Daniel Knitter

From Environment to Landscape. Reconstructing Environment Perception Using Numerical Data

The paper introduces a method that links environment to landscape. The environment-landscape divide appears because of epistemological differences: since studying the landscape involves describing the world as it was perceived by humans, it is difficult to access this dimension through the numerical data that we employ when studying the environment. We approach the issue of non-correspondence between environment and landscape knowledge using fuzzy logic. The numerical data describing two geomorphometric parameters, slope and modified topographic index, are split each into three classes with overlapping borders. The classes are then fused into four qualitative categories: flat wet, steep dry, flat dry, and gradual moist. These four categories have direct correspondence in the real world and can be observed by people through simple perception. The correspondence of such categories to peoples’ perception is checked against evidence of past human settlement in three areas coming from Turkey, Serbia, and Syria. The identified qualitative categories resemble the way people categorized their landscape in all but the second case study. Humans were able to perceive and choose areas which correspond to gradual moist in Turkey and broadly to flat wet in Syria. However, for the Serbian example, the results are inconclusive.

Theory of Modelling

This chapter is dedicated to the term ‘model’ which is explored in the first part. The chapter presents different definitions and characterisations of the term ‘model’ and aims to provide readers with a deeper understanding of differences and similarities in the various concepts of models, as well as their practical applications. A brief discussion of some classifications of models introduces the reader to the most important distinctions in the modelling terminology. Among these classifications, the distinction between theoretical and empirical models plays a dominant role and is pivotal for understanding the methodological chapters. In addition, we present the SQM classification and address the development and reduction of complexity in models. Different purposes and benefits of models are described and may serve as a starting point for describing modelling applications. In order to have practical rules of thumb for the usage of models some modelling directions are given. Finally, examples from archaeology and geography complete this conceptual chapter on modelling.

Critical Physical Geography in Practice: Landscape Archaeology

Landscape archaeology is an emerging interdisciplinary field where researchers from humanities and sciences investigate human/environment interactions and interrelations as well as human perceptions of the environment. It is closely related to Critical Physical Geography except that the investigated humans are long gone. Insights about past societies and their relations to the environment are only indirectly accessible via material traces. We present different challenges when interpreting such material remains that are produced by coupled eco-social systems. In order to tackle these challenges, critical physical geographic thinking is necessary that iteratively questions the application of methods from physical and social science, their data selection and interpretation of results in order to arrive at a comprehensive, transdisciplinary understanding of long gone societies within their environment.

Point Pattern Analysis as Tool for Digital Geoarchaeology: A Case Study of Megalithic Graves in Schleswig-Holstein, Germany

In this contribution, we apply different methods of spatial and geomorphometric analysis in order to present a general approach of data exploration in areas where detailed local information is absent. Our data are based on locations of megalithic graves from Funnel Beaker societies (3700–2800 BCE) in the area of Schleswig-Holstein, Germany. Using these locations, we apply methods of point pattern analysis in order to reconstruct the spatial processes that created the sample: We use density-based measures to show the influence of first-order effects on the dataset. While first-order effects are related to the underlying areal characteristics of the point locations and hence are determinant of their intensity, second-order effects are the result of interactions between points. We conduct distance-related approaches, e.g. focusing on nearest-neighbour characteristics, in order to investigate the interaction between the points. The point pattern analyses are complemented by integrating geomorphometric measures that are indirectly indicative for some general environmental conditions, even in prehistoric times. This helps (a) to relate first-order effects to societal or environmental features and (b) to understand the specific pattern of interactions between the points. The necessary raw data in the form of digital elevation models are freely available for large parts of the globe. All analyses are conducted using free and open-source software in order to provide their limitless application.

Regression and Interpolation

While density analysis only deals with points in a coordinate system, regression and interpolation includes additional information. The points are provided with values. Regression investigates the relationship of the coordinates and the dependent variable. Theoretical functions are used to describe this relationship and these functions are fitted to the empirical data. Among the different techniques of regression, the chapter concentrates on linear models. In the case of interpolation, we are also interested in the relationship between the coordinates and the dependent variable, although we ask for values of the dependent variable where we have no observations. Interpolation means predicting values for points where no measurements are available. The result is a model of the values of the entire area of interest based upon the idea that values are similar when the distance is small. This itself is a conceptual model of the relationship of points in space. Interpolation is an important technique that is applied in many more complex types of models. Inverse distance weighting and kriging are the methods discussed.

Ancient Colonization of Marginal Habitats

The present contribution deals with the concepts of marginal habitats in selected regions of the ancient world, ranging from modern Spain to the Jordanian desert and from Turkey to the Ethiopian highlands. Central to this research is the hypothesis that the occupation of areas beyond the ‘normal’ settlement patterns corresponds to colonization processes which reflect specific social strategies and may have stimulated the development of new technological skills. A review of ‘marginality’ research in various disciplines indicates that there is no comprehensive definition of the concept, which can be approached from a multitude of perspectives and with manifold objectives. A survey of the eight case studies and two more in-depth discussions of the sites of Musawwarat (Sudan) and Ayamonte (Spain) highlight the potentials as well as the limits of the archaeological investigation into past marginalities. Patterns of spatial marginalization are the easiest to detect. The studies also show that we must not limit our analysis to the adverse factors connected to different kinds of marginalities. Instead, our analyses suggest that spatially marginal areas were deliberately chosen for settlement – an integration with core-periphery approaches may help us to understand these scenarios, which have received little attention in ‘marginality’ research in archaeology or elsewhere so far.

Introduction and Mathematics

The emergence and development of modelling in archaeology and geography as well as the relationship of these disciplines is discussed. This narrative is embedded in a broader view on science and humanities. Archaeology and geography are identified as bridging disciplines between the two cultures of science and humanities. This emphasises the interdisciplinary role of archaeology, geography and modelling. In addition, the three sets of data used in the case studies are introduced. These data form a minimal example to exemplify different modelling concepts to highlight their strength and weaknesses. Finally, some mathematical concepts including logic, sets, relations, algebra, graph theory, statistics, stochastic and optimisation are mentioned, since they are vital to understand the content of this book. Brief summaries of concepts and facts as well as references to further reading help to refresh the knowledge and prepare the mathematical background for the following chapters.

Location and Characterisation

Location and characterisation of places is the central topic of landscape archaeology. Places can be characterised according to the landscape in which they are located. The information from a landscape classification can be transferred to places; for example, settlements. This chapter discusses both crisp and fuzzy approaches of landscape classification. The location can be optimised due to providing the place with the best landscape features for a certain purpose. Moreover, the optimisation can be based upon the relationship to other places, which is addressed as structure based optimisation. Predictive modelling draws upon the fact that certain landscape features are optimal for a certain economic and cultural style of life and hence are preferred. Predictive models are regression models showing the likeliness of having a place at a certain location characterised by a certain set of landscape features. Both inductive starting with observations of places with characteristics and deductive approaches starting with a set of theoretical preferences are discussed.