More about The Subjectory

This piece of text has been presented at ELAG 2008 in Wageningen, back in May. I found it interesting enough to post again on my own web site. If you like you can also watch my actual ELAG presentation video.

Subject headings as a closed, professionally managed indexing vocabulary are increasingly under fire. Although their efficiency for searching already indexed documents is undisputed, there always have been problems with subject headings. These have caused many if not most library end users to bypass the subject headings catalog when searching for documents, relying on full text search and other ‘least effort’ methods. But despite their decreasing popularity with the user crowd, subject headings are still widely used – not in the least due to an indexing legacy of decades that simply cannot, and should not, be thrown away.

Social tagging and other, unmanaged, open, free-as-in-freedom systems are gaining momentum. This is largely due to several major, interlinked changes in the way people consume information resources. The advancements in network technology have brought ‘data at the people’s fingertips,’ which has spawned the hyperlink, the Web, Google, user-changeable databases, social tagging, and many more fundamentally networked technologies commonly called ‘Web 2.0.’

So the demand for subject-based access has remained very high. All the disjunct free-tagging systems essentially aim at subject access. And the majority of search engine queries are essentially subject searches: more often than not, people are not interested in one specific document, but in a range.

Can we combine the professional rigour of authoritative subject headings with the popular freedom and extreme flexibility of tagging, either social or private? Can we leverage subject access itself, using a new Web 2.0-like approach?

We can. We should not bother about the differences between subject headings and social tags, or between authorities and the public. Instead, we should take both subject headings and tags seriously. And at this moment we do not take either of them seriously. Both subject headings and tags today are considered just second-class web citizens. They are either imprisoned in authority vaults, or left homeless to wander on the streets. They are not considered valuable. They are typed in and forgotten, quickly replaced by result lists. How can we be so crude to these useful, hard-working web citizens?

Tags and subject headings need a proper home. They demand to become first-class web citizens. They deserve to become content, instead of only access pointers. After all, metadata is also data.

We should give each of our subjects its own web site, with blog, forum, RSS feed, concise and stable URL, maintainers, fans, users… and support development of a subject community. Encourage a number of favourite references to excellent content out there on the web about the subject, and of course also links to our search engine, to content not available on the web. Allow to extend the heading with true content as well, why not? As long as people follow the Wikipedia quality standards and refer to other material and other subjects, what is wrong with concise content directly at the subject?

Objectivity is a noble target, but impractical, as nobody is truly objective. Honest subjectivity lasts longer. Nearly all of Web 2.0 is about subjectivity, about personalization, about personal subjective interpretations. So, just publish your direct subject URLs where possible, and use them. Allow subject access… literally! The URL is your subject heading. It points the world at your subject repository, your subjectory. Subjectories are made to be linked up and to be fostered by dedicated fans. Nobody expects them to define the world.

With our subjectory out on the Web, we can expect new documents and maybe even existing documents to start (back)linking to the subjects. We will have to compete with other subjectories on the web for backlinks, but if you can’t beat them, join them: refer to these other subjectories and we are done. Subject communities should team up with other communities, link to each other, and weave the subject web around the world. Whether or not that will be Semantic Web-compatible, just weave it. If the information is there and it accessible, somebody will eventually tap into it. Remember, nobody sat down to enable Google when creating web pages in the last century.

The subject is dead! Long live Subject 2.0!