Collabulary, Not Folksonomy

via Stephen Downes who points to a very good report capturing some of the main ideas behind Web 2.0 and looking into some of the implications.  He notes that if you are new to Web 2.0, this is an excellent introduction. 

There's always something to be learned by reading from a variety of perspectives and getting out of the silo of a particular field.  So, in this instance, I plunged into the educational technology space.    What caught my eye was the title of the report, What Is Web 2.0? Ideas, Technologies and Implications for Education."    There is a section about tagging. 

Because I'm thinking about tagging from the perspective of online communities of practice, I found this bit in the report interesting.  I'm very interested in the how communities form, the behaviors, the ecosystem of the community as well as the practical how tos.  The one piece of information that was new to me was this:

Folksonomy versus collabulary

One outcome from the practice of tagging has been the rise of the ‘folksonomy’. Unfortunately, the
term has not been used consistently and there is confusion about its application. More will be said
about this in the section on network effects, but for now it is sufficient to note that there is a distinction between a folksonomy (a collection of tags created by an individual for their own personal use) and a collabulary (a collective vocabulary).

Later, a section about tagging, this difference is explained in more detail:

The term folksonomy is generally acknowledged to have been coined by Thomas Vander Wal,
whose ideas on what a folksonomy is stem, in part, from his experience of building taxonomy
systems in commercial environments and finding that successful retrieval was often poor
because users could not ‘guess’ the ‘right’ keyword to use. He has, however, expressed
concern in the recent past about the way the term has been mis-applied and his definition,
taken from a recent blog posting, attempted to clarify some of the issues:

Folksonomy is the result of personal free tagging of information and objects (anything with a URL) for one's own retrival [sic]. The tagging is done in a social environment (shared and open to others). The act of tagging is done by the person consuming the information.'

VanderWal, 2005, blog entry.

The following point makes we wonder about the difference in terms of behaviors and values in tagging communities versus crowd filtering communities (e.g digg). 

Although folksonomy tagging is done in a social environment (shared and open) Vander Wal
emphasises that it is not collaborative and it is not a form of categorisation. He makes the
point that tagging done by one person on behalf of another ('in the Internet space' is implied
here) is not folksonomy and that the value of a folksonomy is derived from people using
their own vocabulary in order to add explicit meaning to the information or object they are
consuming (either as a user or producer):

'The people are not so much categorizing as providing a means to connect items and to provide their meaning in their own understanding.' (Vander Wal, 2005). By aggregating the results of folksonomy production it is possible to see how additional value can be created.

Vander Wal states that the value of a folksonomy is derived from three key data elements: the
person tagging, the object being tagged (as an entity), and the tag being attached to that
object. From these three data elements you only need two in order to find the third. He
provides an example from del.icio.us which demonstrates that if you know the object's URL
(i.e. a webpage) and have a tag for that webpage, you can find other individuals that use the
same tag on that particular object (sometimes known as 'pivot browsing'). This can then
potentially lead to finding another person who has similar interests or shares a similar
vocabulary, and this is one of Vander Wal's key points concerning what he considers to be the
value of folksonomy over taxonomy: that groups of people with a similar vocabulary can
function as a kind of 'human filter' for each other.

I also found this bit extremely important - the whole that actually talks about tagging in the context of "wisdom of crowds" - so what are the differences between, say, tagging and collaborative filtering with digg.  How can they enhance or support?  Does one approach focus more on the information itself without developing the community aspect?  How important is the community aspect to the creating value of the information ... lots to ponder here.

Another key feature of folksonomy is that tags are generated again and again, so that it is
possible to make sense of emerging trends of interest. It is the large number of people
contributing that leads to opportunities to discern contextual information when the tags are
aggregated (Owen et al., 2006), a wisdom of crowds-type scenario. One author describes such
unconstrained tagging, in the overall context of the development of hypertext, as 'feral
hypertext': 'These links are not paths cleared by the professional trail-blazers Vannevar Bush
dreamed of, they are more like sheep paths in the mountains, paths that have formed over
time as many animals and people just happened to use them' (Walker, 2005, p. 3).

Comments

[Stephen Downes]

I noticed this as well and wondered about it.

It is important to distinguish between a network behaviour, such as the folksonomy as described above, and a group behaviour.

A collection of tags may be created in two very distinct ways:

1. people, working independently, just happen to use the same word to describe the same resource

2. people, working together, agree on a term that describes a given (type of) resource

Method number (1) is a folksonomy, and it is a network behaviour. It does not involve collaboration of any sort.

Method number (2) is not, strictly speaking, a folksonomy. It is a method more common to librarians and taxonomers.

We have seen, however, efforts made to organize tags (people will write, "Everybody tag this event 'OCC2007' or whatever).

This sort of organization is arguably no longer a folksonomy, as some people are using a privileged position to instruct other people how to tag (I discuss this in my paper here: http://www.downes.ca/post/14 )

I would not go so far as to use a word like 'collabulary' - that is a ridiculous word, and is not needed to describe something that we already have perfectly good words for, a 'taxonomy' or a 'vocabulary'.

And the author's suggestion that folksonomies ought to be recognized as 'collabularies' is, in my view, a mistake: it either misrepresents what a folksonomy is, or it uses a new word needlessly.

[chris blow]

uh ... Hrm. I like collabulary.

Stephen, I think that your argument against the "spike-based power-law-based Instapundit-based network" is great (delightful article, here: http://www.downes.ca/post/14). But I feel like you're missing the point.

Specifically, in the case of the "nptech" tag we are most certainly using your Method (2) -- and you are right in this view that it is *not* a folksonomy. We decided to use it.

But in the cases I have seen, communities of practice are not using these tags in a way that is formal or even hierarchical. There seems to be just enough organization involved for the tag to provide some higher ratio of signal:noise.

There is no imposition of meaning on the tag "nptech;" it is merely the name of the community. It is an entirely open aggregator. There is not a element of privilege that undermines the value of the network.

I think that "collabulary" connotes (and denotes) this better than

[Steve Dale]

Yes, all very interesting, but getting back to the real world of managing communities of practice and providing the social media environment through which they can collaborate, the definitions of 'folksonomy' or 'collabulary' are completely immaterial. I'm at the practical end of this debate where I spend most of my time explaining to the users of the CoP platform I've deliverd for local government (http://wwww.communities.idea.gov.uk) why it wouldn't be a good idea to provide an IE-style file plan for the document library.

I spent the best part of a week explaining the concepts and benefits of having a flat-file tagging mechanism that would enable all users to see (via a tag cloud) what other's were adding to the library, without the need to understand or maintain a file plan.

These debates about folksonomies and collabularies are strictly for the acedemics and not the practitioners. I'm dealing with the real world and trying as best as I can to de-mystify the jargon so that those new to social networking and social media do not feel threatened by stuff the don't understand.

In troducing words such as 'collabulary' to the English language doesn't help my cause one little bit.