The main evolution of societies in the standard neo-evolutionary sequence is seen as a progression from quite small-scale totally autonomous groups called “bands” towards very large-scale and highly centralized social formations called “states” through still ill-defined intermediary stages labeled “tribe” and “chiefdom.” The whole progression then is defined in terms of political organisation (Yoffee 1993). It is in this context that the long-term history of political organisations in Equatorial Africa (Figure 13.1), described in a recent study (Vansina 1990), is relevant to test the theory. Equatorial Africa designates the northern part of Central Africa – an area as large as the United States east of the Mississippi, mostly covered by various tropical rainforests. Anthropologists have recognized some 450 ethnic groups there, and it is possible to distinguish about twenty-five different political regimes in the area (using the distinctions common in social anthropology and disregarding a special “band” status for foragers such as pygmies since these constituted only a part of a society, the other part being that of their farmer hosts). These political institutions are based on the most diverse ideological principles and range in size from single villages, such as Libinza (Ngiri area, Zaire) comprising as few as a hundred or so inhabitants, to kingdoms encompassing well over 100,000 inhabitants. If one includes the kingdom of Kongo in its centuries-long heyday, there was even a polity of 500,000 people. The range includes the “tribe,” “chiefdom,” and “state” rungs on the neo-evolutionary ladder. In addition, equatorial Africa is especially interesting because all these political regimes developed out of a single ancestral political system, existing 5,000 years ago. Moreover, they did so largely in isolation from the outside world.

As the last chapter illustrated, floodplains have a history – they evolve over time. In order to be able to unravel this history and relate it to other events, be they climatic or cultural, sediments have to be sequenced and dated. Advances in dating have revolutionised not only archaeology (Renfrew and Bahn, 1991) but also recent geology (Vita-Finzi, 1973), and the development of sediment-based techniques will undoubtedly have a major impact on studies of the environmental archaeology of past societies, particularly in alluvial environments.

Dating and alluvial chronologies

The stratigraphy of a floodplain represents a time-sequence, so the development of the floodplain over time can be investigated if the stratigraphy can be dated. The dated alluvial stratigraphy or alluvial chronology can then be correlated with, or even used to date, human activity on the floodplain or in the river catchment. As with other Quaternary sediments, there are two, fundamentally different, groups of dating methods that are commonly used to date alluvial sequences. These are the chronometric methods that provide independent age estimates, either numerical or calibrated, but based upon some kind of physical, chemical or biological clock. Other methods are relative, in the sense that they operate both through stratigraphic order and through matching stratigraphic sequences to a similar sequence elsewhere which may be chronometrically dated. Some relative methods can be so well correlated to chronometrically dated or historical sequences that they can be regarded as a third category: calibrated relative methods.

The origins of this book may help to explain its existence, structure and content. Having undertaken doctoral research jointly supervised by a geomorphologist and a palaeoecologist which had strong archaeological implications, I was by the mid-1980s firmly situated in a multidisciplinary and interdisciplinary approach to fluvial environments. Whilst at Leicester University my contacts grew with researchers in other departments interested in alluvial environments, most notably Zoology, Botany and Archaeology. Increasing contact with archaeologists both at Leicester University and in Leicestershire and Northamptonshire Archaeological Units stimulated my interest in geoarchaeology and the cross-fertilisation of geomorphology and archaeology. This was given a research foundation when I realised that archaeologists were regularly digging large holes in floodplains and valued stratigraphic assistance. I worked first on the Raunds Area Project, Northamptonshire, and since then have worked on many alluvial sites in the Midlands of England and elsewhere. In 1989 I took a joint appointment between Geography and Archaeology at Leicester and began to teach environmental archaeology, which included geomorphology, to archaeologists. Contact with archaeologists on a daily basis has I hope benefited this book, not least through the undermining of the simplistic and naive tendencies concerning culture and society that are all too typically held by natural scientists.

This book stems from this history and a desire to bring together work in a variety of disciplines pertinent to the study of alluvial archaeological sites and floodplain geomorphology and palaeoecology.

This chapter is concerned with the palaeoenvironmental and archaeological record that has been derived from terrace gravels and associated sediments. In these contexts archaeological sites are generally either accumulations of derived artifacts or concentrations of discarded tools associated with temporary camps. The cultures responsible are predominantly non-agricultural and although they undoubtedly had impacts on the environment, floodplains were essentially natural. The combination of geology, geomorphology and archaeology needed to study these sites is typical of the sub-discipline of geoarchaeology and alluvial geoarchaeological studies from both North America and Australia are described here as the clearest examples of this type of archaeology. The multi-disciplinary approach is exemplified by the Thames, England; however, the palaeoenvironmental data have been accumulated over many years, leading, unsurprisingly, to variations in coverage and quality which make unrealistic a definitive Pleistocene and archaeological history of this and other rivers. Instead, the discovery and excavation of new sites will lead to the revision (or abandonment) of current models (Gibbard, 1994; Bridgland, 1994) just as they revised models before them (King and Oakley, 1936).

Palaeolithic terrace sites and floodplain use

Artifacts and bones from river gravels form some of the earliest evidence of the peopling of North America and Australasia. The archaeology of the Palaeolithic in Europe is also predominantly derived from terrace gravels, caves and rock fissures.

The last chapter outlined the impact of humans on the floodplain system through the management and control of floodplain resources. This begs several questions; were floodplains preferentially settled and if so why, and how have people coped with the negative as well as positive aspects of floodplain locations? This chapter explores these questions and the data that archaeology can generate that can be used to test such hypotheses and other hypotheses concerning relationships between culture and environment. Floodplains not only have particular sets of positive and negative resources (in the widest sense) but also may have ritual or mythic significance – all of which is part of the cultural archaeology of floodplains.

Locational data and alluvial environments

Any explanatory theory of the location for any activity, excluding enforcement but including inertia, must relate to the perceived advantages and disadvantages of that location as opposed to other locations. When we are dealing with a single or primary site function (such as mining), this seems obvious, but in reality this is rarely, if ever, the case. Even when there was no relocation decision, as is implied by the occurrence of continuously occupied multi-period sites, the perception of positive and negative factors is important in preventing abandonment. This is especially true in the case of dynamic, or hazardous, environments, such as steep slopes, volcanoes and floodplains.

This chapter considers the alluviation and the burial of floodplain surfaces and sites that is such a common feature of the later Holocene in many different climatic environments. Just as the record is spatially variable so too are the causes and particularly the balance between the human and climatic signal contained within the alluvial record.

Late Bronze Age and Iron Age alluviation in the British Isles

From the Bronze Age onwards in the lowlands of North-West Europe, floodplain and fen-margin sites reveal evidence for increased flooding and alluviation. The picture is much less clear for upland sites (Richards, et al., 1987). This is partly due to the lack of preserved archaeological sites on the valley floors and the more dynamic response of upland rivers to individual storms. Before describing some of the evidence for lowland alluviation, it is worth considering the types of data that may be available, other than a dated increase in overbank-silt deposition. First, pedological data: a lack of soil development and soil structure, caused by a lack of bioturbation and soil development. Soil micromorphology frequently reveals a decrease in pedological fabric and an increase in unbioturbated sedimentary micro-features (Limbrey, 1992). This could be caused by an increase in flood frequency and/or an increase in the sediment loading of overbank flows. The former is probably more important than the latter as fine sediment is supply- rather than transport-limited although both may be the result of climate and land use change.

This chapter examines the ecological characteristics of floodplains, how these are related to environmental parameters and how past environmental conditions can be reconstructed from organic materials preserved within floodplain sediments. The implications for past cultures lie in the resources that these ecologies provide at different times and in different places.

Floodplain productivity

Floodplains are more productive than the land that surrounds them owing to the greater availability of water and nutrients. However, their productivity still varies with latitude and local climatic factors. The net above-ground primary productivity of a forested fen in Minnesota has been measured as 746 g m2 yr–1 while that of a Louisiana bottomland hardwood forest is 1374 g m2 yr–1 (Mitsch and Gosselink, 1993). Net productivity and the difference between floodplain productivity and that of surrounding land is controlled by water supply. In the humid tropics where there is abundant year-round water this difference is minimised and the floodplain margin may not be an obvious ecological feature (Figure 4.1). As seasonal contrasts increase and rainfall decreases in the subtropical climates, the difference is at its maximum because of the greater supply and storage of water in floodplains. In Cool Temperate zones the contrast is smaller and differences in soil type and agricultural potential become more important. In the Boreal zone disturbance of vegetation by highly active rivers complicates the situation. The difference between floodplain net primary productivity and dryland net primary productivity structures the rest of the ecosystem, increasing both animal and plant resources.

Place- and river-name evidence

The study of place- and river-names, which has seen a revival in recent years, has two principal aims. One aim is fundamentally a part of historical linguistics, the study of the history of early languages, elements of which may only survive as place-names from pre-literate languages. The other is archaeological, providing supplementary data on the character and history of landscapes or the perceptions of landscape by prehistoric peoples (Gelling, 1988). River-names are a particularly important sub-set of place-names as it would appear that they may preserve some of the earliest evidence of prehistoric languages and have often remained relatively unchanged or mediated by subsequent languages. In Britain most recent work still uses as its starting-point the classic work by Ekwall (1928), supplemented by Jackson (1953) and Rivet and Smith (1979).

Place- and river-names have a remarkable longevity, often being used long after the local language has been forgotten or suppressed, and so they can provide indications of the form of early languages (Renfrew, 1987). In his classic work on river-names (hydronomy) Hans Krahe (1957) pointed out the similarity of river-names in Central Europe (e.g. Ara in Germany, Holland, England, Scotland and Spain, or Soar, England and Saar, Germany) and he suggested that this indicated a common root in Old European or pre-Indo-European languages. It is also possible that the similarity stems from an early stage in the differentiation of an undifferentiated early Indo-European language spoken in Europe north and west of the Alps that eventually produced the Celtic languages (Renfrew, 1987).

This final chapter further develops the theme of human–environment interactions in alluvial contexts. It aims to show how environmental change at a variety of temporal scales and human response can be viewed not as being deterministic, but, owing to sociocultural factors, as semi-predictable if enough is known about human perceptions of environmental change and the socioeconomic structure of society. A simplistic neo-classical model is used as the springboard for exploring the many social, cultural and political factors which cause this to be so. It is argued that at the root of past simplistic views of human–environment interactions has been a deep-seated conceptualised dichotomy between humans and environment most poignantly illustrated in the false dichotomy of the ‘human versus climate’ debate over the causes of alluviation.

Palaeohydrology, climate and resources

In 1967, Sir Mortimer Wheeler commented caustically on the ‘established belief in recurrent changes of climate, the imprecise usage of the term such as “wetter”, “drier”, “warmer”, “more genial”: “wetter” etc. than what?’ (Wheeler in Raikes, 1967). In chapters 3, 4 and 8 various methods of reconstructing past river flow and climate were introduced. The reasoning employed in palaeohydrology is typical of that used in the interpretation of environment–climate interactions (Figure 10.1). Deductive reasoning is used to determine level 2 information. Implications for levels 3, 4 and 5 are considerably more speculative.

Floodplains are one of the most conspicuous and widespread of all the landforms on the earth. They are the result of both erosional and depositional processes. Over time they develop and change and so they evolve, not to any end-point but to the form that they are today or were at any point in the past. The study of the processes and history of floodplain formation is in part a historical science (i.e. geology) and has much in common with scientific archaeology in both its history and methodology. An understanding of the fundamental processes and products of floodplain evolution is essential for the interpretation of sites in alluvial contexts and can yield fascinating insights into human–environment relationships.

Floodplain evolution: an introduction

Floodplains may be simply defined as the flat areas adjacent to rivers liable to flooding. Floodplains are also complex assemblages of landforms which, as shown in Figure 1.1, include: channel features such as bedforms (ripples and dunes) and bars (e.g. point-bars), channel-edge features such as banks, benches and levees, and floodplain features such as old channels (oxbows), old levees (scroll-bars), backswamps and crevasse-splays. The formation and character of these features will be covered later in this chapter. The existence, development and arrangement of such features is a record of the past history of the river and they may also subtly constrain the current activity of the river.

The aim of these appendices is to provide some theoretical background to parts of the thematic chapters and to provide supplementary information and comment on aspects related to the perception of alluvial environments. Appendix 1 provides the theory behind chapter 3 and palaeohydrology, whilst appendix 2 provides the theoretical background to flood frequency analysis, an important method used frequently throughout the book. Appendix 3 includes material which whilst, not central to the book, does illustrate and elaborate the discussions of the human use and perceptions of alluvial environments in chapters 8 and 9.

What makes riverine or ‘alluvial’ environments different, both from other environments and from each other, and how does this affect the archaeological record? How can we study environmental change in alluvial environments and what impact has it had on human populations? This book aims to answer these questions. It also aims to provide an introduction to the physical and biological aspects of alluvial environments which are central to an understanding of archaeology on, under and near floodplains. Questions of preservation, transportation, burial, environment and subsistence are all intimately related to the characteristics of the landscape and are also essential components in any archaeological interpretation. Another aim of the book is to introduce archaeological aspects of alluvial history to environmental scientists and geographers because the vast majority of, if not all, contemporary floodplains have to a greater or lesser degree been altered by human activity during the last 10,000 years. Indeed some have been so altered as to make them in part artifacts and as such indicators of the impact of humans on the environment. This book is therefore about both the impact of humans on their environment, and the impact of the environment on humans. In order to illustrate this and lead the reader through the complete cycle of the inference of cultural implications from the environmental data a classic example is used: the Nile.