Kvamme, Kenneth L. (1997). Ranter's Corner -- Bringing the Camps Together: GIS and ED. Archaeological Computing Newsletter 47:1-5.

(Copyright is retained by the author; download PDF version, 279 KB)

More about Archaeological Computing Newsletter (ACN)

RANTER'S CORNER BRINGING THE CAMPS TOGETHER: G.I.S. AND E.D.

Perhaps it is a sign of the growing maturity of GIS as a specialty area within archaeology that a number of reflections on what we are doing with the technology--or more accurately what the technology is doing to us--have recently been published (e.g., Gaffney and van Leusen 1995; Gaffney et al. 1995; Gillings and Goodrick 1996; Harris and Lock 1995; Wheatley 1993). Although the current "bandwagon" status of GIS has spewed forth generally upbeat, positive, and even euphoric praise of its potential, for some more seasoned veterans the picture is not all that rosy. In my view a number of overly negative conclusions have arisen from this camp, albeit centered around a kernel of truth. The result is a controversy that borders on name-calling and shows signs of a potential schism that will only serve to fragment our small GIS community. In the following I attempt to clarify misunderstandings and demonstrate that there is unity of purpose and commonality in what we are doing, regardless of perspective.

Physical and Social Environmental Data

The root of the controversy stems generally from the fact that features of the physical environment have taken precedence in archaeological GIS studies. This circumstance is only natural since in most regions maps of the physical environment are relatively easy to obtain. Consequently, through GIS archaeological distributions have been examined for relationships with such environmental factors as soils, geologic, topographic, hydrologic, or biotic conditions.

Yet, it must be emphasized that regions also contain a social environment--the pattern of the distribution of contemporary settlements, the pull and sway of economic forces, the nature of political and social boundaries, the influence of religious or trade centers, and the like. The social landscape has traditionally been much more difficult to get at (both within and outside of GIS), but recent efforts, particularly by European GIS users, are encouraging (Gaffney et al. 1995; Llobera 1996; Ruggles and Medckyj-Scott 1996; Wheatley 1995, 1996). Maps of topography, hydrology, and soils might be relatively easy to obtain in a region, but acquiring maps of contemporary settlements, road networks, political spheres of influence, or socio-economic or religious centers is much more difficult; it requires a great deal of prior research and a good knowledge base of a region. In many contexts the establishment of contemporeneity (a prerequisite for investigation of the social landscape) may be very difficult indeed, as between short-term hunter-gatherer camps of the Stone Age. Lock (1995:16) notes that in general cognitive and social variables are difficult to isolate and measure, particularly when dealing with past peoples. He bluntly points out that "if something cannot be represented...then it cannot be included...in computer-based analysis," an observation that, if more widely heeded, would remove a major source of acrimony and confusion from the on-going debate.

It should be realized at the same time that working with physical environmental variables is not without problems. All too often maps of the present are used as representations for vastly different past environments with potentially disastrous consequences. Reconstruction of past physical landscapes and environments prior to analysis has been sadly lacking in GIS-based studies.

This contrast between the general availability of natural versus cultural data parallels that of the archaeological record where a plethora of evidence in the former category--floral, faunal, and other economic information--are often readily obtained. The recovery of solid evidence of social, cognitive, or ideological factors generally proves much more difficult.

The Specter of Environmental Determinism (E.D.)

In recent years growing displeasure with the preoccupation on physical environmental variables by so many GIS practitioners has been manifested, largely by European scholars who view the social landscape as central to understanding the past. Unfortunately, many in this camp have polarized the debate along "us" versus "them" lines (i.e., those who work with social vs. physical environmental variables). They either reject outright the natural for the cultural or the former is seen as mere background noise that can be filtered out (Gillings and Goodrick 1996). Wheatley (1993) even goes so far as to suggest a "hidden agenda" behind GIS technology that encourages a functionalist perspective owing to the dominance of physical environmental data. Others agree (e.g., Gaffney, in van Leusen and Gaffney 1995; Gillings and Goodrick 1996; Harris and Lock 1995), arguing that as a consequence of this bias use of GIS is not a theoretically neutral undertaking.

Yet, does this sort of bias stem from GIS, or rather from the nature of the data we put into them, and did it not exist long before GIS when archaeologists worked predominantly with paper maps? In my view, GIS should be viewed as only another tool, inherently atheoretical, like other tools or techniques we commonly employ--Maschner (1996) compares it with mass spectrometry, amino acid racimization, and lithic microwear analysis, for example. If GIS analyses are biased toward the natural environment then it is because of the data we feed them, a circumstance little different from microwear analyses that bias our interpretations toward the functions of stone tools simply because they tend to be better preserved than bone or wooden ones. If certain kinds of data tend to be manipulated more often we cannot blame our tools for this shortcoming. Indeed, it seems that GIS is being held responsible for inherent limitations of the physical-material record that forms the basis of our discipline.

Perhaps the worst element of this debate lies in the label of "Environmental Determinism" that has recently been applied to research utilizing physical environmental data (Gaffney and van Leusen 1995; Gaffney et al. 1995; Harris and Lock 1995; Llobera 1996; Lock and Harris 1996). It is certainly an inappropriate one (although admittedly it gained some credence when van Leusen published a defense of physical environmental analysis that actually accepted and utilized the ED label [van Leusen, in Gaffney and van Leusen 1995]).

Environmental Determinism (ED) was a school of thought centered in human geography predominately during the 1920s, but having roots in the previous century. The German scholar Friedrich Ratzel (generally regarded as the founder of human geography), in his early work Anthropogeography, or Outline of the Influences of the Geographical Environment upon History (1882), regarded cultural adaption largely a result determined by natural conditions, a perspective that influenced generations of scholars on both sides of the Atlantic (see Livingstone 1992, Chapter 7). The basic idea behind ED was that the physical environment sets constraints on or determines the nature of possible cultural types and developments. Besides being anthropologically naive (one of the principal critics of this school was none other than Franz Boas, the founder of American Anthropology), there were strong racist themes throughout the ED school. T. Griffith Taylor in Evolution and Distribution of Race, Culture and Language (1921) and Ellsworth Huntington in Environment and Racial Character (1924), for example, invoked their ED perspectives largely to explain the superiority of the European race over others.

No practicing archaeologist today believes that all aspects of the past were environmentally determined and it is unfortunate that this label has been placed on individuals who merely want to ascertain if, for example, agriculturally based settlements in a particular region illustrate statistical associations with specific soils classes.

Let us cease use of this label now.

Nor should we belittle or trivialize one approach as opposed to the other, as Wheatley (1993:) is want to do when he suggests that utilizing physical environment factors in GIS as a basis for explanation "does not tell us much more than we already know.... a very expensive way of stating the obvious." In any domain of archaeology we sometimes find research that employs high-powered statistical or other analyses that illustrate the obvious, and this can happen if GIS demonstrations are repeatedly given that show settlement tendencies on level ground or along rivers, for example. But weak physical environmental relationships with archaeological distributions can sometimes be made clear and new and unforeseen ones can be discovered and documented in regional studies. Past relationships with the natural environment can be just as mysterious, unforeseen, hidden, and difficult-to-discover as cognitive, symbolic, or socio-political circumstances.

A Holistic View

We usually learn in the resolution of controversies that there is truth in both sides and a melding of perspectives is the logical conclusion. How, then, should we view a region as a holistic domain for inquiry into past human behaviors? We should first acknowledge the dynamic relationship between cultural behavior and the natural environment. Maschner (1996:507), for example, observes that although warfare might have large cognitive and symbolic components, the choice of defensible location is largely a function of landform (and that a location selected only for cognitive or symbolic value "quickly leads to a lack of symbolizers"). The physical and social environments both are fundamentally important to understanding the past.

Hilltops might offer defensible locations, valley bottoms greater shelter, river confluences better access for transport, specific soils and hydrologic combinations superior circumstances for crop-growing, and so on. We might regard the physical environment as the backdrop within which the human play is enacted, but one that greatly influences the nature of the acts performed. It sets the boundaries or conditions within which we may assess various socio-cultural phenomena of the past. The nature of relationships between the physical environment and past behaviors will, of course, vary through time, culturally, and from region-to-region.

One way of looking at human locational behavior at the regional level might be as a "proportion of the total variance" explained by various domains (as in a regression equation). The physical environment certainly accounts for a proportion of the variance in location--the nature of landform, water availability, soils, plant and animal communities, the availability of natural resources. Another proportion is accounted for by features of the social environment--the distribution of other settlements, road networks, ceremonial and economic centers, and the like. The remaining variance is unaccounted for or unexplained and derives from idiosyncracies that characterize so much of human behavior.

This scheme recognizes that we can never fully explain all elements of the past; that our behavioral models must be less than complete (as are all models). It also illustrates that we cannot come to a large understanding of human regional behavior if we exclude one or the other of the domains. The physical and social environments both play a role.

Finally, we should note that the foregoing model of regional behavior can be made fully operational in an applied research context simply by working with an archaeological sample distribution and a suite of relevant physical and social environmental variables. Regression or other statistical models could be employed to describe the relationship between each variable and the archaeological distribution, and GIS methods could easily map the results over the region allowing ready visualization of relationships and a means to assess the performance or correspondence of the models with the known distributions (see Warren [1990] for an excellent overview of this methodology). Importantly, proportion of variance explained statistics could be computed for each variable. In such a way we would easily demonstrate (and settle the issue?) that humans past and present respond to both their natural and social worlds. The question is, are we up to the challenge?

References Cited

Gaffney, V., Z. Stan_i_, and H. Watson
1995 The Impact of GIS on Archaeology: A Personal Perspective. In Archaeology and Geographical Information Systems: A European Perspective, edited by G. Lock and Z. Stancic, pp. 211-229, London:Taylor and Francis.

Gaffney, V., and P. M. van Leusen
1995 Postscript - GIS, Environmental Determinism and Archaeology. In Archaeology and Geographical Information Systems: A European Perspective, edited by G. Lock and Z. Stancic, pp.367-382, London:Taylor and Francis.

Gillings, M., and G. T. Goodrick
1996 Sensuous and Reflexive GIS: Exploring Visualization and VRML. Internet Archaeology 1. (http://intarch.ac.uk/journal/issue1/gillings_toc.html)

Harris, T., and G. Lock
1995 Toward an Evaluation of GIS in European Archaeology: The Past, Present and Future of Theory and Applications. In Archaeology and Geographical Information Systems: A European Perspective, edited by G. Lock and Z. Stancic, pp.349-366, Taylor and Francis, London.

Huntington, E.
1924 Environment and Racial Character. In Organic Adaptation to Environment, M.R. Thorpe, ed., pp. 281-99, New Haven:Yale University.

Livingstone, D.M.
1992 The Geographical Tradition. Oxford:Blackwell.

Llobera, Marcus
1996 Exploring the Topography of Mind: GIS, Social Space and Archaeology. Antiquity 70:612-622.

Lock, Gary
1995 Archaeological Computing, Archaeological Theory and Moves Towards Contextualism. In Computer Applications and Quantitative Methods in Archaeology, 1994, Jeremy Huggett and Nick Ryan, eds., pp. 13-18, BAR International Series 600, Oxford:Tempus Reparatum.

Maschner, Herbert D. G.
1996 Review of Archaeology and Geographical Information Systems (Gary Lock and Zoran Stancic, eds., 1995, London:Taylor and Francis). Geoarchaeology 11(6):505-507.

Ratzel, F.
1882 Anthropogeographie, oder Gründzuge der Anwendung der Erdkunde auf die Geschichte. Stuttgart:Engelhorn.

Ruggles, C.L.N., and D.J. Medyckyj-Scott
1996 Site Location, Landscape Visibility, and Symbolic Astronomy: A Scottish Case Study. In New Methods, Old Problems: Geographic Information Systems in Modern Archaeological Research, Herbert D. G. Maschner, ed., pp. 127-146, Center for Archaeological Investigations, Occasional Paper No. 23, Carbondale:Southern Illinois University Press.

Taylor, T.G.
1921 The Evolution and distribution of Race, Culture, and Language. Geographical Review 11:54-119.

Warren, Robert E.
1990 Predictive Modelling in Archaeology: A Primer. In Interpreting Space: GIS and Archaeology, K.M.S. Allen, S.W. Green, and E.B.W. Zubrow, eds., pp. 90-111, London: Taylor and Francis.

Wheatley, David
1993 Going Over Old Ground: GIS, Archaeological Theory and the Act of Perception. In Computing the Past: Computer Applications and Quantitative Methods in Archaeology 1992, J. Andresen, T. Madsen, and I. Scollar, eds., pp. 133-138, Aarhus:Aarhus University Press.

1995 Cumulative Viewshed Analysis: A GIS-Based Method for Investigating Intervisibility, And Its Archaeological Application. In Archaeology and Geographical Information Systems: A European Perspective, edited by G. Lock and Z. Stancic, pp. 171-185, London:Taylor and Francis.

1996 The Use of GIS to Understand Regional Variation in Earlier Neolithic Wessex. In New Methods, Old Problems: Geographic Information Systems in Modern Archaeological Research, Herbert D. G. Maschner, ed., pp. 75-103, Center for Archaeological Investigations, Occasional Paper No. 23, Carbondale:Southern Illinois University Press.

Kenneth L. Kvamme