(2015-04-25) Widner Problems With Rap Genius

Mark Widner (of LacunaStories) sees a number of problems with GeniusCom.

Like Wikipedia, Genius assumes that collaboratively-authored annotations can lead to a stable, monolithic truth about a text, a kind of crowdsourced exegesis. This model fails to respect the possibilities of divergent readings and collapses ambiguity into single narratives authored via consensus... One notable difference between Genius and Wikipedia is that the latter strives for "verifiability, not truth." Wikipedia, in fact, has a wealth of information about the philosophy and principles that guide the site’s editors; these policies have developed over time based on the experiences of the community. Genius provides no such mission statement or philosophical norms for its users, despite explicitly invoking Wikipedia as a model. Rather than distinguish between "verifiability" and "truth," for example, Genius claims that nature of online annotations are self-evident and arise from debate (i.e., consensus)... a simplistic up/down voting system is a naive way to manage a community of annotators.

Is it possible to annotate literature in a digital environment? Certainly, the answer is "yes." Is it possible to do so well and in a way that respects the ambiguity, complexity, and polyvocality of both the texts and the readers’ responses? If the annotations are, like a Wikipedia article, a single text composed through consensus, then I argue in this post that the answer is almost certainly "no." Yet, this model is the one that Genius wants to impose upon literature... EducationGenius maintains a hierarchical divide between teacher and learner... Instructors cannot put entire texts that are under copyright on the site. Instead, they can only use excerpts that meet Fair Use requirements... The entire site is, further, not designed as a reading environment where one can experiment with interpretations or take notes while reading, but as an explanatory resource somewhat akin to Cliffs Notes... Instead of nuanced interpretation that indicates the richness and ambiguity of these lines, we find explanation that aims to resolve complexity like it were a puzzle to be solved.

In recent years, corporations, non-profit foundations, university teams, and open source coders have developed a wide range of annotation technologies and platforms. The most widely used is AnnotatorJs, an open source Javascript library usable on any website and funded by the Shuttleworth Foundation and the Open Knowledge Foundation. A growing number of projects use Annotator, including HypothesIs, EdX, Annotation Studio, and Stanford’s own LacunaStories.. What this means for Lacuna Stories is that the people designing, building, and using it are all humanities scholars focused on classroom-based learning that centers around deep engagements with texts and other materials... Second, Lacuna Stories is different because we design our platform with faculty and students in mind, based on sound pedagogical practice and the above-mentioned ongoing research... Third and finally, we’re not trying to annotate the entire web.1 Instead, we’re focused on annotations in individual courses as a way to expose student reading practices to their instructors and classmates... On Lacuna Stories, there is no "right" annotation and no consensus interpretation... Each student can also select which audience (the student alone, the instructor, or classmates) should be able to see her/his annotation... they can also follow-up on existing annotations and begin a dialog (Discussion Forum)... we want students to make connections across annotations, texts, and other media that they’re studying... one could tie together annotations on any work that engaged with this theme by using Lacuna Stories’ "Sewing Kit" functionality... Threads thus become not only a tool to organize discussions, but also a new affordance for connecting one’s preliminary impressions to the "final" product of a traditional writing-based humanities course... Because, however, research shows that heavy HighLight-ers perform worse on several learning measures, we want each student to pause, write something, and reflect on what type of annotation he or she creates... Although we’ve been providing access to full texts of novels, poems, and other works to students, the process of clearing Copy Right is time-consuming and not scalable.