Theory and Practice: A Matter of Words. Language, Knowledge and Professional Community in Social Work  [1]

Language occupies centre stage in Wittgenstein’s thought, and in the sociology of knowledge, along with critical reflection over the meaning of words, categories, and concepts. The object of inquiry is the relation between language and the world, and the notion that there may be something - a meaning - independent of words, to which words refer and which differentiates them from ‘dead signs’.

Under dispute in this argument is the assumption that our language has, or can have, unquestionable bases, so that words refer to something which exists independently of us in the concrete world or in the abstract realm of ideas (Wittgenstein 1969). Wittgenstein thus invites us to reconsider taken-for-granted beliefs about the relation between the world and words. For instance, we usually assume that there are objects in the world, such as mountains or trees, which are just ‘there’ waiting to be named. Or, in other cases, we assume that there are abstract entities, such as concepts or categories, which have some sort of existence in their own right and capture the ‘essence’ of things. As a consequence, we believe that concepts like ‘child abuse’ grasp the shared essence of a variety of different situations and phenomena in the world. Words, names, merely play a secondary role: they serve solely to indicate something; they express a ‘meaning’ which is independent of us and our practices. This belief is illustrated by the endless discussions in social work on the meaning of crucial terms like ‘theory’ and ‘practice’, as if one could ever identify a correct or proper meaning for them. Most discussions seem to posit the existence of something to which a word applies correctly, or they apparently assume that it is possible to identify abstract rules for the coherent and appropriate use of a word.

Nonetheless, all exercises of this sort seem bound to fail. Whatever words we take, from those typical of everyday language to the most sophisticated and abstract of them, if we seek to identify their meaning in an uncontroversial way, we find ourselves lost in a maze. The more we try to be specific, the greater our bewilderment. The famous statement in relation to the concept of time – ‘When I talk about time I know perfectly what I mean, but if I try to say what time is I feel lost’ - could apply to every term in our language. On the one hand, no set of criteria seems appropriate to all ‘proper’ uses of a term. On the other, verbal explanations of a term entail the definition of criteria; but criteria are just further words which require definition, and the regression from words to other words could continue indefinitely. Thus, for instance, if we say that the word ‘theory’ applies solely to explanatory constructs (Clark 1991), we find ourselves in trouble when trying to explain univocally what ‘explanatory’ means. We would be unable to define clearly and unambiguously terms such as ‘cause’ and ‘effect’ without resorting to other complex concepts.

Similarly challenged in this perspective is the view that one can eventually identify simple objects or facts in the world, ‘primary elements’ which can be described by their names alone (Wittgenstein 1967, 46 - .64). If words refer to objects, or to simple facts, it should be possible to indicate the referent of a name, such as ‘tree’, by pointing to an object. Resorting to a single act of ostension, however, raises the same problems as does a verbal explanation. It will never be clear to what one is referring when pointing a finger at something, or precisely what it is that one’s finger is pointing at: it could be the shape, the colour, the dimensions, a particular or the whole (‘…an ostensive definition can be variously interpreted in every case’: P.I., .28).

Our difficulties stem from the fact that treating languages as a stable source of meaning works extremely well for the purposes of thinking, of understanding each other, of coordinating with other people. We describe situations confident that our alters will be able to figure out what the situation is like; we give verbal orders, we obey them. And we do these things on the assumption that our words ‘refer’ to something, that they have a ‘meaning’. But when we try to identify this meaning, we are lost. Of course, there is no need to be always aware of this difficulty. In most circumstances of our lives we use language unthinkingly; it works, and it is not necessary to consider what the words stand for. The problem arises when we no longer understand each other, when it seems that language does not work as we expect: cases in point being the theory and practice issue or, to consider another salient social work situation, communication in many clients/practitioners encounters.

Why, therefore, are we served so well by language in most circumstances, but find ourselves in trouble when we try to identify the meaning of words? Wittgenstein makes frequent use of the ‘teacher/ student’ situation to exemplify his argument. One way to address the question is to consider how a child learns language (Barnes et al. 1996, 53). Teaching takes place through not one, but repeated acts of ostension in which the teacher spells the word while pointing at something. At the end of this process the child possesses a finite cluster of examples of situations in which the teacher will accept the use of the word. This finite cluster is the starting point for the child when faced by new examples. No understanding whatsoever plays a part in the learning process - or at least no understanding of a ‘meaning’. Rather than teaching, one could define this process as training. To learn a language is to be trained in the use of words and/or to acquire a finite cluster of examples of ‘word use’ accepted by the community. Nor does the teacher know much more than the examples that s/he gives. S/he merely possesses a larger but nonetheless finite cluster of examples.

Clusters of examples cannot be other than concrete groups of cases, and, in Wittgenstein's view, cases are grouped together/classified on the basis of ‘family resemblances’. On examining the different ways in which we use the word game, Wittgenstein comments: ‘And the result of this examination is: we see a complicated network of similarities overlapping and criss-crossing: sometimes overall similarities, sometimes similarities of detail.’ In the next paragraph he adds ‘I can think of no better expression than 'family resemblances'; for the various resemblances between members of a family: build, features, colour of eyes, gait, temperament, etc. overlap and criss-cross in the same way.’ (Wittgenstein 1967, 66-67). As in a family whose members are similar in various respects, examples are grouped together/classified on the basis of different successive evaluations of resemblances.

There is another problem here to be solved: how can we move from a finite cluster of examples of uses of a word, to uses of the same word in any new situation? Each new use of a word involves a judgement: the world is not organised on the basis of our groupings. Every new use of a word starts from our previous experience of a finite number of examples; but previous examples are merely a resource, they cannot determine new uses. We proceed from one example to the next, and a new use of a word is not determined by its past use: it is, in a sense, a creative action. Wittgenstein invites us to conceive of words as ‘tool kits’ which can be put to different uses in different contexts, and the use of which is not predetermined.

When we learn a language, we learn to judge new instances in the same way as other people in the community do. ‘If the language is to be a means of communication there must be agreement not only in definitions but also (queer as this may sound) in judgement.’ (Wittgenstein 1967, 242).

From this perspective, the difficulty of using words confidently and effectively, without being able to clarify their meaning uncontrovertibly, seems to have been dispelled. Of course, we find ourselves in trouble when we use speculative procedures to identify the meaning of words: there is no such thing as an independent meaning; all that we have is the ability to use words in the same way as the other members of our community do, starting from a finite cluster of examples.

Hence Wittgenstein's famous assertion: ‘For a large class of cases - though not for all - in which we employ the word 'meaning' it can be defined thus: the meaning of a word is its use in the language’ (Wittgenstein 1967, 43). And to understand the meaning one should not ‘think’ but look. There is nothing behind words that has to be grasped; everything that is relevant is right before our eyes.

We may now therefore review the debate on the key terms, theory and practice in a new light. It seems possible, for instance, to make sense of discomfort felt by some authors when seeking to ‘uncover’ the meaning of words through speculation (Bailey 1982; Pilalis 1986). On the contrary, the above quotation from Wittgenstein appears to urge empirical research into the use of words. Although Wittgenstein himself was wary and indeed critical of research, the scholars identified with the ‘strong programme’ maintain that it provides the basis for empirical exploration of all forms of knowledge as social products (Bloor 1983)

Another point should be stressed now. Words are never considered in isolation. They acquire and change their meaning (the way in which they are used) in ‘language games’:

The mistake we are liable to make could be expressed thus: we are looking for the use of a sign, but we look for it as though it were an object coexisting with a sign. (One of the reasons for this mistake is again that we are looking for a ‘thing corresponding to a substantive’). The sign (the sentence) gets its significance from a system of signs, from the language to which it belongs. (Wittgenstein 1969, 5)

Barnes uses the reifying metaphor of a fabric for languages: we can think of languages as networks of interrelated concepts and categories; no concept can be conceived in isolation from the entire network: ‘To use a concept is to appraise an instance in terms of an entire fabric’ (Barnes 1982, 71; 73). Languages, therefore, are cognitive organisations within which a concept works like the fragment of a hologram and contains information about the picture as a whole: study of a concept may yield the identification of the entire fabric. However, Barnes warns against taking this reification too seriously: it may be very convenient but something gets lost with it. That ‘something’ is the dynamic character of language, which can be viewed on this account only as an on-going process. What creates stability is continuous movement (the repeated use of words), not a static framework to which we constantly refer.

We can now give a closer look to the issue of different uses, meanings of the same words. This is a crucial point if we want to treat different uses of terms and concepts as a worthwhile subject of inquiry, and not as a mistake to be corrected. The perspective introduced here entails that the world underdetermines whatever system of categorisation and language: the world could sustain many, or even infinite, systems of categorisation. Languages are merely the way in which we organise our interaction with the world. Hence no ‘message’ from reality, no refutation or falsification of our generalisations, performs a role either in language games or in every day language, or in science.

Barnes compare categorisation to cutting a cake: we can always find a different way to cut the slices (Barnes et al. 1996, 55). In fact, because our system of categorisation is conventional, whatever generalisation can be made about categories is bound to meet counter examples. Sooner or later the world will present instances to which current generalisations do not apply. Nonetheless, such instances cannot be said to determine anything within languages, nor do they dictate to human beings how they must be treated. They can be ignored as irrelevant, or looked upon as accidents or monsters; they may lead to the creation of another category, to a radical change in the language, or to restriction of the scope of the generalisation. As Barnes puts it, ‘The unexpected can always be routinely dealt with by adding a further memory to the store. If anomalies evoke crises and revolutions, the question is why those responses are preferred to more conservative alternatives.’ (Barnes 1982, 99).

At the same time, classifications cannot follow abstract universal rules. As we remarked above, criteria cannot be other than words that should be explained. A crucial point in Wittgenstein’s argument, in fact, is that every rule can be interpreted in many ways, even the rule on applying a rule. If neither the world nor rules can provide grounds for language, languages are looked upon as institutions, or self-referential practices. Self- referentiality is defined as lying somewhere in between an independent reference and the total absence of reference: ‘Between reference to an independent reality, and having no reference at all, we have self-reference, i.e. reference to a reality, but a reality which is dependent on the very acts of reference that are directed at it.’ (Bloor 1997, 68)

Hence the relativism of this perspective: languages, as well as all systems of categorisation, are social institutions and they are incomparable; there are no absolute criteria with which to establish the best or most appropriate way of classifying. Moreover, if different groups within a community use words or concepts differently, it is impossible to establish in absolute terms who is right and who is wrong, because the only possible parameter is agreement within the group.

Bloor discerns a close analogy between Wittgenstein's perspective on language and Von Mises' interpretation of price formation in a market economy. He points out that a final or a ‘correct’ price does not exist; prices are constantly re-defined by an infinite number of transactions. Although prices may appear to be external and objective to the individual who participates in these transactions. They are in fact the product of his/her own transactions, and of innumerable other transactions:

‘The notion of 'real meaning' of a concept or a sign deserves the same scorn as economists reserve for the outdated and unscientific notion of 'real' or 'just' price of a commodity. The only real price is the price paid in the course of real transactions as they proceed von Fall zu Fall.’ (Bloor 1997, 76-77)

To push the analogy further, just as the notion of 'just' price is useless because there are no standards by which to calculate the ‘right’ price, so too is the notion of 'wrong' price. Wittgenstein strongly attacks attempts to look for mistakes in the customs of communities. If an entire community uses a concept in a certain way, the only thing that can be said is that this is the way things stand within that community; to talk of mistakes in absolute terms is impossible.

In this view uncertainty about the use of words is not a shortcoming exclusively of social work or of the social sciences. It reminds us that there may be communities with completely different languages from our own, and that there is no absolute standard with which to compare among languages. Of course this is of particular relevance to the social work profession, which is very often – if not always – involved in connecting people from different segments of society, different social classes, different cultures. It helps us make sense of the clash in perspective and power struggle between practitioners and clients, which has been the subject of so much debate (Margolin 1997).

Accordingly, within the social work community, certain terms and concepts are likely to be used by practitioners in ways which academics fail to recognize. Some research has explicitly treated this as a problem (Stevenson and Parsloe 1978; Marsh and Triseliotis 1996; Osmond et al. 2004). In contrast, the perspective presented above suggests that no use of theoretical concepts by practitioners should be regarded as wrong or inappropriate, as long as it is accepted within the practitioners’ community itself. Differences between formal academic definitions and concepts as defined by practitioners simply highlight the differences between the academic and practitioner communities and their respective languages.

Differences though cannot be looked at as casual. Since language is conceived as self-referential, it could be regarded as the outcome of creative processes within specific communities. Proponents of the ‘strong programme’ in the sociology of knowledge (Barnes 1982 and 1983; Bloor 1991; Barnes et al. 1996) offer a different view. They emphasise largely undeveloped themes in Wittgenstein's thought, most notably the part played by interests in language games (Wittgenstein mainly referred to needs). Language games cannot be explained by an external worldly reality, nor by the intrinsic authority of concepts, nor by their nature as collective habits and routines (we change our habits, in fact, and these changes are among the facts that call for explanation). Rather, categorisation as conceptualisation must be explained as determined by the interplay of interests within a community: ‘It is by reference to goals and interests that particular modes of concepts application, selected from and preferred to innumerable alternative options, can be made intelligible’ (Barnes 1982, 102).

In the case of theory and practice, we may start by considering how the different interests at work within the social work community can explain different definitions of these terms.

Before moving to closer consideration of how the theory and practice issue appears from the perspective illustrated above, some specifications are in order. Assumption of this perspective may appear somewhat disconcerting: the importance given to conventional agreements, even in drawing the line between subjectivity and objectivity, seems to involve a nihilist position on reality. If our categories are the product of an agreement within the community - in other words if they are conventional - then reality may not play any part in them. This would be like saying that reality and the world do not exist: there is no world independent of our constructions.

This is not Wittgenstein's position, however; nor, on the whole, is it the position of the interpreters of his thought to whom I have referred. The issue is not whether the world exists but rather whether an absolute and unique order exists independently of our conventional cognitive order. What is maintained is that the world and our experiences present us with an infinite, complex criss-crossing of similarities and differences. Nothing is totally alike to anything else, and nothing is totally different. It can be hypothesised that human beings have a generic disposition to perceive differences and resemblances in a constant way. But neither the world nor our nature dictates the lines (differences and resemblances) along which the world should be cut. The way it is sliced is the product of a certain community agreement in practices, and it can be - and indeed is – constantly revised.

This is a fundamental point in the strong programme. Giving it due importance enables us to counter some of the criticisms of subjectivism brought against the strong programme and Wittgenstein (Lolli 1998; Nagel 1997). Subjectivism and relativism have often been considered high-risk positions in the social sciences, and especially in social work. As a profession, social work entails taking important, at times vital, decisions concerning people’s lives. It seems unacceptable and even unethical to assume that ‘anything goes’ as long as a person accepts it, and that there are no objective standards to which we can at least aspire in order to evaluate situations.

In fact, the perspective presented here can be considered as a form of subjectivism when communities are regarded anthropomorphically as individuals endowed with a perceptive apparatus, a will, and a capacity to select among several possible uses of words. But this is not the case here. What is described is, on the contrary, a natural process: cognitive orders, languages, spring from an infinite number of reciprocal adjustments among individuals (endowed with a perceptive apparatus, etc.); they are not the product of will, whim, or rational decision-making.

In a sense, to say that objectivity arises from a conventional order is not the same as equating subjectivity and objectivity. And to say that the distinction between what is subjective and what is objective, what is true and what is false, is the product of ongoing negotiation and adjustments among the members of a community is neither to say that it is arbitrary nor to reject the distinction (Hughes 1990, 158). Wittgenstein often treats words as ‘objects of comparison’, as akin to measuring rods (P.I. 130, 131): ‘It is one thing to describe methods of measurement and another to obtain and state results of measurement’ (P.I. 241, 242).

One may regard this way the concepts, crucial to social work, from the ones which designate a phenomenon on which social work intervenes such as child abuse. This kind of relativism questions, not the objectivity of the phenomena denoted by these terms as some have suggested (see Sheppard 1995, 1998; Peile et al. 1997), but the bases of this objectivity. These terms can be regarded as measuring rods, as socially constructed standards which create an objective reality.

Considering how reflection on language helps to make sense of the debate, we must now shift our focus from the theory and practice issue to the structure of the professional community and its internal segmentation. That professional communities are internally divided on the basis of conflicting interests has been extensively argued with reference to other professional groups (Freidson 1986; Atkinson 1983). The discussion of key terms which cuts across the entire debate requires us to consider its relevance to segmentations, different interests or different strategies to pursue these interests within the social work community.

Looking at the debate we can see that some key terms are crucial in the discussion. The distinctive feature of the debate is that different definitions are given to a quite specific set of factors, e.g. thinking and acting, or scientific knowledge and common sense, so that these factors are related to each other in different ways.

We may fruitfully consider the entire debate and broad approaches, e.g. one which maintains that professional practice should be guided by theory, and the other which puts at the centre stage practices as processes of knowledge building, in terms of a struggle on boundary definition. In a sense, different definitions and different uses of terms and concepts can be linked to different definitions of the groups concerned (practitioners, academic social workers, employers) and of their relations.

When the different positions in the debate are seen this way, one is struck by the central importance of defining boundaries among groups, associated with the introduction or blurring of distinctions and differences in the terms. The authors who maintain that theory guides practice see a strong divide between scientific theories and common sense. In contrast, those who deny theory such a role also question the sharpness of the latter distinction. One cannot help noticing that these different ways of drawing distinctions impact directly on definitions of the relationship between the academic social work community and the practitioners' community.

They also affect the related, different ways to draw frontiers between the professions, with informal lay helpers on one side, and other professions on the other. The concrete impact of different ways of drawing boundaries, for instance between common sense and scientific theories, is apparent when we think of social workers negotiating over wages or careers on the basis of their possession of an ‘uncommon common sense’; or, on the contrary, on the basis of their esoteric theoretical knowledge.

Given the concrete importance of boundary definition, it is likely that behind the debate lies a complex interplay of interests. Actually several authors have connected the emergence and differentiation of positions with the different interests of the specific groups involved. For instance, Payne (1991) declares that the pragmatic position, and the critical views towards theory and theoretical training in social work, rests on a power struggle for control over practitioners. Although he goes no further than this, one naturally thinks of the struggle among agencies to gain control over the training of social workers (Lee 1982; Dominelli 1997). Employers and agencies are often seen as critical toward theoretical training for social workers, their position being that a good level of practical information would be enough. Some authors (see for instance Dominnelli 1997) maintain that this position is motivated by the interest in employing more manageable and less independent-minded practitioners: in other words, practitioners who are well informed but not influenced by the theoretical ‘bubbles’ and radical ideas which seem hegemonic in the academic context, particularly in sociology.

On the other side, many authors connect the debate over the integration of theory and practice with a struggle by academic social workers to gain acceptance in the academic community and at the same time assume control over the practitioners' community. Albeit in completely different ways, Sheppard (1998) and Sheldon (1978) note that - particularly in order to gain access to the academic community - academic social workers seem to have lost contact with the specificity of social work practice. This combines with the fact that, in order to be accepted, they have assumed a subordinate position with respect to other more established disciplines in the social sciences.

Some authors anyway have clearly focused on the interests and power struggles identified as driving the debate (Karger 1983). Karger’s polemical target is ‘scientism’, which he believes to be increasingly dominant among social work academics. He sees the debate on the transformation of social work into a scientific practice as an undercover struggle between practitioners and academics. Science is merely the ‘symbolic rubric’ which, in a sense, masks the struggle:

It is a struggle between the researcher-academicians and practitioners for control of social work - a struggle between values, beliefs, and the Weltanschauung of the researchers and the practitioners' perspective. (Krager 1983, 202)

Karger remarks that the importance given to science masks a struggle for the definition of a hierarchical relation between different social groups and that there seems to be a wider political dimension in the struggle. First, the division of labour it entails reflects and confirms the division of labour in the wider society. Moreover, science and research are viewed as ways to produce ‘stories’, spectacles with which to see and interpret reality and events in a manner consonant with and confirmatory of existing social paradigms:

The earlier stories were shrouded in religion and today's are scientific, but both make claims of legitimacy. The function of both stories is to reinforce the existing social paradigm in a society. (Karger 1983, 203)

One cannot help thinking of the present debate over evidence based practice, and the quest for scientific social work, which is still very strong. Under the perspective described here, the entire debate can be taken to be part of negotiations by different groups over their reciprocal positions. Different views can be considered different strategies adopted by the academic and practitioner community to negotiate with each other, as well as with third parties such as employers, other related professionals, clients’ groups.

If one examines the two approaches identified in terms of negotiating strategies, it appears that the former entails a quest for control by academics over practitioners and for recognition of social work within the academic context, albeit in a subordinate position in relation to more established disciplines. The advantage of this strategy for practitioners would be elevation to the level of other, more accredited professions. In this approach, boundaries, between thinking and doing, between scientific knowledge and common sense, between professional and lay people, are mainly vertical, and they mark out a hierarchy.

The second approach tends to underline differences and peculiarities more in qualitative terms, but along continuous lines; boundaries are mainly horizontal. Here an alliance between academics and practitioners is crucial, and, in relation to the academic context, the struggle is for social work to be accepted as different but equal among the social sciences (see for instance the emphasis on social work as an autonomous academic discipline in Sheppard 1997).

The ‘debate over words’ can be understood in a different way if it is connected to a struggle for control between different segments of the profession. Negotiating strategies, in fact, are built up through the different uses made of, and the meanings attributed to, the crucial terms. Most of the inconsistencies underlined in the past (Clark 1991) make sense if the positions are seen in terms of strategies for negotiating relationships among groups. Likewise, numerous ambiguities in the different positions can be viewed as concessions to one group or the other, an instance of this dynamic being the ambiguous use of the term ‘practice-theory’.

A review of the debate in terms of different negotiating strategies for boundary definition, whether between academics and practitioners, social work and other professions, or professionals and lay people, illustrates the complexity of research on theory and practice. More specifically, it reminds us of the fact that any new definition of the issue is bound to be one step in a negotiating process, and in doing so, it sets the scene for new critical reflections.

But reflection on languages prompts a further consideration. The endeavour to define theory and practice by means of a speculative exercise - which many regard to be the first step in research - appears destined to create more confusion than clarity. Likewise, individual attempts to create abstract definitions and to draw abstract distinctions among different kinds of knowledge are unlikely to gain empirical relevance: formal rules can always be interpreted in innumerable ways. One can learn to repeat abstract definitions in a manner recognised by the other members of the community, and to discriminate them from other abstract definitions. But this is different from the ability consistently to apply labels to specific situations, namely in the same way as other members of the community do.

Theory and practice, rather than abstract entities which can never be grasped, can be concretely described as families of language games which involve the consistent use of ‘written or spoken words’ (concepts) and of certain connections among words (theories). The point is that it is not necessary to establish and define what the terms mean; we need only look empirically at how they are used in different contexts. Accordingly, the definition given by some students that ‘theory is what one does at school and practice is what one does in placements’, which has been often dismissed as simplistic, is more helpful than abstract classifications.

The issue can now be defined more clearly in terms of the impact exerted by theoretical academic language (learnt at school and through training) on practitioners' language/s; that is, on the cognitive order by which they make sense of what they do. To explore the issue of theory and practice in practitioners’ representations is to explore practitioners’ languages. Accounts and descriptions of work are more than means to understand a reality that lies beyond them; they become the direct object of research. Descriptions can be treated as samples of language games, and it is at this level that the connection between theory and practice can be found.

This approach is not new in social work. Paley (1987) already noted the difficulty of handling the issue of theory and practice within the more traditional frameworks. In his view, the question of whether social workers do or do not use theory should be avoided. With reference to the sociology of knowledge, and ‘the strong programme’ in particular, Paley points out that theories cannot be treated as fixed entities (Paley 1987, 172)

Paley observes that when the theory-practice issue is addressed in terms of languages, most of the outcomes of previous research can be reinterpreted. His hypothesis is that most practitioners' statements about theory express a reluctance to account for their practices in theoretical terms, and at the same time the belief that they should be able to do so. The ‘unconscious use of theory’ or the different approaches to theory (e.g. Barbour 1984) are in Paley's view all variations on the same linguistic theme. He hypothesises that in social work as in other fields (Gilbert and Mulkay 1984), it is possible to identify two linguistic repertoires: the official one, which entails the use of theoretical terms, and the contingent one, which is the one used in daily practice. Paley suggests that it may be more interesting to study the contingent language, rather than ask questions that elicit the official one. This is in tune with many recent studies in which the question of theory and practice has been set aside, and which directly explore practitioners' descriptions of their work.

While assuming the same perspective, I disagree with Paley on many points. It is a matter of fact that practitioners play different ‘language games’ while speaking with their clients, chatting with colleagues, or presenting their work to a meeting or an interviewer. However, Paley treats some language games as more ‘authentic’ than others. This looks like a new version of the dichotomy between language and reality: a certain repertoire becomes the ‘true one’, while the other only performs a legitimising function. Languages are better conceived as ‘forms of life’, i.e., as ways in which reality is constructed. And it is the recurrent, roughly similar use of words and their frequent combinations that enable us to understand the languages of the profession, not nuances or subtle differences.

Secondly, Paley seemingly implies that academic language has no impact whatsoever on the contingent repertoire. What theory provides, in Paley's view, is ‘just an unpopular, and severely limited, form of accounting’. It should be borne in mind, though, that when we talk about theory and practice, we are considering the impact of study on practice. Paley's position implies that years of training and contacts with the academic community are devoid of impact on how practitioners think, or rather, on practitioners' language and frames. If this were the case, it would be better to abandon any reflection on training altogether, viewing professional courses as mere rituals that must be performed in order to acquire the proper designation, namely to qualify as a social worker.

On the contrary, while acknowledging that social workers who account for their practice in terms of specific models are probably something of an exception, the picture changes when we consider the concepts or terms used to describe their work, clients, and so on. One finds that many of the terms used to account for practice are taken from the social work literature. Exploring the use of common terms or concepts in the two different languages may be a fruitful way to investigate the issue of theory and practice. This approach enables us to address a question of relevance to practitioners, without dropping one of the terms of my research question (theory) as other researchers have done.

Incidentally, the role of theoretical concepts in making sense of professional reality has become an interesting object of reflection. As De Montigny suggests, the ethnomethodology approach to language, which underlines how self portrayed objective accounts are indexical, opens up new research on how meaning is constructed in social work (De Montigny 2007). There has been for instance an inspiring study on the concept of ‘acceptation’ by Berlin (2005), who goes back to the roots of social work to trace the different meanings of the concept. Other interesting studies include Furlong’s (2003) on self-determination, the ambiguous use of this term, and its connection to neo-liberal ideology; Kunnemans’ (2005) reflection over the complete different meanings that terms such as empowerment or social justice acquire when they are used within different frameworks of reference, namely neo-liberal ideology; or Lorenz’s (2005) comments on the concept of activation, which in different context can be interpreted within a tradition of social work as a way of helping people to participate in community or, on the contrary in a neo liberal framework, is connected to discriminating the deserving form the undeserving lazy clients.

A relativistic approach to language can also inspire an interesting line of research, as a term considered crucial both in the literature and by practitioners, or which is used across the world, may provide the starting point for exploration of the issue. The use of common terms like ‘acceptance’ or empowerment can be taken as examples of the way in which theoretical language becomes part of how practitioners make sense of what they do in different cultural contexts.

When the issue is addressed in terms of language games, what elsewhere has been treated as a problem - namely, differences in the use of theoretical terms by practitioners, or different uses of words in different contexts - becomes the focus of interest. Interpreting differences in the use of terms and concepts as ‘distortions’ is merely to claim authority for academics over practitioners, an authority which practitioners apparently do not accept. When this claim is set aside, similarities and differences reveal the transformations of use that a term undergoes when used within different groups or segments of the community. In this sense, the analysis of similarities and differences among uses of the same term, and of its linkages with other terms within broader systems of meaning, gives us access to the different languages spoken within the practitioners' community. While abandoning the idea of measuring the correctness of a term’s use against academic standards, this analysis can yield an inside understanding of social work subcultures and frameworks.

At the same time, besides the above considerations, addressing the issue through a particular concrete example may provide practitioners with a chance to express opinions about how they specifically connect theory to practice. Indeed, the vagueness of answers to the question of how theory affects practice can be ascribed to the vagueness of the questions from the practitioners’ perspective. Starting from a specific case renders discussion of the issue more manageable and focused. We are engaging with using a crucial term like ‘contract’ as a fragment of hologram containing information about the picture as a whole. The uses of key term will allow us to explore how, in relation to a defined ‘object’, practitioners connect what they have learnt from books and courses with something learnt from experience, and how they make sense of their practice.