Architecture of the Nasal Complex in Neanderthals: Comparison With Other Hominids and Phylogenetic Significance



Although paranasal sinus configuration has occasionally been the focus in analyses of the phylogenetic relationships of various primates, other elements of the region of the nasal fossa—in particular, the turbinals—have received far less attention. A preliminary study of Neanderthal cranial morphology revealed the presence of an apparently unique configuration of the lateral wall of the nasal cavity: namely, in the region in which in Homo sapiens the anterior extremity of the maxilloturbinal (also referred to as the inferior nasal concha) articulates with the internal surface of the maxilla along a relatively anteroposteriorly long and essentially horizontally oriented conchal crest, there exists a vertically oriented thickening that protrudes medially into the nasal cavity (Schwartz and Tattersall, Proc Natl Acad Sci USA 1996; 93:10852–10854). Subsequent citations of this report either claimed that this “medial projection” in Neanderthals is merely an enlarged maxilloturbinal or mistakenly identified as this structure the base of a maxilloturbinal that had fused to the lateral wall of the nasal cavity and subsequently broken off. In light of the potential significance that any novel configuration of the nasal complex architecture may have for elucidating hominid evolution, we present here a comparative overview of this region in fossil and extant large-bodied hominoids, and demonstrate that Neanderthals do indeed possess a configuration that is unique among hominids. Anat Rec, 291:1517–1534, 2008. © 2008 Wiley-Liss, Inc.

The nasal and paranasal region in mammals and in primates in particular has been the occasional focus of systematic analysis. Such analysis has revealed that some paranasal sinuses are uniquely represented in primates. For instance, maxillary sinuses develop only in extant anthropoid primates (New and Old World monkeys, apes, and humans and their fossil relatives) (Figs. 1–3). Ethmoidally derived frontal sinuses occur in modern humans but not necessarily in all species of the genus Homo [viz. Homo antecessor, as seen in Gran Dolina ATD6-15 (Arsuaga et al., 1999)]. They may not occur in australopiths (see later); and at least in Gorilla and the common chimpanzee they derive from a different region of the ethmoid than in humans (Cave and Haines, 1940; Cave, 1961, 1973). Frontal sinuses have been identified in the fossil anthropoids Aegyptopithecus (early Oligocene) and Proconsul (Miocene), but whether they communicate with an ethmoidal sinus is unknown (Simons and Rasmussen, 1989). Furthermore, polyfollicular ethmoidal sinuses are found in both modern Homo sapiens and in some fossil specimens allocated to other species of the genus Homo (Schwartz and Tattersall, 2002, 2003, 2005), although not in all (e.g. ATD6-15). Less extensively, they are seen in African apes (Cave and Haines, 1940; Cave, 1961, 1973).

Figure 1.

Top: Lagomorph (rabbit; University of Pittsburgh collections). Bottom: A New World monkey, Cebus sp (AMNH l “F”) in which the maxilloturbinal has fused to the inner wall of the nasal cavity. NT, nasoturbinal; ET, ethmoturbinal; MT, maxilloturbinal; CC, conchal crest; SS, sphenoidal sinus; MS, maxillary sinus (meatus of); “FS”, nonethmodially derived frontal sinus/es. Not to scale. Photographs © J. H. Schwartz.

Among anthropoid primates, expanding maxillary sinuses may invade the hard palate (as in chimpanzees, Fig. 3) or may course superiorly through the interorbital region and possibly also partway across each supraorbital region (as in orangutans and at least some of their fossil relatives, e.g. the middle-late Miocene Sivapithecus) (Cave and Haines, 1940; Schultz, 1968; Cave, 1973; Ward and Pilbeam, 1983). CT scans of the australopith cranium Sts 5 from the South African late Pliocene site of Sterkfontein indicate that this hominid (and by implication other australopiths) also developed superiorly expanded maxillary sinuses (Weber et al., 2002). Among hominids, extensive frontal sinus expansion is noted in the fairly complete cranium from Kabwe (Zambia) and to an even greater degree in the complete cranium from Petralona (Greece), both roughly late Pleistocene in age (Seidler et al., 1997).

Other elements of the hominoid nasal region, especially the various mucous membrane-covered bony turbinals (also known as turbinates or conchae) and associated crests, have also been described and discussed in terms of their systematic relevance, although typically with less attention to detail than is the case with the sinuses. Some of these nasal fossa structures develop as lateral outgrowths or extensions of the ethmoid (and are identified generally in mammals as ethmoturbinals or ethmoturbinates), whereas others—the maxillary (or ventral) and nasal (or frontal) turbinals—develop from their own centers of ossification and are thus separate and independent bones in the nasal cavity until and unless they fuse to the bone/s with which they articulate (and from which as a result they derive their names) (e.g. Figs. 1–3) (Cave and Haines, 1940; Sisson and Grossman, 1953; Romer, 1964; Cave, 1973; Starck, 1975; de Beer, 1985; Schwartz and Tattersall, 1996).

Figure 2.

Extant Homo sapiens. Top: left and right sides of the same specimen (AMNH VL/4578); Bottom. San AMNH VL/2464 (left), and AMNH 99/76 (right). FS, ethmoidally derived frontal sinus; SC, spinal crest; SS, sphenoidal sinus; asterisk, region of sunken and concave wall of nasal cavity. Not to scale. Photographs © J. H. Schwartz.

Figure 3.

Top left: juvenile Pan (AIMZ 12497) showing lack of both a maxilloturbinal and a conchal crest (because the MT was poorly developed and free bone). Top right: adult Pan (AMNH CH3 L.15) showing a well-developed conchal crest for articulation with the still-free MT; also note maxillar sinus-derived air cells in palate. Bottom: adult Pongo showing (left) patent articulation between the MT and CC (AIMZ 12515), and (right) complete coalescence of the MT with the wall of the nasal cavity (AIMZ 12509). Not to scale. Photographs © J. H. Schwartz.

Because most mammals develop three or four ethmoturbinals (Romer, 1964), it would be in those taxa that develop fewer that we might impute systematic importance to ethmoturbinal number. In this regard, it is of potential systematic interest that anthropoid primates develop only two ethmoturbinals and the Southeast Asian island-dwelling tarsier develops only one (Starck, 1975, 1984). These character states can be interpreted respectively as further evidence of (1) the phylogenetic unity (= monophyly) of anthropoid primates (i.e. the last common ancestor of anthropoid primates would have had a reduced number of ethmoturbinals) and (2) the morphological uniqueness of Tarsius (at least in contrast to all other extant primates) (Schwartz, 1986, 2003).

As no mammal possesses more than one pair of nasal (frontal) turbinals and one pair of maxillary turbinals (Romer, 1964; Starck, 1975), character–state differences will be reflected either in the modification of these structures or, most dramatically, in their absence. Consequently, it is of potential phylogenetic significance that, among primates, the nasoturbinal is represented only by a ridge in prosimians, and is absent altogether in anthropoids (Starck, 1975). Taken together, these character states can be interpreted as reflecting (1) the monophyly of a group Primates that includes at least all extant prosimians and anthropoids (via reduction of the nasoturbinal) and (2) within Primates, the monophyly of an anthropoid clade united by nondevelopment of the nasoturbinal.

With regard to the maxilloturbinal, curiously little attention has been paid to its morphological expression among mammals beyond noting its presence—if that (Cave and Haines, 1940; Sisson and Grossman, 1953). The most detailed descriptions of this bone come from human anatomy and osteology texts (Clemente, 1984; Schwartz, 2007), in which the general timing of its growth, from onset to completion of ossification, and its ultimate articulations within the nasal cavity with the maxilla, palatine, lacrimal, and ethmoid are presented. The body of the human maxilloturbinal is anteroposteriorly longer than superoinferiorly tall, is essentially horizontally oriented, and partially covers (medially) the meatus of the maxillary sinus, which opens into the nasal cavity above this turbinal (Schwartz, 2007) (Fig. 2). The human maxilloturbinal contacts a portion of the inferior edge of the lacrimal and ethmoid via a superiorly distended process that is only modestly long anteroposteriorly; its more inferiorly situated articulations with the maxilla and palatine are notably longer, and occur along distinctly raised, variably horizontal crests (i.e., the maxillary and palatine conchal crests) that develop along the margin of contact (Schwartz, 2007). The inferomedial extremity of the maxilloturbinal protrudes variably into the nasal cavity and may be pneumaticized and externally rugose (sometimes markedly so in both respects) (Fig. 2). In some human populations (e.g. the San), rather than being typically flat or mildly sunken the lateral wall of the maxillary contribution to the nasal cavity may be notably concave (JHS, personal observations; Fig. 2).

Typically as human and nonhuman mammals age, not only do the cranial (especially splanchnocranial) sutures fuse or synostose but so also does the maxilloturbinal with the bones it contacts (Figs. 1–3). Consequently, without knowing the developmental history of the maxilloturbinal one could incorrectly conclude that a maxilloturbinal that had fused to the lateral wall of the nasal cavity was an outgrowth of the maxillary contribution to the lateral wall of the nasal cavity—and thus at least analogous, if not serially or otherwise homologous—to the ethmoturbinals (which do bud from the ethmoid). This might be the case even though, spatially, the ethmoturbinals are associated with a structure (the ethmoid) that lies centrally within the nasal cavity, whereas the maxilloturbinals eminate from the lateral wall of the nasal cavity (Schwartz, 2007). The false impression that there is some kind of homologous relationship between the ethmoturbinals and the maxilloturbinal is compounded by the language typically used in textbooks and the scholarly literature describing human skeletal morphology, in which the two ethmoturbinals are by tradition referred to as the “superior and middle nasal conchae” and the maxilloturbinal as the “inferior nasal concha” (Clemente, 1984). However, although it may be true that all of these bony protrusions within the nasal cavity are scroll-shaped to some extent (quite intensively so in mammals such as carnivores), and thus can technically be described as conch-like or conchal, it remains true that their developmental histories, as reviewed earlier, are entirely different. Because in mammalian systematics in general (and especially in paleoanthropology) the tradition has been to generate theories of homology and relatedness on the basis of morphological comparisons of specimens of adult individuals, without consideration of the ontogenetic basis of the structure being compared (Schwartz, 1999; Schwartz and Tattersall, 2000), there has thus been much opportunity for error in interpreting the phylogenetic significance of the features under consideration.

Of particular note in this regard is the misunderstanding of a novel configuration of the lateral wall of the nasal cavity that two of us originally identified in specimens of Homo neanderthalensis. As we first noted in the Gibraltar 1 Neanderthal partial cranium (Schwartz and Tattersall, 1996), the region that in H. sapiens bears a typically well-defined and essentially horizontal maxillary conchal crest is here configured entirely differently. In the Neanderthal this region bears a vertically oriented, sometimes terminally superoinferiorly crest-like structure that swells medially midway along its vertical axis. We referred to this structure as a “medial projection” and suggested that it was unique to Neanderthals in comparison not only with other hominids but also with mammals in general.

Subsequent to the initial description and discussion of this nasal cavity structure, we have continued to investigate its representation among hominids. We have done this not only to determine its range of variation and to test our interpretation of the significance of its presence in Homo neanderthalensis as an apomorphy unique to this hominid species but also to address various rejections of our original interpretation of the medial projection as a unique Neanderthal apomorphy. Such objections have, ironically, been based on two diametrically opposed claims: either that this feature in Neanderthals is in fact homologous with the mammalian maxilloturbinal (Churchill et al., 1999; Franciscus, 1999a, b) or that modern humans and non-Neanderthal fossil hominids also develop medial projections (Arsuaga et al., 1997; Frayer et al., 2006; Wolpoff et al., 2006). Neither of these objections is sustainable on closer examination of the evidence.


Fossil specimens reported here are listed in Table 1 and include: Plio-Pleistocene australopiths (n = 8), “modern” Homo sapiens from Europe (n = 5), Homo neanderthalensis (n = 9), Homo spp. from Africa (n = 6), and Homo spp. from Europe (n = 5), Homo spp. from Asia (n = 3). These fossils have been studied over the years as part of our assessment of virtually the entire fossil record, which can be found in our recently completed compendia (see Schwartz and Tattersall, 2002, 2003, 2005). Fossils chosen for this study were based on their exhibiting a relatively intact nasal region. Putative fossil hominids from the Middle Awash, Ethiopia, and Chad are not included, as we did not have access to them. Our extensive, ongoing examination of craniodental material of extant humans (n = 500) from globally representative populations (e.g., Africa, Europe, Asia) were studied in the collections of the division of Anthropology, American Museum of Natural History (AMNH). Also included in our comparative sample are adult, subadult, and infant crania of extant hominoids from the extensive collections of the divisions of Mammalogy and Anthropology, AMNH (Pongo, n = 14; Pan, n = 35; Gorilla, n = 21; hylobatids, n = 52) and that of the Anthropological Institute and Museum, Zürich (AIMZ) (Pongo, n = 18; Pan, n = 17; Gorilla, n = 15). Old and New World monkeys (Cercopithecus, n = 29; Macaca, n = 19; Mandrillus, n = 12; Papio, n = 8; Cebus, n = 20; Ateles, n = 19) and a diversity of nonanthropoid crania (e.g., lagomorphs, insectivores, artiodactyls, equids, carnivores) were also studied at the divisions of Mammalogy and Anthropology, AMNH, and University of Pittsburgh Department of Anthropology collections.

Table 1. Localities of fossil hominid specimens cited in the text, with ages
  1. Ma, million years (ago); Ka, thousand years (ago). Unreferenced ages are very broadly extrapolated from archaeological or faunal associations.

Modern Homo sapiens, Europe
 Mladec∼31 Ka(Wild et al., 2006)
 Cro-Magnon<30 Ka 
 Abri Pataud22 Ka(Bricker and Mellars, 1987)
 Chancelade∼17–14 Ka 
 Grimaldi>25 Ka 
Homo neanderthalensis:  
 Engis 2∼60–40 Ka 
 Gibraltar Forbes' QuarryN/A 
 Krapina130 Ka(Rink, 1995)
 La Chapelle-aux-Saints∼50 Ka(Grün and Stringer, 1991)
 La Ferrassie 1<70 Ka 
 La Quina∼65 Ka 
 Peche de l'Azé55–45 Ka 
 Roc de Marsal>50 Ka 
 Subalyuk∼70–60 Ka 
Homo spp., Africa:  
 Koobi Fora1.9 Ma(Brown and Feibel, 1986)
 Nariokotome1.6 myr(Brown and McDougall, 1993)
 Olduvai Gorge (Bed I, II)1.8–1.5 Ma(Tamrat et al., 1995)
 Jebel Irhoud>130 Ka(Grün and Stringer, 1991)
 Eliye SpringsN/A 
 Ndutu<400 Ka(Leakey and Hay, 1982)
Homo spp., Europe  
 Arago∼450 Ka(Iacumin, 1996)
 Petralona<250 Ka?(Grün, 1996)
 Sima de los Huesos>400 Ka(Bischoff et al., 2003)
 Gran Dolina∼780 Ka(Pares and Pérez-González, 1995)
Homo spp., Asia:  
 Jebel Qafzeh∼92 Ka(Valladas et al., 1988)
 Jinniushan>200 Ka(Chen et al., 1994)
 Liujiang>40 Ka(Yuan et al., 1986)
 Drimolen2.0–1.5 Ma(Keyser et al., 2000)
 Hadar>3.18 Ma(Kimbel et al., 2004)
 Koobi Fora1.7 Ma(Brown and Feibel, 1986)
 Lomekwi2.5 Ma(Walker et al., 1986)
 Makapansgat∼3.0 Ma(McKee et al., 1995)
 Sterkfontein M4 (Sts, StW)2.8–2.6 Ma(Kuman and Clarke, 2000)
 Swartkrans (SK)1.8–1.5 Ma(Brain, 1993)
 Taung∼2.7 Ma(McKee et al., 1995)


A Brief Overview of the Maxilloturbinal in Extant Large-Bodied Hominoids

Regardless of their exact phylogenetic interrelationships, it is fairly certain that the extant large-bodied hominoids [humans, orangutans (Pongo), gorillas (Gorilla), and chimpanzees (Pan)] constitute a monophyletic group (Schwartz, 1986). Consequently, prior to turning to fossil hominids, it is appropriate to review the configurations of the maxilloturbinal and its relation to the lateral wall of the nasal cavity in these extant primates.

Although the timing of onset and of completed ossification of the maxilloturbinal has yet to be determined for the great apes, these primates clearly conform to the pattern seen in Homo sapiens in which a bone arising from its own center of ossification eventually coalesces with elements of the lateral wall of the nasal cavity (Figs. 2 and 3). In Pongo (Fig. 3), as is also the case in Pan (JHS personal observations), the maxilloturbinal is anteroposteriorly longer than tall, and essentially horizontal. It presents a variably long and pointed anterior extension that articulates with the maxilla along a conchal crest; the posterior articulation of this bone with the palatine appears to be more variable, and may be somewhat squared-up rather than tapering. The anterior extremity of the maxilloturbinal (and thus also of the maxillary conchal crest) is situated relatively close to the margin of the nasal aperture in Pan, and quite posteriorly in Pongo (Fig. 3). The maxilloturbinal of Gorilla (Fig. 4) differs from that of the other large-bodied hominoids in being generally much taller superoinferiorly throughout its length, and in having a more vertically oriented and rapidly descending anterior margin whose anteroinferior terminus may be only slightly extended, if at all; thus, no distinct, horizontal maxillary conchal crest is present. As observed in some specimens of Gorilla the maxilloturbinal, when viewed through the nasal aperture may swell medially into the nasal cavity (Fig. 4). But as can be seen in various specimens (e.g. Fig. 4), such swelling results not from a thickening of the bone of the maxilloturbinal but from the development of a smoothly curved flexure (convex medially and concomitantly concave laterally) of the superior portion of the maxilloturbinal. This flexure begins quite anteriorly and extends posteriorly for most if not all of the length of the bone.

Figure 4.

Gorilla. Top left: juvenile, showing partially medially inflated MT that has not yet fully coalesced with the wall of the nasal cavity (AIMZ 6994). Top right: adult showing longitudinal swelling of the MT (AIMZ 12502). Bottom: adult showing steep slope of anterior margin of MT (AIMZ 12501). SMT, swollen maxilloturbinal. Not to scale. Photographs © J. H. Schwartz.

The Region of the Maxilloturbinal in Neanderthals

As the focus of this communication is nasal complex architecture in Neanderthals, we begin with the morphology of this hominid in order to compare it with that of other hominids. As mentioned earlier, the structure we refer to as a “medial projection” was first observed in the Gibraltar 1 partial adult cranium (Schwartz and Tattersall, 1996) (see Table 1 for the approximate age of this specimen and others discussed later). In this specimen, just posterior to the margin of the nasal aperture (the lateral margin) (Gower, 1923) lies a thick crest that arises lateral to the base of the anterior nasal spine (the spinal crest of Gower, 1923). This crest intrudes into the space of the nasal cavity until about midway up its vertical length, beyond which its prominence progressively diminishes superiorly. This structure is present bilaterally in the nasal cavity of the Gibraltar 1 specimen (Fig. 5). In this fossil there is no trace discernible of a structural counterpart of a maxillary conchal crest, which would be expected to lie horizontally and thus at a right angles to this vertically oriented bony prominence.

Figure 5.

Specimens of adult Neanderthals. Top left and middle: Gibraltar 1. Top right and bottom right: La Chapelle-aux-Saints. Bottom left and middle: La Ferrassie 1. MP, medial projection. Not to scale. Photographs © J. H. Schwartz.

The La Ferrassie 1 adult cranium also preserves, bilaterally, vertical swellings that project medially from the lateral walls of the nasal cavity (Fig. 5). These projections are blunter and less protrusive than those in the Gibraltar 1 specimen, and the projection on the left is not as deep anteroposteriorly as the one on the right. The nasal cavity of the Krapina 3 (= C) partial adult cranium, although incomplete, also presents evidence bilaterally of vertical structures emanating from the lateral walls of the nasal cavity, while the somewhat more complete Guattari 1 (= Monte Circeo) adult cranium preserves on the left side an apparently polyfollicular structure that can plausibly be interpreted as the remains of a medial projection (Schwartz and Tattersall, 2002). As with the Gibraltar 1 partial cranium, none of these specimens presents evidence of any structure that might be identified as a maxillary conchal crest. However, the virtually complete cranium from La Chapelle-aux-Saints does exhibit such traces (Fig. 5).

The La Chapelle-aux-Saints cranium preserves both frontal processes, each of which bears internally a blunt and bulbous swelling. On the left side the swelling is much lower than in the La Ferrassie 1 cranium (and thus than in the Gibraltar 1 cranium and probably also the Krapina 3 specimen); and, as in La Ferrassie, its surface lacks addition structural detail. However, on the right side of the La Chapelle-aux-Saints cranium, the low swelling bears an inferiorly and somewhat posteriorly oriented sheet of bone that partially overlaps it; the jagged inferior margin of this structure suggests that its full extent is not preserved. We interpret this sheet-like extension of bone as representing a portion of the maxilloturbinal, which in an earlier report we were unable to situate with confidence in the Neanderthal nasal cavity (Schwartz et al., 1999). If, as we believe, this is part of the maxilloturbinal, the orientation of this ontogenetically once-separate bone is not horizontal but, rather, obliquely up and back. The La Chapelle-aux-Saints specimen differs from the Gibraltar 1 specimen in lacking a spinal crest and in this regard is similar to the other Neanderthal specimens discussed earlier.

Further evidence suggesting the presence of a maxilloturbinal in Neanderthals comes from study of two fragmentary juvenile crania (perhaps 3–4 years old at death) from the sites of Engis and Subalyúk (Fig. 6). The preserved right side of the anterior portion of the nasal cavity of both specimens presents a vertically oriented and medially projecting swelling (much more pronounced in the Subalyuk child) that terminates inferiorly in a ridge-like structure. In the Engis specimen, this feature is vaguely “U” shaped. In the Subalyuk specimen, the posterior aspect of this structure appears to slant upward.

Figure 6.

Specimens of juvenile Neanderthals. Top left: Roc de Marsal. Top middle and right: Engis. Bottom left: Peche de l'Azé. Bottom middle: La Quina. Bottom right: Subalyuk (internal aspect of left maxillary frontal process). Note the verticality and lack of anteroposterior length of the MP. Not to scale. Photographs © J. H. Schwartz.

Other immature Neanderthal specimens display some degree of development of a vertically oriented medial swelling or projection but lack evidence of a maxillary conchal crest. A well-defined vertical swelling is seen on the preserved left side of the nasal cavity of the Roc de Marsal juvenile cranium, whereas a lower swelling is present on the preserved left sides of the nasal cavities of both the Peche de l'Azé and La Quina crania (Fig. 6). Significantly, none of these juvenile Neanderthal specimens possesses a spinal crest; and this, given the absence of this crest in all adult specimens discussed earlier with the exception of Gibraltar 1, leads to the conclusion that the vertically oriented structure we identify as a medial projection is a unique outgrowth of the inner wall of the maxillary frontal process.

The Region of the Maxilloturbinal in Other Fossil Hominids

Among Plio-Pleistocene hominid fossils frequently allocated to species of either Australopithecus or Paranthropus—which we will refer to collectively as “australopiths”—there is a substantial number of specimens that retain at least the anterior part of the nasal cavity on one or both sides. Preservation in these specimens is adequate for us to be able to confirm the absence of structure in this region, although not necessarily the presence of structure. Specifically, from South Africa, the Taung 1 partial child's cranium, Drimolen DNH 7, Makapansgat MLD 6, Sterkfontein Sts 5, Sts 71, StW 505 and StW 53b, and Swartkrans SK 12, SK 48 and SK 52 all present evidence of neither a maxilloturbinal crest nor a medial swelling of any sort (see Figs. 7 and 8). Likewise, there is no morphology—of either a conchal crest or medial swelling—on the partially preserved left side of the anterior nasal cavity of Sterkfontein StW 183a, which is broken partway up. From East Africa, the partial left maxilla AL 486-1 from Hadar, Ethiopia, and the Koobi Fora KNM-ER 406 and Lomekwi KNM-WT 17000 crania from deposits respectively on the east and west shores of Lake Turkana, Kenya, as well as the Olduvai Gorge (Tanzania) hominid cranium OH 5, also lack evidence of either a conchal crest or medial swelling in the preserved anterior part of the nasal cavity (see Figs. 7 and 8). Of particular note in these specimens is that the nasal bones do not project away from the face, and that the frontal processes lie essentially in the same plane as the nasal bones (that is, facing anteriorly rather than laterally). This latter “australopith” configuration is unlike that of Homo sapiens and of fossils commonly allocated to other species of Homo; but it is similar to the configuration in fossil and extant nonhominid hominoids.

Figure 7.

Various australopiths. Top left: Hadar AL 486-1. Top middle: Taung 1. Top right: Makapansgat MLD 6. Bottom right: Swartkrans SK 12. Bottom middle: Swartkrans SK 4. Bottom right: Sterkfontein Sts 71. Note absence of morphology on the internal side of the maxillary frontal process, MFP. Not to scale. Photographs © J. H. Schwartz.

Among specimens that have been referred to the genus Homo, but not to either the species H. sapiens or H. neanderthalensis, we also encounter an absence of structure in the preserved anterior portion of the nasal cavity. Among Plio-Pleistocene specimens in this category, Olduvai Gorge OH 24 (typically assigned to H. habilis) (Fig. 8), Koobi Fora KNM-ER 1470 (which, until it was recently assigned to the genus Kenyanthropus, was regarded either as representing H. habilis or as the holotype of H. rudolfensis) (Schwartz and Tattersall, 2003), Koobi Fora KNM-ER 3733 (variably regarded as either H. erectus or H. ergaster) (Fig. 9), and Swartkrans SK 847 (usually regarded as representing H. erectus) (Fig. 9), there is no evidence of either a maxillary conchal crest or a medial swelling of any kind. In Nariokotome KNM-WT 15000 from Lake Turkana, Kenya, right and left sides of the anterior portion of the nasal cavity are reasonably preserved (Fig. 8). On the left is evidence neither of a maxilloturbinal crest nor of a medial swelling. On the right side, however, is an anteroposteriorly oriented low and roughened area of bone that plausibly provides evidence of a maxilloturbinal that had broken away from the bone of the lateral wall of the nasal cavity to which it had already fused. Of note in most of these specimens is that the nasal bones do project away from the face and the frontal processes of the maxillae are strongly oriented laterally. Only KNM-ER 1470 retains the apparently primitive condition in which the nasal bones, which do not project away from the face, lie in the same plane as the maxillary frontal processes, which face anteriorly rather than laterally (Schwartz and Tattersall, 2003).

Figure 8.

Various australopiths and specimens allocated to Homo. Top left: Koobi Fora KNM-ER 406. Top middle: Olduvai OH 5. Top right: Lomekwi KNM-WT 17000. Bottom left: Olduvai OH 24. Bottom middle: Nariokotome KNM-WT 15000. Bottom right: Koobi Fora KNM-ER 1813. Note absence of morphology on the internal side of the MFP. Not to scale. Photographs © J. H. Schwartz.

Figure 9.

Various specimens allocated to Homo. Top left: Swartkrans SK 847. Top middle: Koobi Fora KNM-ER 3733. Top right: Kabwe. Bottom left and middle: Arago 21. Bottom right: Petralona. With the exception of Petralona, note absence of morphology on internal surface of MFP. Not to scale. Photographs © J. H. Schwartz.

Among middle-late Pleistocene hominid specimens that preserve the anterior portion of the nasal cavity and which are sometimes regarded as representing Homo heidelbergensis, the right and left sides of the partial cranium Arago 21 (France) (Fig. 9) is devoid of notable morphology, whereas the right side of the Petralona skull (Greece) (Fig. 9) presents a well-defined, anteroposteriorly relatively long, and essentially horizontal maxillary conchal crest that is remarkably similar to that observed in H. sapiens. In the Kabwe cranium (Fig. 9), a maxillary conchal crest is better delineated on the right side of the lateral wall of the nasal cavity than on the left. As in Arago 21, the reconstructed cranium from Jinniushan (China), which preserves the left anterior portion of the nasal cavity, presents neither a crest nor a swelling (Schwartz and Tattersall, 2003).

Arsuaga et al. (1997) do not provide information on the presence or absence of either maxilloturbinals or maxillary conchal crests in their specimens from Sima de los Huesos (Atapuerca = AT), northern Spain. They do, however, mention that the Sima Hominid (SH) crania 5 and 6, as well as Sima maxillary fragments AT-767+AT-963 and AT-1217, lack evidence of a “medial projection” or any swelling of the lateral wall of the nasal cavity anteriorly. On the other hand, they describe Sima maxillary fragments AT-1665 and AT-638+AT-772 as presenting a “light swelling” and Sima maxillary fragments AT-1100+AT-1111+AT-1197+AT-1198 and AT-1666 as showing a “more marked” projection (p 268). Accompanying photographs of three of these specimens reveal that SH 6 preserves on the left side a maxilloturbinal that is fused to the lateral wall of the nasal cavity and which, as in extant Homo sapiens, arcs somewhat medially into the space of the nasal cavity before descending toward the nasal cavity floor (Arsuaga et al., 1997, Figure 6). On the preserved left side of AT-767+AT-963 (Figure 11) there does indeed appear to be in frontal view a slight swelling on or just behind the left margin of the nasal aperture. AT-1100+AT-1111+AT-1197+AT-1198 preserves an intact maxillary frontal process on the right side of the nasal aperture, and a partial and glued frontal process on the left (Figure 11). In frontal view, the left frontal process presents a swelling that protrudes into the space of the nasal cavity. Because at least the base and part of a downwardly oriented maxilloturbinal (or perhaps just a naturally short maxilloturbinal) are clearly preserved on the right side of this specimen, this swelling can plausibly be interpreted as the broken base of a maxilloturbinal that had fused to the lateral wall of the nasal cavity. In comparing their Sima specimens to the Petralona cranium Arsuaga et al. (1997) correctly describe the latter as lacking medial projections from the lateral walls of the nasal cavity but do not mention the presence of the maxillary conchal crest (see earlier).

Younger Pleistocene specimens that minimally preserve the anterior portion of the nasal cavity and which have been identified as “archaic H. sapiens” do not, however, present a consistent configuration. In one of the partial crania from Jebel Irhoud, Morocco (JH 1) (Fig. 10), there is a distinct maxillary conchal crest present on both the right and left sides of the nasal cavity; the other cranium from this site does not preserve this region. The preserved left sides of the nasal cavities of the partial crania from Eliye Springs and Ndutu, Kenya (both Fig. 10) present a configuration that can plausibly be interpreted as the base of a now-broken maxilloturbinal that had fused to the inner wall of the maxilla, whereas the preserved anterior portion of the left side of the nasal cavity of the partial cranium LH 18 from Ngaloba, Tanzania (Fig. 10) presents a smooth surface that is devoid of a crest or any other notable morphology.

Figure 10.

Various specimens allocated to Homo. Top left: Jebel Irhoud 1. Top right: Eliye Springs. Bottom left: Ndutu. Bottom right: internal surface of MFP of Ngaloba (LH 18). MT?: plausible base of broken MT. Not to scale. Photographs © J. H. Schwartz.

Specimens that have at one time or another been referred to either as “archaic Homo sapiens,” H. neanderthalensis, “anatomically modern H. sapiens,” or just H. sapiens, do not present a consistently uniform configuration of the anterior portion of the nasal cavity. Such specimens do not, however, differ among themselves simply in either possessing or lacking evidence of a maxillary conchal crest. Among specimens from European sites, the fairly complete Cro-Magnon 1 cranium clearly presents, on the right side of the anterior portion of its nasal cavity, a prominent maxilloturbinal that projects medially away from the lateral wall as it descends vertically toward the floor of the nasal cavity (it is likely that the inferiormost extent of each maxilloturbinal is broken); on the left side is preserved the base of a maxilloturbinal that had fused to the inner wall of the nasal cavity. On the preserved right side of its nasal cavity the less complete Cro-Magnon 2 cranium shows a distinct maxillary conchal crest (Fig. 11), as also does the Chancelade cranium; on the left side of the latter specimen, it appears that the maxilloturbinal had fused to the lateral wall of the nasal cavity but had subsequently broken off, leaving only its base (Schwartz and Tattersall, 2002). The left side of the nasal cavity of the Grimaldi 5 cranium presents what is plausibly the base of a fused but now broken-off maxilloturbinal, whereas the internal surfaces of the Grimaldi 4 and 6 crania appear to lack morphology, even of a maxillary conchal crest. The intact skull from Abri Pataud preserves distinct maxillary conchal crests bilaterally (Schwartz and Tattersall, 2002). Two specimens from the Central European site of Mladeč provide information on nasal cavity structures. Mladeč 2 is a partial and reconstructed cranium; it preserves the right maxillary frontal process, which bears a low maxillary conchal crest on its internal surface (Fig. 12). Mladeč 8, a partial maxilla, also preserves the right maxillary frontal process, which Frayer et al. (2006) (also Wolpoff et al., 2006) described as having a medial projection. However, as in Sima specimens AT-767+AT-963 and AT-1100+AT-1111+AT-1197+AT-1198 (Arsuaga et al., 1997), the horizontal swelling in Mladeč 8 is actually the base of a maxilloturbinal that had fused to the lateral wall of the nasal cavity and then broken off (Fig. 12).

Figure 11.

Various specimens of fossil Homo sapiens. Top left and middle: Cro-Magnon 1. Top right: Cro-Magnon 2. Bottom from left to right: Grimaldi 4, 5, and 6. BMT, base of broken MT fused to internal wall of MPF. Not to scale. Photographs © J. H. Schwartz.

In contrast to these late Pleistocene European specimens, the chronologically older partial cranium from Steinheim, Germany, which preserves the right side of the nasal cavity up to the margin of the nasal aperture, possesses a thickened and essentially vertically oriented strut of bone that is broad inferiorly and tapers superiorly. More or less midway up its length, this bony strut is separated from the internal surface of the maxilla by a vertically oriented foramen that is anteroposteriorly thin but superoinferiorly moderately tall (and is thus slit-like) (Schwartz and Tattersall, 1996). Importantly, this vertical structure is not an extension of a spinal crest, which is lacking altogether.

Of the Levantine cranial specimens from the Mount Carmel caves and Qafzeh, only one—Qafzeh 9 (Fig. 12)—presents any information on the anterior portion of the nasal region. Unfortunately, the nasal cavity is largely filled with wax (no doubt to protect the thin walls of the nasal cavity from cracking further), but on the right side it is evident that there was no medial swelling and that a maxillary conchal crest was probably present. Finally, the very modern-looking Homo sapiens cranium from Liujiang, China, preserves on the left side a maxilloturbinal that had fused with the lateral wall of the nasal cavity; on the right is visible the base of a fused maxilloturbinal that had broken off (Schwartz and Tattersall, 2003).

Figure 12.

Top left: Steinheim (arrow points to thickened vertical strut of bone). Top middle: internal surface of MFP of Mladeč 2. Top right: Qafzeh 9. Bottom: Mladeč 8. Not to scale. Photographs top row © J. H. Schwartz.


The thinness and, thus, the fragility of the bones that circumscribe the lateral and superior walls of the nasal cavity and make up the ethmoid and its turbinate outgrowths and the maxilloturbinal, unquestionably diminishes the likelihood of their preservation in fossil crania. However, as the preceding review clearly shows, a substantial number of specimens preserve sufficient morphology to allow the admission of this region into discussions of hominid systematics and morphological diversity.

On the basis of the descriptions already given, we can delineate four different configurations of the lateral wall of the anterior portion of the hominid nasal cavity: (1) an absence of structural relief; (2) the presence of a maxilloturbinal that articulates anteriorly with a horizontally oriented maxillary conchal crest; (3) the presence of a maxilloturbinal that articulates anteriorly with the maxilla along a vertically oriented zone of contact (as in Gorilla); and (4) the presence of a variably enlarged swelling that projects medially into the space of the nasal cavity, and which may also be associated with a maxilloturbinal or maxillary conchal crest.

The category “absence of morphology” is not confined to any potential clade or species of hominid, and what such “absence of morphology” reflects is not necessarily the absence of a maxilloturbinal. Rather, the possibilities in such cases are that: (1) there was a maxilloturbinal, but a maxillary conchal crest just did not develop; (2) a maxillary conchal crest of the expected configuration and orientation had not developed because the maxilloturbinal was not configured as in, for example, Homo sapiens (i.e., was configured as in Gorilla); or (3) this crest, and thus the associated maxilloturbinal, was situated farther back along the wall of the nasal cavity (as in Pongo). Hominids that fall into the “absence of morphology” category include all relevant australopith crania, OH 24 (putatively Homo habilis), KNM-ER 1470 (putatively H. habilis, H. rudolfensis, or a species of Kenyanthropus), Koobi Fora KNM-ER 3733 (putatively either H. erectus or H. ergaster), and Swartkrans SK 847 (putatively H. erectus). The category also includes Jinniushan (putatively H. heidelbergensis), and Arago (certainly Homo heidelbergensis). This is an oddly assorted group, and as the lack of an anteriorly situated maxillary conchal crest does not necessarily mean that none of these specimens possessed a maxilloturbinal that articulated anteriorly with the lateral wall of the nasal cavity, it would be imprudent at this time to speculate about the phylogenetic significance of “absence of morphology.”

The presence of a horizontally oriented maxillary conchal crest that is situated anteriorly on the lateral wall of the nasal cavity (and with which an anteriorly distended maxilloturbinal is actually associated or is inferred to have been associated) is not confined to extant Homo sapiens. Rather, evidence of this configuration is observed in the fossil crania mentioned from Abri Pataud, Chancelade, Cro-Magnon, Grimaldi, Liujiang, Mladeč, and Qafzeh (all of which can be identified as Homo sapiens); Jebel Irhoud (putatively H. sapiens); Atapuerca (Sima de los Huesos), Kabwe and Petralona (putatively H. heidelbergensis); and Nariokotome KMN-WT 15000 (putatively H. erectus or H. ergaster). Given that the common condition among mammals is the possession of a horizontally oriented maxilloturbinal that achieves some articulation with the lateral wall of the nasal cavity, we are inclined to interpret the articulation in these hominids as primitive for the group. Certainly, without fuller information on the details of maxilloturbinal shape in these hominids, we are reluctant to assume that the configuration in H. sapiens is derived, whether for this species alone or for a clade that subsumes it.

Differing entirely in the configuration of the lateral wall of the anterior portion of the nasal cavity from extant large-bodied apes, from extant and fossil Homo sapiens, and from the fossils cited immediately above, are the specimens from Steinheim, Engis, Gibraltar, Krapina, La Chapelle-aux-Saints, La Ferrassie, La Quina, Pech de l'Azé, Roc de Marsal, and Subalyuk. All of these except Steinheim (which nonetheless bears numerous Neanderthal characters) are routinely identified as Homo neanderthalensis. In all of these specimens the anterior portion of the lateral wall of the nasal cavity presents a vertically oriented three-dimensional structure that may be expressed as either a thickened strut of bone (Steinheim) or as a swelling that may achieve some prominence (e.g. Gibraltar 1). Even in the two Neanderthal specimens that also present morphology that can plausibly be interpreted as representing a maxilloturbinal (the adult from La Chapelle-aux-Saints) or a maxillary conchal crest (the child from Subalyuk), a vertical swelling appears to be present. This vertical three-dimensional structure is not observed in any other fossil hominid that preserves this region, or in any large-bodied ape—or, as far as we know, in any other mammal. We thus feel justified in suggesting that the development of this structure, which lies anteriorly on the lateral wall of the nasal cavity, and which may lie at least partially if not wholly superior to an identifiable maxillary conchal crest, represents a uniquely derived character state. We also suggest that the variably swollen “medial projection” seen in specimens universally regarded as Neanderthal is yet further derived relative to the configuration seen in the Steinheim cranium. The simplest interpretation of these aspects of nasal complex architecture is that they reflect: (1) a sister relationship between whatever taxon the Steinheim specimen represents and the Neanderthals and (2) the uniqueness of Neanderthals compared with all other known hominids. That only two morphs—Steinheim and Neanderthals—can exhibit a cranial structure that has not been reported in any other primate (or mammal for that matter) should not be surprising. Witness the gorilla, which alone among primates develops a lacrimal bulla (Cave, 1973). An obvious corollary of the uniqueness of the medial projection as seen in Neanderthals is that this hominid emerges—once again—as a very autapomorphic species that clearly deserves full recognition as Homo neanderthalensis.


We thank all the curators and other guardians of human fossils for access to specimens in their charge (Schwartz and Tattersall, 2002, 2003, 2005). JHS also thanks C. Zollikofer and M. Ponce de Léon, Anthropological Institute and Museum, Universität Zürich-Irchel (abbr. AIMZ) for access to the collections of extant apes. Finally, we especially thank Samuel Márquez for inviting us to participate in this exceptional issue and to Tim Bromage and an anonymous reviewer for helpful suggestions for revision.