A COGNITIVE CONSTRUCTIONIST

Transcription

1 Fernando G. Alberti A COGNIIVE CONSRUCIONIS APPROACH ON INDUSRIAL DISRICS DSS PAPERS SOC 0206

2

3 ABSRAC Sholars and pratitioners have long reognized geographially defined regional lusters of firms i.e. industrial distrits as relevant soioeonomi engines for the survival, the prosperity and, hene, the ompetitiveness of regions and ountries. A entral issue onerns the understanding of how a population of speialized firms and individuals spatially loated in a geographial area beomes an industrial distrit, so what makes it an industrial distrit, i.e. that speial organizational form whih is wellreognized for enhaning ooperation, oordination, knowledge transfer and innovation. Prior researh on industrial distrits suggests that they are the result of the pathdependent agglomeration of small flexiblyspeialized firms, together with external eonomies, loal resoures and loal labor pools, supported by speialized institutions. However, there may be irumstanes where even assuring the onurrent presene in a geographial area of most of the typial strutural features of an industrial distrit does not make it to at as suh. In this paper, we argue that industrial distrits are not only the result of strutural features, but also the produt of how individuals pereive and evaluate the industrial distrit itself and how they pereive and evaluate their belonging to it. hus, determinants of industrial distrits may be found in individual and olletive ognitions, through whih distrit members assign meaning to their belonging to the distrit and soially onstrut the distrit, enating it. We explore our ideas in an empirial study utilizing a sample of respondents (entrepreneurs, managers, senior publi servants) operating in a soalled typial Italian industrial distrit, i.e. the textile distrit of Como in Northern Italy. We find that individuals in the area of Como mostly lak a sense of belonging to the industrial distrit, and surprisingly to be in an industrial distrit, they show a wide variety of idiosynrati ognitions, whih prevail on olletive ognitions. his has lear drawbaks on everyday behaviors of individuals whih appear to be averse towards an ative partiipation in the olletive enatment of the distrit, thus ausing the deonstrution of the industrial distrit of Como. Keywords: Industrial distrit; ognition; onstrutionism; wayofthinking; ognitive map.

4 he paper is based on my dotoral dissertation «What makes it an industrial distrit? A ognitive onstrutionist approah», defended in November 2002 and published by JIBS Jönköping International Business Shool (Sweden) in the Researh Report Series, no. 1 (2003). Previous versions of this paper were presented at the 8 th Workshop on Managerial and Organizational Cognition, in Paris (2001), the 17 th EGOS Colloquium, in Lyon (2001), the 33 rd EISB Conferene, in Milan (2003) and at the EURAM European Aademy of Management, in Milan (2003). I would like to thank Roland Calori, Leif Melin and Alessandro Sinatra for having ommented previous versions of this paper, guiding my researh with their preious suggestions. I am in debt also to Elaine Mosakowski, Rodolphe Durand, Pino Audia, Luio Biggiero, Alessia Sammarra, Carlo Salvato, Federio Visonti, Mario Minoja for their valuable omments and suggestions during onferene presentations.

5 INRODUCION Sholars and pratitioners have long reognized geographially defined regional lusters of firms i.e. industrial distrits as relevant soioeonomi engines for the survival, the prosperity and, hene, the ompetitiveness of regions and ountries (Porter, 1998). Italian industrial distrits (Markusen, 1996), in partiular, have proven to be an organizational model for loal soioeonomi development (Piore and Sabel, 1984; Pyke et al., 1990), stimulating the debate in several branhes of soial sienes: eonomi geography, eonomi history, applied eonomis and international strategy. he variety of perspetives and definitions about industrial distrits illustrates the multivoied harater of the field. However a entral issue onerns the understanding of how a population of speialized firms and individuals spatially loated in a geographial area beomes an industrial distrit, so what makes it an industrial distrit, i.e. that speial organizational form whih is wellreognized for enhaning ooperation, oordination, knowledge transfer and innovation. Prior researh on industrial distrits (Piore and Sabel, 1984; Beattini, 1990; Saxenian, 1994; Markusen, 1996; Porter, 1998) suggests that they are the result of the pathdependent agglomeration of small flexiblyspeialized firms, together with external eonomies, loal resoures and loal labor pools, supported by speialized institutions. However, there may be irumstanes where even assuring the onurrent presene in a geographial area of most of the typial strutural features of an industrial distrit (Beattini, 1979; 1990) does not make it to at as suh. his proved to be partiularly the ase when industrial distrits have been reated under artifiial onditions instead of letting them emerge spontaneously as the result of trust mehanism, mutual understanding and ooperation, i.e. as the result of individuals ognitions. In this paper, we argue that industrial distrits are not only the result of strutural features, but also the produt of how individuals pereive and evaluate the industrial distrit itself and how they pereive and evaluate their belonging to it (Borroi et al., 1998; Cillo e roilo, 1997; 2002; Viteritti, 2000; Sammarra and Biggiero, 2001a). hus, determinants of industrial distrits may be found in individual and olletive ognitions, through whih distrit members assign meaning to their belonging to the distrit and soially onstrut the distrit, enating it (Smirih and Stubbart, 1985; Weik, 1979; Berger and Lukmann, 1967). We develop these ideas in more detail below, and propose an analysis of individuals ognitions, through whih they interpret the ues from the environment, whih, onversely, is the outome of their own enatment. he literature on managerial and organizational ognition, whih fouses on individual and olletive ognitions in the enatment of environments, suits our task. Aordingly, the onept of wayofthinking, whih onsists of a number of themati sets of values, assumptions, beliefs, ideas and thoughts about a speifi domain (i.e., the industrial distrit), is applied. Individuals Waysofhinking are analyzed through ognitive mapping tehniques, searhing for the sense of belonging and olletive ognitions of individuals, whih are the major onerns of the present study. We explore our ideas in an empirial study utilizing a sample of respondents (entrepreneurs, managers, senior publi servants) operating in a soalled typial Italian industrial distrit, i.e. the textile distrit of Como in Northern Italy. he analysis of respondents ognitive maps suggests that what makes a geographial luster of firms an industrial distrit is a matter of individuals ognitions whih enat it in their everyday soial onstrutions and behaviors. We find that individuals in the area of Como mostly lak a sense of belonging to the industrial distrit, despite they are loated in that geographial area and in the same textile A ognitive onstrutionist approah on industrial distrits 5

6 business. Suh a lak dismantles the impliit assumption widely spread in the literature on industrial distrits that the physial belonging to the loal industry agglomeration implies a ognitive belonging to it. Further, our findings on the lak of sense of belonging have lear drawbaks on everyday behaviors of individuals whih appear to be averse towards an ative partiipation in the olletive onstrution of the distrit. Finally, our empirial analysis explores individuals ognitions, identifying themati sets of onstruts and exploring ore onstruts within the maps. Eah individual proved to have a unique set of ognitions, as suggested in literature (Hellgren and Melin, 1993), but, surprisingly to be in an industrial distrit, with a wide variety of idiosynrati positions, whih prevail on preise olletive ognitions. Yet, assuming a ertain degree of approximation in the omparison of idiosynrati individual ognitive maps two broad olletive ognitions emerge: one vastly diffused whih is haraterized by lak of sense of belonging, detahment from the loal ontext and individualism; the other, less and less diffused, omprises exatly the opposite ognitions of the first one. he remainder of the paper is organized as follows: In the first setion we disuss industrial distrits as enated environments, abandoning a realist view on them and embraing a onstrutionist perspetive. his is followed by a disussion of the method we used and the empirial setting where we onduted the study. Our empirial analysis follows, and we onlude with a disussion of impliations, limitations, and potential extensions of the researh. HEOREICAL BACKGROUND In literature there has been a progressive onvergene around the onept of industrial distrit towards the definition proposed by Beattini (1989a: 112) onsidered as the anonial one. Aording to Beattini, an industrial distrit is «a soioterritorial entity whih is haraterized by the ative presene of both a ommunity of people and a population of firms in one naturally and historially bounded area». Over the past two deades a growing number of sholars has dealt with the topi of industrial distrits, observing them through the lenses of different theoretial frameworks: industrial organization (Vaà, 1986; Beattini, 1979, 1987, 1989a, 1989b, 1990; Bruso, 1989), soiology (Bagnaso, 1977, 1988; Innoenti, 1985; Provasi, 1995; Parri, 1993; Viteritti, 2000), eonomi geography (Sforzi, 1989; Storper and Sott, 1989; inai Mossello, 1990; Vagaggini, 1990, Markusen, 1996), eonomi history (Castronovo, 1980; Sapelli, 1989; Sapelli and Carnevali, 1994; Amatori and Colli, 2001), politial eonomy (Fuà and Zahia, 1983; Nuti, 1992; Bramanti and Senn, 1991; Ciiotti, 1993; Garofoli, 1991a, 1992; Antonelli et al., 1988) and international strategy (Visonti, 1996; Lipparini and Lomi, 1999; Lazerson and Lorenzoni, 1999; Porter, 1998), thus ontributing in making the field a multivoied one. he review of the ontributions about the onept of industrial distrit indiates that industrial distrits have been investigated only as objetive environments, almost forgetting the insightful suggestion, made by Beattini (1989a) in his original formulation, about the onsiousness of individuals and the important role of shared beliefs. he prevailing view of industrial distrits implies the possibility to identify speifi parameters and variables, whose presene and interrelation guarantees the existene of an industrial distrit in a ertain area and the possibility to identify its boundaries (Beattini, 1989a; 1989b; Sforzi, 1990; 1991). 6 A ognitive onstrutionist approah on industrial distrits

7 Nevertheless, suh assumptions do not always permit to shed enough light on the evolution and transformation that some industrial distrits are urrently faing and, thus, «risk to bring sholars to asepti and sterile lassifiations built on parameters that are external to the phenomenon under study» (Beattini, 1987). Staber argues that muh of the previous researh on industrial distrits «has followed a formal struturalist approah without any onsideration of the subjetive aspets [ ] he model [of industrial distrits] must also apture ator s orientations towards one another and towards the soiety in whih relations are embedded» (1996: 172, emphasis added). Likewise, Beattini (2002: 488) very reently has argued that «the study of the industrial distrit as a distint form of organization of prodution has now run its ourse, and that the fous should now shift to the ongruene, in time and spae, among the elementary proesses into whih an evolving distrit an be deomposed». Among suh elementary proesses, Beattini indiates: «the formation and reprodution of a sense of belonging to the distrit» (2002: 489, emphasis added). Eah elementary proess, indiated by Beattini «more or less diretly links hanges in the produtive struture of the loal system with the psyhologial features of its inhabitants, and vie versa» (2002: 489, emphasis added). Suh an approah allows the fat that individuals ould interpret their belonging or not belonging to the industrial distrit in different ways, behaving, aordingly, in ways that may appear oherent only from their own point of view, thus subjetively. he sense of belonging, meant as the onsiousness of loal individuals to belong to a speifi industrial distrit (Beattini, 1987: 39), has an important role in affeting individuals ations, whih day by day will reiterate their belonging to it, assigning a shared meaning to their belonging through their ations. In the field of industrial distrits, there seems to be an impliit assumption (Cillo and roilo, 2002) aording to whih the physial belonging to a distrit brings with it the sense of belonging of individuals as well. In fat, a ommon harateristi of oneptual and empirial ontributions in the field of industrial distrits is their realist stane, aording to whih industrial distrits are objetive environments where the ognitive belonging is redued to a simple physial belonging. Hene, a first researh question emerges: Researh question #1: Does the territorial belonging of a population of firms and a ommunity of people to an industrial distrit imply the sense of belonging of individuals to the distrit? Abandoning a realist perspetive on industrial distrits (objetive environments) and embraing a onstrutionist one, whih reognizes the role of individuals in the onstitution of their environment (enated environments; Smirih and Stubbart, 1985), the topi of the sense of belonging beomes entral. As a matter of fat, soioultural dynamis of industrial distrits (Beattini, 1989a), suh as the industrial atmosphere, interpersonal trust, interfirm ooperation and embeddedness, assume a different onnotation depending on whether the sense of belonging of individuals is present or not. he indiation by Beattini (1987) on the onsiousness of individual distrit members goes in this diretion and merit further onsideration. he analysis of the sense of belonging alls for a ognitive aount of industrial distrits members, taking a subjetive stane (Smirih and Stubbart, 1985; Weik, 1979). his implies an analysis of individuals ognitions, through whih they interpret the ues from the environment, whih is, on the other way round, the outome of their own thinking and ating (enatment; Weik, 1993). Aording to a onstrutionist perspetive, soial reality is the outome of a proess of A ognitive onstrutionist approah on industrial distrits 7

8 sensemaking by individuals (Weik, 1993), who, interatively, soially onstrut it through their ognitions. Aordingly, the anonial oneptualization of industrial distrits offers a partial view of the phenomenon and depits it as an objetive reality. As disussed by Beattini (1987: 36) himself, suh an approah risks bringing sholars to «asepti and sterile lassifiations built on parameters that are external to the phenomenon under study». his paper proposes a view on industrial distrits as soially onstruted systems of shared meanings (Burrell and Morgan, 1979; Pfeffer, 1981; Weik, 1979). Suh approah to environments and organizations, as enated by individuals, relies muh on the influene of interpretive soiology (Shutz, 1967), soiology of knowledge (Berger and Lukmann, 1967) and ognitive soial psyhology (Weik, 1979). From suh interpretive worldview, separate objetive environments, industrial distrits in this ase, simply do not exist (Burrell and Morgan 1979). Instead, organizations and environments are onvenient labels for patterns of ativity. What people refer to as their environment is generated by human ations and aompanying intelletual efforts to make sense out of these ations (ognitions). Enatment theory abandons the idea of onrete, material reality in favor of a largely soiallyreated symboli world (Winh, 1958). If one aepts the notion that people understand the world through braketing and hunking experiene into meaningful units (Shutz, 1967; Weik, 1979), it then follows that industrial distrits provide onvenient, but also arbitrary, labels for some portions of experiene. Researhers often ignore the metaphori and symboli bases of organized life that reate and sustain these organizational ideas. An interpretive perspetive plaes these proesses and symboli entities at the entre of analysis of industrial distrits, whih, then, are defined as the degree to whih a ommunity of people (Beattini, 1989a) share many beliefs, values, and assumptions that enourage them to make mutuallyreinforing interpretations of their own ats and the ats of others. he above disussion suggests fousing researh efforts about the onept of industrial distrit on individual distrit members ognitions. heir olletive ognitions might, in fat, indiate the shared systems of meanings they assign to their belonging to the industrial distrit and to the industrial distrit itself (Gray et al., 1985), thus soially onstruting it (Berger and Lukmann, 1967). Among the different onepts appliable from a ognitive perspetive, the onept whih suits this study best is the one of wayofthinking, developed by Hellgren and Melin (1993) for the ognitive analysis of organizations and soioeonomi ontexts at large. Hellgren and Melin s reason for introduing the onept Waysofhinking is, thus, to understand the omplex but ruial relation between managerial thought and ation. he Waysofhinking onept relates more diretly to ation ompared to other ognitive onstruts, moreover an individual s Waysofhinking omprises all those ognitive shemas an individual has developed with regard to a speifi domain. With the Waysofhinking onept Hellgren and Melin try to apture the individual s ognitive and emotional mind struture as a holisti phenomenon. A wayofthinking onsists of a number of themati sets of values, assumptions, beliefs, ideas and thoughts about a speifi domain. Eah Wayofhinking represents sediments from the total life experiene and the personality of an individual, thus, as Hellgren and Melin (1993) have empirially shown, a Waysof hinking is quite stable over time and it is shaped by an individual s own personality and history. he onept of Waysofhinking may be onveniently applied even to a ommunity of individuals, for instane to industrial distrits. As a matter of fat, the sole aount of individual Waysofhinking is insuffiient to answer the researh questions of this study and to ontribute in a ognitive onstrutionist reoneptualization of industrial distrits. Hene, a refletion on the olletive dimension of Waysofhinking is deemed neessary. A 8 A ognitive onstrutionist approah on industrial distrits

9 olletive Waysofhinking is likely to be the result of the interation among individuals belonging to an organizational ontext (Smirih, 1983; LangfieldSmith, 1992), thus it may be preliminary defined as a partial overlapping among individuals Waysofhinking (Hellgren and Melin, 1993). Hene, a seond researh question emerges: Researh question #2: o what extent do individual distrit members have olletive Waysofhinking about the industrial distrit to whih they physially belong? Studies of industrial distrits through a ognitive approah have been reently arried out. Suh literature, that has foused on individuals as units of analysis drawing on ognitive theories as oneptual frameworks, omprehends, in partiular, the works by Borroi et al., 1998; Viteritti, 2000; Sammarra and Biggiero, 2001a, 2001b; Cillo and roilo, 1997, Borroi et al. (1998) foused on loal strategists pereptions of the ompetitive struture existing within the textilelothing industrial distrit of Carpi. Repliating the researh design of Pora et al. (1995) on the Sottish knitwear industrial distrit, they addressed individual deision makers as a unit of analysis, narrowing down the fous from previous studies on industrial distrits. Viteritti (2000) used a narrative onstrutionist approah to investigate the proess of identifiation of individuals in the industrial distrit of Mirandola, reading the phenomenon of industrial distrit s identity from a subjetive point of view. Likewise, even if moving from a different epistemologial and theoretial standpoint, Sammarra and Biggiero (2001a; 2001b) foused their researh attention towards identifiation proesses of individuals operating within the industrial distrit of Mirandola, hene theorizing on identity onstrution mehanisms at the distrit level. Finally, Cillo and roilo (1997; 2002), who started suh proess of narrowing down the unit of analysis to individuals operating within industrial distrits, ontributed in studying the possibility to onsider the loal musi industry of Naples as an industrial distrit, from a subjetive perspetive. However, still muh researh is needed to understand the role of individual and olletive ognitions in the soial onstrution of an industrial distrit. MEHOD In order to omply with the epistemologial approah of this study, thus to shift the researh fous towards how individuals enat the industrial distrit they are in, ognitive maps are assumed to be a suitable researh instrument, sine they provide a graphial desription of the unique ways in whih individuals view a partiular domain, i.e. a field of thought or ation (Axelrod, 1976; Bougon, 1983; Eden et al., 1983). In partiular, ausal ognitive maps, amongst the others, are the researh instrument that suits best. Causal ognitive maps allow the researher to fous on ation. Moreover, as indiated in the researh note by Cillo and roilo (2002) on how to investigate the sense of belonging and individuals ognitions in industrial distrits, ausal maps might allow temporal interonnetedness as well as a fous on the intertwined nature of thinking and ating of individuals in enating their soioeonomi environment. A ognitive onstrutionist approah on industrial distrits 9

10 Empirial Setting he hoie of the empirial setting of this researh, namely the industrial distrit of Como, has been guided by the purposes of analyzing a typial Italian industrial distrit, widely reognized as one of the most important textile/fashion arenas in the world, whih evolved over enturies in a pathdependent way. Silk reahed Como around the XVI entury, where a muh more onsolidated wool industry was already settled. At the beginning, silk prodution was managed in a handiraft manner by merhants who normally bought yarns and gave them to those small artisans owning a loom. In a few deades, textile fatories flourished not only in the ity of Como, but also in Brianza [the area between Milan and Como] and along the shores of the lake. Under Spanish rule, in the XVII entury, the silk industry of Como almost disappeared, beoming reestablished again after 1720 and onsolidating under the Austrian domination. With the arrival of the industrial revolution, the numerous waterourses running towards the lake from the mountains surrounding Como, favored the onstrution of twisting and spinning mahines in all the territory. After the Seond World War, silk prodution in Como kept on growing, stimulated by lower labor osts and by a weak Japanese rivalry. After 1952, a slowing down proess started, worsened by the spread of polyester, aryli and nylon fibers and by protetionist poliies in ommerial exhanges. In this period, the maturity phase of the distrit was already starting. New quality, standardization and demand prie, renovated the distrit of Como and led to an inreasing speialization of roles among firms. he employment rate in the Italian silk industry then began a slowingdown proess, that is still going on. he reasons for suh remarkable hanges were tehnologial, due to the introdution of lower laborintensive tehnologies, and partly linked to the emergene of soial onflits and to inreased labor osts. For these reasons, the struture of the distrit beame more omplex and single ators more speialized. Some firms began to feel the problem of hoosing whether to evolve towards a higher vertial integration rather than a simple and unstrutured organization. hese hoies were at the same time aused and stimulated by a fierer international ompetition, whih was the preursor of globalization. he filière of Como is omposed by a series of strongly independent stages through whih raw materials (silk yarns) are progressively turned into finished produts: silk fabris hiefly for women s wear (55% of the industry s business), for tie manufaturing (25% of the business) and for ladies apparel, suh as foulards, sarves, stoles, shawls (approximately 15 20% of the setor s business). his kind of organization, together with the omplexity and multipliity of the prodution stages, engendered a progressive fragmentation of the prodution hain and favored the reation of firms with a marked speialization in some (often only one) intermediate proessing stages. During the last deade the industrial distrit of Como has faed a tremendous risis that has negatively affeted the evolution and survival of the entire distrit over time. he determinants of suh a deline are several, omplex and intertwined with eah other. Furthermore, within the distrit, the risis has heterogeneously affeted different produts, phases of the prodution proess and target markets. he deline in Como began in the early 1990s, but initially both analysts and pratitioners pereived it as a business yle of reession, i.e. as a ylial risis and not a strutural deline. hat ylial reession turned out to be strutural as time passed by and the distrit delayed its reation, revealing also the existene of endogenous fators of that risis. 10 A ognitive onstrutionist approah on industrial distrits

11 Data Colletion he ognitive maps, to be used in the aftermath of this researh, were obtained in three stages: the olletion of ognitive material, the oding of it and finally the searh for ore onepts and homogeneous sets of meanings. Considering the high fragmentation of the silk industry in the area of Como it was quite diffiult to selet the respondents for this researh, who ould be onsidered key ators and also enhane the quality and the rihness of the findings. We onstruted a judgmental sample based on the advie of key informants. Individuals to be interviewed were seleted through a sort of snowball sampling tehnique, asking to eah key informant to indiate a list of other prospet ontat persons and then reiterating the proess. Aording to this approah, firstly a reliable ator involved in the industry was ontated in order to have a list of key ators of the distrit of Como. Seondly, a major representative figure, both of the area of Como and of the world of fashion in Italy, namely the President of the Italian Chamber of Fashion, was ontated in order to have another list of key ators to interview. Starting from the overlapping of these two lists of prospet ontats, a few names, ommon to both, appeared to be a good starting point both for their role and relevane in the distrit. he interviews to these ators produed other lists of names that reiterated this sort of snowball sampling tehnique. We intentionally limited our sample when all new names proposed by interviewees were already be onsidered by us, thus when the group of ontat persons was selfreferring. able 1. Main informants profile Interviewees Length of membership in the distrit Interview duration 1. Entrepreneur ~ 40 years 150 min. 2. op manager ~ 15 years 190 min. 3. Entrepreneur ~25 years 130 min. 4. op manager ~ 15 years 170 min. 5. Entrepreneur/Ativist in business assoiation ~ 12 years 100 min. 6. Manager ~25 years 60 min. 7. op manager >30 years 160 min. 8. Entrepreneur 18 years 200 min. 9. Exeutive of a loal business assoiation >20 years 100 min. 10. Entrepreneur >30 years 120 min. 11. Entrepreneur ~20 years 150 min. 12. Entrepreneur >30 years 170 min. 13. Manager >20 years 60 min. 14. Entrepreneur >40 years 100 min. 15. Exeutive of a loal business assoiation ~30 years 200 min. 16. Entrepreneur >20 years 100 min. average ~24 years total of 36 hours As a total, 30 individuals, both from industry and loal institutions, were ontated, even if only 25 of them aepted to be interviewed in depth. In order to omply with the riteria normally applied to ognitive mapping tehniques, interviews were tapereorded, so that the original language, onepts and expressions of ators ould be maintain in the ontent analysis. Nevertheless, even if for this purpose only tapereorded material only was used, the elaboration of findings was omplemented both by interviews or parts of interviews that were not allowed to be reorded and by seondary data, mainly oming from loal press. A ognitive onstrutionist approah on industrial distrits 11

12 Only 16 out of 25 interviewees allowed either to dediate us enough time to omplete the interview on all the important aspets or to be tapereorded. herefore, for a orret and useful appliation of ognitive mapping, only these 16, who allowed making audiotape reordings, entered the proess of analysis (able 1). he proess originated from semistrutured openended interviews, lasting from one to three hours eah with the seleted judgmental sample. Some ore questions or themes, funtional to the purpose of this study, were learly formulated and others were left unstrutured and openended, in order not to fore interviewees in a predetermined path, but to lead them to reveal most of their ognitive map. Moreover, the interviewer, that for homogeneity of interpretation was only one for the all interviews, tried to avoid a tehnial and speifi language/voabulary, in order not to influene the oneptual ategories used by interviewees. For instane, the term industrial distrit was not used as a ategory by the interviewer until the interviewee himself/herself had used it first. As well, the purpose of this researh was not revealed to interviewees so to avoid that this ould influene their answers either overstressing their detahment from the territory or the opposite. For the same reasons, during interviews, the intervention of the interviewer was redued to the minimum, thus to stimulate the interviewees to reveal their thoughts as muh as possible. An interview protool, built on the following themes, was used: a) the struture of the loal business system or value hain, in order to know more about agents involved, roles, ritial points and geographial span of networks; b) loal institutions (business assoiations, trade unions, training and servie entre, publi administration, et.) and their role in the reation, development and evolution of the loal soioeonomi environment; ) networks, relationships and partnerships both at a business level and a soial one, in order to identify the different kinds of exhange of information and knowledge; d) the historial evolution of the loal industrial environment in terms of firms and groups of firms transformation, modifiations of the value hain, inflows and outflows of firms and prospet evolution/revolution of the loal environment. Interviews aimed at undermining the Waysofhinking of eah individual, whih affets their deisionmaking proess and their ations, in order to undermine both their sense of belonging to the industrial distrit and their olletive Waysofhinking through whih they are supposed to soially onstrut the organizational ontext they are in. Data Analysis hese interviews were transribed into omputer readable digitized format and then ontent analyzed to identify all the onepts used by eah interviewee and the onnotation attributed to them. Subsequently, again through ontent analysis of transriptions, relationships amongst onepts and between these ones and their attributes were unraveled. No software support was used for ontent analysis. he oding of the assoiations of onepts was inspired both by Axelrod (1976) and by Huff, Narapareddy and Flether (1990). Four kinds of assoiations were investigated: positive effet, negative effet, belonging and onnotation. he entire text was first read in order to understand the individual s main points. During this phase obvious items were marked and at a seond round onepts and relationships were identified. In order to avoid wording problems, oders stuk to the original wording of the text as lose as possible, even if the exerpts inluded in the present paper had to be translated into English. A fourstepbased oding protool was applied: a) finding of the loation of a relationship in the text; b) plaing of omponents in sequene; ) identifiation of the nature of the relationship and assigning of a linkage ode; d) seletion of the portion of 12 A ognitive onstrutionist approah on industrial distrits

13 the atual text to enter in the software.he latent ontent of the texts was addressed in this analysis, in order to apture the underlying meaning embodied in the text. Less attention was given to word frequeny and syntax analysis. his approah raised a more omplex onern for validity, sine it depends to a greater extent on subjetive interpretation and judgment. Reliability was inreased through rossheking measures among the oders. During the oding phase, a list of interpretative ategories was also established by the oders. At the end of the proess, oders agreed upon the oding of data in 92% of the ases, whih is onsidered a reasonable interrater agreement (Yin, 1989; Kirk and Miller, 1986). Basing on the onepts and the relations identified, ognitive maps were realized using the software Deision Explorer. his software permitted to: enhane the graphial omprehensiveness of eah ognitive map through the use of different olors and fonts for different kinds of onepts and relations; failitate the identifiation of homogeneous areas among onepts, in order to ondut more preise analyses; eliit four kinds of relationships among onepts, assigning linkage odes (ausal, onnotative, temporal and belonging relationships); ondut speifi analyses (entrality and domain, in our ase). Firstly, the domain analysis and the entrality analysis were applied to the entire ognitive map of eah individual. he domain analysis examines eah onept and alulates how many onepts are immediately related to it (i.e. diretly linking in or out of the onept). hrough this, it is possible to be able to identify whih onepts are the best elaborated or have a high density of links around them. he entrality analysis allows a similar alulation to that of domain density to be arried out. However, this analysis alulates the results using more than one level, i.e. not just those onepts that immediately link into speified onepts, but also those whih link through them. his provides some insights into disovering the entrality of the onept in the whole model rather than just its immediate viinity. For eah one of the interpretative ategories established before, the domain analysis highlighted a key onept in eah map, basing on an assessment of the density of eah onept, i.e. their number of ingoing and outgoing relations. A onept in a mental map was onsidered to be denser than another one when it is a node for a high number of ingoing and outgoing relations ompared to others. hen, for eah key onept identified an exploration of its immediate domain was performed through the explore funtion of the software. he omparison of this immediate domain and the interpretative ategories defined in the oding phase permitted to identify several sets of onepts and relations. A set is a method of managing and organising onepts in the map. Sets an omprise any onepts or slies of the map that seem appropriate, the most ommon requirement being that of oding the onepts to relate to partiular aspets of the model. hese homogenous areas, investigated in all the 16 ognitive maps, were oded in speifi sets of onepts, ompliant both with the emphasis attributed by interviewees to some aspets (ategories) and the results of the domain analysis applied to eah individual map. he following sets were identified: weaknesses of the loal business system/territory [weak], strengths of the loal business system/territory [strong], loal information exhange proesses [info], mindset, experiene and bakground of loal ators [mind], lak of sense of belonging to the industrial distrit [nonsb], sense of belonging to the industrial distrit/territory [sb] and emphasis on individual values as opposed to ommunitybased values [indiv]. Centrality analysis was applied to key onepts of eah set to verify their relevane in the whole ognitive map. Some key onepts were present in more than a set, being ruial nodes for the sensemaking proess of agents. he entrality analysis, applied to eah set onfirmed this point. Some ognitive maps did not inlude the totality of sets identified: a few were laking the [nonsb] set, others the [sb] set and some others the [indiv] set. A ognitive onstrutionist approah on industrial distrits 13

14 hrough the use of Deision Explorer several worksheets for eah one of the 16 interviewees were produed: a worksheet representing the entire ognitive map of eah individual, without redutions or uts (Appendix 1 to 16); one summing up the list of onepts emerging from the ontent analysis of texts; one with the list of sets inluded in that speifi map and the onepts relating to every one of them; one with the results of the domain analysis applied to the entire map; one with the results of the entrality analysis applied to the entire map (these last two worksheets were organized as dereasingorder lists of onepts, respetively based on density and entrality.); several worksheets reporting the domain analysis and the entrality analysis, applied to eah set present in that speifi map. Conepts, resulted to be present in more than one set, were analyzed indepth through the onsequenes for a set analysis that enables a hek to be run between speifi sets of onepts to determine whether there are any instanes where the seond set ontains onepts that are onsequenes of the first. his analysis (alled CSE in the software) heks along a path of argument for the onsequenes rather than just examining the immediate onepts. CSE enables a hek to be run between speifi sets of onepts to determine whether there are any instanes where the seond set ontains onepts that are onsequenes of the first. EMPIRICAL FINDINGS he sense of belonging of individuals he exploration of individuals ognitive maps highlighted those showing a sense of belonging towards the industrial distrit and those not showing it. In partiular, it has been possible to split interviewees in three different groups with referene to their sense of belonging to the industrial distrit: those with a strong sense of belonging to the industrial distrit; those with a marked lak of the sense of belonging to the industrial distrit; those with more fuzzy ognitions about their sense of belonging to the distrit, even if with a lear indiation of being loser to those who do not have the sense of belonging than to the others. 14 A ognitive onstrutionist approah on industrial distrits

15 Figure 1. Exerpt from the ognitive map of Interviewee # 3 It affets positively It affets negatively It haraterises It belongs to 96.Diret ontat with problemati situations 88.Ahieving of union servies 101.Meetings among entrepreneurs 86.Ahieving of information 84.Will to at together as a larger body 112.Observane of business morality 93.Networking among entrepreneurs 85.Ahieving of ommerial support 94.Ative partiipation in loal assoiations 97.Awareness of speifi servies 95.Better knowledge of other olleagues 89.Ahieving of training servies 149.I belong to the industrial distrit for ultural reasons 105.Better relationships among entrepreneurs 107.Cultural shift 106.Generation shift 1.Como is a 200 years old industrial distrit 148.I belong to the industrial distrit for geographial reasons 109.Open mind towards the external ontext Cognitive onstruts referring to individuals sense of belonging or lak of sense of belonging were singled out within the themati sets [sb] and [nonsb]. Nevertheless, in some ases, espeially for those individuals omprised in the third of the groups indiated above, indiations regarding their sense of belonging were found in other sets as expressed by the ombination of several onstruts onneted together. Amongst the 16 interviewees, only four of them markedly showed their sense of belonging to the industrial distrit. Even if this is not the ase that quantitative reasoning matter (4 out of 16), the analysis of these four ases, and in partiular of their sets [sb], [info] and [mind], allowed a better understanding of their position. hese individuals, in fat, were members of loal institutions, two of them as exeutives of business assoiations and the other two as entrepreneurs highly involved in business assoiations and unions as eleted representatives. his suggested a relationship between those who at at an institutional level in the industrial distrit and those who show in their ognitive maps a stronger sense of belonging to the territory and the distrit. An exerpt from these ognitive maps, shown in Figure 1, visibly figures out these aspets: the sense of belonging towards the area of Como, the idea that Como is still to be onsidered an industrial distrit and the dominant role of loal assoiations and institutions, in general, in reinforing suh selffulfilling prophey. A ognitive onstrutionist approah on industrial distrits 15

16 Figure 2. Exerpt from the ognitive map of Interviewee # 1 It affets positively It affets negatively It haraterises It belongs to 2.Crisis of the industrial distrit 81.Lak of oordination in the business system 167.Erosion of values 158.Lak of ommon projets 169.Lak of vision and bravery 159.Lak of shared objetives 151.Lak of identifiation in the 16.Lak of selfesteem 165.Lak of solidity in agreements 162.Obsolesent knowhow 106.Mindset 160.People enjoy other people s problems 147.Conservatism 145.here lans among entrepreneurs 144.People from Como are introverted Five interviewees, that during the sampling proess have been identified as being key ators and experts of the industrial distrit of Como showed in their ognitive maps a dramati lak of the sense of belonging to the industrial distrit, undermining at the same time a set of beliefs opposed to the idea of industrial distrit (suh as: individualism, mistrust, detahment from loal issues, et.). An exerpt from these ognitive maps is shown in Figure 2. hrough the analysis of the sets [nonsb], [weak], [info], [mind] and [indiv] identified in the ognitive maps of these five interviewees, the reasons for a lak of their sense of belonging to the industrial distrit have been pointed out (see Figure 2 above, in partiular): sare networking among firms: even if some loalized networks of firms are identifiable, espeially between a leading firm and a group of subontrators, what seems laking, aording to some interviewees, is what Biggiero (1997) alled hypernetworks, i.e. networks of networks that are supposed to onnet the industrial and institutional tissue of an industrial distrit; lak of integration among loal assoiations: their role of atalysts and metaorganizers seems to be vain, due to the fat that they do not manage properly ommuniations among them and with business ators; ultural losure and individualism: while industrial distrits are nowadays faing important hallenges from globalization, open boundaries, new tehnologies, et., the business ators of Como seem to rest upon a historial and ultural sediment formed in hundreds of years, where the revealing ulture is based on individualisti and lose artisan ways of doing business; transationbased relationships: finally, the analysis of the empirial findings from the distrit of Como showed that one ruial determinant of this progressive lak of identifiation with the environment is traeable bak to the prevailing nature of business relationships, that seems mainly transationbased, and not based on stable strategi relationships. ransations are here meant as pure exhanges of goods, 16 A ognitive onstrutionist approah on industrial distrits

17 servies or omponents (lientsupplier relationship), while networking (as widely depited in literature on industrial distrits) is supposed to bring with it exhanges of information, skills, knowledge and shared projets. Only a more sophistiated kind of relations among loal ators is, in fat, assumed both by the epistemologial approah we moved from and by industrial distrits literature to be determinants of shared beliefs systems, and thus anteedents of a soially onstruted environment. Figure 3. Exerpt from the ognitive map of Interviewee # 8 It affets positively It affets negatively It haraterises It belongs to 63.here is the hope of other firms failures to get their lients 46.he important thing is that someone else s business is more problemati than mine 132.Collaboration is only a formal issue 2.Lak of ollaboration 48.We live in a world where assoiations do not matter to muh 45.he important thing is to hamper someone else s business 44.A In Como there is no ommuniation 47.Assoiations have never promoted networking 61.In Como you annot grow 60.In Como people do not tell you things diretly 20.Crisis 59.Deterioration of relationships 21.Several business failures he rest of the interviewees (seven individuals) did not show in their ognitive maps neither a sharp [nonsb] set nor a [sb] set. Nevertheless, through a thorough analysis of the other sets present in their ognitive maps, a sort of detahment from the territory and the distrit, together with a disillusion to at as a system and generalized sense of pessimism were learly reognizable. An example of the Waysofhinking expressed by this group of individuals is shown in Figure 3. For the reason that these seven interviewees resemble muh those who ompletely lak a sense of belonging to the distrit, in the aftermath the two groups will be onsidered together. Colletive Waysofhinking his setion reports some exerpts of the omparison made amongst individual Waysofhinking. he omparison was possible, as illustrated above, first through the identifiation of ore onstruts in eah individual map and seond through the examination of similarity patterns amongst those onstruts. his analytial proess onstituted the basis for a more holisti omparative analysis of single ognitive maps, searhing for the existene of olletive Waysofhinking. Hereafter, some exerpts taken from the most evident shared beliefs and the most evident unshared are presented and then broader onsiderations on the existene of olletive Waysofhinking are derived. hree examples of the very few shared ognitions identified are deemed representative of the ognitive analysis onduted. Quotations taken from transripts of interviews are used as a means to express them. (a) he rule of ondo. A ognitive onstrutionist approah on industrial distrits 17

18 «[ ] the sole relationships that are possible among firms have a ommerial nature, like the ones between suppliers and buyers, no information is shared, neither knowledge [ ]» «[ ] poor relations lay at the basis of the reent risis and in general they are the determinants of a intense and progressive fragmentation of the prodution hain [ ]» «[ ] we are inspired by the golden rule of ondo: don t onfide anything to your neighbors and don t treat them with familiarity [ ] it is like living in the same building, everyone in his flat, everyone fed up with the misbehaving of the others, everyone ignoring the others [ ] and like in a building sometime we meet all together to disuss our problems, but the result is that we are not sinere and we get out of those meeting with more problems than before.» «[ ] our relationships are fored by the fat that we produe not finished goods and so we have to work together, it s like staying in the same flat [ ] you know flat mates sometimes are fored to live together but this doesn t mean that they want to live together [ ]» «how an we ooperate when we ompete? Everyone is opying the serets of the others [ ]» (b) People from Como are individualisti. «People from Como and ooperation are two terms are poles apart [ ]» «I support the individualisti pride of being from Como [ ]» «[ ] everyone is foused exlusively on his business [ ] why should I bother about the others?» «Unfortunately yes! Most of my olleagues are individualisti, but [ ] we are entrepreneurs, this is how it goes [ ] it s our nature even if we should ooperate more [ ]» () Institutions are at enter stage in our ativity. «Despite we all know that our assoiations are useless, I think that an institutional support is deemed neessary in our business [ ]» «Loal institutions, suh the assoiations, strive for establishing strong and durable relations amongst firms [ ]» «Only business assoiations ould help us in inverting the ourse of events we are faing [ ] they have the important role of integrating the prodution hain» he above examples of shared beliefs and the quotations reported qualify shared ognitions amongst individuals. he few ognitions whih are shared (suh as the rule of ondo and the people from Como are individualisti ) qualify the enatment proess in a 18 A ognitive onstrutionist approah on industrial distrits

19 way whih is not benefiial to the fruitful existene of the industrial distrit. As a matter of fat, the thinking and ating of individuals operating in Como will be guided by the belief of being individualisti and not to ooperate or expet ooperation from others. Additionally, if we fous on unshared beliefs, we found a multitude of ases. Five exerpts are hosen here to be representative of the ognitive analysis onduted. In partiular, we hose those ognitions referring to the risis that affeted the textile industry loated in Como and those regarding reipes for the future, so to enhane our sensemaking on the way individuals in Como are onstruting or deonstruting the industrial distrit. As shown above, quotations taken from transripts of interviews are used as a tool to express ognitions. (a) he responsibility of loal business assoiations in the risis. «[ ] our business assoiations are responsible for what has happened, they should have know that!» «[ ] when we ome to disuss the reasons for our inability to at as a ohesive group, we annot ignore the weak support we reeive from our assoiations [ ]» «[ ] business assoiations failed their main goal, i.e. to beome a ommon arena for disussing solutions for the future and for reovering from this risis.» «we are luky beause our assoiation has helped all of us in faing the risis [ ]» «[ ] our assoiations are the expression of our needs, they prevent the misfuntioning of the entire prodution hain [ ]» «Unfortunately not many entrepreneurs onfide in the role of business assoiations operating in Como for the restart of our ompetitiveness [ ]» (b) he importane of pursuing olletive strategies. «[ ] I don t know anything about this projet you are referring to [ ] it should be another way assoiations have found to justify their existene» «If we think about the idea of promoting an integrated network of firms in Como [ ] well, the important think is to know in advane who will benefit from it and who won t [ ]» «I totally support the idea of reating an integrated network of firms in Como, we need it, we are too sparse while we should at together» «[ ] yes yes I have seen the ommerial on the ampaign to reate an integrated network of firms [ ] I just wonder why we didn t think about it before, we need to at olletively to fae the risis» () raining, learning and professional skills. «[ ] workers don t know anymore how to do things» A ognitive onstrutionist approah on industrial distrits 19