There is a great short story by Jorge Luis Borges, called the Library of Babel, in which he describes a fantasy library, containing every possible book. Almost all the books are total gibberish, containing every possible combination of the letters of the alphabet, but a tantalising few contain snatches of meaning. And, logically, there must be a perfect copy of Hamlet in there somewhere, and also a copy where Hamlet is the murderer, or where he is abducted by aliens. Or anything. There must also be books which translate some of the gibberish books into sense. And so on.
It is easy to calculate how many possible books there are in Borges’ library. According to his scheme, he allowed only 22 lower case letters, a period, a comma and a space. Each book was 410 pages long, 40 lines per page, and 80 letters per line.
The answer is described in detail on wikipedia: it boils down to 25 to the power of 1,312,000. You can create one to look through here.
Which brings me to the point. How many possible tweets are there? If we follow Borges’ scheme, there are 25 to the power of 140 possible tweets. Of course, thanks to the miracle of Unicode, we have a lot more than 25 characters available to us. But lets keep things simple. Only a tiny fraction of these posts would make any kind of sense in any language.
Which got me wondering: how many of our tweets are actually unique? How many people have tweeted, for example, about eating a muffin at starbucks? And I thought it would be another fun twitter side project, to mark each non-unique tweet in some way, and to let you know if someone else has made exactly the same post before.
As the years pass, unique tweets would become rarer and rarer. As we posted, we would start expecting to find replicas of our droll little remarks: perhaps we would feel a little less special, or maybe pleasure at the feeling that we were not alone in the universe.
It would also answer the question of which tweet was the most common. My money is on ‘Too drukn to typee’.

