The Role of GIS in the Interdisciplinary Investigations at Olorgesailie, Kenya, a Pleistocene Archaeological Locality

A geographic information system is an ideal tool for use in interdisciplinary studies because it provides automated means of linking and relating different spatial databases. In this paper we discuss GIS applications to ongoing archaeological and paleoecological studies at Olorgesailie, an early hominid archaeological locality in the rift valley of southern Kenya and one of the most noted Acheulian handaxe sites worldwide (Isaac 1977). The questions being asked in early hominid archaeology require thinking beyond individual artifacts and site excavations to broader spatial scales within welldefined time intervals (or chronostratigraphic units) (Blumenschine and Masao 1991; Potts 1991). The sedimentary exposures at Olorgesailie permit the smallest spatial scale of individual artifacts and fossils to be integrated with regional-scale studies. Since many of the GIS applications are still in initial form, the purpose here is largely to illustrate the conceptual framework by which GIS integrates the analysis of spatial data at varying geographic scales in the Olorgesailie basin. Covering over 4000 km in length, the African Rift System trends southward from the Afar Triangle in the Red Sea region to south of the Zambezi River in Zambia. The numerous continental rift basins that make up the rift system have a complex structural and volcanic history. For most of its length, the African Rift traverses Ethiopia, Kenya, and Tanzania. The rift is divisible into eastern and western portions, which merge into a broad faulted region in northern Tanzania (Baker et al. 1972). Between the eastern and western rifts, occupying portions of Uganda, Tanzania, and northern Kenya, is an uplifted plateau 1000 to 1200 m in elevation. Uplifted, elongated domal structures located in Ethiopia and Kenya form the structural base from which the East African Rift System has developed. The rocks that make up this shield complex are Precambrian gneisses, quartzites, and schists. In addition to intrusions by dikes and plutons, these basement rocks have been altered by partial melting and metamorphism. Significant though episodic uplift of the Kenyan dome and its flanks during the late Cretaceous and middle and late Tertiary contributed to the development of a graben structure (Baker 1986; Baker et al. 1972).

Integrating Socioeconomic and Geographic Information Systems: A Methodology for Rural Development and Agricultural Policy Design

Understanding the factors related to destructive ecological processes in the tropics has expanded significantly in the last decade. Much has been learned about heterogeneity in geomorphology, soils, hydrology, and climate and about associated vulnerability to ecological damage. Research on cropping systems has divulged both the suitability and the liability in swidden agricultural practices and has led to recommendations involving alternative cropping and agroforestry complexes (Altieri 1987). At the same time, there has been a growing awareness that a more comprehensive knowledge of tropical ecology and enlarged technological and/or agricultural options will not necessarily affect a sustainable ecology (Altieri and Hecht 1990; Redclift 1984, 1987). Research on peasant economies in Latin America and elsewhere has demonstrated the existence of a highly differentiated peasantry, the vast majority of whom are landless or land-poor and who are more dependent on income earned from off-farm than from on-farm sources (Collins 1986; Deere and Wasserstrom 1981; Stonich 1991b). Such studies have demonstrated that systemic interconnections among family and corporate farmers with landholdings of all sizes promote environmental destruction (Stonich 1989); have established the existence of labor scarcity rather than labor surpluses in many peasant communities and the related environmental consequences (Brush 1977,1987; Collins 1987,1988; Posner and MacPherson 1982; Stonich 1993); and have called for rural and agricultural development policy that takes into account a socially differentiated peasantry and diversified rural poverty (de Janvry and Sadoulet 1989). It is increasingly evident that ecological destruction cannot be fathomed apart from the demographic, institutional, and social factors that influence the agricultural practices and other natural resource management decisions of agricultural producers. This paper describes a multidisciplinary methodology designed to examine the interactions among demographic trends, social processes, agricultural production decisions, and ecological decline in southern Honduras, a region characterized by widespread and worsening human impoverishment and environmental degradation. The methodology integrated the research efforts and databases compiled by anthropologists from the University of Kentucky using a farming systems approach, who were part of the socioeconomic component of the International Sorghum Millet Project (INTSORMIL) with potentially complementary research conducted by the natural and agricultural scientists working as part of the Comprehensive Resource Inventory and Evaluation System Project (CRIES) at Michigan State University.

Land Degradation in the Peruvian Amazon: Applying GIS in Human Ecology Research

Land degradation, a reduction in the productive capacity of land, is a process of increasing concern in the challenge to maintain and enhance global food production. It is an especially critical problem in developing countries faced with the need to increase food availability for growing populations. Billions of dollars are invested in agricultural research and development aimed at increasing the food supply. At the same time, land degradation threatens to reduce production in large areas of agricultural land. While estimates of the magnitude of the problem vary widely (see WCED 1987; WRI/IIED 1988; and Lal and Stewart 1990 for recent reviews), there is a growing consensus that land degradation is a serious and complex problem that merits increased attention from both natural and social scientists. A recent review of this topic by Blaikie and Brookfield (1987) highlights the role of the social sciences in studying land degradation problems. According to these authors, the term “land degradation” refers to a reduction in the actual or potential uses of land due to human activities (1987: 1). The costs of land degradation (“the product of work on degraded lands is less than that on the same land without degradation”) make it a serious social problem for millions of farmers around the world and thus a priority for social science inquiry. A central actor for understanding the causes and consequences of land degradation is the land manager—most often the farmer—who makes the landuse decisions for particular plots of land. Social science has a key role in understanding this process of decision making, including the social and ecological contexts in which decisions are carried out. Anthropology’s emphasis on working with peasants, small farmers, and indigenous people holds out the promise for important empirical and theoretical contributions in understanding land degradation. A human ecology approach that focuses on the adaptive strategies of individuals and groups and the environmental consequences of these behaviors seems particularly well placed to contribute to this topic.

Introduction

Although spatial thinking has long been a part of anthropological inquiry, it has waxed and waned in its perceived utility and centrality to the field. Although the papers in this volume attest to a vigorous tradition of spatial thinking in anthropology and further suggest that, for at least some branches of the field, spatial thinking and analysis are truly central to their definition and mission, it is nevertheless clear that this has not always been the case. Further, despite differences in historical trajectories of development between the two major subfields of anthropology—cultural anthropology and archaeology— in terms of the way space has been used, it is also clear that the two subfields share a number of common interests and themes that deserve discussion and exploration. This exploration is not only interesting from a purely historical perspective, but also has a very practical, down-to-earth dimension. The literature on the history of science is replete with cases of communication failures both within and between scientific disciplines. While in many cases this is merely annoying (different terms used to describe the same procedure, for instance), there are occasions when these failures lead to the creation of a highly idiosyncratic jargon used by small cliques of investigators, which clearly offers the opportunity to inhibit scholarly communication. This, in turn, can lead to redundancy of effort, failure to learn from the mistakes of others, and wasted time and money. By providing a forum in which similarities and differences can be examined, the natural tendency of scientific disciplines to form these cliques can be overcome. I intend this paper to be such a forum for an exploration of the ways in which geographic information systems (GIS) have been employed by anthropologists and archaeologists as represented by the authors of the papers presented in this volume. I will briefly describe the GIS for those readers unfamiliar with it and then turn to a review of the history of spatial thinking and the kinds of tools used to implement this thinking for each of the subdisciplines.

Moving from Catchments to Cognition: Tentative Steps Toward a Larger Archaeological Context for GIS

Geographical informations systems are being used increasingly frequently within archaeological applications. Given the nature of much archaeological data, there can be little doubt that this technology probably represents one of the most flexible and comprehensible tools for the analysis of spatial data presently available. However, there are causes for concern relating to the archaeological context of GIS. This paper suggests that the nature of most GIS is such that they are most readily applied to data that are most conveniently stored in map format and that this may ultimately be restrictive to the natural development of archaeological analysis. In particular it is suggested that the use of GIS modules may lead to the unwitting exposition of an environmentally or functionally deterministic viewpoint of a type that has largely been rejected by most archaeologists. The need to develop cognitive models is emphasized and it is suggested that GIS has an important role to play in the development of such approaches. Particular emphasis is placed on the ability of GIS to incorporate the whole environment within archaeological models and to transform abstract spatial information in order to place it within a cultural domain. Two case studies are presented to support these suggestions. The first involves the re-analysis of a GIS study of late prehistoric settlement and burial data on the island of Hvar by the authors. It is suggested that the original interpretation of these data can be greatly improved through a more thoughtful consideration of the belief systems operating within these communities. The second case study involves prehistoric rock art and other ritual monuments in mid-Argyll in southern Scotland. The GIS-generated viewshed data are used to explore the cognitive context of the monuments within the landscape and to explore the perceived relationship between monuments. The GIS clearly has a lot to offer archaeology. However, there is a need to ensure that we use the technology on the terms of archaeology rather than simply transfer the techniques for which GIS is most commonly used into an archaeological context.

Empirical and Methodological Problems in Developing a GIS Database for Yanomanö Tribesmen Located in Remote Areas

For almost thirty years Chagnon has been studying the settlement patterns of a large cluster of remote Yanomamö communities in southern Venezuela, documenting population growth, mortality patterns, fissioning, dispersal, and pioneering of adjacent virgin areas of tropical forest. Approximately fifteen villages, with a current (1992) population of about 2000 individuals, have been studied. During a period of some 150 years, members of these communities have cleared and subsequently abandoned approximately five hundred sites whose geographical locations are very poorly known. In 1990 and 1991, Charles Brewer Carías, a Venezuelan naturalist, joined Chagnon in this research effort. Recent field research has resulted in geographic and demographic data suggesting that long-term warfare patterns may be contests over the apparently more desirable lowland areas, where economic activities are less costly in terms of energy and resources are more abundant or easier to obtain. Periodic village movements, provoked by hostilities with neighbors, require that relatively large lowland areas must be controlled so that groups can move around within them and maintain maximum distance from enemy groups. To do this, lowland villages must grow large and politically bellicose. When they fission, usually at a size of about 150 to 200 people, some of the resulting smaller groups are driven out and take refuge in more rugged but economically less productive highland terrain, where they adopt a less bellicose political stance toward their neighbors. Rates of mortality due to warfare, frequencies of abduction of women from neighbors, and other sociodemographic attributes distinguish highland from lowland communities in the overall area (Chagnon 1992). Geographic information systems are considered effective methods for organizing and analyzing the variety of spatial information required to test such hypotheses of relationships between environment and social processes. A GIS-based approach would allow maps of parameters relating to resource distribution and environmental characteristics to be compared to a rich and growing record of field observations. Analysis based on GIS would support data management requirements by allowing accurate identification and positioning of cultural and environmental features within a consistent map base.

The Use of GIS to Measure Spatial Patterns of Ethnic Firms in the Los Angeles Garment Industry

The purpose of this chapter is to describe the methodology and present some initial results in our efforts to understand the role of spatial organization in ethnic economies of the Los Angeles garment industry. Our research goal was to provide a spatial dimension to our database in order to test theories maintaining that space is a critical component of economic transactions. To this end, we created analytical grids of various resolutions using three GIS functions: address matching, polygon generation, and identity overlay. The grids were then transformed into matrices and tested for spatial autocorrelations using Statistical Analysis System (SAS) routines proposed by Griffith (1992). The following section presents brief arguments for the importance of space from the perspectives of industrial district and ethnic economy theories. The third section describes the research site and data sources. The fourth section outlines the analytical concerns of spatial autocorrelation and the modifiable areal unit problem (MAUP), which leads to a discussion of the relevant GIS operations in the fifth section. The sixth section presents spatial autocorrelation measures that show a direct relationship between spatial concentration and economic success in the Los Angeles garment district. The concluding section summarizes the importance of GIS as an analytical tool. Both industrial district and ethnic economy perspectives are based on assumptions about the importance of space. Business economists and geographers show that arrangements within spatially concentrated, tightly integrated industrial districts are critical to globally competitive industries, rapid information flow, lowered transaction costs, and increased control over production, permitting quick and flexible responses to changing market demands (Scott and Mattingly 1989; Piore and Sable 1884; Storper and Walker 1992; Storper and Christopherson 1987). In particular, Porter (1990)—a Harvard Business School economist and member of former President Reagan’s Council on Competitiveness— emphasizes that geographic concentration increases local competition as well as fostering such “emotional factors” as trust, pride, and bragging rights. Spatial concentration, while less prominent in discussions of ethnic economies, remains an underlying factor in providing a venue for the exchange of cultural capital.

A Time to Rend, A Time to Sew: New Perspectives on Northern Anasazi Sociopolitical Development in Late Prehistory

Slightly before A.D. 1300, the Four Corners area of the North American Southwest was abandoned by prehistoric agriculturists. By that time, populations had undergone three major cycles of aggregation into large settlements, first constructing relatively large “public” facilities and then redispersing. The reasons for the final abandonment of this area, as well as for the earlier collapse of the Chacoan-related system of the mid-1100s, are classic areas of archaeological inquiry. Recently, the earliest cycle of village formation and dispersal, in the A.D. 800s, has come under increased scrutiny as well (Orcutt et al. 1990; Wilshusen 1991). In this paper we reexamine these phenomena by posing a simple but fundamental question: Under what conditions will farmers find it in their own best interest to share the food they produce? Whatever the particular features of these cycles of aggregation and dispersion, we suggest that periods of increasing complexity in the fabric of sociopolitical organization—which involve the growth of settlements, elaboration of social roles and networks, and heightened cooperation in building, hunting, and exchange—are constructed on top of reliable systems of food sharing beyond that expected among close kin. Such resource pooling has the effect of reducing the impact of variability in agricultural production in an area where great unpredictability surrounds the growing of food. Our thinking about how to approach these systems of food sharing has been influenced by recent analyses of sharing among hunter-foragers (e.g., Kaplan and Hill 1985; Smith 1988) and by current discussions of risk and uncertainty in behavioral ecology and microeconomics (Clark 1990; Stephens 1990). This study focuses on an area in southwestern Colorado about 3 5 km north of the New Mexico border and immediately east of the Utah state line. Notable landmarks include the northward bend of the Dolores River on the northeast, the escarpment of the Mesa Verde in the southeast, and the commanding presence of a volcanic laccolith—Sleeping Ute Mountain—on the south.

Danebury Revisited: A English Iron Age Hillfort in a Digital Landscape

The doyen of British field archaeology, O. G. S. Crawford, noted that “The surface of England is a palimpsest, a document that has been written on and erased over and over again“ (Crawford 1953: 51). Many centuries of detailed observation and recording of the English landscape have resulted in a wealth of archaeological data, covering many thousands of years of human habitation. The need to record and decipher these extensive field data has led to the adoption of methods and techniques developed in other disciplines, including that of geography. Geographic information systems are the latest tools to be adopted in the quest for effective methods of field recording and archaeological analysis (for introductory GIS texts, see Aronoff 1989; Burrough 1986; Star and Estes 1990; Tomlin 1990). The applications of GIS in archaeology can be differentiated according to scale and type, although relatively few mature applications currently exist. Studies that have been undertaken range from intrasite to intersite analyses, and from research-driven applications to inventorying and cultural resource management (see for example Allen et al. 1990; Gaffney and Stancic 1991, 1992; Harris and Lock 1992; Larsen 1992; Lock and Harris 1991). This regional study, based on the Iron Age hillfort of Danebury in England represents a contribution to this growing literature and to the development of GIS use in archaeological analysis. This paper has two main aims. First, it seeks to identify and examine the archaeology of the Danebury region within the context of existing archaeological theory and to refine and add to these interpretations where applicable. Second, the paper seeks to undertake this analysis within a GIS environment. Our goal here is to illustrate how GIS can contribute to archaeological analysis, shed new light on existing knowledge, and enhance our understanding of the prehistoric use of the landscape. In landscape archaeology, there are several well-established themes that are strengthened and augmented by the data handling and analytical capabilities of GIS. This paper elaborates and develops these themes in the context of the ongoing archaeological study of the Danebury hillfort region.

An Analysis of Late-Horizon Settlement Patterns in the Teotihuacan-Temascalapa Basins: A Location-Allocation and GIS-Based Approach

The general interest of linking GIS capabilities and location-allocation (L-A) techniques to investigate certain spatial problems should be evident. The techniques and the technology are often complementary. A GIS can provide, manage, and display data that L-A models require; in turn, L-A models can enhance GIS analytic capabilities. This combination of information management and analysis should have wide appeal. The technique and technology may be especially wellmatched when one considers many of the special requirements of archaeological applications of L-A models. We intend to investigate and illustrate the value of such a combined approach though the example of a regional settlement analysis of the Late Horizon Basin of Mexico. Geographic information systems are increasingly common in archaeology. Their ability to manage, store, manipulate, and present spatial data is of real value, since the spatial relationship between objects is often an archaeological artifact in its own right. Space is central to both archaeological data (Spaulding 1960; Savage 1990a) and theory (Green 1990). Although GIS may not always offer intrinsically new and different manipulations or analyses of the data, they can make certain techniques easier to apply. There is a wide spectrum of GIS-based modeling applications in archaeology (Allen 1990; Savage 1990a). The anchors of this spectrum range from the use of GIS in the public sector in cultural resource management settings to more research-oriented applications. The strongest development of GIS-based archaeological modeling is probably in the former context. Models developed here are predominantly what Warren (1990) identifies as “inductive” predictive models where patterns in the empirical observations are recognized, usually using statistical methods or probability models. This type of application is usually identified with “site location” modeling (Savage 1990a). As defined, these models do not predict the probable locations of individual sites but rather calculate the probability that a geographic area will contain a site, given its environmental characteristics (Carmichael 1990: 218). The primary role of GIS in many of these applications is to manage and integrate spatial information and feed it to some exterior model.