What do capuchin stone tools tell us about human evolution?

A month ago at ESHE and now online in Nature, Proffitt and colleagues describe stone-on-stone smashing behavior among wild bearded capuchin monkeys (Sapajus libidinosus). The online paper includes a great video documenting the action; here’s a screenshot:


Holding the rock with both hands just above head-level, the monkey prepares to crush its enemies. Which in this case are another rock stuck in a pile of more rocks.

In the fairly rare cases where non-human primates use stones, it’s for smashing nuts or something. But when these capuchins see a stone they don’t just see a smasher, they see a world of possibilities* – why use a rock to break a rock, when you could use it to break a heart? So this group of capuchins is unique in part because they’ve been documented to use stones for many purposes.

Now why on earth a monkey would use one rock to break another rock is anyone’s guess. In human evolution, the purpose was to break off small, sharp flakes that could be used to butcher animals or work plants. Proffitt et al. did observe small flakes being removed when capuchins pounded rocks, but the monkeys showed little interest in this debitage, other than using it to continue smashing stuff. More curiously, the monkeys frequently lick the rock after hammering at it:


Mmm, rocks.

Proffitt et al. venture that maybe these monkeys are doing this to ingest lichens or trace elements like silicon. This hypothesis merits further investigation, but what’s clear is that these monkeys’ lithics differ from the hominin archaeological record wherein the express purpose of breaking rocks is to make flakes.

What’s striking to me (pun intended) is the relative size of the rocks. These monkeys that weigh only 2-3 kg are lifting and smashing stones that weigh about half a kilogram on average. Because these stones are fairly large given the monkeys’ body size, they have to be lifted with two hands and brought down on a surface, a “passive hammer” technique. The earliest-known tools made by hominins, from the 3.3 million year old Lomekwi site in Kenya, are also pretty big. Weighing 3 kg on average but topping at 15 kg, these earliest tools would have required the same knapping technique as is used by these little monkeys (Harmand et al., 2015).


Left: Cover of Nature vol. 521 (7552). Right: Bearded capuchin letting a pebble know who’s boss (link).

Why the big stuff at first? Did the earliest hominin tool-makers lack the dexterity to make tools from the smaller rocks comprising the later Oldowan industry? These creative capuchins could lead to predictions about the hand/arm skeleton of the Lomekwian tool-makers (testable, of course, only with fortuitous fossil discoveries). Capuchins are noted for their manual dexterity (Truppa et al., 2016) and have a similar thumb-index finger ratio to humans and early hominins (Feix et al. 2015), although they differ from humans in the insertion of the opponens muscle and resultant mobility of the thumb (Aversi-Ferreira et al., 2014). Maybe these tech-smart monkeys can tell us more about the earliest human tool-makers’ bodies than their brains.


Aversi-Ferreira RA, Souto Maior R, Aziz A, Ziermann JM, Nishijo H, Tomaz C, Tavares MC, & Aversi-Ferreira TA (2014). Anatomical analysis of thumb opponency movement in the capuchin monkey (Sapajus sp). PloS one, 9 (2) PMID: 24498307

Feix T, Kivell TL, Pouydebat E, & Dollar AM (2015). Estimating thumb-index finger precision grip and manipulation potential in extant and fossil primates. Journal of the Royal Society, Interface, 12 (106) PMID: 25878134

Harmand S, Lewis JE, Feibel CS, Lepre CJ, Prat S, Lenoble A, Boës X, Quinn RL, Brenet M, Arroyo A, Taylor N, Clément S, Daver G, Brugal JP, Leakey L, Mortlock RA, Wright JD, Lokorodi S, Kirwa C, Kent DV, & Roche H (2015). 3.3-million-year-old stone tools from Lomekwi 3, West Turkana, Kenya. Nature, 521 (7552), 310-5 PMID: 25993961

Proffitt, T., Luncz, L., Falótico, T., Ottoni, E., de la Torre, I., & Haslam, M. (2016). Wild monkeys flake stone tools Nature DOI: 10.1038/nature20112

Truppa V, Spinozzi G, Laganà T, Piano Mortari E, & Sabbatini G (2016). Versatile grasping ability in power-grip actions by tufted capuchin monkeys (Sapajus spp.). American Journal of Physical Anthropology, 159 (1), 63-72 PMID: 26301957

*well, at least four uses given by Proffitt et al.: mating display, aggression, food-crushing, and digging.


Zach, get over the robust australopithecines

Let’s return to my favorite group (though I think they’re not terribly closely related) in the human fossil record, the robust australopithecines. They have popped fairly frequently in the news this year, most recently regarding their possible use of bone tools <!–[if supportFields]> ADDIN EN.CITE Backwell21221217Backwell, Lucindad’Errico, FrancescoEarly hominid bone tools from Drimolen, South AfricaJournal of Archaeological ScienceJournal of Archaeological ScienceIn Press, Accepted Manuscripthttp://www.sciencedirect.com/science/article/B6WH8-4SPYKKN-2/1/388e9a02101a532d68bfcccbe79ba29b <![endif]–>(Backwell and d’Errico, in press)<!–[if supportFields]><![endif]–>. This group is funny. They first appeared probably some time around 2.7 million years ago, in the controversial form of Australopithecus aethiopicus. No one knows where exactly this enigmatic group came from, save for that the only known, fairly complete cranium (KNM WT 17000) has many primitive features, and is largely similar to the earlier A. afarensis. Around 2.3 million years ago or so, A. aethiopicus appears to be ‘replaced’ by A. boisei, whose face is less protruding, has smaller anterior teeth, and has a derived P3 morphology. This is all in East Africa, mind you. Then, some time probably around 2 million years ago, or a little later, a robust form (A. robustus) appears in South Africa, where erstwhile the only hominin was A. africanus (some argue that there are 2 species in the A. africanus hypodigm). Personally, A. robustus looks like a more ‘robust’ A. africanus: larger posterior teeth, more anteriorly placed cheeks—but there is much overlap in many of traits between these two taxa. And much to the chagrin of many cladists, the South and East African robusts appear fairly different morphologically; a recent study <!–[if supportFields]> ADDIN EN.CITE Gonzalez-Jose200821021017Gonzalez-Jose, RolandoEscapa, IgnacioNeves, Walter A.Cuneo, RubenPucciarelli, Hector M.Cladistic analysis of continuous modularized traits provides phylogenetic signals in Homo evolutionNatureNature775-77845371962008Nature Publishing Grouphttp://dx.doi.org/10.1038/nature06891http://www.nature.com/nature/journal/v453/n7196/suppinfo/nature06891_S1.html <![endif]–>(Gonzalez-Jose et al. 2008)<!–[if supportFields]><![endif]–> that examined hominoids morphometrically (that is, in terms of aspects of cranial shape) found the two robust taxa to be distinct (but that’s a topic for another post . . .). And all the while these buggers lived right alongside Homo, our ancestors! That’s some effed up stuff.

Now for the recent paper. Blackwell and d’Errico (in press) report on an assemblage of bone tools from the site of Drimolen in S. Africa (~2-1.5 million years ago). Drimolen is very near Swartkrans and Kromdraai, two other cave sites with a wealth of A. robustus, and to a lesser extent Homo, material has been recovered. First, how do they know these bones were tools? The tools were compared to other bones, known to have been worn (down) by other processes, like gnawing or carnivore chewing, and the tools appear quite different from these. Also, experimental studies of actually using bones as digging tools (to dig up underground tubers and especially to dig into termite mounds) have produced the same kind of wear as the fossil bone tools. Finally, many bone tools are known from the site of Swartkrans, which as I mentioned above is also laden with A. robustus remains. So it appears pretty likely that A. robustus (or Homo) was using bone tools to dig for foodstuffs around Drimolen and other cave sites in the early Pleistocene. *Note: our friend and colleague Julie Lesnik is currently finishing up her fieldwork, examining bone tools from Swartkrans, and observing chimpanzees digging for termites in Senegal; perhaps we could get her to write a good post on the topic.

Second, how do they know who used these tools? This an excellent question, which similarly plagues the postcranial material from these S. African sites (and even E. African sites, cf. the OH 7 hand and the OH 8 foot, again another topic for another post on another day . . .). The basic argument, similar to that for the postcrania, is that where these bone tools are found, the hominin assemblage is dominated by A. robustus remains; well, at least there are more A. robustus teeth and cranial material relative to comparable Homo fossils. So it’s guilt by association for the bone tools and postcrania here. Of course the only real way to test the hypothesis that most S. African postcrania and bone tools are to be associated with A. robustus and not Homo would be to find a complete skeleton of either, and to find bone tools in strong association (i.e. almost still in the clutches) of a certain taxon. Another argument used in support of robustus as the bone tool user is the fact that stone tools—the hallmark of early humankind?—are relatively absent at these sites. So it’s probable, but certainly not unequivocally proven, that A. robustus was the user of these stone tools (note that the authors never make any claim of such proof).

The authors use the bone tool assemblages from Swartkrans and Drimolen to infer ‘cultural’ behavior patterns for A. robustus. A recent paper claimed that since larger (male) A. robustus cranial specimens tended to be older than smaller ones, that this was evidence for extended male growth in the taxon <!–[if supportFields]> ADDIN EN.CITE Lockwood20079917Lockwood, Charles A.Menter, Colin G.Moggi-Cecchi, JacopoKeyser, Andre W.Extended male growth in a fossil hominin speciesScienceScience1443-1446318Australopithecus robustusdelayed maturation200730 November 2007doi:10.1126/science.1149211<![endif]–>(Lockwood et al. 2007)<!–[if supportFields]><![endif]–>. This, coupled with the pronounced sexual dimorphism in this taxon (and all hominins until only fairly recently), suggested Lockwood et al. that A. robustus had a social structure similar to that of gorillas, in which one or two males associate with a number of females and their offspring (a simpler answer is that selection favored larger males, and so they were living longer, but Mary and I had trouble trying to demonstrate this). So Blackwell and d’Errico figure that this evidence, along with data from chimpanzee termite-foraging behavior, suggests that females were the predominant practitioners of termiting using bone tools.

Why do I think this is interesting? I think this shows a potentially very important ecological divergence between the robust australopithecine lineages on the on hand, and between A. robustus and Homo on the other. Could the disparate toolkits of these hominins have played an important part in the evolution of these lineages? It is debated whether robust australopithecine hands were physically capable of actually making stone tools (which is quite difficult). Part of this stems from the fact that it’s hard to find a complete hand (damned beetles…) and then to attribute it to a specific taxon. For instance, the OH7 partial hand, which was supposed to be part of the H. habilis holotype, was recently found to be most similar morphologically to A. robustus from S. Africa, and functionally not adapted to tool-making <!–[if supportFields]> ADDIN EN.CITE Moyà-Solà200826126117Moyà-Solà, S.Köhler, M.Alba, D. M.Almécija, S.Taxonomic Attribution of the Olduvai Hominid 7 Manual Remains and the Functional Interpretation of Hand Morphology in Robust AustralopithecinesFolia PrimatologicaFolia Primatologica215-2507942008http://www.karger.com/DOI/10.1159/000113458 <![endif]–>(Moyà-Solà et al. 2008)<!–[if supportFields]><![endif]–>. Of course, this relies on assumptions about whether hand material from Swartkrans represents A. robustus or Homo. Nevertheless, it may well be that part of the adaptive divergence of A. robustus and Homo was the former’s use of bone tools to dig for termites, while the latter was able to manipulate stone to exploit higher quality resources. Cool.

Also significantly, I think this points to more evidence against uniting A. robustus and boisei into the genus Paranthropus. Yes, they both had large posterior teeth (though boisei’s were generally larger) and faces built to house large chewing muscles. But other than that they appear pretty different. As noted above, a recent study looking at modularized traits in crania found that the two robust groups were not monophyletic, which is a criterion when making taxonomic decisions (Gonzalez-Jose et al. 2008). A. robustus has striking affinity with the earlier A. africanus, while A. boisei has striking affinity, to the exclusion of A. robustus, with A. aethiopicus, suggesting that A. robustus and boisei didn’t share a common ancestor. It has also been argued that sharing an ecological space is another criterion for generic membership <!–[if supportFields]> ADDIN EN.CITE Wood199926226217Bernard WoodMark CollardDepartment of Anthropology, George Washington University, 2110 G Street NW, Washington, DC 20052, USA; Human Origins Program, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560, USA.; Department of Anthropology, University College London, Gower Street, London WC1E 6BT, UK.The changing face of genus HomoEvolutionary AnthropologyEvol AnthropolEvolutionary Anthropology195-2078619991520-6505http://dx.doi.org/10.1002/(SICI)1520-6505(1999)8:63.0.CO;2-2 10.1002/(SICI)1520-6505(1999)8:63.0.CO;2-2<![endif]–>(Wood and Collard 1999)<!–[if supportFields]><![endif]–>. But isotopic and dental microwear evidence show A. robustus and boisei to have been distinct. Also, (shame on me) I’m not sure the extent to which A. boisei is associated with bone tools, but to my knowledge it is not. So here we have morphological, geological, temporal, and ecological discontinuity between the two main ‘robust’ taxa. That pretty much sinks Paranthropus for me. I don’t know why I am so adamant about these taxa’s paraphyly, but I am. I feel that taxonomy should reflect important biological and phylogenetic reality. And it seems to me that because there isn’t compelling evidence that A. robustus and boisei share the same ancestor, or interbred, that they shouldn’t be taxonomically separated from the other australopithecines. And that’s my two cents.


Backwell L, d’Errico F Early hominid bone tools from Drimolen, South Africa. Journal of Archaeological Science In Press, Accepted Manuscript

Gonzalez-Jose R, Escapa I, Neves WA, Cuneo R, Pucciarelli HM (2008) Cladistic analysis of continuous modularized traits provides phylogenetic signals in Homo evolution. Nature 453(7196):775-778

Lockwood CA, Menter CG, Moggi-Cecchi J, Keyser AW (2007) Extended male growth in a fossil hominin species. Science 318:1443-1446

Moyà-Solà S, Köhler M, Alba DM, Almécija S (2008) Taxonomic Attribution of the Olduvai Hominid 7 Manual Remains and the Functional Interpretation of Hand Morphology in Robust Australopithecines. Folia Primatologica 79(4):215-250

Wood B, Collard M (1999) The changing face of genus Homo. Evolutionary Anthropology 8(6):195-207<!–[if supportFields]><![endif]–>