Conversation: Online/Offline


I received my Branch invitation this weekend, and made my first contribution.

It hasn’t changed my initial assessment that Branch could end up being something rather different from what has come before in very interesting ways, but it did cause a few more ideas to bubble up in my head.

The most interesting ones have been around the characteristics of conversations.

I hadn’t given it much thought before, but face-to-face conversations (at least the worthwhile ones) are largely sequential: only one person can speak at a time, and the other participants are listening. Each contributor in the conversation must take in (or at least sit through) what is being said by others, and wait for the opportunity to make their contribution.

And in a face-to-face situation, the overall direction of a conversation is guided by consensus: if the group cannot achieve a rough consensus on where the conversation should go, that conversation ends.

Online conversations generally don’t follow this model. Whether it’s comments on a blog post or a discussion in a topical forum, multiple participants can and do respond to each contribution simultaneously. In the time it takes to type a thoughtful response, three other people may have already added to the conversation. Participants don’t necessarily have complete information about the overall direction of the conversation when adding their part to it.

There’s also no requirement for consensus on an online discussion. If the participants don’t agree on what the conversation should be about, each sub-group can continue “their” conversation in the same location — they all still share a single thread/post/what-have-you — but it’s no longer really one discussion with a single, directed flow.

It fascinates me that thus far, the Branches that have felt most worthwhile to me have hewed much more closely to the offline model of a conversation than the current online model: each contributor seems to be (metaphorically) waiting their turn and then adding to the conversation as a whole. They feel much more measured, perhaps even “slow,” than most online discussions.

But the catch is how to encourage and facilitate that experience online.

When offline elementary school teachers and management consultants encounter the kinds of issues that online conversations encounter, they often reach for the “talking stick” — you talk while you’ve got the stick, and when you’re done you hand it to the next person. It’s crude, but surprisingly often it can help.

Is the talking stick a metaphor that has value for a class of online discussions? It’s limiting, it can be frustrating or even feel demeaning, but it accomplishes something.

If the goal is to collaborate on creating a directed, coherent discussion online, do the individual participants need to accept some new constraints?

Branch and the Pillars of Social Orthodoxy


Stowe Boyd posted some thoughts on Medium this morning, which included this:

I submit that this early version of Medium is a speculative design intended to challenge us to consider implications of the deep philosophy lurking within, rather than the test of a fully fleshed minimally-viable-product.

It’s interesting to consider the philisophical implications of a product, but I think it’s more interesting to consider the implications of the existence of a product that’s intended to make us consider the philosophical implications of that product. [My theory: it’s early adopters, all the way down.]

But the actual point is that Mr. Boyd reminded me that I have recently been thinking about the philosophical implications of a product: Branch.

I haven’t yet gotten my Branch invitation so take my opinion for what it’s worth, but from what I can see it looks like Branch is breaking with the mainstream philosophy of the “social” web in some interesting ways.

When you look at the social web that has developed in the past five or so years, a rough orthodoxy has been established.

First, in the context of a given service, you “follow” other users that service. There’s a bit of a schism around asymetric vs. symetric following (Twitter or Foursquare model?), but the overarching concept of following — which was relatively new and still a little weird in 2007, mind — is the standard. Even services that started without it (think Disqus, Kickstarter) have moved towards accepting the follow as a core social mechanism.

The second pillar of orthodoxy is the completeness of these follow relationships. When you follow another user of the service, you get everything they put into it. You can’t get just the graffiti pictures that I post to Instagram and not the ones of my kids, nor can you see my music posts in your Tumblr dashboard and not all the other crap: you get the whole person, all the time.

And the final pillar (closely related to the second) is persistence: once you have followed another user of the service they stay followed until you actively remove them. While you are establishing relationships in the context of the service, they are not relationships that require active maintenance. I may never like, heart, star, or otherwise validate your presence on the service, but that doesn’t matter: having stated my interest in you once, I need do nothing else.

So how does this relate to Branch?

Well, Branch is obviously a very social service: you’re asking a group of people to have a discussion on a particular topic. You’re indicating to those people that you’re interested in them and that you value their opinions, which is a key part of most healthy relationships, on or offline. And the discussion itself is visible to a much larger universe: because invitation is required, participation in a Branch discussion is making your relationships visible.

But to accomplish this very social undertaking, Branch is almost completely ignoring the established conventions of the social game. The conversations themselves are put at the center, rather than the participants.

You don’t follow other users, you ask them (or are asked) to participate in a specific discussion. The relationship being created is very focused and explicitly bounded. You don’t have to read the other Branch discussions that someone has participated in, nor they yours. And the relationship has no persistence at all: each new conversation requires a new set of invitations, an explicit renewal.

It’s difficult for me to come up with a more contrarian approach to a social service, but it could actually work. For all the people complaining about social overload and trying to build tools to manage it, very few are doing anything that’s really different. Path limits the number of people involved, but adheres to the social orthodoxy. People build tools to filter, mine, timeshift, spindle, fold, and mutilate your social streams, but they rarely question the principles underlying the construction of those streams.

I could be wrong, but I think that the Branch folks are building something that’s legitimately a part of the social web, but proposing an alternative to the current model. Branch seems very different from Tumblr, or Quora, or Twitter, but I’m not sure that’s a problem. This could get very interesting.