Last week, about 100 technologists, hackers, publishers, scientists, scholars, and librarians from around the world gathered at Fort Mason in San Francisco, CA for the Andrew W. Mellon Foundation funded I Annotate conference. Hosted by Hypothesis, the conference focused on the twin objectives of showcasing progress in the annotation toolbox and exploring how to jumpstart annotation among diverse communities of interest.
One of our attendees, David Streitfeld from the New York Times, wrote an excellent piece that captured the essence of things– “Speak up, the Internet can’t hear you.”
Twenty years after Marc Andreeson and Eric Bina briefly incorporated group annotation into early builds of the Mosaic web browser, the workshop highlighted the substantial progress towards and interest in widely implemented annotation–with a particular focus on efforts leveraging the nascent W3C Open Annotation standard. Many of the principal architects of annotation systems were gathered in the same room, such as the Andreesen-Horowitz funded RapGenius, Domeo, Diigo, Hypothes.is, Annotator, Annotorious and HyperStudio. In addition, key stakeholders were present from domains of application including evolutionary science, medieval manuscripts, religious studies, journalism, astronomy and open government. The wide ranging mix fostered the cross-fertilization of ideas and opportunities; enriched understanding of the insights and shortcomings of past efforts; and enabled participants to establish groundwork for collaborations that would otherwise never have existed. A final day of the event was dedicated to a hackathon.
The workshop agenda was designed to provide a balance of short talks followed by an afternoon of “unconference” sessions organized by attendees and catalyzed by topics from the morning. The first day focused on how we annotate, starting off with a discussion of software stacks such as Hypothes.is, OKFN’s Annotator, and Domeo. These talks were succeeded by discussions of annotation applications for manuscripts, maps, classroom and educational uses, and archaeological inquiry. Two notable talks on resistance to annotation were offered by Simeon Warner of arXiv, and Jennifer Lin of PLoS, who talked about the challenges in adoption and reception that annotation efforts face, and the difficulties of capturing the most constructive forms of criticism and discussion.
The second day was spent in discussions of what we might annotate, ranging from scientific publications to datasets; automated annotation of chemical formula to images of biological materials; and readers’ annotation of ebooks, as well as the use of shared, collaborative annotation to aid their writing and editing. While some of these talks accentuated successful applications of annotation workflows, often involving combinations of manual note-taking with digital content management, many of them stressed the gap between current practices and the potential for richer commentary and engagement. In domains such as journalism or open government, the extension of democratic participation and expert analysis into the web is largely lacking but of obvious need. Much enhancement of online media is currently managed by the organization controlling the material, whether publisher or museum, and the addition of controlled metadata serves as a proxy for annotation. A more open annotation environment, with reputation and identity support, would provide for a far richer contextualization of digital objects and texts, providing not only descriptive information but interpretative elements that are otherwise usually missing.
A wide range of insights were offered. Ian Mulvany of the new open access journal, eLife, noted that “Science sucks at getting better at science” – an observation on process and practice that is true across all human knowledge production systems. Reaching greater understanding of our world through a better grasp of truth and conflict remains cognitively difficult. The availability and promise of networked technologies means that we have to not merely refactor, but rather re-imagine, how we “do” science, write journalism, draft laws, and knit together a global community of informed world citizens. As John Perry Barlow said on the second day’s opening talk, “Mankind needs metabolic ecosystems of meaning” – i.e., that we must build responsive online environments where engagement, identity, and reputation can help sort out responsible reasoning from idle chatter and demagoguery.
The final day was a hackathon, bringing together 45 software developers from journals such as eLife and PeerJ, online annotation systems, identity systems such as ORCID, online communities such as OpenGov, and the W3C annotation standards group. By the end of the day, participants wrote new builds, pushed commits, added features to existing code, and prototyped solutions problems identified the previous two days.
We appreciate the many pictures taken at I Annotate, such as those by Martin Kalfatovic; if other attendees images they’d like to share, please let one of us know. We would also like to call attention to the wonderful drawings of Matt Lumpkin. Jon Voss and many others have directly or indirectly contributed to a Pinboard compilation. Andrew Magliozzi put a hack day ideas and resources page together. The twitter hashtag of #ianno13 was used by many attendees, both in person and those following the livestream, and there is a google spreadsheet archive of all tweets, thanks to Tim McCormick. Finally, Hypothesis is evolving a Storify of the event, with tweets, videos, photos and other media elements.
We hope the workshop advanced our collective understanding not only of how to build an annotation layer on the web, but how to get people using it, once it is built.
Thanks to everyone who participated. We look forward to seeing you next year at I Annotate 2014!