Mata Menge stone tools

Adam Brumm and colleagues (2006) describe the stone artifacts from the Mata Menge archaeological site on Flores. This site is one of several described by Morwood et al. (1998, 1999) dating to the Lower-Middle Pleistocene boundary. This paper places the date for the artifacts between 880,000 and 800,000 years ago.

Mata Menge and other contemporary sites (there are three with stone artifacts dating to before 700,000 years ago) do not preserve any hominid bones, and there is no evidence that H. floresiensis was there making these tools. But the paper is being interpreted in the context of H. floresiensis -- because one scenario has these hominids evolving in situ from earlier Homo erectus, which presumably made these early tools.

Reuters article

"Small-brained or not, Homo floresiensis was capable of making stone tools, and therefore the standard story of the relationship between brain size and behavioral complexity in human evolution may be less straightforward than currently assumed," said the team's leader, Adam Brumm of the Australian National University in Canberra.
Until now it was thought that the larger the brain, the smarter the hominid. Brumm said his findings suggest that may not be the case.
"The causal relationship between brain size and the complexity of tool behavior in humans is assumed, not demonstrated," Brumm said. "Until now stone tools have only been found in association with large and relatively large-brained hominids, but Homo floresiensis changes that, forcing us to rethink the way we associate big brain with sophisticated behavior."

A Nature News article by Michael Hopkin also reviews the find and the recent Flores flap.

The paper itself is relatively short and most consists of description of the 500-some-odd artifact assemblage from Mata Menge. The comparison with the tools from Liang Bua take up relatively little space. That's too bad, because a long description of the Liang Bua artifacts would be very welcome. But that apparently must wait. In the meantime, here is the concluding paragraph from the paper:

The stone artefact assemblages from Mata Menge and the Pleistocene levels of Liang Bua are remarkably similar. We still do not know the species identity of the Mata Menge knappers, as no associated hominin remains have been recovered so far, but the age of the site clearly precludes modern humans. At Liang Bua, however, the skeletal remains of at least nine individuals are represented in finds from the Pleistocene levels, and all diagnostic elements are of H. floresiensis. The most parsimonious explanation for this is that the stone artefacts from Mata Menge and Liang Bua represent a continuous technology made by the same hominin lineage. Pronouncements that H. floresiensis lacked the brain size necessary to make stone artefacts are therefore based on preconceptions rather than actual evidence.

I think that logic bears repeating:

  1. Modern humans could not have made the ca. 800,000-year-old Mata Menge tools.
  2. The Mata Menge tools look similar to the Liang Bua tools.
  3. H. floresiensis is at Liang Bua.
  4. Therefore, H. floresiensis or its ancestors must have made both the Mata Menge tools and the Liang Bua tools.

Item 1 is certainly true. Item 3 is circular -- since it is the existence of H. floresiensis that the tools are supposed to demonstrate, or at least support.

In any event, the mere presence of H. floresiensis (assuming it is real) cannot demonstrate that the species made the tools. We can compare Liang Bua to Swartkrans, which preserved many specimens of Australopithecus robustus and only a small number attributable to Homo. Despite the lopsided proportions of the fossils, there has been no resolution to the question of who made the Swartkrans tools -- far from it, actually. Most of us assume that most of the tools were made by Homo, although it is not possible to exclude the possibility that A. robustus made some of them. But really, this is only a preconception about the abilities of Homo and A. robustus. And, hey, it is only a preconception that prevents us from saying that the tools were made by baboons, which were very abundant in the cave as well.

So the presence of H. floresiensis is relevant only if modern humans were not possible makers of the Liang Bua tools. A bit more on that below.

The key question beneath the conclusion of the paper is whether item 2 is accurate -- are the Mata Menge tools really similar to the Liang Bua tools? This is where the paper seems weakest to me. It is really a stretch to claim that these assemblages were linked by any kind of tradition.

Of course, it would be a stretch in any one place to claim a single tradition spanning 700,000 years or more -- that is a transfer of information across some 35,000 generations. Sure, the Acheulean was maintained for this amount of time or more, but it's not obvious that the Acheulean comprises a single tradition, or that there was much information transfer at all. If the Liang Bua and Mata Menge tools were no more similar than two Acheulean sites of the same age separation, then I would say there was no evidence of a tradition linking them. So even if the tools looked fairly similar, there still might not be a compelling case for descent or isolation based on the artifactual similarities.

But all that presupposes that the tools are similar. Here, I just don't see it. There are two pieces of evidence that supposedly show similarity between the Mata Menge and Liang Bua tools. The first is a similar type of "perforator" -- in other words, a pointy tool that is flaked bifacially to accentuate the point. If you get Nature, you can look at the figure comparing these "perforators" from Liang Bua with the earlier "perforators" from Mata Menge. Go ahead, look. Other than the fact that they both are pointy, I don't see the similarity. The Liang Bua examples are extensively shaped by bifacial flaking; only one of the Mata Menge examples even looks like there was any attempt to shape the "point" unifacially.

Now, consider that these "similar" tools were taken from a sample of over 500 for Mata Menge and a sample of over 3000 from Liang Bua.

The other proposed similarity is this:

For instance, both assemblages show an emphasis on the use of volcanic/metavolcanic fluvial cobbles as raw materials, along with the transportation of flake blanks for use as cores. Core reduction strategies at Mata Menge and Liang Bua are also very similar, with special emphasis on freehand reduction of cores both bifacially and radially. In fact, small, invasively reduced radial cores from the two sites are virtually indistinguishable. In addition, single platform cores, multiplatform cores, cores with 'burination' scars from the production of elongated flakes, 'truncated' flakes and cores indicating anvil-supported percussion and 'perforators' occur in both assemblages (Fig. 4). The maximum dimensions of flake scars on Mata Menge and Liang Bua cores are also very similar (Fig. 5; see also Supplementary Information). This is notable given that Liang Bua cores were more often on flakes, whereas Mata Menge cores were predominantly cobbles and hence tend to be bigger.

"Occur in both assemblages" could link any stone tool assemblages in the world.

Morwood et al. (2004) picture some of the more "advanced" tools from Liang Bua, including blades and microblades (which they suggest were hafted). If these occurred in both assemblages, they might be on to something. But really there are no special similarities between Mata Menge and Liang Bua.

What about those modern humans? They were in Australia by 50,000 years ago. They have to have been passing by Flores by that time, or earlier. The early Liang Bua tools are supposed to be as old as 95,000 years. Maybe modern humans weren't there then, but it's not yet clear from the dates so far presented that any of the artifacts are that old, either. Dates in excess of 70,000 years for modern humans would be credible, and might well explain the site. Whatever is the case, it seems very unlikely that the earliest date for modern humans on Flores could be as recent as this:

In contrast, the first skeletal evidence currently available for modern humans on the island, at Liang Bua around 10.5 kyr bp, is associated with various changes and additions to the stone artefact record, including an increased emphasis on the use of chert and the appearance of new stone artefact types (for example, edge-glossed flakes, grinding stones), as well as the first evidence for symbolic behaviour, such as personal ornaments (for example, beads), pigments and formal disposal of the dead.

So an increased use of chert and deaccentuation of the local volcanic stone is characteristic of the non-floresiensis tools, according to this paper.

But the “perforators” from Liang Bua pictured in the paper are on chert!

Some observations: A cultural transition within modern humans in the region around 10,000 years ago would not be surprising. Cultural transitions also occurred in Australia shortly after that time, possibly related to an influx of new people from Indonesia. That this transition might be associated with greater symbolic behavior also is sensible within the Australasian context, where a Holocene transition to greater symbolic behavior occurred.


Brumm A, Aziz F, van den Burgh GD, Morwood MJ, Moore MW, Kurniawan I, Hobbs DR, Fullagar R. 2006. Early stone technology on Flores and its implications for Homo floresiensis. Nature 441:624-628. DOI link

Hopkin M. 2006. Old tools shed light on hobbit origins. Nature 441:559. DOI link

Morwood MJ, O'Sullivan PB, Aziz F, Raza A. 1998. Fission-track ages of stone tools and fossils on the east Indonesian island of Flores. Nature 392:173-176.

Morwood MJ, Aziz F, Nasruddin, Hobbs DR, O'Sullivan PB, Raza A. 1999. Archaeological and palaeontological research in central Flores, east Indonesia: results of fieldwork, 1997-1998. Antiquity 73:273-286.

Morwood MJ, Soejono RP, Roberts RG, Sutikna T, Turney CSM, Westaway KE, Rink WJ, Zhou J-X, van den Burgh GD, Due RA, Hobbs DR, Moore MW, Bird MI, Fifield LK. 2004. Archaeology and age of a new hominin from Flores in eastern Indonesia. Nature 431:1087-1091.