How to be a Good Institution eg Twitter

Twitter has come in for a good deal of criticism for its recent announcement of new rules for developers using the Twitter api. See the BBC report or John Gruber’s blog on Daring Fireball.

I am a keen admirer of Twitter and have been thinking a lot about Twitter for some months in the context of John Searle’s philosophy of speech acts (yes I am writing a book about it too, since writing is the best way of working out what you think you think). So this post needs a bit of context on Searle’s philosophy of institutions as explained in his last book Making the Social World. If you think that analytic philosophers hammer intricate little issues into the ground you need to take up Searle, Searle’s book has staggering ambition (its subtitle: The Structure of Human Civilisation), and develops a view in which all social institutions are constructed from linguistic action (‘speech acts’) whose provenance and persistence is partially concealed from us:

We live in a sea of human institutional fact. Much of this is invisible to us. Just as it is hard for the fish to see the water in which they swim, so it is hard for us to see the institutionality in which we swim. Institutional facts are without exception constituted by language, but the functioning of language is especially hard to see. This might seem an odd thing to say because we are often conscious of language when we engage in a conversation, receive a telephone call, pay our bills, answer our email and so on.What I mean is that we are not conscious of the role of language in constituting social reality. We are aware of such things as the actual conscious speech acts we perform, and we are often aware of such unimportant things as the accents with which other people speak, but the constitutive role of language in the power relations in which we are immersed is, for the most part, invisible to us. Making the Social World p 90

Searle has a theory which helps to explain how institutions are built from and with language, and it focuses on the key role of what he calls status function declarations (SFDs). This is a pretty ugly term, but SFDs are every where in Twitter, in fact Twitter is a beautiful example of a more or less purely linguistic institution entirely shaped and constituted by SFDs. Twitter is pretty much an ideal example of the way institutions come about through and then grow enormously through the deployment of Status Function Declarations. There are in particular three key SFDs that drive Twitter, the act of joining Twitter, the act of ‘following‘ on Twitter, and of course the act of tweeting. Tweets can be used for almost any kind of speech act (praise, promising, threatening, describing, questioning etc) and they form the ocean of Twitter content, the water in which Twitter fish swim. But for an understanding of Twitter’s structure and its boundaries as an institution, the SFDs involved in following or joining Twitter are even more important and more constitutive. There are more SFDs in Twitter than this basic trinity, for instance in understanding following you need to reckon on unfollowing, blocking and some other lesser moves in the Twitter game, but they are the key to an understanding of the way it works as a purely digital and almost purely linguistic institution.

Well that in a nutshell is the thesis of the book that I have started to write. If the thesis is correct, and I admit that it needs a lot more evaluation and explanation — heck there is a book still to be written — then it puts Twitter’s recent rule changes in an interesting light. In one way Twitter’s new policy barely touches the apparent core of the Twitter institution, because the API and the role of developers in building the Twitter eco-system is barely mentioned in the process of joining Twitter as a user. And one of the striking things about joining Twitter is that the user acknowledges that Twitter (the corporation) has and is deemed to have total control over the way Twitter shall run and the way in which it will develop. Like most private digital institutions Twitter is an absolute monarchy not a democratic institution. But Twitter does have an API and most significant web institutions need an API because web services need to interoperate and work across the digital boundaries of neighbouring institutions. In short Twitter is a digital institution amongst others and it will inevitably interact with other digital services, companies and groupings in ways which are complex and mutual. Its the apparently one-sided nature of the newly announced Twitter rule making for its own developers that is causing comment and concern. Twitter needs to remember that its fish swim in other ponds and adjacent oceans. If Twitter developers start to think in the way of Marco Arment, Twitter stagnates and eventually dies:

Twitter has left themselves a lot of wiggle-room with the rules. Effectively, Twitter can decide your app is breaking a (potentially vague) rule at any time, or they can add a new rule that your app inadvertently breaks, and revoke your API access at any time.

Of course, they’ve always had this power. But now we know that they’ll use it in ways that we really don’t agree with.


Twitter has proven to be unstable and unpredictable, and any assurances they give about whether something will be permitted in the future have zero credibility.

I sure as hell wouldn’t build a business on Twitter, and I don’t think I’ll even build any nontrivial features on it anymore.

And if I were in the Twitter-client business, I’d start working on another product. Interpreting some of Twitter’s API changes.

Twitter may grow out of this phase of ‘over controlled’ API management and too closely supervised developers, I hope so. Twitter needs to behave like a good digital institution, which does not simply mean ruling your users with the supreme confidence of an absolute monarch, but figuring out how Twitter’s obligations and needs fit in with the bigger picture.