The first wall mark changed before Melody finished drawing the fifth column.
Loria saw it.
“Top stroke rotated left.”
Melody wrote under SIGNAL.
Top stroke → left.
Bram was already looking down.
“The floor is doing two things.”
A pale roughness moved in slow bands near the center of the section while a darker granular line pulsed along the right edge. The pale band advanced steadily. The dark line advanced in short bursts, then stopped.
Melody split HAZARD into two rows.
Pale center: steady advance.
Dark right: pulse / stop.
Joss stood nearest the left wall, watching two narrow seams set into the stone at different heights.
“Lower one is moving.”
The lower seam had begun to widen by fractions. Nothing protruded yet, but an internal runner clicked each time the gap increased.
“Upper?” Melody asked.
“Static.”
She wrote:
MECHANISM.
Lower seam active, opening.
Upper seam static.
David had not taken his eyes off the space beyond the two narrow exits ahead.
“Both routes are pulling, but not consistently. The stronger one keeps shifting.”
Melody wrote under ROUTE.
Two pulls. Relative strength unstable.
Lucan checked the floor, walls, exits, and the space between all six of them.
“If everyone tests their own thing, we lose every baseline at once.”
Joss glanced toward the opening seam while Bram kept his eyes on the moving floor and Loria held on the wall mark. No one advanced.
Lucan pointed to the open ground behind Melody. “Hold positions. No one advances. Reports first. One movement test at a time. I am controlling movement, not deciding what the evidence means.”
Joss gave one short nod.
Melody wrote under FORMATION.
Positions fixed by Lucan for comparability.
The wall mark rotated back toward vertical.
Loria tracked it. “Signal reset without movement.”
The pale floor band advanced another inch.
Bram pointed. “Center hazard still independent of us so far.”
The dark line pulsed twice.
“Right edge changed faster.”
Joss kept his hands away from the seam. “Lower mechanism is still opening.”
David shifted only his gaze between the exits.
“Left feels stronger now.”
Melody looked down at the columns.
Five readable states.
For the first time since the conflict at the previous split, none of them had changed the others just by speaking.
“Baseline first,” she said.
Lucan held everyone where they were.
Melody counted three cycles of the dark floor pulse.
The pale band moved at the same slow rate each time.
The lower seam widened on every second dark pulse.
The wall mark changed twice while Loria’s attention stayed fixed on it.
David’s route preference shifted once from left to right without any physical commitment.
Melody wrote each channel separately.
SIGNAL: irregular rotation under fixed positions.
HAZARD: pale steady; dark pulsed.
MECHANISM: lower seam opens every second dark pulse.
ROUTE: preference shifts without movement.
FORMATION: fixed.
She stopped before drawing connections.
Loria leaned slightly toward the mark.
It brightened.
“Attention still matters.”
Melody added:
Signal remains attention-sensitive.
Bram looked at the two floor patterns. “And the pale band is almost at our front line.”
Lucan measured the available retreat.
“We can widen backward without changing the front positions.”
Melody nodded. “Do it.”
Lucan moved Bram and Joss half a pace apart, then shifted himself farther rearward. Melody, Loria, and David held.
The pale band continued forward unchanged.
The dark pulse on the right weakened.
Bram saw it immediately. “Right hazard responds to local spacing. Center one does not, or not at that scale.”
Melody wrote:
Formation test 1.
Right hazard intensity - after rear widening.
Center hazard unchanged.
The lower seam clicked.
Joss lifted a finger without touching it. “Mechanism advanced during the spacing test.”
Melody marked the overlap.
One intervention had already changed more than one timestamp even though Lucan’s movement had not caused the mechanism change.
She drew a bracket between FORMATION and MECHANISM.
Concurrent change. Cause unassigned.
“Next movement?” Lucan asked.
Melody looked at the channels.
David’s route pull had swung right again.
Loria’s mark had split into two short bars.
Bram’s pale band was three feet away.
Joss’s lower seam was now open wide enough to show a dark inner track.
Each problem looked locally solvable.
Together they looked like a trap built out of timing.
Joss could stabilize the lower seam.
Bram could tell them where the pale floor would reach first.
Loria could test which of the two bars responded to directed attention.
David could commit far enough to force one route to answer.
Lucan could stage movement through either exit.
Any one of those actions might change the conditions under which the others were being read.
Melody wrote:
Locally solvable ≠ independently solvable.
Then she looked up.
“Baseline is complete. We keep one action at a time, and the first one belongs to whichever state has the shortest safe wait. Joss, give me the threshold before you act.”
He crouched closer.
The dark floor line pulsed once.
The lower seam opened another fraction.
Joss studied the inner track. “If it reaches the next notch, it will close the left line or lock whatever is behind it. I cannot tell which yet. We have one pulse, maybe two.”
Melody wrote:
Mechanism wait limit: 1–2 dark pulses.
Bram pointed to the pale band. “Center hazard reaches our front line in about four steady advances.”
Loria kept watching the two bars. “Signal test can wait.”
David looked toward the exits. “Route test can wait if neither closes.”
Lucan checked the geometry again. “Spacing holds through two pulses.”
Melody wrote the wait limits beside each channel.
Signal: can wait.
Hazard: ~4 advances.
Mechanism: 1–2 pulses.
Route: can wait conditionally.
Formation: stable through 2 pulses.
The page finally gave them an order without claiming a hierarchy of truth.
“Joss first,” she said.
He did not move yet.
Lucan raised his hand. “Everyone else freezes their channel. No route commitment. No attention test. No spacing change unless safety requires it.”
Melody added:
Controlled intervention 1: mechanism.
Pre-state captured.
The wall mark held.
The pale band advanced.
The dark line pulsed.
The seam clicked.
Joss inhaled once.
“Next notch is the threshold.”
Melody looked at the five columns.
For one brief moment, the section was readable without being solved.
Lucan’s order had done exactly what it was supposed to do: it had kept the problems from destroying one another long enough for Melody to compare them.
She wrote beneath the columns:
Temporary order preserves comparability.
Does not determine which method governs.
The lower seam moved again.
Joss’s hand came up.
“Cannot wait another pulse.”
“Cannot wait another pulse.”
Melody put the pencil against the page and stopped it there.
“Pre-state holds. Joss only.”
Lucan lifted one hand without moving his feet. Loria kept her eyes on the split wall bars. Bram watched both floor patterns. David stayed neutral between the two route pulls.
Joss crouched beside the lower seam.
The dark floor line pulsed.
The seam opened to the notch.
Joss moved.
He set two fingers against the exposed inner track and pressed inward, not against the opening seam itself but against a narrow metal tongue that had begun withdrawing from a slot in the stone.
The tongue resisted.
Then it clicked back into place.
The lower seam stopped opening.
A section of wall beyond it shifted inward by an inch, exposing a narrow line of passage that had been hidden behind the mechanism housing.
“Tongue seated,” Joss said. “Runner stopped.”
Melody wrote under MECHANISM:
J input: tongue returned to slot.
Lower seam stable.
Hidden line open.
Local prediction confirmed.
The upper wall mark went dark.
Loria inhaled sharply.
“The top signal is gone.”
At the same moment, the two short bars near her reorganized. One lengthened downward. The other broke into three points.
Melody moved to SIGNAL.
Upper mark off.
Two-bar pattern → long stroke + three points.
Bram’s dark right-edge pulse reversed direction.
Instead of pressing toward the group, it rolled away from them, crossed the open ground, and gathered beside a patch of floor that had been quiet through the entire baseline.
“Hazard moved,” Bram said. “New accumulation left-center.”
The pale center band kept advancing at its old rate.
Melody wrote under HAZARD:
Pale center unchanged.
Dark pulse reverses → new left-center accumulation.
David’s head turned toward the newly exposed line behind Joss, then away from it. “Right is stronger now. The opening Joss made is present, but it does not feel like the route.”
She added:
ROUTE: right strengthens after J intervention.
Exposed mechanism line not favored by D.
Lucan did not move anyone.
“Formation unchanged,” he said.
Melody wrote that too.
Five columns.
One intervention.
Four changed states.
Only one of them had been the target.
Joss took his fingers off the tongue.
It stayed seated.
The seam did not resume opening.
“My part is stable.”
Melody checked the baseline page.
Before Joss acted, the lower seam had been opening every second dark pulse. Loria’s upper mark had remained active. Bram’s dark right-edge hazard had weakened when rear spacing widened but had stayed on the right. David’s route pull had shifted between left and right without commitment.
After Joss acted, the seam stabilized exactly as he predicted.
Everything else had become a new problem.
She wrote across the bottom of the page:
Correct local intervention changed other live reads.
Then she stopped.
The sentence needed a boundary.
She added:
Does not invalidate local success.
Loria pointed at the reorganized pattern. The upper signal had disappeared completely rather than dimming, and the lower bars had changed form instead of returning to baseline.
Melody copied the distinction exactly.
Joss looked toward Loria’s wall. “The tongue seated inside the same housing. The signal change may be linked.”
“May be,” Melody said.
He nodded.
No one tried to turn may into therefore.
Bram shifted his gaze between the old right edge and the new left-center accumulation.
“The hazard is not just reversing. It is collecting somewhere that was quiet before.”
Melody compared his current floor map with the baseline.
The model she had been starting to form in Beat 1 had depended on two separations: the pale band advanced steadily across the center, while the dark pulse responded to local spacing at the right edge.
That second relation no longer held.
The group had not changed spacing.
Joss had changed the mechanism.
The dark pulse had moved anyway.
She drew a line through:
Dark right intensity tracks local rear spacing.
Not erased. Bounded.
Beside it she wrote:
Valid before J intervention under prior mechanism state.
Current relation unknown.
Lucan saw the revision and understood the distinction: the old result remained part of the record even though it no longer predicted the current floor state.
The pale band advanced another inch.
Bram marked the distance with two fingers. “Three advances to our front line now.”
Melody wrote the new hazard timing.
Loria kept watching the long stroke and three points. One of the points brightened as her attention shifted toward the newly exposed line behind Joss.
She looked back at the main exits.
The point dimmed.
“Fresh signal state,” she said. “Attention-sensitive again, but not in the old pattern.”
Melody added:
Post-intervention signal remains attention-sensitive.
Pattern mapping reset.
Joss stood slowly, keeping his hands clear of the wall.
“The mechanism is solved for now.”
Lucan checked him. “Solved means?”
“The tongue is seated. The runner is stopped. The lower seam is no longer opening. I do not know whether another part of the housing can move.”
Melody wrote the scope under his column.
MECHANISM LOCAL STATE STABLE.
Global consequences unknown.
That distinction felt cleaner than the old instinct to ask Joss what his action had done to every other system.
He had not claimed to know.
He had claimed he could stop the tongue from withdrawing.
He had done that.
The rest belonged to the new state.
“New baseline,” Melody said.
Lucan held the formation.
Nobody tested anything.
Melody called the columns in order.
Loria gave the new signal geometry and the attention response.
Bram gave the pale band’s steady timing and the dark accumulation’s new position.
Joss gave the stable tongue and stopped runner.
David gave the stronger right route pull.
Lucan gave unchanged spacing and two viable movement lines, though the left-center hazard had reduced one of them.
Melody wrote all five without trying to reconnect them.
SIGNAL: long stroke + three points. Attention-sensitive. Old mapping invalid.
HAZARD: pale steady. Dark accumulation left-center. Old right-spacing relation not currently predictive.
MECHANISM: tongue seated. Runner stopped. Lower seam stable.
ROUTE: right stronger. No commitment yet.
FORMATION: unchanged. Left-center floor loss reduces movement options.
She looked at the page.
The old baseline and the new one sat side by side.
Joss’s action formed the boundary between them.
Not an error boundary.
A state boundary.
She wrote:
Sufficient evidence for local action ≠ complete explanation of system consequences.
The wording made Joss glance over.
He tapped the mechanism column with one finger. “The tongue result holds. Everything else changed around it.”
Melody nodded and kept the two claims separate on the page.
Loria raised two fingers.
“One of the points repeats.”
She directed her attention toward the left edge of the newly exposed mechanism line.
The three points lit short-long-short.
She looked away.
They faded.
She repeated the attention shift.
Short-long-short.
Melody wrote the sequence.
Loria did not call it a route.
“The pattern is tied to investigation of that left edge. I do not know what it means yet.”
“Actionable claim?” Melody asked.
“Keep looking there long enough to see whether the sequence develops.”
Melody put the claim under SIGNAL.
Before she could answer, Bram’s head turned sharply toward the floor beyond that same side.
“Do not step there.”
Nothing visible had changed.
The stone looked smooth.
Melody followed his gaze.
Bram pointed to the edge of the quiet floor. “The dark accumulation stopped spreading because its front collapsed inward. The next movement will be there.”
Melody could not see it.
No roughness.
No color change.
No rise.
Only Bram’s read of a pattern that had just changed once already after Joss’s intervention.
“How sure?”
“Enough that I would not put a foot on it.”
Loria’s short-long-short pattern repeated again.
The pale center band advanced.
David looked right.
Joss listened to the stabilized wall.
Lucan held everyone where they were.
Melody looked from Loria’s signal claim to Bram’s invisible hazard boundary.
Neither could be reproduced from her page before the next state change.
Both were scoped.
Both were asking for action.
She drew two boxes.
L: deliberate attention left edge → repeat short-long-short. Claim: continue investigation.
B: dark accumulation front collapsed inward. Claim: keep bodies off visually clear left zone.
Then she looked up.
“Narrowest claims only. Say them again before we move.”
Loria kept her attention on the left edge of the newly exposed line.
“Same claim,” she said. “Short-long-short repeats when we investigate that edge. I am not saying it is a route. I am saying that is the next action that changes the signal in a structured way.”
Bram pointed to the visually clear floor beyond it. “And nobody puts a foot there. The dark front folded inward. That patch is where I expect it to show next.”
Melody wrote both claims again, shorter this time.
L: investigate left edge.
B: keep bodies off left clear zone.
The combination looked contradictory until she stopped asking it to answer one question.
Attention could go left without bodies going there.
She added a third line.
J mechanism stable.
The tongue remained seated. The lower seam had not reopened. The hidden line Joss had exposed still existed, but David had already said it did not feel like the route.
Melody looked at the two main exits.
“We keep the left feature under investigation. Nobody enters Bram’s zone. No route commitment until the signal develops.”
David kept his eyes on the right opening. “Right is already stronger without me moving.”
She wrote:
D: right route pull stronger under current state.
Her current plan did not say left was the route.
It said left was the next justified investigation.
David’s read said the maze already wanted a route decision somewhere else.
Both could be true.
They could also interfere.
Lucan checked the space between the left hazard boundary and the right opening, then established a test lane for David that preserved Bram’s exclusion zone, Loria’s attention state, and Joss’s untouched mechanism.
Melody drew a box around the current state.
“David. Declare the test.”
He did not move yet.
“One step right. No second step unless called. I am testing route response, not safety.”
Melody wrote the scope exactly.
Lucan moved half a pace so David had a clear lane and everyone else retained their position.
“Go.”
David stepped right.
The response was immediate.
A narrow white line appeared at ankle height along the right wall, beginning beside David’s forward foot and extending six feet into the opening.
Loria’s short-long-short sequence remained lit on the left.
Bram’s visually clear danger zone remained visually clear.
Joss’s mechanism did not move.
David stopped after one step.
“Route response,” he said.
Melody wrote:
D commitment 1 → right line appears.
Then she checked the other channels in one pass. Loria’s sequence was unchanged. Bram still saw no visible onset in the warned zone. Joss’s mechanism remained stable, and Lucan reported no spacing loss.
The page now held a clean contradiction.
Left investigation remained informative.
Left movement remained constrained.
Right route commitment had produced a new directional line.
Melody stared at the right mark until she felt the old reflex forming.
Exception.
Noise.
Contamination from Joss’s earlier intervention.
Any of those could protect the sequence she already understood.
None of them was evidence yet.
“Reset David.”
Lucan guided him back to the exact prior foot position.
The right line dimmed, then vanished.
Loria’s three points stayed in the same short-long-short rhythm.
Bram did not move.
Joss did not touch the wall.
Melody waited one dark-floor pulse and one pale-band advance.
“Same test.”
David declared it again. “One step right. Route response only.”
He moved.
The white line appeared again.
Same wall.
Same height.
Nearly the same length.
This time a second short stroke appeared farther ahead after one count.
Melody wrote:
D commitment 2 → right line repeats + second stroke.
Loria called her own state without being asked. “Left sequence still repeating.”
Bram pointed to the clear patch. “Still no visible onset. Still keep bodies off it.”
Joss said, “Mechanism stable.”
Lucan held the formation.
David did not look at Melody.
He looked at the second stroke.
“The route is answering.”
Melody wrote three possible explanations beneath the comparison.
1. Exception.
2. Repeating noise.
3. Missing variable detected by route commitment.
She tested the first against the page.
One response could be an exception.
Two comparable responses under controlled movement made that less useful as an explanation.
She tested the second.
Noise could repeat.
But the response had appeared with the same declared commitment, vanished on reset, and returned when the commitment repeated.
Calling it noise now required ignoring the structure that made Loria’s attention signal worth tracking.
She drew a line through neither option.
Instead she wrote:
Exception possible, weak.
Noise possible, weakening.
Missing variable possible, unproven.
David finally looked at her. “You are preserving the left investigation without calling it the route.”
Melody nodded. That was exactly the distinction the page now required.
The left side held information about what the maze did when they investigated.
The right side held information about what the maze did when David committed.
Bram’s left warning held information about where bodies should not go.
Joss’s mechanism state held information about what not to disturb.
Lucan’s formation held the conditions that let all of those observations coexist.
Nothing required one channel to erase another.
Melody wrote:
Contradiction valid.
No winner assigned.
The pale center band advanced.
Bram raised two fingers. “Two advances before it reaches the old front line.”
Lucan looked at Melody. “We cannot run this test forever.”
She nodded.
The right line was still visible because David remained one step committed.
Loria’s three points continued short-long-short on the left.
Melody set the mixed-state conditions in one instruction: David holds the one-step commitment without anyone following him; Loria keeps attention on the left edge; Bram keeps the left floor exclusion; Joss watches the mechanism; Lucan preserves reversibility.
Lucan repositioned himself half a pace rearward and moved Melody a fraction toward center so neither side had to compress.
The dark accumulation remained hidden beneath smooth-looking stone.
The right route line stayed bright.
The left points repeated.
For three counts, nothing forced the channels to agree.
Melody wrote:
Mixed state preserved.
Left investigation remains informative.
Left movement zone constrained.
Right route response repeats.
Missing variable possible.
The wording was ugly.
It was also more accurate than a clean route arrow.
Loria leaned her head slightly without moving her feet.
“The sequence changed.”
Melody looked left.
Short-long-short had become short-short-long.
“What changed?”
“Attention did not.”
Lucan looked at the group. “Spacing changed slightly when I moved Melody.”
Bram’s gaze dropped to the floor and he pointed at the visually clear zone he had warned them away from. “There. The front is showing now.”
A dark crescent appeared under the clear stone exactly where he had pointed earlier.
It rose in a thin line, then bent away from Melody’s new central position.
Melody wrote both events before interpreting either.
L signal sequence changes after formation adjustment.
B predicted hazard becomes visible in warned zone; boundary curves after spacing shift.
The right route line stayed lit.
David had not moved again.
Joss’s mechanism remained stable.
Melody compared Loria’s unchanged investigative target with the altered symbol sequence. The sequence had changed only after the group’s spacing changed.
Loria studied the points. “Then the signal may be tracking more than attention.”
Bram watched the crescent. “And the floor is definitely tracking spacing.”
Melody separated the notes.
Loria’s channel had changed without a change in investigative target.
Bram’s channel had changed with formation.
That did not make them the same.
It gave them a reason to test the distinction.
She drew two new headings beneath the mixed-state record.
INVESTIGATIVE STATE.
CONFIGURATION STATE.
Lucan saw them. “We are not moving again until you know which test comes first.”
Melody looked at the right route line, the altered left symbols, and Bram’s now-visible crescent.
David’s contradiction had not broken the model.
It had forced the model to admit that it might be watching the wrong dimension.
She underlined the three lines that mattered.
Left investigation remains informative.
Right route response repeats.
Missing variable possible.
Then she added:
Next: separate investigation change from spacing change.
Loria was already watching the symbols.
Bram was already watching the floor boundary.
Lucan held everyone still.
Melody looked from one channel to the other.
“Good. We test those separately.”
“Good. We test those separately.”
Lucan did not move.
David kept the one-step right commitment. The white route line stayed lit along the right wall. Joss’s lower mechanism remained stable. Bram watched the dark crescent that had finally appeared where he had predicted it. Loria kept her eyes on the altered left-side points.
Melody drew a line beneath the mixed-state record.
TEST A — INVESTIGATION CHANGE / FORMATION FIXED.
Melody fixed the target as the left edge of the exposed mechanism line and told Loria to change only what they investigated while keeping her body exactly where it was.
Loria nodded.
Melody looked at Lucan. “Freeze spacing.”
He checked the gaps between all six and gave one short confirmation.
Melody pointed first to the exposed left edge.
“Everyone who can see it, focus there.”
David did not shift his feet. Bram kept his floor sightline. Joss looked without touching the wall. Lucan held the formation.
The three points brightened.
Short.
Short.
Long.
The long point extended downward by a fraction.
Loria’s voice stayed level. “Same sequence. Stronger final element.”
Melody wrote:
Attention fixed on left edge → S-S-L, long intensifies.
Then she redirected their attention to the right route line without letting anyone move.
“Now the right wall.”
David’s route mark was already bright from his commitment.
The left points dimmed.
The long element shortened first.
Then the second short point faded.
Only the first remained.
Loria tracked it. “Left signal weakens when investigation shifts away.”
Melody wrote:
Formation unchanged.
Investigative target left → right.
Left signal response decreases.
David looked at the right line. “My route mark did not change.”
Melody added that too.
Right route line stable under attention shift.
“Back to the left edge.”
Their attention returned.
The three points rebuilt in the same short-short-long sequence.
Not perfectly. The second point took one count longer to appear.
Loria noticed the delay but confirmed the order still held.
Melody recorded the delay instead of smoothing it out.
TEST A result:
Investigative focus changes left symbol state with formation fixed.
Response repeatable, imperfect.
She put the pencil down for one breath.
The distinction was cleaner now.
The test established sensitivity without claiming meaning.
Bram was still watching the dark crescent.
“It moved half an inch while we did that.”
Melody looked down.
The crescent had widened toward the center, but not toward anyone’s feet.
Lucan confirmed the formation had held throughout the attention test.
Melody wrote under CONFIGURATION STATE:
Hazard drift under fixed spacing. Cause unassigned.
Then she turned the page section.
TEST B — FORMATION CHANGE / INVESTIGATION FIXED.
“Loria keeps the left edge as the investigative target. David keeps the route commitment exactly where it is. Joss stays off the mechanism. Bram calls floor only. Lucan changes spacing.”
Lucan looked over the available ground.
The pale center band had advanced again. The dark crescent occupied part of the visually clear left zone. The right route line still marked David’s one-step test state.
“I can widen the rear without changing the front.”
“Use that.”
Lucan moved Bram half a pace left and himself half a pace back. Melody stayed centered. Loria and David did not move. Joss remained beside the stabilized seam.
The left symbol sequence flickered.
Loria spoke immediately. “Still short-short-long, but dimmer.”
Bram pointed at the crescent.
“It bent.”
The dark floor boundary pulled away from Bram’s new position and curved toward the empty space he had left.
Melody wrote both events.
Formation widened rear.
Hazard boundary curves toward vacated compression point.
Left signal dims without investigative-target change.
Loria confirmed her attention had stayed fixed, David’s route commitment had not changed, and Joss reported the mechanism still stable.
Melody added:
Signal also responsive to configuration change.
That complicated the first test without erasing it.
The symbol reacted to investigation.
It also reacted when spacing changed under the same investigation.
Different input dimensions. Same output channel.
Lucan restored the prior rear spacing.
The dark crescent eased back toward its earlier curve, but only partway. Bram marked the partial recovery.
The left symbols brightened again.
Loria watched the long point regain intensity. “Signal strength came back too.”
Melody wrote:
Spacing restore → hazard partial return.
Spacing restore → signal intensity partial return.
She boxed only the part the evidence supported.
CONFIGURATION affects hazard geometry strongly.
CONFIGURATION also affects signal intensity.
INVESTIGATIVE focus affects signal pattern/strength.
No single label was going to own the wall marks anymore.
Melody extended the same comparison instead of starting a new taxonomy. She took Loria’s prediction that attention changes should alter symbol pattern or strength while spacing changes should affect intensity without necessarily destroying the sequence. Bram’s prediction went beneath it: spacing changes should alter floor geometry, while attention-only changes should produce little or no floor response.
“Test attention first. Formation fixed.”
Everyone held.
Melody redirected attention from the left edge to the upper inactive seam.
The left points dimmed in order.
Long first.
Second short next.
First short last.
The floor crescent did not visibly change during the same interval.
Bram waited one more count. “No floor response.”
Loria pointed at the fading marks. “Signal changed as predicted.”
Melody recorded the result under TEST A: the attention shift altered the signal while no floor change appeared in the same observation window.
“Reset attention left.”
The points rebuilt.
“Now spacing. Same target.”
Lucan narrowed the rear gap by one careful half-step without moving anyone into Bram’s danger zone.
The dark crescent responded before the symbols did.
It widened toward the new compression.
Bram’s hand came up. “Floor response.”
One count later, the left points brightened sharply while keeping the same short-short-long order.
Loria nodded once. “Sequence held. Intensity changed.”
Melody recorded the result under TEST B: the spacing shift strongly altered the floor and also changed signal intensity while preserving the sequence.
Lucan restored the spacing.
The crescent pulled back.
The symbols dimmed toward their prior level.
Nothing in this section returned perfectly.
Melody looked over the comparison.
Loria’s method had found an investigation-sensitive signal layer.
Bram’s method had found a configuration-sensitive environmental layer.
Neither one was a disguised version of the other.
And neither superseded David’s right route response or Joss’s stabilized mechanism state.
She rewrote the column labels.
SIGNAL — investigative-state response + configuration sensitivity.
HAZARD — environmental/configuration response.
MECHANISM — local hardware state.
ROUTE — commitment response.
FORMATION — movement geometry / input condition.
The categories were getting less elegant.
They were also getting harder to lie with.
David glanced at the right line. “My route still says right.”
Melody looked over.
The line had not dimmed during either test.
She added:
Route response persists across attention + spacing tests.
Joss listened to the lower seam. “Runner still stopped.”
She added that too.
Mechanism state persists across tests.
Lucan checked the floor. “We have less room now.”
The pale center band was close enough that another long comparison would cost them the front half of the section.
Bram measured it. “One advance, maybe two.”
Loria’s points brightened again as the group’s attention returned to the page and the left feature together.
Melody looked at five different outputs.
Route right.
Investigation left.
Hazard restricting left-center.
Mechanism stable.
Formation time shrinking.
All of them were useful.
None answered the same question.
She turned to a clean line and wrote:
Different methods are detecting different dimensions of the same live state.
Then beneath it:
Do not collapse outputs before comparison.
The pale band advanced.
Lucan raised his hand. “Comparison is over. We need the next move.”
Melody looked at the five channels.
For the first time, she did not ask which one was best.
She asked where they overlapped.
Melody looked at the five channels.
For the first time, she did not ask which one was best.
She asked where they overlapped.
The answer was less satisfying than a hierarchy.
David’s right-route line still held along the wall.
Loria’s short-short-long sequence stayed active on the left as long as the group kept investigating the exposed edge.
Bram’s dark crescent occupied the left-center floor and bent whenever Lucan changed spacing.
Joss’s runner remained stopped.
Lucan had one advance, maybe two, before the pale center band took away the rest of their comfortable formation space.
Melody drew five rows and tried the old question anyway.
Which one controls?
Mechanism first was tempting because Joss could identify a threshold that physically locked or released something.
But his stabilized runner said almost nothing about whether right was the route.
Route first was tempting because David’s line had repeated under controlled commitment.
But it did not tell them whether the floor ahead would remain usable.
Signal first was tempting because Loria’s symbols changed in structured response to investigation.
But they did not tell the group where bodies could safely stand.
Hazard first was tempting because Bram’s predictions had kept them off floor that later declared itself dangerous.
But the floor did not choose the route.
Formation first was tempting because Lucan could keep all the other channels readable.
But readability was not meaning.
Melody wrote beside each one:
Locally strong.
Globally incomplete.
Five times.
David noticed and gave one approving nod at the absence of a ranking.
Lucan looked at the pale band. “It also is not a move.”
Melody turned the page sideways.
RIGHT ROUTE.
LEFT INVESTIGATION.
LEFT-CENTER HAZARD.
MECHANISM STABLE.
FORMATION WINDOW SHRINKING.
She drew no arrows between them.
“What overlaps?” she asked.
Loria answered first. “We can keep investigating left without putting anyone into Bram’s floor zone.”
Bram nodded toward the right half. “And right is still physically cleaner than left-center.”
David pointed along the white line. “Which matches the route.”
Joss listened to the wall. “The mechanism does not require action right now.”
Lucan looked from David’s lane to the floor boundary. “I can shift the group right without collapsing the left sightline. Not everyone. Enough to keep the test active.”
Melody wrote the overlap.
Move formation rightward.
Preserve left investigation.
Avoid left-center hazard.
Do not disturb mechanism.
Keep David’s route response observable.
That was not agreement about what the maze meant.
It was enough agreement about what they could do next.
The pale center band advanced.
Lucan raised two fingers. “One comfortable adjustment left.”
Melody nodded. “Controlled move. Smallest useful shift.”
Lucan set it in pieces.
David held his one-step commitment.
Loria stayed where she could see the left symbols.
Bram moved first, one pace right and half a pace back, staying clear of the dark crescent.
Melody followed into the space Bram opened.
Joss shifted last among the front four so he could keep the stabilized seam in view.
Lucan closed the rear gap without compressing anyone against the right wall.
The route line brightened.
David looked down at it. “Stronger.”
Loria kept her gaze left. “Sequence still active.”
Bram pointed to the floor. “Crescent is redistributing.”
The dark boundary stretched toward their old positions, then curled inward across the space they had just vacated.
Joss said, “Runner still stopped.”
Lucan checked the gaps. “Formation stable.”
Melody wrote all five changes without choosing a winner.
ROUTE +.
SIGNAL persists.
HAZARD shifts into vacated space.
MECHANISM unchanged.
FORMATION viable.
For one count, the move looked almost clean.
Then Loria’s left-side sequence broke.
The first short point vanished.
The second short point lengthened.
The long point split into two.
Loria frowned. “That is new.”
Bram looked down. “Floor did what I expected.”
David looked at the right wall. “Route did what I expected.”
Joss listened to the seam. “Mechanism did nothing.”
Lucan looked at Melody.
The move had succeeded.
One channel had diverged anyway.
Melody stared at Loria’s symbols.
The old reflex came back fast.
If the signal changed unexpectedly, perhaps the move had been wrong.
But the other four channels had not failed.
The route had strengthened.
The hazard had moved into vacated space instead of under a foot.
The mechanism remained stable.
The formation held.
Even Loria’s channel had not disappeared.
It had changed.
Melody wrote:
Unexpected signal change after successful overlap move.
Then below it:
Do not score as failure yet.
Loria shifted her attention slightly along the exposed edge.
The split long point pulsed.
“Still responsive to investigation,” she said. “Different pattern.”
Melody added:
Signal function persists; output changed.
Melody compared the other three channels against Loria’s divergence. Bram’s floor boundary had moved with the formation into the available space, David’s route response had strengthened during the same shift, and Joss’s runner had remained stopped throughout. Their different responses gave her separate sensitivity traces instead of a single verdict.
She drew a new column beside the five outputs.
WHAT CHANGED IT?
ROUTE — David commitment + group shift right.
SIGNAL — investigation persists; configuration changed.
HAZARD — configuration changed.
MECHANISM — no change.
FORMATION — Lucan shift.
Then she looked at the lines again.
The divergence was information.
The signal had changed under a configuration shift even though its investigative target remained active.
The route had strengthened under the same configuration shift while David’s commitment stayed constant.
The hazard had redistributed immediately.
The mechanism had ignored the whole move.
Four different causal tracks inside one shared state.
Melody wrote:
Disagreement between outputs may identify different sensitivities.
Lucan read over her shoulder, then looked at the shrinking floor window. The comparison was useful only if they converted it into a move before the space disappeared.
The pale band answered by advancing again.
Bram looked at it. “Comfortable space is gone.”
Lucan stepped toward the right-side line and measured the remaining geometry. “Then the next move has to take us through, not just reposition us.”
Melody looked at the five outputs.
The right route was the only one currently pointing somewhere they could physically leave.
The left investigation remained informative but did not require bodies to stay left.
The left-center hazard made staying where they were worse with every advance.
The mechanism was stable enough to leave alone.
Lucan could move the group right if they preserved stagger.
This time she did not ask which method had earned authority.
She asked which actions satisfied the most constraints without destroying the other reads.
“Right passage,” she said. “David maintains commitment. Loria keeps the left symbols in sight as long as the angle holds. Bram calls floor redistribution. Joss calls any mechanism change. Lucan owns the stagger.”
Lucan was already placing them.
“David first. Melody second. Loria third until she loses sightline. Bram fourth. Joss fifth. I close.”
Loria checked the angle. “I can keep the symbols through the first half.”
Bram checked the right floor. “Clear now. I will call the boundary, not the surface.”
Joss kept one hand away from the wall. “Runner quiet.”
David moved.
The white line extended around the right bend.
Melody followed.
The left symbols changed again as the formation stretched.
Loria called the change without stopping. “Two long, one short. Still responsive.”
The dark crescent surged across the space they abandoned.
Bram did not watch the visible edge. He watched the direction of its curve.
“Keep going. It is folding behind us.”
Joss passed the stabilized seam.
The runner clicked once.
His head turned.
“Single click. No travel.”
Lucan kept the rear pair from compressing.
The right passage narrowed, then opened after six paces.
David’s route line stopped at the threshold of the widening.
He stopped with it.
Melody nearly ran into him.
“Why stop?”
“Response ended.”
Behind them, Loria lost sight of the left symbols.
The dark crescent filled the old section.
Joss’s seam disappeared around the bend.
Lucan brought the last person through.
For the first time in several minutes, the group stood in a space that was not actively shrinking.
Melody turned back to the page.
Their move had worked because no single channel controlled it.
David provided direction.
Bram constrained where not to place bodies.
Loria preserved a changing investigative output long enough to record how configuration affected it.
Joss confirmed the mechanism did not become the urgent problem.
Lucan made the geometry executable.
The unexpected signal divergence had not invalidated the move.
It had told Melody that signal state and route state were not merely two versions of one answer.
She crossed out the question she had written at the start.
Which one controls?
Under it she wrote:
Compare outputs.
Track what changes each one.
Do not rank by default.
David read the line from beside her.
“That one going to survive the next room?”
“It survives until the evidence says it does not.”
He almost smiled.
Ahead, the widening ended at three waist-high slabs set across the passage.
A narrow seam ran under the center slab.
The floor beyond the right slab looked clean.
Loria turned her attention toward the left slab.
Two pale symbols appeared.
David looked through the right gap.
“I know which one.”
Melody looked at him.
Nothing on her page independently confirmed it yet.
She left the new rule visible.
“Give me the narrowest claim.”