Highlighted Selections from:

The Remixing Dilemma: the Trade-Off Between Generativity and Originality

DOI: 10.1177/0002764212469359

Hill, B M, and A Monroy-Hernandez. “The Remixing Dilemma: the Trade-Off Between Generativity and Originality.” American Behavioral Scientist 57.5 (2013): 643–663. Pre-edited draft. Web.

p.643: In this paper we argue that there is a trade-off between generativity and originality in online communities that support open collaboration. We build on foundational theoretical work in peer production to formulate and test a series of hypotheses suggesting that the generativity of creative works is associated with moderate complexity, prominent authors, and cumulativeness. We also formulate and test three hypotheses that these qualities are associated with decreased originality in resulting derivatives. Our analysis uses a rich data set from the Scratch Online Community – a large website where young people openly share and remix animations and video games. We discuss the implications of this trade-off for the design of peer production systems that support amateur creativity. -- Highlighted apr 19, 2014

p.643: Lessig (2008) has argued that remixing reflects both a broad cultural shift spurred by the Internet and a source of enormous creative potential. -- Highlighted apr 19, 2014

p.643: Benkler (2006) has placed remixing at the core of “peer production” – the organizational form behind free and open source software and articles on Wikipedia – and has argued for the deep cultural importance of remixing. -- Highlighted apr 19, 2014

p.644: For example, Zittrain (2008) argues that some technologies, like the Internet, are generative and important not because they solve problems directly but because they provide rich and unconstrained platforms upon which derivative technologies can be built. -- Highlighted apr 19, 2014

p.644: Jenkins et al. (2006) have argued that educators can work to increase remixing behavior in young people. -- Highlighted apr 19, 2014

p.644: Although remixes are defined by their derivative nature, the promise of remixing is contingent on the originality of derivative works. -- Highlighted apr 19, 2014

p.644: We also know users of remixing communities react negatively to visibly similar remixes of their projects (Hill et al., 2010). -- Highlighted apr 19, 2014

p.644: Moreover, issues of originality are often at the center of moral and legal discussions of remixing (Aufderheide and Jaszi, 2011). -- Highlighted apr 19, 2014

p.645: We attempt to answer two related research questions. First, what makes some creative works more generative than others? Second, what makes some creative works engender more transformative derivatives? -- Highlighted apr 19, 2014

p.645: We suggest that three factors associated with higher levels of generativity – moderate complexity, creator prominence, and cumulativeness – are also associated with decreased originality in the resulting remixes. -- Highlighted apr 19, 2014

p.645: Using data from Scratch – a large online remixing community where young people build, share, and collaborate on interactive animations and video games – we present evidence that supports and extends several widely held theories about the foundations of generativity and originality. Our results suggest that designers of online collaborative communities may face a dilemma obscured by those celebrated exemplars of peer production communities: that system designs that encourage and support increased rates of remixing may also result in more superficial products. -- Highlighted apr 19, 2014

p.646: Zittrain (2008) posits the “Principle of Procrastination” that proposes that generative technologies tend to be designed in a way that leaves most details for later saying, “generative systems are built on the notion that they are never fully complete, that they have many uses yet to be conceived of, and that the public can be trusted to invent and share good uses.” For example, Zittrain suggests that the Internet was a more effective platform for innovation than corporate networks like Prodigy and Compuserv because its relative simplicity offered fewer constraints for potential innovators. -- Highlighted apr 19, 2014

p.647: In Lessig’s account, the act of remixing is often understood as a social statement of parody or critique. Jenkins (2008) documents how youth use fan fiction to create remixes of popular and culturally salient products and symbols. -- Highlighted apr 19, 2014

p.647: Cheliotis and Yew (2009), Healy and Schussman (2003), and others have shown that activity in remixing communities is distributed unequally and that only a very small number of peer production projects incorporate the work of a large number of individuals building on one another’s efforts. The majority of efforts are largely, or even entirely, uncollaborative. -- Highlighted apr 19, 2014

p.647: Cheliotis and Yew (2009) have suggested that highly unequal rates of collaboration among projects in the ccMixter community is driven by a process of “preferential attachment” (Barabási and Albert, 1999) or cumulative advantage (DiPrete and Eirich, 2006) where, “works exhibiting a high degree of reuse become more attractive for further reuse.” -- Highlighted apr 19, 2014

p.648: we hypothesized that one possible mechanism behind “release early, release often” and the Principle of Procrastination is that simple projects are easier for new contributors to build on. -- Highlighted apr 19, 2014

p.650: Scratch was designed by the Lifelong Kindergarten Group at the MIT Media Lab as a platform for constructionist learning (e.g. Papert, 1980) and aims to introduce young people to computer programming. Scholars have located much of the practice and promise of remixing in communities of young technology users (e.g., Lessig, 2008; Jenkins, 2008; Palfrey and Gasser, 2008). With a large community of young users, Scratch represents an ideal platform to study remixing. -- Highlighted apr 19, 2014

p.650: As of April 2012, more than one million users had created accounts on the website and more than one third of these users had shared at least one of more than 2.3 million total projects. As the only web community built around sharing Scratch projects, the community contains virtually all Scratch projects shared online. -- Highlighted apr 19, 2014

p.651: A majority of the community’s users self-report their ages ranging between 8 to 17 years old with 13 being the median age for new accounts. Thirty-five percent of users of the online community self-report as female. -- Highlighted apr 19, 2014

p.651: Influenced by theories of constructionist learning in communities (Papert, 1980) and communities of practice (Lave andWenger, 1991), the community seeks to help users learn through exposure to, and engagement with one another’s work. The commitment to remixing is deep and visible in Scratch. -- Highlighted apr 19, 2014

p.653: Issues of generativity and originality play a prominent role in the Scratch community. Although Scratch is designed as a remixing site, experience suggests that only around one tenth of all projects of Scratch projects are likely to be ever be remixed. Approximately 2% of remixes are flagged as inappropriate – often with accusations of unoriginality like, “This is MY own artwork he has uploaded without an ounce of originality.” On the other hand, Scratch creators often explicitly encourage others to remix their works and even request help creating new features and solving bugs. Scratch users frequently respond to these requests with remixes but also frequently remix without prompting or communication. -- Highlighted apr 19, 2014

p.663: Although we cannot speak to the generalizability of these results to other remixing communities or peer production projects, we believe that remixing in Scratch provides insight into the behavior of young creators more generally. The degree to which these results will generalize to adults, to other communities, or to activities other than the creation of animations and games, remain largely open questions for future research. -- Highlighted apr 19, 2014

p.663: This paper’s primary contribution for system design theory is the proposal of a critical trade-off between the quality and quantity of remixes. To the extent that these results generalize, designers may need to trade-off deeper remixing with increased collaboration. -- Highlighted apr 19, 2014

p.664: For example, designers of a new peer-production system in need of more content might want to build features that further emphasize the salience of author prominence and remix “chains” in order to encourage generative content. However, our findings suggest that these designs might come at a cost in terms of the originality of the derivative works. -- Highlighted apr 19, 2014

p.664: Surprisingly, our weakest support is for the hypotheses about complexity that stem from our elaboration of Zittrain’s “Principle of Procrastination” and Raymond’s exhortation to “release early and release often” – the most widely cited theories of generativity. We find support for our hypothesis that the most complex projects will be less generative than projects of moderate complexity, but only if we consider the very most complex examples. In general, we find largely positive relationships between complexity and both generativity and originality over most of our data. This may point in the direction of one potential solution to the “remixing dilemma” we propose. It may also be that the young users of Scratch are unlikely to create projects that are complex enough to trigger the effect suggested by theory. It may also be that complexity is simply a poor measure of completeness, earliness, or open-endedness as it is theorized by Zittrain and Raymond. More research is needed to clarify this relationship. -- Highlighted apr 19, 2014