Can a community write a novel? Wiki-novels and and other collaborative writing projects

CAN A COMMUNITY WRITE A NOVEL? WIKI-NOVELS AND OTHER COLLABORATIVE WRITING PROJECTS A dissertation submitted to The University of Manchester for the degree of Master of Arts in the Faculty of Humanities 2009 Veronica Periani SCHOOL OF EDUCATION CONTENTS Page ABSTRACT ................................................................................................................. 4 DECLARATION AND COPYRIGHT ........................................................................ 5 1. INTRODUCTION 1.1 Thematic introduction ........................................................................................ 6 1.1.1. Understanding Web 2.0........................................................................ 6 1.1.2.Collaborative writing ............................................................................ 8 1.1.3 Web 2.0 meets Collaborative writing.................................................. 10 1.2 Problem statement and research question ........................................................ 11 2. LITERATURE REVIEW 2.1 The culture of the Internet: from Web 1.0 to Web 2.0..................................... 13 2.2 Changes within patterns of participation: The expansion of SNS activity and transmediality ........................................ 17 2.3 The convergence between UGC and SNS: collaborative narrative ................. 19 2.4 Summary of the literature review..................................................................... 23 3. METHODOLOGY AND METHODS................................................................... 24 3.1 Contextualising the research question: Methods and choice of data................ 24 3.2 Presentation of methods ................................................................................... 27 3.3 Thematic analysis and Grounded Theory......................................................... 28 4. ANALYSIS OF SELECTED PLATFORMS ........................................................ 32 4.1 Structural analysis of the platforms: Structural features ................................. 32 4.2 Thematic analysis: Writing features................................................................ 37 5. FINDINGS 5.1 Analytical findings ........................................................................................... 52 5.2 Overall findings................................................................................................ 56 5.3 Validity limits................................................................................................... 59 6. CONCLUSIONS AND RECOMMENDATIONS................................................. 60 BIBLIOGRAPHY ..................................................................................................... 63 1 LIST OF TABLES Table 3.1: 3.2: 4.1: 4.2: 4.3: 4.4: 4.5: 4.6: 4.7: 4.8: 4.9: 4.10: 4.11: 5.1: 5.2: Page Relations between codes ................................................................................ 28 Operations performed on codes...................................................................... 30 Structural features........................................................................................... 32 Writing methods ............................................................................................. 37 Contribution constraints. ................................................................................ 42 Registered members and actual contributors to each project ......................... 44 Frequent contributors...................................................................................... 44 Code 5............................................................................................................. 46 Narrative material and edits............................................................................ 47 Blog, forum and other posts – SIC ............................................................ 48 Blog, forum and other posts – MLO .............................................................. 48 Blog, forum and other posts – AMP............................................................... 48 Web 2.0 technologies ..................................................................................... 49 Thematic analysis codes ................................................................................. 52 The code family .............................................................................................. 53 2 ABSTRACT In recent years, collaborative writing on the Internet has been the object of much scholarly attention. However, to date, there have been no specific studies of collaborative narrative writing on the Internet. The present research aims to fill this gap by studying three different platforms of collaborative narrative writing in the context of both Web 2.0 and new participatory trends fostered by user-generated content and social networking sites. In order to lay out the framework of the analysis, the characteristics of both Web 2.0 and collaborative writing are scrutinised and defined: the former in the context of the lively debate on participatory culture and the latter, first in the light of the notions of authorship and texts, then according to the distinction between narrative and non-narrative writing. Next, this work focuses on the convergence between collaborative narrative writing and Web 2.0, represented by different Web platforms for collaborative writing, which are analysed from both structural and thematic perspectives, showing how their overall architectures influence users’ participation. Findings of structural and thematic analysis reveal that the setting of both a fixed writing method and constraints on contributions is essential to the achievement of the platforms’ objectives. Moreover, an important role is played by discussions between co-authoring users by means of tools such as blogs provided by Web 2.0 applications. Such tools greatly increase users’ engagement in the common project. The findings of the structural and thematic analysis carried out in the present work illuminate possible new directions of enquiry. Specifically, it would be useful to consider the intrinsic aesthetic and literary value of collaborative narrative work, with a view to understanding how this is affected by both the medium (Web 2.0) and the specific modalities of creation (collaborative writing). 3 DECLARATION No portion of the work referred to in the dissertation has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning. COPYRIGHT STATEMENT i. Copyright in text of this dissertation rests with the author. Copies (by any process) either in full, or of extracts, may be made only in accordance with instructions given by the author. Details may be obtained from the appropriate Graduate Office. This page must form part of any such copies made. Further copies (by any process) of copies made in accordance with such instructions may not be made without the permission (in writing) of the author. ii. The ownership of any intellectual property rights which may be described in this dissertation is vested in the University of Manchester, subject to any prior agreement to the contrary, and may not be made available for use by third parties without the written permission of the University, which will prescribe the terms and conditions of any such agreement. iii. Further information on the conditions under which disclosures and exploitation may take place is available from the Head of the School of Education. 4 1. INTRODUCTION 1.1 Thematic Introduction In 2006, some American magazines dedicated their cover pages to a very peculiar subject: the User. The following are some of the headlines: ‘The Power of Us’, Business Week, June 20, 2006; “Putting the ‘We’ in the Web”, Newsweek, April 3, 2006; “Time’s Person of the Year: You”, Time, December 13, 2006. What were these cover pages about? At that time it was not clear to what they were referring, but nowadays it is immediately obvious: the advent of Web 2.0 with its famous and controversial emanations, user-generated content (UGC) and social network sites (SNSs). The context of the present work is the Web 2.0 galaxy, more precisely collaborative narrative writing on the Internet. I shall present and discuss three different experiences of collaborative narrative writing, paying particular attention to their structural and organisational features. To this end I shall clarify a number of general concepts which constitute the framework of the present research. I shall first define the notion of Web 2.0, in its origins and its elements of continuity and discontinuity with Web 1.0. Secondly, I shall introduce the general concept of collaborative writing, as well as the two notions of authorship and text. This will allow me to discuss the issues arising from the intersection between Web 2.0 and collaborative writing, focusing my attention on narrative (as opposed to nonnarrative) writing. 1.1.1. Understanding Web 2.0 At first glance, the term ‘Web 2.0’ may suggest the idea of ‘a new and improved internet network operating on a separate backbone’, but a closer examination reveals that this is not the case, as stated by Web researchers Mary Madden and Susannah Fox (2006, p. 6). Rather, Web 2.0 should be seen as a ‘conceptual umbrella’ accounting for the recent diffusion of several Internet applications and tools, which enable and foster wider interactivity and user participation in the creation of Web content. 5 According to Wikipedia, the first appearance of the term can be traced back to an article by the web designer Darcy DiNucci, where the Internet of the future is defined ‘not as screenfuls of text and graphics but as a transport mechanism, the ether through which interactivity happens’ (DiNucci, 1999, p. 89). At any rate, its large scale use began after the 2004 Web 2.0 Conference held by O’Reilly Media, where its characteristics were defined as affecting users’ behaviour rather than in technical terms.1 In fact, according to Tim O’Reilly, one of the most cited technology gurus, the novelty of Web 2.0 consists mainly in the following five elements: 1) using the Web as a platform, 2) harnessing collective intelligence, 3) individual data management, 4) releasing software as services, not as products, 5) lightweight programming models. All of these features, as they converge, have led towards the facilitation of communication, information sharing, interoperability and collaboration between users (O’Reilly, 2005). The best known example of a Web 2.0 application is Wikipedia, the Internet encyclopaedia that can be edited by anyone and which aims to provide free encyclopaedic information to its readers.2 Since its launch,3 Wikipedia has not only reached increasing numbers of readers and users, amounting to about 10 million to date,4 but also given birth to the ‘wiki phenomenon’. This expression alludes to the rapidly increasing number of websites relying on the wiki technology, whose content is referred to as UGC. It should be borne in mind that UGC refers not only to wikis; actually the Web 2.0 galaxy contains a myriad of websites and applications whose main activity is hosting users’ materials (written texts, images and music), as well as offering tools to manipulate them. To have an idea of the vastness of this galaxy, one need only think of the Google constellation, which includes Gmail, Google Document, YouTube, Google Video and so on. 1 Nevertheless, a variety of technical standards and innovations can be said to have fostered the success of the Web 2.0 idea: the main ones are the standardisation of XML, a platform-independent language, and the diffusion of JavaScript and Ajax, respectively allowing interactivity and background updating in web pages. 2 3 4 http://en.wikipedia.org/wiki/History_of_Wikipedia Jimmy Wales and Larry Sanger launched Wikipedia on 15 January 2001. Data related to the English version. 6 But UGC is just one of the elements combining to define Web 2.0; the second, as mentioned above, is SNSs, which are ‘websites with multiple users where one user can publish contents himself/herself and connect with others sharing personal or professional interests’.5 SNS applications are so numerous and varied that it would not be possible to list them all; however, Facebook and Twitter are perhaps currently the most representative, in terms of popularity and as examples of possible development. It is precisely the combination of UGC and SNS features which fostered the participatory approach to the Net; not only can we now interact with the Net, but we can do it collectively, sharing ideas and content amongst users. This convergence has implications in numerous fields, such as connectivism and transmediality. As a theory of learning, connectivism is concerned with the broadening of the concepts of teaching and learning, with predicted great benefits in the near future from these new networking opportunities (Wellburn, in press). But according to its formal proponent, George Siemens, connectivism has implications in many aspects of life; as an example, Siemens cites how it is effecting changes in management and leadership, media, news and information, the design of learning environments and personal knowledge management. Media analyst and MIT professor Henry Jenkins describes the new frontier of the participatory trend as transmediality or ‘transmedial storytelling’ (2006), referring to the transfer and reuse of content from one medium to another, which also implies a convergence of content created by ‘traditional’ producers/authors and by readers/users. To summarise, what is deemed to be the real revolution brought by Web 2.0 is the possibility for users to manage information, to communicate with other users and to create, share and publish, individually but more often collaboratively, every kind of content on the Web platform. The medium has not changed; it is the ways in which we interact with it and amongst users that have changed. 5 http://en.wikipedia.org/wiki/SNS 7 1.1.2.Collaborative writing6 Collaborative writing is often defined as ‘projects where written works are created by multiple people together (collaboratively) rather than individually’.7 Within collaborative writing, the basic distinction is between narrative and non-narrative texts. Non-narrative writing commonly refers to the collective editing of academic work such as research studies, reports, manuals and encyclopaedias: in general, texts having some kind of informative content, not necessarily written simultaneously. The main studies of such activities place stress on the modalities of collaboration, as in the monumental bibliographical work of Speck (1999) and in the taxonomy proposed by Lowry (2003). In particular, Lowry suggests a classification into 1) practices, 2) strategies, 3) activities, 4) content check, 5) roles and 6) organization; within activities, which represent the core of the process of collaborative writing, Lowry further distinguishes between brainstorming, outline, draft, revision and editing. Turning to collaborative writing with a narrative purpose, things become more complicated, if only because the actual act of ‘writing together’ implies the making of compromises between the individual artistic views of the people involved. Moreover, the mere idea of collaborative narrative writing seems to many a sort of oxymoron; in fact, in recent centuries the act of writing has generally been considered an individual act, performed in solitude by a gifted person, the author. Despite the lively debate concerning the ‘author function’ since the 1960s, in modern times collaborative narrative has been practised far less than individual authoring: the Western literary canon comprises single authored works. Thus the question arises as to whether this individualistic inclination rests on peculiar characteristics of narrative writing, or whether it is dependant on historical and cultural contingencies8. Although some authors use the term ‘collective’, I prefer the adjective ‘collaborative’; according to the Oxford Dictionary, the former means ‘done by or belonging to all the members of a group’, whereas the latter refers to working ‘jointly on an activity or project’, where ‘jointly’ indicates more precisely a process of simultaneous co-authorship. Conversely, the notion of ‘collective’ often makes reference to works created by a plurality of people but not necessarily simultaneously, such as edited books or anthologies. 7 http://en.wikipedia.org/wiki/Collaborative_writing 8 6 Some of these will be considered in the literature review. 8 However, the practicalities of both narrative and non-narrative collaborative writing entail the collaborative activities listed by Lowry. Moreover, as pointed out by a professor of new media studies (Rettberg, 2005), both have always also involved a plurality of people working together to edit, design, produce and distribute the artefacts called books; and this is where the participatory features of Web 2.0 can make a difference. 1.1.3 Web 2.0 meets collaborative writing All the above activities have traditionally been carried on at different moments, linearly and hierarchically, due to technical constraints, whereas Web 2.0 tools now allow multiple authors to work simultaneously on the same text in a variety of ways, constructing an entity that can be developed in several directions. But how do these new tools manage to bring together a plurality of individuals coping with specific narrative issues such as individual views on how to develop storylines, characters and situations? We will see that Web 2.0 already provides the tools to make collaborative writing on the Net a reality, through the ability to create content (UGC) and the communicative power of SNSs; in fact, collaborative narrative is being authored on the Net now. But does the mere existence of technical possibility imply the achievement of a goal? In other words, can Web 2.0 and related tools ensure success in a field such as narrative writing, considering all its specific features? All these issues will be taken into account when formulating my research question in the next section. 9 1.2 Problem statement and research question As shown in the previous section, in the context of Web 2.0 studies, scholars deal with collaborative writing on the Net as a medium and as a technology, whereas there is still a scarcity of reports on the use of these media and technologies to create collaborative narrative work. As I shall explain below, I believe that narrative writing projects on the Internet represent per se an interesting topic to explore, allowing me to pose some interesting questions. The first element of interest rests on the peculiarity of this kind of collaborative writing, i.e. its narrative nature. As I will show in the literature review, narrative writing raises inherently difficult issues such as the nature of authorship and text, which become even more complex when applied to collaborative writing. Thus, it will be interesting to see how these issues are dealt with in collaborative writing and how collaboration is practised by means of Web 2.0 tools. The second element of interest concerns the varying success of such collaborations: some attract very many participants and succeed in their purposes, while others strive to find enough users, or lose most of them in a short period. According to recent studies, several factors account for these different outcomes, including the variety of collaborative writing methods (Rettberg, 2005), the organisational structure, technical features and governance policies (Roth, 2007). Depending on the combination of these elements, collaborative narrative writing platforms can result in different architectures, whose complexity goes well beyond the mere sum of their structural features. In this sense, the fact that some platforms rely on wiki technology, allowing users both to submit and to edit all the content, whereas websites do not permit users to edit, is incidental to the different development paths of such platforms. In the light of all the above, my research question is: How do different architectures of collaborative narrative influence user participation? In order to clarify this question, let me now explain the terminology I shall use to articulate the core concepts. By ‘participation’ in general I refer not only to figures, but also to typologies of participants’ activities, as they consist in submitting narrative material or discussing issues related to the projects. To distinguish these two elements within participation I will refer to ‘consistency of contribution’ and ‘engagement’. Specifically, the former alludes to kinds of user contribution, be they 10 text submissions, edits or discussion posts, while engagement refers to users’ commitment to a project, in terms of the regularity and continuity of their participation. Therefore, my aim is to scrutinise the above elements to discover how platform architecture boosts user participation in order to produce collaborative pieces of narrative. 11 2. LITERATURE REVIEW 2.1 The culture of the Internet: from Web 1.0 to Web 2.0 The participatory culture arising from the Web 2.0 through the diffusion of UGC and SNSs is quite a recent phenomenon, but already complex enough to have elicited a wide debate about how it is affecting the social sphere. This debate has become polarised into two opposed stances. On one hand, the possibility for everyone to collaborate, create and publish content on the Net is deemed to have the capability to foster public participation in the media and consequently to change the social, cultural and political landscape. On the other hand, some scholars highlight the limitations of the participatory culture and argue that the processes of differentiation, hierarchization and control which structure offline human interactions have migrated online along with social networks (Cammaerts, 2008; O’Neil, 2005). Before entering the core of this debate, it is useful to set it in the wider context of the literature on Internet development, which can provide a conceptual framework to better understand the more recent issues. Most of the studies I have surveyed consider the following three features to be crucial in understanding the expansion of the Net: the fast pace of its development, the peculiarity of its sovereignty and its nature as a network. These three elements were already present in ‘A Declaration of the Independence of Cyberspace’ (Barlow, 1996), which calls cyberspace ‘the new home of the Mind’, characterised by the free circulation of thought, and argues that it should not be subject to the laws of politics and economics. Although Barlow was writing before the term ‘Web 2.0’ was coined, the way he defines this brave new world anticipates some of the features now attributed to it: ‘We (the users) have no elected government, nor are we likely to have one […]; Cyberspace consists of transactions, relationships, and thought itself, arrayed like a standing wave in the web of our communications. Ours is a world that is both everywhere and nowhere, but it is not where bodies live’ (Barlow, 1996). This emphasis on freedom finds a partial confirmation in Manuel Castells, one of the first Internet analysts, who states that the Internet was purposely designed as a 12 technology of free communication among government institutions, universities and research centres. The culture of its early designers shaped it as a malleable technology, whose openness has been determined especially by a techno-meritocratic culture rooted in academia and science: open source software is the key feature in the development of the Internet (Castells, 2001). The said techno-meritocratic culture is just one of the four components of this early Internet culture described by Castells, together with ‘the hacker culture, the virtual communitarian culture and the entrepreneurial culture’ (2001, p. 37). These elements are not to be seen as acting independently, but as influencing each other and socially determined, as shown also by Social Shape of Technologies studies: ‘Technology does not develop according to an inner technical logic but is instead a social product, patterned by the conditions of its creation and use. Every stage in the generation and implementation of new technologies involves a set of choices between different technical options.’ (Williams, 1996, p. 858) In turn, as argued recently by Joe Karaganis, ‘social construction feeds back into processes of technical innovation, shaping research priorities and design. In the end there is no simple causality: no chickens, no eggs’ (2008, p. 6). According to this developmental paradigm, technological determinism cannot be invoked to explain the shift from Web 1.0 to Web 2.0: new technical instruments and new patterns of usage have to be considered as shaping each other, in a circular process. The mutual shaping process between technology and society has been analysed amongst others by Thurlow, Lengel and Tomic, who define online interaction as ‘a struggle between invention and appropriation’ (2004, p. 27). Producers develop and publish an application carefully designed for a specific purpose, but users will often exploit it in a different manner, according to their needs and circumstances. An example of this ‘struggle’ is provided by the case of the social network Friendster, whose participants subverted the rules, creating an intra-network of fake characters (Fakesters), eventually suppressed by the network’s owners (Boyd, 2008). In this view, the recent participatory development of the Internet known as Web 2.0 can be seen as a product of the original openness of software sources: the 13 diffusion of UGC represents for most scholars a further and natural expansion of its original nature as a network. The future development of the Web is the object of a complex debate dealing with a plurality of issues, which could be grouped into three main dimensions: 1) technical/ structural,9 2) social/cultural, 3) legal/political/economic. These issues have generated two opposite views, the first mostly positive and the second quite critical. Let us see how these two schools of thought deal with the three dimensions of debate. On the technical/structural and social/cultural dimensions, enthusiasts of Web 2.0 claim as its main benefit that ‘it allows more people to share ideas with more people in more ways’ (Leadbeater, 2008, p. 3). Rather than being provided with packaged contents by a few professional experts and creators, the participatory Web offers everybody the opportunity to be involved in the cultural process, communicating with experts and each other, and proposing ideas. ‘The web’s underlying culture of sharing, decentralization and democracy makes it an ideal platform for groups to self-organize, combining their ideas and know-how, to create together games, encyclopedias, software, social networks, video sharing sites or entire parallel universes’ (Leadbeater, 2008, p. 10). This positive view is shared by Clay Shirky, who lists many examples of collaborative and interactive activities on the Net, regarding the so-called ‘culture of the amateur’ as a positive element. He also coined the phrase ‘publish then filter’ to explain this shift in the cultural scenario, questioning the traditional notion of expertise characterized by the opposite approach: ‘filter then publish’. In this view, knowledge is no longer created only by the few and distributed to the many, but is created by many single intelligences which multiply their power by combining (see also Downes, 2007; Jenkins, 2006; Levy, 1997). As for the critical approach to the first two lines, Andrew Keen, entrepreneur and author, deplores the loss of cultural authority on the part of professionals because of amateurs’ contributions, following the logic of ‘publish then filter’. In his view, this trend will undermine the foundations of the pre-existing cultural framework: ‘What the Web 2.0 gives us is an infinitely fragmented culture in which we are 9 Specific technical aspects of the Web will not be dealt with in the present study, with the exception of some occasional general hints. 14 hopelessly lost as to how to focus our attention and spend our limited time’ (Keen, 2008, p. 60). The varied and numerous issues linked to the filtering of information have been recently termed ‘information obesity’ by Whitworth (2009), who defines this as ‘a failure to turn information into knowledge, and thus use it to sustain our minds, bodies, lives and communities’. According to Whitworth, educational and political solutions would entail a critical approach, articulated in a radical rethinking of ICT skills education, information literacy and the politics of organizations. In Keen’s opinion, the popularity of tools such as Google, YouTube and blogs is to be blamed, because ‘rather than using the Net to seek the news, we use it to BE the news’ (Keen, 2008, p. 7). In particular, Keen criticizes the phenomenon of ‘Google bombing’, which ranks websites according to numbers of visitors and links, claiming that in this way we just perpetuate one another’s biases. As for the third dimension, the main issue seems to be whether Web 2.0 is destined to foster freedom of expression, or whether it will provide a more subtle form of control over citizens. As a champion of the positive standpoint, the above-cited Charles Leadbeater, author and advisor of the former British Prime Minister Tony Blair, claims that ‘the culture of sharing also makes the web difficult for governments to control and hard for corporations to make money from’ (Leadbeater, 2008, p. 15). But these ideas have been opposed by scholars such as Andrew Shapiro (1999) and Lawrence Lessig (1999), who were among the early supporters of the critical approach. Lessig warns that ‘there is no reason to believe that the grounding for liberty in cyberspace will simply emerge’: the argument here is that ‘left to itself, cyberspace will become a perfect tool of control’ (Lessig, 1999, p. 6). Other scholars share this fear, acknowledging the relevance of monitoring and studying the potentialities of these new social patterns, also in order to prevent their appropriation by political and economic power systems, which could blunt it (Cammaerts, 2008; O’Neil, 2005). Echoing these words, Lessig argues that this control does not necessarily refer to governments or to ‘some evil, fascist end’: the setting he and other scholars depict is not that of an Orwellian Big Brother. Rather, 15 he suggests that ‘the invisible hand of cyberspace, through commerce, is building an architecture that perfects control’.10 The above three groups of issues merge in another much debated question: the inequality of participation, not to be confused with the issue of the digital divide, which is actually just one aspect of it. In fact, if it is certainly undeniable that the lack of infrastructure automatically excludes people from access to the Internet and consequently from any involvement in the digital society, it is equally true that the mere availability of an Internet connection and basic literacy do not automatically grant real participation. The results of an education project sponsored in the UK by New Opportunities Funding contests the general assumption that ‘simply by introducing technology to disadvantaged communities, the digital divide will be removed and people will go online’ (Keeble, 2003). On the same topic, Hargittai (2007) argues that ‘when specific site usage is considered, statistically significant relationships emerge between race and ethnicity and social network sites uses, in addition to the predictive power of parental education’. Support for the above arguments can be found in a 2006 study by Jacob Nielsen, a scholar of Web issues, revealing that in most online communities 90% of users are ‘lurkers’ who never contribute, 9% of users contribute a little, and 1% of users account for almost all the contribution (Nielsen, 2006). Data like this demonstrate that participation is quite a faceted issue and warn against merely mechanical explanations. 2.2 Changes in patterns of participation: The expansion of SNS activity and transmediality 10 Lessig first published Code and Other Laws of Cyberspace in 1999. After five years in print and five years of changes in law, technology and the context in which they reside, Code needed an update. To begin it, in 2005, Jotspot ported Code v1 to a wiki and for the rest of the year people edited Code v1 by adding ideas, questions, or new material. In January 2006, Lessig took the product of that wikiedit and edited it to produce Code v.2. 16 Whatever the approach, enthusiastic or critical, the key words for understanding Web 2.0 seem to be participation, community, connectivity, networks, sharing and collaboration. While each of these terms bears a different nuance of meaning, they all make reference to the same basic concept, ‘the participatory Web’, which has emerged as ‘a governing concept, albeit one surrounded by conflicting expectations’ (Jenkins, 2006, p. 175). Currently, the most significant manifestation of this participatory trend is represented by SNSs, which can be considered a development of the online communities born during the 1990s. A famous definition by Howard Rheingold, who was involved for seven years in an early community developed in the San Francisco Bay area, states that ‘We do everything people do when people get together, but we do it with words on computer screens, leaving our bodies behind’ (Rheingold, 1994). Amongst the many definitions of online community, I would make reference to the one proposed by Jenny Preece, purposely broad in scope: ‘An online community consists of: People, who interact socially as they strive to satisfy their own needs or perform special roles, such as leading or moderating A shared purpose, such an interest, need, information exchange or service that provides a reason for the community Policies, in the form of tacit assumptions, rituals protocols, rules, and laws that guide people’s interactions Computer systems, to support and mediate social interaction and facilitate a sense of togetherness’ (Preece and Maloney-Krichmar, 2005). Although this definition was elaborated in 2005, its components are still crucial when talking about participatory activities typical of Web 2.0 such as social networks, where people interact with a purpose, according to rituals and rules, by means of computer systems. The Web 2.0 element of novelty rests on the merging of different communitarian activities and purposes, by means of new technical possibilities. In fact, if it is true, as pointed out by Madden and Fox (2006, p. 6), that ‘the beating heart of the internet has always been its ability to leverage our social connections’, it is true as well that keeping in contact with old friends and finding new ones does not exhaust the range of social networking activities. SNSs are often 17 alleged to be superficial and a waste of time, but on closer inspection they are seen to be increasingly used to perform artistic activities, such as taking ‘material found online—like songs, text or images—and remix[ing] it into your own artistic creation’ (ibid). Such activities—and many others which involve merging different media— entail the mastering of what Sue Thomas (2007) has termed ‘transliteracy’, i.e. ‘the ability to read, write and interact across a range of platforms, tools and media from signing and orality through handwriting, print, TV, radio and film, to digital social networks’. According to Thomas, transliteracy refers to the literacy of the transmedial process, which is specifically examined by Henry Jenkins in Convergence Culture (2006). This book explores phenomena such as the so-called ‘wisdom of crowds’ or collective intelligence, which emerge from communities of fans collaborating to anticipate developments in the storylines of their favourite series. A transmedial product such as The Matrix has been conceived not as a film trilogy but as a complex system of games, comics and other multimedia materials, most of them produced by fans and users, all feeding the intricate storyline (ibid). Jenkins goes further, showing how these new transliteracies that we are learning by means of apparently ‘futile’ activities can be transferred to other fields, such as political activity (ibid). To sum up, these recent studies state that we are in the presence of a radical change in users’ behaviour: more precisely, what Jenkins and other media scholars argue is that these new Web 2.0 instruments are progressively blurring the line between producers and users, offering users the means to create content, rather than just consuming it. However, as seen in the previous section, the implications of this ongoing cultural shift remain to be fully explored. 2.3 The convergence between UGC and SNS: Collaborative narrative The convergence between SNS and UGC has fostered more and more types of collaborative activity on the Net. Within these activities, collaborative narrative writing poses some specific issues. One of the main questions to be considered is the relation between collaborative narrative and the notions of authorship and text. 18 These two concepts have changed throughout history and have shaped each other according to technological, historical, literary and cultural contingences; Michel Foucault remarks that ‘there was a time when the texts we today call “literary” were accepted, put into circulation and valorised without any question about the identity of their author’ (1994, p. 245). Immense corpora of founding myths, legends and religious texts such as the Bible, the Iliad and the Odyssey were told orally and passed from generation to generation in anonymity, before the concept of the written fixed text arose. Naturally, this is not to say that the concept of author did not exist in ancient times, but it was certainly a more elusive notion, due to the flexible nature of oral literature. Nevertheless, ancient literature cannot be defined as collaboratively authored as we understand it nowadays; myths and legend were generated by a plurality of people, but this does not imply simultaneous cooperation in the creative process, which rather consisted in an endless process of addition of variants to the corpus of tradition, linking one text to another in order to form an uninterrupted sequence. Such literature would be better termed collective than collaborative. The passage from oral to written literature marks another milestone in the emergence of the notion of authorship. It is important to stress the mutual shaping of these two elements: the fixity of the text is an essential condition for the establishment of the notion of author. According to Marx, the invention of printing and the consequent establishment of copyright11 gave birth to the modern author and reader (Marx, 1866, cited by Hadjiafxendi, 2007, p. 2). A further change took place in the late eighteenth century, when the figure of the Romantic artist provided a stable affirmation of the notion of literary authorship, reinforced by the image of the solitary genius expressing his individuality. This idea remained fixed with no ontological change until the late 1960s, when the debate about the notion of text and authorship flourished within structuralism and deconstructionism. Roland Barthes’ The Death of the Author (1968) and Michel Foucault’s What is an author? (1969) ‘…initiated the movement against the author through a theory of authorial absence’, and ‘claimed that the text 11 Copyright or ‘creative commons’, or t is now a much debated issue: in the spirit of freedom which marks Web 2.0, many authors choose to release their work as ‘copyleft’ o put it in the public domain (Lessig, 2007). 19 could no longer be considered as an expression of an author’s unique personality’ (Hadjiafxendi, 2007, p. 3). In the light of the above, we could say that criticism and literary theory prepared the cultural ground for the participatory developments we are witnessing today; once the uniqueness of the author and the fixity of the text were questioned, new and different ways of authoring a text became possible. As an example, I would cite the work of the Italian collective of five novelists Wu Ming, who in the last decade have published several collaboratively written novels, including Q and 54.12 A review of a Wu Ming book in the Independent recalls the themes dealt with in the present section, claiming that ‘In contrast with those Structuralist and Marxist academics who have produced reams of speculative theories on the death of the solitary auteur, Wu Ming has acted’.13 But Wu Ming represents an exception in a landscape still dominated by individual authors: in point of fact, notwithstanding the abovementioned shift in critical paradigms, we had to wait for the expansion of the electronic media before seeing changes in authorial practice. The first modern experiments were based on the technology of the hypertext, conceptualised in 1965 by Ted Nelson as ‘a system of interconnected writings’ (Nelson, 2003, p. 450), not based on a predetermined hierarchy of contents, but working ‘as a collective pool of knowledge, which we can access, view, and reorganize in a variety of ways’ (Rettberg, 2005, p. 2). Some examples of these early literary hypertexts are novels such as Michael Joyce’s Afternoon. A story (1990) and Hypertext Hotel, launched by Robert Coover (19931996). In 2002, Wu Ming launched a collaborative project on its website called ‘I Shall Call You Russell’: the collective wrote the first chapter of a sci-fi novel and invited anyone who wished to contribute to write subsequent ones and submit them by email. Interestingly, some participants in the ‘Russell experiment’ are continuing the experience independently in a workshop on the Net: they have created another 12 13 http://www.wumingfoundation.com/english/biography.html The Independent, 11 July 2005. 20 collective of novelists, Kai Zen, and have already published books, some of them born of collaboration with other writers.14 In 2005, the Web started to offer more participatory tools and Scott Rettberg noted that, ‘As the web moves towards systems of organising material based on tags and collectively defined “folksonomies”, as access to bandwidth expands, as more different types of media sharing become available and web applications become more powerful, collectively15 written constructive hypertext has become everyday practise on the web’ (2005, p. 8). Beside hypertext, a variety of applications and platforms are now available for such projects, including wikis, websites and blogs: Henry Jenkins describes how thirteenyear-old Heather Lawver launched a Web-based newspaper which is, in the words of its young creator, ‘…an organisation dedicated to bringing the world of literature to life… By creating an online “newspaper” with articles that lead the readers to believe this fanciful world of Harry Potter to be real, this opens the mind to exploring books, diving into the characters, and analysing great literature’ (2006, p. 179). The publication has a staff of 102 children all over the world and a large number of contributors, who create profiles (usually presenting themselves as relatives of Potter characters) and subsequently report fictional ‘events’ at Hogwarts School (Jenkins, 2006, p. 178). Jenkins lists the skills needed by these children to become full participants in this experience: they must be able to pool knowledge with others in a collaborative enterprise, share and compare value systems by evaluating ethical dramas, make connections across scattered pieces of information, express interpretations and feelings, circulate what they create via the Internet, and last but not least, engage in role-playing, as a means both of exploring a fictional realm and of developing a richer understanding of themselves and the culture around them. This, concludes Jenkins (2006, p. 185), ‘is a kind of intellectual mastery that comes only through active participation’. 14 http://henryjenkins.org/2006/10/how_slapshot_inspired_a_cultur_1.html. Kai Zen’s novel: The Rettberg uses ‘collaborative’ and ‘collective’ interchangeably. Powerness of Eymerich, Bacchilega, Imola (Bologna), 2005. 15 21 Although the number and variety of collaborative narrative experiences is increasing, to date the attention of scholars has been almost entirely turned towards non-narrative collaborative writing. The most analysed and discussed collaborative platform is still Wikipedia (see Bryant, 2005; Giles, 2005; Viegas, 2007), although scholars are extending their interest to wikis in general, exploring their features and dynamics. Recently, Roth (2007) has studied the factors and features that influence life and death of wikis, finding a strict correlation between population dynamics and content dynamics by analysing functional, organisational and demographic factors. More recently, Bonk et al. (2009) have reported on a project where students in higher education create educational wikibooks; this study concentrates on issues of instruction, collaboration, technology, constructivism and sense of community, as well as on wikibooks themselves. Even scholars of collaborative narrative share this emphasis on wikis; concluding his 2005 study, Rettberg imagines ‘… a writing community with the robustness of Wikipedia, dedicated to a collective vision of writing a novel […] written according to sets of specific constraints to ensure a degree of formal unity, and tagged with metadata that would make it possible to easily remix novels in thousands of structured configurations. Such a project would be performance, game and literature. What we do today with our collective references and photographs we could soon do, together, in collective literature. It is well within our reach’ (2005, p. 9).. This overview shows that the world of collaborative narrative is expanding, increasingly triggering questions which need to be studied from a wider perspective to include not only technological aspects but also narrative specificities. 2.4 Summary of the literature review Considering the breadth of issues encountered within this literature review, I consider it useful to summarise the main ones. From section 2.1, it is clear that the diffusion of user-generated content represents a development of the original open-network nature of the Web, originating in the culture of its early creators and shaped by social changes. Consequences of this original openness, i.e. UGC and SNSs, have prompted a wide debate and opposing views about the necessity of some form of regulated governance. 22 Section 2.2 offered an insight into the development of Web 2.0 SNSs, as they are rapidly expanding their sphere of application and merging different activities, media and languages. Transmediality seems to be the new keyword in this emerging scenario. The final section noted that within the new participatory and transmedial activities, collaborative writing poses specific issues, especially when producing narrative texts. Considering these issues, can collaborative narrative writing be successfully performed? Is the mere existence of the appropriate Web tools a sufficient condition for its successful realisation? 23 3. METHODOLOGY AND METHODS This chapter describes and justifies the methodology chosen to conduct the research. This methodology involves multiple steps, beginning with the literature review, which influenced the selection of the methods of analysis and the choice of data. I shall set out the methodology as follows: section 3.1 is concerned with a contextualisation of my research question, connecting it with the choice of methods and data; section 3.2 provides details of the methods used to perform the actual analysis of the selected data; and section 3.3 puts these methods in a broader methodological frame of reference. 3.1 Contextualising the research question: Methods and choice of data As anticipated in the problem statement (section 1.2), the aim of the present work is to answer to the following research question: How do different architectures of collaborative narrative influence users’ participation? I have previously defined each element of this question as follows: by ‘architectures’ I mean the ensemble of means and practices used by Internet platforms to perform collaborative writing, resulting from different combinations of organisational structures, technical features and governance policies. As for users’ participation, I distinguish two characteristics: consistency, i.e. types of contribution, and engagement, i.e. regularity and continuity of contribution. I shall now contextualise this question in the light of the literature review, also describing the path that led me to the choice of methods and data. As platform architecture is the first element at issue here, I decided to start with an examination of the structural components of the platforms, applying to my data the analytical model proposed by Camille Roth (see section 2.2), which scrutinises a selection of wikis according to their structural features, as specifically detailed in the next section. However, the literature review made clear that the narrative aspect of my object of study also requires me to consider writing specificities, whereas Roth deals with the structure of platforms regardless of their scope. Secondly, platform structures represent only one element of my research; in fact, what I aim to discover is how the overall architectures affect users’ participation, identifiable as consistency 24 of contribution and engagement. Roth’s model does not account for the actual processes at work; indeed, Roth’s research focuses on platform structures ‘at some static timepoint’ (2007, p. 124), whereas I am interested in the process of the realisation of narrative projects. Thus, it was necessary to integrate Roth’s structural analysis with a qualitative analytical approach, suitable to capture the dynamics of participation and to describe the ‘development paths’ of the platforms (ibid., p. 124). The chosen qualitative method was thematic analysis, which is based on the creation of categories or ‘codes’ from the actual data, allowing the grouping together of all the data which relate to a particular theme or process. As for the choice of data, my analytical strategy involves the scrutiny of different kinds of platform, to illustrate how different architectures determine different ways for users to contribute. Considering also the dynamic dimensions inherent in my analysis, I have chosen three different kinds of platform, at different stages of development: the first has already accomplished its project, whereas in the second and third work is still in progress, although they display many dissimilarities as to development modalities. The detailed selection criteria are these:  Narrative nature of the writing projects  Different types of platform (one wiki, two websites)  Presence of Web 2.0 technologies  Minimum number of registered users (at least 100)  Similar launch time (platforms relying on comparable technological capabilities)  Different time points of development  Variety of origins of the platforms (different countries). The following are short presentations of the platforms—A Million Penguins, SIC and Make Literature Online—which I selected after a survey of the Net. 1) In 2007 Penguin Books and De Montfort University (Leicester, UK) launched the project of a collaborative novel on the Net, called ‘A Million Penguins’ and open to everyone. According to the research report by Bruce Mason and Sue Thomas from De Montfort University, the presence of a famous publisher in the background is an important element. The choice of the Wikimedia platform was determined by the 25 popularity of Wikipedia, although the result, according to Mason and Thomas (2008, p. 2), was ‘neither a wiki nor a novel as the terms are commonly understood’. The experiment has been concluded and is available at: http://www.amillionpenguins.com/wiki/index.php/Main_Page. 2) Scrittura Industriale Collettiva (Industrial collective writing; SIC) was launched in 2007 by two Italian novelists, Gregorio Magini and Vanni Santoni. The name refers both to the platform and to the writing method, which constitutes the basis of the work performed on the platform,16 a website. The SIC writing method has been applied to five published novels, available at: http://www.scritturacollettiva.org/biblioteca. Presently the platform is running a project called Grande Romanzo Aperto (Big Open Novel; GRA), which started in February 2009 and is due to finish in March 2010: http://www.scritturacollettiva.org. 3) Make Literature Online runs on a website opened in 2007. The site is owned and operated by Mina System d.o.o, which set up the platform as a business enterprise.17 Nevertheless, the website is free and open to all users to create literature collaboratively. Like SIC, it presents a particular writing method. There are several novel-writing projects belonging to several subgenres running at the same time, as well as a collective poetry project. Amongst these, I have chosen to focus on the collective novel People, the project which has the most visibility on the platform. The URL is: http://www.makeliterature.com/. With regard to the three projects, it has to be said that it is not possible to provide hints about the content of the Million Penguins novel, because several storylines are included. The SIC novel is set in various locations in Italy during the Second World War, and has three main characters plus various supporting ones. As for Make Literature Online, the story is set in the UK and concerns a group of fourteen main characters, struggling with the problems of everyday life today. 16 The SIC method is discussed by Gregorio Magini (2007/2008) in a thesis on theories and techniques of mass communication, within a course in Media and Journalism at the Faculty of Political Sciences of the University of Florence: http://www.scritturacollettiva.org/blog/tesi-la-scrittura-industriale-collettiva 17 http://www.makeliterature.com/system/about-us 26 3.2 Presentation of methods I will first place the platforms in the analytical categories used by Camille Roth in ‘Viable Wikis’ (2007, p. 119), namely: 1) platform, 2) licence, 3) scope, 4) language, 5) users, 6) project pages, 7) date, 8) governance. Although Roth’s analysis is based on wiki platforms, I think her categories are broad and general enough to be applied to other collaborative writing platforms, as all of them include the same basic building elements listed above. Roth’s categories are designed to explore the mechanisms determining the ‘viability’ of wikis, which refers to the ‘dynamic stability of both population and quality content (Roth, 2007, p. 119). Findings are obtained through both a quantitative and a qualitative scrutiny of collaborative platforms, delineating some of the most salient patterns modelling wiki platforms. I will use these categories to describe the platforms’ macroscopic structure, in order to identify issues in need of in-depth examination. The results of this first scrutiny will be termed ‘structural features’, to distinguish them from the features resulting from the thematic analysis. As seen above, the narrative specificity of my research question requires the platforms to be submitted to further analysis, leaving aside Roth’s structural categories. Since I am interested in capturing data from specific processes at work on the three selected platforms, I have opted for an inductive approach, which allows the features of writing and participation to emerge, without the imposition of predetermined labels. Therefore, I will submit the platforms to a thematic analysis, focusing on the three single projects mentioned above. I call the codes obtained ‘writing features’, since unlike structural ones, they can account for specific collaborative writing activities. I will group together different instances of data under an umbrella term that will allow them to be regarded as ‘of the same type’ (Andretta, 2000-2009). As prescribed by the method, I will start with a focussed ‘reading’ to find ‘paths’ through the information, reading for themes and sub-themes and searching for recurring elements. In concrete terms, it will be necessary to ‘break up’ data and rebuild them into thematic categories, which will emerge from the actual data. 27 Among the different mechanisms of code creation, I have opted for ‘relational coding’, which implies the relating of those categories to each other. This process instantly draws connections between the features of platforms, which will be discussed more widely and systematically in section 5.1 (Findings). Thus, the analysis section should be read not only as a presentation of data, but as a first map of links between features providing and initial and partial perspective on the issues being studied in this work. Table 3.1 is a list of relations between codes of which I made use (Andretta, 2000-2009), showing examples of ways in which my categories are related. Table 3.1: Relations between codes Cause – Code A causes Code B Property – Code A is a property of Code B Aspect – Code A is an aspect of Code B Association – Code A is associated with Code B Result – Code A results from Code B Contrast – Code A contrasts with Code B As Andretta points out, these are just some of the possible relations between codes; virtually any relational noun or verb could express such a relation. These relations show how the two attributes of participation, i.e. users’ contribution and engagement, are determined by the platform architecture. In the next section, I shall put thematic analysis in the broader context of Grounded Theory, which ‘clearly lays out a framework for carrying out this type of code-related analysis’ (Andretta, 2000-2009). 3.3 Thematic analysis and Grounded Theory Thematic analysis is a qualitative research approach; more precisely, it is ‘part of the early procedures in data analysis in Grounded Theory’ (Ezzy, 2002, p. 87). Therefore, I will provide an overview of the methodology and procedures of Grounded theory, signalling dissimilarities between the two methods as appropriate. 28 The rationale of Grounded Theory is based on the discovery of theory from data to avoid—or at least minimise—the influence of presumptions and preconceptions; the same applies to thematic analysis. When Barney Glaser and Anselm Strauss presented their theory, they claimed that methods of social research had ‘focused mainly on how to verify theories. This suggest an overemphasis [...] on the verification of theory, and a resultant deemphasis on the prior step of discovering what concepts and hypotheses are relevant for the area that one wishes to research’ (Glaser and Strauss, 1967, p. 3). Thus, their main purpose was to find a method to move from data to theory, and not vice versa: moreover, they also intended to provide qualitative analysis with a rigorous procedure of verification, strong enough to compete with quantitative analysis methods in matters of validity, trustworthiness and generalisation. Coding is almost universally considered the most fundamental process in Grounded Theory and the main link between collecting data and developing an emergent theory (Willig, 2001; Charmaz, 2006; Gibbs, 2007. Since even in thematic analysis it represents the core procedure, I will describe it in detail, marking the difference between the two methods. The process of coding consists in gradually identifying categories within the data, first at a descriptive level (open coding or initial coding), then at a more interpretive and abstract level (focused coding, according to Charmaz, 2006, p. 21). Thirdly, categories are grouped in meaningful units, which should be labelled ‘in vivo’18 in order to avoid importing existing theory into the analysis process (Willig, 2001). Charmaz defines this latter step as ‘theoretical coding’, with reference to Glaser (1998). In thematic analysis all these coding steps are not necessarily undertaken: coding is usually divided into open and relational, as described above (Andretta, 2000-2009). In developing codes, researchers usually perform the operations listed in Table 3.2. As coding is the core of the justification of Grounded Theory, it is not surprising that it has given rise to different theoretical interpretations and practices, which I shall outline here. Strauss and Corbin (1990) present a new formulation of Grounded Theory, whose main feature is a more deductive approach to data analysis, through a different coding paradigm focused on manifestations of ‘process’ and 18 I.e. utilising the actual words used in samples. 29 ‘change’ in data. Strauss and Corbin term this process ‘axial coding’, since it builds relationships between categories and subcategories around their ‘axes’, to bring data back together after the initial coding, which is, by contrast, a process of separation and distinction. Table 3.2: Operations performed on codes Combining or ‘merging’ Codes – Everything that had been coded as ‘A’ or ‘B’ now becomes ‘A’ Splitting a Code into different parts – Code ‘A’ becomes ‘A1’ and ‘A2’ Creating a new code (sometimes called a ‘super code’) from two existing codes – anything that has been coded as ‘A’ AND ‘B’ is coded as ‘C’, while ‘A’ and ‘B’ are maintained as distinctive codes. Creating a Code Family – Placing a group of codes into a code grouping (without actually using that grouping as a new code, as with super codes). However, well before Strauss and Corbin, Glaser had tempered his pure inductive approach, introducing ‘theoretical coding’ (Glaser, 1978). This consists in 18 theoretical coding families or ‘possible relationships between categories developed in focused coding’ (Charmaz, 2006, p. 16), providing a sort of regathering frame. Whatever version of Grounded Theory is considered, coding represents one of its three main building blocks (Willig, 2001, p. 36), the other two being constant comparative analysis and theoretical sampling. Here lies another distinction between the two methods. I shall first describe the procedures of Grounded Theory, then indicate its dissimilarities to thematic analysis. Constant comparative analysis should ensure that the complexity and diversity of the data could be recognised, avoiding the impulse to homogenise them. Theoretical sampling consists in continuing to collect data according to the progressive emergence of new characteristics; this should lead to the continual elaboration and refinement of the categories, developing their properties until new ones emerge. In thematic analysis, constant comparative analysis is performed, whereas theoretical sampling is not, because thematic analysis is conducted on data already entirely collected, removing the need for further data collection. In fact, the final purpose of theoretical sampling is ‘theoretical saturation’, when no further properties emerge from the data and there are no more data to 30 collect. However, even in Grounded Theory, saturation ‘functions more as a goal than a reality’ (Willig, 2001, p. 46), as it is often unfeasible for the researcher to continue collecting new data, even though Glaser and Strauss recommended considering the published work ‘not the final one, but only a pause in the neverending process of generating theory’(Glaser & Strauss, 1999, p. 198). Amongst the main features common to Grounded Theory and thematic analysis is memo-writing. Glaser defines memos as ‘the theorising write-up of ideas about codes and their relationships as they strike the analyst while coding... It can be a sentence, a paragraph or a few pages... it exhausts the analyst’s momentary ideation based on data, with perhaps little conceptual elaboration’ (Glaser, 1978, p. 62). Charmaz advises researchers to keep memos from the very first, as they can be very useful during interviews or data collection in general, to crystallise impressions, ideas to develop and so on. They become even more pivotal when the analyst begins to read the data and to conduct the initial coding. If regularly kept, implemented and compared with data, memos can indicate to the researcher a new way to consider his data, a new line to develop, or even a change of direction for the research itself. Hence, the main difference between the two methods may be expressed as follows: Grounded Theory is fully applied when a researcher aims to generate a theory concerning a complex issue which it would be difficult to synthesise into a single research question, since this analytical method implies a continuous repositioning of the purpose of the study itself. Conversely, thematic analysis is better suited to addressing a specific research question, not to be answered by verifying an existing theory, but by drawing a theoretical explanation from the data themselves. 31 4. ANALYSIS OF SELECTED PLATFORMS This chapter presents an analysis of the structural elements of the three selected platforms. Specifically, I shall set out their main structural features and highlight those aspects which deserve in-depth examination. However, I shall not indicate inferences, nor shall I anticipate the connections with the thematic codes arising from the structural features. In fact, implications and connections will be discussed in the next chapter, after the completion of the structural and thematic analyses, because of overall inductive approach of this research, which precludes the drawing of implications before having scrutinised all data. 4.1 Structural analysis of the platforms: Structural features As anticipated in the previous chapter, the first step of the present analysis is to place the three platforms in the analytical categories proposed by Roth (2007, pp. 7-10). To allow an easy overall comparison of the data, I first present them in Table 4.1. Table 4.1: Structural Features Name A Million Penguins (AMP) SIC Make Literature Online (MLO) Licence Penguin Books Ltd. CC22 Platform Scope Topical20 Language Registered users 1476 Project pages19 1031 2200 To date From Governance Mediawiki English 2007 RR21 RR, admins,23 moderated24 RR, admins, moderated Website Website Topical Italian 406 2007 CC Multipurpose25 English 802 ± 10 To date 2007 19 ‘Project pages’ refers to the number of pages of contributions for a single writing project. In the Project devoted to the creation of a single novel. Registration required to contribute. Creative Commons. ‘Admins’ denotes the enforcement of community decisions by administrators. ‘Moderated’ means that submissions will be reviewed before being published. Project devoted to the creation of more than a single narrative work at a time. case of AMP and SIC (but not MLO), this number is the total for the whole platform. 20 21 22 23 24 25 32 As I shall show in detail below, the websites SIC and Make Literature Online (MLO) seem to share many structural features, whereas the wiki A Million Penguins (AMP) differs in a number of respects. 1) Licence SIC and MLO publish their material according to the Creative Commons26 licence, which, as stated on the home page of the website, aims ‘to mark creative work with the freedom the creator wants it to carry, so others can share, remix, use commercially, or any combination thereof’.27 Conversely, all materials on the AMP platform, which was launched by the commercial publishing house Penguin, are subject to copyright. Hence, Penguin is the actual owner of all the material submitted by the contributors and is therefore entitled to exploit it commercially without restriction. The question of licence is dealt with very differently on the three selected platforms. Actually, only in SIC is this question publicly addressed and discussed by users, whereas in AMP and MLO there is no public discussion of it. While this is not unexpected in the case of the commercially run AMP, I find it surprising that such a relevant issue is not addressed at all in MLO, which like SIC has a forum for public discussion. 2) Platform SIC and MLO utilize websites, whereas AMP uses a wiki tool. This marks an important first difference in participation. AMP users are in fact entitled to decide how to contribute in terms of content, organization of the materials, storylines, characters and so on, editing their own and other contributors’ texts. They may submit many pages or just edit the punctuation of other texts. Conversely, SIC and MLO users can only send their contributions according to the conditions established by the websites’ creators. 26 Creative Commons is a nonprofit corporation dedicated to making it easier for people to share and http://creativecommons.org/about/ build upon the work of others, consistent with the rules of copyright. 27 33 3) Scope In Roth’s research, scope distinguishes ‘all-purpose projects, aiming at building broad encyclopedias, [from] topical projects based on narrower matters, e.g. precise technical or geographical areas, particular political flavors’ (2007, p. 2). In the present work, I have kept the same basic distinction. However, since the selected platforms all deal with collaborative narrative, Roth’s distinction does not refer here to broader or narrower subject matters, but rather to the number of projects undertaken. AMP and SIC are dedicated to the accomplishment of a single narrative project, whereas MLO is designed to carry on more than one project at a time, more precisely one for each sub-genre allowed by the project.28 4) Language AMP is a British initiative and not surprisingly has English as its language, as does MLO, which was launched in Serbia, while SIC, launched in Italy, uses Italian. Thus, different necessities and aims motivate the choice of language. Obviously, MLO’s choice of English was intended to allow the involvement of a virtually unlimited number of people well outside Serbia’s borders, whereas by choosing Italian, SIC consciously limited its potential users to Italian speakers. However, this choice finds its rationale in the fact that SIC novels are deeply rooted in the Italian political, historical and social context, thus making it difficult for members of other cultures to participate in the writing process. 5) Users Registered & 6) Project Pages On the dynamics of users and content, Roth states that ‘[wiki] population [roughly] correlates with content size’ (2007, p. 123). In other words, successful wikis29 show a virtuous demographic path, with content and contributions coevolving. But if we look at data for MLO, we see that co-evolution is not happening: the number of users is high, while the number of pages is quite low. This shows a lack of proportionality 28 There are six categories of subgenre: Science Fiction and Fantasy, Mystery and Horror, Crime and According to Roth (2007,p. 1), ‘successful’ equates to viable, i.e. a wiki that can survive large Thriller, Romance and Comedy, History and Adventure, Children’s and Social. 29 variations in the stock of users and articles, maintaining a regular activity. 34 when compared to the other two platforms. Thus, quoting Roth’s terminology, MLO should be classified as a ‘not viable wiki’, destined to failure. But since I am interested in studying single selected projects for each platform, I have quoted the actual number of pages submitted by users for each novel considered here. Conversely, the term ‘users’ refers to the totality of registered members of the platforms, as in Roth’s table. The introduction of this variant obviously limits the applicability to my data of Roth’s findings. More precisely, these are applicable to AMP and SIC, where both users and pages refer to the whole platform, since the number of platform pages is the same as that of project pages. Therefore, in the case of MLO, what can be defined as ‘not viable’ is only the project, not the whole platform. However, another factor of a different order could be taken to account for the scarcity of project pages, i.e. the multipurpose scope of the MLO platform, which may tend to disperse its users amongst the numerous different novel projects running at the same time. There is another interesting aspect of page numbers. From the perspective of a qualitative research, the total number of pages is just a rough datum which needs to be further scrutinized, since it does not account for elements such as the content of pages. For example, in wikis like AMP, many pages are automatically generated by the platform, containing ‘information about the wiki and redirects’ (Mason and Thomas, 2008, p. 25). Nevertheless, this important distinction is not considered in statistics, which include system-generated pages in the final count. Websites such as SIC and MLO do not have this automatic feature, so ‘total page number’ stands for submissions by users, although in our specific case this does not necessarily correspond to the content of the final novel, as will be seen later. 7) (Date) From As already remarked in the presentation of the three platforms, the fact that they were all launched in the same year indicates that they enjoyed equal technical opportunities and could rely on similar hardware and software when commencing their activities. Thus, performance depends on the founders’ choices and evaluation of the same virtual possibilities. But the mere ‘date from’ does not tell us anything about the platforms’ subsequent development paths, since it captures a static feature, as remarked by 35 Roth. To obtain information about their development, I will need to take into consideration the timelines of the projects. 8) Governance As Roth insightfully remarks, ‘hierarchies in wikis often reduce to a simple dichotomy between contributors as suppliers of content and administrators as enforcers of community decisions’ (2007, p. 122). This perfectly matches the AMP structure of governance, where contributors may freely submit content without any editing by administrators.30 Conversely, SIC and MLO seem to share a hierarchical structure, being organized around a central administration which decides policies and rules. Not only do users need to register in order to participate, but the content they propose is also submitted to a certain degree of ‘evaluation’ before being published. But similarities end here: in fact, in MLO the administrators31 do not interfere in the writing and editing project, limiting their contributions to the policy forum and to guidelines. In SIC, the two administrators actively participate in the editing process. They are at the same time administrators and editors, in that they both run the platform and have reserved to themselves an important position within the organizational structure: the Art Director (AD) role. This does not mean that the platform has a fixed organization: on the contrary, from the forum and blog discussions it emerges that some users have been promoted to the AD position, a dynamic observed also by Roth: ‘a small subset of most active users is granted administrator status’ (2007, p. 124). To grasp the implication of this last dynamic and those emerging from the other structural features, it will be necessary to consider the writing methods. In fact, the structural features dealt with above describe only the static aspects of the platform, which does not allow us to understand fully the platform architectures and 30 In fact, AMP gave its users the widest freedom, which sometimes posed dilemmas for the administrators. For example, the user Yellow Banana edited other users’ pages, inserting the word ‘banana’ and replacing with it other motifs and words, creating a debate among the Penguin administrators and those responsible for the project about the opportunity to moderate this user’s contributions. In the end, administrators decided to leave it to the other users to deal with Yellow Banana, with a positive result (Mason and Thomas, 2008, p. 9). 31 Actually, there seems to be just one administrator.
 36 their influence on user participation. To this end it is necessary to tackle collaborative writing features, analysing them from a different perspective. 37 4.2 Thematic analysis: Writing features In this chapter I shall present the codes I have created when studying the platforms and their writing activities by grouping instances of data and phenomena by theme. The themes have to be considered as ‘umbrella terms’ that can enable data to be regarded as ‘of the same type’ (Andretta, 2000-2009). As will be seen, rather than representing a mere list of features, the relations between codes provide a first key to understanding the dynamics under study. Indeed, beginning with the first code, I have generated each subsequent one to describe aspects or themes of the previous code or codes, or which can be analysed from a different perspective. Considering the inductive approach I have chosen, I prefer not to list the codes here, but to present them sequentially, in order to show their development during the process of analysis. The first code is concerned with writing methods, an element recurring quite often during the structural analysis as a decisive factor to be considered. CODE 1: WRITING METHOD As seen from the above analysis of structural features (Chapter 3), the experimental nature of AMP, supported by the choice of a wiki platform, entailed that the development of the novel was left entirely to users, with no suggestion of a writing strategy to be followed, whereas both the SIC and MLO websites asked their contributors to abide by a writing method already elaborated and established. Thus, in what follows I will expose and compare the writing methods of each platform, in order to identify the differences amongst them and their effect on user participation. Table 4,2 offers a general overview of the main subcategories of data collected within the ‘writing methods’ code. Table 4.2: Writing methods Stages SIC MLO AMP Yes Yes No Role hierarchy Yes No No Editing Yes No Yes Contribution form Yes No No Classification by genre No Yes Yes Contribution rating No Yes No 38 I will now discuss the methods and their features in a narrative manner. a) The SIC method is laid out by its authors on the home page of the website32 as follows: 1. Each writer writes every component of the novel. 2. One or more Art Directors (ADs) select and standardize the written material submitted by the writers. 3. The writing is performed by the compilation of ‘forms’, dedicated to developing the various elements of the storyline (characters, places, scenes and so on). It seems evident that the method pivots around the AD, who: 1. Writes a brief subject 2. Elicits from the subject the elements of the storyline (e.g. the characters) 3. Submits the character forms to the writers 4. Composes what the writers send back. Composition consists in marking in green the approved parts and in red what is not approved. All the approved material is then amalgamated into the definitive character forms. 5. Updates the subject if new elements emerge from the forms. 6. Repeats points 2 to 5 for location, style and situation forms. Drawing on forms (which represent the raw material of the final draft), the AD performs the editing, possibly integrating useful hints from other forms. All these operations are discussed on the SIC website, specifically in the forum and in the blog; nevertheless, the editing process is carried out by ADs. In the novels and tales already completed the two founders and administrators played the AD role, while in the present project other users have performed that role as well, as a result of a specific request by them. Another reason to expand the number of 32 All translations are mine. From http://www.scritturacollettiva.org/documentazione/metodo-sic - sic- comic 39 people endowed with AD responsibilities was the larger number of people involved in this project. Dealing with one novel at a time, the method entails neither a classification of genres nor a system of rating submissions. b) MLO. What follows is the exposition of the MLO method given on the website, available from the clickable icon ‘People – The novel’:33 ‘1. Project framework - The main unit in [the] Make Literature conception is not the chapter (which is usually the case in interactive fiction schemes), but the New Fiction Book Project. To make a structural foundation for solid project management, we defined a framework in which all literature genres are sorted in six general categories: Science Fiction and Fantasy Mystery and Horror Crime and Thriller Romance and Comedy History and Adventure Children’s and Social At one time, there may be only one active project in each genre. That makes six active projects simultaneously overall. In other words, all community efforts are focused on writing one particular fiction book in each genre. 2. Project management - Each project comprises of stages that are organized like separate writing contests: Storyline - Idea submissions Opening chapter Middle chapter 1, 2, 3... Closing chapter Book cover design Book finalization tasks Selection cycles are organized around the forum topics attached to each stage, with review submissions, user rankings and discussion. Writers [at] this stage may receive valuable feedback [on] their writing. The goal is to select the best 33 http://www.makeliterature.com/system/about-us. I have amended [in brackets] quoted passages from the MLO website, in order to facilitate the reading. 40 option for [each] new chapter, one that reflects [the] general desire of the community. 3. Advanced rating algorithm - To be sure that only [the highest] quality content will [make] the official draft, a complex rating algorithm has to be developed. It takes in[to] consideration many different factors, some of which are: Submission time Popularity - how many unique views the item received Ranking - how members vote Number and quality of reviews Compliance [with] the main idea and storyline It is important to stress that despite its complexity, this rating system has its transparent interpretation, which can be easily reviewed at any time for each particular item. 4. Project Micromanagement - One major problem related to dealing with collaborative writing projects is the fact that you cannot actually impose some firm timetable in [an] environment in which so many things depend on other people’s behavior and [where there are] so many other limitations [...]. The only [solution] is to establish [...] Project Micromanagement, [a] set of additional rules and criteria that automatically handles each of these particular situations. Some of these criteria are: Minim[um] submissions per stage to start [the] voting process Minim[um] number of ranking[s] received before an item is considered Maxim[um] number of submissions per [...] member Maxim[um] number of votes per member per stage Minim[um] rank limit for [each] selected item ... 5. Project Timeline - [This] defines each project dynamic by marking out significant checkpoints and [the] timeframe in which each of them should be reached. Stage launching time Submission period Voting period Results consolidation period Project idle period - time between two consequent projects 41 In conjunction with Project Management, Project Timeline determines transitional procedures between consequent stages and [forms] the spine of the project development’. Unlike SIC, in MLO all users have the same role: all are in fact entitled to submit, rate and vote on contributions, with no distinction between users and administrators. Decisions about the novel’s development have to pass through community discussion, which draws a picture of non-hierarchical governance, contrary to the first impression rendered by structural analysis. In light of the description of the SIC method and of Lowry’s depiction of collaborative writing stages (section 1.1) what is really surprising is the lack of an editing stage. The method entails that each chapter submitted will be rated, voted on and discussed by users, in a sort of contest where the one with the most votes will be chosen to continue the storyline. But there are no scheduled times or spaces for editing; apparently, a contribution is considered in its entirety, without being submitted to changes or corrections. Albeit not the object of the present research, I would comment that the lack of editing calls seriously into question the aesthetic value of the whole project. c) AMP writing method Mason and Thomas (2008, p. 3) report that ‘The wiki was opened to the public on Thursday 1st 2006 with a line from Charlotte Bronte’s Jane Eyre: “There was no possibility of taking a walk that day”’. No other indication as to the writing was given to users, nor did the platform originally carry an internal articulation of genres. During the writing process, a user proposed the introduction of a distinction according to literary genres, the ‘choose your own adventure’ formula. Since this generated an alternative path of its own (2008, p. 7), I decided to consider it as a present feature when compiling Table 4.2. This subdivision by genre also testifies to the lack of a hierarchy of roles amongst users, since anyone could submit texts, edit and structure the whole text, with no restrictions at all.34 The Penguin administrators followed this users’ suggestion. In fact, after the writing stage, in order to facilitate it, they subdivided the body of the novel into seven sections, plus another one containing the most ‘bizarre elements’, such as the ‘Banana version’ and ‘choose your own Adventure’ (Mason and Thomas, 2008, p. 7). 34 42 CODE 2: CONTRIBUTION CONSTRAINTS The writing methods described above determine some constraints with regard to modalities of contribution. It is important to specify that some of them are determined by the platforms’ technical capabilities, whereas others are a consequence of the founders’ choices. Since these elements are all interconnected, I will consider them as attributes of the unique code CONTRIBUTION CONSTRAINTS. To facilitate the comparison, I present them in Table 4.3. Table 4.3: Contribution constraints SIC Time Limit Word Limit per Submission Project Participant Limit 1 year Determined by the forms provided Restricted MLO Not scheduled Open Open AMP 5 weeks Open Open • Time Limit It is interesting to notice a first implication of contribution constraints. In MLO, the chosen storyline for the People project was released on 14th September 2008, then commented on and rated by a few users in the following days, but no other progress to the next stage is visible; proposals for the second chapter were not submitted by two other users until 2009, in April and August respectively. Thus, in a whole year, not much progress was made in the creation of this collective novel, at least none comparable to the SIC and AMP platforms, discussed later in this section. What could be the reasons for this difference? I believe that it can be mainly attributed to low user participation, although the website has a very large number of registered members (see Table 4.1); for instance, another MLO project, TwiHaiku – Twitter Poetry,35 is followed by more than five hundred people. Thus, the problem is not the popularity of the website itself. Rather I think that the real reason is an excess of conditions to be met in order for a chapter to advance to the next stage, which are these: 35 http://www.makeliterature.com/twihaiku/twitter-poetry 43  Minimum submissions per stage to start voting process  Minimum number of rankings received before an item is considered  Maximum number of submissions per member  Maximum number of votes per member per stage  Minimum rank limit for selected items. This means that if a submission does not reach a minimum number of rankings and votes, it will not be allowed to continue in the project. Consequently, if no chapter meets the conditions, the novel will not actually proceed. By contrast the constraints applied by SIC and AMP seemed to work. With AMP, the rigid timeline allowed the project to be conducted to its completion, while also helping to manage a high number of contributors and contributions. As for SIC, despite the long time accorded to the project, the role of the ADs assures a fast progression to the next step, as can be observed on the platform. Word Limit • A shared feature of MLO and AMP is the unrestricted number of words per submission, whereas in SIC users can contribute only by completing the forms provided, which obviously limits greatly their freedom, in relation both to the number of words and to content. The nature of actual contributions to the novel will be specifically analysed through a further code, concerning the types of contribution, in other words the consistency of contributions. Participant Limit • On the SIC platform, the number of participants is restricted and they were chosen by a temporal criterion: the first candidates who answered by email when the project was launched were automatically enrolled in it, until the predetermined number of one hundred was reached.36 As for MLO and AMP, there is no restriction on the number of participants and they are hence open to everyone. Does this condition of openness, common to the two platforms, result in equal data in relation to actual user participation? In order to address this question I created another code. 36 Actually, the number of enrolled participant to the project corresponds to 236, but actual contributors who produced at least one form are 90. 44 CODE 3: REGISTERED USERS AND ACTUAL CONTRIBUTORS In order to understand more about the participation, it is necessary to compare the number of registered members with the actual contributors to the project. In fact, registered members include people who just signed into the platform to look at it, but did not really participate. These have to be distinguished from actual contributors, i.e. members who actually contributed to writing the novel. Table 4.4 compares these data. Table 4.4: Registered members and actual contributors to each project SIC Registered members of platform Actual contributors to project 406 90 MLO 815 10 AMP 1476 570 At first glance, it seems that many registered members have made no actual contribution to the writing process. Approximate proportions of the total membership actually contributing were 1/4 in SIC, 1/8 in MLO and 1/3 in AMP. It is interesting to note that these proportions again show a gap between MLO and the other two platforms, as already observed in the matter of project pages (Table 4.1); these numbers suggest a weak commitment of MLO users to the project. The next code explores this datum further. CODE 4: FREQUENT CONTRIBUTORS The rationale for code 4 is based on the report of Mason and Thomas (2008) on AMP, where they differentiate participants according to the number of edits they performed during the life of the project. Table 4.5: Frequent contributors 45 USERS Edited at least once Edited 2-5 times Edited 6+ times SIC 90 90 90 MLO 10 0 0 AMP 570 67 18 It can be seen that all SIC contributors edited frequently; in AMP, 570 members edited at least once, 12% of these did so on 2-5 occasions and only 3% edited more than 6 times. As for MLO’s People, no user edited more than once. These data have important implications. First, the frequency of editing, which measures the users’ engagement, shows a high level of engagement for SIC, whereas in AMP the frequent contributors, i.e. the ones who really write the novel, are a minority group. As for MLO, there were no frequent contributors to this project, which indicates a total lack of engagement. Second, these numbers reveal significant patterns in user engagement, but to better grasp their implications it is important to define what the term ‘edit’ refers to. Table 4.5 is quite specific in terms of quantity, but does not account for the nature of these edits. Can we equate editing with making a contribution? This is addressed by the next code. CODE 5: TYPES OF CONTRIBUTION The definition of ‘edit’ in the Oxford English Dictionary reads: ‘a change or correction made as a result of editing’.37 But Mason and Thomas (2008, p. 25) use it in a broader sense, referring in fact both to contributions and to changes or corrections to contributions. This depends on the fact that wikis register all users’ actions as edits, without distinguishing between them. As a result, an edit can either correspond to the first submission of a chapter, or to the change of just one line in a text already submitted, or even to the correction of punctuation.38 All of these are recorded in history pages as edits, at the same level. 37 38 http://www.askoxford.com:80/concise_oed/edit?view=uk Technically, the first contribution could not be termed an ‘edit’, since it does not represent ‘a change or correction’; nevertheless, it is certainly an ‘action’, and this is why in wikis it is statistically included within ‘edits’. 46 Conversely, in websites, editing is not a feature commonly performed by all users, nor can edits be assimilated to contributions. As seen in the analysis of writing methods, in SIC contribution and editing are separate tasks, accomplished by different users; in MLO users are allowed only to send contributions, discuss and vote on them, but there is no edit stage. Moreover, in a Web 2.0 context, i.e. a collaborative context, all the platforms also host discussions and communications related to the novel’s creation which thus have to be considered as contributions. Considering all the above-mentioned typologies and dissimilarities, I believe that contributions can be more usefully coded according to their functions. Narrative material to be included in the main body of the project and edits to it can be coded as ‘narrative material and edits’, with opportune distinctions between wikis and websites. Conversely, non-narrative contributions by users, such as discussions and other communications related to the novel, need to be separately classified and analysed. They are made by means of blog and forum posts concerning discussions amongst platform users about methods and procedures, which we could term ‘metaposts’. Then we also have other posts referring to communication both internal and external to the platforms, not necessarily aimed to the novel’s creation, usually performed by means of other technologies, most of them SNSs. In light of the above, I have created two sub-codes for contributions, shown in Table 4.6. Table 4.6: Code 5 Code 5: Types Of Contribution Sub-Code 5a: Sub-Code 5b: Narrative material & Blog, Forum & Other Edits Posts SUB-CODE 5A: NARRATIVE MATERIAL & EDITS This code groups narrative material and edits because they share a narrative function, but takes into account the differences pointed out above, pertaining to the modalities of submission of the contributions, which appear clearly linked to the constraints delineated in Table 4.3, particularly the word limit. 47 In SIC, contributions are made via storyline forms provided by the administrators to each project participant. The administrators forward the completed email forms to the ADs, who perform the editing. Only edited texts are published on a summary board, which is updated during the various stages of the process. In MLO users are supplied with a text editor embedded in the novel page to submit their contributions, representing entire chapters, which are redirected to the public ‘catalogue’ section for voting and rating. No edit stage is scheduled, according to the writing method set out above. In AMP, contributions are submitted through the main novel pages and are immediately published with no filter. Users can edit contributions, move and remove materials, although ethical guidelines require them to take into consideration elementary principles of respect and politeness towards other users’ contributions. These contribution modalities are schematised in Table 4.7. Table 4.7: Narrative material and edits NARRATIVE MATERIAL SIC MLO AMP Emailed forms Embedded text editor Main page submission EDITS By ADs No Yes, by all users SUB-CODE 5B: BLOG, FORUM AND OTHER POSTS Since SIC and MLO both run on websites, they could be supposed to share a similar organisation of discussion spaces. In fact, these spaces are structured in contrasting ways: in SIC the blog is used for communication by administrators on more general issues, whereas the forum serves as a space to discuss more specific aspects of the novel, with threads created from actual users’ posts. In MLO the blog is intended to discuss issues related to the novel, while the forum is dedicated to general issues, originally threaded by the site administrators according to general themes. In AMP, there is no actual forum. However, Mediawiki generates for each registered user specific ‘talk’ pages, indicated in the table, which in fact function as forum pages. As a rough distinction, the blog hosts reading reports by the Penguin editor Jon Elek, along with subsequent discussions with other administrators and 48 users. On ‘forum’ pages—or perhaps one should say ‘talk’ pages—users generate discussions in which administrators can take part. Therefore, the main difference is in the source of the discourse: talk pages are generated from specific pages of the novel and by users, which means that they are more related to specific elements of the novel, whereas the blog usually deals with more general issues triggered by administrators, such as reflections on the wiki-novel experiment in progress. Tables 4.8 to 4.10 schematise the above observations; because of the large amount of data involved, there is a separate table for each platform. Table 4.8: Blog, forum and other posts – SIC SIC Function Contributors Threads BLOG POSTS General issues about the method, literary and cultural discussions Users and administrators Created by posts FORUM POSTS Discussions and comments on the novel Users and administrators Created by posts OTHER POSTS Comments or discussions by means of Web 2.0 technologies Users and administrators None Table 4.9: Blog, forum and other posts – MLO MLO Function Contributors BLOG POSTS Discussions and comments on the novel Users FORUM POSTS General issues, not related to the novel Users and administrators Created by admin: Your writing, Writing issues, Reading and literary topics, Writing contests, Community, Site info OTHER POSTS Comments or discussions linked to SNS groups Users and administrators Threads None None Table 4.10: Blog, Forum & Other Posts – AMP AMP BLOG POSTS FORUM POSTS OTHER POSTS Talk pages, User pages, User Talk pages: Discussions and comments Discussions and comments on the whole novel on the whole novel and on specific elements of it Users and administrators Users and administrators Function Contributors 49 Threads None Created by posts CODE 6: WEB 2.0 TECHNOLOGIES From the analysis of code 5 it can be seen that when present, ‘other posts’ refers to communication both internal and external to the platforms, performed by means of other technologies, most of them SNSs. Since I am interested in the role of Web 2.0 technologies within collaborative writing, I think it is important to study at closer range the use of these technologies within the projects. I will distinguish between embedded technologies, i.e. integrated in the platform, and simply linked, thus connecting to another Internet page outside the platform. The data are presented in Table 4.11. Table 4.11: Web 2.0 technologies Embedded SIC Twitter Blog Forum Twitter Blog Forum Vimeo RSS Feed Linked Facebook My Space Del.icio.us ANobii Twitter Friend Feed Facebook Stumble Upon My Bloglog DIGG Del.icio.us Technorati Favorites Wordpress.org MLO AMP Blog It is immediately clear that AMP has the smallest number of technologies and MLO the largest, with SIC between the two. Does the number of technologies displayed influence the platforms’ activities in terms of project progress? It would be reasonable to anticipate that the answer would differ according to the technologies being considered. With regard to embedded technologies, we could answer 50 positively. I will start by analysing the technologies common to the three platforms, i.e. blog and forum. In AMP both blog and talk pages (fulfilling forum functions) correspond to quite a large number of pages, which testifies to a high volume of discussion between users about the writing of the novel. Similarly, in SIC both blog and forum are densely populated, as they have an important function, i.e. discussion among users about the novel’s development and the methods used. In MLO blog and forum are also embedded, but the data are quite different compared to the other platforms: the blog concerned with the novel is almost deserted, whereas there is a huge number of forum posts, but these are not related to the novel, as seen in Table 4.9. Again, this datum shows the poor vitality of the People project. I will now consider the use by SIC and MLO of Twitter as embedded. In SIC it is used to provide users with fast updates on project news and deadlines, whereas in MLO it hosts another platform’s activity, Twitter Haiku, i.e. the composition of short verse poetry in haiku style.39 Besides proposing their poems, users comment and discuss; one thread of comments is called ‘Twitter Poetry: a new genre?’ At first glance, TwitHaiku seems presently to be the most successful activity on the platform. As for the remaining embedded technologies, MLO houses VIMEO and RSS feeds: the former shows a video which has little to do with the writing process, whereas RSS is used to provide users with frequently updated content from their favourites websites or blogs, with no indication of a connection to the project. Let us now turn to linked technologies, which can be observed on all the platforms. Amongst these we find a free publishing platform present uniquely in AMP (Wordpress.org) and some Internet search engines for blogs (Technorati Favorites, Del.icio.us.), but most of them are SNSs, thus primarily aimed at connecting users to one another; again, what is of interest here is if they have some influence on the writing project. As for the Wordpress link in AMP, there is no real connection with the novel; therefore, for AMP, I can immediately answer the above question negatively. 39 Haiku is a form of Japanese poetry, consisting of 17 moras, in three metrical phrases of 5, 7, and 5 moras respectively. [… ] In English they usually appear in three lines, to parallel the three metrical phrases of Japanese haiku. From http://en.wikipedia.org/wiki/Haiku. 51 In SIC three of the linked SNSs, namely Facebook, MySpace and aNobii, are not directly connected with the novel project, but rather with discussion groups interested in collaborative writing, or engaged in similar projects. Interestingly, the link to Del.icio.us presents a list of bookmarks leading to web pages whose content can serve as documentation for the project. Hence, the SNSs make no actual contribution to the novel, whereas Del.icio.us serves as a depository of materials providing detailed information on topics raised in the novel. In MLO the technologies listed are linked from two different positions. In the homepage menu, under the headline: ‘Follow us online’, they link to groups on other networks discussing literature; on the novel pages they are displayed as icons above each user contribution, inviting people to add the latter to the named SNS. Thus, they can serve to hand out texts created by platform users in other circles. In AMP there are no links to SNSs, thus no provision of external communication. Hence, these Web 2.0 technologies serve the following purposes: • • • To share an interest in writing and literature with people outside the platform (SIC, MLO) To give more visibility to the project (SIC, MLO) To share between users information and material related to the novel (SIC). To answer the above question about the influence of technologies on the projects, we can say that embedded technologies such as blogs and forums are vital elements of the projects, whereas linked technologies seem not to add real value to the creation of novels. Other implications will be drawn in the conclusion, when discussing the findings of this work from an overall perspective. 52 5. FINDINGS 5.1 Analytical findings In this chapter, I shall discuss the implications of the structural and thematic analysis carried out above. First, I shall examine the connections between the different structural features that I have identified in the structural analysis (section 4.1) and the writing features, expressed by codes created through the thematic analysis (section 4.2). I shall then discuss the interrelationships between the different codes in order to highlight the hierarchy of the various features composing the overall platform architecture and to analyse how these features shape participation according to its main attributes, i.e. consistency of contribution and engagement.40 As anticipated in sections 3.1 and 3.2, a structural analysis considers only static features, not dynamic elements. Hence, this analytical method alone is not able to provide an overall picture encompassing the full range of elements and interplays which have an impact on the three platforms constituting the object of the present research. In fact, this first analytical stage triggered new issues, which could be effectively dealt with only through an in-depth examination of writing specificities, as I remarked at the beginning of Chapter 4. As noted there, structural features describe static aspects of the platforms, but they do not provide details, either singularly nor combined, of their activities. The issues arising from the structural analysis categories are dealt with by codes that structure the thematic analysis, as shown in Table 5.1. Table 5.1: Thematic analysis codes Licence  Frequent users CODE 4 Platform  Contribution constraints, CODE 2 Scope  Writing method, CODE 1 Language  Contribution constraints, CODE 2 Users registered  Contribution constraints, CODE 2 Project pages  Writing method, CODE 1 Governance  Writing method, CODE 1 From  Contribution constraints, CODE 2 40 Implications of a more general nature will be examined in the concluding chapter. 53 The frequent recurrence of the first two codes, writing methods and contribution constraints, means that most issues raised by structural features can be addressed mainly through these. The central positions of these two codes will appear more clearly in the light of the map of relations between codes, which I shall present and discuss below. Writing methods (code 1) cause Contribution constraints (code 2). Code 2 is constituted by three specific constraints, on time, words and participation. Two of these attributes, namely limits to words and participation, have further generated other codes: Registered users and Actual contributors (code 3), Frequent contributors (code 4), and Types of contribution (code 5). Code 3 expands the notion of participation limit, distinguishing between registered and actual contributors; Code 4 further expands code 3, obtaining from it the numbers of contributors according to their editing frequency. Code 5 has been generated from two different codes: on one hand it represents an aspect of ‘word limit’ (code 2), in that it defines what kinds of contribution users are allowed to submit. On the other hand, it expands the notion of ‘edit’ resulting from code 4, which poses both terminological and substantial issues. Considering the variety and number of elements to take into account, code 5 has been split into two, 5A and 5B. The latter generated code 6, which differentiates 5B from the perspective of the technologies involved. The code family can now be schematised as shown in Table 5.2. Table 5.2: The code family CODE 1  CODE 2  CODE 3, CODE 4, CODE 5  CODE 5A/ CODE 5B  CODE 6 54 This map of relations has a first important implication: the architecture of collaborative writing platforms rests primarily upon writing methods and contribution constraints, since the other codes derive from them. I shall now summarise the findings of these first two codes, depicting platform architectures. Code 1: Writing Method Architecture Writing methods involve a certain number of elements or ingredients: this code displays the level of complexity of each writing method, according to the number of the following ingredients: stages, role hierarchy, editing, form of contribution, classification by genre and contribution rating. As a result, the platform displaying the greatest level of complexity of method is SIC (stages, role hierarchy, editing, contribution form), followed respectively by MLO (stages, classification by genre, contribution rating) and AMP (editing, classification by genre). Code 2: Contribution Constraints Architecture This code shows that the most restrictive organisation is that of SIC: not only does it have a fixed deadline for the project, but it also prescribes fixed forms of contribution. In addition, the participation is restricted by a temporal criterion. AMP also has time constraints, while MLO has none. The findings of codes 1 and, i.e. the two related to platform architecture, show that the platform displaying the most complex architecture is SIC, which also sets the highest number of constraints. Code 3: Registered Users and Actual Contributors User engagement The first general result concerning engagement is that very few registered users make a substantial contribution to the realisation of these projects. In fact, the large majority of users, after an initial interest and participation, fall away. The platform with the greatest number of actual contributors is AMP, followed by SIC and then by MLO, which has very few. Code 4 Frequent Contributors  User engagement The analysis of this code provides an in-depth examination of the previous one. SIC has highest number of actual contributors, whereas in AMP frequent contributors, i.e. 55 the ones who really ‘write the novel’, are a minority group. As for MLO, the platform did not manage to create a group of frequent contributors at all. Code 5A: Narrative Material & Edits  Contribution Consistency This code deals with the first type of user contribution, i.e. the actual submission of narrative material and edits, analysed according to their modalities of submission. When narrative material is submitted to SIC, the administrators filter it, whereas contributions to AMP and MLO are immediately published on the platforms. As for editing, in SIC it is performed by the Art Directors and in AMP all users can both submit and edit, whereas in MLO there is no scheduled editing. Code 5B: Blog, Forum & Other Posts  Contribution Consistency This code deals with the second type of user contribution, i.e. to the discussion and communication posts. More specifically, it displays discussion modalities and rates for the three platforms. Differently from the previous code, SIC has the same number of participatory elements as the other platforms. The main difference between platforms is represented by the presence or absence of post threads: in SIC they are present and generated by users’ posts, whereas in MLO they are predetermined by administrators and in AMP discussions are not threaded. As for rates of discussion, these are distinguished according to topic, i.e. whether they concern general issues or the project itself. SIC and AMP have a higher volume of discussion than MLO in the latter category. General issues are tackled on all the platforms, but have different cores: in SIC and AMP they are related to the collaborative project, whereas in MLO they pertain to more varied themes. Code 6 Web 2.0 Technologies  User engagement This code analyses how the engagement of users is influenced by the technologies used on the different platforms. Findings differ between embedded and linked technologies. With regard to the former, examination of the posts reveals that the technologies most frequently used are blogs and forums, which are generally very useful in project realisation, since they allow users to discuss and manage the composition of the novel. As for linked technologies, most of them are SNSs. MLO has the highest number, followed by SIC and then by AMP, which has just one external link. The 56 analysis shows that the function of technologies is not directly associated with the realisation of the project, with the exception of Internet search engines, which in SIC are used to find and share useful documentation for the novel. Hence, MLO has the highest number of technologies, but almost all, both embedded and linked, are hardly ever used. The analysis also reveals a low volume of posts. SIC occupies a middle position, using a good range of technologies, with some of them playing a central role in the novel. The volume of posts is very high. AMP has only one blog, which is used to discuss aspects of the novel. It shows a dense activity of users’ posts. 5.2 Overall findings The three platforms display many architectural differences, which result in different levels of achievement: AMP has accomplished its task (i.e. produced a narrative work), SIC’s work is successfully progressing, whereas AMP’s People is struggling to take off. I shall now summarise the overall implications of the findings, then at the end of this section I shall spell out the nature of these achievements in more detail. We have already seen that structural features are only one component of architectures and they do not account for the main differences among the three platforms. This is demonstrated by the fact that the differences between the two platforms that run on websites, i.e. SIC and MLO, are more numerous and more relevant than those between either of them and AMP, which is a wiki. The architectural component that really influences outcomes is the writing method. In fact, although both websites specify a writing method, unlike the wiki, the effects of these two methods are quite different: SIC has succeeded in recruiting members who feel involved in the common project, whereas the MLO project has very few frequent and regular contributors. The wiki AMP attracted many people to register and the novel was completed despite the absence of an actual writing method; its success is mainly attributable to a high level of user engagement obtained by frequent and animated discussions among users and even more importantly between users and administrators, who succeeded in keeping a number of users engrossed in the experience. Since user engagement will be discussed in more detail below, I will now turn back to the implications of writing methods. 57 The complexity of methods is not a sufficient reason per se to explain these varying outcomes. In fact, difficulties in MLO cannot be attributed to a method with few stages, but to the lack of two specific writing features, connected to each other: role hierarchy and an editing stage. These two features are associated in that role hierarchy leads to a clear subdivision of tasks between users, which seems to help the SIC project greatly, whereas its absence in AMP led to seven different storylines, plus other pieces of content not integral to the main novel. Within writing methods, the choice between a single project (topical scope; AMP and SIC) or on several novels at the same time (multipurpose scope; MLO) also plays an important role; the multipurpose scope tends to spread users over different novel projects or other activities. The architectures here delineated, referring to combinations of writing method and constraints, influence user participation through rules for consistency of contribution and user engagement. The first of these is variably regulated on the three platforms, according to the extent of prescriptions. Considering the outcomes of these different systems on the three platforms, it should be noted that the presence of a contribution form and of a prepublication filter, as in SIC, speeds up the work. Otherwise, a platform has to cope with very diverse kinds of contribution, resulting in too many different storylines to manage easily, as in AMP, or simply in contributions which are difficult to integrate because of different writing styles, as in MLO. The editing stage also appears to be decisive: where it is present, as in SIC and AMP,41 it promotes the progress of the project, providing the opportunity to work on the contributions submitted; where it is missing, as in MLO, progress is poor. Obviously, constraints in forms, filtering and editing also have downsides: decisions are devolved to a minority of users or to administrators, limiting the participatory element. I shall return to this issue in the final chapter, from a more general perspective. Constraints on contributions also play an important role in maintaining a high level of user engagement: time limits help users to focus their efforts on a clear goal, whereas word and participation limits prevent them from submitting too little or too much material, which would pose difficulties in managing it. The usefulness of these 41 Although with very different characteristics: on AMP the editing is an inherent function of the wiki platforms, whereas on SIC the editing is performed by the AD outside the platform. 58 three constraints is dependent on their balance, which differs among the platforms: on SIC the time constraint is less strict, but it is compensated by very rigid word and participation limits, whereas on AMP it is the exact opposite. MLO does not provide any constraints, which I consider the most likely reason for its difficulties. To support this statement, I make reference to another project on the same platform, TwitHaiku, based on the Twitter technology, which imposes a strict word limit: in this case, MLO has succeeded in attracting users and maintaining their engagement. Both contribution consistency and user engagement are influenced by the use of Web 2.0 technologies. These allow users to contribute not only narrative material, but also their opinions and ideas about the novel’s development. The extent of discussion of the project is a measure of engagement, which strongly affects chances of success: on SIC and AMP discussions are lively and regular, involving both users and administrators, whereas on MLO there is little consultation between users about the novel. Still, the role of SNSs seems secondary or peripheral: not all the platforms analysed make use of them. Moreover, it is not directly connected to the novel project, with few exceptions. Hence, not all Web 2.0 technologies can enable collaborative writing platforms to achieve their purpose. In light of all the above, SIC turns out to be the platform best able to achieve the objective of writing a collaborative novel. Its designers have chosen to limit not only the number of participants, but also their freedom, in order to reach their goal. Nevertheless, there is a lively debate between users and administrators on the website related to the method, which its creators consider open to improvement. From the data analysed it is possible to infer that the final product will result in a linear narrative development, making it possible to read it both electronically and in printed form. By contrast, a year after its launch, MLO’s People is still striving to find its way through a method which leaves almost every choice to users. The result is that the project is not taking off: the three submissions of first chapter proposals have been commented on and voted by a few people, but they are still waiting to reach an unspecified ‘minimum number of ratings’ to advance to the next stage. Therefore, no users have reached the status of ‘frequent contributors’. There is also very little discussion on the website by users of issues related to the novel. Although scarce, the contributions so far submitted present a linear storyline, readable both electronically and printed, as in the case of the SIC novel. 59 As for AMP, this wiki project has been completed. It has the largest number of registered members and many frequent contributors, although far fewer than SIC. Discussions on blog and talk pages are numerous, involving both users and administrators. Nevertheless, the resultant text is quite laborious to read even electronically, let alone on paper. This is mainly because a lack of constraints on narrative contributions, which can be endlessly edited by all users with no restrictions. This has caused the proliferation of storylines, characters and situations, eventually subdivided into seven parts by the administrators in an attempt to make reading easier. 5.3 Validity limits Research methods may be compared to lenses used to analyse a given set of data. Hence, provided that they are applied with enough rigor and hermeneutic awareness, their heuristic validity depends very much on how they fit the kind of research questions they try to answer. Qualitative research such as mine, which by definition involves the active engagement of the researcher, needs to take into account the subjective element in the research process. This means that I am aware that the way I approach and interpret the data is somewhat influenced by my personal worldview and interests. Yet, I look at the data with both the awareness and openness to the possibility that my analytical work will influence me and change my own perspective. According to Gubrium and Koro-Ljungber (2005), the collection of data is heavily influenced by the so-called ‘border-making’ process between researcher and respondent: there are no questions without assumptions. From a social-constructivist perspective, questions reflect a way of perceiving reality, and every data collection method is a co-construction between the agendas of researcher and respondents. There is no such a thing as a fully objective method of data collection, or a fully objective analysis and theorisation. That being the premise of every scientific endeavour, it behoves the researcher to handle the data with hermeneutic prudence and epistemological awareness, continuously testing her findings against the background of her own assumptions. 60 Coming now to the limits stemming from the specific research question that guides the present work and to the methods used to analyse the data, I am aware that these allow me to infer reasons for diversity among platforms of collaborative narrative writing on the Internet. However, I am also aware that they do not allow me to apply its findings to all the architectures of collaborative writing. Furthermore, being based on a limited sample, an analysis such as this is also limited in terms of generalisation and theoretical inference. Nevertheless, the inclusion of quantitative data within codes reveals patterns of behaviour which can be extended to similar experiences of collaborative narrative authoring. Numeric data offer the opportunity to compare the features of platforms, helping to elucidate the results of their activities. 61 6. CONCLUSIONS AND RECOMMENDATIONS The earlier chapters have analysed the impact of the specific architectures of the selected platforms on user participation and therefore on the accomplishment of the platforms’ objectives. We have also seen that user participation is a function of the constraints set by the platform and that it benefits from tools enabling discussion among users, which foster engagement. More precisely, the stronger the constraints, the higher the level of participation, and therefore the greater the chances of bringing the project to a positive conclusion. Undeniably, SIC is the platform whose design was found to be the most effective in the accomplishment of the task of composing a collaborative novel on Web 2.0. As already remarked, the main reason for this is the degree of constraint set by the project’s designers. From this, it seems possible to conclude that ‘the wisdom of crowds’ needs to be opportunely regulated and limited in order for that wisdom to manifest itself. The above conclusion brings us to the much debated issue of the need to set rules and constraints on Internet users’ participation. In fact, the debate about Web 2.0 is polarised into two competing ideological stances: the first supports the selfregulation of the Internet, whereas the second argues for the need to introduce rules. Both stances have the same final objective: to create the conditions for a real participation in the Internet sphere for all its users. The difference between the two positions mainly concerns the route to this goal. According to the first and more optimistic position, the participatory Web will succeed in limiting the attempts of corporations and governments to control the Internet: through the use of Web 2.0 tools, users will become aware of their power by learning to cooperate in order to take full control of their lives. Not only are supporters of the second stance sceptical about the realisation of this prophecy, but they also question the abovementioned ‘wisdom of crowds’, alleging it to cause ‘an endless digital forest of mediocrity: uninformed political commentary, unseemly home videos, embarrassingly amateurish music, unreadable poems, essays and novels’ (Keen, 2008, p. 60). Proponents of the more restrictive view also argue that the supposed democratisation of the Web through UGC ‘is undermining truth, souring civic discourse, and belittling expertise, experience and talent’ (Ibid, p. 64). 62 Considering my findings in the light of these two stances, one would be inclined towards the second. In fact, this research has demonstrated that in collaborative narrative writing the presence of rules and constraints turns out to be essential for the accomplishment of the collaborative novel. To quote Scott Rettberg, ‘If truly collective web narratives, open to contributions from anyone on the network, are to be successful as stories, they either need to be edited and structured by some subset of the contributors, or need to be structured by the system used to create the work. The larger the scale of the collaboration, the more important it is that contributors’ roles in the writing of the project are clearly defined, as are the constraints under which individual contributions should be written’ (Rettberg, 2005, p. 4). In my opinion, Rettberg’s view does not necessarily entail that the creative ferment on the participatory Web needs to be stopped. On the contrary, participatory creativity will probably flourish and yield better results by distinguishing those activities which can benefit from collaborative efforts and UGC from those which cannot. Having established that rules and constraints do not limit users’ creativity but rather allow for that creativity to be effectively directed toward the creation of a coherent narrative work, I would like to raise two further questions that may well constitute the starting point for future research. On the basis of the findings of the present work, I believe it important first to address the question of the relationships among the several co-authors of a collaborative narrative work against the background of thematic and stylistic coherence. It is also important to consider the intrinsic aesthetic and literary value of collaborative narrative works with a view to understanding how this is affected by both the medium (Web 2.0) and the specific modalities of creation (collaborative writing). Studying these problems would allow us to give an objective and unbiased answer to a question posed in the introduction and which in the light of the results of my analysis I would now reformulate as follows: Should we praise collaborative authoring only because it is based on UGC? Can we really call a collaborative novel successful merely because it has been completed, without calling into question its literary value? 63 BIBLIOGRAPHY Andretta, S. (2000-2009). Thematic analysis. Applied Information Research, 2009, from http://www.ilit.org/. Barlow, J. P. (1996, February 8, 1996). A Declaration of the Independence of Cyberspace. At http://homes.eff.org/~barlow/Declaration-Final.html Bonk, C. J. et al. (2009). The tensions of transformation in three cross-institutional wikibook projects, Internet and Higher Education. Retrieved from doi:10.1016/j.iheduc.2009.04.002. Boyd, D. (2008). None of this is Real. In J. Karaganis (Ed.), Structures of Participation in Digital Culture (New York: Social Science Research Council, pp. 132-157). Bryant, S. et al. (2005). Becoming Wikipedian: Transformation of participation in a collaborative online encyclopedia. Group '05. Cammaerts, B. (2008). Critiques on the Participatory Potentials of Web 2.0. Communication, Culture & Critique (1), 358–377. Castells, M. (2001). The Internet Galaxy: Oxford University Press. Charmaz, K. (2006). Constructing grounded theory: a practical guide through qualitative analysis. London ; Thousand Oaks, Calif.: Sage Publications. DiNucci, D. (1999). Fragmented Future. Print, 53(4), 32. Downes, S. (2007, November 27, 2007). An Introduction to Connective Knowledge. Paper presented at the Media, Knowledge & Education Conference – Exploring new Spaces, Relations and Dynamics in Digital Media Ecologies. Ezzy, D. (2002). Qualitative analysis: practice and innovation. London: Routledge. Foucault, M. (1994). What is an author? (Dits et Écrits, Vol. 1: 1954-1975, Paris: Gallimard). Gibbs, G., & Flick, U. (2007). Analyzing qualitative data. London: Sage Publications. 64 Giles, J. (2005). Internet encyclopaedias go head to head. Nature, 438, 900-901. Glaser, B. G. & Strauss, A. L. (1999). The discovery of grounded theory: strategies for qualitative research. New Brunswick, N.J.: Aldine. Gubrium, & Koro-Ljungberg. (2005). Contending With Border Making in the Social Constructionist Interview. Qualitative Inquiry, 11(5), 689-715. Hadjiafxendi, K. and MacKay, P. (Ed.). (2007). Authorship in Context - From the Theoretical to the Material. New York: Palgrave Macmillan. Hargittai, E. (2007). Whose Space? Differences among Users and Non-Users of Social Network Sites. Journal of Computer-Mediated Communication 13(1), article 14. Jenkins, H. (2006). Convergence Theory. New York University Press. Karaganis, J. (Ed.). (2008). Structures of Participation in Digital Culture. Columbia University Press. Keeble, L. (2003). Why create? A Critical Review of a Community Informatics Project. Journal of Computer-Mediated Communication, 8(3). Keen, A. (2008). The cult of the amateur: how blogs, MySpace, YouTube, and the rest of today's user-generated media are destroying our economy, our culture, and our values (Rev. ed.). London: Nicholas Brealey. Leadbeater, C. (2008). We-think: The power of mass creativity. London: Profile. Lessig, L. (1999). Code and other laws of cyberspace. New York: Basic Books. Lessig, L. (2007). Code: version 2.0 (2nd ed.). London: Basic Books. Levy, P. (1997). Collective Intelligence: Mankind’s Emerging World in Cyberspace. Boston: Cambridge. Lovink, G. (2005). The Principle of Notworking. Paper presented at the Inaugural speech, Hogeschool van Amsterdam. Retrieved from <http://www.networkcultures.org/>. Lowry, P. B. et al. (2003). Building a Taxonomy and Nomenclature of Collaborative Writing to Improve Interdisciplinary Research and Practice. The Journal of Business Communication, 41(1), 1-21. 65 Madden, M., Fox, S. (2006). Riding the Waves of “Web 2.0” - More than a buzzword, but still not easily defined, Pew Internet Project. Mason B. and Thomas, S. (2008). A Million Penguins Research Report: Institute of Creative Technologies, De Montfort University, Leicester. Nelson, T. (2003). Proposal for a Universal Electronic Publishing System and Archive: MIT Press, Cambridge. Nielsen, J. (2006). Participation Inequality: Encouraging More Users to Contribute. Alertbox. O'Reilly, T. (2005). What Is Web 2.0? Design Patterns and Business Models for the Next Generation of Software. From http://oreilly.com/pub/a/oreilly/tim/news/2005/09/30/what-is-web-20.html O’Neil, M. (2005). Weblogs and authority. Paper presented at the Blogtalk Downunder. Retrieved from http://incsub.org/blogtalk/?page_id=107" Preece, J. (2000). Online communities : supporting sociability, designing usability. New York: John Wiley. Preece, J., and Maloney-Krichmar, D. (2005). Online communities: Design, theory, and practice. Journal of Computer-Mediated Communication, 10 (4/1). Rettberg, S. (2005). All together now: Collective knowledge, collective narratives and architectures of participation. Paper presented at the DAC 2005. Rheingold, H. (1994). Virtual community: Finding connections in a computerised world. London: Secker & Warburg. Roth, C. (2007). ViableWikis: Struggle for Life in the Wikisphere. Paper presented at the WikiSym '07: International Symposium on Wikis, Montreal. Shapiro, A. L. (1999). The control revolution: How the Internet is putting individuals in charge and changing the world we know. New York: PublicAffairs. Shirky, C. (2008). Here Comes Everybody: The Power of Organizing Without Organizations: Penguin Group. Speck, B. W. (1999). Collaborative writing : an annotated bibliography. London: Greenwood Press. 66 Thomas, S., Joseph, C. et al. (2007). Transliteracy: Crossing divides. First Monday [Online], 12(12). Thurlow, C., Lengel, L. & Tomic, A. (2004). Computer mediated communication: Social interaction and the internet. London: Sage. Viegas, F. et al. (2007). Talk before you type: Coordination in Wikipedia. Paper presented at the 40th Hawaii International Conference on System Sciences. Wellburn E. B. (In press). Imagining multi-roles in Web 2.0 Distance Education. In G. Veletsianos (Ed.), Emerging technologies in distance education. Athabasca University Press. Williams, R. and Edge D. (1996). The social shaping of technology. Research Policy, 25, pp. 856-899. Willig, C. (2008). Introducing qualitative research in psychology: Adventures in theory and method (2nd ed.). Maidenhead: McGraw Hill/ Open University Press. Winner, L. (1986). Do Artefacts have Politics? The Whale and the Reactor: A Search for Limits in an Age of High Technology (pp. pp.19-39). Chicago: University of Chicago Press. Whitworth, A., (2009). Information Obesity. Oxford: Chandos. 67