Classification, culture & the Flickr.com tag

On: October 15, 2009
Print Friendly, PDF & Email
About Maarten Hoogvliet
I am a MA student of the Media and Culture master New Media at the University of Amsterdam and I have a BA degree in Communication and Multimedia Design at the HRO in Rotterdam, formerly a part of the Willem de Kooning Academy of Art. Next to doing my masters I am a graphic designer/illustrator.

Website
http://www.one3rd.nl/blog    

Web 2.0 tagging systems like Flickr’s categorize the website’s content bottom-up.

The classification is powered by users applying their common sense and intuition; wisdom-of-the-crowd, resulting in a folk taxonomy of everything that is to be found in the Flickr database.

A folksonomy is contextual, ambigious, gradual, adaptive, easily accessible, open and flexible. A folksonomy connects with the reality of its users, uses the idiom of their native language and therefore, has a plane learning curve. As said, a folksonomy is built bottom-up, instead of top-down classic classification (e.g. Aristotle’s classification of living things, the Dewey Decimal system, the periodic table of elements).

What things exist and how are they related? This is the main question classic ontology asks. Through the centuries, man has always longed for a universal classification system, as to have the feeling of everything that exists in the universe being under human control. And, next to that, that system also needs to be flexible; capable of incorporating future anomalies.

Sadly, this seems to be impossible. Take for instance Aristotle’s classification of living things, in which he placed emphasis on the type(s) of soul an organism possessed, asserting that plants possess avegetative soul, responsible for reproduction and growth, animals a vegetative and a sen

sitive soul, responsible for mobility and sensation, and humans a vegetative, a sensitive, and a rational soul, capable of thought and reflection. Or second, the Dewey Decimal system, in which world religion is divided into nine categories: (210) Natural theology, (220) Bible, (230) Christian theology, (240) Christian moral & devotional theology, (250) Christian orders & local church, (260) Christian social theology, (270) Christian church history, (280) Christian sects & denominations, (290) Other religions. [1]

These two examples illustrate that classification systems have always been historically and culturally embedded and imposed upon us by scientists/experts. In the spirit of Karl Popper; (contemporary) science seems to be incapable of conjuring a classification system that will not be falsified at some point.

“It is clear that there is no classification of the Universe not being arbitrary and full of conjectures. The reason for that is very simple; we do not know what thing the universe is.” [2]

Systems of classification always are culturally embedded, built on a stable assumption of what the world is at a particular time. In medieval science, for instance, that assumption was based on the four directions of the wind, the four seasons, the four bodily fluids of Galenus, the earth as the centre of the universe, the church as the central power in society, and so on. If a culture develops and knowledge increases, there inevitably comes a moment when a classification system has to be heavily adjusted or repudiated.

Bowker and Star write that categorization is a natural phenomenon. People don’t just see the relations that define units within a culture, people learn them by being a part of that culture and within that culture of different “communities of practice” [3]

“In the simplest seeming action, such as picking an article of clothing to wear, is embedded our complex knowledge of situations. (Where will I go today? What should I look like for the variety of activities in which I will participate?) These situations involve multiple memberships and how objects are used differently across communities. (Will this shirt ‘do’ for a meeting with the dean, lunch with a prospective lover, and an appointment with the doctor at the end of the day?) Many of these choices become standardized and built into the environment around us; for example, the range of clothing we select is institutionalized by the retail stores to which we have access, traditions of costuming, and so forth. To think of this formally, the institutionalization of categorical work across multiple communities of practice, over time, produces the structures of our lives.” [3]

One community uses a unit differently than the other. Units get coded en decoded when they transfer the borders of a “community of practice”. They become meaningful because they reside in different contexts. To be meaningful, the different contexts need to be connected by system of comparison and appraisal; categorization.

Bowker and Star point out that classification systems are cultural embedded. In a classic classification, units are captured in a certain vocabulary, which, at a certain time, turns out to misfit the units it describes. Units develop continuously within a culture, different communities are using them differently, which ever influences their meaning.

We need a richer vocabulary than that of standardization or formalization with which to characterize the heterogeneity and the processual nature of information ecologies.” [3]

Now, folksonomy could be that rich vocabulary. Folksonomy is automatically culturally embedded, simply because all authors are. This means that folksonomy automatically adapts to whatever changes units go through. Because folksonomy is bottom-up instead of top-down, it always stays up-to-date. What about classifying the whole web by wisdom-of-the-crowd and folksonomy?

Of course, there are a lot of problems with classifying by folksonomy. Take for instance polysemy (multiple meanings for the same tag), homonymy (one tag, but multiple unrelated tags), synonymy (multiple tags with the same meaning), different levels of categorization (mammal, pet, cat), spelling errors and different ways of writing (nyc, New York, NY) [4]

Also more deep problems arise. A photo tagged with ‘England’ for example; does this mean the photo shows something in or from England or that the photo is taken in England? Tags can have many different relations with the objects they describe. [4]

However, looking through these practical problems, there seems to be a possibility for a differentiated system of search- and findable units.

“Folk wisdom, however, is not necesserily impaired by ambiguity and inexactness. It is true that tagging systems like Flickr’s do not allow users to specify meanings by expressing structural relations, be they linguistic or taxonomic. […] But at the collective level of the tag cloud one can observe a process of subtle differentiation and speculation.” [4]

Folksonomy does have the potential to grow and become a more defined system, where units become ever more ‘natural’, because the meaning of units changes with their use within a cultural practice; Gadamer’s hermeneutic circle, embedded in a classification system. A classification system, created by culturally situated subjects, develops along the same line as the meaning and use of phenomena used in a culture.

This maybe is a bit abstract, and of course there are a lot of practical problems yet to be solved, but folksonomy has a connection with daily life and common sense that every classic classification system lacks, which is the main reason why those systems collapse eventually.

To end this piece, a quote which does not argues for folksonomy directly, but if you read between the lines…

“Any working infrastructure serves multiple communities of practice simultaneously, be these within a single organization or distributed across multiple organizations. […] To do so, it must bring into play stable regimes of […] objects such that any given community of practice can interface with the information system and pull out the kinds of information objects it needs.” [3]

  1. Shirky, Clay. ‘Ontology Is Overrated’ Clay Shirky’s Internet Writings, 2005.
  2. Borges, Jorge Luis. ‘The Analytical Language of John Wilkins’. Trans. Paul Perry, 1941.
  3. Bowker Geoffrey C., Star Susan L. ‘Sorting Things Out: Classification and Its Consequences’. The MIT Press, Cambridge, Massachusetts, (1999): p. 285-326.
  4. Simons, Jan. ‘Tagalese – or the Language of Tags’, Fibreculture Journal, Issue 12, Metamodels and Contemporary. 2008

Comments are closed.