Photo Essay: The Culinary and Graphic Art of Homo Erectus at Trinil

of 06
500,000 Year Old Graphic Art

Engraved Fossil Pseudodon Shell, Homo Erectus Site at Trinil
Engraved Fossil Pseudodon Shell, Homo Erectus Site at Trinil. Wim Lustenhouwer, VU University Amsterdam

Re-analysis of the extensive freshwater clam shell collection recovered from the Trinil site, a Homo erectus site located on the island of Java in Indonesia, has rewritten what people understand about early modern behavior, setting the date of the first glimmerings of artistic expression back 300,000 years.

Trinil was discovered and excavated in 1891 by Dutch army surgeon and amateur paleontologist Eugène Dubois. Dubois recovered over 400,000 marine and terrestrial fossil vertebrates from the main bone layer (Hauptknochenschicht in German, abbreviated HK) at Trinil, and brought them back to his home university of Leiden in the Netherlands. Among those fossils, he discovered partial skeletons of at least three Homo erectus individuals, including a skull cap, two teeth and five femora. Although the site is currently underwater, Dubois' collection is still at Leiden University. That collection has been the focus of scholarly analysis during the 21st century.

This photo essay discusses the recent findings of the analysis of freshwater clam shells within the Trinil collection at Leiden published in Nature in December 2014: that Homo erectus consumed (presumably raw) shellfish, that they made and used shell tools, and, most surprisingly, that they carved or etched geometric grids on those clamshells, all about 500,000 years ago.

Analytical techniques used on the Trinil collections have included paleoenvironmental reconstruction and stable isotope analysis: but the most recent and astonishing evidence of modern human behaviors has been identified within the freshwater clamshell assemblage from the site. A team led by Josephine C.A. Joordens and Wil Roebroeks of Leiden University in the Netherlands has found evidence of consumption of freshwater clams, use of their shells as tools, and, if the team is right, the earliest evidence of geometric engravings--abstract art in its rawest sense--known on the planet.

of 06
Characteristics of the Faunal Collection

Buffaloes Getting Bathed in the Solo River Near Trinil (1864)
Buffaloes Getting Bathed in the Solo River Near Trinil (1864). Dr. W.G.N. (Wicher Gosen Nicolaas) van der Sleen (Fotograaf/photographer) - Tropenmuseum, Leiden

While Dubois collected all or nearly all the artifacts in the HK, and drew careful maps of the site deposits, the context of specific artifacts wasn't recorded. Further, scholars believe that the artifacts were likely overbank deposits, eroded out of their original location and dumped on the river bank during a series of floods. That makes interpretation somewhat difficult but not impossible.

The shell assemblage from Trinil includes examples from 11 different freshwater clamshell species, including a minumum of 166 individuals of extinct Pseudodon. The Pseudodon clams include 143 articulated pairs of valves (both sides, still connected to one another), 23 single valves and 24 fragments, representing a minimum number of 166 animals. The appearance of the shells, and their deposit apparently above the water line and with the bones of other animals, does not appear to have resulted from the inadvertent burial of a living population.

Instead, argue Joordens et al., they represent a shell midden--a purposeful dumping of used shells after the meat had been consumed--and the consumer must have been Homo erectus, based on the presence of holes drilled into the living shell by a tool such as a shark tooth. Thus, say researchers, the shell assemblage at Trinil could represent the remains of a purposeful shellfish collecting and processing event by H. erectus along the banks of the Solo River.

of 06
Evidence for Shellfish Consumption

Perforated Pseudodon Shell, Trinil Homo Erectus Site, Java
Inside of the fossil Pseudodon shell (DUB7923-bL) showing that the hole made by Homo erectus is exactly at the spot where the adductor muscle is attached to the shell. Credit: Henk Caspers, Naturalis, Leiden, The Netherlands

Evidence for Homo erectus having consumed the freshwater clam meat is the presence of holes perforating the shells. In about 1/3 of the total Pseudodon clams, holes had been pierced through the shell, most (73 of 92 holes) at the location outside of where the anterior adductor muscle attachment lies. Modern clam eaters know that muscle is what keeps the shell closed, and if you pierce the muscle in a living animal, the shell will open. The holes generally have a diameter of ~5-10 millimeters (or .1-.2 inches), larger than those drilled by carnivorous snails, more regularly shaped than those made by marine gastropods.

Shellfish dinners are enjoyed by many species, and other possible predators include otters, rats, monkeys, macaques, and birds. All of these predators have developed ways to get freshwater mussels open, but none use a pointed tool to pierce through the shell and cut the anterior muscle--only humans.

Shark Tooth Tools

Joordens et al. conducted experiments on living mussels, using a shark tooth--shark teeth were found in the Trinil faunal assemblages, but no stone tools. They first punctured a hole by striking the tooth with a hammerstone, but that resulted in breakage of the tooth and shell. But "drilling" a hole, by applying a shark tooth to the shell and rotating it (no haft required) produced a hole in the right place with shell damage similar to that seen in the fossil specimens. The main difference between the experimental tests and the fossil evidence is the lack of faint circular striations in the fossil examples. Joordens et al. suggest that may have been weathered away.

Examination of the shark teeth recovered from the Trinil site showed that 12 of 16 teeth recovered were damaged, but it was unclear how that damage came about.

of 06
Using Clam Shells as Tools

Shell Tool Made by Homo Erectus at Trinil
a. Shell tool made by Homo erectus by modifying the ventral margin of a Pseudodon shell (DUB5234-dL). b. Detail of the ventral margin that forms a sharp edge for cutting or scraping. Credit: Francesco d’Errico, Bordeaux University

A single shell valve, labeled DUB5234-dL, exhibits signs of modification by retouch--the careful pressure on the interior rim of the shell to reshape and thin the outer edge. The ventral margin features a string of contiguous flake scars exposing the nacreous (mother of pearl) interior layer which had been smoothed and polished. Shallow striations on the tool are present in lines running parallel to the retouched edge, and an elongated triangular pit and scoring mark is also seen.

As to what this tool might have been used for, Joordens et al. do not speculate, but at the nearby Homo erectus site of Sangiran (dated between 1.5 and 1.6 million years ago, but like Trinil the date is somewhat in debate), Choi and Driwantoro (2007) identified 18 cut marks on a bovid (extinct cow), which had been made by a sharpened clamshell.

of 06
500,000 Year Old Graphic Engravings

Detail of Engraved Fossil Pseudodon Shell from Trinil Homo Erectus Site
Detail of Engraved Fossil Pseudodon Shell from Trinil Homo Erectus Site. Wim Lustenhouwer, VU University Amsterdam

Finally, and most interestingly, the outer exterior of one clamshell from Trinil, DUB1006-fL, has been carved with a geometric pattern of grooves. Some of the lines are connected zigzags, created by turning the tool. The grooves are smooth and rounded, and experiments show that they could only have been made on fresh shell with a sharp and pointed object.

Joordens and colleagues conducted additional experiments to reproduce the grooves with a shark tooth, a pointed flint tool and a surgical steel scalpel (something Dubois might have had on hand). The experimental grooves made with a shark tooth matched best: with a shark tooth, there were no striations inside either fossil or experimental grooves, and the grooves had, like the fossil example, an asymmetrical cross-section.

Incident Light

The shell was photographed under incident light at different angles and directions, and lines that were unambiguously verified as having been engraved were traced and captured in the image on page six, produced by an Alicona 3D Infinite Focus imaging microscope.

The previous earliest geometric engravings known by the human species were on ochre and ostrich shell by early modern humans at several caves in South Africa such as Diepkloof and Blombos Caves, assigned to Howiesons Poort and Stillbay industries between 70,000-110,000 years ago.

of 06
Scholar Resources for Clamshell Use at Trinil

Infinite Focus of an Engraved Line in Psuedodon Shell
Infinite Focus image of a line engraved by Homo erectus in Pseudodon shell DUB1006-f. Scale bar is 1 mm. Joordens et al.

Choi K, and Driwantoro D. 2007. Shell tool use by early members of Homo erectus in Sangiran, central Java, Indonesia: cut mark evidence. Journal of Archaeological Science 34(1):48-58. doi: 10.1016/j.jas.2006.03.013

de Vos J, and Sondaar P. 1994. Dating Hominid Sites in Indonesia. Science 266(5191):1726-1727. doi: 10.1126/science.266.5191.1726-a

Indriati E, Swisher CC III, Lepre C, Quinn RL, Suriyanto RA, Hascaryo AT, Grün R, Feibel CS, Pobiner BL, Aubert M et al. 2011. The Age of the 20 Meter Solo River Terrace, Java, Indonesia and the Survival of Homo erectus in Asia. PLoS ONE 6(6):e21562. doi: 10.1371/journal.pone.0021562

Joordens JCA, Wesselingh FP, de Vos J, Vonhof HB, and Kroon D. 2009. Relevance of aquatic environments for hominins: a case study from Trinil (Java, Indonesia). Journal of Human Evolution 57(6):656-671. doi: 10.1016/j.jhevol.2009.06.003

Joordens JCA, d’Errico F, Wesselingh FP, Munro S, de Vos J, Wallinga J, Ankjærgaard C, Reimann T, Wijbrans JR, Kuiper KF et al. 2014. Homo erectus at Trinil on Java used shells for tool production and engraving. Nature in press. doi: 10.1038/nature13962

Szabó K, and Amesbury JR. 2011. Molluscs in a world of islands: The use of shellfish as a food resource in the tropical island Asia-Pacific region. Quaternary International 239(1–2):8-18. doi: 10.1016/j.quaint.2011.02.033