MEANING IN ACTION

All political scientists offer us interpretations. Interpretive approaches differ in offering interpretations of interpretations.² They concentrate on meanings, beliefs, and discourses, as opposed to laws and rules, correlations between social categories, or deductive models. Of course, the distinction between interpretive approaches and others is fuzzy. After all, laws, social categories, and models are, as proponents of an interpretive approach would point out, matters of belief or language. Sensible institutionalists, behaviouralists and rational choice theorists recognise that typologies, correlations, and models do explanatory work only when unpacked in terms of the beliefs and desires of the actors. Nonetheless, there is a family of interpretive approaches to political science that stand out in their focus on meanings. This family includes at least decentred theory, poststructuralism, and social constructivism (see, for example, Bevir and Rhodes, 2003; Burchell et al, 1991; Berger and Luckman, 1971). It overlaps with other approaches, including those strands of the new institutionalism concerned with the impact of ideas (see, for example, Berman, 2001; Hay, 2000; Lieberman, 2002; Finnemore and Sikkink, 2001).

An interpretive approach is not alone in paying attention to meanings. It is distinctive because of the extent to which it privileges meanings as ways to grasp actions. Its proponents privilege meanings because they hold, first, beliefs have a constitutive relationship to actions and, second, beliefs are inherently holistic (cf. Taylor, 1971).

First, an interpretive approach holds that beliefs and practices are constitutive

'An interpretive approach is ... distinctive because of the extent to which it privileges meanings as ways to grasp actions.'

of each other. When other political scientists study voting behaviour using attitude surveys or models of rational action, they separate beliefs from actions to find a correlation or deductive link between the two. In contrast, an interpretive approach suggests such surveys and models cannot tell us why, say, raising one's hand should amount to voting, or why there would be uproar if someone forced someone else to raise their hand against their will. We can explain such behaviour only if we appeal to the inter-subjective beliefs that underpin the practice. We need to know that voting is associated with free choice and so with a particular concept of the self. Practices could not exist if people did not have the appropriate beliefs. Beliefs or meanings would not make sense without the practices to which they refer.

Second, an interpretive approach argues that meanings or beliefs are holistic (on holism see Fodor and LePore, 1992). We can make sense of someone's beliefs only by locating them in the wider web of other beliefs that provide the reasons for their holding them. So, even if political scientists found a correlation between a positive attitude to social justice and voting Labour, they could not properly explain people's voting Labour by reference to this attitude. After all, people who have a positive attitude to social justice might vote Conservative if, say, they believe Labour will not implement policies promoting social justice. To explain why someone with a positive attitude to social justice votes Labour, we have to unpack the other relevant beliefs that link the attitude to the vote. To explain an action, we cannot merely correlate it with an isolated attitude. Rather, we must interpret it as part of a web of beliefs.

Many political scientists typically treat beliefs, meanings, ideas, and norms as if they can be differentiated from actions and related individually to actions. In contrast, an interpretive approach holds that meanings or beliefs form webs that are constitutive of actions and practices. This philosophical analysis of meaning in action informs other aspects of an interpretive approach, including bottom-up modes of inquiry and critiques that expose unquestioned assumptions and inconsistencies.

Proponents of an interpretive approach incline to bottom-up forms of social inquiry. They usually believe that people in the same situation can hold different beliefs because their experiences of that situation can be laden with different prior theories. No abstract concept, such as a class or institution, can explain people's beliefs, interests, or actions. Such a concept can represent only an abstract proxy for the multiple, complex beliefs and actions of all the individuals we classify under it. So, for these reasons, practices need bottom-up studies of the actions and beliefs out of which they emerge (see, for example, Bang and Sørensen, 1999). An interpretive approach explores the ways in which social practices are created, sustained, and transformed through the interplay and contest of the beliefs embedded in human activity.

Another shared interpretive theme is an emphasis on the contingency of political life. Typically, an interpretive approach holds that people in any given situation can interpret that situation and their interests in many ways. So, political scientists must allow that no practice or norm can fix the ways its participants will act, let alone how its participants will innovate in new circumstances. An interpretive approach thus concludes that our practices are radically contingent. Our practices lack a fixed essence or given path of development. An emphasis on contingency explains why an interpretive approach often questions alternative theories. Its proponents believe political scientists efface the contingency of social life when they attempt to ground their theories in apparently given facts about the nature of reasoning, the path dependence of institutions, or the inexorability of social developments. They thus try to expose the contingency of those facets of political life that other political scientists mistakenly represent as natural or inexorable (see, for example, Kass and Catron, 1990).

SITUATED AGENCY

Interpretivism itself consists of a diverse cluster of traditions. There are important differences among its several proponents, in particular over aggregating practices. Proponents of an interpretive approach can seem confused about the nature of the meanings that inform practices. Poststructuralists sometimes imply that meanings exist as quasi-structures governed by a semiotic code or random fluctuations of power.³ Others analyse meanings as the beliefs of individuals; they take ideology, discourse, or language to refer only to a cluster of inter-subjective beliefs.

When poststructuralists imply that meanings stem from quasi-structures, they usually do so because they want to stress how beliefs and subjectivity are constructed out of social backgrounds. They want to reject a strong notion of autonomy. However, we can distinguish between autonomy and agency. Autonomous individuals can, at least in principle, have experiences, reason, adopt beliefs, and act, outside all contexts. Agents can reason and act in novel ways but they can do so only in the context of a discourse or tradition. Most poststructuralists reject autonomy because they believe all experiences and reasoning embody theories. Thus, people can adopt beliefs only against the background of a prior set of theories, which at least initially must be made available to them by a discourse or tradition. However, a rejection of autonomy does not entail a rejection of agency. We can accept that people always start with a discourse or tradition and still see them as agents who can act and reason in novel ways to vary this background. Proponents of an interpretive approach have no reason to throw agency out with autonomy. When they defend a capacity for agency, however, they might recognise that it always occurs in a social context that influences it. Agency is not autonomous – it is situated.

The notion of situated agency resolves confusion among proponents of an interpretive approach about aggregating studies of practices.⁴ At the moment, poststructuralists sometimes rely on concepts such as discourse to aggregate their accounts of practices. These concepts appear to treat meanings as if fixed by quasi-structures. The idea that quasi-structures fix meanings surely falls foul, however, of the poststructuralists' own emphasis on contingency and particularity. The greater the stress we place on the contingency and particularity of beliefs, actions and practices, the harder it is to explain them with aggregate concepts. Indeed, if an interpretive approach relies on discourse to do explanatory work, this concept can suggest a worrying neglect of agency. If a discourse claims to explain patterns of belief or speech, the implication is that the discourse fixes the content of the beliefs or intentions people hold. What is more, if poststructuralists use discourse as an explanatory concept, they adopt a determinism that cannot account for change. If individuals arrive at beliefs by a fixed and disembodied ideology, they lack the capacity to change that ideology. So any such changes will seem inexplicable.

'If a discourse claims to explain patterns of belief or speech, the implication is that the discourse fixes the content of the beliefs or intentions people hold.'

Of course, poststructuralists often criticise structuralism for displaying just such determinism, while arguing that they themselves view such transformations as instabilities inherent in structures. Alas, however, this claim merely elides the question of whether we are to understand instabilities, contradictions and transformations as necessary qualities of a disembodied discourse or as contingent properties and products of individual subjects, their beliefs, and their actions.

An interpretive approach often struggles to aggregate accounts of practices in ways that have explanatory power. The problem can be resolved by the idea of situated agency (Bevir, 1999: chapters 5 and 6). To reject autonomy is to accept that traditions and discourses influence individuals. Explanatory concepts must suggest, therefore, how social influences permeate beliefs and actions even when actors do not recognise such influence. To accept agency is, however, to imply that people have the capacity to adopt beliefs and actions, even novel ones, for reasons of their own. In so doing they can transform the social background. The idea of tradition covers both inheriting beliefs and transforming them as they are handed down from generation to generation. It is evocative of a social structure in which individuals are born, which then acts as the background to their beliefs and actions even while they might adapt, develop, and reject much of this inheritance. Similarly, an interpretive approach could usefully explore change by focusing on dilemmas. Change arises as situated agents respond to novel ideas or problems. It is a result of people's ability to adopt beliefs and perform actions through a reasoning that is embedded in the tradition they inherit.

INTERPRETATION AND COMMON SENSE

An interpretive approach rests, first, on a philosophical analysis of meaning in action. An analysis of the constitutive relation of meanings to actions implies that we can grasp actions properly only by examining the beliefs embodied in them. It prompts us to offer interpretations of interpretations. An interpretive approach rests, second, on a philosophical analysis of the holistic nature of meanings. An analysis of meanings as holistic, rather than tied individually to referents, implies we can grasp beliefs properly only as part of the wider webs of which they are part. We have suggested that it prompts us to explain beliefs by reference to webs of belief, traditions, and dilemmas.

One criticism of an interpretive approach is to say that it is mere common sense. Indeed, in a sense, interpretivism is common sense. It derives from a philosophical analysis of the theories that make up our everyday way of discussing actions. Wittgenstein (1972: 109) argued that 'philosophy is a battle against the bewitchment of our intelligence by means of language'. Similarly, we use philosophical analysis to dispel the bewitching effects of other ways of discussing actions. Such analysis returns us to our everyday concepts to challenge positivist attempts to discuss actions as if they were akin to the physical phenomena studied by natural scientists. It undermines claims there is a superior scientific language.

The bewitching effects of allegedly scientific languages have led critics to reject an interpretive approach by contrasting its others. Sometimes critics wrongly identify an interpretive approach with a particular object of inquiry rather than a philosophical analysis of meaning in action. They set up dichotomies between those objects allegedly studied by proponents of an interpretive approach and those studied by other political scientists. They contrast interpretation with several others. Thus, they suggest interpretation focuses on meanings not practices, beliefs not rhetoric, or discourse not power. Alternatively, critics wrongly equate an interpretive approach with a particular mode of inquiry rather than a philosophical analysis of meaning in action. They set up dichotomies between interpretive modes of inquiry and those adopted by other political scientists. Once again, they contrast interpretation with spurious others; interpretation is about understanding not explanation, elucidation not critique, or empathy not rigour. We consider these several misconceptions in some detail.

PRACTICES

One common misconception about an interpretive approach is that it concerns only beliefs or discourses, not actions or practices. This misconception implies that an interpretive approach might be a reasonable way of recovering the froth of political ideas but that it does not help us to understand the real word lurking underneath such froth. This misconception only makes sense, however, if we draw a false dichotomy between beliefs and actions. If beliefs and actions were unrelated to each other, it might make sense to suggest we could recover one without exploring the other. In contrast, an interpretive approach rests on the claim that beliefs are constitutive of actions. Interpretivism implies we cannot properly understand actions except by recovering the beliefs that animate them. Far from neglecting practices, an interpretive approach typically explores meanings or beliefs precisely to grasp better the practices that embody them.

Critics might still suggest that ideas such as belief, tradition, and dilemma are too abstract. They ignore the way meanings are always embedded in habits and social interactions. However, we introduced the notion of tradition precisely to capture the embedded nature of individuals and their beliefs. What is more, although tradition refers mainly to beliefs, these beliefs need not be especially conscious or rational.⁵ An interpretive approach allows that beliefs and traditions do not exist as disembodied ideologies or discourses but become concrete in actions and practices. It suggests we can ascribe beliefs to people, including ourselves, only by interpreting actions, including, of course, speech-acts.

Although an interpretive approach explores practices by unpacking the relevant beliefs, it does perhaps conceive of practices in a different way from other political scientists. The difference appears in the way other political scientists often prefer to see practices as institutions (cf. March and Olsen, 1989). One difference arises over what it means to say that practices or institutions are concrete social realities. Proponents of an interpretive approach rarely see practices as natural or discrete chunks of social reality. Practices do not have boundaries that make them discrete entities. They do not have natural or given limits by which we might separate them out from the general flux of human life. For example, the boundary of a political party does not clearly lie with those who attend weekly committee meetings, those who attend once a year for the annual general meeting, those who go to fund-raising events organised by the party, or those who participate in direct action over a political grievance. For a researcher using an interpretive approach, the limits of a practice are decided pragmatically, justified by the purposes of their inquiry. Practices are concrete social realities, but they are not natural kinds. So it is political scientists as observers who separate particular practices, and they do so to suit their research purposes.

Perhaps proponents of an interpretive approach also differ from other political scientists in their analysis of conventions, shared understandings, or interactions in practices or institutions. Although practices display conventions, this does not mean conventions constitute the practices. No doubt many participants often seek to conform to the conventions of a practice. Even so, first, they do not always do so, and, second, even when they do, they might misunderstand the conventions. So conventions cannot be constitutive of practices. The situated agency of participants constitutes practices, and such agency is creative, not fixed by rules. Individuals are situated agents who necessarily interpret the conventions that characterise the practices in which they are engaged, and who can vary the conventions. This appeal to situated agency does not imply that all people are heroic individuals who have great impact on the historical direction of a practice. It implies only that they have the capacity to adapt their inheritance and act in novel ways. When they do, they are unlikely significantly to alter a practice unless others also adjust their beliefs and actions in a related fashion. Even then, the changes in the practice are unlikely to correspond to any they might have intended. Practices rarely, if ever, depend directly on the actions of any given individual. They do consist of nothing but the changing actions of various individuals.

STRUCTURES

For many political scientists, this analysis of practices contrasts sharply with approaches that rely on concepts of social structure. At issue here is how political scientists should think about the nature of social contexts and their impact on people. We have emphasised situated agency, arguing that traditions only influence but do not define the beliefs individuals come to adopt and the actions they attempt to perform. We have also emphasised that traditions are not natural, arguing that observers construct them out of an undifferentiated context to explain whatever interests them. Critical realists worry that these emphases neglect the influence and the constraining effect cultural schemes or structures exercise on people (see, for example, McAnulla, 2004; Reckwitz, 2002).

An interpretive approach might allow for the influence and the constraining effects of social contexts. It just will refuse to reify practices or traditions by treating them as structures or cultural schemes. To begin, although an interpretive approach might defend the capacity of the individual for situated agency, commonly it rejects the idea of autonomy. We have suggested that people only ever come to hold beliefs or perform actions against the background of a tradition that influences them. Appeals to traditions go a long way to explaining why individuals hold the beliefs they do and act in the ways they do. In addition, although proponents of an interpretive approach argue that tradition does not constrain beliefs, they recognise that practices can have a limiting effect on actions. Individuals are situated agents in that they have a creative ability to adopt beliefs or attempt actions for reasons of their own. However, they do not necessarily succeed in the actions they attempt. The results of their actions typically depend on how others act. Practices thus constrain the actions people can successfully make.

An interpretive approach can allow that traditions influence people, and that practices constrain the actions people can perform successfully. Where proponents of an interpretive approach still might differ from critical realists is in the logical content they attribute to such concepts. So, we might prefer the terms 'tradition' and 'practice' to 'cultural scheme' and 'structure' precisely because we might worry that the latter two neglect situated agency and reify social contexts. The term 'tradition' captures an analysis of individuals who inherit a set of beliefs that forms the background to their later reasoning. It inevitably influences them even though they might transform it over time through their local reasoning. On the other hand, the term 'cultural scheme' suggests a disembodied structure of ideas that sets clear limits to the beliefs and agency of individuals by fixing the ways they experience the world.⁶ Similarly, the term 'practice' captures an analysis of how social contexts constrain actions. Practices constrain the actions people attempt to perform if they enter the subjective reasoning of the actors. Practices also constrain the effectiveness of actions because they consist of the actions of others: politicians might try to lower inflation only to find that the actions of business organisations and citizens prevent them. While an interpretive approach can allow practices act as constraints, it does so in ways that make practices reducible to the contingent actions of other individuals. In contrast, the term 'structure' invokes a physical object that constrains people in its own right, rather as the Atlantic Ocean stops us driving back and forwards between London and New York.

EXPLANATION

Another related misconception is that interpretive approaches aim only to understand actions and practices, not to explain them. The dichotomy between understanding and explanation again makes sense only if we falsely separate actions from beliefs. An interpretive approach rests on a philosophical analysis of actions as constituted by beliefs. This analysis implies that other political scientists go awry when they try to explain actions in ways that do not appeal to beliefs. Any satisfactory explanation of actions or practices must refer to the beliefs that animate them. To understand the relevant beliefs is to explain the action or practice. What is more, when proponents of an interpretive approach argue that beliefs are inherently holistic, they imply that we can explain them by locating them as part of the web of meanings or beliefs that give them their character. To locate beliefs in webs of belief, and to locate webs of belief against the background of traditions and dilemmas, is to explain those beliefs and the actions and practices they inspire.

The philosophical analysis of meaning in action that informs an interpretive approach suggests, however, that human sciences rely on a distinctive form of explanation, which we describe as narrative (see Bevir, 1999; Roe, 1994). When we explain actions by beliefs and desires, we rely on a concept of choice and on criteria of reasonableness that have no place in natural science (see Davidson, 1980). So, the natural and human sciences use different concepts of causation. This difference does not mean the human sciences have no interest in causal analysis. To the contrary, the human sciences explain actions and practices in narratives that point to the beliefs and desires that cause the actions.

Narratives distinguish an interpretive approach from those approaches that treat meanings or beliefs merely as 'ideational variables' alongside other factors (as in, for example, Gerring, 1999; Wendt, 1999). An interpretive approach suggests other variables do explanatory work only if they are unpacked as beliefs. Equally, its proponents argue that it is a mistake to ask how they would specify the precise

'Any satisfactory explanation of actions or practices must refer to the beliefs that animate them.'

links between independent variables. Critics might say that actions and beliefs, or beliefs and traditions, cannot be identified independently as they should be in explanations. All proponents of an interpretive approach offer, they might conclude, are re-descriptions. However, an interpretive approach rests on a philosophical analysis of meaning in action that invalidates the methodological rigour – the specification of independent variables – that prompts the criticism. This philosophical analysis implies actions are intentional, which means they are necessarily performed for reasons or beliefs. Similarly, this philosophical analysis implies that people are not autonomous, so they necessarily reach beliefs against the background of tradition. These philosophical arguments provide the causal mechanisms at work in a narrative. They indicate that actions and beliefs, and beliefs and traditions, are entwined. Thus, when political scientists try to specify them independently of one another, they are misled by a spurious concept of scientific rigour into adopting a form of explanation that is inappropriate for political science.

METHOD

Yet another misconception equates an interpretive approach with certain techniques of data generation (and on the misleading distinction between 'qualitative' and 'quantitative' methods, see Schwartz-Shea and Yanow, 2002). An interpretive approach is said to be limited to textual readings and small-scale observations, excluding survey research and quantitative studies. However, a concern to offer interpretations of interpretations does not necessarily favour particular methods. To the contrary, proponents of an interpretive approach might construct their interpretations using data generated by various techniques. They can draw on participant observation, interviews, questionnaires, mass surveys, statistical analysis, and formal models as well as reading memoirs, newspapers, and official and unofficial documents. The philosophical analysis underpinning an interpretive approach does not prescribe a particular methodological toolkit for producing data. Instead, it prescribes a particular way of treating data of any type. Proponents of an interpretive approach argue that political scientists should treat data in ways consistent with the task of interpreting interpretations. They should treat data as evidence of the meanings or beliefs embedded in actions. Political scientists should not try to bypass meanings or beliefs by reducing them to principles of rationality, fixed norms, or social categories.

The interpretive view of how we should treat data does, of course, have some implications for methods of data collection. It leads, in particular, to greater emphasis on qualitative methods than is usual among political scientists. Suppose that the data provided by models, formal constitutions, or large-scale surveys leads us to assign certain beliefs to a group of people. Since such data typically abstracts from individual circumstances to find patterns, it elides differences between people, lumping together individuals who act in broadly similar ways for different reasons. Therefore, an interpretive approach often favours more detailed studies of the beliefs of the relevant people using textual analysis, participant observation and interviews. Much present-day political science prefers the latter 'scientific' techniques and ignores,

'People always construct the content of their experiences through the prior theories they bring to bear on them.'

or even denigrates the other methods. In contrast, an interpretive approach does not require an exclusive use of any one method. However, it does redress the balance to the qualitative analysis more often associated with anthropology and history than with political science.

RHETORIC

The case for anthropological and historical studies should not be confused with the claim that political scientists must accept actors' own accounts of their beliefs. Obviously people's statements about what they believe offer significant evidence about what they believe. Equally, however, people can be deliberately misleading. Admittedly, people do act sometimes on political commitments they have agonised over. However, they also act on habitual, unreflective beliefs about the nature of the world and about what is right in a given context. So, we might explain an action using beliefs other than the stated beliefs of the actors.

Another misconception is, therefore, that interpretive approaches cannot deal adequately with rhetoric (see Dowding, 2004). We can explore rhetoric using forms of explanation based on the concepts of belief, tradition, and dilemma. When people use a rhetorical pattern, they do so because they believe it will get a suitable response to their ideas. So a political scientist can explain people's choice of rhetoric by identifying their relevant beliefs and preferences about different patterns of rhetoric, their appropriateness, and their probable effectiveness. This analysis will involve placing people's beliefs about rhetoric in their wider webs of belief before relating these wider webs of belief to traditions and dilemmas.

Critics worry that if we are to invoke beliefs other than those stated by the actors, we need criteria for identifying beliefs (see Brown, 2002). They worry that an interpretive approach guesses people's beliefs rather than finding hard evidence of them. Proponents of an interpretive approach might reply that all experiences, not just experiences of others' beliefs, are guesses in that they are theory-laden. People always construct the content of their experiences through the prior theories they bring to bear on them. All too often, however, this insistence on the constructed nature of experience leads critics of an interpretive approach to assimilate it to a postmodern denial of any object outside the 'text'. Most supporters of an interpretive approach would deny entrapment in texts. For instance, we would propose using philosophical reasoning to defend a commitment to the existence of general classes of objects, including beliefs. We would then use inference to the best explanation to defend a commitment to the existence of a particular case of such objects.

Whenever we act, we commit ourselves to certain concepts. For example, if we use a pen to fill in our tax form, take it to the tax office, and pay by cheque, we commit ourselves to beliefs about the existence of certain objects, such as forms and money. We also commit ourselves to beliefs about the nature of these objects – for example, that paying tax avoids interest and even fines for late or non-payment, and that others accept authorised cheques as discharging our liabilities. Finally, we often commit ourselves to beliefs about ourselves – for example, that we can attempt to pay, or not to pay, taxes. Philosophy can go to work on the concepts we thus commit ourselves to in our actions. It can analyse the implications of these concepts to provide an account of the classes of objects with which we populate the world and the forms of reasoning suitable for such objects. For example, our acceptance of tax forms and use of pens suggests we populate the world with physical objects. Our convictions about the utility of money suggest we populate the world with objects that gain significance through inter-subjective beliefs. Our convictions about our ability to act for reasons of our own suggest we populate the world with beliefs.

While philosophical reflection on the ideas embedded in our actions provides us with good reasons for proposing the existence of beliefs, it cannot justify ascribing particular beliefs in any particular case. Nonetheless, an interpretive approach can justify attributing particular beliefs to people by claiming that doing so best explains facts on which we agree. Although political scientists do not have direct access to people's beliefs, they can justify ascribing beliefs to people by saying that doing so best explains the evidence on which we agree.

POWER

Poststructuralists sometimes imply that other interpretive approaches are insensitive to the ways in which relations of power constitute individuals including their beliefs. However, the concept of tradition can do much the same work as does the poststructuralist one of power. Tradition asserts that individuals, far from being autonomous, always come into being in a social context, which influences the beliefs they come to hold. People inherit concepts, values, and practices. They can reflect on this inheritance and even modify it, but they can do so only in the context of other beliefs they adopt against the background of a social inheritance. So, if 'power' is the influence society inevitably exerts on individuals, then a concept of tradition similarly covers the effects of society. We prefer the concept of tradition for two reasons.

First, if we use the term 'power' here, we deprive it of explanatory and critical force. If power is everywhere, to point to its presence in any given case fails to provide any critical or explanatory leverage. Second, the notion of tradition emphasises a commitment to situated agency. Appeals to power as constitutive of subjectivity can seem to deny the agency of the subject. We are unsure whether particular poststructuralists oppose agency as well as autonomy. If they do, we would argue that it is a mistake to conceive of traditions as reified quasi-structures that somehow determine the beliefs people can come to hold.

Tradition need not be conceived of as uniform. Rather, we can disaggregate it into conflicting strands. Nor need we think it is ever natural. Rather, we can seek to question the unquestioned and show how any tradition arises as a contingent product of struggles over different ways of conceiving of and responding to constructed dilemmas. These political conflicts and contests are not confined to government. Rather, we might use the word governance to stress that such contests take place throughout society.

If we so conceive of tradition, then our narratives often will be critiques. Our narratives often will unmask the partiality of a political interpretation by showing how it arose against the background of a particular tradition. And our narratives often will unmask the contingency of traditions by showing them to be just one among several historical possibilities. They might seek to reveal the contingency and contestability of narratives that present themselves as natural and fixed.

POLICY ADVICE

Yet another misconception about an interpretive approach is that it cannot produce policy-relevant knowledge. Critics suggest policy-relevant knowledge comes from prediction based on models or correlations between independent variables. Before addressing this misconception directly, we need to confront the notion that scientific expertise and prediction are the correct way of thinking about the advice political scientists might offer practitioners. An interpretive approach typically rejects the possibility of prediction – as opposed to the looser idea of informed conjecture – since it is incompatible with the narrative form of explanation. Its proponents usually portray change as a product of the ways in which people change inherited traditions and practices, and the ways in which they adapt them are open-ended and so not amenable to prediction.

Since traditions, and practices do not fix the ways people might develop them when confronted with new circumstances, we cannot know in advance how people will develop their beliefs and actions in response to a dilemma. Therefore, political scientists cannot predict how people will respond to a dilemma. Whatever limits they built into their predictions, people could always arrive at new beliefs and actions outside those limits. Political scientists cannot predict. However, they can offer informed conjectures that seek to explain practices and actions by pointing to the conditional connections between actions, beliefs, traditions and dilemmas. Their conjectures are stories, understood as provisional narratives about possible futures.

At this point we can directly address the issue of how an interpretive approach contributes to policy advice. Most policy-oriented work on governance seeks to improve the ability of the state to manage the markets, bureaucracies and networks that have flourished since the 1980s. Typically, this work treats hierarchies, markets, and networks as fixed structures that governments can manipulate if they use the right tools. An interpretive approach undercuts this idea of a set of tools that we can use to manage governance. As governance is constructed differently, contingently, and continuously, we cannot have tool kits with which to manage it. So an interpretive approach encourages us to foreswear management techniques and strategies. Crucially, it replaces such tools with learning by telling stories and listening to them.⁷

Other commentators have traced the rediscovery of storytelling in the subfield of public administration (Van Eeten et al, 1996). They sometimes distinguish between storytelling by administrators and storytelling by scholars to make the important point that this intellectual fashion has its feet firmly on the ground. In both public and private organisations managers use stories not only to gain and pass on information and to inspire involvement but also as the repository of the organisation's institutional memory. Rein (1973: 266) points out that the central thread in a policy narrative is metaphor, which makes the unfamiliar analogous to familiar situations: 'The simplest stories are proverbs and parables, used to justify policy-relevant stories'.

While statistics, models, and claims to expertise all have a place in such stories, we should not become too preoccupied with them. We should recognise that they too are narratives about how people have acted or will react given their beliefs and desires. No matter what rigour or expertise we bring to bear, all we can do is tell a story and judge what the future might bring.

CONCLUSION

When critics contrast an interpretive approach with others, they are often groping

'Critics of interpretivism rarely avow positivism. Surely, though, their idols of hard data, experimental tests, and methodological rigour lose all allure once one renounces a positivist faith in pure experience?'

for a way of expressing their sense that an interpretive approach lacks rigour. They invoke the same basic dichotomy. They want to dismiss interpretation as fuzzy, subjective, and impressionistic. They want to defend a political science that relies on hard data, experimental testing, and methodological rigour. In this article, we have challenged this dichotomy by giving details of the data, methods and even epistemology associated with an interpretive approach. More importantly, we have given reasons to renounce the false idols of hard data, experimental tests, and rigorous methods.⁸

Critics of interpretivism rarely avow positivism. Surely, though, their idols of hard data, experimental tests, and methodological rigour lose all allure once one renounces a positivist faith in pure experience? If we cannot have pure experiences, all data are soft because they presuppose prior theories that are themselves contestable. If all data are soft, we cannot evaluate particular narratives or theories using experiments. All knowledge arises, rather, from comparisons between rival theories or narratives that are based on at least partly constructed facts. Also, we can challenge the idol of methodological rigour. Often methodological rigour is held up as a way of producing secure facts that others can replicate and accept. In contrast, we might suggest that methods and the facts they construct should be evaluated together as parts of larger narratives or theories. We will accept methods as 'rigorous' – or to use a more accurate term, 'appropriate' – only if we adopt philosophical theories that imply that the relevant methods are suitable for the objects to which they are applied. Judgements about methodological rigour or appropriateness always depend on logically prior judgements about philosophical rigour or appropriateness.

The idol of methodological rigour typically acts to obscure prior philosophical issues or even to prejudge such issues to support positivism. An interpretive approach, in contrast, gives primary importance to philosophical rigour. It highlights the importance of political science meeting the logical requirements of our concepts. It rejects the stress on methodological rigour as a bewitching effect of the positivist philosophy of the natural sciences.

Notes

¹ A longer version of this paper first appeared in the Australian Journal of Political Science. Earlier versions were presented at the American Political Science Association Annual Conference, Chicago, 1–3 September 2004; and the Australasian Political Studies Association, University of Adelaide, 29 September–1 October 2004.

² Although this paper concentrates on political science, interpretive approaches are widespread across the human sciences. Useful collections include Rabinow and Sullivan (1979), Rabinow and Sullivan (1987) and Scott and Keates (2001). Much of the movement charted by these collections derives from the philosophical repudiation of positivism in the 1960s and 1970s. See Bernstein (1976) and Fay (1975).

³ See Foucault (1972, 1980). For varied assessments of the continuing impact of structuralism upon poststructuralism see Dreyfus and Rabinow (1982), Gutting (1989) and Harland (1988).

⁴ When we follow the logic of disaggregating concepts like voting or policy network, we end up with micro-level stories of individual actions based on one person's set of beliefs. Although such stories are interesting as cases, there are times when we want to tell more general stories, for example about governance. To do so, we need aggregate concepts like tradition and dilemma.

⁵ That said, we could make sense of someone's beliefs only by postulating them as a web that exhibits some kind of consistency and rationality. For discussion of various principles of charity according to which we do thus ascribe some kind of conceptual priority to rational beliefs see Bevir (1999: 158–171), Davidson (1984b) and McGinn (1977).

⁶ We would draw attention, more generally, to the difficulties that confront any dualism of 'scheme' and 'content', or 'paradigm' and 'experience', given the implausibility of an uninterpreted reality (see Davidson, 1984a). Such difficulties affect even those who emphasise meanings only to conceive of them as schemes, paradigms, or frames, including, for example, Rein and Schon (1995).

⁷ There is an extensive literature that explicitly applies an interpretive approach to policy analysis. Examples include Healy (1986), Hummel (1991), Jennings (1987), Van Eeten et al (1996), Weick (1995) and Yanow (1999).

⁸ For details of our own preferred epistemology see Bevir (1999: 78–126).

References

1. Bang, H. and Sørensen, E. (1999) 'The everyday maker: a new challenge to democratic governance', Administrative Theory and Praxis 21: 225–241.
2. Berger, P. and Luckman, T. (1971) The Social Construction of Reality: A Treatise in the Sociology of Knowledge, Harmondsworth: Penguin.
3. Berman, S. (2001) 'Ideas, norms and culture in political analysis', Comparative Politics 33: 231–250.
4. Bernstein, R. (1976) The Restructuring of Social and Political Theory, Philadelphia: University of Pennsylvania Press.
5. Bevir, M. (1999) The Logic of the History of Ideas, Cambridge: Cambridge University Press.
6. Bevir, M. and Rhodes, R.A.W. (2003) Interpreting British Governance, London: Routledge.
7. Brown, V. (2002) 'On some problems with weak intentionalism for intellectual history', History and Theory 41: 198–208. | Article |
8. Burchell, G., Gordon, C. and Miller, P. (eds.) (1991) The Foucault Effect: Studies in Governmentality, London: Harvester Wheatsheaf.
9. Davidson, D. (1980) Essays on Actions and Events, Oxford: Clarendon Press.
10. Davidson, D. (1984a) 'On the Very Idea of a Conceptual Scheme', in Inquiries into Truth and Interpretation, Oxford: Clarendon Press, pp. 183–199.
11. Davidson, D. (1984b) 'Radical Interpretation', in Inquiries into Truth and Interpretation, Oxford: Clarendon Press, pp. 125–141.
12. Dowding, K. (2004) 'Interpretation, truth and investigation', British Journal of Politics and International Relations 6: 136–142.
13. Dreyfus, H. and Rabinow, P. (1982) Michel Foucault: Beyond Structuralism and Hermeneutics, Chicago: Chicago University Press.
14. Fay, B. (1975) Social Theory and Political Practice, London: George Allen and Unwin.
15. Finnemore, M. and Sikkink, K. (2001) 'Taking stock: the constructivist research program in international relations and comparative politics', Annual Review of Political Science 4: 391–416. | Article |
16. Fodor, J. and LePore, E. (1992) Holism: A Shopper's Guide, Oxford: Blackwell.
17. Foucault, M. (1972) The Archaeology of Knowledge, translated by A Sheridan-Smith, London: Tavistock.
18. Foucault, M. (1980) C. Gordon (ed.) Power/Knowledge: Selected Interviews and Other Writings, 1972–1977, Brighton: Harvester.
19. Gerring, J. (1999) 'Does ideology matter? a roll-call analysis of key congressional votes, 1833–1992', Journal of Policy History 11: 399–442.
20. Gutting, G. (1989) Michel Foucault's Archaeology of Scientific Reason, Cambridge: Cambridge University Press.
21. Harland, R. (1988) Superstructuralism: The Philosophy of Structuralism and Poststructuralism, London: Routledge.
22. Hay, C. (2000) Political Analysis, Houndmills, Basingstoke: Palgrave-Macmillan.
23. Healy, P. (1986) 'Interpretive policy inquiry: a response to the limitations of the received view', Policy Sciences 19: 381–396. | Article |
24. Hummel, R.P. (1991) 'Stories managers tell: why they are as valid as science', Public Administration Review 51: 31–41.
25. Jennings, B. (1987) 'Interpretation and the practice of policy analysis', in F. Fischer and J. Forester (eds.) Confronting Values in Policy Analysis, Newbury Park, CA: Sage, pp. 128–152.
26. Kass, H. and Catron, B. (eds.) (1990) Images and Identities in Public Administration, London: Sage.
27. Lieberman, R. (2002) 'Ideas, institutions, and political order: explaining political change', American Political Science Review 96: 697–712.
28. McAnulla, S. (2004) 'Paving the rhodes to post-positivism? a critique of the new interpretive approach to British politics – towards a realist alternative', paper presented to the 54th Annual Conference of the UK Political Studies Association, University of Lincoln, 6–8 April.
29. March, J.G. and Olsen, J.P. (1989) Rediscovering Institutions: The Organizational Basis of Politics, New York: Free Press.
30. McGinn, C. (1977) 'Charity, interpretation, and belief', Journal of Philosophy 74: 521–535.
31. Rabinow, P. and Sullivan, W. (eds.) (1979) Interpretive Social Science, Berkeley: University of California Press.
32. Rabinow, P. and Sullivan, W. (eds.) (1987) Interpretive Social Science: A Second Look, Berkeley: University of California Press.
33. Reckwitz, A. (2002) 'The constraining power of cultural schemes and the liberal model of beliefs', History of Human Sciences 15: 115–124. | Article |
34. Rein, M. (1973) Social Science and Public Policy, Harmondsworth: Penguin Books.
35. Rein, M. and Schon, D. (1995) Frame Reflection, New York: Basic Books.
36. Roe, E. (1994) Narrative Policy Analysis, Durham, NC: Duke University Press.
37. Schwartz-Shea, P. and Yanow, D. (2002) '"Reading", "Methods", "Texts": How research methods texts construct political science', Political Research Quarterly 55: 457–486. | Article |
38. Scott, J. and Keates, D. (eds.) (2001) Schools of Thought: Twenty Five Years of Interpretive Social Science, Princeton: Princeton University Press.
39. Taylor, C. (1971) 'Interpretation and the sciences of man', Review of Metaphysics 25: 3–51.
40. Van Eeten, M.J.G., Van Twist, M.J.W. and Kalders, P.R. (1996) 'Van een narratieve bestuurskunde naar een postmoderne beweerkunde?' Bestuurskunde 5: 168–189.
41. Weick, K.E. (1995) Sensemaking in Organizations, London: Sage.
42. Wendt, A. (1999) Social Theory of International Politics, Cambridge: Cambridge University Press.
43. Wittgenstein, L. (1972) Philosophical Investigations, translated by G. Anscombe, Oxford: Basil Blackwell.
44. Yanow, D. (1999) Conducting Interpretive Policy Analysis, Newbury Park, CA: Sage.

About the Authors

Mark Bevir is a Professor in the Department of Political Science, University of California, Berkeley. His recent publications include The Logic of the History of Ideas (Cambridge University Press, 1999); New Labour: A Critique (Routledge, 2005) and, with R.A.W. Rhodes, Interpreting British Governance (Routledge, 2003), and Governance Stories (Routledge, 2006).

R.A.W. Rhodes is Professor of Political Science and Head of Program in the Research School of Social Sciences, Australian National University, and Emeritus Professor of Politics at the University of Newcastle, UK. He is the author or editor of 25 books. As well as the books with Mark Bevir, his recent publications include: Transforming British Government Volume 1 Changing Institutions; Volume 2 Changing Roles and Relationships (Macmillan, 2000); and (with Patrick Weller), The Changing World of Top Officials: Mandarins or Valets? (Open University Press, 2001). He has been editor of Public Administration since 1986. He is a life Vice-President of the Political Studies Association of the United Kingdom. He is a fellow of both the Australian and the British Academies of Social Sciences.