The barely adequate site

Dr Jack Fenner

Perfect is not a word that comes up very often in archaeology. Trying to understand people is never easy, and it doesn’t get any easier when they lived a long, long time ago and all you’ve got to go on is some of the more durable rubbish they left behind. So you won’t find archaeologists claiming perfect understanding, perfect analysis or perfect data sets; a search of the hundreds of scholarly archaeology articles on my computer didn’t turn up a single such claim of perfection.

There is, however, a related claim that turns up irritatingly often: the Ideal Site. The claim of having the Ideal Site to answer some specific archaeological question turns up regularly in otherwise restrained and scholarly, not to say boring, articles from academics who generally would not assert anything more definitive than that data indicates the sun tends to come up in the east.

Of course, they don’t really mean Ideal in the sense of the dictionary definition: “conceived or regarded as perfect or supremely excellent in its kind”. No, they mean Ideal in the archaeological sense: Barely Adequate. Archaeological sites and their artefact assemblages always have issues and uncertainties. Sometimes they are obvious problems of the ‘someone-dug-a-well-through-it’ sort and sometimes the issues are more subtle, such as a lack of the patterning needed to distinguish a hunting camp from a long-term residence. So when we have a site with characteristics suitable for answering an archaeological question and with issues and uncertainties that don’t seem overwhelming – a Barely Adequate site – there’s a tendency to get over-excited and call it Ideal. This is one of my pet peeves, and I always cross it out when doing peer review. Professional archaeologists should know better.

Its most obviously great feature was a series of stone lines that had once been ovens in which the Macassans used to boil trepang before preserving it in smokehouses and shipping it back to Sulawesi or elsewhere.

So imagine my horror when a search through the collection of articles on my computer for an Ideal claim to make fun of turned up one of my own articles. I couldn’t believe it, but there it was: “Therefore it is the ideal site at which we can test accepted chronological models for South-East Asian – Australian cultural contact”. D’oh! It’s an article I co‑authored about work a group of us did at a site called Anuru Bay A in Northern Australia that showed evidence for the presence of Macassans (people from the island of Sulawesi in modern Indonesia, or thereabouts).

Anuru Bay A is a nice site with many fine qualities, but it isn’t Ideal. We knew that – in fact, the article itself lists many issues and uncertainties. For instance, to find out when a site was occupied, archaeologists can sometimes use artefacts like coins with dates printed on them or pottery designs that were only made for a short, known time period. Unfortunately no such artefacts were found at Anuru Bay, so the site occupation date must be determined using ‘carbon-14 dating’. This method relies on the natural decay of carbon atoms in once-living things to determine when they died. This is a highly reliable technique, but is not as precise as the date found on a coin.

The article also notes other problems. It rains a lot during the wet season, which can move things around above and below the surface. Besides a previous archaeological excavation, pigs and other animals probably dug up parts of the site. In recent years it has been used for recreational boat launches, leading to modern rubbish as well as soil disturbance.

The list goes on. Other issues didn’t come up in the article but are common in archaeological work. For instance, Anuru Bay A is in Arnhem Land, which is a long way from our home institution in Canberra. Distance made it expensive to excavate and limited how much work we could do at the site. Plus it has a lot, and I mean a lot, of flies and spiders.

Given all that, you may be wondering why we would be at all excited to excavate Anuru Bay A, much less wind up calling it ‘Ideal’. We were excited because of the site’s good qualities. It was relatively undisturbed (nobody had built a house or a Bunnings on top of it). One of our team members knew many people in the local Aboriginal community, and they were happy to allow our excavations and even to host our campsite. Above all, Anuru Bay A was known to have extensive artefact evidence of Macassan presence. Earlier archaeological work suggested it may be one of the first places they visited.

Its most obviously great feature was a series of stone lines that had once been ovens in which the Macassans used to boil trepang before preserving it in smokehouses and shipping it back to Sulawesi or elsewhere. A trepang is a marine animal sometimes called a sea cucumber. It has a tough skin but cooks up into a slippery goo considered a delicacy by people with the money and inclination to try it. Trepang are, or at least were, abundant in Northern Australian bays and inlets.

That was why the Macassans came, and continued to come, to Australia for centuries. The question at hand was how many centuries. Some archaeologists and historians argue for a relatively short duration with the trepanging visits starting in the late eighteenth century. Others suggest a long contact model with the visits extending back a further 200 or more years. We hoped our excavations and subsequent analyses would help resolve which of these two ideas is right.

In the end, we were able to work around or minimise the impact of many of Anuru Bay A’s issues. We analysed extra carbon-14 samples and did some fancy statistics on the results to minimise the occupation date uncertainties. We avoided disturbance impacts as much as possible through careful sample selection, such as charcoal that was wedged between rocks within a trepang oven stone line and thus unlikely to have moved over time. We argued that other issues do not substantially affect our conclusions. For instance, we were able to recognise a buried shell midden of likely Aboriginal origin and exclude it from our assessment. In the end, we argued that the longer contact period is correct: Macassan trepanging visits to Anuru Bay almost certainly began by the 1630s, and they may have started in the 1500s.

Thus, Anuru Bay A was not by any means Ideal, and we were silly to suggest it was. But, in our view, the site was adequate to answer the question we were asking. Not everyone agrees with us, and that’s okay. A healthy scepticism of other peoples’ (and your own) analyses is part and parcel of archaeology and of scholarly work in general. The social sciences and humanities are particularly susceptible to erroneous interpretation of data because we are rarely able to control the circumstances of our work. Researchers have nevertheless been able to piece together convincing descriptions of the human past and of the overall human condition. We’re not living in an ideal world and there are no ideal sites, but a bunch of barely adequate sites combined with careful analysis and a big dollop of humility has taken us far. Sometimes barely adequate is pretty darn good.