Erik Trinkhaus, Tianyuan 1 and Sunghir 1, and the Earliest ‘Evidence’ of Footwear

Erik Trinkhaus published a study, along with Hong Shang, in the July issue of the Journal of Archaeological Science on what they consider the earliest evidence of footwear of modern humans. The paper is titled, “Anatomical evidence for the antiquity of human footwear: Tianyuan and Sunghir.” They compared and contrasted the morphology of a couple foot bones of humans to human ancestors to see if they can find the effects of shoe wearing on foot bone anatomy.

Before we get into this, I want you to be aware of a couple things. First, the earliest direct evidence of footwear, and by direct I mean actual artifacts of shoes, that I know of are mostly complete sandals from California that date to 9,000 years ago. Other evidence comes from fossilized footprints. Looking at the effects of footwear on foot bones is indirect.

Second, this is not Trinkhaus’ first attempt at establishing the earliest record of footwear. In 2005, he published, “Anatomical evidence for the antiquity of human footwear use,” in the same journal as this current paper. In that paper he sampled foot bones from humans as old as 100,000 years to 10,000 years. He specifically looked for changes in human foot anatomy.

This earlier paper relied on an assumption that shoe wearing causes phalanges or toe bones to become more delicate, or gracile. In other words, the four smaller toes on each foot of people that walk barefoot, flex to allow better traction. This torsional flexion promotes robusticity and sturdy phalanges. In contrast contrast, supportive footwear, like sandals and tennis shoes, lessen the load and force on the four small toes, thus ‘weakening’ them. I’ve seen it myself, in people from the Caribbean and Africa. Often times these people do not wear shoes, or they wear poor shoes. Their feet are much wider than mine and other shoe wearing people, so are their toes.

In the 2005 paper, Trinkaus looked at the toes of western Eurasian human skeletons from the Middle and Upper Paleolithic. He saw that the anatomy of the foot began to change around as early as 30,000 years ago, becoming more gracile than robust Neandertals and early Homo sapiens.

As you can tell from the title of the current study, the two subjects are Tianyuan and Sungir. The specimen from Tianyuan, Tianyuan 1, was actually published by Hong Shang as well as other authors (link is dead, I know). The Tianyuan 1 specimen is represented by a partial mandible and dentition, limited axial remains, portions of all long bones as well as some hand and foot bones. It was discovered in the Tianyuandong, a.k.a. Tianyuan Cave near Zhoukoudian, China. Radiocarbon dating of faunal remains from the same stratigraphic level indicate the Tianyuan 1 remains are at most 39,000 years old.

Sunghir 1 is also a partial skeleton, but it is not from China. In fact, Sunghir was found in a site on the outskirts of a Russian city called Vladimir. It has also been radiocarbon dated. It is roughly 23,000 years old. The interesting thing about Sunghir 1 is that it was excavated with an arrangements of beads sewn onto clothing that was wrapped around the feet — which is interpreted as evidence of footwear.

Trinkhaus and Shang compared Tianyuan 1 and Sunghir 1 foot phalanges to phalanges of archaic humans: Neandertals, humans from the Middle and Upper Paleolithic. They also compared them to foot bones from modern humans: Pueblo Indians, Inuits, and European Americans. The comparison involved the measurement of the cross-sectional area of the phalanges, and looked for the amount of torsion caused by the force of not wearing shoes. This value was coined the polar moment.

The results of the polar moments of area versus the natural log of phalangeal length times body mass is represented to your right. As you can see Tianyuan 1 and Sunghir 1 fall within the range of Inuit and European Americans. But, some European Americans and Inuits also exhibit the same polar moment as Neandertals and archaic humans from the Middle and Upper Paleolithic. Those that do have robust toes could have been non-shoe wearers, they still exist — believe it or not!

If the graph is making your mind spin, check out this photographic array of the Tianyuan 1 and Sunghir 1 phalanges compared to robust, presumably non-shoe wearing human ancestors, Qafzeh 8 and Kiik-Koba 1, a Neandertal. You don’t need fancy pants charts to see how narrow and delicate the Tianyuan 1 and Sunghir 1 bones are compared to Qafzeh 8 and Kiik-Koba 1. But note that bones are not resized to the same scale. Why not?

It is really hard to infer a behavior from one measurement, but I’m pretty convinced. So people were wearing shoes 10,000 years earlier than previously assumed from indirect evidence — definately plausible. People were living in extreme environments for hundreds of thousands of years and needed to protect their feet from the elements!

My most pressing issue is why the bones weren’t adjust to scale in this photo. If they were, we would have a better comparative framework to make the visual assessment ourselves. What if the two TY1 phalanges are longer than the KK1 phalanges? They sure appear to be. If they scaled them down to the length of KK1 phalanges, and the TY1 phalanges appeared to be just as wide as KK1, then Trinkhaus and Shang have a problem. It’s really simple to do. A couple minutes on Photoshop would do the trick. So long as Trinkhaus and Shang  provided the normal photo alongside the scaled photo, there would be no worries of manipulation.

    TRINKAUS, E., SHANG, H. (2008). Anatomical evidence for the antiquity of human footwear: Tianyuan and Sunghir. Journal of Archaeological Science, 35(7), 1928-1933. DOI: 10.1016/j.jas.2007.12.002
    TRINKAUS, E. (2005). Anatomical evidence for the antiquity of human footwear use. Journal of Archaeological Science, 32(10), 1515-1526. DOI: 10.1016/j.jas.2005.04.006

25 thoughts on “Erik Trinkhaus, Tianyuan 1 and Sunghir 1, and the Earliest ‘Evidence’ of Footwear

  1. Excuse me? How else would you explain a change in the “anatomy of the foot” due to the use of footwear if not through Darwinian adaptation, which takes place via biological inheritance, no?

  2. Victor,

    What’s with the ‘excuse me’ tone? The effects of shoe wearing on the anatomy of the foot is not discussed as an inherited trait. It is discussed as an induced trait.

    Furthermore, this paper doesn’t discuss if people are inheriting a gracile toe bone trait throughout generations. This paper discusses that footwear, a cultural trait, effects the morphology of feet.

    If I decide not to wear shoes, my toe bones will be under more torsional force and will restructure themselves to cope with the added pressure. If you decide not to wear shoes, your toe bones will do the same. Wearing shoes relaxes pressure on the bones of the toe, causing them to become more gracile.

    The same discussion could be made for other types cultural modifications on the body — most notably Mayan skull modifications. By placing a board on the developing skull of a child to mold the skull into a flattened or cone shape. No genetic modification was necessarily inherited in this scenario. But the pressure of the board affected the growth and development of the skull — much like shoes affect the growth and development of the foot.


  3. Sorry, Kambiz, but my tone reflected an extreme puzzlement and consequent skepticism regarding what was apparently being claimed. It wasn’t addressed to you but to Trinkhaus, because it seems that he is indeed implying a Darwinian process at work.

    What YOU say makes a lot more sense. Certainly a lifetime of shoe wearing can reshape the foot just as a child’s skull can be reconfigured through use of a board.

    I’m not about to pay $30 to read either of these articles in full, but here is a quote from the abstract of one of them: “This interpretation is based principally on the marked reduction in the robusticity of the lesser toes in the context of little or no reduction in overall lower limb locomotor robusticity by the time of the middle Upper Paleolithic.” Can a reduction in the “robusticity” of certain bones be achieved so easily, just by wearing shoes for a few years? And if so, wouldn’t there be marked differences between specimens from younger and older individuals?

    If he’d referred to changes of position or shape I’d understand. But “robusticity” implies an inborn condition, does it not?

  4. Can a reduction in the “robusticity” of certain bones be achieved so easily, just by wearing shoes for a few years?

    AFAIK, yes. Even if it’s not for just a few years but rather a lifetime. I understand that this issue of the robusticity of toes related to shoe-wearing has been known for some time now, not from archaeological remains but from actual living people. I am not going to google on this for you but it’s not anything new in any case.

    And if so, wouldn’t there be marked differences between specimens from younger and older individuals?

    Yes. If you compare the foot of a baby of a shoe-wearing group with the foot of a baby of a barefooted tribe, you should notice no differences.

    If he’d referred to changes of position or shape I’d understand. But “robusticity” implies an inborn condition, does it not?

    No. Not necesarily. There can be individual variation for sure but there are general patterns too, just like people who walk a lot develope certain traits or people who ride on horse a lot do as well. In this case the lack of use of some parts of our body weakens them. Toes do not need to work the same with shoes and specially if these are hard-soled shoes, the effect in the small toe seems to be the most noticeable one.

  5. “Yes. If you compare the foot of a baby of a shoe-wearing group with the foot of a baby of a barefooted tribe, you should notice no differences.”

    Sorry to be so persistent on this point, or maybe I’m just being stubborn. That’s been known to happen.

    But I’ve never seen a human baby or child with “robust” toes. If little children now have gracile toes is that how Trinkhaus things we got them, by wearing shoes?

  6. I think you are misconstruing the meaning of robust and gracile. In point of fact, yes, some babaies have more robust or more gracile toes than others. What Trinkaus is talking about relates to human plasticity. Bone only has two responses to stress. It can absorb bone or lay down more bone. Put more stress on one area and bone will be deposited. Relieve stress and bone will be resorbed. The interplay between the two can remodel bone in interesting fashions. Such changes being accrued during the lifetime of an individual are not genetic and are not passed on to the offspring via the genes, rather they are the result of a common cause having a common effect.

  7. actually, I misspoke. In thinking about it and consulting several juvenile osteology books I have to agree with the statement that in terms of robust or gracile, babies toes are about the same. To follow up on the robusticity comment. Robusticity does not describe an inborn trait, rather it has to do with size and shape.

  8. Sorry but it still looks to me as though Trinkaus has been making unwarranted assumptions — or is simply confused. In all my years of walking with shoes I have noticed exactly NO changes in the anatomy of my feet, except for the occasional callous. Shoes are designed to preserve the feet, NOT alter their bone structure.

    If people were to continually walk barefoot over rough terrain, then I’d imagine the “anatomy” of their feet might well be altered and might well wind up looking relatively “robust,” i.e., rough, over time. But that’s NOT what Trinkhaus claims. He claims the opposite, that our feet begin as “robust” and then become gradually more “gracile” over time due to the use of footwear.

    Actually it’s not at all clear what he’s claiming because frankly he seems confused about what he’s claiming. Judging from the abstract I read, it sounds as though he actually IS at least suggesting that human feet became more “gracile” over time (i.e., thousands of years) due to the use of footwear. He describes this as a gradual process, implying a gradual process over many generations. If it were a gradual process during the course of a single person’s lifetime, then that would be very easy to check, simply by consulting with a foot doctor!

    In other words, if Trinkaus’s argument hinges on anatomical changes occurring during the course of each individual’s lifetime, then it should be just as possible to track those changes among living humans as among fossil remains. And young humans today should have toes that are every bit as “robust” as the toes of young Neanderthals, prior to their use of footwear (assuming they used it at all).

  9. It is possible to track those changes in living populations and Trinkaus draws on some living populations in his study. It is important to note as well, that Trinkaus creates a biomechanical model to draw predictions from, the predictions are then tested against the skeletons drawn from the archaeological record. I think part of your misunderstanding comes from not reading the articles in question. It also seems like you are trying to shoehorn the study into an area where it does not fit.

  10. Another misconception that Victor isn’t explicitly making, but many do, is concept that bone morphology doesn’t change after puberty. I know you addressed this, Afarensis, in one of your earlier comments.

    But to reiterate, bone is constantly adapting and changing to the forces like gravity and pressure as well as the nutrition of the individual. Individuals lucky enough to go into space, lose their bone density because their bodies don’t feel the effects of Earth’s gravity as much as we do, on land. Their bodies almost immediately stop laying bone tissue down to help cope with the force of gravity.

    Contrary to the popular understanding of bone tissue, it is always changing and adapting to the stresses of life — be it not wearing shoes or life in outer space.


  11. It makes more sense to assume that bone tissue will be distorted as a result of walking barefoot over rough terrain then that it will be distorted as a result of being “coddled” by the use of footwear. Because the footwear will tend to preserve, rather than distort, the original condition of the toes at birth. (I’m talking normal footwear, obviously, not foot-binding, which is a whole other issue.)

    So when Trinkhaus writes about the “marked reduction in the robusticity of the lesser toes in the context of little or no reduction in overall lower limb locomotor robusticity by the time of the middle Upper Paleolithic,” it’s hard for me to understand what he’s getting at. If he’s arguing that people’s toes were originally robost as a result of distortion from a lifetime of walking barefoot (rather than at birth), and that this process of distortion stopped after their feet were protected by footwear, that would make some sense, for sure. But that’s not what he’s saying. At least not in the abstract of his 2005 paper.

    What he appears to be saying is that the inborn robustness of Neanderthal and early “modern” feet (as reflected by the inborn robustness of the “lower limbs” to which he refers) would have been preserved through walking barefoot, but then distorted into gracile form after the introduction of footwear.

    Part of the problem I’m having with all this has to do with the terminology, which I find misleading. “Robust” vs. “gracile” is usually used in reference to inborn characteristics of early and later human forms, respectively. If the intention were to point to how certain bones changed during one individual’s lifetime, I would think the proper term would be “distorted,” rather than “gracile.”

    I’m wondering if there’s anyone out there who’s actually read Trinkaus’ paper(s) and might find a specific quotation that could clarify his position on all this? Because as I say there seems something either very wrong or at least very ambiguous about his thinking, as far as I can see.

  12. I’ve read the paper and I want to clarify a couple things. I termed the characteristics as gracile but that nor robust does not imply inborn characteristics. Gracile and robust are just descriptions of the morphology of the bone. An individual can use the description in a genetic/inheritance context, i.e. my checkbones are robust like my father’s or they can use the description in a environment/behavior = phenotype context. i.e. repetitive strain on the knee joint lead to a robust restructuring of bone.

    I see where you maybe getting a bit mixed up, the last sentence in the 2005 abstract has this phrase, ‘by the time of the middle Upper Paleolithic,’ which implies there was some sort of inheritance/selection. But he’s saying that Middle Paleolithic foot bones are robust and presumably of individuals that didn’t wear footwear, but the lack of robusticity became more frequent by the middle Upper Paleolithic as a result more individuals wearing footwear.


  13. Victor – the problem is you are using idiosyncratic definitions of the words gracility and robusticity and trying to force the Trinkaus study into those definitions. As Kambiz points out, the definitions of these terms used in paleoanthroplogy have nothing to do with “inborn characteristics”. Typically, robusticity is defined as a ration of length to diameter – at least in reference to long bones. Another definition has to do with the size of the articular surfaces. A third has to do with the prominence of muscle markings (the term rugosity is also used in this connection). Incidentally, although both terms have been applied to various and sundry hominins (gracile australopiths vs. robust australopiths, etc) robust does not equal early and gracile does not equal late human. In both studies Trinkaus used articular length of the metatarsals and both midshaft diameters (anterior-posterior and mediolateral) scaled to body mass. At any rate, here is one quotation that makes things pretty clear:

    Since the foot provides the contact between the body and the substrate, and since the use of footwear with a semi-rigid sole will alter the distribution of mechanical forces through the foot, it might be possible to perceive differences in the relative hypertrophy of portions of the foot in response to changes in habitual biomechanical loads through the pedal skeleton.

    This next quote comes from the Trinkaus and Shang paper:

    There is marked decrease in the robusticity of those bones, in the context of little change in overall lower limb robusticity (Trinkaus, 2006a), between the Middle Paleolithic early modern and late archaic humans on the one hand, and Middle Upper Paleolithic (Gravettian sensu lato) humans on the other hand. Since there is no meaningful change in the overall biomechanical load levels placed on these Late Pleistocene human lower limbs, through mobility and burden carrying (Trinkaus, 2006a), this reduction in lesser toe robusticity can only be interpreted as indicating localized mechanical insulation from
    ground reaction forces during heel-off and toe-off. Since the pedal digital flexor muscles plantarflex the toes into the substrate in these latter portions of the stance phase (Reese et al., 1983), an artificial reduction, or dispersal, of that ground reaction force must have taken place.

    So what Trinkause is saying is that wearing shoes caused an alteration in the biomechanical forces affecting the metatarsals. The metatarsals responded to the changed stress by altering the patterns of bone deposition and resorption.

    Look at it this way. If you were to measure your heart rate, pulse, respiration, etc., now, then run a mile and measure them again they would be completely different. Gradually, they would return to the measurements obtained while you were at rest. Such a change is neither transmissible to your offspring nor a distortion of your basal metabolism, rather they represent a normal physiological reaction of your body to increased stress and its subsequent reduction. Same deal with wearing shoes, or not wearing shoes. Your metatarsals undergo a physiological change and respond to that stress and/or lack thereof.

  14. Thanks Kambiz and Afarensis for your patience and your explanations. But I’m still troubled by Trinkhaus’ failure to clearly and unambiguously distinguish — as you have — between the two very different modes of “evolution,” the one genetic, the other cultural. Not that I accept for a moment the inheritance of acquired traits — but that I wish Trinkhaus had made his position clearer, because judging from the secondary literature I’m seeing on the Internet regarding these articles, there does seem to be some confusion out there regarding that issue.

    What interests me more, however, than the question of the origin of footwear, is the broader issue that’s been raised in our discussion regarding the difference between three very different modes of assessing physical remains:

    1. as representing a condition determined purely by genetic inheritance.

    2. as representing a condition determined largely by environmental conditions.

    3. as representing a condition determined largely by culture.

    This may already be old news to you guys, but I’ve never thought of it this way until now. Not only the feet, but many parts of the body could very well be altered by either the environment or cultural practices or both, to the point that it could be all but impossible to meaningfully distinguish among the three possibilities. For example the “robust” condition of the lower limbs that Trinkhaus finds in the Neanderthal samples could have been caused by stresses induced through walking or running day in day out over very rough terrain. Hands could very quickly become gnarled in similar fashion over time, giving the bones a more “robust” appearance than they’d have had if these people had lived under less stressful conditions. And who knows what the Neanderthal toes might have looked like if Neanderthals had been farmers living relatively quiet lives in sedentary communities.

    It seems as though the only way to make an assessment in purely genetic terms — i.e., strictly in terms of inheritance ruled by Darwinian adaptation — would be to examine the skeletal remains of small children.

  15. Actually, if you reread the two quotes I provided, which came from the papers in question, you will find that Trinkaus did, in point of fact, make the kinds of distinctions you are talking about.

  16. afarensis, I’m not sufficiently acquainted with the technical terminology to completely follow either of those quotations. However, I AM aware of the general nature of the argument he appears to be making and I still see ambiguity, if not confusion, there. Clearly he is referring to alterations in bone structure due to the presence or absence of footwear, I have no problem getting that point.

    What is not completely clear to me is whether he sees such alterations as occuring over thousands of years due to some form of gradual “adaptation,” or over the lifetime of a single individual. I was hoping that you or Kambiz might have found some quote from one of the papers that makes THAT point clear, but apparently not. (Unless there is something in the quotes you provided that does make that particular point unambiguously but is over my head.)

    It seems to me that if he really wanted to make the point you guys are seeing, he’d have taken a look at the way human feet develop over time today. In other words, the simplest and most logical way to assess what happens to feet during a single lifetime of wearing shoes is to compare the feet of very young humans with those of older humans from the same general population. And IMO I really doubt very much that you would find dramatic differences of the sort he’s referring to. Because what that implies is that the feet of very young children in, say, New York City or LA or Boston or wherever, would have a more “robust” bone structure than those of adults. That makes no sense to me.

    I’m also bothered by the assumption that the earlier, “robust” state of these toe bones represents their default condition, because clearly if footwear can distort those bones then certainly a lifetime of walking barefoot over rough terrain can also distort them. Which means that the correlation between the “robust” state of the leg bones and the “robust” state of the foot bones that he finds in the older samples is very unlikely to mean what he implies it means.

    I realize I’m over my head here as far as the technical stuff is concerned, so I could be totally off base. But so could be the many thousands of people reading all the many reports on this research, who might very well get the totally wrong idea that human feet “evolved” from “robust” to “gracile” over thousands of years due to the use of footwear.

  17. What seems to be lacking in this discussion — other than a reading of the original article in full by all participants — is an understanding of the metabolic/energetic cost of maintaining a robust body part. Let us suppose that a hitch-hiking ancestor of ours, as yet undiscovered, had a grossly enlarged Sissy Hankshaw-like thumb that was essential to life. Let us then suppose that hitch-hiking as a behavior became extinct (say, due to unavailability of fuel). Those descendants of our hitch-hiking ancestor who continued to produce babies with giant thumbs, which required energy to build and maintain, would be at an evolutionary disadvantage. Those who had the genes for smaller thumbs — who were once AT a disadvantage when hitch-hiking was all — would after the cessation of hitch-hiking be an an advantage. Over time, the behavioral change ought to be reflected in an anatomical change (reduction of the thumb) because the giant thumb was no longer advantageous.
    Similarly, Trinkhaus’ show hypothesis suggests that robust toes, once no longer needed, would slowly evolve toward less robust toes because there was no pay-off for the cost of making them and maintaining them.
    Think about it. This IS biological evolution as well as cultural evolution.

  18. Pat the Anthropologist,

    With all due respect, I know who you are and do not appreciate your shameless jab, where you said there was a “lack of reading of the original article.” I read the article and find no explicit nor implicit discussion on the genetics of robust toe development in this paper, nor is there any inference to the heritability or selection of robust toe development throughout evolutionary time. That being said, your argument is out of the scope of this paper.

    Trinkhaus examines the observation that footwear affects the robusticitiy of toe development in the fossil record. In his studies, he notes that footwear reduces strain from the toes, and the bones of the feet deposit less bone in response. The toe bones ultimately appear to be more gracile and have a different morphology than toe bones of individuals who were not habitual footwearers. He looks at the fossil record and investigates samples to see if they resemble the bones of individuals who wear footwear, or if they resemble the bones of individuals who walk barefoot. He finds that in two samples, the bones resemble that of a footwearing individual and suggests footwear was present 40,000 years ago.

    From what I can tell, you’re making a completely different claim, again out of the scope of this paper, which suggests that gracile toes, an inherited trait, possibly fit into shoes better and affected the survivability of the individual… Offering a selective advantage.

    Furthermore, your argument suggests there has been a positive selection towards gracile toes — but looking at the figure from the paper (which I linked above) we see that in modern populations, individuals exhibit a wide variation phalange morphology. This doesn’t support your hypothesis. If gracile toes were at one period more advantageous than robust ones, we wouldn’t be seeing a disproportionate amount of robust toes, would we?

    I believe this is your first time commenting here, and since you decided to post behind an alias, I decided to respect your decision to remain pseudonymous. But, like I said, I know who you are and in the future, will not hesitate to call you out if you say that all participants didn’t read the paper and then offer up an argument that is completely out of the scope of the paper being discussed.


  19. I do not know who Pat is, but I take exception to his characterization of us as people who have not read the paper – especially when I quoted from both the papers in question in a previous response. Trinkaus was quite clear that both studies were directed towards the question of when people started wearing shoes and with the morphological correlates thereof.

  20. Pat wrote as follows:
    “What seems to be lacking in this discussion — other than a reading of the original article in full by all participants —”

    In other words the problem is that ALL participants haven’t read the article, not that none have read it. So calm down, guys, I’m sure Pat’s comment was not directed at you. It was directed at me — and I don’t mind.

    As far as Pat’s theory is concerned, as I see it, that’s pretty close to the sort of thing Trinkaus may have had in the back of his mind. And it’s also more or less the impression I get from reading the media reports. But maybe the worst aspect of this whole matter is that the difference between changes wrought during a single person’s lifetime and changes wrought over thousands of years appears to have been left ambiguous.

    In any case, assuming Trinkaus was referring to the former and not the latter, then we are left with the assumption that all human toe bones, down to the present time are innately “robust,” and only become “gracile” through the wearing of footwear, which looks to me like a huge boner.

    And if the latter, then, with all due respect to Pat, the adaption would have had to take place within a very short time, not from the Paleolithic to the present, but from an earlier point in the Paleolithic to a somewhat later one. Far too short a time for the process Pat describes to have taken place.

    Sorry but Trinkaus leaves me unimpressed.

  21. This discussion is fascinating, even to a lay person such as myself. I’m wondering if there is any consensus of thought about when and where early man first fashioned sandals or some kind of wrap to protect his/her footsies. Thanks.

Comments are closed.

A Website.

Up ↑

%d bloggers like this: