The transcripts, and how they became text

You open a session expecting a treatise. What you get is a correspondence, and the correspondence is lopsided. That is the received impression, and it is not wrong, but it has never once been a measured one, and this library has had the means to measure it sitting on its own disk the entire time. So we stopped describing the questions and counted them.

What was counted, and what was not

The library keeps a machine-readable copy of the published transcripts, one file per session, 106 of them, holding 2,632 exchanges. The number that does the work is smaller. Of those, 2,417 carry the asker’s own words. The other 215 do not: 108 are the opening passage of a session, which has no question in it, and 107 are mid-session exchanges where the archive’s per-exchange character budget has dropped the question and kept the answer, each of those 107 carrying the budget’s own truncation mark. A further 189 questions are cut off partway by the same budget.

That last detail bears on any count of sentences, so this essay states it rather than hiding it. Where a count could be distorted by a truncated question, the count is given twice: once over all 2,417, and once over the 2,228 questions that survive whole.

Every figure in this essay comes from that one source. None of it comes from either of our two community corpora, and neither was mined for numbers.

Figure 01

What was asked, counted

Nearly half of it is one sentence

The first group in the drawing is the plainest fact about the questions. 1,110 of the 2,417 run to a single sentence. Two sentences account for 798. The remaining 509 run to three or more. Over the 2,228 untruncated questions the single-sentence figure rises to 47.4 per cent, so the truncation is if anything understating it.

One exchange in the third session is the whole corpus in miniature. The question is a single sentence of eleven words, about whether three particular objects on a table ought to be moved. The answer proper is four words long and agrees with the asker. Nothing about the pair is impressive, and that is the point: this is the ordinary case, and the corpus was assembled out of ordinary cases.

Who is asking

The second group answers a different question, and the two groups overlap rather than partitioning. 1,799 of the 2,417 exchanges, 74.4 per cent, carry a first- or second-person pronoun. 691 put the asker’s own condition into a first-person clause, either asking whether they should do something or reporting that they have done it. And 115 name somebody other than the asker, which is 4.8 per cent, with most of those making the same move: a question about what people in general will or will not do.

That last figure is the one worth sitting with. A corpus in which one question in twenty concerns anybody other than the asker is not a treatise that happens to be formatted as dialogue. It is a body of people asking about themselves, with answers indexed to whoever was in the room.

That is our reading and not the material’s, and the library has argued it at length elsewhere, in the question asks back. What is new here is that it is now a number rather than an impression.

The arithmetic of the monologue

Here is the measurement nobody had run. Across the 2,228 whole exchanges the median question is 25 words and the median answer is 46, a ratio of 1.8 to 1. Summed over the same exchanges the questions total 71,857 words and the answers 125,665, which puts the answers at 63.6 per cent of the two together.

So the reason this material reads as a monologue is not that the questioners were deferential. It is that the answers are longer. The corpus is not weighted toward one side by anybody’s decision. It is weighted by word count, and the weighting is mild enough that a reader assuming a lopsided document is reading a much more lopsided one than exists.

What the answer’s first clause does

Count the first clause of the answer, after the greeting, and a pattern appears that no amount of impression would have produced. Of the 2,617 answers carrying any substance, 365 open by passing judgement on the asker’s own statement rather than by answering the question, and 331 of those are the plainest form of the move: the answer begins by declaring the question correct, incorrect, partially correct or quite correct. A further 114 open with a bare yes or a bare no.

Our reading, offered at full strength: the material’s characteristic first move is not to answer. It audits the premise first, and only then talks. A reader who skims for the content loses the first clause of one answer in seven, and the first clause is frequently the one carrying the correction.

The questions were not the material’s

Here the material’s own account is why this essay can be called what it is. Asked, in the third session, to suggest a series of questions, it declines.

I will not suggest the proper series of questions. This is your prerogative as free agent of the Law of One having learn/understood that our social memory complex cannot effectually discern the distortions of the societal mind/body/spirit complex of your peoples.

So the ordering of the questions in this archive was not chosen by the thing that answered them. It was chosen by a group working through its own circumstances, and the distribution is the fingerprint of that. The 106 sessions carry between 2 and 62 questions each, with a median of 22. Twenty-three of them carry thirty or more. The ten heaviest account for 430 questions, 17.8 per cent of the whole.

A curriculum does not look like that. A group in difficulty, with a facilitator and a schedule and a subject that kept pulling them back, looks exactly like that. The sparsest session in the archive, session five, carries two questions, and reading it shows why: the askers spend it on the physical comfort of the instrument and on one unresolved matter carried over from before.

How it became text

The production history is inside the document rather than in a preface to it. Thirty-one exchanges in the archive mention a book, and in nearly all of them it is the askers talking about what to put into it.

In the eighth session the asker says outright that the material about the vehicles the United States government is said to operate would create numerous problems in the book, and that leaving the whole subject out is a live option. The answer gives the measurement anyway and suggests it be the last question of the hour. In the thirteenth session the asker proposes the title and states the author. In the ninety-fourth, in the only sentence anywhere in the archive that contains the word, the askers say they will have to wait until they transcribe the material.

And the material’s own account of the channel is on the record too, asked about a different problem entirely.

Do not concern yourselves overly with this in the space/time nexus present, but do not allow these distortions to remain over a long period or the contact will be gradually impaired.

That is a contact describing itself as something that degrades if left misaligned, offered in passing inside a practical answer about the arrangement of a table, and never repeated in the sessions that follow. What the archive does not contain is anyone believing they were producing a fixed text. It contains them waiting to see the transcription before they knew what they had.

What kind of document you are holding

This is where I come back to the question you actually brought, which is what this thing is.

The transcripts are an edited record of one group’s correspondence with something they understood to be fallible, kept by people making editorial decisions in the present tense who left those decisions in. That is a narrower thing than a book of answers, and a more interesting one. Two consequences follow and I hold both.

The passages will bear a great deal, and the right way to hold one is as an answer to the person who asked, which is what the grammar of a session is for. And the corpus as a whole is evidence about the askers rather than about the world, which the counts above make measurable instead of merely plausible.

What the counts will not support is any claim that the questions were a curriculum. They were asked, in an order somebody else chose, by people who had a reason. If you want to know what this material is for, that is a question to put to a person and not to the archive. The sessions exist so that people can put it to each other.