Australia: The Land Where Time Began

A biography of the Australian continent 

Fern Cave

This is a large cave with 2 high-domed chambers, with an occupation site going back at least 26,000 years. There are a few heavily painted peckings on the wall adjacent to the excavation, that includes a series of loosely clustered pits, a star shape, and 3- to 4- pronged motifs, resembling 'tridents' or bird tracks. These have been demonstrated to be similar to other peckings from Chillagoe, Mitchell-Palmer Region, Laura Region and Koolburra region, which are believed to be very old, based on the degree of patination and the nature of the superimpositions.

The lowest layer in this deposit has been dated to 26,010 +/- 410 BP, but it is believed the deposit actually extends back to about 30,000 BP according to extrapolation based on the age-depth curve (David, 1991).

According to Laura Lamb her paper presents the results of a technological analysis of the assemblages from Test Pit 4 in Fern Cave in the southeast of Cape York Peninsula. Investigation of David’s 1991 claim that the stone artefact deposition rate at Fern Cave increased during the peak of the Last Glacial Maximum (LGM), about 20,890-17,200 BP was her specific aim in this study. Other claims made by David (1991), that deposition rates of other cultural components, such as mussel shell, remains of fauna, burnt earth, or ochre, were not investigated further in this study. More intense use of Fern Cave during the LGM was suggested by David (1991) to be the reason for the peak period of deposition and sedimentation.

Questioning of the use of artefact densities for the inference of site use intensity (Hiscock, 1981; Ross, 1985), their argument being that increased deposition rates may be the result of factors that are not related to the intensity of occupation of the site, such as changes in the processes of manufacture. They argue that because of this an attempt should be made to determine the nature of the systems of lithic production by lithic analyses. This analysis has been designed by Lamb to test a number of issues associated with the increased production of flaked stone at Fern Cave. See Lamb, 1993.

Fern Cave is in Spring Tower, part of the limestone karst region of Chillagoe, north Queensland. There is permanent water, springs, within 2 km and 4 km of the cave (observations by Lamb; Robinson, 1982). Lamb suggests these springs may be important for understanding the occupation patterns of Fern Cave, as well as the Chillagoe region as a whole. The landscape where these water sources are located is at least 80 million years old (Willmot & Tresize, 1989). Lamb suggests this landscape appears to have been subjected to slow change during the human occupation of the continent, and the assumption that springs were in existence during the LGM seems reasonable to her (B. Bultitude, Queensland Department of Minerals and Energy, pers. Com. To Lamb). The age of the springs was investigated by David et al.

David excavated Fern Cave over 2 field seasons, 1985 and 1989. The 1985 excavation was part of his Masters project (David, 1987). The aim of his research was to obtain a general sequence of occupation to ascertain temporal changes in 2 cultural phenomena, stone artefact characteristics and strategies of faunal exploitation in the Chillagoe region, the study becoming part of a broader regional archaeological project (David, 1987).

Fern Cave was included in David’s doctoral research, because it contained engravings that were believed to date from the Late Pleistocene/Early Holocene period (B. David, pers. Comm. to Lamb). David obtained a radiocarbon date of 25,710 ± 400 BP on shells of land snails that were recovered from the 1985 excavation (Lab.no. Beta 30403). Later in 1985 the cave was re-excavated to find more evidence of the age of the occupation at the site. In 1989 David excavated 4 50x50 pits, one of which TP4) was analysed in this paper. According to Lamb there are 2 reasons for this analysis being restricted to TP4, there is a heavy encrustation of CaCO₃ on most of the material from the other excavated squares, and there were dates for material from other squares that were not encrusted in CaCO₃.

Stone Working Techniques

Strategies for the reduction of cores recovered from Fern Cave underwent changes over time, though the apparent rates of increases of deposition of stone artefacts cannot be said to result from these changes during the LGM. There are notable decreases in the bipolar technique use that is evident at this time and the evidence of core rotation also declines. It is implied by a decreased use of these reduction strategies at this time that people were actually obtaining fewer flakes per core in Phase 2 (about 20,890-17,200) than in Phase 3 (about 25,710-20,890 BP), rather than more that would be necessary for them to contribute to the apparent increases in rates of deposition of stone artefacts in Phase 2. This trend was coincident with a shift in the procurement patterns of raw material, with the selection of chert increasing and quartzite decreasing from about 25,710-20,890 BP to about 20,890-17,200. Chert is available locally, being present within 3 km of the cave, while the nearest quartzite source is about 11 km away}.

Aspects of Site Use

There is an increase in the proportions of artefacts that are heat treated from Phase 3 to Phase 2, and this trend is associated with an increase of flakes that were transversely snapped and flakes that exhibit edge damage (snap fractures). According to Lamb the increase in flake production as well as the incidence of edge damage and snapping, as the effective tensile strength of homogeneous stone. When a flake snaps, transversely of otherwise, as a result of the flake being subjected to post-depositional forces, the archaeological record effectively contains 2 or more artefacts.

Core decortication, the initial stages of core reduction appear to have been carried out at Fern Cave during Phase 2. Previously, decortication was performed elsewhere. The steep increase in the number of flakes with cortex intact at that time, in spite of no significant change in the average or range of artefact sizes, is the strongest evidence for this view. It is implied by this that reduction behaviour changes and Fern Cave’s position in the flaking systems’ organisation may have contributed to the increase in the rates of stone artefact deposition at this site, especially when coupled with the possible effects of burning. Deposition rates still peak in Phase 2 even if heat-treated artefacts are eliminated from the sample. Increased rates of deposition of stone artefacts during Phase 2 are likely to be caused by other factors.

Discussion

Significance Testing

When small sample sizes are used for tests of significance there is a tendency to obtain poor levels of statistical significance. Many of the results obtained using Fisher’s Exact Test were not statistically significant. Statistically significant patterning is shown by only patterns of raw material (for statistical calculations see Lamb, 1993). Lamb draws some conclusions about site use trends that rely on ‘substantive’ trends, instead of statistically significant trends (D. Chant, Department of Education, University of Queensland, pers. Comm. to Lamb).

When an attribute or characteristic shows numerical variation between analytical units substantive trends exist. Until it is proven to be statistically significant any trend is substantive by virtue of quantitative variation. Lamb argues that it is valid to use substantive trends only when it is possible to demonstrate that a series of such trends has an association with each other and with a set of statistically significant trends. The result of this is that the interpretations and conclusions concerning aspects of site use are of a preliminary nature only.

According to Lamb substantive trends were used in this study with the proviso that in regards to Fern Cave more work needs to be carried out on the lithic assemblage.

Results – Discussion

During the peak of the LGM, about 20,890-17,200 BP, to local lithic resources from those that were relatively distant implies a size reduction of the catchment of raw materials, the size of the territory over which the inhabitants of the cave collected stone, and/or there was a reduction in the frequency of visits to the more remote parts of their country.

It is implied by the accompanying change is strategies of reduction that there was no shortage of resources or needing to be conserved as it appears the cores were reduced less in Phase 2. In Phase 2 chert was the predominant material being used, and there were large, exposed ridges from which the material could be collected relatively close to Fern Cave. Therefore by considering the increased use of chert and the accompanying change in reduction strategies, Lamb argues that the cave inhabitants increasingly used the abundant chert resources that are present near the cave at the height of the LGM, also suggesting that it may have been as a result of water being less easily obtained at that time.

A very small proportion of the lithic assemblage of each XU is made up by complete flakes in TP4. The XUs between radiocarbon dates were combined into ‘dated phases’, which were the basic analytical units used in this analysis, for the purpose of the analysis. This was done to increase the sample size of complete flakes, thereby making the observable trends suitable for significance testing. The trends that are finer-grained between XUs will not be observed, which is the limitation of this method (David, 1991).

Changes in the location at which the initial stages of core reduction took place is related to a second factor that contributed to the increase in stone artefact deposition rates. It is suggested by the evidence that the initial stage of raw material for stone artefact took place somewhere other than the site during Phase 3, though in Phase 2 the initial stage of reduction was being carried out in Fern Cave. Assuming that initial processing of stone cores took place near the raw material source, the pattern is consistent with the quartzite sources being more distant from Fern Cave than those of chert. This kind of use of the site would effectively result in more artefacts being produced in the site within Phase 2 than in Phase 3, if it is assumed that subsequent stages of core reduction were consistently performed within the site in both temporal phases, though this is not sufficient to account for the general trends that were observed (David, 1991).

He Chillagoe region is suggested by the data to have been relatively dry over the LGM apart from that were 2 and 4 km from Fern Cave where there were springs. Lamb suggests there was likely to have been less seasonal water flow in the region during Phase 2, as a result of reduced precipitation, and that settlement/subsistence patterns in the region may have been affected by this (Also David, 1994; Hiscock 1988: 67; Lamb, 1993). Lamb suggests, for this reason, that the inhabitants of Fern Cave collected their stone raw materials from sources close to the cave, such as the sources of chert at this time, and therefore the processing of the chert raw materials, the initial reduction of the cores, was carried out in the cave.

If the springs near Fern Cave continued producing water over the LGM, 20,890-17,200 BP, increased intensities of human activity in the vicinity of the waterhole, including Fern Cave, could be expected of have occurred. Lamb suggests such concentrated activity would manifest itself archaeologically as increased rates of deposition of cultural materials within key sites, and at more marginal sites, as decreases or abandonment.

Sources & Further reading

1. Josephine Flood, Archaeology of the Dreamtime, J. B. Publishing
2. Phillip J. Habgood & Natilie R. Franklin, The revolution that didn't arrive: A review of Pleistocene Sahul, Journal of Human Evolution, 55, 2008

3.      Lamb, Lara. "Investigating Changing Stone Technologies, Site Use and Occupational Intensities at Fern Cave, North Queensland." Australian Archaeology, no. 42 (1996): 1-7.