WorksComputer Science
Causal Information Exchange as a Necessary Condition for Adaptive Regulation: A Perturbational Framework for Life and Agency
Abstract
Information exchange can support living organisation without providing a sufficient criterion for life. We develop a perturbational framework that distinguishes causal transmission, directional imbalance, internal retention, response structure and functional regulation. Randomised interventions define Shannon information channels; for deterministic binary recipients, their context-conditioned mutual information is exactly a paired perturbation-response indicator. Spatial and temporal averages of these local channels measure information-exposure density, rather than joint information or conserved information flux. Complementary Coding Theorem Method and Block Decomposition Method (BDM) estimates describe response and trajectory structure. A same-shaped zero-mask reference separates raw BDM from a structural contrast; neither is a channel mutual information. Retained exploratory BDM plots are distinguished from the normalised causal-channel results. Retention-weighted coupling and directionality obey algebraic bounds, and ensemble reciprocity need not imply reciprocity within individual contexts. For a specified binary sensor–internal-channel–actuator task, we derive a positive information lower bound from successful regulation and construct controllers with identical information profiles but opposite functional outcomes. A study of 2,000 ordered elementary-cellular-automaton rule pairs, six detailed examples, recurrence and lossless coding analyses, and a signed-network extension demonstrates dissociations among perturbational responsiveness, structural regularity and recurrence. The results establish causal information as a necessary resource for the stated form of adaptive regulation, while showing why neither causal coupling nor response complexity alone measures agency or life.
Introduction
An informational account of life must explain more than the presence of complex patterns. A system can display statistical dependence, persistent dynamics and sensitivity to perturbations without maintaining itself, regulating a relevant variable or acting beneficially on its surroundings. Conversely, effective regulation can rely on simple and highly compressible mechanisms. The relevant distinction is therefore not simply between low and high information content, but between information that is present, information that can be transmitted through causal mechanisms, and information whose use contributes to a specified function.
A natural starting point is Schrödinger’s negentropy argument that living systems maintain organised states despite the tendency toward disorder [27], together with Prigogine’s account of dissipative structures maintained far from equilibrium [26]. Shannon provided the mathematical theory of communication [28], while Eigen connected information, replication and molecular evolution [21]. Statistical-physics approaches subsequently related information flow, replication and thermodynamic constraints [22, 25, 46, 19]. Causal-entropic work has also explored how access to future states can generate apparently adaptive behaviour [30].
The artificial-life tradition provides a complementary computational perspective. Barricelli’s early numerical evolutionary systems [18], Conway’s Game of Life [20] and Langton’s formulation of artificial life [23] established that life-like organisation can be investigated in fully specified computational worlds. Algorithmic Information Theory supplies a further language for generative structure through Solomonoff probability and Kolmogorov complexity [29, 24]. These lines of work motivate asking not merely whether a pattern is complex, but whether its organisation has identifiable causal consequences.
Thermodynamic and algorithmic approaches have related living organisation to the exploitation of environmental structure rather than to indiscriminate maximisation of randomness [1]. Information-theoretic accounts of individuality ask how a system maintains predictive autonomy [2], while semantic-information approaches assess which correlations are causally necessary for continued viability [3]. These perspectives motivate an important separation: a causal communication channel is a structural resource, whereas its contribution to survival, regulation or agency is a further question.
A useful counterexample to attempts to define life directly from a scalar complexity measure is provided by Assembly Theory. Assembly Theory has been proposed as a way to identify biosignatures, selection and life-like organisation from the complexity and multiplicity of assembled objects. However, subsequent analyses have shown that its central quantities largely reduce to established statistical compression principles, do not uniquely distinguish biological from abiotic structure, and do not by themselves establish selection, causality or viability [15, 16, 17]. This illustrates the central distinction pursued here: structural complexity, copy number or compressibility may characterise an object, but life requires additional evidence about the causal organisation that maintains, regulates and couples that object to its environment.
The perturbational programme developed in earlier work provides a direct methodological foundation for this distinction. Early computational implementations used compressibility to characterise cellular-automaton dynamics [32] and later examined their sensitivity to controlled changes in initial conditions [33]. Interacting cellular automata were subsequently studied under global rules, providing an early framework for analysing how distinct dynamical subsystems influence one another [9]. This perspective was developed further in studies of programmability, persistent and colliding structures in Conway’s Game of Life, and systematic Algorithmic Information Dynamics analyses of one- and two-dimensional cellular automata [37, 38, 39].
A particularly important conceptual development is the mutual-perturbation formulation of systems and observers introduced by Abrahão and Zenil [40]. In that framework, observation is itself an interaction between dynamical systems, and the appearance of emergent algorithmic information can depend on the observer’s formal knowledge while becoming robust under broader classes of description. The present work adopts this perspective by treating the candidate agent and environment initially as observer-specified interacting subsystems. The basic object is therefore not an isolated agent but a pair of directed perturbational relations,
whose asymmetry, reciprocity, persistence and functional consequences can be measured.
Following this mutual-perturbation framework, we treat the agent–environment partition as part of the observational description and identify robustness across plausible boundaries as an important future criterion for causal individuality.
Closely related in conceptual and methodological scope are recent algorithmic approaches to persistence, regulation and agency developed by Ruffini. His recent Game of Life study examines persistence, regulation and the disappearance of organised patterns, while his algorithmic-regulator framework describes regulation in terms of algorithmic complexity and model-based interaction between a regulator and its environment [45, 44, 43]. These developments are closely aligned with the broader AIT/AID programme, but the emphasis of the present study is different: perturbational causal information exchange across an explicit agent–environment boundary is made the primary object of analysis, and the same intervention logic is evaluated using both classical and algorithmic information-theoretic tools.
There is also a methodological distinction between prediction and intervention. Mutual information is symmetric and can be positive because of common input. Transfer entropy conditions on the recipient’s history and quantifies additional predictive information [4], but predictive transfer and causal effect are not generally interchangeable [5]. Intervention-based information flow instead asks how recipient distributions change when a source is externally varied [6]. Empowerment addresses the capacity of an agent’s actuation channel and therefore concerns possible control, not necessarily beneficial realised behaviour [7]. We build on these distinctions rather than identifying every statistically dependent subsystem as an agent.
Algorithmic Information Dynamics combines perturbations with the analysis of candidate generative mechanisms and their information content [8, 10]. AID is intervention-centred rather than committed to a single numerical estimator. We distinguish two uses of perturbational data. The first uses classical Shannon information to quantify how reliably an intervention is transmitted across a boundary. The second uses algorithmic-information approximations: the Coding Theorem Method (CTM) estimates the algorithmic complexity of short blocks from algorithmic probability [31], and the Block Decomposition Method (BDM) combines those local estimates to evaluate larger objects [14]. The two approaches answer different questions. Shannon information measures an intervention channel, while CTM/BDM characterise descriptional structure under a specified representation. Raw BDM of an all-zero response mask is generally positive, so a matched zero-mask reference is required for the structural contrast; causal presence is checked separately from the mask itself. The normalised channel experiments, the network mask descriptors and the retained exploratory ECA panels are identified separately. Agreement between estimators is not assumed from their shared use of perturbations.
The present study extends perturbational cellular-automaton analysis to an explicit agent–environment partition. We ask how interventions separate outward, inward and internally retained responses; what follows from combining them; and what makes those quantities relevant to adaptive regulation. Hybrid elementary cellular automata provide known local mechanisms [11], while a separate binary regulatory task supplies an explicit success criterion that visual complexity cannot.
Our claim is conditional: successful challenge-dependent regulation needs informative pathways when challenges are unpredictable, require different responses and have no alternative causal route to the output. It does not require positive outward dominance or a high instantaneous score in every living state. The contribution is a protocol for measuring causal organisation and structure, and identifying the assumptions needed for functional interpretation.
There are three levels of inference. The channel identities and the task-specific necessity result are mathematical statements. The numerical experiments characterise responsiveness, structural regularity and recurrence under stated protocols. Organisational persistence, selective reciprocity and viability are proposed extensions that require additional observations. In particular, a positive reciprocal score constructed from ensemble means does not establish reciprocal interaction within the same context.
An intervention-based causal information profile
System, boundary and intervention ensemble
Following the mutual-perturbation perspective of Abrahão and Zenil [40], we regard the agent and environment not as intrinsically privileged ontological categories, but initially as two observer-specified interacting dynamical subsystems. Their interaction can therefore be represented as a pair of directed perturbational relations,
The quantities introduced below operationalise these two directions. Agency is consequently not identified with interaction itself, but with properties of the asymmetry, persistence, selectivity and functional organisation of this mutual perturbational relation. This framing also connects directly to earlier open-ended-evolution work in which a finite subsystem can exhibit innovation only relative to counterfactual isolated dynamics and interaction with an environment [35, 36].
Let a dynamical system have state , where is a designated focal subsystem and its complement. The dynamics are
The variable represents exogenous noise when present. The boundary, observation scale and intervention horizon are parts of the experimental specification. They are not inferred merely by calling an agent.
A context contains a pre-intervention state, the update mechanism and the experimental settings. For a binary source cell , draw independently of the context and intervene at time by setting
The two arms are an unchanged state () and a single-bit flip (). For a recipient observable , the experimental distribution is
Define the context-specific channel score
All logarithms are base two. Randomisation supplies the causal interpretation: differences between the two response laws arise from the specified intervention, subject to consistency of the simulator and isolation of other experimental changes. The score measures the distinguishability of experimentally induced response distributions, not an observational association between and . Equivalently, it is the Jensen-Shannon divergence between the two intervention-arm response laws: , where . It vanishes exactly when those two laws coincide. Future unobserved noise is marginalised in (3), not silently included in the conditioning context. This distinction matters: for with independent fair noise, , even though paired outputs always differ if the same realised noise is held fixed. Consequently, the deterministic estimator introduced next must not be transferred unchanged to stochastic systems.
Exact relation to deterministic perturbation responses
Proposition 1 (binary deterministic channel identity). For a fixed context and a deterministic binary recipient ,
Proof. Determinism gives . If both intervention arms produce the same value, . If they produce different binary values, the fair input makes the output equiprobable and bit. Subtracting the conditional entropy proves the result.
Thus, after fixing a context, no empirical probability table or finite-sample mutual-information estimator is needed for a binary deterministic ECA. We evaluate both intervention arms exactly. Sampling uncertainty arises from the distribution of contexts, not from estimating the two-point channel within a context.
For a cell at lag , write
Equation (5) makes this damage indicator an exact local conditional-information measurement. The relation to damage-spreading analyses is substantive: trajectory separation has long been used to study the sensitivity of automata [12]. We do not claim that damage spreading itself is new. The present construction specifies its causal experiment, recipient normalisation and separate self-retention control.
A collection of affected cells does not contain that many independent transmitted bits. For the complete deterministic response vector under a single binary intervention, is still at most one bit. Local scores count the spatial and temporal availability of the same perturbation information. Their average is therefore called an information-exposure density, not a joint mutual information, a conserved information flux or an entropy-production rate.
Outward transmission, inward transmission and internal retention
For a source and recipient region , define the geometrically eligible site-time set
where is periodic lattice distance. The last condition is the radius-one ECA light-cone bound. It depends on geometry, not on whether a particular rule actually transmits a perturbation. With a specified distribution over source cells, define
measures outward exposure and inward exposure. Each source receives equal weight; division occurs before averaging over sources. An empty eligible set is assigned zero and must be reported as an unreachable probe, not interpreted as evidence of biological inactivity. The primary experiment has no such empty cross-boundary sets.
To distinguish internal retention from an effect that leaves the subsystem and returns through the environment, run an additional intervention arm. At every update, overwrite all environment cells with their unperturbed reference values, while allowing to evolve from its perturbed state. Denote its damage indicator by . Using the late-half set
we define
This is internally mediated retention of the tested perturbation, not self-maintenance, robustness or the ability to repair damage. A corrective mechanism can intentionally erase a single-bit perturbation. Such erasure is not evidence that the mechanism lacks agency. The late-half convention tests retention over a stated finite timescale; it is not a universal timescale of life. We retain the corresponding open-environment score, , as a diagnostic of the clamping assumption.
Each local information value is normalised by the one-bit intervention entropy before aggregation; its numerical value is the indicator in (5). Thus , , and their aggregate scores are dimensionless, with . Opportunity normalisation prevents the direct counting of unreachable recipient cells, but it does not remove dependence on the boundary, source distribution, accessible states or observation horizon.
Write , , for the source-averaged primitive quantities within context , so that , and . This specifies the aggregation level before any nonlinear product is formed. The clamped experiment preserves the unperturbed environmental drive; measures internally mediated retention under that drive, not autonomy without external input.
Coupling, directionality and reciprocal coupling
Define three derived scores:
is retention-weighted coupling, signed directional influence, and an ensemble reciprocal-coupling summary. These are descriptive products of the primitive means, not established measures of life or adaptive agency. In particular, compares the two cross-boundary directions; it does not compare agent influence with the environment’s influence on itself. No term occurs.
The dimensionless directionality ratio is
When , is undefined and is not assigned a misleading zero. If but , remains a description of unweighted cross-boundary asymmetry, while . The ratio equals only when .
Positive or indicates outward dominance under the selected protocol. When , zero indicates ensemble directional balance, not absence of influence. Zero similarly indicates equality of the two positive cross-boundary means. If , however, regardless of their imbalance. Negative values indicate inward dominance, not absence of agency. holds exactly when the three primitive ensemble means are positive. It does not imply that they are positive in the same context. Similarly, alone does not establish bidirectional interaction.
For example, an equal mixture of contexts with
has , although every context has zero reciprocal score. The distinct quantities
would respectively describe mean within-context reciprocity and its prevalence. Both vanish in this example. These definitions clarify the inference; no unreported values of them are claimed for the simulations. The numerical tables retain , computed after averaging primitive scores. Even concerns separately probed channels within a finite context and window, not proof of an operating feedback loop, simultaneous bidirectional action or beneficial function.
Figure 1. A paired perturbation response for rule 30 in a rule-18 environment. Black cells differ from the unperturbed trajectory after a single flip at position 104. Dashed vertical lines delimit the ten-cell focal region; time runs downwards. This is the first stored initial condition after 64 updates, not a selected high-scoring example. The plotted 64-step response illustrates localisation; the primary score uses a 32-step horizon.
Algorithmic descriptions and the zero-mask reference
For a finite object , Kolmogorov complexity is uncomputable. The algorithmic coding theorem motivates CTM estimates for short objects from output-frequency distributions of finite computational experiments [31]. BDM combines local CTM estimates with logarithmic multiplicity terms [14]. Write
where ranges over distinct blocks and specifies the CTM table, block dimensions, partition alignment, remainder convention and implementation. Fixed-block BDM approximates aspects of local descriptional structure; it does not provide an exact shortest program for the full object.
For paired recipient arrays with identical dimensions, define
The superscript raw identifies a computable descriptor rather than exact Kolmogorov complexity. A null response has , but its raw BDM is generally positive. The same-shaped zero-mask reference is
With the same estimator settings and array convention, define
This signed structural contrast vanishes for a null response. The converse is not assumed: different masks can share a BDM value. The contrast is not guaranteed nonnegative for arbitrary settings and is not a divergence, mutual information or exact causal-effect magnitude.
Presence of a deterministic response is established separately by
A structural report can therefore distinguish response occurrence, exposure density, raw BDM, the matched zero reference and the signed contrast. Conditioning BDM summaries on would further characterise non-null responses. Neither that analysis nor new baseline-corrected values are asserted where they have not been calculated.
The baseline also matters for directional composites. With matched array shapes and fixed , source- and context-averaged outward and inward raw descriptors share a reference , giving
The sum retains an offset; the difference cancels it only when the references match. Unequal recipient sizes or different block conventions invalidate this cancellation. If references differ across samples, they must be subtracted at sample level before applying the specified weighting. Because the contrasts can be signed, they do not inherit the nonnegative-component bounds or square-root reciprocity formula of the Shannon scores.
Unperturbed layer complexity, response-mask complexity and the joint description of perturbed and reference trajectories are different objects. In particular, BDM() is not generally or algorithmic mutual information. A zero-array reference for an unperturbed layer is a regularity reference, not a no-intervention control. The retained ECA panels include layer descriptors and native-scale composites; their role is specified separately from the network mask analysis.
The common AID principle is controlled perturbation. A structural label such as coupling, life or agency provides neither null calibration nor functional validation.
Composition and collective interpretation
Combining subsystems requires fixed update rules, source weights, recipient sets and clamping. Exposure totals can obey accounting identities, but normalised densities and nonlinear products are not simply additive. Score cross-terms alone do not demonstrate emergent causation or collective function. The complete derivation and its assumptions are retained in Appendix § 9.
Mathematical consequences and a conditional necessity result
Algebraic geometry is not empirical validation
Proposition 2 (bounds and factorisation). For ,
where the ratio is defined, and
Proof. Non-negativity gives , hence . Also , which gives the sharper bound . Expanding proves (15). The ratio bound follows from the same non-negativity argument.
For , . These are dependent summaries of the same three components, not orthogonal empirical measurements. A scatterplot of against must lie within the bounds; observing that high absolute requires high therefore supplies no independent evidence that life precedes agency.
The stronger bound also exposes a conceptual problem with equating agency to imbalance. Maximal coupling, , forces and therefore . A perfectly reliable, balanced controller would receive the largest coupling score and zero directional imbalance. Conversely, implies that one cross-boundary direction is absent and . Large imbalance is not a general objective for living organisation.
What information is genuinely necessary for regulation?
A useful necessity claim must specify the task rather than define success by a high score. Consider a fair binary challenge independent of a background context . Let , and be binary sensor-memory, action and external response variables, with causal chain
Assume independent channel noise conditional on , no direct challenge-to-response bypass, and no unmeasured common cause between adjacent channel variables. The action is evaluated against the externally specified requirement . There is no challenge-independent response that exceeds one-half success.
Theorem 3 (task-conditioned causal-information necessity). Let . With the binary entropy function and ,
If adjacent-channel interventions use the task-induced input distributions, define the task-aligned scores , and . Information values are expressed relative to the same one-bit reference. They then obey
Proof. Conditional binary Fano bounds give . Since is fair and independent of , . Conditional data processing along (16) bounds this end-to-end information above by the information across each of the three cuts. The causal-chain assumptions make the adjacent interventional channels coincide with these conditional distributions when their input distributions match. Applying the three lower bounds to the two products yields (18). These steps use standard information inequalities [13].
This is a task-specific necessity result, not an algebraic restatement of (12). Success is defined independently, and the lower bound depends on measured error. It does not require ; the benchmark below has while regulating successfully.
The information quantities in (17) average over , so the resulting is an aggregate statement, not a guarantee for every context. For a particular whose task error , the same proof applies within that context because remains fair given . Overall success does not imply that every context meets that error condition.
Several conditions are essential. A direct environmental pathway to the response can invalidate the chain. Predictable challenges, pre-existing information or a single response that succeeds under every challenge remove the one-half open-loop bound. A different task requires a corresponding decision-theoretic bound. Most importantly, the task-aligned whole-channel quantities in (17) are not automatically equal to spatially averaged single-bit ECA scores. The latter are experimental diagnostics, not universal necessary conditions for every possible regulator.
Here measures information across the internal sensor-to-action channel; it does not itself measure temporal memory duration or organisational survival. Connecting the theorem to an ECA would require an explicit task implementation, task-induced intervention distributions and recipient representations that retain the relevant information.
An explicit failure of sufficiency
Proposition 4 (identical information, opposite function). There are controllers with identical but opposite performance on the same regulatory task.
Proof by construction. Let be a fair bit and introduce independent sensor, memory and actuator errors with probabilities . Define
where policy is corrective and inverts the action. In both policies,
The policy flip is bijective and cannot change these information quantities. However, with
task success is for the corrective policy and for the inverting policy. At zero noise, both have and , while success is respectively one and zero. No function of this information profile alone can distinguish their functional correctness.
Response-mask structure does not automatically repair this ambiguity. Complementing corresponding binary outputs in both intervention arms leaves their XOR mask unchanged:
Both its raw BDM and matched zero-mask contrast are therefore unchanged by that relabelling, although task correctness can reverse. This statement concerns matched paired masks; it is not an assertion that every possible algorithmic descriptor is invariant to every policy change.
The counterexample does not make causal information irrelevant. It identifies what is missing: a relation between the causal mechanism and an independently specified regulatory requirement. A biological extension must similarly test the effect of causal pathways on viability rather than rename their existence as life. This is consistent with, but not equivalent to, the viability-intervention approach to semantic information [3].
Boundary dependence and non-additivity
With matched source and recipient protocols on equal-sized complementary regions, exchanging the two regions exchanges and , so changes sign. The focal retention term generally changes too; need not simply negate when is recomputed for the new focal subsystem. Padding the lattice with cells outside the relevant causal ancestry cannot change a score, provided both the perturbation cones and their required initial-state ancestry are preserved. This is a locality property, not unrestricted size invariance.
Subagent scores are not generally additive. If two subagents act on the same recipient, adding exposure counts can double-count that recipient. Under a simultaneous intervention ensemble, let independent fair inputs produce . Then but . Joint interventions can therefore reveal synergistic influence absent from marginal channels. Single-source interventions that condition on the other input constitute a different experiment and correctly give a different answer.
Computational methods
Experimental design
The primary experiment uses a periodic 210-cell lattice: ten focal cells and 200 environmental cells. Each region follows its assigned ECA rule, with synchronous updates and actual cross-boundary neighbour states. For rule number ,
The scan samples 2,000 distinct ordered rule pairs uniformly without replacement, with 16 independent Bernoulli-one-half initial configurations per pair. Interventions occur after 64 updates, probing every focal cell and the five environmental cells adjoining each interface. The horizon is 32 updates; clamped retention uses lags 17–32.
Six detailed pairs, (30, 18), (250, 30), (30, 2), (30, 45), (110, 60), (45, 2), are followed from 128 initial configurations at ages 0, 16, 64, 128, 256 and 512. The same configurations are followed across ages. Homogeneous rules, sham and boundary-cut interventions, geometry changes, role exchange, locality-preserving padding and a scalar implementation provide controls. Full settings and validation procedures are in Appendix § 10; the study was not preregistered.
Structural and network measurements
Site-frequency entropy and level-nine raw DEFLATE describe fixed 64-update unperturbed ECA windows. Complete-state recurrence is sought through update 4,096; undetected recurrence is censored. Neither code length nor recurrence is a channel-information estimate.
The separate network experiment uses 32 binary nodes on a degree-four ring, split into equal regions, with symmetric edge weights in . Five positive-edge probabilities from zero to one each have 64 independent network–initial-state contexts. Threshold updates retain the previous state at ties. Boundary probes begin at age 16, with a 16-update horizon and clamped retention at lags 9–16. The full update and sampling protocol is in Appendix § 10.§ 5.
Network BDM describes full XOR response masks. Its reported is raw; the matched structural contrast would subtract , with . No numerical baseline-corrected values are claimed. Estimator settings and shape must match when using this reference. Exploratory ECA panels use different native encodings and are retained separately in Appendix § 13.
Regulation and uncertainty
The binary task uses sensor, internal-channel and actuator errors (0.10, 0.05, 0.10), plus seven equal-error settings from zero to one-half, with 100,000 challenges per policy and setting. Corrective and inverting policies share challenges and noise; an independent fair-action control succeeds at chance. Wilson intervals describe individual success probabilities.
Primitive context scores are averaged before forming , , , . Thus is an ensemble summary, not within-context reciprocity. Intervals use 5,000 bootstrap resamples of complete initial-condition units, or complete network–initial-state units; cells, lags and probes are not independent replicates. Temporal contrasts are paired across ages. Ordinary intervals are 95%; six designated endpoint contrasts use 99.167% marginal intervals for nominal Bonferroni familywise coverage of 95%, subject to bootstrap approximation. Sign counts and split-half ranks describe the sampled ensembles. Full statistical conventions, seeds and software details are in Appendices B and G.
Results
The rule-pair scan separates coupling from directional dominance
The 2,000-pair scan contains 1,625 positive-coupling pair means and 375 zero-coupling means. Positive signed directionality occurs in 725 pairs (36.25%), negative directionality in 900 (45.00%), and zero in 375 (18.75%). Median coupling is 0.00390 and mean coupling is 0.02261. There is therefore no predominance of positive directional influence in this protocol. Unequal compartment geometry and the chosen intervention ensemble remain part of the interpretation; the sign distribution is not evidence for a membrane-like mechanism.
All points satisfy and the factorisation to numerical precision. These observations verify the implementation of the definitions, not the biological meaning of the scores. Split-half Spearman correlations are 0.975 for across all pairs and 0.952 for across the 1,956 pairs with the ratio defined in both halves. The high repeatability supports stability of these ensemble summaries within the sampled protocol; it does not remove bias from an inappropriate boundary or task definition.
Detailed cases and controls
Rule 250 in rule 30 has outward exposure but zero inward exposure and late retention at age 64. Consequently , despite : a large directional ratio need not signify internally sustained organisation. Rule 30 in rule 2 instead has positive primitive means and aggregate , without establishing reciprocity within every context. Rule 30 in rule 45 and rule 110 in rule 60 are inward dominated.
Figure 2. Retention-weighted coupling and signed directional influence for 2,000 distinct ordered rule pairs. Each point is computed from the mean primitive scores over 16 independent initial conditions after 64 updates. Dashed lines show the algebraic bounds in the displayed range. Accumulation near the origin and the bounded wedge are not independent evidence for a life-agency hierarchy.
The controls reinforce this distinction. Constant rule 0 has zero scores; identity rule 204 retains the altered bit but has no boundary transmission; shift rule 170 transmits without late focal retention. All have . Additive rule 90 nevertheless has without a regulatory task. Complete six-case estimates, intervals, trajectories and homogeneous controls appear in Appendices § 11 and § 17.
Transient changes separate coupling from directionality
Between ages zero and 512, falls from 0.0749 to 0.0558 for rule 110/60 and from 0.0442 to 0.0147 for rule 45/2. Both decreases survive the multiplicity adjustment, while neither signed- change is resolved by its ordinary 95% interval. Rule 30/45 has no clearly resolved endpoint change in either score; rule 30/18 increases in coupling. There is no general monotonic age–coupling law. A negative directional score approaching zero increases in signed value while decreasing in magnitude. Full paired contrasts and additional trajectories are in Appendix § 11.§ 2.
Figure 3. Coupling at successive intervention ages for the three transient comparisons. Points are calculated from 128 initial-condition units and error bars are pointwise 95% bootstrap intervals. The horizontal scale is linear through age 16 and logarithmic thereafter. Different initial conditions are not substituted at later ages.
Structure, recurrence and causal responsiveness
For rule 30/2, focal site-frequency entropy remains near one bit, while the DEFLATE ratio falls from 0.6183 to 0.1396 between ages zero and 512. Coupling changes only from 0.0757 to 0.0727, with a paired interval spanning zero. Structural regularisation therefore need not remove causal responsiveness. Rule 110/60 shows the converse contrast: its code ratio changes little while coupling decreases.
Complete-state recurrence occurs in all tested contexts for four of the six detailed pairs. Rule 110/60 has median transient 44.5 and period 256, illustrating why short-window irregularity need not mean nonperiodicity. The full recurrence table and interpretation are in Appendix § 11.§ 3.
Geometry and intervention controls
Changing focal width can reverse directionality: rule 45/2 changes from at width six to +0.115 at width twenty. This changes the update-rule allocation as well as the boundary, so it tests architecture dependence, not solely observational repartitioning. Padding, matched role exchange and scalar checks support the stated invariances. Sham and boundary-cut controls give zero Shannon effects; a sham BDM mask instead has raw reference and zero structural contrast. Clamping may either increase or suppress a response. Full settings, estimates and the geometry figure are in Appendix § 11.§ 4.
Figure 4. Signed directional influence for the same initial-condition ensembles. The rule-110/rule-60 and rule-45/rule-2 coupling decreases in the preceding figure do not establish corresponding decreases in signed directionality. Error bars are pointwise 95% bootstrap intervals and do not constitute a test of the entire time-course shape.
Equal causal information can support corrective or harmful responses
| Policy | Exact success | Simulated success [95% interval] | |
| Corrective | 0.7579 | 0.7880 | 0.7880 [0.7855, 0.7905] |
| Inverting | 0.7579 | 0.2120 | 0.2120 [0.2095, 0.2145] |
| Open-loop | 0 | 0.5000 | 0.5009 [0.4978, 0.5040] |
Table 1. Regulatory benchmark at sensor, memory and actuator errors of 0.10, 0.05 and 0.10. Corrective and inverting policies have identical causal-information profiles. The open-loop action is independent of the memory, so its task-aligned retention channel and coupling are zero. Each row uses 100,000 trials.
At , both task-responsive policies have , , and . Nevertheless, their theoretical task-success probabilities are 0.788 and 0.212. The independent-action control is at chance. These differences are not attributable to different amounts of channel information: the corrective and inverting systems differ only in the policy’s relation to the task requirement.
Figure 5. Structural regularity and causal coupling for rule 30 in rule 2. Binary entropy, raw DEFLATE bits per original agent cell, and normalised coupling are displayed as separately defined quantities; they are not alternative estimators of the same object. Entropy remains close to one, code length declines and coupling is comparatively stable. The code and entropy use 64-update windows; the causal score uses the fixed 32-update protocol.
The equal-error sweep confirms the same separation across the entire tested noise range. As channel information increases, corrective performance improves and inverting performance worsens, despite identical information profiles at every matched noise level. The example directly refutes sufficiency of the task-aligned (, , ) and their derived scores for useful regulation; it is not a direct task test of the ECA exposure densities. It also refutes positive directional dominance as a necessary feature of a successful regulator.
Scope of the exploratory information and BDM comparisons
Appendix E retains all exploratory ECA MI-density, BDM-composite and layer-level panels with their full interpretation. Their native scales and incompletely matched protocols do not identify them with the normalised channel scores. Some directional patterns differ between MI and BDM, so the panels do not establish estimator agreement. Their values remain raw: zero-mask references must be applied to the encoded primitives under the actual aggregation formula, not inferred by subtracting an arbitrary constant from a plotted product. Layer regularity is distinct from a null causal effect, and stable structural curves do not establish homeostasis or equilibrium.
Figure 6. Opposite functional outcomes at identical coupling. Curves are the exact equal-error predictions from equations (20)–(21); points and Wilson error bars show 100,000 simulated challenges per policy and error setting. Corrective and inverting policies have the same information profile at each horizontal position. The dashed line is the expected marginal-matched open-loop success probability, shown as a chance reference; the open-loop controller itself has zero coupling.
Signed-network proof-of-concept
In the deterministic 32-node extension, aggregate coupling at was , with 95% context-bootstrap interval [0.001353, 0.002223]. At , , with interval [0.000508, 0.001045]. Across the five separately sampled ensembles, the point estimates of decreased monotonically. This is a descriptive ordering, not an established monotonic population law. All aggregate intervals included zero: the estimates do not resolve an ensemble directional imbalance, but neither prove equality of the channels nor demonstrate balanced interaction within contexts.
The reported BDM response-mask sum was 70.45 [67.93, 72.96] at and 60.28 [58.16, 62.43] at . These are raw description-length estimates, including the contribution of unchanged response masks, not zero-referenced effect sizes. With the common 16 × 16 shape and fixed estimator , the corresponding structural contrast is , where . A common subtraction preserves the between-ensemble ordering and differences, but absolute corrected values cannot be supplied without the numerical reference and complete estimator settings. No corrected values or corrected intervals are claimed here. In particular, a positive raw sum does not by itself show a nonzero causal response: that evidence comes from the paired masks or Shannon channels.
Recurrence was observed in all 64 contexts at each mixture within the 512-update search window, with detected periods of one or two updates (period one only at ). The channels and raw structural descriptors nevertheless varied across the sign ensembles. The experiment thus extends the specified perturbational protocol beyond cellular automata and illustrates the non-equivalence of recurrence, channel exposure and response description. It does not estimate within-context reciprocity, learning, beneficial agency or a physical critical transition.

Figure 7. Signed-network proof-of-concept: 32 nodes on a degree-four ring, equal 16-node regions, symmetric random edge signs and 64 independent network–initial-state contexts per positive-edge probability . Panels A and B show and , computed from primitive ensemble means. Balance of means is not within-context reciprocity. Panel C’s BDM proxy denotes the raw mean of XOR-mask descriptions, not causal coupling or a zero-referenced result. With the fixed encoding, the structural contrast would subtract . Bars are 95% percentile intervals from 5,000 complete-context bootstrap resamples, without estimator uncertainty.
The complete five-ensemble numerical table is retained in Appendix § 12.
Discussion
The central result is a separation between causal access, structural description and functional accomplishment. Theorem 3 derives positive information across each cut of a specified challenge–sensor–internal channel–response chain from better-than-chance regulation. Its force comes from an independently defined task, not from naming a positive score life or agency. Proposition 4 then supplies the complementary limitation: controllers with identical task-aligned information profiles can implement correct or inverted responses. Thus causal information is necessary under the theorem’s assumptions, but its amount does not determine its use.
This necessity statement must not be transferred unchanged to the ECA densities. The theorem concerns whole task-aligned channels under task-induced input distributions. The ECA experiment averages local binary response indicators over specified sources, recipients, lags and contexts. Those averages measure causal information-exposure density, not joint information, conserved flux or a demonstrated regulatory capacity. The numerical work establishes properties of these diagnostic profiles, whereas the theorem establishes a conditional constraint on a separate explicit task. An ECA test of that constraint would first need an implemented task and the corresponding whole-channel variables.
Aggregation also limits inference. Reported can be positive when different one-way contexts supply the inward and outward means. Near-zero can likewise reflect cancellation rather than within-context balance. Claims about individual contexts require their joint scores, , prevalence or distributions. Even those separately probed directions do not demonstrate operating feedback; that requires temporal and task evidence. The present bootstrap describes ensemble summaries, not these unmeasured properties.
BDM requires a parallel distinction. An unchanged mask has raw reference , not necessarily zero. Its matched contrast vanishes for a null effect but may be signed or vanish for other masks. Causal presence is checked separately. Moreover, complementing both intervention outputs preserves their XOR mask while potentially reversing task correctness; structural calibration does not resolve functional ambiguity.
The network table therefore retains its raw values. Common-baseline subtraction preserves matched-ensemble differences, not a causal interpretation of positive raw levels. The exploratory ECA panels use different native scales and encodings; some MI-density and BDM directional patterns even disagree. They cannot be treated as interchangeable estimators or quantitative corroboration of the normalised experiment. Layer-level trends describe displayed structure, not homeostasis or equilibration. Numerical harmonisation requires the arrays and full configuration; defining its baseline is not a completed reanalysis.
The network experiment transfers the specified intervention logic and normalised channels beyond ECA. Variation in those channels and raw mask descriptions alongside short recurrence illustrates non-equivalence, not statistical independence or a universal edge-mixture law. Learning, biological viability and multi-agent objectives are not tested.
Boundary choice is another part of the experiment rather than a nuisance removed by normalisation. Padding outside the preserved causal ancestry and matched role exchange test well-defined invariances. Changing the width of a region governed by a different rule also changes the physical update map. The resulting directionality reversals therefore demonstrate architecture dependence; they do not isolate the effect of an observer repartitioning fixed dynamics. The mutual-perturbation perspective [40] motivates testing the latter question separately.
Appendix § 14 develops the proposed organisational-persistence and interaction-timescale tests. A possible future summary of boundary robustness uses a justified family of partitions of the same fixed mechanism and a matched context ensemble. Let contain prespecified primitive and within-context descriptors and let be a stated reference profile. One illustrative summary is
The normalisation, weights and treatment of undefined ratios must be fixed in advance. Including both primitive scores and algebraically dependent composites can double-weight the same evidence. High robustness is not sufficient for individuality: an inert system may be uniformly unresponsive under every partition. Boundary robustness must therefore be combined with nontrivial organisation and independent function, and evaluated on held-out contexts if boundaries were selected using the data. This proposed statistic was not measured here.
Biological applications require an independent viability observable, such as survival in a specified set over a fixed challenge horizon. Compare intact and channel-disrupting conditions:
A causal interpretation requires a controlled intervention, relevant marginal-input matching and checks that disruption does not merely damage the organism or remove essential material resources. A positive would support a viability-relevant role for the tested channel, not automatically establish life. This is a proposed bridge to semantic-information research [3], not a result of the ECA simulations.
Neither dormancy nor passive protection is excluded by the theorem. A predictable challenge, pre-existing information or one universally successful response removes its chance-performance premise. Error-correcting systems can also maintain organisation while suppressing every response in a narrow perturbation ensemble. Measured absence therefore concerns the chosen probes and horizon, not an unconditional absence of agency. Conversely, outward dominance is not required: the balanced corrective controller regulates successfully at . Raw channels, context structure, persistence and task benefit should be reported separately before an aggregate receives a functional label.
Relation to AID and to information-theoretic causality
AID supplies intervention, counterfactual comparison and response localisation. Known ECA mechanisms identify effects by construction, without discovering unknown programs or computing algorithmic mutual information.
Regular and irregular responses can reach similar recipient fractions yet have different descriptions. CTM/BDM and DEFLATE provide structural proxies, not exact Kolmogorov complexity. Analyses of full response profiles or model changes must declare encodings and approximation limits. The zero-mask reference calibrates a specific structural contrast; it does not supply algorithmic conditional information.
Randomised intervention inputs connect the experiment to causal information theory [5, 6]. The method is not transfer entropy under a different name, nor does its causal interpretation require Shannon information to be intrinsically noncausal. The intervention distribution identifies the experimental channel; mutual information quantifies that channel. Algorithmic and Shannon observables answer different questions within a common perturbational methodology.
Ruffini’s work raises related questions about persistence, regulation and algorithmic environmental models [45, 44]. Its Game of Life construction is close in experimental spirit to cellular-automaton studies here. The historical context also includes earlier algorithmic approaches to life and evolution, Game of Life programmability, persistent and colliding structures, AID perturbations, open-ended evolution and observer dependence [1, 34, 37, 35, 36, 38, 40, 39]. The present contribution is an operational synthesis with explicit channel definitions, a conditional theorem, counterexamples and a separation of demonstrated measurements from proposed extensions.
Limits of inference and generalisation
The simulations use imposed boundaries, a finite rule family, limited contexts and specific perturbations. The 2,000-pair scan does not average over all primary initial states. Its bootstrap resamples complete initial-condition units; it does not cover all choices of boundary, estimator, encoding or intervention distribution. No living system is classified, no sensitivity or specificity for life is estimated, and no universal threshold is proposed. The regulatory benchmark is not a model of metabolism, reproduction or biological individuality.
Marginal channels can miss joint information. With independent fair , let and . Each recipient alone has zero information about , whereas the pair has one bit. A sum of marginal information scores cannot recover that synergy. Likewise, single-bit interventions can miss coordinated state transitions. Joint interventions and task-aligned recipient representations are needed when those possibilities matter. Composition identities under fixed source, recipient and clamping conventions do not turn nonlinear score cross-terms into evidence of emergence or collective function.
Exact recurrence, stationarity, equilibrium and homeostasis are distinct. A finite deterministic ECA eventually repeats, but the simulator has no calibrated temperature, work, heat or entropy-production variables. Thermodynamic information-flow results require a physical model [46]. Stable BDM, compression or channel curves cannot supply that missing model. Homeostasis requires a regulated observable and a restoring mechanism, neither of which follows from a flat curve.
The remaining empirical limitations are explicit. Within-context reciprocity and episode-level feedback have not been quantified; zero-referenced numerical BDM results have not been reconstructed; and the retained exploratory ECA descriptors have not been harmonised with the primary protocol. These are substantive analyses requiring their underlying data and configurations, not repairs that can be achieved by notation alone. They constrain what the figures establish without invalidating the specified Shannon measurements or the task-conditional theorem.
Conclusions
This paper develops a perturbational framework connecting causal information to adaptive regulation without identifying either causal coupling or complexity with life. In deterministic binary channels, context-conditioned mutual information has an exact paired-response representation. Averaging these local quantities produces outward, inward and clamped internal information-exposure densities. Their retention-weighted composites satisfy
an algebraic identity rather than empirical evidence for a life-agency hierarchy. The reciprocal composite concerns ensemble means; reciprocity within contexts, coordinated episodes and functioning feedback require separate evidence.
The 2,000-pair scan, detailed ECA cases, recurrence and coding analyses, geometry controls and signed-network extension demonstrate protocol-dependent differences among responsiveness, directional imbalance, structural regularity and recurrence. CTM/BDM add descriptions of encoded objects, not a second estimate of the same Shannon channel. For paired masks, the relevant structural contrast subtracts the BDM of a same-shaped zero mask under identical settings. Existing numerical BDM values remain explicitly raw, and the exploratory plots remain on their native scales; neither is presented as a completed baseline-corrected causal analysis. A common reference preserves matched-estimator comparisons, but raw positivity alone does not establish an effect.
The strongest theoretical conclusion is conditional and functional. Successful regulation of the specified unpredictable binary challenge requires information across each causal cut, while correct and inverted policies can have identical information profiles and opposite outcomes. Neither ensemble reciprocity nor a response-mask complexity descriptor resolves that ambiguity. Information is a necessary resource under the stated task and causal-chain assumptions, not a sufficient criterion for agency or a universal measured requirement for every living system.
The proposed extension to life (Appendix § 14) therefore combines independent tests: persistence of a specified organisation, context- or episode-resolved interaction, causal contribution to viability, and robustness across justified boundaries of fixed dynamics. Perturbation retention is not organisational survival, architecture variation is not solely observational repartitioning, and a stable complexity curve is not equilibrium. Selective sensing and action may be sequential and need not produce permanently positive outward dominance.
These distinctions preserve the operational role of information exchange and algorithmic response analysis while making their limits testable. They connect the study to earlier work on open-ended evolution and mutual perturbation [35, 40], without claiming that the present automata establish biological function. The framework delivers measurable causal profiles, a task-specific necessity result and explicit failures of sufficiency; its broader account of life and agency is a programme for functional validation, not a scalar verdict.
Data and code availability
A reproducibility package containing the complete simulation and analysis code, explicit configuration and seeds, initial states for the primary scan and detailed cases, context-level numerical arrays, plain-text summaries, validation outputs and all figure-generation scripts are posted. The supplementary package is the data source for every newly generated numerical claim in this article.
Collections of subsystems: composition under a fixed experiment
Consider a focal subsystem , with disjoint components and external environment . A valid composition law must keep the update mechanism, context distribution, source weighting and clamping intervention fixed. Measuring inside is different from declaring the entire focal system and clamping as part of its environment.
To make the accounting explicit, first use unnormalised exposure totals, with sums rather than averages over source cells. Define
and analogously
For disjoint source and recipient sets,
These are accounting identities for separately indexed intervention–recipient observations, not additivity laws for independent information bits.
For retention, clamp only the common external region , allowing both and to evolve. Define
where . Then
These terms classify intervention origins and recipient locations. A within-component term may contain a path that passes through the other component before returning. It is not an isolated self-retention measurement, and clamping the other component would change the experiment. Write , , , and . With and ,
These division-free identities remain valid when a within-component retention term is zero. The last two terms can remain nonzero when ; their presence is therefore not evidence of emergent inter-component causation.
Source averaging changes these relationships. If are recipient counts averaged over uniformly sampled sources and , then
not . With opportunity normalisation, inward densities must also be combined using eligible-recipient counts. Consequently, the normalised do not inherit simple total-count additivity. Claims of collective function require an independent functional criterion and an intact-versus-disrupted comparison; joint informational synergy requires a joint intervention distribution and joint-channel analysis.
Extended computational methods
Hybrid cellular automata and sampling
We use a periodic one-dimensional lattice with 210 binary cells. The central ten cells form ; the remaining 200 form . Each cell updates synchronously using its own region’s ECA rule and the actual states of its two neighbours, including neighbours across the boundary. For rule number , the update is the rule-table expression in equation (22) of the main text, which extracts the rule-table bit indexed by the three-cell neighbourhood. The implementation uses integer right shift followed by a one-bit mask. This convention is tested against constant, identity, shift and XOR rules.
The primary scan samples 2,000 distinct ordered pairs uniformly without replacement from the possibilities. Sixteen independent Bernoulli-one-half initial configurations are generated per pair and evolved for 64 updates before intervention. Every focal cell is probed. Environment probes occupy the five cells immediately outside each boundary. The two directions therefore have equal numbers of one-bit interventions and matched distances from their nearest interface. The 32-step primary horizon uses lags 17–32 for internal retention.
The six detailed combinations are (30, 18), (250, 30), (30, 2), (30, 45), (110, 60) and (45, 2). They retain propagation, contracting-region, structured-environment and transient cases within a common protocol. Each is evaluated on 128 independent initial configurations at ages 0, 16, 64, 128, 256 and 512. The same initial configurations are followed across ages, enabling paired temporal comparisons. The cases were specified before the simulations reported here, but neither the study nor its analyses were preregistered.
| Element | Specification |
| Primary ensemble | 2,000 distinct ordered ECA rule pairs; 16 initial configurations per pair |
| Primary geometry | 10 focal cells and 200 environment cells; periodic boundary |
| Interventions | One-bit flips; 10 focal and 10 boundary-matched environment sources |
| Primary timing | Age 64; horizon 32; internal retention evaluated at lags 17–32 |
| Detailed cases | Six detailed combinations plus five homogeneous controls; 128 configurations; six ages |
| Geometry sensitivity | Seven geometries; horizons 16, 32, 64; 48 configurations per pair and geometry |
| Uncertainty | 5,000 initial-condition bootstrap resamples; paired resampling across ages |
Table B.1. Computational design. All simulation settings and initial-state arrays are included in the reproducibility package.
| Element | Specification |
| Regulatory benchmark | Eight error settings; 100,000 challenges per policy and setting |
Table 2.
Negative controls and geometry checks
Five homogeneous comparisons are added: constant rule 0, identity rule 204, additive rule 90, shift rule 170 and rule 30. These distinguish extinction, static retention, regular transport and sensitive nonlinear dynamics. They do not supply biological positive or negative labels.
A sham intervention leaves the source unchanged. A boundary-cut control replaces each cross-boundary neighbour input by its reference-trajectory value at every update. The latter preserves the reference trajectory but prevents a perturbation from being transmitted across the cut. Both controls are evaluated for all six detailed pairs on 32 contexts each.
The zero-effect assertions for these controls concern paired differences and Shannon channel scores. For raw BDM, a sham response instead returns the same-shaped zero-mask reference ; its defined structural contrast is zero. No numerical BDM control results are inferred from the Shannon checks.
Geometry sensitivity uses , , , , , and and horizons 16, 32 and 64. There are 48 independent initial configurations per pair and geometry, shared across horizons within that geometry. At most five sources are sampled at each side of either interface; when , three are used. All available sources are enumerated, not randomly subsampled. Initial configurations differ across geometries, so these comparisons concern ensemble means rather than paired individual trajectories.
Additional exact tests use 50 random contexts for equal-region role exchange, and 50 for padding a 210-cell configuration to 410 cells while preserving its relevant initial-state ancestry. The latter are performed at intervention age zero with horizon 32, not after independently evolving the two differently sized lattices. A separate scalar implementation validates the numerical estimator on 24 small-lattice contexts.
Structural descriptors, BDM conventions and recurrence
Causal sensitivity, compressibility and recurrence are measured separately. For each detailed case and age, binary site-frequency entropy and raw DEFLATE length are calculated from a fixed 64-update unperturbed spacetime window, treating focal and environmental regions separately. Matrices are flattened in time-major order, bit-packed most-significant-bit first and compressed with level-nine raw DEFLATE (negative window-bits setting, without a zlib header). Code length is reported as compressed bits per original binary cell. Site-frequency entropy describes the fraction of zeros and ones, not the entropy rate of the full spatiotemporal process.
DEFLATE is a fixed-code structural descriptor, not an estimate of mutual information or exact Kolmogorov complexity. Finite-format costs can give ratios above one. Within-focal-region comparisons across ages use the same representation; focal and environmental code lengths are not treated as directly interchangeable.
The network analysis describes rectangular XOR masks; exploratory ECA panels describe layers and native-scale composites. The latter are not assumed to share the primary scan’s clamping, weighting or normalisation. They support qualitative comparisons, not reconstructed normalised scores or cross-estimator statistics.
Exact BDM reproduction requires array shape and ordering; software and CTM table; block dimensions, tiling and remainder handling; and source/context weights [14]. Defaults cannot be recovered from plot axes. Calculate each mask’s zero reference under identical settings before aggregation (§ 2.5). The network summaries supply raw values, not the numerical reference or corrected results; none are invented here. A common reference translates means and percentile intervals without changing between-group differences. Varying shapes or settings require sample-specific references. This subtraction does not remove finite-block approximation error. For an unperturbed layer, a zero-array reference concerns regularity, not causal absence.
We also search for an exact repeat of the complete 210-cell state through update 4,096. A repeated complete state proves entry into a cycle because the finite update map is deterministic. The first repeated state’s transient length and cycle period are reported. Failure to observe recurrence within the window is censored observation, not evidence of aperiodicity, chaos or thermodynamic nonequilibrium.
Regulatory benchmark and statistical analysis
The binary regulatory benchmark uses (19) at and at seven equal-error settings from zero to one-half. For each setting, 100,000 independent challenges are evaluated for each policy. Corrective and inverting policies share the challenge and noise realisations, so their paired successes are complementary. A marginal-matched open-loop control draws the action independently as a fair bit; it preserves the actuator’s input distribution but destroys the sensor-to-action dependence. Its success probability is one-half. This is an abstract regulatory task, not a biological experiment.
For each ECA pair, we first average the context scores to obtain , then calculate . We do not interchange these nonlinear operations. In particular, reported is ensemble reciprocity, not the mean within-context reciprocal score or the prevalence . The latter require the joint context-level scores and are not estimated here. Percentile bootstrap intervals use 5,000 resamples of complete initial-condition units; cells, time points and source interventions are not treated as independent replicates. Temporal intervals resample the same initial conditions jointly at both ages. Ordinary 95% intervals are descriptive. For the three designated transient comparisons (30, 45), (110, 60), (45, 2), the six endpoint contrasts involving and additionally use 99.167% marginal bootstrap intervals, providing a Bonferroni-adjusted nominal familywise level of 95%, subject to bootstrap approximation.
Scan sign counts describe estimated pair means, with zero defined by absolute value at most . They are not exact counts of population signs after integrating every possible initial state. Split-half rank correlations compare the first and last eight contexts per pair and quantify repeatability, not biological validity. For the benchmark, binomial Wilson intervals accompany individual policy-success estimates; the policies are not treated as independent when comparing them.
All numerical experiments use the explicit seeds and software versions recorded in the accompanying code and configuration files. The implementation evaluates conditional information exactly within deterministic contexts; its uncertainty statements concern the sampled ensembles. No published or manuscript figure was digitised to obtain a numerical result.
Signed-network generalisation using the same perturbational observables
As a separate proof of concept, the intervention protocol was applied to a deterministic network of 32 binary nodes partitioned into contiguous 16-node regions on an undirected degree-four ring, with connections to the two closest nodes on either side. Each undirected edge was independently assigned symmetric weight , with positive-weight probability . For each , 64 independent network-and-initial-state contexts were generated. Synchronous dynamics used
with for , for , and for ties. Networks evolved for 16 updates before intervention. The four boundary-connected sources in each region were probed by single-bit flips. Exposure was normalised over recipient–lag pairs reachable within graph distance at a 16-update horizon. Retention used lags 9–16 with the environment clamped to its reference trajectory.
Unperturbed structural measurements comprised binary entropy and level-nine raw DEFLATE bits per cell over a 32-update window. Complete-network recurrence was sought over the next 512 updates. The response-mask descriptor was two-dimensional pybdm BDM of perturbed XOR reference recipient arrays. Both directions used a full 16-lag by 16-node recipient array, including zero entries outside actual response support. These rectangular descriptors differ from the geometrically normalised channel densities.
For the numerical tables and existing figure labels, abbreviate source- and context-averaged . They are not baseline-subtracted. quantities. For fixed estimator settings, let
The corresponding baseline-aware sum would be . Its value is not claimed here because the numerical zero-mask reference is not reported. The same reference applies across the five mixtures, so comparisons between their raw sums are unaffected by this constant subtraction. Its absolute size still cannot be interpreted as independent transmitted information or functional coupling.
For each , , , were averaged over the 64 contexts before constructing , , , , matching the ECA convention. This estimates ensemble products, not , or the average of context-level products. Percentile intervals used 5,000 complete-context bootstrap resamples, treating a sampled network and initial state as one unit. The sign-mixture ensembles were independently generated, not paired across . Bootstrap uncertainty concerns context sampling; it does not quantify uncertainty from BDM block conventions.
Paired intervention visualisation

Figure B.1. (a) Unperturbed control case. (b) Perturbation of the focal subsystem. (c) Perturbation of the environment. Perturbational visualisation for rule 30 in a rule-18 environment. The panels show the control trajectory and perturbations originating on either side of the imposed boundary. They are retained to make explicit the common intervention logic underlying both the Shannon and structural analyses. The panels alone do not specify a BDM estimator, zero-mask reference or functional criterion.
Extended ECA results and controls
Detailed cases distinguish outward effects from retained coupling
Table C.1: Primitive and derived scores at intervention age 64. Means are over 128 independent initial configurations; intervals resample those configurations. A ratio of one with zero coupling, as for 250 / 30, is not evidence of retained agency.
| Rules A / E | [95% interval] | |||||
| 30 / 18 | 0.260 | 0.362 | 0.416 | 0.259 [0.251, 0.268] | -0.042 | -0.164 |
| 250 / 30 | 0.059 | 0 | 0 | 0 [0, 0] | 0 | 1.000 |
| 30 / 2 | 0.134 | 0.101 | 0.299 | 0.070 [0.066, 0.074] | 0.010 | 0.139 |
| 30 / 45 | 0.311 | 0.353 | 0.388 | 0.258 [0.248, 0.268] | -0.016 | -0.063 |
| 110 / 60 | 0.026 | 0.148 | 0.326 | 0.057 [0.054, 0.059] | -0.040 | -0.699 |
| 45 / 2 | 0.024 | 0.035 | 0.272 | 0.016 [0.015, 0.017] | -0.003 | -0.197 |
Table C.1. Primitive and derived scores at intervention age 64. Means are over 128 independent initial configurations; intervals resample those configurations. A ratio of one with zero coupling, as for 250 / 30, is not evidence of retained agency.
The rule-250 focal region has nonzero outward exposure, , but vanishing late internal retention and inward exposure at age 64. Consequently its coupling is zero even though its unweighted directionality ratio is one. This case demonstrates why a large directional ratio alone is not evidence of internally sustained organisation.

Figure C.1. Cellular-automaton evolution for rule 250 as the focal subsystem in a rule-30 environment.
Rule 30 in rule 2 has positive aggregate reciprocity and modest outward dominance: its mean inward, outward and retention scores are positive. This does not establish reciprocal transmission within every context, or report how many contexts have all three positive scores. Rule 30 in rule 45 shows larger coupling but slight inward dominance. Rule 110 in rule 60 is more strongly inward dominated. These are descriptions of causal responsiveness under specific boundary probes, not rankings of which automaton is more alive.

Figure C.2. Cellular-automaton evolution for rule 30 as the focal subsystem in a rule-2 environment.
The homogeneous controls further delimit interpretation. Rule 0 has . Rule 204 retains the altered focal bit, giving , but transmits nothing across the boundary, so . Rule 170 transports perturbations across the boundary but retains none within the focal region over the late window, again giving . The additive rule-90 lattice has despite having no implemented regulatory objective. Homogeneous rule 30 has and : even identical update rules on both sides do not force zero asymmetry when the regions have unequal geometry.
Transient changes in coupling do not imply monotonic changes in directionality
| Rules A / E | [95% interval] | [95% interval] |
| 30 / 18 | 0.0207 [0.0076, 0.0334] | -0.0245 [-0.0316, -0.0172] |
| 250 / 30 | 0 [0, 0] | 0 [0, 0] |
| Rules A / E | [95% interval] | [95% interval] |
| 30 / 2 | -0.0030 [-0.0082, 0.0022] | 0.0070 [0.0025, 0.0114] |
| 30 / 45 | -0.0100 [-0.0226, 0.0031] | 0.0035 [-0.0012, 0.0080] |
| 110 / 60 | -0.0191 [-0.0228, -0.0154] | -0.0009 [-0.0036, 0.0018] |
| 45 / 2 | -0.0294 [-0.0334, -0.0255] | 0.0012 [-0.0013, 0.0037] |
Table C.2. Paired changes from intervention age zero to 512. These are differences between aggregate scores constructed from primitive ensemble means, not averages of per-context aggregate-score differences. Multiplicity-adjusted intervals for the three designated transient comparisons are stored in the numerical summary.
For rule 110 in rule 60, coupling falls from 0.0749 to 0.0558 between ages zero and 512. For rule 45 in rule 2, it falls from 0.0442 to 0.0147. Both decreases remain separated from zero by the multiplicity-adjusted bootstrap intervals. Their signed directional changes, however, are not separated from zero by ordinary 95% intervals. In particular, a negative score moving closer to zero would be an increase in the signed score and a decrease in its magnitude; these must not be conflated.
For rule 30 in rule 45, the coupling difference is -0.0100 with an interval spanning zero, and the signed directional difference also spans zero. These data support the absence of a clearly resolved endpoint change at the available precision, not a proof of stationarity or equilibrium. For rule 30 in rule 18, coupling instead increases. There is no general monotonic law linking age, complexity and coupling across the six examples.

Figure C.3. Cellular-automaton evolution for rule 30 as the focal subsystem in a rule-45 environment.
The aggregate coupling and directionality time courses are displayed in the main text; the table above retains every paired endpoint contrast.
Exact recurrence and structural regularity are different from causal responsiveness

Figure C.4. Cellular-automaton evolution for rule 110 as the focal subsystem in a rule-60 environment.

Figure C.5. Cellular-automaton evolution for rule 45 as the focal subsystem in a rule-2 environment.
| Rules A / E | Observed cycles | Median transient | Median period | Period range |
| 30 / 18 | 0 / 128 | - | - | - |
| 250 / 30 | 128 / 128 | 515 | 2 | 1–2 |
| 30 / 2 | 128 / 128 | 397 | 4 | 3–8 |
| 30 / 45 | 0 / 128 | - | - | - |
| 110 / 60 | 128 / 128 | 44.5 | 256 | 256–512 |
Table C.3. Exact recurrence of the complete lattice state through update 4,096. A dash denotes no detected recurrence, not evidence of an infinite nonperiodic trajectory. Transient and period summaries are conditional on recurrence being observed.
Exact complete-state cycles are detected in every tested initial condition for four of the six detailed pairs. For rule 110 in rule 60, the median transient is only 44.5 updates, but the median cycle period is 256 updates. A short local window can therefore look irregular even after exact recurrence has been reached. Neither high short-window entropy nor stable code length establishes absence of periodic organisation.
Rule 30 in rule 2 provides a clear separation of pattern regularity and causal response. Its mean agent binary entropy remains close to one bit per site (0.9981 at age zero and 0.9992 at age 512), while the 64-step DEFLATE ratio falls from 0.6183 to 0.1396. Coupling changes only from 0.0757 to 0.0727, with its paired confidence interval spanning zero. The system can therefore become much more compressible while maintaining a comparable response to intervention. This is not a paradox: the code measures the regularity of the unperturbed spacetime array, whereas measures sensitivity to controlled changes.
For rule 110 in rule 60, the same short-window code ratio changes much less, from 0.9169 to 0.9040, even though coupling decreases. The two cases rule out a simple identification of coding complexity with causal coupling in either direction. We make no claim that these fixed-code results reproduce a BDM estimate or measure algorithmic mutual information.
Geometry and intervention controls
Geometry can alter both the magnitude and the sign of directionality. With a 200-cell environment and a 32-step horizon, rule 45 in rule 2 changes from at focal width six to -0.238 at width ten and +0.115 at width twenty. Rule 30 in rule 2 changes from -0.077 at width six to +0.151 at width ten. Therefore dimensionless normalisation is not a licence to treat the score as an intrinsic, scale-free property of a rule pair.
Changing focal width changes the allocation of update rules and therefore the update map as well as the intervention geometry. These reversals demonstrate architecture-dependent causal attribution, not a change in agency caused solely by an observer moving a boundary around an otherwise unchanged system. The mutual-perturbation framework makes the observer–system decomposition explicit [40], but a direct test of observational-boundary robustness would keep the physical update map, underlying trajectories and context distribution fixed while recomputing matched profiles for several justified partitions. That separate test has not been performed here.
In contrast, adding distant cells while preserving the relevant local initial-state ancestry leaves the scores exactly unchanged in all 50 padding tests. Equal-region role exchange swaps and within a maximum discrepancy of in 50 tests. This distinguishes desirable invariances of a fixed experiment from genuine changes to the experiment itself.
Every sham intervention yields zero Shannon effect scores. This does not mean a zero raw BDM: the corresponding unchanged-response mask has , and only its defined matched contrast is zero. Every boundary-cut control yields exactly zero and , across 192 pair-context combinations. The independent scalar implementation agrees exactly with the accelerated implementation in all 24 checks. Across the scan, the largest factorisation residual is and no bound violations occur.
Clamping also changes interpretation. At age 64, and for rule 30 in rule 45, whereas for rule 45 in rule 2 the corresponding values are 0.1716 and 0.2717. Clamping need not reduce the response: environmental feedback can either sustain or suppress a difference. Accordingly, is not assumed to be a nonnegative decomposition of causal influence.
Figure C.6. Directionality depends on the chosen focal width. The environment has 200 cells and the horizon is 32 updates. Error bars are 95% initial-condition bootstrap intervals from 48 configurations per pair and width. The two regions of a partition cannot be treated as interchangeable merely because both scores are densities; the primary effect of this plot is to expose, not eliminate, scale dependence.
Complete signed-network estimates
| [95% CI] | [95% CI] | [95% CI] | |
| 0.00 | 0.001750 [0.001353, 0.002223] | [, 0.000204] | 70.45 [67.93, 72.96] |
| 0.25 | 0.001421 [0.000972, 0.001983] | [, 0.000114] | 69.51 [65.51, 73.95] |
| 0.50 | 0.001118 [0.000744, 0.001544] | 0.000005 [, 0.000199] | 67.00 [63.42, 70.81] |
| 0.75 | 0.001054 [0.000701, 0.001522] | 0.000015 [, 0.000211] | 64.06 [61.05, 67.35] |
| 1.00 | 0.000749 [0.000508, 0.001045] | [, 0.000059] | 60.28 [58.16, 62.43] |
Table D.1. Signed-network ensemble summaries with 95% context-bootstrap intervals (64 contexts per mixture). abbreviates the sum of raw BDM descriptors; the zero-mask offset has not been subtracted. These values are not channel mutual information. The table does not estimate within-context reciprocity.
Exploratory ECA information and BDM comparisons
The retained panels examine related ECA examples using information-density and BDM constructions in their native units. They preserve the structural comparisons that motivated the present separation of causal responsiveness, descriptional structure and function. They are not additional measurements of the normalised , in equation (12).
This distinction is visible in the axes: the exploratory MI scan is labelled non-normalised and its values exceed ; the BDM composites have still larger numerical scales. In this subsection the plotted , are denoted , . The original image labels are retained, but they do not identify the plotted values with the primary causal scores.
The panels do not specify enough of the intervention ensemble, BDM configuration, clamping and aggregation for a matched quantitative reconstruction. Their visual observations are therefore kept separate from the context-array results, without inferred correlations or intervals.
Native-scale rule-pair comparisons
Associations between sum-like and difference-like composites of common non-negative components can arise algebraically. They do not show that life causes agency or that both estimators measure the same quantity. Cross-estimator comparisons require matched contexts and explicit formulas.
Zero-mask offsets must pass through the actual BDM aggregation: subtracting a constant from a product is not equivalent to replacing its primitive terms with structural contrasts. Neither composites nor signs are retrospectively recalibrated from image pixels.
Contrasting focal and environmental dynamics
The cases contrast a simple focal rule in a complex environment with a complex focal rule in a more structured environment. Wolfram classes here describe dynamical examples, not biological classes.
For rule 250 in rule 30, the displayed MI distributions for and overlap substantially and are positive; the BDM distributions also overlap. They do not support describing directionality as uniformly weak relative to coupling in that example. These distributions are distinct from the age-64 primary result , . In that protocol, the retention multiplier makes coupling zero despite outward exposure. Positive native-scale exploratory values cannot be presented as replication of the normalised result.
For rule 30 in rule 2, the exploratory MI panel separates a large coupling-like distribution from a directional distribution near zero. The BDM panel displays its own native-scale distribution. These observations describe the plotted constructions; neither establishes functional regulation or measured within-context reciprocity.

Figure E.1. Retained exploratory MI-density scan in native nonnormalised units. The original , labels denote , , not the normalised scores of the primary 2,000-pair scan. No numerical conversion between the constructions is inferred from this panel.
Structural descriptions through dynamical transients
The layer-level panels describe unperturbed focal and environmental arrays. They are distinct from the network XOR masks. Within-layer temporal comparisons use the plotted representation; absolute cross-layer comparisons remain sensitive to size and decomposition. An all-zero reference for a layer describes structural regularity, not absence of causal transmission.
For rule 30/45, both layer BDM curves fluctuate around comparatively stable levels. The native MI directional curve is largely negative, whereas the native BDM directional curve is largely positive. This visible sign difference rules out treating the panels as interchangeable measurements of the same signed causal quantity. A stable envelope does not establish equilibrium.
For rule 110/60, focal-layer BDM decreases while the environmental curve remains comparatively stable. The BDM coupling-like composite falls, while its negative directional composite approaches zero: an increase in signed value and a decrease in magnitude. The separately specified normalised analysis resolves a decrease in , but not an endpoint change in signed at the reported precision.
For rule 45/2, focal BDM shows a pronounced stepped decline; environmental BDM drops early and subsequently remains more stable. The composites also change, with directional values generally moving towards zero. The two layers should not be described as showing identical sustained declines. These are descriptive observations of structural regularisation, not statistical tests based on digitised trajectories.

Figure E.2. Retained exploratory BDM composites in native units. The displayed values are not baseline-corrected channels or algorithmic mutual information. Zero-mask references would have to be applied to the actual primitive objects and aggregation formula before interpreting structural contrasts.
The structural panels preserve useful comparisons but do not establish statistical independence, a universal complexity trend or equivalence of estimators. Quantitative claims about the normalised causal profile rest on its specified simulations; the DEFLATE and recurrence results provide separate numerical dissociations; the task construction establishes the functional distinction. A harmonised BDM comparison would require the stored trajectories, exact estimator settings and matched zero references.

Figure E.3. Exploratory MI-density distributions for rule 250 in rule 30. Native-scale , overlap. They are not the normalised age-64 scores, for which zero retention gives .

Figure E.4. Exploratory BDM composites for rule 250 in rule 30, without a reported zero-mask correction. Their absolute positivity is not evidence of positive retained causal coupling.

Figure E.5. Exploratory MI-density distributions for rule 30 in rule 2. The native-scale separation is distinct from the primitive means and normalised directionality reported in the primary analysis.

Figure E.6. Exploratory BDM composites for rule 30 in rule 2. These are raw structural composites, not zero-baseline causal scores. Comparing estimators quantitatively would require matched definitions, settings and contexts.

Figure E.7. Retained layer-level BDM for rule 30 in rule 45. Both layers fluctuate around comparatively stable levels. The historical asset filename does not establish thermodynamic equilibrium; these are unperturbed-layer descriptors rather than response-mask contrasts.

Figure E.8. Exploratory MI-density time series for rule 30 in rule 45. The original units and time convention are retained. These are not the ensemble-normalised time courses from the primary protocol.

Figure E.9. Exploratory BDM composites for rule 30 in rule 45. The predominantly positive directional curve differs from the predominantly negative MI-density curve in Figure E.8. No unreported baseline correction is assumed. The corresponding CA spacetime example is retained in the preceding results.

Figure E.10. Layer-level BDM for rule 110 in rule 60. Focal structural description length decreases under the plotted representation while environmental BDM remains comparatively stable. No thermodynamic state is inferred.

Figure E.11. Exploratory MI-density composites for rule 110 in rule 60. Native values and fluctuations do not replace the normalised paired endpoint estimates or their context-bootstrap intervals.

Figure E.12. Exploratory BDM composites for rule 110 in rule 60. The coupling-like curve declines and negative directionality approaches zero. Decreasing magnitude is not decreasing signed value. These native composites are not reported as baseline-corrected responses.

Figure E.13. Layer-level BDM for rule 45 in rule 2. The focal curve shows stepped simplification; the environmental curve drops early and then stabilises more closely. The layers do not exhibit identical temporal changes.

Figure E.14. Exploratory MI-density composites for rule 45 in rule 2. Their native-scale changes are descriptive observations under that construction, distinct from the normalised ensemble comparison.

Figure E.15. Exploratory BDM composites for rule 45 in rule 2. Structural regularisation and composite changes do not establish functional loss, baseline-corrected causal transmission or a biological classification.
Proposed dynamical classification of life-like causal organisation
The results motivate an operational research programme, not a classification of the simulated automata as living or nonliving. Permanently positive outward dominance is not required by the regulation theorem. Sensing and acting can involve different directions of influence, and a successful controller can have zero directional imbalance. Three questions must instead be kept separate: whether an organisation persists, whether its boundary supports appropriately coordinated interaction, and whether that interaction contributes to an independently defined function.
Persistence of a perturbation is not persistence of an organisation. The measured describes a late, clamped internal response to a specified probe; an error-correcting system can preserve its organisation by erasing that response. To measure organisational persistence, specify an observable , an admissible set , and a context ensemble, then estimate a survival function such as
For a prespecified survival level , one possible timescale is . The observable, threshold, sampling interval and censoring horizon are part of the definition. Choosing a trivial admissible set would make persistence uninformative.
A separate causal-persistence experiment could randomly prepare an internal feature using an intervention input and measure
where contains only the specified pre-intervention context and the environmental-control protocol is stated. This is a channel about the prepared feature, not a survival probability. Its decay time need not equal : stable organisation can actively remove the preparation. Neither nor has been measured in the present experiments.
For interaction, define and as the first lags at which outward and inward effects meet prespecified detection criteria. If the hypothesis concerns access in both directions, the relevant completion time is
The minimum would time only the faster direction and could misleadingly classify a one-way channel as rapid reciprocal access. A direction not detected within the observation window must be reported as undetected or censored; it must not be assigned the other direction’s latency. Estimates of require compatible time units, finite positive denominators and a sufficiently long observation window. No value of this ratio is inferred from the existing fixed-window or recurrence measurements.
A large , where estimable, would describe relatively persistent organisation with faster two-way causal access. It would not by itself establish selective or beneficial interaction. For that purpose, use matched contexts and report both the aggregate and a within-context summary such as , together with . Positive aggregate can arise by mixing one-way contexts. Even positive within-context scores establish two separately probed directions, not necessarily a simultaneously operating feedback loop. Where sensing and action occur sequentially, define the episode, lag ordering and task relevance before testing their coordination.
The following regimes organise these questions; their qualitative thresholds are not fitted diagnostic criteria.
| Regime | Persistence of organisation | Boundary interaction | Interpretation |
| Decoupled or inert | Low | Low | Little measured persistence or cross-boundary response |
| Persistent isolated | High | Low | Persistent structure; maintenance mechanism remains to be tested |
| Inward-driven | Assessed separately | Inward-dominated response, not necessarily absence of agency | |
| Outward-led | Assessed separately | Source or actuator; task benefit is undetermined | |
| Highly coupled | Assessed separately | High | Strong ensemble exposure; selectivity and individuality unresolved |
| Candidate adaptive | High under challenge | Coordinated, task-relevant | Requires within-context or episode evidence and demonstrated function |
Table 3.
Algorithmic descriptors add a structural axis to this programme, not an alternative persistence or viability criterion. Raw BDM must identify its encoded object; paired response masks require the same-shaped zero-mask reference, whereas an unperturbed trajectory describes state structure. Neither a low description length nor a flat curve implies maintained organisation, and a nonzero zero-referenced contrast does not establish beneficial regulation.
This programme is compatible with earlier work on open-ended evolution. Innovation there is defined relative to counterfactual isolated dynamics, and scalable open-ended evolution was associated with state-dependent dynamics [35]. The complementary question here is whether an interacting subsystem preserves an independently specified organisation while its dynamics change. Neither innovation nor persistence alone demonstrates biological life.
The directional ratio can also be followed through time,
A hypothesised sensing–integration–action sequence might move from negative through approximately zero to positive . Such switching can also occur in passive or externally driven systems; its adaptive interpretation requires task-aligned interventions and functional evidence. Moreover, averaging different contexts at each age can manufacture an apparent sequence absent from any individual trajectory. Matched-context time series or explicitly aligned episodes are needed to test the hypothesis.
Thus persistent organisation, selective reciprocal access and functional benefit are proposed complementary tests, not consequences of , , a BDM trend or exact recurrence. The present paper supplies specified channel measurements, a task-conditional theorem and explicit counterexamples; the broader classification remains a falsifiable extension rather than an empirical finding.
Estimation algorithm and verification trail
For each rule pair and initial configuration, evolve the unperturbed system to the intervention age and store its reference trajectory through the required horizon. For every eligible source in , flip that source alone, evolve the open system and calculate the eligible environment response. Run the corresponding environment-clamped arm and calculate late internal response. Repeat open-system interventions for the matched environment sources and record the focal response. Average the source-normalised quantities within the context, then average primitive quantities across initial conditions. Only then calculate the aggregate scores.
The primary scan evaluates 32,000 pair-context units. Each unit contains 20 open-system single-source interventions and ten additional clamped interventions, giving 640,000 open-system and 320,000 clamped counterfactual trajectories, plus their references. The simulation code stores mean changed-site counts and geometric opportunities as well as densities. These counts permit verification of normalisation but are not interpreted as numbers of independent information bits.
The main experiment seed is 20260905. Statistical and benchmark calculations use 20260906; independent validation uses 20260907; graphical interval calculations use 20260908. The recorded numerical environment is Python 3.13.5, NumPy 2.3.5 and Numba 0.65.1. The source package records additional library versions. Bootstrap replicates are 5,000 throughout. Code reports rather than conceals ratios that are undefined because both cross-boundary scores are zero.
A separate pure scalar update function checks the ECA bit-order convention and evaluates the output entropy of both intervention arms directly. It agrees with the accelerated damage-indicator estimator in all tested contexts. Analytic rule checks cover rules 0, 255, 204, 170, 240 and 90. Further tests examine the algebraic bounds, exact factorisation, Shannon null interventions, boundary cuts, role exchange and locality-preserving padding. These checks do not supply a numerical BDM baseline: a separate structural audit must record each encoded object’s shape, estimator settings, raw value and matched zero-array reference.
Exact counterexamples and scope of the causal claim
Common input without transmission. Let a fair bit set both and , with no causal arrow from to . Observational , but intervening on leaves unchanged and the intervention channel has zero information. Correlation alone therefore fails the causal test.
Simple transmission without complexity. A perfect copy channel has one bit of causal information for a fair input despite its trivial generating rule. High response complexity is not necessary for a functioning channel.
Noise masking a microscopic perturbation. For with independent fair future noise, although the two outputs differ if one artificially fixes across intervention arms. Conditioning on future noise changes the estimand. The ECA damage identity is exact because those simulations are deterministic, not because paired Hamming differences estimate every stochastic causal channel.
Joint information invisible to marginal recipients. For and , both and vanish, but . A scalar sum of marginal information scores cannot replace a joint-channel analysis in general.
Stored state without external interaction. The identity rule retains a focal bit indefinitely but has no cross-boundary transmission. Requiring cross-boundary transmission prevents this memory from being called reciprocal interaction; self-retention alone is not evidence of self-maintenance.
Opposite task meaning under a bijection. Inverting the response preserves all channel mutual informations in the regulatory construction but reverses its relation to the task requirement. Hence sufficiency cannot be repaired by simply increasing a threshold on , or .
Complete example trajectories and additional results
The supplementary arrays retain every detailed case at every intervention age, including the constant, identity, shift, additive and homogeneous-rule-30 comparisons. The plain-text context table permits reconstruction of each reported mean without re-running the simulator. The sensitivity table contains all 126 rule-pair/geometry/horizon combinations, not only those displayed in the geometry figure in Appendix § 11.§ 4. Exact recurrence outputs include the transient and period of every observed cycle and an explicit censoring marker otherwise.
| Rules A / E | ||||||
| 0 / 0 | 0 | 0 | 0 | 0 | 0 | Undefined |
| 204 / 204 | 0 | 0 | 0.1000 | 0 | 0 | Undefined |
| 90 / 90 | 0.2485 | 0.2165 | 0.2638 | 0.1226 | 0.0084 | 0.0688 |
| 170 / 170 | 0.0347 | 0.0197 | 0 | 0 | 0 | 0.2768 |
| 30 / 30 | 0.3151 | 0.3749 | 0.4035 | 0.2784 | -0.0241 | -0.0866 |
Table I.1. Homogeneous controls at age 64 under the primary geometry and horizon. Zero coupling can reflect absent transmission or absent late internal retention; the primitive components distinguish these possibilities.
Figure I.1. Coupling for all six detailed rule pairs across the six intervention ages. This supplementary view includes the propagation example and the contracting focal region as well as the three principal transient comparisons. Numerical outputs and context-level uncertainty are retained in the data package; the full curves do not establish a universal monotonic trend.
References
- [1]Zenil, H., Gershenson, C., Marshall, J. A. R., and Rosenblueth, D. A. (2012). Life as thermodynamic evidence of algorithmic structure in natural environments. Entropy, 14, 2173–2191. doi:10.3390/e14112173.DOI
- [2]Krakauer, D., Bertschinger, N., Olbrich, E., Flack, J. C., and Ay, N. (2020). The information theory of individuality. Theory in Biosciences, 139, 209–223. doi:10.1007/s12064-020-00313-7.DOI
- [3]Kolchinsky, A. and Wolpert, D. H. (2018). Semantic information, autonomous agency and non-equilibrium statistical physics. Interface Focus, 8, 20180041. doi:10.1098/rsfs.2018.0041.DOI
- [4]Schreiber, T. (2000). Measuring information transfer. Physical Review Letters, 85, 461–464. doi:10.1103/PhysRevLett.85.461.DOI
- [5]Lizier, J. T. and Prokopenko, M. (2010). Differentiating information transfer and causal effect. The European Physical Journal B, 73, 605–615. doi:10.1140/epjb/e2010-00034-5.DOI
- [6]Ay, N. and Polani, D. (2008). Information flows in causal networks. Advances in Complex Systems, 11, 17–41. doi:10.1142/S0219525908001465.DOI
- [7]Klyubin, A. S., Polani, D., and Nehaniv, C. L. (2005). All else being equal be empowered. Advances in artificial life, 3630. Springer. doi:10.1007/11553090_75.DOI
- [8]Zenil, H., Kiani, N. A., Marabita, F., Deng, Y., Elias, S., Schmidt, A., Ball, G., and Tegnér, J. (2019). An algorithmic information calculus for causal discovery and reprogramming systems. iScience, 19, 1160–1172. doi:10.1016/j.isci.2019.07.043.DOI
- [9]Adams, A., Zenil, H., Reyes, E., and Joosten, J. (2015). The Effects of Global Rules on Interacting Cellular Automata. In J. Kari, I. Törmä, and M. Szabados (Eds.), 21st International Workshop on Cellular Automata and Discrete Complex Systems (AUTOMATA 2015). TUCS Lecture Notes.
- [10]Zenil, H., Kiani, N. A., Abrahão, F. S., and Tegnér, J. (2020). Algorithmic information dynamics. Scholarpedia, 15, 53143. doi:10.4249/scholarpedia.53143.DOI
- [11]Wolfram, S. (1984). Universality and complexity in cellular automata. Physica D, 10, 1–35. doi:10.1016/0167-2789(84)90245-8.DOI
- [12]Derrida, B. and Pomeau, Y. (1986). Random networks of automata: A simple annealed approximation. Europhysics Letters, 1, 45–49. doi:10.1209/0295-5075/1/2/001.DOI
- [13]Cover, T. M. and Thomas, J. A. (2006). Elements of information theory. Wiley.
- [14]Zenil, H., Hernández-O, A., and Tegnér, J. (2018). A decomposition method for. A decomposition method for global evaluation of Shannon Entropy and local estimations of Algorithmic Complexity. Entropy, 20, 605. doi:10.3390/e20080605.
- [15]Utham acumaran, A., Abrahão, F. S., Kiani, N. A., and Zenil, H. (2024). On the salient limitations of the methods of assembly theory and their classification of molecular biosignatures. npj Systems Biology and Applications, 10, 82. doi:10.1038/s41540-024-00403-y.DOI
- [16]Abrahão, F. S., Hernández-Orozco, S., Kiani, N. A., Tegnér, J., and Zenil, H. (2024). Assembly Theory is an approximation to algorithmic complexity based on LZ compression that does not explain selection or evolution. PLOS Complex Systems, 1(1), e0000014. doi:10.1371/journal.pcsy.0000014.DOI
- [17]Ozelim, L., Utham acumaran, A., Abrahão, F. S., Hernández-Orozco, S., Kiani, N. A., Tegnér, J., Hernández-Quiroz, F., and Zenil, H. (2026). Assembly theory collapses to dictionary compression and is rendered redundant by common statistical algorithms. npj Complexity. doi:10.1038/s44260-026-00088-w.
- [18]Barricelli, N. A. (1954). Esempi numerici di processi di evoluzione. Methodos, 6, 45–68.
- [19]Bialek, W. and Schwab, D. J. (2021). The physics of life. Annual Review of Condensed Matter Physics, 12, 1–24.
- [20]Gardner, M. (1970). Mathematical games: The fantastic combinations of John Conway’s new solitaire game “Life”. Scientific American, 223(4), 120–123.
- [21]Eigen, M. (1971). Selforganisation of matter and the evolution of biological macromolecules. Naturwissenschaften, 58, 465–523.DOI
- [22]England, J. L. (2013). Statistical physics of self-replication. Journal of Chemical Physics, 139, 121923.
- [23]Langton, C. G. (1989). Artificial life. In Artificial Life I, pp. 1–47. Addison-Wesley.
- [24]Li, M. and Vitányi, P. (2019). An Introduction to Kolmogorov Complexity and Its Applications, 4th ed. Springer.
- [25]Parrondo, J. M. R., Horowitz, J. M., and Sag awa, T. (2015). Thermodynamics of information. Nature Physics, 11(2), 131–139.
- [26]Prigogine, I. (1980). From Being to Becoming: Time and Complexity in the Physical Sciences. W. H. Freeman.
- [27]Schrödinger, E. (1944). What Is Life? Cambridge University Press.
- [28]Shannon, C. E. (1948). A mathematical theory of communication. Bell System Technical Journal, 27, 379–423 and 623–656.
- [29]Solomonoff, R. J. (1964). A formal theory of inductive inference, Parts I and II. Information and Control, 7, 1–22 and 224–254.DOI
- [30]Wissner-Gross, A. D. and Freer, C. E. (2013). Causal entropic forces. Physical Review Letters, 110, 168702.
- [31]Soler-Toscano, F., Zenil, H., Delahaye, J.-P., and Gauvrit, N. (2014). Calculating Kolmogorov complexity from the output frequency distributions of small Turing machines. PLOS ONE, 9(5), e96223. doi:10.1371/journal.pone.0096223.
- [32]Zenil, H. (2007). Cellular Automaton Compressibility. Wolfram Demonstrations Project. Wolfram Research.
- [33]Zenil, H. and Villarreal, E. (2014). Cellular Automata Sensitivity to Initial Conditions. Wolfram Demonstrations Project. Wolfram Research.
- [34]Zenil, H. and Marshall, J. A. R. (2013). Some Aspects of Computation Essential to Evolution and Life. ACM Ubiquity, 2013(April), 1–16. doi:10.1145/2480352.2480353.
- [35]Adams, A., Zenil, H., Davies, P. C. W., and Walker, S. I. (2017). Formal Definitions of Unbounded Evolution and Innovation Reveal Universal Mechanisms for Open-Ended Evolution in Dynamical Systems. Scientific Reports, 7, 997. doi:10.1038/s41598-017-00810-8.DOI
- [36]Hernández-Orozco, S., Hernández-Quiroz, F., and Zenil, H. (2018). The Limits of Decidable States on Open-Ended Evolution and Emergence. Artificial Life, 24(1), 56–70.
- [37]Zenil, H. (2015). Algorithmicity and Programmability in Natural Computing with the Game of Life as an In Silico Case Study. Journal of Experimental & Theoretical Artificial Intelligence, 27(1), 109–121. doi:10.1080/0952813X.2014.940686.DOI
- [38]Zenil, H., Kiani, N. A., and Tegnér, J. (2019). Algorithmic Information Dynamics of Emergent, Persistent, and Colliding Particles in the Game of Life. In From Parallel to Emergent Computing, pp. 367–384. CRC Press. doi:10.1201/9781315167084-17.DOI
- [39]Zenil, H. and Adams, A. (2022). Algorithmic Information Dynamics of Cellular Automata. In The Mathematical Artist: A Tribute to John Horton Conway, pp. 171–190. Springer. doi:10.1007/978-3-031-03986-7_8.DOI
- [40]Abrahão, F. S. and Zenil, H. (2022). Emergence and Algorithmic Information Dynamics of Systems and Observers. Philosophical Transactions of the Royal Society A, 380(2227), 20200429. doi:10.1098/rsta.2020.0429.DOI
- [43]Ruffini, G. (2017). An Algorithmic Information Theory of Consciousness. Neuroscience of Consciousness, 2017(1), nix019. doi:10.1093/nc/nix019.DOI
- [44]Ruffini, G. (2026). The Algorithmic Regulator. Entropy, 28(3), 257. doi:10.3390/e28030257.DOI
- [45]Ruffini, G. and Klaus (2026). When Patterns Persist: Reach, Regulation, and Death in the Game of Life. BCOM Working Paper WP019, companion to WP0162. Zenodo. doi:10.5281/zenodo.21008860.DOI
- [46]Horowitz, J. M. and Esposito, M. (2014). Thermodynamics with continuous information flow. Physical Review X, 4, 031015. doi:10.1103/PhysRevX.4.031015.DOI







