Chapter 1: Introduction
In this book, we propose new responses to old questions about the nature of the state and how to study it. We argue that political scientists should think of the state as a series of contingent and unstable cultural practices, where cultural practices consist of the political activity of specific human agents. We also argue that political scientists should explain these cultural practices by reference to the meanings embedded in them, where these meanings arise against the background of contingent historical traditions and dilemmas. This chapter sketches the roots of theories about the state. We trace the shift from idealism to modernist-empiricism, and to a radical historicism and anti-foundationalism.
The study of politics has long concentrated on the state as a sovereign authority. This concept of the state arose gradually and contingently during the Renaissance and Reformation, culminating in the great texts of Bodin and Hobbes. However, once the concept of the state as sovereign authority had arisen, it proved to be remarkably powerful and remarkably resilient. It inspired political actors to remake the world in its image, most famously in the Treaty of Westphalia, which enshrined it as a principle of international relations. Moreover, as this concept of the state became more and more entrenched in political life, so many students of politics began to take it for granted, treating it as a natural development.
The theory of the state dominated political science in the nineteenth century. In America and much of Europe, the central place given to the state was in part a response to, and a catalyst of, various forms of state-building. The important place given to the state also reflected its role as one of the key principles in the developmental historicism that then dominated the social sciences. The state made sense of historical developments: statehood was the consummation of the history of nations that were held together by ties of race, language, character, and culture. The state played this guiding role in traditions of scholarship as diverse as Hegelian idealism, Comtean positivism, evolutionary theory, and Whig historiography.
Proponents of all these different forms of developmental historicism typically agreed on several key points. First, they believed the state expressed, or at least could express, the common good of a people bound by cultural and ethnic ties. Second, they thought the social sciences could explain the character of any particular state through a historical narrative about the emerging political organization of a cultural and ethnic nation. Third, they believed that representative institutions, perhaps alongside a constitutional monarchy, could enable citizens to hold accountable political actors who embodied, acted on, and safeguarded the common good of the nation.
Developmental historicism often led to a view of the state as akin to an organism. A state embodied a spirit or culture that itself had a progressive teleological history. Thus, for example, Bosanquet (1965: 172) famously argued the state was not just an institutional reality but also an ethical one. Following Hegel, he saw the state as a nest of interconnected social relations and institutions of which it was the ethical core and guiding spirit. The state was greater than a particular set of political institutions. It embodied the general will of its people. For Bosanquet, to ask whether the state was just a set of institutions was like asking whether a person was just a set of physical organs. Just as an individual’s will and intentions guided and governed his body, so the public interest and general will of the people informed the public policies and historical development of a state.
Bluntschli is a neglected figure. Yet his classic The Theory of the State was probably the single most influential work of political science in the nineteenth century, acting as a founding text for the discipline in Germany and America. Bluntschli, like Bosanquet, believed in a developmental historicism in which the state was a ‘moral and spiritual organism’ (Bluntschli 1885: 54). He began by rejecting those existing theories of the state that gave it only a partial role. He attacked the legal conception of the state, arguing that it limits the role of the state to the protection of the rights of citizens; it ignores the state’s ethical role in promoting social justice and public welfare. He also attacked those concepts of the state, notably the ‘socialist’ one, that he thought gave it too much power. Then he turned to detailed discussions of representative democracy and constitutional monarchy. In both cases, he traced the historical development of a set of political institutions that arose as the expression of the moral organism. He paid particular attention to the link between these institutions and the historical cultures of specific nations.
Today the idea of the state as an ethical organism is alien to most of us. It was decisively undermined by the First World War, which eroded faith in progress and, so, in developmental historicism. The war led American and British scholars to distance themselves from the German philosophy and political science that was the basis for much of this theory of the state.
If the state is not an ethical organism, how should we think about it? If political scientists can no longer believe in developmental historicism, how can they make sense of the state and governance? Throughout the twentieth century most political scientists responded to these questions in ways indebted to modernist-empiricism and positivism. Modernist-empiricists see the state as a set of formal and informal institutions and behaviour, and they try to explain these institutions by appealing to ahistorical structures, classifications, correlations, and mechanisms.
Behaviouralists are far more critical of the concept of the state, often rejecting it for sociological terms such as ‘system’.
Modernist-empiricism and positivism
Since the collapse of developmental historicism, the study of the state in political science has been dominated by a modernist-empiricism mainstream that shades into positivism. We will briefly review these approaches to the state, looking at the forms of explanation and empirical topics which they favour.
Bluntschli’s decline from a pre-eminent voice to a neglected one reflects two overlapping but separate trends in political science. Although these two trends tended to reinforce one another, each had its own roots, and belief in one did not require belief in the other. The first trend saw modernist-empiricism replace developmental historicism as the dominant approach in the social sciences. Developmental historicism had located actions, cultures, and institutions in broad temporal narratives governed by largely fixed principles such as those of liberty, reason, and statehood. This developmental historicism was undermined by the crisis of faith in reason and progress associated with the First World War. This crisis encouraged the rise of modernist-empiricism. Modernist-empiricism treats institutions such as legislatures, constitutions and policy networks as discrete, atomized objects to be compared, measured and classified. It adopts comparisons across time and space as a means of uncovering regularities and probabilistic explanations to be tested against neutral evidence (see Bevir 2001). Political scientists such as Herman Finer (1932) and Friedrich (1937) broke the state into its constituent formal-legal institutions to provide comparative analyses of institutions and politics across modern states.
The early twentieth century witnessed a second, overlapping trend in political science: the rise of new empirical topics. Political scientists began to look beyond the topics associated with institutional history, constitutional law, and the general theory of the state. They believed that these older agendas reflected a pre-democratic Europe, and were no longer relevant to the mass-based politics that developed with an expanding mass suffrage. In their view, modern societies could be understood only if the dynamics of policy making, mass-based political parties, and public opinion were studied alongside formal government institutions. Political scientists wanted to study how policy making, parties and public opinion worked. The most important such study was Bryce’s The American Commonwealth (1888), which moved relatively quickly through historical and legal issues to devote hundreds of pages to parties and public opinion. Bryce’s work inspired other political scientists, including, most notably, Harvard’s Lowell, who later repaid the trans-Atlantic debt with The Government of England (1908).
It was when political scientists turned to modernist-empiricism to examine the new empirical topics that they crafted what would now be called the ‘old institutionalism’. Many of them continued to understand the state as both a set of formal and legal institutions and the ideas embedded in these institutions. Today many people treat this institutional approach to the state (staatswissenschaft) as the historic core of political science. They tout the study of formal and legal institutions as the defining feature of political science. Eckstein (1979: 2) is a critic of the old institutionalism, objecting that its practitioners were ‘almost entirely silent about all of their suppositions’. Nonetheless, he recognizes the importance of this ‘science of the state’ with its focus on ‘the study of public laws that concern formal governmental organizations’ (Eckstein 1979: 1, 2). Indeed, in an earlier article, he pointed out that:
If there is any subject matter at all which political scientists can claim exclusively for their own, a subject matter that does not require acquisition of the analytical tools of sister-fields and that sustains their claim to autonomous existence, it is, of course, formal-legal political structure. (Eckstein 1963: 10–11)
Similarly, Greenleaf (1983: 7–9) argued that constitutional law, constitutional history and the study of institutions constitute the ‘traditional’ approach to political science.
As the old institutionalism broke with developmental historicism, so it turned towards both modernist-empiricism and new empirical topics. First, the main forms of explanation typically became increasingly ahistorical. They were often cast in the form of cross-temporal and cross-cultural regularities or patterns. Legal rules and procedures often served as the independent variable with the functioning of democracies acting as the dependent variable. For example, Duverger (1959) criticized electoral laws on proportional representation because they fragment party systems and undermine representative democracy. Second, old institutionalists expanded the range of terms such as ‘state’ and ‘constitution’ far beyond the older focus on textual documentation and attendant legal judgements. For example, Herman Finer (1932: 181) defined a constitution as ‘the system of fundamental political institutions’. The old institutionalism came to cover not only the study of written constitutional documents but also the associated ‘customs’ (Lowell 1908: 1–15). The distinction between constitution and custom has since recurred in many related guises, most notably in the distinction between formal and informal organization.
The old institutionalism promoted comparative and inductive studies of states and their histories (Rhodes 1995: 43–6; and for the usual caricature see Thelen and Steinmo 1992: 3). Finer (1932) is a fine exponent of the comparative approach (and see Eckstein 1963: 18–23 and Bogdanor 1999 for more examples). While Finer broke away from the older state-by-state approach to compare institution-by-institution across states, he still located his institutional analysis within a clear theory of the state. For Finer (1932: 20–2), the defining characteristic of the state is its legitimate monopoly of coercive power (see also Sait 1938, chapter 5). He surveys the main political institutions ‘not only in their legal form, but in their operation’ (1932: viii), as they evolved. Political institutions are ‘instrumentalities’ which embody the ‘power-relationship between the state’s individual and associated constituents’ (Finer 1932: 181). Then, and only then, Finer compares the political institutions of America, Britain, France and Germany. His analysis covers the parts of state organization, including: democracy, separation of powers, constitutions, central-local territorial relations and federalism. Finally, he turns to ‘the principal parts of modern political machinery, namely, the Electorate, the Parties, Parliament, the Cabinet, the Chief of State, the Civil Service and the Judiciary’ (1932: 949). His approach is grounded in a theory of the state and explores key institutions and their operation.
The old institutionalism was also inductive. Its proponents claimed that the great virtue of institutions was that we could ‘turn to the concreteness of institutions, the facts of their existence, the character of their actions and the exercise of their power (Landau 1979: 181, emphasis in the original). We can draw inferences from repeated observations of these objects by ‘letting the facts speak for themselves’ (Landau 1979: 133).
Finally, while the old institutionalism turned away from historicist forms of explanation, it often continued to take a historical perspective, looking at the historical development of states and forms of government. History is extolled as ‘the great teacher of wisdom’ because it ‘enlarges the horizon, improves the perspective’ and we ‘appreciate … that the roots of the present lie buried deep in the past, and … that history is past politics and politics is present history’ (Sait 1938: 49). Because political institutions are ‘like coral reefs’ which have been ‘erected without conscious design’, and grow by ‘slow accretions’, the historical approach is essential (Sait 1938: 16).
In America and Britain, the formal-legal analysis of the old institutionalism remains alive and well. It dominates many textbooks, handbooks and encyclopaedias. Major works are still written in the idiom. Sammy Finer’s (1997) three-volume history of government combines a sensitivity to history with a modernist-empiricist belief in comparisons across space and time, regularities and neutral evidence. He attempts to explain how states came to be what they are with a specific emphasis on the modern European nation state. He searches for regularities across time and countries in an exercise in diachronic comparison. He sets out to establish the distribution of the selected forms of government throughout history, and to compare their general character, strengths and weaknesses using a standardized typology. He provides a history of government from ancient monarchies (about 1700BC) to AD1875.
Formal-legal analysis also remains prominent across continental Europe. It was the dominant tradition in Germany, although challenged after 1945. The challenge is yet to succeed in, for example, Italy, France and Spain. Here we can only give a flavour of the variety that is French political science and establish that it is a distinctive endeavour that runs at times in a different direction to, and at times parallel with, Anglo-American political science.
There is a strong French tradition of constitutionalism. It is descriptive, normative and legalistic, focusing on the formal-legal aspects of institutions, but not on case law. Chevalier (1996: 67) argues that ‘the growth of the French liberal state in the nineteenth century led to the predominance of the law and lawyers stressing the guarantee of citizen’s rights and limits on state power’. These jurists monopolized the field for nearly a century and it remains a major influence (see, for example, Chevallier 2002). So, despite various challenges, the 1980s witnessed ‘the resurgence’ of ‘legal dogma’ with its focus on the state’s structures and functions (Chevalier 1996: 73).
Outside the tradition of constitutionalism, the French approach to the study of institutions remains distinctive and does not engage with the Anglo-American literature. An early example is Duverger (1954; 1980). Although his work on electoral systems and semi-presidentialism is probably better known outside France than inside, nonetheless it was a major challenge to the academic lawyers and it influenced a younger generation of scholars. Latterly, ‘the strategic analysis of institutions’ focuses on electoral systems and core political institutions (such as the presidency) and tries to identify how institutions, singly and in combination, affect behaviour (for citations see Elgie 1996). Its main proponents include Duhamel and Parodi (1985). Parodi explains the changing nature of the Fifth Republic’s political system by identifying, for example, how the direct election of the president with a majoritarian electoral system for the National Assembly bipolarized the party system.
Bryce and Lowell introduced new empirical topics associated with mass-suffrage societies, but they typically continued to conceive of the state as the sovereignty of a collective will. This concept of the state began to lose ground to pluralist alternatives only after the First World War. The new trend arose through trans-Atlantic exchanges, for while American pluralism later developed a distinctive hue, it initially owed much to British scholars, especially Laski who spent several years lecturing at Harvard and Yale. Laski brought the term ‘pluralism’ and British debates about sovereignty into the American academy. Equally, his time in America made it central to his democratic theory (Gunnell 2007, chapter 7). In the first decades of the twentieth century, pluralism combined with the new empirical topics to transform the theory of the state. The older idea of the state as a sovereign authority became less popular. In its place, there arose studies of pressure groups within which sovereignty and authority were dispersed among various organizations. A string of future Presidents of the American Political Science Association (APSA) built their careers on studies of pressure groups – Herring (1929), Odegard (1928) and Schattschneider (1935). By the 1950s, the concept of a pressure group was also applied to British politics by the recent Harvard PhDs: Beer (1956 and 1963) and Eckstein (1960) alongside Finer (1958) and Mackenzie (1955). During the inter-war period empirical research on public opinion, parties, and pressure groups was known collectively as the study of ‘political behaviour’. The American Political Science Association’s (APSA) review of the state of the discipline in the 1940s stressed the pervasiveness of this phrase. It even claimed that ‘political behaviour has largely replaced legal structures as the cardinal point of emphasis among political scientists’ (Griffith 1948: 224). Even if this claim was an overstatement, it rightly suggests that the study of political behaviour was prominent in American political science beforethe onset of the ‘behavioural revolution’.
The behavioural revolution thus represented less a change in empirical topics than the rise of a new theory that stood in sharp contrast to modernist-empiricism as well as the earlier developmental historicism. In other words, to specify what was new about behaviouralism, we must remember the new empirical topics had arisen long before. Many behaviouralists had substantive interests in public opinion, pressure groups, and topics other than formal government institutions, but these interests simply extended an intellectual trend dating from the turn of the century. Behaviouralism got much support from the Committee on Political Behaviour, which was established in 1949 by the Social Science Research Council (SSRC). The Committee’s main goals were not, despite its name, to promote the study of behavioural topics. Rather, the common theme of the articles arising from the Committee on Political Behaviour was the call for a political science that was ‘systematic’ (Eldersveldet al. 1952; Leiserson 1951: 29–32; Garceau 1951: 69–85; Truman 1951: 37–39). The focus was on how to study politics, not what topics to study.
The Committee on Political Behaviour (1950: 20) had two declared goals: ‘development of theory’ and ‘improvement in methods’. Although behaviouralism is often remembered for the second goal, the first was as important. The failure of the specific theories advanced by behaviouralists in the 1950s and 1960s to win lasting support should not obscure the revolutionary impact of behaviouralism on conceptions of the character and role of theory, especially in American political science. Behavioural theory was broadly positivist, ethically neutral, and orientated to empirical science more than political practice.
Up into the 1940s there was little dispute about what theory was or should be. Political scientists on both sides of the Atlantic understood ‘theory’ principally to mean the study of the history of the political thought that had shaped modern institutions. Most political theorists spent more time studying, teaching, and writing about historic texts than attempting to produce novel theories. Political theory combined moral philosophy and constitutional history with the theory of the state. From moral philosophy, it inherited the idea of training young elites to take their place in the world by teaching them a canon of great texts. From constitutional history, it inherited a concern to study law, authority, and institutions through the study of legal and historical documents. From the theory of the state, it inherited a concern with classifications of types of government and institutions.
The rise of modernist-empiricism and the new empirical topics in the first half of the twentieth century did not challenge this approach to political theory. A view of science as fact gathering and objective reporting gave political scientists no reason to question the epistemic standing of historical research. While fewer and fewer political scientists outside the subfield of political theory did historical research, they did not see their theory colleagues as obstacles to a scientific discipline. To the contrary, political theory complemented the rest of political science. Political theory gave students both a historical survey of ideas set against the backcloth of evolving institutions, and a guide to the concepts used by contemporary political scientists. Thus, the generalist scholar still flourished. Barker (1915, 1944 and 1956) in Britain and Friedrich (1941 and 1952) in America wrote about both political ideas and institutions. They produced comparative institutional studies, histories of ideas, translations of canonical texts (Aristotle 1946; Kant 1949), and many other works. Modernist-empiricism offered little warning of the theoretical departures that positivism later inspired in American political science.
Positivist theory took varied forms in the different subfields of political science. In the study of American politics, the most famous theoretical product of behaviouralism was a new version of pluralism, best illustrated by Dahl’s classic A Preface to Democratic Theory (1956). But it was in the subfield of comparative politics that positivist hopes were most ambitious. Comparative politics underwent a dramatic expansion in size, scope, prestige, and funding as America became a superpower engaged in the cold war, competing for the allegiance of the new nations that emerged from decolonization in Africa and Asia. When Friedrich and Finer crafted new analytic frameworks for comparative study in the inter-war decades, they grounded their categories in the historical experiences of Europe and America. Their categories had a restricted scope. In contrast, the SSRC’s Committee on Comparative Politics brought young scholars together in the 1950s in a bid to remake the subfield. The positivism of these scholars led them to take abstraction to a whole new level. They sought a general set of concepts that could bring comparative studies of all countries within a single framework. This quest for a general theory was pitched at the macro-societal level, drawing on functionalism and systems theory (see, for example, Almond 1960). The members of the Committee on Comparative Politics were influenced by Parsons and his systems approach. They also owed a debt to Easton and British social anthropologists, which meant they adopted ‘political system’, rather than Parsons’ ‘social system’, as their core concept. When crafting a general theoretical scheme for analysing changes in political systems, they made ‘political development’ their key concept, rather than the more sociological and reductionist ‘modernization’ (see, for example, Almond 1965).
If the behaviouralists brought anything new to the critique of the state, it derived precisely from their positivist concept of general theory based on observable phenomena. The behaviouralists argued that the core life of a society should not be sought in the empirically mysterious and dubiously intelligible entity called the state. They thought the social sciences should study empirically observable activities and behaviour. Easton (1953: 108), a doyen of the movement, argued that the state is dispensible in political science since it is empirically unknowable. Similarly, Lasswell and Kaplan (1950: 184) claimed that ‘every proposition about the abstraction “state” can be replaced by a set of propositions referring only to the concrete acts of certain persons and groups’.
Behavioural theory, especially in its most aggressively positivist forms, barely appeared in Britain. To the contrary, even when British political scientists adopted techniques such as survey research or cross-national aggregate analysis, they remained tied to the older modernist-empiricism. The possible exceptions include Rose and Blondel, but they concentrated on empirical work, writing little about theory. Ironically, therefore, the main impact of positivist theory on British political scientists was that they began to define themselves against it. The biggest clichés about British political science define it in contrast to American positivism. This contrast enables British scholars to forget their modernist-empiricist inheritance, including the extent to which their search for formal explanations based on classifications and correlations among types of institutions embodied a modernist revolt against historicism. They define themselves instead as peculiarly sensitive to history, context, and agency. As Bogdanor (1999: 150) writes, typically unaware of the historical context of the contrast he is making, ‘if there is a central tendency to the discipline as it has developed in Britain in the twentieth century, it lies in aversion to positivism’. This position does not distinguish British political science. It just aligns the mainstream of the British discipline on one side of the conflict between modernist-empiricism and behaviouralism that defines the mainstream of the American discipline.
Historicism and interpretation
Modernist-empiricism, positivism, and neo-Marxism dominated political science for much of the twentieth century.
But they were by no means the only responses to the collapse of developmental historicism. One alternative was an emphasis on historical understanding and the interpretive nature of social science.
Hermeneutic theorists continued to conceive of action as meaningful even as they retreated from idealist notions of the absolute. They also began to craft a more radical historicism, replacing assumptions about a unified and rational progress with an emphasis on diversity, contingency, and ruptures. At the turn of the nineteenth century, German philosophers drew on the rich traditions of idealism and hermeneutics to discuss the philosophical foundations of the social sciences and their relationship to the natural sciences. These discussions produced some of the classic statements on historical understanding (verstehen). Dilthey’s work had a profound impact on approaches to the social sciences within the hermeneutic tradition. Moreover, the German discussions helped to inspire similar ones throughout Europe, including, through the Italian Croce, the work of the British philosopher, Collingwood.
Historical understanding has to do with grasping the intentional content attached to human actions. It is a distinctive approach to the social sciences, typically based on the idea that the meaningful nature of human action requires a distinct epistemology in the form of historical understanding. Philosophers who study historical understanding focus on three interwoven but analytically distinct concerns. The first concern is to lay out the ontological conditions for the possibility and grounds of historical understanding: what features of human nature support interpretation? The second is to clarify the epistemological nature of historical understanding itself: what does it mean to interpret actions that took place in the past? The final concern is to specify the role of understanding in historical explanation: how and to what extent does interpretation explain social phenomena? In stressing these issues, we challenge two myths about historical understanding. The first myth is that historical understanding is just a method for producing historical facts. The second myth is that it is inherently an intuitive, quasi-mystical act that resembles the work of an artist more than that of a scientist. In place of these myths, we highlight the philosophical content of historical understanding seen as an analysis of what it means to base the social sciences on the meaningful character of human action.
Given that historical understanding concerns the intentionality of human action, it is not surprising that discussions of its philosophical principles arose against the background of idealism and hermeneutics. The idealists stressed the primacy of mind over matter in human existence including efforts to understand that existence. Hermeneutics was preoccupied with questions of understanding linguistic utterances by grasping the mental content expressed within them.
Dilthey was the first major thinker in both the idealist and hermeneutic traditions to attempt a systematic theory of historical understanding. Dilthey anchors both the possibility and the necessity of historical understanding in ontological arguments. According to his ontology, human life differs from the rest of the natural world in that it consists of an inner world expressed in outer manifestations such as gestures, words, music, poetry, churches, and universities. Understanding is the process of grasping the inner content to which outer expressions refer. It is only by such understanding that humans are able to live, act, and communicate in a society. Understanding is a fundamental presupposition of individual and social life. This ontology leads Dilthey to define the social sciences as dependent on historical understanding. The social sciences, like human life, are about understanding expressions. The social sciences grasp the inner content of outer manifestations.
How does Dilthey see the process of historical understanding? At times Dilthey (1996: 229) identifies understanding with an intuitive and imaginative re-creation of inner contents in one’s psyche. He based this re-creation on an analogy with one’s own lived experiences and enhanced by ‘sympathy with everything human’. Scholars should ‘transfer’ themselves ‘into a given complex of manifestations of life’ in a way that enables them to experience the temporal flow of historical events within their own psyche (Dilthey 2002b: 235). In his later writings, however, Dilthey downplayed the importance of individual psychic processes. Instead he placed ‘objective spirit’ at the centre of historical understanding. Objective spirit refers to the ensemble of objectifications (outer manifestations) that comprise a society at a certain epoch. For Dilthey (2002a: 229), all these objectifications share ‘a common stock of ideas, attitudes, and ideals’, which ‘even the work of a genius will reflect’. Our everyday understanding of the objectifications that surround us needs these common conventions. So, historical understanding can only be grasped by using these social commonalities; that is, by ‘locating the individual manifestation of life within a common context’ (Dilthey 2002a: 230).
The process of understanding is important for Dilthey’s epistemology. It leads him to raise a question that remains prominent in many later discussions of historical understanding. How can empathetic historical understanding have a ‘universally valid’ objectivity (Dilthey 1996: 253)? On what grounds can a mental process occurring in the subjective experience of the historian be recognized as a valid interpretation of a given historical phenomenon? Dilthey treats the question of objectivity as one not of methodology but of philosophy. He does not answer it by offering a procedure that if followed by historians would secure the validity of their interpretations. He answers it by an ontological argument that, in his view, guarantees the validity of the re-created experiences in a historian’s mind. Dilthey (2002a: 298) posits the identity of the subject and object of knowledge:
The primary condition for the possibility of historical science is contained in the fact that I am myself a historical being and that the one who investigates history is the same as the one who makes history. … Lived experience contains the totality of our being. It is this that we re-create in understanding.
While this response to the problem of the validity of historical understanding may seem vague, underdeveloped, even quasi-mystical, Dilthey’s work did much to raise issues that still concern philosophers interested in historical understanding.
Late in life Collingwood (1946 and 1978) suggested that any statement or action was an answer to a question, where questions are posed in particular contexts that consist of historically specific webs of beliefs. History is a series of particular ideas and epochs uninformed by any real continuity, let alone any concept of the universal. However, Collingwood did not see history purely as the particular. To the contrary, much of his philosophy retains idealist echoes. In particular, his famous concept of re-enactment keeps a form of historical continuity and arguably drifts toward some of the religious and mystical ideas he held in his youth.
Collingwood argued that historians know past thoughts when they re-enact them; that is, when they have the same thoughts but in the present (see Van der Dussen, 1995; Bates, 1996). At times, Collingwood treats re-enactment as a method or formal requirement of understanding. At other times, he leans toward a more metaphysical understanding of re-enactment. Re-enactment here involves a mind becoming directly aware of another that existed in the past in a way that suggests that thought has the transcendent quality of persisting through time and across contexts. Collingwood’s claim of a transcendent link with past thought is most clear in his early writings where he relates it to a mystical and immediate insight into the unity of the whole. He wrote:
The necessity of the mystical experience lies in the principle that we discover new truths by an act of the mind which reaches out beyond the given, grasps new thought as it were in the dark, and only after that consolidates its new conquest. (Bates 1996: 36)
Although Collingwood (1946: 287) later dropped the reference to mystical experience, he still argued that historians can have thoughts identical with those of their subjects. He argued that, although ‘acts of thought happen at definite times’, ‘thought itself is not involved in the flow of immediate consciousness’, but rather ‘stands outside time’. Thought has the peculiar property of not only ‘occurring here and now in this context’, but of being able ‘to sustain itself through a change in context and revive in a different one’ (Collingwood 1946: 297).
According to Collingwood (1946: 306), thought alone has some existence outside time: it ‘is not merely immediate’. The transcendent quality of thought is what makes re-enactment possible. Because thought can reappear in different contexts while still keeping its identity, historians can re-enact it as it was in the past. Things other than thought are immediate, which means that they cannot be re-enacted. So, they are not amenable to historical study. As Collingwood (1946: 304) explained, ‘of everything other than thought, there can be no history’.
An alternative to positivism
Between the wars, Collingwood was almost a lone voice in Oxford drawing on an older idealism to address hermeneutic themes. In the 1950s, however, hermeneutic themes and the idea of historical understanding became a prominent alternative to positivism in Anglophone philosophy. By the 1960s, a distinctive Anglophone approach to historical understanding appears, especially in Britain and Canada. Philosophers such as Dray, Winch, and Taylor drew on Collingwood, Wittgenstein, and a range of hermeneutic and phenomenological thinkers to counter influential attempts to develop a positivist analysis of the social sciences. Instead, they insisted on the interpretive nature of social science.
The positivist challenge
A second wave of verstehen theory emerged in reaction to the positivist attempt to reunify the natural and social sciences on a naturalist basis. The positivist project, which had middle European roots, consisted of an attack, led by Hempel and Abel, on the idea that historical understanding could be a legitimate form of explanation.
Hempel (1965 and 1966) claimed that empathy is not a basis for explanation in the social sciences, but, at most, a source of hunches. He argued that historical explanations are always implicitly nomological in that they involve a deduction from premises which include at least one natural law. Explanations that appeal to the motives of actors presuppose psychological generalizations.
Abel’s (1974: 85) approach is more complex than Hempel’s. On the one hand, he claims that historical understanding is ‘necessary and indispensable’ for the ‘study of social behaviour’; but, on the other, he restricts the role of historical understandingto the generation of hypotheses in which meanings are reduced to intervening variables. He writes, ‘verstehen … consists of the act of bringing to the foreground the inner-organic sequence intervening between a stimulus and a response’ (Abel 1948: 214). Abel reduces historical understanding from a philosophy to a method. He makes historical understanding an ‘operation’ within a positivist philosophy.
The positivist challenge provoked a lively debate about historical understandingin Anglophone social sciences, and a number of scholars rose to defend its importance (see Truzzi 1974). The most systematic defence of historical understanding, however, came not from social scientists, but from philosophers such as Dray, Winch, and Taylor.
Dray (1995), following Collingwood, insisted that historical explanation depends on making sense of the inner world of actors. Dray departs from Collingwood’s perspective, however, in his attempt to integrate causal explanation and historical understanding. In particular, Dray argues that the causal explanation of any action requires an understanding of the purposive reasoning, whether conscious or unconscious, spoken or latent, that caused the actor to perform that action.
Dray (1957: 125) describes interpretive explanation as follows:
Rational explanation may be regarded as an attempt to reach a kind of logical equilibrium at which point an action is matched with a calculation [attributed to the actor]. A demand for explanation arises when the equilibrium is upset – when from the ‘considerations’ obvious to the investigator it is impossible to see the point of what was done. The function of the historian’s explanatory story will in many cases be to sketch in the corrections to these ‘obvious’ considerations which require to be made if the reader is to be able to say: ‘now I understand what he was about’.
The historian has an initial expectation about what a ‘normal’ course of events would look like in a given situation. A deviation from that expectation prompts the historian to piece together those elements of the actor’s reasoning ‘which are different from those we might have assumed in absence of evidence to the contrary’ (Dray 1957: 125). The investigation is completed once the amassed knowledge of the actor’s reasoning is enough to account for the ‘deviant’ course of action. Reconstructing the inner reasoning of historical actors is not an arbitrary, intuitive product of pure empathy. Rather, it is built up ‘from the evidence’ (Dray 1957: 129).
Winch (1958: 23 and 123) argues that social life is fundamentally different from the rest of the natural world in that social relations are ‘internal relations’. Social relations exist through ideas held in people’s minds rather than through the external, physical aspects of human interaction. Thus all ‘specifically human behaviour’ is meaningful (Winch 1958: 52). Winch then adds that all meaningful behaviour is a matter of applying intersubjective, socially established rules in specific social contexts. For Winch, this view of human society is ‘logically incompatible with the kinds of explanation offered in the natural sciences’ (Winch 1958: 72). ‘Historical explanation’, writes Winch, ‘is not the application of generalizations and theories to particular instances: it is the tracing of internal relations’ (Winch 1958: 133). Explanation is a matter of understanding the social actors’ own understanding of their reality.
Winch, like Dray, restores the constitutive role of meaning in social life and in our knowledge of it. But, unlike Dray, he rejects any association of historical inquiry with causal analysis. He attacks Weber’s view that to get scientific validity, historical understanding needs to be supplemented by a probabilistic or statistical causal analysis. Statistical analysis, according to Winch, can never get us closer to understanding social action. For example, ‘[a] man who understands Chinese … is not a man who has a firm grasp of the statistical probabilities for the occurrence of the various words in the Chinese language’ (Winch 1958: 115). Winch also rejects psychological forms of understanding directed at grasping internal mental processes. In his view, all understanding (of oneself, of one’s reality, of the objects of historical inquiry) is mediated by socially established concepts. All understanding occurs within an intersubjective linguistic medium, rather than in a private unarticulated experience.
Taylor studied at Balliol – the birthplace and stronghold of British idealism. After that, he was appointed successively a Fellow of All Souls College, Oxford, and Chichele Professor of Social and Political Theory at Oxford, before he later returned to his native Canada. His doctoral thesis reworked the idealist concern to defend a vitalist analysis of human behaviour against the formalist and positivist approach of behaviouralists. It drew on the emerging analytic and linguistic philosophy in Oxford to rewrite this concern as a conceptual and empirical issue. This issue could not be decided by an appeal to metaphysical categories such as absolute mind (Taylor 1961 and 1964). After completing his doctorate, Taylor wrote a series of essays explicitly challenging behaviouralism and its core beliefs. He argued, in ‘Interpretation and the Sciences of Man’, that beliefs, meanings, and language were constitutive of human actions and practices. The social sciences were unavoidably hermeneutical. His argument entailed a break with ‘mainstream social science’ and its empiricist and positivist epistemology. In particular, ‘we cannot measure such sciences against the requirements of a science of verification; we cannot judge them by their predictive capacity’ (Taylor 1971: 51).
In ‘Neutrality in Political Science’, Taylor extended his argument to take direct aim at ‘the cult of neutrality’ (Taylor 1967: 48, 27, and 46). Behaviouralists defended the superiority of their approach by arguing that older approaches were always permeated by value positions. So, their frameworks were never scientific. They always served the interests of a normative or ideological theory. Behaviouralists proposed to turn the study of politics into a technocratic ‘policy science’, akin to engineering or medicine, which would ‘show us how to attain our goals’. However, Taylor pointed out that when behaviouralists constructed theoretical frameworks to delimit the proper area of scientific inquiry, they, too, made fundamental choices that entailed normative commitments. The work of Lasswell, Easton, and Almond hid their norms. Taylor suggested that conceptual frameworks always depended on theory, and theory cannot be constructed apart from values. The ties binding theoretical frameworks and values also opened the possibility of seeing some values as especially meaningful responses to particular empirical contexts.
Stivers (2008: 943) suggests our argument for an interpretive approach was ‘weakened by the failure to include its philosophical roots. Nowhere is there a citation to any of those who constructed the foundations of interpretivism’. Hay (personal correspondence) urged us to take account of ‘similarly minded figures in the literature’ and to situate our views ‘with respect to the existing literatures on the state’. Initially, we were reluctant to provide yet another critical review of the present-day political science literature. Engaging with the literature is one of those ‘rules’ of the political science community that can lead to endless citations and much tedium. It is also a rule suffused with notions of cumulating scientific knowledge. As should now be obvious, we do not draw on present-day political science to any significant degree. So, to the extent we draw on a literature unfamiliar to many political scientists, we have now identified our intellectual forebears. When relevant, as in Chapter 5, we will use endnotes to identify where our concerns overlap with those of some of our contemporaries. Our primary task, however, is to spell out our preferred approach to studying the state, and Part 1 concentrates on the general theory of the state.
A summary of the argument
The theory of the state is the historic heart of political science. As we show briefly in this chapter, during the twentieth century, the main approach to the state was modernist-empiricism with its associated empirical topics. Positivists, including many behaviouralists, tried to replace the concept of the state with other concepts more amenable to their general theories, but modernist-empiricists nonetheless ensured that the concept kept a key place in the lexicon of political science. We looked at present-day versions of modernist-empiricism, unpacking the distinctive philosophies, concepts, and empirical topics. We did so to identify the distinctiveness of our approach. Our aim is to define, defend, and illustrate an alternative theory of the state. It is not based on modernist-empiricism but on hermeneutics and historicism; on Dilthey and Collingwood rather than Weber or Marx.
Chapter 2 examines the leading present-day expression of modernist-empiricism: the new institutionalism. First, we challenge the belief of new institutionalists that they have a shared new paradigm. Instead, we suggest that it is varied, containing strands that can scarcely be reconciled with one another. Second, we argue that historical institutionalism exists primarily as a counter to rational choice institutionalism, but its inductive, modernist-empiricist logic lacks a micro-level theory for explaining historical change and awareness grows that it is not a distinctive approach. Third, we argue that ideational or constructivist institutionalism offers a promising way forward provided it develops a micro-theory that appeals to the beliefs and desires of actors and it breaks with the ahistorical forms of explanation of modernist-empiricism.
Chapter 3 explores what many observers suggest are the leading present-day forms of interpretive social science: postmodern theories of discourse. First, we challenge the belief that these approaches have a shared paradigm stemming from anti-foundational philosophy. Instead, we suggest that anti-foundationalism is compatible with all kinds of approaches to political science, and that these approaches arose out of distinctive socialist or radical traditions, not mainstream political science. Second, we argue that much discourse theory still relies on structuralist and occasionally even determinist forms of explanation associated with modernist-empiricism. There is a clear tension between this lingering structuralism and historicist genealogies. Third, we suggest that discourse introduces several new topics to the theory of the state. These topics include governmentality, collective identities, ideologies, and resistance in governance.
In Chapter 4, we present our own theory of the state and highlight its distinct and distinctive features. Our approach differs significantly from the ideational and interpretive approaches found respectively within the new institutionalism and discourse theory. In particular, we defend a historical approach to meaning as situated agency, explicitly opposing structural concepts of signifiers and regimes. From this perspective, the state appears as a differentiated cultural practice composed of all kinds of contingent and shifting beliefs and actions, where these beliefs and actions can be explained through a historical understanding. We then provide the set of aggregate concepts appropriate to historical studies of the state: situated agency, practice, power, narrative, tradition and dilemma.
We explain our approach to studying the state as an introduction to our main concern in this book: empirical shifts in, and debates about, the changing nature of ‘the state’. We want to show how our anti-foundationalism leads us to see both the state and the reforms that produce changes in the state differently. In Chapter 5, we take our aggregate concepts and decentre present-day approaches to the changing state. Such changes are typically described as a shift to governance. First-wave studies of governance typically describe changes in the state and its links to civil society as akin to a differentiated polity. This polity was characterized by a hollowed-out state, a core executive fumbling to pull rubber levers of control, and, most notably, a massive spread of networks. Second-wave studies of governance herald a return to the state with the idea of metagovernance. We pronounce the death of both narratives, because we argue that there is no single account or theory of contemporary governance, only the differing constructions of several traditions. We counterpose these narratives with a third-wave of governance that announces the arrival of the stateless state. We argue that the state arises out of the diverse actions and practices inspired by varied beliefs and traditions. The state, or pattern of rule, is the contingent product of diverse actions and political struggles informed by the beliefs of agents rooted in traditions. We identify new research topics suggested by our decentred analysis under the headings of the ‘3Rs’ of rule, rationalities and resistance, and we provide examples of each in Part II.
There is no one-to-one correspondence between the ‘3Rs’ and the individual chapters, but there is a steady shift of emphasis from rule to rationalities to resistance. We start by looking at patterns of rule as they are understood by elite actors. Chapter 6 describes the everyday life of ministers through an account of their daily life in government. We provide our interpretation of their interpretation of what the world looks like through their eyes. We watched people in their everyday office life, reporting on what they said and did, on their reasons for their actions. We understand ‘ruling’ and, therefore, ‘the state’ through their eyes.
We then turn to the rationalities or technologies used by government, specifically the rationality associated with the new public management, or managerialism, with its emphasis on targets and performance measurement. In Chapter 7, we decentre the Blair government’s putative reforms of a central feature of the British state as understood by Westminster elites, notably the civil service. We tell three stories about the civil service and public service delivery. First, there is the centralisation story, which claims the changes sought to increase the power of the prime minister at the expense of cabinet and the departments. Second, we tell the management story, which claims the reforms of the civil service foundered on Blair’s lack of policy making and management skills. Finally, we recount the governance story, which argues that the Prime Minister is locked into webs of dependence that undermined his initiatives.
Chapter 8 looks at the way senior civil servants coped with managerial rationalities: the continuous reform of the public service. The rise of the new public management in the 1980s led to recurring challenges to the administrative traditions of the public service in Australia, Canada and the United Kingdom. We analyse how the heads of the public service articulate the traditions of ‘constitutional bureaucracy’ found in Westminster systems of parliamentary government and selectively draw on past understandings to understand present-day changes. We describe living traditions under challenge that reshape reforms as reforms reshape them. We conclude that the heads of the public services have found ‘space’ or ‘voice’ to identify innovative ways of combining past traditions with new organizing principles of governance. In each case, it is not a question of ‘in with the new, out with the old’, but of ‘in with the new alongside key parts of the old’. The myths and legends of yore remain germane to the modern public service.
So far our focus has been on elite narratives and elite understandings of the state and its rationalities. In Chapter 9, we turn to the question of how middle-level managers, employees, and citizens’ experience the state and its rationalities. Again, we employ ethnographic methods to understand the ways in which they construct ‘the state’ in their everyday beliefs and practices. We show there are significant differences in the way the rationality of managerialism is experienced and understood. These different narratives are stories of resistance and are central to understanding what elites term the ‘implementation gap’ in the reform of public services.
Finally, in Chapter 10, we summarize our core argument on the stateless state. We highlight our key findings. We reflect on what we have learnt from observation compared with more conventional methods. We discuss the strengths and the weaknesses of our methods. We answer the ‘so what’ question by enumerating the benefits of our approach: theoretical and methodological. Finally, we reflect on the normative implications of our analysis for the Westminster state and its tradition of representative democracy. We examine briefly the case for supplementing representative institutions with participatory democracy, collaborative forms of governance, and dialogic policy-making.