Читать книгу The Commodification Gap - Matthias Bernt - Страница 11

The ambition of this book is thus directed towards theory building, which will be achieved through a comparison of empirical materials. But how can gentrifications happening in different cities and at different times be meaningfully compared? Answering this question demands some reflection on the ways in which comparisons are composed, why they are done and how the similarities and differences observed can be brought together. Against this background, this chapter now lays out the basic etymological and ontological orientations of this study and presents its methodological design.

Separating the shared essence of a phenomenon from its various expressions is an epistemological problem that puzzled even ancient philosophers. It has proven to be so fundamental to our understanding of the world that wrestling with it has resulted in fundamentally different axiomatic positions and practices in scientific research. In this sense, the difficulty of coming to grips with the applicability of the gentrification concept beyond the classical cases of the UK and USA described above reflects a deep‐seated problem faced by social science in general.

Crucially, in conceptualising gentrification as either a universal phenomenon or a specific local experience only relevant to a handful of cities, the proponents of the described debate have entered the rocky waters of comparative methodologies. While doing so, they have necessarily taken on board a series of long‐established methodological problems that come along with the comparison of complex social phenomena.

The first is the use of comparison. Why are comparisons done? What is the point of comparing urban change in New York’s Lower East Side in the 1980s with something that is going on in, say, Bangalore today? In fact, there are many reasons for carrying out a comparative study. The aim could be to explore whether a theory developed about the causes of gentrification (e.g. middle‐class invasion, the rent gap, or cultural upgrading) holds true when some of the variables it is built upon vary. Alternatively, one could examine a small number of cases holistically to see whether similarities or differences observed between the cases can be related to causal conditions. These two ways of proceeding can be termed as variable‐oriented vs. case‐oriented strategies of comparison (Ragin 1987, pp. 54–55). Some authors have also argued for comparison as a mode of thought, enabling ‘defamiliarisation’ and assisting in uncovering the hidden assumptions a theory is built upon (Robinson 2006; McFarlane 2010). In this view, studying a process in a non‐familiar environment (e.g. gentrification in a shrinking city) could function as an eye‐opener and allow factors and connections that are hidden elsewhere to be revealed.

No matter the particular motivation, the essential goal of all comparison is not to describe something, but to understand and explain it. The point here is to ask the reasons why a particular comparison is made. If apples and bananas are compared with donkeys, we need to have an idea about what constitutes fruits in contrast to animals. Yet, what exactly is the link between a theory, say about fruits, and the need to compare? While most comparative researchers agree that comparisons refer to theory, there is no privileged stage of theory building in which it is most appropriate. Comparing can, thus, contribute to uncovering the limits of a theory, it can help to draw a contrast between cases and it can be used to suggest testable hypotheses. These hypotheses, in turn, can create demand for a new theory that can then be tested again by including new cases (Skopcol and Somers 1980). The relation between theory building and empirical comparative work is, thus, cyclical and there is no hierarchical (but rather a mutually reinforcing) relationship between theory and comparative empirical observation.

The current treatment of gentrification in urban studies as either a particular local experience, or a planetary urban phenomenon, occupies an uneasy position when examined against these potentials (see also Bernt 2016a). While individualising accounts have proven to work well for attacking the usefulness of gentrification theory for a specific case, they more often than not leave one wondering whether the problem is just a misclassification of an individual case (e.g. when the gentrification concept is applied to cases where gentrification doesn’t happen), a linguistic issue or something that is more deep‐seated and inherent in the concept of gentrification. Universalising accounts, on the contrary, often relegate differences to the role of contextual factors, local specificities or contingencies, seen as negligible for theory building. Here, the consequence is an immunisation of established theories, but hardly their advancement. In sum, both ways of comparison fall short. This results in a stalemate in which both sides employ more or less convincing evidence to support their claims, but leave existing theories untouched and fail to frame a way forward.

One reason for this lies in the particular limits that are specific to either of these forms of comparison. There have been elaborate discussions about the difficulties of comparing complex social phenomena (like revolutions or cities) that are characterised by ‘small n, many variables’ (Lijpart 1971) in the past, especially within the field of comparative history. As a consequence, manifold useful strategies for conducting comparisons have been proposed (see for example Przeworski and Teune 1970; Lijphart 1971; Sartori 1970, 1991; Smelser 1976; Tilly 1984; Skocpol 1979; Skocpol and Somers 1980; Ragin 1987). In this context, Charles Tilly (1984) has developed a widely cited typology that distinguishes four types of comparison along the two dimensions of scope and number (see Figure 1.1). Thereby, in scope, comparisons can range from quite particular (getting the case right) to quite general (getting the characteristics of all cases right) and in number, comparisons can range from single to multiple.

On this basis, Tilly (1984, pp. 82f.) distinguished between individualising, universalising, generalising (or variation finding) and encompassing comparison. In this scheme, individualising and universalising comparisons are antidotes: individualising comparisons contrast specific instances of a phenomenon as a means of grasping the particularities of each case, whereas universalising comparisons aim to establish that each instance of a phenomenon follows the same rule (ibid.). The point is that both types of comparison can be made, but each has a different function for different purposes. Individualising comparisons work well to illustrate a given theory. They can also be very effectively used to challenge the validity of a theory. The problem, however, is that they leave much to be desired when it comes to developing a new theory. Universalising comparisons, in contrast, are an effective means to go beyond the surface of similarities and dissimilarities and produce models and generalisations that allow for theorising. More often than not, the price for this is a neglect of differences.

FIGURE 1.1 Types of comparison (Tilly 1984).

Expressed differently, both types of comparison have their particular strengths and weaknesses – but only in relation to a specific strategy in making a theoretical argument. What counts is not the comparison as such, but the theoretical argument. It is in this field that both individualising and universalising approaches have their Achilles’ heel.