Ilkka Tuomi Networks of Innovation Change and Meaning in the Age of the Internet 2006

Networks of Innovation

This page intentionally left blank

Networks of Innovation

Change and Meaning in the Age of the Internet

ILKKA TUOMI

1

Great Clarendon Street, Oxford OX DP

Oxford University Press is a department of the University of Oxford.It furthers the Universitys objective of excellence in research, scholarship,and education by publishing worldwide in

Oxford New York

Auckland Cape Town Dar es Salaam Hong Kong KarachiKuala Lumpur Madrid Melbourne Mexico City NairobiNew Delhi Shanghai Taipei Toronto

With ofces in

Argentina Austria Brazil Chile Czech Republic France GreeceGuatemala Hungary Italy Japan Poland Portugal SingaporeSouth Korea Switzerland Thailand Turkey Ukraine Vietnam

Oxford is a registered trade mark of Oxford University Pressin the UK and in certain other countries

Published in the United Statesby Oxford University Press Inc., New York

Helsinki,

Sitras Publication Series, publication no. ,ISSN (Sitra)Sitra the Finnish National Fund for Research and Development

The moral rights of the authors have been asserted

Database right Oxford University Press (maker)

First published First published in paperback

All rights reserved. No part of this publication may be reproduced,stored in a retrieval system, or transmitted, in any form or by any means,without the prior permission in writing of Oxford University Press,or as expressly permitted by law, or under terms agreed with the appropriatereprographics rights organization. Enquiries concerning reproductionoutside the scope of the above should be sent to the Rights Department,Oxford University Press, at the address above

You must not circulate this book in any other binding or coverand you must impose the same condition on any acquirer

British Library Cataloguing in Publication Data

Data available

Library of Congress Cataloging in Publication DataTuomi, Ilkka.

Networks of innovation : change and meaning in the age of the Internet / Ilkka Tuomi.p. cm.

Includes bibliographical references and indexes.. Technological innovations. . Computer networks. . Internet. . Linux. I. Title.

T. T .dc21

Typeset by Newgen Imaging Systems (P) Ltd., Chennai, IndiaPrinted in Great Britainon acid-free paper by Biddles Ltd, Kings Lynn, Norfolk

ISBN ISBN X (Pbk.) (Pbk.)

1

ACKNOWLEDGEMENTS

Many friends and colleagues provided ideas, comments, and support while I wasworking on this book. I wrote the rst version of the manuscript during my year-and-a-half stay at the University of California, Berkeley. Sitra, the Finnish NationalFund for Research and Development, kindly supported my research at Berkeley, asa part of their innovation research program. My former employer Nokia providedpartial funding for my research project.

Finished books are often the result of years of work and they can contain manylayers of ideas built one over another. It may be difcult to acknowledge all thoseintellectual debts that go into writing a book. Sometimes, however, it is easy to recognize the importance of a specic individual. Without the personal and intel-lectual support of Manuel Castells I could not have written this book. Manuel invited me to Berkeley, commented on my work and the manuscript at differentphases, and, together with Emma Kiselyova, helped me to feel at home there.

Berkeley is one of the intellectual hot spots in the world. During my stay there, I was able to discuss my ideas in various seminars, workshops, and conferences,and greatly enjoyed the opportunity to talk with many brilliant thinkers andresearchers. I would like to thank in particular John Canny and Jerry Feldman, whoboth gave feedback on my work and linked me with many key researchers in theSan Francisco Bay area. Jerry also organized excellent working conditions for me atthe International Computer Science Institute. I gained much from discussions witha great number of people. I would like to thank Guy Benveniste, Martin Carnoy,Robert Cole, Paul Dourish, Claude Fischer, Blanca Gordo, Bronwyn Hall, MartyHearst, Nalini Kotamraju, Martin Kenney, Jennifer Kuan, Andrew Leonard, AaronMarcus, David Mowery, Bonnie Nardi, Richard Nelson, Christos Papandimitriou,Walter Powell, Laurence Prusak, Richard Rosenbloom, Minna and Nir Ruckenstein,Annina Ruottu, Warren Sack, Pam Samuelson, Annalee Saxenian, Claus OttoScharmer, Susan Stucky, Nancy Van House, Hal Varian, Bill Verplank, Georg vonKrogh, Jack Whalen, Matthew Zook, John Zysman, and my friend and colleaguefrom Nokia Research Center, Jukka-Pekka Salmenkaita.

Ikujiro Nonaka provided me with the opportunity to present an early version ofchapter 7 at the Berkeley Knowledge Forum, and I have gained much from discuss-ions with him in various places around the globe. Paul Duguid read carefully anearly version of the manuscript and made many extremely useful, inspiring, and

knowledgeable comments. John Seely Brown provided intellectual support andencouragement at various phases of the project.

Many of the ideas in this book were discussed among the members of the MenloCircle, originally organized around the Stanford KNEXUS program. I would like tothank in particular Liisa Vlikangas, George Campbell, Syed Shariq, Eilif Trondsen,Peter Coughlan, Renee Chin, Mahnoush Haririfar, and Helga Wild.

The Institute for the Future kindly invited me to join their Global InnovationOutlook program, which provided me with the opportunity to study the SiliconValley innovation culture and get feedback on my work-in-progress. I would like tothank in particular Marina Gorbis, Patric Carlsson, Rod Falcon, Paul Saffo, JanEnglish-Lueck, and Magnus Karlsson for useful discussions.

I am also thankful for all the intensive discussions among the members of Sitrasinnovation research program, and in particular for the comments provided by GerdSchienstock, Timo Hmlinen, J.-C. Spender, Reijo Miettinen, Aija Leiponen, PeterMcGrory, Tanja Kotro, Rogers Hollingsworth, and Antti Hautamki. In Europe,Japan, and the US, I have also beneted from discussions with Aleksi Aaltonen,Marko Ahtisaari, Jean-Claude Burgelman, Kathy Curley, Kirsten Foot, Kinji Gonda,Sara Heinmaa, Jeremy Hunsinger, Juha Huuskonen, Yrj Engestrm, JyriEngestrm, Sirkka Jarvenpaa, Petri Kasper, Kari Kuutti, Dorothy Leonard, TarmoLemola, Irma Levomki, Tuomas Lukka, Teija Lytnen, Ian Miles, Hajime Oniki,Mika Pantzar, Matthew Ratto, Tuomas Toivonen, Linus Torvalds, Ryoko Toyama,Paavo Tuomi, and Jaakko Virkkunen.

Indeed, without interaction and networks no innovation could happen and noknowledge could be created.

I.T. February

Helsinki

ACKNOWLEDGEMENTS

CONTENTS

List of Figures ix

List of Tables xi

. Introduction

. Innovation as Multifocal Development of Social Practice

. Putting the User in Focus . Use as Meaningful Practice . Production as an End . Investment and Invention of Meaning . Community as the Locus of Practice . Interpretative Flexibility and Ecology of Social Practices . Social Drivers of Innovation . Individual Exploration . Spaces of Novelty

. Dynamics of Networked Innovation Spaces

. Inventing the Web

. The First WorldWideWeb Proposal . State of the Art: KMS . Architecture of the WorldWideWeb . Mobilizing Resources . The Vision of Xanadu . Sources of Success

. The Making of the Internet

. Laying the Infrastructure . Networking the World . Competing Technologies . Message-packets and Resilient Networks: Innovation at RAND . Time-sharing and Network Society: Work at NPL . Interactive Computing: Augmenting the Human Mind . Time-sharing and On-line Communities . IPTO: Translating Ideas into Money and Technology

. Analysis of the Early Phase of Internet Development

. Technological Frames . Resource Mobility in the Early Phases of Internet History

. Socio-Cognitive Spaces of Innovation and Meaning Creation

. Thought Collectives . Speech Genre and Chronotope . Communities of Practice . Social Learning in Communities of Practice . The Concept of ba

. Breaking through a Technological Frame

. Two Evolutionary Paths of Communities . Development of Specialization, Division of Labour, and

New Technological Frames . Combinatorial Innovation in an Ecology of Communities . Layered ba and Combinatorial Innovation

. Combination and Specialization in the Evolution of the Internet

. Email as a Combinatorial Innovation . ARPANET Ecology and the Evolution of

the Network Working Group

. Retrospection and Attribution in the History ofArpanet and the Internet

. The First Paper on Packet-switching Theory . Reconstructing the Internet

. Learning from Linux

. The Evolution of Linux . The Linux Developer Community . Sedimentation, Translation, and Reduction of Complexity . Quality Control, Linuss Law, and the Ecology of Bugs . Rules, Regulations, and Intellectual Property . Developer Incentives and Resource Allocation

. Concluding Remarks

. Linux as Modern Economy . The Hierarchy of Innovation . The New Economy . The Road Ahead

References

Name Index

Subject Index

CONTENTS

LIST OF FIGURES

. WWW servers connected to the Internet . Changing productionconsumption linkages . How the World Wide Web became the web . Investment cost of transatlantic cable per minute of use . Total number of transatlantic voice circuits . Revenues from international services in the US, . Total share of international revenues in the US, . The US Postal Service, pieces of mail handled . Some key loci of innovation in the evolution of the

Internet, . The general structure of thought collectives . The basic SECI-model of knowledge creation . Bakelite as a structural coupler between communities . A layered ba . Layers of a systematization ba . The waves of Internet technology development . The division of labour in the early phases of Internet development . The production of fractal community structure . The Morris worm and Internet articles in

newspapers, . The Morris worm in major newspapers, . Early Internet awareness . Internet articles in the Financial Times,

. Layers of a Unix operating system . Linux kernel distributions . New source code directories . Removed directories as a function of time . Number of people in the Linux CREDITS le, . Location of active contributors, March . Geographical expansion of development activity,

March March . Linux architecture . The Growth of Linux core kernel components

. Growth of architecture-dependent and extensible components . Resource and object development paths . The standard interface as a multifaceted translation mechanism . The basic bug removal cycle . Mediated interactions in the bug removal process . Licence type distribution in the Linux Software Map

database,

LIST OF FIGURES

LIST OF TABLES

. The augmentation frame, . The on-line communities frame, . The communication networks frame, . The electronic services frame,

. Kernel bug management resources . Licence types in the Linux Software Map database,

This page intentionally left blank

CHAPTER

1Introduction

According to user surveys, the Linux operating system is rated as the best operatingsystem available. It is considered to be more reliable than its main competitors. Itsfunctionality is claimed to be better, and according to many experts, new releasesof Linux implement innovative ideas faster than its competitors. In other words, itis argued that Linux development creates complex new technology better and fasterthan the biggest rms in the software industry.

Yet, Linux also seems to break many conventional assumptions that underlieresearch on innovation and technological change. Linux is developed by an informalself-organizing social community. There is no well-dened market or hierarchyassociated with it. Most of Linux development occurs without economic transac-tions. Instead of getting paid for their efforts, the developers often spend a lot ofmoney and effort to be able to contribute to the advancement of the developmentproject.

The open source development model, which underlies Linux, has attractedincreasing attention in recent years. Today, Linux is considered to be a seriousthreat to Microsofts market dominance in operating systems. More generally, opensource development projects have in recent years had a major impact in softwareand internet-based industries. For example, almost per cent of Internet con-nected Web servers were open source Apache servers in October . As can be seen from Fig. ., the second most popular Microsoft servers were about onethird as popular with per cent. Although Microsoft has gained market share withits Internet Information Server, at the end of about per cent of active web sites were running Apache. The most common operating system in the webserver machines was Linux. Some open source projects, such as Sendmail, Perl,and Emacs, have achieved large user bases, making it difcult for commercialenterprises to enter the market.

Linux has been developed in the open source mode to a large extent because theInternet itself was to a large extent developed in this same mode. The collaborative

http://www.uk.linux.org/LxReport.html. Source: Netcraft, http://www.netcraft.com/survey/. For a discussion on server market shares,

see Netcraft and Peeling and Satchell (Peeling and Satchell, ).

and participatory development model gained visibility in the mid-s, when theearly users of time-shared computers realized that collaboration often producedunexpected benets. The predecessor of the modern Internet, ARPANET, was cre-ated in this mode, and many critical contributions, such as Internet email, Usenetnews, and the World Wide Web emerged as a result of open collaboration. TheInternet Engineering Task Force, which denes standards for the Internet, has alsoused an open source approach since its formation in (Bradner, ).

Several commercial software rms have recently tried to adopt aspects of theopen source model. For example, Netscape announced in that it would dis-tribute the source code of Netscape Communicator with open source licence. IBMdecided to use the open source Apache server as the core of its Web server offers.Red Hat, SuSE, Caldera, and other new economy rms, in turn, make their businesson packaging Linux distributions and by producing added value for Linux users.Sun Microsystems has used a version of the open source model to support devel-opment of its Java and Jini platforms. After launching an attack on Linux in ,Microsoft declared that it will have its own Shared Source Philosophy, which wasaimed at making open source development possible without losing intellectualproperty rights. In all these cases, business rms are experimenting with ways tobenet from innovation that occurs in the open source communities. Instead of traditional economic competition, such initiatives rely on symbiotic relationships,and on the willingness of developer communities to collaborate.

In much of the innovation literature, innovation is dened as something that haseconomic impact. Linux and other open source initiatives show that this denition is problematic and possibly misleading in important practical cases. For example,during its history, most Linux development has occurred independently of direct eco-nomic concerns. It would be tempting to argue that Linux development is differentfrom economic activity and something that, strictly speaking, should not be calledinnovation. Indeed, in its early history Linux development was not in any obviousway associated with changes in production functions, market competition, or appro-priation of economic investment and surplus. Yet, obviously Linux developers collect-ively produce new technology. If economy is about collective production, this is it.

INTRODUCTION

Nov. 1995 Nov. 1996 Nov. 1997 Nov. 1998 Nov. 1999 Oct. 2000

Apache

Microsoft

iPlanetNCSA

Other

35

70

0

Per

cen

tage

Fig. 1.1. WWW servers connected to the Internet

Source: Netcraft, http://www.netcraft.com/survey/.

Linux, therefore, is an interesting test case for economic theories of innovation andtechnology development. For example, the history of Linux allows one to question to what extent existing economic models of innovation and technological develop-ment capture phenomena that underlie collective production of new technologies.

In very practical terms, Linux is an economically important phenomenon.Indirectly, the success of many new businesses, venture capitalists, investment funds,and individual investors critically depends on the productive activities of the Linuxcommunity. Today, many corporations, governments, public sector organizations,and individual developers are starting to deploy Linux to cut costs, promote inter-operability, and avoid lock-in to proprietary systems. Yet, when we consider theentire history of Linux, the economic impact seems to appear almost as an after-thought and as a side effect of a long period of technology creation. Linux, therefore,provides an interesting history of globally networked innovation, illustrating thesubstance that underlies the discussions on the new economy. If the new economyis about global Internet-enabled and software-driven production, this is it.

More generally, the history of Internet-related innovations enables us to discussthose social and cognitive phenomena that underlie technological change. Bystudying such innovations, we can open some black boxes of innovation theory,including such widely used concepts as learning, capability, utility, and consump-tion. By observing the development of the Internet, we can describe the microstruc-ture of innovation, and transcend the boundary between invention and innovation.

Although such studies have obvious consequences for innovation research ingeneral, Internet-related innovations are, however, also special. On the Internet theproducts of innovative activity are externalized as technological artefacts and docu-ments that can be studied relatively easily. Never before has innovation and itsresults been recorded in such historical detail. On the net we live in dog years, butour memory is that of an elephant. There exists sufcient documentation so that wecanat least tentativelydescribe some key principles that underlie the develop-ment of Internet related innovations. For a researcher on technological change, thisis an exciting opportunity.

Internet related innovations are obviously important as the Internet has becomea key technology in many areas of our everyday life. Below I will argue, however,that these innovations reveal important aspects of all innovative activity. Indeed,my key message is that the traditional models of innovation are often misleading,and that they will become increasingly misleading in the future. In practice, wehave to move beyond abstract descriptions and ask what makes novelty meaning-ful. This leads to social and cognitive theories of innovation.

From a practical point of view, Internet related innovations also provide test cases for analysing product development models and proposals for organizing for innovation. For example, the extensive use of modern communication and collaboration technologies in Linux development highlights some aspects of tech-nology development that were not easy to see in earlier studies on innovation.Although I will not explicitly discuss organizational or policy implications below, I believe that the following chapters highlight several points which have such implications.

INTRODUCTION

Linux, open source projects, and Internet-related innovations may have develop-mental histories where collaboration and networking are more visible than in someearlier innovations. The open source model, however, obviously goes beyond soft-ware programming projects. As many commentators have observed, the process ofscience itself is very much based on peer-review, incremental development, non-economic motives, and geographically distributed collaboration. Indeed, tradi-tional models of innovation often assumed that basic research generates ideas andtechnologies that are appropriated by entrepreneurs who turn them into productsand money. The history of Linux and Internet-related innovations enables us to seehow the boundaries between basic and applied research are being transformed.Indeed, I will argue below that the distinction between basic and applied researchneeds to be reconsidered.

From the very beginning, the Internet has been used to distribute work and itsresults. Division of labour is the foundation of all societies; the Internet, however,makes it possible in qualitatively new ways. A study on Internet-related innova-tions, therefore, has implications when we try to understand the ongoing socialtransformation towards the network society. To give just one example: when NASAran its Clickworkers pilot where volunteer Internet users could mark craters on pictures of Mars, between December and June people marked over . million craters. Although each volunteer only marked a few craters, collectivelytheir results were indistinguishable from those of a well-trained expert. This exam-ple is interesting as it shows that trivial individual effort may lead to a high-qualitycollective outcome. In a very concise form it shows one way by which a new balancemay emerge in the network society between increasing specialization and network-enabled participatory decision-making. Internet-related innovations, therefore,have relevance both when we try to understand how new technologies are devel-oped but also when we try to understand how technological development andsocial change could be linked in the future.

History is always constructed from the perspective and for the purposes of thepresent. A useful history, however, provides opportunities for more than one inter-pretation. Historical description, therefore, has to be rich enough in detail and ithas to give room for multiple voices. Yet, a balance has to be found between detailsand conciseness. Reality is always richer than any of its descriptions. I have tried tosolve this problem by combining relatively general conceptual arguments with out-lines of specic innovation histories and more detailed in-depth case studies. Somechapters make rather controversial theoretical claims without extensive empiricalsupport for these claims. Subsequent chapters, hopefully, ll in some of the details.

The next chapter introduces some main concepts and assumptions that underliethe present work. In effect, it tries to set the reader in a position where the subse-quent discussion can make sense. It points out that innovation is fundamentallyabout social change, and that innovations emerge and become articulated whenthey are taken into meaningful use in social practice. It argues that meaningful

INTRODUCTION

http://clickworkers.arc.nasa.gov/documents/crater-marking.pdf: Clickworkers results: cratermarking activity, July .

useas well as the meaning of technology itselfis grounded on social groups that can be called practice-related communities. As a result, innovation and tech-nological change can be studied as phenomena that occur within an ecology ofsuch communities. Construction of technology requires construction of meaning,and new technology is much more than improved functionality. Instead of theupstream of the traditional linear model of innovation, we have to focus on thedownstream where social communication and change occurs. All innovation issocial innovation. Innovation does not happen out there in the world of objects,but in society and in minds. More particularly, it happens in the minds of the users,which are intrinsically integrated with the activities of the users. Those cultural andmaterial resources that are available for the users, therefore, become key resourcesin the innovation process.

The third chapter is a quick rst take on making these concepts more concrete. Itillustrates the nature of innovation by outlining the history of the World Wide Web.It asks who invented the Web, what were the resources used in its invention, andwhat actually was invented in the process. Many of the details of this history arewell known. Many accounts of the history of the World Wide Web, however, alsoshow that some details of the story are often missed. These details become import-ant when we try to understand innovations such as the World Wide Web.

The fourth chapter moves from recent history back in time, describing the earlyphases of the evolution of the Internet. More exactly, the focus is on that point of time when computer networking was only an idea. The chapter introduces thehistorical data that will be used in subsequent chapters. Although there now existsexcellent histories of the Internet, such as those written by Abbate () andNaughton (), it is necessary to provide enough historical detail to make the origins of the Internet understandable. In the process, I will also make some notesthat hopefully complement existing histories in interesting ways. The chapterdescribes how electronic communication systems evolved and laid conceptual andmaterial foundations for computer networks. It also introduces leading actors whoplayed key roles in the early phases of computer networking.

The fth chapter summarizes the early history of the Internet and describes thevarious technological frames that generated the basic innovations of computer net-working. In other words, it puts history in the context of technology and innovationstudies. It also discusses resource mobility in the early phases of the Internet devel-opment. One main claim in the book is that innovation occurs when social practicechanges. The mobility of resources, therefore, is a key factor in enabling and con-straining innovation.

The sixth chapter returns to the topic of communities. It discusses several alternative theoretical traditions that have described the social basis of meaning,knowing, and knowledge creation. It starts by introducing the concept of thoughtcommunity that was originally introduced by Ludwik Fleck () in the s.Flecks historical study described many of those social processes that underlie the emergence of new scientic knowledge and new technologies. The chapter further discusses Bakhtins speech genres, cultural-historical activity theory, sociallearning in communities of practice, and the concept of ba. Ikujiro Nonaka and his

INTRODUCTION

colleagues have argued that innovation and knowledge creation occur in know-ledge creation spaces, or bas. The chapter discusses the nature of bas, and links thisconcept back to its origins in the epistemological theory of Kitaro Nishida and theKyoto School. The sixth chapter, therefore, introduces a set of alternative theoreticalviews that can be used to understand the cognitive and social basis of innovation.

One of the main arguments below will be that innovation can properly be under-stood only by studying the social basis of innovation. The heroic individual innova-tor is not a good model when we try to understand the evolution and developmentof technology. If knowledge and the meaning of technology is grounded in com-munities that reproduce existing social practice, as this book argues, it may seem,however, that innovation is a contradiction in terms. How is it possible that newsocial practices emerge when communities more or less by denition reproducetheir current practices? How do we break technological frames and how are newtechnological frames created? Chapter argues that there are two distinctive waysthat new communities and new technological practices can emerge. One is basedon increasing specialization, and the other on combination of existing resources. Inother words, there exist two qualitatively different dynamics of innovation, andtheir analysis requires two different theoretical approaches. As a result of these twodifferent modes of socio-technical evolution, the concept of ba can therefore beredened. The chapter links the concept of ba to the sociocultural basis of know-ledge, and proposes a new interpretation of Nonakas knowledge creation model.

Using these theoretical concepts, Chapter then returns to the history of theInternet. It briey discusses email as an example of combinatorial innovation, and describes the evolution of the social structure that provided the basis for thecreation of ARPANET and the Internet. It shows, for example, that both resourcecombination and evolution of specialization have played important roles in thedevelopment of social structure of Internet-related innovation communities. The current Internet community is in many ways rooted to the Network WorkingGroup, which started in as an informal group of computer students. Internet,itself, however, would not have been possible without a combination of resourcesthat came from outside this nucleus or the Internet culture.

Chapter picks up one aspect of this history, which is an interesting topic forboth innovation studies and policy. This is the question of retrospection and attri-bution of authorship. If innovations are to an important part created by their usersand the meaning of innovation is reconstructed from the present position, howshould we read historical accounts that describe evolution of technology? And towhom should the credit go? Did Al Gore really invent the Internet? Or was he justdoing what Rembrandt did: signing off works that, strictly speaking, were producedby others, but which could not have existed without him? Should Linus Torvalds geta patent on Linux? What, indeed, does intellectual property mean when technologydevelopment uses resources that are networked, cumulative, often unintended,and when adaptation of new technological opportunities depends on institutionalchange and competence development in the downstream? Should we reconsiderthe author, or is the confusion created by a wrong conceptualization of the productsthemselves? By analysing newspaper articles that have discussed the Internet

INTRODUCTION

during the last fteen years, we show how the common understanding of theInternet has evolved. As Chapter shows, the heroes of innovation are mentalreconstructions, but so is the technology itself.

Chapter , nally, returns to the case of Linux. It describes both social and tech-nological evolution of Linux and its development community. For example, it showshow technological architecture and social structure co-evolve as technical prob-lems are solved in the social domain and social problems are solved in the tech-nical domain. By analysing in detail the evolution of the structure of Linux sourcecode over a period of years, it shows how social control and coordination becomeembedded in a technological artefact. It also shows how social interaction can betranslated into resources by black-boxing some of the underlying complexitybehind technological interfaces. The chapter argues that one reason why the opensource development model has been successful is that the social translation mech-anisms it uses allow several communities to interface simultaneously to a commontechnological artefact. Moreover, the open source model guarantees that when soft-ware fails, it fails gracefully, at least in the social sense. In open source, black boxeshave transparent and penetrable walls. The chapter also discusses the bug removalprocess in Linux and highlights some trade-offs that are needed to make distributedinnovation and technology development effective.

The last chapter puts the open source model of technology development in abroader perspective, and discusses the cultural and value system that underliesopen source. Indeed, it argues that a study on socio-cognitive basis of innovationleads to a new approach in economic theory, where the concept of value has toaccommodate the idea that in innovation processes new meaning is created andnew domains of social practice are generated. Such expansive theory of econom-ics may lead to new insights when we formulate and study technology and innova-tion policy. The chapter also points out that the networked mode of production that underlies open source may lead to new dynamics in socio-economic develop-ment as the social institutions that usually provide stability in socio-economic sys-tems are constantly renegotiated in the network mode of development. The chap-ter also discusses the differences and similarities between the open source modeland the Silicon Valley innovation system. The chapter nally points out some areasfor further study, and ends with some concluding remarks.

INTRODUCTION

CHAPTER

2Innovation as Multifocal Development of Social Practice

Popular accounts and histories on innovation often focus on inventors and inven-tions. The creative genius of an inventor is commonly viewed as a force that producesnew technologies and reveals hidden laws of nature. Such popular accounts, there-fore, tell us, for example, how James Watt invented the steam engine, how Thomas Edison developed electric lighting, and how Tim Berners-Lee created the World Wide Web.

Research on innovation has conceptually rened this model by separating innova-tion from invention. Invention has generally been understood as it was described inthe popular accounts, as a process of creative insight and heroic efforts in problemsolving. Innovation, in contrast, has been dened as a process that renes inven-tions and translates them into usable products.

This traditional view led to a linear model of innovation. According to this model,innovations are rst invented and then developed, packaged, marketed, and,nally, taken into use. Following Schumpeter (), Usher (), and others, manyauthors dened the process of innovation as sequential phases of idea generation,invention, research and development, application, and diffusion. Many productdevelopment and innovation management models have been based on this linearmodel. Similarly, many theoretical models have been developed to describe andpredict the adoption and diffusion of new products generated in this process.

Since the s it has often been noted that the linear model is too simplied (e.g.Kelly, Kranzberg et al., ; Kline and Rosenberg, ; Padmore, Schuetze, andGibson, ). In practice, innovations emerge in a complex iterative process wherecommunication, learning, and social interaction play important roles. Allen (;Allen and Cohen, ) and others observed that communication and ow ofknowledge is critical in the innovation process. Rogers (), in turn, noted thatcommunication among users is necessary for the diffusion of innovations. VonHippel (; ) emphasized that users often play an important role in theprocess of innovation by modifying and improving products. Cohen and Levinthal(; ) argued that adoption of new innovations requires learning and devel-opment of competences by the potential adopters, whereas Nonaka (),

INNOVATION AND SOCIAL PRACTICE

Dougherty (), and others (Brown and Eisenhardt, ) noted that internaliza-tion of customer and market knowledge is critical for successful product creation.

Today product development processes are often iterative and technologies arerened while products are designed. Product development requires multidiscipli-nary integration and creation of knowledge. Kodama () argued that the conven-tional product development pipeline is increasingly being replaced by dynamicdemand articulation where product concepts are created for non-existent virtualmarkets. In demand articulation, potential user needs are integrated into a productconcept, and the emerging product concept, in turn, is decomposed into develop-ment agendas for its individual component technologies. Many product developersnow use focus groups, study lead users (Grifn, ; Cooper and Kleinschmidt,; Urban and Von Hippel, ), use active exploration and experimentation(Lynn, Morone, and Paulson, ; Thomke, Von Hippel, and Franke, ), and cre-ate alliance networks to improve new product designs (Doz and Hamel, ).

Although innovation research has made impressive progress, it still often relieson a basic assumption that also directed the traditional view. The conventional viewassumed that the product of the invention process has well-dened characteristics.For example, the patent system was designed on the assumption that the function-ality of each invention could be described in detail, so that its novelty could beunambiguously decided. It was also assumed that invention had a well-denedauthorthe inventorand a well-dened moment of birth.

In the conventional view both the inventor and the invention were unproblem-atic and easy to dene. The moment of invention created simultaneously both theinventor and the invention. Although it was well understood that the primaryinsight often required development before an invention was articulated as a work-ing prototype and could be produced, the exact details of the process of inventionwere often considered to be irrelevant.

Economists opened the black box of technology when they realized that innova-tion is a driver for economic growth. At the same time, however, invention was putin its own opaque box, veiled under impenetrable layers of creativity and insight. Asa result, technological development was conceptualized as consisting of two qual-itatively different phases: invention and its subsequent development into a product.

Science and technology studies, however, provide ample historical evidence thatthis fundamental assumption is not valid in general. New technologies do not comeinto the world ready-made. Instead, they are actively interpreted and appropriated by existing actors, in the context of their existing practices. A single technologicalartefact can have multiple uses, and new uses may be invented for old artefacts.Often a product is used in unanticipated ways, and perhaps no one uses it the wayits designers expected it to be used.

Where, then, can we nd the author of an innovation? When, exactly, is a new innovation born?

When patent systems were increasingly deployed in the th century, these assumptions were,however, often contested, cf. Machlup and Penrose ().

One way to see the limits of the conventional view is simply to turn it around.Instead of a heroic inventor we can focus on a heroic user. The traditional viewassumed that invention happens when a new concrete artefact or mental insight is created. The alternative view starts from a different assumption. Innovation happens when social practice changes. If new technology is not used by anyone, itmay be a promising idea but, strictly speaking, it is not technology. Similarly, if newknowledge has no impact on anyones way of doing thingsin other words, if itdoesnt make any differenceit is not knowledge. Only when the way things aredone changes, an innovation emerges. Therefore we can say that invention occursonly when social practice changes.

This view is a useful starting point and it is compatible with historical evidence.It allows us to rethink some common assumptions that have become so central inthe traditional view that they have become quite invisible. Careful historical studyof innovations also shows that there has always been a great abundance of ideasand visions, only a few of which ever change everyday life and social practice.

Let us, then, at least for a while, assume that this user-centered model is usefuland provides new insights on the process of innovation. If we give up the idea oftechnological innovations as something xed, we can note that technologies andtechnical products have interpretative exibility, to use a term proposed by Bijker(). Different user groups and stakeholders impute different meanings to a giventechnological artefact. A given technological artefact can play several different roles in different social practices. Instead of being a well-dened objective artefact,with characteristics that could be described without reference to social practice, theartefact in question has many, and possibly incompatible, articulations. Thesemeaningful products may develop independently of each other, and one techno-logical artefact can embed several meaningful products simultaneously.

If we adopt this user- and practice-centered model of innovation, it is easy to seethat innovation has many agents and that the process of innovation is distributedin time, space, and across groups that use technology for different purposes. Thetraditional model of innovation focused on a very special case of innovation. Thiswas the case where the user was well-dened, predictable, and whose needs couldbe taken for granted. As will become clear below, it never accurately described howinnovation happens.

. PUTTING THE USER IN FOCUS

By dening innovation as something that generates and facilitates change in socialpractice, we put the user in a central place in the process of innovation. In a veryfundamental sense, it is the user who invents the product.

For example, for many decades after the telephone was invented, it was marketedmainly for business use in the US. When the telephone was not used for businesstransactions, it was often understood as a broadcast medium. Telephone entre-preneurs tried to use the telephone to broadcast news, concerts, church services,



weather reports, and stores sales announcements. The telephone was also expectedto be used for voting campaigns, long-distance Christian Science healing, and tobroadcast lullabies to put babies to sleep (Fischer, : ). Interactive social use oftelephone was neglected for a long time by the inventors and the industry. Social con-versations and visiting over the telephone were not uses that telephone was sup-posed to serve, and industry sometimes resisted such use. As Claude Fischer notes:

The story of how and why the telephone industry discovered sociability provides a few lessonsin the nature of technological diffusion. It suggests that the promoters of a technology do notnecessarily know or decide its nal uses; that they seek problems or needs for which theirtechnology is the answer, but that consumers themselves develop new uses and ultimatelydecide which will predominate. The story suggests that in promoting a technology, vendorsare constrained not only by its technical and economic attributes but also by an interpretationof its uses that is shaped by its and their histories, a cultural constraint that can persist overmany years. (Fischer, : )

The telephone plays a very different role in the different communication cultures ofJapan, China, the US, Spain, Finland, or Bangladesh. In a very fundamental andpractical sense, the telephone is a very different thing in these different cultures.Moreover, although a technological artefact may remain similar in shape and func-tionality, it is constantly created by its users. To give a very simple and concreteexample, much of the revenue and most of the prots of telecom operators inEurope originate today from SMS text messages. When this technology was denedas a part of the GSM standard, no one imagined the various ways the users of thistechnology would appropriate it.

Innovation, therefore, is not generated only by scientists or engineers, and oftenthey are not critical sources of innovation. In many cases, we can take the availabil-ity of science and engineering for granted. In many ways, the modern world is full ofideas and new technologies are rarely, strictly speaking, new. The traditional modelof a heroic inventor is therefore losing some of its obviousness and descriptive value.The emergence of new innovations depends to a large extent on resources that areavailable for the potential users, as well as on constraints that limit change in theircurrent practices. Therefore, to understand innovation, we need to understand tech-nologies in use. How, indeed, does technology become part of social practice?

. USE AS MEANINGFUL PRACTICE

The traditional model of technological innovation was based on the idea that theinventor and the invention are unproblematic. According to this view, a typicalinvention is a well-dened artefact with well-dened characteristics. Consequently,new uses are new ways of using this given artefact. In this view, the telephone, forexample, remains the same even when new ways are found to use the phone.

SMS was mainly intended to be used to notify phone users that they had voice messages waiting.

Use of technology, however, is not something that we can understand as a spe-cic use of a given technology. To talk about something as technology means thatwe already assume some uses. The concept of technology doesnt exist without animplicit model of use. Technological objects are not something that we can discoverfrom nature: they exist as material artefacts that embed uses. Technological arte-facts are artefacts full of meaning. If this meaning is taken away, we are not left withthe objective object without subjective interpretations: instead, we are left with apile of undifferentiated matter.

This is sometimes difcult to realize. We rarely meet technological objects with-out some interpretation of their meaning. Even when we have no clue what theobject is supposed to do, we still normally assume that someone knows. Most current technologies come packaged with standard ways of using the technology,and in everyday conversation we take these standard uses for granted.

For example, it is quite difcult for us to talk about a telephone without inter-preting it as enabler of common everyday practices in which the telephone plays acentral role. When it is used as a hammer or a weight in a shing net, for example,we may nd this amusing or exceptional. Such exceptional uses, however, implythat there is a normal way of using the phone, and a corresponding standard inter-pretation of the meaning of the thing. In some circumstances a telephone may be aperfect hammer; most of the time, however, we are not supposed to use it as such.The reasons are complex and fundamentally social: for example, a telephone mayhave economically more efcient uses in a given culture. There may be cheaperhammers lying around. A telephone embeds a complex system of economic andsocial relationships and if we break the phone by hitting a nail with it, these rela-tionships may break as well.

Technology enters our life as a way to conduct meaningful social practice.Therefore technology does not exist in a pure objective form outside the context ofsocial practice. Technology always exists as technology-in-use, and it is, in general,impossible to nd a stable core use which would well dene the nature of a tech-nological artefact. There is no hard core of technology that would provide a xedfoundation for different variations of use; instead, there are multiple ways a giventechnology can be appropriated by different actors, and different ways these actorscan integrate technological products in their everyday life.

Technology-in-use refers to meaningful use of technology. Meaningful use, inturn, is rooted in social practice. In social life, completely idiosyncratic and uniqueevents make no sense, and they appear as random noise in the social sphere of


Material artefacts, of course, have some uses that are more natural than others. According toGibson (), the world presents itself to us as affordances. For example, a ladder affords ascent ordescent, and a chair affords sitting. We, therefore, see ladders and chairs as such, instead of seeingmeaningless objects, which only after information processing become infused with meaning(Tuomi, : ). More generally, affordances characterize the possible uses of things. For example,it is difcult to push with a rope. A rope affords pulling. But, although material objects afford somethings and do not afford others, their meaning is not xed by a specic given use. In some circum-stances it is even possible to push with a rope. If the pushing, however, becomes common enough,we nd a new name for ropes that are used for pushing.

interaction. Social practice therefore grounds collective meaning. Meaningful use oftechnology, consequently, is always inherently social and related to social practices.

The traditional view on innovation was based on the idea that innovation is wellrepresented by the material object that embeds the invention. A prototypical innovation, therefore, was something like a steam engine or a light bulb. The user-centric view on innovation leads to a very different prototypical innovation. Insteadof material it focuses on mental. In the user-centric view, a prototypical innovationis more like a word or a concept. A word acquires its meaning through the ways it isused in communication. The same word can be used in many ways and it can playdifferent roles in different types of conversations.

Innovation is, therefore, as much about creating new meanings as it is about creating novel material artefacts. Ormore exactlyit is more about creatingmeanings than it is about creating artefacts. A steam engine remains a pile of metalwithout someone making sense of it in the context of ongoing social practice.

Artefacts enter social practice as meaningful objects and we have no way of talkingabout material objects except as meaningful objects. Whereas the traditional viewsaw technological innovation as something that generated functional objects, theuser-centric view sees these objects as carriers of social practice and as artefactsthat embed theories of meaningful use.

Such theories of meaningful use are open for reinterpretation and for new appli-cations. Even though a designer of a new product may have a theory of use, there isno guarantee that this theory works in practice. Moreover, a given technical artefactcan be understood through different theoretical and conceptual frameworks. Theseframeworks can evolve in various directions, long after the design of the artefact is frozen.

One relevant actor in all this, of course, is the producer of a technological prod-uct. In the traditional industrial mode of production, the producer was a manu-facturer. The product naturally played an important role in the manufacturers practices and often was the focus of activity. But even in the most prototypicalexamples of mass production, such as Fords Model-T, the producer had only a limited control of the ways the product was integrated into the everyday life prac-tices of car users. The meaning of a car was not invented by Henry Ford and hisengineers. Instead, it was created by the users. In a very important sense the usersproduced the car, and without their active production, Model-T would haveremained a working prototype. So, while the practices of the manufacturer areimportant when we try to understand the emergence of an innovation, they do notdetermine the evolution of the innovation. The essence of a specic technologicalartefact cannot be found only by asking what its manufacturer believes the artefactto be. Technical specications of a telephone, for example, tell very little about theuses for which telephones are put in different cultures and contexts.


Of course, already the idea of pile of metal itself assumes that there are practices where metalsare differentiated from formless matter and non-metals. A steam engine becomes a steam enginein relation to uses where it is different from just a pile of metal.


. PRODUCTION AS AN END

One way to view the role of manufacturer is to see it as one user of a technologicalproduct among others. The manufacturer tries to make money with the product.Although many users may also try to make money with the technologyfor exam-ple, by cutting costs in their productionthe manufacturer has a special intentbecause this is essentially the only motive for the manufacturer. Whereas for themanufacturer the technological product is an end in itself, for the other users it is ameans of getting something done.

This manufacturers perspective generates the modern economic view of theworld and, on the other hand, only makes sense within it. As innovation researchoften focuses on the economic aspects of innovation, it is important to understandthe character and limitations of this economic view.

The economic view of the world is a very special view. Indeed, it is a special caseof the utilitarian view that underlies also the modern concept of technology. In theutilitarian view, the meaning of objects is their use value. In a consistent utilitarianworld, there are no ends but only means.

Such a world without ends risks becoming a world without meaning. As a reac-tion to the emerging modern utilitarian views, Kant argued that to make values,ethics, and politics possible we have to break the innite regression of means in theutilitarian system. In Kants philosophy, the world is an instrument for our interestsand needs but we have to regard human beings as special cases, and treat them asends in themselves, never only as means for other ends.

Classical economic theory solved the problem of innite regression of means inanother, ingenious, way. It assumed that meaning is fully reected in prices. Insteadof grounding value to something outside the economic system, value became oneaspect of the system itself. The innite chain of meansends relations was con-nected by prot, which became the end of all means.

As a result, the meansends system of the economic view of the world is a closedsystem. Use value is conceptually and without a residue linked to exchange value.Structurally, the economic world parallels the Newtonian closed universe, wherenovelty is a contradiction in terms. Scientically, it all looked very nice, two cen-turies ago.

The economic view is important in the modern world and many actors operateunder its conceptual and practical constraints. For economic actors, the economicsystem is a perpetuum mobile, where quality can only appear in terms of quantity

For example, Craig Mundie (), Senior Vice President of Microsoft, argued in May thatthe software industry needs intellectual property rights because that is the only way prots can bereinvested in research and development, and that it is only by such reinvestment that successfultechnology corporations are able to grow and continue to provide societal benets. Mundie, there-fore, adopts the later Schumpeterian () model, where corporate R&D units become central in theproduction of new knowledge and innovation. According to Schumpeter himself, this leads to thegrowth of the size of rms. Schumpeter, however, also noted that the underlying logic eventuallyleads to the collapse of capitalism.


and where economic growth is the mother of all ends. It is, however, also importantto understand the particular limitations of this view. Perhaps the most fundamen-tal assumption that is built into the structure of economic explanations is that themeaning of products and production can be understood without reference to thesocial and material reality. In a conceptually coherent and completely closed self-referential system of meansends, there can be no loose ends. As the economic viewimplies in this sense a perfect utilitarian world, it is conceptually blind to anythingthat does not operate within a utilitarian logic. Values therefore remain inherentlyexternal to the economic world and make no sense within this world. The meaningof economic life is prot. All other meanings are unlinked from the system, irrele-vant for it, and fundamentally meaningless.

It is always possible to introduce new theoretical concepts that expand any closedtheoretical system so that any particular phenomenon can be explained within thesystem. The economic concepts of utility and preference, indeed, try to do some-thing like this, and link the economic system back to the world. The reasonablenessof such extensions, however, is also an empirical question. At some point we mayend up with a complex Ptolemaic system and may ask whether new conceptualstarting points could make our explanations more powerful and compact.

Turning back to Kant, for example, we might ask whether there are non-utilitarianends. More fundamentally, we can rethink what makes products and productionmeaningful. Such a quest for meaningfulness of technology, products, and innova-tion may at rst look remote from the concerns of entrepreneurs, managers, andmodern policymakers. There are, however, many ways to make prot. In the futuregood business may require understanding how meaning is produced.

. INVESTMENT AND INVENTION OF MEANING

It is useful to illustrate these concepts by looking for innovations that are all aroundus, yet rarely discussed in innovation literature. During the last two decades, theactive role of users has been one central theme in research on fashion. Fashion is aninteresting reference point to studies on innovation as fashion is novelty withouttechnological function. In other words, whereas we often assume that new techno-logy is adopted because it in some sense works better than existing technology,fashion has no obvious link between novelty and function.

Fashion is innovation without progress but also obviously social, and full ofmeaning. According to Grant McCracken (), the meaning of material goods hasthree loci, and there are several different ways meaning can be transferred from onelocus to another. The original location of meaning is the culturally constituted

The inherent destruction of meaning in the utilitarian world was discussed already by Nietzsche.Arendt () has provided an insightful account of the conceptual structure and history of themodern economic world-view in her classic work, The Human Condition, originally publishedin .


world:

This world has been constituted by culture in two ways. Culture is the lens through which allphenomena are seen. It determines how these phenomena will be apprehended and assimil-ated. Second, culture is the blueprint of human activity. It determines the co-ordinates ofsocial action and productive activity, specifying the behaviors and objects that issue fromboth. As a lens, culture determines how the world is seen. As a blueprint, it determines how the world will be fashioned by human effort. In short, culture constitutes the world by supplying it with meaning. (McCracken, : )

According to McCracken, this meaning can be characterized in terms of two con-cepts: cultural categories and cultural principles. Cultural categories represent thebasic distinctions with which a culture divides up the phenomenal world. Thesecategories include categories of time, such as years, centuries, leisure time andwork time, and sacred and profane time; and categories of space, ora, fauna, socialclasses, status, gender, age, and occupation, for example. Cultural principles, inturn, provide the principles that allow us to evaluate, rank, construe, distinguish,and interrelate phenomena in the world. Together, these categories and principlescreate a system of distinctions that organizes the world. Culture, therefore, consti-tutes a world by investing it with its own particular meanings.

It is thus that each culture establishes its own special vision of the world and thus that it renders the understandings and rules appropriate to one cultural context preposterouslyinappropriate in the next. Culture makes itself a privileged set of terms within which virtuallynothing appears alien or unintelligible to the individual and outside of which there is no order,no system, no safe assumption, no ready comprehension. (McCracken, : )

Cultural categories cannot be seen as such in the world. Instead, they remain invis-ible, at the same time providing a structure of distinctions that is continuouslyreproduced in social life. According to McCracken, consumer goods have an import-ant role in materializing and substantiating these categories. Through goods themeanings that organize the world are made visible.

Cultural categories can be embedded in artefacts through the systems of advert-ising and fashion. Advertising brings a consumer good and a representation of theculturally constituted world together, so that the known properties of the world canbecome resident in the unknown properties of the consumer good (McCracken,: ). Similarly, the fashion system invests and divests of meaning in goods.

McCracken noted that the fashion world works in three distinct ways in transfer-ring meaning to goods. One is related to advertising. The fashion system takes newstyles and associates them with existing cultural categories. The fashion system,however, can also invent new cultural meanings:

This invention is undertaken by opinion leaders who help shape and rene existing culturalmeaning, encouraging the reform of cultural categories and principles. These are distant

Research on categorization has, of course, been a key theme in cognitive science. Gardner ()and Lakoff () provide good introductions to the history of categorization research. More socio-logical studies on categorization and its consequences include Bowker and Star (), and Foucault(e.g. ).


opinion leaders: individuals who by virtue of birth, beauty, celebrity, or accomplishment, areheld in high esteem. (McCracken, : )

The third way the fashion system works is when it reforms cultural categories andvalues that dene cultural principles. The groups that are responsible for radicalreform of cultural categories usually exist at the margin of society. These groupsmay adopt cultural categories and principles that differ fundamentally from thetraditional ones. For example, hippies and punks redened the categories of ageand status, and gays redened the category of gender. At the same time thesegroups became meaning suppliers for the mainstream culture.

In the fashion system, journalists, social observers, and market analysts act asgatekeepers, ltering aesthetic, social, and cultural innovations, judging some asimportant and some as trivial. It is their responsibility to observe, as best as theycan, the whirling mass of innovation and decide what is fad and what is fashion,what is ephemeral and what will endure (McCracken, : ). After making theirselections, they engage in a process of dissemination with which they make theirchoices known. These cultural innovations are then invested in products by design-ers who create fashion products for mass consumption.

By consuming the products, the meaning can transfer to the consumer.McCracken proposed that there are four ways this transfer can occur. First, thereexist various possession rituals. Consumers may clean, discuss, compare, reect,show off, and, for example, photograph their new possessions. While all these activ-ities have an overt functionality, they also enable the consumer to claim the pos-session as his or her own. In addition to such possession rituals, there exist alsoexchange rituals, where a consumer chooses, purchases, and presents goods asgifts. The gift giver, for example, may choose a gift because it possesses meaningfulproperties that the giver would like to see transferred to the gift-taker.

A third type of meaning transfer ritual is grooming. Some meaning is in its natureperishable and has to be constantly recreated. The purpose of grooming is to takethe special pains necessary to insure that special and perishable properties residentin certain clothes, certain hair styles, certain looks, are, as it were, coaxed out of their resident goods and made to live, however briey and precariously, in theindividual consumer (McCracken, : ). The object of grooming can also be the good itself. For example, cars, computer games, and tamagotchi may becomesupercharged with meaning when they are groomed. This meaning, in turn, can be transferred to the person who grooms the car, game, or tamagotch, redeningthe person as a proud owner of the consumer item in question.

When meaning becomes invested in a material artefact it may become part of thepersonality of the possessor. Divestment rituals may therefore be needed to separ-ate the object from its consumer. For example, when someone buys an object thathas previously been part of the personality of someone else, divestment rituals mayerase the meaning associated with the previous owner. According to McCracken, thecleaning and redecorating of a newly purchased home may be seen as such an effort.

McCracken noted that in North American culture, cultural categories and principlesare exceptionally indeterminate. People can dene and choose their age, gender,


social class, and other category memberships in ways that are often strictly speciedand policed in other cultures (McCracken, : ). Indeed, much time and energygoes to managing such memberships in a culture where they are not given.

In a modern world, cultural diversity supplies huge amounts of potential mean-ings for everyday use. The resulting instability of cultural categories creates a para-doxical situation. Individuals have to consume mass produced goods to differentiatethemselves. This is not only an economic decision, or an attempt to minimize thecost of constructing ones identity. For differentiation to make sense, it has to be based on commonly known cultural distinctions. Mass consumption, in otherwords, requires mass media. Cultural contingency leads therefore to a situationwhere cultural categories tend to be highly visible. Such a world looks open to innova-tion and individual expression, but it also constrains novelty and expression. Torephrase Henry Ford in the age of electronic highways: you can select any age, aslong as it looks young.

As Thompson and Haytko () noted, current fashion is also often seen as acommon generalized other against which differences can be made. For example,in the interviews conducted by Thomson and Haytko, some students actively triedto avoid becoming a statistic and dressing just like the others. The perceived indi-viduating and transformative power of clothing, however, is ultimately contingentupon a belief that others will notice and care about ones appearance.

Thompson and Haytko argued that clothing often has a metonymic role.Metonymy, used in its normal linguistic context, means that a part, association or aproperty stands for a whole, for example, as in the phrase Wall Street focuses on tech-nology stocks, in Paris wears yellow this spring, or he drank the whole bottle.Clothing and material objects have a similar metonymic function when they situate anindividual as a member of a particular social sphere. A tie, jewellery, baggy jeans, a yar-mulke, or a turban all may stand for a whole lifestyle and a complex system of values.

Simple reproduction of cultural categories provided by the fashion-system is,however, an antithesis for a fashion-conscious consumer. Instead, the highly pro-moted brands and looks are used in developing ones own style. The culturallyavailable resources are combined and adapted to create something new. Indeed, inthe world of US university students, the capability to create coherent ensemblesfrom a range of brands and styles is taken to signify a number of positive meaningssuch as creativity, organization, competence, and conscientiousness (Thompsonand Haytko, ).

The contingent nature of fashion means that the only way its different expres-sions can acquire meaning is through active discourse where these meanings arenegotiated. There is no such thing as a personal fashion, as clothing derives its symbolic capital from cultural categories and principles. But neither can fashion be

In practice, the system of fashion works well because it secures its own foundation. Fashionrequires that we construct our identities under the objectifying gaze of imagined others. As a con-sequence, the attitude that underlies fashion guarantees that we dont need to know what othersreally think. Although identity is fundamentally social, the fashion system enables identity con-struction that is based on mediated representations of others. For the same reason, it is also possi-ble to dress fashionably just for oneself even when there are no others present.

universal. Distinctions become invisible if everyone wears a similar dress.Similarity, therefore, denes an interpretative community, which in turn deneswhat it considers to be similar and where it nds differences.

The culturally constituted world, as described by McCracken, is therefore not ahomogeneous world. Although generic cultural categories provide the backdropagainst which distinctions can be projected, individuals negotiate and recreate the meanings of these distinctions as members of interpretative communities.Sometimes the community may limit the way new interpretations can be invented.For example, a similar uniform or gown may signal that some cultural categories areexcluded from the ongoing discourse.

A closer look at the ways fashion is consumed therefore reveals that goods areactively produced by the consumers. Fashion items are used in a discourse whichdenes and comments cultural categories. Pure fashion lacks functionality andtherefore its use is purely communicative. The consumer is never a passive sink ofgoods or an end point of a production chain. Instead, products that are consumedare used in the production of the user. Only if we assume that these uses of the con-sumer are irrele-vant or uninteresting, can we forget the processes by which theconsumer makes a product meaningful.

In industrial products fashion is often regarded as irrelevant and the fashionindustry usually regards functionality as irrelevant. In both cases, the adoption ofnew products, however, is based on the users capability to make sense of the prod-uct and to integrate it into ongoing practice. This practice, in turn, denes a com-munity that sustains and reproduces the practice in question. Research on fashionhighlights the point that the resources needed to create new social practices andcategories are socially distributed and produced, and that, for example, commun-ication, advertisement, and the reputation of opinion leaders play an important role in their change.

In effect, we therefore redene the producer. The traditional view on innovationassumed that a producer is either an inventor or an entrepreneur who produces anew innovative product and develops a market for it. A user-centric view on innova-tion, in contrast, sees the traditional inventor and the entrepreneurial innovator as users among other users. They have specic roles, competences, and motives but in that regard they do not fundamentally differ from other actors that collec-tively co-produce innovations as meaningful products. Innovations are producedthrough interaction between the different users, and innovation therefore cannotbe localized within a single business rm or in the head of a single inventor. As the following chapters show, this has important implications for the evolution oftechnologies, as well as the ways innovative activity can be organized.

. COMMUNITY AS THE LOCUS OF PRACTICE

The locus of innovation is a group of people who reproduce a specic social practice. Social practice does not exist in a vacuum, and it is not something that an



individual can invent on her own. There are no more private social practices thanthere are private words. Language and practice are both inherently social. A newword can be created by an individual but it becomes a meaningful word only if it istaken into use in language. Similarly, innovations become innovations only whenthey start to play a role in meaningful social practice.

Social practice consists of reproduced forms of action. Technological artefactsoften play an important role in the formation of social practice as they externalizeaspects of practice and transform parts of it from the mental sphere to the concretematerial world. Practices, therefore, exist as complex networks of tools, concepts,and expectations.

When we talk about a given social practice, we therefore assume that there is arecurrent form of activity that has some stability. Social practice structures andorganizes social life, and provides a foundation for collective meaning processing.This foundation is not xed but it provides a practical basis for interpreting theworld. As practices comprise complex heterogeneous networks of artefacts, con-cepts, and ongoing social activity, reconguration and evolution of practice hasmany constraints.

Meaning is not something that can be grounded on individual decisions or cog-nitions about the world. But neither it is something that can be derived from someabstract structure of society. Instead, meaning is grounded on specic commun-ities that produce and reproduce meaning and their unique ways of knowing theworld. Meaning has its origins in collaborative practical activity and the communitythat reproduces specic meanings is the community that reproduces the relatedpractices. The foundation and carrier of social meaning can therefore be called acommunity of practice.

There exist different proposals on how we should conceptualize such commun-ities of practice. Some authors have focused on communities of identity and inter-pretation, others on communities of production, competence development, orcommunication and knowledge creation. These proposals are discussed in moredetail in subsequent chapters. At this point, we may simply note that a communitycreates specic potential uses of technology. The user of technology, therefore, isnot an individual person but a member of a community with a practice that usesthe technology in question. The individual user is engaged in the practices of thecommunity and makes sense of technology in the context of these practices. Wheninnovation changes these practices, new ways of doing things create new inter-pretations of the world. If innovation is technological, technology becomes integratedin social practice in new ways, and acquires new meaning.

As a user of a given technological product, an individual is a carrier of social prac-tice. In other words, the user is conceptually more accurately described not as aperson but as a practice.

This is an abstract conceptual point, but it is also a very important one. Innovation studies often adopt an individualistic and object-centric view. This subjectobject dichotomy is deeply ingrained in our language and conceptual systems. When we try to describe alternatives for it, our concepts easily start to


look imprecise and not well dened. This is because concepts make sense only as parts of larger conceptual systems. Therefore it is impossible just to dene a concept such as a user in a new way, without simultaneously changing the rela-tionships between many other concepts.

For the purposes of the current work, it is not necessary to deal in any great detailwith these conceptual issues. Instead of dening concepts we can describe con-crete examples and historical cases of innovation. By illustrating the processes of innovation we can provide material that allows us to redene the way we useexisting concepts. It is, however, important to note that to the extent that the meaning of technology cannot be grounded on interpretations of individual actors,models of innovation cannot be based on individual users.

The strong theoretical claim that underlies the present work is that technologyexists as technology-in-use, in a context of a specic practice. If we abstract awaythis social practice, we end up with an individual user or aggregate groups of simi-lar users. At the same time, however, we also abstract away technology itself. Theconceptual starting point for innovation studies, therefore, has to be at the level ofsocial practice.

. INTERPRETATIVE FLEXIBILITY AND ECOLOGY OF SOCIAL PRACTICES

The user-centric view on innovation means that the various stocks of meaningavailable to the different users provide the basis from which the innovation is arti-culated. Innovation, therefore, is not created just by using the resources availablefor the producer. In this sense, the innovation process is distributed among a num-ber of stakeholders. Innovations are generated in the interaction between the various users and the artefact that embeds the innovation in a concrete form.

There are many ways of using a given technological product and there are severalcommunities of practice that have their idiosyncratic views on the meaning of the product. The traditional view on innovation often noted only two of these: theproducer and the consumer. In practice, there are many different consumers whoconsume the product in their own productive practices. If we follow a given arte-fact and register all the different communities where it is used, we therefore may

As Shapin () noted in his overview on the sociology of scientic knowledge, any discussionon the alternatives to the object-centric conceptualization of the world has to rely on object-centricconcepts if it wants to be intelligible.

When we talk about the user as a social practice, we therefore implicitly switch to a new view ofthe world where many concepts acquire new meaning. This is, more generally, the key characteris-tic of radical innovations. Radical innovation requires that we change some of the central conceptsand practices in a given community and reorganize the system of activities and meanings in a discontinuous way.


nd many such communities. In each community, the meaning of the artefact is different. Only rarely is there just a single community of users.

For example, word-processing software is an object of development and manu-facturing for people who work in the rm that makes the product. Within the rmthere may be many different communities that deal with the product in differentways. For example, product developers may see it as software code, marketing people may see it as a solution to customer problems, and the nance departmentmay see it as a source of costs and revenue. For store managers it is an object thatneeds to be shelved and sold. For technical support people it is a collection of doc-umented and non-documented features and bugs. For others, a word-processorcan be a tool to write poems, business letters, and technical documents. What aword processor is depends on how you use it. How you use it depends on what prac-tices you are engaged in. A word processor, therefore, can be many different thingsin different contexts of use.

In a sense, these different uses may look trivial. A single product obviously hasdifferent functions for its manufacturer and customer, for example. The point here,however, is more radical. Strictly speaking, there is no single product. Such a thingsimply doesnt exist. This becomes important when we try to understand the evolu-tion and life cycles of technical innovations.

Although software products may seem exceptionally generic and multipurposeproducts, they are not fundamentally different from any other technological products. Wheels, forks, cars, light bulbs, telephones, computers, email, and voicerecognition systems are all multipurpose products and have multiple communitiesof practice using them. Some technological products are, of course, difcult toadapt to different uses. For example, a nuclear power plant may produce heat andelectricity, and there may be constraints on the ways the plant can be used for alter-native purposes. It may, for instance, be difcult to turn a nuclear plant into a play-ground for children. But although there are limits to the ways a given technologicalproduct can be usedand many practices exist where the product is not usedingeneral there are many user communities and many interpretations of the product.

A given technological product evolves in the context of these different interpreta-tions. Different stakeholders make different claims concerning the product. Somestakeholders may dominate this process and make some interpretations sociallymore legitimate and visible than others. Some uses and user communities maybecome dominant, others may become peripheral, and some users may remainmarginal. In the course of evolution of a given technological product the centralityof different communities may change and latent uses, for example, may becomedominant. Although the product itself may remain similar in design, its meaningmay change.

An innovative product, therefore, is about co-development of practices and ameaningful product that plays a role in those practices. Sometimes innovation can occur simply when the meaning of an existing technological artefact is reinter-preted and appropriated in a social practice where the artefact was not previ-ously used. Innovation does not necessarily require change in the design of theproduct.


The development of a given technological product is therefore a continuousprocess. The evolution of its functional characteristics reects a process of socialdifferentiation and negotiation of interests. When the design of the artefact changes,some of the tensions in the underlying social processes become concretely embed-ded in the product. At the same time, the artefact may freeze some of this under-lying social structure. Some technological designs provide ample opportunities forcontinuous innovation whereas some designs effectively implement a xed view on the use of technology, thus limiting the opportunities for innovation.

We should therefore understand innovation as a multifocal process of develop-ment where an ecology of communities develops new uses for existing technologi-cal artefacts, at the same time changing both characteristics of these technologiesand their own practices. Some of these communities have, of course, a more promin-ent position in this process than others. Indeed, the traditional model assumed thatthe only relevant communities are the producer community and the primary usercommunity. In many industrial products this assumption was a reasonable one. Aniron plough, for example, may have its most relevant use in agriculture, and manyof its uses are relatively easy to predict. Many modern products, however, have ex-ible uses and many user communities. The drivers for innovation cannot easily befound by looking at a single group of users or by searching the source of inventionfrom the deep well-springs of individual creation. Innovation is a social phenom-enon. It is generated in complex interactions between several communities, eachwith their own stocks of knowledge and meaning. Technological designs and socialpractices co-evolve. Therefore all innovation is fundamentally social innovation.

. SOCIAL DRIVERS OF INNOVATION

As was noted above, in the multifocal and practice-centered innovation model,innovation occurs when social practice changes. Drivers for innovation can there-fore often be found by looking for tensions and contradictions in existing socialpractice. Social practices form a complex network of interlinked practices and thisnetwork is continuously evolving. Technology addresses a need when it releases orreduces some of the tensions generated in this process.

The problem of exibility, of course, is becoming an increasingly important challenge as prod-uct life cycles decrease. Research on product development models has traditionally focused onprocess exibility and product exibility (Adler, ). Process exibility means that manufacturingcapability can be easily reorganized and product exibility refers to the ability to create product variations. More recently, researchers have studied ways the product creation process itself can bemade more exible (e.g. McKee, ; Mullins and Sutherland, ; Bhattacharya, Krishnan, andMahajan, ; Verganti, ). Bhattacharya et al., for example, discuss the problem of creatingproduct denitions in highly dynamic environments where customers cannot easily articulate theirneeds and where these needs may rapidly change. New product development models, however, donot consider interpretative exibility, or user meaning creation and innovation.

Entrepreneurs develop new technological products with the explicit purpose ofaddressing needs. An entrepreneur interprets the meaning of technology from thepoint of view of a potential user and designs a product that addresses the usersneed. When the product addresses a need that articulates an important tension inthe underlying network of practices, the product can be quickly adopted.

The entrepreneur may also address latent needs that have not yet been articu-lated as well-dened needs. In this case the entrepreneur invents both the need andthe product that addresses the need. This is the process that Kodama () calleddemand articulation.

Of course, there are limits within which needs can be invented. It is difcult or impossible to articulate needs that have no basis in the current social life. A successful entrepreneur is in this sense like a popular poet who puts in words whateveryone was thinking but no one had said before. If a new product gives a compactand coherent expression to something that was not expressed before, the productcan become an important element in the construction of social life.

Often such innovations, however, fail. The entrepreneur is rarely able to see allthe constraints that underlie the forms of current practices and how they limit theways practice can change. In other words, the entrepreneur may have a wrongmodel of the potential use. This happens easily because potential uses are poten-tial: there is no simple way to observe such potential uses in their actual form.Although they can sometimes be simulated and tested they cannot be observedin their ecological context. As a result, the entrepreneur has limited possibilities toimprove the imagined models of use.

Historical analysis of important innovations shows, however, that even when theentrepreneur has a wrong model of use, innovations often succeed. The entrepreneurproduces a product for purposes that look relevant and important. Frequently, how-ever, the product is used for different purposes, and these unintended uses maybecome key drivers in the evolution of the product. Sometimes the producer makes a good guess on how the users will understand the product, but often the guess goeswrong. Often the producer neglects some potential user communities entirely, andthese forgotten communities may become main users of the product. Fundamentally,however, it is the users who either succeed or fail in making the product meaningful.

Entrepreneurial activity thus both addresses existing needs by reducing tensionsin the system of social practices and creates possibilities for new forms of practice.In the latter case, we often talk about a solution looking for a problem. In suchcases, the entrepreneur may have a model of use that doesnt resonate with anypotential user group. Many of the proposed early uses for telephone, television, andcomputer, for example, tried to articulate needs that didnt exist. People really werenot that interested in listening to concerts using a telephone or maintaining recipesin computersespecially when they would have lled a kitchen.

In such cases, the locus of innovation moves away from the producer. Tech-nological products become more like Rorschach ink-blots, and it is up to the usersto gure out what they mean. The telephone can be appropriated for social conver-sations, television can become a medium for advertising, and the computer maybecome a communication machine. Technology creates interpretative exibility



and makes new forms of practice possible. Technology itself can therefore promotechange. It can destabilize existing forms of practice and create contingency whereit did not exist before. Technology is, as it were, thrown to the world for someone topick it up and gure out what to do with it.

More often, however, technology is thrown to the world for specic expected uses.Also then it can be picked up and used in practices where it was not designed to beused. As will be shown below, this is quite a common event. In many ways

Ilkka Tuomi Networks of Innovation Change and Meaning in the Age of the Internet 2006

Documents

Transcript of Ilkka Tuomi Networks of Innovation Change and Meaning in the Age of the Internet 2006