The world’s oldest graffiti: by Homo erectus! (maybe)

December 4, 2014 • 11:47 am

Look at the shell below, which has been dated to 500,000 years ago. See the scratches? Those represent the oldest human etchings, or graffiti, ever found, preceding the next oldest by 300,000 years! Some anthropologists dissent, but more on that at the end of this post.


As Science reports, this shell was found by a graduate student Stephen Monro on Java in 2007, and now, 7 years later (yes, they studied it intensively), he and his colleagues have concluded not only that this was a human production—perhaps made by Homo erectus, who lived on Java at that time—but that they also used the shells as tools.

The full paper is in Nature, with the link and reference below (I believe it’s free). I’ll summarize what’s in the paper, and then go back to the Science report for a dissent by scientists.

The shell was part of an apparent cache of fossil mussel shells excavated in 1890 in Trinil, Java, which happens to be the type locality for Homo erectus, discovered by Eugène Dubois in 1891-1892 and originally called “Java Man“.  (We’re not quite sure what happened to H. erectus: it could have interbred with expanding Homo sapiens populations coming from Africa beginning about 60,000 years ago, or it could have gone extinct without leaving descendants.) The dating of the shells from sediments enclosed in them gives an age of about 540,000 ± 10,000 years. Only H. erectus was in Java then.

The shells, too, were collected by Dubois, and, based on their uniform large size and the holes in them (see below), as well as other suggestive human modifications, were probably a midden of sorts left by H. erectus individuals who had eaten the mussels.

Now, the shells show three signs of modification by humans, including the “graffiti” above.

  • They have holes drilled in them, likely by a human using a shark tooth. They are consistent in location and size, near the rear adductor muscle, and aren’t consistent with holes made by other predators like marine snails or otters.  The authors found that you can make these holes with a shark tooth, and if it’s driven into just that spot in the rear, the mussel can’t use its muscle to keep the shell closed and opens, presenting its contents for food. Here’s a photo of the shell holes, presumably made by hominins (all species in our clade after we diverged from the lineage that produced modern chimps). The figure and caption are from the original paper. Similar holes were made by the native (pre-European) inhabitants of the Caribbean to open gastropods.

Screen Shot 2014-12-04 at 12.23.44 PM

  •  Some of the shells appear to have been modified for use as cutting or scraping tools. Their outer layer was scraped off down to the nacreous (“pearl-like”) inner layer, making them sharp. These shells show wear patterns consistent with them having been used as tools.
  • Most interesting, of course, is the graffiti: a poignant remnant of an early human (and yes, they suggest it was a single “person”). The journal’s description of this follows, along with another photo:

“One of the Pseudodon shells, specimen DUB1006-fL, displays a geometric pattern of grooves on the central part of the left valve (Fig. 2). The pattern consists, from posterior to anterior, of a zigzag line with three sharp turns producing an ‘M’ shape, a set of more superficial parallel lines, and a zigzag with two turns producing a mirrored ‘N’ shape. Our study of the morphology of the zigzags, internal morphology of the grooves, and differential roughness of the surrounding shell area demonstrates that the grooves were deliberately engraved and pre-date shell burial and weathering (Extended Data Fig. 5). Comparison with experimentally made grooves on a fossil Pseudodon fragment reveals that the Trinil grooves are most similar to the experimental grooves made with a shark tooth; these experimental grooves also feature an asymmetrical cross- section with one ridge and no striations inside the groove. We conclude that the grooves in DUB1006-fL were made with a pointed hard object, such as a fossil or a fresh shark tooth, present in the Trinil palaeoenvironment. The engraving was probably made on a fresh shell specimen still retaining its brown periostracum, which would have produced a striking pattern of white lines on a dark ‘canvas’. Experimental engraving of a fresh unionid shell revealed that considerable force is needed to penetrate the periostracum and the underlying prismatic aragonite layers. If the engraving of DUB1006-fL only superficially affected the aragonite layers, lines may easily have disappeared through weathering after loss of the outer organic layer. In addition, substantial manual control is required to produce straight deep lines and sharp turns as on DUB1006-fL. There are no gaps between the lines at the turning points, suggesting that attention was paid to make a con- sistent pattern. Together with the morphological similarity of all grooves, this indicates that a single individual made the whole pattern in a single session with the same tool.”

Here’s another figure showing this great specimen, along with its caption:

Figure 2 | The geometric pattern on Pseudodon DUB1006-fL. a, Overview. b, Schematic representation. c, Detail of main engraving area. d, Detail of posterior engravings. Scale bars, 1 cm in a and c; 1 mm in d. See also Extended Data Figs 5 and 6.

And finally, the paper’s conclusion, adding that the authors see this as the earliest use of a natural material to make tools.

The combined evidence for high-dexterity opening of shells, use of shell as a raw material to make tools, and engraving of an abstract pattern on a shell with a minimum age of 0.43 ± 0.05 Myr indicates that H. erectus was the agent responsible for the exploitation of freshwater mussels at Trinil described here. The inclusion of mussels in the diet of H. erectus is not surprising, as predation on aquatic molluscs is observed for many terrestrial mammals, including primates. The reported use of shells as raw material for tool production is the earliest known in the history of hominin technology. It may explain the absence of unambiguous stone artefacts in the Early and Middle Pleistocene of Java, possibly the result of poor local availability of lithic raw material, as also suggested for the much younger (about 110,000 years old) Neanderthal shell tools from Italy and Greece. Our discovery of an engraving on shell substrate is unexpected, because the earliest previously known undisputable engravings are at least 300,000 years younger.

Leaving aside the question of whether these “graffiti” are anything other than a hominin fooling around, or have a more symbolic meaning, there’s at least one anthropologist who takes issue with the claim that these patterns were made by H. erectus. As the Science report notes, the dating could be off, and the scratches made by more recent individuals of H. sapiens:

Yet even if ancient humans engraved the shell, says Russell Ciochon, a paleoanthropologist at the University of Iowa in Iowa City, the team has not shown that H. erectus did it. Ciochon, who has spent many years working at sites in Java, agrees with criticisms that the shells have been taken out of context, because Trinil was not an occupation site where early humans actually lived. Rather, Ciochon argues, the human fossils found there (which include a skullcap widely agreed to be H. erectus and a thigh bone that could belong to either H. erectus or H. sapiens,a matter of sharp debate) were washed into the site by a powerful flood, and nothing found with them—including the shells—can be assumed to have been associated with them originally. Although the team dated four of the shells in the collection, including the engraved shell, to about 500,000 years ago using two different techniques on sediments of sand and clay found inside them, Ciochon says that those sediments could have entered the shells during the earlier flood event that created the site, and that H. sapiens still could have come along much later and performed the etching.

I don’t have the expertise to judge how powerful this criticism is, but if it’s accurate, then the whole story goes out the window pending further study and excavation. But I like the H. erectus story, for, as Jake says to Brett in the last line of The Sun Also Rises, “Isn’t it pretty to think so?”


h/t: Gravelinspector

J. C. A. Joordens et al. 2014. Homo erectus at Trinil on Java used shells for tool production and engraving. Nature, published online, doi:10.1038/nature13962

83 thoughts on “The world’s oldest graffiti: by Homo erectus! (maybe)

  1. I am also skeptical, but more important is that even if the shell is as old as they think and the scratches were made by Homo erectus, it seems to be a one-off. Any kind of evidence of any symbolic behavior is nonexistent for Homo erectus. One shell does not indicate they were capable of art in any way like moder humans are. It is an anecdote. Interesting, but this does not significantly change any views on the intelligence of H. erectus.

    1. Me too. It is nice to think that H. erectus made those etchings but if it is indeed a one off, it’s somewhat speculative unless I’m missing a big swath of evidence. And that it is out of context is a compelling argument to suggest that we just don’t know if H. erectus carved it up.

        1. Homo erectus had control of fire. Had done for the thick end of a half million years before this artefact is dated to.

    2. No, I think you are ignoring the implications of a McDonald’s logo being found at such an early date. It indicates not just the presence of art, but Happy Meal Toys.

      For the latter, note the shells with holes drilled in them. That and a strip of seaweed and a kid back then could happily entertain themselves for hours. They didn’t need all these fancy computers and expensive electronics. They knew how to play.

    3. The problem, of course, is that’s the only one that we’ve found. That does not mean that’s it’s a one off. It is very difficult to prove that this is a one off, but it might be easy to falsify that idea. I’ll bet there are folks scouring museum archives this week looking for more. Even if there are ‘t any more there, there could be more out in the field waiting to be found.

  2. Do I understand there is no controversy about the holes used for disabling the clam’s muscles being 1/2MY old? I’m impressed with the cleverness; I’ve opened scads of live clams and did not know that trick.

    Looking forward to seeing “How to Open a Clam Like a Caveman” on YouTube.

    1. I’m not surprised by this (my background is paleoanth). Chimps cracking nuts with rocks usually start out pretty clumsily, but it doesn’t take long for them to master the technique. Homo erectus and other early hominins were clever, no doubt since they invaded new environments, were often skilled hunters of dangerous big game and thrived for a very long time. The suggestion of symbolic art is what stands out here. That is a whole new ballgame for Homo erectus.

      1. Thank you – and I’m still wowed. I’m reminded of Darwin’s assertion that human intelligence vs that of other “higher” animals is a matter of degree, not of kind. Very exciting to hear of any observation that adds detail and color to that lovely idea.

      2. I’ve always thought that invention comes by accident or association or play and not by shear brain exertion. It just requires observation of cause and effect.
        For example, as a poor student (who minored in anthropology) I had to work in a plywood mill to earn my way. One night I invented the atlatl quite by accident (and thousands of years too late). While offbearing veneer I threw one piece into the cart and it split at the throwing edge and hinged on a knot increasing the velosity so much that it shattered as it hit the backboard.
        Voila, a way to throw harder with the same effort.
        I’m sure that the original version was invented quite by accident too. Maybe youth playing with sticks pretending to be hunters threw a stick that acted in much the same way as my veneer did.
        The thing is, you have to be smart enough to recognize a variation on a theme and take that discovery in a new direction, like stringed music coming from playing with a hunting bow when hearing the tone of the vibrating string.
        What’s interesting to me in this case is not so much the graphics as holding meaning or not, but the adaptation of the boring tool into a marking tool. Their minds were working and seem to have been able to use things in new ways and the act of scratching shows this. Whether it is or isn’t symbolic art is not as interesting to me as seeing that they had the minds capable of producing it, if not then, not too long after.

    2. The discussion on the Science site suggests that the wear on the edges of the groves and opening marks shows that the shell was rattling around in the sediment for a considerable time. Therefore the shell (and grooves) pre-dates the sediment. A combination of argon-argon dating and luminescence dating give the limits on age of the sediments : maximum age of 0.54 ± 0.10 million years and a minimum age of 0.43 ± 0.05 million years.

  3. The journal article avers that the scratches predate burial and weathering and cite “extended data.” I have no basis for remarking on that extended data, but if it holds up, the claim gains legitimacy. And it is nice to think so….

  4. it could have interbred with expanding Homo sapiens populations coming from Africa beginning about 600,000 years ago

    Homo sapiens did not exist 600,000 years ago. Did you mean “60,000”?

      1. That is a likely explanation.

        But we don’t know how long the Denisovan population were located in Asia. They split before 600 kyrs ago IIRC, so if Jerry on the grounds of proven introgression places them within H. sapiens

    1. Yes, this bothered me a lot, too. I cannot find how homo sapiens would have left Africa that long ago–60K seems right.

  5. Alas, the paper is not free, so I can’t really comment on it, but they say they used the luminescence dating method, which gives the last time a grain of quartz or feldspar was exposed to sunlight.

    So the hypothetical H. sapiens who “could have come along much later and performed the etching” must have done it in the dark.

    That would likely be my criticism of the criticism, but unless I see the methods section of the paper, I won’t sign it.

    PS: The full reference link doesn’t seem to work.

      1. Well thanks, but I had already found it. Problem is, it’s still behind a paywall.

        I was in a hurry earlier and now see I left something unclear: In the abstract they say that they performed the dating on sediment still contained in the shell.

        So there was a lump, big enough that its exterior shielded the interior from sunlight. So if really some Homo sapiens came along a few hundred thousand years later and said “Why, what a handy shell” – they managed to engrave the signs and use it as a tool, without cleaning (that is, completely cleaning) the shell’s interior. Is that likely?

        1. Does this mean that the shell was buried, still articulated at the hinge? Someone could have dug it up thousands of years later and carved on it. Now, it is possible that the original sediment in the shell was still kept inside, with the two valves sealing it tightly inside. I have an very old (but cenozoic) clam fossil w/ the two halves still together. The sand inside is hardened into a soft sandstone. If I carved on the shell, then dated the sand, I could get an extraordinary result!

          1. Only if it’s less than 200.000 years old. 🙂

            But good point, I didn’t think of that. Precisely that’s what I meant about having to look at their methodology, not just the abstract.

            I also have no idea on what part they made the second dating with Ar39/Ar40. My university will have the license, I hope.

            1. I’m talking nonsense, that is, talking Wikipedia, which gives the upper limit for that method at 200.000 years. But obviously, it’s more than that.

    1. Thanks for the info on the luminescence dating method. While reading the piece, I was wondering how/if they could date the actual engraving. Now I know…clever humans.

  6. So, “wanna come back to the cave and see my etchings” may have started out as a Paleolithic pick-up line? (I knew it was ancient, but not that ancient.)

  7. I want to share that I am now experiencing a sense of powerful coincidence! Powerful forces have aligned to make me look rather silly.
    Just this morning I had wrapped up my evolution lecture with the later stages of evolution of humans, describing the (then) known behaviors of H erectus up through H. sapiens, etc. I emphasized that the earliest signs of symbolic art are possibly from neanderthals, and definitely from H. sapiens — those living in Africa.
    Now there is this discovery! Well, maybe. Clearly the universe has aligned this research and publication and my history of visiting WEIT so that I will have my lecture up-ended right after I gave it.
    Thanks a lot, universe!

        1. Yeah I meant mussels. Though when I ate the oysters in the dream, I referred to them as, “clams” but they were definitely oysters.

          Silly subconscious not getting words right!

  8. A silly question, they suggest that the hole and carvings were made with a shark tooth. That seems very specific, are shark teeth readily available or commonly used by H. Erectus?

  9. “…washed into the site by a powerful flood.”

    We may have sighted a future quote mine in its natural habitat.

  10. This is intriguing to say the least. It supports the idea that the things that make us human, like symbolic thinking, did not arise suddenly, but evolved gradually over an undetermined period of time.

    As for the issue of the flood altering the date, I would really like to hear what some geologists have to say about that. It’s way outside my ability to make a judgement.

    1. Without the full paper, that context was my big concern too.
      Unfortunately, Dubois wasn’t as careful about his fieldwork or his records of his field work as might have been. And the danger of archaeological excavation is that once a site has been dug, you can’t go back and re-dig it. Which is part of the reason for doing digs with those walls of earth left in place between the excavated areas.

          1. Hell, no! they spoke the King’s English. King James, to be precise. How else could the BuyBull be written in it?

  11. I’m not a palaeontologist but…
    If i understand the criticism correctly, to me that seems to require a rather extraordinary sequence of events.

    The mussel is eaten and begraffitied at say, 200kY ago, then gets washed away by a flood and just so happens to get buried in 500kY old sediment, which also just so happens to mess up the dating, then exactly this mussel gets found again by modern scientists.

    I don’t know if that’s what the criticism is meant to say but if so I’ve got a razor here made by Ockham that is looking to cut out some superfluous fat from that theory.

    It seems much more parsimonious to assume the whole thing was used, scratched and buried at -500kY.

    1. If OTOH they mean that the mussel was eaten at -500kY, buried, then found again at like -200kY, engraved, then buried again (at the same 500kY old site?), and then found again, i find that even more extraordinary.

      Would that not make for the first evidence of paleontological activity in early humans 😉?

    2. I’m not seeing where the 200,000 years comes from. The dating FTFA is “a maximum age of 0.54 ± 0.10 million years and a minimum age of 0.43 ± 0.05 million years”
      My memory may be slipping, or I’ve missed something, but I thought that the next oldest evidence for “symbolic thought” (or graffiti) was Henshilwood et al’s
      Oh, how sad – just watching an Attenborough programme. Goslings doing a 400ft jump from next down to feeding grounds … just in time to become Arctic Fox food. Of dear. What a pity. Never. Mind.
      Sorry, where was I? Henshilwood, C. S., d’Errico, F. & Watts, I. Engraved ochres from the Middle Stone Age levels at Blombos Cave, South Africa. J. Hum. Evol. 57, 27–47 (2009)75,000 to 100,000 years old. So this discovery, if the dating is valid, is a considerable step further out. On the other hand, the existence of such artefacts from long, long before (say) cave painting is precisely what one would expect from it taking a long time to develops such skills.
      Bloody foxes are head-butting lemmings now. Distracting bloody animals.

      1. Oh, now I see the mention of 300,000 years younger. In a quote from the paper, I take it. Which implies that there has been another find since the Henshilwood discovery, and about as much older again.
        So, if you’re looking at dates for painting and decorating, you have :
        us, contemporary ;
        villages and agriculture, 10,000 years ;
        cave painting, 20-40,000 years (maybe 50,000);
        Henshilwood’s ochre, 75-100,000 years ;
        something I’ve missed, 200,000 years
        this artefact at 540,000 years.
        That is starting to look a sensibly filled-out progression.
        I note that the dating of the sediment revises the age given for the Trinil deposit, and thus “Java Man”, from around 700,000~1Myr to 540,000.

        1. Henshilwood’s ochre, 75-100,000 years ;
          something I’ve missed, 200,000 years
          this artefact at 540,000 years.

          FTFSI :

          12. The Trinil engraving in broader archaeological context
          In Extended Data Figure 10 (which I don’t have) we present the engraved shell DUB1006-f from Trinil in broader archeological context, against the backdrop of the marine isotope record70. With a minimum age of 0.43 ± 0.05 Ma, the engraved shell is considerably older than other geometric engravings documented thus far (Extended Data Fig. 10).
          An enigmatic series of grooves on a bone is known from the hominin site Bilzingsleben (ref. 71) in Germany, but the intentionality of that c. 350 ky old object is debated (72). The oldest (c. 100 ka) undisputed engravings were made on pieces of ochre and derive from Klasies River Cave 1 (73) (Extended Data Fig. 10c) and from Blombos Cave (74) (Extended Data Fig. 10d) in South Africa. Of about the same age is the
          engraving on a core from Qafzeh (75) in Israel (Extended Data Fig. 10e). The famous – and more
          elaborate- Blombos engraving (76) (Extended Data Fig. 10f) is dated to c. 75 ka. In addition,
          well-known are the at least 60 ky old engravings on ostrich egg shell from Diepkloof (77) in
          South Africa (Extended Data Fig. 10g), and the engraving on cortex from Quneitra (75) (Syria)
          attributed to Homo neanderthalensis (Extended Data Fig. 10b).

          I’d heard about the engraved ochre, and the engraved ostrich eggs. The German bones were news, and I think the one about the “emergence of music refers to a flute carved from a birds arm bone.
          71 Mania, D. & Mania, U. Deliberate engravings on bone artefacts of Homo erectus. Rock Art Research 5, 91-97 (1988).
          72 Davidson, I. Bilzingsleben and early marking. Rock Art Research 7, 52-56 (1990).
          73 d’Errico, F., García Moreno, R. & Rifkin, R. F. Technological, elemental and colorimetric analysis of an engraved ochre fragment from the Middle Stone Age levels of Klasies River Cave 1, South Africa. J. Archaeol. Sci. 39, 942-952 (2012).
          74 Henshilwood, C. S., d’Errico, F. & Watts, I. Engraved ochres from the Middle Stone Age levels at Blombos Cave, South Africa. J. Hum. Evol. 57, 27-47 (2009).
          75 d’Errico, F. et al. Archaeological evidence for the emergence of language, symbolism, and music – an alternative multidisciplinary perspective. J. World Prehist. 17, 1-70 (2003).
          76 Henshilwood, C. S. et al. Emergence of modern human behavior: Middle Stone Age engravings from South Africa. Science 295, 1278-1280 (2002).
          77 Texier, P. J. et al. A Howiesons Poort tradition of engraving ostrich eggshell containers dated to 60,000 years ago at Diepkloof Rock Shelter, South Africa. Proc. Natl. Acad. Sci. 107, 6180-6185 (2010).

      2. Oh yeah, the goslings plummeting to their doom. As me and my girlfriend said, “nature is f****d in the head”.
        The 200k years i thought i had remembered from the post as being the next oldest date for graffiti, but as it’s clumsy to scroll up and down while commenting that may be incorrect.
        It doesn’t matter in a way, it’s just a placeholder number for the argument. Works equally well with 75 to 100k years.

        1. I found a mention of some suspect marks on bone “350k younger” (which would be about 190k) from Germany, but those are of “suspected intentionality”.
          There’s a reason that the team go to some effort to describe the shape, depth, skewness and kurtosis of the groove at different points : it’s to establish that similar amounts of work were being put into displacing prismatic aragonite crystals from the shell at different points in the mark ; therefore the mark’s maker intended to make the mark as much at the end as they did at the start.
          There’s a more subtle sub-text there too: “we’ve learned more about the microstructure of pelycypod shells than most people really care to know ; wow, they’re complex little marvels of natural engineering, to get such toughness out of brittle materials”. Which is very true.
          A few days ago I was having to explain to Ben Goren the distinction between toughness and hardness. This is another side of that (literally) multifaceted coin. “Shell” is a complex material, much under-appreciated.

  12. I’m puzzled by the comment that similar holes as those drilled in these mussels were used by inhabitants of the Caribbean to open gastropods.

    As gastropods are not bivalves how did this work?

        1. Thanks for the link. Drilling bivalves in this way makes sense (although it seems quite laborious and I would imagine a sharp blow with a rock would enable a sharp edge – such as a fragment of the creatures own shell – to then cut the adductor muscle with much less effort if you don’t mind picking a few shell fragments out of the meat!).

          I was actually more puzzled about the drilling of gastropods in the Caribbean, though.

    1. I suspect that is from commentary, not the paper. The bits of the paper I can see (just the Supplementary Info) only mention Unionid (Unio-related) pelycypods collected by Dubois as having these marks.

    2. Ha, good point!

      Perhaps they were referring to some practice of drilling through the operculum in order to pry it out and get to the snail…

  13. Discovered by grad student Stephen Monro. Squinting at the scratches, I find myself wondering if his middle name starts with ‘H’. That would be a coincidence, maybe a message from a H. erectus ancestor.

  14. As a person who likes geometry and prefers squares/lines instead of circles/curves, I am surprised why this “caveman” etched straight lines instead of random curves/scribbles.

    Is fascination with straight lines a human trait – longing for something that does not occur naturally?

  15. I heard a Dutch professor discussing these shells on CBC radio. They also found small holes, apparently drilled. I have seen clam shell with holes about 2 mm diameter, and photos of Murex shells with such holes. I was told some predator made the clam shell holes (starfish? I forget) and the Murexes were cannibals. I wonder how you could tell these apart from homo-erectus drilled holes.

Leave a Reply