COMPETENCE — THE SYMBOLIC PASSE-PARTOUT TO CHANGE IN A LEARNING ORGANIZATION

Scand. J. Mgmt. Vol. 14, No. 4, pp. 373—393, 1998( 1998 Elsevier Science Ltd

All rights reserved. Printed in Great Britain0956—5221/98/$19.00#0.00

PII: S0956-5221(98)00017–7

COMPETENCE — THE SYMBOLIC PASSE-PARTOUT TOCHANGE IN A LEARNING ORGANIZATION

SILVIA GHERARDI

Research ºnit on Organizational ¸earning and Cognition,Dipartimento di Sociologia e Ricerca Sociale, ºniversita di ¹rento,

via »erdi 26, 38100 ¹rento, Italy

Abstract — The methodology of social representation theory was applied to the concept of competenceduring the follow-up phase of an action-research project in the public sector. The competence ofauthority — a key concept in legal culture and in the bureaucratic model — was contrasted with thecompetence to solve problems, as one of the cornerstones of the learning organization.When verifieda posteriori the social representation of competence among middle managers proved to be an interestinghybrid, which anchored the new in a legal culture in the form of the capacity to interpret problems,coupled with the ability to conceptualize problems. The objectivization of competence is based on thedynamic between these two cognitive processes: from the particular to the general, from the abstract tothe concrete. ( 1998 Elsevier Science Ltd. All rights reserved

INTRODUCTION

Studies of social cognition in organizations share, beyond their great differences, at leastthree theoretical premises, namely:

f that the behaviour of people in organizations is a function of their interpretations ofreality;

f that their behaviour is influenced by interaction with others, real or imaginary;f that people are thinking subjects.

Underpinning these premises are markedly diverse theoretical and disciplinary schemes.Still widely influential is the neo-behaviourist tradition that derives the cognitive map(Tolman, 1951) from the cognitive need to establish a relationship between the end pursuedand the means for achieving it. Also, the usual warning of this orientation — that the map isnot the territory — again proposes the metaphor of faithful reproduction on a reduced,simplified and conventional scale.

Another theoretical tradition, that of social constructionism, assumes that all reality issocial reality, and that it is created and re-created in human experience. The strand ofcognitive studies that relies most closely on this ontological presupposition is socialrepresentation analysis, which although not widespread among organization studies is wellestablished in social psychology. The best-known surveys of studies on organizationalcognition (Schneider and Angelmar, 1993; Walsch, 1995; Strati and Nicolini, 1996), do noteven mention this theoretical and methodological tradition of analysis.

It is not the purpose of this article to remedy this shortcoming. Instead, its aim is todescribe an application of the methodology of social representations in the specific domain

373

of the follow-up phase of an action-research project conducted in an Italian comune ormunicipal administration. The project was named Progetto Sviluppo delle Competenze(Competence Development Project) and hinged on the concept of competence — asa prerequisite for genuine participation (Heller, 1992) — and its ambiguity and ambivalencewhen applied to administrative competence. At the end of the intervention phase, aninvestigation was subsequently made of the social representation of competence possessedby the middle managers who had taken part in the research as leaders of the work groupsexperimenting with innovative problem-solving practices.

The sections that follow first illustrate the scenario in which the problematic of compe-tence problem arises, and its connection with the processes of organizational learning. Thetheoretical assumptions regarding social representations are then presented. The articlecontinues with a discussion of the results of a content and similarity analysis, results whichillustrate the attitudes of an occupational community towards what they deem to becompetence. Finally, the conclusions discuss the diffusion and pervasiveness (or otherwise)of social representations, and then return to a theme central to the discussion of socialcognition in organizations: the active role of the subject in developing interpretations and inmastering socially re-presented reality.

The principal argument advanced in favour of using social representations is that theyprovide a theoretical and methodological framework better able to handle complex phe-nomena than cognitive maps. They yield a description which comprises the levels ofinformation (contents) and of attitudes, and also the strategies used to cope with there-presented social object.

Analysis of the social representation of an object reveals the degree of mastery of a socialsituation enjoyed by those who have helped to create it.

A LEARNING BUREAUCRACY?

At first sight a learning bureaucracy appears to be an oxymoron, a contradiction in terms.The paradox lies in the fact that the concept of the learning organization has beendeveloped in the literature in implicit opposition to the bureaucracy. The metaphor oforganizational learning (Gherardi, 1995b) stands as a valid alternative to the image of therational organization, because it depicts an organization grappling not only with trial anderror but also with the ambiguity of interpretative processes, of experience, of history, ofconflict, and of power. Learning suggests that a sequence of experiences — and not theabstract processes of rational thought — is the basis of action, and that the organizationpossesses a further resource — knowledge — which is an asset, an investment and a goodthat must be maintained. In this case one may say that the organization is a place oflearning; it comprises processes of organizational learning. However, one can also say thatthe organization learns because the activity of organizing is a form of practical knowledge:a bureaucracy is an organizational model based on the separation of thought fromexecution; a learning organization is an organizational model based at the same time onknowledge in action, on working, learning and innovating (Brown and Duguid, 1991).Organizational learning is now giving way to the learning organization, which consequentlyorganizes itself in order to equip itself with attention rules which govern the manner inwhich the actions and behaviours of individuals and communities of practice (Lave andWenger, 1991) interrelate to form systems of collective cognition and action.

374 S. GHERARDI

The learning organization is an organizational form that pays less attention to problemsof efficiency, order and stability, and more attention to the reliability of its performance, toadaptive change and to flexibility.

However, there is no reason for a bureaucracy to transform itself into a learning organiza-tion, even though there is a social expectation that the transformation of the public adminis-tration will move in this direction. The question — theoretical and empirical — can thereforebe formulated as follows: what conditions can favour transition to a learning bureaucracy?

This question is certainly not new to political science, as for instance, when it addressesthe problem of an administrative system’s performance (Freddi, 1989) and argues thata satisfactory level of efficiency is a function of the political and ideological attitudes ofsenior bureaucrats and of the organizational properties of administrative structures. Theimportance of behavioural, motivational, cognitive and affective factors has been summedup in the concept of ‘‘receptivity’’, which is ‘‘the ability of a bureaucracy to react positively,with promptness, reliability and efficiency, to the demands and needs expressed by societyand its political representatives, and at the same time to focus its concern more onprogrammes and problems than on formal norms and procedures’’ (Freddi, 1989; p. 29). Asearly as 1944 Kingsley argued that the bureaucratic responsibility of the modern state liesnot so much in the presumed and largely fictitious impartiality of functionaries, as in theirloyalty to goals. Receptivity may therefore be translated into a variable. However, this is ofless interest here than to stress that the literature has, for some time, addressed the factors ofsocial cognition and the content of its change, in the awareness that organizational changeis based on both cognitive and structural change.

The historical context of this problem — at least as regards the European bureaucraticmodel based on the political neutrality of the public administration and on the executivenature of administrative action — is the change in the function of public administrationsfrom guarantors of the law to providers of services. This change in the historical circumstan-ces that gave rise to the bureaucratic organizational model is interpreted by organizationalscholars as a change in the ends pursued by the organization, a change entailing a series ofadjustments to structures and operational routines.

Bureaucracy — as a form of organizational knowledge — belongs to the technology ofwriting, of mechanization, of two-way correspondences. It is an organizational formsystematically opposed to the development of learning processes because, as Morgan (1986)argues, it imposes fragmented thought structures and well-defined patterns of thought andresponsibility. Bureaucracy prevents information from circulating freely and thus encour-ages organizational sub-units to develop representations at variance with reality and toregard the sub-goals assigned to them as final objectives, with the consequence that they failto comprehend their contribution to the overall objective.

This description of bureaucracy as a system which obstructs learning was assumed inorder to develop a model for action research in the comune of Trento. The aim was toremove this obstacle and to introduce methods for the analysis and resolution of organiza-tional problems on a basis of group work, whereby individual work was posited asstructuring thought and action on the basis of fragmented information.

For a detailed description of the project the reader is referred to other works (Gherardi1995a; 1996). Here I restrict myself to analysing whether and how the intervention in workpractices was accompanied by a change in the system of organizational thought.

The research was carried out using a symbolic approach which sought to identifya symbolic system central to the organizational culture and influencing the actions of its

CHANGE IN A LEARNING ORGANIZATION 375

members. The aim was to exploit the ambiguity of symbols to produce a semantic shift inthe representational field that could alter work practices and their meanings. This symbolicsystem was identified in competence and in the ambiguity of the term in phrases such as‘‘having competence’’ (i.e. jurisdiction) or ‘‘being competent’’ (i.e. proficient). For the sym-bolic approach to organizational analysis, ambiguity is a space of interpretative freedom forthe comparison, contrast and negotiation of meanings, and of the implications of thesemeanings for actions and for negotiation among actors about whether or not to take action(Gherardi, 1995c).

Ambiguity is ambivalence, indeterminateness and a psychological resource for flexiblebehaviour (Levine, 1985). In organizations, the ambiguity and incompleteness of rules areresources with which actors can activate one of the processes of organizational change(March and Olsen, 1989).

The choice of the construct ‘‘competence’’ enables one to enter the core of a bureaucraticculture and its change. Bureaucracy, as a Weberian ideal type, is based on hierarchy andcompetence. The latter is predominantly the competence of authority, defined as the powerallotted by law to each jurisdictional or administrative body. In this connotation, it issynonymous with legal authority or empowerment, and belongs to the semantic field ofhaving competence and therefore responsibility. But Weber’s bureaucracy is also a profes-sional bureaucracy, in the sense that competence is assigned according to professionalism,i.e. according to specialization, qualifications and the possession of objectively assessedcapabilities. Functionaries must therefore be competent and their careers are defined interms of their possession of a set of capacities that have been ascertained externally. Theyare therefore both jurisdictionally competent and professionally competent. Competence issynonymous with proficiency when it denotes the ability to perform a given activity or toaccomplish a given task. It is accordingly a symbol of expertise, cognition or skill (Strati,1985).

This conception of competence proper to the legal sciences is matched by the conceptiondeveloped by ethnomethodology, social cognition, and sociolinguistics of communicativecompetence, all of which envisage the subject’s interpretative ability to obey or not to obeyrules, and to know which rules are appropriate to the context and when. Competence is thusdepicted in Keats’s terms as ‘‘negative capability’’ (Lanzara, 1993), as the capacity to live inuncertainty.

In the bureaucratic system, the ambiguity between possessing competence and beingcompetent gives rise to two cultural models.

The legal conception of competence expresses that formal logic which has become thepowerful instrument of an ideology of law and a technique of social organization — legalpositivism — dominant in European legal culture and which conceives the function ofadministration as merely executive (Tarello, 1980). In legal positivism, the application of thelaw does not produce new law. It is simply the application of legislation to concrete cases.This is a practical syllogism which takes a norm as its major premise, the assertion of a factas its minor premise, and has the imperative declaration of the law in the concrete case as itsconsequence or conclusion. If the function of bureaucracy is to apply the law, then indelimited spheres of authority competent functionaries have simple logical tasks to perform,and their competence is constituted by their ability to apply the law and to behaveaccording to the rules. It is this legal culture that dominates the organizationalculture of the public administration in Italy (Gherardi and Mortara, 1987) and governs thetraining and education of Italian civil servants. According to this formal logic, competent

376 S. GHERARDI

behaviour rests on two assumptions: (a) that the corpus of rules is clear and does not requireparticular interpretation (indeed, if such interpretation is necessary, the competent legislat-ive organ is asked to pronounce on the matter); (b) that it is coherent, i.e. that it does notcontain conflicting provisions which require choices to be made. Under this conception ofcompetence, the requisite cognitive ability is that of accomplishing a logical task, of takinga decision in conditions of certainty, and of reducing discretion or subjective interpretationto the minimum. In contraposition to it is competence as the ability to solve problems, tointerpret problems, to devise alternatives, to choose among them, to exercise discretion andtherefore responsibility.

The meaning of our organizational intervention, and of the collaboration between theUniversity of Trento and the comune of Trento, turned on this transition from ‘‘possessingcompetence’’ as the prerogative of the ‘‘old’’ bureaucratic system, to ‘‘being competent’’ asthe ability to solve problems typical of a bureaucracy that delivers services of good quality.Now that the action-research project has been completed, I intend to verify which of thesocial representations of competence was held by the senior officials of the comune. To doso, it is necessary to clarify the theoretical perspective that informs the social representationanalysis.

THE THEORY OF SOCIAL REPRESENTATION

I shall adopt the theoretical perspective constructed around the particular phenomenaknown as ‘‘social representations’’. Serge Moscovici, the French social psychologist whofirst introduced the concept (1969, 1981), argues that in social representations we discernnot only ‘‘opinions about’’, ‘‘images of’’, or ‘‘attitudes towards’’, but ‘‘theories’’, ‘‘sciences suigeneris’’, devoted to the discovery of reality and its order. The role and importance of socialrepresentations emerge in the form of ‘‘commonsense knowledge’’. In fact, a social repres-entation is a system of values, notions and practices which has a dual purpose. First, toestablish an order which enables individuals to orient themselves in the social and materialenvironment and to dominate it. Then to ensure communication among the members ofa community by providing them with a code for their exchanges, one with which they canunequivocally name and classify the parts of their world, of their individual and collectivehistory (Moscovici, 1969).

The adjective ‘‘social’’ attaches to representations because it is possible to representan extraneous object by drawing on a universe of shared meanings. Emphasizing thissocial character is the practical value attributed to social representations. The use andtherefore the spread of a social representation depends on the extent to which it is deemedable to solve concrete problems of everyday life, restoring its orderly and unproblematicnature.

The autonomy of meaning and the possibility of determining the actions of thoseindividuals who have ‘‘constructed’’ it, assigns the status of concrete and distinctivephenomena to social representations. Once they have formed, representations acquire anindependent capacity to develop; they turn into a new reality, one different from the objectrepresented. What was originally a cognitive relation between a thinking individual and anexternal object becomes a social object with its own autonomy.

The theory of social representations therefore suggests two simultaneous lines of re-search: one concerned with the cognitive act of representing, the other with the content of


such activity. The act of representing consists in the establishment of a relationship betweenthe subject and the object of the representation, not in a cognitive replication of reality. Therepresentation comprises both the mental replication of something else (in this it is akin tothe sign or symbol) and ‘‘representing is re-presenting, making present to the mind, toconsciousness’’ (Jodelet, 1989; p. 41). In this sense, a representation has been defined as thegeneral form of an item of knowledge, in that its content becomes stable once the relationbetween subject and object has consolidated itself. In the case of social representations, italso becomes social in character insofar as it is shared by the minds of a collectivity.A representation is not a replica of reality at the level of consciousness, but a modalitywhereby reality is constructed: it is autonomy and creativity that distinguish the humanexperience of the world.

In short, social representations can be defined as the elaborations of an object bya community which enable its members to behave in a comprehensible manner and tocommunicate. They are cognitive systems with a logic and language that enable themembers of a community to organize the conditions and contexts of their interactions. Onthe one hand, social representations allow individuals and groups to construct a coherentvision of reality which they use to orient their behaviour. On the other, they are the outcomeof mental activity modulated by the features of the social situation in which they areproduced (Romagnoli and Sarchielli, 1983).

I have briefly described the conditions in which social representations are produced inorder to indicate that they might be of interest to organization scholars. From my point ofview, the aim of conducting an action-research project intended to introduce an organiza-tional change by exploiting the ambiguous meaning of competence, was to effect a cognitiveand behavioural shift that, hypothetically, could be discerned in its social representation bya community undertaking change in its work practices.

More generally, social representations are of interest to organization scholars because ofthe relationship established between two forms of knowledge: the abstract scientific know-ledge that obeys general rules and is expressed in the specialized language of a communityof scholars, and the concrete knowledge elaborated in the consensual sphere of everyday lifewith its predominantly practical, contingent and commonsense knowledge derived from thevulgarization of science in order to deal with problems. The connection between these twoforms of knowledge, as evidenced and analysed by the methodology that accompanies thetheory of social representations, sheds light on the conditions in which theoretical andapplied knowledge is produced, given that organization scholars claim to change theobjects that they study.

We may now move on to a discussion of the methodological implications of the theory ofsocial representations, before looking at the methodology used for the research project.

The content of a representation can be analysed along specific dimensions, since a socialrepresentation is involved whenever individuals express value judgements about theirenvironment, organizing the information available on the subject of their representationand their attitudes towards it (Abric, 1989; Di Giacomo, 1985). Starting from content,therefore, it will be possible to identify an information component in social representationsas well as an attitudinal component related to the emotive-affective domain of the relationwith the object.

The organization of social representations is ‘‘polar’’. It articulates itself around anucleus surrounded by peripheral elements. The nucleus, which consists of the attitudinalcomponent, can be called the ‘‘site of coherence’’. It is the chief organizer of the

378 S. GHERARDI

representation, the discriminating element with respect to which all the representativeelements are organized and endowed with meaning. Given these characteristics, the nucleusis the most stable part of a social representation, ‘‘that is, the one most opposed to change’’(Abric, 1989, p. 213).

The network of meanings attributed to the representation, as organized by its nucleusaccording to the criteria of inclusion/exclusion and centrality/periphery, gives rise to a ‘‘fieldof representation’’. This is an outright ‘‘map’’ which restores the object of the representationreconstructed according to the nature, needs and beliefs of the group of individuals whoseinterest it has aroused. The particular configuration of the field evidences the coherence ofthe field of the representation via the evaluation and selection of the information gatheredon the object represented. Hence, individual strategies of action acquire a strategic directionand a very precise meaning.

It is still necessary to clarify how social representations come into being. In his earlyresearch on the spread of psychoanalysis in postwar France, Moscovici (1961) identified twoprocesses which steer and account for the procedure set out above: anchoring and objectivi-zation.

The former process attaches ideas, objects and unusual events to the closest and mostfamiliar context of the consensual universe. It refers to the networks of categories andordinary images already possessed by human beings. By means of comparison witha prototype or ‘‘typical member’’ of the category selected, the object or idea can be classifiedand then labelled. Assigning a name to an anonymous and therefore ambiguous andpotentially threatening object is to give it an identity and a meaning which, because they areshared by several individuals, are social in character. As soon as a ‘‘thing’’ is given a nameand a specific place in the organized system of real phenomena, it assumes a meaning and itsown ‘‘real’’ value, albeit with reference to a model of familiarity and continuity. Theassimilation and placement of a newly named element in ‘‘the identity matrix of our culture’’facilitates understanding of the interactions and motivations that underlie human actions(Moscovici, 1988, p. 222).

In the process of objectivization, familiarization with an alien phenomenon attributesa quasi-physical reality to something that previously appeared abstract, renderingit immediately accessible and ready for use in the everyday world. The objectivization ofan idea, which is a typical phase of the process of representation, creates a different levelof reality. The conceptual and abstract nature of objects that do not belong tothe consensual universe is replaced and given material form in a conventional reality.The social representation is made autonomous and therefore free to circulate in the socialworld.

On completion of this process, cognitive activity of a mental and social origin is no longerdistinguishable from the objects to which it refers and applies. If reality is a socialconstruction, social representations in their endless sequence are the most efficacious andimmediate manifestation of it.

A METHODOLOGICAL NOTE

A reconstruction of the polysemous meagings that attach to the concept of com-petence was achieved by means of a set of methodological choices, which I shall now brieflyoutline.


A year after the intervention phase ended, we interviewed all the group leaders who werestill in service and who had taken part in the project on a regular basis from its beginning.Allowing for the fact that many of them had retired in the meantime and that someinterviews had been lost, the number of valid interviews, 27, can be considered very close tothe population of group leaders (46). The interviews selected and analysed were those ofinformants with equivalent organizational positions, i.e. they were second-level managers,heads of an administrative service (15 interviewees) or of a technical service (11 inter-viewees). Unexpectedly, we found no significant difference between the technical andadministrative managers when we came to analyse their interviews. Consequently thedistinction was dropped, as was that of gender (of the 27 interviewees, only five werewomen).

The interview was structured to encourage the interviewees to explain the sub-jective criteria upon which they judged the competence of a politician, a colleague, auser, a manager at their own level, or the overall administration. The reason for this choicewas our desire to locate competence in an interactive context in which both actions andactors were defined and clearly identifiable, the aim being to encourage the expounding ofconcrete strategies of action rather than generic observations on the abstract notion ofcompetence.

The interviews lasted 45 min on average. They were recorded and transcribed. It isimportant to note that not only were the interviewers known to the interviewees; they werealso the same researchers who had assisted the group leaders to start up their groups.Apparently banal, this fact is nevertheless very important when one considers an interviewto be not solely a simple act of data-gathering but a context of interaction (Antaki, 1985).The conception presupposes awareness: (a) of social construction (of the map, of thediscourse topic, of the interaction, of mutual understanding and misunderstanding); (b) ofthe contextuality (of the interaction and of its contingent features but also of the historicalmoment at which the interaction takes place); (c) of language as a medium of expressionwhich contains its own limits.

It was particularly important that interviewee and interviewer should be reciprocally‘‘accountable’’, that both should be familiar with the operational environment, that bothshould be able to rely on a minimum of reciprocal trust and on the will to maintain therelationship. This meant therefore that the subject of the discourse could not be new, that itwas important for both parties, and that there was a shared communicative space in whichthe parties could discuss and evaluate an experience a posteriori. In assessing the quality ofthe ‘‘data’’ analysed it is important to know if and to what extent the persons interviewedwere motivated to make judgements and to draw inferences in a correct and accuratemanner. As Higgins and Bargh (1987) and Depolo (1988) point out, researchers oftenattribute inaccurate information or systematic errors to the ‘‘cognitive’’ limitations of theirinterviewees, rather than to motivational or affective factors relating to the interaction ofthe interview. In our case, the discourse topic was ‘‘objectively’’ important, and whether ornot the individuals concerned submitted themselves to interview was entirely at their owndiscretion, although none of them refused.

The content of the interviews was analysed using a methodology based on groundedtheory (Glaser and Strauss, 1967; Turner, 1983). Accordingly, the system of categories wasdeveloped on a basis of all the interview material, and the quantitative analysis necessary forsubsequent similarity analysis was combined with a qualitative analysis that sought touncover the theory of competence implicit in that occupational community.

380 S. GHERARDI

WHAT IS COMPETENCE AND WHAT IS NOT?

This section describes the categories that shape the managers’ discourse on competenceand its opposite.

Thirty-nine categories were identified when the interviews were analysed (Exhibit 1).Since the frequency of occurrence of these categories was not particularly high, theirnumerical distribution indicates more about the breadth of the concepts used to describethe notion of competence than about the existence of a small number of categories sharedby all the interviewees.

Only two categories — ‘‘definition of the problem’’ and ‘‘sensitivity to context’’ — werepresent in two-thirds of the interviews. Both these categories relate to the discretion of thesubject in interpreting reality and defining what is meant by the term ‘‘problem’’, and whatimplications the context in which a ‘‘problem’’ arises has for its definition and for thepractices that ensue from it. I shall examine the meaning of these two categories, whilesimultaneously explaining how the concept of competence belongs among the cognitivecapacities in the area of human experience where subjects — with lesser or greaterawareness — decide the connection between what they think and the action or non-actionto be undertaken in consequence of this.

Competence as **definition of the problem++The capacity to define a problem relies on a form of general knowledge — a general

model — which is adapted to the specific situation. Competence is thus the possession ofa specialist type of knowledge which provides the background for the decision to define theindividual problem as a manifestation of a broader range of cases, which are already knownand for which there exist established rules and procedures and, consequently, solutions. Thecognitive operation denoted by the category ‘‘definition of the problem’’ consists in the priorknowledge that ‘‘appropriate’’ solutions exist, according to the way the problem is defined.Understanding a problem involves the activation of specific technical knowledge in order toselect a label with which to identify a situation, such that it constitutes a specific case ofa more general category for which there already exist rules for action and resolution.

The ability to relate a poorly structured, new, unknown, ambiguous or complex situationto one already known, already covered by the law, by regulations, by organizationalcustoms and routines is one of the milestones of competence.

Competence as **sensitivity to context++The second milestone of competence is ‘‘sensitivity to context’’, by which is meant the

ability to select from the repertoire of work practices those that are most appropriate to thesituation at hand. Being context-sensitive means being able to self-adjust, to choose amongseveral strategies, to assess the action best suited to the contingent moment. In this case,choice among possible actions — all equally correct — is based on a criterion of appro-priateness (MacIntyre, 1980—90), not of consequentiality. It is on this principle that ‘‘situ-ated’’ decisions are made; that is, decisions which depend on the context in which theyoccur. Competence in situated decision making is the ability to adapt working practices tothe requirements of the context.

The logic of appropriateness is associated with normatively compulsory action, as opposedto the logic of consequentiality, which is associated with prediction-based choice. The logicof appropriateness seeks to answer the following questions (March and Olsen, 1989):


382 S. GHERARDI

What type of situation is this? Who am I? How appropriate are various actions to thesituation? And it suggests that the action should follow the choice most appropriate to thesituation. The coupling of these two categories in the interviews — problem definition andcontext sensitivity — marks out a semantic field closer to appropriateness than to con-sequentiality, for the logic of consequentiality is the logic of rational decision making, ofproblem-solving centred on the devising of alternatives and on the decision maker’s values.

Competence as **knowledge-in-action++I now turn to the categories cited by at least half the interviewees. This frequency class

(50—66%) comprised heterogeneous categories, although there are some similarities amongthem. First was ‘‘knowledge in action’’, i.e. the ability to translate theoretical knowledge intoaction, followed by three categories concerned with relational capabilities: that of themanager or functionary in establishing relationships (‘‘networking’’), that of users incooperating (‘‘user maturation’’), and that of the politician in performing his/her role (‘the‘‘good’’ administrator’ category). There were then two further categories — ‘‘knowing howto’’ and ‘‘definition of responsibilities’’ — which specifically related to managerial compet-ence. These last two categories will be illustrated in the next section, when I discuss thesocial representation of competence.

The first of these various categories — knowledge in action—highlights the gap that mayarise between abstract or specialist disciplinary knowledge, the body of technical expertisepossessed by a person, and his or her capacity to translate this into practical knowledge, i.e.into knowledge that informs action and renders such action competent. Under this defini-tion, competence is the ability to access a corpus of theoretical knowledge and to select andmodify it according to the purpose to which it is to be put. On the one hand, therefore,specialist knowledge is assumed, indeed it is taken for granted; on the other, however, thecompetent use of this knowledge depends on mastery of it and its practical implementation.

Accordingly, all the knowledge acquired through practice in its turn becomes theknowledge required to narrow the gap between abstract and practical knowledge.

Competence as **networking++Familiarity with organizational practices is presupposed by the ability cited by half the

interviewees as the height of competent behaviour: networking. Administrative competencethus involved the remedying of an administrative lack of coordination. Networking occur-red in the comune of Trento as the service heads wove a web of relations and contacts witheach other, creating cross connections to lubricate the bureaucratic machinery. Beinga competent networker thus required the ability to interact with others, an ability acquiredover time from the experience and knowledge of organizational practices that presupposeinterdependence among several services. Competence in networking related principally tothe manager, whose competence consisted in the ability to create a network of relations andto navigate through it, in order to accomplish that coordination among services which theformal channels envisaged but did not promote. Competent networking requires relationalskills, and it was this capacity that was cited by the two principal interlocutors andco-authors of work relationships: the users and the politicians.

Competence as **the Other+s desired behaviour++Whereas networking involves a relational skill inter pares, the ‘‘user maturation’’ category

concerns a relation in which the Other behaves in a manner desired by the One: coopera-tion, participation, information, civic propriety, interaction.


On the one hand, this category expresses awareness that the quality of a relationshipdepends on the actions and attitudes of all the partners involved; on the other, it expressesnot only an idealization of the Other but also an endeavour to socialize and discipline theOther to the desires of the One. Bureaucratic competence therefore also comprises therelational capacity to induce users to mature in the civic sense, in that their compliance isa function of their exercise of power. This pedagogic function also appeared in therelationship between politicians and managers in the Trento comune; that is, where greaterinterdependencies reduced the margins of freedom. The competence of the manager wasenhanced by the competence of the politician, and if the latter was a ‘‘good’’ administrator,the two roles operated in synergy. As in the case of users, the idealization of the Otheroperated here. The ability — for the politician — to be a ‘‘good’’ administrator depended ona separation of roles which required that a sharp distinction be drawn between the planningand the execution and organization of a service.

This category expresses the Weberian notion that the distinction between politics andadministration corresponds to the distinction between planning and execution. Paradoxi-cally, the awareness of reciprocal dependence expressed by our interviewees regarding usersand their participation was completely lost, when they described interaction situationswhere the power was with the other part. The pedagogic purpose in this case was not toinvite participation but to separate roles, so that relational capacity was the ability to createboundaries and to conceive them as trenches.

This category of the ‘‘good’’ administrator in terms of a purported separation of rolescontains a paradox — i.e. the negation of reciprocal dependence — of which tracesremained in certain statements that would have been overlooked if attention had focusedsolely on the quantitative analysis of the interviews. In fact, among categories witha frequency of less 30%, there are some that contradict the monolithic view of the goodadministrator as someone who knows his or her place. They define the behaviour of theupper levels of a public administration according to a typology which labels politicians as‘‘entrepreneurs’’ (one-third frequency), managers or technicians (less than one-third).

One should not be surprised at the co-presence in the same discourse of a ‘‘trench’’mentality, combined with finer distinctions which are drawn between a range of non-stereotypical behaviours attributed to the politician and which are matched by behavioursmarked by greater reciprocity among managers. In this way the veil that conceals copingstrategies is partially lifted. Politician-entrepreneurs recall the Schumpeterian innovator, inthat they impose a research and development strategy on their department in order toinnovate bureaucratic routine. In the case of this type of politician, managerial competencewas viewed in terms of his/her participation in the planning and management of resources.Another type are politician managers who act as mediators between the demands advancedby the users of a service — which they collect — and the service’s capacity to satisfy them.They are therefore politicians whose competence is measured in terms of their knowledge ofthe bureaucratic structure and in terms of their ability to conciliate external and internaldemands. Their managerial skill consists in their ability to negotiate the amount ofinnovation that the structure is able to absorb, and of the conditions under which theexchange will be regarded as fair. The third type is the politician—technician. Thesepoliticians usually have the same technical competence as the manager of the sector forwhich they are responsible. But they also have the administrative competence of a politicianfamiliar with the environment in which he or she moves. They were implicitly attributedgreater technical competence than the managers. The asymmetrical relationship was thus

384 S. GHERARDI

legitimated, and the collaboration was endorsed by acknowledgement of the co-presence ofhierarchy and competence.

**Non-competence++I have outlined these categories, although they do not belong to the nucleus of the

discourse on bureaucratic competence, because the indicators of a contradiction betweenwhat the interviewees said that they thought, and the way in which they describedbehaviour that contradicted what they had said, were in danger of being lost. The coherenceof accounts is a product of discourse obtained by suppressing other accounts that competewith them, so that the rhetoric of the discourse exhibits a main line of argument. The mostfrequently used categories in the discourse may be taken as quantitative indicators of thismain argument. But everything contributes meaning to a discourse, even if it is only hintedat, implied or concealed (Derrida, 1967, 1971; Billig, 1991).

Post-structuralism (Cooper and Burrell, 1988; Martin, 1990; Linstead, 1994) applied toorganizational discourse, has revealed the contribution made to the meaning of a term bythe term to which it is contraposed. The first term has been expunged and silenced by thesecond, but still exists in a supplementary relationship to it. In our case the first term iscompetence, and the term with which it has a supplementary relationship, and which wouldbe overlooked in an analysis of content which describes the contents according to fre-quency, is non-competence. The analysis here first describes the nucleus around whichdiscourse on competence articulates itself (see next section) and then examines the socialrepresentation of competence in terms of its opposite.

Six categories were used to describe non-competence. Their frequency was less than 30%and they implicitly divided non-competence between two levels: that of the organizationand that of the individual.

Non-competence depends on organizational factors when the following categories areemployed: ‘‘paper mountain’’, ‘‘watertight compartments’’, ‘‘waste of time’’. The first termindicates the complexity of regulations, which are produced in a constant stream. Deliber-ately ambiguous, they generate a hyper officiousness which translates into cumbersome andlong-drawn-out practices: a paper mountain indeed. The cause of this non-competence liesmainly outside the administration, in the regulations that, at the organizational level, breedfurther regulations because of the obscurity of the law. Although relatively stable, institu-tional rules and routines are, nevertheless, incomplete. And as the neo-institutional school(March and Olsen, 1989) has pointed out, the incompleteness of rules is a source of bothambiguity and change. The ‘‘watertight compartments’’ category denotes a lack of coord-ination among sectors and services. The hierarchical principle was the only form ofcoordination in the comune of Trento, and it produced centralization that preventedcommunication between its various services. Non-competence therefore stemmed from thefragmentation of tasks and from a failure to connect them, or from an attempt to connectthem only mechanically. Thirdly, non-competence is ‘‘time-wasting’’: the sluggishness ofbureaucratic procedures, the delay between the moment when a need is expressed and themoment when it is addressed. The phenomena to which this category was applied shed aninteresting light on the way competence is lost. Time-wasting comprises the slowness ofprocedures, the time that elapses between the decision and its implementation, the time lostas neglected files lie unattended to, the absence of economic appraisal of the time factor inthe public sector, the delay incurred while services labour to coordinate their action, andfinally the time lost because a standard procedure is lacking.


Interestingly, these elements of non-competence were viewed by the interviewees asphysiological, not pathological. They constituted the ‘‘norm’’ against which people meas-ured their skill in navigating through the bureaucracy and coping with these obstacles,which they regarded as a ‘‘natural’’ part of bureaucratic work. This view was reinforced bythe picture of non-competence at the individual level, which was yielded by the categories‘‘routine behaviour’’, ‘‘demotivation’’ and ‘‘logic of execution’’. Routine becomes a factor ofnon-competence when individuals allow themselves to be entirely absorbed by routinetasks; when they forgo the planning and innovation of what they do, stop thinking about it,and how they could do it otherwise; when they are uninterested in why certain things aredone in certain ways; and when they do only what is required of them, and passively. Thiscategory bears close similarities with that of ‘‘logic of execution’’, a term that has enteredItalian bureaucratic jargon to indicate that the law imposes a set of tasks to be performedon schedule and according to explicit criteria, which activate sanctions, formal checks andthe apportioning of responsibility. Following the logic of execution means respectingbehaviourial criteria according to the rules, and ensuring that others respect them, butwithout assuming personal risks and using the rules to justify ‘‘prudent’’ behaviour anda lack of initiative—behaviour which forbids rather than enables. It is the instrumental useof rules and regulations to evade responsibility and to defend oneself against uncertaintyand risk, reducing operationality to mere compliance with the law. Following rules,procedures and routines, interpreting them to the letter, is stigmatized as non-competentbehaviour even when one discerns the many ‘‘good reasons’’ for it. One of these isdemotivation as an ex ante cause (the absence of incentives, of a system of rewards andpunishments, of motivation factors in the work) which produces work behaviours defined aslack of commitment, of self-esteem and of satisfaction. This identification of the non-competent person with the demotivated person is curious. It apparently indicates thatdemotivated workers are tired of representing their work role ‘‘with competence’’, and areno longer concerned to display a competent representation of it, in that demotivation isa socially valid justification, a credible account which partially releases them from respons-ibility for non-competent behaviour. Knowing how to be competent and wanting to becompetent are therefore distinct phenomena, and they are flanked by the representation ofcompetence as skill in performing a social role.

The categories used to carry out the analysis of the content of competence and non-competence can be grouped under three headings: competence as capacity, i.e. as a potentialability, as knowing-how; competence as realized knowledge, knowledge in action; andcompetence as relational skill.

THE SOCIAL REPRESENTATION OF COMPETENCE

This section presents the similarity analysis (Degenne and Verges, 1973; Flament, 1981)conducted on 24 of the 39 initial categories, after a frequency threshold of the use ofa category by at least one-third of interviewees had been established. The outcome of theanalysis is a graph (the similarity tree, Exhibit 2), which indicates the simultaneousappearance of the pairs of categories (co-occurrence matrix in Exhibit 3) but does notdenote the direction of the relationship. The tree, however, presents only the highestco-occurences of certain categories. Lower co-occurences cannot be deduced from this wayof representing the data, but Exhibit 3 allows for other elaborations.

386 S. GHERARDI

The aim of the similarity analysis was to verify the presence of a social representation ofcompetence within a small occupational community. I have already said that socialrepresentations are more likely to develop when a group must familiarize itself with a newevent of importance to it, and knowledge of which will enable the group to improve its


individual and collective performance. The involvement of all the interviewees in an actionresearch which lasted three years was the precondition for the development of a socialrepresentation of competence defined as a theory of organizational action based oncompetence and not on rule-bound behaviour.

The literature suggests (Nigro et al., 1989; Depolo et al., 1996) that if the situation requiresthe generation of a social representation, it will come about very rapidly. It was therefore tobe expected that the subjects who had acted as group leaders would have accomplished thecognitive and behavioural change necessary for the assimilation of the work-groups’methods and of a different principle of organizational behaviour.

Inspection of the structure of the tree of similarity shows that the closest connection isbetween the categories ‘‘definition of the problem’’ and ‘‘sensitivity to context’’, which havealready be described in detail. These two categories correlate with a further 13 categories(54. 2% of the total).

I shall first discuss the categories linked with competence defined as the capacity to definethe problem, and then those comprising competence as sensitivity to context.

The category referring to the ability to define the problem was closely related in theinterviews with two other categories — ‘‘knowing how’’ and ‘‘defining responsibilities’’— also occurring with the same frequence. These are categories which comprise themanagerial capacity of a section head, his/her handling of authority and its delegation.‘‘Knowing how’’ involves various abilities in dealing with people without directly invokingauthority or issuing orders: encouraging cooperation, being authoritative, giving security andtrust in collaborators, showing respect for them, having an organizational sense, knowinghow to plan work, having a broad vision of it. The competence of officials is thereforeconnected on the one hand with their ability to interpret the ‘‘problem’’, and on the other withtheir ability to mobilize the work of their collaborators while still applying the hierarchicalprinciple that assigns differentiated responsibility and authority. This category thereforesignifies that competent officials are able to assume responsibility, but they are also able todiscern when something is their responsibility and when it attaches to someone else. It is thecategory that views competence as a personal sphere of responsibility in which one isaccountable for action or non-action. Put otherwise: competent behaviour can/must beactivated within the organizational ambits defined by official duties; anything that lies beyondtheir boundaries is the responsibility of someone else, or of nobody if there is no already-defined container for them. This category encapsulates the classical bureaucratic concep-tion of competence as synonymous with responsibility; and this meaning is further rein-forced by the category connected with it: ‘‘problematic case management’’, which denotesthe ability to take exceptional action when a problem is not covered by routine procedure.

It should be borne in mind that the interviewees belonged mostly to middle management.The extent to which they articulated their discourse along the dimension ‘‘having responsib-ility/assigning responsibility’’ therefore came as no surprise. Indeed, two other categories— ‘‘politician—entrepreneur’’ and ‘‘negotiating the managerial role’’ — make explicit refer-ence to the middle managers’ margins of freedom compared with senior managers orpoliticians. Obviously, problems category, depending on how they were formulated, eithercould be the responsibility of either of them when wanted to bring the problems within theirown range of action or when they wanted to off-load responsibility for them. Conversely,the category ‘‘capacity for autonomy’’ relates principally to competent workers; that is, tothose who do not only execute but know how to act independently within the constraints ofdependence imposed on them.

388 S. GHERARDI

Overall, this part of the tree of similarity suggests that competence is knowledge of wherethe boundaries of a person’s autonomy lie.

I now analyse the categories correlated with sensitivity to context, pointing out thatcontextualization is bound up with relational skills (categories 28, 36 and 30), which havealready been discussed. Interestingly, as many as four of the categories associated withcontext sensitivity refer to the presence of an external interlocutor: the users. Takentogether, however, these four categories are the mirror image of category 20: maturation ofusers to participation. If category 20 represents the citizen as desired by the bureaucrat, theother four depict the hostile bureaucratic view of clients. Users are ‘‘disoriented’’ becausethey are uninformed and they find it difficult to acquire information. They are kept ignorantof bureaucratic procedures and are therefore unaware of what is being done for them andhow much time is needed to do it. The functionary sees users as a set of citizens who merelywant their needs to be satisfied, and rapidly, and are uninterested in more general concernsor in the interdependence between individual and collective needs (note that also category25 deals with the selfish interests of users). Finally, users are attributed a set of prejudices(category 23) which consist in a negative image of a public official and the work of a publicagency and in the generalization of an unfortunate experience with a particular publicservice to the administration as a whole. This is the relational dynamic that I have elsewherecalled the ‘‘game of mirrors’’ (Gherardi and Strati, 1990), whereby groups project imagesand representations of their actions onto each other without empirically testing theirveracity, thus giving rise to self-fulfilling prophecies.

Confirmation of this was provided by the richness of the interviewees’ descriptions of thepresence (real or imaginary) of the citizenry, although they failed to mention copingstrategies. Our interviewees did not mention the skill and contrivances with which theysought to break the vicious circle between bureaucracy and citizenry. Some evidence of thisconcealed capacity was provided by the categories co-occurring with ‘‘networking’’. Asalready mentioned, this strategy was frequently cited in relation to users in a hoped-forcooperative relationship. Equally, frequently mentioned was the category ‘‘placing yourselfin someone else’s shoes’’. This label denotes the commonplace situation in which officialsand users are equally recipients of public services, and the bureaucrat, when and if willing,recognizes a set of shared interests and is able to see him/herself on the other side of thecounter; or when he or she is the recipient of services delivered by other colleagues.References to these situations were laden with ambivalence: on the one hand, those workingwithin the bureaucracy declared that, unlike other citizens, they knew the organizationaldifficulties that hampered the satisfaction of needs; on the other, as citizens they could notjustify or tolerate the fact that citizenship rights were not respected. This consideration wasdepicted as an ethical principle that motivated competent officials to activate their relation-al network internally to the organization to resolve users’ problems. ‘‘Enjoying one’s work’’(category 11) and ‘‘feeling responsible to citizens’’ (category 12) were themes that denoted anactive relationship with work and also the assumption of individual responsibilities. Note,in fact, that the other two categories (32 and 14) correlating with this theme refer to thesupport and protection afforded to collaborators so that they can count on their superior,and to a relational capacity relating mainly to users and the ability actively to listen to them.

The ambivalence and ambiguity of competence as the capacity to situate one’s conduct inconcrete situations apparently hinges on the relationship with the user and on the twofoldimage of the user as selfish and prejudiced but also as participative and cooperative (Pipan,1996).


In concluding this description of the similarity tree, I would emphasise that the analysishas revealed competence in bureaucratic work to be an elaborate and rich phenomenon.Numerous analytical elements have been used, and subtle reasoning has been required todistinguish between what is competent behaviour and what is not. It is interesting to notethat the transition from the hetero-defined bureaucratic notion of ‘‘having competence’’ tothe auto-defined conception of ‘‘being competent’’ gives rise to ambivalence, and that thisambivalence is coagulated by the presence of the user in the discourse. Similarity analysis,used in its quantitative dimension, has taken a photograph of an object in movement, butthe same analysis from a qualitative point of view shows that this blurred image has its ownaesthetic quality because it captures the movement of competence as an equilibriumbetween the relation of the local to the universal (expressed by category 7) on the one hand,and the situating of the universal in a local setting on the other (expressed by category 8).

CONCLUSIONS

The structure of the field of the social representation of competence allows a number ofconclusions to be drawn, on a basis of the quantitative elements intrinsic to the methodo-logy used and of the qualitative elements that interpretation of the interviews has high-lighted. The conclusions answer the following three questions:

1. Is there a social representation of competence among the officials of the comune ofTrento?

2. If there is, to what extent does it resemble the social representation of bureaucraticbehaviour to be found in the literature?

3. What has the empirical analysis of competence taught us about the relationshipbetween ideas and action?

The answer to the first question is tentative. The fact that the interviews were so rich andvaried in content led to the identification of 39 categories with low frequencies (only eight ofthem were cited by at least half the interviewees), and the similarity analysis found theco-occurrence of few categories (only six with a co-occurence higher than 10).

This outcome seems to suggest the urgency of assimilation and familiarization with animportant and new social object, rather than indicating the consolidation of a discoursethat has already assumed a definite semantic space and an univocal identity. The distribu-tion of the categories and the subtlety of distinctions are instead indicative of an ongoingsocial process of collectively elaborated transition from value-elements ‘‘taken for granted’’to new ones that must still find their legitimacy of discourse expression and a capacity toinform coping strategies.

Competence in this social group does not hinge on certainty, nor on stereotypicalconcepts belonging to some ideology or justificatory discourse. The ambivalence of rela-tions is apparent, so that the rhetoric of the argumentation assumes the character ofthinking out loud together with other persons. The ambivalence is particularly evident inthose discourse contents with internal incoherence: the relation with the politician was bothdenied by invoking the separation of roles and affirmed in the typology of reciprocalcollaboration. The relationship with users exhibits explicit love/hate dynamics. In bothcases transition in work practices requires transition in discourse practices. This is an

390 S. GHERARDI

ongoing and still open process of negotiation in the particular occupational communitystudied. If we look for an answer to the question of whether and how the representation ofcompetence differs from its representation in the ideal type of bureaucracy, we find supportfor this conclusion.

The Weberian model was based on an image of competence as solely the ability to obeythe duties of office, as normatively constrained behaviour, as impersonality and impartial-ity. If we then consider Merton’s (1957) thesis that the manifest functions of a bureaucracyare intimately bound up with its latent functions, we should expect the emergence ofconcepts which recall ritualism, trained incapacity, e& sprit de corps. And we may alsoconsider the keywords in the conception of the Rechtsstaat: certainty, predictability andreliability (Freddi, 1989, p. 46). And yet a conception of competence as ‘‘having competence’’is, nevertheless, present in the tree of similarity (see categories 10, 36 and those relating toa negative image of users). It is expressed principally in the category ‘‘definition of theproblem’’. This activity involves both the application of the concrete case to the generalnorm, using the legal syllogism, and the relating of a situation to an already-known casefollowing the procedure that Goffman (1961) called labelling.

What is of interest here is not so much the descriptive content of competence, whichI assume to be the application of rules, as the description of competence from various pointsof view. My analysis has provided a description of legal competence from within (Schwartzand Jacobs, 1979), from the point of view of those who believe that their skills and those ofothers consist in the prior choice of which law (or case) to apply. What is insinuated orconcealed is that the capacity to update rules to resolve problems derives from interpreta-tive ability. Accordingly, we may also conclude that the interviewees viewed competence asthe capacity to solve problems within a context of action constituted by norms.

Resolving this ambiguity in managers’ discourse by imposing a univocal interpretation(of the researcher) would have meant jettisoning the richness of a polysemic discourse.Conversely, we can adopt an alternative interpretation which answers the question: Whatimage of managerial competence is proposed by this occupational community representingthe middle management of an average Italian comune? Their social representation ofcompetence speaks through silences and omissions. If we are to assume that the parametersof managerial action are efficacy, efficiency and equity, then we note the implicit presence ofreferences to the efficacy of action (the definition of problems prior to their solution), as wellas to the equity to which it gives rise (impact on the rights of citizens). Almost entirelyabsent, however, are references to cost cutting, to the alternative use of resources, to thequality of the services supplied (except with reference to users). The only managerialcompetences explicitly referred to are those relating to human resources management.

We saw at the outset that social representations arise in close relation with the environ-ment and with human experience of this environment. We must therefore deduce that anorganization containing this type of social representation is one that rewards legal compet-ence, which does not require managerial skills, which regards the resources available asgiven resources, and which selects its members according to these values. We may alsodeduce that change in bureaucratic practices is of necessity very slow because ambiguity isinherent to the complexity of administrative action, with its co-presence of regulatory taskswhich follow the logic of the law and tasks of service delivery which follow an economic andnegotiative logic.

The final conclusion concerns the information that the theoretical scheme of socialrepresentations gives us about the relationship between ideas and action. First, it enables us


to consider the circular and dialogic nature of the relationship between the sphere ofthought and the sphere of action, without having to assume that one precedes the other, orthat one orients the behaviour of the other, or that one is congruent with the other. A socialrepresentation is not a plan of action but an explanatory map. It delivers the image of anactive subject, of a social being, and of a subjectivity not reduced to individualism whilegaining mastery of the socially constructed world by means of coping strategies elaboratedand negotiated by a plurality of subjects.

In particular, the social representation of competence in the occupational communityanalysed had a stable nucleus centred on two typically cognitive processes. Competence wasprincipally cognitive competence: the ability to relate the particular to the general and,simultaneously, the ability to assess the particular in relation to the contexts in which itappears and which gives it meaning. In other words, reality is defined by an interpretativeactivity which searches for meaning in terms of its anchoring in more general situations,principles or norms, and uses a logic of situationality to evaluate this anchoring process inrelation to the features of the context.

Obviously, every methodology has its strengths and weaknesses. I must acknowledgethat the above analysis of the social representation of competence has produced a staticimage of the phenomenon. This is because the methodology is designed to verify thepresence and strength of connections. It is less able to depict the processes, rhetorical on theone hand and negotiative on the other, that induce the representation of the object viadialectical and rhetorical argument, both purely intra-individual and social. This criticismhas been well expressed by Billig (1991), and it should be taken into account so that thismethodology can be integrated with other methods of rhetorical analysis or ethnographicinquiry. In fact, any attempt at representation is always one sided and partial, since themarking of distinctions through language use is both a necessary condition of social life andyet one that inevitably limits alternative distinctions (Alvesson and Deetz, 1996).

Acknowledgements — I incurred too many intellectual debts in the course of the research that gave rise to this articlefor me to list them in detail. The institutions that have financed the project are the Department of Sociology andSocial Research of the University of Trento, the municipality of Trento and CNR (contribution no. 9501870 CT 10).I am extremely grateful to the research team and especially to Flaviano Zandonai, who processed the data. I amindebted to my colleagues and friends Bruno Bolognini and Hans Maria Schadee for their discussion of the results.

REFERENCES

Abric, J. C., Un approccio teorico e sperimentale allo studio delle rappresentazioni sociali in una situazione diinterazione. In: Farr, R. M. and Moscovici, S. (Eds), Rappresentazioni Sociali (Bologna: Il Mulino, 1989).

Alvesson, M. and Deetz, S., Critical theory and postmodernism approaches to organizational studies. In: Clegg, S.,Hardy, C. and Nord, W. (Eds), Handbook of Organization Studies (London: Sage, 1996).

Antaki, C., Ordinary explanations in conversation: causal structures and their defence, European Journal of SocialPsychology (1985), Vol. 15, pp. 213—230.

Billig, M., Ideology and Opinions: Studies in Rhetorical Psycology (London: Sage, 1991).Brown, J. and Duguid, P., Organizational learning and communities of practice: toward a unified view of working,

learning and burocratization, Organization Science (1991), Vol. 2, No. 1, pp. 40—57.Cooper, R. and Burrell, G., Modernism, postmodernism and organizational analysis: an introduction, Organiza-

tion Studies (1988), Vol. 9, No. 1, pp. 91—112.Degenne, A., and Verges, P., Introduction a l’analyse de similitude, Revue Francaise de Sociologie (1973), Vol. 16,

pp. 471—512.Depolo, M., Entrare nelle Organizzazioni (Bologna: Il Mulino, 1988).Depolo, M., Fraccaroli, F., Sarchielli G. and Zandonai, F., Il contributo psico-sociale alla rappresentazione di un

cambiamento amministrativo, Amministrare (1996), Vol. xxvi, No. 1, pp. 143—152.

392 S. GHERARDI

Derrida, J., De la Grammatologie (Paris: de Seuil, 1967).Derrida, J., ¸’Ecriture et la Difference (Paris: de Seuil, 1971).Di Giacomo, J. P., Rappresentazioni Sociali e Movimenti collettivi (Napoli: Liguori, 1985).Flament, C., L’analyse de similitude: une technique pour le recherches sur les representations sociales, Cahier de

Psychologie Cognitive (1981), Vol. 18, No. 1, pp. 423—429.Freddi, G., Burocrazia, democrazia e governabilita. In: Freddi, G. (ed.), Scienza dell’amministrazione e Politiche

Pubbliche (Roma: La Nuova Italia Scientifica, 1989).Gherardi, S., Conoscere ed intervenire nella Pubblica Amministrazione secondo un approccio simbolico, Rivista

¹rimestrale di Scienza dell’Amministrazione (1995a), Vol. 42, No. 1, pp. 211—239.Gherardi, S., Organizational Learning. In: Warner, M. (Ed), International Encyclopedia of Business and Manage-

ment (London: Routledge, 1995b).Gherardi, S., When will he say: ‘‘Today the plates are soft’’? The management of ambiguity and situated

decision-making, Studies in Cultures, Organizations and Societies (1995c), Vol. 1 No. 1, pp. 9—27.Gherardi, S., Knowing and intervening in the public administration according to the symbolic approach. Paper

presented at the 3rd PeSCE workshop Internazionalization, Europeanization and the Public Sector (Bodo: 16—19june 1996), pp. 1—29.

Gherardi, S. and Mortara, V., Puo la cultura organizzativa essere uno strumento per l’analisi della amminis-trazione pubblica italiana? Rivista trimestrale di Scienza dell’Amministrazione (1987), Vol. 33, No. 1, pp. 39—58.

Gherardi, S. and Strati, A., The texture of organizing an italian academic department, Journal of ManagementStudies (1990), Vol. 27, No. 6, pp. 605—618.

Glaser, B. and Strauss, A., ¹he Discovery of Grounded ¹heory (Chicago: Aldine 1967).Goffman, E., Asylums (Garden City: Doubleday, 1961).Heller, F., Decision-making and the utilization of competence. In: Heller, F. (Ed.), Decision-Making and ¸eadership

(New York: Cambridge University Press, 1992).Higgins, E. T. and Bargh, J. A., Social cognition and social perception, Annual Review of Psychology (1987), Vol. 38,

pp. 369—425.Jodelet, D., Les representations sociales (Paris: Presse Universitaires de France, 1989).Kingsley, J. K., Representative Bureaucracy (Yellow Springs: The Antioch Press, 1944).Lanzara, G. F., Capacita% Negativa (Bologna: Il Mulino, 1993).Lave, J. and Wenger, E., Situated ¸earning: ¸egitimate Peripheral Participation (New York: Cambridge University

Press, 1991).Levine, D., ¹he Flight from Ambiguity (Chicago: The University of Chicago Press 1985).Linstead, S., Objectivity, reflexivity, and fiction: humanity, inhumanity, and the science of the social, Human

Relations (1994), Vol. 47, No. 11, pp. 1321—1346.Macintyre, A., After »irtue (London: Duckworth 1980—90).March, J. and Olsen, J. Rediscovering Institutions (New York: Free Press, 1989).Martin, J., Deconstructuring organizational taboos: the suppression of gender conflict in organizations, Organiza-

tion and Science (1990), Vol. 4, pp. 339—359.Merton, R. K., Social ¹heory and Social Structure (New York: Free Press, 1957).Morgan, G., Images of Organizations (Beverly Hills: Sage, 1986).Moscovici, S., ¸a Psycanalyse, Son Image et Son Public (Paris: PUF, 1961).Moscovici, S., Preface. In: Herzlich, C. (Ed.), Sante& et Maladie: Analyse d’une Repre& sentation Sociale (Paris:

Mouton, 1969).Moscovici, S., On Social Representations. In: Forgas, J. P. (Ed.), Social Cognition (London: Academic Press, 1981).Moscovici, S., Description of social representations, Journal of Social Psychology (1988), Vol. 18, pp. 211—250.Nigro, G., Galli, I. and Poderico C., I Bambini e il nucleare: genesi ed evoluzione di una rappresentazione sociale

(Milano: Cortina 1989).Pipan, T., Il ¸abirinto dei Servizi (Milano: Cortina, 1996).Romagnoli, G. and Sarchielli, G., Immagini del lavoro (Bari: De Donato, 1983).Schneider, S. C. and Angelmar, R., Cognition in organizational analysis, Organization Studies (1993), Vol 14, No. 3,

pp. 347—374.Schwartz, H. and Jacobs J., Qualitative Sociology. A Method to Madeness (New York: The Free Press, 1979).Strati, A., ¹he symbolics of skill (Trento: Quaderni del dipartimento di Politica Sociale, 1985).Strati, A. and Nicolini, D., Cognitivism in organization studies. In: Ortmann,G. Sydow e J. Turk, K. (Eds), ¹heorien

der Organisation (Opladen: Westdeutschen Verlag, 1996).Tarello, G., ¸’Interpretazione della ¸egge (Milano: Giuffre, 1980).Tolman, E. C., Behavior and Psychological Man (Berkeley: University of California Press, 1951).Turner, B., The use of grounded theory for the qualitative analisys of organizational behaviour, Journal of

Management Studies (1983), Vol. 20, pp. 225—247.Walsh, J. P., Managerial and organizational cognition: notes from a trip down memory lane, Organization Science

(1995), Vol 6, No. 3, pp. 280—321.


COMPETENCE — THE SYMBOLIC PASSE-PARTOUT TO CHANGE IN A LEARNING ORGANIZATION

Documents