NoFollowing Wikiality
The introduction of the NoFollow attribute introduces a whole new set of questions about User Ethics on the internet. In the age of social bookmarking, however, the question of whether the NoFollow itself is ethical underscores the question of whether Wikipedia is a source of credible information.
When the NoFollow tag is included in a link, search engines will not count that link toward the linked site's search ranking. Consequently, individual webmasters have a lot more power in effecting search results. Of course, there are concerns that while many will abuse this power by blanketing all their links with the NoFollow tag, others will forego their so-called search-vote.
Like any democratic process, however, abuse and apathy is an inherent part of the system, and it is probably better that some abuse the system while other forego it, than if there was no system at all. After all, the NoFollow tag mean power, and power can be corrupting. It is not, however, better to be powerless.
In a recent post titled NoFollow Leverages Mistrust Among People, Carsten Cumbrowski voices concerns about many of the dangers inherent in the NoFollow tag. If anything, though, Cumbrowski's concerns illustrate the inadequacy of search algorithms writ large--not that of the NoFollow. Indeed, the NoFollow seems to be a quick fix for an already defective system. As he explains:
Next to the extreme cases that lead to paranoia or too much trust on the other side are the majority of daily cases where the nofollow attribute causes “bad blood” among people. “What? You are using nofollow to link to my site? You don’t trust me? Okay, now I know what you really think about me.”.
The reason for somebody who uses “nofollow” intentionally for a specific link might not be “mistrust” of the site the link points to, but rather “not complete trust”.
Unfortunately is there no way to indicate in shades of grey how much somebody trusts somebody else. The nofollow attribute is black and white, absolute, yes or no and that does not reflect real live situations at all.
[...]
For that reasons do I believe that nofollow does not and can not work properly in the world of automated ranking algorithms that are completely unaffected by emotions, because search engines can not and will never be able to account for the persons emotional reasons to add or not add the attribute.
Of course, search engine algorithms are ever-evolving phenomena. They were developed for a much smaller internet, and are continually struggling to adapt to rapidly changing online landscape. Therefore, they are imperfect and it is unreasonable to expect them to ever be perfect. Indeed, in the age of social bookmarking, using links as the primary indicator of a site's popularity seems somewhat ridiculous. After all, a popular page does not necessarily mean an authoritative one. Nevertheless, Cumbrowski seems to think that back links are the only indicator that search engines will ever be able to rely on to gage the relevance of a page:
Based on the context the link was placed does it become clear to the human visitor if the link reflects a positive opinion, negative opinion or neutral opinion with a lot of shades in between.
The same is true when it comes to intention. Why do want a visitor to follow that link? Backup your claims? Show without being biased the source of your citation? Find additional content to the subject that goes beyond the content of the current page? Buy something? Register for something?
[...]
For intention could be an attribute introduced with a number of values that are understood by search engines and can be extended as needed, e.g intend=purchase or intend=reference. [...]
How far Search Engines trust the attributes set by the webmaster is a complete different story. I would not recommend to accept them blindly and be biased.
It seems to me that for a search engine to be able to gage intention, it would require a degree of AI built into its algorithm. In the bare minimum, it would require some kind of interpretive capability similar to that which translation software works. It would need to develop a semantic understanding of each sentence and compare it to previous and susequent sentences to try and determine whether elements such as sarcasm are being employed by a webmaster. All in all, it seems like a long day off and, therefore, it is probably best to run with the NoFollow for now and let abuses of online democracy exist alongside those of real life democracy.
