вход по аккаунту


Metrical reconsideration of the Skhul IV and IX and Border Cave 1 crania in the context of modern human origins.

код для вставкиСкачать
Metrical Reconsideration of the Skhul IV and IX and Border Cave
1 Crania in the Context of Modern Human Origins
Department of Anthropology, Southern Illinois University, Carbondale,
Illinois 62901-4502
Skhul, Crania, Neandertals, Border Cave
The “out-of-Africa”models for origins of modern Homo supiens incorporate Skhul as one site documenting that early origination. However, only Skhul V is usually considered in the comparative craniology of the
question, neglecting the other substantial crania, Skhul IV and IX. Craniometric comparison demonstrates that IV and IX amplify the picture of continuous gradations of Neandertal-to-modern variations throughout the Levant;
much variation is thus represented within this one site, raising serious questions about Neandertals and moderns being discrete and long-separated species. Qafzeh 6 too is craniophenetically closer to Neandertals than to the true
anatomically modern people of the European Upper Paleolithic. Proper distance analysis of Border Cave 1cranium shows it is actually far removed from
modern African populations. References t o Qafzeh, Skhul, and Border Cave as
“fully anatomically modern’’ require reconsideration.
Many ideas go around more than once in
paleoanthropology. The idea has re-emerged
that anatomically modern Homo supiens
(hereafter AMHS) originated very early, and
long coexisted as a distinct species alongside
Neandertals. This idea, long identified as
the “presapiens” theory of modern human
origins (Howell, 1957; Vallois, 1958), has
been around a long time and has come and
gone in anthropological fashion through this
century. The current surge of interest in a
non-European extension of the idea (e.g.,
Stringer and Andrews, 1988; Mellars, 1988;
Delson, 1988) has not been given the same
name, but it follows the same idea. The
idea’s new proponents have not always acknowledged and detailed the parallels and
differences with the historical schools. As
reviewed by Smith et al. (1989), this “out of
Africa” model postulates an early evolution
of anatomical modernity in south or central
Africa, while archaic or Neandertal humans
existed elsewhere. While one version of the
model admits t o possible gradual change, intermediate forms, and interbreeding with
Neandertals upon contact, the stricter view
denies gradualism or interbreeding in keeping with the presapiens legacy.
Found between 1940 and 1942, the fragmentary Border Cave 1 cranium has some
modern features and derives from a site
with some Middle Paleolithic deposits. However, that the specimen’s provenance is with
the early deposits can never be certain (Day,
1986). Historically, Rightmire’s (1979) analyses of the Border Cave cranium, and ensuing comments, were instrumental in advancing the notion of very early AMHS from
South Africa. Yet, several of those same (and
later) commentators also made the point
that the fossil was interpolated into a discriminant space defined by 8 modern African population samples, and Border Cave’s
position among those samples can only be
Received November 30,1990; accepted November 2,1991
validly interpreted if it indeed intrinsically
belongs to one of them. This is hardly likely.
In fact, such a research design maximizes
the chance of Border Cave’s appearing t o
conform to the multivariate confidence limits of some of the modern samples, since
their characteristics define the discriminant
space. Subsequent statistical consideration
has been provided by Campbell (19841, who
alludes t o the possibility of Border Cave being statistically atypical for all the living
groups, and by Ambergen and Schaafsma
(19841, who state “Border Cave is certainly
not a random drawing from any of the populations involved in Rightmire’s comparison.
Rightmire (1981)concedes that Border Cave
is significantly different from all the reference samples but still particularly like a
“large Bushman-like” population. Van Vark
et al. (1989:43) conclude from a more extensive treatment that “the Border Cave 1cranium does not fit in with any of the populations with which it was compared. . . . It is
not definitely more like a recent African
population than it is like other recent populations . . . [therefore] the evidence obtained
suggests that it is less obvious t o call it ‘anatomically modern’ as is usually done.”
In eyeballing the univariate statistics provided by Rightmire (1979), one sees that
Border Cave stands apart in its glabellar
protrusion, orbital breadth, supraorbital
projection, frontal subtense, and biorbital
chord, all of which by reference to archaic H .
sapiens make Border Cave seem more primitive. At the same time the fossil is uniquely
small in nasion angle and subtense, suggesting the combination of outlying large and
small variables will construe this form as
very different in both size and shape.
less complete, and in fact V itself required
considerable midface reconstruction.
Has the undue attention to V somehow
grown from its more modern appearance
among an actually quite variable population? McCown and Keith (1939) chose V as a
sort of type specimen because it was the tallest individual and (only in the sense of the
entire skeleton) most complete, and also because it was very distinct from Neandertals
thus “foreshadowing” the Cro-Magnons.
However, IV was said to be the “most representative of the Skhul type of ancient Palestinian.” They documented the sharp occipital angulation and the long flat face of IV in
contrast to V. In Skhul M, Neandertal features predominate over “Neanthropic” ones
(unlike IV) according to McCown and Keith,
and IX is the most massive, prompting comparison with La Chapelle. Both IV and IX,
and other Skhul fragments, have a relatively low position of maximum breadth on
the parietals. Thus McCown and Keith
(1939) called variation within these and
other Skhul individuals “unexpectedly
great” and Howell (1957) referred to an anatomical “range of variation which overlapped that of early Neandertal peoples” in
Skhul as a whole. These assessments (and
the measurements I have borrowed from
them) were based on crania reconstructed
by McCown and Keith to be as close as
manageable to anatomical truth; although a
new reconstruction is under way on Skhul
IV (P.Andrews, pers. commun.), it is not yet
possible to assimilate it or the contrasts it
might provide to McCown and Keith’s version.
For McCown and Keith the great variation within Skhul material indicated the
“throes of change” of a species near the diTHE FORGOTTEN SKHUL CRANIA
chotomy between Neandertal and modern
One tendency in this Neandertal-AMHS lineages. At that time there seemed to be no
perspective has been to use Skhul V to char- evidence for other than a “homogeneousdeacterize early AMHS in the Middle East posit’’ signifying a single population in the
(along with Qafzeh crania), while paying Skhul burial level. A few workers have subscant comparative attention to the other re- sequently been alive to the implications of
mains from Skhul. This point is documented Skhul crania other than V in formulations
by Corruccini (1974:lOO)in a comprehensive about modern human origins (Coon, 1962;
calvariometric study of fossil hominids. The Brose and Wolpoff, 1971; Wolpoff and Caspoint bears repeating today. The crania pari, 1990; Kidder et al., in press). In the
numbers IV and IX from Skhul are scarcely current context, it is worthwhile examining
the revived presapiens ideas in terms of
variation present within Skhul alone.
shape) variables. What is important is the
difference in pattern.
Border cave
Skhul crania
Let us take a different and more appropriate approach to assessment of Border Cave
by using it as the mean of one of 9 reference
samples in Rightmire’s (1979:33) data, and
performing distance analyses on the 11provided measurements: glabella protrusion,
supraorbital projection, biorbital chord, nasion subtense, nasion angle, malar height,
frontal chord, frontal subtense, frontal angle, orbital breadth, and mastoid length. To
satisfy various proponents of the virtues of
various methods, distances will be of the following varieties:
1. Simple (Pythagorean or “Sakal’s’’d).
2. Shape vector: Mosimann’s suggestion
that measurements be divided through by
the geometric mean so each case is converted to the same mean size. Refer to Corruccini (1987).
3. Generalized (Mahalonobis’ covariance
inverted) distance. The Generalized D is calculated using Goodman’s (1972) shortcut,
essentially the painvise Pythagorean distance between cases’ principal component
positions inversely weighted by each component’s eigenvalue.
I emphasize that in calculating and comparing both the raw (size) d and the shape
vector d, this does not entail “throwing out”
the size information but merely attempting
to separate it. Darroch and Mosimann
(1985) demonstrate that different sorts of
shape vectors all decompose into one another by a constant factor, and that total
discrimination equals shape difference plus
a constantly scaled version of the difference
in the size variable (the geometric mean
mentioned above). Therefore if shape distances or discrimination eigenvalues are
smaller for shape variables than for original
raw (total) variables, the difference in size
variables is significant. The smaller quantities do not denigrate the shape statistics for
they can never exceed the corresponding
statistics generated from raw tie., size plus
Similarly, I have here undertaken a new
analysis of sizehhape patterning in later
Pleistocene hominid crania, with focus on
how Skhul IV and IX reflect on the affinities
assumed for Skhul V. The 27 craniometric
variables used (listed with Fig. 4) are just
those measurements that can be reliably
taken or have been published on all three of
these Skhul crania, especially in McCown
and Keith (1939) but supplemented by other
sources including casts calibrated for expansion or shrinkage. For full details on sources
of data (including my own measurements),
error, and other matters, see Corruccini
(1974). The methods again include simple
distances based on raw dimensions (“size”),
distances following Mosimann-style shape
vector alteration of the raw dimensions, and
Generalized distances individually calculated from inversely weighted principal components following Goodman’sprocedure.
In this analysis, missing values crop up
for some important crania. Therefore the
shape vectors have been calculated after
first normalizing variables (columns) by dividing through by the mean, then figuring
shape variables by dividing cases (rows)
through by the geometric mean. This prevents a missing trait which has an unusually large or small mean value from unduly
biasing the geometric mean. The d’s are calculated as the root mean square of difference
between measurements present in both of
each compared pair of crania. For the Generalized D, missing values had to be substituted from the preceding cranium (in terms
of rank order similarity according to the
simple d) having that measurement, prior to
the principal component computations.
The comparative samples are very crudely
composed, owing to the varying definitions
and classifications. First, there is a broadly
defined Neandertal grouping (Western European Classic, Middle Eastern, and earlier
Pre- or Progressive Neandertals). This consists of La Chapelle, La Ferrassie 1, Le
Moustier (a late juvenile), Spy I and 11, La
Quina adult, Neandertal, Circeo 1,Gibraltar
(Forbes Quarry), Teshik-Tash (although an
early juvenile), Steinheim, Ehringsdorf, Saccopastore 1 and 2 , Krapina C, Shanidar 1,
Tabun I, Amud 1,and Zuttiyeh.
The putative early African and Levant
collection of precocious anatomical moderns
is defined by crania thus allocated by at
least one proponent in the review by Smith
et al. (1989),or similarly indicated as “transitional” from archaic to modern morphology, and for which there exist or can be
found adequate data. Apart from the Skhul
material, this group includes Florisbad,
. Qafzeh 6, Omo 2 , and Djebel Ighoud 1and 2 .
The grouping is justified post-hoc since
these specimens form a primary cluster
when crania are considered individually
(see Results). For want of a more agreedupon term, I shall refer to these, by analogy,
as the “presapiens”to distinguish them from
the established AMHS below.
True AMHS, that is, Upper Paleolithic
Eastern European crania (Predmostl, 3, 4,
7, 9, and 10, Oberkassel 1 and 2 , Mladec
(Lautsch) 1, Brno 2 , Brux, Dolni Vestonice 3)
constitute my comparative sample of true,
undisputed early AMHS. Restriction of the
sample to the non-Western European Upper
Paleolithic crania gives the broadest possible benefit of the doubt to AMHS affinities of
earlier specimens, for the Eastern grouping,
in part, has frequently been referred to earlier dates and a less modernized transitional
morphology (cf. Corruccini, 1974).
8 Modern African Samples
s.d.= 0.82 within
Fig. 1. Multivariate dispersion among male and female means of 8 African samples from 5 populations for
11 craniometric variables provided by Rightmire
(1979:33). The Mahalonobis D is calculated from
weighted principal components following Goodman’s
(1971) PCA-based formula; Border Cave 1 is treated as
the mean of a ninth sample so it can express its own
uniqueness, impossible when it is just interpolated into
the discriminant space of the 8 living samples. The circles around Border Cave represent one standard deviation and (stippled area) one standard error for the Generalized Distances from the 8 modern samples. The
circles around the modern Africans represent one standard deviation and standard error of the Ds among
those samples.
most four times as dissimilar from living Africans than the average dispersion within
those diverse Africans, and almost 8 times
as distant as its standard deviation of pairwise distance to the Africans. In cluster
analysis, according to any of the different
methods, all modern populations cluster together at a final distance level half as large
as their inclusive distance to Border Cave.
Even with univariate analysis, differences
are clear (pointing up the old truism that
understanding the univariate differences is
Border cave
a necessary first step to the multivariate difResults concerning Border Cave are very ferences). The size variable for Border Cave
consistent with each of the 3 distance ap- is 2.34 standard deviations larger than for
proaches employed (size d, shape d, and Ma- living Africans (P < .02);the most contrasthalonobis D all have cophenetic r = +0.96) ing shape variable is supraorbital projecbecause gross size differences are not large. tion, 8.57 standard deviations larger in BorThat is, a difference in a variable will cause der Cave. The Border Cave 1 specimen thus
both a size and a shape difference. Border constitutes a classic statistical outlier to the
Cave (Fig. 1:the Generalized version of dis- distributions among modern Africans.
tance is given here) falls far outside the morSkhul crania
phometric “envelope”of all modern African
populations, so contrary to Rightmire’s
The dendrogram in Figure 2 is a weighted
analysis underlying the current presapiens pair-group phenogram, clustering together
enthusiasm, the fossil in no wise can be specimens at increasing levels of shape disequated with those particular living African tance (nearest neighbors are clustered, then
humans. The probability for the null hy- their mutual distances to remaining groups
pothesis is off the scale; Border Cave is al- averaged). The similarity between the raw
43 7
Fig. 2. Pairwise cluster analysis of simple shape distances using the Weighted Pair-group Method.
The 12 Eastern European Upper Paleolithic crania at the bottom all cluster together in a dendritic
configuration not shown, between d values of 0.61 and 1.34. The d is an average, and must be multiplied
by 27 (the number of variables) to be analogous to Mahalonobis D.
size and the shape-modified distances is
marked, but the latter produces clusters
more recognizable within existing frameworks of relationships and similarity (the
other, size-based phenogram will be supplied upon request). At the top of Figure 2
Classic (Western European Wurm) Neandertals cluster, then join a group of 5 Middle
Eastern and other less classic (perhaps “progressive”?)Neandertals. Skhul V joins some
of the earlier and more generalized of European specimens (Steinheim, Ehringsdorf)
and then the broadly constituted Neandertal cluster.
Skhul IV and IX, meanwhile, although
sharing a relatively small d from V, join the
cluster of supposedly early African/
Mideastern AMHS (my “presapiens”), and
then the above-described, increasingly heterogeneous Neandertal grouping. The Upper Paleolithic true AMHS exclusively cluster with one another, relatively far removed
from these other groupings. Thus, the picture is one of overriding affinity among all
the crania earlier than the European Upper
Paleolithic, whether they be considered
Classic Neandertal, Progressive Neandertal, AMHS, presapiens, or whatever.
Neanderlals 4
IV + I X
- **
, 7
v to IV+ I X
1 - 1
Fig. 3. Top: Dispersion of samples along the major
principal component calculated from raw craniometric
values and, judging from uniformly positive eigenvector
loadings (except for parietal sagittal arc and chord), representing general size increasing to the right. Component one represents 35.5% of total variance. Middle:
The major component based on shape variables, representing 37.0% of total variance. Bottom: Mean and
range of pairwise D values from Skhul V and from IVplus-M to the other samples.
Fig. 4. Graph of the variables with at least a correlation of k0.60 with the first or second PCA axes, based on
the shape variables. The major axis separates the Neandertal and AMHS samples maximally; a relatively small
upper facial height (and other features) combined with
relatively high calvaria distinguishes the AMHS, for example. The second (vertical) axis (25.6% variance) primarily acts t o separate the early “presapiens” grouping
from both Neandertals and AMHS. The craniometric
variables: 1,maximum length; 2, nasion-basion; 3, maximum breadth; 4, least frontal diameter; 5 , maximum
frontal diameter; 6, basibregmatic height; 7, auricular
height; 8, auriculo-bregmatic arc; 9, total cranial sagittal arc; 10, frontal arc; 11, frontal chord; 12, parietal arc;
13, parietal chord; 14, occipital arc; 15, occipital chord;
16, cranial capacity; 17, basion-prosthion; 18, superior
facial breadth; 19, inner biorbital breadth; 20, bizygomatic breadth; 21, bimaxillary breadth 22, upper facial
height; 23, interorbital breadth 24, orbital width; 25,
orbital height; 26, nasal width; 27, nasal height.
The spatial picture of relationships is expressed as the major principal component
mean (with one standard deviation) for raw sive forms could be intermediate and ancessize and shape analyses (Fig. 3). In the tral to both the divergent lineages of
former (top),the unrealistically variable na- Neandertals and AMHS (this is the Preneture of “Neandertals,” broadly defined, be- andertal theory); the “presapiens” assemcomes evident; this variation is strongly blage of crania, plus Skhul, could better be
time-progressive from left to right. That is, called preneandertal since they are metriSteinheim, Saccopastore, and other pre- cally closer to Neandertals than to true (UpWurm specimens are to the left and the lat- per Paleolithic) AMHS.
Looking at the pairwise distances beest and neurocranially largest Classic Neandertals t o the right of the mean. The sample tween crania, Skhul IX is nearer Shanidar,
construed to represent early non-European Zuttiyeh, La Quina, and Spy 1 and 2 than is
AMHS is nearer to Neandertals than to the Skhul V. Both IV and IX are nearer Amud,
true AMHS of the Upper Paleolithic. Skhul Ehringsdorf, and the original Neandertal
crania IV and M, in contrast to the cele- than is V. Skhul IX has a smaller distance to
brated V, are more within the earlier nean- Amud, Zuttiyeh, and Neandertal than it has
dertaloid than the later, modern range.
to Skhul V. Paradoxically, and typifying the
In the shape major axis, similarities con- mosaic of cranial similarities amongst Upform to the older idea that earlier progres- per Pleistocene fossils, Skhul IX (but not IV)
TABLE 1. Linear generalized distances (D)among Skhul crania and other aggregate samples defined both by
attributions in the literature and by clustering during the present investigation: “presapiens” (Qafzeh, Omo,
Ighoud, Florisbad), Neandertals (broad sense), and AMHS (Eastern European Upper Paleolithic only)’
Raw D
Skhul V
Skhul IV,V,IX
Covariance standardized
Skhul V
Skhul IV,V,IX
‘Both the raw D and a forcibly covariance-homogenized D (setting unit variance around the mean D) are presented
is also closer than is Skhul V to several of
the Predmost specimens.
The Generalized Distance provides a different approach from shape vectors to neutralizing covariance (which is strongly associated with the major size axis of covariance).
As before, Skhul V actually shows strongly
generalized rather than clearly AMHS affinity but is slightly more similar to Neandertals than to the Upper Paleolithic sample
(Fig. 3, bottom).
The presapiens, Neandertal, and AMHS
aggregates form a roughly triangular arrangement of averaged individual Generalized Distances. Skhul V is very nearly the
center of this triangle. Let us combine the
three Skhul crania with the “presapiens”
sample as is their standard assignment, to
quantify this triangular distance. For presapiens and Neandertals D = 2.34, presapiens-AMHS D = 2.61, and NeandertalAMHS D = 2.65. Table 1gives distances for
Skhul V alone, then for all three Skhul crania combined for the perspective given to
relationships when V is considered in isolation. The 3 Skhul crania combined are relatively much less similar to the Upper Paleolithic group. Skhul IV plus M on average are
further removed from both Neandertals and
AMHS than is the very neutral Skhul V, yet
they are much more variable in these distances.
The trait-by-trait pattern of diversity
within Skhul is tracked in Table 2, which
gives the measurements and shape vector
values wherein Skhul IV and/or M are intermediate t o or approximate Neandertals
rather than their stratum-mate Skhul V.
Skhul V has a relatively shorter (hence relatively rounder), more flexed, frontally narrowed, higher (in the sagittal aspect) yet less
inflated (in the coronal aspect) cranium
TABLE 2. Selected craniometric values in mm and
corresponding shape variables (column-standardized
shape vector values centered on 1.00) compared among
Skhul V, IV, and IX, and the mean of the more
complete and classic Neandertals (Tabun, Shanidar,
Amud, La Chapelle, La Ferrassie, Le Moustier, Circeo,
LaQuina, Spy I and II, Neanderthal)
Max. length
Max. frontal B 114
Basion-bregma 129
Auric.-Breg. arc 305
Frontal arc
Orbit width
Orbit height
Nasal width
than have Skhul IV, IX, and Neandertals.
The frontal arc is shorter due to less glabellar protrusion rather than a less rounded
forehead; the difference between the frontal
arc and chord reveals a clearer progression
from Skhul V (12 mm) to IV (14 mm) to IX
(16 mm) to Neandertals (18.5 mm). In the
well-known orbital shape and nasal width
traits, crania IV and IX again are intermediate or closer to Neandertals.
Among other traits, only Skhul IX has the
sagittal parietal flatness of Neandertals,
and only IV their vertically longer face.
Skhul IV and IX are different from the other
fossils considered here in interorbital wideness, possibly due to inadequate adjustment
in McCown and Keith’s (1939) reconstruction estimates for postmortem distortion.
Ehringsdorf I ---------
--I Amud
- 1 La Ferrassie
Qafzeh I-----
----- I
------------ I
Predmost 3
8Ep3EA@E Fa@Aa& Bm&h
Predmost 1
---I Amud
Ferrassie I--Qafzeh I---I Prectmost 3
Fig. 5. Values (each dash represents ,011 above and below the grand mean of 1.0 for shape variables
compared for selected diverse “Neandertals”(Ehringsdorf, Amud, La Ferrassie) against Qafzeh 6 and two
selected true AMHS, Brno and Predmost 1 (Predmost 3 substituted in case of missing value).
Is Qafzeh 6 anatomically modern?
As a digression it is informative to look
also at the Qafzeh 6 cranium which is invariably joined with Skhul V as a Levantine representative of early arriving African AMHS.
Like Skhul V, Qafzeh crania are all too often
blithely referred to as anatomically modern
or “fully modern” without supporting argumentation (e.g., Schwarcz et al., 1988). Average shape d from Qafzeh 6 is 1.44 k 0.22 t o
Neandertals and 1.46 0.30 to Upper Paleolithic crania in the present study. This is
not mere coefficient trickery, for the raw
measurement d and Generalized D give ex-
actly the same result, a virtually equal distance of Qafzeh 6 to Neandertals or to true
AMHS (Fig. 5). Thus Qafzeh shares with
Skhul V a very central craniophenetic position vis-a-vis Neandertals (in the broad
sense) versus these Eastern European crania of the Upper Paleolithic. Corruccini
(1974) documents more thoroughly that
Qafzeh and Skhul are even nearer the broad
range of neandertal or archaic H. sapiens
crania than they are to modern human varieties o r to the earliest Western European
AMHS. When including more facial measurements (Corruccini, 1976) in a sample of
more complete hominids, Skhul V becomes a
shade closer to Upper Paleolithic hominids
rather than, as here, a shade closer to Neandertals.
drews, 19881, the Skhul population should
represent a distinct species from Neandertals (whether the latter are broadly or narrowly defined) that is indistinguishable
from modern humans (unless Neandertal
gene flow is incorporated). Taken altogether, however, the results above raise perplexing questions about the interpretation
of two supposedly long-separated species.
The variation within Skhul alone spans
much of the phenetic spectrum between
archaidneandertal and modern cranial form,
while the remainder of specimens considered here fall not into two well-defined clusters but three variable and indistinct ones.
Such results are hardly unprecedentedStringer himself earlier (1974) showed in a
similar craniometric study that Skhul V is
more than twice as close as Upper PaleoDISCUSSION
lithic samples are to Petralona, and in many
The Border Cave cranium, so central to ways follows a neandertal pattern of disthe course of “out-of-Africa”thinking despite tances. Kidder et al. (in press) demonstrate
its uncertain age, can support no special re- Qafzeh 6 as well as Skhul IV and V are well
lationship to living African Homo sapiens. separated from later European AMHS. This
This does not rule out a closer relation to calls into question blithe assumptions that
AMHS than to Neandertals or archaic Homo Skhul and Qafzeh are, cranially, anatomical
sapiens in general, but it does contradict moderns. On the other hand, Trinkaus
much of the specific language encountered (1976, 1986) asserts that cranial and postin support of the punctuated, African emer- cranial evolution were somewhat uncoupled
gence of the earliest anatomical moderns. during those times and the respective postVan Vark et al. (1989) corroborate that Bor- crania do fit a modern pattern in Skhul and
der Cave does not particularly resemble any Qafzeh.
Currently there seem to be similar
modern human population, is less similar to
Africans than to other contemporary people, thoughts emerging. Arensburg (1991) deand is further from both Upper Paleolithic nies that any of the Levantine specimens are
and modern AMHS than those two groups true Neandertals, and Trinkaus (1991) supports the reasoning that only very late, isoare from each other.
The considerable diversity within Skhul lated Western European specimens fit the
points up the problem with puzzling levels of Neandertal stereotype; both authors refer to
variation throughout Middle Eastern sites clinal distributions of characters. Arensburg
and indeed throughout the geographic and (1991) reports “the morphological variabiltime spans involved in the Neandertal prob- ity within sites (e.g., Qafzeh VI vs. Qafzeh
lem. Although the “out-of-Africa”idea is one M, Tabun C vs. Tabun B, Skhul V vs. Skhul
reductionist attempt to simplify this diver- M) makes a separation between Neandersity, it is phenetically an unrealistic division tals and Homo sapiens in these localities iminto two species according to these cranio- possible and demonstrates a great range of
metric data. This does not necessarily imply variation within a unique population”. Vanthat qualitative as opposed to quantitative dermeersch (1991) seemingly agrees, statmorphological traits might not demonstrate ing the “link to the neandertals is not evia different pattern, unifying the Skhul spec- dent” in characteristics of Tabun 2 in
contrast to Tabun 1. Rak (1991) refers to
According to the new sudden-replacement Shanidar 11-IV as “less differentiated Nemodel of AMHS origins (Stringer and An- andertals coexisting with the “classic”
Shanidar I and V. Wolpoff and Caspari
(1990) and Kidder et al. (in press), in addition to confirmatory reference to withinSkhul variations, find Qafzeh 3 more archaic than 6 and 9. Simmons (1991) finds
major aspects of similarity between the Ighoud and Qafzeh frontal bones.
For purposes of general discussion, let me
cite from amongst the plethora of current
writings the following reviews of the new
models for modern human origins: Smith
et al. (1989),Trinkaus (1986), Smith (19841,
Stringer and Andrews (1988), Spencer
(1984), Wolpoff et al. (19841, Mellars (19881,
Clark and Lindley (1989)) and in a popular
vein, Gould (1988a,b). In general, the reasoning for the stricter “out-of-Africa”model
is as follows:
1. New mitochondrial DNA evidence
points to a single African common ancestry
of great antiquity for living humans. The
fallacy of treating the mitochondrial evidence as specifically relating to AMHS has
been pointed out from the start (Cann et al.,
1987) and repeatedly, since there is no relation between selectively neutral DNA and
morphology; nevertheless, investigators as
well as popular writers are increasingly falling into the habit of referring to the mtDNA
claims as evidence in support of the timing
and geographic source of anatomical modernity. Furthermore, other molecular interpretations and phylogenies (see Spuhler,
1989, for review) allow other geographic origins and timings to be considered about
equally probable. Severe problems have recently arisen concerning the mtDNA clock;
Ward (1991) and Valencia and Ward (1991)
find mtDNA too diverse within populations
to provide credible dates between populations. They find mate separations of
50-75,000 yr and Asian subpopulations separated by 110-140,000 yr, weakening my
confidence in the calibration of lineage antiquity.
2. Fossils now exist of very early AMHS in
South Africa. As with earlier periodic popularizations of the idea, the new vigor of this
presapiens thinking relies too heavily on
fragmentary remains of uncertain provenience or dating. These are now Klasies
River Mouth and Border Cave fossils; previ-
ously they have been Fontechevade, Swanscombe, and even Piltdown in the earlier
conceptualizations of the presapiens fossil
lineage. The highlighting of extraordinarily
poorly preserved or questionably contextualized remains thus appears to be a centurylong tendency in theories of this nature. Another widely noted paradox concerns the
later Levantine presapiens (Qafzeh and
Skhul) carrying quite typical Mousterian
(i.e., Neandertal) industries. Clark and
Lindly (1989) demonstrate that Levantine
stone tools fit the model of a nexus of transitional variability to moderns. Somehow replacement of Western Asian Neandertals by
African presapiens did not incur or result
from technological replacement.
3. Intermediary forms between Neandertals and AMHS are lacking, whereas a sudden appearance of AMHS is consistent with
current thinking about punctuated equilibria. According to that thinking the earliest
AMHS, perhaps 200,000 yr in age, should be
“essentially us’’ (Gould, 1988b) immediately
as they appear, rather than gradually approaching the human tendency. Several
hitches in this reasoning have been frequently pointed out and can be briefly reiterated here. First, complete (hence interpretable) and context-rich later Pleistocene
remains (up to c. 40,000 BPI seemingly always contain appreciable Neandertal-like
traits (including Qafzeh and Skhul). The
earliest putative AMHS (Klasies and Border
Cave) are the most fragmentary (Wolpoff
and Caspari, 1990). The pattern is nonrandom enough to require an explanation that
sparsely represented presapiens populations were of lower density, andor not practicing interment (whereas the Neandertals
probably were), hardly consistent with the
former’s alleged technological superiority or
behavioral and psychological equivalence to
modern humans.
Why did artwork begin merely 20,000 yr
BP when AMHS have been just like us for
200,000 yr? And why does the Epipaleolithic
dwarf point to social support (Gould,
1988a,b) more than do the elderly of Shanidar and La Chapelle? Marshack (1989) argues that “the capacity for creating different
types of symbolic marking on both stone and
Fig. 6 . An impressionistic picture synthesizing location, time, and morphology in Upper Pleistocene crania illustrating interaction among evident geo-,chrono-,
and morpho-clines. This is consonant with a generalization of the “Preneandertal/Progessive Neandertal”
model. For further supporting data see Corruccini
(1974). There is a cline of increasing Classic Neandertal
specializationin the western “Wurmian”zone, a reverse
cline in primitive holdover traits among earliest AMHS,
and a particularly marked resultant cline in the morphological gap and suddenness of change between latest
Neandertal (broad sense) and earliest appearing Upper
Paleolithic crania.
bone (and with ochre perhaps on the body)
was clearly present before the Aurignacians. . . . A number of researchers have indicated that the Neandertals did in fact
have conceptual models and maps as well as
problem-solving capacities comparable to, if
not equal to, those found among anatomically modern humans.”
In some ways the new competing “out-ofAfrica” versus multiregional continuity hypotheses are geographic generalizations and
recirculations of the old presapiens versus
Neandertal theories of human origins, respectively (e.g., Spencer, 1984). What are
the prospects of geographically generalizing
and retreading the old compromise “Preneandertal” theory (Howell, 1951, 1957) as a
preferable alternative to the two newer
models above, that are increasingly being
mentioned as the two sole alternatives? The
still too Eurocentric idea would be that Progressive Neandertal populations are the
only confidently known predecessors both of
AMHS and of glacially isolated extreme
Western European Neandertals of the later
Wurmian time, and that anatomical modernity evolved in hominids of the general tricontinental crossroads of North Africa,
Western Asia, and Southeastern Europe
prior to 30,000 yr BP-again an old idea
(Howell, 1952; Birdsell, 1963) but one I believe deserves much more circulation among
the other revivals currently provoking debate (See Fig. 6). A revived Preneandertal
formulation would incorporate a gradualistic version of the single origin model but
deny its earliness and exclusively South African Eden.
Howells (1989) has issued the furthestranging multivariate cranial study of modern humanity. His results share many similarities with what I have here presented. He
finds that “such marginal anatomical moderns as Irhoud 1or Skhul5 ally themselves
consistently, if not closely, with Neanderthals rather than with moderns.” Further-
more, concerning the idea of gradual multiregional change Howells shows that among
Neandertals “no real indication appears of a
closer link to one living population group
(e.g., Europeans) than t o others” and that
“modern populations taken all together are
indeed limited in the geographical variation
of both shape and size of the cranium. . . .
Some early anatomically modern specimens
may tend to fall a little out of this range. The
impression, true or not, is that the modern
unity is fairly recent.” Howells’ results,
based on unprecedentedly good samples,
numbers of measures, and sound methodology, indicate no support for the multiregional continuity idea resurrected and
adapted from Coon (1962), or the early African origination of current human varieties
suggested by mtDNA. Remember that Coon
was ostracized from the bioanthropological
community for the racist implications of his
model of deeply-rooted human races (Birdsell, 1963). Kamminga and Wright’s (1988)
and van Vark and Dijkema’s (1989) recent
craniometric reconsiderations of Zhoukoudian Upper Cave hominids are also very important in this regard. Using what I regard
as convincing methodology, they show that
there is no “mongoloid quality whatsoever
to those hominids (but some vague similarity t o Ainu and Australian aboriginals), denying the great antiquity of Asian races.
Howells (1989) also rightly denies
Stringer and Andrews’ (1988) proposition
that greatest modern population differences
will be between Africans and all others:
“there is no sign that, cranially, Africans are
set off as most distant from other populations-quite the reverse”; there is “no support for a sub-Saharan first source for anatomical moderns.” This confirms that fossil
regional skull shape differences resembling
the modern distribution appear only near
the Holocene boundary (Washburn, 1944).
In summary, the substantial craniometric
diversity within the Skhul remains parallels
the general problem of gradational variation
among Middle Eastern and other Upper
Pleistocene sites. Results presented above
do not fit the reductionist “out-of-Africa”
model and raise anew the perpetual difficulties with the presapiens interpretation of
Neandertals (broad sense) and AMHS as
two long-separate species. Skhul crania as a
group are, along with Qafzeh 6, quite phenetically intermediate between Upper Paleolithic samples and neandertals. Continued
facile reference t o Skhul and Qafzeh as craniometrically “fully anatomically modern” is
not responsible to the craniometric data.
The old McCownKeith idea of Mount Carme1 variation signaling a species in the
throes of evolution may lack any neontological theoretical backing. but if we cannot expect evolution to follow uniformitarian principles then this idea fits the current data at
least as well as the “out-of-Africa”model.
I thank G.M. van Vark, Fred Smith, M.
Wolpoff, D. Frayer and P. Andrews for reprints and suggestions.
Ambergen AW, and Schaafsma W (1984) Interval estimates for posterior probabilities, applications to Border Cave. In GN van Vark and WW Howells (eds.)
Multivariate Statistical Methods in Physical Anthropology. Dordrecht: Reidel, pp. 115-134.
Arensburg B (1991) From sapiens t o neandertals: rethinking the Middle East (abstract). Am. J. Phys. Anthropol. [Suppl.] 12:44.
Birdsell J B (1963) The origin of human races. Q. Rev.
Biol. 38:17%185.
Brose DS, and Wolpoff MH (1971) Early Upper Paleolithic man and late Middle Paleolithic tools. Am. Anthropol. 73:1156-1194.
Campbell NA (1984)Some aspects of discrimination and
allocation. In GN van Vark and WW Howells (eds.):
Multivariate Statistical Methods in Physical Anthropology. Dordrecht: Reidel, pp. 177-192.
Cann R, Stoneking M, and Wilson A (1987) Mitochondrial DNA and human evolution. Nature, 325:31-37.
Clark GA, and Lindly JM (1989) Modern human origins
in the Levant and Western Asia: the fossil and archeological evidence. Am. Anthropol., 91:962-985.
Coon CS (1962) The Origin of Races. New York: Knopf.
Corruccini RS (1974) Calvarial shape relationships between fossil hominids. Yearb. Phys. Anthropol. 18:89109.
Corruccini RS (1976) The interaction between neurocranial and facial shape in hominid evolution, Homo
Corruccini RS (1987) Shape in morphometrics: comparative analyses. Am. J . Phys. Anthropol. 73:289-303.
Darroch JN, and Mosimann J E (1985) Canonical and
principal components of shape. Biometrika 72:241252,
Day MH (1986) Guide to Fossil Man, 4th Ed. Chicago:
University of Chicago.
Delson E (1988) One source not many. Nature 332206.
Goodman MM (1972) Distance analysis in biology. Syst.
ZOO^. 21:174-186.
Gould SJ (1988a) Honorable men and women. Nat. Hist.
Gould SJ (198813) A novel notion of neanderthal. Nat.
Hist. 97(6):16-21.
Howell FC (1951) The place of Neanderthal man in human evolution. Am. J. Phys. Anthropol. 9:379416.
Howell FC (1952) Pleistocene glacial ecology and the
evolution of “Classic Neandertal” man. Southwest. J .
Anthropol. 8:377-410.
Howell FC (1957) The evolutionary significance of variation and varieties of “Neanderthal” man. Q. Rev.
Biol. 32t330-347.
Howells WW (1989) Skull shapes and the map: craniometric analyses in the dispersion of modern Homo.
Pap. Peabody Mus. Archaeol. Ethnol. 79.
Kamminga J, and Wright RVS (1988) The Upper Cave
at Zhoukoudian and the origins of the Mongoloids. J .
Hum. Evol. 17:739-767.
Kidder JH, Jantz RL, and Smith FH (in press) Defining
modern humans: a multivariate approach. In G
Brauer and FH Smith (eds.): Continuity or Replacement? Controversies in Homo sapiens Evolution.
Balkema: Rotterdam.
Marshack A (1989) Evolution of the human capacity: the
symbolic evidence. Yearb. Phys. Anthropol. 32:l-34.
McCown TD, and Keith A (1939) The Stone Age of
Mount Carmel. 11. The Fossil Human Remains From
the Levalloiso-Mousterian.Oxford: Clarendon.
Mellars P (1988) The origins and dispersal of modern
humans. Curr. Anthropol. 29r186-188.
Rak Y (1991) A model for morphologic and taxonomic
variation in Neandertals and early Homo sapiens (abstract). Am. J . Phys. Anthropol. [Suppl.] 12:147-148.
Rightmire GP (1979) Implications of Border Cave skeletal remains for later Pleistocene human evolution.
Curr. Anthropol. 20:23-35.
Rightmire GP (1981) More on the study of the Border
Cave remains. Curr. Anthropol. 22:199-200.
Schwarcz HP, Grun R, Vandermeersch B, Bar-Yosef 0,
Valladas H, and Tchernov E (1988) ESR dates for the
hominid burial site of Qafzeh in Israel. J. Hum. Evol.
Simmons T (1991) North African and Levantine hominid affinities: frontal squama and browridge morphometry (abstract).Am. J. Phys. Anthropol. [Suppl.]
Smith FH (1984) Fossil hominids from the upper Pleistocene of Central Europe and the origin of modern
Europeans. In FH Smith and F Spencer (eds.): The
Origins of Modern Humans: A World Survey of the
Fossil Evidence. New York: Alan R. Liss, Inc., pp.
Smith FH, Falsetti AB, and Donnelly SM (1989) Modern
human origins. Yearb. Phys. Anthropol., 32:3548.
Spencer F (1984)The Neanderthals and their Evolutionary sequence: a brief historical survey. In FH Smith
and F Spencer (eds.): The Origins of Modern Humans:
A World Survey of the Fossil Evidence. New York
Alan R. Liss, Inc., pp. 1-50.
Spuhler J N (1989) Evolution of mitochondrial DNA in
human and other organisms. Am. J. Hum. Biol.
Stringer CB (1974) A multivariate study of the Petralona Skull. J. Hum. Evol. 3r397-404.
Stringer CB, and Andrews P (1988) Genetic and fossil
evidence for the origin of modern humans. Science
Trinkaus E (1976) The morphology of European and
southwest Asian Neandertal pubic bones. Am. J.
Phys. Anthropol. 44t95-104.
Trinkaus E (1986) The Neandertals and modern human
origins. Annu. Rev. Anthropol. 15:193-218.
Trinkaus E (1991) Would the real Neandertal please
stand u p t h e search for Neandertal apomorphies
(abstract). Am. J. Phys. Anthropol. [Suppl.] 12t174175.
Valencia D, and Ward RH (1991) Using mitochondrial
sequence data to estimate evolutionary distance between mates (abstract). Am. J. Phys. Anthropol.
[Suppl.] 12:176-177.
Vallois HV (1958) La Grotte de Fontechevade. Arch.
Inst. Paleont. Hum. 29(2).
Vandermeersch B (1991) Contemporaneity of Homo sapiens and neandertals in the Near East? (abstract).
Am. J. Phys. Anthropol. [Suppl.] 12:177.
Van Vark GN, Bilsborough A, and Dijkema J (1989) A
further study of the morphological affinities of the
Border Cave 1 cranium, with special reference to the
origin of modern man. Anthropol. Prehist. 100:43-56.
Van Vark GN, and Dijkema J (1989) Some notes on the
origin of the Chinese people. Homo 39:143-148.
Ward RH (1991) On the age of our mitochondrial ancestors: evidence for deep lineages in “mongoloid populations” (abstract). Am. J. Phys. Anthropol. [Suppl.]
Washburn SL (1944) Thinlung about race. Sci. Educ.
Wolpoff MH, and Caspari R (1990) On middle
paleolithidmiddle stone age hominid taxonomy. Curr.
Anthropol. 31:394-395.
Wolpoff MH, Wu X, and Thorne AG (1984) Modern
Homo sapiens origins: a general theory of hominid
evolution involving the fossil evidence from east Asia.
In FH Smith and F Spencer (eds.) The Origins of Modern Humans: A World Survey of the Fossil Evidence.
New York Alan R. Liss, Inc., pp. 411-483.
Без категории
Размер файла
1 063 Кб
metrical, cranial, context, cave, skhul, modern, origin, border, reconsiderations, human
Пожаловаться на содержимое документа