Filter, Feed and Funnel: Social media participation

Nearly six weeks ago I promised to post these speaking notes from FutureEverything, and now that I’m at another conference doing another talk, I finally have.

None of these ideas are really new – what I wanted to do with this piece was think through some of the complexities of participation in what I call “the politics OF the network.”  It was a fun talk to give with a great audience, and here is, more or less, what I said:


This conference is about the future, and I think, in an unspoken way, about technology’s impact on the future. I want to shift our attention, for a few minutes, to the past. Don’t worry, I’ll get back to the present soon, and maybe even to the future. But I’m skeptical of presuming that the past has nothing to tell us, and that we should stride forward in the expectation of perpetual progress. There are many thing about our present social media landscape that are different than what we’ve experienced in the past. People get much more information much more easily, and this information is mediated in very different ways than it was in the past. They have faster access to other people as well. And an increasing number of people have access to technical tools that they can build and change in order to take advantage of these other factors (what social scientists call “affordances”). So as citizens, we in the privileged West are in a position to share information (which we do at an unprecedented rate) but also to collaborate to make change.

In any case, to start out I’m going to talk about our historical models for citizenship, and the media spaces that they are associated with. Then I’m going to talk about the media spaces of the present, and the way that filter, feed and funnel shape our opportunities for participation. I’ll talk about some of the problems of social media participation, and then suggest things we can think about – and DO – to use the opportunity that our networked communication provides.

Part 1: Our historical models for citizenship: Spaces and Media

It’s only been since the 19th century that westerners have had an understanding of people outside of the elite as citizens, who could discuss and debate opinions about how the world operates. I’m sure many of you are familiar with the concept of the “public sphere” which is the space of democratic discussions – deliberative democracy, if you will.

1. In the beginning, there was the cafe, and the newspaper. This was the Habermasian model for deliberative democracy. Guys get together and talk about the news of the day. Sometimes their voices amplified by the press, and then perhaps a response to the press from the elected representatives. The emergence of the press was huge! All of a sudden people were aware of the decisions elites were making, and able to read comments on them and discuss.

2. Then there was the street, and the zine, which are spaces and media more associated with Nancy Fraser’s view of the public sphere as also producing counterpublics. Obviously this conception of the “public sphere” was a very limited one: it didn’t include the resistance, all those counterpublics with different ideas. So there were other ways of occupying the street (public spaces) and alternative ways of producing media.

3. There are technological publics too, as Christopher Kelty (and others) describe. The development of free software licenses has meant that code can also act as a way of deliberating the issues of the day. GNU public license stipulated that the software could be freely copied, modified and distributed that every piece of software using the license had to also be subject to the license.

Part 2: Social media models for citizenship

Ok, so now we have a networked set of publics, supported by social media. The great thing about social media is that it’s a set of functions that can work in all kinds of different ways. It’s not the newspaper with the public opinion. It’s not the radical zine. It’s both, and it’s more. I’m going to talk about 3 things that social media does that are significant for participation, and some examples of how they contribute to citizenship

Filter – if we think about the movement from the cafe and the street to the network, what’s one of the most significant media transformations? Quite simply, the end of scarcity of information. Now we have the opposite problem of the early newspaper reporters. We have to make sense of this. That’s where the filter logic comes in. We choose what to attend to, determine what conversations to respond to. It’s no longer a mass media situation, instead its an algorithmically sorted feed of information that can also be used participatively.

Feed – remember the cafe? The feed (Twitter, Facebook) is the public sphere we choose, or, as Eli Pariser points out, is chosen for us based on algorithms that our participation determines. The level of deliberation on feeds and blogs rivals that of the public sphere, but it’s a public that’s chosen and refined

Funnel – This is where individual practices change the larger structures. The difference between previous public spheres and media was that the power coming from the top down was not met by the power of the resistance. Now we have the ability to amplify our views – although we’re not always perfectly in control of how they are amplified. Various projects like MySociety’s Fix my street aim to take advantage of this possibility for amplification.We also have opportunities to use the features of social media progressively to work together. What useful things could we do with the funnel? How about, monitoring air and water quality locally? Using motion sensors to map safe cycling routes from the perspectives of cyclists?

Part 3: Dark sides. The echo chamber

1. social media networks can be elitist publics

When we think about the publics that are made easier by social media, we have to keep in mind that filter, feed and funnel are ways of connecting data to participation. We also have to understand the complexities of the situation.  Young people are less likely to use Twitter than adults aged 25-40 (although teenage girls are an exception). They are also less likely to blog. This finding should remind us that participating in social media is not a unified experience. The relationships that committed Twitterers of a certain age construct (your author included) may be more representative of our age and demographic than indicative of social media itself.

In other words, a feed of people you’ve chosen is a public, and it can be full of exciting political discussion. But it might be just an elite a space as the cafe in the 19th century. Furthermore, Facebook feeds are full of people with whom you have reciprocity, while Twitter does not. At the moment we have a pretty narrow set of opportunities for engagement.

  1. Funnel processes means that we are generating data, and value for others.

The processes of participation offered by social media mean that we can amplify and aggregate our views and our data and provide them to those in power. They are powerful but the question is, who is in control? Centralized social networks like Facebook provide platforms for engagement, but their cost is that they collect an enormous amount of information.

Our relationships with social media infrastructures influence our participation in ways that aren’t easy to see. Filters are part of what makes the platforms and infrastructure opaque. We don’t see the algorithms that sort our relationships.

  1. We create relationships with infrastructures

One big difference between social media participation and participation in other types of politics ishow these applications are now becoming infrastructures for participation. To understand them, we need to know more about how they are built, how they work, and who controls them. Yes, we want to make things together, and we want to make relationships with people. It’s easier to do this using applications like Facebook Twitter, and YouTube. But this also means creating a relationship with the platform itself. The algorithms to which we’ve delegated the work of connecting and communicating also have agency. We don’t know much about them and their relationship to our participation is opaque. Sometimes, we get a small view into the algorithms of certain systems – but generally we have so little understanding of the ways that our participation is mediated through the experiences of filter, feed and funnel.

The question of social media use and agency is not just a question of knowing or being able to understand the design process. If different generations or social groups want to relate to each other in different ways, then there’s social interest in understanding how different infrastructures shape and are shaped by those relationships. But we are participating more – and this needs to be more progressive than simply turning icons green, or saving the Brazilian Galvao (see Ethan Zuckerman on this inside joke).

Part 4: So what do we do?


  1. build our own infrastructures. This is the most direct form of participation. There are lots of good examples of this, from community broadband infrastructures to local wireless networks like Ile Sans Fil in my hometown Montreal or the Friefunk network in Berlin. Hardware hacking and the developoment of open source hardware is also part of this participation. These are access networks but they are also participation networks – getting people together in real places to apply technology to a problem. And local infrastructures can turn around feed, filter and funnel. Local WiFi hotspots can provide information using the location as a filter, inviting engagement in local art or in local politics. But they are again limited by the insularity of their
  2. Open the code: local networks are made possible by both free software and by the network. Now there is recognition that centralized social networks are using funnel to generate too much value for too few people.  Crabgrass and possibly the Diaspora project (when it’s built) are platforms that use the beneficial affordances of social networks but are based on the principles of GNU Social, and on individual control of privacy.
  3. Figure out how feed, filter and funnel can work transformatively. It’s tempting to argue that “we” the technically-savvy should remake social media, but network effects mean that common platforms will be used by more people, so creating an alternative infrastructure may not be the most effective way of working for transformation. Participation means everyone – and the logic of social media makes it harder for everyone to be heard The challenge is to make things BESIDES the tools open – like organizations, innovation cycles, and structures of participation.