Image source: flickr
David Weinberger gave a description of a link in a panel last year at the Hyperlinked Society Conference. A link is a conscious act of generosity. These acts is moral, and they form the architecture of the web. He goes on to explain that the syntax of a link (i.e. the href HTML tag) has no meaning within itself, it is merely an instruction which points to another location. The meaning of the link, which can be agreement or disagreement, is found in the text surrounding the link.
While these links have no meaning, they do have value, which is the reason by creating a link performing generosity. Google ranks pages by the number of links other sites point to a page. Appearing early in a search result clearly has value over a later listing. You can only have a reputation if other people can find you. A page and her owner’s reputation then relies on the generosity of others linking to her page. If an author disagrees with the contents of page and wishes to dispute it, linking to the page adds to its value and reputation. The author is then left to not link. However, this practice which the status quo forces people to use still leaves the reader at a disadvantage.
There have been suggestions to create a newer kind of syntax and link taxonomy which would add to the current binary options of link or no link. The simplest system would be to have three choices, positive link, negative link and no link. This system would actually be very easy to for users. All you need to do is add a tag to the link.
Flipping forward one year, I was struck when Jonathan Zittrain pointed out in his talk last Saturday, the use robot.txt files for telling search engines not to spider a file or directory started in the early age of the web as an adhoc measure by individual which became an internet standard. Today, it is much harder to get a standard adopted, but the story of robot.txt reminds us that it is possible to create grassroots change in internet standards. Endorsement links allude to aspects of the Semantic Web, but frankly, I’m not sure if it will every come. Contextual syntax might evolve over time with gradual implementations.
The idea of rated links get even more interesting when you consider how search engines might use links that interpret reputation and authority. Of course, gaming the system would occur, but that happens now and should not deter the implementation of a link taxonomy. It might also encourage search engines to become open to annotating listings, as Frank Pasquale has suggested. Generally, search results are given by relevance or time of creation. New categories could be ranked in terms by agreement, disagreement or even controversy. The end result would be better ways for author to link, for readers to under the context of the link, and for searcher to usage links in the aggregate.
I know this doesn’t completely answer the question, but rel=”nofollow” does start to address the issue, although it carries it’s own set of issues. While it does offer a way to link without adding value, the way Google and other engines have prescribed it’s use strikes me as manipulative (most blog engines use nofollow automatically to link to commenter names — I wrote a bit about it awhile back). Also, while you’re on the topic, check out this article.
Thanks Noah. You read my mind, as one of my next posts was going to be on no follow. Interesting links that I’ll try to incorporate too.