on the Ethiopian border
in the Atbara drainage, mirror those at Jebel
Moya. C-Group and Meroitic Nubian influences
in pottery are also evident (Clark, 1973, 1984;
Clark & Stemler, 1975). The presence of Napatan
and Meroitic grave items has already been
mentioned. However, in concert with Addison
(1949), Clark (1973, 1984) asserted that the
people of Jebel Moya and nearby sites were, at
least culturally, distinctive from these outside
groups. Thus, the cultural manifestation of this
place and time received a separate designation
termed the Jebel Moya Complex (Clark, 1984).
According to Gerharz (1994) the appearance
of this culture coincided with the advent of his
Phase II (ca. 3000–800 BC) and continued
through Phase III (800–100 BC). This sequence
corresponds with the recovered archaeological
and skeletal remains. Like other workers, Gerharz
regards it as a distinctive heterogeneous culture
that combined elements of various outside
groups. The lack of uniformity in grave types,
orientation and inventories (see also Addison,
1949) is especially supportive of this premise.
Specifically, he sees Jebel Moya as having been an
‘annual meeting place of widely distributed
segmentary family units, the common identity
of which was maintained by their periodical
cohabitation there’ (Gerharz, 1994: 330).
In light of these archaeological interpretations
the craniodental affinities make sense. In both
cases Jebel Moyans exhibit traits characteristic of
many groups (i.e. ‘intermediate’), while concur-
rently demonstrating their own distinct biocul-
tural composition. Mukherjee et al. (1955)
specifically investigated the issue of population
make-up. Non-metric traits in crania recovered
from different levels were thought to suggest
variation over time. Evidence for cultural change
between strata was also reported. However, the
same stratigraphic control problems (Caneva,
1991; Gerharz, 1994) affecting the original site
dating rendered these findings useless as well;
additionally, calculations of craniometric stan-
dard deviations and, in particular, coefficients of
variation yielded contrary findings. Despite some
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
J. D. Irish and L. Konigsberg

internal variation, the series was not found to be
significantly more variable than the Egyptian ‘E’
sample. In the end, it was concluded that ‘physical
characters of the Jebel Moyans are reliably
represented by the mean values for the sample’
(Mukherjee et al., 1955: 64).
The same interpretation is implied by the
comparisons of rank-scale dental trait standard
deviations among samples. Jebel Moya does not
exhibit higher levels of internal dental variability
than a number of comparative samples, some of
which have been assumed to represent relatively
homogeneous populations. An assessment of
dental affinity between time-successive Jebel
Moya subsamples also provides support. Specifi-
cally, using Gerharz’ (1994) age indicators
(above) with the grave inventory of Addison
(1949), it was determined that 31 of the 58
dentitions date to Phase II, whereas the other 27
are definite or probable Phase III. The original
trait complement was then compared between
these subsamples with the MMD. Use of all 36
traits permits a more complete intra-sample
comparison, and the MMD is most effective
when comparing the larger number of traits, as
long as they are not correlated. The statistic also
has a significance test, where MMD >2SD
indicates that the null hypothesis of population
equality is rejected at the 0.025 level (Sjøvold,
1977). To test for pairwise correlation, the
subsamples’ non-dichotomised data were sub-
mitted to Kendall’s tau-b. Only 30 of 629 pairwise
comparisons are strongly (0.5) correlated. The
resulting MMD of 0.00 patently denotes a lack of
significant difference between subsamples.
Although the population was probably affected
by outside influences, it appears that it may have
displayed, to employ an oxymoron, uniform
heterogeneity through time. Thus, like the
craniometric means, Jebel Moya dental frequen-
cies are probably representative of a population
that, although unique, was relatively uniform and
stable in its composition over a span of some
3000 years.
Pairwise affinities
Beyond the general relationship of Jebel Moya to
the sub-Saharan and North African groupings,
specific between-sample comparisons can be used
to address briefly the several affiliations suggested
by the archaeological evidence. Craniometric
affinities support the unlikely concept of a far
west African link (Addison, 1949; Connah, 1981);
individual D
2
values show Jebel Moya as most
akin to the historic Ibo and Cameroons samples,
and least like the spatially proximate Tigrean,
Taita, and X-Group (Mukherjee et al., 1955). On
the other hand, dental D
2
values (Table 4) and
Figure 3 seem more intuitively plausible. Jebel
Moya is distinct from west Africans, but nearest
Ethiopia (ETH) and the A-Group Nubians
(AGR). The latter two affinities, in particular,
support some of the archaeological links (Clark,
1973, 1984; Clark & Stemler, 1975); the Ethiopia
sample, from the northern half of that country, is
adjacent to the setting of the Jebel Moya-like
Butana Industry, and the A-Group is of a similar
age to Butana and early Phase II Jebel Moya.
However, Jebel Moya is least like Meroitic (MER)
and, to a lesser extent, C-Group Nubians (CGR),
which seems counterintuitive based on this same
evidence (Clark, 1973, 1984; Clark & Stemler,
1975). The reasons for such discrepancies are not
apparent based on the dental data, but it is clear,
at least in this case, that cultural influence does
not necessarily translate into biological affinity.
Lastly, none of the comparative samples are
conducive to exploring the Central Saharan link
suggested by Caneva (1991); however, the first
author studied a 19th–20th century dental sample
(n¼29) from the Tibesti, Borkou and Ennedi
regions of Chad (Irish, 1993) that was used in a
second 14-trait D
2
comparison (not shown).
Although it is too recent to explore directly a 5th
millennium BC association, the results may,
nevertheless, be suggestive. The Chad sample
is wholly unlike Jebel Moya and most closely akin
to the west Africans (esp. Nigeria–Cameroon).
All other individual D
2
distances and inter-sample
relationships, as illustrated by Figure 3, remain
essentially constant.
Summary and conclusions
A half-century after The Ancient Inhabitants of Jebel
Moya (Sudan) was written by Mukherjee et al.
(1955), the craniometric findings correspond
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
Ancient Inhabitants of Jebel Moya

with the new dental affinities and previous
archaeological data to portray the site’s popu-
lation as a biocultural amalgam. The application
of the Mahalanobis D
2
statistic to cranial
measurements suggests a sub-Saharan link,
particularly with far-west Africans; yet, as
Mukherjee and associates observe, the sample
still ‘deserves a special position in relation to all
the series under consideration’ (p. 88). The dental
study, comparing many of the same samples from
the craniometric report, finds some concordance
between methods (i.e. ‘intermediate-yet-dis-
tinct’); yet, the overall place of the Jebel Moyans
appears to be within a greater northeast African
sphere of biological influence, based on the
phenetic affinities to various samples past (e.g. A-
Group) and present (e.g. Ethiopia). Lastly,
cultural attributes suggest a mosaic of northern,
southern and western influence; yet, the site
inhabitants incorporated all to yield their own
distinct Jebel Moya Complex.
In the end, despite myriad obstacles affecting
completion of the Mukherjee et al. (1955)
report—related to problems during and after
the fieldwork, and a subsequent lack of physical
anthropological and, to a lesser extent, archae-
ological study of the collections, it appears that
new information can still be gleaned from the old
site. The inhabitants were cranially sub-Saharan,
dentally North African, culturally aligned with
both regions yet, in all instances, distinct.
Although ostensibly starting out as a diverse
group of individuals, the Jebel Moyans apparently
came to comprise what may best be described as a
uniformly heterogeneous population that exhib-
ited its own distinctive biocultural identity.
Acknowledgements
Thanks are extended to the individuals at those
institutions from which the data were collected:
Christy Turner, Charles Merbs and Donald Mor-
ris from Arizona State University; Douglas Ube-
laker and David Hunt, National Museum of
Natural History; Ian Tattersall, Jaymie Brauer,
Gary Sawyer and Ken Mowbray, American
Museum of Natural History; Andre Langaney,
Frances Roville-Sausse, Miya Awazu Periera da
Silva, Phillippe Mennecier and Alain Froment,
Muse
´e de l’Homme; Neils Lynnerup from the
Panum Instituttet; and Rob Foley, Marta Lahr
and Maggie Bellatti, University of Cambridge.
Dana Thomas, University of Alaska Fairbanks,
provided useful statistical advice. The Mahalano-
bis distances for the non-metric dental data were
calculated using a program written by the second
author, as modified by Steve Byers. The research
was supported by the National Science Founda-
tion (BNS-9013942 and BNS-0104731), Arizona
State University Research Development Pro-
gram, and the American Museum of Natural
History. Lastly, the comments of IJO editor Terry
O’Connor and the anonymous reviewers were
insightful and greatly appreciated.
References
Adams WY. 1977. Nubia: Corridor to Africa. Princeton
University Press: Princeton.
Addison F. 1949. Jebel Moya. Oxford University Press:
London.
Addison F. 1956. Second thoughts on Jebel Moya.
Kush 4: 4–18.
Bedrick EJ, Lapidus J, Powell JF. 2000. Estimating the
Mahalanobis Distance from mixed continuous and
discrete data. Biometrics 56: 394–401.
Benze
´cri JP. 1992. Correspondence Analysis Handbook.
Dekker: New York.
Bermudez de Castro JM. 1989. The Carabelli trait in
human prehistoric populations of the Canary
Islands. Human Biology 61: 117–131.
Berry AC, Berry RJ. 1967. Epigenetic variation in
the human cranium. Journal of Anatomy 101: 361–
379.
Berry AC, Berry RJ. 1972. Origins and relationships of
the ancient Egyptians. Based on a study of non-
metrical variations in the skull. Journal of Human
Evolution 1: 199–208.
Berry AC, Berry RJ. 1973. Origins and relations of the
ancient Egyptians. In Population Biology of the Ancient
Egyptians, Brothwell DR, Chiarelli BA (eds). Aca-
demic Press: New York; 200–208.
Brace CL, Tracer DP, Yaroch LA, Robb J, Brandt K,
Nelson AR. 1993. Clines and clusters versus Arace:
A test in ancient Egypt and the case of a death
on the Nile. Yearbook of Physical Anthropology 36:
1–31.
Calcagno JM. 1986. Dental reduction in post-
Pleistocene Nubia. American Journal of Physical Anthro-
pology 70: 349–363.
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
J. D. Irish and L. Konigsberg

Caneva I. 1991. Jebel Moya revisited: A settlement of
the 5th millennium BC in the middle Nile Basin.
Antiquity 63: 262–268.
Clark JD. 1973. The University of California,
Berkeley, expedition to the Central Sudan. Nyame
Akuma 3: 56–64.
Clark JD. 1984. Prehistoric cultural continuity and
economic change in the Central Sudan in the Early
Holocene. In From Hunters to Farmers: The Causes and
Consequences of Food Production in Africa, Clark JD,
Brandt SA (eds). University of California Press:
Berkeley; 113–126.
Clark JD, Stemler A. 1975. Early domesticated sor-
ghum from Central Sudan. Nature 254: 588–591.
Clausen SE. 1988. Applied Correspondence Analysis: An
Introduction. Sage Publications: Thousand Oaks, CA.
Connah G. 1981. Three Thousand Years in Africa: Man and
His Environment in the Lake Chad Region of Nigeria.
Cambridge University Press: Cambridge.
Coppa A, Cucina A, Mancinelli D, Vargiu R, Calcagno
JM. 1998. Dental anthropology of central-southern,
Iron Age Italy: The evidence of metric versus non-
metric traits. American Journal of Physical Anthropology
107: 371–386.
Donlon DA. 2000. The value of infracranial nonmetric
variation in studies of modern Homo sapiens:An
Australian focus. American Journal of Physical Anthro-
pology113: 349–368.
Eliason SR. 1993. Maximum Likelihood Estimation: Logic and
Practice. Sage Publications: Newbury Park.
Gerharz RG. 1994. Jebel Moya. Meroitica 14. Akademie
Verlag: Berlin.
Green R, Suchey J. 1976. The use of inverse sine
transformation in the analysis of non-metrical data.
American Journal of Physical Anthropology 45: 61–68.
Greenacre MJ, Degos L. 1977. Correspondence
analysis of HLA gene frequency data from 124
population samples. American Journal of Human Genetics
29: 60–75.
Greenberg JH. 1966. The Languages of Africa. Indiana
University: Bloomington.
Greene DL. 1972. Dental anthropology of early Egypt
and Nubia. Journal of Human Evolution 1: 315–
324.
Greene DL. 1982. Discrete dental variations and
biological distances of Nubian populations. American
Journal of Physical Anthropology 58: 75–79.
Haaland R. 1987. Socio-economic Differentiation in the
Neolithic Sudan. Cambridge Monographs in African
Archaeology 20. BAR International: Oxford.
Haeussler AM, Turner CGII, Irish JD. 1988. Concor-
dance of American and Soviet methods in dental
anthropology. American Journal of Physical Anthropol-
ogy 75: 218.
Hanihara T. 1992. Dental and cranial affinities among
populations of East Asia and the Pacific: The basic
populations in East Asia, IV. American Journal of
Physical Anthropology 88: 163–182.
Hemphill BE. 1998. Biological affinities and adap-
tations of Bronze Age Bactrians: III. An initial
craniometric assessment. American Journal of Physical
Anthropology 106: 329–348.
Hiernaux J. 1975. The People of Africa. Charles Scribner’s
Sons: New York.
Hofmann I. 1967. Die Kulturen des Niltals von Aswan bis
Sennar: Vom Mesolithikum bis Zum Ende der Christlichen
Epoche. Kommissionsverlag Cram, de Gruyter & Co:
Hamburg.
Howells WW. 1989. Skull Shapes and the Map: Cranio-
metric Analyses in the Dispersion of Modern Homo. Papers
of the Peabody Museum of Archaeology and
Ethnology, Vol. 79. Harvard University: Cam-
bridge, MA.
Irish JD. 1993. Biological Affinities of Late Pleistocene through
Modern African Aboriginal Populations: The Dental
Evidence. PhD dissertation, Arizona State University,
Tempe.
Irish JD. 1994. The African dental complex: Diagnos-
tic morphological variants of modern sub-Saharan
populations. American Journal of Physical Anthropology
18 (Suppl.):112.
Irish JD. 1997. Characteristic high- and low-frequency
dental traits in sub-Saharan African populations.
American Journal of Physical Anthropology 102:
455–467.
Irish JD. 1998a. Dental morphological indications of
population discontinuity and Egyptian gene flow in
post-Paleolithic Nubia. In Human Dental Development,
Morphology, and Pathology: A Tribute to Albert A.
Dahlberg, Lukacs JR (ed.). University of Oregon
Anthropological Papers 54. University of Oregon
Press: Eugene; 155–172.
Irish JD. 1998b. Dental morphological affinities of
late Pleistocene through recent sub-Saharan
and North African peoples. Bulletins et Memoires
de la Societe
´d’Anthropologie de Paris Nouvelle serie 10:
237–272.
Irish JD. 1998c. Diachronic and synchronic dental
trait affinities of Late and post-Pleistocene peoples
from North Africa. Homo 49: 138–155.
Irish JD. 1998d. Ancestral dental traits in recent sub-
Saharan Africans and the origins of modern humans.
Journal of Human Evolution 34: 81–98.
Irish JD. 2000. The Iberomaurusian enigma: North
African progenitor or dead end? Journal of Human
Evolution 39: 393–410.
Irish JD. 2005. Population continuity versus disconti-
nuity revisited: Dental affinities among Late
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
Ancient Inhabitants of Jebel Moya

Paleolithic through Christian era Nubians. American
Journal of Physical Anthropology 128: 520–535.
Irish JD. 2006. Who were the ancient Egyptians?
Dental affinities among Neolithic through post-
dynastic peoples. American Journal of Physical Anthro-
pology 129: 529–543.
Irish JD, Guatelli-Steinberg D. 2003. Ancient teeth
and modern human origins: an expanded compari-
son of African Plio-Pleistocene and Recent world
dental samples. Journal of Human Evolution 45:
113–144.
Irish JD, Turner CGII. 1990. West African dental
affinity of late Pleistocene Nubians: Peopling of
the Eurafrican–South Asian triangle II. Homo 41:
42–53.
Ishida H, Dodo Y. 1997. Cranial variation in prehis-
toric human skeletal remains from the Marianas.
American Journal of Physical Anthropology 104:
399–410.
Jackes M, Silva AM, Irish JD. 2001. Dental
morphology: a valuable contribution to our under-
standing of prehistory. Journal of Iberian Archaeology 3:
97–119.
Johnson AL, Lovell NC. 1994. Biological differen-
tiation at predynastic Naqada, Egypt: an analysis
of dental morphological traits. American Journal of
Physical Anthropology 93: 427–433.
Keita SOY. 1990. Studies of ancient crania from
northern Africa. American Journal of Physical Anthro-
pology 83: 35–48.
Keita SOY. 1992. Further studies of crania from
ancient northern Africa: an analysis of crania from
First Dynasty Egyptian tombs. American Journal of
Physical Anthropology 87: 245–254.
Keita SOY. 1996. Analysis of Naqada predynastic
crania: a brief report. In Interregional Contacts in the
Later Prehistory of Northeastern Africa, Krzyzaniak L,
Kroeper K, Kobusiewicz M (eds). Poznan Archae-
ological Museum: Poznan; 203–213.
Kitagawa Y, Manabe Y, Oyamada J, Rokutanda A.
1995. Deciduous dental morphology of the prehis-
toric Jomon people of Japan: comparison of non-
metric characters. American Journal of Physical
Anthropology 97: 101–111.
Kitson E. 1931. A study of the Negro skull with special
reference to the crania from Kenya Colony. Biome-
trika 23: 271–314.
Konigsberg LW. 1990. Analysis of prehistoric bio-
logical variation under a model of isolation by
geographic and temporal distance. Human Biology
62: 49–70.
Konigsberg LW, Kohn LAP, Cheverud JM. 1993.
Cranial deformation and nonmetric trait variation.
American Journal of Physical Anthropology 90: 35–48.
Kruskal JB, Wish M. 1978. Multidimensional Scaling. Sage
Publications: Beverly Hills, CA.
Larsen CS. 1997. Bioarchaeology. Cambridge University
Press: Cambridge.
Lukacs JR, Hemphill BE. 1991. The dental anthropol-
ogy of prehistoric Baluchistan: a morphometric
approach to the peopling of South Asia. In Advances
in Dental Anthropology, Kelley MA, Larsen CL (eds).
Wiley-Liss: New York; 77–119.
Martin PF. 1921. The Sudan in Evolution: A Study of the
Economic, Financial and Administrative Conditions of the
Anglo-Egyptian Sudan. Constable and Co. Ltd:
London.
Mourant AE. 1983. Blood Relations: Blood Groups and
Anthropology. Oxford University Press: Oxford.
Mukherjee R, Rao CR, Trevor JC. 1955. The Ancient
Inhabitants of Jebel Moya (Sudan). Cambridge Univer-
sity Press: Cambridge.
Murdock GP. 1959. Africa: Its Peoples and Their Culture
History. McGraw-Hill: New York.
Nielsen OV. 1970. Human Remains: Metrical and Non-
metrical Anatomical Variations. The Scandinavian Joint
Expedition to Sudanese Nubia: Odense, Denmark.
Nurse GT, Weiner JS, Jenkins T. 1985. The Peoples of
Southern Africa and Their Affinities. Clarendon Press:
Oxford.
Phillips D. 1995. Correspondence analysis. Social
Research Update 7: 1–8.
Pietrusewsky M. 2004. Multivariate comparisons of
female cranial series from the Ryukyu Islands and
Japan. Anthropological Science 1–13; published online
13 July 2004 in J-STAGE, www.jstage.jst.go.jp
Prowse TL, Lovell NC. 1996. Concordance of cranial
and dental morphological traits and evidence for
endogamy in ancient Egypt. American Journal of
Physical Anthropology 101: 237–246.
Rightmire GP. 1999. Dental variation and human
history. Review of Archaeology 20: 1–3.
Roseman CC, Weaver TD. 2003. Multivariate appor-
tionment of global human craniometric diversity.
American Journal of Physical Anthropology 125: 257–263.
Roychoudhury AK, Nei M. 1988. Human Polymorphic
Genes World Distribution. Oxford University Press:
New York.
Sakuma M, Ogata T. 1987. Sixth and seventh cusp on
lower molar teeth of Malawians in east-central Africa.
Japanese Journal of Oral Biology 29: 738–745.
Schneider KN. 1986. Dental caries, enamel compo-
sition, and subsistence among prehistoric Amerin-
dians of Ohio. American Journal of Physical
Anthropology 71: 95–102.
Sciulli PW. 1990. Deciduous dentition of a Late
Archaic population of Ohio. Human Biology 62:
221–245.
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
J. D. Irish and L. Konigsberg

Scott GR. 1973. Dental Morphology: A Genetic Study of
American White Families and Variation in Living Southwest
Indians. PhD dissertation, Arizona State University,
Tempe.
Scott GR. 1980. Population variation of Carabelli’s
trait. Human Biology 52: 63–78.
Scott GR, Turner CGII. 1997. The Anthropology of
Modern Human Teeth: Dental Morphology and its Variation
in Recent Human Populations. Cambridge University
Press: Cambridge.
Scott GR, Yap Potter RH, Noss JF, Dahlberg AA,
Dahlberg T. 1983. The dental morphology of Pima
Indians. AmericanJournal of Physical Anthropology61: 13–31.
Sjøvold T. 1973. Occurrence of minor non-metrical
variants in the skeleton and their quantitative treat-
ment for population comparisons. Homo 24: 204–233.
Sjøvold T. 1977. Non-metrical divergence between
skeletal populations: The theoretical foundation
and biological importance of C.A.B. Smith’s mean
measure of divergence. Ossa 4(Suppl.):1–133.
Smith P, Shegev M. 1988. The dentition of Nubians
from Wadi Halfa, Sudan: an evolutionary perspect-
ive. Journal of the Dental Association of South Africa 43:
539–541.
Turner CGII. 1984. Advances in the dental search for
Native American origins. Acta Anthropogenetica 8: 23–
78.
Turner CGII. 1985a. The dental search for Native
American origins. In Out of Asia, Kirk R, Szathmary
E (eds). Journal of Pacific History: Canberra; 31–78.
Turner CGII. 1985b. Expression count: a method for
calculating morphological dental trait frequencies
by using adjustable weighting coefficients. American
Journal of Physical Anthropology 68: 263–268.
Turner CGII. 1987. Late Pleistocene and Holocene
population history of East Asia based on dental
variation. American Journal of Physical Anthropology
73: 305–322.
Turner CGII. 1990. Major features of Sundadonty and
Sinodonty, including suggestions about East Asian
microevolution, population history, and late Pleis-
tocene relationships with Australian Aboriginals.
American Journal of Physical Anthropology 82: 295–
318.
Turner CGII. 1992. The dental bridge between
Australia and Asia: following Macintosh into
the East Asian hearth of humanity. Perspectives
of Human Biology 2/Archaeologica Oceania 27: 120–127.
Turner CGII, Markowitz M. 1990. Dental disconti-
nuity between late Pleistocene and recent Nubians.
I. Peopling of the Eurafrican–South Asian triangle.
Homo 41: 42–53.
Turner CGII, Scott GR. 1977. Dentition of Easter
Islanders. In Orofacial Growth and Development, Dahl-
berg AA, Graber TM (eds). Mouton: The Hague;
229–249.
Turner CGII, Nichol CR, Scott GR. 1991. Scoring
procedures for key morphological traits of the
permanent dentition: the Arizona State University
dental anthropology system. In Advances in Dental
Anthropology, Kelley MA, Larsen CS (eds). Wiley-
Liss: New York; 13–32.
Williams MAJ, Adamson DA, Abdulla HH. 1982.
Landforms and soils of the Gezira: a Quaternary
legacy of the Blue and White Nile rivers. In A Land
Between Two Niles: Quaternary, Geology and Biology of the
Central Sudan, Williams MAJ, Adamson DA (eds).
A.A. Balkema: Rotterdam; 111–142.
Copyright #2006 John Wiley & Sons, Ltd. Int. J. Osteoarchaeol. (in press)
Ancient Inhabitants of Jebel Moya