The mobile Neandertal

Beauval et al. (2005) report on the anatomy, mtDNA genetics, and archaeological associations of a femur shaft fragment from Les Rochers-de-Villeneuve, France.

Two aspects of the specimen have attracted interest. The first is the connection with the hyenas. The National Geographic news report focuses on this aspect:

"The Neandertals and large carnivores occupied the cave in rapid succession," said Erik Trinkaus, a paleoanthropologist at Washington University in St. Louis, Missouri. "We have the bones of herbivores like bison and deer being chewed or processed by both Neandertals and hyenas, and they're both only going to do that if the meat is reasonably fresh, and if there's still something on there to get off."
"We have this idea that once humans became reasonably successful as hunters that they walked with impunity on the landscape, and that's just not so," Trinkaus said. "I'm not saying they were having fights at the mouths of caves with the hyenas, but I'm sure there were plenty of times when the hyenas came and, not being stupid, the Neandertals said 'see ya later, guys.'"

About this, there is little to be said. Pleistocene humans generally were in strong competition with carnivores. For the Neandertals this competition may have been more important than elsewhere, because we infer that their diets were strongly dominated by meat. On the other hand, finding human cutmarks and carnivore tooth marks on the same bones does not indicate that humans were frequently (or ever) on the losing end of their encounters with carnivores. Lions and hyenas both exploit the same carcasses today, and although hyenas can be formidable opponents for lions, the lions are generally successful both in defending carcasses for as long as they want them, and in stealing carcasses from hyenas. We don't know that Neandertals had this kind of success, less, or more. Of course, the fact that the hyenas ate part of this individual (toothmarks on the femur itself) goes farther than most of the evidence toward the idea that humans were more than rarely on the losing end.

The other point of interest is the inference of increasing mobility based upon the anatomy of the femur shaft. The following is from the Washington University press release:

"In Europe, with the transition from Neandertals to modern humans, anthropologists have long argued that major behavioral changes and major improvements in adaptation began to take shape with modern humans," said Erik Trinkaus, Ph.D., Mary Tileston Hemenway Professor of Anthropology at Washington University and co-author of the paper.
"One of the changes that has been documented with the transition from Neandertals to modern humans was that people became more mobile and their territories became much larger. They became less locally focused and more regionally focused," Trinkaus said.
It's been assumed that this happened in the Upper Paleolithic which is associated with some very late Neandertals and early modern humans. However, this is a femur bone from a Middle Paleolithic Neandertal. It shows in the shape of the femur that a shift to greater mobility had already begun prior to the transition to the Upper Paleolithic, prior to any appearance of modern humans in Europe.

The pattern observed in Les Rochers-de-Villeneuve is not unique. Trinkaus and colleagues (1998) presented the cross-sectional geometry and biomechanical analysis of the femur of the Saint-Césaire skeleton. This specimen is a late Neandertal dating to around 36,000 years ago. The Saint-Césaire right femur is similar to other Neandertals in the cross-sectional geometry of its subtrochanteric region, but is different from Neandertals in its midshaft -- the same region of the bone that Les Rochers-de-Villeneuve preserves. From Trinkaus et al. (1998:5838-5839):

[T]he degree of structural anteroposterior reinforcement of the femoral midshaft of at least recent humans parallels degrees of mobility, such that, on average, males exhibit greater midshaft anteroposterior reinforcement than females, including among Pleistocene Homo, and pre-industrial populations inhabiting more accentuated terrains exhibit an emphasis on anteroposterior femoral strength. The greater femoral midshaft anteroposterior strength in the early modern human sample therefore suggests a higher average level of habitual mobility (contrasts in terrain per se are not relevant because the two reference groups occupied most of the same regions, and, if there are any differences in habitual terrain, it is more likely that the Neandertal sample experienced a steeper topography because many of the early modern human remains derive from the more open central European plains). With respect to this, Saint-Césaire 1 falls clearly with the early modern humans and separate from the Neandertals.

That paper concludes with the hypothesis that Saint-Césaire 1 represents a late Neandertal population in which the hyperarctic body proportions are maintained but mobility patterns have begun to change. Both bones are fragmentary, so the estimation of limb length (and thereby of leg length and proportions) is imprecise. But assuming that both bones represent individuals with short, cold-adapted legs, they would appear together to be significantly different from other Neandertals in their cross-sectional geometry at midshaft. According to Trinkaus et al. (1998), the Feldhofer 1 specimen is similar to these in approaching the midshaft cross-sectional geometry of Upper Paleolithic modern humans, although not to the extent of the Saint-Césaire femur. Feldhofer is also a relatively late Neandertal, now thought to date between 40,000 and 45,000 years. So there is a suggestion that later Neandertals differ from earlier ones in their femur diaphyseal anatomy, and that this difference may reflect activity to some extent.

On the other hand, the later Neandertals would also include Spy, which do not have greater anteroposterior bending strength. And it is unclear whether the variation in the late Neandertal sample taken as a group is significantly different from earlier Neandertals. What we see in these latest specimens is the hint of a change, but not strong evidence of one. This problem is ultimately one of sample size, and it would be interesting to consider the kind of variation within samples of recent humans as well as between groups with different activity patterns to see how much variation is consistent with a single activity pattern. More on this below.

What is mobility?

The mobility pattern is a way to broadly talk about the home range size, daily movement distances, annual range and long-distance movements, speed, and locomotor power. The femur reflects the influence of all of these, through their biomechanical requirements, such as the lever arms of muscles, the muscle force as a correlate of muscle size, and the joint reaction forces and biomechanical loads related to moving the mass of the body. Thus, the femur anatomy is a compromise among the many requirements placed on it.

Inferring all of these things from the geometric distribution of cortical bone in the femur diaphysis is -- at least on the face of the problem -- statistically impossible. There is really only a single variable: femur shaft breadth/depth ratio. Cortical bone area arguably is a second variable but in the absence of information about body size its utility is questionable. The variables are influenced by many parameters, including all those listed above and perhaps several more. A single variable can explain at most a single independent parameter. Thus, we have a problem: there are simply too many possible combinations of parameters that could influence shaft anatomy.

Research into mobility patterns deals with this problem by attempting to reduce the number of parameters. The most promising way to accomplish this is to show that the parameters are not independent -- that changing one of them tends to change all the others as well. Given the small number of recent foraging groups whose skeletal remains are available for study, this research effectively condenses to a small number of contrasts between "mobile" and "less mobile" or "sedentary" groups. For example, Holt (2003:204, citations elided) describes the relationship between cross-sectional geometry, muscle activity, and forager mobility:

Femoral and tibial diaphyseal morphology reflects variability in levels of locomotion. Studies of human gait show that anterior-posterior (a-p) bending moments resulting from the combined effects of hamstring and quadriceps muscle groups are highest in the knee region (midshaft femur to midshaft tibia), reflecting activities such as running and climbing. The transition from highly mobile hunter-gatherer subsistence to a more sedentary one produces important changes in lower limb bones, namely inward contraction of cortex, reduced a-p bending rigidity of the midshaft femur and tibia, resulting in reduced directionality of greatest bending strength, and increased circularity of midshaft femoral and tibial cross sections. This was elegantly demonstrated in a recent comparison of limb bone strength between Holocene southern African and Andaman Islands foragers (Stock and Pfeiffer, 2001). The highly mobile prehistoric South African foragers had significantly stronger and less circular lower limb bone shafts, while the protohistoric Andaman Islanders, who rely primarily on exploitation of marine resources by boat, exhibited more robust clavicles and humeri, in keeping with expectations from their low level of terrestrial mobility.

Thus, "mobility" is a conceptual category that includes essentially every activity that might make people walk farther, move faster, exercise more, and stay in one place less. In certain comparisons between "mobile" groups and "less mobile" groups, the more mobile groups have a cross-sectional anatomy with a greater anteroposterior diameter. In modern humans those people also tend to have a pronounced pilaster on their posterior femora, and have greater cortical bone area. As long as these components of the overall activity pattern can be grouped together into a contrast between two populations, the argument that bone anatomy responds to something called "mobility" can be supported.

There is a more extensive literature on this issue that can be reviewed from the references in these papers; I am skeptical of the relationship but please read further into these references if you are interested in the way mobility is assessed and compared between populations.

A problem with this approach is that some relatively uncorrelated parameters still remain. The most important are body mass and stature. Since recent hunter-gatherers vary substantially in both these values, studies of mobility and its effects on bone anatomy have attempted to "control for" them. But what such a control might mean when extrapolated to an extinct population with very different mass and stature than most living foragers is basically not known. It is through these values that the degree of cold adaptation becomes important. A short limb and a broad trunk may be expected to exert a greater mediolateral force on the limb bones, as the body requires more powerful muscular support during one-legged stance. Weaver (2003) asserts that much of the femur anatomy of the Neandertals is a consequence of their relatively short legs and broad pelvis. At the least, cold adaptation is a potent confounding factor that complicates the interpretation of mobility patterns from bone cross-sectional anatomy; at the worst it makes such inference impossible. The problem is also addressed by Pearson (2000) who attempts to relate robusticity to activity patterns and climate, finding that climate explains much in the robusticity patterns (rough measures of bone strength) of living human populations.

In this vein, Lieberman's (2000:595) comment is pertinent:

Since some recent humans (e.g., Australians) have levels of residual robusticity as high as those of Neanderthals, simplistic models of behavioral change which predict that technological advances from the Middle to the Upper Paleolithic led to reductions in stress levels over time are untenable. Life was apparently pretty tough for everyone in the Paleolithic but possibly in substantially different ways.

Until this kind of concern can be answered, it is not clear that any test of activity patterns in Pleistocene fossil samples will be possible. Unlike Lieberman, I don't think these answers will come from controlled laboratory experiments on living human or animal subjects. The kinds of activities that contribute to the anatomy of a hunter-gatherer femur over a lifetime are a very complex system. Reducing activity to different components and testing separate effects may have much to offer in terms of showing the relative importance of different activities within a population. But for Neandertals, the problem is inferring an unknown pattern from a single variable; and from that perspective, a fuller conception of activity patterns in living populations only further problematizes the issue. In other words, the more we learn, the less we know. For the scientist, this is how it should be, but it does not bode well for our ability to interpret changes in later Neandertals.

The multiple explanations of Neandertal mobility

But suppose we dismiss my skepticism about whether we can infer anything about activity patterns from femur cross-sectional diameter. In any event, the anatomy of the later Neandertals appears to make an interesting contrast with that of the earlier ones. Does this contrast point to behavioral changes in the Neandertal population?

We don't yet know what selective balance acted on the mobility pattern of Neandertals. Further, we don't know to what extent environmental changes (such as mobility patterns induced by cultural changes) may have affected the expression of the relevant anatomies. It is often assumed that these two mechanisms of change -- selective and environmental -- are concordant in direction, and so may be explicable in terms of the same ultimate causes. Thus, a change in mobility might both initiate direct environmental changes due to a changing pattern of muscle action on the bone; at the same time, the bone may be under selection to better match the pattern of muscle forces upon it. But there is really no justification for such an assumption; one might as easily argue that a decrease in muscle force induced by nongenetic causes results in a random change in selection on the cross-sectional anatomy. It does seem likely that developmental plasticity is the first explanation for changes; which puts the muscle forces themselves to the fore.

For most Neandertals, the femur is relatively short, has relatively large proximal and distal articular surfaces, has an anteroposteriorly curved shaft, and has a shaft cross section that is broader mediolaterally than anteroposteriorly, without a pronounced pilaster. I would suggest two hypotheses to explain the Neandertal bone geometry:

  1. Neandertal mobility was highly constrained by their short legs, which were themselves primarily adapted to climate. The restriction in mobility was mediated by a higher energetic cost of locomotion, which favored a residence pattern with less annual and daily movement.
  2. The requirements of hunting or other activities (including the possibility of conflict) in Neandertals demanded powerful limbs, loaded in many different directions, and did not accentuate long-distance daily or annual movements.

Although these explanations are not mutually exclusive, I tend to think the second hypothesis is the more likely one. With their relatively extreme body proportions, Neandertals were certainly at an energetic disadvantage in locomotion, but much of that energetic expenditure would have been inevitably lost to heat production, whether the Neandertals were moving around or not. Just living in a cold place exerts a high energetic cost, and the adaptive value of short legs is not merely in reducing surface area but also in increasing heat production. So the very high energy budgets of Neandertals give little reason to think that energy constraints were a decisive influence on their activity patterns.

The late Neandertals Les-Rochers-de-Villeneuve and Saint-Césaire appear to have retained the short femora and curvature, but have a cross-sectional geometry at midshaft that is more circular, with a relatively greater anteroposterior diameter. If the first hypothesis were true, then we have two options to explain the change:

  1. Climate change or competition with modern humans forced late Neandertals to become more mobile than their ancestors. This change in activity pattern would have imposed a high energetic cost because of their short legs.
  2. Cultural change enabled a more efficient foraging pattern (e.g. one based on Chatelperronian technology). This change allowed greater mobility because it relaxed the energetic constraints on Neandertal groups.

If in contrast, the second hypothesis were true (as I think likely) then the change apparent in later Neandertals would be explained by a change in foraging (specifically hunting) style. We may relate this change to the archaeological changes evidenced for the latest Neandertals, and indeed these technical changes may also have had energetic consequences, particularly if the technologies of the initial Upper Paleolithic helped to reduce the impact of cold, or if they enabled more energetically efficient foraging or hunting.

All this must be considered extremely tenuous, since it is not really testable with the information at hand. For the latest Neandertals, this is more or less always the case: the number of specimens is pretty small, and for most comparisons the sample size is therefore insufficient to show significant differences. But there are exceptions to this (supraorbital torus size and projection are two, presence of a chin is another), and with the right method, we might find that the postcranial evidence for change in this later sample is stronger than one might think.


Beauval C. et al. 2005. A late Neandertal femur from Les Rochers-de-Villeneuve, France. Proc Nat Acad Sci USA PNAS Online

Holt BM. 2003. Mobility in Upper Paleolithic and Mesolithic Europe: Evidence from the lower limb. Am J Phys Anthropol 122:200-215. Wiley InterScience

Lieberman D. 2000. Comment on Pearson, OM, "Activity, climate, and postcranial robusticity." Curr Anthropol 594-595.

Pearson OM. 2000. Activity, climate, and postcranial robusticity. Curr Anthropol 41:569-607. Current Anthropology

Stock J and Pfeiffer S. 2001. Linking structural variability in long-bone diaphyses to habitual behaviors: foragers from the southern African Later Stone Age and the Andaman Islands. Am J Phys Anthropol 115:337-348.

Trinkaus E, Ruff CB, Churchill SE, Vandermeersch B. 1998. Locomotion and body proportions of the Saint-Césaire 1 Chatelperronian Neandertal. Proc Nat Acad Sci USA 95:5836-5840.

Weaver TD. 2003. The shape of the Neandertal femur is primarily the consequence of a hyperpolar body form. Proc Nat Acad Sci USA 100:6926-6929.