‘A Life of Their Own’? Traditions,Power and ‘As If Realism’ in Political Analysis

Introduction

In a foreshadowing of a future debate in political science, the historian Eric Hobsbawm discussed the open question of the viability of traditions. In a paper which focused on the ‘invention’ of tradition, Hobsbawm (1977: 2) noted that the ‘extent to which invented traditions . . . develop a life of their own’, and why some remained in existence while others fell away, would have to remain points unaddressed for the time being. ¹ In more contemporary work, while these critical questions are beginning to be addressed, tradition remains ambiguous – simultaneously under-theorised, yet much discussed as a concept seemingly overcome by intractable ontological differences within political science. Across different approaches to political analysis, traditions have been seen as ideational backgrounds that do not determine an actor’s beliefs going forward (Bevir and Rhodes, 2018: 17), part of an ‘ongoing social inheritance’ for actors (Bevir and Blakely, 2018: 54), things that are ‘inscribed’ within institutions (Marsh et al., 2014: 341) and beliefs that can be transmitted and developed not just by agents, but by ‘structures and discourses’ too (Hall, 2011: 73). While traditions have long featured in political science, this article argues that the ‘interpretive’ turn (Bevir and Blakely, 2018: 2), and a renewed focus on traditions (Hay, 2011), requires further theoretical work to clarify a key ideational concept. In so doing, I also aim to draw upon insights from different analytical approaches, theorising tradition while avoiding ontological debates becoming ‘ontological evangelism’ (Hay, 2011: 175).

It is by attempting an answer to Hobsbawm’s question, one more recently articulated by Emily Robinson (2016: 123) around ‘why certain traditions “stick”’, that this article seeks to contribute to theorising tradition. Furthermore, and building on the suggestion that to understand why some traditions ‘acquire and retain resonance’ we need to consider synthesising interpretivist and constructivist insights (Hay, 2011: 167), this article seeks to build on the suggestion of a synthesis. Working with an interpretive theoretical framework, my argument also draws upon contributions from social constructivism, critical realism and work exploring institutionalism. The argument’s distinctiveness lies in an understanding of tradition to be what Colin Hay has termed ‘as if real’: both as an analytical device and as part of a ‘separate ontological category’ where a conceptual abstraction can still be understood as ‘at least partially generative of the practices and processes which we can directly observe’ (Hay, 2014: 467). While all traditions are legitimated by people and therefore contingent upon them (Berger and Luckmann, 1975: 111), the actions of those who uphold traditions can also lead to a level of social or institutional authority (Robinson, 2016: 121). Actors orient from (on the basis of inherited beliefs) and around (on the basis of the beliefs of others) traditions, and these ‘effects’ can be usefully understood and analysed through an ‘as if real’ ontology (Hay, 2014: 465). Traditions are abstractions of particular beliefs and practices, and so not ‘real’. Nor are they ‘real’ in the sense of being capable of agency. But they can appear ‘as if real’ on the basis of the actions of groups of actors within historically contingent contexts. The resonance ² of traditions, I argue, can therefore be understood as ‘socially contingent’: that is, contingent upon power relations and the ideational maintenance of traditions by their upholders.

Both are essential for understanding the contingency and relative resonance of traditions, and both are understated in the literature on tradition and interpretive political science. From this ontological position comes the argument that traditions are partly constitutive of both an actor’s beliefs and of their institutional context – they help us understand why a person believes what they believe and how a person’s strategic calculations may be affected by perceptions of what others believe. Traditions are not only an initial foregrounding of inherited beliefs, but also evolving and ongoing orientation tools in dynamic institutional and political contexts (Hay, 2014; Robinson, 2016). This helps us understand the relevance of traditions to any actor in an institutional context, particularly where there is contestation over the ‘right way’ of doing things. It is an incredibly important insight of interpretivism, effectively and convincingly set out in the work of Bevir and Rhodes, that the stories people construct about the past, present and future are integral to understanding beliefs, motivations and actions. Epistemologically, this means ‘grasping the intentional content attached to human actions’ (Bevir and Rhodes, 2012). Building on these insights requires us to ask why some stories prove to be so successful and to move towards a greater understanding of how competing stories (and therefore competing traditions) affect political debates and decision-making.

Tradition is a key explanatory concept within interpretive political science, yet the issues raised will be familiar to any political scientist: the ontological debate about ‘structure’ and ‘agency’. If tradition appears to ‘compel’, interpretive political science would consider it to be too structural a concept. If tradition is merely an inheritance that any actor can discard, many outside of interpretivism would query its analytical utility. As such, the role of tradition can be unclear to scholars across different approaches. Social constructivism has an important role to play here, particularly the insight that contexts – partly constituted by traditions – favour certain strategic choices over others (Hay, 2008: 63–64). Engaging with this emerging interpretivist/constructivist institutionalist synthesis, my argument adds something more. Scholars will be familiar with the idea that people are ‘socialised’ into traditions, but the process of people seeking to socialise beliefs – to create, maintain and uphold traditions – is overlooked. Similarly, why somebody would also seek out such traditions, to help orient themselves, is a point rarely made in the literature. As the philosopher Mary Midgely (2006: 21) argued, ‘humans are bond-forming animals’. People make use of the social world they inhabit. Traditions are a part of that social world, and what makes some traditions ‘stick’ are both the actions of people and the attraction of traditions in our social world. After all, ‘politics is an expressive form of human activity’ where ‘who we are and what we want to be and how we want other people to view us’ are highly relevant (Hindmoor, 2018: 221).

Tradition, Hannah Arendt (2006: 5) wrote, ‘selects and names . . . hands down and preserves . . . indicates where the treasures are and what their worth is’. ³ Tradition involves an inheritance, and therefore is strongly connected to experiences and stories of the past. Traditions ‘bind actors together in often unacknowledged inter-subjective communities’ (Hay, 2011: 170). This, I argue, requires an understanding of power which springs from a group getting together and acting together (Arendt, 1970: 52) legitimating, maintaining and upholding a particular aspect of social knowledge (Berger and Luckmann, 1975). Utilising this analytical framework, analysts can unpack why some ‘inter-subjective’ communities are stronger than others. Where there are conflicting narratives over ‘how things should be done’, we can witness the effect of competing traditions on the strategies of political actors. This effect stems from an actor’s perception of power relations and the relative support they perceive for competing traditions. An actor’s chosen strategies, and so actions, may be affected by their perception of the resonance of traditions within their institutional context. Perhaps they will adjust their rhetoric or postpone a reform, hoping or even expecting the balance of power to change. On the contrary, traditions can reinforce an actor’s actions, perhaps bolstering a reform or enhancing the effect of rhetoric.

These arguments suggest that individuals need not be forced to alter their beliefs, when encountering a ‘dilemma’ (Bevir and Rhodes, 2018: 17), nor that individuals or groups of individuals are necessarily transforming ‘a tradition’ when advocating for an alternative. Instead, it can be observed that people argue for and uphold competing traditions in a contest of interpretation, persuasion and relative resonance. The first part of this article briefly summarises and then critically analyses the ontological debates ‘about whether meanings and traditions exist independently of individual subjects’ (Diamond and Richards, 2012: 181). Here, I intentionally focus on the most recent arguments on tradition from Bevir and Rhodes (2012, 2018), recognising that the theoretical and philosophical underpinnings for interpretivism continue to be fleshed out. I then move to the substantive theoretical contribution of the article, arguing that the resonance of a tradition is contingent upon power and ideational maintenance from upholders, and that ‘as if realism’ is an appropriate ontological position for tradition as a concept. Finally, I offer a substantive example of how traditions affect the beliefs, strategic deliberations and actions of political actors by analysing the British Labour Party leader Neil Kinnock and his appraisal of the traditions within his party during the 1970s and 1980s. In the conclusion to the article, I set out some preliminary conclusions on the role of tradition in political analysis.

Contingent, Reified, Inscribed and ‘As If Real’ – Traditions in Political Science

Traditions, defined by Bevir and Rhodes (2003: 33) as beliefs and practices developed by individuals and passed on to the next generation, are critical to interpretivism. They are the ‘background’ against which a person comes ‘to adopt an initial web of beliefs’ that influences their beliefs and actions in the future (Bevir and Rhodes, 2018: 17). When institutions – from a constitution to a political party – are understood as patterns of beliefs and practices partly constituted by traditions, we see how integral the concept is to interpretive political science. A feature of the work of Bevir and Rhodes is consistency of terms, which for a concept such as tradition is important for explanatory power. This is not always the case with work that features patterns of beliefs and practices. Tradition is at times ‘lumped in’ with a whole host of other terms, many of which Bevir and Rhodes consider to be emblematic of positivism, such as ‘norms’, rules and customs. Tradition is preferred because, unlike norms or rules, accepting that people ‘reach the beliefs they do against the background of a social tradition’ need not entail the giving up of intentionality (Bevir, 2000: 33). Dilemmas – experiences or ideas that conflict with an individual’s belief – force people to alter the beliefs they inherited by way of tradition. These adapted beliefs are then passed on, meaning tradition can be constantly altered and is contingent on the beliefs of individuals (Bevir and Rhodes, 2012: 79).

I accept the definition of traditions, defined as beliefs and practices, which are inherited – even if a person actively seeks such an inheritance by, for instance, joining a political party. While in some literature tradition is enveloped in negative description – where, as Williams (1988: 320) noted, traditions are ‘inconvenient to virtually any innovation’ – work that seeks to understand tradition as a concept in social science typically adopts the more positive, dynamic understanding which Bevir and Rhodes use. Yet other definitions do add some important details about the concept – and these are particularly relevant to comprehending the attractiveness and usefulness of traditions to people navigating the social world. These additions are about the kind of social knowledge upholders present and pass on from one generation to the next. As with the definition offered by Arendt, quoted above, scholarship from different disciplines has focused on traditions as ‘knowledge with authority’. For example, Giddens (1998: 46) focused not on a long past – which traditions need not have – but rather that traditions presume ‘an idea of ritual or revealed truth – and this defining trait is also the origin of its authority’. Retellings and interpretations of the past are, of course, essential for explaining why a tradition has this ‘truth’, but a practice simply existing for a long period need not make it a tradition.

This is similar to Hobsbawm’s (2012: 4) argument about invented traditions, where one is observing ‘a process of formalization and ritualization, characterised by reference to the past, if only by imposing repetition’. Hobsbawm added a Marxist delineation, too, when separating tradition from convention and routine. The latter are not invented traditions ‘since their functions, and therefore their justifications, are technical rather than ideological (in Marxian terms they belong to “base” rather than “superstructure”)’ (Hobsbawm, 2012: 3). This shared insight – that repetition of beliefs and practices over generations is a part of tradition, but not its entirety – is important for understanding the nature of tradition and why it is worth studying. Traditions, as with myths, ‘have an extraordinary power to rally’ (Samuel and Thompson, 1990: 18). Within an institutional context, traditions can represent who a person is, and in a contestation of competing beliefs and practices, which side a person is on (Samuel and Thompson, 1990: 18). With this comes a sense of being part of something ‘bigger’. An individual is situated not only within a historical background but also within a defined part of the social world, with symbols and meanings that become a part of a person’s identity.

On the use of tradition in interpretive political analysis, Hay (2011: 181) has both summarised a key ontological dispute and suggested there are ‘synergies with cognate perspectives, notably the constructivist institutionalism with which it shares so much’. On the dispute, it relates, he argued:

to the status of the aggregate concepts used to describe the context within which actors are situated. Put simply, for critical realists, the structures to which they point . . . are ontologically real; for interpretivists, the equivalent aggregate concepts, notably the traditions with respect to which they contextualise actors’ beliefs and meanings, are not (Hay, 2011: 177).

From the perspective of scholars challenging some aspects of interpretivism, the degree to which individuals can transform traditions has also been overemphasised. Traditions should instead be understood as being ‘inscribed’ in institutions, processes and narratives, thereby constraining actors (Marsh et al., 2014: 341). Furthermore, it has been suggested that what tradition actually does within an interpretive theoretical framework is difficult to see: ‘the causal relationship between traditions and beliefs is vague. Traditions exist but do not determine. They are strong in that they socialise, but weak in not preventing other forms of beliefs’ (Smith, 2008: 146). In particular, Smith noted an absence of an analysis of power in the work of Bevir and Rhodes (2008: 145), suggesting that ‘when humans are placed within a web of power relations, actions and institutions may depend on beliefs, or beliefs may be a consequence of webs of power relations’. To Bevir and Rhodes (2018: 14), much of this criticism is the ‘reification’ of concepts like tradition, or beliefs, into ‘structural’ forces that lose the contingency of social constructions.

For Hay (2011: 175), the concept of ‘situated agency’ in Bevir and Rhodes’ interpretivism means there is the ‘at least implicit conception of structure at work within’. This, Hay argued, showed the distance between a critical realist and an interpretivist ontology is smaller than both sides have suggested. Actors are ‘situated’ in a given context, one ‘understood in terms of the structuring role of inter-subjective traditions as interpretive resources’ (Hay, 2011: 176). This ‘structuring’ role for inter-subjective meanings leads Hay (2011: 177) to suggest the workability of an ‘as if real’ ontology for interpretivism. As consistent beliefs and/or practices are real, the ‘structure’ or ‘context’ they present through habituation can appear ‘as if’ real, influencing individual conduct. Individual practices and action can, in turn, affect that context. It is dialectical. Bevir and Rhodes, like Hay, consider ideational contexts as historically contingent and temporal. Yet in the case of Bevir and Rhodes, tradition is a beginning. For Hay (2018: 102), it forms a part of an actor’s context which cannot be ‘willed away’.. On the basis of ‘situated agency’, one can see how traditions can be an ‘enabler’ of action (I am not acting alone, or in isolation, but as part of a shared tradition) and a constraint on action (my beliefs are not those of the majority of my colleagues, but I have to work with them). As Robinson (2016: 121) argued, traditions are ‘a way of understanding the world and of orienting one’s own position in relation to it – whether as an upholder, reclaimer, challenger or destroyer’.

It is here where the ‘synergies’ with constructivist institutionalism are present. For if traditions are constitutive both of an actor’s beliefs and of their institutional context, then as with constructivist approaches

actors are strategic, seeking to realise certain complex, contingent, and constantly changing goals. They do so in a context which favours certain strategies over others and must rely upon perceptions of that context which are at best incomplete and which may very often prove to have been inaccurate after the event (Hay, 2008: 63–64).

This is, of course, a distinctive ‘new’ institutionalism – one that considers a more structural, normative/sociological institutionalism, where rules are ‘followed because they are seen as natural, rightful, expected and legitimate’ (March and Olsen, 2008: 7) to be weighted far too greatly towards institutional stasis (Hay, 2008: 60). Instead, path dependency and the historically contingent nature of institutions – and inherited traditions – are partnered with the ‘identification of moments of path-shaping institutional change’ (Hay, 2008: 65), when the relationship between actor and context (both internal and external) is the focus in explaining and understanding political change. Thus, constructivist institutionalism ‘emphasises institutional innovation, dynamism, and transformation, as well as the need for a consideration of processes of change over a significant period of time’ (Hay, 2008: 65).

‘The Many Bound Together’: Power and Traditions

Bevir and Rhodes (2018: 16) argued – understandably in light of an interpretivist or constructivist ontology – that theorists should reject as reifications ‘concepts of power that refer to social relations based on the allegedly given interests of classes or other social groups’. Instead, interpretive political science involved an understanding of power as the way ‘historical backgrounds impact on individuals, influencing their subjectivity and their actions. Power refers here to the constitutive role played by tradition in giving people their beliefs and actions, and so making the political world’ (Bevir and Rhodes, 2018: 16). Furthermore:

power can refer to the restrictive consequences of the actions [my emphasis] of others in defining what people can and cannot do. Restrictive power works across intricate webs . . . in these terms, interpretive political science may show how various actors restrict what others can do in ways that thwart the intentions of policy actors (Bevir and Rhodes, 2018: 16).

These are important points about power, tradition and its constitutive role – the starting point for the beliefs and practices of political actors, and so for the political world more widely (being patterns of the beliefs and practices of individual actors). Yet it remains an open question, when theorising the relationship between power and tradition within an interpretive framework, as to why ‘particular narratives gain traction, why certain traditions . . . come to seem natural and reducible to “common sense”’ (Robinson, 2016: 123).

Such scenarios are present in the work of Bevir and Rhodes (2018: 23), for example, with the suggestion that ‘street-level police officers are often influenced by an organisational tradition that stresses combating crime even when the new police commissioner wants to ginger up the troops by promoting community policing’. What are the power dynamics at play here? We have what appears to be a tradition with enhanced institutional resonance – that of ‘crime-fighting’ – comprised of associated beliefs and practices, influencing the actions of individual officers despite a competing tradition – that of ‘neighbourhood relations’ – being pushed by senior officials. Why does this ‘organisational tradition’ have enhanced resonance? Here, we must seek to understand how the actions of people have contributed to the institutional authority of meanings and practices. In other words, if we are considering power, we need to consider how actors have – through persuasion, argument, organisation and other actions – contributed to particular stories and meanings achieving enhanced resonance. Hannah Arendt’s work on power, in both her essay On Violence and the book The Human Condition, helps greatly in underpinning a simple, though important point about power and traditions: they rise or fall depending on the support lent to them by people and the subsequent effort of legitimation through shared (inter-subjective) meanings. ‘If sovereignty is in the realm of action and human affairs what mastership is in the realm of making and the world of things’, Arendt (2018: 245) argued, ‘then their chief distinction is that the one can only be achieved by the many bound together, whereas the other is conceivable only in isolation’.

Put differently, ‘all political institutions are manifestations and materialisations of power; they petrify and decay as soon as the living power of the people ceases to uphold them’ (Arendt, 1970: 41). This rather people-centred conception of power was disregarded in the ‘community power debate’. Lukes (2005: 34) criticised Arendt’s focus on the ‘locution “power to,” ignoring “power over,”’ yet the structuralism of the third and fourth dimensions of power is less useful for work which, in the language of Bevir and Rhodes, seeks to decentre concepts like tradition. If one accepts, as interpretive political scientists must, that actions are contingent on beliefs and that those beliefs are shaped, but not restricted by an ideational background, then power itself is contingent upon beliefs and the people who hold them. Arendt’s understanding of power is most fitting. We can understand a tradition in terms of power relations, becoming ever more present in an actor’s context when ‘people get together and act in concert’ (Arendt, 1970: 52) from the point of coming together to the ongoing ideational maintenance of a tradition.

Acting together, or acting individually in the knowledge that one is acting in conjunction with others, is vital to understanding the power of groups who uphold traditions. Traditions are a part of the social world. People are socialised into traditions. To ‘change’ a tradition, to ‘reinvent’ tradition, is a social act which in turn requires significant intervention within the social world. Michael Oakeshott’s articulation of ‘technical’ and ‘practical’ knowledge is instructive here. ‘Practical’ knowledge is less susceptible to change through precise tinkering with rules (Oakeshott, 1991: 12) – something that, for example, the leadership of an organisation, perhaps even one person, could swiftly change. Instead, ‘practical’ knowledge exists in ‘method’ (or practice) and becomes akin to a ‘common knowledge’ (Oakeshott, 1991: 12). This is very reminiscent of the work of Henry Drucker (1979: 1), who when studying the British Labour Party’s ‘ethos’ separated ‘doctrine’ from the party’s traditions, beliefs and practices. A process of socialisation means that these beliefs and practices are embodied by people, unlike the detail of policy programmes, and thus their ‘power’ is found within those people. The argument from Arendt on theories of power, and our understanding of the type of ‘knowledge’ tradition represents from Oakeshott, is brought together in a similar way by Berger and Luckmann (1975: 27) in The Social Construction of Reality. The authors were concerned with a research agenda that examined ‘what people ‘know’ as ‘reality’ in their everyday’. What people ‘know’ comes from the social world, they are socialised in it, and ‘what is more, I know that others share at least part of this knowledge, and they know that I know this. My interaction with others in everyday life is, therefore, constantly affected by our common participation’ in this knowledge (Berger and Luckmann, 1975: 56). This helps people orient themselves (Berger and Luckmann, 1975: 59).

There are two points I wish to highlight from Berger and Luckmann’s work which are particularly relevant to my argument regarding the role of tradition in political analysis. The first is how people perceive patterns of beliefs and practices in the social world – how ‘real’ they seem. Knowledge, Berger and Luckmann (1975: 76) argued, can be experienced by people as something

existing over and beyond the individuals who ‘happen to’ embody them at the moment. In other words, the institutions are now experienced as possessing a reality of their own, a reality that confronts the individual as an external and coercive fact.

Their example is children and the parental home, or ‘institution’, where a world of parental authority and routine ‘attains a firmness in consciousness’ (Berger and Luckmann, 1975: 77) because that is the way things are and the way things are done. Note, however, the use of ‘experienced as’ by Berger and Luckmann. Ontologically, this ‘firmness’ remains a human activity, contingent on human beings. The relationship between a person and the social world is dialectical (Berger and Luckmann, 1975: 78). There is no status to institutions independent of human activity (Berger and Luckmann, 1975: 107). The second important point from Berger and Luckmann returns to power. Social knowledge, they argued, required legitimation, the ‘ascribing [of] cognitive validity to its objectivated meanings’ (Berger and Luckmann, 1975: 11). Legitimation, particularly at high levels of social knowledge seeking to ‘explain’ ever greater depths of human experience, required ‘conceptual machineries’ to continue, and the ‘success of particular conceptual machineries is related to the power possessed by those who operate them’ (Berger and Luckmann, 1975: 126).

Conflicting or competing forms of social knowledge, therefore, ‘implies a problem of power – which of the conflicting definitions will be “made to stick” in the society’ (Berger and Luckmann, 1975: 126). Power, in Berger and Luckmann’s (1975: 137) argument, ‘includes the power to determine socialisation processes’. When understood alongside Arendt’s theory of power – that of the many bound together bestowing legitimacy – we have the beginnings of an understanding of the varied power relations behind competing traditions. Traditions, formed by groups of people in institutional settings and competing with other traditions, are products of power (a group acting together) and ongoing tools and strategies for legitimation (ideational maintenance). Their viability and efficacy can – in part – be understood through power relations and contestation within institutional contexts. The power lies in the initial grouping together – the forging of shared meaning – and the continued legitimation of those shared meanings by the upholders of a tradition. These insights are very relevant to considering traditions to be ‘as if real’. For with traditions, not only are the beliefs and practices themselves real, but the collective nature of them, the habituation and passing on from one generation to the next, adds to the sense that within institutional contexts traditions appear ‘as if real’. And while they are contingent on individuals, the chances of traditions attaining resonance are contingent upon groups of people – something which could be termed ‘socially contingent’.

‘Accents of Menace’ – Traditions and Division within 1980s Labour

As I noted above, tradition as a concept in political analysis is under-theorised. As such, the role of tradition in the literature on the Labour Party’s ‘modernisation’ period is often static rather than dynamic. Traditions are understood as either baggage to be dumped or an organisational/collective encumbrance, often on the basis of a rationalist assumption for why political parties change. ⁴ They are presented as ‘old’ views that can be discarded or attacked by actors (Chadwick and Heffernan, 2003; Cronin, 2004; Heffernan, 2000) or considered to be more sociological/normative, where a ‘mentality’ must be overcome (Cronin, 2004; Panitch and Leys, 2001). An interpretive approach to Labour’s modernisation would be to ‘decentre’, to

focus on the social construction of a practice through the ability of individuals to create and act on meanings. It is to unpack a practice as the disparate and contingent beliefs and actions of individuals . . . Thus, decentred theory involves challenging the idea that inexorable or impersonal forces, norms, or laws define patterns and regularities in politics (Bevir and Rhodes, 2012: 73).

Through a decentred approach, traditions are understood not as packages of ‘old things’, but ideational backgrounds that root the historically contingent beliefs held by actors. The analysis which follows builds on this approach. I use the theory of tradition I have set out above to show the importance of a contestation of meanings to a process of political change – relevant to the motivations of actors, their strategic calculations and the outcomes we can observe (Hay, 2002). With a particular focus on the question of power relations and institutional resonance, I show how traditions are key to understanding the dynamic relationship between strategic actors and their institutional contexts. Understood in this way, Labour’s ‘modernisation’ – and the party’s ‘zig-zagging’ political trajectory during the 1980s and 1990s – was not something ‘done’ to an ‘old’ party by the ‘new’, but a period of political change greatly affected by competing traditions. ⁵

Neil Kinnock became Labour’s leader early after his own political journey from a parliamentary ‘trouble maker’ on the Labour left to a moderate centre-left MP who resisted the power of Tony Benn – the charismatic, prolific talisman of the Labour left who, despite being initially positive of Kinnock, dismissed him as a careerist (Pike, 2019). The journey for Kinnock did not end after becoming leader. He had been elected in 1983 as the candidate of the ‘soft left’ against Roy Hattersley of the Labour right. He had committed himself to the ideas and policies the Labour left had fought hard to achieve in Labour’s programme from the late 1970s to 1983. Yet by the beginning of the 1990s, following a ‘policy review’, the left’s policy gains had gone. Kinnock had revolutionised the party’s campaigns and communications infrastructure, enlisted the help of advertising and polling professionals, and pursued the hard left ‘Militant Tendency’ with unrelenting energy. By the time he resigned following Labour’s 1992 election defeat, many on the Labour left felt angry and betrayed by him. ‘Anything and everything was sacrificed to so-called “electability”’, the left-wing Labour MP Diane Abbott (1992) wrote for The Guardian after Kinnock’s resignation. He was, Abbott (1992) claimed, ‘vindictive against his erstwhile left-wing colleagues’.

Yet, Kinnock also proceeded as leader with a strategy of incremental change, conscious that any move deemed to be too sudden would look unprincipled and illegitimate. In some areas of policy – nuclear weapons, for example – Kinnock retained commitments he was told would lose Labour public support, partly on the basis of his own beliefs and partly because he did not believe any change was possible without serious discord. His denunciation of Thatcherism – repeated vociferously – was based on a firm rejection of Conservative politics. He was aggressively attacked by sections of the press, deliberately embarrassed by the Thatcher-friendly US administration of Ronald Reagan and presented by the Conservative Party as a big state, high tax, weak on defence Labour ‘lefty’. A different kind of moderniser emerged at this time, with Peter Mandelson (2010: 104) reflecting that Kinnock was ‘too much of a socialist . . . [who] hates the idea of being seen by the party as anything different’. Kinnock became a figure somewhat disquieted by what followed him: New Labour and Tony Blair’s three terms in office. While Kinnock was a leader known for his pragmatism, and the singular goal of winning an election, the direction Blair took in government – particularly in his second and third terms – was not the culmination of Kinnock’s modernisation, but a different kind of modernisation.

These are two very different takes on Neil Kinnock, from groups of actors who uphold competing traditions within the Labour Party. Importantly, though, they not only represent appraisals of Kinnock’s period as leader, but snapshots of the narratives Kinnock perceived and heard from competing Labour traditions while he was leader. Kinnock’s ‘story’ as leader is wrought by the painful compromises and very public rows of a process of party change – and the narratives that emerged contain rich material for understanding how traditions constitute both political actors and their context. In a 1993 Institute of Historical Research seminar, a paper given by Kinnock summarised his view of the 1983 Labour Party. ‘Labour was increasingly seen to be a party slipping towards impossibilism’, Kinnock (1994: 535) wrote, ‘succumbing to fads, riven by vicious divisions, speaking the language of sloganized dogma – and usually voicing it in the accents of menace’. It was as if, he wrote,

sections of the party measured the purity of their socialism by the distance which they could put between it and the minds of the British people. These characteristics . . . were not, of course, typical of the great majority of party members. But it is an inevitability of politics that the nature of a party is judged not so much by the modulated voice of the many as by the braying of the few (Kinnock, 1994: 535).

It is clear from Kinnock’s remarks that while he did not believe the majority of Labour party members subscribed to what he called ‘impossibilism’ – read ‘Bennism’ and narratives resisting modernisation – there remained a powerful tradition within Labour’s ranks that received attention and would fight many of the decisions he took. Importantly, Kinnock perceived this power as emanating from particularly committed activists in positions of authority.

Labour’s institutional structures, from local branches upwards, were – he believed – filled with those reluctant to embrace change. Kinnock (1994: 537) noted:

I was aware from wide personal contact that there was a body of opinion in the Labour Party that, in the wake of the defeat of 1983, would either embrace change eagerly or – at worst – give it the benefit of the doubt.

However, this

asset was not . . . readily available. Its supply was blocked to some extent by those who thought of themselves as guardians of the soul of Labour. Many of the people, sitting on General Committees and other decision-making bodies had armoured themselves against public opinion and changing realities and were constantly on the look-out for what they considered to be ‘deviation’ . . . Some could, as time passed, be persuaded by argument and they were. Others were going to have to be superseded by the more general realism of party members (Kinnock, 1994: 537).

Kinnock viewed two competing traditions here, at least in terms of attitudes to policy change. They can be understood as the expressive and instrumental traditions and are a core tension of Labour Party politics (Parkin, 1968: 35). Instrumental politics places less emphasis on ‘gestures felt to be morally right’, while expressive politics places greater emphasis on ‘the defence of principles’ (Parkin, 1968: 36). Labour Party people rarely take an absolutist position, despite some of the characterisations of expressive politics that tend to come from the Labour right (‘the politics of protest’) and of instrumental politics that tend to come from the Labour left (the ‘abandonment of principle’). Instead, Labour Party people have a different interpretation of the blend of these traditions, giving rise to more expressive or instrumental political styles.

Perception is key, and perception is – of course – an imperfect barometer of party traditions. Kinnock (1994: 539) ‘began a series of regular meetings with leading trade unionists’ as well as

periodic regional meetings with ordinary members of the party . . . they were conducted as question and answer sessions and, although the attendances always ran into hundreds, none of the confidences that were frankly – and sometimes acrimoniously – exchanged in such meetings, over a period of nearly eight years, ever became public.

From these meetings, Kinnock would himself detect the mood and feelings of his fellow members, gauging levels of support or discomfort. How much change, and when, was a strategy in a state of constant calibration, with information feeding the judgements of Kinnock and his team? This was not only a matter of Kinnock’s beliefs here, and the tradition he represented, but the relative resonance of competing traditions too. He was ‘reading’ these beliefs, imperfectly but often, in what was an ongoing contestation of competing Labour traditions. He hoped some fellow party members would change their views, but also believed that some would simply have to be superseded by a larger body of people who would give him the support he needed – in other words, that the balance of power would change, tipped a little more towards the ‘instrumental’ and pragmatic.

In addition to understanding how Kinnock perceived competing traditions within Labour, we can also consider the beliefs and practices Kinnock viewed himself to embody and the meanings he assigned to his own actions. His then deputy, Roy Hattersley (2003: 273–274), reflected that while Kinnock wrestled with electoral viability and long-standing policy commitments, ‘from the start he was determined to dispose of what he called the “illegitimate left”’. While Kinnock generated a reputation for ‘taking on the left’, he did so very much from a basis he believed to be more ‘traditionally Labour’. To Kinnock, challenging the ‘hard’ or ‘extreme’ left was not challenging Labour’s ‘legitimate’ traditions, it was restoring them. In a revealing preface to a popular history of the Tolpuddle Martyrs, Kinnock (1985: 9) wrote in 1985 that the

six Dorset labourers now honoured by democrats everywhere did not have their eye on posterity. They were profoundly moderate and pitifully honest. They did not seek martyrdom or self-glorification . . . It is the simplicity of their case and the propriety and patience with which they put it as much as the elementary justice of their demands which has given strength and resonance to their message down the years.

‘Patience’, ‘honesty’, ‘profoundly moderate’. To decentre Kinnock’s reform agenda is to find beliefs and practices which he felt were legitimated by the history of the British working classes, and indeed the communities of the South Wales valleys in which he grew up – traditions which partly constituted his beliefs. To ‘be Labour’ came with a significant dose of local, practical electoral politics – hard campaigning work, elections and marginal progress. These were the beliefs Kinnock himself sought to legitimate through his rhetoric and political strategy. They are not too dissimilar to those associated with ‘Labourism’ (Marquand, 1999; Shaw, 2004), although with a decentred approach we can better understand the individual beliefs and meanings Kinnock attached to his actions, rather than applying a more ‘catch-all’ term.

Conclusion

This article has sought to answer the question about why some traditions ‘stick’. Building on the ontological insight of ‘as if realism’, I have argued that some traditions cannot be ignored, and that from an initial moment of legitimation by a group of actors, traditions achieve a level of social or institutional authority that actors perceive and are affected by. This authority is subject to maintenance and enhancement by upholders of traditions – something I term ‘socially contingent’ or contingent upon the actions of a group. Traditions present a ‘right way’ of doing things. They are, on that basis, powerful ideational concepts – not simply collections of ‘old things’. Different traditions have different power relations behind them and are contingent upon the groups of people who uphold them. As such, traditions are partly constitutive of both an actor’s beliefs and of their institutional context. They can exert a powerful pull to political actors not only as orientation tools in complex social settings, but also through the persuasive arguments, symbolism and meanings that are attached to traditions by those who uphold them. While traditions are entirely contingent on people’s beliefs, it is ‘as if’ they have a life of their own. Of course, actors are ‘not analytically substitutable’, nor are their preferences simply ‘derived’ from an institutional setting (Hay, 2008: 64). As such, I contend only that an explanation of political change requires comprehension of the beliefs and strategic considerations of the actors involved. Both of these are affected by traditions – but those affects are both historically contingent and shaped by the actions of those who uphold traditions.

As such, the role of traditions in political analysis will never be uniform. In different contexts, the relative importance assigned to contestation between competing traditions, and the amount of effort actors make to maintain traditions, will vary. For example, a focus on the contestation of competing traditions – which I applied in the example of Labour’s modernisation – feels particularly suited to political parties, but is perhaps less important for traditions within institutions like legislatures or the courts. The composition of the traditions of a nation, perhaps consistent over many generations, may vary greatly to traditions within communities or families. It is useful to reflect here on Blyth’s (2003: 701) argument that there can be different ‘levels’ of analysis for different ideational concepts, for example the ‘social imaginaries’ and ‘large-scale worlds of meaning’ which can be understood as operating ‘above’ the level of organisational or institutional traditions (Bevir and Blakely, 2018: 146). In addition to different ‘levels’ of ideational concepts, there are complementary analytical tools when analysing tradition too, as with myth which I discussed above (Samuel and Thompson, 1990).

What does the analysis – both the theoretical contribution and the substantive example – offer to interpretivism, the study of Labour and other political parties, and political analysis more generally? For interpretive political science, the theory of tradition I presented here both offers an answer to why some traditions acquire enhanced resonance (contingent upon power and ideational maintenance) and shows how exploring synergies with complementary analytical approaches can offer ways forward in ontological debates. Tradition as a concept, I have argued, is well suited to attempting a synthesis between different approaches to political science, and in particular to elucidating an ‘as if real’ ontological stance. Such an approach offers a philosophical basis for exploring how a concept can simultaneously constitute both actors and their contexts. When applied to a political party, this greater specificity for what tradition is ‘doing’ allows for a clearer focus on the contestation of different meanings at the heart of political change – helping to explain the beliefs, motivations, perceptions and strategies of situated political actors. Such an interpretive framework also necessitates a broader understanding of the ‘ideational’ as an ‘explanatory strategy’ (Randall, 2003: 10–11). In the case of the 1980s, the current literature has either underplayed the role of the ideational or restricted it too narrowly to the ideational as doctrinal – for example, ideological paradigm or policy idea.

For political parties and political analysis, traditions are of clear importance to literature on party change. How long political parties take to ‘get the message’ from successive election defeats (Bale, 2016), for example, cannot be understood fully without considering the patterns of beliefs and practices that have achieved resonance over time. The significance of traditions means a potentially lengthy contestation between competing beliefs and practices cannot be ruled out when political actors are faced with an electoral ‘dilemma’. With the Labour Party, after successive election defeats a large body of actors made the case that the party needed to change across a number of areas. Yet it wasn’t and isn’t ‘obvious’ that this ‘factor’ – electoral defeat – should have necessarily triggered such a reaction across the entire Labour movement. Why? Because some actors held to competing beliefs and practices. Traditions reveal the diversity of a party’s beliefs. And, far from traditions being ‘transformed’ or ditched by Labour’s modernisers, some have found greater resonance in more recent times. Traditions do not easily ‘disappear’ – whether or not they do can be explained through their ‘social contingency’.

My final point comes back to political identity. Traditions are relevant to what has been termed the ‘partisan claim’ of political parties (White and Ypi, 2016: 213). Undoubtedly, political actors can unite, often passionately, behind a particular policy or programme. Yet, the beliefs and meanings which those policies represent are of deep importance to that unity and to the coherence of a political movement more widely. Traditions, I have argued, partly constitute identity and position an actor as being part of ‘something bigger’. Individual policies do not do this. When we critique political parties today, including expressions of doubt about ‘what they believe’, rarely does any perceived failing come down to an absence of policy solutions in areas such as health, education, or law and order. Instead, we perceive an absence of an underlying force. Resonant traditions which bind partisans together, constituting political identities and providing linkages between ‘politics from below’ and ‘high politics’ are a part of what we perceive as absent. Social democratic parties, for example, may well have policies that would boost low-income households, yet shorn of resonant traditions which – from origin – speak to the representation of the working classes, the strength of the party’s ‘partisan claim’ may be weaker. One could fruitfully ask, therefore, whether ‘technocratic’ traditions can speak to a party’s partisan identity as effectively as those with a more ‘ideological’ worldview.