The Archaeological Potential of Secondary Contexts

Project Conclusions

1. Introduction

The focus of this project has been the UK’s archaeological secondary context resource, with specific reference to the Lower and Middle Palaeolithic periods (c. 500–40,000 years BP). The archaeological value and potential of this resource has been demonstrated through a series of investigations exploring:

a. The spatio-temporal structure of the resource, emphasising the chronology of the secondary contexts (fluvial sedimentary deposits) and the derivation of the archaeology (stone tool assemblages).

b. The range and types of archaeological data which can be extracted from secondary context assemblages.

c. The relationships between archaeological secondary contexts and extant analytical and interpretive frameworks.

d. The management of the secondary context resource.

2. Summary

The key points raised by this project are summarised below:

2.1 The temporal structure of fluvial sedimentary deposits (Modules 1 & 2)

2.1.1 Fluvial activity during the Middle and Late Pleistocene (787–11,000 years BP) is associated with periods of climatic change.

2.1.2 These periods of climatic change operate both at high magnitudes and low frequencies (the 100,000 year glacial/interglacial (Milankovitch) cycles) and at low magnitudes and higher frequencies (sub-Milankovitch climatic variations, occurring over centennial or millennial timescales).

2.1.3 Fluvial studies from the last glacial/interglacial cycle (127–11,000 years BP) and the Holocene (11,000–present) indicate that:
a. phases of fluvial activity occurring in response to climatic change are relatively rapid, typically lasting hundreds rather than thousands of years.
b. all rivers typically respond to the 100,000 year Milankovitch cycle climatic events, while fluvial response to the sub-Milankovitch climatic variations is less universal.

2.1.4. Fluvial activity is not continuous during these periods (e.g. sedimentation is likely to have occurred only in response to spring floods during an annual cycle), but current geochronological tools do not permit such high resolution dating.

2.1.5 There is only partial preservation of fluvial sedimentary features from the Middle and Late Pleistocene periods. Typically, the larger-scale features associated with the Milankovitch cycles are favourably preserved, while the small-scale features associated with the sub-Milankovitch events are vulnerable to subsequent erosion.

2.1.6 Sedimentary units can be dated to individual marine isotope stages and it is predicted that they represent short periods of time (probably hundreds of years). However, current geochronological tools do not permit the dating of sedimentary units to specific episodes of climatic change.

2.2 The temporal structure of archaeological assemblages occurring within fluvial sedimentary deposits (Modules 5 & 8)

2.2.1 The artefacts are derived and have therefore been removed from their original place of discard by processes of soil erosion, solifluction and flooding. The artefacts were incorporated into fluvial sedimentary deposits by stream flow. These processes have two implications:
a. the artefacts are therefore older than the formation of the sedimentary deposit. The sediments therefore only provide a terminus ante quem. Assessing the relative magnitude of the age discrepancy between the archaeology and the fluvial sediments was therefore a major focus of this research (sections 2.2.2–2.2.5 below).
b. the findspot locations associated with secondary context assemblages cannot be regarded as ‘sites’ in the same way as in situ contexts from the Palaeolithic (e.g. Boxgrove) and later archaeological periods.

2.2.2 The frequency of sub-Milankovitch climatic variations (interstadial events occurred every 3–4,000 years over the last 80,000 years) indicates that artefacts discarded upon floodplain surfaces would have been regularly exposed to significant fluvial activity. While preservation of the fluvial sediments associated with high frequency climatic events is variable (see above), their widespread occurrence within European Late Glacial and Holocene river systems indicates that artefacts are unlikely to have lain unmodified upon floodplains for tens and hundreds of thousands of years.

2.2.3 Once artefacts have been incorporated within fluvial sedimentary deposits they are potentially vulnerable to subsequent erosion of those deposits. This can result in the artefacts being re-worked into younger deposits. The age discrepancy between the younger and older deposits can range between hundreds of years (reflecting localised erosion in response to sub-Milankovitch climatic events) and hundreds of thousands of years (reflecting river downcutting in response to Milankovitch cycles).

2.2.4 A model is therefore presented for assessing the degree of temporal re-working undergone by artefact assemblages. The model emphasises:
a. the river morphology and its implications for the preservation/erosion of sedimentary (river terrace) deposits.
b. the local and regional geological bedrock controls and their implications for preservation/erosion of the deposits.
c. the stratigraphic position of the artefact assemblage, its relationship to fluvial activity across a Milankovitch-scale glacial/interglacial cycle, and its probable chronology of re-working.
d. the physical condition of the artefacts (Section 2.3).

2.2.5 The model provides a relative measure of the degree of re-working undergone, and the importance of sedimentary evidence from the field is emphasised for the evaluation of the model.

2.3 The spatial structure of archaeological artefact assemblages (Modules 4, 5 & 8)

2.3.1 The physical condition of all the stone tool assemblages examined from archaeological secondary contexts indicates that they have undergone degrees of fluvial transport.

2.3.2 Experimental flume research has documented distinctive patterns of artefact damage, sustained during fluvial transport. Damage was assessed in terms of the état physique of the artefacts, incorporating arête ridge widths, edge micro-flaking and gross factors of artefact morphology. The distinctive patterns indicate different modes of bed-load transport and enable the differentiation of damage sustained during phases of movement and damage sustained during static periods.

2.3.3 Modelled transport distances were based upon comparisons between the damage sustained by experimental artefacts over known transportation distances and that displayed by artefacts from archaeological secondary contexts.

2.3.4 The modelling allowed the source areas of derived artefact assemblages to be mapped and inter-assemblage comparisons to be undertaken.

2.4 The range and types of archaeological data in secondary context assemblages (Modules 4, 7 & 8)

2.4.1 The vast majority of data consists of fluvial sedimentary deposits and stone tools.

2.4.2 Although palaeoenvironmental material occurs within archaeological secondary contexts, the contrasting spatio-temporal scales of the data do not permit their direct equation with the derived artefact assemblages. When dealing with derived, secondary contexts, it cannot be assumed that the flora, fauna and artefacts were ever associated in time and/or space prior to their deposition within the same sedimentary deposit. Reconstructed palaeoenvironments are therefore only specific examples of some of the range of habitats that existed prior to and during deposition, but they cannot be explicitly populated with either hominids or artefacts.

2.4.3. The spatio-temporal models of the structure of the archaeological secondary contexts indicated nine distinct analytical scales for the interpretation of the artefact data. Chronological scales ranged from sub-MI stages (100’s and 1,000’s of years) to MIS cycles (100,000’s years). Spatial scales ranged from local (10’s and 100’s of metres) to regional, river system basins (1,000’s of metres).

2.4.4 Research questions and techniques vary between the spatio-temporal scales (e.g. the analysis of technological trends at local, sub-MI stage scales, and demographic modelling at regional, MIS-cycle scales). Case studies were presented for the secondary context assemblage at Broom (Module 4) and the Axe and Test valley basins (Module 10).

2.5 Relationships between archaeological secondary contexts and extant analytical and interpretive frameworks (Module 10)

2.5.1 Archaeological research operates at a variety of data scales and resolutions. The project therefore sought to map the secondary context archaeological resource against appropriately scaled research questions.

2.5.2 It was clear that there are specific research questions that cannot be answered from archaeological secondary contexts (e.g. on-site analysis of spatial patterning in artefact distributions and subsistence practices).

2.5.3 It was also clear that there are specific research questions that are best addressed through archaeological secondary contexts (e.g. demographic patterns over MIS cycles and regional variations in artefact densities).

2.5.4 In order to fully investigate hominid behaviour, it is necessary to integrate primary and secondary context data. The frameworks proposed here have therefore highlighted the scope of secondary context investigations and demonstrated their incorporation within existing research frameworks more traditionally associated with primary context data-sets.

2.6 With respect to the management of the secondary context resource (Module 9)

2.6.1 The aggregates industry has facilitated the recording of the majority of UK Palaeolithic data, due to the deeply buried nature of the resource.

2.6.2 Therefore, archaeological watching briefs are the principal available mechanism for the monitoring of archaeological secondary contexts and the recording of data.

2.6.3 Current archaeological watching brief practice offers a range of contingency measures should in situ archaeology be discovered and suggests that monitoring is primarily concerned with such discoveries. This practice is unhelpful as it both downplays the value of the secondary context resource and presents an unfavourable image of archaeological interventions to the aggregates industry.

2.6.4 The spatio-temporal models have highlighted the importance of geochronological and sedimentary data from archaeological secondary contexts. It is therefore proposed that archaeological watching briefs need to:
a. Explicitly state that Palaeolithic archaeological data from secondary contexts is not restricted to lithic artefacts. This principle should be explicit within both project design and practice.
b. Develop cost-effective strategies for the systematic geochronological sampling of fluvial sedimentary sequences, utilising recent developments in OSL and amino-acid techniques. Developing securely dated geochronological sequences is important both for regions with and without a rich artefact heritage.
c. Develop cost-effective strategies for the systematic lithostratigraphic recording of fluvial sedimentary sequences. These recording processes can be streamlined using new technologies (e.g. digital data-capture devices).
d. In regions with rich artefact assemblages, reduce the focus upon the collection of lithic artefacts.
e. However, in regions without demonstrated Palaeolithic occupation, sampling for lithic artefacts should be afforded a higher priority.

2.6.5 It is stressed that these proposals for watching brief practice would need to be developed in conjunction with the aggregates industry, and that the archaeological community needs to emphasise:
a. That the primary goal of the watching brief process must still be preservation by record, given the destructive nature of the aggregates extraction process.
b. That the principle foci of this preservation by record are the secondary context data (sedimentary sequences, geochronological sampling, and artefact recovery from exposed sections).
c. That the secondary goal of the watching brief process must be the streamlining of on-site time and the minimising of disruption to industrial process.
d. That primary context, in situ discoveries are extremely rare and that the profitable recovery of valuable secondary context data can be rapid and cost-efficient.

3. Future Directions

The project has also highlighted a number of areas towards which future research should be profitably directed:

a. The development of regional geochronological frameworks for fluvial sedimentary sequences.

b. Refinement of geochronological techniques with respect to their application and resolution over Pleistocene timescales.

c. Experimental research exploring artefact behaviour within fluvial systems and the formation of secondary context assemblages.

d. Re-investigation of historic secondary context artefact assemblages, utilising current models, recording methodologies and techniques.

4. Conclusion

In conclusion, this investigation into the archaeological potential of secondary contexts has demonstrated that:

a. Due to their widespread geographical and temporal coverage, archaeological secondary contexts from the British Palaeolithic are a unique resource.

b. The taphonomic processes associated with the spatio-temporal structure of the resource can be successfully modelled using extant data and new techniques.

c. Valuable data for the reconstruction of early human behaviour can be extracted from secondary context artefact assemblages, at a range of different spatio-temporal scales.

d. The development of appropriate strategies for the continuing management and recording of the secondary context resource during aggregates extraction are therefore vital.

e. The integration of the primary and secondary context resource is critical to gaining a fuller understanding of early human behaviour during the British Palaeolithic.

NB: the project outputs are summarised at product outputs.

Return to top