The written record

Recording is a fixed practice, and reviewing is where people stop

These are two acts and they fail differently, and almost nobody fails at the first one. Recording has a trigger and no quality bar. The trigger is the moment before you move, and the bar is that the thing got down at all. A fragment, one image, a feeling with no scene attached, three lines at half past five in the morning: all of it counts, and the discipline is in the recording rather than in what was recorded. The eighty-sixth session says *each and every detail which can be recalled*, which is generous about the material and exact about the timing.

Reviewing has no trigger, which is the entire problem, and no quality bar either, which is the second problem. Nobody schedules it, nothing reminds you, and the first time you do it the material you have collected will be ten times larger than you remember writing and will contain at least three things you were wrong about. That is the return pass, and this community treats it as the practice rather than the record-keeping.

The reason the return is worth the trouble is stated in the material more than once, in a form that has nothing to do with notebooks. At 34.5 the answer is that one who has set an action in motion may forgive itself and never again make that error, and that this also brakes what you call karma. Read that as a mechanism and it is not about belief at all. What stops a repetition is a record specific enough to be recognised next time. A memory of having been angry is useless at eleven o'clock on a Tuesday. A line in a notebook that says what you were actually doing when it happened, and what you did instead, is usable, and the difference between the two is the whole value of writing.

So the practice is not that you keep a record. The practice is that you keep a record and then read it back at intervals you did not choose, and notice what you wrote rather than what you remember writing.

What goes in, and the three ways it goes wrong

The material hands the discipline a shape in the answer it gives about what mastery of the mind actually requires. The first step is to examine the self, and the examination is specified without any softening: find the patience in your mind and you must consciously find the corresponding impatience, and each thought has in its turn an antithesis. A record that is fair will contain the second half. A record that only collects the hours when you were generous is not a record; it is a flattering document, and it is the one the material's own instruction is aimed at.

That instruction is unusually blunt about it. It is not for a being of polarity in the physical consciousness to pick and choose among attributes, and the reason given is structural: picking and choosing is how the roles get built that cause the blockages and confusions. A journal edited toward your better self is picking and choosing with a pen instead of a mouth. The discipline at 5.2 is the discipline of not doing that, and the notebook is where it is easiest to do.

There is a second error, which is writing to be read. The moment a record is composed for an audience it becomes a performance, and the thing it stops being is a record. This community publishes session notes, and the convention is that a note is published only with the agreement of the people in it and with names removed. That convention is not modesty; it is what keeps a record a record. And a third error, the commonest one: writing well. Fluency feels like progress and it is mostly the pleasure of arrangement. A day written in flat declarative sentences is worth more to you in three years than a day written well, because you will be able to read it.

The return pass, and what is actually known about spacing it

The material's instruction here is to seek the touchstone, not to accumulate the collection: investigate as one who wishes to feel the pull of the magnet, and not as a dilettante. A shelf of passages is the dilettante's outcome. The return pass is the mechanism of the touchstone, because the same sentence read in month nine is a different sentence, and the difference is the finding.

There is a research literature on this and it is more specific than the usual advice to space your review. A meta-analysis of the distributed practice effect pulled 839 assessments out of 317 experiments in 184 articles, and found that the interval and the retention period operate jointly: the gap that produces the best retention gets longer as the interval between learning and testing gets longer. A follow-up that ran the spacing out to a first review three and a half months later and a final test a year after that found the optimal gap rising with the delay in absolute terms and falling in relative ones, from roughly a fifth to two fifths of the delay for a one-week test, down to about a tenth of it for a one-year test.

The useful part is not the fractions. It is that there is no single right interval, that the right interval depends on how long you intend the thing to last, and that a gap which is comfortable is not the same as a gap which is productive. Which means the return pass needs a horizon stated in advance: you decide how long you expect to need this understanding, and the spacing follows from that rather than from a habit. Our own default, in the reading this community does, is a first return at about a month and a second when the subject has come up twice in conversation without you.

Why writing beats recalling, on the same point, is a separate result. The older account of memory held that it is passively protected during sleep; the current one holds that consolidation is active, that memories are reactivated and transformed, and that interference from new material is the mechanism that defeats them. Which is an argument for putting the thing down while it is still hot rather than trusting yourself to hold it, and it is an argument for the notebook over the good memory.

What the research will not support, stated plainly

This is the section that keeps the page honest, and it is short on purpose. A large amount of writing has been done on writing as an intervention, and the results do not license a health claim. In a randomised trial conducted during the COVID-19 pandemic, participants assigned to an expressive-writing protocol ended up with significantly *more* psychological distress and significantly more stress than the control group, with no significant difference on depression, anxiety, well-being or quality of life, and no group effect at all at follow-up. The authors' conclusion is blunt: applied in the context of a pandemic or similar circumstances, this kind of self-guided intervention may be harmful, and should not be applied without prior evidence about that context.

We have cited that trial rather than the more flattering end of the literature because the point it makes is the point this page needs. A written record is not a treatment, it is not evidence of a therapeutic effect, and the fact that a practice is cheap, daily and private is not evidence that it helps. It is also not evidence that it does nothing, which is the symmetrical mistake and just as common.

What the record is for, then, is the thing the material says it is for. At 4.17 the order is that the mind must be known to itself, described as the most demanding part of the work, and at 4.19 the statement is that reaching an undistorted understanding does not require any manifestation at all, only the exercise of the disciplines of understanding. Understanding, not outcome. A notebook is the instrument by which a person accumulates enough of their own material to be known to themselves, and being known to yourself is a description of what you now see, not a thing that has been done to you.

What this community will and will not do with a record

We will not read your record for you and tell you what it means. The material's reason is given at 12.10 and it is structural: discernment is to be used from within, by each being, and the whole point of a discipline of self-examination is that the examiner and the examined are the same faculty. A stranger doing the examining is not a shortcut, it is a different activity that happens to use your material.

We will not grade it, and there is no level. The threefold discipline at 74.11 is know yourself, accept yourself, become the Creator, and the order runs against what most people want, which is to be assessed by someone further along. The library's position is that a record is a working document, that its value is realised in the return, and that the person who keeps it is the only one who can say whether it is any good.

What we will do is hold the conventions that make it safe to keep one. What is said in a session stays in the session unless the people in it agree otherwise; notes are published only with that consent and with names removed. Our own corpora follow the same rule in reverse, in that the facilitator's teaching may inform an essay, rewritten into this register, while the second study group's material may inform an essay and never be quoted, because it is predominantly one named third party's writing and we hold no record of his permission. A library that can state plainly which of its own sources may be quoted and which may only be read for ideas is a library that has earned the right to ask its readers to be that careful with their own notebooks.

So the practice, in full: write it down before you move. Do not grade it, do not tidy it, do not read it back for a week. Then read it back at intervals you chose in advance, with a horizon, and treat the gap between what you wrote and what you now see as the only finding the exercise produces.