The ten scenes this paper read, in the order of its case list. Each is a GAEKJU synthetic scene: written for comparison, in GAEKJU's words, and not testimony from its trade; written in Korean, and the English is this study's translation, labelled as such in each case. For each scene the appendix gives its title, setting, scene, lines of talk, evidence, three readings, judgment, counter-condition and proposal, word for word. Some sentences are left out, and the sentences kept stand as the scene has them.
The first coil
A GAEKJU synthetic scene, a longer scene with three lines of talk, set in the 1990s, in this study's English.
Why is the first coil discarded after a roll change not called a mistake?
Roll-change crew at a steelworks' hot-rolling mill
In a hot-rolling mill of the 1990s, once the work rolls are changed, the first coil is booked in advance as a downgraded product, for the time until the thickness settles. That loss is written in the production plan, so no one calls it a misstep. The same week, when the roll gap slipped a little and one more coil was downgraded, a third-year member of the roll-change crew wrote an incident report. The roll-change foreman went into the shift meeting holding the two coils' thickness sheets side by side.
The first coil is the plan and my coil is an accident. The steel went in just the same.
Both are the price of the rolls settling. Yours just had no line in the plan. Let's take the two sheets and ask for a spare slot up to the second coil at the next change.
You caught the gap slipping as soon as the first coil was done. If you'd missed it and let three go through, that would have been a real accident.
Whether the two downgrades were actually the same size was not observed. Here the same downgrade becomes a 'change loss' if it is in the plan and an 'adjustment error' if it is not.
The second coil is not the same thing as the first. A planned loss is the price of the rolls settling, but setting the gap wrong is an error of the hand, so putting the two in one slot blurs what can be learned from the error.
That the third-year caught the gap at the end of the first coil and kept it from spreading to a third coil is the day's output. The incident report records only the one coil lost, not the two coils kept.
The name of a loss is set by the plan, not by its size. The first coil is not a mistake because it was written down in advance, and the third-year's coil is a mistake because that line was missing.
The hot-rolling mill gives the same downgrade two names. What divided the names was not the size of the loss but whether the plan had that line. If there are days when the adjustment after a change is not all over by the first coil, part of the rolls' unsteadiness is passing into a person's incident report. (as of 2026-09-25)
If the downgrade rate of the second coil after a roll change is no different from that of coils on days without a change, the third-year's coil is indeed an error of the hand and not the unsteadiness of the change. This is checked by gathering the thickness records and the downgrade records coil by coil just after changes.
Leave one more slot free in the production plan for the adjustment margin up to the second coil after a roll change. The unsteadiness of the change is then borne in a slot of the plan before it passes into an incident report.
The empty doorstep
A GAEKJU synthetic scene, a longer scene with three lines of talk, set in the 2020s, in this study's English.
Whose schedule pays the price of the third wasted trip to an empty home?
Office of city-gas safety inspectors
Routine city-gas inspections are scheduled on weekdays between 9 and 6 in the daytime. In apartment complexes with many dual-earner households, a door may stay shut even after three visits to the same home. An inspector in the eighth year sticks on a third absence sticker and, going down the stairs, counts the floors visited in vain today. The office gathers the uninspected households by inspector and sends them up at the end of the month.
That's the third sticker. It's not that I got it wrong; my hours and this household's hours don't match.
Senior, I wrote a Saturday-morning number on the sticker, and two households got in touch.
That's good. If they call, let's do the round together on Saturday. And on the month-end sheet we'll each put down half.
How heavily that month-end sheet comes back on the inspector is an assumption this scene sets up.
Some of the doors shut on weekday daytimes are the choice of residents who put off the inspection. If all three wasted trips are put down to the schedule, the office takes on even the share it should split with the residents.
The three wasted trips are the price of the office timetable and the residents' day not matching. As that mismatch is written on the sheet of uninspected households by inspector, the price the schedule should pay moved onto one person's legs.
The Saturday number the new inspector wrote on the sticker is this scene's discovery. The wasted trip was information telling when residents are at home, and the office form simply had no box to take it down.
In the third wasted trip no one got anything wrong. Yet the price is paid only by the inspector's legs and the month-end sheet. That two homes opened through the Saturday number means the wasted trip was a signal telling of mismatched hours. (as of 2026-09-25)
If uninspected households in the same complex do not fall even after Saturday-morning visits are opened, the wasted trips are the price of residents avoiding the inspection, not of mismatched hours. This is checked by dividing each complex's record of uninspected households into before and after the change of visiting hours.
Move part of the inspections in complexes with many dual-earner households to Saturday mornings and weekday evenings, and gather the record of uninspected households under the visiting hours rather than under inspectors' names, returning the price of wasted trips to the side of the schedule.
The room desk
A GAEKJU synthetic scene, a short scene with two lines of talk, set in the 2020s, in this study's English.
What blank does the front desk's 'it worked yesterday' point to?
The room-assignment counter at a hotel front desk
The front-desk lead says a room change that was possible the day before is blocked today. A colleague points out that 'cleaning done' and 'available for sale' are different marks. Before telling the guest a time, the two bring the room's status into line again.
It worked yesterday and today it has stopped. Where did it change?
'Cleaning done' and 'available for sale' show up separately. Let's check them together.
No actual room error or harm to a guest was assumed.
The rule for room changes may itself have changed.
The difference between yesterday and today is a confusion on the part of the person in charge.
The problem is that the cleaning mark and the sale mark moved separately.
The lead's words are a clue for finding where things are blocked. Only by checking whether the cause is that 'cleaning done' and 'available for sale' were merged into one status does it not end as a person's fault. (as of 2026-09-25)
If room changes keep failing even though the two marks always change together, the conditions for changes outside the cleaning and sale marks must be compared.
Let the front desk see together the times at which the cleaning and sale statuses change. Put the time for informing guests in a schedule shared by the front desk and the rooms clerk.
The two dates
A GAEKJU synthetic scene, a short scene with two lines of talk, set in a year supposed for 2030, in this study's English.
Which appointment does a confusion behind automatic entry set awry?
A clinical-trial scheduling room in 2030
An automated judgment copies visit schedules from the documents. The trial coordinator sees that the dates the participant said were possible differ from the dates the hospital can book. The automatic entry was accurate, but reading the two appointments as the same thing is the problem.
The dates written down are right. But these are days that are possible, not days that are booked.
I'll look at the two dates separately. I'll also set aside time to ask the participant again.
In a clinical trial the participant's available time and the institution's booking are separate. It is a supposition about the future and not the record of an actual trial.
The responsibility for rechecking the participant's time cannot be left with the coordinator alone.
If the entry is accurate, there is no scheduling problem either.
The interpretation that bound two different appointments into one went awry.
Even if the letters were copied well, merging two dates of different meaning throws the schedule off. The coordinator's discovery makes it possible to tell apart not how well the automatic entry performs but whose appointment each is. (as of 2026-09-25)
If the participant's possible dates and the institution's booked dates are always the same and no further adjustment is needed, the judgment that the meanings were confused weakens.
The automated judgment sets the two dates as separate appointments and picks a possible overlap. The time for rechecking is divided between the coordinator and the booking clerk.
The makeshift rest
A GAEKJU synthetic scene, a short scene with two lines of talk, set in the 2020s, in this study's English.
Whose invention does a makeshift rest remain?
A surface-mount line for electronic components
On a surface-mount line, the jig that holds parts does not fit the size of a new part. A technician carries on the work with a makeshift rest. The line lead acknowledges the skill and suggests letting the designers see the difference in dimensions.
The rest didn't fit, so I used something else for a moment. The dimensions seem to have changed.
Well saved. Let's check the quality too and show the design side the difference.
In surface mounting, the fit between a part's dimensions and its rest bears on quality. This invention and its quality result are not actual observations.
Before it is verified, it cannot be settled as either a risk or a discovery.
The makeshift rest is a choice outside the specification, and so it is risky.
The makeshift rest reveals a difference in size the design missed.
The technician's skill kept the day's work going and can show a difference the design missed. Whether it can be used in the product needs a separate check, so a conclusion praising or condemning the invention is put off. (as of 2026-09-25)
If quality problems recur in parts made with the rest and the dimensional difference has nothing to do with the design, the reading that it is a sign of invention must be set aside.
Allocate the quality check of the makeshift work and the review of the jig's dimensions to different times. Let the designers too see the difference the technician found.
The fire lookout
A GAEKJU synthetic scene, a longer scene with three lines of talk, set in the 2010s, in this study's English.
When does a lookout who reported a false smoke and was chided report the next smoke?
A forest-fire lookout post
In the dry spring season, the lookout on duty saw smoke beyond the valley and radioed it in. When the firefighting crew went up, it was smoke from burning the ridges of a paddy, and a single 'Again?' came back over the radio. The next day, when similar smoke rose, the lookout reported it only after watching for ten more minutes through binoculars. That evening the veteran at the next post radioed first.
Yesterday's was paddy ridges, wasn't it. Today I'm going to call when I'm sure.
By the time you're sure, it's already a big fire. Even if you call ten false smokes, if you call the real one first, that's our job. I called in plastic being burned twice last week myself.
While we were up there yesterday, I told the owner of that paddy how to report a burn. It wasn't a wasted trip.
What this scene does not answer is how much the ten minutes' delay held back the discovery of an actual fire.
The number of false reports is another name for how keenly the lookout watches. If a single 'Again?' raised the threshold for reporting by ten minutes, words meant to cut missteps moved the price onto seeing a real fire late.
Watching ten minutes more was judgment, not fear. If, in a season of frequent burning, the call is made after seeing the colour of the smoke and the way it spreads, the firefighting crew's wasted trips fall and people are left for a real fire.
As the crew leader said, the wasted trip was not only a wasted trip. As with telling the paddy's owner about reporting burns while up there, a false report can also open a way to know of the next burning in advance.
The lookout's false report is closer to the price of keenness than to being wrong. If a word of reproach raises the threshold for reporting, that price moves to the most expensive place, the first ten minutes of a real fire. What guarded the eyes of this post was the veteran's talk, which brought up the veteran's own false reports first and set the threshold back. (as of 2026-09-25)
If, under the same smoke conditions, the time of the first radio call shows no difference between days after a reproach and other days, the reading that the threshold rose is wrong. This is checked by setting the times in the lookout's radio log against the firefighting crew's call-out records.
The imbalance at this post is that the information for telling false smoke apart reaches the lookout late. Hand the village's list of reported ridge and rubbish burnings to the lookout first, and share the price of wasted trips with the side that takes the burning reports.
The pencilled drawing
A GAEKJU synthetic scene, a short scene with two lines of talk, set in the 1980s, in this study's English.
Who remembers at the next build a dimension corrected in pencil?
The drawing table of a generator-building crew in the 1980s
The crew foreman corrects a dimension on the drawing in pencil and fits the part. A colleague asks whether the same line will remain on the drawing for the next build. The foreman realises that the reworked part and the correction of the drawing are different jobs.
Today's part is fitted. The next build will need to know this line too.
I'll go with you to the person who will change the drawing. We can't leave the pencil to work alone.
Pencil corrections on paper drawings are a particular condition of the medium on the building floor. It is not an actual record of a past factory.
The part fitted, so the correction is finished too.
A pencil mark can hardly become the standard for the next build.
It must be checked who can change the official drawing.
The foreman's hand saved today's part. To save it in the next build too, the dimension has to reach not one person's memory but the place that holds the right to make corrections. (as of 2026-09-25)
If pencil corrections are always carried into the official drawing the same day and there is no rework in the next build, the memory of the correction was not broken.
Link the time of corrections on the floor with the time of reviewing the official drawing. Do not leave the foreman's pencil mark as the sole memory for the next build.
The buried notice
A GAEKJU synthetic scene, a longer scene with three lines of talk, set in the 2000s and the 2020s, in this study's English.
What did the error of mishearing on the radio become in the group chat?
The site office of a new factory construction site
On a new factory construction site in the 2000s, instructions from the site office went out by radio. A foreman who misheard 'column 3' as 'number 7' asked back 'Which number?' on the spot, and the error was cleared up in a few seconds. On the same company's sites in the 2020s, instructions go to a group chat holding forty subcontractors' foremen. A change in the pouring order posted in the morning was buried by lunchtime under twenty photos and thirty 'Confirmed' messages, and one foreman closed the formwork in the old order.
I didn't see that notice. On my phone there were just twenty photos above it.
In the radio days, if you didn't hear, you at least asked back. In the group chat you don't even know you missed it, so there's no chance to ask back. For instructions that change the order, I'll phone the foremen one by one once more.
Forty calls on your own is too much, manager. I'll make half. For anyone who doesn't pick up, I'll walk over and tell them.
Whether errors actually increased on this site was not observed. An error in the radio days showed at the moment it was heard, and an error in the group-chat days shows after the formwork is closed, so what changed may be the time an error comes to light rather than the amount of error.
When the channel changed, the shape of confusion changed. A radio error was a mishearing and was cleared up by asking back on the spot; a group-chat error is a not-seeing and gives no occasion even to ask back.
The group chat did not increase errors; it left a record. Among the mishearings of the radio days there must have been many that passed without anyone asking back, and now at least who posted what and when remains.
The manager's offer to phone each one brings the vanished asking-back in again by voice. If that labour is loaded onto the phones of one or two people in the site office, a few people end up paying for forty foremen's asking-back.
The radio's error and the group chat's error are not the same 'error'. The first was misheard and asked back on the spot; the second was not seen, so no one knew to ask back. As the channel changed, the time a confusion comes to light was pushed back from the moment of hearing to after the formwork is closed. (as of 2026-09-25)
If work in the old order turns up as often even after instructions changing the order are given once more by phone, the reason lies not in burying but in the instruction itself being late. This is checked by setting the site's instruction records against its rework records channel by channel.
Split off instructions that change the order alone so that, apart from the group chat, they reach each foreman once more by voice, and let two people in the site office take turns with that share of calls.
The swapped container
A GAEKJU synthetic scene, a short scene with two lines of talk, set in the 2020s, in this study's English.
Does one swapped container shorten the wait and blur the trail?
The dispensing room of a hospital pharmacy
The deputy head of the pharmacy uses another storage container for the time being to get around a supply delay.
The deputy head looks for a place to keep both pieces of information together.
We have to cut the wait.
As long as the link is clear, the next shift won't get lost. I'll check it.
In a pharmacy, storage and tracing bear on safety. It presents no actual dispensing or patient outcome.
If the link for tracing is missing, the cost of checking later grows.
A different allocation is needed to support waiting and tracing at once.
Temporary storage cuts the wait, so it is enough.
The gain of the stopgap is acknowledged, but the person who traces must not be made to link the information again alone later. The deputy head's decision can be assessed when the two times are allocated together. (as of 2026-09-25)
Divide dealing with the wait and checking the link among different times and people. Do not leave the aftermath of a hurried choice to one pharmacist.
The missed handover
A GAEKJU synthetic scene, a short scene with two lines of talk, set in the 2020s, in this study's English.
What does a manager's admission of an omission leave for the next shift?
The store manager says first that a message for a customer with a booking was missed. The staff member on the next shift joins in asking whether the customer will be kept waiting again. The manager writes down clearly what was missed and secures a turn to explain to the customer.
I missed passing it on. I'll write it up now.
Before we ask the customer, let's first line up what we know.
In a store, the promise to the customer and the explanation at the shift change remain together. No actual inconvenience to customers or sales outcome was set.
As long as the admission comes quickly, the promise to the customer is restored.
Only if time for explaining is allotted after the admission does the shift carry on.
The burden of asking the customer again is hard to avoid.
The manager admitting it first let the next shift know the situation. But the time needed to keep the promise to the customer has to be set aside separately. (as of 2026-09-25)
If, even when the omission is admitted at once, the next shift does not know the same information and the customer's waiting recurs, the linking effect of the admission is small.
Schedule the time for explaining to the customer side by side with the shift handover. Do not pile all the later explaining on the manager who found the omission.