On Day Zero, a fictional shared workshop announces the abolition of its reservation system. Previously, users had to register in advance, enter according to assigned time slots, and return tools when finished. Some found the procedures cumbersome and were unable to use the equipment even when free time appeared on short notice. The new rule consists of a single sentence: beginning tomorrow, all members may come and go freely.
On Day One, the scope of access has indeed expanded. Time slots that were previously off-limits can now be entered. At the same time, the duties of handover, maintenance, and material replenishment have also been abolished, on the grounds that they belonged to the same old system.
On Day Five, some people arrive on site and cannot find the tools they need; others do not know whether the equipment has passed inspection and dare not continue using it; two jobs require the same workbench at the same time, the old schedule is no longer valid, and no new arrangement has yet taken shape. Viewed individually, each change removes one constraint; taken together, they begin to alter what the system can accomplish.
From Day Zero to Day Five, more people "may come," yet the conditions under which "coming leads to getting things done" are not necessarily better. The opening of the entrance is a real change, and the impairment of the capacity to act is a real change as well. Both hold simultaneously. The five phases that follow all advance along this clock: a deviation that could still be locally corrected on the previous day acquires new connections on the next, until correction itself loses its conditions.
This stress test does not argue for keeping all reservations, nor does it use the workshop to allude to actual events. The clock tests only one seemingly natural inference: whenever existing constraints decrease, the subject's possibilities increase. Before the test, we must first specify what it is that we are counting — permission, reachable action, or the capacity to keep choosing after acting.
Permission Does Not Complete the Conditions Required for Action
For a definite task, permission is only one condition among several. A member who wants to build a bookshelf also needs materials, tools, workspace, and sufficient time. If some irreplaceable condition is missing, nominal permission cannot, for the moment, be converted into the capacity to build. This does not mean that every action must be fully guaranteed by an organization, nor that freedom ceases to exist whenever conditions fall slightly short. Different tasks demand different conditions, and individuals can fill some gaps on their own. The role of analysis is to see the gaps and who bears them, and to prevent mistaking "the task has not yet been forbidden" for "the task can now be carried out."
In the thought experiment, some people bring their own tools, while others depend on what the workshop provides. The abolition of common maintenance affects the former less and the latter more. If one describes openness only from the former's position, one arrives at an incomplete judgment: since everyone is no longer constrained, everyone has gained the same opportunity. A change in common rules produces different outcomes through different relations of dependence.
RC discusses the available margin under the unity of subject and object and under sustainable decision-making. This chapter applies it as a question: how else can a subject alter its path under concrete conditions? We do not treat the Ground of Possibility in the philosophical sense as an inventory of social resources, nor do we infer from a "conservation of margin" that the tools each member loses must be compensated somewhere else.
Calling the refusal of premature fixation and the deliberate suspension of judgment "reverse observation" can expose premature convergence, but it can also misdescribe it as a discipline for making reality possible again. Yet observation cannot withdraw an injury that has already occurred back into a state of not having occurred, and suspension cannot excuse imminent action. Reverse observation can therefore serve only as a stress test — asking what a given fixation has excluded and whether it can be reopened — and cannot become a method for dissolving objects, meanings, or the capacity of others to judge.
Possibility that is not yet locked in at the level of ontology, and the choices a concrete person can bear at this moment, are not concepts that may be interchanged without explanation. If the intervening conditions of body, skill, relationships, and materials are ignored, abstract openness may be taken as an answer to concrete loss. When a person cannot complete his work, "the future still holds unlimited possibility" says nothing about how he is to continue today.
Some Constraints Create Shared Capacity
The old reservation did constrain the member. He could not use the workbench at all times, nor occupy it indefinitely. But other members were subject to similar constraints, and it was therefore possible for the slot he booked to mean anything at all. Surrendering some arbitrariness and gaining a stretch of comparatively reliable access to work are two aspects of one and the same arrangement.
We should not conclude from this that reservations are naturally fair. If a few people hold priority over the long term while other members can choose only the slots no one wants, the rule creates capacity while distributing it unevenly. A proper analysis must state separately the cooperative function and the mode of distribution; the necessity of the former does not excuse the latter.
By the same token, one cannot infer from unfair distribution that the workbench no longer needs any scheduling. Modifying the order of priority, limiting consecutive occupation, and reserving room for ad hoc needs can each be directed at a specific problem. Whether such measures apply requires testing, but they at least show that between keeping everything as it is and abolishing everything there are discussable changes.
RC's finite order emphasizes determinacy that can be coordinated and operated. In the workshop, a stretch of predictable working time is a local manifestation of just such determinacy. It need not remain forever unchanged, but it must allow users to arrange their next steps around it. With no commitment to rely on, the subject must renegotiate all the conditions of every action from scratch.
Renegotiation is sometimes worth its cost and sometimes too expensive. For long-term allocations that affect many people, renewed discussion may be a necessary correction; for routine handovers that already have boundaries and can be reliably executed, repeated demolition and rebuilding may devour time that ought to have gone into actual work. The value of openness cannot be detached from the cost of re-coordination.
Chaos Is First of All Someone's Experience
After the reservations were abolished, there are more voices in the workshop. Members who rarely appeared before begin to demand better slots; the original users find the disputes exhausting. Does this show that the institution is destabilizing? An increase in argument alone does not yet warrant the conclusion.
Suppose the earlier quiet came from some members having no entrance through which to speak at all; then the disputes may bring hitherto obscured needs into common observation for the first time. Greater difficulty in reaching agreement in the short term does not necessarily mean a decline in the capacity for long-term cooperation. The time it takes for problems to surface cannot all be booked as meaningless loss. But neither can every disorder be celebrated for this reason as the release of voices. If the people responsible for equipment inspection can no longer enter the site and necessary maintenance is interrupted consecutively, this consequence must be dealt with on its own. Growing participation in discussion and impairment of basic functions can occur at the same time. Respecting the former does not require denying the latter, and restoring the latter need not be conditioned on canceling the former.
"Chaos" is therefore a description that needs to be taken apart. One person may mean that he has lost his priority; another, that there is too much information to understand; a third, that the work originally planned can no longer proceed. Each of these experiences deserves recognition, yet they cannot be merged, undifferentiated, into a judgment about the whole.
This chapter discusses destabilization in terms of expectations for action, the alignment of responsibilities, and the capacity for correction; it does not quantify social situations by thermodynamic entropy or mathematical chaos. The word in the title supplies no method of calculation. Unless the system, the variables, and the relations are specified, the resonance of the terminology may conceal the parts of the analysis that remain undone.
Abolishing a Duty Does Not Abolish the Problem It Handled
Suppose the workshop previously had a person on duty who checked the return of tools. After the duty is abolished, one item of input disappears from the administrative ledger. But whether the tools are complete still affects the work that follows. Now every person on site must search and confirm on their own; if something cannot be found, they must ask previous users. Part of the work the organization shed has turned into work dispersed across the users.
We cannot assert in advance that such a transfer is necessarily worse. Perhaps the old checking was too burdensome and members' self-service suffices; perhaps the range of tools is so small that a dedicated role is unnecessary. What should be compared are the actual tasks and their consequences, not the preservation of every old procedure for the sake of praising maintenance. Yet if users each repeat the same searching while no one is able to update the common record, everyone who comes after must still begin again. The problem is not only that total input has increased, but that the input forms no result available to the next person. There is a great deal of activity, but collective capacity does not necessarily accumulate accordingly.
This gives another layer of meaning to "chaos is not an increase in possibility." People may be busier than before and take more improvised actions, yet merely keep replenishing conditions of coordination that already existed. An increase in the number of actions is not an increase in the range of tasks that can be completed, still less an enrichment of future choices.
RC's analysis of pressure flow requires us here to ask: who can delete a duty from their own ledger, and who must fill it back in within their life? If the drafters count only their own reduced expenditure while the members who depend on the tools bear all the searching and waiting, then the claimed gain in efficiency needs to have its observational boundary redrawn.
From a Single Gap to Consecutive Failures
Now restrict the problem to a single tool that is temporarily missing. If a spare exists, the user can continue and trace the matter afterwards; if there is no spare but the order of work can be rearranged, the impact may still be limited; if every subsequent task must first use this tool, then the same gap halts more and more work.
The consequences of a local loss thus depend on the mode of connection. One cannot declare a systemic collapse on the ground that a tool has gone missing, nor decide the loss is not worth addressing because it looks small. What matters is which tasks depend on it, whether substitutes can be brought into use in time, and whether waiting will cause other conditions to lapse as well.
Suppose, for example, that a member can move the finished product only on this day. If the making is delayed, the moving must be rescheduled. That the tool is later found does not automatically restore all the conditions lost in the meantime. Repairing one node and restoring a complete path of action are not always the same thing. This remains a deduction under stipulated conditions and offers no probability of failure for real systems. It prompts us, when evaluating change, to track the sequential dependencies among functions. If an evaluation asks only whether each resource eventually returns, it may underestimate the lasting impact of missed timing.
RC stresses moderate redundancy and multiple paths for critical parts. Here "moderate" demands differentiation: not every tool must be backed up without limit, but the question is how wide a range of action the absence of a single point would interrupt. Under finite resources, one should state the reasons for retaining a given substitute while admitting that other positions remain at risk.
Counterexample: The Old Order, Too, Can Exhaust People
At this point the reader may feel that this chapter is forever defending rules. We need to alter the thought experiment: suppose the old reservation system allowed a few members to renew indefinitely, while others, though bearing the same maintenance costs, could long gain no access to the equipment; their appeals met only with the advice to wait patiently. For the latter, the so-called stability provided no usable path at all.
Under such conditions, continuing the old arrangement has its costs too. Some people hold time open for waiting yet cannot begin work; some abandon projects they could otherwise have completed. If we tally only the losses newly added after a change while leaving untallied the losses of continuation, the comparison has already been tilted toward the status quo before it began.
Abolishing exclusive priority may be a necessary change. It costs the former beneficiaries some of their predictability while granting the excluded entry. The value of this redistributive change cannot be offset by pointing to more disputes overall. What this chapter therefore argues for preserving is the conditions that allow subjects to continue living and to amend arrangements — not an inviolable right attached to existing positions. Maintenance of common equipment may continue while exclusive distribution is changed; reliable records may be kept while the authority to monopolize their interpretation is adjusted. Which continuities are worth protecting requires concrete identification.
If the old mode of control is tightly bound to necessary functions, change will indeed be harder. But difficulty does not prove the old mode forever reasonable. It calls for a clearer design of transition: which functions need to be carried over, who is in a position to take them over, and what information and support the successors need. One cannot close the discussion with "there will be chaos," nor skip these questions with "change is necessary."
Transition Is Not a Blank Page
The workshop decides to rewrite the reservation rules. The old schedule cannot simply be cleared, because some members have already purchased materials and arranged collaborations around it; nor can new applications wait indefinitely, or the reform merely postpones entry under a new name. Between the two sets of commitments there is a real conflict, and no abstract principle makes them automatically compatible.
One discussable handling is to distinguish tasks with substantial investment already made and close to execution from reservations that have yet to involve concrete outlay, stating the conditions of conversion for each. This distinction, too, can be abused, so the reasons must be kept comparable and the disputes reviewable. A transitional arrangement must both recognize dependencies and prevent anyone from perpetually retaining privilege on the strength of claimed investment.
The handover of information likewise belongs to the transition. The former person on duty knows which tools await repair and which materials are already committed; if this knowledge is lost along with the change of role, new participants bear a groping that could have been avoided. Demanding that this knowledge be preserved is not the same as demanding that the old role retain full authority of decision.
RC's processual completeness means that a revision can still be met by the next observation. The new rules should therefore state what the trial period is to observe. If the aim is to enlarge new members' opportunities to use the workshop, it will not do to count only total hours of use; if the aim is to reduce waiting, one must distinguish for whom waiting has decreased and who has taken on a more uncertain schedule.
Nor is a trial naturally reversible. Materials may have been consumed, time has passed, and members' understanding of the commitments may have shifted. To say "we can always go back" requires checking which things can in fact be restored. Being able to change a form back does not mean being able to undo all the consequences the trial has produced.
Who Can Wait for the Next Round
For someone with an alternative workplace, a period of transition may be a temporary inconvenience; for someone who depends chiefly on this workshop, the same period may make an entire plan impossible to continue. Even if the average recovery time is identical, the participants' capacities to withstand interruption differ. This means that transitional responsibility cannot simply be divided evenly by headcount. It is not that those with less capacity to bear are thereby entitled to decide every arrangement, but that their constraints must enter the assessment. Otherwise those with the most fallback options can choose the longest trials, while those with the fewest are asked to pay all their opportunities up front for the sake of common learning.
Support can take many forms: reserving necessary hours of use, providing temporary substitutes, changing responsibilities in stages. We do not prescribe any one of these as the universal answer. One should show, in light of the dependencies already identified, how the support keeps some path reachable, and by whom its costs are jointly borne.
We also need to admit that some losses cannot be adequately compensated. Admitting this does not make choosing easier, but it prevents a future aggregate gain from being used to declare, in advance, that all present costs have already been offset. Gains and losses fall on different people at different times; they must pass through further judgment and cannot be settled automatically by the language of addition.
How Openness Turns into a Transfer of Costs
Is the preservation of optionality, by itself, already enough to constitute a good decision? As long as whose optionality is meant remains unspecified, the answer may be unsettling. A subject can strive to keep several retreats of its own open while pushing the costs of adjustment onto others. Viewed locally, its decision is quite sustainable; viewed from the position of those onto whom the costs have been shifted, the same decision may keep narrowing a life. This is not an extraneous problem lying outside RC. A theory that understands power as a capacity to converge possibility into structural concentration must face the fact that this capacity can be used to preserve margin selectively. An advantaged position is able to keep itself free to rebook, withdraw, or switch plans at any time, while requiring others to remain on call indefinitely. One person's flexibility is conditioned on another group's inability to change.
In the workshop thought experiment, the manager can switch to another venue at any time yet requires the members who depend on this site to accept an interruption without term. So long as the manager can still keep experimenting, the reform is declared still to have a chance. Such a judgment takes one's own repeatability of attempt as proof that all participants can still afford another round. This dark usage must be named directly as the asymmetric distribution of optionality; it cannot be smoothed over with "the whole still has possibility." The concept of the whole can conceal who has already dropped out, who cannot last until the improvement arrives, and who bears no equal consequences when the trial fails. A sustainable decision-making that does not address the scope of its subjects may become a language by which the advantaged preserve their advantage.
This book's critique of it also needs an explicit normative basis: those affected by a decision and bearing its continuing losses must have their subsequent lives enter the assessment. This is not a good that follows automatically from a system's mere capacity to persist. The descriptive capacity of the theory and the ethical demands we are willing to defend must be established separately.
Why Accelerated Collapse May Be a Position That Transfers Costs
Another dark corollary declares that since the old order locked possibility in, the faster it is brought to failure, the more readily a new world opens. This claim joins the removal of constraints directly to the creation of conditions for action without saying who takes over once basic functions are interrupted. Its appeal is intelligible: when gradual change has been thwarted again and again while the old arrangement continues to exact its losses, people come to doubt whether revision means anything anymore. But disappointment with gradual paths does not by itself prove the rupturing path better. The comparison must still include the losses, the capacities for taking over, and the irreversible consequences of both the old path and the new.
The position of those issuing the claim especially demands scrutiny. People with alternative resources can endure a longer chaos, yet may mistake this capacity for endurance for a condition shared by all; those who depend most on existing basic functions may be the first to lose the chance of living to see the new arrangement. The future openness is promised to everyone, but the losses of the transition do not arrive evenly.
Criticizing such a position requires no assumption that its advocates are necessarily malicious. The conviction can be sincere, and the transfer of costs can exist all the same. To stress only sincerity misses the consequences; to infer malice from consequences alone overruns the evidence. What we must test is whether the claim states who takes over and how the burdens are distributed — not replace the comparison of paths with moral labels.
The more dangerous interpretation calls every loss a necessary clearing and every demand to lessen loss a defense of the old order. Failure then no longer weakens the claim; it becomes proof that the transformation has not been thorough enough. A theory that announces the opening of possibility has thereby canceled the possibility of correcting its own direction.
Crisis, Too, Can Be Used to Freeze Correction
Opposite to accelerated collapse, the preservation of order can also become a dark usage. A position can bind its own decisions to basic functions and claim that any challenge to it will cost everyone their existing security. Whether the crisis is real and whether authority may be expanded without limit are two different questions.
A real interruption may call for centralized coordination, but the centralization should be stated around the corresponding tasks, time limits, and responsibilities. If the scope of the task keeps expanding, the interim measures have no conditions for review, and the return to normal recedes ever further, the crisis may pass from being a problem to be solved into being a rationale for maintaining an exceptional position.
In the workshop, that an equipment failure requires suspending a certain use does not automatically support forbidding discussion of the fairness of reservations; that the maintenance staff need to schedule inspections does not thereby grant them authority over every qualification of the members. Bundling these different matters together makes every local correction look like a threat to common survival.
RC's analysis of continuance must be able to recognize this binding. Maintaining basic functions and maintaining the current configuration of power are not always the same aim. Without differentiation, the language of stability may supply a theoretical garb for control beyond appeal, in conflict with the direction of continuing correction that it originally emphasized.
The Reverse Demands a Theory Must Bear
These corollaries do not prove that RC necessarily leads to domination, but they do prove that the consistent appearance of concepts cannot substitute for ethical and factual judgment. The preservation of margin can be used selectively, the rebuilding of order can be used to tolerate others' irreversible losses, and the maintenance of stability can be used to cancel correction. This book therefore cannot merely describe malicious actors distorting a theory that was risk-free in itself. Once a theory explains how the conditions of observation, structures of dependence, and paths of action are changed, it also exposes the power seated in these positions. It must submit to examination by those affected, and it must distinguish which conclusions merely describe how some mechanism may operate from which conclusions still require independent grounds of legitimacy.
The most direct test is not whether a scheme uses the words openness, resilience, or sustainability, but who still has a way out when failure arrives; who must answer when the injured point to the costs; and whether the existing decision can still change when an alternative interpretation appears. If the answers always protect the same position, then however complete the theoretical language, it cannot exempt the actual distribution from critique.
Halting a Trial Should Also Be an Available Option
If correction can occur only in the direction of further enlarging the original scheme, the trial has not preserved a full capacity to learn. When the workshop finds that a new handover procedure keeps losing track of tools, it may revise the procedure — and it should also be able to suspend it and use a workable alternative already at hand. What costs a suspension entails should be stated, but it cannot be ruled out merely because it makes the initial decision look unsuccessful. The alternative here must still be executable. When the old records are already lost and the former successors have already left, announcing a return to the past does not mean there is a path back. The input needed to preserve a usable alternative is precisely one of the costs that should be counted before a trial begins.
This requirement constrains the advocates of reform and the defenders of the status quo alike. The former cannot treat stopping as betrayal; the latter cannot take one suspension as proof that no further change is ever to be allowed. The value of a bounded failure lies in being able to alter the judgment about a given measure without having to seal off the entire future.
Preserving for Change the People Who Will Carry It
At the end of the thought experiment, the workshop abolishes exclusive renewal, keeps the maintenance of equipment, publishes an entrance for handover, and sets up a space, adjustable, for ad hoc needs. These arrangements may still be flawed, but they no longer treat all rules as one thing to be preserved or destroyed.
To evaluate a change, one must first know which conditions of action disappear, which come into being, and which are merely transferred to others to bear. Next, one must know who, in the passage from one set of conditions to another, can keep working and whose path is cut off. Finally, one must know whether, once a deviation appears, the people, information, and resources needed for correction are still there.
Chaos may accompany new opportunities and may equally accompany the loss of old capacities; the relation between the two must be spelled out concretely. An openness genuinely worth striving for is not exhausted by the removal of the prohibition on the door; it includes being able to work after entering, being able to propose changes, and being able to continue acting after a misstep.
When we speak of the future in the name of possibility, we cannot omit the people who will live in that future. They are not conditions that may be temporarily shelved while a new order takes shape. The meaning of the change must be tested by their capacity to act thereafter.