Primary Information Survives Only When the View at the Time Is Separated From the Later Commentary

A first commission, quoted at twenty hours, actually took fifty-eight. This event can be written two ways.

The first: Early on after going independent, my estimates were poor and I had a hard time. Estimating hours comes down to experience, so it is better to allow extra at the start. The second: I counted the work, applied the hours a past job of similar size had actually taken, and quoted twenty. The client agreed on the spot. It took fifty-eight. The gap came not from the work itself but from the back and forth of checks and rework.

Both are written from the same event, and the first retains not one piece of information that only that person holds. Estimates are poor, experience matters, allow extra. Anybody who has not lived through that event can write all of them. The experience was used as an input and did not remain on the output side.

The division runs as follows. First, acquiring and describing are separate stages, and that acquiring alone does not put primary information onto the page. Second, why that conversion occurs without the writer seeing it. Third, what operation stops the conversion.

Things that do not get settled here also go down first. How many characters to give the experience and how many to the commentary does not emerge; it varies with the subject and with how much material the reader already holds. And whether the primary information you wrote holds value for the reader cannot be confirmed from inside this process either, since writing that has stopped the conversion does not thereby become important writing.

The direction of the misdiagnosis settles the harm. Somebody who feels their primary information is short moves toward increasing experience — rarer experiences, more unusual jobs. Yet if what is short sits somewhere other than experience, no amount of increase changes the output. Increasing has no end, and while you are increasing, no occasion arrives on which to check where the jam sits.

And compressed writing is easy to read and looks well gathered. Looking orderly, it does not become an object of inspection. The look of the finished piece does not tell you whether a conversion occurred.

Nor is it visible from the reader’s side. A sentence in the shape of a lesson can be accepted on the spot, because it coheres and has no rough edges. That it cannot be moved into their own situation becomes clear when they try to move it, and by then they have finished reading. No information about the failure to move it comes back to the writer.

How Is the Difference Between Primary and Secondary Information Explained

Primary information is information obtained through your own direct experience, or acquired by investigating or testing yourself. Information obtained by reading what somebody else compiled is called secondary, and the two are distinguished by distance from the source.

The reasons given for its value are distinctiveness and reliability. Rearranging information that anybody can find by searching yields the same thing as everybody else, whereas information you acquired yourself exists nowhere else, so a reason to read it arises.

Ways of gathering get shown concretely. Test something yourself and record the result. Ask clients or users directly. Run a questionnaire. Go to the site. Measure figures. All share the point that you go and get it.

The point that you must test while verifying, rather than merely experience, comes attached: settle what you wanted to confirm by doing it, and the result stays as information.

Advice to connect it to the reader’s problem also appears: rather than putting your material out as is, combine it with the problem the reader carries before moving to structure. Including an angle few people look at is recommended too: even on a subject everybody experiences, distinctiveness emerges from handling a side that goes unregarded.

That distance from the source settles value is a fact, and that holding a verifying angle in advance leaves a record is true as well. Connecting to the reader’s problem does function. As a record of practice, nothing here is off.

All five accounts share one premise. Primary information is organised as a problem of acquiring. Once you have it, all that is left is to write, runs the premise.

None of this argues that acquiring is unnecessary. Nothing can be written from having experienced nothing, and without records the details do not survive. The stage of going to get it is genuinely required.

What goes blank once it is organised as acquiring is the stage of turning experience in hand into writing. With that stage missing, talk proceeds on the premise that what is in hand is already primary information — that only arrangement and expression remain.

Sometimes it is not. The experience is in hand, and what emerges at the moment of writing wears the face of secondary information. With that stage blank, the account of the jam flows toward I need rarer experiences.

There is no end at the far side of that flow. Rarity is a relative quantity, so it never becomes sufficient however far you go. And going to acquire rare experiences has more of a feel to it than writing does. The work with a feel to it gets chosen, so the jam continues a long time.

The Same Experience Becomes Secondary Information Depending on How It Is Written

Acquiring and describing are separate stages. What was acquired does not automatically arrive on the page as primary information.

The two ways of writing the twenty-hour estimate serve as a concrete example of primary information as it stands. What occurred in the first is that the event was compressed into a lesson. What dropped out in the compression is the specific figures, what was looked at when judging, and what was not known at the moment of judging.

You can read it as meaning that putting it in the form of a lesson is what makes it useful to the reader. There are cases where it is useful. The lessons themselves, however, are already in circulation. Allow extra on estimates comes up endlessly in a search. What the reader receives from that writing is a re-confirmation of what they already know.

What does not get received is the conditions under which the lesson holds. In what situation, overlooking what, by how much does it go wrong. That sits only in the part the compression dropped.

The compressed form is correct in some settings. In short communication to somebody with whom the premises are already shared, explaining the conditions is redundant. Allow extra gets through where the other party already knows the circumstances under which it goes wrong.

The two part on whether the reader can move that lesson into their own situation. Moving it requires conditions, and conditions drop out in compression. So in writing addressed to somebody reading for the first time, the compressed form cannot be used.

And the dividing line is hard to see from the writer’s side. The writer knows the conditions, so reading the compressed sentence they read it with the conditions supplied. That it makes sense on your own reread is not evidence that the conditions are written.

People who say they have no primary information may in fact have it. They convert what they have into secondary information every time, at the stage of writing. In this case, increasing experience does not change the situation. So long as the habit of conversion is unchanged, the added experience gets compressed in the same form.

The conversion happens without awareness, because writing in the form of a lesson is easier to read and looks better gathered. Where the finished piece looks orderly, the likelihood that compression has occurred is what gets looked at first.

Compression has degrees. The form that dropped only the figures. The form that dropped the figures and the order of judgement. The form that dropped the event itself and kept the lesson. The more that dropped, the shorter, the more correct, and the closer to something read elsewhere the remaining sentence becomes. And a short, correct sentence does not get selected as a candidate for rewriting, because no reason to fix it appears on the sentence’s side.

Once the Conclusion Is Known, the Self Who Did Not Know It Cannot Be Recalled

Compression is not a problem of the writer’s posture. There is a working on the side of memory.

Once a result is known, the degree to which one feels it could have been predicted rises. This is called hindsight bias, a phenomenon Fischhoff reported (Fischhoff, 1975, Journal of Experimental Psychology: Human Perception and Performance, 1(3), 288–299).

There is also an experiment that checked this on the side of memory (Fischhoff & Beyth, 1975, Organizational Behavior and Human Performance, 13(1), 1–16). Its structure is simple. Have people estimate probabilities for an event before it occurs. Then tell them the result and have them recall what they answered before knowing it. They report values shifted toward the result, away from what they actually answered.

The important point is that they are not lying. They are trying honestly to recall. What they recalled has been overwritten by information obtained afterwards.

What this produces at the stage of describing: writing about an estimate that went wrong, you write from a state in which you already know it goes wrong. The you of the time did not know. The you doing the writing does.

As a result, information the you of the time did not hold gets mixed into the description of the time. I was estimating too optimistically is such a sentence. At the time you did not think you were estimating optimistically; you thought it reasonable and sent it. Optimistic is an assessment obtained afterwards.

Memory does not disappear. Ask about details and much can be recalled. What gets overwritten is the sense of how certain each thing looked at the time. Not the memory of the facts but the memory of the certainty is what moves.

The prospect of avoiding it by staying aware is hard to sustain. Replications have accumulated in which explaining the phenomenon and warning people about it barely narrows the shift. Knowing something as knowledge and changing how recall works are different things.

It is avoidable only where memory is not relied on. A record written immediately after the fact belongs to a moment when the result was not yet fixed, so it does not get overwritten afterwards. The reliable way of holding the position of the time is to write from that position at the time.

Writing of the form I should have done this back then presupposes information that did not exist at the time. So the reader cannot move that judgement into their own situation, because the reader stands on the side that does not know the result. That it cannot be moved is not apparent while reading. The sentence holds, so it passes with the reader satisfied.

Separate the View at the Time From the Position of Commentary

There is one remedy. Separate the paragraph that writes the experience from the paragraph that comments. Do not mix them.

In the experience paragraph, write only what was visible at the time and what was not known at the time. Put in no assessment words. Optimistic, inexperienced, looking back now are all words from a position obtained afterwards.

Separated into experience and commentary, the estimate example looks like this.

The experience paragraph runs like this. The commission was of a kind already handled before. I counted the work, applied the hours a past job of similar size had actually taken, and quoted twenty. The client agreed on the spot and I began. Fifty-eight was the actual figure. The gap came not from the work itself but from the back and forth of checks and rework, which had not occurred in the past jobs.

The commentary paragraph runs like this. What produced the gap was not a misreading of hours but the condition that the number of checks is settled by circumstances on the client’s side. In past jobs there was one contact point; in this one, three people entered the checking. It is a variable invisible from the volume of work.

Separated, what is happening becomes visible. From the experience paragraph alone, the reader feels that quoting twenty hours looks like a reasonable judgement. So the figure fifty-eight carries meaning. The information that a gap of nearly three times can come out of a reasonable judgement gets through.

The point that readers want the conclusion first holds. At the level of the whole article, the claim comes first. Separating does not mean reversing the order. It means placing the conclusion first and then, in the description of experience that supports it, holding the position of the time.

The two do not conflict. What the reader wants first is what the piece is about, not how the writer finally assessed things partway through the experience. What goes first is the claim, not the impression.

The unit of separation is not necessarily the paragraph. In short passages, the position can switch inside one sentence. The adjudication is not by form but by which side information not known at the time has been placed on.

Separate and the weakness of the commentary gets exposed. While the two are mixed, the concreteness of the experience hides the vagueness of the commentary. The moment they are separated, it becomes visible that the commentary paragraph can write nothing but so experience matters. This is not a fault produced by separating; what was already there has merely become visible.

The remedy for what got exposed is not to write more. The commentary is thin because the structure extractable from that experience has not yet been identified. Write more before identifying it and what gets added is generality.

Immediacy Rises Out of the Reader’s Memory

That writing from the position of the time gets through has a reason in what happens on the reader’s side.

Somebody reading is not receiving the written image as it stands. Using the written material, they assemble an image out of their own memory. Reading fifty-eight hours, what the reader calls up is a long stretch of work of their own.

Memory has kinds. Tulving distinguished the memory of events themselves from memory as knowledge (Endel Tulving, “Episodic and Semantic Memory,” in Tulving & Donaldson eds., Organization of Memory, Academic Press, 1972). The former contains when, where and what happened; the latter is held in the form of in general, it is like this.

Writing compressed into a lesson reaches only the reader’s latter kind. Allow extra on estimates gets registered as knowledge and mixes with knowledge of the same kind already held, until it cannot be told apart.

A description written from the position of the time touches the former. The reader calls up a similar scene they lived through and places the written figure inside it. Only then does something move on the reader’s side.

The point that this depends on the reader’s memory and cannot be controlled by the writer holds. Control is impossible. Design is possible. Which memory gets called up cannot be specified, but which region of memory gets used is settled by the choice of material. Write a length of time and the memory of time gets called; write an amount of money and the memory of paying gets called.

So you select material of the kind that corresponds to the structure you want to convey, and the vague task of writing so it reaches the reader turns here into the choice of which kind of memory to use.

Where the reader holds no material of their own, this does not work. A description of a wholly unfamiliar territory passes without an image rising. Here you build a bridge to other material the reader does hold. That method is handled in how to use analogies.

Write too much detail and the room for the reader’s image disappears. Write in the colour of the room and the weather and the reader can no longer use their own memory; they merely watch the writer’s image. What is needed is enough material for the reader to call up their own memory, not a complete reproduction.

The quantity of detail therefore has two directions of error. Too little and no image rises; too much and the reader’s memory cannot enter. What can be adjudicated is whether that detail bore on the judgement. Detail that bore stays; detail that did not gets dropped.

Separate Observation From Assessment

Removing assessment words from the experience paragraph is the most concrete operation here.

An observation is what anybody present could record identically. The time, the amount, the count, the words spoken, the order. An assessment is the judgement passed by whoever saw it. Hard, easy, good, bad.

Writing composed only of assessments can be written by somebody who was not there. It was a hard job can be written without knowing the job. Three revisions and it passed on the fourth can be written only by somebody who was present.

The point that emotion is part of the experience holds. Writing emotion is not what is being argued against. The trouble is that replacing an emotion with an assessment word makes the emotion itself disappear.

Write I was anxious and no anxiety rises in the reader. Write four days passed with no reply and on the fifth I spent thirty minutes deciding whether to send a chasing message and the reader calls up anxiety from their own memory. The way to convey an emotion is to write the conditions under which it arose, not the name of the emotion.

Banning assessment words entirely is not the argument. In the commentary paragraph, assessment is required as the conclusion itself. The separation is needed only in the stretch describing experience.

It can be adjudicated by whether writing that word requires knowing the result. Three revisions can be written without knowing the result. A hard job cannot be written until it is over. That gap overlays the boundary between observation and assessment exactly.

Try to write only observations and the writer discovers afterwards what they were feeling, because the part the assessment words had capped opens up. And the assessment first held and the assessment emerging from the observations may not agree. Something remembered as a hard job turns out, on the record, to have been a job with many rounds of checking.

That disagreement is itself information worth writing, because the gap between the assessment left in memory and the assessment emerging from the record occurs on the reader’s side in the same way.

And writing observations comes out shorter. Assessment words demand explanation. Write it was hard and you must go on to say why. Three revisions completes on its own. Separation is not an operation that lengthens writing.

Coming out shorter has another consequence. A paragraph written in observations leaves the reader room to place a different reading. Write a hard job and the reader receives that assessment; write three revisions and the reader can decide, against their own experience, whether it is heavy or light. Being able to decide means a reading that does not agree with the writer’s assessment also holds. Whether that room stays open or closes is settled by whether an assessment word is placed.

When You Feel You Have No Primary Information, What Is Missing Is the Record or the Question

The state of having nothing to write comes apart. What is missing is not experience but the record or the question.

No record means the details did not survive. How many hours it took, how many exchanges there were, how much it cost. These do not come accurately out of memory. What comes out is the assessment, and an assessment is already a compressed form.

No question means it was unsettled what you were trying to confirm by doing it. Carry it out unsettled and the result survives only as the binary of it worked or it did not. A binary carries little information, so trying to write it yields nothing but the form of a lesson.

The remedy is to settle one line before carrying it out. What do I want to confirm this time. With that in place, what remains at the end is no longer a binary.

The line works when put in the form of a prediction. Write the checks will probably be done in two rounds and what remains at the end is the manner of the miss. The manner of a miss carries more information than a hit. A hit confirms what was already known; a miss identifies where the reading failed.

The point that doing this every time is unrealistic holds. It need not be attached to every execution. Attach it only to the first job of a kind and the records accumulate sufficiently.

There are also stages of genuinely shallow experience. At three jobs handled, you cannot write about a tendency across ten. Here you write that there were three. In two of three, the same thing happened is accurate primary information. Hide the count and generalise, and only there does it drop to secondary information.

On the recording side, the resolution of the experience itself rises. With what to record settled, what becomes visible during execution changes. This is a by-product of recording rather than its purpose, and the effect is sometimes larger than the purpose.

Once records accumulate, the separate problem of which to write rises up. The criterion for selecting here is not interest but the degree to which prediction and result diverged. The larger the divergence, the more information unknown to the reader as well it contains. A record with no divergence is a record you need not write, because going as predicted means an account exists within the shared understanding.

The size of a divergence, however, also grows with the coarseness of your prediction. If the prediction is careless, everything you measure will miss. Not every missed record carries value, so in looking at divergence you also look at whether grounds for the prediction were written. Only a grounded prediction that missed contains information about structure.

Experience With Other People in It Carries a Judgement About What May Be Written

Much primary information does not arise alone. Clients, participants, suppliers appear in it. There is a question here of what may be written.

The judgement divides into three. First, permission. Where you write in an identifiable form, confirm in advance. Confirming after the fact creates a situation in which the other party finds it hard to refuse, so it is not confirmation.

Second, identifiability. Write the sector, the region and the size and the three in combination can sometimes identify the party. A state that is identifiable while intended as anonymous is not anonymous. Write it as something that occurred in common across several cases rather than as an individual case and the problem resolves in most instances.

Third, excerpting. Take out part of what the other party said and the meaning can change. The only person able to adjudicate whether it changed is you, who know the whole. The other party cannot verify from their side, so the judgement sits entirely on yours.

You can read it as meaning that excerpting poses no problem since you are writing facts. Being a fact and having the meaning preserved are different things. At first I thought it was expensive can be quoted as a fact. Drop what followed and the same sentence becomes the opposite information. Excerpting becomes a problem not when you write a lie but when you write only true things.

Nor does keeping the three become permission to write. Whether the other party, reading it as their own case, can agree with what was written. That question survives the procedure.

The procedure does not make the judgement for you. Take permission, render it unidentifiable, quote without dropping the surrounds, and the question of how using that case as material appears to the other party still remains.

What can be placed as a rough marker is whether you could show it to them. Where you are writing in a way you could not show them before publication, the trouble sits on the side of handling rather than of procedure.

Settle the range of what may be written in advance and the manner of asking changes. Free of concern about later usability, you can ask out of interest in the moment alone. Talk heard out of interest alone turns out, in the end, to yield things worth writing. The asking itself is handled in how to conduct a customer interview.

Settling the range has a reverse effect too. Where the unwritable range is large, experience in that territory falls out of the scope of recording, since a judgement operates that what cannot be written is not worth recording. Once the habit of not recording the unusable sets in, no record survives to go back to when the range later widens. The range of recording and the range of publishing therefore get settled separately.

The Place Where the Conversion Occurred Can Be Identified Before Publication

Three checks show whether what you wrote has been converted. All can be carried out on your own before publication.

The first confirmation is whether the paragraph could be pasted under another person’s name in the same field.

If it holds after pasting, that paragraph contains no information only you hold. If it does not hold, it does. The adjudication takes seconds.

The second confirmation is to delete every assessment word.

Erase hard, demanding, excellent, good. If nothing remains after erasing, the paragraph was made of assessment alone. If something remains, observation is in it. If the erased state reads better, the assessment words were unnecessary from the start.

The third confirmation is to see whether information not known at the time has entered the description of the experience.

Passages beginning looking back now, as it turned out, in fact are those. All are words from a position that knows the result. Rather than deleting them, move them into the commentary paragraph. Moved, the experience paragraph returns to the position of the time.

Writing quality is examined by none of the three. Whether a conversion occurred is a question of position rather than of expression.

There is a range the three cannot reach. All of them are usable only after writing. To hold the position before writing, there is something to do at the stage of recording. Immediately after execution, leave the figures and the order without writing an assessment. Immediately after, the result has not yet hardened into a whole picture, so the position of the time stays as it is.

How many characters for the experience and how many for the commentary. How many figures to include. No numbers get placed on these. They vary with the subject handled and with how much material the reader already holds. To hand over a number would be to settle the allocation on your behalf without looking at those conditions.

What divided the two ways of writing the same event was neither the quantity of experience nor skill in writing. It is the position the writer stands in. Write from a position that knows the result and a lesson comes out; write from a position before knowing it and an event comes out. And the position is invisible to the person writing, because what the you of the time did not know is buried inside what the you of now knows.

A paragraph somebody who does not know the result could also have written stands at the position of the time. A paragraph only somebody who knows the result could write is a candidate for moving to the commentary side. Which of the two you are holding comes out of three checks: swapping in another writer, deleting assessment words, and looking for what was not known at the time. Where it goes after that is settled by the writer.

Bridging experience into the reader’s memory is handled in how to use analogies, and the order for presenting experience in narrative form is in writing from personal experience. Where what to write does not settle in the first place, what to do when you do not know what to write sits upstream. Separating the position at the time from the position of the commentary is, before it is a matter of craft, a matter of whether that distinction exists inside you. The order itself — the distinction first and the words after — is handled in what verbalisation is.