Variant issue19    www.variant.org.uk    [email protected]    back to issue list


Evaluating the social impact of participation in arts activities
A critical review of François Matarasso's Use or Ornament?
Paola Merli


Introduction
In 1993 the independent research organisation Comedia, on behalf of the Arts Council of Great Britain, produced a discussion document on the social impact of the arts (Landry et al., 1993). The study was followed in 1995-1996 by an empirical research project focused on the social impact of participatory arts programmes, co-ordinated by François Matarasso for Comedia, producing the influential report Use or Ornament? The Social Impact of Participation in the Arts (Matarasso, 1997). The research programme included, among others, the working paper How the Arts Measure Up (Williams, 1997), an updating of a research project on the social impact of community arts (Creating Social Capital, 1996) carried out in Australia in 1994-1995 with support by the Australia Council for the Arts. Williams' study pioneered the methodology developed by Matarasso in his research.
While earlier publications on the social impact of the arts had attracted relatively little attention, Matarasso's study has played an important role in establishing a near-consensus in Britain among cultural policy-makers. The research was cited by the then Secretary of State for Culture, Media and Sport, Chris Smith, in speeches at the Fabian Society conference at the Playhouse Theatre, London, on 19th September 1997, and at the University of Hertfordshire in Hatfield on 14th January 1998 (Smith, 1998).
Matarasso's research has thus become politically important and worthy of critical analysis. It develops the complex theme of the social impact of the arts from a particular ideological perspective. This is partly due to the author's strong desire to be relevant and useful to the policy process and to contribute to decision-making, but such relevance seems to have been achieved to the detriment of the quality of the research work.
In the first part of this paper I will concentrate on analysing the quality of Matarasso's research. My critique will focus on methodological issues and will try to show that the research project is flawed in its design, execution and conceptual basis. I will then deal with political issues such as whether using participatory arts as a form of governance, under the heading of promoting social cohesion, is actually worthwhile and desirable. Finally, I will frame some suggestions for possible future research.

A short description of Matarasso's study
The subject of Matarasso's study is the social impact of participatory arts programmes "because it is to this area of the arts that social benefits are most commonly attributed in policy discussion" (Matarasso,1997, p. iii). The researcher has deliberately avoided studying the social impact of the professional arts. In fact, in the foreword to his research project's final report he recommends reversing the traditional logic of funding: "Britain deserves better than the exhausted prejudices of post-war debates over state support for the arts"; the new objective should be "to start talking about what the arts can do for society, rather than what society can do for the arts" (Matarasso, 1997, p. iv).
The objective of the research is to advocate for the funding of participatory arts programmes on the grounds that they can produce positive social effects which are "out of proportion to their cost" (Matarasso, 1997, p. 81). Such objective is achieved by generalising the results of the cases examined.
The study covers 60 projects in different contexts (rural, urban and metropolitan). The main survey was conducted through a questionnaire made up of 24 questions with a set of three response categories (yes/no/I don't know).1 The questionnaire was given out to 513 participants.
The 50 findings (i.e. the 50 hypotheses which, according to Matarasso, have been proved right by the research) are structured in six principal themes taken from the eight areas of social impact identified in the discussion document by Landry et al. (1993): personal development, social cohesion, community empowerment and self-determination, local image and identity, imagination and vision, health and well being.2

Methodological problems
My first criticism is that Matarasso's study has no internal validity, i.e. the data collected cannot support conclusions about the hypotheses of the research project.
Many of the 50 hypotheses are expressed as relationship between abstract concepts which are not observable, nor measurable. For example: participation in the arts "can give people influence over how they are seen by others", or "can help validate the contribution of a whole community", or "can help people extend control over their own lives", or "can help community groups raise their vision beyond the immediate". The author has not explained what people are expected to do when, for example, they have gained influence over how they are seen by others, or when they have extended control over their own lives.3 Moreover, the 24 questions which make up the participants' questionnaire are not related to the hypotheses. Therefore it is impossible to guess what legitimates Matarasso to say that if people answer "yes" to some specific questions, it means that, because of participation, they have raised their vision beyond the immediate, rather than having gained influence over how they are seen by others.
The reader of Use or Ornament? is informed from the very beginning that objectivity should be set aside, because it is an "inappropriate aspiration in evaluation of social policy" (Matarasso, 1997, p. 4), while the only interesting data are the subjective points of view of participants, which are "an appropriate response to the nature of the arts and the complexity of its social outcomes" (Matarasso, 1997, p. 4). Yet there is no systematic record of such subjective points of view in the final report, where we can only read "yes/no/I don't know" answers to the researcher's predetermined questions.4
Moreover, to what extent is the subjective perception of the participants reliable or, on the contrary, does it reflect only their (unsatisfied) desires? As Bourdieu (1979) argues, it is possible that the researcher,
"because of the dissymmetry of the survey situation and his social position, is invested with an authority which encourages the imposition of legitimacy (...). The imposition of legitimacy in the course of the survey is such that, if one is not careful, one may, as many cultural surveys have done, produce declarations of principle which correspond to no real practice. (...) it is no accident that it is the culturally most deprived, the oldest, those further from Paris, in short those least likely really to go to the theatre, who most often acknowledge that 'the theatre elevates the mind'" (pp. 318-9).
The wording of questions in Matarasso's questionnaire may actually have led respondents to biased answers. For example, the question "Was being able to express your ideas important to you?" (Matarasso, 1977, p. 101) implicitly assumes that everybody had a chance and was able to express his or her ideas. In addition, from the respondent's point of view, answering "no" to this question means either not having been able to express one's ideas (this is bad in terms of personal development, thus some respondents may answer "yes" only in order not to look or feel undeveloped) or not appreciating the fact of having been able to express one's ideas (this is socially undesirable, thus some participants may answer "yes" even if it is not necessarily what they think). The same happens with other questions, for example: "Was doing something creative important to you?" and: "Since being involved did you become more confident about what you can do?" (If one answers "no" it might seem that one was not able to grasp the opportunity of doing something creative and becoming more confident, thus some respondents may answer "yes" just to protect their own sense of self-worth).
Under these circumstances, it is by no means certain that Matarasso is really measuring what he claims to measure. There is a distinct possibility that he is measuring something else, such as the social desirability of the abstract concepts of "happiness", "empowerment" and "confidence" used in the questionnaire. Social desirability can in fact behave as a moderator variable which modifies the intensity of the dependent variables. Its influence can even make it the decisive factor in the research findings. However, the author of the research fails to note the possibility of the existence of social desirability bias and consequently does not control its extent.
In addition, some of the questions unintentionally force answers to unfold from indifference to positive values only. For example: "Since being involved have you felt better or healthier?" (even by answering "no" the respondent cannot at any rate mean that he is feeling worse, but only that he has not experienced any change); and also: "Since being involved have you been happier?" (even answering "no" the respondent cannot in any case mean that he is more unhappy or miserable, but, at most, that he is just as happy as before attendance). The consequence is that the author of the research can rule out possible negative impacts.
The reason why Matarasso does not evaluate the negative side of activities is, he claims, that the artistic experience is so important for the individual (according to whose criteria?), that it is always worth taking the risk (however, this should have been decided by the participants and not by the researcher). Thus he judges other people's quality of life according to his own standard. In a multi-ethnic society and in projects which involve people with different cultural backgrounds, this way of working seems rather inappropriate.
Factors which appear to be fairly secondary in relation to the hypotheses are examined by at least four of the 24 questions: "Has taking part encouraged you to try anything else?" (is there any social value in just trying anything else? What about trying a murder?); "Do you feel different about the place where you live?" (is there any social merit in just changing one's feeling about where he or she lives, or does this question imply that before attendance people must have felt necessarily negative about their place of residence?); "Has the project changed your ideas about anything?" (is there any social worth in just changing one's ideas about, for example, lipstick colour?). These questions do not seem to refer to indicators linked to the social issues and problems which, according to Matarasso, participatory arts should address. This shows quite clearly that Matarasso has not devised the questions in relation to the hypotheses.
On the whole, Matarasso's survey only allows him to know some of the ideas, attitudes and intentions of respondents, not to evaluate real modifications in their daily conditions of existence in relation to specific social issues. Therefore the author has only measured a change in the ideas and the values of participants, a change which seems, at least in part, to coincide with the degree of acculturation of participants to his own ideas. Strangely enough, Matarasso has devoted a whole chapter of his methodological working paper (Matarasso, 1996) to ethical issues, but has just touched upon the possibility that a patronising attitude is developed on the part of the evaluator. In any case, in the empirical research he seems to have done nothing to counter the risk of imposition of the researcher's values and ideas on participants.
The research design also has no control groups, nor a longitudinal dimension, which would have allowed the researcher to control possible extraneous variables and to ascertain the existence of a casual link or a correlation between the activities examined and the measured effects. The independent variable (arts participation) has not been manipulated along time, because the questionnaire was only distributed once. Matarasso actually worries about the fact that the phrasing of questions such as "Do you feel differently about where you live?" did not take into consideration how people might have felt before they took part in the project (Matarasso, 1977, p. 96). He notices that something has not gone well, only because some participants have answered "no" to such question. However, apart from the fact that just feeling "different" does not mean anything precise in terms of social impact, the underlying structural problem is clearly the absolute lack of before-after comparisons in the research design.5
The author of the research does not even attempt to establish a causal link because, he asserts, the difficulty is so deeply rooted in social research methods, that it is not possible to solve it. Against possible criticism to his methodological choices, Matarasso puts forward the rather evasive argument that in any case the determination of causal links does not answer the question "Why the project has been a success or a failure", while, he points out, it would be extremely important to be able to do so by analysing "the causal mechanism triggered by a given programme" (Matarasso, 1966, p. 20). However, he himself does not analyse such mechanism.
Not having tackled the problem of the causal link, or at least of a correlation, Matarasso lists the results of the research with the sentence: "the study shows that participation in the arts can:" (a list of 50 effect follows). It is like reading the label on a bottle of mineral water: "it can be diuretic". Nobody knows if this is a threat or a wish, but in any case the advertised effect cannot be guaranteed.
To sum up, it is quite clear that the data collected by Matarasso cannot support conclusions about his own hypotheses. In other words, his research has no internal validity.
How does the author deal with this problem? He justifies himself with the argument that internal validity is unattainable in the evaluation of artistic programmes because "creative initiative cannot have internal validity (...). But that is their strength - it is in the creative unpredictability of their outcomes that arts project add an essential tool to the range of social action" (Matarasso, 1996, p. 21, my italics). Leaving aside the fact that internal validity is not an attribute of real life initiatives but of pieces of research only, the author should have inferred, from his own premises, that the impact of arts programmes cannot be studied by using predefined indicators. In fact, predetermined indicators are methodological tools not suited to the task of discovering the unpredictable results of activities. Yet Matarasso does use predefined indicators - i.e. in this case, indicators which are constructed by the researcher and agreed with project partners without preliminary discussions with the people who have taken part in the activities.
Let us now suppose that all the findings of the research are valid. There are still some problems. The author does not investigate whether he is dealing with lasting results or rather with only transitory or even evanescent effects. For example, what will people do when the arts programmes are over? Also, the author does not attempt to understand whether participants in the activities under examination are actually economically deprived and socially excluded people, rather than affluent people who perhaps were feeling somewhat bored before participating.
My second criticism of Matarasso's research is that even if we assume that the findings are valid in relation to the specific activities examined, the research has no external validity. The results cannot legitimately be generalised because the sample is not representative of the wider population and of all participatory arts activities. It simply includes all participants in initiatives, which, in terms of research methodology, are selected in an accidental way. This would not be a problem in itself, if the author of the report clearly stated that his sampling procedure only provides a very weak basis for generalisation - which Matarasso fails to do. Although he may consider his sample to be typical of participatory arts activities, he does not provide any explanation of why the sample itself is a typical one. Participatory arts and the people who attend them (and thus, presumably, their impacts) are not the same in different places, times and contexts.
Moreover, in 75% of the activities studied the questionnaires were not returned. In this situation, the already weak basis for generalisation is further undermined. If it were precisely the failed projects (and hence the dissatisfied individuals) not to return the questionnaires, quantitative evidence could be interpreted as a 75% negative result. These underlying doubts are strengthened by the author's statement: "Efforts were made to extend the net beyond the enthusiasts, and speak with those who, so to speak, were inclined to slip away silently. Neutral and dissenting views were actively sought from (ex-)participants, members of the local community and professionals with knowledge of the projects" (Matarasso, 1997, p. 98). Still, the negative points of view are not discussed in the final report. They are only mentioned briefly in Chapter 9, under the title "Counterweight", where unfavourable criticism is tellingly categorised as "dissenting voices" and dissatisfaction is mistaken for cynicism: programmes where "initial optimism had dissolved and been replaced by a further layer of cynicism about the possibility of change, because nothing had happened since" (Matarasso, 1997, p. 75) are safely restricted by the author to "one or two cases".
On balance, it is fairly obvious that the "results" of the research on the specific activities examined cannot be generalised.
At the same time, Matarasso does not examine critically the projects which possibly failed to achieve their objectives, but ascribes failure to the circumstances that the "employment conditions of artists who work with people are lamentable", materials and equipment are limited, and "once or twice problems seemed to arise from inexperience of the artist(s) concerned" (Matarasso, 1997, p. 74). The scarce research material concerning the difficulties and negative effects of participation is placed by Matarasso in a section of the text under the (again very telling) title "The cost of change", where the author simply reassures the reader that "costs of participation in the arts, as in life, may be indicators of richness and engagement" (Matarasso, 1997, p. 76).
Comedia's discussion document (Landry et al., 1993) drew attention to the concept of opportunity cost used in cost-benefit analysis. The opportunity cost is the value of the benefits which would have been offered by foregone alternative projects with the same costs. In a study on the social impact of the arts, the application of cost-benefits analysis might show, for instance, that the value of the benefits produced by participatory arts activities is higher than the value of the benefits produced by other types of intervention with the same costs. On the other hand, it might show that participatory arts programmes have lower costs if compared with other types of intervention, which produce the same benefits. But Matarasso cannot develop this issue, because he has not proved that such benefits exist. Still, he asserts that "arts project can provide cost-effective solutions (...). They represent an insignificant financial risk to public services, but can produce impacts (social and economic) out of proportion to their cost" (Matarasso, 1997, p. 81). He just considers this assertion to be self-evident. Nevertheless, for impacts to be out of proportion to their cost when such impacts are not proved to exist, the cost of participatory arts programmes should be just zero. This is not a very good result for Matarasso's advocacy.

The theory of social change implicit in Matarasso's research
My last point of criticism is that the research is flawed in conceptual terms. In brief, I will try to show that the theory of social change implicit in Matarasso's research is questionable.
The interest in participatory arts which Matarasso shares with some policy-makers seems to be the expression of a particular philosophical attitude towards society. Many intellectuals have started looking at society as a mere fact: they do not venture questions, hard criticism and struggle any more; they increasingly behave like "new missionaries", who play guitar with marginalised youth, the disabled and the unemployed, aiming at mitigating the perception which they have of their own exclusion. However, this benevolent attitude does not seem to be capable of solving problems. Indeed, it does not seem that "feeling differently" (Matarasso, 1997, p. 101) about the place where one lives will transform slums into wonderful places, nor that just helping "transform the image of public bodies" (Matarasso, 1997, p. x) will transform the reality of public bodies, nor that having "a positive impact on how people feel" (Matarasso, 1997, p. x) will change people's daily conditions of existence - it will only "help" people to accept them.6 However, making deprivation more acceptable is a tool to endlessly reproduce it. Social deprivation and exclusion arguably can be removed only by fighting the structural conditions which cause them. Such conditions will not be removed by benevolent arts programmes.
The "new missionaries" also think that social problems can be tackled through top-down social cohesion and integration strategies. In a publication by the Centre for Analysis of Social Exclusion at the London School of Economics, Mark Kleinman (1998) criticises the fashionable recommendation of fighting exclusion by simply promoting inclusion:
"the current vogue is that (...) socially excluded areas don't just need jobs and better homes - apparently they need community centres, self-help groups, voluntary organisations and community businesses. I don't quite follow the logic here, as these seem to be things which better off areas don't have, or at least don't have that much (...). The danger is that the emerging ideology of social inclusion will lead to the imposition of modes of behaviour on the poor, which the rest of society has rejected (Kleinman 1998, pp. 10-12)."7
Moreover, for Kleinman a social cohesion strategy as a tool for preventing potential social conflicts is only a "chimera". To prove this, he reports the results of a study carried out by the American sociologist Mary Baumgartner on the calm suburbs around New York (The Moral Order of the Suburb, 1988), which concludes that the sense of social cohesion is actually opposed to social order and lack of crime:
"the order (...) does not arise from intimacy and connectedness, but rather from some of the very things more often presumed to bring about conflict and violence - transiency, fragmentation, isolation, atomisation, and indifference among people (...). If people in such places cannot be bothered to take action against those who offend them or to engage in conflicts, neither can they be bothered to help those in need" (Baumgartner, quoted in Kleinman, 1998, p. 11).
On the other hand, the "new missionaries" have a notion of their work as an instrument to transform the culture of the studied communities and make it more similar to their own culture and values. They know what is good for people, what their "deep sources of enjoyment" (Matarasso, 1997, p. 68) should be, and how such sources should be provided. They know what levels of "personal development" and "confidence" (Matarasso, 1997, p. 14) people should possess and what should be done in order to raise them. They even claim that people should "widen their horizons" (Matarasso, 1997, p. 16) and explain how this should be accomplished.8 Such a commitment to changing people's ideas and behaviour does not solve problems because it leaves the structural conditions of deprivation untouched. Andrew Sayer (1992) rightly questions research approaches where the critical attitude is aimed at influencing and changing the people whom it studies, because this is not, in itself a sufficient condition for social change and disillusionment might lead to unexpected negative effects:
"in primarily leading a life of reflection, it is easy for the researchers to forget that changing people's thinking may leave the world of practice largely unchanged, although a relation of dissonance may be induced between the two" (Sayer, 1992, p. 254).
Research aimed at the emancipation of the researched, continues Sayer, would involve an elimination of the division between the researcher and the researched. However, such division is deeply rooted in our society and thus the interests of the researcher are far from compatible with the interests of the researched; on the contrary,
"in these circumstances the development of certain types of knowledge may (and often does) have the effect of reinforcing domination and subordination and hence opposing a general emancipation" (Sayer, 1992, pp. 254-255).

Social stability as "new" objective
At a first reading, Matarasso's final report awakens the enthusiasm, typical of the seventies, about the idea of active participation of people in cultural activities. Nevertheless, from a deeper investigation a fundamental difference emerges between Matarasso's conception of participatory attitudes and the original phenomenon of community arts.
Matarasso claims that the "real purpose of the arts" is "to contribute to a stable, confident and creative society" (Matarasso, 1997, p. v). This point is crucial, because it reveals the new trend in the revival. While the original phenomenon was a spontaneous movement, its revival is a device "offered" by the government. While the former was directed to the expression of conflicts, Matarasso's vision is directed to social stability obtained by means of "peaceful" popular consensus, the underlying inspiration seemingly being that whereas the rich are doing the "right" things, the poor should be soothed through "therapeutic" artistic activities. While in the seventies the aim was emancipation and liberation from any form of social control, also (and above all) by means of artistic creativity, in the revival of interest in participatory arts advocated by Matarasso the aim is the restoration of social control using the same tools, although otherwise directed.
Hence, the issue of what the arts can do for society, proposed by Matarasso as an innovation, is simply a new way of achieving the old "civilising" objective of cultural policy.9 This is quite evident in some of his hypotheses, such as: participation in the arts "can promote tolerance and contribute to conflict resolution", or "can provide a route to rehabilitation and integration for offenders", or "can help people feel a sense of belonging and involvement", or "can be an effective means of health education" or " can extend involvement in social activity". However, the old "civilising" objective is used for new aims: "to develop a society of responsible risk takers" (Habermas,1999, p. 53) who are willing to accept, in a constructive and self-reliant way, the process of dismantling the welfare state. This is clear in some of Matarasso's hypotheses, such as: participation in the arts "can provide a forum to explore personal rights and responsibilities" (whereas the welfare state - according to its critics - had accustomed people to focus on rights only), or "can encourage people to accept risk positively" (while the welfare state had tended to minimise risks for people).

Conclusions and suggestions for possible future research
I have tried to show why and how Matarasso's study does not produce a well-founded understanding of the social impact of the arts. The research design is flawed, research methods are not applied in a rigorous way and the conceptual bases are questionable.
Research on the social impact of participatory arts should thus aim to develop new approaches. Publications on social impact assessment can be especially interesting even though they do not concentrate on participatory arts programmes, but, more generally, on the social consequences of policy formulation and implementation. For example, Finsterbusch (1980) provides a conceptual framework for social impact assessment, devoting as many as 25 pages to definitions of the concepts of "community cohesion" and "neighbourhood", to a historical analysis of the functions of neighbourhoods and their transformations in the twentieth century, and to methods of measurement of community cohesion and neighbourhood attachment. Finsterbusch, Llewellyn and Wolf (1983) have edited a collection of essays on social impact assessment methods, including a stimulating discussion on causes and correctives for errors of judgement in social impact research. Finally, the chapter "Towards a methodological framework" in Comedia's discussion document (Landry et al., 1993) offers methodological guidelines and caveats which can be very useful in designing an empirical research project about the social impact of arts activities.10 The document lists five different methodological approaches, explaining the contexts in which each of them would be more appropriate, and stresses the need to give relative weights to each variable that is going to be evaluated, arguing, for example, that in order to legitimately declare that an artistic programme has improved the quality of life of participants it is necessary - first of all - to know what are, in the opinions of participants, the main constituents of "quality of life" and the relative weights attributed to them.
However, one of the major problems of research into the social impact of participation in arts activities is that it has no strong theoretical grounding. The argument that the arts do have social effects (which therefore just need be measured) is far from being substantiated. Furthermore, it is not tenable that any kind of participatory arts activity in any kind of community and culture would have identical social impacts. Differences are expected to exist and it is necessary to know more about this. Without knowing what the real, specific effects of the arts are, and in which circumstances they occur, the researchers are only going to measure what they would like to be there, for example - as in Matarasso - a reduction of crime and vandalism: "the community police officer argued that active participation of residents in the life of their community was essential to maintaining order on the estate" (Matarasso, 1997, p. 35). We should therefore capitalise on - and develop further - through an interdisciplinary approach, the contributions of other important fields of research. Moreover, we should proceed along clear lines and make explicit the theories underpinning our research.
Relevant contributions include, for example, psychological and sociological theories of creativity and art perception and empirical studies in the field of cognitive psychology on the effect of the arts on individuals.
The psychologist Lev Vygotskij researched into the phenomenon of creativity as a social process. He was interested in understanding how individual creativity affected innovation at a wider social level. In his fundamental work on imagination and creativity in childhood (1930), he rejected the romantic conception of creativity as the product of a sudden inspiration. He also refuted the idealisation of children's creativity, based on a typically romantic negative attitude towards education. On the contrary, for Vygotskij the products of authentic creative imagination, in any field, were exactly the results of education and mature imagination. He showed that children have less creativity than adults, but they believe more in the products of their imagination because they have less control and critical judgement over such products - this is why children do not feel frustrated with their creative achievements. However, as rational thought gains control over imagination in the process of growing up, children who do not have a chance to learn the cultural and technical factors which make mature creativity possible, gradually become frustrated about their creative accomplishments and stop engaging with imaginative activities. Creativity - in the arts, science and technology - is, for Vygotskij, an historical and consequential process. Inventors, even geniuses, are always creatures of their times and of their environments. Their creative capacities are prompted by the needs for innovation which had been formed in earlier times and by the opportunities offered by the context in which they work. All this accounts for the disproportion in the number of innovators in different social classes. The privileged classes, according to Vygotskij, produce a much higher percentage of innovators because their members enjoy all the necessary conditions of creation.
Vygotskij's notion of creativity can raise significant research questions. For example, what is the role of participatory arts programmes in encouraging this type of creativity, compared with other forms of education? In this framework, are the social effects of participatory arts activities attended by children different from the effects of the same activities attended by adults? In what ways? Also, are the impacts of creative activities implemented in exceptionally disadvantaged areas different from the impacts of the same activities implemented in other areas? In what ways?
John Sloboda has conducted important research in the field of the cognitive psychology of music. His book The Musical Mind (1985) is an enquiry into music as a cognitive skill. Sloboda is interested in understanding "how music is able to affect people". He argues that music has to do with emotional factors, its social functions and motivations being only secondary to the emotional factors; but, he writes: "the cognitive stage is a necessary precursor of the affective stage":
"a listener cannot find a joke funny unless he understand it. However, the affective stage does not necessarily follow the cognitive state. A person may understand a joke perfectly well without being moved to laughter by it. So it is with music. A person may understand the music he hears without being moved by it. If he is moved by it then he must have passed through the cognitive stage, which involves forming an abstract or symbolic internal representation of the music (...). Composition and improvisation require the generation of such representations, and perception involves the listener constructing them" (Sloboda, 1985, p. 3).
Thus given the "conceptually-mediated nature of experience" (Sayer, 1992, p. 54), does the cultural background of participants influence the social impact of the activities? Bourdieu's Outline of a Sociological Theory of Art Perception, first published in 1968, argues that the arts exist only for those who are able to decipher them:
"the recapturing of the work's objective meaning (...) is completely adequate and immediately effected in the case - and only in the case - where the culture that the originator puts into the work is identical with the culture or, more accurately, the artistic competence which the beholder brings to the deciphering of the work (...). Whenever these specific conditions are not fulfilled, misunderstanding is inevitable: the illusion of immediate comprehension leads to an illusory comprehension based on a mistaken code" (Bourdieu, 1968; 1993, p. 216).
The first consequence of Bourdieu's theory might be that a real, deep impact of the arts can exist only for a very small élite of educated people who are able to decipher their codes, while for all those who do not have artistic competence, the main function of the arts is to legitimate social differences: "art and cultural consumption are predisposed, consciously and deliberately or not, to fulfil a social function of legitimating social differences" (Bourdieu, 1979; 1984, p. 7). This might be taken as a grounding theory for an interesting research on the social impact of the arts. For example, do participatory arts overcome such drawback?
From Bourdieu's theory we can also derive ideas which have implications on the choice of research methods. As the work of art only exists for those who can decipher it,
"the satisfactions attached to this perception (...) are only accessible to those who are disposed to appropriate them because they attribute a value to them, it being understood that they can do this only if they have the means to appropriate them. Consequently, the need to appropriate (...) cultural goods (...) can appear only in those who can satisfy it, and can be satisfied as soon as it appears. It follows on the one hand that, unlike "primary" needs, the "cultural" need as a cultivated need increases in proportion as it is satisfied, because each new appropriation tends to strengthen the mastery of the instruments of appropriation and (...) consequently, the satisfaction attached to a new appropriation; on the other hand, it also follows that the awareness of deprivation decreases in proportion as the deprivation increases" (Bourdieu, 1968; 1993, p. 227).
Also involvement in participatory arts is a cultivated cultural need and not a primary need, thus asking people whether they are satisfied with participatory arts programmes is arguably not fair unless those who are being surveyed are fully aware of their cultural deprivation. Yet because of the particular characteristics of awareness of deprivation identified by Bourdieu, it is probably not correct to use questionnaire surveys to assess whether socially deprived people are satisfied with participatory arts programmes. In-depth interviews might prove to be a better tool because they offer chances to compensate, though only in part, for distortions in communication, allowing the interviewee to ask questions and obtain information from the researcher, and enabling the researcher to understand - and not simply to measure - the ideas and the feelings of the interviewee. In-depth interviews also allow the researcher to control the effects of the research relationship, "to perceive and monitor on the spot, as the interview is actually taking place, the effects of the social structure within which it is occurring" (Bourdieu, 1993; 1999, p. 608). Finally, such methodological tool can make interviewees feel free to express and explain ideas and opinions which are not being asked to them, thus revealing aspects unforeseen by the researcher.

Acknowledgements
I would like to thank Franco Bianchini, Stuart Price, Mike Walford and two anonymous referees for their help.

Notes
1 Matarasso asserts that besides the questionnaires completed by participants, he has also made use of other tools (questionnaires completed by observers, field visits, participant observation, formal interviews, discussion groups, observer groups and evaluation through indicators agreed with project partners). However, these tools have been used only in some projects and the results are not mentioned in the research report.
2 After completing the research under consideration in this article, Matarasso has continued to work extensively on the social impact of participatory arts programmes, drawing on the same methodological framework, with only minor changes. The hypotheses and findings in his more recent works are very similar to the ones in Use or Ornament? Some of the results are published in the reports Poverty and Oysters (Matarasso, 1998a) and Vital Signs (Matarasso, 1998b).
3 In Vital Signs Matarasso (1998b) includes another indicator which is not observable, nor measurable: the "creation of new positive symbols" (p. 34). Who decides, and according to what criteria, that the symbols created are positive?
4 Apart from sporadic, isolated quotations from participants' comments, such as: "It made me realise that I'm capable of doing anything I put my mind to, whereas before I never thought that I could do anything" (Matarasso, 1997, p.15), or "We used to be worst - now we are the best" (Matarasso, 1997, p. 50).
5 In the research report Vital Signs, Matarasso (1998b) introduces a longitudinal dimension in the analysis of data on cultural infrastructures, but not in the participants' questionnaires.
6 Towards the end of the final report Matarasso surprisingly claims that the arts can "reduce public expenditure by alleviating social problems which the state would otherwise be obliged to put right" (Matarasso, 1997, p. 93, my italics). A slip of the pen?
7 In Poverty & Oysters, Matarasso (1998a) actually identifies social inclusion as one of the main hypotheses of impact.
8 In Vital Signs (Matarasso, 1998b) the objective of changing people's ideas becomes explicit. One of the key indicators identified by Matarasso of such change is, revealingly, the "increased appreciation of public authorities" (p. 34).
9 For an account of the debates on culture's civilising effect, see Bennett (1998).
10 It is quite strange that Matarasso seems to have almost completely ignored the suggestions given there.

Works Cited
Bennett, T. (1998) Culture: A Reformer's Science (Sage, London).
Bourdieu, P. (1968) "Outline of a sociological theory of art perception", in Bourdieu, P. (1993) The Field of Cultural Production. Essays on Art and Literature edited by R. Johnson (Polity Press, Cambridge).
Bourdieu, P. (1979) La distinction. Critique sociale du judgement (Les éditions de Minuit, Paris), (English translation: Distinction. A Social Critique of the Judgement of Taste, Routledge and Kegan, London, 1984).
Bourdieu, P. (1993) "Comprendre", in Bourdieu, P. et al., La misère du monde (Éditions du Seuil, Paris), (English translation: "Understanding", in Bourdieu, P. et al., The Weight of the World. Social Suffering in Contemporary Society, Polity Press, Cambridge, 1999).
Finsterbusch, K. (1980) Understanding Social Impacts. Assessing the Effects of Public Projects (Sage, Beverly Hills).
Finsterbusch, K., Llewellyn, L.G. and Wolf, C.P. (1983) Social Impact Assessment Methods (Sage, Beverly Hills).
Habermas, J. (1999) "The European nation-state and the pressures of globalization" in New Left Review, 235, May-June 1999, pp. 46-59.
Kleinman, M. (1998) Include Me Out? London School of Economics, Centre for Analysis of Social Exclusion, Casepaper 11.
Landry, C., Bianchini, F., Maguire, M. and Worpole, K. (1993) The Social Impact of the Arts: A Discussion Document (Comedia, Bournes Green, Stroud).
Matarasso, F. (1996) Defining Values: Evaluating Arts Programmes, Working Paper (Comedia, Bournes Green, Stroud).
Matarasso, F. (1997) Use or Ornament? The Social Impact of Participation in the Arts (Comedia, Bournes Green, Stroud).
Matarasso, F. (1998a) Poverty & Oysters. The Social Impact of Local Arts Developments in Portsmouth (Comedia, Bournes Green, Stroud).
Matarasso, F. with Chell, J. (1998b) Vital Signs. Mapping Community Arts in Belfast (Comedia, Bournes Green, Stroud).
Sayer, A. (1992) Method in Social Science. A Realist Approach, 2nd ed. (Routledge, London).
Sloboda, J.A. (1985) The Musical Mind. The Cognitive Psychology of Music (Clarendon Press, Oxford).
Smith, C. (1998) Creative Britain (Faber and Faber, London).
Vygotskij, S.L. (1930) Voobraûenie i tvorèstvo v öko'nom vozraste (Gosudartsvennoe Izdatel'stvo, Moska-Leningrad), (Italian translation: Immaginazione e creatività nell'età infantile, Editori Riuniti, Roma, 1972).
Williams, D. (1996) Creating Social Capital (Community Arts Network of South Australia, Adelaide).
Williams, D. (1997) How the Arts Measure Up, Working Paper (Comedia, Bournes Green, Stroud).

This article originally appeared in the International Journal of Cultural Policy, 2002 Vol. 8 (1), pp. 107-118. Reprinted with the kind permission of Taylor & Francis Ltd. ©
For a response by François Matarasso to this article see International Journal of Cultural Policy, 2003 Vol. 9 (3), pp. 337-346.

Paola Merli, De Montfort University, Faculty of Humanities, School of Media and Cultural Production, Clephan Building, The Gateway, Leicester LE1 9BH
email: [email protected]