Tag Archives: Wallis Heritage Consulting

What you see is (not always) what you get? A final reflection on the analysis of the stone artefacts from Gledswood Shelter 1

Who would have thought a box of artefacts from a remote shelter in northwest Queensland could contain such a complex story of the past 10,000 years?  Well at least it has proved to be a much more complex story than I could have imagined and I have only managed to scratch the surface as part of my directed study. The project has proven to be more demanding in time and effort than any other subject I have undertaken and still there is so much that seems incomplete and in need of further research. There is more than a thesis waiting in just one square of excavation from Gledswood Shelter 1 (GS1).

Part of my study involved age-depth modelling. This is the process of using the radiocarbon dates obtained from the spits throughout the excavation to understand the history of sedimentation at the site. Age-depth modelling is a science in its own right and what I learned from my study was that GS1 warrants a thorough modelling of its history of sedimentation using some of the modern techniques available, such as linear regression, splines and interpolation. This work alone would be enough for a directed study project.

I also learned that what you think you see is probably not what you can see. My tendency was to see trends in the spits in terms of artefacts numbers and to believe that these trends were real. However, once these numbers were correlated with time in the age depth model a very different picture emerged. What appeared to be a peak in artefact numbers was not, and what did not appear to be a peak in artefact numbers was.


Evidence of the use of stone axes such as the axe pictured is seen as small basalt fragments throughout the excavation at GS1

Once the trends were understood it was time to make sense of this through researching the available literature. A trend towards increased activity at the site seems to correlate with a wider trend that occurred across northern Australia during the mid to late Holocene, where populations moved into more marginal areas exploiting food sources not previously used, such as the toxic seeds of cycads. These changes were believed to be responses to rapid climatic changes that required innovations in the way people lived off and used the land, and the technologies available to them. There are many more questions to be answered in relation to these responses to change. For example, there are reports that some sites show evidence of responses to change in the mid Holocene, whilst others show evidence later in the Holocene, sometimes a couple of thousands of years apart. The site specific nature of these responses is a complex question and GS1 still has many questions to be explored.

Perhaps this project taught me more about research than it did about GS1. Sometimes, when I thought I saw clear evidence of a pattern, there was a tendency to search for evidence that would support it. However, this had the effect of excluding information that might challenge my hypotheses. When I became aware of this behaviour I could adjust my approach to research and sure enough the result would often be quite different to what I believed I was observing.

This directed study has been a great journey and I have learned more from this topic than any other. Thanks to Lynley Wallis my Industry Partner who has assisted me throughout my project.

An analysis of the stone artefacts from Gledswood Shelter 1 – update 3

After sorting through, describing and categorising 936 artefacts I can definitely say that my skills in identifying and describing artefacts have improved. With more than 800 of the artefacts comprising flakes and broken flakes, each one had to be studied to identify the features that would enable me to determine how it had been made. Most of these flakes were made of quartz and, owing to the intrinsic properties of that raw material, determining the reduction process proved challenging at times.

After identifying the types, artefacts were then categorised by raw material type, counted and weighed.  The total numbers and weight of artefacts from each spit were volumetrically adjusted to account for differences in the amount of sediment removed from each spit. The results are shown in the graphs below. The graphs show total weight of artefacts in grams per kilogram and total count of artefacts per kilogram. Viewing both graphs shows clearly that, when a large artefact substantially affects the weight—such as in Spit 16 which includes a pestle—the artefact count is not affected and we are not misled by the results.

Artefact Numbers
Artefact Weight

The next phase of the project is to interpret the graphs and other recorded data, which will require consideration of several factors. Firstly, sediment deposition in the shelter is not consistent through time, so the spits do not represent equal periods of time. For example, Spits 20 to 16 represent almost 5000 years of time, meaning the bottom 20% of the excavation represents 50% of the time line. What appears in the graphs as a peaks in artefact counts and weights for the lowest two spits may actually be non-existent when the timeframe of deposition is considered relative to the other spits.

Even when volumetrically adjusted, artefact weights for each spit also need to be carefully considered, as it cannot be assumed that a high weight represents a greater level of activity in the shelter. For example, Spit 16 shows the greatest weight of artefacts for all spits; however, one of the artefacts it contains—a pestle—weighs more than 400g. In contrast, Spit 6 includes 130 flakes that only weigh 174g. The pestle may have only been used once, whilst the flakes may have been used multiple times. So in this case a lower weight could represent greater activity at the site on the basis of the artefact types found.

DSC_0011 A stone used for grinding found in Spit 16

Patterns that need to be explained also include changes in time through the raw materials present. In the lower spits (i.e Spits 20 to16) flakes are equally likely to be made of both quartz and quartzite. After this there is a gradual change, with an increase in quartz flakes and a reduction in the number of quartzite flakes. This change reaches a peak in Spit 6, with a ratio of 97:3 of quartz to quartzite flakes. Trying to determine what might have caused people to change their preferred raw material is one of the things I am exploring.

Selection of Quartz Flakes A selection of quartz flakes from Spit 6

Further afield there are different changes seen in the assemblages of other shelters in northwest Queensland that are not seen in Square B0 at GS1. For example, some shelters have shown changes in raw materials used at a specific time, and increases in the use and presence of grinding stones. These changes have not been seen in Square B0.

Once I’ve explored the patterns seen in  Square B0 I’ll then consider this in relation to the results from the excavation of the adjacent squares in GS1, other shelters on the Middle Park Station, and other shelters in the northwest Queensland region. This will help to understand whether the patterns observed are specific to the Square B0 assemblage, the GS1 shelter, the Middle Park Station area or whether they are part of a wider sequence of changes occurring in Queensland and beyond at the time.