Annotation Summit at the New York Times

By jeremydean | 29 May, 2015

Poynter Annotation Summit leaflet.
Poynter Annotation Summit leaflet.

Together with the Poynter Institute, and with funding from the Knight Foundation and Craigconnects, Hypothesis hosted an “Annotation Summit” at The New York Times Building last week. The event brought together approximately fifty technologists, publishers, and writers from a range of publications including The New York Times, The Washington Post, The LA Times, the New Yorker, First Look, Time Magazine and others. Our goal was to bring these industry leaders up to speed on the state of the technology and the standards efforts afoot, and to learn from them how annotation might play a role in the future of online journalism.

Silde from Dr. John Unsworth 's presentation.
Silde from Dr. John Unsworth ‘s presentation.

Dr. John Unsworth, Vice Provost, Chief Information Officer, University Librarian, and Professor of English at Brandeis University (!), opened the Summit by offering a scholarly definition of annotation in its traditional forms. For Unsworth, those working on collaborative annotation must attend to this history, because, unlike other analog to digital metaphors (desktop, for example), the key motivations to annotate remain the same online as they do on paper. Given Unsworth’s academic focus, how widely these motivations are shared remains to be seen, but the primary standards organization for the Internet is working to develop protocols for web annotation in anticipation that such commentary could become a part of the everyday practice of “webizens.”

Home page for the w3c working group on web annotation.
Home page for the w3c working group on web annotation.

In his presentation, Doug Schepers announced that the World Wide Web Consortium (w3c) had a dedicated working group for annotation and gave an impassioned presentation on the importance of open standards for web annotation. For Schepers, and we here at hypothes.is agree with him, individual annotation services will fail to annotate the Internet; such functionality must be part of the underlying infrastructure of the Web itself. We don’t want a Facebook for annotation. But we do want the ability to bring relevant insights, from Facebook and elsewhere, to every page on the Internet. The good news for publishers is that in this vision, annotation would be a distributed feature, not something handled by a single site or provider. This question hung over the proceedings for the rest of the day: What would the consequences be (for users? for publishers?) if a single proprietary service was to emerge as common annotation application?

Next up, Bassey Etim (New York Times), Amy Hollyfield (PolitiFact), and Andy Carvin (First Look Media) provided a sense of the landscape of annotation within the publishing world. What was clear from the panel is that comments are broken in more ways than one. There was much talk of trolling throughout the day and whether annotation systems could save us from the Dark Tetrad of personality that trolls bring to the Internet (Greg Barber of The Washington Post brought up a study using this phrase/concept in relation to trolling). Etim also made the interesting point that bottom page comments are a particularly poor experience on mobile and of course any emergent comment/annotation system will have to be optimal on phones and tablets and, perhaps, watches. But it was not at all clear from the panel’s point of view that there is actually widespread demand for annotation among readers of mainstream media. In his experiments with annotation as a means of engagement, Carvin has simply not observed participation in great numbers. However, this isn’t an issue for PolitiFact, where Hollyfield is more interested in using annotation to bring expert commentary to primary documents in the news, like for example, speeches and other artifacts from the 2016 election. Both Genius and Hypothesis have shown that expert annotation is a clear value add to news content.

PolitiFact annotating the 2015 State of the Union using Genius.
PolitiFact annotating the 2015 State of the Union using Genius.

This conversation was followed by “Ignite” demos by a variety of annotation providers. Two services not usually associated with annotation per se, LiveFyre and Instapaper, were on hand to discuss the practice nonetheless. LiveFyre does have a new(ish) annotation product in their suite of commentary tools for publishers: SideNote. It was particularly gratifying to hear how versed Benjamin Goering of LiveFyre was about the open standards movement in annotation–he indicated that once the standards were completed by the w3c, he planned to advocate for LiveFyre’s participation. It remains unclear, however, how private models like that of LiveFyre will interact with a public annotation on the Web. Brian Donohue of Instapaper noted that ever since the introduction of a highlighting tool to the mobile bookmarking application, their users have been clamoring for annotation as an added feature. Stay tuned for more from them.

Annotating on Medium.
Annotating on Medium.

By the time Stephen Levy of Medium took the stage, his platform’s slick annotation functionality had already been invoked more than once. There was a broad sentiment at the summit that for annotation to succeed in journalism at scale, journalists would have to become more involved in commenting and responding to comments on their articles, and Medium systematizes this by giving author control over what annotations appear on their articles publicly. If Schepers had said something to the effect of, “Medium annotation is great, but it’s only for Medium,” Christopher Glazek of Genius essentially made the argument that Genius’s “beta” offsite annotation functionality was Medium for every page on the Internet. Glazek, a journalist himself, shared an annotated, “DVD extras”  version of his own New York Times Magazine piece, which included annotations from the subject of the article and a larger community of commenters (note: Genius has currently turned off their proxy service on the Times so you can’t view the annotated article that way). With very much the same technology as Hypothesis (extension, bookmarklet, proxy), Genius, in Glazek’s words, wants to be both Facebook and Wikipedia for every page on the Internet–both a commentary and an encyclopedia layer for the Web.

In his presentation, our own Dan Whaley closely aligned the vision of Hypothesis with the open annotation movement, historicizing its origins at Los Alamos through the establishment of AnnotatorJS as the foundation for shared standards in annotation. For Whaley, as for Schepers, the origins of the web are where the lessons must be learned for annotation: the model for an interoperable ecosystem of annotation is HTTP . Dan also showcased several projects using the Hypothesis client both in journalism and education. And he shared a bit of our product roadmap, particularly the imminent launch of a private groups feature.

The Coral Project's blog on Medium.
The Coral Project’s blog on Medium.

Perhaps the biggest buzz at the event was inspired by the presence of the Coral Project, Mozilla’s Knight Foundation-funded project in partnership with The New York Times and The Washington Post to “improve community” on news sites through open source software. In his presentation, Greg Barber of The Post made it clear that improving community was really about “civil discussion” (or in other words, silencing trolls) and so this effort could very likely involve annotation among the federated series of apps the project plans to produce. Andrew Losowsky, the new Project Lead for Coral, was also on hand, listening attentively and proving with his contributions that Coral is thinking about these problems in the right way. The folks from Coral Project were certainly eager participants in the day, hoping to learn from previous efforts to improve comment culture. Hopefully they will build on the infrastructure developed through the open annotation movement.

Bringing writers, publishers, and technologists together in one room to discuss the future of online commentary, the Poynter Annotation Summit no doubt marks an important moment in the history of web annotation particularly as it relates to journalism, perhaps the most fruitful field for this emergent technology. It was clear that in addition to important conversations that were had, many key relationships were begun that will likely become collaborative partnerships in building the new infrastructures of Internet comment culture. Though big questions about moderation and reputation remain, if, as Jeff Jarvis said at some point, “the problem with comments is in their structure,” leaving the public as an afterthought to the journalistic process, then the inline commentary of annotation could indeed be a potential solution.

Dan Gilmor, one of the final respondents who offered their thoughts on the proceedings of the day, took this idea further, asking the participants to think not only about annotation as an “end product”–that is as post-publication commentary–but as “part of the process” of news publishing itself. What, for example, is the role of annotation in news gathering? Could we imagine new production as iterative, including the annotations of the public during the process or journalistic composition? And how might annotations become news themselves, that is embedded within articles by journalists? It seems a shock to the system to imagine writers and readers as collaborators in the production of the news, but perhaps its the revival that online journalism needs.

Share this article