Critical Realism In Geography Research Paper

1. A Definition

Critical realism argues that the natural and social worlds we study exist largely independently of the researcher, and that there is necessity in the world which enables and constrains what can happen. Causation is to be understood in terms of the powers that objects possess rather than in terms of empirical regularities among events; if and when these ‘causal powers’ are activated, the results depend on contexts, and are unlikely to be regular. Since social phenomena such as actions and ideas are intrinsically meaningful, unlike the subject matter of the physical sciences, their meaning also has to be interpreted. The world can only be understood in terms of available discourses. While there are no privileged routes for discovering the truth about the world, and empirical tests are themselves fallible, it is often possible to distinguish better from worse accounts in terms of their adequacy for informing material practice.

2. Intellectual Context

Critical realist philosophy was developed by Roy Bhaskar out of his earlier work on the philosophy of natural science (‘transcendental realism’) (Bhaskar 1978, 1989). The latter sought an alternative to both those positions which treat knowledge as straight-forwardly based on observation and experience (empiricism) and for which truth is a simple matter of correspondence between statements and the objects to which they refer, and various relativist positions which treat the truth of theories as purely relative to discourses. Critical realism also steers a course be-tween positions which regard social science as the same as natural science and those which treat it as entirely different.

3. Realism And Fallibilism

The most fundamental realist assumption is that much of the world can exist independently of our knowledge of it. The plausibility of this assumption rests upon the fallibility of knowledge; the fact that knowledge claims sometimes prove to be mistaken and generate false expectations implies that the world is not a product of our thought but is whatever it is regardless of what we think about it. Realists reject the empiricist view that reality is whatever we can experience since this implies an improbable coincidence between the world and the limits of our sensory powers, and implies that either knowledge is already complete, or that new knowledge creates new objects. Critical realists accept that our observations and knowledge are always formed through available discourses or are ‘theory laden’, so that we can never escape from our conceptual frame-works to see how they compare with the world they seek to represent. However, that observation is theory laden does not necessarily mean that it is entirely determined by theories and so it can still be possible to compare the adequacy of different theories from within discourses and find observations which contradict expectations. This offers a way of avoiding the relativist view that truth is merely relative to theoretical systems. It also implies that science can make progress in explaining the world even though absolute truth is not an intelligible goal.

Critical realism therefore does not involve claims to have a privileged access to the truth about the world, a uniquely ‘realistic’ view, for the very independence of the world from our knowledge of it renders such claims problematic. It must necessarily view all knowledge as in principle fallible, although this does not mean, of course, that it is all false or all equally true or untrue.

4. Key Arguments And Their Influence On Geography

Critical realism became influential in geography partly for the above reasons, for geographers, like other scientists, were seeking a way of understanding science that avoided the unappealing alternatives of empiricism and relativism. However, a further important factor was probably its view of causation and systems. Some of the first discussions of critical realism in geography emphasized this issue and its implications for the nature of theorizing in geography (Sayer 1982, 1985).

The scientific revolution in geography, exemplified by location theory and spatial analysis, had largely assumed that causation was a matter of empirical regularities or ‘constant conjunctions’ among causes and effects. It was expected that understanding of the regularities in spatial organization of phenomena would grow to the point where it would be possible to formulate laws about them. This regularity-based view of causation had dominated the philosophy of science despite the fact that, particularly in social sciences like human geography, there appear to be few enduring precise empirical regularities, so that the search for laws of social and spatial behavior has been singularly unsuccessful. For realists this is unsurprising. For them, causation is a matter of what produces change, not a matter of empirical regularities in sequences of events. Objects have causal powers, that is, the power to act in certain ways. A seed has the power to germinate, a person the power to work. These objects have these powers in virtue of their structures, such as the biochemical constitution of the seed, or the anatomical and mental structures of a person. Whether these powers are ever activated depends on conditions whose presence is contingent, that is, neither necessary nor impossible, such as moisture and solar energy in the case of the seed or a need for food in the case of the person. Furthermore, the particular effects which follow when these causal powers are activated depends on conditions which again are contingent. The germination of the seed may be encouraged by fertilizer or halted by herbicides, the attempt of the person to work may be hampered by lack of tools or encouraged by their availability.

This alternative conception of causation has far-reaching implications. First, since it dissociates causation from empirical regularities, it enables understanding of how unique events, as much as repeated ones, can be caused. Second, it focuses on necessity rather than regularity or the search for empirical order. Third, in view of this, explanation does not require repeated observations: what makes something happen has nothing to do with the number of times it has or has not been observed to happen. It follows from this that the old debate in geography between idiographic and nomothetic approaches—the former a supposedly nonscientific study of the unique, the latter a supposedly scientific search for laws—was falsely based. Regardless of whether phenomena were unique or widely replicated, they could be explained in the same way.

The uniqueness of many geographical phenomena no longer posed a threat to geography’s scientific status. Fourth, it became possible to see that the search for generalizations is distinct from explaining how causes work; generalizations, as the term suggests, attempt to tell us how common or extensive certain phenomena are, and this is different from explaining how they are produced. The greater part of standard literature on the philosophy and methodology of science had confused these two things, and many geographers had done so too. Fifth, the realist account offers a way of understanding how it is possible for the same cause to produce different effects and for different causes to produce the same effects, since the relationship between causal powers and empirical outcomes is not one of automatic regularity but is dependent on the conditions within which they are located.

An example of this can be given from a kind of economic geography which was popular at the same time critical realism became influential. The geography of industrial restructuring looked at changes in economic geography in terms of how competitive pressures upon firms to restructure produced responses (e.g., Massey and Meegan 1982). In realist terms, competitive pressures on profit margins were considered as the main causal power of interest here, with their effects being mediated by a wide range of contextual circumstances, such as the particular cost structures of the firms under pressure, the scope for technological change, and the availability of various kinds of labor. Thus, in an industry such as clothing, which had little scope for further automation or speeding up work, and in which wages were a major element of costs, a common response was to seek out cheaper sources of labor, sometimes by relocating.

By contrast, in industries which were more technology intensive and had scope for further automation, a more common response was technological innovation. By such means, it was possible to explain how the same or similar causes could produce radically different results according to context, whereas approaches which rested upon the discovery of empirical regularities foundered. Similar research also illustrated how a single kind of outcome—job losses— could have completely different causes, e.g., a drop in demand for the firm’s products, or the introduction of automation during steady or rising demand, or reorganization of work routines to eliminate idle time and intensify work (see also Henry 1992).

One of the most important and lasting influences of critical realism on geography has been on its empirical research designs (Sayer 1992). Hitherto, the dominance of conceptions of explanation as tied to generalizations about empirical regularities had been used to privilege ‘extensive’ research designs which sought to uncover such regularities or patterns of covariation. Although these are, as the term indicates, useful for establishing how extensive certain phenomena are, they do not necessarily identify causes, as opposed to spurious correlations. The latter are likely to be thrown up by extensive research because it preselects groups of objects taxonomically, that is, on the basis of similarities (e.g., industries of a certain kind), rather than functional connection. By contrast, intensive research designs, including ethnographic case studies, focus on causal rather than taxonomic groups and follow up causal connections wherever they lead— often between heterogeneous objects, such as firms of different kinds. Intensive research is typically more exploratory, since the purpose is to discover the relevant causal groups. It is often time consuming to do this for large numbers of cases, and therefore it is not always a good basis for generalization. However, as we have seen, this lack of generalizability does not undermine the status of causal explanations of the particular cases which are studied intensively. Consequently, having formerly been regarded as unscientific, intensive studies have now become more accepted in geographical research.

Critical realism acknowledges that the methodology of social science must diverge from that of the natural sciences because human, social phenomena are intrinsically meaningful. While the objects that, say, geologists study, are what they are regardless of any meaning people may give to them, objects such as socially segregated urban areas depend on the meanings actors give to them. To understand what these are, one has to know what their meaning is in society, whereas the nature of a rock does not depend on how its constituents understand one another, as they do not. Consequently, an important part of the methodology of human geography, no less than other social sciences, is interpretive understanding. For example, to understand the geography of the Middle East, we would have to understand what significance its land had for Jews and Arabs in terms of their identities.

Although this interpretive understanding has also been the focus of humanistic geography, critical realists make certain qualifications which the latter might not. The three most important are as follows. First, critical realists argue that although we have to interpret the understandings that actors have of their situations in order to make sense of how they act, their under-standing is not necessarily a good one, indeed it may be systematically flawed. Moreover, these misunderstandings may have important effects which actors cannot explain. Thus a mysogynist may believe that the proper spatial location for women is in the home, on the false grounds that they are incapable of coping with life in the public sphere; the falsity of this understanding makes a difference to what happens and therefore has to be acknowledged in scientific accounts. Further, different actors may give contradictory accounts of the same situation: where this is the case the researcher cannot sit on the fence and say both are correct. Actors’ accounts of situations there-fore face social scientists’ accounts both as objects of study and rivals. It is therefore likely that researchers’ accounts will be at least implicitly critical of actors’ accounts—a conclusion which humanistic geographers are generally unwilling to accept. It is primarily for this reason that Bhaskar’s philosophy of social science is called critical realism.

Second, among the intrinsically meaningful aspects of social phenomena are actors’ reasons. Critical realists argue that reasons can be causes, since they can be responsible for producing change, by prompting actions. This, of course, implies that realists do not limit causation to physical processes. To deny that reasons are causes is to imply that the giving of reasons by actors is ineffectual and redundant, which of course makes it unclear why they are given. Third, an insistence on the necessity of interpretive understanding should not allow us to forget that social life is also material, embodied, and spatially situated, and that things happen to people, so that, for example, they can be dominated, sometimes regardless of their under-standing of the situation.

The realist facility for explaining irregular processes was particularly attractive to a discipline such as geography, concerned as it is with spatial differentiation and uneven development, and hence unable to ignore—as, say, economics tries to do—the problems of applying equilibrium models or making generalizations which are supposed to hold across time and space. The systems which geography and other social sciences study are open, that is, their elements can themselves be undergoing internal qualitative change—actors aging and learning, for example—and the relations among the elements are not constant, with the result that any empirically regular behaviors are unlikely to be durable or widespread. Evolutionary change and disequilibrium rather than equilibrium are the norm.

Where previous attempts at scientific explanation in geography had tried to evade the prevalence of differentiation and irregularity, on the mistaken grounds that explanation required the discovery of regularity, human geography began to re-embrace differentiation, nowhere more strikingly than in Doreen Massey’s influential Spatial Divisions of Labour (Massey 1984). This was closely followed by a burgeoning literature on locality studies, which typically focussed on how economic restructuring worked out in different localities, and how those differences impacted back on the restructuring process (e.g., Cooke 1986).

5. Critical Reception

Some geographers feared that such concerns heralded an abandonment of theory and generalization and a slide towards the empiricist documentation of facts (e.g., Harvey and Scott 1989), indicating a resistance to the critical realist de-emphasis of generalization. These fears were not merely methodological but partly about a weakening attachment in economic geography to a particular kind of theory—Marxism—and also a tendency to treat critical realism as a source of substantive theory rather than as a philosophy, when it had no such pretensions. At the same time, some of the attempts to use critical realism in localities studies invited justified criticism by arbitrarily translating it into the study of general, supralocal causal processes interacting with local contingent conditions. However, critical realism did not license such an extraordinary alignment of scale and necessity: the supralocal is no more or less the realm of necessity than the local, nor is the local any more the realm of contingency than the supralocal.

Towards the end of the 1980s a newer generation of geographers became influenced by postmodernism. In some ways this continued the emphasis on differentiation—only phrased as ‘difference’ which had connotations of social differences of identity. Societies were seen as inherently messy, and skepticism towards grand universalizing accounts of the world grew. One of the positive features of this literature was that it gave greater concern to the textual nature of accounts of the world (Sayer 1989, Stones 1996). At the same time, some postmodernist writers rekindled relativist, anti-realist tendencies (e.g., Barnes 1996, Graham 1990), although sometimes by mischaracterizing realism as claiming privileged access to the truth rather than as having a fallibilist view of knowledge. The debate continues (e.g., Peet 1998).

Bibliography:

1. Barnes T J 1996 Logics of Dislocation. Guilford Press, New York
2. Bhaskar R 1978 A Realist Theory of Science. Harvester Press, Hassocks, Sussex
3. Bhaskar R 1989 The Possibility of Naturalism, 2nd edn. Harvester Wheatsheaf, Hemel Hempstead, UK
4. Cooke P 1986 Localities: the Changing Face of Urban Britain. Unwin Hyman, London
5. Graham J 1990 Theory and essentialism in Marxist geography. Antipode 22: 53–66
6. Harvey D, Scott A J 1989 The practice of human geography: theory and empirical specificity in the transition from Fordism to flexible accumulation. In: Macmillan B (ed.) Remodelling Geography. Blackwell, Oxford, UK, pp. 217–29
7. Henry N 1992 The new industrial spaces: locational logic of a new production era. International Journal of Urban and Regional Research 16: 375–96
8. Massey D 1984 Spatial Divisions of Labour. Macmillan, London
9. Massey D, Meegan R 1982 Anatomy of Job Loss. Methuen, London
10. Peet R 1998 Modern Geographic Thought. Blackwell, Oxford, UK
11. Pratt A C 1994 Uneven Re-production: Industry, Space and Society, 1st edn. Pergamon, Oxford, UK
12. Sayer A 1982 Explanation in economic geography. Progress in Human Geography 6: 68–88
13. Sayer A 1985 Realism and geography. In: Johnston R J (ed.) The Future of Geography. Methuen, London, pp. 159–73
14. Sayer A 1989 The ‘new’ regional geography and problems of narrative. Environment and Planning D: Society and Space 7: 253–76
15. Sayer A 1992 Method in Social Science: a Realist Approach, 2nd edn. Routledge, London
16. Stones R 1996 Sociological Reasoning: Towards a Post-modern Sociology. St Martin’s Press, New York