Relevancia teórica y empírica de los documentos en la sociología del deporte guracional Documents and detachment in the gurational sociology of sport

This article discusses the methodological implications of using a #gurational approach to the study of Gaelic games (the Irish sports of hurling and Gaelic football) and their organisation and governance from the late nineteenth century onwards. Following Norbert Elias, we argue that data generation guided by theories, and potentially modifying such theories, are essential components of research endeavours. The problem of generalisation or representativeness of historical documents can be addressed by treating documents as part of a #guration of evidence, thereby constituting a mutually supporting network of data that serves as a scaffold or structure for theoretical and empirical #ndings. Based on the experience of using documents such as newspaper and referee reports of matches, and also organisational documents such as the minutes of meetings at various levels of governance, we contribute to the discussion concerning borders and boundary maintenance between history, sociology and historical sociology. While we elaborate on the Eliasian perspective on emotional detachment, and secondary involvement, when conducting research, we also note the dif#culties in avoiding ‘preferred’ readings of national history given that Gaelic games have been intertwined with the narrative and project of national self-determination in Ireland. Researchers are compelled to recognise the theoretical signi#cance of sports as a motor and symbol of emotional identi#cations, but must also create distance from their own national habitus in order to prioritise participation in intergenerational knowledge generation over further national myth-making.


INTRODUCTION
The purpose of this paper is to discuss the role of methods in gurational sociology. We do this by interrogating the methods used in our recent analyses of the development of various aspects of sport under the auspices of the Gaelic Athletic Association (GAA, established 1884) in Ireland (Dolan and Connolly, 2009, Connolly and Dolan, 2010, Dolan and Connolly, 2014, Connolly and Dolan, 2011, Connolly and Dolan, 2013a, Connolly and Dolan, 2013b, Connolly and Dolan, 2012. We have focussed on the sports of hurling and Gaelic football, both standardised and codi ed in the late nineteenth century, though an ancient indigenous pedigree is often claimed for hurling in particular. Both sports are team eld games involving fteen against fteen. As these studies have taken the form of journal articles, there has been little scope for extended methodological discussions. The articulation of methods adopted and adapted for any study are clearly relevant for the status of any consequent explanations, but, following Elias, we are reluctant to espouse a speci c methodology that others should follow in the pursuit of knowledge. As Dunning (Dunning and Hughes, 2013: 175) recalls in a note Elias wrote in response to an early draft of the preface to Quest for Excitement, Elias refused to state a particular methodology to follow; rather he used methods in the plural depending on the research questions pursued. It is clear enough from his body of work that he used what would be called 'mixed methods' in the current methodological parlance, but as his concern was generally to analyse and synthesise various social processes over time, the use of historical documents were indispensable. His implicit (through occasionally explicit) mistrust of highly statistical inquiries sprang from their tendency to construct variables whose meaning, function and direction were often stripped of context (in terms of social situations), culture (in terms of historically-speci c meaning), and structure (in terms of the interrelated functions and meanings of processes, events and part-structures that can be understood only as parts of an integrating whole).
The structure of the paper begins with our account of the theory-guided necessity of data generation and analysis. In our case we have attempted to use (though not, we hope, in an hagiographic way) the ideas and theories of Norbert Elias, and his followers such as Eric Dunning, in order to explain the development of sporting organisation, the emotional control of players and spectators (by themselves and by others), and the changing constitution of sports games themselves, all within the context of broader changes in social structures. Throughout this research we have been mostly reliant on the use of historical documents in forms such as newspaper accounts of matches, referee reports, and minutes from organisational meetings, that relate to the organisation and playing of Gaelic games in Ireland from the 1880s to the present. Though we have also generated and used qualitative interviews (see especially, Connolly and Dolan, 2012), they are not our main concern in this paper. The next section examines the relative focus and eld of inquiry of the related social science disciplines, Sociology and History, and also the emergent subdiscipline of Historical Sociology that in many respects constitutes a sequestration of historical perspectives from mainstream sociological concerns. Then we examine the key issue of emotional involvement on the part of researchers (often depicted as re exivity in the methodological literature) and indeed the people portrayed in our research ndings. Finally we deal with the status of documents in the construction of actual social processes. Throughout this we attempt to relate these 'methodological' concerns with our actual research practices in developing and publishing our ndings and analyses. DOCUMENTS AND DETACHMENT..

GUIDANCE BY THEORY
In order to understand various sporting practices and processes, including spectator and player violence (though not considered 'sporting' in the moral sense, such violence did occur in the spaces where games were played), we nd it necessary to employ theoretical frameworks in order to recognise the relevance of data that might relate to such processes and practices. However, we are not suggesting that theories should remain sacrosanct in the face of evidence derived from historical documents and other sources. Elias continually stresses the importance of developing 'testable' models and explanations; in other words, data from other social contexts and situations could be used to revise the prevailing models as long as such revisions accounted for the evidence (facts) in such diverse contexts. In that way, we have sought to use gurational theory as a resource to explain Gaelic games in Ireland, and insofar as the data derived from that particular national context suggests modi cations of gurational theories as they pertain to sport or the contexts of its production and change over time. One of the key themes in gurational sociology is interdependence, and this applies as much to the research process as it does to relations between people: Without the constant interdependence of the development of bodies of detailed knowledge and synthesizing models, of empirical exploration and theory, analysis and synthesis, the production of detailed knowledge, no matter how careful the method of production, remains uncertain, often enough misleading and scienti cally irrelevant. (Elias, 1997: 367-8) So our approach towards data generation was already shaped by our knowledge of figurational theory. Naturally researchers choose theoretical resources and models that they nd most compelling. This is a choice structured by the career trajectory of the researcher, so it is in fact a choice limited by opportunity, combining planned and unplanned aspects. The rst author was introduced to the work of Elias while studying for a postgraduate sociology degree; the second author was introduced to his work through conversations with the rst author. This of course is an example of social interdependence; we depend on others over the course of our lives, and such dependencies are not entirely controlled by the protagonists. By guration Elias (2012a: 525) means 'a structure of mutually orientated and dependent people … the network of interdependences formed by individuals'. The nuances and subtleties of gurational sociology are beyond the scope of this paper, but it is appropriate to suggest some key aspects which informed the theoretical lens through which we searched for and made sense of data. Firstly, we recognised that Gaelic games developed into more formal encounters during the late nineteenth century. The rst author had already used Elias's theories to examine the development of consumer culture in Ireland over the course of the twentieth century (Dolan, 2005, Dolan, 2009a, Dolan, 2009b. The choice of a gurational approach for that study was not immediate or automatic, but followed dissatisfaction with an earlier inclination towards a Foucauldian analysis. This concern stemmed from the unsuitability of Foucault's archaeological or genealogical mode of analysis to connect different normative frameworks surrounding consumer culture. As Burke (1992: 151) notes, Foucault's work demonstrates a 'failure to discuss the mechanics of change'.
Elias, on the other hand, stresses the sequential order of social change without resorting to teleology or notions of progress. The dynamic of change consists of the unstable tensions between contending groups and individuals comprising any particular social unit, and indeed tensions between such units at higher levels of integration or interdependence. As actions have unforeseen and unintended consequences, the order of change is largely unplanned, but the relative contribution of planned and unplanned aspects of social change depends upon the power balances between the various groups and individuals. In the context of Irish sport, it was clear that the formation of the Gaelic Athletic Association in 1884 was related to broader changes involving the rise of cultural nationalism in Ireland during this late nineteenth-century period (Mandle, 1987). Of course as researchers we are reliant on secondary historiographic literature for this knowledge. As such, we exist in a network of interdependences with other researchers, in this case historians who have collected and analysed primary documents concerning organisational formation and change, at the level of both states and sporting organisations.
Broadly speaking, Elias argues that as people become more constrained and enabled by denser, more diverse, longer, and multiplying webs of interdependence, they must learn and exercise more even and all-round selfsteering capacities to navigate this complex social terrain. This occurs over many generations, and new standards of social conduct and new symbolic repertoires develop to re ect and also to channel these social changes. Part of these greater requirements for more exible and even self-steering included the capacity to control (not simply constrain, but to manage in a more re ned way) emotions in dealing with others. This newer concern for emotional self-control combined with decreasing social inequality between classes and genders led to the development of a conception of self oriented to what Elias (2010) labelled homo clausus (closed person). People increasingly felt closed off or separate from others, as if some barrier kept them apart. Though this is an inaccurate perception of social reality, it was based on a real feeling of personal separation, and therefore was real in its consequences. In fact Elias argues that this perceived duality is the basis for much epistemological theory in the philosophy of science and knowledge (Elias, 2007b, Elias, 2007a, Elias, 2012b, Kilminster, 2007. According to Elias, the 'subject' of knowledge does not rst deduce hypotheses or conclusions from laws in his or her imagination or 'mind' and then seek support in the 'objective', 'external' world. Nor does the researcher simply observe facts devoid of prior synthetic re ection, and then attempt to link observations together to formulate theory. There is a constant 'two-way traf c' between two layers of knowledge: that of general ideas, theories or models and that of observations and perceptions of speci c events' (Elias, 2007b: 89). Historical research in particular is sometimes conceived as free from theory, DOCUMENTS AND DETACHMENT.. EMPIRIA. Revista de Metodología de Ciencias Sociales. N. o 30, enero-abril, 2015, pp. 33-52. ISSN: 1139-5737, DOI/empiria.30.2015 following E.P. Thompson's famous treatise The Poverty of Theory: But despite an ideology of 'nothing but the facts,' historians seldom engage in radical inductivism. They do not simply walk arbitrarily into archives and demand documents, storing up facts as the experience dictates. They seek to test existing claims and extend understanding to new (or hitherto unstudied) events. They treat studies of specific towns and social movements, fertility rates and royal succession crises as cases to be compared as well as appreciated in their specificity. They have theories, even when they choose to express them implicitly rather than explicitly. Above all, in light of the present debate, they write narratives that are hardly exercises in induction or indeed in pure empiricism. (Calhoun, 1998: 855) Tough we do not self-identify as 'historians', we engaged with newspaper and organisational archives informed and guided by an existing theoretical framework, or what Elias often refers to as a 'fund of social knowledge' passed down intergenerationally producing a chain of interdependence linking researchers over many decades and even centuries (see Gabriel and Mennell, 2011). When we look at relations between players on the eld, or between players and spectators, and between team managers and organisational functionaries within the Gaelic Athletic Association we have the concepts of guration and habitus at our disposal. The dif cult challenge for following generations of researchers, and particularly those following a sociologist of the calibre of Norbert Elias, is not to treat such concepts with 'too much respect', to use a sporting analogy. We endeavour to interrogate theories and concepts against found and emerging evidence, but we could not even recognise evidence without the symbolic repertoire of concepts that other social scientists have bestowed on the research community. To facilitate this, we endeavour to avoid seeing the social world as if from the position of homo clausus, and instead recognise the changing and multiple forms of connections between many people over time and place. In other words, gurational researchers, whether investigating sports or other social practices, attempt to bypass the dualistic tendencies of philosophical theories of knowledge (see Kilminster (1998Kilminster ( , 2007.
In the next section, we address the vexed issue of using history for sociological inquiry, and indeed the value of the boundaries between the disciplines of history and sociology. One apparent solution of these border disputes has been the emergence of a relatively new subdiscipline titled historical sociology.

HISTORY, SOCIOLOGY AND HISTORICAL SOCIOLOGY
Though we use historical evidence in our examination of Gaelic games in Ireland, we do not adopt a traditional historiographic strategy: 'The sociologist's history is not the historian's history' (Elias in Elias and Dunning, 2008: 152). By this we mean that we are less concerned with the imagined inventors and creators of Irish sports and sporting organisations. This is not to say that such people were not important in the development and popularisation of the sports, but that their actions and orientations were framed and enabled in the complex network of social relations which they formed with others. This guration included not just those they interacted with face-to-face but also distant others, and also previous generations who inherited, upheld and modi ed the traditions of Irishness, and in particular the ideal of cultural distinction. It is only in this social context that leaders of the new sports movement could even conceive of the need for sports distinct from British sports that were proving to be enormously successful across the social classes in Britain, Ireland and beyond. In that sense, as Elias argues, sport has a relative, but limited, autonomy (Elias in Elias and Dunning, 2008: 158). Once people conceive of new sports, they generally coalesce in the form of an organisation to frame rules, set standards of play and adjudication, develop competitions at various levels (in terms of age, gender, ability and geographic region), and to co-ordinate the activities of the organisation itself as it grows and changes in partly unplanned ways (by virtue of the growing numbers of people with different interests and priorities that become involved in the sports and their organisation). So while historians of Gaelic games in Ireland have tended to focus on origins, narrative, and key individuals in the development of sports, we have tried to focus on the changing structure of relations between Irish and British social groups. This includes relations between sporting organisations more or less loyal to central sporting organisations in the United Kingdom (as it was then in the late nineteenth century). While gurational sociologists do not dismiss the signi cance of important historical gures, they are placed in the uid context of power relations between groups, including social classes and ethnic groups. Such individuals are formed in the uid intersections between many social processes and relations over time. In terms of the Gaelic Athletic Association, the speci c actions of Michael Cusack (usually recognised as the key founder of the GAA) become less important than shifting power relations within Ireland and between Irish and British groups. This does not diminish his uniqueness, in that the complex, changing gurations comprising millions of people allow for the formation of differentiated people depending on their initial position and subsequent trajectories through these gurations. Though history and sociology as disciplines have sought to maintain their independence, their objects of investigation are more or less the same -societies (Elias, 1997: 363-4). Elias suggests that the division of history and sociology as 'independent subject areas' is a function of their organisational demarcation in universities, and therefore a myth, though in earlier work (The Court Society) he presents a clearer distinction between these disciplinary perspectives: 'The former [historical perspective] throws light on particular individuals, in this case individual kings, while the latter [sociological perspective] illuminates social positions, in this case the development of the royal position' (Elias, 2006: 4 rather had, in absolute monarchies) greater scope for decision making, but this is by virtue of their relative power ratio with other social positions, and the fact that the monarchy as an historically developed social institution brings with it privileges largely irrespective of the current occupant as an individual with speci c strengths, abilities and aws.
Similarly, in our work, we have sought to treat historical gures as occupants of particular social positions within the broader organisational guration of the GAA, which itself exists within national and international gurations. This does not mean that such gures are inactive, passive instruments of inexorable structures colliding with other structures. They have been enabled through their habitus formation and relate to others according to particular sets of emotional dispositions and strategising capabilities. These resources are afforded by their nodal positions and their trajectory within the social networks they form with others. Individual gures with the GAA are also not entirely coherent and consistent in their wishes and deeds, as their locations in gurations at different levels of integration produce contradictions and dilemmas. The same person can be acting on behalf of a small club, a larger provincial unit, the central organisation working at national level, or even within an international context in the form of the recent development of the hybrid sport of International Rules devised to enable games between Ireland and Australia. But it makes little sense to locate the source of action and motivation in the individual him or herself, especially when our concerns have included the cultural signi cance of national identi cation for the development of sport and the fact that these sports are team games. Even individual sportspeople tend to have opponents and must learn the rules of the game from experienced and knowledgeable others. But in team sports, and in relations between players and spectators, it is entirely obvious that people do not act as isolated individuals intent on pursuing their own individually-derived ends. The whole game constitutes an interdependent, uid pattern (Elias and Dunning, 1966). Elias (2006: 9-10) does argue that differences between history and sociology do not constitute absolute opposites. He contends that at least sociology as a discipline has expectations of relating emerging or generated facts to theoretical models or synthetic frameworks that protect, to some degree, against the incursion of present controversies and moralities. Elias (2006: 8, original emphasis) also criticises historians for attributing blame (though they are not the only social scientists open to that particular charge): Ranke pointed to the heart of the problem: the historian apportions praise and blame. … Contemporary circumstances decide how he sees 'history', and even what he sees as 'history'. He makes his selection from the events of the past in the light of what he approves or disapproves in the present.
The personal latitude afforded to individual historians to pursue their interests in events and polities of the past unencumbered by theoretical models has become a badge of honour amongst historians, but this lack of structure in narratives of events can easily serve as propaganda for particular groups (Elias, 1997: 368). These 'extra-scienti c issues of the day' lead to the constant rewriting of history (Elias, 2006: 37).
So, in principle at least, Elias argues that historical evidence should be used by sociologists because they are tasked with explaining how and why people act, think and feel as they do, and all the evidence for this has already happened -it's in the past. The main reason to treat the recent past as the everyday, recurring present is based on the belief that social structures, cultural traditions, values, standards of conduct and emotional displays, self-experience, social contours of belonging and identi cation, and so on, do not really change. As Elias nds this implausible, sociologists must therefore construct their explanations (or theories) to account for such changes in the myriad aspects of societies. For Elias this is not so much a researcher preference as it is a requirement dictated by the nature of the subject matter of sociological inquiry -societies themselves, up to and including global society. Of course change can happen over a relatively short time period so the framing of any research problem in terms of time period depends upon the actual degree of change in the processes, practices and structures of interest. In relation to Gaelic games in Ireland, it became clear early in the project that rule changes were a signi cant factor in their development, and also an indicator of changing standards of emotional control concerning violence and crowd safety. Following Elias, we never sought absolute origins, but attempted to construct processes from the evidence regarding rule changes and organisational development. We also used newspaper reports (see below for further detail) to trace the conduct of both players and spectators over time.
The question of time periods is a dif cult one when drawing on historical evidence. While social processes can initially appear to be going in a more or less consistent direction, closer inspection often revealed quali cations, exceptions, reverses, and at the very least uneven diffusion and acceptance of newer standards of conduct. This is why we have been reluctant to x change to any particular date. This is consistent with Elias's own periodisation of social change. While there are occasional spurts (even civilising spurts) in one direction, change tends to be gradual and uneven. There is a succession of events, but no single great event that serves as an imagined watershed of historical change.
The 'independence' of the disciplines of history and sociology has led, somewhat paradoxically, to the emergence of the subdiscipline of historical sociology. But this development is largely confined to the large research community within North America. Debates on the usefulness of historical sociology centre around the status of theory. Calhoun (1998) rebuts Kiser and Hechter's insistence on the construction of covering laws as the basis for a more scienti c historical sociology: 'Thinking of explanation only in terms of causality, and of theory only in terms of the general theories that generate DOCUMENTS AND DETACHMENT..
EMPIRIA. Revista de Metodología de Ciencias Sociales. N. o 30, enero-abril, 2015, pp. 33-52. ISSN: 1139-5737, DOI/empiria.30.2015 causal explanations, leaves out many important efforts to achieve relatively general understandings of social life' (Calhoun, 1998: 865). The use of narrative explanations are considered important by Calhoun in building explanations of how certain events and processes unfold in particular societies over particular time periods. Elias (2007b: 20) also argues that the emphasis on nding singular causes tends to lead to a misguided search for origins of social change. Searches for scienti c laws also elevate the status of recurring events or conditions that could 'explain' changes in something else, but the natural sciences also deal with 'unrepeatable and unique phenomena' (Elias, 2006: 12).
The concern with finding some order (in the non-normative sense) or structure to social change does not constitute naïve optimism in relation to the future. The concept of 'social development' is treated suspiciously because of connotations of nineteenth century notions of moral progress, or twentieth century reductions of long-term social processes to simple stage models like 'modernisation'. As Elias states: The notion of a development ordained by nature which leads towards a happier coexistence of human beings is entirely alien to me. The progress some of our fathers believed in, universal and automatic progress, does not exist. But forward steps, just like backward steps in certain regards, can be observed. (Elias, 2008b: 10) In our studies of sport in Ireland we have sought to identify order in the apparent discontinuities of social structure. Elias sees the order of change in terms of changing densities and complexities of multiple forms of dependencies between people, which have unintended consequences. Thus, there is continuity through the recurring events and ruptures (Dolan, 2010). This is not the same as arguing that there is a spirit or law guiding diverse changes across many social spheres. Elias does not rely upon some 'base', economic or otherwise, to explain superstructural changes. As a contrast, we can see that Foucault is explicitly suspicious of all claims to historical continuity, on the grounds that it is linked with an 'originating subject of all knowledge and practice' (Foucault, 2000a: 301), and that 'the notion of development' assumes that event sequences must be 'the manifestation of one and the same organizing principle' (p. 302). For Foucault (2000b: 429) , 'History appears then not as a great continuity underneath an apparent discontinuity, but as a tangle of superimposed discontinuities.' But it is precisely this tangle, or intertwined plans and actions of interdependent people over many generations, that constitutes the structure of social change. Once Foucault refuses the challenge of deciphering order then the fragments of history left to us as documents are elevated as a sealed discursive formation. Paradoxically Foucault does occasionally state that the discursive can be related to the non-discursive, but this is left largely unexplained. However, Elias does not rely on a single subject of knowledge, nor does he reduce any social process to a single organising principle. Processes become interdependent, like the monarchical state's capacity to generate income through taxation and the related capacity to fund a standing army and claim the monopoly of physical violence within a particular territory. Each process reinforced the other, but there is no origin of change, no single moment from which the engines of history rst began.
Sport has a relative autonomy through history, but it is not immune to the vagaries of other social processes and institutions. We show how Gaelic games in Ireland, primarily through the standardisation of hurling and Gaelic football rules, were organised on a national scale partly in response to the actions, ambitions and exclusions of other sporting bodies. While this form of national cultural resistance could be conceived as a deliberate plan, such planning could only occur in the context of pre-existing folk games suf ciently different from rugby, soccer and cricket to constitute a national heritage. The competitive figuration between rival sports organisations also shaped the actions and feelings of sports administrators towards each other. Exclusions on one side were countered by exclusions on another. And this sporting guration makes little sense unless examined as part of a broader Irish-British guration that was changing rapidly due to the heightened commercialisation of agricultural production in Ireland, the intensi cation of industrialisation in Britain, and the consequent shifts in power ratios between classes in both nations, but within the same state in the late nineteenth century.
While some may ask what have these historical processes and events got to do with understanding sport in Ireland today, well the we-feelings of supporters and administrators are passed on from one generation to the next. Children in any nation develop that national identity through pedagogical practices at home and in school. In Ireland at least, the development of the national sporting organisation that is the GAA is taught as part of Irish history. It is presented as not only a manifestation of changing British-Irish relations, but also a driver of change. It takes time for such we-feelings to change, and to understand the operations of collective emotions and memories they must be examined in time, not devoid of a temporal context in order to establish some imagined, and misguided, notion of recurring laws or eternal features of the human condition. Also, the present should not be reduced to a new rupture characterised as epochs like 'second modernity', 'late modernity', 'risk society' or 'liquid modernity' (Inglis, 2014).
Indeed Inglis echoes Elias's (1987) concerns about the discipline of sociology decades after warning about the 'retreat of sociologists into the present': If sociologists today wish to avoid the condescension of future generations, then it is a vision of a profoundly historically-sensitive sociology they should be defending, not the kind of sociology that was already in large part enslaved to the present. (Inglis, 2014: 114) DOCUMENTS AND DETACHMENT.. EMPIRIA. Revista de Metodología de Ciencias Sociales. N. o 30, enero-abril, 2015, pp. 33-52. ISSN: 1139-5737, DOI/empiria.30.2015 Another major theme of Elias's methods is the balance of emotional involvement and detachment in research, and we address below our approach to detachment.

INVOLVED DETACHMENT
One of the main arguments in Elias's Involvement and Detachment is that the development of more reality-congruent knowledge develops in a spiral process with greater capacity for detachment on the part of people towards objects, events, and processes that they encounter in life. Elias is not claiming that researchers should adopt a perspective of value-neutrality in respect of their objects of investigation, simply because following a more emotionally detached perspective is a value of scienti c inquiry. Also involvement and detachment are not polar opposites, nor are people predisposed to one inclination or the other: 'One cannot say of a person's outlook in any absolute sense that it is detached or involved … Normally adult behaviour lies on a scale somewhere between these two extremes' (Elias, 2007b: 68-9). Kilminster (2007: 122-4) also argues that sociological research entails involved detachment, in that a passionate commitment to scienti c discovery is required, not a dif dent, indifferent emotional staleness to the process. However, detachment does mean distancing oneself from preferred outcomes of research, and ensuring that the political or moral perspectives of the researcher do not prejudice the ndings. This is a considerable challenge; as sociologists participate in the 'objects' of their investigations, they are highly susceptible to feelings of danger, anger or outrage (Elias, 2007b: 13). Such participation is also necessary to generate understanding: [social scientists] cannot cease to take part in, and to be affected by, the social and political affairs o f their groups and their time. Their own participation and involvement, moreover, are themselves conditions for comprehending the problems they try to solve as scientists. For while, in order to understand the structure of molecules, one need not know what it feels like to be one o f its atoms, in order to understand the functioning o f human groups one needs to know, as it were, from the inside how human beings experience their own and other groups, and one cannot know without active participation and involvement. (Elias, 2007b: 84) In our studies of the development of sport it has of course been dif cult to access the experience of administrators, of cials, players and spectators over time. We have attempted this through the examination of historical documents contained in the GAA archives and also newspaper reports of matches and organisational changes. But it is important to empathise with those participants in terms of a kind of sociological and methodological imagination whereby one's developing knowledge of immediate and broader gurations, the dynamics of organisations and matches themselves, affords an understanding of participants' perspectives. This of course does not mean that researchers can experience the exact emotions and calculations of historical gures, but one can, so to speak, imagine oneself in their uid positions within dynamic gurations. In fact, Kilminster claims that Elias's method of writing allows 'the reader to experience, or perhaps to re-experience to a degree, the dominant behavioural codes of the different historical stages, from the unavoidable point of view of the current one' (Kilminster, 2007: 86, original emphasis). Similarly, through reading historical documents, researchers can imagine the prevailing standards and social situations of particular historical periods. Researchers also face the emotional difficulty of detaching themselves somewhat (and perhaps temporarily) from their own national habitus. Learning about Gaelic games and the GAA is part of Irish people's socialisation process (one could even say nationalising process, in that children are encouraged to become emotionally attached to the sports as a source of national pride, given that sport is increasingly globalised). Consequently we must forgo 'preferred' readings of history from a nationalistic perspective. The development of Gaelic games must be seen less as a phoenix rising from the ames of national revolution and destiny towards its fateful dominance in the people's practice and imagination, and more an outcome of interdependencies and interactions between competing sporting activities and organisations.
Doing figurational research can also produce 'secondary forms of involvement' (Elias, 2007b: 40) in that nding connections between processes and structures produces its own enjoyment. Discovering the strangeness of historical events and practices from a modern perspective is an early form of involvement, not without its pleasures, for the researcher, but a detour via detachment requires reducing the distance between past and present by locating such events and practices within the gurations, standards of conduct, and traditions of the past. Much of early gurational work on sport by Elias and Dunning focussed on violence -by players and spectators in particular, but also by hooligan groups with tenuous connections to actual sports (in Dunning's work with colleagues at the University of Leicester; for example, see Dunning (1994Dunning ( , 1996, Dunning, Murphy, and Williams (1988), Dunning and Sheard (2005)). Similarly, we nd relatively high levels in the early development of Gaelic games in Ireland, and also different forms or structures of violent events over time. This can be particularly challenging from the perspective of maintaining 'involved detachment' in the scienti c sense described above. We came across accounts of death and severe bodily assault in the early decades of the GAA. But it is important to eschew a moralising perspective on those participants involved in such violence at the time. This is not to excuse it, but simply to acknowledge that moralising can interfere with developing explanations of the dynamics of violent encounters and the formation of a social habitus quick to anger and violent assault. Remaining relatively detached involves focussing on the actions of participants in context rather than the painful and humiliating effects on victims. It also means avoiding blame attribution, at least initially. Secondary reinvolvement may lead to blame EMPIRIA. Revista de Metodología de Ciencias Sociales. N. o 30, enero-abril, 2015, pp. 33-52. ISSN: 1139-5737, DOI/empiria.30.2015 attribution in terms of the exercise of state monopolies of violence control within stadia, but blaming speci c individuals who seem to perpetrate violence according to newspaper reports runs the risk of seeking the origin of action within the imagined isolated subject -a very homo clausus 'solution'. Of course journalists often do attribute blame to particular participants in violent actions, and this is useful in order to understand the moral standards of the time period in question.
During the actual discovery of evidence particular events can evoke emotions on the part of researchers, and these responses can be useful in building explanations and descriptions. There is no strict opposition between rationality and irrationality when it comes to connecting events and processes; Elias (1991: 75) notes that rationality and irrationality should not be treated as opposites, as the path towards more reality-congruent knowledge often involves fruitful detours and milestones which are more fantasy-oriented. Researcher emotions can also be evoked by documentary evidence even though violent events occurred in the distant past. Seal (2012) discusses her feelings of empathy, but also negative emotions, towards women incarcerated for murder in the mid-twentieth century. Her evidence comprised of case le documents which included details of the women's early lives. While she does not use a gurational approach, Seal is imagining the lives these women led in terms of their limited life chances given the constraints and opportunities afforded by the gurations they formed. Seal (2012: 690) argues: 'To remain indifferent to the pain that is an integral part of homicide prosecutions seems amoral and ethically troubling, especially as the research was largely conducted for my bene t as part of my academic career.' This seems like a moral argument against taking a detached perspective, but gurational researchers attempt to move beyond the initial emotional involvement. Also, Elias stresses that research is an intergenerational pursuit, and this realisation actually assists in the capacity to suspend initial emotional involvement. We did not pursue our studies into Irish sport primarily for the bene t of our academic careers, but simply to make a small contribution to a gurational body of work on sport. The concern of individual researchers for their careers is entirely understandable, especially in the context of the increasingly pressurised environment of research metrics and ' exible' or shortterm employment contracts in universities. But when researchers locate their work within a ow of intergenerational work, then the prospect of 'their' work producing feelings of guilt or unease for using the experiences of others becomes less pressing: In fact, I can say that one of the af ictions of our time originates from the fact that the chain of generations has been loosened or broken. One no longer has the feeling that this chain still exists. People believe they can seek the ful lment of their own life here and now in our own time. They do not see that none of their tasks is an end, that it becomes meaningless if is not taken up by the next generation, whose task is also taken up by the generation following them. (Elias, 2008a: 265) For us, part of the relative detachment process involved a degree of depersonalising of sports participants. We analysed people as spectators, players, referees, of cials, administrators, and so on, but not as individual life stories. Many newspaper reports did not carry such evidence in any event, though more recently there has been a more pronounced personalisation of news stories, itself a manifestation of the individualisation of our times. If researchers can view themselves as part of intergenerational links of knowledge development, then the emotional pressure of imagining what impact the research outcomes mean for the individual researcher him or herself can recede to some extent.
We now turn to the status of such documents in gurational research, and the practicalities of nding and using them for our studies of Irish sport.

DOCUMENTS AND FIGURATIONAL DYNAMICS
We used documents in the form of newspaper reports, referees reports, organisational minutes of meetings, and organisational reports to reconstruct practices, events and moral standards of earlier periods. The newspaper reports were mainly used as accounts of matches, in order to get access to player and spectator behaviour. Referees reports from the GAA archives were used to supplement newspaper reports and in particular to get a sense of the changing 'shame thresholds' (Elias, 2012a) as evidenced by referees' accounts of spectator and player transgressions of codes of conduct. The GAA archives also contained documents pertaining to the minutes of meetings at various levels with the overall organisation; these were used to build models of the organisational guration. Though qualitative interviews with journalists and broadcasters were also generated, our main reliance was on historical documents in various forms. In other words, we treated documents as fragments and legacies of events that really happened. This does not mean we acted as naïve empiricists gullible to every journalist's written word in a newspaper. Such accounts were treated critically and sceptically in the light of other evidence available. So the documents themselves constituted a kind of symbolic guration enabling realistic interpretation. A gurational account of the use of documents in understanding the development of consumer culture addresses the use of parliamentary debates as documentary sources (Dolan, 2009c), but here we focus mainly on newspapers. A recurring theme in the historiographical literature is the status of documents as reliable accounts of events depicted in such documents. Much of this has been propelled by the 'postmodern' or 'poststructural' turn in history (see Munslow, 2006). Foucault's work has been highly in uential in this 'new history' focussing as it does on discourse rather than events or social conditions. Chartier (1997: 77) argues that discourse does not simply invent events and other non-discursive practices because the processes of discursive construction of events and practices are themselves socially formed by people according to the material and symbolic resources at their disposal. For Chartier, these social processes are 'objective' and 'external to discourse'. Now, the postmodern approach to historiography could posit that knowledge of such social processes, properties, communities and organisations could only be ascertained through the existence of fragments such as documents. So we are still left with a constellation of texts from which to narrate versions or stories of the past. But, for example, if several journalists attend the same match and write reports from their perspective for their newspapers, and such accounts largely overlap in the narrative and sequence of actions, then we, as researchers, can be reasonably con dent that such actions did in fact occur. Of course people can only make sense of actions and objects through symbols developed over time by people, but this does not mean that nothing happened. Elias presents the matter rather succinctly when he states: 'If the symbols of a language were not to some extent congruent with reality, with the data they represent, humans could not survive … But side by side with reality-congruent representations communal fantasies abound in languages past and present' (Elias, 1991: 97).
Generally we have found that when journalists used words and phrases like 'riot', 'crowd disorder', 'free ghts', 'remarkable scenes', and 'disgraceful', they provided details of various participant actions to warrant those descriptions. Our focus was on descriptions of individual and collective actions, such as hitting, punching, kicking, striking with sticks, throwing stones, slapping, headbutting, and so on. Of course words can change meaning, but this can usually be ascertained by the association of words with other words. If a constellation of words, regularly associated with each other, change their meaning simultaneously then deciphering meaning can be dif cult, but this rarely happens abruptly because new meanings create uncertainty amongst people communicating, so they elaborate with more words. Journalists existed in a guration with newspaper editors and counterparts in other newspapers, as well as readers, so the words used had to make sense to a broader guration. This is not to say that mistakes could not have been made, but research ndings are always open to modi cation or correction by further evidence. Dibble (1963) states that historians generally use documents to make sense of other documents, and also interpret events depicted in documents. Dibble also suggests that historians tend to, implicitly if not explicitly, follow established rules on the interpretation of documents. These rules include: the greater reliability of testimony in documents containing speci c details rather than general conditions (on this point, see also Earl et al., 2004); testimony recorded shortly after events tends to be more accurate than witness accounts written long after the event. But, more critically for our purposes, Dibble (1963: 207) refers to the need for historians to develop a 'sociology of documents' to support the reliability of evidence. This is akin to treating documentary evidence as a symbolic guration (see above). Bryant (2000: 498) also argues that artifacts, including documents, from the past were produced and functioned 'in a larger social matrix'. Knowledge of such a social guration lends understanding to the meaning and functions of documents. Referees reports could lead to further action against players and clubs, and they could be contested. Typical defences included provocation rather than the denial of any particular actions, so players sought to in uence moral interpretations rather than effectively claim referees had fabricated stories to wilfully punish players. Similarly newspaper journalists wrote reports knowing that some of those present at matches would read their reports. A journalist's professional reputation and livelihood depended on his or her credibility. In other words their writing was enabled and constrained by the guration they formed with others -these interdependencies placed an onus of accuracy on their accounts of matches. Booth (2006) is right to argue that media reports often contain silences, as not everything can be reported. So reports are invariably partial. But newspaper reports do not have to be complete in order to be useful. We have used such reports to understand sports violence. No doubt many spectators ate sandwiches at matches, and this is rarely reported. This does not mean the report cannot be trusted. It simply means that the journalist supposed that such details were irrelevant or of little interest to his or her readers. Also, if accounts are compared chronologically over time, silences can be as revealing as details of events. For example, early newspaper reports sometimes commented that the crowd were well-behaved; this rarely if ever occurs in recent decades. The implication is not that crowds have become increasingly rowdy, but that their proper conduct is no longer worthy of comment. It is taken for granted. Similarly, Elias largely constructs his analysis of advancing thresholds of shame based on what is no longer written in etiquette texts. The silences are meaningful precisely because they re ect changing reality.
We followed several strategies for generating newspaper reports. These included manually searching for newspaper accounts in the National Library of Ireland based on reading histories of Gaelic games in Ireland, as well online searches of The Irish Times historical archive (providing Irish Times articles from 1859), and the broader Irish Newspaper Archive, which provides access to many national and provincial newspapers, some of which date from the eighteenth century. These digital databases allow word searches within the content of entire articles, so are more exible than newspaper archives based on keyword indexes constructed manually by librarians (see Deacon, 2007, Upchurch, 2012. In constructing our account of social changes, we tried to demonstrate the shifting structural character of violent events at sports matches, and to relate such changes to changing habitus formation in Ireland over the long term. Newspaper reports can be subject to selection bias in terms of focussing on violent acts (Earl et al., 2004), but we did not seek empirical generalisation in terms of quanti cation. We sought to relate gurational dynamics at various levels of social integration with the changing structural orders of violent events. We gain access to these events mainly through accounts left by journalists, but as witnesses rather than participants, they are normally less biased (McCullagh, 2000: 59). An exception we found concerned local newspaper accounts of matches involving a team from that locality. Reports tended to have a more partisan tone, but violence was usually described and then justi ed (due to provocation) rather than signi cant details omitted (based on reports of the same matches in national newspapers, or newspapers based in the opposing team's locality). So we critically interpreted witness accounts, rather than selecting those suiting a preordained narrative.

CONCLUSION
In this paper, we describe the process of using gurational theory for the study of Gaelic games in Ireland. Though we address various choices regarding methods employed, we stress the need for a theory-informed approach to data generation, analysis and synthesis. However, following Elias, we argue against the pursuit of laws or deductive schemes in the pursuit of knowledge. Rather, in our studies of sport we have sought to uncover the connections between multiple social processes over time, including the changing nature of the social habitus for players, spectators and administrators alike. Following a figurational approach, this entails an involved detachment in examining historical documents, and the realisation that we are but a link in the chain of researchers over many generations.