OR KI ZA HEX

This is a set of field notes rather than a method, because the interesting material is in the months where the working did not work. The aim was creativity, in the specific sense of reliably producing finished pages rather than occasionally producing good sentences. A sigil was built, charged, and kept in a notebook for roughly eight months. The machine ran the whole time. What it produced was a large volume of unpublishable writing, and the reason traces back to the aim itself, which was a noun where it should have been a number. The doctrine underneath this is the standard one: the muse is treated as an external source of inspiration, and the sigil replaces that source with an internal mechanism of commitment, which removes the excuse of waiting for permission. Austin Osman Spare's contribution was to promote specular writing and a fixed daily output, both of which are constraint rather than inspiration. Constraint of that kind reliably increases creative production, which is the same principle a sigil imposes. The most common diagnosis of a creative block is that the block is a judgement about the quality of the work, not a shortage of capacity. That diagnosis turned out to be correct here, and the eight months were spent learning how to work with a process that produced mediocre output on purpose.

The working, and what it was supposed to produce

The written aim was to write more. It sat in the notebook for eight months, charged once and reinforced every fortnight, and it is the single clearest example in these notes of a sentence that could not fail. Nothing about writing more can ever be shown to have not happened. There was always more, in the sense that the word itself was never in question.

Rewriting the aim as a countable output is the single change that produced results, and it should have happened in week one. A workable version names a number, a form and a condition: five hundred words, unedited, before reading anything I had written that day. Nothing about that can be satisfied by intention. Either the words exist or they do not.

Month one, and the output that was not good

The first month produced roughly the expected word count and almost nothing publishable. I had assumed the two would arrive together, which is the assumption the muse model encourages. They do not. The count goes up and the quality stays flat, sometimes for months, and the flatness is usually read as evidence that the practice is not working.

What the flatness actually was, I think, is the absence of a selection step. Without a decision about which of the day's pages to keep, everything is equal and nothing is finished. The count was never the hard part. Deciding what to keep was the hard part, and no sigil made that decision for me.

The aim that could not fail

An aim that cannot fail is worse than no aim, because it produces a feeling of progress while nothing changes. The test is simple: describe the evidence that would show you had not done it. If you cannot describe that evidence, the aim is decorative. Write more cannot fail. Miss a day of five hundred words can fail, which is exactly what makes it useful as a specification.

There is a second version of the same problem that is subtler. An aim can be countable and still not matter, as with the month I spent producing a thousand words of notes on a subject I had abandoned. The count was met every single day. The number was accurate and the activity was worthless, and nothing in the mechanics flagged it.

Why the constraint only worked when it was external

Self-imposed constraints failed about half the time. The same constraints, imposed by a deadline set by somebody else, worked almost every time. This is not a magical observation; it is the ordinary result of a commitment device that has teeth, and it is worth naming because the sigil is supposed to be exactly such a device and mostly is not.

The fix that eventually worked was to hand the specification to someone. A writing partner who asked for the pages on Friday, or a client who had already paid for them, created the external condition that the sigil alone could not. This is the honest version of what the sigil does: it holds a decision in place that your own attention keeps renegotiating. When something else holds it too, the renegotiation stops.

Monetising, and the month that taught me nothing

Partway through the eight months I retargeted the working at income, which is the most common reaim in this material and the one I would now rank lowest. Money is a strong symbol precisely because it is countable, which means it cannot fail either, and it added a second unmeasurable quantity to a practice that was already short of measurements.

The month produced no useful information. I could not tell whether the writing improved, whether the aim was working, or whether the two were related, because I had replaced the countable output with an uncountable one. Reverting to the word count after that month felt like relief. The symbolic target had removed the only instrument I had.

The muse never arrived, and the work did

Waiting for a mood of inspiration was the habit I started with and the habit I finished without. The substitution that the sigil performs is exactly this: an internal, scheduled mechanism takes the place of an external, unschedulable source. Nothing arrived from anywhere. The pages appeared anyway, at the same time each day, which is a considerably less romantic result and a more useful one.

The specific discovery was that the quality of a session correlates with whether I knew what the page was for before sitting down, not with whether I felt inspired. Inspiration is a mood that arrives after a decision has already been made, and mistaking the mood for the cause is what made the first six months unproductive.

Three failures with the same root

Three separate things failed during the eight months, and reviewing them together showed one cause. The sigil stopped being consulted. The word count was quietly abandoned for a fortnight. The output stayed unedited for six weeks. All three are the same event: nothing external required any of it.

The common root is that the sigil had no trigger. The original specification had a time of day and a page count but no external event, so its only activation was my own intention to activate it, and intention is exactly the resource the practice was supposed to be replacing. A trigger attached to something outside my attention, such as arriving at the desk, was the change that made the rest hold together.

What the field notes actually show

Counting only what can be counted: over eight months, approximately one hundred and thirty sessions logged, one hundred and nineteen met the word count, twenty-one produced something I would now call finished, and four produced something I was willing to send out. The gap between finished and sent out is the interesting number and it is not small.

The first hundred and thirty days produced nothing publishable. Everything publishable came after the seventh month, and the seventh month is when the trigger was added and the aim was rewritten as a countable output. Both changes landed together, so I cannot separate them, and I would not trust anyone who claims to.

A number that could be checked

The version of the aim that finally held was: five hundred words, sitting down, before any reading of my own work, with a title written at the top of the page beforehand. Every element is checkable by an observer. The title is the least obvious and probably the most important, because it forces a decision about what the page is for before any words exist.

For eight months the sigil was attached to a mood and a topic. It is now attached to a number, a time, a condition and a title. The mechanical description of the sigil as a compressed statement of intent turns out to be a description of a specification, and a specification is a list of checkable conditions. The compression is the decoration; the conditions are the machine.

The one practice that survived the failures

The only element that was never abandoned was the written page, and specifically the act of starting a new document rather than appending to an old one. A fresh file has no precedent in it, which is what makes the sitting possible on a day when nothing is forthcoming. Appending requires continuing something, and continuing is the step that fails.

This is the smallest of the practices in these notes and it did more work than any correspondence or timing scheme I tried. The rule became: new document, title first, five hundred words, no reading first. Everything else in the working is optional and most of it was optional because I wanted it to be necessary.

The aim I would write now

If I were writing this specification again, the aim would name the artifact rather than the activity. Not five hundred words but one finished piece, defined in advance as between eight hundred and twelve hundred words, drafted in three sessions, sent to one person by a date. The compression to a sigil is then trivial and the check is whether the piece exists in someone else's inbox.

I would also write the retirement date into the first page, which I did not do the first time. The working ran eight months because nothing told it to stop, and the last three of those months were maintenance of a habit rather than any kind of practice.

What I would tell someone starting

Do not start with a feeling. Start with a number, a time, a condition and a title. Expect the first hundred days to produce nothing you would show anyone, and decide in advance that this is acceptable, because the alternative is deciding it in week three when the first flat fortnight arrives.

The mechanism is not mysterious and it is not mystical. It is a written specification, a trigger attached to something outside your attention, and a repetition that makes the specification cheaper to follow each morning until the following costs less than not following. That is the whole of it. Everything else is a way of describing the same arrangement in language that sounds older.

Frequently Asked Questions

How long before creative work produced by a sigil is any good?

In these notes, roughly one hundred and thirty sessions, spread over about eight months, before anything publishable appeared, with the first usable output in the seventh month. The count was being met from the first fortnight, so the delay was not a failure of volume. Anyone promising finished work in the first month is describing a different kind of project or is describing luck.

Should the aim be a feeling or a number?

A number, with a form and a condition attached. An aim phrased as a feeling cannot fail and therefore provides no information, and an aim phrased as a number can be checked by an observer. The sigil's description as a compressed statement of intent works best when the statement underneath it is a list of conditions that could be verified.

Why did my self-imposed constraints stop working?

Because they had no external component. Self-imposed constraints failed about half the time in these notes, and the same constraints set by someone else's deadline worked almost every time. The sigil holds a decision in place while your own attention renegotiates it; adding a second holder, such as a person who expects the pages, is what stopped the renegotiation.

Is it worth aiming a sigil at money?

These notes say no, and specifically say that the change produced a month with no usable information. Money is strongly symbolic largely because it is countable, which is the same property that makes it unable to fail as an aim. Replacing a measurable output with an uncountable one removes the only instrument available for telling whether anything is happening.

What do I do when the output is consistently poor?

Keep the count and add a selection step, which was the actual missing piece in the first month. Producing a daily volume with no decision about which pages to keep leaves everything equal and nothing finished. The quality gap usually closes when the practice includes deciding what to keep, not when it includes writing more.