When it began in 2000, the Transliteracies Project took on the broad topic of “the technological, social, and cultural practices of online reading.” Several years of work now leads Transliteracies to focus on a specific, high-value research-and-development direction—one that positions online reading as a new kind of research activity positioned in a sweet spot between academic and mainstream information practices. This direction is challenging enough to drive next-generation research on online reading environments, and is distinct from (even as it complements) related research projects.
Transliteracies is building RoSE, a research-oriented social network. RoSE is a combined intellectual and practical experiment—a trial balloon—designed to allow interdisciplinary researchers to think about the future of online research when it will overlap with such online activities blurring the divide between agency and reception as reading, searching, sampling, bookmarking, linking, annotating, commenting, authoring, friending, and following (as in a Twitter “follower”). Currently, RoSE is being developed with seed funding as a working platform (in a Ruby on Rails implementation, wrapped around a relational database) that is stocked with a representative set of metadata about people and documents and a sample set of interfaces, textual outputs, data visualizations, and other end-user experiences.
While RoSE does not itself contain the full texts of documents and other primary resources, it will be able to link to such resources, thus allowing it to partner with major data repositories—e.g. the REKn or EBBA archives of Renaissance materials assembled by two of RoSE’s partners: the Electronic Textual Cultures Lab at the University of Victoria and the English Broadside Ballad Archive at UC Santa Barbara. RoSE is thus positioned to be a next-generation social-network research platform that can sit on top of—or, though its externally accessible XML datapath, provide social data to—external repositories that otherwise have only limited, ad hoc, or imitative social-network dimensions (e.g., a digital archive with folksonimical “tagging” added on the side).
At present, the limited goal of RoSE is to be a system robust enough to suggest what is possible, but still malleable enough to be open to critique and revision. It is a platform for thinking about next-generation online reading/research.
The Main Idea:
Underlying RoSE is the question: what do academic, professional, or lay users (“lay” meaning people who, whatever their expertise, are starting research on something outside their expertise) want when they begin researching online? RoSE’s answer is that people seeking knowledge do not necessarily want to go to either a document (a “document-centric” approach) or a person (a “social- network” approach) as their first point of access—though they will take either. More ideal is an online environment that allows them to seek out documents and people in the context of relationships between the two (e.g., of authorship, reception, affiliation, recommendation, sponsorship, commentary, rebuttal, etc.). In such an environment, there would be no documents sitting in virtual “libraries” as opposed to people joining “communities.” Instead, documents, authors, editors, publishers, readers, annotators, and other documents will be interlinked in combined, dynamic orbits of knowledge. Such an online environment would reveal clustered, evolving relationships between people-and-documents, people-and-people, and documents-and-documents (as well as, recursively, people-and-’people-and-documents,’ etc.), where the and in these formulations is some variable relationship more fine-grained than the standard “author of” or “friend of.” Facilitated by data-mining and data-visualization embedded in RoSE, people doing research would thus be able to see, and move, through the dense topographies and histories of knowledge—to say, in essence, “aha! That’s how that thought-stream started and who/what the major players are. Bring me back to that point in the combined document-social graph—to that particular interaction of a person with a document—so I can drill down.”
What RoSe is Not:
Because it does not itself include full-text documents (though it can link to them), RoSE is not a digital document portal, text base, archive, edition, encyclopedia, or “professional reading environment” (e.g., “PreE,” the name of the project being built by its Electronic Textual Cultures Lab [ETCL] partner)—though it complements such other resources and paradigms. (Recently, Transliteracies collaborated with the ETCL and EBBA on a proposal for the NEH/SSHRC/JISC multinational “Digging into Data” challenge grant competition, which, if successful, would allow ProSE to be integrated in add-on fashion with the REKn knowledgebase, the EBBA archive, and the PreE professional reading and annotation environment.)
Because it sees both documents and people as active knowledge agents, RoSE is also not just a social-network system like Facebook, LinkedIn, Academia.edu, and so on (though it is also such a network). The most prevalent mainstream, professional, or academic social-network systems focus on people and their social relations (friends, contacts). While they often also allow users to upload or link to fuller documents, the resulting documents are poorly, if at all, integrated in the social system (they are just “baggage” carried around by people)—even if, in reality, some documents do have meaningful social agency (their own “friends,” and so on).
What RoSE Is:
RoSE is a system for tracking and integrating relations between people and documents (as well as groups) in a combined “social-document graph” (not just a “social graph”). It allows users to learn from the relationships between people-and-documents, people-and-people, and documents-and-documents even before, or simultaneous with, drilling down to full-text and full-profile-page resources (through future links to external repositories).
Unique features of RoSE include:
- RoSE accommodates fine-grained relationship types. Not every person or every document is just a “friend” or “digg.” RoSE’s fluid, extensible typology of relationships accommodates the complexity of negotiations between local and global knowledges today, not to mention the rain forest of historical knowledges. (By contrast, how do mainstream social-network systems deal with people from other parts of the world or from the distant past who do not follow normative naming conventions and who are related to each other by standards foreign to today’s developed-nation typologies of social relations—e.g., “tribe of,” “house of,” “apprentice of,” “slave of,” “patron of,” and so on?)
- RoSE’s social network includes historical or “dead” people. Confucius, Queen Elizabeth I, Shakespeare, Wordsworth, and Einstein, for example (as well as people less canonical), will have profile pages and “walls.” After all, the nature of true research is that it does not respect a natural divide between the living and the dead. Its “longitudinal” horizon is much more extensive. The dead not only continue to influence living knowledge but, reciprocally, themselves seem to change as a result of later research.
- Though not possible without implementation-scale funding (or partnering with other projects), RoSE envisions allowing the above-described features—and other emergent features—to be enhanced through automated data mining. Ideally, the project would be able to data mine primary documents and such secondary resources as WorldCat, National Dictionary of Biography, Wikipedia, etc., to harvest initial person-and-person, person-and-document, and document-and-document metadata. Beyond harvesting elementary metadata, a full-on data-mining approach would allow for the following capstone feature. It would simulate agency on the profile pages and “walls” of historical people (as well as of such influential documents as major religious, governmental, literary, or other works). People and documents from the past would appear online as what in reality they still are: intellectually alive. Imagine, for instance, that Shakespeare’s page could be refreshed with content drawn from his writings or those of his followers, scholars, etc., in response to breaking new research or recent news events. Or imagine that the U.S. Constitution had its own profile page and wall, so that yesterday’s Supreme Court decision could trigger on the Constitution’s page—through data-mining of related people and documents—“learned commentary” about that decision.
You want to know something. You look in RoSE to spot a cluster of evolving relationships (whether of authorship, trust, influence, controversy, etc.) between the people and the documents that seems to have defined the current knowledge-scape on that topic. So you go there—to that hive of knowledge relationships—to learn. If you want, you can move the time slider back and forward to see how the relationships looked then and now. Then it comes clear: what you know and still do not know, and who or what next to research, to read.