I work a lot with teachers that don’t want to expose themselves to risk, argument, or harassment on the web. They create crazy good stuff that would be useful to other people for their classes, but the thought of sharing it on the web where they would be prone to lawsuits or complaints from wingnuts is unappealing to them or their bosses, so into the Blackboard Learning Management System it goes to die behind a password.
What sort of things are they afraid of? Well, maybe they are afraid that the readings of their class on gender identity will get picked up by wingnut media, and their legislators will be pushing for them to be fired. Maybe they are afraid that the page they wrote on the economics of medical care, in which they use a graphic under fair use, will get hit by a copyright suit. Maybe they are afraid that some unfriendly pedant out there will see hastily composed lecture notes and scoff at a mistake they make in presenting kin selection algorithms.
All these things have low probability overall, but present enough risk that people won’t share. As open educators we’ve fought, I think, a prolonged battle to cut through the FUD and get people to embrace the small level of risk.
I’ve lately been wondering if there are other solutions. Almost all the risks come from people you don’t know encountering your stuff and making your life miserable. So the solution has been to cut off access to your stuff by people you don’t know, which reduces network effects.
But another way to do this is to let anyone access the stuff you create (worksheets, explanations, activities, etc) but only let people who know the creator see the identity and institution of the creator. In other words, you sort of password protect the identity of the person doing it, not the asset itself.
I think (maybe) with hashing this might not even be hard — identity is stored in a hash, when I find materials I like they are checked against my list of email contacts to see if the hashes could be results of any of those, and if they are I can see the person’s identity. If not then they are Anon689aef6, or whatever.
Just playing around with the idea. It would be a scheme mostly for people wanting to release Creative Commons licensed material for other to fork while minimizing exposure.
Hapgood 
I know we just put into play a “federated” identity for our Wireless network. Eduroam is for visiting academics who want access to the network, but don’t want to put up with the limits/vagaries of our public wireless network. It’s based within Shibboleth, but at least has a tie back to the home institution through the eduRoam federation.
Forking might address the scared-to-share anxiety. The anxiety is understandable. It is not easy to cast s rhetorically effective position w/o stirring up someone. (Students tell me this all the time. ‘How to I critique this on my blog / on the wiki w/o pissing off the author?’ ‘Learn decorum.’ And we talk abot how to do that.) But forking might anonymize the statement enough so that those who might get wound up won’t know where to address their spit. Might frustrate them enough that they go elsewhere. Forking spreads that authorship around.
I’ve been working on revocable pseudonymous identity—I’ll add this to the use cases.
I’d be really interested to hear more about this.
How about this. You use standard PKI to leave a signature. Anyone who knows your public key can verify the signature, anyone who doesn’t has to try every key in the world randomly to figure out who you are.
Yes! This is exactly the sort of thing I’m talking about. So a class could all work together in the open, but no one knows who they are unless individuals want that. If later they want to show an employer their work, they could share their key. If they’d prefer not to show that work they don’t. And if they want to claim the work for the whole world to see they can publish the key to a public space.
Hmm, if people are working together, then they’re probably going to want to think in terms of a static nym for each person. If some of that working is f2f, they’re going to want to map the nym to the real ID. So, if you really care, this smells like a recipe for leakage (e.g. someone oversharing a page mapping real names to nyms, or even just accidentally referring to someone by name instead of nym in a public post).
Why just not share-with-identity internally, and release to the wild completely anonymously? You can’t get credit then, of course….
Peter Wayner’s book on “Translucent Databases” addresses a variety of interesting designs along these lines. Latest edition: http://www.amazon.com/dp/1441421343/