It was announced today that the Library of Congress will digitally archive all public tweets. Every public tweet that’s been sent since 2006 and every public tweet in the future will be stored on their servers. Of course, your private and reserved messages will remain private, but there is now a serious public archive of our messages to each other. It’s uncertain whether the Twitter archive will be the *Epic of Gilgamesh* of the year 6000 A.D., but it does have interesting implications for the stories we’re tweeting at the moment.
It’s likely this Library of Congress archive will be able to survive a major catastrophe and remain searchable for generations. Not that the United States government isn’t capable of losing massive amounts of data or that Twitter isn’t up to the job, but having some very smart archivists working on the problem is a big help. There is a national strategy being formulated right now by the National Digital Information Infrastructure and Preservation Program to:
collect, preserve and make available significant digital content, especially information that is created in digital form only, for current and future generations.
There are a lot of great tools for digital preservation, such as the LOCKSS system from Stanford University Libraries. I trust the Library of Congress is using the best techniques possible to insure our great-great-great-great grandchildren have access to their ancestors digital writing. However, just in case, you should store all of your tweets in the cloud *and* locally so if the Library of Congress loses all of the data in 2,000 years we can piece it together from everyone’s local cache. FYI.
So what kind of story are we writing about ourselves now that will be studied by future archeologists? Is there a more structured way of writing our tweets to make their job easier?
Tweets, I’m sure you know, are limited to 140 characters or less. This may seem like an arbitrarily short character string, but it’s the standard length of most text messages transmitted over cellular networks. Billions of people, including those in developing countries, have access to cell phones and 140 character messages. It is quickly becoming the standard size to encode one human thought.
This is interesting from a literary standpoint as well. Literature is essentially made from human thoughts encoded into character strings then transcribed in some manner. Poets have been able to manipulate these in elegant and nuanced ways to evoke deep emotion. Take, for instance, this poem:
so much depends
a red wheel
glazed with rain
beside the white
A beautiful and terse piece of literature encapsulated into exactly 102 characters, leaving enough room for metadata. If you count up all of the *letters and spaces* you’ll notice it only adds up to 92 characters. This is because at the end of each line there is one or two line breaks, adding an additional 20 characters total to represent William Carlos Williams’ very deliberate and thoughtful typewriter carriage returns. This is a crucial part of *The Red Wheelbarrow *and the poem fails without those 20 characters of code. Therefore we need to encode and pay special attention to subtle changes intended by the author. Here’s what the poem looks like if we use the newline character (from the C programming language) to represent his typewriter carriage returns:
so much depends\nupon\n\na red wheel\nbarrow\n\nglazed with rain\nwater\n\nbeside the white\nchickens.
It is now in a convenient format to send over the network. Whatever is used to render the text on the other end will move the text to the next line whenever it encounters
A deliberately strict use of language can be described as constrained writing, a literary technique in which the writer is bound by some condition that forbids certain things or imposes a pattern. This was used by groups such as Oulipo, a gathering of French poets and mathematicians who devised a “workshop of potential literature” in the 1960’s to invent new structures and patterns. For example there was:
A poem in which each line is a single word, and each successive word is one letter longer.
Working within this constraint requires creativity and wit to pull off successfully. As does this constraint:
A message less than or equal to 140 characters in length.
You can see where I’m going with this. By using a tweet as a form of constrained writing we have created the “potential literature” to describe a link between human thought and machine language. Micro-format fiction and haikus are nothing new and have been explored thoroughly in Japan where phone novels, known as keitai shousetsu, account for many of the top-selling titles. These individual chapters (text messages) are typically much larger than the 140 character limit of Twitter, however, so more information can be transmitted at once.
If computers were communicating with each other via tweet they could easily encode much more information in the same character length. For example, it’s been demonstrated that a recognizable image of the Mona Lisa can be encoded using Chinese characters. The tweet looks like this:
It’s a highly compact, elegantly restrained piece that could transmit flags or icons, but impossible for a human to write quickly on their own. We know that we need special codes to convey moods, ineffable thoughts, and complex emotions (Williams’ Carriage Return), so some level of encoding is needed. Hash tags work great for marking up objects, topics, and locations but are not computational. A glue language, such as Inform, would be needed to add programmability to the hash tags. You could obviously send short snippets of Python, each tweet acting as a shared module, but this is still too obfuscated for many readers. Inform is a language developed specifically for both writers and programmers. From the Inform manual:
Interactive fiction is a literary form which involves programming a computer so that it presents a reader with a text which can be explored. Inform aims to make the burden of learning to program such texts as light as possible. It is a tool for writers intrigued by computing, and computer programmers intrigued by writing. Perhaps these are not so very different pursuits, in their rewards and pleasures.
Inform 7 source code has a syntax which resembles natural english, so it’s fairly easy to read and write. However, it compiles into code understood by a virtual machine (either Z-Machine or Glulx) and can then be run as interactive fiction. The author is responsible for establishing objects, sets, relationships, containers, labels, descriptions, time, objects, and everything else that goes into the story.
By using Inform to structure tweets, its possible to write collaborative fiction easily accessed by future archive computers. This is not to say everything on Twitter should be written in Inform, only that it would be very interesting to have a piece of interactive fiction added to this archive. Since tweets are 140 characters or less, they are roughly the same size as one line of code. Organizing a jumble of code could be extremely challenging, so they would need to be kept fairly basic.
To start, the hash tag #inform7 should be used at the beginning of your tweet as a sort of shebang to indicate the following text is to be interpreted . If you use Twitter this shouldn’t be out of the ordinary for you.
Inform code can be quite complex and describe very sophisticated relationships. For the purposes of using it for tweets it should be kept minimal, so I’ll only go over the very basics. However if you want to learn more advanced techniques, you can always read through the Inform 7 manual for a complete guide to this interesting language.
Inform resembles English sentences. Each line of Inform code is an “assertion” or a rule about the world you’re creating. The world starts as a formless, empty void; you must create every rule for how your story works. There are, however, built-in standard rules so you don’t have to describe the laws of physics, etc. for every story. The primitive data types in Inform are present tense verbs. The built-in primitives are the following:
* to be
* to have
* to carry
* to wear
* to contain
* to support
You can create your own verbs, but for now you should only use these types.
So let’s write a “hello world” Inform 7 tweet! Remember to start with the hash tag, use the primitive data types (for now), keep it present tense, and keep it under 140 characters total.
#inform7 Iowa City is a room.
Not very exciting, but it’s a great start. Keeping with the interactive fiction convention, a ‘room’ is the term for an area of space. You can use the term ‘region’ to describe a larger area that contains a group of rooms.