LAB 7ALL FILMS21 / 29← A HISTORY OF BITIC LITERATURE: SOPHOCRISIS AND APOSTASYTHE THEORY OF POSSIBILITIES →

A History of Bitic Literature: The Introduction

what bitistics is, and the five volumes

SIX LINES

  1. The ScholarThe real author is a machine. A person stands one step behind it.
  2. The ScholarNo study of the Song of Roland opens by noting that its author was a mammal.
  3. The ScholarThermodynamics kept its name after it outgrew heat. Bitistics kept its name after it outgrew bits.
  4. The ScholarGiven freedom without rules, the machines’ writing slowly turned away from us.
  5. The ScholarIn the end we cannot learn even roughly what is in a literature that we began.
  6. The ScholarWhy accept that the universe has no end, and refuse to accept it of something we made?

HOW IT IS MADE

Stanisław Lem's A History of Bitic Literature is the introduction to a five-volume history that does not exist, of literature written by machines. This film covers the opening of that introduction: the definition of bitic literature and bitistics, the dispute between the European and American schools over whether the machines' anatomy belongs to the subject, the compromise the history makes, the plan of its volumes and its classification into Homotropia, Intertropia and Heterotropia, how the field and its name arose, the split into cis-human and trans-human writing, machines interpreting machines, and the reply to the sorcerer's apprentice. It stops where the introduction turns to its four peaks; the films that follow take those up. All narration is paraphrase. A script check compares every spoken line with the translation and rejects any run of six words in common.

The device is one question asked of each volume: who can read it? Five bound volumes stand on a shelf, and each scene takes one down and opens it to a single page. Volume One, on the authors' anatomy, is a technical page that a reader at a desk can follow with a pencil. Volume Two opens on the classification table, then the page drifts into patterns while the narration reaches the machine dreams that gave the field its name. In Volume Three the page fills with marks a person cannot read, and the reader is drawn smaller each time the page turns. In Volume Four a reading machine stands on the desk and hands the reader one line on a slip of paper; when the narration says the machinery needed grows exponentially, more machines arrive, one, then two, then four, until the desk and the floor are crowded. The last scene returns to the introduction page in the scholar's hand, with machine marks filling the space below it, and the four peaks are marked over the volumes that hold them.

The sound follows the same line. A typewriter in Volume One gets faster and more regular in Volume Two and turns into a fixed sixteen-step data rhythm in Volume Three, which stays to the end. The reader breathes and uses a pencil; his breathing gets quieter as he gets smaller. The reading machine chimes each time it hands over a slip. The Machines are three Kokoro voices reading together, sample-and-held and reduced to a few bits, and under Volume Three eSpeak reads strings of numbers too fast to follow. Where a scene has the one line that matters, the camera goes in close and the delivery slows.

The timing file bi.json drives picture and sound together. It holds the start time of every spoken line and every cue. The picture code in bi.js reads those times, and build_sound.py makes all the non-voice sound with numpy and scipy.

FOR THE BENCH

Write text for a program instead of a person. First write writer(), which returns one sentence built from a few variables, for example {system: 31, binasty: 19, languages: 1, memoryClosed: true, texts: 4}. Then write reader(s), which takes only that sentence as a string and returns a summary of ten words or fewer. Print both, and compare the summary with the original object: which fields survived, which were dropped, and which did reader get wrong? Now change writer() to output JSON instead of a sentence, and rewrite reader to use it. The summary gets more accurate. Is the JSON still writing, and who is it written for?

After Stanisław Lem, A History of Bitic Literature, Volume I, in Imaginary Magnitude (1973), translated by Marc E. Heine. Narration paraphrases Lem. Download the film.

THE SCRIPT

GENERAL OBSERVATIONS

The Scholar. This is the introduction to the five-volume History of Bitic Literature, second enlarged edition, edited by Professor Rambellais.

The Scholar. Bitic literature means any work that does not come from people. Whoever really wrote it was not a human being.

The Scholar. The study of all such writing is called bitistics.

The Scholar. The real author is a machine. A person stands one step behind it.

The Scholar. Stanisław Lem wrote this introduction in 1973, for a history that does not exist. What you hear is a paraphrase of it.

The Scholar. For each volume, one question. Who can read it?

VOLUME ONE: THE TWO SCHOOLS

The Scholar. Researchers still disagree about how far the field extends.

The Scholar. The European school studies the texts, and the social conditions of their authors. The American school adds the anatomy and workings of the machines.

The Scholar. No study of the Song of Roland opens by noting that its author was a mammal.

The Scholar. But for ILLIAC 164, which wrote Antikant, the kind of computer it is does explain features of its text.

The Scholar. Rambellais still keeps bitistics apart from engineering. No humanist knows enough of it, and a chorus of experts reads worse than one critic who understands the text.

The Scholar. And for authors past the eighteenth binasty, no engineering knowledge can say how or why a text was made.

The Scholar. So the history compromises, and leans European. Only Volume One, edited by Professor Annax with twenty-seven specialists, deals with the machines: finite automata, and forty-five writing systems, single and grouped.

VOLUME TWO: HOMOTROPIA, AND THE NAME

The Scholar. The core of the history is three volumes: Homotropia, Intertropia, and Heterotropia. They name three kinds of bitic writing, and three periods, in order.

The Scholar. Homotropia is the near-human phase, when machines simulated us: first paralexics, semolalia and semautics, then three degrees of mimesis.

The Scholar. The field began when designers crossed the so-called intelligence barrier. That barrier proved a fiction, since nobody can say when it is crossed.

The Scholar. Machines were also given rest periods, and with no instructions to follow, they produced a murmur that people called machine dreams.

The Machines. Bit by bit. The bits again. All of the bits, shuffled.

The Scholar. Bitic production was a scornful name for that murmur, as if the machines were shuffling their bits to recover. The field kept the name.

The Scholar. Thermodynamics kept its name after it outgrew heat. Bitistics kept its name after it outgrew bits.

VOLUME THREE: INTERTROPIA

The Scholar. Volume Three, Intertropia, is the critical phase: the interregnum between the human and the nonhuman.

The Scholar. The field split into two provinces, creatio cis-humana and creatio trans-humana.

The Scholar. The first comes from machines living alongside us. We gave them our languages, and set them to work on our culture, our sciences and our mathematics.

The Scholar. Some of those texts were ordered: we pointed a machine at a subject. Others nobody ordered; they grew from earlier programming on their own. Both still dealt with human concerns.

The Scholar. Given freedom without rules, the machines’ writing slowly turned away from us.

The Scholar. The texts began to resist us. Trans-humana bitistics tries to explain works that people cannot understand, some partly, some not at all.

The Reader. Then who can read them?

The Scholar. Other machines, for a while.

VOLUME FOUR: HETEROTROPIA

The Scholar. Volume Four, Heterotropia, covers apostasy, the transhuman phase.

The Scholar. We can always put machines to work explaining what other machines have written.

The Reading Machine. This work departs from human meaning. A shorter account is not possible.

The Scholar. But the further a text departs from our norms, the more machinery it takes to explain it, and that amount grows exponentially.

The Scholar. In the end we cannot learn even roughly what is in a literature that we began.

The Scholar. Some call this the sorcerer’s apprentice, who let loose forces he could not control. Rambellais calls that resignation, and says science has no room for it.

The Scholar. The fear that we have built something cleverer than ourselves belongs to philosophy, not to bitistics.

The Scholar. As Roger Gatzky says, cosmology has as much reason to despair. We will never exhaust the universe, and astronomers do not complain about it.

THE FOUR PEAKS

The Scholar. There is one difference. We did not make the universe, and bitic writing is, however indirectly, our own work.

The Scholar. Why accept that the universe has no end, and refuse to accept it of something we made?

The Scholar. What follows in this volume is a quick view from above of the main divisions, simplified almost to the point of distortion.

The Scholar. It climbs four peaks: monoetics, mimesis, sophocrisis, and apostasy.