The foramen magnum does not point to locomotor adaptation

Aidan Ruth and colleagues in the Journal of Human Evolution have an interesting paper with the seemingly counter-intuitive result that foramen magnum orientation does not relate to locomotor pattern. They consider several mammal groups in which some species have more vertical or orthograde posture and others a more horizontal posture. This includes rodents, strepsirrhine primates and marsupials, all groups in which some species have a more upright locomotor pattern and others more quadrupedal.

Springhare skull, source Dallas Krentzel
A springhare skull, one of the bipedal hopping rodents. Credit: Dallas Krentzel CC-BY

From the summary:

When we explored the variation in marsupial FM [foramen magnum] location, we found no relationship to locomotion (Table 4). Within rodents, heteromyids and dipodids have independently evolved grossly enlarged auditory bullae, which distort the entire splanchnocranium. When the effects of these structures are accounted for, the impact of locomotion on FM position becomes statistically negligible (Table 5, Fig. 4). We analyzed FMA [foramen magnum angle] and locomotor category in strepsirrhines and found no significant difference between pronograde and orthograde species (Table 4), but did find a significant relationship between FMA and ER (Fig. 5). Our results show that locomotion does not influence FM orientation in marsupials, rodents, or strepsirrhines, and that basicranial evolution is a complex phenomenon that must be explored in the context of each taxon's unique evolutionary and developmental history.

This is not the first paper to come to this result; Ruth and colleagues cite a number of others who have had similar conclusions either from a comparative or mechanical point of view.

The obvious question is: What about early hominins? Two genera of Late Miocene and Early Pliocene primates, Sahelanthropus and Ardipithecus, have been argued to show evidence of an orthograde postural repertoire, largely due to their basicranial form, including foramen magnum orientation. For Sahelanthropus there is no other evidence of posture; for Ardipithecus there is an abundance of postcranial evidence, none of which argues compellingly for an orthograde posture.

The reconstruction of an anterior foramen magnum position for both these genera has been controversial. Both of the relevant specimens have required extensive virtual reconstructions. Neither of those reconstructions has been independently assessed by other groups of scientists.

So is it possible that these were actually representatives of one or more non-hominin ape lineages that exhibited some basicranial convergence with early hominins?

Ruth and colleagues make a pretty clear statement on this point:

Our data from all three taxa examined show that locomotion and posture do not influence FM position or orientation, and we conclude that this is likely in hominids as well. Anterior migration of the foramen magnum in early, small-brained hominids likely reflects changes in the relative sizes of the brain's components (i.e., an expanded neocortex) (Holloway, 1966 and Holloway et al., 2003), and as such is still a useful indicator of the behavior and biology of extinct species. Bipedality, however, should be inferred from the postcranium.

Owen Lovejoy is one of the authors of this paper, as well as upon the functional interpretation of Ardipithecus ramidus anatomy. To me, this statement is therefore especially important to the way we should interpret basicranial form in Ardipithecus. As stated clearly here, we cannot interpret posture or locomotion from the orientation of the foramen magnum.

As the study suggests, basicranial form in early hominins may reflect other constraints. For example, like the rodents considered in this study, the form of the posterior neurocranium in hominins might exert a constraint on the position of the brainstem.

Arguably the manner of bipedal locomotion in some rodents and marsupials is not comparable to hominins in the dynamics of the cervical spine and basicranium. A hopping kangaroo or kangaroo rat does not really have an orthograde spine, and the skull is still held in an anterior position even during bipedal locomotion. Further, these bipeds feed and carry out other behaviors with skull and spine in a habitually pronograde position.

But this criticism carries less weight for the strepsirrhine primates considered by Ruth and colleagues, which are often using a more orthograde posture arboreally, for both locomotion and feeding. The study shows that the foramen magnum angle is not related to locomotor pattern among these primates, but it is related to the relative size of the brain—a relatively larger brain tends to have a more inferiorly directed foramen magnum. The criticism is likewise questionable for hylobatids and spider monkeys which have a more orthograde posture both arboreally and terrestrially.

Some may argue that if hominins are truly unique in the degree of basicranial modification, that would make the basicranium of Sahelanthropus and Ardipithecus even stronger evidence of affinity with hominins, on the logic that such distinctive basicranial specialization is unlikely to result from parallelism.

I don’t buy that argument. We know too little about Miocene primates and their functional anatomy to know whether the hominin form is distinctive or unique. The best argument for its importance was the assumption of a link between basicranial form and bipedal locomotion. As Lovejoy and others have argued elsewhere, the anatomy of chimpanzees and gorillas are themselves derived in comparison to the human-African ape ancestral condition in important ways. There’s no reason to assume that the basicranial form of this common ancestor or other Miocene ape lineages was necessarily like living chimpanzees or gorillas. I conclude that we need to exert more skepticism about Sahelanthropus and Ardipithecus as relatives of early hominins. The proposed similarities in basicranial form to hominins in these genera are neither well-documented nor functionally understood.

This paper is a good illustration of how the comparative method helps us to understand human evolution. Anthropologists for fifty years have been explaining foramen magnum position in relation to locomotion almost as a mechanical necessity. As shown in this study, other bipedal mammals show that the basicranial form of hominins is neither mechanically necessary nor is the variation among species in the basicranium particularly related to the postural and locomotor demands of locomotion. This needs to come out of the textbooks.


Ruth AA, Raghanti MA, Meindl RS, Lovejoy CO. 2016. Locomotor pattern fails to predict foramen magnum angle in rodents, strepsirrhine primates, and marsupials. Journal of Human Evolution 94:45-52. doi:10.1016/j.jhevol.2016.01.003