Conversation curation

In a couple of good posts, JP Rangaswami reflects on the need and opportunity for democratized curation. He cites Google CEO Eric Schmidt quantifying the incredible amount of information being generated on the internet – these days, 5 exabytes of information is created every two days, as much as all the information created between the dawn of civilisation and 2003. JP writes about the need for curation of text, music, image, and video. I’d like to focus on a new opportunity – curating conversation.

The last few years has seen the rise of the realtime web, so-called status updates in Facebook, Twitter and other services, much of which is really conversation. The stream flies by quickly. If you missed it, it’s gone. Search of stream content is getting better, but even so, if you find a single message, you don’t really get the gist of a conversation. This is where curation comes in. This is different but closely related to “tummeling”, which is the art of facilitating a live conversation in process. Conversation curation is the art of representing and summarizing a conversation, so others can see it later, and the conversation can pick up again from a new starting point.

Conversational curation isn’t needed or wanted for many conversations – sometimes the conversation is truly transient – for example, nobody needs an edited record of people cheering their team through a hockey championship. But sometimes conversation does have longer-lasting value. For example, there was a fascinating Twitter conversation between Howard Rheingold and his Twitter followers about attention and distraction. This discussion contained information and arguments that seemed worth preserving, so I wrote it up as a post, which has continued to get references well after the original discussion. People have been using the practice of summarizing conversations in mailing lists and forums for years. The realtime web makes this practice more important because conversations can be even more transient and hard to piece together without a curated record.

There are some very old, pre-modern examples of the form of curated conversation – found in the Talmud and, I’m told, ancient Chinese traditions also. In the Jewish tradition, the form of curated conversation comes from attempting to preserve some of the texture of an oral tradition of dialog and debate, as that tradition was being represented in written form.

This is one of the reasons why I’ve been interested recently in modern takes on the representation of multi-voiced discourse in ancient works – because I think that this old form has lessons for a new need in quite a different cultural context. An edited conversation, with multiple voices assembled by an editor, is not identical as a live conversation in which participants speak for themselves. Scholars looking at the old forms debate how much the edited conversation is actually conversational. Daniel Boyarin argues, building on Bakhtin, that the editor’s hand smooths out differences in the represented voices and turns the dialog into a monolog. But David Frank contrasts the dialog in Plato, where the conversational partner is represented merely as a foil to reach a foregone conclusion, with dialog in the Talmud, where the different voices carry different ideas, and the whole picture includes multiple voices.

Another distinction – and something that may be important for the future genre – is how readers are brought into the picture. With the Talmud, says Marc-Alain Ouaknin, the dialog is represented – and culturally presented – in a way such that readers are drawn in to converse together in realtime to carry on the conversation, in debate with each other, adding their own contributions. By contrast, in Socratic dialog, the reader is expected to understand, assimilate, and agree with the presented conclusions.

In a new book that looks at these ancient forms of represented dialog (that comes to different conclusions than David Frank, and than I do agreeing with Frank), Daniel Boyarin makes an important point. Representing a conversation doesn’t freeze it, it just pauses it. The transition between speech and writing is a repeated cycle – “written culture becomes transmuted into oral culture and then back… over and over and over again.” Part of the form of curating conversation will be representing it in a way that people will find it welcoming and interesting to continue the conversation in realtime, and continue the cycle again.

Another important difference from the pre-modern forms is the boundary of the conversation. Daniel Boyarin notes astutely that the conversation represented in the Talmud is open with respect to ideas seen as within the community of the Talmud’s rabbis, but closed with respect to ideas seen as outside that framework. In modern settings, people create boundaries for conversations in very different ways – but those boundaries still exist, often as informal social norms. In communities of fan fiction, participants decide what works fit into the canon they will remix. In political communities, participants decide which opinions are legitimate for debate in a given community, and which positions are out of bounds. The editors / curators will play key and controversial roles in maintaining these norms.

There are some emerging technical components that will make the practice of curating conversation easier – activitystrea.ms to conduct conversation across services, and Salmon to pull together the comments. Plus, perhaps, there is a need for visual editing tools to pull the pieces of a conversation together.

In the world JP Rangaswami envisions, where curation is an important part of improving the ratio of signal to noise, conversational curation will be an important art, and the cycle between live conversation and the edited representation of dialog will become important once again.