Scientific specialisation has made it possible to enter very deeply into the substance of the real. A cell, a chemical reaction, a brain, a planet or an institution each call for different instruments and methods. This division has produced considerable precision. Certain difficulties appear when the observed behaviour depends strongly on the relations between several elements, on their organisation and on their history.
The living makes this difficulty particularly visible. Describing the organs of an organism yields an immense quantity of information without fully restoring the way it maintains itself. Knowing the molecules present in a cell hardly suffices to explain its organisation. Counting the trees of a forest tells us little about the soils, the water, the fungi, the interactions between species or the capacity to regenerate after a disturbance. Part of the information lies in the components, another in the relations that make them functional within a milieu.
Ludwig von Bertalanffy placed this question at the centre of general systems theory. The various disciplines already had powerful models in their respective fields. Certain phenomena additionally required studying the interactions, the exchanges and the organisation of the whole. Specialised knowledge keeps all its value; bringing it into relation becomes necessary when the observed properties depend on several levels at once.
An organism illustrates this organised continuity. It breathes, feeds, transforms energy, eliminates waste, renews its cells, repairs certain damage and modifies its responses according to its environment. Its matter changes ceaselessly while its organisation retains enough continuity to hold the whole together.
Ilya Prigogine's work on non-equilibrium systems gave an important place to flows in understanding certain organised forms. The dissipative structures studied in thermodynamics can maintain themselves thanks to the exchanges of energy and matter that traverse them. They describe only a limited part of what characterises the living. They nevertheless show that an organisation can depend on a permanent movement.
Heart rate varies, breathing adjusts, the immune system modulates its responses, and organisms modify their behaviour according to the conditions encountered. Biological continuity rests on these changes being contained within ranges compatible with the functioning of the whole. A stable appearance tells us rather little about the effort needed to maintain it. An organism can continue its activity through a period of prolonged pressure while its recovery deteriorates. A forest can keep a substantial canopy while losing certain capacities of renewal. An institution can keep functioning thanks to people who continually compensate for its defects.
The visible state is a partial snapshot. The trajectory tells us what had to be mobilised to arrive there.
Stability, variation and the cost of maintenance
C. S. Holling's work on ecological resilience helped shift attention toward the capacity of a whole to absorb disturbances while preserving its essential functions. An ecosystem can change a great deal and keep a viable organisation. Another can appear stable over a long period while its possibilities of response diminish.
Recovery time provides precious information. Two systems exposed to a comparable disturbance may return to their usual regime at very different speeds. When that return progressively demands more time or more resources, the apparent state no longer tells the whole story.
Variation deserves the same attention. A living organism rarely produces a perfectly identical response from one moment to the next. Some fluctuations accompany a good capacity for adjustment. Excessive regularity can signal a loss of suppleness in certain contexts. Fluctuations that amplify, persist or become hard to resorb may indicate a deterioration of regulation.
Amplitude alone remains insufficient for interpreting this behaviour. The duration of the disturbance, the time needed to return and the evolution of these parameters across observations give more solid information. A rapidly absorbed variation means something different from a comparable phenomenon that leaves a trace over several cycles.
This distinction makes it possible to separate useful variability from disorder. Movement belongs to the living. The problem appears when the available responses become fewer, more costly or less able to bring the system back into a zone compatible with its functioning.
The term margin here denotes what the system can still mobilise when a difficulty arises. This margin takes different forms according to the object studied: energy reserve, available time, redundancy, functional diversity, mobilisable competence, alternative pathway or capacity for reorganisation. A diversified forest, an electricity network and a work team have margins of very different natures.
Their evolution remains comparable on one point: the number of accessible responses can diminish. An alternative pathway disappears, a recovery time lengthens, a resource once available becomes indispensable to daily functioning, a team no longer has anyone to absorb the unforeseen. The margin then describes the space remaining before a disturbance forces the system to alter its functioning substantially.
The cost required to maintain the present state completes this observation. Some costs are fairly easy to measure: money, energy, materials, overtime, staff mobilised or frequency of intervention. Others appear less readily in the accounts. Trust erodes, attention fragments, know-how disappears with departures, relationships deteriorate or a collective memory is lost.
These losses can take much longer to rebuild than certain material resources. Replacing a piece of equipment or recruiting a person can be relatively quick. Reconstituting an experienced team, a professional culture or a trust accumulated over years belongs to another temporality.
The cost of maintenance then becomes information in its own right. Producing the same result with more corrections, more resources, more overtime or more fatigue indicates a change in functioning even before any fall in performance appears.
The timescale chosen strongly influences this reading. An organism combines responses occurring within seconds, recoveries taking several days and slower adaptations. A forest follows seasonal cycles while transforming over several decades. An institution may absorb a one-off crisis within a few weeks and progressively lose competences over years.
Too short an observation can turn a normal fluctuation into a worrying signal. An average built over too long a period can efface a recent deterioration. ORI-C must therefore adapt its observation windows to the phenomenon studied and retain several scales when rapid and slow developments intersect.
Thresholds, memory and bifurcations
Research on regime shifts has shown that gradual modifications can produce rapid transformations. Some lakes can absorb an increase in nutrient inputs for a time before feedbacks favour a turbid state. Pressure rises progressively while the visible change appears much later.
This non-linearity makes trajectories deceptive when attention stays fixed on the state of the moment. A long period of compensation may precede a rapid change. The last disturbance then attracts all the attention, although it meets a system already modified by what preceded it.
The return may follow a different path. Some systems fairly easily recover a functioning close to the previous one when pressure decreases. Others durably retain the mark of the tipping. Internal relations have changed, certain resources have been consumed, species have disappeared, or new feedbacks sustain the configuration reached.
Hysteresis describes this dependence on the path travelled. Reducing the initial pressure may be insufficient to restore the previous regime. History modifies present possibilities.
Not all losses have the same reach. A structural irreversibility concerns an element genuinely lost at the scale considered. An extinct species or a consumed non-renewable resource does not reappear with the mere return of earlier conditions. Other transformations close off above all the previous path. A function may be rebuilt differently, an organisation may develop new competences, an ecosystem may evolve toward a different composition able to ensure part of the earlier functions.
This second situation is rather an irreversibility of trajectory. The old path has disappeared, while others remain accessible. The difference prevents every threshold from being treated as a definitive disappearance and obliges us to look at what can still be rebuilt, with which resources and within what time.
Viability, moreover, imposes no systematic return to the initial state. A whole can pass through a significant disturbance and build a different organisation that remains able to function under the new conditions. Populations may shift their range, ecological communities may recompose, and human organisations may profoundly alter their practices after a crisis.
A transformation becomes worrying when it durably reduces the possibilities of response, increases the cost required for maintenance, or makes the system dependent on compensations that are hard to renew. A viable bifurcation produces another result: it reorganises the functions while preserving enough resources, recovery and future possibilities.
The past then serves as a historical reference rather than as a model that must be restored.
This distinction matters for ORI-C. Observing the proximity of a threshold yields information. One must also look at what crossing it risks making inaccessible and at what a new organisation might still make possible.
When fragility circulates
The systems studied often depend on other systems. Flows of energy, matter, money, information or labour cross their boundaries. A local stability may rest on resources drawn from elsewhere or on constraints absorbed by other wholes.
A company may preserve its results by increasing pressure on its subcontractors. An economy sustains part of its consumption thanks to resources extracted in other territories. An administration keeps complex procedures because its staff or its users do the work needed to work around certain inconsistencies. A family sometimes takes on what a collective arrangement can no longer absorb.
The burden changes destination without disappearing.
The boundary chosen by the observer then becomes decisive. A rise in productivity within a company can coexist with more overtime, suppliers placed under pressure and increased resource consumption. Each measure describes a real part of the situation. Bringing them together reveals how the cost circulates.
The viability of a system also depends on what it demands of the wholes from which it draws its resources or which absorb its constraints. A stability obtained by continually reducing their own margins progressively weakens the relations on which it depends.
Robert May showed, in his work on ecological modelling, that more interactions did not automatically lead to greater stability. The structure of the relations, their intensity and their distribution strongly modify the behaviour of the network.
This distinction remains useful for understanding contemporary networks. High connectivity accelerates exchanges and multiplies the available pathways. It also accelerates the propagation of certain disturbances. The quality of the network then depends on its capacity to process what circulates.
Integration presupposes differentiated functions. Some information can be filtered locally, some disturbances contained; several pathways can ensure a function when one link disappears. Modularity sometimes limits the propagation of a problem while preserving the relations necessary for general coordination.
The living provides many examples of this distributed organisation. Functions are specialised, barriers filter exchanges and certain responses remain local. Faster transmission improves nothing when the system loses the capacity to select what deserves to be propagated.
Digital networks make this difference very visible. They transmit an immense quantity of information almost instantaneously. This power facilitates communication and coordination at a distance. It also accelerates errors, rumours, emotional reactions and certain economic disturbances. Human capacities for verification and contextualisation progress far more slowly than the volume of signals received.
Saturation appears when the system continues to receive information while becoming less able to assign it a proportionate importance.
Observing without forgetting the observer
Indicators pose a particular difficulty in human systems. People often adapt their behaviour to the criteria used to evaluate them.
A school may alter certain priorities according to the results that will be measured. A company orients part of the work toward the numerical objectives used to judge its teams. An administration may reduce an official processing time by shifting more procedures onto the user. The measurement continues to describe something real, while its use progressively modifies practices.
Observation gains in soundness when it also tracks these effects. Accumulating further indicators risks merely increasing the noise if each reproduces the same blind spot. Cross-checking different measures makes it possible to confront their results and to spot their contradictions.
A rise in productivity accompanied by an increase in absences calls for a different reading from a gain obtained with a stable workload. A reduction in administrative delays takes on another meaning when the time demanded of users increases. The gap between several indicators can become more instructive than their separate evolution.
Some losses remain hard to convert into figures. Trust, collective memory, the quality of a relationship or the progressive disappearance of a skill can be detected through field observation, interviews or the tracking of events that seem secondary taken in isolation. The difficulty of measuring them does not reduce their role in actual functioning.
ORI-C can then observe what the instrument makes visible, the behaviours it encourages and the phenomena it leaves out of frame. An indicator relevant at one moment may lose part of its value when actors learn to optimise their practices around it.
In human systems, the point of view adopted must also be explicit. A management board, an employee, a user and a territory do not experience the same costs and do not have the same margins. A policy can improve a general result by concentrating its constraints on a part of the population.
An assessment of viability must then specify the boundary studied, those who bear the costs and those who benefit from the functioning obtained. This precaution prevents a system from being declared viable merely because its principal functions continue while some of its members exhaust themselves to maintain them.
The model retains an important limit here. It can help make tensions visible, compare developments or signal the reduction of certain capacities. The choice of what should be protected, transformed or accepted remains a human decision.
Seeing before the break
Maturana and Varela placed self-maintenance at the centre of their work on autopoiesis. A living system continually produces the components that take part in its own organisation. What arrives from the milieu meets an already present structure, which partly explains why a similar disturbance can have different effects depending on the state of the system.
Heavy rain may be absorbed by a soil in good condition and cause far more run-off on a heavily degraded soil. A professional workload bearable after a period of rest may become difficult after several months of exhaustion. The event takes on its meaning within the trajectory it meets.
Costs of maintenance, available margins, recovery times, thresholds, possibilities of return and transferred constraints each give partial information. Their convergence becomes more interesting than any one of them taken alone.
An isolated signal deserves to be followed. Several persistent signs evolving in the same direction give the observation more weight. A recovery that is slowing, costs that are rising, alternative pathways that are disappearing and compensations that grow ever more frequent describe a situation different from a one-off fluctuation.
Alert can escalate with the quality of the available information. A first signal calls for closer observation. A persistent trend justifies looking for its causes and its extent. The simultaneous deterioration of several capacities can make an intervention reasonable before a precise threshold is known. Once the tipping has been observed, attention shifts to the possibilities of recovery or reorganisation.
This gradation keeps a place for uncertainty. It avoids announcing a rupture on the basis of an isolated phenomenon and limits the risk of waiting for an unattainable certainty before reacting.
Communicating a fragility demands the same precision. A useful observation states what has been measured or noted, for how long, with what limits and which capacities appear to be changing. Uncertainty is part of the information.
ORI-C can work within this space without reducing a forest, a cell, an institution and an economy to the same functioning. Their structures, their temporalities and their constraints remain proper to each domain. Comparisons become useful when they bear on sufficiently precise properties: cost of maintenance, recovery time, reduction of margins, change in fluctuations, transfers of burden, closure of certain possibilities, or capacity to build a new viable organisation.
Observation then bears on the way a system continues to function as much as on its immediate result. Some systems maintain their activity by progressively consuming their reserves and those of the wholes around them. Others pass through a significant disturbance, change profoundly and recover possibilities of response in another form.
The difference appears in what they retain after the effort: available resources, capacity for recovery, diversity of responses and the possibility of reorganising when conditions change.
The first signs of fragility often appear before the visible break. They lie in what must be mobilised more and more often to maintain the same functioning, in what takes longer to rebuild, and in the possibilities that disappear over time.