Eric Hobsbawm (1998: 164) observed that although there has been a great deal of discussion about the nature, or even the possibility, of objectivity in the social sciences, there has been far less discussion about the problem of 'partisanship'. Certainly this has been true of the academic field of industrial relations (IR). On the one hand, based on the predominant 'pluralist' IR paradigm that there are fundamental differences of interest between employees and employers, many IR researchers and theorists have long been concerned with the problem of balancing competing objectives of different stakeholders--to use Budd's (2004) phrase, 'balancing equity, efficiency and voice'--and the advocacy of a normative view of the requirements for improving labour-market institutions, policies and practices (Kochan, 1998: 37). On the other hand, IR academics have generally also tended to posit a strong belief in the value of critical social science research premised on the notion of academic impartiality and objectivity that is not aligned to the economic or political priorities of either employers or unions (Bain and Clegg 1974; Winchester 1983; Berrill, 1983; Sisson 1991; BUIRA 2008). Although it is in Britain now largely taught in business schools, it could be argued that this differentiates IR from other 'management' subjects such as mainstream human resource management (HRM) and professional courses leading to membership of the Chartered Institute for Personnel and Development (CIPD), as well as from shop steward and trade union studies courses. Instead, academic IR is about generating understanding through scholarship that investigates the employment relationship from an impartial vantage point and in a critically questioning fashion towards all actors and parties. Such objectivity suggests the exclusion of partisanship (of 'taking sides'). In the process it also raises the related (albeit not synonymous) issue of whether social research science research generally can, or should, be value-free, or at least value-neutral.
The question of whether partisanship is an unavoidable feature of social science research is especially pertinent for a field of study as potentially value-laden as industrial relations. This article re-examines some classic philosophy of science dilemmas to demonstrate that much industrial relations research, far from being completely impartial or value-free, is often effectively partisan, albeit this is usually not explicitly acknowledged. Focusing on the 'radical/critical' contribution to IR scholarship, it goes on to argue that IR can at one and the same time be both partisan and objective, and provides a defence of partisanship that is underpinned by rigorous scholarly research methodology.
Objectivity, value neutrality, partisanship and bias
One of the key assumptions during the first half of the 20th century was that social science could mirror the natural sciences and produce 'objective' knowledge, which uncovered the true nature of the social world in exactly the same way as scientists had discovered the 'laws' of physics, chemistry and biology. The task of the social scientist was to hold up a mirror to society, producing a 'warts-and-all' image that may not be acceptable to the viewer (Rothschild 1982). It was widely believed that this necessitated a commitment on the part of social scientists to the ideal of 'value neutrality' (Hammersley 2000: I). For some, 'value-neutrality' means that research should be wholly independent of all values, and concerned with the pursuit of theoretical or factual knowledge for its own sake; while for others, value-neutrality is treated as a principle (or ideal) that guides the behaviour of researchers, so that whilst not renouncing their values, they must set them aside and not seek to promote them through their research. Defenders of value-neutrality often draw a distinction between the objective factual evidence of their research and any subjective evaluation of its implications and consequences, drawing no relationship whatsoever between the two.
Yet research can never be value-free, or even completely impartial, because it is simply implausible to suppose that a researcher is not influenced by his or her own established and preconceived values, which reflect a complex cocktail of the person's political and social beliefs, moulded by their life experiences. It is inevitable that these will feed through into research findings by influencing research focus, the questions asked, the way evidence is selected, the interpretation of facts, and subsequent evaluative analysis. But this raises the question of whether research can ever be objective if a researcher is not value-free or impartial, and whether research has to be value-free or impartial in order to be objective. Arguably, 'objective' research is objective not because it produces knowledge of phenomena as a direct result of the researcher's eschewing any personal values, or disavowing any standpoint favourable to one group's interests as opposed to another. It is objective because it produces knowledge on the basis of research which is scholarly and rigorous in its methodology, in terms of measures such as reliability, validity, representativeness and verification, and which systematically assesses any formulated hypotheses in the light of the evidence, as opposed to 'biased' research in which subjective evaluations are expressed independent of such evidence.
'Partisanship' is a more difficult term to define, albeit there have been some important sociological contributions on the question. At one end of the spectrum, Howard Becker's famous article 'Whose side are we on?' (1967) has sometimes mistakenly been interpreted as suggesting that all research should be explicitly partisan, with researchers aligning themselves from the outset with the struggle of particular social or political 'underdog' groups, with the aim of producing policy recommendations that serve these groups interests. However, Hammersley (2000: 60-89) points out that Becker actually argued that researchers cannot avoid becoming more sympathetic to some groups they study than others. This does not mean that they should set out to be partisan per se, but that they cannot avoid being viewed as taking sides, because in taking account of society's underdogs, the research will necessarily be politically partisan in its effect, challenging dominant power structures, and thereby contributing to the goal of freedom and equality. Alvin Gouldner, in his equally famous article 'Anti-Minotaur: The myth of a value-free sociology' (1962), argued for what he termed 'objective partisanship', with research explicitly directed towards serving the goal of social improvement, with a commitment to a particular set of values (such as greater equality, democracy and social justice) rather than commitment to any particular social or political group. As we shall see, such underlying sentiments probably inform most industrial relations researchers, with partisanship referring to research which is underpinned by values and beliefs that reflect the researchers' view of a progressive, more equal society, and which therefore lead them to be more sympathetic to the interests of employees than to employers, when studying the employment relationship.
The critical question is whether partisanship and objectivity are mutually exclusive. It follows from what has been said so far that--in common with the conclusions of critical realist (CR) philosophy (Sayer 1992; 2000; Carter and New 2004; Fleetwood and Ackroyd 2004)--while we reject the objectivist approach of positivism, with its notion of a completely value-free/impartial and objective social science, we also reject the subjectivist position, which assumes (as with postmodernism and social constructivism) that objective knowledge of a 'real' social world is not possible, and that all research is necessarily selective, interpretative and biased according to the actors' and/or researchers' subjective interpretations, and therefore all versions of events are equally valid. Instead, as the historians Carr (1987: 24) and Evans (1997: 224) have explained, because no 'fact' or 'truth' can be found outside a context of meaning, interpretation and judgment, it follows that the relationship between researchers and the social phenomena they investigate is mutually interdependent. It is impossible to assign primacy to facts or interpretation; both subject and object are in a continuous process in which the researcher moulds facts to interpretation and interpretation to facts. The researcher brings a distinct 'perspective' to an enquiry--a perspective that does not deny the possibility of achieving a degree of objectivity in investigation, but one which equally does not deny the presence and significance of the values and subjectivity of the observer. Hence the need for meta-theories that attempt to account for the interplay of social structures and agency.
While an objective social world does largely exist independently of the subjective interpretation of the researcher, since our understanding of it can only be refracted through the partisan interpretation of any individual researcher, some theories, descriptions and accounts of...