AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 91:325-347 (1993)

Zuttiyeh Face: A View From the East

SONGY SOHN AND MILFORD H. WOLPOFF Paleoanthropology Laboratory, Department of Anthropology, University of Michigan, Ann Arbor, Michigan 48109 KEY WORDS Zuttiyeh, Paleoanthropology, Asia, Neandertal, Levant, Zhoukoudian

ABSTRACT We analyze the phylogenetic position of the frontofacial frag- ment from Zuttiyeh, Israel. This specimen is dated to the Middle Pleistocene (the latest estimate is between 250 and 350 kyr) and is associated with the Acheulo-Yabrudian, which makes it the oldest cranium from the region. It has been previously regarded as a Neandertal, and early “anatomically modern Homo sapiens,” and a generalized specimen ancestral to both. These different phylogenetic interpretations of its features have a historic basis but in our view also result from a confusion of grade and intraspecies clade as valid sources of variation. We show here that generally the differences that distin- guish Zuttiyeh from Neandertals are similarities it shares with the Zhou- koudian remains. These similarities involve a unique combination of features, and suggest the possibility of an ancestral relationship. It is less likely that Middle Pleistocene remains from Europe or sub-Saharan Africa are uniquely or significantly ancestral to Zuttiyeh. An accurate understanding of the rela- tionship between populations of eastern and western Asia is important for resolving the more general questions surrounding the position of the Upper Pleistocene Levant populations in human evolution, including (1)whether there are significantly different contemporary Mousterian populations in the Upper Pleistocene, (2) whether Neandertals are clearly intrusive in the re- gion, and (3) whether there is an early appearance of (what many workers call) “anatomically modern Homo sapiens.” The hypothesis of a recent unique African ancestry for all modern humans is disproved by our study, which shows Asia as a significant source area for at least some living populations. 0 1993 Wiley-Liss, Inc.

The critical geographic position and early evolution in the Levant, because their solu- date of the Zuttiyeh frontofacial fragment tions are central to the hypotheses about combine to make it pivotal in clarifying a modern human origins. We will examine the number of problems surrounding the evolu- question of different Late Pleistocene lin- tion of Homo in Asia. In this paper we eages in the Levant, and the confusion of present the context and associations of the grade and intraspecies clade in interpreting 1925 discovery, discuss the historic basis for these hominids-especially over the prob- the differing interpretations of its phylogeny lem of whether “non-Neandertal” necessar- (including why it was never systematically ily means “modern.”Our use of the intraspe- compared with the Zhoukoudian remains), cies clade concept is unconventional, resting and establish the best case for its ancestry on Huxley’s (1958) older definition of clades and relationships by comparing it with the as groups sharing common genetic descent. Zhoukoudian sample as well as with Middle Because we use clade to describe a regional Pleistocene humans from Europe and sub- Saharan Africa. It is our intent to examine the issues surrounding local ancestry and Received August 27, 1990; accepted November 17, 1992.

Q 1993 WILEY-LISS, INC 326 S. SOHN AND M.H. WOLPOFF aspect of intraspecies variation, monophyly and below the Acheulo-Yabrudian in the is not implied, and indeed cannot apply. Zuttiyeh cave. These were 95 * 10 and 97 * 13 kyr (from the pre-Mousterian GEOARCHAEOLOGICAL CONTEXT above), 148 * 6 from the Acheulo-Yabru- The Wadi el Amud enters the Sea of Gali- dian, and 164 * 21 kyr (from the underlying lee on its northwestern bank. A number of travertine). Following this, Copeland and rich archaeological sites from the Late Pleis- Hours (1983) proposed another date for the tocene, such as Amud, are found along the Acheulo-Yabrudian at Zuttiyeh of 140-150 gorge. Where it cuts through a low range of kyr. Based on these publications, and his limestone hills, just 200 meters upstream analysis of the Levant archaeological se- from the Mugharet el Emireh, F. Turville- quence, Bar-Yosef (1988, 1989) conserva- Petre discovered the Mugharet el Zuttiyeh tively bracketed the date of the Zuttiyeh face (the Cave of the Robbers) in the spring of to between 110 and 150 kyr by uranium se- 1925. Subsequent excavation at this site ries, the most common age estimate to be during the month of June yielded a large widely cited. However, more recently he has human craniofacial fragment which became gathered support for an even earlier date known as the Galilee, or Zuttiyeh, skull. (Bar-Yosef, 1992). TL dates from Yabrud The archaeological sequence of the Middle and ESR dates from Tabun D suggest a ter- to earlier Late Pleistocene in the Levant mination for the Acheulo-Yabrudian in ex- consists of three major complexes-the Up- cess of 200 kyr, and this is supported at least per or Late Acheulean, the Mugharan Tradi- in part by the ESR dates for the Acheulo- tion (Jelinek, 1981), and the Mousterian (re- Yabrudian from Tabun Ea-Ed: from 151 * viewed by Bar-Yosef, 1989). The Mugharan 21 to 182 * 31 kyr (early uptake) and Tradition, or the Acheulo-Yabrudian, is geo- 168 15 to 213 * 46 kyr (linear uptake). graphically limited in its range to the cen- The correlation of Tabun E, F, and G with tral and northern Levant. It can be further Isotope Stages 11-13 extends the base of the subdivided into three facies: (1)a facies with Acheulo-Yabrudian into the 400-500 kyr a high percentages of bifaces and flake age range. Bar-Yosef concludes that the Zut- scrapers (the Acheulean facies); (2) a facies tiyeh specimen is most likely bracketed be- in which scrapers are abundant and bifaces tween 250 and 350 kyr. It is the earliest rare (Yabrudian); and (3) a facies rich in Levantine cranial fragment. blades, end-scrapers, and burins (Amudian or “Pre-Aurignacian”).The Mugharan tradition ZUTTIYEH AS A LEVANTINE is followed by the Mousterian tradition, which NEANDERTAL is characterized by the Levallois technique. Zuttiyeh was uncovered at the base of the First conclusions cave’s archaeological sequence, in a shallow Keith (1927) first presented a description pit dug into the sterile sandy deposit which of the specimen, as a contribution to a vol- filled the lower part of the cave (Turville- ume on the cave edited by Turville-Petre. He Petre, 1927). Later, Garrod (1962) noted later included part of this description in the that the human specimen was under a layer 1931 edition of New Discoveries Relating to which contained an Acheulo-Yabrudian the Antiquity of Man. In this work he sug- (Mugharan) assemblage, similar to the in- gested that the specimen was a young fe- dustry of Tabun E (Bar-Yosef, 1988). This male. His age diagnosis emphasized the un- was later confirmed by the excavation of the fused frontal and sphenoidal sutures. Sex remaining brecciated deposits (Gisis and diagnosis was based on frontal size and fa- Bar-Yosef, 1974). The association of Zut- cial features. Keith regarded the cranium as tiyeh with the Mugharan Tradition shows it that of a Neandertal variant, distinguished to be archaeologically older than any of the from the European Neandertals mainly by a Mousterian associated Levant specimens. high but narrow forehead and an unusually Schwarcz et al. (1980) provided ThKJ broad supraorbital region (especially rela- dates for various stalagmitic layers, above tive to the maximum breadth of the frontal). ZUTTIYEH FACE 327 Hrdlitka examined the Zuttiyeh specimen the specimen represented an advanced Ne- in 1927, and reviewed its morphology in his andertal variety. 1930 monograph (pp. 303-310). He con- As for why McCown and Keith did not in- cluded that it was male because of the prom- clude the “Sinanthropus” remains in their inent supraorbitals. HrdliCka also consid- comparisons, it probably had little to do with ered the specimen a Neandertal variant, the question of how near or far they might specifically regarding this variety as “more have considered the relationship of the east advanced” than the Europeans. In compari- Asians to be. Like Weidenreich, they were son with the European specimens he focused not really familiar with distant, poorly pub- on the narrower, higher forehead and the lished specimens. For instance, over the pe- facial gracility (given his sex determina- riod when they were preparing the mono- tion), the modern (squared) orbital shape, graph, only a cast of the El adolescent was and the narrower nasal root, concluding in the British Museum (or generally avail- that Zuttiyeh “belongs probably well for- able). The Locus L “Sinanthropus” crania ward in this group.” Indeed, when discuss- had not yet been described in detail. Thus, ing his opinion that the newly described while they indeed did decline to use the Zhoukoudian El cranium could also be in- Zhoukoudian material in their comparisons, corporated in the Neandertal paradigm, he it is unlikely that the question of whether or proclaimed (p. 368): not they should be making explicit compari- sons between Zhoukoudian and Zuttiyeh fig- It is not like the lowest type of the Neanderthaler, but ured prominently in their considerations of corresponds rather to the better developed specimens how to limit the sample they would use for of that class such as the Galilee skull [our italics]. comparisons in their description of the Comparison with Zhoukoudian Mount Carmel specimens. This omission, in other words, did not necessarily reflect a This contention, because it concerned studied phylogenetic interpretation. Zhoukoudian and classified one of its speci- mens as a Neandertal, seemed to have been particularly irritating to Weidenreich, who The relation to Tabun discussed its dismissal at length (1943:214- In fact, McCown and Keith drew few con- 215 j, As part of his defense of the contention clusions about Zuttiyeh at all, except in the that the “Sinanthropus” remains were more discussion of the Tabun female, in which primitive than those of the Neandertal they contended (p. 13) that Zuttiyeh is “of group, he focused on the specific comparison the same type,” linked to the Neandertals by Hrdlicka had made with Zuttiyeh: characteristics of the frontal, zygomatic, and sphenoid bones (see Table 1j. Never was a statement more astray than Hrdlicka’s which brought Sinanthropus into the field of the Gal- What is her [Tabun relationship to the Galilee ilee skull. How little Sinanthropus had in common 11 skull? We believe that they belong to the same race, with this Palestine skull has been shown by McCown the differences, so far as they are revealed from the and Keith (1939). In their monograph on the Mount frontal bone, malar, and sphenoid, being due to sex Carmel Group, in which they include the Galilee and not to race. (P. 265) skull, they expressly decline to compare it with Sinanthropus because of their “distant relationship.” McCown and Keith reversed Keith‘s original But Weidenreich had not seen a cast (or the diagnosis of female sex. Their agreement original specimen) of Zuttiyeh, and based with Hrdlicka’s sex determination was for his conclusions about it on the published de- the same reasons as his, the emphasis on the scription, and especially on the discussions size and robustness of the supraorbital torus in the McCown and Keith monograph. His (p. 256). Paradoxically, they therefore re- concerns were really focused on the status of garded the narrower forehead and greater ‘Sinanthropus” and did not directly involve height of the Zuttiyeh vault, the only signif- the Zuttiyeh skull. In fact, as discussed be- icant differences they drew from their com- low he agreed with Keith’s contention that parisons with the female from Tabun, as a 328 S. SOHN AND M.H. WOLPOFF

TABLE 1. Comparisons abstracted from the text of the Mount Carmel monograph Zuttiyeh Skhul5 Gibraltar Tabun Supraorbital Earlier robust form Earlier form Earlier robust form Frontal height High High Low Low above FH Frontal angle High (64) High (63) Low (55) High (62.5) Minimum frontal Low (97 mm) Low (99 mm) High (107 mm) Low (98 mm) breadth Supraorbital torus Large (22 mm) Large (23 mm) Small (11 mm) Small (15 mm) overhang Maximum frontal Small Small Large Large breadth Angle of inferior Straight, angle Incipiently notched, Straight, angle zygomatic border to marked angle less marked marked the horizontal Body of the Small Small Small zygomatic relative to orbital pillar’ Zygoproptosis’ Slight Slight None Alignment of orbital Straight Straight Moderate bending pillar Projection of lateral Greater (39 mm) Lesser (37 mm) Lesser (37 mm) orbit border in front of greater wing of the sphenoid Extent of lateral Slight Marked Absent pterygoid attachment on the greater wing of the sphenoid Nasomalar angle High (159) High (156) Low (134) Intermediate (140) (M 77) Length difference Slight (3.0 mm) Marked (10.5 mm) Slight (3.0 mm) Slight (3 0 mm) between glahella and nasion horizontals Recession of nasal Shallow Deep Shallow Shallow root Superior internasal Fused Fused Open (in spite of age) suture Projection of nasals Small Small Large anterior to the frontal process of the maxilla Superior breadth Intermediate Small Intermediate Very large across nasals Breadth of nasal root Intermediate Intermediate Small Large (bilachrymal) Orbit width (M 51) Less Slightly greater Slightly greater Less Large mm) Intermediate (33 mm) Orbit height (M 52) Large (37 mm) Small (30 mm) - (39 ‘The zygomatic is quartered with a horizontal line connecting the superior points on the zygotemporal and zygomaxillary sutures, and a line orthogonal to it extending from thefmo point. According to McCown and Keith, the “modern”condition is to have a large zygomatic body (the inferoposteriorquarter) relative to the size of the orbital pillar. ‘Zygoproptosis, according to McCown and Keith (1939:365), is the inferior projection of the masseteric attachment on the lower edge of the temporal process of the zygomatic bone. This contrasts with the straight lower border of Neandertals as described by Rak (1986).

reflection of male sex rather than of a more As our investigations proceeded we encountered so distant relationship. many characters which linked the Skhul to the Tabun type that we were ultimately obliged to pre- The emphasis On a ‘lose to sume that we had hefore us the remains of a single the Tabun woman iS more phylogenetically people, the Skhul and the Tabun types being but the ambiguous than it might seem, since Mc- extremes of the same series. (P. 12) Cown and Keith clearly regard this woman as a member of the same population as the There is no contradiction here. With a single Skhul hominids: Levant population represented in the Mount ZUTTIYEH FACE 329 Carmel caves, Zuttiyeh could logically be re- either. Zuttiyeh was systematically com- garded as coming closer to the “Tabun end pared to Tabun, and otherwise the specimen of the range. was discussed only briefly in numerous dif- McCown and Keith were far from alone in ferent and often unrelated sections. regarding Zuttiyeh as a Tabun-like Nean- dertal specimen. Coon (1963567) linked The relation to Skhul Zuttiyeh to the central European Neander- To better examine the data McCown and tals of the last interglacial and suggested Keith dealt with, we have abstracted all of that it could be a descendant of populations the comparisons with Zuttiyeh they pre- such as represented at Krapina and sented throughout the monograph, in Table Ehringsdorf. In their description of the 1. These have their basis in the discussions Amud cranium, Suzuki and Takai of the three specimens Zuttiyeh was com- (1970:189)wrote that Zuttiyeh is unlike the pared with Skhul 5, Gibraltar, and Tabun. “more advanced” Amud male in having a We note that the sex of Skhul5 is male, and more continuously developed (not divided of Tabun and Gibraltar female. Dimorphism into medial and lateral elements) and is therefore a potential source of variation in thicker supraorbital torus, and in its these comparisons. Nevertheless, it is of in- rounded (Shanidar-like) orbital margin. terest that even a liberal interpretation of They regarded the combination of a dispro- the comparisons in this table suggests a portionately small zygomatic and robust su- quite different conclusion, that Zuttiyeh praorbital as particularly Neandertal-like, shares by far the most features uniquely and concluded that Zuttiyeh is more archaic with Skhul5 (9 in all) and only one uniquely than Amud and most closely resembles with Tabun. Three features are uniquely Shanidar and Tabun. Trinkaus (1983,1984) shared with Gibraltar, and three more with also regarded Zuttiyeh as an ancestral Ne- both Gibraltar and Tabun. Only two fea- andertal earlier in his analyses of western tures are shared uniquely with both Tabun Asian hominid evolution (but see below). His and Skhul5. No doubt, McCown and Keith position was that the specimen is generally overlooked this aspect of their work because archaic and similar to the earlier of the their focus was on the Mount Carmel re- Shanidar specimens, with specific links to mains and they are unlikely to have con- the later Levant Neandertals. Comparing densed and reviewed their specific compari- Zuttiyeh with Shanidar 2 and 4 (1984:285), sons with Zuttiyeh, as we have. However, he stated: our review is hardly supportive of the con- clusion that Zuttiyeh and Tabun were espe- . , , these individuals all have relatively long and cially close in their relationship. robust faces, but they lack the midfacial prog- nathism of the more recent specimens. This is re- We are not the first to recognize a close flected in high zygomaxillary angles, anterior posi- relation of Zuttiyeh and Skhul hominids (al- tions of the anterior zygomatic roots, . . . absence of though the reader should note that this “rec- nasion projection in Zuttiyeh 1, and angled zygomatic ognition” is based only on the comparisons bones of all three. The Tabun C1 specimen. . . may made in McCown and Keith). Weidenreich well belong to this group. (P. 281) believed that Zuttiyeh was more like Skhul 5 (1943:234) than like Tabun. He thought These citations are taken from publica- there were no unique links between Zut- tions in which the authors were not focused tiyeh and the Zhoukoudian hominids, and particularly on Zuttiyeh, and none formally considered it a member of “the Neandertal analyzed the relationships of Zuttiyeh with group” (as he did Skhul5): the other Levant specimens. They seem to have been influenced by McCown and It would perhaps never have entered anybody’s mind Keith’s conclusion that it was Tabun-like. to attribute the Galilee fragment or, more recently, However, the basis for McCown and Keith’s the Skhul calvaria to the Neandertaloid group were they not equipped with typical supraorbitals. This conclusions is not easy to establish because feature is one of the most characteristic peculiarities systematic comparisons were not made with of all the forerunners of modern man. In Paleolithic Zuttiyeh in the Mount Carmel monograph Man of modern type mentioned above (Cro Magnon, 330 S. SOHN AND M.H. WOLPOFF

Predmosti, Upper Cave of Choukoutien-the man TABLE 2. Comparisons inuoluing Zuttiyeh abstracted Obercassel may be added) the supraorbital region from the text of the Zhoukoudian monograph bulges more than in modern man, but typical su- Zhoukoudian fernaleb) Zuttiyeh praorbitals like those developed in the Neanderthal group are missing. (P 235) Frontal sinus small Large as in many Neandertals No sphenoidal sinus Sphenoidal sinus In fact, Weidenreich believed there were extending laterally to four varieties of Neandertals (p. 237). The the pterygoid process, as in Ehringsdorf “Rhodesian Group,” most “primitive” in his Distinct frontal boss Distinct frontal boss, also view, included only the Kabwe skull. The in Ehringsdorf and the “Spy Group” included the European Wurm Mount Carmel crania Supraorbital tori are Same general Neandertals (and the Saccopastore crania). heavy and projecting, supraorbital The “Ehringsdorf Group” consisted of the continuously connected characteristics, also pre-Wurm Neandertals, as well as Stein- by a robustly shared with the Mount developed glabellar Carmel crania heim, Tabun, and Qafzeh (sic). Zuttiyeh and torus, separated from the Skhul crania comprised the fourth the anterior face of the frontal squama by a group, which he believed were “intermediate well defined supratoral between the Neanderthalians and modern sulcus man.” Like McCown and Keith in their in- The facies cerebrulis is The facies cerebralis is transversely small and large, both in terpretation of the Mount Carmel remains, low, not exhibiting any transverse and vertical Weidenreich did not conceive of these variet- special relief directions, and shows ies as being dramatically different. This is relief. Gibraltar resembles Zuttiyeh, suggested by his often stated opinion (for while the Ehringsdorf instance, p. 239) that “the Ehringsdorf skull and Wurm comes very close to the Galilee Skull.” Neandertals resemble Zhoukoudian Zuttiyeh the Neandertal Therefore since Weidenreich regarded found the specimen most like the other Le- Zuttiyeh as a Neandertal variant, he in- vant Neandertal then known, Tabun, while cluded the Zuttiyeh cranium in many of his others regarded it as most similar to the comparisons. This provides the opportunity Skhul remains. At that time this was not to examine some of the relationships with seen as contradictory, since as noted above these earlier Asians reported by Weiden- McCown and Keith as well as Weidenreich reich, the only systematic comparisons be- regarded Tabun and Skhul as members of tween Zuttiyeh and the Zhoukoudian re- the same “paleoanthropic” or “Neanderta- mains that have ever been published to loid” sample. McCown and Keith considered date. These are presented in Table 2. Each of Skhul and Tabun a population “in the throes the five comparisons he made showed Zut- of evolutionary change.” In specific, they re- tiyeh to be similar to at least some of the garded it to be a Neandertal collateral to Neandertals, while three of the five were modern humans. unlike the Zhoukoudian sample. The only two Zuttiyeh features resembling the All who believe in evolution are agreed that Neander- Zhoukoudian hominids are the distinct fron- thal man and modern man are descendants of a com- tal boss and the projecting supraorbitals mon human stock. There must have been a time in which are separated from the frontal the history of that ancestral stock when individuals were undergoing differentiation along, at the least, squama by a deep sulcus. two direction-towards the purely Paleoanthropic While the descriptions and comparisons of (Neanderthal) type and towards Neanthropic type Zuttiyeh from the first half of the century represented by the early people of Cro Magnon. (Pp. almost invariably lead to the conclusion that 13-14). it was a Neandertal, two opinions about its Our theory. . . assumes that the Mount Carmel peo- relations to the other Levant populations ple are not the actual ancestors of the Cromagnons emerged from the initial description and the but Neanderthaloid collaterals or cousins of the an- Mount Carmel comparisons; some workers cestors of that type. (P. 17). ZUTTTYEH FACE 331

TABLE 3. Comuarisons of Zuttiveh morpholom as abstracted from Hublin (19761 Zuttiveh is most similar to European Skhul5 Neandertal Kabwe “Sinanthroous” The cranial height (from bregma to “subcerebral X X plan”)’ Curvature of the frontal, high in the middle of the X frontal forming a frontal boss Nasion-Bregma X Supraorbital torus X X Orbital height X Malar bone X X Nasomalar angle X Sphenoid X Subcerebral plane is the plane which is defined by the following three most medial points. on the frontomalar suture, on the parietomastoidal suture, and on the superior surface of the lesser wing of the sphenoid. According to Hublin, this plane was used earlier by Keith instead of Frankfort or Schwalbe planes because of their inapplicability to the Zuttiyeh fragment. This analysis shows the estimated height of the Zuttiyeh skull to be close to the higher part of the modern European range.

These opinions became more distinct later greater antiquity did not reinforce the ear- as the framework for interpreting the Le- lier case for its Neandertal status. Instead, vant hominids shifted. Largely due to the Zuttiyeh came to be viewed as a potential work of Howell (1958), the Tabun woman ancestor either of all of the Middle Paleo- came to be regarded as earlier than the lithic Levant hominids or uniquely of the Skhul hominids and her more “Neanderta- so-called “modern” ones, and it became in- loid” characteristics consequently could be creasingly clear that it was relevant to con- explained as reflecting a difference in evolu- sider how it may have contributed to the tionary stage instead of as a populational later western Asian populations. variant. With a more distinct boundary en- In 1976 Hublin considered this question, visioned between the Neandertals of the Le- reviewing the information then available re- vant and the other Levant hominids (i.e., garding the dating, morphology and affini- Skhul and Qafzeh), the issue of which group ties of Zuttiyeh. He agreed with Keith’s ini- the Zuttiyeh skull more closely resembled tial diagnosis of female sex, and was the became a more serious one. first to note that some of the Zuttiyeh fea- tures are more archaic than any of the other ZUTTIYEH AS EARLY “MODERN Levant specimens: HOMO SAPIENS” As Tabun came increasingly to be thought Zuttiyeh presents more archaic characters than those of as a European-type Neandertal, without of the neanderthalians, which could be explained by any new discoveries or significant reanaly- its Riss-Wurmian (?) age. The specimen also presents sis Zuttiyeh was shifted phylogenetically to characteristics that are apparently more modern. In spite of the fragmentary nature of the fossil and the become a non-Neandertal, which meant al- problem of how to orient it, it seems to me that the most by default a unique ancestor of “mod- majority of the characters constitute an intermediate ern Homo sapiens.” This is how it happened. morphology between Homo erectus and the earliest The most important development affect- modern humans whose antecedent populations are ing the interpretation of Zuttiyeh came de- known only from Palestine. (Hublin, 1976:9) cades after Weidenreich, with an accurate assessment of the position of the specimen We summarize Hublin’s detailed observa- in the Levant chronology, and eventually tions that are said to support these conclu- with the determination of an absolute date, sions in Table 3, comparing Zuttiyeh to the as discussed above. The archaeological asso- same specimens that Weidenreich discussed ciations and the date, showed the specimen in his 1943 monograph. We find it difficult to be older than any of the other Levant cra- to draw the same inferences that Hublin did nia, although curiously the recognition of from the data we abstracted from his text. 332 S. SOHN AND M.H. WOLPOFF As Table 3 shows, the Zuttiyeh specimen re- TABLE 4. Vandermeersch’s (1989) analysis of the sembles both Kabwe in four out of eight fea- Zuttiyeh “Modern”morphological pattern tures, and the Zhoukoudian crania in four Modern Homo sapiens out of eight features. Two of these four are Features condition in Zuttiveh’ the same, but for the other two resem- Length of the frontal Long (125 mm) squama blances the comparisons involve different, Nasion-bregma chord/arc High: 90.4 unique characteristics. These results do not index Lateral thinning of Exists (cf. Skhul 4, support Hublin’s conclusion (1976, 1983) brow-ridge PFedmosti 3) that the specimen is transitional between Orientation of the frontal Set vertically Homo erectus and the “first modern humans squama Separation of browridge By supratoral sulcus of Palestine,” because the comparisons show and frontal squama’ no special similarities to the specimen best Minimum frontal width Low (cf. Skhul 5, representing the latter sample, Skhul5. In- Neandertals) Predmosti, unlike Orientation of nasion Set back very little from terestingly, the results also do not support the glabella Weidenreich‘s contention that there are no Nasal notch Very weak (unlike special resemblances between Zuttiyeh and Neandertals) Maximum frontal width3 Smaller than the Zhoukoudian remains. Hublin noted Neandertals such resemblances, stating: Frontal sinus Differently shaped and thinner than Neandertals Except for orbital height, most of the “Neandertal” Nasal bones Flat, and not shaped as characters of Galilee are found in the anatomy of Neandertals Sinanthropus, the best known of the Homo erectus Orbits Rectangular (wider than specimens. high) Zygomatic bone More robust than Neandertals In fact, Hublin’s observations (1976) are Position of zygomatic Anteriorly facing unlike close to what McCown and Keith (1939) re- body Neandertals ported, which is not surprising since he re- ‘By “Modern Homo sapiens,” we believe that Vandermeersch actually lied heavily on their monograph. Yet, unlike means “non-Neandertal.” He does relate these features to what he regards as “proto-Cro-Magnons”(Skhul and Qafzeh specimens) and to them he finally concluded that Zuttiyeh has at least a few of the early post-Neandertal Europeans (Predmosti 1 no special relationship to the Neandertals, and 31. However, by not comparing to other groups it is unclear whether these uniquely link the specimens he compared. Vander- and no particular relationship to Tabun meersch also notes Neandertal-like features in Zuttiyeh such as the (1987). large supraorbital torus or robust zygomatic. These features are com- pared with the Amud 1, Shanidar 1 and 2, and sometimes Tabun 1 in Vandermeersch (1989), who expressed a his analysis. ‘On the left side of the face there is a classic continuous supraorbital similar position on the affinity of Zuttiyeh, torus. On the right side a healed wound is present, perhaps contribut- expanded on the details supporting the con- ing to the separation of the supraorbital into a supercilhary arch and trigone by a shallow supraorbital groove. tention that Zuttiyeh should be considered 3Maximum frontal length in Zuttiyeh is also smaller than “Proto-Cro- uniquely ancestral to a “modern” lineage, Magnons” and “Cro-Magnons,” except for Qafzeh 9 and Skhul 5, ac- and not ancestral to the Neandertals (1989). cording to Vandermeersch. This is part of the support he provided for his contention that the Levant Neandertals are intrusive, refugees from European gla- ary arches-see our discussion of the speci- cial conditions (1981). His analysis is sum- men, and Table 5. marized in Table 4. We find Vandermeer- Vandermeersch stated that the many sch’s analysis of characteristics to be mostly “modern Homo sapiens-like” features similar to ours, and only disagree with his shared by Zuttiyeh and the later Levant assessment of two features. For the length of specimens from Skhul and Qafzeh provide the frontal squama, we find it to be within evidence of a single lineage. In contrast, be- the range of the Levantine sample and spe- cause only a few “Neandertal-like” features cifically similar to (or smaller than) one are shared by Zuttiyeh and the Amud and specimen in each subgroup (similar to Skhul Tabun specimens, there is no closeness of 9, smaller than Amud). Secondly, our as- relationship or corresponding intraspecies sessment is that there is no separation of the clade, in his view. Therefore he did not envi- frontal torus into lateral tori and supercili- sion Zuttiyeh as a common ancestor €or two ZUTTIYEH FACE 333

TABLE 5. Distribution within the Levant hominids of nonmetric features shared by Zuttiyeh and the Zhoukoudian crania‘

-~Qafzeh Skhul Tabun Amud 3 5 6 9 4 5 9 ~~ .. Frontal curvature 0 X - - x- 000 Narrow boss X 0 0x0000- Short temporal notch - 0 - - xxoo- Supratoral sulcus: size and X 0 oxxooxx morphology Glabella depression X 0 x- 0- xo- Flat nasal root X 0 x- 0- xo- Nasal bones transversely flat - 0 0- 0- - - - Tubercle on zygomatic anterior face - X - - xxoo- Anterolateral orientation of - 0 - - xxoox zygomati Unangled superior nasal bones - 0 x- 0- - - - (vertical orientation) ‘The character states are defined by the shared features of the Zuttiyeh-Zhoukoudian comparison and evaluated for the purpose of this comparison and not on an absolute scale. Thus, the table does not provide an in dependent definition of frontal curvature, but rather reports that Amud and Qafzeh 6 have frontals that are curved in the sagittal plane in a manner similar to the frontal curvature common to Zuttiyeh and the Zhoukoudian sample. X = present; 0 = absent; - = part not preserved. distinct Levantine demes, but rather consid- ubiquitous among Levant populations that ered it a unique ancestor for only one of are later than Zuttiyeh, for instance Amud these. and Tabun, but are only rarely and sporadi- Vandermeersch recognized that there are cally present in earlier hominids? What numerous regional features shared among makes this problematic is the fact that the all of the Levant specimens, including Zut- shared unique features of a hypothetical tiyeh. However, he emphasized the unique modern Homo sapiens clade that are defined links between this early specimen and the by the assumption that the Neandertal-like SkhuYQafzeh remains in his analysis of evo- features are plesiomorphic, are the very fea- lution in the region. In summarizing Van- tures that have the best case for geologic dermeersch’s observations, Table 4 shows precedence. If geologic precedence is the cri- that his model of a “Zuttiyeh-modern Homo terion for plesiomorphy (an assumption sapiens” lineage is based on the “non-Nean- Vandermeersch is willing to make in dis- derthal like” features. This model therefore carding the similarities of Zuttiyeh and the assumes that “non-Neandertal-like” must later Neandertals), then the attributions of mean more “modern-like” and synapomor- character states must be reversed. In this phic. The lack of a similar set of shared case, Zuttiyeh and the SkhuVQafzeh speci- unique resemblances linking Zuttiyeh and mens would have to belong to a “modern the Levant Neandertals is significant in clade” that is not defined by synapomor- Vandermeersch’s phylogenetic reconstruc- phies! tions. Thus, according to Vandermeersch‘s We agree with the suggestion that “it is phylogeny, the few “Neandertal-like” fea- possible that some of the facial features that tures in Zuttiyeh must be interpreted as ple- separate [Zuttiyeh] from Neandertal speci- siomorphic, and their presence is explained mens are plesiomorphies rather than mod- as a consequence of Zuttiyeh’s more ancient ern apomorphies” (Stringer et al., 1984:118). age. But if true for some of the Levantine folk, ZUTTIYEH AS A GENERALIZED this interpretation must be true for all, and LEVANT HOMlNlD the fact is that this interpretation does not The phylogeny suggested by Hublin and validly describe character states in the Le- Vandermeersch takes on added significance vant. Why should the Neandertal-like fea- with the contention that the split between tures found in the Levant be considered ple- Levant Neandertals and what they (follow- siomorphic, given the fact that they are ing Howell, 1958) call “proto-CroMagnoids” 334 S. SOHN AND M.H. WOLPOFF is at the species level (Gould, 1989; Rak, Finally, in a recent study Rak (1986) sug- 1990; Stringer, 1990). But not all authors gested that the Zuttiyeh face fit his model believe that there is a split in the Levant for a “generalized (i.e., unmodified) face” (p. hominids, instead accepting Howell’s (1958) 153). By “generalized face” Rak meant the model of a continuous evolving human line facial morphology “shared. . . by many pri- in the Upper Pleistocene of western Asia. mates, including modern humans and fossil Others, accepting a split, believe it occurred species of the genus Homo. He characterized later than Zuttiyeh. If there is no split, or if the “generalized” face in great detail, but in it postdates Zuttiyeh, there is no problem of so far as the regions preserved in Zuttiyeh which side Zuttiyeh lies on. are concerned only two of the criteria can be Trinkaus, after first regarding Zuttiyeh applied: as an ancestral Neandertal, later came to consider it as a generalized ancestor of all a peripheral region oriented approximately on the coronal plane with the infraorbital plates facing for- the more recent west Asian hominids, as he wards. Their surface slopes down and slightly back- did the specimens he considered it most sim- wards. ilar to, Shanidar 2 and 4. He believes (1989:45)“Zuttiyeh . . . preserve[sl m unique- According to Rak these specifics are shared ly derived (autapomorphous) features of ei- with the Zhoukoudian faces, which if true ther modern humans or of Neandertals.” would provide a link between Zhoukoudian However, he bases this assessment on a and Zuttiyeh lacking phylogenetic signifi- number of comparisons we believe to be in- cance since it would be based on plesiomor- correct (also see Trinkaus, 1991). These in- phies. clude the claim that its frontal morphology We concur that there is such a link, but is archaic (compared to what? The frontal is question whether it is actually established similar to the Zhoukoudian hominids but on features plesiomorphic (or “generalized”) quite unlike those from Sangiran), the as- for Homo. Our concern is based on two facts. sertion that frontal is “clearly distinct from First, we have discussed (and discuss below) those of early modern humans” (in fact it is a number of additional similarities between very similar to Qafzeh 3, and see below), and Zuttiyeh and the Zhoukoudian remains that the assessment that its anterior orbital pro- Rak does not include in his concept of the jection is particularly elevated relative to “generalized face. Many of these are unique the cranial base. to Asian Homo, even if the two features Rak Similarly, Smith (1985) regards Zuttiyeh discussed are not. Second, we believe that of as an early (or the earliest) member of the all the features Rak discussed, the two that Levant hominid line, ancestral to all of the pertain to Zuttiyeh actually are not “gener- later hominids from the region and there- alized” in the sense he meant. For instance, fore not uniquely ancestral to either a mod- the male face from Sangiran 17 has infra- ern or a Neandertal lineage. In a principal orbital plates that slope down and markedly components analysis of the Levant sample forwards. If one believed there was such as (Simmons et al., 1991) it was concluded that thing as a single “generalized”facial pattern Zuttiyeh is not particularly similar to the for Homo, the Indonesian specimen is much Shanidar sample, but is most similar to more complete than any from Zhoukoudian Amud. These authors show that and would be at least as good a candidate for ascertaining the “generalized morphology arguments for a greater phenetic or cladistic similar- ity between Zuttiyeh and SkhuVQafzeh than between character state. Further, if we were to use Zuttiyeh and the Levantine Neandertals are simply geologic precedence to establish “general- not supported by available data. . . . Zuttiyeh is prim- ized,” among the earlier pre-dispersal faces itive compared to both Levantine Neandertals and of Lower Pleistocene Africans, ER 3733 re- the SkhuVQafzeh hominids sembles the Zhoukoudian reconstruction in Stringer (1985:294) would seem to concur, these two features, but the infraorbital concluding that Zuttiyeh “does not appear to plates of ER 3883 have an anteroinferior ori- this author to be transitional between ar- entation (similar to Sangiran 17) and a lat- chaic and modern humans.” eral orientation to the peripheral aspects of ZU’R’IYEH FACE 335 the plates that begins in a more medial posi- tiyeh. Such a phylogenetic model, while pos- tion than any of the Asian Homo faces dis- sible, has never been suggested for the Le- cussed above. One of us (MHW) has studied vantines. the original specimen and asserts that this If nothing else, it is evident from the above anatomy is not the result of distortion in the that the morphology of the Zuttiyeh speci- infraorbital region. We conclude that there men is neither an obvious reflection of evo- is no single pattern of plesiomorphic (or lutionary grade nor of cladogenesis. We “generalized”)infraorbital anatomy that can question why so many different interpreta- be established within known specimens of tions of the features of Zuttiyeh have been Lower and earlier Middle Pleistocene Homo. retained? In our opinion, its phylogenetic In sum, we reject the contention that the position in the region and the explanation of relation of the Zuttiyeh face to the Zhou- why there has historically been so much con- koudian specimens is irrelevant to phylog- fusion about it, is not a simple consequence eny because it is based on plesiomorphic fea- of incompleteness. It is, at least in part, a tures. consequence of confusing grade and in- traspecies clade features. WHY SO MANY OPINIONS? Historically it is clear that virtually every EAST ASIAN EARLIER MIDDLE opinion possible has been expressed with re- PLEISTOCENE FEATURES IN ZUTTIYEH gard to Zuttiyehs phylogeny, with different It is this suggestion that we will explore hypotheses often expressed by different below. If a confusion of grade and intraspe- combinations of the same authors [for in- cies clade results from analyzing the later stance, compare Stringer (1985), Hublin Levant specimens, perhaps information (19761, Vandermeersch (19891, and Stringer about earlier hominids would be useful. et al., (198411. Among those who believe Such information must come from other re- there were two lines, Zuttiyeh has been allo- gions since Zuttiyeh is the earliest Levan- cated to each side. In the past, whether Le- tine cranium. We find that visual inspec- vant hominids were regarded as evolving to- tion, if not the conclusions that can be drawn ward Neandertals (of the sort found in from Hublin’s (1976) comparisons, indicate western Europe), or toward post-Neandertal the possibility of a special relationship with European populations such as seen at Cro- the earlier east Asian folk. Therefore, we Magnon, Zuttiyeh has been considered to be begin by examining the relationship of this part of each process. More currently, for specimen with the remains from Zhoukoud- those who believe that early modern hu- ian. If this comparison shows that Zuttiyeh’s mans appear first in the Levant, and that unique features can be linked with these the Levant Neandertals are later refugees much earlier east Asians, by definition they from glaciated Europe, Zuttiyeh is seen as cannot be considered modern apomorphies. the unique ancestor of the modern humans. [As an aside, we do not choose to also com- For others who regard the Levant sequence pare Zuttiyeh with later Middle Pleistocene as reflecting regional continuity and the Asians (Dali, Maba, Jinniushan, Yunxian) gradual evolution from Neandertals to mod- for several reasons. This comparison raises ern humans, Zuttiyeh is seen at its base. different issues that we intend on covering While the question of two Levantine lin- in a forthcoming paper and which do not eages has been critical in evaluating Zut- effect the question of a link between Zut- tiyeh, we believe that the morphology of Zut- tiyeh and Zhoukoudian. Moreover, three of tiyeh cannot be used to determine the the four specimens in question are not yet converse, whether one or two lineages are available for a detailed study that can be represented by the later Levantine homi- published (M.H.W.’s observations, however, nids. This could only be a useful approach if suggest that these fit the model quite well). there were clearly two subsequent samples Finally, it is quite possible that Zuttiyeh is of Levant hominids defined by two different of comparable age to these specimens, which sets of apomorphic features, and if one of would make the question of potential ances- these were linked to synapomorphies in Zut- try irrelevant.] 336 S. SOHN AND M.H. WOLPOFF We believe that the obvious resemblance proach because these are likely to do little of Zuttiyeh to the frontofacial portions of the more than continue the confusions of grade Zhoukoudian specimens is a reflection of an and intraspecies clade features that already intraspecies clade relationship which has are well published (see above, and for a re- never been examined systematically-this cent example Holt and Waddle, 1991).More- is the hypothesis we propose to test. Perhaps over, we have not attempted to once again this hypothesis was never tested before be- fully describe Zuttiyeh. Rather we focus on cause Weidenreich was so adamant in deny- two ways of viewing the problem. These are ing any special relation existed, but we be- the questions of (1)how Zuttiyeh fits into lieve Weidenreich’s opinion was more in comparisons between east Asian earlier response to Hrdlicka’s claim that “Sinan- Middle Pleistocene remains and the later thropus” was a Neandertal variant, than the Levant crania, and (2) whether comparisons result of any systematic analysis of Zuttiyeh with other geographic regions show the fea- on his part. In our view, the intraspecies tures that link Zuttiyeh and the east Asians clade relationships reflected in the non-Ne- to be regionally unique. andertal features of the western Asian sam- ples have been misinterpreted to suggest a METRIC ANALYSiS “modern Homo sapiens” grade (the assump- If our hypothesis is correct, we would ex- tion that non-Neandertal means modern), pect many if not most of the differences be- when their explanation is actually quite dif- tween the later Levant crania and Zuttiyeh ferent. It is also possible that the hypothesis to place Zuttiyeh in the direction of the has been ignored because of potentially Zhoukoudian frontofacial morphology. Con- confusing descriptions of character states versely, if Zuttiyeh does not fit between in fossil hominids. For instance, Habgood these east Asians and the subsequent Le- (1992:279) describes Zuttiyeh as exhibiting vant folk, it would argue against the conten- a Mongoloid-like “flat upper face with for- tion of a unique phylogenetic link connect- ward jutting malar bones,” but he also at- ing the Upper Pleistocene hominids from tributes this morphology to specimens such the Levant and the earlier east Asians. diverse as Bod0 and Sangiran 17. Such attri- In our metric analysis of Zuttiyeh, we ex- butions suggest a quite different definition amined 77 chord and arc measurements of of a Mongoloid-like flat face than the classic the specimen and compared these with the morphological one (cf. Coon, 1963) we have corresponding data for the other Levant cra- applied here to the incomplete Zuttiyeh face nia (Amud, Tabun, and the full samples (see below). from Skhul and Qafzeh). All of these mea- Here we take several approaches to the surements were taken by one of the authors question of whether the Zuttiyeh morphol- (M.H.W.) on the original Levant specimens, ogy includes those elements of the Zhou- using the same techniques. The east Asian koudian frontofacial complex that can be Zhoukoudian measurements were taken by expected in the anatomical regions it pre- the same author on the primary cast set at serves (Weidenreich, 1943; Coon, 1963; Wol- the American Museum of Natural History, poff et al., 1984). Most of the comparative and on the original H3 specimen in Beijing. information comes from ZKT L2 and L3, The two other earlier Middle Pleistocene since few other specimens preserve parts crania from the region cannot be included in corresponding to Zuttiyeh, and no other east the metric analysis. Gongwangling pre- Asian earlier Middle Pleistocene remains serves virtually nothing that remains in preserve comparable elements of the face. Zuttiyeh except the frontal and superior na- Limited comparisons are also made with the sal, and this is far too distorted for accurate less complete Zhoukoudian vaults, Gong- measurements. Hexian is not yet available wangling (Woo, 1966), and Hexian (Wu, for measurements. 1985; Wu and Dong, 1982; Etler, 1990). To The most obvious conclusion we found in examine the potential refutations for this the preliminary univariate metric compari- hypothesis we decline to use either a dis- sons is an unsurprising one; for the vast ma- crimination type analysis or a clustering ap- jority of linear measurements and indices, ZU’ITIYEH FACE 337

0 115 0 0 r/] z 9, Q 0 5 110- 0 € 0 0 0 k 0 5 105- 0 0 cm 9, 4 0 100 - Legend 0 Zuttiyeh Po Levant Neandertalr 0 Skhul/Oafzeh 0 Zhoukoudlon

Fig. 1. Relative size of the Zuttiyeh frontal bone in the Zhoukoudian and Levant samples.

Zuttiyeh closely resembles the other Levant zygomatic, or of the middle face, clearly sep- crania (especially the Tabun and Qafzeh 3 arate the specimen from the other Levant females), falling within the range of varia- crania. Zuttiyeh‘s uniquenesses involve ele- tion of the Levant sample. Moreover, for vir- ments of frontal size (Fig. 11, frontal shape tually all of these measurements the Levant (Fig. 21, and the dimensions of the lateral samples themselves overlap, so that at least aspect of the supraorbital torus (Fig. 3). as far as the parts preserved on the Zuttiyeh The unique aspects of frontal bone size fragment are concerned, few details sepa- can be seen in the comparison of the bone’s rate the so-called “Levant Neandertals” sagittal length from glabella plotted against from the “moderns.” The implications of a maximum frontal breadth (Fig. 1). The Zut- similar observation were discussed at length tiyeh bone is narrower than any other of the by McCown and Keith (1939) and Dobzhan- Levant crania, but longer than all but Amud sky (19441, who recognized a close relation- (it is about the same length as Skhul9).The ship between the Skhul and Tabun remains. scatter plot shows Zuttiyeh totally within In our analysis, we find no purpose in focus- the Zhoukoudian cluster. In terms of size, ing on the features that overlap between the then, the Zuttiyeh frontal is unlike the later post-Zuttiyeh populations of the Levant. Levant crania and indistinguishable from Only a few metric features were found to the Zhoukoudian remains. separate Zuttiyeh from the other Levant The distinguishing aspects of frontal crania, although these do not necessarily shape (Fig. 2) are found in the frontal curva- distinguish Levant samples from each ture (arclchord) index as calculated from na- other. We review all of these in the scatter sion, and the relative projection of nasion plots shown in Figures 1 through 3. The dis- anterior to the bi-fmt line (the relative pro- tinguishing features are all associated with jection is an index of the nasion projection the frontal bone. No metrics of the Zuttiyeh and the nasion-bregma length). These pro- 338 S. SOHN AND M.H. WOLPOFF 301

0

0 0 0 0

Legend 0 Zuttiyeh Levant Neond*rtalr 0 Skhul/Qafzrh 0 Zhoukoudion

105 110 115 120 125 frontal Curvature Index

Fig. 2. Relative shape of the Zuttiyeh frontal bone in the Zhoukoudian and Levant samples. vide measures of the frontal curvature and crania in medial and mid-orbital positions, of what we describe as the flatness of the and indeed the Levant hominids cannot be upper face. [This is not the same as a much separated into different samples on the ba- more widely cited measure, facial flatness, sis of these measurements. However, in its which is a description of the zygomaxillary most lateral aspect (Fig. 3) Zuttiyeh is verti- region that is incomplete in this specimen.] cally thinner than any specimen except Zuttiyeh has a relatively uncurved (i.e., sag- Amud, but its projection anterior to the en- ittally flattened) frontal, very flat as com- docranial surface is greater than all of the pared with the Levant Neandertals. In the Levant crania. It lies between the Levant paracoronal plane, its upper face is also frontals and the Zhoukoudian sample in the transversely flat. Taken together, the small scatter plot showing the relation of these nasion projection and the large frontal cur- two variables. vature measures fall very close to the Perhaps the most unexpected result of fo- Zhoukoudian cluster. We note that in con- cusing on the features €or which Zuttiyeh is trast, upper facial flatness is not a European unique is that these seem to either com- Neandertal feature (see Wolpoff et al., 1981; pletely align the specimen with the Zhou- Trinkaus, 19911, just as it does not charac- koudian sample, or show Zuttiyeh to be terize the Levant Neandertals. The presence intermediate between these earlier east of transverse upper facial flatness is not Asians and both of the later Levant samples. necessarily a marker of “modern humans” either, since the Zhoukoudian specimens, MORPHOLOGICAL ANALYSIS and as far as one can determine visually To further examine the possibility of a Gongwangling and Hexian as well, also link between Zuttiyeh and the earlier east have flat upper faces. Asians, we undertook a series of morpholog- Supraorbital dimensions of Zuttiyeh ical comparisons. Our conclusion, as de- closely resemble those of the other Levant tailed below, is that the morphological simi- ZUTTIYEH FACE 339

14 0

13

v, (0 : 12 & .o, E 0 < 0 s 00 .o, 11 0 i 0

10

iemnt ffeanderklr 0 Skhul/Oafzmh 0 Zhoukoudh

Fig. 3. Relative dimension of lateral supraorbital torus of the Zuttiyeh frontal bone in the Zhoukoud- ian and Levant samples.

Fig. 4. Zuttiyeh (left) and a cast of Zhoukoudian 12 in lateral view, not to the same scale.

larities of Zuttiyeh with the Zhoukoudian koudian specimens and Hexian there is no crania-and, where comparisons are possi- frontal keel and the squama thickness is ble, also with Gongwangling and Hexian- generally less. The temporal notches are are striking (Figs. 4, 5). There are obvious equally short in an anteroposterior direc- similarities in the frontal shape (curva- tion, while the internal wall is more vertical ture-also as indicated in the metric com- in Zuttiyeh than in the Zhoukoudian speci- parisons-and narrowness of the centrally mens. The size, shape, and depth of the well located boss), although unlike the Zhou- excavated supratoral sulcus is identical to 340 S. SOHN AND M.H. WOLPOFF

Fig. 5. Zuttiyeh (left) and a cast of Zhoukoudian 12 in frontal view, not to the same scale. all of the Asian specimens, and this identity Thus, we interpret the orientation of the me- extends to the transversely flattened supra- dial aspect of this bone as reflecting marked nasal region. mid-facial flatness similar to that in the A suite of upper facial features is shared Gongwangling midface (in so far as it is pre- with all of the east Asian specimens. For served) and even flatter than the ZKT L2 instance, Zuttiyeh resembles all in exhibit- maxilla, and we question why Keith was ing a rounded superior orbital contour and able to describe this region as “Neandertal- supraorbital configuration above it. Gla- like.” The isolated L1 zygomatic, a male, is bella is not prominent; in fact, from above more robust than Zuttiyeh where compara- the central supraorbital region is slightly ble. However, Zuttiyeh is similar to it in the depressed at the midline. Directly below gla- marked development of a tubercle along the bella, when preserved the superior nasal re- inferior third of the anterior face, above the gion is extremely flattened, both above and tuberosity for the masseter attachment below the frontonasal suture where the na- (which is missing on Zuttiyeh). sal bones of Zuttiyeh are transversely as On the whole, Zuttiyeh shares many fea- flattened as ZKT D1 and L1, and even flat- tures with the comparable portions of Gong- ter than H3, L2 (which is identical to Gong- wangling, Hexian (although there is not wangling), and L3. The Zuttiyeh nasals are much basis for comparison) and the Zhou- also similar to the above in their lack of a koudian crania, most of which are regionally sagittal keel, the very flat (undepressed) na- specific to this north China sample. Most sal root, and the vertical orientation of the striking is the contribution of the superior preserved superior aspect of the nasal orbital border, the nasal root and region just bones. The zygomatic is somewhat more lat- above it, and the zygomatic, to the combina- erally oriented than in L3, and is more simi- tion of upper facial flatness and flatness of lar to the L2 reconstruction in its anterolat- the mid-face described above, as these com- era1 angulation (for instance as seen from bine features that are most common in liv- above). It differs from a number of speci- ing Mongoloid populations and (those who mens sometimes described as “flat-faced,” we perceive to be) their ancestors. We do not but whose flatness only involves the ante- mean to imply that Zuttiyeh is no more than rior maxillary face, contrasting with a much a west Asian version of the Zhoukoudian more laterally oriented zygomatic which folk, for there are differences as well as the sharply angles along the zygomaxillary su- similarities we mention above. Never the ture. In Zuttiyeh the zygomatic contributes less, we find that the contention of a fairly significantly to midfacial flatness of the close relationship between these is well sup- classic Mongoloid form (cf. Coon, 1963). ported by the morphological data and not at ZUTTIYEH FACE 341 all contradicted by the revised Middle Pleis- is the same age as Zuttiyeh, which would tocene age estimate. make the question of even potential ances- try irrelevant. OTHER REGIONAL COMPARISONS The significance of the morphological Middle Pleistocene Europeans comparisons is linked to the distribution of Most of the frontofacial features Zuttiyeh these features in other samples that could, shares with the Zhoukoudian sample do not alternatively, be ancestral to Zuttiyeh. Such on the whole characterize Middle Pleis- samples are found in Europe and Africa. tocene samples from other areas, although they are occasionally found in isolation. For Lower Pleistocene Africans instance, considering the European sample, The earliest Homo sapiens remains are with the possible exception of frontal nar- African-this is where to the best of our rowing, our comparisons show that none of knowledge the species originated at the be- the metric distinctzons of Zuttiyeh or its ginning of the Pleistocene. Some workers specific detailed morphological similarities [for instance, Simmons et al. (1991)l have to the Zhoukoudian remains characterize suggested that the flat facial morphology Steinheim or Arago 21 (contra Wolpoff 1980, shared by Zuttiyeh and the Zhoukoudian re- who reported similarities between Zuttiyeh mains is also shared by the Lower Pleis- and Steinheim in a comparison that was not tocene Africans. If true, such a similarity metric and did not include the Zhoukoudian could potentially remove facial flatness remains). Nor are these metric and morpho- from the suite of derived characters logical distinctions linking Zuttiyeh and the uniquely linking Zuttiyeh with the Zhou- Zhoukoudian hominids shared with the koudian specimens (although it would not male vault from Petralona. If these mem- remove the other shared unique features bers of the Middle Pleistocene European in- from this list). In fact, however, it is not the traspecies clade are regarded as an out- case. In ER 3733 and 3883 there is variabil- group for comparative purposes, the ity in the orientation of the zygomatic, as we similarities Zuttiyeh shares with the Zhouk- describe above, but in both specimens the oudian crania are clearly highlighted as be- bulk of the zygomatic’s face is oriented in a ing different and thereby unique. much more lateral direction than in Zut- tiyeh. We believe that the description of Middle and Early Upper Pleistocene these early Africans as “relatively flat- Africans: the “Eve” theory faced is not accurate and does not match With regard to the later African remains, our delineation of east Asian facial flatness according to some hypotheses (for instance, detailed above. the “Eve”theory) Middle and/or early Upper These Lower Pleistocene Africans are Pleistocene Africans are uniquely ancestral common ancestors to both Zuttiyeh and to all modern humans (Stringer, 1990; Zhoukoudian. The absence of Mongoloid- Stringer and Andrews, 1988). Presumably like middle and upper facial flatness in this this means that Middle and/or early Upper sample supports the contention of a Zhou- Pleistocene Africans are related to Zuttiyeh, koudian-Zuttiyeh relationship, but because unless this specimen is not thought to be the sample is much earlier it does not ad- ancestral to the SkhuIlQafzeh hominids-a dress the issue of whether this relationship contention no author has supported. We, is particularly unique. For this reason, it is with others, accept the hypothesis of this useful to examine the Middle Pleistocene ancestral relationship. Therefore the de- humans of Europe and Africa, since they mands of the “Eve” theory are relevant to represent populations who may be different the interpretation of Zuttiyeh. If Zuttiyeh is potential ancestors of Zuttiyeh, but not of not a Neandertal, but is ancestral to the their east Asian contemporaries or prede- SkhuVQafzeh hominids, it must be closely cessors from Zhoukoudian. We recognize the related to the Middle and/or early Upper possibility that one or both of these samples Pleistocene Africans who are also ancestral 342 S. SOHN AND M.H. WOLPOFF

TABLE 6. African distribution of nonmetric features shared bv Zuttiveh and the Zhoukoudian crania Eliye Features Bod0 Kabwe Ndutu Florisbad Ngaloba Springs Frontal curvature 0 0 - - 0 X Narrow boss 0 0 0 0 0 0 Short temporal notch 0 X - 0 - X Supratoral sulcus: size & 0 0 - 0 0 0 morphology Glabella depression 0 0 - 0 X - Flat nasal root 0 0 0 0 0 - Nasal bones flat (transversely or 0 0 0 X - 0 superiorly) Tubercle on zygomatic anterior face 0 X - X - 0 Anterolateral orientation of 0 0 - X - x zygomatic Unangled superior nasal bones X 0 0 0 - X (vertical orientation) 'The character states are defined by the shared features of the Zuttiyeh-Zhoukoudian comparison. For instance, the table indicates that Eliye Springs has a frontal that is curved in the sagittal plane in a manner similar to the frontal curvature of Zuttiyeh and the Zhoukoudian sample. X = present; 0 = absent; - = part not preserved.

to the SkhuVQafzeh hominids according to the comparisons to include the males, Ka- this theory. bwe and Bodo, still results in only a few sim- To examine the possibility of an African- ilarities-none for all three specimens. Zuttiyeh relationship, comparisons with the If the later remains from Ngaloba (Laetoli African specimens we regard as having the Hominid 18) and Florisbad are also consid- best chance of dating to a timespan earlier ered potential ancestors, of the 10 morpho- than Zuttiyeh are necessary. These include logical comparisons that uniquely link Zut- Kabwe, Bodo, and Ndutu (earlier specimens), tiyeh and the Zhoukoudian sample, only and to be conservative also Ngaloba and Flo- three link Zuttiyeh with Florisbad, and just risbad (specimens that are almost certainly one links it with Ngaloba. In fact, Zuttiyeh too late to be ancestral) and the undated most closely resembles Eliye Springs (ES Eliye Springs cranium. We did not include 11693) in these features (four are shared). the Jebel Irhoud crania (as urged by two However, only one of these four is also reviewers of this paper) as neither they nor shared with the next most similar specimen, the stratigraphic levels they are said to have Florisbad. Moreover, Eliye Springs may be been discovered in are directly dated, and in much too recent to be ancestral. We find any event their provenience is uncertain. many details of this specimen to resemble Grun and Stringer (1991:185)report that other Africans and not Zuttiyeh; for in- stance, the frontal keel extending to the su- the unsystematic collecting procedures used earlier praorbitals and interrupting the supratoral at the site mean that a precise context cannot be sulcus, the flattening of this sulcus at the provided for Irhoud 1-3. temporal lines, the length of the frontal from the lateral orbit to the stephanion position Moreover, there is little reason to be confi- on the coronal suture, the keeling of the in- dent of the ESR dates reported for the cave. ter nasal suture, and the depression of the The ESR date range for three animal teeth nasal root. We are also concerned that two of from a single level, 90-190 kyr, throws the the observations suggesting similarity with accuracy of the ESR date estimate in this Zuttiyeh, the vertical orientation of the su- cave into serious doubt. perior nasals and the flattened infraorbital In the comparison with the single female region, may be affected by crushing. in the earlier African group, Ndutu, none of In all, while gene flow from Africa is a the features linking Zuttiyeh and Zhou- likely source of at least some variation in koudian can be found (Table 6). Expanding Levant populations, we find no case for a ZUTTIYEH FACE 343 unique or even significant pattern of African discuss, but would have to discuss to ascer- ancestry for the western Asians in these tain the pattern of relationship among the data. Levant specimens. The shared east Asian features in the Tabun and Qafzeh 3 females EAST ASIAN FEATURES IN LATER and the Qafzeh 6 male show that if one ac- LEVANT HOMlNlDS cepts the division of the Levant hominids This leaves the question of whether the into Neandertal and “modern”samples, both features linking Zuttiyeh and the Zhou- samples show east Asian features and there- koudian specimens can be found in the later fore are unlikely to be uniquely African in Levantine folk, as one would expect if they their origin. Instead, a case based on the reflect an ancestral relationship and Zut- east Asian features of Zuttiyeh can be made tiyeh (or her peers) had offspring. While not for significant local continuity in the Le- all comparable parts are preserved, and in vant. spite of the broader frontal in the later fe- Moreover, the focus on the east Asian fea- male, we find most of the similarities be- tures of Zuttiyeh helps explain one more dif- tween Zuttiyeh and the Zhoukoudian sam- ference between this study and two earlier ple shared by the Tabun woman. This is ones; mainly, the suggested relationship to especially striking in the region surround- Skhul 5. McCown and Keith (1939) never ing the root of the nose, where Tabun differs specifically addressed the relationship of dramatically from the European Neander- Zuttiyeh to the Levant sample, but the con- tals (female and male). However, before this clusions we abstracted from the compari- is used to make Zuttiyeh into a Neandertal sons they made in their text (i.e., Table 1) again, we note that in those cases where suggest that the closest phenetic relation of Zuttiyeh is similar to Tabun, the other Le- Zuttiyeh is, in fact, with Skhul 5. However, vant Neandertal, Amud, is quite different. if instead of examining all the possible com- To quantify the comparisons of Zuttiyeh parisons, we limit the comparison to only with the other Levant specimens for the east east Asian intraspecies clade features pre- Asian features Zuttiyeh shares with the served in Zuttiyeh, this specimen is proba- Zhoukoudian crania, we have prepared Ta- bly in fact least like Skhul 5-a conclusion ble 5. This table shows that some of the east that may have been reflected in Coon’s Asian features of Zuttiyeh are to be found in (1963574) comments about this Skhul spec- both the Levant Neandertal sample and the imen, in which he specifically likened it to non-Neandertals (the so-called “moderns”). Australasians. Cognizant of the uncertainties due to miss- In the second study, Hublin also related ing data, we nonetheless believe that these Zuttiyeh to Skhul5. In Hublin’s (1976) sub- similarities are not unique to either of these sequent analysis, there is a contradiction be- samples, although both the number and the tween the conclusions in his text and the combination in each specimen is different. systematic comparisons that he provided. Thus, for instance, the similarities sur- His stated conclusion is that Zuttiyeh is rounding the nasal root region shared with much like Skhul5, while the comparisons he Tabun are all absent in Amud. These partic- presents (abstracted in our Table 3) show it ular similarities are, however, shared with to be least like Skhul 5. We wonder if this Qafzeh 3 and 6. As another example, the difference between what he thought and specific form of the Zuttiyeh supraorbitals what he showed may not reflect the confu- and of the supratoral sulcus are shared with sion of grade (i.e., “modern”) and intraspe- both Tabun and Qafzeh 3. cies clade (i.e., Asian) features. His analysis We do not mean to imply that Tabun surely suggests that the features in which shares a unique relationship with Qafzeh 3 Zuttiyeh is similar to the Zhoukoudian hom- and 6. This paper, after all, is not an analy- inids (a relationship he supports) distin- sis of the other Levant hominids, and these guish it from Skhul 5, and therefore what specimens have many features not pre- Hublin saw in this specimen may not be too served in Zuttiyeh that we therefore do not different from our own understanding. 344 S. SOHN AND M.H. WOLPOFF However, with the collapse of the idea (because of an earlier date?). Moreover, nu- that there are numerous unique Skhul merous east Asian characteristics are also 5-like features shared with Zuttiyeh, we found in Skhul and Qafzeh specimens, again question what evidence exists to suggest in some more than in others. In many cases that Zuttiyeh is particularly like modern these are the features characterized as Homo sapiens. Here we differ from Vander- “modern” in earlier publications, but we meersch. In fact, some of the specific “mod- have shown them to be east Asian rather ern” features Vandermeersch cites (Table 4) than “modern,” thereby substituting an in- are not modern at all and we suspect he may traspecies clade explanation for a grade one. have confused “modern”with “non-Neander- Zuttiyeh shows the highest number of tal.”For instance, Figure 1shows the frontal Zhoukoudian-like features of all the Levant bone length, which Vandermeersch empha- specimens, perhaps because it is the earli- sized, to be both within the Zhoukoudian est; its range of estimated dates, after all, and the Levant Neandertal ranges, but at could bring it quite close to Zhoukoudian in the very edge of the Levant non-Neandertal terms of its age. Our hypothesis of a Zhou- range. This is hardly a unique “modern”fea- koudian ancestry best explains the origin of ture for Zuttiyeh. Similarly, Vandermeersch the non-Neandertal features in the Levant regards the frontal curvature index from na- hominids. The source is geographic, and sion as “modern,” while our data show this east Asian rather than African. index to be totally out of the Levant hominid These data do not obviously demonstrate range, but in the Zhoukoudian range. The that the Levant Neandertals are intrusive, orbital marginal tubercle he cites as modern refugees from the severe conditions of the is also found in the Zhoukoudian sample. We last European glaciation as Vandermeersch find none of his “modern” features to be (19811, Bar-Yosef (19881, and others have unique aspects linking Zuttiyeh to a pre-sa- suggested (see also Trinkaus, 1991). In fact piens intraspecies clade. it is clear to us that if migration into the region was the case, differences with the DISCUSSION AND CONCLUSIONS cold-adapted Neandertals of Europe suggest We conclude that Zuttiyeh preserves a that at least some of the Levant Neandertals number of intraspecies clade features from are not the refugees themselves, but rather east Asia, many of which appear in the later are their descendants. These descendants, populations of the Levant-both the Nean- in this interpretation, show evidence of mix- dertals and the so-called “modern Homo sa- ture with the local populations because piens” remains. As several have said before many of the (local) east Asian features ap- us, there is no convincing evidence to sug- pear in the Levant Neandertals. gest that Zuttiyeh is uniquely or more The (local) east Asian features we have closely related to one of these than it is to identified also appear, to a greater or lesser the other, whatever the exact level of rela- extent, in the non-Neandertal Levantine tionship exists between the later Levant sample, and we believe there is clear evi- hominid samples. Even given the most dence in these data to support the following liberal possible definition of “anatomically two contentions. (1) The Levant Neander- modern Homo sapiens” (cf. Delson, 19881, tals are not more than racially distinct from that includes archaic specimens such as the non-Neandertal populations of the re- Jebel Irhoud 1, we do not believe it is pos- gion. The evidence of interbreeding between sible to regard Zuttiyeh as “anatomically any intrusive Neandertal population and modern Homo sapiens,” unless the entire the descendants of Zuttiyeh (i.e., hybridiza- later Levant sample is also characterized tion) strongly suggests they cannot be differ- this way. ent species. (2) If we can assume that any of Clearly east Asian features are in the Le- these Levant populations (Neandertal, non- vant Neandertal sample. But the two most Neandertal, or both) are ancestral to living complete Levant Neandertal adult crania do people, the significant elements of earlier not have the same east Asian features; east Asian ancestry in the late Pleistocene Tabun retains many more than does Amud populations of the Levant demonstrates ZU’ITIYEH FACE 345 that no known Africanpopulation can be the tal characteristics in the Levant popula- unique ancestor of all modern populations. tions, but non-Neandertal is not necessarily In sum, Zuttiyeh plays an important role modern, and unless one believes that all in understanding the evolutionary pattern modern human populations have a unique in the Middle and Upper Pleistocene of the recent origin in Asia, it is clear that region Levant. It cannot be shown to uniquely un- as well as evolutionary stage could profit- derlie either a local Neandertal or a local ably be included as a valid source of Pleis- “anatomically modern” intraspecies clade in tocene human variation. the region. It seems to link both of these later Levant samples with the earlier re- ADDENDUM mains from east Asia, thereby invalidating Yoel Rak has been re-examining the de- any contention of a unique 200 kyr or younger African ancestry for modern popu- posits at the base of the Zuttiyeh cave, and lations. It would appear that features re- several years ago wrote us that he had made garded as “modern” were actually “non-Ne- an exciting new discovery. When he sent a cast of the Kebara pelvis, included in it was andertal,” in many cases because they were east Asian. It is this potential for confusing a a box labeled “Newly Discovered at the Base grade with an intraspecies clade explana- of the Zuttiyeh Cave.” In the box was a for- tune cookie. tion that in our opinion underlies the lack of earlier agreement about the relations of Zut- tiyeh-a lack of agreement at least twice ACKNOWLEDGMENTS characterizing the difference between the This paper was first prepared for the In- analysis and conclusions of the same paper. ternational Symposium on Palaeoanthro- But for this confusion, many who have re- pology in Commemoration of the 60th Anni- ported on the specimen would seem to con- versary of the Discovery of the First Skull of cur that it presents a reasonable and not Peking Man, held in Beijing (October 19 unexpected morphology for an early Levant through 24, 1989). A short version was pre- hominid. sented at the 1990 AAPA meetings. T. Sim- We add to this understanding a link with mons, A.B. Falsetti, and F.H. Smith gra- earlier east Asian morphology, especially ciously made available their unpublished as expressed at Zhoukoudian, where the material on Zuttiyeh. 0. Bar-Yosef, E. more complete specimens in the region are Trinkaus, and G. Pope and B.A. Wood pro- found. Perhaps the Late Pleistocene exten- vided a number of helpful comments and sion of the European (i.e., Neandertal) mor- suggestions. We also thank the following in- phology into the Crimea and even further dividuals who made the original specimens into central Asia, such as represented at Za- available for our study: J. Zias of the Rock- skalnaya and Teshik Tash, was a tempo- efeller Institute, C.B. Stringer of the Natu- rally limited and misleading (at least for to- ral History Museum, Y. Rak of the Univer- day’s paleoanthropologists) interruption of a sity of Tel Aviv, W.W. Howells of the different, more normal Pleistocene distribu- Peabody Museum, Harvard University, B. tion of populations. The east Asian features Vandermeersch of the Universite de Bor- in Zuttiyeh, combined with what we regard deaux, and J.-L. Heim of the Institute de as the east Asian relations of the (probably) Paleontologie Humaine, Paris. earlier Narmada vault from India, suggest a broader Pan-Asian distribution of regional LITERATURE CITED features than is normally recognized. We Bar-Yosef 0 (1988) The date of South-West Asian Nean- think that the links between Zuttiyeh and dertals. In M Otte (ed.): LHomme de Neandertal. Vol. the east Asian hominids are not expected, 3: L’Anatomie. Liege: Etudes et Recherches Arche- because there are probably more general ologiques de 1’Universite de Liege. 30:31-38. links between the Levant hominids and Bar-Yosef 0 (1989)Geochronology of the Levantine Mid- dle Paleolithic. In P Mellars and C Stringer (eds.): The many populations just to the east. It is this Human Revolution: Behavioral and Biological Per- set of east Asian features that provides the spectives on the Origins of Modern Humans. Edin- genesis of at least some of the non-Neander- burgh: Edinburgh University Press, pp. 589-610. 346 S. SOHN AND M.H. WOLPOFF

Bar-Yosef 0 (1992) The role of Western Asia in modern Mount Carmel: The Human Remains from the Leval- human origins. Philos. Trans. R. SOC.Lond. [Biol.] lois-Mousterian. Vol. 11. Oxford: The Clarendon Press. 337:193-200. RakY (1986)The Neanderthal: a new look at an old face. Coon C (1963) The Origin of Races. New York: Knopf. J. Hum. Evol. 15(31:151-164. Copeland L, and Hours F (1983) Le Yabroudien #El- Rak Y (1990) On the difference between two pelvises of Kowm (Syrie) et sa place dans le Paleolithiques du Mousterian context From the Qafzeh and Kebara Levant. Paleorient 9:21-38. Caves, Israel. Am. J. Phys. Anthropol. 18:323-332. Delson E (1988) One source not many. Nature 318:107- Schwarcz H, Goldberg P, and Blackwell B (1980) Ura- 108. nium series dating of archaeological sites in Israel. J. Dobzhansky T (1944)One species and races of living and Earth Sci. 29:157-165. fossil man. Am. J. Phys. Anthropol. 2(31:251-265. Simmons T, Falsetti AB, and Smith FH (1991) Frontal Etler, DA (1990) A case study of the "erectus"-"sapiens" bone morphometrics of Southwest Asian Pleistocene transition in Asia: Hominid remains from Hexian and hominids. J. Hum. Evol. 20:249-269. Chaoxian Counties, Anhui Provence, China. Kroeber Smith FH (1985) Continuity and change in the origin of Anthropological Society Papers 71-72:l-19. modern Homo sapiens. 2. Morphol. Anthropol. Garrod D (1962)The Middle Paleolithic of the Near East 75:197-222. and the problem of Carmel men. J. R. Anthropol. Inst. Stringer C (1985) Middle Pleistocene hominid variabil- 92:232-259. ity and the origin of Late Pleistocene humans. In E Gisis I, and Bar-Yosef 0 (1974) New excavation in Zut- Delson (ed.): Ancestors: The Hard Evidence. New tiyeh Cave, Wadi Amud, Israel. Paleorient 2:175-180. York: Alan R. Liss, Inc. pp. 289-295. Gould SJ (1989) Grimm's greatest tale. Nat. Hist. Stringer C (1990) The emergence of modern humans. 2/89:20-28. Sci. Am. 263(61:98-104. Grun R, and Stringer CB (1991)Electron spin resonance Stringer C, and Andrews P (1988) Modern human ori- dating and the evolution of modern humans. Archae- gins. Science 241:773-774. ometry 33(2):153-199. Stringer C, Hublin J-J, and Vandermeersch B (1984) Habgood, PJ (1992) The origin of anatomically modern The origin of anatomically modern humans in western humans in East Asia. In G Brauer and FH Smith Europe. In FH Smith and F Spencer (eds.): The Ori- (eds.): Continuity or Replacement? Controversies in gins of Modern Humans: A World Survey of the Fossil Homo supiens Evolution. Rotterdam: Balkema, pp. Evidence. New York: Alan R. Liss, Inc., pp. 51-135. 273-288. Suzuki H, and Takai F (eds) (1970) The Amud Man and Holt BM, and Waddle DM (1991) The affinities of the His Cave Site. Tokyo: University of Tokyo Press. Zuttiyeh frontal reconsidered (abstract). Am. J. Phys. Trinkaus E (1983) The Shanidar Neandertals. New Anthropol. [Suppl.] 12:95. York: Academic Press. Howell FC (1953) Upper Pleistocene men of the South- Trinkaus E (1984) Western Asia. In FH Smith and F west Asian Mousterian. In GHR von Koenigswald Spencer (eds.): The Origins of Modern Humans: A (ed.): Hundert Jahre Neanderthaler. Utrecht: Ke- World Survey of the Fossil Evidence. New York: Alan mink en zoon, pp. 185-198. R. Liss, Inc., pp. 251-293. HrdliEka A (1930) The skeletal remains of early man. In Trinkaus E (1989) Issues concerning human emergence Smithsonian Institution Annual Report for 1928. pp. in the later Pleistocene. In Trinkaus E (ed.): The 593-623. Emergence of Modern Humans: Biological Adapta- Hublin J-J (1976) L'homme de Galilee, Memoire de tions in the Later Pleistocene. Cambridge: Cambridge D.E.A. de Paleontologie.Paris: Universite de Paris VI. University Press, pp. 1-17. Hublin J-J (1983) Les origins de l'hommes de type mod- Trinkaus E (1991) Les Hommes Fossile de la Grotte de erne en Europe. Science 62t62-71. Shanidar, Irak: Evolution et Continuite parmi les Hublin J-J (1987) Qui fut l'Ancestre de l'Homo sapiens? Hommes Archaiques Tardifs du Proche Orient. Science 113:2&35. L'Anthropologie (in press). Huxley JS (1958) Evolutionary processes and taxonomy, Turville-Petre F (ed.) (1927) Researches in Prehistoric with special reference to grades. Uppsala Univ. Galilee, 1925-26. London: Council of the British Arssks. 1958:21-38. School of Archaeology in Jerusalem. Jelinek A (1981)The Middle Paleolithic in the Southern Vandermeersch B (1981) Les premiers Homo supiens au Levant from the perspective of the Tabun Cave. In J Proche Orient. In D Ferembach (ed.):Le Processus de Cauvin and P Sanlaville (eds.): Prehistoire du Levant. l'hominisation. Paris: Centre National de la Recher- Paris: Centre National de la Recherche Scientifique, che Scientifique, pp. 97-100. pp. 265-280. Vandermeersch B (1989) The evolution modern hu- Keith A (1927)A report on the Galilee skull. In Turville- mans: Recent evidence from Southwest Asia. In P Petre F (ed.): Researches in Prehistoric Galilee, 1925- Mellars and C Stringer (eds.):The Human Revolution. 1926. London: Council of the British School of Archae- Edinburgh: Edinburgh University Press, pp. 155-164. ology in Jerusalem, pp. 593423. Weidenreich F (1943) The skull of Sinanthropus Pekin- Keith A (1931) New Discoveries Relating to the Antiq- ensisr a comparative study of a primitive hominid uity of Man. London: Williams and Norgate. skull. In PaleontoloPia Sinica (n.s.D) 10 (Whole Series McCown TD, and Keith A (1939) The Stone Age of 127). Beijing: Geological Survey of China. ZUTTIYEH FACE 347

Wolpoff MH (1980) Paleoanthropology. New York Knopf. Woo J (1966) The skull of Lantian man. Curr. Anthro- Wolpoff MH, Smith FH, Malez M, Radovcic J, and pol. 7t83-86. Rukavina D (1981) Upper Pleistocene human remains Wu R, and Dong X (1982) Homo erectus in China. In R from Vindija Cave, Croatia, Yugoslavia. Am. J. Phys. Wu and JW Olsen (eds.):Paleoanthropology and Pale- Anthropol. 54r499-545. olithic Archeology in the People’s Republic of China. Wolpoff MH, Wu X, and Thorne AG (1984)Modern homo New York: Academic Press, pp. 79-89. supiens origins: A general theory of hominid evolution Wu X, and Wu M (1985) Early Homo sapiens in China. involving the fossil evidence from East Asia. In FH In R Wu and JW Olsen (eds.): Paleoanthropology and Smith and F Spencer (edsj: The Origins of Modern Paleolithic Archaeology in the People’s Republic of Humans: A World Survey of the Fossil Evidence. New China. New York: Academic Press, pp. 91-106. York: Liss, pp. 411483.