The Same Balance Does Not Mean a Return to the Same Position
At 15:20 Thursday, Chengwan Repair Station still has thirteen units of working capital. Judged only by the balance, this is one more than the twelve it had at the start on Saturday. Yet the team has not returned to a simple state of being “one unit ahead of the starting point.”
The three-unit initial payment has arrived. Two units have been used for G1 and the Tuesday segment of G2, while one unit remains subject to V2 conditions. Three half-day labor slots have gone to G1, G2’s first segment, and the conditional investigation. Wednesday’s window has closed, and both the inspection record and the relationship with the other party, who is waiting for an explanation, now exist.
What can be done today depends on more than what remains today. Earlier choices changed the uses of resources, the commitments between the parties, the evidence that can be retrieved, and the entrances to the next step. This chapter records those changes in a path ledger.
Chengwan, D17, labor slots, units of working capital, h, and d remain elements of a fictional case. “Path” here describes traceable relations within the case. It does not claim that every real project must produce the same consequences.
A Path Is More Than a Chronological Retelling
A running log says what happened Monday and what happened Tuesday. A path ledger must also answer which later options a choice eliminated, made more expensive, revealed, or newly created.
If something happened without changing the future feasible set, it may remain in the process log without being exaggerated into a critical path. If a small action changes a version, deadline, or authority, it may be a path node even when it costs no money.
Gu Ning therefore does not draw every message and reading with the same heavy arrow. She checks each node through five kinds of transmission: resources, commitments, technical state, information, and time windows.
The ledger is not meant to prove that the present is completely determined by the past. It lets the team see which constraints come from what has already happened, which choices remain open, and which supposed impossibilities can still be negotiated.
Define the Two Moments Being Compared
The starting point is Saturday before discussion of D17 began. Six of the following week’s ten half-day slots were already committed and four remained schedulable. Cash was twelve, of which at least eight was protected and four was available. R17’s first version was closed, and V2 did not yet exist.
The current point is 15:20 Thursday. Six of the ten slots remain committed elsewhere; G1, Tuesday’s G2 segment, and the investigation have each used one slot, leaving only one uncommitted. Cash is thirteen: eight protected, four from the original unallocated balance, and one unit of V2 conditional balance.
At neither moment is there an established new receipt window. The difference is that the present contains a V2 history, S-3 results, material on changing conditions, and one failure to deliver within the original window. The comparison must preserve both quantity and state.
Comparing total balances alone loses their uses. Comparing action labels alone loses remaining capacity. The path ledger places the starting point and the current point in the same column structure.
The First Path Item Is a Time-Limited Inquiry
The four-question message sent at 09:00 Monday used twenty-five minutes. It occupied no half-day slot and spent no working-capital unit. Those twenty-five minutes have passed and cannot be recovered, but they are not production capacity within the week’s ten-slot model.
This small action changed the information path. The replies at 09:42 and 09:50 jointly supplied scope, window, fee, payment, and acceptance proposals, turning “the other party may want to continue” into discussable V2 conditions.
Had the team not asked, G0, the initial payment, G1, and the later investigation would not have arisen along the present path. This counterfactual says only that they depended on conditions formed through the inquiry. It does not prove that the inquiry was the only possible starting point or that sending a message must produce similar replies.
The ledger marks this node as low resource consumption, completed, with information that cannot be canceled and no later commitment yet created. This avoids rewriting “the message was small” as “the message had no consequences.”
Conditional Confirmation Changed Authority
At 11:20 Monday, the parties confirmed the G0-stage version. Confirmation did not directly use a half-day slot, but it changed later actions from unilateral guesses into execution under a retrievable version.
It also created constraints: scope continued to cite the uniquely identifiable list in V1; receipt was limited to 14:00–16:00 Wednesday; the fee had two payment triggers; and when scheduling was impossible, materials could not be prepared in advance. The team cannot inherit the favorable initial payment and window while ignoring the acceptance and no-preparation conditions.
Version confirmation is not forever immutable. The parties can renegotiate future conditions, but they cannot pretend the old version never existed. History can be retained while the future is reopened. That is the difference between a path leaving a trace and a path becoming permanently locked.
In the ledger, G0 enters the commitment column rather than the cash-expenditure column. What it changes is who has authority to do what, and under which conditions.
The Initial Payment Did Not Make All Money Free
The three units received at 11:35 raised the bank-style total from twelve to fifteen. After one unit each was spent on G1 and Tuesday’s G2 segment, the current total is thirteen.
If money is treated as unlabeled water, the team appears to have one unit more than at the start. Classified by use, the remaining unit still comes from V2’s initial payment, and its later disposition is unresolved. Without confirmed conditions, it cannot automatically be merged into the original four units of free capacity.
The other two units have left the cash ledger and cannot be recovered by canceling future G2 work. Their work products, records, and learning may still have value, but “the expenditure cannot return” and “the expenditure left nothing” are different judgments.
The path ledger separately records the initial payment as received, used, remaining, and subject to disposition authority. This prevents the net balance of thirteen from concealing the source of the funds.
G1 Left a Passing Result and a Scope of Applicability
G1 used one slot and one unit. Under Monday’s low-h conditions, S-3 produced d=1.6. That slot and unit have been consumed. The result left behind qualified G2 for subsequent review and supplied a retained sample and baseline for Thursday’s comparison.
The 1.6 cannot be recovered as labor, nor expanded into “all conditions pass.” Its value depends on the object, procedure, equipment, time, and h position at that moment. Scope of applicability is part of the learning asset, not an added inconvenience.
Without G1, the team would not possess this low-range evidence. If it had stopped after G1, some limited learning might still remain. This node both consumed resources and opened a branch for comparing conditional change.
Gu Ning separates the “pass” label from the raw reading and its conditions. The former once triggered stage eligibility; the latter later helped explain the change.
G2’s First Segment Created a Partially Completed State
After passing the entry check current at 10:10 Tuesday, G2’s first segment used one slot and one unit that afternoon. It completed the planned front-end work but did not form the full result deliverable during Wednesday’s window.
The completed portion does not disappear automatically when the window closes, nor does “half done” by itself require continuation. Under a new version, it may be reusable, require rechecking or rework, or cease to apply because time and conditions changed. Current material does not decide among these cases.
The ledger therefore does not assign it a vague “half value.” Gu Ning records the completed scope, version, resources used, presently unknown usability, and validation entrance.
A partial result readily creates a psychological straight line: since the team has come this far, only the last step remains. In practice, changes in window, conditions, and capacity may have turned that last step into a different action. It cannot be obtained by simple subtraction from the old estimate.
Waiting Protected E4 but Consumed a Calendar Opportunity
Chapter 12 chose to wait until 11:00 Wednesday so that E4 could have priority use of Ye Cheng and C1. E4 finished at 10:20, so the existing commitment was not displaced by G2 and A_D did not occur.
Waiting itself did not register a labor slot as used or occupy a working-capital unit. It did concentrate G2’s executable time immediately before the receipt window, reducing the margin for handling an anomaly once discovered.
This does not mean waiting was necessarily wrong at the time. It protected a higher-priority item and preserved review at a defined node. The path ledger records actual consequences; it does not infer backward from B_D that every act of waiting was unreasonable.
Calendar opportunities cannot be inventoried. Once Wednesday morning passed, even a nominally unused labor slot could not be moved back to a moment before the original window. Resource and time ledgers must stand side by side.
The Pause Closed One Route and Preserved Another
Wednesday’s entry check first produced 2.3 and then 2.2 on rereading the same item. Neither met d no greater than 2.0. The team paused before releasing that day’s G2 segment and its expenditure.
The pause closed the route for completing V2 within Wednesday’s window under the original version. After 16:00, the original B_D had occurred; a new result could not flow backward into the missed delivery. At the same time, the pause preserved one unreleased slot and one planned unit of expenditure, avoiding greater exposure when the entry condition was unmet.
One action can produce both loss and protection. If the ledger records only “missed window,” it omits the avoided consumption. If it records only “successful loss containment,” it omits that the other party received no result and the final payment was not triggered.
Gu Ning places the two consequences side by side rather than covering them with one good-or-bad label.
Thursday’s Investigation Exchanged Remaining Capacity for Explanatory Material
Thursday’s low–high–low investigation used one half-day slot and no additional working-capital expenditure. It reduced uncommitted capacity from two slots to one, in exchange for paired readings from two objects under Q1 conditions, an R0 state record, and E0/E1 operating boundaries.
The investigation generated no new receipt window and did not authorize G2 to resume. It narrowed some explanations: neither an anomaly confined to b nor a single copying error could cover the common direction. It left Q1 influence, the measurement chain, execution, and other conditions unresolved.
Even if V2 never continues, this material may improve future condition records, although its transferability to other tasks remains untested. Learning is not naturally convertible into cash, yet it is not zero simply because it cannot be sold.
The ledger records the investigation as “one slot consumed, evidence formed, no increase in external delivery, transfer scope pending.”
Five States Cannot Be Compressed into Recoverable and Irrecoverable
The first state is consumed: three labor slots, two units of expenditure, and the elapsed window. Canceling future action cannot return them to the ledger.
The second remains but is protected: eight safeguard units and the last recovery slot. They physically exist, but existence alone does not release them to V2.
The third remains but has unsettled ownership or use: the one-unit balance from the initial payment. It is included in total cash but cannot yet be mixed with the original four units of unallocated capacity.
The fourth consists of results that can be carried forward within limits: S-3 records, h logs, Q1 pairs, version pages, and the communication interface. The fifth consists of changed relationships and deadlines: the original window is closed, the other party knows about the pause, and any continuation requires new conditions.
A binary classification would mistake the third state for a free resource, reduce the fourth to an intangible slogan, and omit the fifth from the ledger. The path ledger preserves all five.
Recoverability Must Specify a Recovery Action
“The material can be reused” becomes executable only after stating how it will be verified, how long verification takes, and who authorizes it. Without those elements, it is merely hope.
G2’s first segment is currently marked “retained; reusability pending review,” not “directly deliverable.” If a new version is formed, the team must at least check the object, version, exposure to conditions, and acceptance method for the new window. Those actions may themselves consume the final slot.
Likewise, refunding the one-unit conditional balance, crediting it to a new version, or recognizing it as settled income requires the other party’s terms or joint confirmation. The path ledger does not replace financial or contractual judgment. It only makes clear that no current evidence authorizes unrestricted use.
A strict recoverability entry includes the object, recovery steps, deadline, cost, and authority. If any element is missing, it remains in a pending-review state.
A Closed Branch Needs a Reason for Closure
“Complete V2 within the original Wednesday window” is closed because the entry condition failed that day and 16:00 has passed. The reason is not insufficient cash or a failure of E4: A_D did not occur, and cash did not cross the eight-unit line.
“Continue Thursday under the old plan and finish in one slot” is likewise unavailable directly, because the old receipt window and the conditions then in force have changed. The existence of the final slot cannot restore the old version’s time condition.
“Reclassify R17’s first version as pending” had closed even earlier. V2 activity does not alter the original Thursday report of 0.5 or Friday’s Y=0.
Writing a reason for every closed branch prevents the team from attributing all failures to one cause. It also identifies precisely what is missing if a branch is to be reopened.
A Newly Open Branch Is Not Necessarily the One to Choose
The Q1 result opened a design branch for a revised record and entry condition for operating conditions. The one remaining slot opened a branch for one limited recovery check or for handling exit. An existing communication interface may still support a proposal for a new window.
These branches are discussable possibilities, not authorizations. Each may require a new object, deadline, budget, and review of its worst consequence.
Learning is often presented as a reason to keep investing, on the theory that only continuation can “cash in the learning.” That reasoning turns learning into a debt. The material already exists; whether to use it in the future must be judged by the new action itself.
The path ledger therefore marks open branches as “candidates,” not as a to-do list.
Path Dependence Is Not Destiny
Current options do depend on the past: without G0 there would be no current V2 version; without G1 there would be no low-range baseline; without Wednesday’s pause, the unreleased resources from that day would not have been preserved. Each dependency can be identified.
Yet the team can still exit, refund or settle separately, or perhaps reopen under new conditions. The required cost and authority remain to be checked; this does not mean only one direction of action exists.
Saying “once we took this path, there was no way back” conceals negotiable and recoverable parts. Saying the past does not matter ignores real consumption and deadlines.
This chapter uses one operational meaning only: through resources, commitments, technical states, information, or time, past choices have made the current feasible set different from the set that would exist had those choices not occurred.
Temporal Order Is Not Necessarily Path Dependence
Monday preceding Tuesday is merely temporal order. The sequence enters a path explanation only when the present state requires a version or result formed Monday.
The h log rises over time, but sequence alone cannot prove that earlier conditions caused later conditions. Chapter 14 retained the common process, the Q1 setup, and other candidates.
Likewise, G2 failing after G1 passed does not show that passing caused failure. G1 is an authority entrance into a later stage; the relation between environment and reading belongs to another generative chain.
The path diagram uses different arrows for authority, resources, evidence, and possible mechanisms, so that every earlier-and-later event is not drawn as the same causal line.
Three Counterfactuals Check the Path Arrows
The first counterfactual asks what happens without the four-question message: the current V2 commitment chain would not form. It checks the information entrance. The second asks what follows if Wednesday’s entry passes: under the conditions then in force, G2 could be released, but final delivery would still not be guaranteed because the execution result remained unknown.
The third asks what happens without Thursday’s investigation: two slots rather than one would remain, but there would be no Q1 paired material. It checks the exchange between resources and learning.
These counterfactuals rely on explicit rules and do not invent complete outcomes for unobserved worlds. The team does not know what the other party would have done without the investigation or whether passing the entry check would necessarily have produced delivery.
A sound counterfactual changes one node and states how far the implication reaches. Writing an entire alternative history in excessive detail turns imagination into evidence.
Historical Cost Cannot Price Future Action
The team has already used two units, one investigation slot, and two production slots. Adding them together and saying “if we do not continue, all of it is wasted” uses past losses to pressure a future action.
Historical cost belongs in the overall project review, but it does not automatically become the return on the next unit spent. Future judgment must compare additional resources, consequences, and obtainable information from the present forward.
Conversely, reusable results from the past cannot simply be ignored. They may reduce the additional cost of a new version, but reuse must first be verified. History changes the inputs; it does not make the future conclusion.
Chapter 17 will separate invested, recoverable, and future incremental items into columns. This chapter first preserves the source of each item so that the next judgment does not count it twice.
Relational Commitments Also Leave a Path
At 11:25 Wednesday, the other party knew that delivery would not occur that day and was waiting for an explanation. The team had sent a limited account and had not received a new window. This shared knowledge changed the starting point for communication.
The next contact cannot proceed as though Wednesday never happened, and it should not claim unilaterally that an extension has been accepted. The available facts are only that the original window was missed, a bounded explanation for the pause was sent, and a response remains pending.
A timely explanation may preserve room for negotiation, but there is no usable scale for “how much the relationship improved.” The path ledger records what was sent, what the other party knows, and the absence of a reply. It does not force the relationship into working-capital units.
Commitment paths often lack inventory numbers but still alter which future actions are acceptable. They therefore need versions and timestamps rather than fictional valuations.
Capability Growth Must Be Separated from a Single Result
The team now knows how to execute S-3 before release, retain the raw h log, and perform a limited low–high–low comparison under Q1. These process capabilities are more concrete than they were Saturday.
One successful execution does not show that the capability is stable across all people, objects, and conditions. Ye Cheng both performed the work and knew the conditions; C2 still lacks procedure and usable-state records. A capability page must include applicable people, equipment, and unverified scope.
Capability can reduce the startup cost of designing a later investigation, and it also creates maintenance cost. If its records lose their versions, supposed experience decays into impressions.
Gu Ning uses three levels: “has been done,” “can be reproduced,” and “can be transferred.” Only some steps currently reach the first two levels; transfer is not checked automatically.
Action Itself Reshapes Later Possibility
In RC’s practical cycle, a complete world is not first given for a subject merely to choose among its contents. The subject’s inquiries, confirmations, inputs, waiting, and pauses intervene in the ground of possibility. They lock some paths, make others more expensive or close them, and generate new observational material.
The path ledger reduces this co-constructive relation to a traceable scale of action. It does not portray the past as immutable destiny. It shows how historical lock-in and remaining margin coexist: what has happened forms present reality, while what can still be recovered, negotiated, or reconstructed preserves room for renewed convergence.
How One Path Ledger Is Organized
Each row has seven columns: node and time, basis at the time, irreversible consumption, resources remaining, commitments formed, material obtained, and branches changed. The final payment that did not occur and Wednesday’s unreleased segment do not enter the used column.
Different arrow types connect the rows. Solid authority arrows show G0 permitting G1 and G1’s pass allowing G2 to enter review. Resource arrows show reductions in labor and cash. Evidence arrows show G1, Wednesday’s readings, and the Q1 material jointly supporting a judgment of conditional sensitivity.
Dashed lines indicate only candidate mechanisms or candidate reuse. The closed original window ends at a terminator; no arrow is drawn back into Wednesday.
Together, the table and diagram keep net amounts, temporal order, and causal relations from collapsing into one thing.
Current Summary of the D17 Path Ledger
The four-question message consumed twenty-five minutes and formed V2’s information entrance. G0 formed the version commitment. Three units of initial payment arrived; two have been used for G1 and G2’s first segment, while one awaits settlement of its use. G1 used one slot and left the 1.6 reading and low-range baseline. G2’s first segment used one slot and left a stage result whose reusability remains pending review.
Waiting protected E4 and compressed the margin before the original window. Wednesday’s failed entry closed the original delivery branch while avoiding release of the next slot and unit. Thursday’s investigation used another slot, formed evidence of conditional sensitivity, and reduced uncommitted capacity to one slot.
Current cash is thirteen: eight protected, four originally unallocated, and one subject to V2 conditions. Of the week’s ten slots, only one remains uncommitted. The three-unit final payment has not arrived, no new window has formed, and G2 remains paused.
This summary does not judge the historical actions as wholly right or wrong. It only gives the next judgment a current state in which nothing is counted twice.
Questions for Validating the Path Ledger
Tang Ke asks five questions row by row: if future action is canceled, what returns automatically; whose confirmation is required for anything else to return; which record can still be traced to its original version; which branch has actually expired; and which candidate branch exists only in the team’s imagination.
The two units of expenditure and three labor slots will not return automatically. The conditional balance needs explicit disposition. The S-3, h, and Q1 records can all be traced to their original pages. The original Wednesday window has expired. A new window remains only possible.
She then checks for double counting. The investigation slot cannot be both “consumed” and “available for recovery.” The initial-payment balance cannot be both conditional funds and free funds. The G1 reading cannot be both a historical result and a new delivery.
Only after these checks may the ledger enter the continuation decision. Otherwise it is merely a decorative project timeline.
The Pause Page at 16:00 Thursday
At 16:00 Thursday, Gu Ning freezes path ledger P17-1. The freeze fixes the history only through that moment; it does not prevent new rows from being appended. Any new reply, window, or refund arrangement must be added at its own time.
She sets no continuation percentage for “how far we have already come,” and she does not combine three labor slots and two units of expenditure into an emotional total. Different resources cannot be added at will, and historical value has not yet been compared with future increments.
What can be established is that D17 has left its Saturday starting point: the original V2 window is closed, three slots have been used, two units have been spent, limited evidence and a collaboration history have formed, and one recovery slot and eight safeguard units remain protected.
The next chapter begins from this frozen page and addresses the question most easily distorted by the path: after investing this much, should the team continue? The answer cannot come from the amount already invested. It must come from the additional action after 16:00 Thursday, its cost, recoverable results, future consequences, and exit conditions.