KEYNOTE � Folksonomies: What Are They Good For?

Matt Locke, Head Of Innovation at BBC New Media, explored the issues around folksonomies as a type of user-generated classification and its relevance to content providers in the Web 2.0 era at the Content 2.0 conference on June 6th 2006...

Report by Deirdre Molloy

[Register and post your own comments on this article below...]

Download this session from the Content 2.0 Podcasts!

Stressing its early-days nature, Matt noted that in Google Trends "folksonomies" only registered on the radar in late 2005, although people have been writing about and working in this field for over 2 years.

Matt wanted to locate folksonomies in the broader context of classification. He cited the book �Sorting Things Out: Classification And Its Consequences� by Geoffrey Bowker and Susan Leigh Star as a good primer for anyone serious about the area. It was written in 1999 but by using very anecdotal case studies about how classifications have impacted on people, their social lives and communities, and it gives us a lot of valuable tools outside of a purely digital conversation about what tagging is as a technology. Matt wanted to talk about people, not technology, and he particularly wanted to stress the differences between and motivations of people, emphasising that we needed a more granular view of what we mean by users, consumers, prosumers, Pro-Ams, the people formerly called the audience, etc.

The book explores how classifications make things visible and invisible, and explores the case of nurses who got together to make a taxonomy classification system about their work that included a lot of things that their employers and mangers had overlooked. Bowker & Star used this idea of how the nurses used the vocabulary database that they developed (and what it both achieved and overlooked) in order to draw out three important factors in tension in any form of classification.

"Visibility is what folksonomies have really excelled at, they�ve made participation open to everyone"
- Matt Locke

There are 3 things in contention with folksonomies, Matt noted � comparability (the stated goal of most classification systems, in a way what Tim Berners-Lee talks about with the semantic web, the ability to transfer terms from context to context and for them still to have some kind of meaning). Control (which is implied in the use of a classification system) � it�s a continuum of classification, as soon as you have any form of classification over any community or any group of resources you are implying some kind of control over it.

What's more, comparability and control are always in tension. If you have an awful lot of control, you get a lot of comparability, but you also make a lot of things invisible. And the notion of visibility is where the politics lie. If comparability is a stated goal, and control is a reality, visibility is a kind of fantasy of classifications, it�s what every classification system sets out to do: to make the previously invisible, visible. Their ability to succeed in that is what this tripartite tension organises.

Visibility is what folksonomies have really excelled at, Matt observed. They�ve made participation open to everyone, lowering the barriers to entry radically in this sphere, which has hugely fore-grounded the role of visibility in the system. But as a result, comparability and control have become less important. The foregrounding of visibility also foregrounds play and other behaviours that we wouldn�t normally see in classification, and from this Matt extrapolates that folksonomies are only useful when there is nothing at stake. He stressed he didn�t mean this pejoratively, because the areas in which nothing is at stake are huge and lead you into some of the most interesting areas, like the world of social media.

"eBay would not use folksonomies as a form of reputation management."
- Matt Locke

Where folksonomies aren�t that useful is in areas where there are very, very hard transactions to get done. eBay would not use folksonomies as a form of reputation management as there are simply aren�t those factors of comparability and control to the level that you would expect them to have. However there are huge domains of information that we can categorise using folksonomies where nothing being at stake is part of our expectations of them. This applies generally this area of social media that we�re talking about: my pictures, my websites, my TV viewing habits, my music preferences, my community. These are gigantic spaces and spaces where folksonomies really excel.

But Matt still felt that these tensions are at play. And that arises from the problem of how the interface between environments where nothing is at stake interacts with environments where there are scarcities � scarcities of economics, scarcities of attention, etc. How do folksonomies interact with areas where there are scarcities? We don�t know quite where the edges are yet to these boundaries, and that�s part of what they�re trying to sort out in companies like Yahoo, Google, BBC and the like.

"Open standards are crucial to this notion of the �tearable web�"
- Matt Locke

He moved onto what motivates people to tag. Today�s webizens, Matt reckoned, are what Bruce Sterling calls �data wranglers� � they love wrangling data, linking things together, putting their profiles out there and fishing to see what kind of value they can get back; but he was unsure if this would become the majority view outside of early-adopters. However, people have been tagging things for years. The capacity to link objects from other places into one scrapbook goes back a century and Matt thought Marc Canter�s mention of open standards crucial to this notion of the �tearable web�. The Flickr badge is a perfect example of the tearable web, although overall it�s still some way off, he reckoned.

Referencing a document on motivation for tagging systems from the WWW Conference 2006 in Edinburgh http://www.rawsugar.com/www2006/29.pdf Matt moved onto these motivations. They are :
(1) organizational (filing) and (2) social (scrapbooking) and he broke these down into the following sub-motivations

(i) Future retrieval (as it exists in a contextual space, eg Delicious)
(ii) Contribution & sharing
(iii) Attract attention (seen in any social network)
(iv) Play & competition (as we saw with the espgame Bradley Hrotowitz referenced, potentially a very interesting and fruitful way of designing tagging systems)
(v) Self-presentation
(vi) Opinion expression (value-based judgements)

Matt strongly believes that for a social system to be successful it needs to serve the selfish individual motive. People will never be motivated to use these networks unless there�s a specific value add back.

"The most successful folksonomy services will be content agnostic, and the same goes for social media services as a whole."
- Matt Locke

The consequence of this is that the most successful folksonomy services will be content agnostic, Matt said, adding that the same goes for social media services as a whole. Flickr doesn�t care what photos you�re posting as long as they�re not pornographic, YouTube doesn�t really care what you�re putting up videos of, Delicious doesn�t actually care what websites you�re looking at. But the BBC really, really cares to know what you�re watching � that�s our business. And that�s the biggest problem for any type of major content business working in this space. You�ve got to understand that you can�t just build something for your stuff because actually people�s motivation for using it is about their stuff, it�s absolutely about them controlling their own content, their user-generated content and the content they�re interested in and hence the service itself has to be content agnostic.

He would go so far as to say all social media services full-stop are content agnostic. This puts the BBC in the space where � as Clayton M. Christensen describes in The Innovator's Dilemma � we are the incumbent being disintermediated massively. And this is a big problem for the BBC � how can we design and deliver services which are about more than just our content? There are so many rights issues, and fair trading issues and market issues.

"Can a company with a stake in content successfully employ folksonomies?"
- Matt Locke

For content agnostic writ large, type �BBC� into YouTube.com and the first and second results are not the UK BBC. YouTube doesn�t mind that the tags don�t align with what we think BBC is. And that�s the issue for the BBC in working with folksonomies. The organisation wants too much to hold onto the power of its brand. For a lot of people in the organisation this is really threatening and a big issue, although Matt doesn't feel that way.

Can a company with a stake in content successfully employ folksonomies? The BBC Innovation Labs that Matt ran in 2006 explored this idea. Out of 170 ideas submitted they picked 29, and had workshops with them and are developing ideas out of that now � including ideas from Poke and Amberlight who were present at Content 2.0. They will be trialling some of them on the BBC site later on.

And the BBC Programme Catalogue Tom Loosemore has done with Matt Biddulph and Ben Hammerlsey, has exposed the programme data of content that the BBC doesn�t know what people want to do with. The Creative Archive project sprang from a meeting with Larry Lessig who Matt brought over a few years ago. Backstage BBC led by Ben Metcalfe was another such initiative. Rights issues are holding back the opening up of the actual content, but they do have all the metadata, which is what the Programme Catalogue so successfully opens up through its API that will allow you to go on and play with it and mix it up. At the minute, this is the best way they have to see what people�s motivations are.

"Flickr are fantastic at building more value into their system without ruining that relationship of being playful"
- Matt Locke

Throwing this data out there and getting people involved with using it is the best way they can start to test that interface between formal and existing taxonomies and emerging folksonomies. This is where that relationship and space between the three tensions of comparability, visibility and control can be experimented with. Matt wants to see services built that play with those tensions, and create value services for people.

The most successful services will be playful, he reiterated, because they will have nothing at stake. What Flickr are fantastic at is building more value into their system without ruining that relationship of being playful. It�s really important that �interestingness� isn�t an overt transaction because as soon as they did that something would be at state and people would start gaming it. The fact that it�s this kind of back-end thing that�s a picture of the system � that�s really about preserving that playfulness in taxonomies.

Making them selfish and content agnostic are the other challenges. Flickr understood brilliantly that people have an emotional investment in their photographs which makes them want to talk about them. Also there�s no existing metadata around them so people add it that themselves. But how do people feel about telly � not user-generated TV � but the BBC�s programming? How do they want to link to it, index it, what kind of things do they want to say about it, what do they want to link it with and cut and paste it with?

He signed off by giving us a priviledged glimpse of the Rodney Trotter latex mask from YouTube.

Content 2.0 - 2006 conference Website:
http://www.content2point0.com/2006/

About Matt Locke::
Matt is Head of Innovation for BBC New Media & Technology. He is responsible for developing and running research programmes within the BBC and with external partners, including developing academic and industry partnerships, and developing open innovation initiatives like Backstage BBC and The BBC Open Programme Catalogue. Before that, he was Head of Creative Research & Development within the same division. Matt is a regular participant on working groups and advisory boards for other public sector projects and organisations. This has included being on the working group for the Office of the E-Envoy's Digital Inclusion Report, The Advisory Board for Ars Electronica's first Digital Communities award, and the jury for the mobile technology strand of ISEA 2005. He also chairs the External Advisory Panel for the Creative Archive License Group and is on the Board of the Brighton Photo Biennale. Prior to joining the BBC, Matt ran a centre for research in digital culture based in Yorkshire. Before that, he worked as a curator and writer, specialising in the social adoption of technology and the cultural impact of digital technology, and still continues to write regularly about these themes for journals, websites and his own site at http://www.test.org.uk

OTHER CONTENT 2.0 SESSION REPORTS

Content 2.0: Mesh Up - Connecting Content To People

Content 2.0: Goodbye New Media Hello Social Media

Content 2.0: Marketing 2.0 Forum

Content 2.0: Can Brands Be Trusted?

Content 2.0: The Future Of Web Search

Content 2.0: Search & Enjoy Forum

Content 2.0: The Invisible Culture

Beers & Innovation (music special) @ Content 2.0