I got some offline feedback on my last entry, with the effect that I rethought a few things. Here are some of the new thoughts:
Anonymity. The way I defined this previously was “being out in public without being notable”. This isn’t a very good definition, because, as Gavin pointed out, anonymity actually has a more technical definition that’s important to preserve, namely: being in public without being known. So works of art can be anonymous, in that they are well-known but no one knows who made them (they are completely unsigned). Or a person can be anonymous by being in disguise or otherwise completely unrecognized. Or information can be made anonymous, “unconnected to an identity”, by purging it of identifying information, like aggregated web search data unconnected to IP address or other similar identifiers.
Gavin suggested that the concept I defined previously could be described as being “unnotable” or “unnoticed”. Perhaps a better word is needed, but having both concepts is certainly more useful.
Another concept that I didn’t define explicitly, but left under the umbrella of privacy, is pseudonymity. This is a very important concept in modern web communications since so much information these days is attached to usernames. When is a pseudonym truly unconnected to a person’s “real identity”? This can be a challenge to determine, and a lot of pseudonymous information is poorly protected because of subtle identifiers in the information or interconnection between pseudonymous information and information filed under a “real name”; it can also become an issue when pseudonym or username is used for multiple sites, services, or types of works. It’s often easier to find a person’s data on the web once you know one of their common usernames than it is when you know their name. Usernames are, by their nature as keys to a specific record, more unique than names.
I also am not that fond of my definition of notability. It doesn’t seem to me to require numbers, but only a certain level of significant interest. However, that’s pretty hard to describe and define.
Finally, Dave wrote me an extensive discussion of yet another concept relating to accessibility: risk.
Risk is what you have when information is accessible to some people, but not others, because there’s a risk of failure of the safeguards that prevent it from being accessible to everyone (loss or deliberate breakage), as well as a risk of legal decision that the safeguards must be removed (search warrants, subpoenas).
Dave sums his discussion up thus: “Heightened accessibility, even if it is well-understood under normal conditions, still creates the prospect of lowered privacy.”
This is, I think, one of the big deals about accessibility that makes people pitch a fit about sudden increases in it.