Dmanisi

The announcement of the Malapa skeletons has many of us going back to descriptions of early Homo. After the paper by Berger and colleagues came out last month, I wrote up some notes on KNM-ER 1813. This is another skull, often argued to be Homo, that has struck many people as similar to the samples from Sterkfontein and Makapansgat.

KNM-ER 1813 has some of the smallest postcanine teeth of any maxillary specimen attributed to Homo habilis. The Malapa MH1 specimen is also small compared to the Homo habilis sample, but not KNM-ER 1813 -- MH1 is larger than KNM-ER 1813 in at least one dimension of all its maxillary teeth.

Over the past 20 years, led by Wood (1991), most commentators have placed KNM-ER 1813 in Homo. But that assignment reversed many of the opinions on the skull's morphology that had been expressed since its discovery. Richard Leakey (1974) emphasized the differences between KNM-ER 1813 and KNM-ER 1470, which he had earlier attributed to Homo habilis. KNM-ER 1813 is quite a lot smaller in its endocranial volume -- only 509 ml, where KNM-ER 1470 is 752 ml (Holloway 1983).

Wood framed his discussion of the relationships of the specimen by explicitly listing the features in which KNM-ER 1813 differs from Australopithecus africanus. He began by acknowledging that the overall size and shape of the skull aligns it with A. africanus, whether we consider metric or nonmetric traits. Then he discusses several derived similarities that detract from that simple picture:

However, detailed differences between KNM-ER 1813 and A. africanus suggest that the general phenetic resemblance may be misleading. One of these detailed differences involves the frontal, and two concern the morphology of the occipital. The frontal of KNM-ER 1813, unlike that of A. africanus, shows a modest, but unmistakable post-toral sulcus and it is also braoder thn that of A. africanus. The occipital of KNM-ER 1813 makes a relatively greater contribution to the sagittal profile than in A. africanus, no matter which of the two locations of lambda is used. In addition, it bears an incipient torus, the shape of which has been interpreted by two independent authors as being reminiscent of H. erectus. All three of these detailed differences are such that the condition in KNM-ER 1813 is derived in the direction of H. erectus (Wood 1991: 92-93, citations omitted).

The extent of the occiput may be a simple correlate of a larger vault, but the other two characters -- supratoral sulcus and occipital torus, are not.

When the morphological features of the cranial base of KNM-ER 1813 are assessed against this comparative background, the evidence suggests that the cranial base of KNM-ER 1813 differs from that of A. africanus. Without exception, the expressions of these characters in KNM-ER 1813 are more derived in the direction of Homo than are the homologous characters of A. africanus. The angulation of the petrous temporal and the inclination of the foramen magnum are two particularly crucial indicators of the very different arrangement of the cranial base of KNM-ER 1813 and A. africanus. Further evidence of the relative shortness of the base of KNM-ER 1813 comes from the position of the porion with respect to the anteroposterior axis of the cranium, and the evidence of a relatively wide sphenoid (e.g. the location of the foramen ovale and spinosum and the make-up of the entoglenoid) must also be taken into account. This is not to say that all features of the cranial base of KNM-ER 1813 are Homo-like, for Dean (1984) has shown that the relative sizes of the insertions of the nuchal muscles are still remarkably pongid-like (as in deed they are for many early hominids). However, for several features for which we have good comparative evidence, their expression in KNM-ER 1813 must lead us to reject a close association between it and A. africanus. The anatomy of the mandibular fossa region is also derived with respect to the australopithecines, and Picq has claimed to see in KNM-ER 1813 the basis of a temporomandibular joint morphology that can be traced through KNM-ER 3733 to the condition seen in H. erectus from Asia (Wood 1991: 93, citations omitted).

Wood here went into a lot of detail about the cranial base because of his earlier work on this part of the anatomy (for example, Dean and Wood 1982). But even this description may seem cursory in light of the variability within A. africanus of these cranial base characters. KNM-ER 1813 is not ideally preserved, missing most of the basisphenoid and basiocciput, and without a good join along the midline back cracking across the occiput to the right asterion. The petrous orientation is very different from Sts 5, which also has a more posteriorly placed foramen magnum. But Sts 19 is not nearly so different in these respects from KNM-ER 1813.

Now, the cranial base of the MH1 skull from Malapa is still embedded in matrix, so we can't do this comparison yet with that skull. Will it look Homo-like in the ways that KNM-ER 1813 apparently does, or will it fit squarely within the range of Sterkfontein sample? If I were going to put money on the question, I would bet that the cranial base is influenced by endocranial volume. If so, then the small brain of MH1 will determine an essentially australopithecine-like cranial base. We'll see when the scans are examined.

The point of discussing this anatomy is not because the cranial base is itself intrinsically important. It fades next to more familiar traits such as brain size and dental morphology. But brain and teeth can't answer the question alone -- they need corroborating evidence from other characters. Particularly in cases like KNM-ER 1813, and MH1, with a more Homo-like dentition than brain, we want to find a phylogenetic hypothesis that maximizes consistency across the entire skeleton.

To understand why, consider another fossil: D2700 from Dmanisi. Rightmire, Lordkipanidze and Vekua (2006) explicitly noted the similarity of the subadult D2700 skull with KNM-ER 1813, including the size, the contours of the facial and vault profile, the size, shape and depth of the palate. The picture reflects the broad similarities noted in the text of that paper:

KNM-ER 1813 (left) and D2700 (right), from Rightmire et al. 2006.

Rightmire and colleagues (2006) note that KNM-ER 1813 is more like H. erectus in some respects than is the Dmanisi specimen -- chiefly, D2700 has little if any development of an occipital torus and has a longer clivus and wider interorbital distance than D2700. Rightmire and colleagues (2006) assume that KNM-ER 1813 represents H. habilis for the purposes of this comparison, and they then show that most of the resemblances between this specimen and D2700 are primitive. That amounts to an argument that D2700 is not H. habilis.

It's funny that this key point in human evolution is best documented by three skeletons that could all represent 11-year-old boys -- D2700, MH1, and KNM-WT 15000. In the case of D2700, the contrast with KNM-WT 15000 is possibly great, but hard to interpret because of the imperfect state of our developmental knowledge. Along those lines, it becomes clear just how much there is yet to learn about MH1.

References:

Dean MC, Wood BA. 1982. Basicranial anatomy of Plio-Pleistocene hominids from East and South Africa. Am J Phys Anthropol 59:157-174.

Rightmire GP, Lordkipanidze D, Vekua A. 2006. Anatomical descriptions, comparative studies and evolutionary significance of the hominin skulls from Dmanisi, Republic of Georgia. J Hum Evol 50:115-141. doi:10.1016/j.jhevol.2005.07.009

Wood B. 1991. Koobi Fora Research Project. Vol. 4. Hominid Cranial Remains. Clarendon Press, Oxford.

Gretchen picked up a partial set of Time-Life volumes, from 1973, part of the series "The Emergence of Man". She found them at a garage sale. There's a lot of fun stuff in them, and some very useful illustrations.

For example, I'm looking through the volume titled, The First Men, which is basically about Homo erectus. The meat of the book is a series of descriptions of fossil and archaeological finds -- Dubois on Java, Terra Amata, Torralba and Ambrona. No surprise, each of these has a very different theme than we would give them today!

Here's a fun comparison:

This is a two-page spread in the book; really a fine illustration by Roger Hane.

Homo erectus, in the middle, is reconstructed with a stature pretty much right in between Australopithecus and Homo sapiens.

If you open up most recent textbooks, you'll find Homo erectus illustrated as the same height or taller than us. This is mostly due to the KNM-WT 15000 (Nariokotome) skeleton, discovered in 1984. This skeleton was estimated to have a moderately tall adult stature -- around 185 cm (6 feet 1 inch). There are three or four other femora from the Lower Pleistocene that also correspond to stature estimates up around 180 cm, in particular KNM-ER 736 and KNM-ER 1808.

Now one might reasonably wonder, what's the big deal about 185 cm? The Nariokotome skeleton hardly represents a giant -- at 6'1" he would have been an inch shorter than me, for goodness' sake! And with only one fossil specimen within 2 inches, it shouldn't be churlish to point out that the Nariokotome estimate is not based on a real femur length -- it's an estimate based on an estimate. Most Lower Pleistocene fossil femora were much shorter, and yield stature estimates well under 180 cm. So why did anthropologists so eagerly cling to the tallest estimates for Homo erectus?

Few Lower Pleistocene postcranial bones are associated with skulls, so it's difficult or impossible to assign smaller bones to a species. How do we know whether a short femur belongs to Homo erectus or Homo habilis -- which we know from OH 62 is much smaller in body size? Or A. boisei, which it would appear from KNM-ER 1500 is also smaller? We don't really know -- so the bones that correspond to mid-range stature estimates, say around 160 cm, might belong to any of the above. But the tall ones -- well, we know that those must represent the largest-bodied hominin. So there was a tendency to assume that the tall specimens were near the average for Homo erectus.

It was a hypothesis. It has turned out to be false.

The illustration in the Time-Life book is based on entirely different fossils. The Turkana fossils were unearthed during and after the early 1970's. Before that, Homo erectus stature could be estimated from the Trinil (Java) and Zhoukoudian (China) femora. These are later than the early African Lower Pleistocene sample. The Zhoukoudian femora in particular give stature estimates at or under 5 feet (152 cm). We might read it as a decline, and some people did as recently as 5 years ago. But the Dmanisi postcrania are also short, a bit shorter than the Zhoukoudian femora. And they're earlier than Nariokotome. And we now know of smaller crania of Homo erectus in the East African Lower Pleistocene. So the tallest statures aren't the average; they're the tallest.

It ain't rocket science, I know. But this is progress.

Today, I think it's fair to say that the variation of stature in Homo erectus was more or less like the variation within living people. There are short and tall populations today, varied in ecology and latitude. The average stature of young men in the Netherlands today is 184 cm. Adult women in the Philippines average only 150 cm. So the best way to compare statures is to illustrate the range.

That being said, I don't think we know how stature has evolved over time. We do have some data points -- the Neandertals were shorter than Upper Paleolithic Europeans, for example, but seem to have been around the same height as Mesolithic people (and a shade taller than Neolithic Europeans). The Dmanisi people were on the short end of the human range, but not unusually so. The variability within Lower Pleistocene East Africans seems high, but I'd want to see a serious test compared to human populations.

It's a case in the fossil record where discovering more seems to have resulted in us knowing less. But that's just because we can now reject several categorial statements that people used to accept uncritically.

Related articles:

"News flash: Dmanisi hominids were not short"

"Body size in Holocene South Africa"

Body mass in ancient humans and high-latitude populations

After my Q and A with paleoanthropologist Mica Glantz, I got a lot of great response -- people really liked reading about work in the field from somebody other than me!

So, I'm going to make these interviews a regular feature. When I was in Michigan last week, I got a chance to talk with Adam Van Arsdale, who graciously agreed to answer some questions about his work.

UPDATE(11/29/2007): After posting, I heard from a reader who reminded me that I omitted Adam's affiliation and info! Adam is a lecturer in anthropology at the University of Michigan. You can find out more about his interests on his webpage.

Hawks: You were lucky enough to work at one of today's most exciting paleoanthropological sites, Dmanisi. What can you tell us about your experience there?

Van Arsdale: Dmanisi is a wonderful place and I can't say enough positive things about the site and all the people I have worked with through the project. To begin, the site itself is just a nirvana for anyone with an interest in history or prehistory. The primary excavation area is in the middle of a ruined medieval citadel complex which rose to prominence as a trading town along the silk road; down from the promontory are the tombs of Mongols who sacked the city in the 12th century; further down are early Christian burials, and along the river are the remains of bath houses for travelers along the Silk Road. It is a literally a place where time seeps out of the ground.

Leaving the setting aside, the people associated with the project have been wonderful to work with. The size of the excavation team would vary but there would be times when, at the end of a long excavation day, I would find myself sitting at a long dinner table surrounded by 40 people speaking more than half a dozen languages. In the years I worked there as a graduate student I think we had students and researchers from 15 different countries (and I'm probably missing a few). Everyone who works at the site, including the local residents of Patara Dmanisi, adds their own character to the project. As a graduate student, my summers at Dmanisi served as something of a Paleoanthropology bootcamp, with regular discussions and debates between all of us with very different training and different theoretical perspectives on the issues of human evolution.

And then on top of all of this there are, of course, a remarkable set of fossils and archaeological materials.

Hawks: Do you want to give a shout-out to anybody in Georgia?

Van Arsdale: There are too many to name, but certainly David Lordkipanidze, who first invited me to Dmanisi in 2001, deserves recognition. I'll also add Gocha Kiladze, Teona Shelia and Dato Zhvania, who began working at Dmanisi in 1991 as students and who continue to play a significant role in the operation of the site today. One of the great things about the site is that it has served as a tremendous springboard for Georgian students interested in paleoanthropology. I think it is a safe bet we will be hearing a lot from our Georgian colleagues in the years ahead.

Hawks: Your dissertation work focused on the Dmanisi mandibles. I know that you still have publications coming out on these, so feel free to keep quiet about anything you're saving for print. What can you tell us about the sample?

Van Arsdale: The Dmanisi mandibles are a remarkable sample. They show a huge amount of morphological variation in a set of fossils derived from a temporally and geographically constrained set of deposits. One of the mandibles is in many characters the largest mandible assigned to the genus Homo. Two of the others are quite small, with variably large and small teeth. And the fourth specimen is one of the earliest edentulous mandibles in the hominid record. Given the current season, it is perhaps appropriate to describe the sample as a real cornucopia of variation. And the location and date of the site itself is surprising. Dated to 1.8 million years and about 2000 miles from the outlet of the rift valley in northeast Africa, the site is a long way from the contemporaneous and well-known deposits from the Turkana Basin in Kenya and Olduvai Gorge in Tanzania.

So how do we account for all this variation? That was basically the question of my dissertation. I sought to answer this question by testing a series of hypotheses focused first on sources of intraspecific variation, particularly age and sexual dimorphism, then secondarily on hypotheses of interspecific differentiation (i.e. multiple species). I then evaluated the results of these quantitative tests in the context of the comparative anatomy of the Dmanisi sample. Sparing you all the details, I think there are strong reasons to consider the Dmanisi hominid sample as that of a single species, but one displaying considerable amount of variation associated with age and possibly elevated levels of sexual dimorphism relative to what we observe in contemporary and recent human populations.

Hawks: Of course, your work required a lot of comparisons with other samples, and mandibles are among the most common skeletal elements represented in the fossil record. How did you handle your comparative work?

Van Arsdale: Paleoanthropology is at its root a comparative discipline. It is difficult to interpret any set of fossils outside of some comparative model. My work is no different. In asking questions about variation associated with age and sex, my dissertation is really asking how strange (or not strange) does the variation in the Dmanisi sample look if we treat it like a mixed age and sex sample of humans? Of chimpanzees? Of gorillas? Each of these species possess somewhat differing patterns of variation so that our final understanding of the Dmanisi specimens is based on a combination of similarities and differences with these different comparative models.

You can also try to understand the sample from the perspective of other fossils. These comparisons are more challenging because we have less certainty regarding the things we think we know about fossils. For example, in my dissertation I also make a series of comparisons between the Dmanisi mandibles and a sample of Australopithecus boisei mandibles from East Africa. It is much more difficult to say for certain whether any given fossil specimen is male or female, and in the absence of well preserved teeth, young or old. That uncertainty limits the power of the hypothesis tests we can bring to the question by limiting the amount of information we have to work with.

One of the exciting aspects of Paleoanthropology's comparative perspective is that new fossils give us new ways of looking at old fossils. Possibly the most exciting aspect of the Dmanisi fossils is that they provide us a tremendous platform from which to look back at these large samples from East and Southern Africa that we have known about for a long time and reexamine questions which had either previously been unanswerable or whose accepted answers no longer seem so clear.

Hawks: Any stories you can share about your travels?

Van Arsdale: One of the more unique experiences from my travels occurred while I was tagging along with a graduate student from Yale on her project involving 4.5 million year old fossil exposures in the Tugen Hills of the Central Rift Valley, Kenya. I was off on my own one day, walking along one of the exposures when I came across what appeared to be part of a fossilized crocodile skull just barely sticking out of the ground. I sat down and began very carefully exposing its boundaries so that it could be properly prepared and taken out. After about 20 minutes of this, a young Tugen boy came out of the bushes and sat down next me and began watching me work. I tried to say a few words of greeting in my very rudimentary Kiswahili, but either my pronunciation was too terrible to be understand (quite likely) or he was too young to have yet learned Kiswahili (he looked like he was between 8 and 10). After a few more minutes the boy, who had been carrying a small bow and set of arrows, took out one of his arrows and began using its steel tip as a mini-trowel. I would have discouraged him out of fear he might damage the fossil or go on trying to dig up other fossils in the area, but as I watched him he was exceedingly careful and seemed completely enraptured by the work. It was just one of those moments where, while the event was going on, I recognized how amazingly unique it was. Here we were, a graduate student from the University of Michigan with twenty plus years of formal education and a young Tugen boy with at most a few years of schooling, sitting side by side on a hillside in the middle of Kenya carefully exposing a 4.5 million year old fossil. The only common language between us was the action of my Marshalltown trowel and his handmade arrow point and a basic curiosity in this fossil.

Hawks: It's a story you hear from students a lot: teeth and mandibles are "bor-ing". But of course, they're the best representatives of variation we have through much of human evolution -- if you want to study evolution, you'll be studying jaws and teeth. What keeps these questions exciting for you?

Van Arsdale: One of the reasons I enjoy looking at mandibles and teeth are that they can potentially provide a window into numerous aspects of human evolution. As you point out, they are the most abundant element in the fossil record and therefore provide a large set of data with which to address questions of evolutionary relationships and evolutionary change. They can also tell you something about the ecology and diet of the individual specimen. Finally, they tell us something about how an organism develops throughout life and ages.

This also means that questions regarding variation in jaws and teeth can be difficult to answer because many different processes might account for the observed variations. When testing hypotheses about mandibular variation it is important to keep this in mind. It is always striking to me how many hominid type specimens are or have served at some time as type specimens for a new species. This is in part a reflection of their relative abundance, but I think it also reflects how difficult it is to adequately address all the potential sources of variation in mandibles. If you accept the conclusions of my research, the Dmanisi mandibles serve as a cautionary tale in this regard.

Hawks: Some readers may know that you and I share the same graduate advisor, Milford Wolpoff, who has certainly been a strong influence on the way I approach evolutionary questions. But I also find myself going back to other people who influenced my training. Who/what really got you interested in the field, or shaped the way you think about evolution?

Van Arsdale: I initially entered anthropology by happy circumstance. Entering college (Emory University) I was interested in majoring in both English Literature and Evolutionary Biology. My first year two things happened; I realized Emory's biology department was primarily focused on microbiology and full of pre-med students (something I was not interested in) and I took my first Anthropology course to fulfill a distribution requirement. I was immediately hooked. Here I could have the best of both worlds... an integrative approach towards understanding what it means to be human and a careful examination of the evolutionary processes which have shaped the pattern of human evolution. I owe a huge part of my perspective to Milford and the other faculty and students I worked with as a graduate student, but I don't think I fully realized the influence my undergraduate teachers had on my perspective till the AAA meetings last year when I was able to attend a session honoring the graduate advisor (Jack Kelso) of my undergraduate advisor (George Armelagos). I listened to talks by people I had never met, but with whom I share some of my academic phylogeny, and what I heard were familiar themes on the interaction of human biological and cultural processes. This bio-cultural perspective is something I carry with me from Emory and is evident in the approach I take towards questions of Pleistocene human evolution, where changes in human skeletal form cannot be understood outside of the context of our ever-expanding brains and the increasingly complex ways in which we interact with the people and environments around us. Now that I am teaching, it is something I am aware of when I am in front of the undergraduates in my own classes.

Hawks:Some of your current research involves a lot of genetic modeling. How did you get into this area? Can you tell us about some of your thoughts?

Van Arsdale: My interest in genetic modeling first began as an undergraduate. In part it reflects my status as an admitted math nerd. I like numbers, I like using computationally intense models and simulations to address specific hypotheses, and I like understanding how evolutionary and cultural processes interact in dynamic ways. But when I was an undergrad my interest in genetic models stemmed out of my interest in modern human origins and the belief that any really good model should be able to simultaneously explain the pattern of fossil, archaeological, and genetic evidence. At the time there was quite a bit of discussion not just about how the increasing amount of genetic data related to previously held understandings of the fossil and archaeological record, but also how compatible data from different genetic systems were with each other. In particular, data from non-recombinant genetic systems (mtDNA and parts of the Y-chromosome) seemed to provide a different picture of human evolution than data from recombinant genetic systems. My attempt to understand these differences is what really drew me into aspects of genetic modeling.

Since that time my interest genetic modeling has really developed out of what I consider an anthropological approach towards understanding genetic systems. I like to quote one of the take-away messages from the dissertation defense of another Michigan graduate, Keith Hunley, who modeled genetic aspects of South American population structure in his dissertation. As Keith said in his defense, what people do matters. Most genetic models are dependent on a variety of demographic parameters (population size, structure, etc.), all those things that people do. And yet most geneticists do not, or simply cannot directly address these demographic parameters with the data available to them. As a paleoanthropologist, one role my research serves is to provide better understandings of what people did and the ways in which they interacted in the past so as to better inform such genetic models.

On a more theoretical level I am very much interested in exploring how the unique ways in which humans shape and interact with our evolutionary landscape serves to structure genetic variation and the evolutionary forces which shape it.

Hawks: What's the next step for you? Where do you go from here with your research?

Van Arsdale: Most of the questions I am working on now reflect my current thinking that the basic pattern which characterizes Pleistocene human evolution; the complex interaction between increasing cultural complexity, expanding ecological niches, and basic anatomical changes (encephalization, dental reduction); establishes itself early in the Pleistocene if not prior than that. Essentially, that sometime around 2-2.5 million years ago a group of hominids stopped acting like bipedal apes (the Australopithecines) and started acting human. This basic human pattern then continued to develop and characterize Pleistocene hominids until about 10-20,000 years ago when we stopped acting like humans and started acting like domesticated humans.

By understanding how this pattern manifests itself early in the Pleistocene, for example, by considering how, why and with what changes human populations expanded into places like Southern Georgia as early as 1.8 million years ago, you can develop broader understandings of the Pleistocene as a whole. I am just finishing up two projects related to this broad topic, one examining the Habiline-Erectine transition in the Lower Pleistocene and another attempting to characterize broad demographic changes within the Pleistocene.

I also want to continue my involvement in paleoanthropological field work and would like to continue examining Plio-Pleistocene deposits in Western and Central Asia. Dmanisi is an incredible site and has provided a great amount of detailed evidence to address questions of human evolution from this time period. But the detailed picture it provides encompasses only a narrow range of time and space...the more we can expand that window the better we can understand the broad patterns of change which characterize humans in the Plio-Pleistocene.