Distant Neighbors: Collective Learning Beyond the Cluster
Abstract
EconStor is a publication server for scholarly economic literature, provided as a non-commercial public service by the ZBW.
Full text
Cole, Alexander Doctoral Thesis Distant Neighbors: Collective Learning Beyond the Cluster PhD Series, No. 20.2014 Provided in Cooperation with: Copenhagen Business School (CBS) Suggested Citation: Cole, Alexander (2014) : Distant Neighbors: Collective Learning Beyond the Cluster, PhD Series, No. 20.2014, ISBN 9788793155411, Copenhagen Business School (CBS), Frederiksberg, https://hdl.handle.net/10398/8966 This Version is available at: https://hdl.handle.net/10419/208896 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/3.0/
Alexander Cole The PhD School of Economics and Management PhD Series 20.2014 PhD Series 20.2014 Distant neighbors copenhagen business school handelshøjskolen solbjerg plads 3 dk-2000 frederiksberg danmark www.cbs.dk ISSN 0906-6934 Print ISBN: 978-87-93155-40-4 Online ISBN: 978-87-93155-41-1 Distant neighbors Collective learning beyond the cluster
!"" ! "$ &"#!" !! ""& % !" #!!! #"" " "
Alexander Cole Distant neighbors Collective learning beyond the cluster 1st edition 2014 PhD Series 20.2014 © The Author ISSN 0906-6934 Print ISBN: 978-87-93155-40-4 Online ISBN: 978-87-93155-41-1 “The Doctoral School of Economics and Management is an active national and international research environment at CBS for research degree students who deal with economics and management at business, industry and country level in a theoretical and empirical manner”. All rights reserved. No parts of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage or retrieval system, without permission in writing from the publisher.
-i- The journey to completing this dissertation has at times been thrilling, but often arduous and difficult. At times it has been hard to discern just how I would arrive at this day. The fact that I have done so is a credit to the friendship and support of a large network of people who stood behind me or next to me for parts of this journey. The journey began in 2002, when I was invited to pursue this opportunity by Peter Maskell and Mark Lorenzen, who thought it would be interesting to bring an American scholar to Denmark, and arranged to support me through the DRUID. In Copenhagen, I discovered an amazing city, and in the Department of Innovation and Organizational Economics an equally amazing working environment. There I found a group of colleagues – Lars Frederiksen, Henrik Sornn-Friese, Lars Bo and many others-- who made academic research fun. The liveliness of the intellectual discussions in that department, the ease with which people shared ideas, and the general enjoyment that one felt at being able to work with these people and on these problems every day was something unmatched in my academic experience. In addition to this general appreciation, I owe particular thanks to Christian Knudsen, who guided me through a large part of my dissertation work and Peter Lotz, who helped me navigate both the bureaucracy and my self-doubts. During my PhD fellowship I also had the good fortune to visit INGENIO at the Polytechnic University of Valencia, in Spain, where I was hosted and helped by Jordi Molas and met David Barbera, who became an intellectual confidant and sometime collaborator, helping me immensely during the later stages of this work. I owe special thanks to Lars Håkanson, who believed in my work when I was ready to give up on it and guided me through the later stages of the dissertation. Thank you Lars. This would not have happened without your help. I also owe thanks to the many people in the Danish and European animation industry who shared their time and thoughts with me. Without them, this research would not
-iihave been possible. Irene Sparre, now of Wil Film, was particularly generous with her knowledge and in introducing me within the industry. Finally, there is my family, who has been absolutely amazing in their love and support. Thank you Leila, who spent a large part of her childhood watching me work on this project. Thank you Sonia, who spent many years by my side and happily moved to Copenhagen so that I could pursue this and supported me through many ups and downs. Thank you Andrea, who supported me and kept me going through the later part of the project. My Aunt Faggie, who was always there to talk to me. The biggest thanks, however, are reserved for my parents, Mike and Sheila Cole, who have supported me and believed in me throughout. This is for you.
-iii- The co-location of industry in agglomerations of similar and related firms is one of the salient features of the contemporary global economy. Over the last thirty years, a large body of theory and case-literature has addressed this phenomenon and sought to understand the advantages that accrue when industries are spatially clustered. Contemporary scholars in this tradition have focused on the advantages of face-toface interaction and the access to spatially sticky information in the form of buzz available to cluster agents. They have further suggested that the development of local conventions and a local idiom facilitate knowledge circulation and collaboration within clusters while perhaps frustrating access to outsiders. The resultant learning views of agglomeration have become dominant within the field of economic geography. In the past decade, however, this dominant view has been challenged by a counterview challenging the idea that physical proximity is neither necessary nor sufficient for economic learning. First, it has been noted that much of the learning that occurs in clusters may actually be organized through various forms temporary proximity. Secondly, it has been argued that knowledge circulates not by virtue of spatial proximity, but through participation in knowledge communities that share a basic epistemological framework and common purpose. These communities may be spatially clustered or may be widely dispersed. The research presented in this dissertation aims to contribute to this debate on the relative importance of physical and relational proximity to processes of economic learning. It does so through a qualitative study of the European Animation Industry and its attempt to build supportive networks and institutions resembling those found in successful geographic clusters, but in the context of a spatially dispersed industry. It demonstrates how through the extensive use of temporary proximity in the form of conferences, market places, and workshops, European animation was able to create a dense social fabric supporting learning and collaboration among firms that were both geographically and culturally distant.
-ivThe dissertation is developed in three distinct articles written for journal publication. These are followed by Appendix One discussing methodological issue related to the research, Appendix Two providing empirical introduction to European Animation, the object of the dissertation’s case study. They are followed by a brief conclusion discussing the dissertations findings. The first article, Distant Neighbors (also the title chosen for the dissertation) refers to the idea that relational proximity may facilitate collective learning processes even when actors are not permanently co-located. Against a growing literature that has emphasized the advantages of clustering, especially in ‘creative’ industries such as film, advertising, and music, the case of European animation is used as a ‘veto case’ to the general idea that the tacit or ‘symbolic’ nature of knowledge in such industries, means they are best organized in clusters. The case describes the means by which European animators have created the institutions of a pan-European project-network and demonstrates the positive relationship between the accumulation of local animation firms and participation in this network. The second article, Negotiating Conventions and Creating Community: The Case of Cartoon and European Animation, extends this analysis, providing an up-close account of the industry association Cartoon, and the bridging practices it developed to weave local productive communities in the animation industry into a ‘network of practice’. The paper demonstrates how this community developed the repertoire to effectively negotiate issues of different geographic and institutional contexts and collaborate in the creation of animation products. The account shows how through conversation, interaction and networking, particularly in the context of temporary gatherings, entrepreneurs in European animation have been able to overcome geographical fragmentation and build a coherent ‘world of production’. However, it also illustrates the tensions that are inherent in this project of spatial bridging. The issues raised in the first two articles through an examination of European animation are more deeply theorized in the third article, The Disjunctive Geographies of Knowledge as Skill and Knowledge as Context: Why Knowledge is both Locally
-vSticky and Globally Mobile. This theoretical article seeks to build on insights developed in the first two article and address a paradox that the same learning dynamics that are used to explain localization may also be found locally. The article suggests debates on learning have often failed to distinguish between two types of learning: knowledge diffusion, which implies the learning and adaptation of knowledge developed in different contexts; and knowledge creation through the combination of different kinds of knowledge around a concrete problem. The article argues that the literature on knowledge communities has been more concerned with the first kind of learning – the ability of a community to circulate knowledge from place to place and the geographic and organizational topographies in which this happens. Clusters, on the other hand, may be better explained through the second kind of learning, knowledge creation through combination, which is facilitated by the ease with which workers can be assembled within the region. The contribution of this dissertation is to challenge the ontological assumption that learning and learning relationships can usefully be divided between the local and global. First, the dissertation suggests that learning processes that are generally considered proper to clusters can be found at other spatial scales provided there is an incentive to interact and the proper institutional structures facilitating interaction have been created. In doing so, it also contributes to the emerging literature on transnational knowledge communities, illustrating empirically how buzz and knowledgecirculation may be organized in a spatially extended knowledge community. The case studies of the European animation industry and Cartoon, the association that is largely responsible for organizing this industry, provide evidence concerning the many contextual issues that afflict attempts at sharing knowledge between actors in different social and institutional environments. However, they also demonstrate how, in practice, through the use of periodic meetings and the practical experience of collaboration, these contextual issues can be superseded and an effective knowledge community forged.
-xiiTABLE OF CONTENTS "& "& # ! !"! $" %$ *$ " $ & ' ' %$# " $ ! " ($) $$ $"%$ # %$ " $# ##"$$ "!) ! !! ! $ & ! "! " $" %$ $ "!) %$%" !" %$ %" !E# $ %#$") "! " '$ F #$"%$ $ "$ " !%$# "* !" %$ "$&$) $) "# #$ " #$"$) $' " #$"%$%" # $"# #$ $ # %# # $ %$%" %" !E# $ %#$") "!) ! !! #! ! "!& ! ! " ! $" %$ "* " "!) #!"# %$# " % "$ $ %" ! $ %#$") $ "$ #$"* ' " !"$# !" $# %** #" %"#$# # #!# $' " $ $ & %$) $ '# "$ "$# $" % !" %$ #%## %# # "!)
-xiii- ! ! "!# $ $ !%! $ & $ ! & !& & $" %$ $$ ' # # $$$ ' # $($ "!) ' # # #!$ $($ $ * " & ' $ $' "# & ' # $"$ ' " $ #!# $ " % # " $($# $ "!) ## !$ $ %$ $ "$%$ F " $ ' "! $ $ %#$" "!# " ) $ " !$ # "!) ! " " "#%$# #%## $$ # %$%" "#" "!) % ! " ! ! " ! !! # #$%# "#" ($" &$) # #$ $ $ $" &$) $"&'# ! "$!$ #"&$ ##%# "! # "!) % !"! ! ! " ! ! " $ $ $ !" %$ !" ## &% $ "# $ $ #$ " %" ! $ " $ &# $%" $ &"$ #%!! "$ #%## "!)
CHAPTER 1 _______________________________________ INTRODUCTION
-21. INTRODUCTION Debates about economic geography, the geographies of knowledge, and innovation in economic geography have highlighted the importance of physical proximity and localization in the production and circulation of knowledge (GERTLER 2008). The large literature on ‘territorial innovation models’ and the numerous ‘success stories’ associated with clusters have seemed to offer strong evidence of the importance of co-location and proximity on processes of knowledge creation and circulation (MOULAERT and SEKIA 2003, ASHEIM and GERTLER 2005). The regional scale, or cluster, has seemed to play a prominent role in innovation and economic learning because spatial proximity creates opportunities for faceto-face interaction and the rich information exchanges this enables. Frequent face-to-face interaction has been said to intensify the circulation of information, ideas, rumors and shared narratives. This cluster “buzz” has been assumed to help keep cluster members abreast of developments in their industry and anticipate coming changes in the market (STORPER and VENABLES 2004, MALMBERG and MASKELL 2006). Over time these interactions may also stimulate the development of shared conventions, a common language and compatible interpretative schemes – referred to as relational proximity— facilitating the circulation of ‘tacit’ knowledge and promoting learning (STORPER 1997, LAWSON and LORENZ 1999, MASKELL and MALMBERG 2007). This canonical view on proximity, localization and learning is the starting point for the work undertaken in this dissertation. 1.1 LOCALIZATION AND LEARNING: THE CANONICAL VIEW The idea that there is a relationship between industrial structure, spatial relationships, and innovation can be traced back at least to the writings of Marshall on industrial districts in the late-19th century (MARSHALL 1890). During the late 1980s, following a period of debate regarding the on-going restructuring of Western industrial economies after ‘Fordism,’ the topic of industrial districts and economic agglomeration made a notable return to the agenda, as scholars asked if flexibility and clustering might hold the key to sustained economic growth and good jobs in advanced economies (PIORE and SABEL 1984, SCOTT 1986, STORPER 1989). In the early 1990s, economists began to join the debates regarding the causes and consequences of agglomeration (PORTER 1990, KRUGMAN 1991).
-3During the 1990’s a ‘learning perspective’ emerged suggesting that advantages of clustering lay in 'learning' or innovation; geographic proximity enhances knowledge creation and innovation within clusters in ways not available to agents located outside them. Territorial innovation models were developed based on the idea that innovation is a collective activity that emerges from the ‘exchange’ or ‘sharing’ of ‘tacit knowledge’ in intensive face-to-face interaction. In contrast to earlier views that generally conceived of learning as the fruits of an individual’s direct engagement with the world (an engineer, inventor, or even an R&D department were treated as individuals in this discussion), the localized learning perspective drew on a relational view of knowledge that conceived of innovation as the product of social and practical activity (ROSENBERG 1963). Through this lens, the economy can be understood as a social learning system in which dispersed, practical know-how, or capabilities, are constantly tested and refined (HAYEK 1945). Innovation requires coordinated action by different actors in the economy, so locally generated institutions and conventions play a fundamental role in smoothing collective action problems and facilitating entrepreneurship (STORPER 1997, LORENZEN and FOSS 2002). Spatial clustering facilitates information flows both because co-location makes possible inperson, face-to-face interaction and because the constant circulation of people and ideas within the cluster facilitates the emergence of a dense information ecology (GRABHER 2001). Marshall, of course, famously referred to this on-going conversation as ‘atmosphere,’ and contemporary scholars have referred to ‘buzz’ (STORPER and VENABLES 2004) or ‘noise’ (GRABHER 2002) the constant stream of information, stories, rumors, and conjecture that help inform and situate local agents within their industry. MASKELL and MALMBERG (2006, p. 6) suggest that spatial proximity “increases the likelihood of fruitful unanticipated opinions, and ideas from a broader community of informed observers, not all of whom are necessarily directly involved in the current rent-seeking activities.” The focus on territorial innovations also directed emphasis to the importance of local institutions in facilitating learning within the region. According to this view, as people create new knowledge with which to transform the world, they also create a set of tools -- the codes, institutions, and shared understandings-- without which communicative action and social learning could not happen. These accounts suggested that ‘institutional thickness’
-4facilitates information flows and interactions that generate collective learning (AMIN and THRIFT 1995), encouraging institution-building as a recipe for realizing ‘synergies’ within the cluster (ASHEIM 1996, COOKE and MORGAN 1998). This institutionalism was premised on the idea that co-location alone may be an insufficient condition for realizing synergies or external economies of learning within the region. For these synergies to occur, a proper institutional set up is needed to encourage co-operation and help ease information flows among potential rivals (STORPER 1997). By the late 1990s, the various ideas about ‘localized learning’ in clusters had become somewhat of a dominant paradigm within economic geography and were widely adopted by neighboring academic disciplines, such as economic sociology and innovation studies (GRABHER 2006) and even mainstream economics (MORETTI 2012). Given that “The Death of Distance” as a result of new broadband Internet technologies was one of the leading memes of that decade (CAIRNCROSS 2001), establishing the continued importance of physical proximity for innovation, entrepreneurship and competitiveness was an important victory. 1.2 KNOWLEDGE COMMUNITIES AND RELATIONAL PROXIMITY By the early 2000s, however, the prevailing view became increasingly questioned. The focus on learning as a largely localized phenomenon, it was argued, overestimated the importance of local relationships while under-estimating the significance of long-distance ties (OINAS 1998, RALLET and TORRE 1999, AMIN and COHENDET 2004, LAGENDIJK and OINAS 2005). The emphasis on localized learning, it was argued, had largely been developed by ‘sampling on the dependent variable’ (successful clusters), in disregard of proper comparisons with other types of regions and firms outside of clusters (STABER 1996, HÅKANSON 2005, STABER 2009). During the following decade, a growing literature sought to formulate a more differentiated notion of physical proximity, pointing out, for example, that face-to-face interaction can be organized without permanent co-location (TORRE 2008). This point is clear in the case of business travel – a fact that was never really questioned by scholars of the region—but the
-5newer literature also highlighted other forms of temporary proximity, such as conferences, trade fairs, and workshops. MASKELL, BATHELT, et al. (2006) described these meetings as ‘temporary clusters’, highlighting the similarities in function with how they believed permanent clusters operate. Later research has documented how such meetings can enable multiplexed forms of interaction, offer opportunities for observing rivals, scoping out partners, and tapping into industry ‘buzz’ in ways not unlike those found in the most lively learning regions (POWER and JANSSON 2008, BATHELT and SCHULDT 2010, SCHULDT and BATHELT 2011). Drawing on the recent literature in the sociology of knowledge, the dissenters argued that knowledge circulates through knowledge communities – variously referred to as ‘communities of practice’ (BROWN and DUGUID 1991, WENGER 1998, BROWN and DUGUID 2000),‘epistemic communities’ (BRESCHI and LISSONI 2001, AMIN and COHENDET 2004, HÅKANSON 2005) or ‘networks of practice’ (DUGUID 2008). Although the idea of communities of practice was originally proposed to describe smallscale, localized communities, organizational scholars could see analogies with larger-scale and sometimes spatially-stretched networks where people share a common enterprise, face common situations in their daily working life and speak a common ‘language’ (AMIN and ROBERTS 2008, DUGUID 2008).These communities are defined by cognitive and relational proximity; their members share a common base knowledge and epistemological culture, as well as a common understanding of the institutions and conventions that govern their behavior and interaction (BLANC and SIERRA 1999, RALLET and TORRE 1999, BOSCHMA 2005, GRABHER 2006). Like the idea of localized learning, this argument also had its problems. The ‘communities of practice’ concept, as formulated by LAVE and WENGER (1991), referred to small, localized communities and work groups. The simple substitution of ‘relational’ for ‘physical’ proximity begged the question of just where relational proximity came from and how it could be maintained without regular interaction (STORPER and VENABLES 2004).There was also considerable room for overlap and synthesis between the two concepts, as the cluster was reinterpreted as more of an open knowledge system
-6characterized by both internal knowledge exchange through ‘buzz’ and through ‘pipelines’ to distant knowledge sources (BATHELT, MALMBERG, et al. 2004). 1.3 CONTENT AND CONTRIBUTIONS The aim of this dissertation is to evaluate the literature regarding the power of spatially stretched knowledge communities to circulate knowledge and facilitate learning using a particular problematic case: the European animation industry. Creative industries draw largely on ‘symbolic’ knowledge bases rooted in local culture and conventions and are often characterized by high degrees of market uncertainty. Geographically, they tend to be organized in tight agglomerations characterized by intense social interactions (CHRISTOPHERSON and STORPER 1986, SCOTT 1999, CAVES 2000, SCOTT 2000). European animation, however, had seemingly organized these networks at a much larger spatial scale – linking different local communities in a trans-national network that seemed to mimic much of the functionality of a geographic cluster. The case of European animation thus provided a vehicle for examining and testing claims regarding the adequacy of temporary proximity as a substitute for permanent proximity and geographical clustering in organizing knowledge circulation and learning in an industrial system. Through the examination and analysis of this case, this dissertation has sought to examine and provide insight in to three related sub-questions: 1. To what extent can a project ecology organized around long-distance networks reproduce the advantages attributed to localized networks in successful clusters? 2. If knowledge communities are increasingly able to circulate knowledge and facilitate learning at a distance, in what ways might permanent proximity in clusters enhance knowledge creation and circulation in ways that a non-clustered network cannot reproduce?
-73. If the circulation of knowledge is not strongly bound by cluster geography, then what are the proper micro-foundations for explaining the clustering observed in the contemporary knowledge economy? As I will elaborate below, the evidence provided though the European animation case seemed ambiguous. In some aspects it provided evidence that trans-local knowledge communities can effectively organized the circulation of knowledge and encourage learning in ways not dissimilar to successful clusters. In other aspects, regarding the overall performance of the non-clustered network examined proved problematic. This differentiated result provided an impetus for further refining and developing theory regarding just how ‘relational proximity’ is generated in networks and the relative importance of permanent and temporary proximity in this process. 1.4 GUIDE TO READING THIS DISSERTATION The three articles that constitute the core of the dissertation are not, and were not intended to form, a monograph. However, they do reflect a learning trajectory in my understanding of the relationships between physical and relational proximity and how these affect the way an industry such animation becomes spatially organized. The first article, Distant Neighbours: The New Geography of Animated Filmmaking in Europe, is the most exploratory of the three. I here use the case of European animation to learn to what extent a network or ‘project ecology’ organized through temporary proximity might be functionally equivalent to the spatially clustered project ecologies that are typical of other creative industries. While most creative industries are organized in tight spatial agglomerations, the European animation sector consists of dozens of studios, broadcasters, distributors, producers, schools, and of course, artists, spread among the various member states of the European Union. This widely distributed community interacts regularly around a series of events – film marketplaces, workshops, etc. – and collaborates regularly in financing and producing animated films. The article investigates how productions are organized across locality, looks at how talent is sourced, and how the networking takes place. It argues that the creation of relationships beyond the local has been instrumental in
-14STORPER, M. (1989). "The transition to flexible specialisation in the US Film Industry: External economies, the division of labour and the crossing of industrial divides." Cambridge Journal of Economics 13(2): 273-305. STORPER, M. (1997). The Regional World: Territorial Development in a Global Economy. New York, Gilford Press. STORPER, M. and A. J. VENABLES (2004). "Buzz: face-to-face contact and the urban economy." Journal of Economic Geography 4(4): 351-370. TORRE, A. (2008). "On the Role Played by Temporary Geographical Proximity in Knowledge Transmission." Regional Studies 42: 869-889. WENGER, E. (1998). Communities of Practice: Learning, Meaning and Identity. Cambridge, UK, Cambridge University Press.
CHAPTER 2 _______________________________________ DISTANT NEIGHBORS: THE NEW GEOGRAPHY OF ANIMATED FILM PRODUCTION IN EUROPE !>,53<2/.37 *,-10'. 56)-*4 &85>6/ >6,/;>5B
Abstract. A growing literature on the organization of cultural products industries has highlighted their tendency to cluster in tight agglomerations. This paper explores the implications of a case, animated feature-film production in Europe, which offers a notable exception to this tendency. This case is used to more deeply explore the logic of agglomeration in cultural production and probe exceptions to this logic. Specific strategies are suggested for institutional strategies to help firms that are 'born in the wrong place' create a more supportive ecology for themselves. Key words: animation, cultural production, project ecologies, socio-spatial networks 2.1 INTRODUCTION With few exceptions, literature on the geography of cultural industries has focused on their tendency to cluster in tight agglomerations characterized by intense social interactions. 1 There are, however, other geographies of cultural production. In this article I examine one of these 'other' geographies, scrutinizing the case of animated film production in Europe. It is argued that the recent growth of animated feature-film making in Europe is largely due to the creation of an institutional framework that has encouraged and enabled cooperation and learning between geographically distant studios and allowed filmmakers to transcend the constraints of the local resource base. In effect, Europe's animation industry has created a spatially-extended ‘project ecology’ that shares many qualities with those tightly agglomerated clusters that populate much of the geography literature. To understand how feature-animation in Europe came to exhibit its peculiar spatial and organizational features and to explain why it deviates from the ideal-typical clusters described by other scholars, my analysis focuses on the feedback mechanisms between the organizational requirements of production, firm strategy and geographical patterns of production. Scholarship on cultural production has pointed to the pervasive uncertainty of product markets, the ambiguity of productive outcomes, and the urgency in production schedules, arguing that these make spatial proximity 1NORCLIFFE & REDANCE (2003) and COE (2000, 2001) are notable exceptions.
advantageous if not essential when organizing cultural production processes. The usefulness of proximity in facilitating factor markets and organizing production processes is thought to place strict boundaries on what kind of geographies of production are likely to be viable. From the present investigation of the animation industry, particularly its European variant, the author has concluded that the organizational importance of proximity has been somewhat overstated. While colocation and frequent face-to-face meetings may represent something of a best organizational practice, they do not represent the only viable one. Although, the dominant 'neo-Marshallian framework of the last twenty years has emphasized issues of coordination and learning when assessing the role of geographical proximity, firms often have other concerns such as access to particular market niches or sources of finance. In developing and exploiting organizational opportunities that do not require proximity, Europe's animation firms have opened up new strategic opportunities that both draw upon and reinforce the creation of extralocal project ecologies. The case of Europe's animated filmmakers raises important questions for how one understands the geography of cultural production and in particular the forces that lead firms in these industries to cluster. Can agglomeration be explained entirely by reference to the inherent logic of organization and coordination in these industries? If so, does the relatively dispersed geography of Europe's animation industry represent at best a 'second best' institutional arrangement appropriate under given conditions, or might dispersal have real advantages as an institutional arrangement in an age of modern telecommunications? The research strategy used in this paper is to examine critically the dominant idealtypical explanation by contrasting it with a case study, European animation firms, that does not conform to certain key predictions of the ideal-type. Divergences between the ideal-type and the actual outcomes of the case are then examined and processes generating these divergent outcomes are suggested. The case of European animation proved particularly difficult to study because the industry consists of shifting networks of small firms, many with a short life-span,
spread across an indistinct and expanding geography. Defining the exact extent of the network (e.g., which firms to include) and finding systematic data on these firms is a task daunting enough to convince this author why scholars prefer to study geographically distinct clusters of firms. Because the sector is characterized by fuzz boundaries and highly heterogeneous practices internally, the core research consisted of interviews with 22 key actors – producers, line-producers, directors, and studio heads at several medium-sized studios in Denmark and Spain, as well as the director of the Cartoon Media Program. These actors were chosen specifically because of their ability to lend insights into new developments in the field and the entrepreneurial opportunities opened by new organizing strategies. In addition, to learn about the industry the author has relied heavily on a vibrant secondary literature, particularly the insightful articles and interviews available in Animation World Magazine and other publications written by and for the world of independent animators, and has attended several ‘masters courses’ such as Cartoon Feature and Cartoon Future that aim to educate and inform people in the field, as well as conferences such as SIGGRAPH (in Los Angeles) and others more narrowly focused on European animation, particularly Cartoon Forum. These courses and conferences have provided an occasion for innumerable informal conversations, which mostly reinforced by occasionally gave reason to question the information gained from interviewees. Finally, Tim Westcott of Screen Digest has done the most comprehensive surveys of European animation and the paper has relied heavily on his research. The paper begins with a literature review that sets out what the author believes to be the dominant ideal-type in the geographical literature on cultural production, explains the logic of this ideal type, and examines some of the challenges that have been posed to it. Against this background, the case study is then developed in four sections: a general background on the sector and the competitive position of the firms studied is followed by two sections that explain how geographical constraints are overcome in organizing projects and input markets, and how they are overcome in the production process. These two sections are followed by a section illustrating how firm strategies both shape and are shaped by the emerging networked structure of the industry. This analysis suggests that the specific spatial pattern adopted by flexible networks of
firms may be path dependent and highly sensitive to initial conditions. The conclusions return to the theme of multiple geographies and discuss the merits of a strategy of linking disparate pockets of resources for firms that do not have access to a critical mass of resources locally. 2.2 THE GEOGRAPHY OF CULTURAL PRODUCTION This study seeks to add to a bourgeoning literature on the geography of cultural industries, particularly those that are focused on producing media content such as film, recorded music, advertising, and video-gaming. There are important reasons why these industries have become the object of increased scholarly scrutiny. To begin with, growing demand, fuelled by the fact that consumers have more money to spend and time to dedicate to leisure activities, means that these industries are growing relatively quickly. Perhaps more significantly, the importance of creative labour in these sectors – scholars such as CAVES (2000) actually call them ‘creative industries’2-- means that they embody organizational practices and face organizational dilemmas that are becoming increasingly common in other sectors as the knowledge economy spreads (SCOTT 2000). This dual agenda is evident in the seminal research of Michael Storper and Susan Christopherson on the Hollywood film industry (STORPER 1989, STORPER and CHRISTOPHERSON 1987, CHRISTOPHERSON and STORPER 1986). Their work, which was part of a broader research program on the geographical implications of flexible specialization, focused on how, from the 1950s onward, the vertically 2Matters of definition have become somewhat tricky when writing about these industries. While Scott uses the terms ‘cultural industries’ (a definition that emphasizes the importance of culture as both an input and output) other authors such as Caves use ‘creative industries’ (a definition that emphasizes the role of creativity in the productive process), ‘entertainment industries’ (see HESMODNHALGH D. (2002) The Cultural Industries. Sage Publishers, London.), or ‘content industries’ (a definition that points towards the markets that they compete in). For the purposes of this article I will use the term ‘cultural industries’, however, these can be seen as a subset of ‘creative industries’ since it is the problem of organizing creative labour processes that is supposed to provide the greatest constraints on their geographical organization. Should the reader wish to pursue definitional matters further, an excellent discussion is available in MARCUS C. (2005) Future of Creative Industries: Implications for Research Policy. European Commission..
integrated studio system of Hollywood’s classic period was gradually replaced by a vertically disintegrated production system characterized by producers and service providers temporarily collaborating around particular projects. This new organization of production exhibited what they dubbed a ‘split locational pattern’. On the one hand, the restructuring of the industry into numerous specialized firms collaborating on specific film projects meant that deal-making took on new importance in the industry. This restructuring created considerable advantages to locating in and around Los Angeles where one could keep tabs on the ever-shifting coalitions making key decisions about projects and use ‘face time’ both to gather important information and to negotiate the details of deals. On the other hand, vertical disintegration and the ability to recombine specific resources according to the needs of each project combined with new, more mobile equipment to make it easier to move filming and production activities to sites outside of Los Angeles. Successive studies of different cultural industries have reaffirmed these findings. In numerous case studies and more general theoretical reflections, SCOTT (2000, 1999, and 1997) shows how the spatial processes identified by Christopherson and Storper are common to a number of ‘cultural industries’. GRABHER’s (2002a, 2002b) work on the London advertising industry suggests that much the same is true for creative industries more generally. Scott and Grabher point to the same factors of localization first described by Alfred Marshall and now familiar through the relentless emphasis they are given in the neo-Marshallian discourse that currently dominates economic geography: the creation of a local pool of labour, the availability of specialized inputs that firms can easily access, and an environment that is conducive to information sharing (MARSHALL 1890). Because cultural production often takes place in temporary organizations that assemble unique constellations of resources only to dissolve when the project is finished, agglomeration is fundamental to providing labour-market flexibility (LORENZEN and FREDERIKSEN 2005). Following this line of reasoning, both Scott and Grabher emphasize the important learning effects that emerge from the localization. Scott argues that it is within placebased communities, aided and re-enforced by institutional infrastructure such as
schools, training establishments and apprentice programs that mutually complementary skills are developed and the norms and conventions that tie them together are reproduced. The constant social intercourse in these places facilitates communication and creates an industrial ‘atmosphere’, which encourages innovative activities. In a similar vein, Grabher discusses the way that clusters of localized communities generate ‘buzz’ or ‘noise’, a kind of information that people are aware of without really having to search for it consciously. Borrowing a key idea from the literature on situated learning (LAVE and WENGER 1991), he notes how localization facilitates the kind of ‘hanging out’ and ‘peripheral participation’ that allow newcomers to become acculturated into the norms and conventions of a creative community. Because of its emphasis on local sources of competitive advantage, the importance of non-local ties has until recently been somewhat neglected in this literature. In a scathing critique of Storper and Christopherson’s interpretation of Hollywood as a Marshallian industrial district, ASKOY and ROBINS (1992) argued that the major film studios, which have integrated into multi-national media conglomerates, continue to exert effective power over the industry by effectively monopolizing distribution outlets and thus exerting effective financial and creative control over content producers. This view has largely been accepted, and SCOTT (2002) has argued that the geography of the “New Hollywood” derives from the overlap of localized productive networks with the centres of control for global networks of finance and distribution. According to Scott, rather than undermining the power of a dominant agglomeration such as Hollywood, multinational distribution extends its geographic reach by moving the cultural products it creates to ever-wider markets. Borrowing a notion proposed by AMIN and THRIFT (1992), KRATKE (2003) suggests that the geography of such industries consists of Marshallian nodes of cultural production articulated within Global Networks of distribution. In line with neo-Marshallian theories of clusters more generally, recent contributions have also emphasized the importance of local productive systems having strong links to non-local sources of knowledge. This point was made emphatically by
GRABHER (1993), who pointed out that a local production and communications systems risk ‘lock-in’ and stagnation when it cuts itself off from outside sources of ideas and innovation. A common refrain now is that successful regional economies are characterized by both by strong local processes of imitation, adaptation, and learning and strong connections to non-regional knowledge sources. NACHUM and KEEBLE (2002) have used the Soho (London) media cluster to argue “Why Being Local Just Isn’t Enough.” BATHELT, et al. (2006) have suggested that a healthy cluster requires both ‘buzz’, characterized as information that is available by just being there, and ‘pipelines,’ specific investments made to access knowledge from distant partners. The problem with such a metaphor is that it conflates geographical proximity with social proximity defined by networks (TORRE and RALLET 2005; BOSCHMA 2005). This is a bit like claiming that my close friends all live close to me while the people that I left in California are distant relationships. While proximity may facilitate social closeness at a particular historical juncture, in a world where local relationships can be quite heterogeneous and people move around, they are not always the same thing. Using Marshallian language, we can say that firms benefit from external economies, particularly input-output relationships and knowledge spillovers that are not localized. It would seem that this outcome is particularly likely for less-favoured regional economies for the simple region that exogenous resources are likely to be of higher quality than anything that the region can generate endogenously. COE’s (2000, 2001) work on the Vancouver film industry has described just such a situation. Vancouver’s film and television industry emerged largely as a peripheral site where Hollywood producers, enticed by lower labour costs and a cheap Canadian dollar, could locate ‘runaway’ TV and film productions. Since most production was organized in Hollywood, and the industry also benefits from various national schemes to promote the industry, any analysis of the world that film producers live in necessarily require that local, national, and international scales must all be taken into consideration. According to Coe, Vancouver can be seen as a ‘hybrid agglomeration’ that combines the qualities of an export-platform, totally dependent on outsiders to organized production, financing and distribution, with the greater local autonomy of a
Marshallian industrial district. Coe’s case study exemplifies a point made by PHELPS (2004) that we should be careful not to collapse the idea of external economies into localization economies because the relationship between these is likely to change with the development of new social, institutional, and technical infrastructures facilitating communication. In terms of the case examined in this paper, the author claims is that key actors in the animation industry have both incidentally and intentionally created a degree of relational closeness and that this relational proximity is actually increasing, despite the fact that they often live in different countries. There are few precedents in recent literature for describing such spatially dispersed project ecologies although NORCLIFFE and REDANCE’s (2003) work on artisanal comic book producers describes a networked production system that strongly resembles that formed by European animators. Because European animators lack the financial resources to make a product of the scale and quality necessary to compete with top Hollywood productions, these producers rely on their intimate cultural connection with national and linguistic niche markets and their relationships to national sources of financing as their main source of competitive advantage 3 . However, this means that they must organize other externalities – specialized inputs, sources of cutting-edge knowledge, and to a lesser, but not insignificant degree, labour markets – from a distance. The recent growth in animated film production across Europe is a testament to their ability to do this. 3Both TSCHANG T. (2003) The Effects of Product Development and Cultural Sourcing on the Location of Creative Industry: The Case of the US Computer Game Industry. Singapore Management University, Wharton-SMU Research Center of Singapore Management University. and AOYAMA Y. and IZUSHI H. (2002) Hardware gimmick or cultural innovation? Technological, cultural, and socia foundations of the Japanese Video Game Industry, Research Policy 32, 423-44. elaborate the thesis that the national cultural milieu is the most important locational factor for the creators of culture.
among animators with producing a feature film cannot be underestimated. Many filmmakers, producers and other creative talents were fulfilling lifelong dreams by moving into film production. Although it is a complicated and financially risky undertaking, feature films are also a good basis for building up a studio because they keep hundreds of people employed over a relatively long period of time and are likely to draw in the best talent. 2.4 CONSTRUCTING THE MARKET FOR INPUTS Scholars of the post-Chandlerian network economy have tended to hold contradictory views about the factors enabling a switch from corporate to more market-based forms of governance. Some have pointed to the widespread adoption of marketsupporting institutions such as formal specifications, which allow for a measure of modularity in the productive process (LANGLOIS 2002, STURGEON 2002). Modularity rests on the possibility of adopting standard interfaces between different parts of the productive process, greatly reducing the cost of exchanging information, thus allowing customers and suppliers to interact almost as if they were operating on spot markets. Others have argued that vertical disintegration rests critically on an increased use of social mechanisms and relationships that facilitate the exchange of information and the formation of trust between transacting parties (see SABEL and ZEITLIN (2004)). Studies of creative industries have almost uniformly fallen into the later camp. Inputs into creative parts of the productive process are characterized by what CAVES (2000) calls “infinite variety”: they differ along many different dimensions of quality and may not be evaluated by all consumers in just the same way. Therefore, they are only imperfect substitutes for each other. Pervasive uncertainty in output markets and the creative nature of the labour process mean that coordination requires a great deal of reflexivity. Tasks are constantly modified in light of contributions from other workers. This situation may frustrate attempts at imposing modularity. If so, the informational complexity and high levels of reflexivity that characterize these markets explain why cultural producers tend to agglomerate in dense clusters close to final
consumers or in the case of mass-media industries, the distribution agents who get to decide what consumers want. This standard argument, however, seems only partially to capture the ways that coordination is achieved in creative industries. First, while cultural industries are characterized by the widespread use of social mechanisms in governing the market, the social relationships underpinning these markets may be stretched across great distances and need not be confined to certain localities, even if in practice they often are. This distance spanning is achieved by embedding transactions in ongoing relationships, arranging periodic face-to-face meetings, and where possible, using modern communications technologies as a partial substitute for such meetings. Secondly, the line between ‘strategic’ interactions that require face-to-face communication and those that do not is to some extent subject to organizational choices. By separating the core decisions that require reflexivity, decision-making process can be simplified. For these types of interactions, actors in the animation industry do draw on standardized understandings of roles and shared metrics in structuring their relationships, thus reducing the amount of information that parties must exchange. This process can be seen in the thriving global market for animation services, particularly in Asia and Eastern Europe. After a long period of absence from the geographical literature, the role of periodic markets in tying together spatially dispersed actors is again receiving renewed attention. 7 Drawing on notions from time-geography developed by THRIFT (1977) and PRED (1981), NORCLIFFE and REDANCE (2004) describe how comic book artists gather from different rural and urban location in North America to form a ‘periodic social economy’, meeting up with each other at annual comic conventions or weekly readings the local comic shop where they may engage in intense periods of sociability before dispersing again to take up their creative labour in relative solitude. 7BOGGS J. S. (2005) Theorizing the Frankfurt Book Fair (Or: Why transaction cost analylsis still matters), Regional Studies Association Conference on Regional Growth Agendas, Aalborg, Denmark.(2005, p.8) notes that “The role of periodic markets was once a staple in Economic Geography,” and offers a review of the thriving literature on the subject that, until 1942, was produced by this field.
These authors point out that generating the social interactions that underpin economic transactions in the cultural industries does not require that comic book artists be permanently co-located, only that they co-locate at some point in time. Similarly, MASKELL et al. (2006), three scholars who have explored the knowledge generating aspects of firm clusters extensively, have noted that ‘temporary clusters’ and ‘permanent clusters’ are functional substitutes for each other in many respects. In short, trade fairs, markets, and other periodic gatherings seem to be an essential aspect of many industrial ecologies, particularly those for which the need to be close to dispersed customers makes it impossible for producers to co-locate. Members of Europe’s animation community meet, share ideas, and negotiate deals at markets such as MIPTV, Cartoon Forum, and Cartoon Movie. Cartoon Forum, which is focused on animation for television, and Cartoon Movie are smaller gatherings focused exclusively on animation where several hundred potential investors come together with the aim of uniting animation producers together with potential distributors and investors in order to negotiate financing for new projects. These meetings combine intensive work sessions in which projects at various stages are pitched to potential investors and co-production partners with business meetings, socialising and sightseeing. As such, these temporary meetings create places where the kinds of ‘strategic information’ that Storper and Christopherson point to as the anchor of the Hollywood agglomeration can be exchanged. Such socializing allows producers to ‘sound out’ projects and learn about trends from distributors such as what the demand for new television shows is likely to be in the coming year. One commentator at the Cartoon Forum noted: The most effective part of the Cartoon Forum is probably the bar. The nicest thing about Cartoon Forum is the one thing that everybody rails against; they always choose some God-forsaken remote place that takes you a whole day to get to! You can
guarantee the hotel doesn’t have email points and has faxes that turn into some sort of scroll when they finally get them delivered to your room. 8 What is interesting about these comments is that they stress that ‘being away’ from somewhere else may be just as important as ‘being there’ in that it intensifies social experiences by taking people out of the normal patterns of their work and home life. Periodic markets are supplemented by other meetings such as film festivals where artists, fans, critics and producers show and discuss work, Masters’ courses that introduce new skills and technology to professionals are offered, and student exchange programs that encourage that cement relationships between future professionals from different parts of Europe operate. Together these kinds of programs ensure that a vibrant animation community can stay in touch, share ideas and its members can inspire each other. Recently, economic geographers have begun to pay more attention to the important role played by longer distance networks that are formed as people move from place to place in tying together labour markets and diffusing information about entrepreneurial opportunities (COE and BRUNELL 2003, AMIN and COHENDET 1999, SAXENIAN and HSU 2001). Such networks can constitute ‘small worlds’ in which people enjoy the informational benefits of relational proximity even over large geographic distances. The more artistically ambitious segments of the animation industry have always constituted a fairly small world. During the 1970s and 1980s the slack demand for animators even in the United States created a generation of ‘gypsies’ who moved from city to city and country to country working on any project that would keep them employed for a while (SITO 2004). With the animation boom of the late 1980s and 1990s, many animators in the United States were able to find permanent employment and to settle down. However, in Europe it is still common for animators to move from place to place following jobs and to spend long periods away from home. Such migrations have been a powerful force creating a shared sense of 8John Bullivant, cited in KENYON H. (2001) A Look at Europe's Cartoon Forum with John Bullivant, Animation World Magazine..
community among animators from different countries; several people in the industry who I interviewed referred to the animation industry as ‘a family’, united by its shared love for their art. Ironically, given the emphasis on localization in the current geography literature, it is the experience of being ‘alien’ or out of place that often cements a common bond among animators. Because travel removes them from other social obligations such as going home to their families, their emotional dependence on each other is intensified. While informal relationships obviously play an important role in forming the market, particularly when it comes to knitting together deals and defining projects, it is important to recognize that other, more formal mechanisms are also used to structure transactions in the animation industry. Rather than the opacity and indeterminacy that an emphasis on reflexivity suggests, less strategic interactions may in fact be relatively straightforward. Using a combination of institutionalized understandings about how jobs are done and pragmatic instruments to monitor each other’s compliance, animators seem able to collaborate even in situations where there has been little time to generate shared understandings and their appears to be little basis for trust. When pulling a project team together, actors in the animation industry are able to draw on a well institutionalized set of roles, each of which is responsible for certain tasks. The diverse skill set and roles that make up an animated production are fairly standardized and, with some exceptions, these standards tend to be the same from place to place. The standardization of roles on a project team makes it much easier for employers to evaluate the skills and experience levels of employees and to describe the requirements of a given job (CHRISTOPHERSON 2003). Much of the technology, and hence the techniques that go along with this technology is also quite standardized. As one producer told me, “What is behind all of this [outsourcing] is not technology. It is that we speak the same language. English, yes. But we also speak animation.”
Where standard descriptions fail to provide the full range of information necessary for a transaction, pragmatic measures may also be used. In particular, as with other creative industries, it is not uncommon for a producer to ask for a work sample before outsourcing work to a new studio. The relationship between partners will then typically develop slowly, with more or more complicated tasks being sent out if earlier ones are completed in a satisfactory way. Such practices are far from perfect and it is commonplace for relationships to dissolve in the middle of a production because the work submitted by a service studio isn’t deemed of high enough quality or because the contracting studio is asking for more work than they pay for. However, combining social mechanisms such as reputations or face-to-face interaction, standardized metrics, and pragmatic instruments such as screen tests, the market seems to work well enough. 2.5 ORGANIZING PRODUCTION: CREATIVITY, TAYLORISM AND DISTANCING Once the necessary resources have been assembled, the animation producer must then effectively coordinate their use in the productive process. Presumably creative labour processes are particularly difficult to organize over long distances. Since tasks cannot be specified precisely, organizing tasks may require a great deal of negotiation as well as back and forth interaction, both of which will raise transaction costs. These interactions may be considerably easier to bring to a satisfying conclusion when the parties can meet face-to-face, where misunderstandings can quickly be cleared up and ruffled feathers smoothed over, and feasible solutions easily be demonstrated. The possibility for producing animated film in widely dispersed production sites also rests on the ability to separate the creative, iterative parts of the process in which the project is conceived and given shape from the more routine tasks of ‘rendering’ this creative vision in animated footage. As has long been understood, the ability to separate creative or conceptual tasks from routine production is a basic pre-condition facilitating outsourcing.
Similar to other audio-visual products, the process of making an animated film is divided between creative development, pre-production, production, and postproduction, after which the finished product is sent to a distributor. 9 In animation the normal procedure is for some or all of the production process to be outsourced, depending on the film’s budget and the quality aimed for by the producers. The easy divisibility of the labour process, in which creative work is separated from more routine tasks, assures the creative control can effectively be maintained even when the work is done in a distant location. Thus, for lower budget productions such as madefor-video films and television serials, a common procedure is to prepare a preproduction package meticulously and then outsource the entire production process to some lower cost producer, often in Asia. For higher quality productions, certain parts of the animation such as key animation may be kept in-house while intertwining and ink-and-paint work are outsourced. Even within the pre-production processes, a certain amount of distancing is possible, if not desirable. Irish scriptwriters, for instance, may be employed to write Danish feature animation while living in Dublin and only be physically co-present with the rest of the pre-production team on occasion. Because this kind of creative work requires a great deal of solitary labour, the benefits of co-location are not clear-cut. The articulation and coordination of creative and more routine tasks occurs in two ways: through parameter-setting, or specification (LENT 2001); and by using supervision, which requires both direct observation and dialogue in which tasks are redefined locally. Specification involves the creative worker in setting parameters for other workers such that the latter will have sufficient guidance in executing the task. Sometimes these specifications leave some room for creativity, as when a key animator has to use his artistic talents to bring the lead character to life, but often they define routine tasks, such as ink and paint work that can be easily executed by a worker with little understanding of their general significance. 9Readers who are interested in knowing the details of the animation production process can find a reasonably complete description in Pixar’s annual report to investors (10K), which is available through their website, www.pixar.com. WINDER C. and DOWLATABADI Z. (2001) Producing Animation. Focal Press, Amsterdam. also offer an excellent description of the production process.
Specifications are embodied in the animation in a number of boundary objects (STAR and GRIESEMER 1989), documents and artefacts that are used to communicate the parameters in the absence of the directing artist. The script, the story-board, colour keys, timing sheets showing the precise timing of certain scenes, and exemplary pictures are all physical artefacts that enable the spatial and temporal disarticulation of the production process. In some cases, such as timing sheets, the parameters strictly determine the actions of the directed worker. In other cases, such as when a creative director includes examples of an art work that is supposed to be inspiring or to exemplify certain stylistic elements she requires, the boundary object is merely directive, providing a ‘good enough’ sense of what is needed that the artist can fill in the rest. However, there are limits to the amount of coordination that can be achieved simply through the setting of specifications. Coordination in the context of an animation project usually requires a large amount of managerial oversight and interaction between workers. Direct communications is important not only to assure specifications are met, but also to negotiate and adjust in those situations where they don’t meet local contingencies. More importantly, direct communications are necessary for communicating commitment and intent and making sure that everyone understands their contribution to the overall goal of the project. Obviously, this may be achieved in different ways. Sometimes a phone call is sufficient. In other situations there is no substitute for getting on an airplane and visiting distant production sites. Current developments in computing and communications have made distributed production easier and cheaper, and more importantly widened the range of tasks that can be outsourced. However, given the ease with which creativity and rendering can be separated in animation, outsourcing was heavily used well before the development of the latest generation of communications technologies (SCOTT 1984). What has changed with new computing and communications technologies also is that these technologies make it possible to circulate the artefacts that bind and guide their labour. Thus the use of FTP sites and mirrored servers which allow a producer and director in two different places to look at the same piece of work in real time have
recently replaced the fax machine as a key instrument for moving documents around. Whereas a few years ago the ‘bible’ containing the storyboard, colour keys and visual guides for animators was literally a book, today it might well be a database which is updated as different scenes are completed, thus allowing animators located in different parts of the world to reference each other’s work and achieve greater continuity. 10 Thus the greater ease of circulating artefacts, particularly those artefacts that play a central role in defining and coordinating the tasks of different labourers, complements and extends the well-known features of communication technologies such as email and video conferencing in allowing people to communicate across time and space. 11 The upshot of this is that geographically distributed production, already a wellestablished practice in the animation industry, is only likely to increase with the intensive use of communications technologies and the ever-decreasing cost of airtravel. For the European feature film industry where productions are often funded with budgets one-tenth the size of the average Disney feature film, the ability to outsource large parts of the production process easily as well as the existence of competent and inexpensive subcontractors around the world created by previous rounds of outsourcing are necessary conditions allowing for economical productions. Where today’s independent producers both in Europe and in North America and Asia go beyond earlier generations is in the variety and quality of the work that they 10 For an account of how the Toronto based animation company, DKP, is using high-powered mirrored servers and databases see SHACHTMAN N. (2004) The Secret Life of Off-Duty Characters, The New York Times, New York.. 11 See (BAKER, et al. 1999) for a more in depth discussion of creative collaboration using communication networks in the media industries. They claim that: Studios and companies involved in animation work are another industry segment that are early adopters of network technologies. The main reason is that suitable artist-technologists are not available in sufficient numbers in the primary work locations, so these organizations are setting up work groups where the talent is. For example, a separate group of animators based in San Francisco will be linked with the main animator group in LA, allowing for more of the production to be carried out in parallel. One of our participant organizations had a project that involved a lot of model work on a spacecraft. Part of the work was done in London, part in Ardmore in Ireland, and part in Los Angeles. As one UK post-production company executive noted, “It’s a question of being able to work where the talent is rather than being frustrated by the physical limitations." (p. 320)
outsource. While outsourcing began in the 1950s as an extension of Taylorist work practices, in the new century co-development, the collaboration between distant parties on the more creative tasks that define the production, is quickly taking hold. 2.6 FIRM STRATEGY AND NETWORK STRUCTURE: SKILL CONTAINERS AND SHIFTING COALITIONS The key players in feature animation projects are a number of small and mediumsized studios, most located in major cities such as London, Paris, Copenhagen and Munich, but many located in smaller urban centers such as Santiago de Compostela in Spain and Galway in Ireland. The strategic problem these firms face is to minimize the risks inherent in making large, sunk investments when demand is uncertain while at the same time accumulating the capital, skills and reputation that will allow them to compete for more ambitious projects such as feature films. For such firms, maintaining the full employment of a core group of workers and finding challenging projects that develop their capabilities and enhance the firm’s reputation are balanced against short-term profit motives in taking on projects. Following KRISTENSEN (1994), these studios can be described as ‘skill containers’, in that they are loosely structured collections of artistic and managerial workers whose skills are readily adaptable to the requirements of different projects. The standard accumulation and growth strategy for these firms consists of leveraging success and recognition into new and better project opportunities. Success can be defined in different ways: a wonderfully creative project or the managerial savvy to bring a project in on-time and budget, for example. What matters is that the firm gains a reputation that can differentiate it from competitors. 12 The prospect of better pay and more interesting work is then used to attract and retain a better labour force. Ideally a single success can trigger a virtuous circle by helping the studio build its competencies while providing a reputation that is visible to distributors, financiers 12 FAULKNER’s (1983) Music on Demand: Composers and Careers in the Hollywood Film Industry. Transaction Books, New Brunswick, New Jersey. work on Hollywood studio composers is the classic reference on the accumulation of reputation in free-lance markets.
BIBLIOGRAPHY AMIN, A. and P. COHENDET (1999). “Learning and adaptation in decentralised business networks.” Environment and Planning D: Society and Space 17(1): 87-104. AMIN, A. and N. THRIFT (1992). “Neo-Marshallian nodes in global networks.” International Journal of Urban and Regional Research 16(4): 571-587. AOYAMA, Y. and H. IZUSHI (2002). “Hardware gimmick or cultural innovation? Technological, cultural, and socia foundations of the Japanese Video Game Industry.” Research Policy 32(3): 423-444. ASKOY, A. and K. ROBINS (1992). “Hollywood for the 21st century: global competition for critical mass in image markets.” Cambridge Journal of Economics 16(1): 1-22. BOGGS, J. S. (2005). Theorizing the Frankfurt Book Fair (Or: Why transaction cost analylsis still matters). Regional Studies Association Conference on Regional Growth Agendas. Aalborg, Denmark. BOSCHMA, R. A. (2005). “Proximity and Innovation: A Critical Assessment.” Regional Studies 39(1): 61-74. CAVES, R. (2000). Creative Industries: contracts between Art and Commerce. Harvard University Press, Cambridge, Massachusetts and London, England. CHRISTOPHERSON, S. (2003). Project work in context: regulatory change and the new geography of media. Environment and Planning A: Society and Space 34(11): 2003-15. CHRISTOPHERSON, S. and M. STORPER (1986). “The city as studio; the world as back lot: the impact of vertical disintegration on the location of the motion picture industry.” Environment and Planning D: Society and Space 4(3): 305-320. COE, N. (2000). “The View from out West: embeddedness, inter-personal relations and the development of an indigenous film industry in Vancouver.” Geoforum 31(4): 391-407. COE, N. (2001). “A hybrid agglomeration? The development of a satelliteMarshallian industrial district in Vancouver’s film industry.” Urban Studies 38(10): 1753-75. COE, N. and T. BRUNELL (2003). “’Spatializing’ knowledge communities: towards a conceptutalization of transnational innovation networks.” Global Networks 3(4): 437456.
FAULKNER, R. R. (1983). Music on Demand: Composers and Careers in the Hollywood Film Industry. New Brunswick, New Jersey, Transaction Books. GRABHER, G. (1993). The weakness of strong ties. The lock-in of regional development in the Ruhr area. The Embedded Firm. On the Socioeconomics of Industrial Networks. G. Grabher. London and New York, Routledge. GRABHER, G. (2002a). "Cool projects, boring institutions: Temporary collaboration in social context." Regional Studies 36(3): 205-214. GRABHER, G. (2002b). “Feeble sector, robust practice: project organizing in the new media.” Environment and Planning A: Theme Issue 34(11): 1911-26. HESMODNHALGH, D. (2002). The Cultural Industries. London, Sage Publishers. KENYON, H. (2001). A Look at Europe's Cartoon Forum with John Bullivant. Animation World Magazine. KRATKE, S. (2003). “Global media cities in a world-wide urban network.” European Planning Studies 11(6): 605-628. KRISTENSEN, P. H. (1994). “Strategies in a volatile world.” Economy and Society 23(3): 305-334. LANGLOIS, R. (2002). “Modularity in technology and organization.” Journal of Economic Behavior and Organization 49(1): 19-37. LAVE, J. and E. WENGER (1991). Situated Learning: Legitimate Peripheral Participation. Cambridge, Cambridge University Press. LENT, J. A. (2001). Overseas animation production in Asia. Animation in Asia and the Pacific. J.A. Lent. London, John Libbey Publishing. LORENZEN, M. and L. FREDERIKSEN (2005). "The management of projects and product experimentation: examples from the music industry." European Management Review 2(3): 198-211. MARCUS, C. (2005). Future of Creative Industries: Implications for Research Policy, European Commission. MARSHALL, A. (1890). Principles of Economics. Toronto, MacMillan and Company Limited. MASKELL, P., H. BATHELT and A. MALMBERG (2006). “Building global knowledge pipelines: The role of temporary clusters.” European Planning Studies 14(8): 997-1013.
MOINS, P. (2003). Panorama of European Animated Feature Film. Animated World Magazine. NACHUM, L. and D. KEEBLE (2002). “Why being local just isn’t enough.” Business Strategy Review 13(1), 37-42. NORCLIFFE, G. and O. RENDACE (2003). "New geographies of comic book production in North America: the new artisan, distancing, and the periodic social economy." Economic Geography 79(3): 241. PHELPS, N. A. (2004). “Clusters, dispersion and the spaces in between: for an economic geography of the banal.” Urban Studies 41(5/6): 971-989. PRED, A. (Ed) (1981). Space and Time in Geography: Essays dedicated to Throsten Hagerstrand. Lund, Sweden, CWK Gleerup. SABEL, C. and J. ZEITLIN (2004). “Neither modularity nor relational contracting: inter-firm collaboration in the new economy. A critique of Langlois and Lamoreaux, Raff, and Temin.” Enterprise and Society 5(3): 388-403. SAXENIAN, A. and J.-Y. HSU (2001). “The Silicon Valley-Hsinchu Connection: technical communities and industrial upgrading.” Industrial and Corporate Change 10(4): 893-920. SCOTT, A. J. (1984). “Territorial Reproduction and transformation in a local labor market: the animated film workers of Los Angeles.” Environment and Planning D: Society and Space 2(3): 277-307. SCOTT, A. J. (1997). "The cultural economy of cities." International Journal of Urban and Regional Research 21(2): 323-339. SCOTT, A. J. (1999). “The cultural economy: geography and the creative field.” Media, Culture and Society 21: 807-817. SCOTT, A. J. (2000). The Cultural Economy of Cities : Essays on the Geography of Image-Producing Industries. London and Thousand Oaks, SAGE. SCOTT, A. J. (2002). “A new map of Hollywood: The production and distribution of American motion pictures.” Regional Studies 36(9): 957-975. SHACHTMAN, N. (2004). The Secret Life of Off-Duty Characters. The New York Times. New York. SITO, T. (2004). Will There Still Be an Animation Business When I Graduate? Animation World Magazine.
STAR, S. L. and J. R. GRIESEMER (1989). “Institutional ecology, ’translations’ and boundary objects: amatuers and professionals in Berkeley's Museum of Vertebrate Zoology, 1907-39.” Social Studies of Science 19(3): 387-420. STORPER, M. (1989). “The transition to flexible specialisation in the US Film Industry: External economies, the division of labour and the crossing of industrial divides.” Cambridge Journal of Economics 13(2): 273-305. STORPER, M. and S. CHRISTOPHERSON (1987). "Flexible specialisation and regional industrial agglomerations." Annals of the Association of American Geographers 77(1): 104-117. STORPER, M. and R. WALKER (1989). The Capitalist Imperative. Oxford, UK and Cambridge, USA, Blackwell. STURGEON, T. J. (2002). “Modular production networks: a new American model of industrial organization.” Industrial and Corporate Change 11(3): 451-496. THRIFT, N. (1977). “Time and theory in human geography.” Progress in Human Geography 1: 65-101. TORRE, A. and A. RALLET (2005). "Proximity and Localization." Regional Studies 39(1): 47-60. TSCHANG, T. (2003). The Effects of Product Development and Cultural Sourcing on the Location of Creative Industry: The Case of the US Computer Game Industry. Wharton-SMU Research Center of Singapore Management University, Singapore Management University. WESCOTT, T. (2002). European Feature Animation. Cartoon: European Association of Animated Film. WINDER, C. and Z. DOWLATABADI Z. (2001). Producing Animation. Amsterdam, Focal Press.
CHAPTER 3 _______________________________________ NEGOTIATING CONVENTIONS AND CREATING COMMUNITY: THE CASE OF CARTOON AND EUROPEAN ANIMATION
-52Abstract. This article examines the construction of transnational ‘learning communities’ in an industry that is fragmented by various local and national institutional environments. It examines the process through which the European animation industry created a coherent “world of production” despite its origins as separate national industries shaped by different institutional environments. The paper highlights the important role played by an industry association, Cartoon, in developing common conventions and facilitating learning among European animation firms across geographic and institutional contexts. In particular, the analysis highlights Cartoon’s use of temporary meetings as a space where common understandings could be negotiated, new norms developed, and transnational relationships worked-out. 3.1 INTRODUCTION Every art, as indeed, every collective endeavor, is underpinned by a set of shared rules and conventions that help guide participants’ actions, make them mutually intelligible, and suggest common solutions or ways of doing things (BECKER 1982). These conventions may emerge organically through the give and take of daily collective work or may be created quite consciously and imposed by managerial fiat. Over time, conventions become embedded in routines and practices; they come to define “the way things are done” in a workplace or company, and through the movement of workers and managers may spread and become generalized beyond the boundaries of the firm to an entire industry. They form an essential part of the architecture of collective action. Conventions and their close cousins, institutions, have played an important role in contemporary debates in economic geography. As conventions arise in a given social, economic, and institutional context, they are shaped by unspoken assumptions about power-relationships, the regulatory environment or the market. Conventions that make sense in one place may be hard to adapt in another place and may, indeed, seem quite irrational or counter-productive there. Scholars such as STORPER and SALAIS (1997), GERTLER (1995, 2003, 2010) and LORENZEN and FOSS (2002) have argued that the development of localized conventions often facilitates the coordination and knowledge development between local actors while frustrating close
-53collaboration and learning between actors from different regions or countries. This is particularly important in entrepreneurial or creative endeavors, where the inherent uncertainty and inventive nature of the enterprise means that shared but usually tacit understandings and reference points are needed to ground effective communication, coordination and mutual endeavor. By the early 2000s, however, the idea that knowhow mostly develops through local or regional relationships and the focus on localized learning was increasingly being challenged; scholars such as ALLEN (2000), COE and BUNNEL (2001), and AMIN and COHENDET (2004) pointed out that knowledge creation occurred within knowledge communities that to an increasing extent were being organized globally. They pointed to the ways that even tacit, highly contextual forms of knowledge could circulate globally within and between ‘communities of practice’, ‘epistemic communities’, or professional communities located in different parts of the world. The literature on global knowledge communities has built on the idea that people engaged in similar practices – engineers, musicians, or psychologists, for example— also share the codes, frameworks, tools and practices of their profession that facilitate their understanding of each other’s work and engagement in productive interactions. While some interpretations of this view seem to suggest simply that location does not matter and that knowledge will easily flow within such communities regardless of where the members are, a more nuanced version of this approach acknowledges that in reality, different conventions and institutions may fracture communities of practice, but argue that under the right conditions differences these fractured communities may be bridged by a variety of mechanisms. This article contributes to the debate on the geography of contemporary knowledge communities through a case study of the European animation industry and of the industry association, Cartoon. We argue that Cartoon has acted as a bridging organization, helping local animation communities successfully to align themselves and to form a transnational network of practice. The study of Cartoon is particularly interesting for two reasons. First, existing theory tells us that creativity and inventiveness in an industry producing cultural products like animation should be
-54particularly dependent on locally shared and often-tacit cultural understandings and institutional norms. Indeed, the academic literature on the geography of cultural industries has focused on their tendency to cluster geographically in tight agglomerations where industry players can easily engage each other, interact and form localized networks of learning (LORENZEN and FREDERIKSEN 2005, SCOTT 1997, STORPER and CHRISTOPHERSON 1987). Secondly, the European animation industry represents an especially challenging environment for studying these issues because it is constituted by firms from relatively small nation-states, each with their own language and culture, posing unusually challenging barriers to generating the relational proximity to successfully collaborate or share knowledge. As a consequence, Cartoon provides a useful case for exploring the middle ground in which the local contexts in which knowledge is embedded are negotiated and for understanding successful intermediation within cultural industries. While the literature on communities of practice has often focused on the role of global firms as intermediaries in creating common conventions and connecting knowledge communities (HILDRETH, KIMBLE and WRIGHT 2000; HILDRETH and KIMBLE 2004), this case highlights the role played by an industry association and suggests that geographers need to pay attention to the global associational economy and the role that associations may play in structuring global learning communities (HOWELLS 2006). 3.2 ORGANIZING LEARNING IN GEOGRAPHICALLY DISPERSED COMMUNITIES In recent decades, the importance of knowledge as a competitive asset has come to be recognized across a range of academic fields, from economics to organization and strategy to geography (COWAN, et al. 2000). Within this context, one of the fundamental contributions of economic geography has been to document the localized nature of much knowledge creation. This finding has found support in econometric studies of innovation and patent studies, but most importantly, the importance of ‘localized’ knowledge has emerged from the hundreds of studies of industrial districts, clusters, learning regions and other kinds of localities that have been the
-55mainstay of economic geography over the last two decades (see BRENNER and MUHLIG (2007) for an overview). In explaining the localized nature of knowledge production and the apparent paradox that much competitively valuable knowledge in today’s global economy is ‘sticky’ (e.g., it adheres to its setting of origin), it has been common to invoke the concept of ‘tacit knowledge’. This line of argument emphasized that while codified knowledge such as industrial recipes, programs, and blueprints could circulate far and wide (and thus become ‘ubiquitous’) the tacit knowledge required to make competitive use of this codified knowledge would tend to stick to the people and places where it was developed (MASKELL and MALMBERG 1999a). AMIN and COHENDET (2004) call this an ‘islands of innovation’ view of knowledge in that it suggests that knowledge creation, or innovation, mostly happens on local islands of dense interaction and that this knowledge is then codified at which point it becomes globally mobile. These authors challenge the implicit dualism in this view between locally sticky, ‘tacit’ knowledge versus globally mobile codified knowledge (also see ALLEN 2000).As HÅKANSON (2005) points out, the dualism between locally tacit and global codified knowledge rests on a fuzzy concept of knowledge, and particularly the use of ‘tacit’ knowledge as a discursive catch-all, defined mostly in contrast to its supposed opposite, codified knowledge. When using the concept of tacit knowledge, reference was often made to Michael Polanyi, and particularly his writings on processes of scientific discovery (POLANYI 1966). Polanyi, of course, noted that knowledge may emerge as a hunch before the knower is able to articulate this knowledge and fully communicate it. Focus on the process of articulation highlighted by Polanyi (as contrasted with already articulated ‘codification’) may actually offer important insights into the spatial organization of creative or inventive industries such as animation, in which products first take shape when collectivities of artists -- writers, character designers, musicians, and programmers-- may literally engage in acts of creation that are analogous to those described by Polanyi (HÅKANSON 2007). Such collective creativity may well benefit from the close interactions of participants including physical gestures and the
-62The success of these early initiatives can be seen in the rapid rise of a significant industry focused on producing animation for television. From just 80 hours of programming in 1988 when Cartoon was formed, the industry grew to produce over 1200 hours in 2003. As the industry grew, Cartoon began to sponsor new initiatives. Among these were quarterly masters’ classes where professionals and industry newcomers could learn about the latest developments and trends in the industry. They covered topics such as film financing, feature production, and the changing technology of animated filmmaking. In late-1997, Cartoon also started a new co-production market for animated feature films called Cartoon Feature. As with television, co-productions in which studios from different countries work together were widely adopted in financing and producing these films; over half of the animated films produced in Europe between 2000 and 2010 were international co-productions. Figure 1: # of European Feature Film Productions (1990-2010) Figure 1 shows the large increase in feature films starting in the later half of the 1990s.
-63The frequent use of co-productions is one of the key features of European audiovisual production and is particularly pronounced in animation, where the visual and narrative characteristics of the medium allow the product to travel well. Another important characteristic is the extent to which programming relies either directly on public funding or on mandates that require broadcasters to purchase some part of their programming locally. The combination of these mechanisms means that European animation is often produced through a coalition of studios, each of which is able to tap into local financing sources to fund their work. While many European coproductions involve interesting, creative collaborations, they are almost inevitably motivated by a search for territorially specific funding sources. Producers in France find co-producers in Belgium or Luxemburg to make up part of their budget. Danish producers look for Irish, German, or Norwegian partners. And so on. The result has been an extraordinary internationalization of the animation industry in Europe and a very heavy reliance on multi-studio, multi-location productions. This increased internationalism was accompanied by an increased institutionalization of the industry as well, with Cartoon playing an important role as mediator/facilitator. Prominent in this regard has been Cartoon’s role in enculturating new participants into common “European ways” of making and financing animation. It has also been pushed along by a general climate of globalization, and the increased opportunities to learn of standard industry practices. Finally, it has emerged somewhat organically, although not accidentally, through Cartoon’s efforts to create cross-cultural dialogue within the industry. The following three sections discuss the work of Cartoon in detail, focusing on the tools it used to forge a European animation community and how these facilitated cooperation and learning across the industry. 3.4 THE CREATION OF STANDARDIZED WORK PRACTICES One problem that frustrated a greater cooperation between European animation professionals in the late 1980s was that local animation studios, which had grown up
-64in different national environments, used widely different practices to organize production. Creating a truly European animation industry required that local studios find a way to overcome local differences and work with studios using very different creative and cultural conventions. Some of these conventions are cultural and aesthetic. For example, European cultures have different ideas on issues such as what kind of humor is appropriate for a five year old. They may also have an attachment to different visual styles. This kind of cultural difference, while presenting a barrier of types, is also a font of creativity and aesthetic expression. It is the very Englishness of Nick Parks’ creations, such as Wallace and Grommit, that make his animation appeal to audiences around the world. Attempts to erase the cultural specificity and aesthetic quirks in European animation have largely been derided for creating a bland, American-style of animated storytelling that yet isn’t quite as good as the American ideal-type it aspires to emulate. The issue of creating international appeal while staying true to one’s local stories and ideals continues to bedevil the industry, as it does European filmmaking more generally. When Cartoon formed in the late 1980s, a whole range of practical issues needed addressing. As noted above, many studios in Europe simply didn’t have the managerial and organizational practices in place to deal with large-scale production, such as the production of a regular, 22-minute TV series. This would be a particular problem when companies sought to cooperate in or co-production arrangements where work would be moving from one studio to another. Remembering those days, one participant of one of the few co-production projects carried on told me “In 1981 we had a huge debate about how many minutes long each episode of x should be and how many episodes should be in a series.” (Interview with R.W., 9/12/2011) Industrial scale animation as practiced by U.S. producers had a well-established set of conventions regarding how things should be organized and managed. Animation in the United States had long been organized using a detailed division of labor first created for the mass production of footage by Joseph Bray, who was directly inspired by Taylorist practices of a fixed and rigid division of tasks (SITO 2006).
-65In the early 1990s, Cartoon sponsored the publication of the European Animation Industry’s Production Handbook, a 400-page guide to various practices for producing and managing the production process. To research the book, three experienced animators were sent around Europe to visit studios over the course of two years before writing began. The stated aim of the book was to create “a springboard for the harmonization of standards in search of future European standard. (ERNEUX 1991)” The very attempt to create these standards is revealing both of the state of the industry as it existed when Cartoon was formed in the late-1980s and of the kinds of conflicts that a project to unify the industry would entail. As one of the authors recounted, What was interesting was that the British animation industry seemed to me to be organized very much like a cottage industry. I was appalled at the lack of organization of some quite big-name companies. (Interview with R.W., 9/12/2011) In both Britain and in Spain, which was a major site for low-cost service work at that time, putting out systems were widely used and workers would only show at the studio to collect assignments or submit approved work. Basic tools such as sign-off sheets to track artwork and make sure that sequences were completed were often not utilized. “In early 90’s … usually studios would print their own folders and there would be something like a table on the front of it in which each person filled in and you signed your name, so you took some responsibility for what was in that folder. And I know at a later stage, in Hahn Film, they even added a barcode system as well. He invented a sort of barcode thing so he knew where everything was at any given time. But when you went to British studios, everything was in the folder, but there was no writing on it. And it was just left to someone to remember who got what and what stage it was at. (Interview with R.W., 9/12/2011)” This situation seemed fairly typical of European animation except in Ireland and France. In Ireland, the animation industry had absorbed workers from a hand-full of large American productions that located there during the 1980s. In France, industrial
-66habits seem to have diffused when the French animation house, DiC, moved to California, exposing many French animation professionals to the American system. “At the time we were putting together The Technical Bible, there was some sort of industrial standard, as practiced by US companies like DIC, and adopted by many French companies. Some French producers would say that they came up with it in the first place. This system was very unpopular in the UK, where it was felt to be rigid, industrial, and anti-creative. Animation in the UK has always clung to its artistic pretensions, and this was reflected in its (not too efficient) production methods. (Personal email from R.W.)” ‘The Bible’, or Handbook, as it was often called, went on to describe the various systems of charts, folders, dope-sheets, bar-sheets and other standard tools for coordinating large production in animation at that time. However, according to contemporary accounts, later confirmed in interviews, there were also disagreements regarding the extent to which standards should be imposed. One solution, preferred by one of the Handbook’s authors, was to codify clear guidelines and prescribe best practices. A different idea, which ultimately made it into the book, was to describe the different practices in use and give multiple examples of how a given issue might be addressed, in hopes that professionals could understand their purpose and adopt them in part or whole according to their local needs and preferences. As one author stated, “Attempting to impose standards from the outset is also running the risk that under these circumstances some people may have no use for the Bible. I moreover do not believe that we must necessarily harmonize everything, both techniques and working methods, in order to achieve our aims, namely much more intense cooperation between European production companies. (ERNEUX 1991)” As a result, part of the book was written as a dialogue, discussing the merits of different systems instead of presenting a single standard. Ultimately, this frontal attack on the issue of standardizing conventions may not have changed anyone’s behavior. However, the book did provide detailed information on the practices in use, so that studios did not have to reinvent these every time they entered in production.
-67Managerial and workflow practices have improved over time and the industry has become “professional” since the early 1990s. The main form for this change, however, has been organic and gradual change, rather than a radical shift prompted by the publication of standards. As competition has increased, studios with efficient production methods have prospered at the expense of those that did not. This is particularly true for service-providers, for whom efficiency is a core selling point. The vast increase in co-production has also helped diffuse better practices as studios have opportunities to learn both skills and working methods from their production partners. “Our studio grouping,” one informant told me, “brought us into close company with studios with quite different ways of working. We adopted several ideas from some studios (the French and Belgian) and tried to put the Germans on a more efficient course. (Interview with R.W., 9/12/2001)” The circulation of labor during the past 20 years has also played an important role in exposing animators to the best contemporary practices. This circulation has occurred in Europe, but it also includes large numbers of animators who have worked in the United States before returning home. Currently, the recruitment of animators from California’s large studios by studios in developing countries attests to the important role of individuals as carriers of these cultural practices. More than 15 years after the publication of ‘the Bible’, as this book was called, considerable differences between practices were still in evidence, as a conversation with a Danish Production Coordinator reveals: There are a lot of national differences. In some places you would pay people for footage meaning you would pay them according to how much they get a piece. That’s not something that is usual in Denmark. That means that you would have a difference. You would have the same job needing to be done and depending on whether it was done in Germany in Ireland or in Denmark the same job would be paid differently. … The same hierarchy was still apparent…but when you work with other people whom you hadn’t worked with before you will find out that one drawing is not necessarily defined as one drawing in two different studios. The definition of one type of work process might vary from one place to another. So we needed to tune into each other’s ways of how the work was defined in different places. (Interview with I.D., 6/10/2003)
-68The spread of Digital Asset Management systems has helped assure standard managerial practices among studios in different locations. STEINMULLER (2000) has written about the important role of the software industry in creating large incentives to codification and standardization. While their use is still far from universal, programs for managing digital assets are now fairly standard on larger projects such as feature films. These systems allow for a vast reduction in managerial resources and the easier management of digital assets, even when these are being produced at studios in different locations or even different continents. Actor network theory has emphasized the importance of ‘non-human actors’ in networks and the ways in which interacting with similar non-human actors may exercise an isomorphic effect on practices in different contexts (FAULCONBRIDGE 2010; TAKTEYEV 2009). In a very real sense, the use of Digital Asset Management and other software actually enforces a change in local practices as local studios adapt their practices to work more effectively with the software. The tensions between standardization and local adaptation that have played out in European animation are a miniature version of those that afflict globalizing capitalism more generally. In corporate systems, this struggle is resolved by the power of management to impose standards (or refrain from doing so). In a networked system, such as the heterogeneous networks that define European animation, a less direct method seems to have been more effective. This has involved the creation of an environment where people are exposed to new knowledge, to encourage their cooperation, and to ultimately trust in their ability to take and adapt what they need or find themselves at a disadvantage on the market. 3.5 PERIODIC MEETINGS, BUZZ, AND SHARED UNDERSTANDINGS In contemporary geography of creative industries, the social aspects of agglomeration have come to play an important role in explaining ‘competitive advantage’ and have been seen to provide an important explanation for why such industries agglomerate. Beyond the ease and convenience that living in a place like Hollywood (SCOTT 2002, 2004) or London’s advertising village (GRABHER 2002b) might provide,
-69professionals in the film or advertising industry choose these places because of the access to fresh information and rumor they provide and the desire to participate in the ‘conversations’ that define the bleeding, innovative edge of the industry. To be in an ‘it’ place is to know things before people elsewhere do, get a sense for how others perceive this information, and to be able to anticipate the future (STORPER and VENABLES 2002). Economically valuable information has a ‘social life’ (BROWN and DUGUID 2000) and being around the people who form the center of an industry affords the opportunity to take part in this social economy. Information becomes socially sticky when it is either highly contextual to a given a social circumstance – so it only makes sense in that context – or perhaps will only be exchanged when the receiving party is known and their reactions can be gauged. Taken individually, most such information is of little value. However, collectively, this concatenation of information, rumor, opinion – this ongoing conversation about the ‘who’, ‘what’, ‘why’ and ‘how’ of local industry – may provide important contextual clues that help an individual understand and navigate a creative field. Alfred MARSHALL (1890), in his famous passage on industrial districts, poetically wrote that “the secrets of industry are not secrets at all, but exist, as it were, “in the air.” Contemporary followers of Marshall have used ‘buzz’ or ‘noise’ to describe much the same phenomenon (BATHELT, et al. 2004; GRABHER 2004). In industries that are only emerging or where the race for novelty is relentless, buzz or noise are valuable guides helping individuals to focus and invest their energy and resources. Confusion about the nature of the knowledge exchanged through buzz has led some scholars to argue that important innovations don’t rely on the haphazard, non-directed communications. This misses the point, I believe, conflating directly rentable knowledge – such as patents or know-how— with the broad awareness of the creative field that is necessary for an individual to effectively engage with others. This later kind of knowledge, which is highly contextual, is part of the knowledge that the individual uses to more effectively make sense of the field.
-70However, buzz is not simply a local phenomenon. Information circulates through many routes in today’s economy. Periodic meetings and markets where professionals gather from various places to meet, exchange information, learn about developments in their profession or industry or trade, play a particularly important role in the circulation of information and the question has arisen to what extent this kind of ‘temporary proximity’ may substitute for the more permanent co-location of industry in a given locale (BATHELT, et al. 2004). In their study of comic book writers and artists, NORCLIFFE and RENDACE (2003), use the terms “periodic social economy” to describe how artisans use meetings and conventions to stay in the loop, engage in professional networking, and catch up on their industry. SKOV (2005) has illustrated how fashion fairs create the conditions for framing and comparison between products that help participants makes sense of their market environment. BATHELT and SCHULDT (2010) note that the co-presence of people from all parts of industry – suppliers, producers, users, retailers, media representatives, and interested experts— stimulates Buzz, which helps orient actors around common conventions and understandings. The opportunity for intensive face-to-face interaction, create an opportunity for very rapid and open-ended communications. The possibilities for observing and listening-in offer contextual clues and stimulate questions that might not have otherwise come to mind. Conversations are fomented in which different interpretations of events in the industry are put forward. At the center of Cartoon Media’s work with the animation sector are a series of meetings where the industry from around Europe meets up to learn, do business, but also play together. The most important event over the years has been the Cartoon Forum and Cartoon Movie, where producers, financiers, distributors and some members of the press gather to discuss the financing and co-production arrangements for animated television and feature films respectively. The Cartoon Forum was first held in 1991 and was first major initiative undertaken by Cartoon. It was formed at a time when the expansion of European television was creating increased demand for animated programming. However, European producers, locked into financing film projects in their home territories, were unable to meet the budgets to compete with
-71American and Japanese programming, which was widely available in syndication. Cartoon Movie, which was first held in 1998, aimed to replicate the success of the Forum and encourage early investment and co-production agreements in feature films. Cartoon Forum and Cartoon Movie are both structured around a public pitching process. Distributors, financiers and producers gather in lecture rooms, and producers with projects are allotted a 30-minute time slot to make a pitch. The public nature of pitching provides an opportunity for producers to both broadcast their own creations to a large audience and observe how that audience responds to other pitches. This is different from how most film markets work, where the meetings are all in private. The public pitch creates the possibility of generating buzz around a project; when a broadcaster stands up in front of others and shows enthusiasm for a show her enthusiasm immediately signals to others to take note. As they do, the chance of more investors coming on board is increased. The very public discussions of the merits or drawbacks of a given idea also provide important information for everyone in the room, and often filter out to other participants in the industry through subsequent conversations. As LAMPEL and MEYER put it (2008), in the emergence phase of fields and industries the focus of field configuring events such as Cartoon meetings, “is on the processes that transform a disparate set of organizations and individuals into a community of organizations that partake of a common meaning system.” Indeed, over the years, one effect of regular meetings has been the emergence of shared standards within the industry around technical issues such as the length of broadcasts. Producers who wanted to export their product in the 1980s faced the issue that different national broadcasters followed different conventions regarding the length of individual broadcasts and number of broadcasts in a series. What worked for one market wouldn’t necessarily work for another. The Cartoon Forum is often credited for changing this situation. In the early days of Cartoon … there was absolutely no coordination among broadcasters in those slots at all. I think one of the results of the Cartoon
-78reinforced by the continual circulation of the same faces through different events throughout the calendar year. POWER and JANSSON (2008) have pointed out that trade fairs are not simply oneoff events; that often the same people meet up over and over at different fairs over the course of the year. The yearly calendar for many European animation professionals — especially the producers and distributors who are involved in the ‘deal-making’ side of the business — are filled with meetings, where often the same familiar people meet up and re-acquaint at a regular basis. Several producers who met at the Cartoon Forum mentioned that they would see each other again at MipCom two weeks later. One producer described the European circuit as follows: February is the Berlinale; March is Cartoon Movie; April is MipTV, May is Cannes Film Festival, June is Annecy, September is Cartoon Forum then October is MipCom. Seven events. Then you have Cartoon Masters. Sometimes I go and sometimes I don’t. And some other things might spring up. So basically my work as I see it is travelling around, meeting people, mingling. Once you’re in the loop with the proper guys, you can jump from one to another. (Interview with G.S., 9/8/2011) Such an intensive circuit of meetings clearly has a cost. During the recession of 2008 and 2009 most companies cut back on their attendance or sent smaller groups of people. The ‘buzz’ or purely social aspects of meeting were notably reduced as companies focused more intently on getting deals done. Exploitation of opportunity became the order of the day; exploration was at least temporarily, pushed to the side. Importantly, however, the meetings have, over time, had the effect of creating a shared sense of relational proximity. Scholars of industry districts have often focused on the ‘trust’ that is built up through constant interaction and the sense that one is dealing in a community where one’s reputation will catch up. This is also the case in European animation. As one informant told me: This is a small world. So if you f*** up once you can be sure that you have jeopardized your own reputation for further possibilities. So it’s in
-79everybody’s interest to live up to standards. So it’s not that you have to put everything in writing. Many of us would prefer to put things in writing just to be on the safe side. But there is a much more dangerous consequence of not delivering: It is that basically you will not get any more jobs. That’s the risk. You will be the last one of anybody’s studio’s choice. If you haven’t delivered on time you be the last one anyone calls next time. (Interview with I.D., 6/10/2003) Nonetheless, although repeated businesses are quite normal, the animation industry seems to be characterized by ‘fast trust’ as well; people meet, feel an affinity for each other’s style of work, and decide to collaborate, staking a great deal on the commitment of partners who are barely known. Fast trust, as GRABHER (2002a) points out, is fairly typical of industries where freshness and variety are highly valued and job roles are well defined. Fast trust, however, rests on a sense of shared values and understandings – membership in an imagined community. This community has been built up through the constant circulation and interaction of its members over many years. The most notable kind of circulation have been the temporary meeting places of markets and masters courses and the familiarity that repeated interactions has bred among the animation profession. These temporary meetings have given a strong impetus to the networks that have come to define the industry. Labor mobility has also played an important role in building a common set of understandings, and many animators spend time working in studios abroad. This too has been encouraged by Cartoon through the development of internship programs placing students in studios across Europe. Thus, for example, an award winning Irish studio now finds most of its labor not in Ireland, but from leading animation schools in Denmark, France and Italy (Interview with P.I., 9/22/11). Labor mobility has also been set in motion by the simple fact that co-productions and collaborations between studios give professionals experience of working together. Often, when projects are over, professionals from partner studios in other countries are called to work on the next project. Because of this, the animation industry is filled with migrants, or as one professional put it, ‘gypsies’ who are willing to move where the work is. This is the
-80temporal shadow that has been cast by temporary meetings and the networking that occurs in them. In this way, once the process of international collaboration was set in motion, it has tended to sustain itself. It’s a very little business. So people know each other and there is a tendency that … people within the business travel with the projects…which is of course not very good for people if they want to have a family and if they want to be home in their place, wherever that might be. But we know each other, we know people all over Europe. The rumour spreads that “now there is a project coming up here and they seek good talent. People apply and we might request people because we know them from previous projects. So it’s a small business. There is also a tradition among studios in Europe that when the deadline gets close and people might be running a little bit late they will start looking for companies that they can outsource work to. And then those companies will say “Sure, we can take five seconds of animation and clean it up for you.” And then we will get people in Copenhagen Area. And then those guys in say Italy or Germany will get to know those guys in the Copenhagen area because they will be working on the project through A-Film. So it’s a small family in many ways and we know each other. If you look at the credits for many of the bigger European and International productions there will be names repeating themselves. (Interview with I.D., 6/10/2003) In this way, meetings at the Cartoon Forum, Cartoon Movie or Masters class are not isolated events; the networks that are formed at these events may cast long shadows that shift the spatial structure of the industry. 3.8 THE DOWNSIDE OF CARTOON: RENT-SEEKING THROUGH COPRODUCTION A great deal of the academic literature on networking and community formation has assumed a neat, functionalist form, eliding the conflicts that often exist within both networks and communities (GRABHER and IBERT 2006). While I have held Cartoon up as an example of a business association that has helped generate common institutional understandings and facilitate network formation across national contexts, I did encounter people who cast doubt on the value of events such as the Cartoon Forum and on Cartoon’s ability to create a truly viable animation industry in Europe.
-81All of these critics had directly benefited in their careers and businesses from Cartoon’s programs. The main issue that critics point to is the poor record of European animated feature films at the box office when compared to American feature animation or even coproductions between American distributors and European animation studios. Of the top eight European animated films in terms of box office during the last decade, seven are co-productions between European studios and American distributors (Arthur is the only exception). All of these films were made in Europe, but each of them was made outside of the “European system”. Although some of them benefited from publicsupport schemes, for the most part they were privately financed; and none of them appeared at Cartoon Movie to look for financing. In contrast, the large majority of films that have appeared at Cartoon Movie have performed poorly at the box-office. With a few notable exceptions, they have also traveled poorly between different European countries and few managed to earn back production costs. One criticism that was commonly voiced was that the work-splits created by the international co-productions impose onerous costs on projects, both in terms of extraspending and lost creativity in the workplace. These costs can usefully be described using Scott’s concept of spatial transaction-costs (SCOTT 1986). Territorial support schemes mean that co-productions are usually motivated not by the access that coproducing partners might have to support or tax shelters in a given area. So, for example, a Danish producer might find a German co-producer in order to gain access to the support offered by the Bavarian government. Although most experienced professionals in the industry emphasized the importance of finding the right partners – one whose capabilities and vision complement and align well— financial considerations are clearly primary. The results are that typically 20 to 25 percent of a production’s budget is lost in coordination costs. These costs include duplication of managerial systems, frequent communications and the time and expense to travel back and forth between studios. They also include the need to re-do work because of communications problems that
-82are inevitable when studios in different countries are involved. New software managing the large number of drawings or digital assets that are required to make a film can help minimize these costs by reducing the number of people required to manage a project and automating coordination between studios. However, problems remain in terms of getting artists up to speed on a project and overcoming differences in creative vision or in practical priorities. In addition to direct, financial costs, the distributed production model may have negative consequences on the ultimate quality of the project. One of the common complaints in the European co-production community is the problem of “Europuddings”, films that have no clear cultural or creative identity (interview with C.J., 6/22/2003). As one producer told me: I think co-production is a pain in the ***... Co-production, when you have animation when you normally have a certain size of project it is almost a must. But the problem is that it sometimes takes the soul away from the project and when you have too many studios working, it becomes a problem. So you have two countries, OK. If you have more, it’s a real pain. (Interview with M.S., 01/11/2011) The views expressed by Petteri Pasenen, the producer of the much-acclaimed 2008 movie, Niko and the Way to the Stars, are fairly typical: We should challenge the old financing system. With the co-production system you can lose up to 30% of your budget just because of the work split. Then the approving system, even though it really made our movie better, it takes ages. You have lots of Skype conferences with ten people or more, trying to say something about something and you speak for hours... I’m not a big fan of this old financing system. I’m really looking for something else. (Cited in Cineuropa, 19/07/2010) Of course, while electronic communications tools like email and Skype can help overcome some of the frictions of distance, most informants (but not all) agreed that traveling and personal visits are still a necessary part of working together. Often, the problems that arise during production are ones of trade-off. For example, a foreign studio is asked to work overtime when a production falls behind. They may have
-83other projects that they are working on as well and may resent the request or decide to put less-qualified talent on the job. Although the tasks could easily be communicated via Skype, a personal visit, and a dinner, still serves to build trust and commitment in a way that remote communications cannot. (Interview with I.D., 9/28/2011) Although the above comments are typical, not all the subjects I interviewed agreed that remote production was a problem. Some saw collaborating with foreign studios as a learning opportunity, while others simply saw no difference between working with people in a common studio and working with them at a distance. In part, of course, it depends on the complexity of the project involved and comes down to personal experience. A second criticism that dovetails with the issues of distributed production is that most European films simply don’t make enough money to pay for themselves. They are highly dependent on subsidies to make their budget and reduce risk for private investors. This criticism, which is spelled out in length in The Movie Game, Martin Dale’s excoriation of what he sees as the corrupt, elitist practices of European state support for film, was generally whispered sotto voce (DALE 1997). The bulk of European films, I was told, are simply not good enough to compete and therefore should not be made. “€6 million is a lot of money to pay for someone for a vanity project,” one informant told me (Interview with B.D., 9/9/2011). The fact that money is available simply for locating productive activities in a given country or territory, he argued, led to decisions that worked against the quality and creativity of projects. Many producers, in their zeal to make up a budget, will find co-production partners who can bring in money, but assign them only low-skill or relatively unimportant tasks on the film. The result not only undermines the purpose of support schemes to build up local industry, but also changes the logic of filmmaking towards one of subsidy-chasing (also see (MORAWITZ, et al. 2007). The obvious alternative to this model is the higher-quality, higher-budget films financed and distributed by American majors and mini-majors. While this model may also produce flops, it is notable that all of the top-earning European films are either
-84American co-productions or European productions that are made on production budgets that offer something of a discount from large American productions while emulating the marketing and production organization of the American model. These films may occasionally subcontract work to service studios, but as a rule, they keep work in house. As TSCHANG and GOLDSTEIN (2004) have carefully explained in their analysis of Pixar, American producers use very different models for producing low-quality, television animation than they do for the higher-quality animation required for top feature-films. While television animation is almost always outsourced to Asia for production, after the script and pre-production have been prepared in the U.S., feature films generally keep all of the work in house. This in-house model of production allows for creative synergies and rapid adjustments in the creative process that help account for the high quality of the product ultimately produced (also see PRICE (2009)). One of the problems with European films might therefore be that they are using a model of production that is not appropriate for the high-quality media they are seeking to produce. While outsourcing may be appropriate for television and for low-budget films, it is not appropriate for films that want to compete as massentertainment. The fact that American film studios are now looking to Europe as a location for feature films speaks well for how much the industry has developed over the last 20 years. Many European artists who had previously emigrated to the U.S. in order to work at top studios are now returning to Europe (FORDE 2011). However, if a switch to a model of filmmaking with fewer, high-budget films were to make significant inroads and public subsidies were to be cut, it would likely lead to the extinction of many of the smaller studios in Europe and the concentration of animation talent in fewer locations. The currently distributed nature of animated filmmaking in Europe, partially an artifact of institutional conditions and especially the way that financial resources have been politically mobilized, may then no longer be viable.
-853.9 DISCUSSION AND CONCLUSIONS This paper has discussed the role of Cartoon in promoting the development and growth of Europe’s animation industry over the last 20 years. Cartoon has via a longterm project promoted interaction and collaboration between various players in the animation industry, generating relational proximity, shared understandings, and personal trust among professionals, broadcasters and financiers.” Since its creation in 1988, Cartoon has played a central role in stimulating multi-locational production practices. In the late-1980s, the liberalization of television in several European countries created a demand for animated content. However, little trans-European trade existed and station programmers were more likely to buy content from the U.S. or Japan than from another European country. Cartoon was formed to unify the European animation industry by creating a common set of institutions and understandings to create and facilitate cooperation and within both formal and informal industry networks. The institutional practices created by Cartoon have helped producers break out of their national markets and expand internationally by and brought institutional coherence to the notion of “European Animation.” This case study makes a couple of distinct contributions to our understanding of how learning communities can be fostered in creative industries. First, Cartoon offers a concrete example of how learning communities may be generated both across and within regions, and provides concrete evidence on the types of learning-by-interaction that are necessary to build a successful industry. Through this case, I have tried to illustrate an expanded from a narrow emphasis on innovation our understanding of what ‘learning’ looks like in a system where the collaboration of different actors with different roles must be achieved. I have tried to show that while actors may occasionally gain specific and actionable knowledge through casual interaction, their main benefit of such interaction is the chance to explore their environment and to develop shared understandings, conventions and awareness of the thoughts, opinions and goals of others in their field. In short, this kind of learning allows people to expand out from one specific geographic, organizational or cultural context and
-86become competent at collaborating with others who may bring different contextual assumptions to their work. These findings contribute to the literature on transnational knowledge communities; illustrating empirically how buzz and knowledge-circulation may be organized in a spatially extended community (COE and BRUNELL 2003). In AMIN and COHENDET’s (2004) seminal account, these communities seem to cut across local geographies, thereby, questioning the idea that highly “tacit” forms of knowledge will generally circulate only locally. According to their argument, such communities consist of people who share a high degree of ‘relational proximity’ as well as a common purpose or professional orientation, enabling mutual communication regardless of geographical distance. The counter-argument suggests that knowledge in such communities is – perhaps inevitably – fragmented by the different ways that practices are articulated in different local contexts, making the sharing of knowledge and collaboration in production across distances more difficult (GERTLER 1995; LAM 1997). At times, this literature has presented a pessimistic picture of the possibilities of knowledge exchanges outside of closely-knit geographic communities, a picture that feels quite at odds with how many people experience contemporary globalization. The evidence of Cartoon is better in line with the more nuanced view evident in GERTLER’s more recent writings (2008), which depict the contextual nature of knowledge as potentially problematic but surmountable provided the proper context and motivation (FAULCONBRIDGE 2007, 2010). In this framework, Cartoon well illustrates the kinds of activities that foster learning across contexts. Cartoon’s institutional entrepreneurs attempted several ways to promote shared conventions, such as the “Bible” and the temporary gatherings happening in Cartoon Forum and Masters. However, these modes of intervention have been unevenly successful: whereas a more radical shift prompted by the publication of the “Bible” has been largely forgotten, the more organic and gradual mode of change stimulated by the temporary gatherings seems to have been more effective.
-87I am far from suggesting a strong normative implication derived from this fact and related with the convenience of celebrating this kind of meetings. Indeed, recent institutional entrepreneurship research has stressed the controversies that can be created around these sorts of events. For example, in his study of the emergence of a new medical technology, GARUD (2008) highlighted the bitter disagreements about standards happening in several multi-stakeholder conferences. Thus, instead of pointing to narrow normative implications, I believe the lessons to be extracted are more related to recent research on institutional work (LAWRENCE, et al. 2009, 2011; LAWRENCE and SUDDABY 2006), which –in its effort to overcome previous “heroic” conceptions of entrepreneurshiphighlights the existence of unintended consequences and failures in the various means used by institutional entrepreneurs. The story of the struggles of Cartoon in finding an effective way to promote shared conventions in the European Animation Industry powerfully resonates with the “muddles, misunderstandings, false starts and loose ends” (LAWRENCE, et al. 2009, 2011) which characterizes a more realistic and non-linear perspective of the everyday effortful practices of institutional entrepreneurs. These findings are relevant for industrial policy beyond the animation industry, especially to policy-making efforts to promote ‘learning’. Regional policies have often focused on thickening relationships within geographic ‘clusters’ of competing and related industries, on the assumption that this will lead to more robust endogenous innovation dynamics. The arguments invoked to support this assumption were often based on the notion that ‘tacit knowledge’ can only be exchanged in ‘localized learning’ processes in especially privileged ‘learning regions’, the emulation of which should be supported by regional policy. Later versions suggested that it was also important to build ‘pipelines’ with outside sources of knowledge (BATHELT, MALMBERG and MASKELL 2004), but the precise meaning of this metaphor has remained somewhat unclear. The case of Cartoon and the analysis presented here suggests that, at least in a European context, the promotion of transnational and trans-regional networks of practice, furthering competitiveness through scale, novel combinations, variety, and
-94RALLET, A. and TORRE A. (1999). "Is geographical proximity necessary in the innovation networks in the era of global economy?" GeoJournal 49(4): 373-380. SCHULDT, N. A. and H. BATHELT (2011). "International Trade Fair and Global Buzz. Part II: Practices of Global Buzz." European Planning Studies 19(1): 1-22. SCOTT, A. J. (1986). "Industrial Organization and Location: Division of Labor, the Firm, and Spatial Process." Economic Geography 62(3): 215-231. SCOTT, A. J. (1997). "The cultural economy of cities." International Journal of Urban and Regional Research 21(2): 323-339. SCOTT, A. J. (2002). "A New Map of Hollywood: The Production and Distribution of American Motion Pictures." Regional Studies 36(9): 957-975. SCOTT, A. J. (2004). "Hollywood and the world: the geography of motion-picture distribution and marketing." Review of International Political Economy 11(1): 33-61. SITO, T. (2006). Drawing the Line, The Untold Story of Animation Unions from Bosko to Bart Simpson. Lexington, Kentucky, The Univesity Press of Kentucky. SKOV, L. (2005). The Role of Trade Fairs in the Global Fashion Business. SLAGER, R., J.P. GOND, and J. MOON (2012). "Standardization as institutional work: The regulatory power of a responsible investment standard." Organization Studies 33(5/6): 763-790. STEINMULLER, W. E. (2000). "Will New Information and Communication Technologies Improve the 'Codification' of Knowledge?" Industrial and Corporate Change 9(2): 361-376. STORPER, M. and S. CHRISTOPHERSON (1987). "Flexible specialisation and regional industrial agglomerations." Annals of the Association of American Geographers 77(1): 104-117. STORPER, M. and R. SALAIS (1997). Worlds of Production:The Action Frameworks of the Economy. Cambridge, Mass, Harvard University Press. STORPER, M. and A.J. VENABLES (2002). Buzz: The Economic Force of the City. DRUID Summer Conference 2003. Copenhagen/Elsinore. TAKTEYEV, Y. (2009). "Networks of Practice as Heterogeneous Actor-Networks." Information, Communications & Society 12(4): 566-583. TORRE, A. and A. RALLET (2005). "Proximity and Localization." Regional Studies 39(1): 47-60.
-95TSCHANG, T. and A. GOLDSTEIN (2004). Production and Political Economy in the Animation Industry: Why Insourcing and Outsourcing Occur. DRUID, Elsinore, Denmark.
-97-
CHAPTER 4 _______________________________________ THE DISJUNCTIVE GEOGRAPHIES OF KNOWLEDGE AS SKILL AND KNOWLEDGE AS CONTEXT: WHY KNOWLEDGE IS BOTH LOCALLY STICKY AND GLOBALLY MOBILE
-99Abstract. Over the last decade, claims within geography about the necessity of localization and proximity for the production and circulation of tacit knowledge have been challenged by a ‘networks’ perspective that has sought to correct a perceived overemphasis on local knowledge and localized learning. The resulting conversation has opened the field up, and it is now accepted that both local and non-local relationships may be important for the creation and dissemination of knowledge. In this paper, I reconsider the issue, introducing a distinction between knowledge creation, which often, although not always, requires sustained face-to-face interaction between people with different expertise, and knowledge circulation, which requires that people with similar expertise and experience translate and adapt knowledge and skills in new social, economic and geographic contexts. The paper investigates factors creating proximity or dispersion in each of these distinct processes and suggests that they may describe divergent geographies, with the one requiring localization while the other does not. In a final section, four potential combinations of localization and dispersion in knowledge creation and knowledge circulation are explored. 4.1 INTRODUCTION Debates about economic geography and the geographies of knowledge and innovation over the last several decades have centered on the importance of physical proximity in the production and circulation of knowledge (GERTLER 2008). In this regard, the innumerable success stories associated with clusters and other ‘territorial innovation models’ seemed to offer irrefutable evidence of the importance of co-location and proximity to processes of knowledge creation and circulation (MOULAERT and SEKIA 2003, ASHEIM and GERTLER 2005). The regional scale, or cluster, was highlighted because of the rich opportunities afforded for face-to-face interaction and the intensive circulation of information, or cluster buzz, keeping cluster members abreast of developments in their industry (STORPER and VENABLES 2004, MALMBERG and MASKELL 2006). Furthermore, spatial proximity was thought to stimulate the development of shared conventions, a common language and compatible interpretative schemes – in short, relational proximity— facilitating the circulation of ‘tacit’ knowledge and promoting learning (STORPER 1997, LAWSON and LORENZ 1999, MASKELL and MALMBERG 2007). The connection between regionalization
-100and learning became embodied in concepts such as ‘the learning region’ and ‘localized learning’ and became a dominant paradigm within economic geography (GRABHER 2006). However, increasingly this dominant line regarding the causal relationships between localization and learning has been challenged (AMIN and COHENDET 2004). Critics of the prevailing consensus have argued that geographical proximity is neither necessary nor a sufficient guarantee for tacit knowledge exchange (OINAS 1998, RALLET and TORRE 1999, TORRE and RALLET 2005). First, it was pointed out that in place of permanent and continual proximity offered by co-location, temporary physical co-presence might be sufficient for collaboration (TORRE 2008). For example, the role of trade fairs, conferences, and other forms of ‘temporary clustering’ in supporting knowledge circulation seemed to mimic those of the permanent cluster (NORCLIFFE and RENDACE 2003, MASKELL, BATHELT, et al. 2006, BATHELT and SCHULDT 2010, BATHELT and TURI 2013). Second, scholars pointed to important structures of informal collaboration, fueled by travel and Internet connectivity, that reached well beyond the cluster (BATHELT and TURI 2013). These networks were conceived of using a language of ‘knowledge communities’ – reference often being made to ‘communities of practice’ (BROWN and DUGUID 1991, WENGER 1998, AMIN and ROBERTS 2008) or ‘epistemic communities’ (BRESCHI and LISSONI 2001, AMIN and COHENDET 2004, HÅKANSON 2005). Knowledge, it was argued, is likely to flow easily within these communities regardless of their geography, but will move only with difficulty between them. So, whereas MARSHALL (1920) had famously noted that information in localized industrial districts moves so freely that ‘the secrets of industry are seemingly in the air’ and his neo-Marshallian followers had celebrated ‘local buzz’ as a source of information that kept entrepreneurs and firms within clusters a step ahead of their competitors elsewhere, BRESCHI and LISSONI (2001) argued that knowledge spillovers within industrial districts are not "in the air" but “in the network.” Against the ontology of place that was at the center of the clusters story,
-101these scholars asserted the primacy of the ontology of the network, without focusing a priori on any one spatial scale (BUNNELL and COE 2001). During the decade since this ‘knowledge communities’ view was first articulated, the geography of knowledge has proceeded in an eclectic way. The field has generally recognized the importance of trans-local or global networks in circulating knowledge, and research has expanded on this topic. A somewhat uneasy synthesis between seemingly opposed paradigms has gradually become accepted, notable in concepts such as ‘buzz and pipelines’ (BATHELT, MALMBERG, et al. 2004). It is generally acknowledged that even if economic transactions are not necessarily localized, localization is still important (BATHELT and TURI 2013) and that each perspective has a valid contribution to make. Rather than challenging each other’s accounts, the respective proponents focus on different cases, contexts and empirical evidence. While this approach may help guard against an overly narrow fixation on only one dimension of the geographical puzzle – the benefits of localization – it leaves unanswered the basic question as to the importance of proximity in the creation, circulation, and sharing of knowledge and how this issue relates to the geographic clustering of competing and complementary industries. If the production and circulation of knowledge can occur non-locally as well as locally, then how do we explain agglomeration and regional learning? In this article, I try to shed some light on these questions through a re-reading of the debate on tacit knowledge, clustering and knowledge networks. I suggest that the debate around localized learning and learning in non-localized networks has been muddied by a failure to specify the kind of knowledge that is being acquired and the different kinds of learning taking place. I contend that the discussion about the geography of learning has failed to clearly separate two dimensions of practical knowledge and understand how they are acquired. The first kind concerns knowledge of a particular domain – such as that of a discipline or a practice – that can be relatively invariant and consistent across contexts. ‘Domain knowledge’ includes not only theoretical knowledge but also practical skills and know-how developed through experience in the exercise of a profession or practice. The second kind relates to
-102- ‘contextual knowledge’ of the social and economic environment in which these skills are utilized. This kind of contextual knowledge is often acquired through processes of ‘acculturation’ into the workforce. The argument developed in this paper is based on the observation that much knowledge creation takes place through combinations of knowledge from both overlapping and different domains; it often requires intensive negotiations around the potential and feasible combination of knowledge elements, the framing and definition of problems, and the alignment of the criteria that the participants bring to the process (GRANT 1996, MUTHUSAMY and WHITE 2005, HÅKANSON 2010). Knowledge creation – particularly when it involves complex and tacit forms of knowledge – often requires intensive face-to-face interaction. The geography of innovation through complex combinations of knowledge is therefore shaped by the spatial transaction costs of getting knowledgeable people to meet in the same place over a sustained period of time. Knowledge diffusion, on the other hand, typically involves people with overlapping domain knowledge and therefore may occur as a byproduct of contacts through daily work, but can also be organized between individuals who only see each other occasionally, or know each other only through their virtual presence. Marshall’s seminal writings on industrial districts offered a version of this distinction, attributing the success of industrial districts both to the advantages of pecuniary externalities such as localized labor markets and the availability of specialized inputs and to non-pecuniary, informational externalities, the ‘industrial atmosphere’ that clung to such places and gave people there a sense of what was happening in local industry well before the rest of the world (MARSHALL 1920). Theories of localized learning and learning regions have tended to focus on the latter point, non-pecuniary externalities – atmosphere or buzz – recommending that policy-makers facilitate their creation and exploitation through, for example, local institution building and encouraging more exchange within the cluster (AMIN and THRIFT 1995, BATHELT, MALMBERG, et al. 2004, ASHEIM and GERTLER 2005). In contrast, I place a reduced emphasis on the institutional aspects of clustering. I suggest that knowledge communities may be increasingly cosmopolitan so that even highly tacit
-103developments in their knowledge base may circulate widely. However, the creation of new knowledge often requires flexible and timely collaboration, which facilitates different domain knowledge to be recombined. The clustering of specialized talent greatly facilitates this process. In the following section, I develop the theoretical distinction between domain and contextual knowledge.Section Three discusses the issue of knowledge circulation and diffusion and how the geography of knowledge communities is determined by issues of context and the ability of such communities to adapt knowledge from context to context. Section Four returns to traditional arguments regarding agglomeration, discussing how the need for face-to-face interaction when combining tacit and context dependent knowledge may lead to agglomeration. In Section Five, I trace the consequences of the distinction between the geography of knowledge circulation in a knowledge community and the geography of knowledge creation in teams and highlight four geographies of knowledge that might emerge from their combination. While most of the debate on the localization of knowledge has focused on situations where both processes are localized, I demonstrate that other feasible combinations need to be explored. 4.2 TACIT KNOWLEDGE AS SKILL AND TACTIT KNOWLEDGE AS CONTEXT The standard view on the localization of knowledge circa the year 2000 was based on fairly simple presuppositions about the nature of knowledge. The first of these was the idea that knowledge can usefully be divided into that which is codified in or codifiable into some shared language, code or system of notation while ‘tacit’ knowledge is that which is not or cannot be codified in this manner 13 .Tacit 13 The distinction between tacit and codified knowledge was introduced into current discourse most prominently in the writings of NELSON and WINTER (1982) through their extensive referencing of the philosopher of science, Michael POLANYI (1966). In subsequent usage, ‘codified knowledge’ has conventionally come to refer to all knowledge that can adequately be recorded and communicated through a recipe, blueprint or instruction manual, permitting its near costless dissemination, both through voluntary transfer and involuntary imitation. By contrast, tacit knowledge is residually defined as all knowledge that cannot be communicated in such a direct way.
-110economy. This is what MARSHALL (1920), meant by ‘industrial atmosphere’. Contemporary scholars have used the concept of ‘Noise’ (GRABHER 2002) and ‘Buzz’ to refer to the same phenomenon. BATHELT, MALMBERG and MASKELL (2004, p. 38) tell us that: Buzz refers to the information and communication ecology created by faceto-face contacts, co-presence and co-location of people and firms within the same industry and place or region. This buzz consists of specific information and continuous updates of this information; intended and unanticipated learning processes organized in accidental meetings, the application of the same interpretive schemes and mutual understanding of new knowledge and technologies, as well as shared cultural traditions and habits within a particular technology field, which stimulate the establishment of conventions and other institutional arrangements. Actors continuously contribute to and benefit from the diffusion of information, gossip and news by just 'being there.' STORPER and VENABLES (2004, p. 9) describe Buzz as "a highly efficient technology, a means of overcoming coordination and incentive problems in uncertain environments." Taken individually, most such information might be of little value. However, collectively, this concatenation of information, rumor, opinion – this ongoing conversation about the ‘who’, ‘what’, ‘why’ and ‘how’ of local industry – may provide important contextual clues that help an individual understand and navigate a creative field. As BATHELT, MALMBERG, et al. (2004) note, the constant exchange of information also plays a role in shaping a set of conventions and other institutional arrangements. The ephemeral, informational aspects of buzz sit upon, but also feed into, more long-standing arrangements and understandings that bind a community or network together. Within the localized learning story, the constant exposure to information combined with the shared conventional knowledge to make sense of this information, create a powerful mechanism for learning in the cluster. POUDER and JOHN (1996) suggest that the density of social interaction in ‘hot spots’ leads entrepreneurs to converge around ‘shared cognitive frameworks.’ Such shared frameworks make the
-111entrepreneur’s actions intelligible to collaborators and hence provide legitimacy in their interactions with other cluster agents such as venture capitalists. They also allow practices to be benchmarked and imitated by entrepreneurs who have little problem interpreting each other’s actions. Similarly, MALMBERG and MASKELL (2002) lay out a theory of learning in which firms within clusters learn by observing and imitating or improving on innovations put forward by rivals. Proximity, it is argued, is important for the collective learning processes outlined above for two reasons: First, because frequent in-person interactions facilitate the flow of information within the cluster; second, because people who live and work in the same community are more likely to understand and interpret this information in a mutually intelligible manner. The information is valuable because the person or people interpreting it can fill in the context and act appropriately. 4.3.2 The Global Networks View Against this view, scholars such OINAS (1998), COE and BUNNELL (2003), and AMIN and COHENDET (2004) have argued that the idea of co-location as a prerequisite for learning has been taken too far, becoming something of a spatial fetishism. While the ‘localized learning school’ sought to generalize from the case of successful clusters, these scholars argued that in the cosmopolitan world, mere physical proximity actually says little about the relationships between people. To give an illustration of this point offered by RALLET and TORRE (1999), two neighbors who share a common wall between their apartments may never talk and may indeed have little in common, but may have friends and colleagues scattered all over the world. These authors suggest a distinction between geographical proximity – essentially co-localization in space —and relational or organizational proximity, which measures the similarity or difference between two people on some socialinteractional scale. While geographical proximity denotes the physical distances between actors, relational and organizational proximity refers to the closeness of actors in relational or organizational terms (BOSCHMA 2005).
-112While these scholars agreed with the general institutionalist approach to knowledge espoused by the learning regions and localized learning literature, they suggested that knowledge is not generally available simply by virtue of location, but travels among networks of people who have the same general training and knowledge orientation. In his critique of the common practice of invoking ‘tacit knowledge’ to explain localization, HÅKANSON (2005, p. 442) writes: All knowledge is context-dependent. It requires for its meaningful interpretation and effective application mastery of the codes (language, vocabulary, symbols, etc.) in which it is expressed, at least intuitive understanding of the theories (implicit and explicit cognitive frames, beliefs, mental maps, etc.) to which it refers, as well as familiarity in the use of tools of the associated practice. Neither tacit nor explicit knowledge can therefore be defined without reference to the context of the social community where it resides. So, contra ideas of ‘localized learning,’ which rest on the ways local contexts can become the springboard for engaging in rounds of learning, the ‘learning communities’ literature argues that local contexts are less important than the context and common dispositions that are shared among people that engage in common practices. These criticisms on the specificity and irreproducibility of local contexts found support in literature from the sociology of knowledge on ‘communities of practice’ (BROWN and DUGUID 1991, WENGER 1998), ‘epistemic communities’ (COWAN, DAVID, et al. 2000, AMIN and COHENDET 2004, HÅKANSON 2005), and ‘occupational communities’ (BECHKY 2003a, 2003b). While the three terms have different histories and different uses, they all suggest that knowledge is organized and shared primarily among people who share a common knowledge practice, cognitive frameworks, and identity (WENGER 1998); in short, people who have invested in a particular branch of knowledge. People who are geographically distant but who share a common professional disposition or are part of the same organization will have the needed context to
-113engage, communicate and learn from each other. Such learning communities, which may or may not be defined by geography, share the codes, theories and tacit knowledge that go along with their trade. However, the learning communities view can also be criticized. In particular it often appears that scholars in this tradition assume ‘relational proximity’ – that there is sufficient similarity in the organizational and institutional environments within which people are working – and fail to give proper recognition to local contextual factors that may frustrate efforts to collaborate or draw useful lessons. LAM (1997), for example, has written about the immense differences in the way engineering is practiced and institutionalized in England and Japan and the problems this caused for efforts to share knowledge and collaborate between the English and Japanese branches of a multinational firm. Research like this suggests that behind the appearance of a professional community, in reality knowledge practices may be deeply fractured by institutional and cultural differences. It is useful to remember that the concept ‘communities of practice’ was developed through observations of highly localized work groups constituted through regular face-to-face interaction (LAVE and WENGER 1991). The term has since been generalized to allow for groups that share a similar outlook and share knowledge and insights, but may never actually meet in person. However, the degree to which geographic context matters for the ‘sharing’ or ‘co-creation’ of knowledge within these various knowledge networks is an issue that geographers are just beginning to come to grips with. Two factors are likely to determine the answer to this question and thus the degree to which knowledge networks form around specific local contexts or are able to spread trans-locally. The first factor is the degree to which a given knowledge-practice is context-sensitive. The second is the incentives and means available to actors to overcome differences in contextual knowledge and learn to translate their knowledge and insights in new places.
-1144.3.3 Stretching Knowledge Relationships Through Disembedding While recognizing the context-dependent nature of knowledge-in-practice has provided profound insights into the geography of knowledge, it is important to also recognize that not all knowledge practices are equally context-sensitive. Writing about economic development the 1950’s, HIRSCHMAN (1988) noted that airline accidents were exceedingly rare, even in Third World countries where one was likely to find peasants driving donkey-carts right outside the boundaries of the airport runway. While the airports were located in a Third World environment, the practices around air-travel were disembedded from the norms and practical work culture of the societies around them. They were, in essence, islands of modernity in the middle of pre-modern cultures. Many knowledge practices may take this form, existing in semi-isolation from local practices around them. This is particularly the case when the practice requires a highdegree of internal consistency from place-to-place. Epistemic communities may then circulate knowledge fairly easily without regard to distance because members can be fairly confident that the knowhow they are conveying will land in an epistemic environment that is sufficiently similar to the one it is being sent from. TAKTEYEV (2009), for example, discusses how similar socio-technical environments enable Java programmers in Brazil to learn their trade through practical problem solving and the aid of written manuals – codified knowledge – imported from the U.S. Drawing on actor-network theory, he suggests that knowledge circulation is facilitated by the fact that the knowledge context within which Java is used is common both to the Americans writing the manuals and to the Brazilians using them. This insight into context also explains ASHEIM, COENEN, et al.'s (2007) finding that ‘analytical knowledge communities’ such as the hard-sciences communities may rely little on buzz. The authors make a distinction between ‘analytic’, ‘synthetic’, and ‘symbolic’ knowledge, which corresponds roughly with scientific, engineering, and creative industries. Their argument is that industries based on analytical knowledge – scientific labs, for example – depend less on buzz than industries based on synthetic
-115or symbolic knowledge. One may question the underlying distinction; some ‘symbolic’ industries, such as classical music or jazz, for example, are able to isolate their practices from their local cultural surroundings, thereby permitting largely consistent expression around the world, However, their basic insight can be explained by the fact that industries based on analytic knowledge are, by design, islanded or disembedded in order to gain internal consistency. 4.3.4 Bridging Local Contexts In spite of the obstacles, there is evidence to suggest that – given the right circumstances – even heavily context-dependent kinds of knowledge may also be geographically mobile, although the process here is altogether more difficult and uncertain. Although highly context-dependent knowledge may be more spatially (and organizationally) sticky, there are often strong economic reasons for actors to participate also in non-local circles of knowledge-exchange. The advantages of doing so in order to gain access to a wider pool of knowledge exchanges can provide a powerful incentive to engage in codification and articulation of knowledge (COWAN, DAVID, et al. 2000). DUGUID (2008), who argues strongly that all knowledge is context-dependent, nonetheless suggests that membership in a community of practice in which knowledge is generated and shared within a very specific, local context, may become a springboard for membership in a spatially dispersed ‘network of practice’. Duguid’s account of how communities of practice connect to spatially extended ‘networks of practice’ rests squarely on an insistence that tacit knowledge matters and that ‘knowing how’ always has a tacit dimension based on immersion in practical activity. However, participation in local communities of practice provides the background against which relationships of knowledge circulation and knowledge creation can be stretched to encompass a wider network of practice. It is not that tacit knowledge doesn’t matter in these networks; rather it is the case that practitioners from place to place share enough of their practice – the tools, cognitive models, and common purpose – to be able to translate what they know and communicate effectively across
-116geographic context. So, for example, an economist trained in California may coauthor a paper with colleague from Europe or Asia. A Shakespearean actor becomes involved in theatre locally, but then enters the relational ‘world’ of Elizabethan theatre and becomes part of a network tying her to the national or international ‘theater world’. Institutionalization is at least partially endogenous with respect to the process of relationship building. When two partners form a relationship, routines and conventions emerge as they build up a history of interacting (EGIDI and NARDUZZO 1997, LAZARIC and LORENZ 1998). At first, communication may be difficult and the partners may prefer to keep interactions routine, but with time, they often develop shared understandings of their interaction and, possibly, a specialized ‘codebook’ to facilitate more complex and open ended types of exchanges (MAGUIRE and HARDY 2006, ETZION and FERRARO 2010, SLAGER, GOND, et al. 2012). Eventually, such institutionalized behavior may be replicated throughout a wider network of actors, as new actors find it increasingly desirable to adopt similar institutionalized understandings in order to gain access to the resources that the network commands (COWAN and FORAY 1997). Geographers have provided some important caveats to this vision of an easy progression from localized to cosmopolitan knowledge relationships. The process of spatial dis-embedding does not always proceed smoothly and in some cases actors may remain fettered to localized knowledge worlds. GERTLER (1995, 2003) and LAM (1997) have documented the problems that multinational companies have in transferring best practices or organizing social learning processes between subsidiaries in different countries (SZULANSKI 1996). The smooth functioning of communities across space cannot simply be assumed by stating that the actors involved enjoy a high degree of cognitive, institutional or social proximity. Embeddedness in local institutions may obstruct the formation of common practices, thereby frustrating the transmission of knowledge. However, the success of a network is often more dependent on its ability to adapt embedded knowledge from one context to another (HELPER, et al. 2000).
-117The topic of how knowledge is translated from place to place, the contextual work that is required and the social and institutional arrangements that facilitate this work all need to be researched more deeply. Current examples of this research come from studies by Faulconbridge of knowledge circulation and creation among architects, advertising executives, and legal professionals in global (FAULCONBRIDGE 2006, 2007, 2010). In these studies, Faulconbridge emphasizes the ways that knowledge circulation does not happen through the transfer of best practices, but through ‘conversation’ (KROGH and ROSS 1995) with foreign colleagues and foreign artifacts such as texts, photographs or architectural plans, that stimulate the production of new knowledge. This circulation happens as the people involved grapple with issues of context and which aspects of their knowledge are useful or applicable as they move to other contexts. Current research on ‘Global Buzz’ supported by global trade-fairs, conferences and other forms of periodic gatherings also offers insights into this process (BATHELT and SCHULDT 2010, SCHULDT and BATHELT 2011, BATHELT and GIBSON 2013). The argument is that these events provide the information ecology through which global knowledge communities may become more connected. As Bathelt and Schultdt have argued, these events create multiple learning opportunities for members of a knowledge community, helping them to engage in the type of learning by observation and imitation believed to be the hallmark of localized learning. The gathering of large segments of a community at these events may also make them important sites for the emergence of institutionalized norms and conventions throughout the community. My own work suggests that through regular attendance at such events, member may learn other ‘idioms’ and become better at translating their knowledge and communicating with colleagues who work in distinct knowledge local contexts. As participants to these events engage in periodic exchange, they are able to overcome distanced contexts and build a transnational ‘learning community’ based around a common knowledge base (COLE 2008). This kind of temporary face-to-face interaction seems adequate for exploring relationships and building a network,
-118particularly when the contacts are renewed periodically through repeated meetings in different contexts (POWER and JANSSON 2008). Through selective dis-embedding, institutionalization of non-local networks, and a learned ability to translate between different contexts, knowledge communities are likely to become much more effective at creating a shared body of knowledge across different geographic and organizational contexts. The arguments outlined above suggest a pragmatic approach towards the spatial reach of knowledge communities. Some kinds of knowledge practice are sufficiently insulated from the local environment that moving knowledge practices from place to place is trivial. However, it appears that communities may also effectively transfer and translate strongly context-dependent knowledge. How this occurs has lately attracted much research interest; travel, international conferences and trade fairs, participation in online communities all seem to play a part in the increasingly global circulation of knowledge in the contemporary era. 4.4 THE GEOGRAPHY OF SKILLS AND COMPETENCE As argued above, knowledge is not always ‘locally sticky’ and there is good reason to think that the literature on localized learning exaggerates the role of local context as a barrier to knowledge diffusion. However, that does not mean that geographical clustering of competing and complementary industries is unimportant or unrelated to the accumulation of know-how. First, while the local context in which know-how is applied may not be a significant barrier to knowledge diffusion or may be overcome through learning and adaption, that is different from saying that context does not matter at all. Second, when workers with the same domain knowledge are able to figure out issues of context, the actual use of this knowledge in the production process, particularly when this production is innovative and not routine, will often require proximity. Traditionally, co-location was driven by indivisibilities in capital goods – buildings, machinery, and other tools – but as cases such as Wall Street (BEUNZA and STARK
-1192004) or Silicon Valley make evident, agglomeration may be even more important in industries whose outputs are largely immaterial and where work is largely screenbased – advanced services, intellectual property, and management, for example. In general, geographers agree that agglomeration can be explained with reference to the importance of skilled labor and the usefulness of co-location in solving complex coordination issues when applying tacit knowledge to create products and services. 4.4.1 Face-to-Face Communication in the Articulation of Complex Tacit Knowledge Knowledge creation in the contemporary economy is often a collaborative process involving interactions between people. It may often involve the combination of different kinds of knowledge, rooted in different epistemological frameworks 16 .These interactions can be routine and unproblematic, following well-defined and accepted rules and conventions or they may be quite complex, requiring the intensive negotiation around the potential and feasible combination of knowledge elements, the framing and definition of problem, and the alignment of the criteria that the participants bring to the process. Such knowledge transactions – particularly when they involve complex and tacit forms of knowledge – often require intensive and sustained face-to-face interaction (BECHKY 2003a, 2003b). While physical co-presence is not always necessary for processes of co-development and articulation of complex tacit knowledge, the power of face-to-face interaction Although ‘recombination’ is considered the source of novelty in canonical accounts of the relevance of knowledge in contemporary capitalism (NELSON and WINTER 1982, NONAKA and TAKEUCHI 1995), the question about the ‘nature’ of these recombined knowledge elements is still not well solved. Whereas some authors limit recombination to elements from different knowledge domains, others highlight the opportunities of recombination in ‘restless’ knowledge domains (METCALFE 2002), in which new elements are constantly added to one domain through time, thus creating ‘space’ for more recombination between ‘temporary distant’ elements (O'DRISCOLL, RIZZO et al. 1996). I keep agnostic on this point: although the creative potential of bridging different knowledge domains is evident, accounts of creativity at the workplace highlight the ability of more ‘closed’ communities to innovate ‘restlessly’ (LATOUR and WOOLGAR 1979, BROWN and DUGUID 1991, METCALFE 2002). See ARTHUR (2007) for a nice illustrations of novel combinations of ‘temporary distant’ knowledge in the domain of electric engineering.
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
-127Across the top of the matrix, two situations are described. On the left, knowledge creation is localized as complex interdependencies in the workplace are facilitated by spatial proximity. This localization could occur at the level of the individual firm or workplace; for example, the highly innovative animation firm, Pixar, which does not form part of any animation cluster. Typically, of course, geographers have been concerned about proximity at the scale of the region. Clustering allows for the rapid assembly of project teams when it is difficult to engage in long-term employment relations. On the right, individuals, firms or other organizations in similar industries or workplaces are able to perform their work in various locations. This is often the case in knowledge-intensive industries such as management consulting, architecture, medicine, or teaching, where the need to locate close to consumers or patrons dictates a more dispersed locational pattern. Moving vertically down the matrix, the top row represents situations where knowledge circulation is strongly localized, either because information is held within a local network or because this knowledge is embedded in local practices and conventions in a way that makes it hard for outsiders to interpret and use. On the bottom row, knowledge circulates effectively between localities and distant members of the relevant knowledge community are able to effectively bridge local contexts. Real geographies of learning, of course, always consist places of sustained interaction (workplaces) and places of occasional contact (the local bar, the conference center, the airport, the corporate meeting room) linked together at multiple spatial and temporal scales. However, in organizing places of knowledge creation and places of knowledge circulation in a matrix, as we have above, four ideal-typical situations are revealed. Upper Left: When spatial transaction costs in production are high and the conventions that define the knowledge community are local, then we are dealing with a learning region or the classic Marshallian industrial district described by localized learning theorists and exemplified by leading as Silicon Valley, Wall Street, or the Hollywood film industry. These places host strongly localized project networks,
-128making it easier to create a team using top industry talent than it might be elsewhere. In addition, the rapid evolution of industry networks in response to competitive pressures means that entrepreneurs are often crossing un-trodden ground, creating the conventions and shared understandings that allow teams to co-ordinate at the same time that they create new industry knowledge. The ambiguity that faces outsiders who are not participants in the process may slow the rate at which knowledge diffuses outwards, giving a competitive edge to those who are quickly learn from and incorporate the innovations of their neighbors. Lower Left: Moving downwards, we come to the situation in which a localized cluster of firms facilitates interactions among teams of collaborators, but people who make up these teams are also embedded in knowledge communities that go beyond the cluster. This combination of localized networks for knowledge production and non-localized networks for knowledge circulation seems characteristic of many industries and has been theorized in terms of ‘buzz and pipelines’ (BATHELT, MALMBERG et al. 2004). Of course, such a cluster may exist with no buzz at all if local actors find little reason to interact beyond the founding of new firms (BATHELT 2005, BOSCHMA and TER WAL 2007). The combination of localized and non-localized relationships, of course, rests on the different frequency and intensity of in-person interaction that are required to sustain these two kinds of activities. Whereas knowledge creation is mostly occurring in specific, specialized domains among colleagues who need to interact frequently, sparser, less-frequent interactions are sufficient for circulating and updating knowledge among distant colleagues who have strongly overlapping knowledge-bases and epistemic attitudes (STORPER and SCOTT 1995). The fact that individuals belong to a strong ‘epistemic community’ means that collaboration on specific projects even among distant parties is at least possible, and such communities may give birth to knots of activity (ENGESTRÖM 2008) as ‘latent relationships’ that evolve into closer, working relationships in different places, becoming a key path of globalization. However, much work is still done within the cluster.
-129Lower Right: The lower right quadrant represents a situation where knowledge creation may be carried out in multiple localities. This may be the case because the interactions involved are more routine and can be embedded in longer-distant collaboration. Or it may be that firms and organizations are able organize work internally, recruiting specialized labor through long-distance networks. As the knowledge community is dispersed in different locations, institutions have been developed to effectively circulate knowledge either through extensive codification or through regularized travel and face-to-face contact. This is the ‘archetypical’ situation forwarded by the global networks literature (COE and BUNNELL 2003) and ‘proximities school’ (RALLET and TORRE 1999). This situation is typical of all kinds of knowledge-intensive industries where the need to locate near to clients means that clustering is not viable. The requirement to educate means that academics in any given discipline find themselves dispersed at different universities; doctors need to work at hospitals located in different cities; management consultants consult at corporate headquarters all around the world. In this regard, the cases for which this ideal-type is suitable have little to do with those for which localized learning is suitable since the need to locate close to customers or clients precludes clustering. However, even in industries where customer-relationships are arms length, such as most manufacturing, firms or small clusters may be able to manage the transactions involved in daily knowledge production internally. This is the case of Pixar, the highly innovative and creative animation studio. While Pixar is strongly embedded in international technological communities and sources its talent internationally, it is not locally embedded in any meaningful way (PRICE 2009). This also suggests a large grey area between the lower left and lower right quadrants in which networks of distant collaborators may try to engage in knowledge creation beyond their region despite the complexities of doing so. As communications technologies within the workplace steadily improve the possibilities of creating
-130knowledge through combinations of proximate and distant collaborations become greater, making this area highly dynamic. Upper Right: Finally, it may be that a given industry or activity can be performed relatively autonomously by individuals or organizations with little need to interact around production. However, there may still be local knowledge spillovers, and a unique, localized knowledge community may emerge. This situation represents a different kind of cluster such as the classic fine-arts community where writers or visual artists congregate in a given place. Some of the cultural or creative industries are ‘transaction intensive’ in that large groups of professionals are required to work on projects. However, others require only small groups or solitary individuals who may only need to interact occasionally around specific tasks. The Danish landscape painters who gathered in Skågen during the later-19 th century are an example of this, as are groups of intellectuals and writers, who gathered in centers such as Paris or New York. While SCOTT (1999) has argued that cultural clusters are often tied together through transacting relationship—proximity to galleries, printers, and publishers, for example – it is clear that many of these transactions are infrequent enough to have little material effect on how work is carried out. Rather, the main locational draw is the ability to be part of a place-based ‘conversation’ that defines a knowledge sub-community. Biographical of individuals in artist communities often highlight the importance of meetings in well-known restaurants of bars, where permanent familiarity allows to “go there without an appointment” to exchange ideas with other fellow artists or writers 17 . These places are examples of temporary and local proximity, where the two elements are equally necessary to foster idea exchange: proximity must be temporary to allow writers and painters to work alone; at the same time, locality allows familiarity without planning. “Artists residencies”, where creative individuals spend various months and join fellow artists and writers in specific moments of the day (at 17 “Hitch-22”, by the writer Christopher Hitchens, is a nice and recent example of this works. Hitchens devote excellent pages to his meetings in a London bar with a large group that included the best British writers of his generation.
-131dinner, for example) are also contemporary examples of places joining temporary (during dinners) and geographic (during some months) proximity for fostering idea exchange in jobs which have to be performed mainly alone. 4.6 IMPLICATIONS The analysis in this article has aimed to show the usefulness of considering knowledge circulation and knowledge creation as distinct, if interrelated, problems in the geography of knowledge. Knowledge creation, I have argued, often requires intensive negotiations around the potential and feasible combination of knowledge elements, the framing and definition of problems, and the alignment of the criteria that the participants bring to the process (GRANT 1996, MUTHUSAMY and WHITE 2005). It draws on knowledge and skills that are embodied, and so give rise to geographies governed by the spatial transaction costs of bringing interacting parties together or over a sustained period in an efficient manner. The issue with knowledge circulation, on the other hand, is that knowledge produced in one context may travel poorly across institutional cultural and organizational boundaries and be difficult to interpret outside of the socio-spatial context in which it originates. This circumstance may create a situation where local observers learn through imitation and adaptations of successful innovations, that others outside the cluster are unable to understand as effectively (MALMBERG and MASKELL 2002). However, I have also pragmatically suggested that, contra ideas of localized learning, the contextual issues posed by the contextdependent nature of much new knowledge can often be overcome when there is a strong desire to do so. Adopting the view that these two knowledge issues are distinct and may play out at different spatial scales allows for a more differentiated, and I believe, realistic view of knowledge geographies. The classic debate regarding ‘proximities’ was set up as a dualism: knowledge circulation either requires permanent proximity or it does not. I have suggested that some activities (knowledge creation when it involves complex exchanges) may benefit from localization while others (knowledge sharing) can be arranged through more occasional contact. However, the opposite geographical
-132situation might also be true, in which productive activities hardly require proximity, but a local knowledge conversation nonetheless creates localization. A failure to recognize the distinction between knowledge creation and knowledge circulation in order to understand what might be particular to clusters and what instead might occur through ‘temporary proximity’ at other spatial scales, can lead to policy failures. Regional policies have often focused on thickening relationships within the geographic cluster, by building institutional relationships that encourage feedbacks and learning and encouraging cluster buzz (ASHEIM 1996). It is assumed that this will lead to more robust endogenous innovation dynamics (BATHELT 2005). However, these ideas of building institutional connections within the cluster are now being re-evaluated (HASSINK and KLAERDING 2012, RUTTEN and BOEKEMA 2012) as the value of tapping into knowledge networks organized at a much larger scale beyond the region is recognized. On the other hand, the assumption that the availability of ‘contextual knowledge’ and the ability to form part of a knowledge community at a distance means that ‘being there’ no longer matters may also lead to inappropriate policies. My research into the European animation industry investigated the extent to which the ‘temporary clusters’ and other temporary gatherings that European animation had organized could substitute for the fact that many of the firms involved were not permanently colocated, or clustered (COLE 2008). I found that these meetings were quite successful at bridging institutional differences and allowing learning and collaboration between studios in different localities. However, in terms of the creative work of developing high-quality animation content, the spatial dispersion caused problems. Attempts to distribute production among multiple locations often led to high spatial transaction costs –expressed in terms of time spent traveling, missed communications, and wasted work-- as teams tried to coordinate complex knowledge work through spatially distributed collaborations. More importantly, attempts to organize productions in a way that didn’t require co-presence often inappropriately reduced the scope to which team members could apply and use tacit creative skills in the production process. Aspects of production that could have benefited from the creativity and judgment of
-133the workers involved were often simplified to accommodate the limited ability of workers to engage in multi-dimensional conversations. Referring back to the matrix presented in Table One, the policy of European animation assumed that animation belonged in the lower-right quadrant, but in fact the lower left quadrant, in which cosmopolitan knowledge flows are combined with localized production, would be more appropriate. Finally, the distinction between knowledge creation and circulation may help to overcome a puzzle regarding the common assertion that localization can be explained with reference to the idea that tacit knowledge can only be transferred locally. As the literature on global knowledge networks has developed, and scholars have begun to investigate the importance of temporary places such as global conferences and marketplaces in knowledge circulation, the unique suitability of the cluster for knowledge circulation has come in doubt. Writing about trade fairs, or what they call ‘temporary clusters’, MASKELL, BATHELT, et al. (2006, p. 997) have asked: “ !!# ! !$ "!# "# ! !! ! ! ! !
-134BIBLIOGRAPHY ALLEN, J. (2000). Power/economic knowledges: symbolic and spatial formations. Knoweldge, Space, Economy. J. Bryson, P. W. Daniels, N. Henry and J. Pollard. London, Routledge: 15-33. AMIN, A. (1999). "An Institutionalist Perspective on Regional Economic Development." International Journal of Urban and Regional Research 23(2): 365-378. AMIN, A. and P. COHENDET (2004). Architectures of Knowledge: Firms, Capabilities, and Communities. Oxford; New York, Oxford University Press. AMIN, A. and J. ROBERTS (2008). "Knowing in action: Beyond communities of practice." Research policy 37(2): 353-369. AMIN, A. and N. THRIFT (1995). Globalisation, Institutional 'Thickness' and the Local Economy. Managing Cities: The New Urban Context. H. P. ARTHUR, W. B. (2007). ‘The structure of invention.’ Research Policy 36(2): 274287. ASHEIM, B. (1996). "Industrial Districts as 'Learning Regions': A Condition for Prosperity?" European Planning Studies 4: 329-400. ASHEIM, B., L. COENEN and J. VANG (2007). "Face-to-face, buzz, and knowledge bases: sociospatial implications for learning, innovation, and innovation policy." Environment and Planning C 25(5): 655. ASHEIM, B. and M. GERTLER (2005). "The geography of innovation." The Oxford handbook of innovation: 291-317. BATHELT, H. (2005). "Cluster Relations in the Media Industry: Exploring the 'Distanced Neighbor' Paradox in Leipzig." Regional Studies 39(1): 105-127. BATHELT, H. and R. GIBSON (2013). "Learning in "Organized Anarchies"." Technological Search Processes and Knowledge Floses at International Trade Fairs Regional Studies(47). BATHELT, H., A. MALMBERG and P. MASKELL (2004). "Clusters and Knowledge: local buzz, global pipelines and the process of knowledge creation." Progress in Human Geography 28(1): 31-56. BATHELT, H. and N. A. SCHULDT (2010). "International Trade Fairs and Global Buzz I: Ecology of Global Buzz." European Planning Studies 18(2): 1957-1974. BATHELT, H. and P. G. TURI (2013). Knowledge Creation and the Geographies of Local, Global, and Virtual Buzz. Knowledge and the Economy, Springer: 61-78.
-135BECHKY, B. A. (2003a). "Object Lessons: Workplace Artifacts as Representations of Occupational Jurisdiction1." American Journal of Sociology 109(3): 720-752. BECHKY, B. A. (2003b). "Sharing meaning across occupational communities: The transformation of understanding on a production floor." Organization Science 14(3): 312-330. BEST, M. H. (2001). The New Competetive Advantage: The Renewal of American Industry. USA, Oxford University Press. BOSCHMA, R. A. (2005). "Proximity and Innovation: A Critical Assessment." Regional Studies 39(1): 61-74. BOSCHMA, R. A. and A. L. J. TER WAL (2007). "Knowledge networks and innovative performance in an industrial district: the case of a footwear district in the South of Italy." Industry and Innovation 14(2): 177-199. BRESCHI, S. and F. LISSONI (2001). "Localised Knowledge Spillovers vs. Innovative Milieux: Knowledge "Tacitness" Reconsidered." Papers in Regional Science 80: 255-273. BROWN, J. S. and P. DUGUID (1991). "Organizational learning and communities of practice: towards a unified view of working, learning and innovation." Organization Science 2(1): 40-57. BRUSCO, S. (1996). Global systems and local systems. Local and Regional Response to Global Pressure: The Case of Italy and Its Industrial Districts. F. Cossentino, F. Pyke and W. Sengenberger. Geneva, ILO-International Institute for Labor Studies. BEUNZA, D. and D. STARK (2004). "Tools of the trade: the socio-technology of arbitrage in a Wall Street trading room." Industrial and Corporate Change, 13(2), 369400 BUNNELL, T. G. and N. M. COE (2001). "Spaces and scales of innovation." Progress in Human geography 25(4): 569-589. CHOI, Y. B. (1993). Paradigms and Conventions: Uncertainty, Decision Making, and Entrepreneurship. Ann Arbor, The University of Michigan Press. COE, N. M. and T. G. BUNNELL (2003). "‘Spatializing’knowledge communities: towards a conceptualization of transnational innovation networks." Global networks 3(4): 437-456. COLE, A. (2008). "Distant neighbours: the new geography of animated film production in Europe." Regional Studies 42(6): 891-904. COLLINS, R. (1987). Interaction Ritual Chains, Power and Property: The MicroMacro Connection as an Empirically Based Theoretical Problem. The micro-macro link. J. C. Alexander. Berkeley, University of California Press.