Living systems: when flows leave traces
How far-from-equilibrium organisation makes regulation, memory and adaptation possible
Didier Daloze · ori-c.be · ORI-C
A living cell and a cell whose activity has ceased can still contain, for a certain time, a large part of the same molecules. Yet their futures rapidly diverge. The first sustains ionic gradients, renews its components, regulates its exchanges and repairs part of its damage. In the second, these processes progressively cease to be coordinated. Gradients dissipate, reactions are no longer functionally coupled and the organisation that held the whole together loses its coherence.
The difference therefore does not lie only in the matter present. It lies in a dynamic organisation able to use continuous flows of matter and energy to maintain its own conditions of existence.
This observation nevertheless seems to raise a paradox. If the second law of thermodynamics implies that irreversible transformations are accompanied by a global increase of entropy, how can structures as organised as cells, organisms or ecosystems appear and maintain themselves?
The answer presupposes no exception to physical laws.
Living systems are not isolated systems. They continually exchange matter and energy with their environment. They capture free energy from light, nutrients or chemical gradients, convert part of it into biological work, then dissipate heat and release metabolic products. Their local organisation thus remains compatible with the global increase of entropy.
Living systems therefore do not stand above thermodynamics. They organise the flows and transformations it makes possible.
Their singularity lies not in the creation of a costless order, but in the emergence of structures able to channel flows, preserve differences, detect certain perturbations and, in some cases, keep traces that will modify their future responses.
Living order is not opposed to entropy
Entropy is often presented as a measure of disorder. That image can be useful as an introduction, but it quickly becomes insufficient. In statistical thermodynamics, entropy is related notably to the number of microscopic configurations compatible with a macroscopic state. It also makes it possible to characterise the dispersion of energy and the share of it that is no longer available to produce work under given conditions.
In an isolated system, the differences liable to produce work tend to fade. A temperature difference decreases, a concentration diffuses and a gradient progressively loses its capacity to drive an exploitable flow.
Living systems function in another regime because they exchange constantly with their environment.
A cell imports resources, releases products, sustains electrochemical gradients and renews its constituents. It does not reduce total entropy. It maintains a local organisation thanks to transformations accompanied by a global dissipation.
The work associated with Ilya Prigogine showed, in certain classes of physicochemical systems held far from equilibrium, that organised structures can emerge and persist thanks to the flows that run through them. He called them dissipative structures to stress that they do not exist independently of the exchanges that sustain them.
A flame, a vortex or certain chemical structures can also be described as dissipative regimes. Their form persists as long as the conditions that feed them remain present.
A cell nevertheless has additional capacities. It renews the components required for its maintenance, controls part of its exchanges, coordinates reactions, detects certain deviations and repairs certain damage.
Living systems are therefore not merely a form produced by a flow. They are an organisation that uses flows to preserve the relations on which their own continuity depends.
Living systems are not a state. They are a regime of maintenance continually rebuilt.
Maintaining differences far from equilibrium
Biological viability does not correspond to a motionless equilibrium.
A living cell maintains differences of concentration between its interior and its environment. It preserves electrical potentials across its membranes, regulates its acidity, renews its proteins and distributes its resources among many functions.
These differences are not imperfections that living systems would seek to remove. They make their activity possible.
A cell membrane, for example, is not merely a boundary. It makes it possible to maintain different compositions on either side of its surface. Proteins control exchanges, transport ions and help produce electrochemical gradients.
These gradients contain a free energy liable to be converted into work.
Cellular respiration provides one of the most remarkable examples. In mitochondria, a chain of reactions helps establish a proton-motive force across a membrane. ATP synthase then uses that gradient to produce ATP. This molecular machine has a rotary mechanism powered by the electrochemical gradient of protons.
ATP is often presented as the energy currency of the cell. The image is useful, but it is not an abstract reserve. ATP takes part in chemical couplings that make possible processes such as molecular synthesis, active transport, movement or the modification of proteins.
Living systems therefore do not passively resist equilibrium. They continually spend energy to maintain the differences that make their activity possible.
From homeostasis to homeodynamics
Homeostasis designates the capacity of an organism to keep certain variables, such as temperature, pH or blood sugar, within ranges compatible with life. It remains indispensable for describing many regulatory loops. But it does not on its own sum up the continuity of living systems.
An organism does not preserve its viability by constantly returning to an identical state. Its components are renewed, its rhythms vary, its priorities change and some experiences durably modify its capacities of response. Even when a variable returns to a functional range, the system may have been transformed by what it has just gone through.
The term homeodynamics, already used in some work on ageing and on the maintenance, repair and adaptation capacities of biological systems, makes it possible to describe this continuity produced through change. It is therefore not a creation of the present essay, but an existing conceptual framework whose use is specified here.
Homeodynamics can be understood along two complementary dimensions. As a phenomenon, it designates the capacity of a living system to maintain its viability while renewing its components, modifying its rhythms and reorganising its responses. As an angle of observation, it consists in attending to the flows, delays, margins, costs and transformations that make that continuity possible.
Homeostasis thus describes the maintenance of certain variables. Allostasis describes the adjustment of regulations and priorities according to present or anticipated needs. Homeodynamics widens the view to the permanent reconstruction of the living system as a whole.
These notions are not mutually exclusive. They illuminate different dimensions of a single functioning. A homeostatic regulation can take part in an allostatic response, itself inscribed in a wider dynamic of renewal, repair, plasticity and adaptation.
Living systems therefore do not necessarily maintain a fixed state. They maintain a viable trajectory through continually organised changes. That dynamic organisation rests on precise material architectures able to channel flows rather than simply undergo them.
Channelling flows rather than undergoing them
The sophistication of living systems lies largely in the precision with which they organise transformations.
Membranes control certain exchanges. Enzymes accelerate certain reactions. Metabolic networks distribute resources. Signalling systems coordinate responses. Feedback loops stabilise certain variables or temporarily amplify certain processes.
Energy therefore does not circulate freely within an organism. It crosses architectures that modify the accessible pathways and make some transformations more probable than others.
Enzymes clearly illustrate this principle. They lower the kinetic barriers of particular reactions without modifying the thermodynamic balance that distinguishes favourable from unfavourable transformations. They allow biologically useful reactions to occur at a rate compatible with life.
Molecular motors likewise show how an organisation can produce directed movement in an environment dominated by thermal fluctuations.
At the molecular scale, Brownian agitation is permanent. Biological structures do not remove it. Molecular motors couple conformational changes, structural asymmetries and inputs of free energy in order to produce an oriented transport.
Directed movement therefore does not arise from the suppression of chance. It results from an architecture able to exploit certain fluctuations while consuming energy to maintain a privileged direction.
The image of a biological Maxwell's demon can then become illuminating, provided it remains an analogy.
In Maxwell's thought experiment, a small agent sorts molecules by their speed and thus seems to decrease entropy at no cost. Living systems do indeed carry out countless operations of selection. Receptors recognise signals, membranes filter certain exchanges, enzymes distinguish substrates and cellular systems transport molecules to precise destinations.
But these operations rest on material mechanisms. Recognition, sorting, control and the maintenance of differences require an expenditure of energy and are accompanied by dissipation.
Living systems resemble less a demon escaping thermodynamics than a demon continually paying the material cost of the distinctions it maintains.
Biological information has a material support
Biological information does not exist independently of the structures that carry it.
DNA must be synthesised, copied, repaired and transmitted. Proteins must be produced and then renewed. Cellular states are maintained by networks of reactions. Immune and neuronal memories rest on physical modifications that must themselves be sustained.
Biological memory is therefore not an immaterial property added to matter. It corresponds to the persistence of certain differences in structures able to influence future responses.
Landauer's principle is often invoked to link information and thermodynamics. Its reach must nevertheless be stated precisely. Landauer showed that a logically irreversible operation, such as erasing information, implies a minimal dissipation into the environment. He did not establish that every copy, every synthesis or every biological memory directly had the same thermodynamic cost.
The real cost of biological memory comes mainly from the processes required for building, copying, stabilising, repairing and renewing its supports.
Landauer recalls a more general idea: physically embodied information cannot be separated from the material transformations that allow it to be processed.
But it does not by itself constitute an energetic theory of DNA, of immunity or of neuronal memory.
Living systems therefore do not spend energy only to erase information. They spend it continually to maintain structures far from equilibrium and to prevent the functional differences those structures contain from disappearing.
Memory nevertheless brings a new capacity. It allows a past experience to influence a future response. It avoids every adaptation having to be rebuilt entirely from scratch.
Not every regulation is a memory
The passage from perturbation to signal must be described with caution.
A cell can detect a change and produce a response without durably preserving the trace of that event. An osmotic perturbation can, for example, trigger compensation mechanisms that bring certain variables back into a functional range. When the perturbation disappears, the system can return close to its earlier regime.
There is then a regulation, but not necessarily a durable memory.
Three levels can be distinguished.
Reactive regulation
A perturbation is detected and triggers a response that helps preserve or restore a function.
The system corrects the deviation without the future relation between stimulus and response being necessarily modified.
The sequence can be summarised as follows:
Perturbation → detection → regulation → functional return
Memory
The experience leaves a modification that persists after the perturbation has disappeared.
The system does not return exactly to its earlier state. A molecular, cellular, immune, neuronal or structural trace remains.
Immune memory illustrates this persistence. Some responses produce cell populations able to modify the reaction upon a later exposure.
Learning, or acquired adaptation
The trace left by experience modifies the way the system will respond to a future situation.
Learning therefore does not consist merely in preserving a mark of the past. It presupposes that this mark transforms a later response, a threshold, a sensitivity or a capacity.
Within the functional framework adopted here, we shall speak of learning when this trace durably modifies a future response. That definition is deliberately broader than those which reserve learning for behavioural changes issuing from experience. Depending on the discipline, some persistent modifications will rather be described as sensitisation, acclimatisation or physiological adaptation.
A regulation helps to maintain or restore a variable or a function within a viable range. A memory preserves a modification after the perturbation has disappeared. Learning appears when that trace durably transforms the way the system responds to a future situation.
Not every regulation is therefore a memory.
Not every memory necessarily constitutes a learning.
And not every adaptation presupposes a conscious representation of what has been lived through.
From perturbation to signal, with no necessary trajectory
The continuity between physics, biology and cognition must not be presented as an inevitable progression.
Most physical gradients do not produce life.
Not all biological perturbations become signals.
Not all signals leave a durable trace.
Not all traces modify future responses.
It is therefore preferable to speak of a conditional widening of possibilities.
In a simple physical system:
Gradient → flow → dissipation
In certain living organisations:
Constraint → detection → signalling → mobilisation → regulation
When mechanisms of memory are present:
Perturbation → response → persistent trace → possible modification of a future response
In cognitive and social systems:
Perturbation → interpretation → decision → mobilisation → reorganisation → collective learning
These sequences describe neither a universal law nor a necessary march towards greater complexity. They indicate that certain organisations add new capacities for processing a perturbation.
A gradient can feed a flow.
In certain living structures, a perturbation can be converted into a signal.
Some signals can leave traces.
Some traces can modify future responses.
The continuity lies in material transformations. The difference lies in the capacities of organisation that become possible at each level.
Living systems as a permanent trade-off
The energy available is never unlimited.
When an organism meets a constraint, it must redistribute its resources among several functions that may compete.
It must maintain essential functions, respond to danger, repair damage, preserve its reserves, support its growth, ensure its reproduction and maintain certain capacities of exploration or learning.
The distribution depends on the nature of the constraint, its intensity, its duration and the earlier state of the system.
An intense or prolonged pressure can concentrate resources on immediate protection. Some functions are then slowed, deferred or reduced. Short-term maintenance can come at the price of a decrease in future capacities for growth, reproduction, repair or exploration.
It would nevertheless be too simple to claim that biological innovation is a mere energetic residue.
An energy margin can make certain forms of plasticity or exploration possible, but it is not enough to produce them.
Adaptation also depends on the diversity of responses available, on the architecture that makes their exploration possible, on the mechanisms of variation and on the processes able to stabilise certain transformations.
Energy funds the possibilities. Organisation determines which of them can be explored. Memory allows some transformations to become durable.
This logic can be brought closer, with caution, to certain social dynamics.
When a society goes through a deep crisis, a growing share of its resources may be devoted to supply, security or the maintenance of essential infrastructures. Long-term investment and experimentation may then decrease.
The comparison does not mean that a society is an organism. It reveals a constraint common to systems with limited resources: when immediate maintenance absorbs almost all available capacities, the space of exploration shrinks.
This analogy nevertheless has a limit that is not merely one of degree. In a cell, resources, functions and viability have a relatively determinate physiological definition. In a society, what we call a resource, a function, viability or adaptation also depends on conflicts of interest, relations of power, political decisions and the way costs are distributed.
A society can remain materially functional while severely degrading the situation of part of its population. Social viability therefore cannot be deduced from systemic efficiency alone, and the analogy must remain strictly limited.
Can constraint increase future capacities?
Not every constraint is necessarily destructive.
Some limited perturbations can activate mechanisms of protection, repair or compensation. The term hormesis designates responses in which a low dose or a moderate intensity produces an effect different from, and sometimes opposite to, the one observed at high dose.
But hormesis does not mean that every difficulty makes one stronger.
The effect depends on the type of constraint, its duration, its frequency, the state of the system and the time available for recovery.
A constraint that is tolerable for one organism can be excessive for another.
A stress that is useful in the short term can become damaging when it is prolonged.
An improvement in one function can have a cost in another.
Constraint is therefore not creative in itself. It can become fruitful when it activates mechanisms of response without exhausting the resources required for them to work.
We then find a possible relation between intensity and adaptation:
Insufficient constraint → weak mobilisation
Constraint compatible with the capacities available → adaptation possible
Excessive or prolonged constraint → saturation, loss of function or degradation
This zone is neither fixed nor universal. It depends on the system, on its history and on its margins.
Five dimensions for analysing viability under constraint
Modularity, redundancy, the fruitful zone of constraint, tempo and margin do not constitute a general law of living systems.
They are neither universally necessary nor sufficient to ensure viability.
They form a heuristic grid intended to organise observation and to produce conditional hypotheses. Their influence can be assessed only by specifying the system studied, the function one seeks to preserve, the nature of the perturbation, its intensity and the temporal scale considered.
Each of these dimensions can then be associated with a proposition liable to be examined empirically, without claiming that it applies identically to all living systems.
Modularity
Living systems are organised into cells, tissues, organs and partly differentiated functional networks.
This organisation can allow certain responses to stay local. A perturbation can be contained or compensated without immediately imposing a global reorganisation.
Modularity therefore protects space by limiting the propagation of damage.
It can also protect time, since a local response can begin before the whole system is affected.
At comparable function and perturbation, a more modular organisation should, under certain conditions, further limit the propagation of effects or allow more independent local adjustments.
That advantage is nevertheless not guaranteed.
Modules that are too isolated may coordinate their responses badly, duplicate certain functions or slow the circulation of important information.
Nor is modularity the only route to robustness: some strongly distributed architectures remain robust without being clearly modular.
Redundancy
Several mechanisms can contribute to maintaining a similar function.
Alternative metabolic pathways, compensatory capacities or functional reserves can preserve continuity when an element becomes less effective.
But a redundancy is protective only if the alternative pathways do not all share the same critical point.
Several mechanisms depending on the same resource, the same organ or the same central regulator can give the appearance of security while remaining vulnerable to a common failure.
At comparable perturbation, several functionally distinct and sufficiently independent pathways should increase the probability of maintaining a function after one of them fails.
That effect decreases when the pathways share the same critical dependency.
Redundancy also has an energetic and material cost: maintaining surplus pathways mobilises resources that are no longer available for other functions.
The fruitful zone of constraint
For certain constraints and certain systems, the relation between intensity and adaptation can be non-linear.
An intermediate range can activate mechanisms of protection or reorganisation, while a stronger intensity can produce saturation or loss of function.
For a given system and constraint, the adaptive response could then be non-monotonic: an intermediate intensity would produce more adaptation than the absence of constraint or than an excessive constraint.
That prediction nevertheless holds only locally. The existence, the shape and the width of the window must be established in each context, and no universal level of beneficial constraint can be deduced from it.
Tempo
A response must occur before the damage irreversibly reduces the capacities of the system.
In a simplified form:
The time for detection, mobilisation, regulation and repair must remain shorter than the time the perturbation needs to cause an irreversible degradation.
This relation is not a predictive equation.
The stages are not always strictly successive. Some can unfold in parallel. Their duration also depends on the scale observed.
The formula is an instrument of diagnosis.
The time of irreversible degradation does not, moreover, always correspond to a single threshold. It can take the form of a succession of partial losses, with cumulative damage, incomplete recoveries and different possibilities of restoration depending on the functions considered.
A system may detect too late.
It may detect quickly but mobilise its resources insufficiently.
It may temporarily regulate a perturbation without repairing the damage.
It may, finally, have effective mechanisms that become active after the threshold of reversibility has already been crossed.
When the time required for the response approaches or exceeds the time of irreversible degradation, the probability of functional loss should increase.
Margin
Energy reserves, plasticity, alternative pathways, latent capacities and periods of recovery constitute margins of reorganisation.
Glycogen and lipid reserves can temporarily support energy needs.
Compensatory capacities can maintain a function.
Sleep, rest and repair mechanisms help restore mobilised resources.
But margin is not reducible to a stock.
It also includes the possibilities that remain open and the time available before a perturbation becomes irreversible.
At comparable perturbation, accessible resources, latent capacities and substitution pathways should extend the period during which a reorganisation remains possible.
That advantage disappears when the reserves are inaccessible, impossible to convert into action or dependent on the same critical point as the threatened function.
Margin is the energetic, structural and temporal space in which another organisation still remains possible.
Living systems articulate several timescales
Biological tempo is not reducible to a race between a response and a degradation.
Living systems articulate several temporalities.
A rapid perturbation can trigger a rapid response, and then leave a modification that persists far longer.
A wound can cause coagulation within minutes, be repaired over the following days and leave a durable scar.
An infection can trigger an immediate response and then modify certain immune capacities for a long time.
A brief neuronal experience can produce transformations that consolidate progressively and influence future behaviour.
Two complementary dynamics can then be distinguished.
The fast loop protects immediate viability:
Detection → mobilisation → regulation → limitation of damage
The slow loop transforms part of the experience:
Trace → consolidation → structural modification → different future response
The first can help preserve immediate viability.
The second can durably inscribe certain effects of experience and thereby modify future responses.
The fast response helps protect continuity. The slow transformation can inscribe part of the experience over time.
Long-term adaptation thus depends on a coupling between several timescales.
A brief event can become a durable structure.
A one-off response can modify a future capacity.
A present perturbation can transform the space of responses to come.
Changing, then modifying the conditions of change
An adaptive system does not always modify only its state or its response.
Some experiences can transform the conditions under which new modifications will become possible.
In neuroscience, metaplasticity designates the durable influence of earlier activity on the capacity of a synapse or a network to express later plasticity.
Past experience therefore does not necessarily modify the present response directly. It can shift the thresholds of induction, modify the amplitude or direction of a future change and make certain forms of plasticity more or less accessible.
Plasticity modifies a response or a structure. Metaplasticity modifies the conditions under which new transformations will be able to take place.
It must nevertheless be avoided to extend this concept directly to all living systems.
Metaplasticity has a precise meaning in neuroscience.
To designate a more general property, the expression second-order plasticity can be used as a conceptual hypothesis, provided it is reserved for situations in which the history of the system genuinely modifies its rules of transformation.
In the conceptual grid proposed here, we shall distinguish a first-order plasticity, which modifies the state or the response of the system, and a second-order plasticity, which modifies the conditions of its later transformations. This pair is an analytical instrument, not a universally stabilised biological taxonomy: it does not everywhere have the same status as synaptic metaplasticity, whose meaning is precise.
A durable modification of the response to a constraint then belongs to first-order plasticity.
A modification of the capacity to produce new adaptations, of their amplitude, of their direction or of the conditions of their appearance belongs to a higher level.
Memory preserves a trace. Plasticity transforms a response. Second-order plasticity modifies the space of transformations that will remain accessible in the future.
Three complementary angles on adaptation
The distinctions between regulation, memory and learning, between fast loop and slow loop, and then between first- and second-order plasticity do not constitute three competing hierarchies.
The first two describe a single dynamic under two aspects: the function of the modification and its temporal inscription. The third does not merely change viewpoint. It adds a level, since it distinguishes what is transformed: the response of the system, or the very conditions of its future transformations.
The distinction between regulation, memory and learning bears on the function of the modification.
The distinction between fast loop and slow loop describes its temporal inscription.
The distinction between first-order and second-order plasticity separates a modification of the response from a modification of the rules by which that response will still be able to change.
These categories can overlap without being equivalent.
A slow loop is not necessarily a learning.
A memory can persist without modifying the future rules of plasticity.
A learning can mobilise several timescales.
A second-order plasticity can modify the capacity to learn without preserving the detailed content of the event that triggered it.
These distinctions therefore do not seek to multiply categories. They make it possible to specify what changes, for how long and with what consequences for future transformations.
From the cell to the biosphere
The same principles take different forms depending on the scale observed.
Photosynthesis converts part of the energy carried by solar radiation into chemical energy. That energy then contributes to carbon fixation and to the production of organic matter.
Cellular respiration transforms the chemical energy of nutrients into electrochemical gradients, and then into forms usable by the cell.
Immunity adds mechanisms of recognition, of regulation and, in some cases, of memory.
Organisms also modify their environment.
Over the history of the Earth, living systems have profoundly influenced the cycles of carbon, nitrogen and oxygen. Interactions between organisms, atmosphere, oceans and rocks take part in the global functioning of the Earth system.
The Gaia hypothesis proposed that certain feedbacks issuing from living systems might help maintain conditions compatible with habitability.
The existence of feedbacks between life and environment must nevertheless be distinguished from the stronger idea of a biosphere functioning as a unified organism.
Living systems do effectively modify their environment.
Some interactions can produce stabilising effects.
Others can amplify changes or degrade the conditions on which certain species depend.
No global intention should be attributed to the biosphere.
An open hypothesis: maximum entropy production
Some theories propose that systems far from equilibrium tend, under certain conditions, towards regimes characterised by a high production of entropy.
The principle of maximum entropy production, often abbreviated MEP, has been used in the study of climate, ecosystems and the energy exchanges of the Earth system.
This hypothesis remains stimulating, but its general status is still debated.
Formulations differ, the domains of application are not always clearly delimited and quantitative validations in ecology remain limited.
Even when a biological or ecological system seems to display an increased dissipation, that does not demonstrate that entropy production is the variable directly selected.
Natural selection acts on differences of survival and reproduction in particular environments.
The regimes of dissipation observed may be consequences, constraints or emergent properties of biological organisation without necessarily constituting its objective or its local criterion of optimisation.
Dissipation is an unavoidable material condition of the activity of living systems.
Nothing allows us to make it, in general, the purpose of evolution.
Evolution aims at neither complexity nor consciousness
Thermodynamics defines constraints and makes certain transformations possible.
It does not on its own determine the trajectories of evolution.
The history of living systems does not follow an obligatory progression from simple organisms towards complex ones, and then towards cognition and consciousness.
Evolution can produce increases in complexity, but also specialisations, losses of function and simplifications.
A more complex structure is not necessarily better adapted.
Its value depends on the conditions under which it functions, on the resources it requires and on the possibilities it opens.
Consciousness is therefore not a destination contained in the second law.
It is one of the properties that emerged in certain lineages in the course of a history made of variations, constraints, transmissions and selections.
Thermodynamics makes it materially possible to sustain the structures required for memory and cognition.
It is not enough to explain their content, their function or their history.
From dissipation to anticipation
The most interesting trajectory is perhaps not that of an ever-increasing complexity.
It is that of a widening of the ways of responding to constraint.
A physical system can be transformed by a gradient.
An organism can detect certain perturbations and regulate certain variables.
A system endowed with memory can preserve a modification after the stimulus has disappeared.
A system capable of learning can modify its future responses.
Some forms of regulation even become anticipatory. Allostasis notably describes adjustments that prepare the organism for expected needs rather than merely correcting a deviation that is already present.
This anticipation requires no conscious cognition. It can emerge at the physiological level, from inherited or learned regularities, without an explicit representation of the future being formed. Anticipation is therefore not reserved for cognitive systems: it already appears in the regulations of the body, before any representation.
With cognition, a system can produce representations and compare several possibilities.
With societies, memory can be externalised into narratives, techniques, institutions, archives and models.
Each level remains subject to the physical constraints of the previous levels.
But some levels also add new ways of detecting, preserving, interpreting and transmitting the effects of experience.
This continuity is not a necessary progression.
It describes a succession of conditional possibilities.
A gradient can feed a flow. In certain living organisations, a perturbation can become a signal. Some signals leave traces. Some traces modify future responses. And some organisations become able to anticipate perturbations that are not yet fully present.
Conclusion
The matter that keeps certain possibilities open
Living systems do not defy the arrow of time.
They maintain themselves through irreversible processes.
They capture gradients of free energy, channel their transformations and dissipate part of that energy into their environment. They use these flows to renew their components, preserve differences, repair damage and maintain the relations on which their viability depends.
But their singularity does not lie only in the complexity of that dissipation.
It appears when certain structures become able to detect perturbations, regulate their exchanges, preserve modifications and transform their future responses.
Regulation helps preserve immediate viability.
Memory preserves part of the experience.
Learning uses that trace to modify a future response.
Second-order plasticity can modify the very conditions under which new adaptations will become possible.
Homeodynamics is not an additional level in this succession. It designates the wider dynamic within which these processes are articulated: the continuity of a living system that renews its components, adjusts its regulations and lets certain experiences transform its future possibilities.
Living systems therefore do not always merely respond faster than degradation.
They can sometimes translate a rapid perturbation into a durable modification and move part of the experience from one timescale to another.
Across dissipative physical systems, living organisms and then cognitive and social systems, certain organisational dynamics reappear. Constraints modify flows, call on the resources available and can impose a reorganisation.
But the mechanisms are not identical.
A physical system is transformed.
An organism can regulate.
A memory can preserve.
A learning can modify a future response.
A cognitive or social organisation can interpret, anticipate and transmit.
When capacities of regulation, memory and learning are present, adaptation can be favoured by the existence of accessible resources, a diversity of responses, alternative pathways and enough time to turn the perturbation into a new form of viability rather than into irreversible degradation.
Thermodynamics therefore contains neither life, nor intelligence, nor consciousness as results already written.
It defines a world of gradients, transformations and irreversibility in which certain organisations have acquired the capacity to maintain their conditions of existence and to let their history modify their future possibilities.
Energy makes transformations possible. Organisation determines which of them can be explored. Regulation preserves immediate viability. Memory inscribes part of the experience over time. And time decides whether reorganisation can still outrun degradation.
Bibliographical landmarks
This essay is conceived as a position paper: the argument governs the organisation, not the reverse. The references below therefore do not function as individual footnote calls, but as entry points to the work that grounds the notions mobilised. They are grouped by theme and can be consulted independently.
Non-equilibrium thermodynamics and dissipative structures
Schrödinger, E. (1944). What is Life? Cambridge University Press.
Nicolis, G. & Prigogine, I. (1977). Self-Organization in Nonequilibrium Systems. Wiley. (Prigogine, Nobel Prize in Chemistry, 1977.)
Seifert, U. (2012). Stochastic thermodynamics, fluctuation theorems and molecular machines. Reports on Progress in Physics, 75, 126001.
Molecular motors and the rectification of thermal noise
Astumian, R. D. (1997). Thermodynamics and kinetics of a Brownian motor. Science, 276, 917–922.
Jülicher, F., Ajdari, A. & Prost, J. (1997). Modeling molecular motors. Reviews of Modern Physics, 69, 1269–1282.
Noji, H., Yasuda, R., Yoshida, M. & Kinosita, K. (1997). Direct observation of the rotation of F1-ATPase. Nature, 386, 299–302.
Information, erasure and thermodynamic cost
Landauer, R. (1961). Irreversibility and heat generation in the computing process. IBM Journal of Research and Development, 5, 183–191.
Bérut, A., Arakelyan, A., Petrosyan, A., Ciliberto, S., Dillenschneider, R. & Lutz, E. (2012). Experimental verification of Landauer’s principle linking information and thermodynamics. Nature, 483, 187–189.
Constraint, hormesis and anticipatory regulation
Ristow, M. & Schmeisser, K. (2014). Mitohormesis: promoting health and lifespan by increased levels of reactive oxygen species (ROS). Dose-Response, 12(2), 288–341.
Sterling, P. & Eyer, J. (1988). Allostasis: a new paradigm to explain arousal pathology. In S. Fisher & J. Reason (eds.), Handbook of Life Stress, Cognition and Health. Wiley.
Sterling, P. (2012). Allostasis: a model of predictive regulation. Physiology & Behavior, 106(1), 5–15.
Plasticity and metaplasticity
Abraham, W. C. & Bear, M. F. (1996). Metaplasticity: the plasticity of synaptic plasticity. Trends in Neurosciences, 19(4), 126–130.
Homeostasis, regulation and biological memory
Rattan, S. I. S. (2006). Homeostasis, homeodynamics, and aging. In J. E. Birren (ed.), Encyclopedia of Gerontology (2nd ed.). Elsevier.
Cannon, W. B. (1929). Organization for physiological homeostasis. Physiological Reviews, 9(3), 399–431.
Ahmed, R. & Gray, D. (1996). Immunological memory and protective immunity: understanding their relation. Science, 272(5258), 54–60.
Modularity, robustness and functional redundancy
Hartwell, L. H., Hopfield, J. J., Leibler, S. & Murray, A. W. (1999). From molecular to modular cell biology. Nature, 402, C47–C52.
Kitano, H. (2004). Biological robustness. Nature Reviews Genetics, 5(11), 826–837.
Edelman, G. M. & Gally, J. A. (2001). Degeneracy and complexity in biological systems. Proceedings of the National Academy of Sciences, 98(24), 13763–13768.
Energetic trade-offs, resilience and niche construction
Stearns, S. C. (1992). The Evolution of Life Histories. Oxford University Press.
Holling, C. S. (1973). Resilience and stability of ecological systems. Annual Review of Ecology and Systematics, 4, 1–23.
Odling-Smee, F. J., Laland, K. N. & Feldman, M. W. (2003). Niche Construction: The Neglected Process in Evolution. Princeton University Press.
Critical transitions, thresholds and hysteresis
Scheffer, M., Carpenter, S., Foley, J. A., Folke, C. & Walker, B. (2001). Catastrophic shifts in ecosystems. Nature, 413, 591–596.
Scheffer, M., Bascompte, J., Brock, W. A., Brovkin, V., Carpenter, S. R., Dakos, V., Held, H., van Nes, E. H., Rietkerk, M. & Sugihara, G. (2009). Early-warning signals for critical transitions. Nature, 461, 53–59.
Planetary scale: feedbacks from living systems and entropy production
Lenton, T. M. (1998). Gaia and natural selection. Nature, 394, 439–447.
Martyushev, L. M. & Seleznev, V. D. (2006). Maximum entropy production principle in physics, chemistry and biology. Physics Reports, 426(1), 1–45.
Kleidon, A., Malhi, Y. & Cox, P. M. (2010). Maximum entropy production in environmental and ecological systems. Philosophical Transactions of the Royal Society B, 365(1545), 1297–1302.