Yesterday I published one of those unholy been-in-the-drafts-forever posts on issues of linking. Here I see if I can make the point a bit more cleanly.
When we started getting people to use federated wiki in December, I thought the default sort of editing would be of the main article. You write a piece on Dylan’s 1966 Royal Albert Hall show, I come by and add more facts to it.
When we initially put people into federated wiki people didn’t do that. Instead, they commented, by adding endless signed thoughts to the bottom of the page. This is bad for reasons I won’t go into here. So I told people stop commenting, and some people complained, but most complied.
The next behavior that emerged was more interesting. People started extending articles not by editing them, but by linking to older articles or writing new ones. They’d do this at the bottom of the page. For instance:
This seems related to [[Walking the Lines]]
Where Walking the Lines was a page detailing a concept, theory, anecdote, or example of something related to the main page. Over time I started to formalize this pattern in my own writing on wiki. Here’s the bottom of my page on the concept of the Moral Cascade:
[[Normal Accident Theory]] posits that error should be seen as a normal occurrence, and systems should be designed to avoid cascading behavior.
[[Moral Cascade in the Classroom]] details how similar patterns play out in classroom management.
Broken Windows Theory described a moral cascade where small offenses led to large offenses, but it may be overhyped. See [[Broken Windows Theory Broken]] for a rejection of Moral Cascade patterns.
This started to become the most enjoyable part of the process for me, and the most profitable.
As I started to think about this I realized how natural the behavior we were seeing was. Say a person writes a page on — well, Dylan at Royal Albert Hall in 1966. The page is short and focused. Other people come by and read it.
The chances that any given reader knows more about that subject than the person who just wrote it are slim. It happens, but its not the median experience. On the other hand, the chances that any given reader knows something related to that subject are very high. So (in our fictional example here) the links accrue:
Stress from the tour would later be cited as a cause of [[Dylan’s Motorcycle Accident]].
For explanations of why people would go to a concert just to heckle, see [[Psychology of Heckling]]
Royal Albert Hall’s acoustics were not particularly suited for rock music. See [[Acoustics of Royal Albert Hall]]
And what starts to occur to me is that this is actually the Vannevar Bush vision, where the median user is neither a full-fledged writer or a simple reader but a linker. And part of the reason the reader can be a linker is that
they have their own copies of the document
they can add supplementary documents very easily and link to them, and
they don’t have to rewrite the main document to add links
Now you do get these capabilities in some annotation systems, but talking about problems with annotation systems is perhaps for another post.
In any case, this is the core idea of the last post: we can recapture, perhaps, this vision of the reader as the primary link author, but it requires us to think of links in a different way.