How Your Private Tweets Might Be Included in the Library of Congress Public Archive

Today’s announcement that the Library of Congress will be archiving all public tweets since March 2006 prompts many questions. But most people, I suspect, are comfortable with the concept since the LOC is only archiving public tweets; those who decided to restrict the visibility of their Twitter traffic can rest assured that their chatter won’t be included in this mass collection of public utterances.

Or can they?

Consider this scenario:

1. You decide to protect your privacy/visibility and keep your tweet stream protected.
2. I send a request to follow you. You accept. I now receive your private tweet stream.
3. I retweet one of your private tweets. (Note: Twitter restricts the ability to retweet private tweets, but it can be done manually, or in most 3rd party apps.)
4. My steam is public. Your tweet is now public, embedded in my stream.
5. The Library of Congress archives my public stream, including your private tweet that I had retweeted.

So, a few reflections on this (common) scenario. First, it is ethically-questionable whether users with public streams should be retweeting private tweets without express consent.

Second, and more to the point, people’s attempts to restrict access to their tweets can be easily thwarted, and ultimately those private tweets can end up in the Library of Congress’s archives with the rest of the public tweets.

It’s time to brush off the contextual integrity lecture notes.

UPDATE:  See this article in the New York Times, where a LOC spokesperson notes:

Some online commentators raised the question of whether the library’s Twitter archive could threaten the privacy of users. Mr. Raymond said that the archive would be available only for scholarly and research purposes. Besides, he added, the vast majority of Twitter messages that would be archived are publicly published on the Web.

This is the classic “but the information is already public” argument that, while technically true, presumes a false dichotomy that information is either strictly public or private, ignoring any contextual norms that might have guided the initial release of information or how a person expects that information to flow.

This is Nissenbaum’s theory of contextual integrity, which Fred Stutzman has already invoked related to this case.

Further, it is interesting that the LOC seems to acknowledge that there are non-public tweets within the archive: “…the vast majority…are publicly published on the Web”. I will reach out to Mr. Raymond to seek clarification.