What are we producing knowledge for?

Joan Orme

Professor of Social Work

GlasgowSchool of Social Work

Introduction

This paper explores the imperatives for undertaking research for the purposes of knowledge production in social work. This is particularly important in an era where we are producing more and more information, but this does not necessarily contribute to knowledge and understanding, but is often driven by other agendas. In this context of increasing information it is apparent that there has been little or no strategic direction for thisto contribute to knowledge production.Such assertions suggest a distinction between knowledge and information. They also beg the question of whether there should be ‘strategic direction’ or whether knowledge should be produced for the sake of knowing.

The paper draws on theories of knowledge to explore knowledge production in and for social work. Drawing on Foucault’s premise that knowledge is produced out of antagonism it suggests that there are a number of competing imperatives for producing knowledge, or more accurately, undertaking research, producing evidence and/or information. These are deliberately polarised as negative and positive imperativesto explore the phenomena of knowledge production. Of course these polarizations are false dichotomies, but drawing on Foucault’s archaeological approach it could be argued that rearranging disparate elements of knowledge production may give different perspectives of knowledge production (Chambon, 1999).

Ideas about knowledge

Marsh & Fisher (2005) in their analysis of the need for support for developing the evidence base for social work draw on Lewis’s work in producing a formula for knowledge. They suggest that knowledge in social work is the product of combining evidence with practice wisdom and user and carer views. What is significant about this definition is that it challenges the orthodox view of what constitutes knowledge. The first component represents what is commonly thought of as knowledge in social sciences. Research based evidence is significant in policy making because it is rigorous and generalisable: it has wide applicability. It gives some of the certainties that are often required from ‘knowledge’. But combining this with practice wisdom draws on tacit knowledge and validates practitioners and their perceptions: knowledge developed from interventions. These are seen by some as subjective and less scientific (Hammersley, 1995). The inclusion of user and carer views validates service users’ and carers’ knowledge, which has been previously subjugated.

Accepting such a definition of knowledge recognises the transformative power of knowledge. It accepts that transformative knowledge ‘rattles certainties’, shakes complacency’ and ‘unhinges us from secure meanings’ (Chambon, 1999 p. 53). However this definition should be placed on one side of the dichotomy and is counteracted by imperatives for knowledge production that seek certainty and truth.

There are other ways of polarising knowledge. Gibbons et al in their exploration of the means of knowledge production define differences between what they term mode 1 and mode 2 knowledge. Mode 1knowledge is produced to ensure compliance with ‘sound scientific practice’ while mode 2 knowledge is transdisciplinary, transient, socially accountable and reflexive (1994 p.3). This description of mode 2 knowledge fits what is aspired to in social work research (Orme, 2000).Significantly,in the light of the opening statements of this paper, they argue that it is the abundance of information that has brought about a redefinition of competence:

When information is plentiful, perhaps too plentiful competence does not derive from being able to generate yet more but from insights gained by arranging what exists in novel ways (Gibbons et al, 1994 p. 64)

Hence it might be argued that in Social Work it is important to look at the information being generated in novel ways and see what insights emerge.

Setting up the antagonisms

Here then is the tension for Social Work. On the one hand we are required to produce information, to undertake research and to contribute to knowledge. Often this is driven by what I would argue are negative imperatives, where knowledge is prescriptive. These include:

  • The academic imperative
  • The policy imperative
  • The managerial imperative

On the other, as Marsh & Fisher’s definition suggests, there is the potential for knowledge production to be transformative in the Foucauldian sense, to disrupt commonly held understandings and to contribute to ‘socially distributed knowledge’ that has the potential for application (Gibbons et al, 1994), but is not prescriptive. These positive imperatives include:

  • The imperative to enhance understanding of social problems
  • The imperative to improve practice
  • The imperative to empower service users and carers

Negative Imperatives

The academic imperatives

Social work has had an uneasy relationship with research. This is related to both fragmentation as a discipline, but also the fragmentation of the discipline of Social Work with the profession of social work (Karvinen & Satka, 1999; Lorenz, 2004). Much of this fragmentation could be contributed to Social Work’s contentious place in universities. Throughout Europe the arrangements for Social Work education differ. That it is not always taught in universities, and when it is it sometimes has an uneasy position in those universities, is evidence of continuing and global tensions around the place of Social Work in the academic project. The academic imperatives for knowledge production are manifest in a variety of ways but include research quality, disciplinary identity and research funding.

Assessment of research

Social Work’s contentious position in the academy is emphasised by processes such as the Research Assessment Exercise (RAE) in the UK. In this assessment all subject areas in all Universities are required to submit their research quality profiles which are given a score. A major implication is that the score ultimately influences the funding that goes to academic departments:‘to inform the selective distribution of public funds for research’ (HEFC, 2006).The amount of funding is influenced both by the quality of research and the number of people undertaking research. Academic Social Work departmentsare under great pressure to undertake research in order to ensure funding to continue teaching Social Work. Having said that the status of a discipline is also judged by how well the discipline performs overall[i].

Here is a conundrum. Many would argue that Social Work must continue to be taught in universities in order for practice to be informed by research. For some this is based on the assumption that only by rigorously testing and evaluating practice will we ensure that service users and carers receive the best services. For others, universities are the sites of ‘new knowledge’that facilitates the ‘discovery’ of truths about human behaviour and/or the causes of social problems. In the social sciences we are unlikely to achieve the equivalent to splitting the atom or human cloning, but this does not mean our endeavours are worthless. Social Work and other social sciences need to argue that theories of human nature, theories of welfare and understanding interpersonal relationships are as important to society as scientific discoveries. Protecting children from abuse, working with people to manage dementia and assisting offenders to desist from criminal activities can involve decisions about life and death, certainly involve issues of human rights and as such are as important as ‘rocket science’, which is often held up as the gold standard of research.

If research in academia is so important why is it a negative imperative? The problem is that the assessment exercise judges the research using criteria for excellence which are linked to Mode 1 forms of knowledge production (Gibbons et al, 1994). ‘Research’ for the purpose of the RAE is to be understood as ‘original investigation undertaken in order to gain knowledge and understanding’ (HEFC, 2005).It is within the mechanics of the exercise, how the quality of knowledge is quantified, that there are challenges to Social Work knowledge. The need to address the quality of ‘applied’ or ‘practice based research’ has become crucial not just for Social Work (Furlong & Oancea, 2005). The guidelines for the RAE have suggested that those who submit practice based research should be assessed fairly against appropriate criteria. However in past assessment exercises concerns have been that outputs such as refereed journal articles and research reports ‘score’ more highly than guidance or changes brought about in practice that are not supported by ‘academic’ reports.

While it is vital that knowledge produced is of high quality the negative imperative of the RAE is that knowledge produced becomes more academic treatise. The RAE was introduced in the early 1980s as a policy of the Thatcher government. There are suspicions that constraining academics to publish research in journals that were only read (if at all) by other academics was a way of de-radicalising academics, especially those in the social sciences. A consequence of the assessment exercise was to constrain criticism of government policy or produce arguments for alternative social, economic and governmental structures.

The impact is measured by citation indexes. A different definition of high quality practice based research is that it is meaningful and useful; it has an impact on the quality of practice.

Disciplinary identity

A further negative aspect of the academic imperative as represented by the RAE is that it encourages tension and competition between disciplines. In a competitive environment each discipline claims bodies of knowledge and ways of knowing as either unique to their discipline or that they have made the major contribution to the development of bodies of knowledge. This is problematic for Social Work. Teaching the practice of social work often draws other disciplines e.g. Sociology, Social Psychology, Politics and Economics. This has led to Social Work being described as merely ‘the appliance of social science’. In concentrating on the practice of social work we have not built up a body of knowledge that is central and/or unique to Social Work. This tension has also been characterised as the distinction between Social Work as a discipline and as a profession (Lovelock et al 2004).

While it is acknowledged that Social Work has much in common with other practice based disciplines (e.g. Education) the imperative for a disciplinary identity has led to some significant activities in the UK. An exploration of the relationship of Social Work with the UK research funding council (Shaw et al, 2004) led to further funded work on the kinds and qualities of Social Work research in UK universities (Shaw, forthcoming). Also the Joint University Council Social Work Education Committee (JUC SWEC) has developed a research strategy for Social Work research. These activities are designed to ensure that Social Work knowledge production is validated and supported, in the way that other disciplines are recognised. However a negative consequence is that the need to establish disciplinary identity works against knowledge production that is transdisciplinary. Also, efforts have been made to identify distinctive characteristics of Social Work (e.g. research that make links with practice and involve partnerships with service users). However this argument for distinctiveness might lead to Social Work research being identified with mode 1 knowledge, foundational theoretical knowledge that cannot be produced outside disciplines (Gibbons et al, 1994). The irony here is that a distinctive characteristic of Social Work research is that it is committed to the socially distributed and participative nature of mode 2 knowledge production. Hence we can see that the academic imperatives may produce unintended negative consequences.

Allocating funds

The need for Social Work to be a sustainable activity in universities makes it subject to other negative aspects of academic imperatives for the way knowledge is produced, and the nature of that knowledge. In the UK the social science research council has produced large scale datasets on various aspects of the organisation of society. Secondary analysis of these data sets is encouraged by funding mechanisms that privilege research projects which utilise these. That is not to say that Social Work cannot and should not undertake analysis of large scale datasets, but it does raise the question of the nature of the analysis, what the analysis will tell us and for whom this knowledge will be useful. This links to longstanding debates about appropriate methodologies for researching the social world.The focus here is not the paradigm war between quantitative and qualitative methods, but the inter-relationship between information and knowledge production. The information available from datasets, which are now larger and more complex through the use of technology, is vital, but knowledge comes from ‘interrogating’ or mining’ such datasets in the wider context of rigorous qualitative methods that reflect the social dimensions.

A final problem for knowledge creation in Social Work raised by academia the ‘taxation’ system that currently operates in UK Universities. Research grants have to be calculated to include ‘overheads’ to pay for the infrastructure of universities. This often makes the cost of research prohibitive and puts pressure on individuals to concentrate on achieving large grants. This is to the detriment of smaller scale, more local studies that may be meaningful to particular communities, non-governmental agencies and special interest groups. The knowledge accessed by such smaller scale, local projects and the knowledge produced can be vital to our understandings of the social world, but it is often rejected as being not economical, or not meaningful.

Significantly ‘taxation’ systems also have implications at a more global level. European funding does not attract ‘overheads’ and researchers may be discouraged from working collaboratively across borders to share and transfer knowledge. This situates knowledge production in a contentious ‘space’.One characteristic of this ‘space’ is that with increasing marketisation of research it is being populated by researchers who are not situated in universities. Increasingly funding issues have led to the development of private organisations which undertake research, produce knowledge ‘for profit’. Marketisation of research has implications for knowledge production not only in terms of the cost of research, but also in terms of making knowledge a commodity. As a commodity it is not a good in its own right, but is judged only by its capacity to generate income

Constraining critique

This then leads to many constraints on researchers in the academy and on Social Work in particular. Knowledge has to be generated and presented in particular ways to achieve funding. Findings have to be disseminated in particular outputs (peer refereed, international, academic journals) for it be positively assessed. All of which leads us to question in the academic imperatives: What are we producing knowledge for?

The policy imperatives

Since the 1990s policy makers in the UKhave espoused a policy of Evidence Based Practice (EBP). It arose out of concerns that ‘nothing works’ or more accurately that there was no systematic collection of evidence that social work intervention (or indeed other professional practices) made any positive impact. The impetus was for the social professions to gather data, evaluate services to ensure that they were having a positive impact and meeting their intended outcomes. Evidence based practice has ‘making decisions’ as its central concern (Webb, 2001; 61). Decisions will be based on the evidence of ‘what works?’

What works?

The problem for knowledge production is that assumptions differ about what intended outcomes of social work intervention might be; about what works. This may be because professionals disagree with policy outcomes. Or, and this is more likely, the way that the policy outcomes are interpreted reflect different constructions of both the policy and the intended outcomes by different people in the system

These latter concerns have informed the development of methodological approaches in the constructivist tradition. In the UKrealistic evaluation (Pawson & Tilley, 1995) has been identified as providing a holistic approach to evaluation of practice and policy(Kazi, 2000). The complexities of knowledge creation and its use in an evidence based agenda are demonstrated in a research project that utilised realistic evaluation to evaluate a policy initiative within the criminal justice (Dutton et al, 2004). Through key informant interviews and focus groups with stakeholdersit became apparent that the motivation for the policy was not understood by managers and practitioners. It also emerged that the espoused outcomes for the policy were also underlaid by requirements that had not been made transparent, but were guessed at by those carrying it out. The purpose of the evaluation was supposedly to demonstrate if the policy had worked: to produce knowledge about how to implement it successfully more widely. How could this be done when there were multiple perceptions of what the required outcomes were?

Will it work?

The development of the ‘What Works?’ agenda in the domain of politics has moved from post hoc evaluations of policy to ideas and initiatives that become policy. Researchers and academics are asked to find evidence to support knee jerk policy development that is more about gaining votes than about systematic and rigorous evaluation of human behaviour and/or social policy. In the UK policies to deal with ‘anti-social behaviour’ have been dominated by such approaches. Here the requirement is for guess work (or more technically ‘predictions’) and not knowledge. Where knowledge is utilised it is commandeered to inform polices in ways that are controversial. For example, recent research findings suggest it is possible to predict at the age of 3 whether a child will become involved in antisocial behaviour and/or delinquency. Politicians wanted to utilise these findings to argue for intervention in the families of such children at a very early age.

This is not to decry the need for the systematic evaluation of research or the need for research to inform policy. That is a positive imperative discussed later. But the ‘will it work agenda’ is about grand claims and the need for easy answers to complex social situations. Often these are about changing individuals’ behaviour and not reviewing the social and economic factors that contribute to that behaviour.

Make it work!

One consequence of the above policies is that those in the social professions are then responsible for implementing the practices that arise from policies and for demonstrating that these are effective: that is to make them work.