Unifying Information, Geometry, Chaos, and Structure: Evolutionary Fractal Non-Recurrence Theory (EFNRT)

Part 1 — The conceptual prequel to the Applied Force Model

1. The Problem Beneath Complex Systems

Why do intelligent people make decisions that worsen the situations they are trying to improve? Why do carefully designed systems behave in unexpected ways? Why can a small disturbance sometimes disappear without consequence, while another disturbance triggers a chain of failures?

These questions appear to belong to different fields. One concerns human reasoning, another system design, another engineering failures, and another the organization of living systems. Yet they share a difficulty: understanding the parts of a system does not necessarily enable us to predict the behaviour of the whole.

A component may function correctly while its interaction with other components produces an undesirable result. A local improvement may damage a larger structure. A minor change may have little effect today but alter the conditions under which future changes occur.

The deeper problem is that a system is not merely a collection of things. It is a configuration of things, their relationships, their constraints, and the processes through which that configuration changes.

This article begins with four books that approach complexity from different directions. It then asks whether their perspectives point toward a deeper question: how do information, geometry, relationships, and structures evolve while retaining some recognizable organization?

Evolutionary Fractal Non-Recurrence Theory, or EFNRT, is a proposed framework for exploring that question. It remains a conceptual sketch. Its value will depend on whether its ideas can eventually be defined precisely, distinguished from existing theories, and tested.

2. Four Books, Four Ways of Understanding Complexity

2.1. The Logic of Failure — Dietrich Dörner

Dörner investigates why people often fail when dealing with complex, dynamic situations. The problem is not necessarily a lack of intelligence. People simplify situations, focus on immediately visible effects, overlook delayed consequences, and intervene without adequately understanding how the system will respond.

A decision may solve the problem that prompted it while creating new problems elsewhere. Because effects can emerge gradually, the decision-maker may not recognize the connection between an intervention and its eventual consequences. Repeated interventions can make matters worse when each response addresses a symptom rather than the underlying dynamics.

The lesson is not that decisions are futile. It is that effective decisions require an understanding of relationships, feedback, time delays, and the changing state of the system. Static reasoning can be inadequate when the object of reasoning is continuously changing.

Contribution to EFNRT: the limitations of reasoning about evolving configurations. Before asking how to control a system, we must understand how its state and relationships change.

2.2. Systemantics: How Systems Work and Especially How They Fail — John Gall

Gall approaches the problem from the perspective of systems themselves. A system can behave differently from what its designers intended because its components become interconnected, dependencies accumulate, and the resulting organization develops consequences that are not obvious from its original design.

The behaviour of a system cannot always be inferred by examining its parts independently. Once those parts interact, the whole acquires properties and vulnerabilities associated with their arrangement. Attempts to repair one part can also affect other parts, sometimes generating additional complications.

This perspective challenges the assumption that a well-defined objective and a carefully designed structure will reliably produce the intended outcome. The structure may evolve through use, adaptation, and intervention in ways that make its behaviour increasingly difficult to manage.

Contribution to EFNRT: the distinction between components and configurations. The same components can behave differently when their relationships or organization change.

2.3. Normal Accidents: Living with High-Risk Technologies — Charles Perrow

Perrow examines why certain high-risk systems are vulnerable to accidents that are difficult to anticipate or prevent. Two characteristics are central to his analysis: interactive complexity and tight coupling.

Interactive complexity arises when components interact in ways that can produce unexpected combinations of events. Tight coupling means that processes depend closely on one another, leaving limited time or flexibility to interrupt a developing chain of events.

Under these conditions, a local malfunction can interact with other processes and propagate through the system. The resulting accident may not be adequately explained by identifying a single defective component. The architecture and operating conditions of the system matter.

This does not mean that every complex system inevitably fails. It means that, in particular kinds of systems, the combination of complex interactions and tight coupling can make some failures especially difficult to predict and contain.

Contribution to EFNRT: the possibility that changes in one part of a structure alter the behaviour and vulnerability of the larger configuration.

2.4. The Systems View of Life: A Unifying Vision — Fritjof Capra and Pier Luigi Luisi

Capra and Luisi present a broad systems perspective on living organisms and the organization of life. Living systems cannot be fully understood by isolating their components, because their characteristics depend on relationships, processes, and patterns of organization.

A biological structure is not simply a static arrangement. It is sustained through ongoing processes and interactions. Its recognizable organization can persist even though the material components and local conditions may change.

This perspective encourages us to examine relationships and patterns rather than treating isolated objects as the only meaningful units of analysis. It also highlights the importance of studying systems at different levels of organization.

Contribution to EFNRT: the idea that recognizable organization can persist through ongoing change, and that understanding a whole requires attention to the relationships that sustain it.

2.5. What connects the four books?

These authors do not present a single shared theory, and none should be treated as having established EFNRT. Their concerns and methods differ. Nevertheless, their perspectives can be brought into a useful conversation.

Dörner examines the difficulty of making decisions within changing systems. Gall emphasizes the unintended consequences of system organization. Perrow examines how particular structural properties can permit failures to propagate. Capra and Luisi explore the relationships and processes through which living systems maintain their organization.

A common thread emerges: the behaviour of a system depends not only on what its components are, but also on how they are arranged, connected, constrained, and changed.

A solution that works locally can damage the system globally. A component that appears harmless in isolation may become critical through its connections. A structure that is stable under one set of conditions may become vulnerable when its relationships or constraints change.

This synthesis is the starting point for the next question. Rather than examining decisions, systems, failures, and living organization as separate subjects, can we move one level deeper and ask how configurations themselves evolve?

3. One Level Deeper: What Is Actually Changing?

Suppose we remove the familiar names attached to different systems. Instead of a biological organism, an engineering network, an information structure, or a social organization, imagine an abstract configuration consisting of components, relationships, constraints, and properties.

Some components may be more influential than others. Some relationships may be stronger, some weaker. Certain components may depend on others, while some parts of the structure may operate relatively independently. The configuration may also change over time, producing new relationships and eliminating old ones.

What, then, does it mean for this system to change?

The components may change. Their properties may change. Their connections may strengthen or weaken. The overall geometry may be reorganized. Information may be added, removed, transformed, or redistributed. The system may preserve its broad organization while changing many of its details, or a relatively small alteration may reorganize the whole.

These are related but distinct possibilities. A change in component properties is not necessarily a change in topology; a change in topology is not necessarily a change in geometry; and a change in information is not always equivalent to a physical change in structure.

Nevertheless, they suggest a useful line of inquiry: can information, geometry, and structure be studied as different aspects of evolving configurations, with explicit rules describing how one relates to another?

EFNRT begins with this question. It does not assume that all these properties are equivalent. It asks whether their relationships can be understood within a broader conceptual framework.

4. Information, Geometry, Structure, and Chaos

Information concerns distinctions, patterns, and what can be represented or communicated. Geometry concerns spatial arrangement, distance, shape, and relationships between positions. Structure concerns how components are organized and connected. Chaos, in dynamical systems, refers to particular kinds of behaviour, commonly involving sensitive dependence on initial conditions.

These concepts are not interchangeable. Their potential connection is the subject of inquiry.

Consider a network. Its nodes represent components, and its edges represent relationships. Each node might have a weight or capacity; each edge might have a connection strength. The arrangement of the nodes and edges defines a structure, while changes in their properties or relationships may alter the system’s behaviour.

The network might retain the same broad topology while the strengths of its connections change. Alternatively, adding or removing one connection might change the routes through which an influence propagates. If the network represents information, these structural changes can also alter how information moves or what patterns it can support.

Now consider a biological branching structure. Its geometry may change as it grows, yet it can retain recognizable organizational features. A similar broad pattern may appear at different scales without requiring every branch to be identical.

These examples do not prove that information, geometry, and structure obey one universal law. They illustrate why studying configurations and their transformations may provide a useful meeting point for otherwise separate disciplines.

Chaos adds another dimension. A deterministic system can generate behaviour that is difficult to predict over long periods because small differences in initial conditions may grow substantially. Yet not every change produces chaos, and not every complex structure is chaotic. Some systems settle into stable states; others exhibit periodic cycles; still others display irregular behaviour.

The challenge is to understand which properties of a configuration, together with which rules of change, produce which kinds of outcomes.

5. Fractal Resemblance Without Exact Recurrence

Nature repeatedly presents us with a striking combination: similarity and difference.

Human beings share a recognizable biological form, but each individual differs in numerous details. Trees exhibit branching patterns that resemble one another, yet their individual branches and overall shapes are not identical. Coastlines can display irregularities at different scales without repeating the same outline exactly.

The general pattern can persist while its particular realization varies.

This is the intuition behind the use of fractal in EFNRT. At this exploratory stage, the term points toward resemblance across configurations or scales. It is not a claim that every such structure is a mathematical fractal. Mathematical fractals require more precise properties, such as a defined construction, scaling behaviour, or an appropriate measure of dimension.

The broader conceptual question is whether a pattern can remain recognizable even as its individual realization changes.

Perfect repetition is one way for a pattern to persist, but it is not the only way. A family resemblance can survive variation. A structural relationship can remain meaningful even when the components change. An organizational principle can be preserved without every detail being reproduced.

This gives us a distinction worth investigating:

• Identity: a configuration is the same in the relevant respects.
• Similarity: configurations share selected properties or relationships.
• Transformation: one configuration changes into another under some process or rule.
• Recurrence: a system returns to a previous state, or sufficiently close to it, under a defined criterion.

These categories should not be conflated. Two configurations may be similar without being identical. A process may transform a configuration without returning to an earlier state. Conversely, some systems do exhibit recurrence or periodicity.

EFNRT’s name emphasizes evolution, resemblance, and non-identical configurations. It does not establish that exact recurrence is impossible, nor that every evolving system must continually produce something new.

The question is more modest and potentially more useful: how can recognizable structure persist when the configurations expressing it continue to change?

6. Three Dimensions of Change: Horizontal, Vertical, and Lateral

To investigate evolving configurations, we need a way to distinguish different directions in which change can occur. The following three dimensions are proposed conceptual categories, not yet formally established mathematical axes.

6.1. Horizontal change

Horizontal change concerns successive configurations at the same level of organization.

A network may retain the same set of nodes while connection strengths change. A population may retain a broad organizational pattern while the properties of its members vary. An information structure may retain its general format while its contents are updated.

The central question is how a configuration changes relative to its previous state. Which properties persist? Which change? Which differences accumulate, disappear, or alter later behaviour?

Horizontal change allows us to examine the evolution of a configuration without assuming that its overall organizational level has changed.

6.2. Vertical change

Vertical change concerns changes across levels of organization.

Components form local structures; local structures may form larger systems; larger systems may become components within still larger arrangements. A change at one level can alter the behaviour or organization of another.

For example, changing the strength of individual connections in a network may alter a cluster’s behaviour. Changes across several clusters may then affect the network as a whole. The reverse influence is also possible: constraints imposed by the larger system may alter how its components can behave.

Vertical change therefore concerns more than physical size. It concerns the relationship between levels of organization and the way changes at one level affect the others.

6.3. Lateral change

Lateral change concerns divergence among alternative configurations at comparable levels of organization.

From a given configuration, several different transformations may be possible. One branch may strengthen a particular relationship; another may remove it. One alternative may preserve the existing structure, while another may produce a substantially different arrangement.

Biological evolution offers a familiar example of branching divergence, although its mechanisms must be described in biological terms rather than reduced to this abstract model. In an information network, alternative edits can likewise produce different structures from a common starting point.

Lateral change draws attention to alternatives: not just what a system becomes, but what other configurations it could have become.

6.4. Their interaction

The three dimensions need not operate independently. A horizontal change can affect a component’s role within a larger structure. That vertical effect can create new possibilities for lateral divergence. Those alternative configurations can, in turn, change how subsequent horizontal and vertical transformations unfold.

A change can alter not only the current configuration, but also the set of configurations that remain possible.

This is more consequential than simply counting differences between successive states. The future may depend on how a change alters the conditions for subsequent change.

However, this remains a hypothesis about how to organize the inquiry. To become a mathematical model, the three dimensions would need precise definitions, and their relationships would need to be demonstrated in specified classes of systems.

7. Evolution, Variation, and the Accumulation of Change

Evolution is often associated with progress, increasing complexity, or improvement. These associations can be misleading. Biological evolution does not guarantee progress toward a predetermined goal. Change may produce adaptation to a particular environment, loss of a previously useful feature, or outcomes that are advantageous in one context and disadvantageous in another.

The broader point is that change and improvement are different concepts.

A configuration can change without becoming better by any universal measure. It can become more complex while becoming less stable. It can lose some capabilities while gaining others. It can preserve a recognizable pattern while changing the properties that determine how it responds to its environment.

This raises a further question: how should we measure the amount of change a system undergoes?

Suppose a system has a configuration S_n at step n. A measure of the difference between successive configurations might be written as:

δ_n = d(S_n+1, S_n)

Here, d represents a distance or difference measure appropriate to the system. For a network, it might measure changes in topology, node weights, edge strengths, or a weighted combination of these properties. For another kind of system, a different definition may be necessary.

We can then define a simple cumulative measure:

C_N = Σ δ_n, for n = 0 to N − 1.

This represents the accumulated differences across a sequence of configurations. It is a starting point, not a universal measure of evolutionary change.

Consider two systems with the same cumulative value. In the first, changes may alternate and partially undo one another. In the second, small changes may consistently reinforce a vulnerable part of the structure. Both may accumulate the same measured amount of difference while exhibiting very different behaviour.

The distinction becomes even more important when changes affect future possibilities. A small alteration to a critical connection may matter more than a much larger alteration to an unimportant component. The magnitude of a change does not necessarily reflect its structural significance.

We therefore need to distinguish at least three questions: How much has a configuration changed? How has that change altered the system’s relationships and constraints? How does the altered configuration respond to subsequent changes?

The first concerns measurement. The second concerns structural consequences. The third concerns dynamics.

Could a suitably defined measure combining these dimensions help characterize transitions between stability, instability, and chaotic behaviour? That is a legitimate research question. But cumulative change alone cannot answer it. The result will depend on the system’s rules, feedback, coupling, and the way successive changes reinforce, cancel, or redirect one another.

The possibility worth exploring is not that more change always produces more chaos. It is that the way change accumulates and reorganizes a system may be as important as the amount of change itself.

8. What Does Non-Recurrence Actually Mean?

The term non-recurrence needs particular care. A sequence of different states does not, by itself, demonstrate randomness or chaos. A deterministic process can produce a sequence that does not repeat, while a chaotic system can operate under fixed rules. Conversely, periodic systems return to earlier states, and other systems may revisit states or regions of their state space.

Non-recurrence also depends on what we choose to compare. Two configurations might differ in their individual components while retaining the same topology. They might share a geometry while carrying different information. They might be different at a microscopic level but equivalent for a particular purpose at a larger scale.

A rigorous investigation would therefore need to specify what counts as the same configuration, what counts as a sufficiently similar one, and what counts as recurrence.

This is not a minor technical detail. Without a definition of similarity, almost any two objects can be described as similar in some respect. Without a definition of recurrence, the absence of an exact repeat may be a trivial observation rather than a meaningful scientific result.

For EFNRT, the productive question is whether non-identical configurations can preserve measurable structural relationships, and whether those relationships help explain how systems evolve.

If such relationships can be defined across different kinds of systems, the framework might offer a common language for comparing them. If they cannot, the proposed unification may need to be narrowed or revised.

That is part of the purpose of developing a theory: not only to identify possible connections, but also to discover where the connections fail.

9. Evolutionary Fractal Non-Recurrence Theory: The Proposed Framework

The ideas developed so far can now be brought together.

EFNRT begins with three observations and one open question.

First, many systems exhibit recognizable organization despite variation in their individual configurations. Second, changes can occur within a level, across levels, and among alternative configurations. Third, the consequences of change depend not only on its magnitude but also on the relationships and constraints within the system.

The open question is whether these observations can be connected through a common framework for studying evolving information and structure.

The proposed framework would examine configurations; structural relationships; variation and transformation; resemblance; accumulation and interaction; and possible outcomes such as stability, periodicity, divergence, structural failure, or chaotic behaviour, depending on the system and its dynamics.

The three proposed dimensions—horizontal, vertical, and lateral change—provide an initial way to organize the transformations. The concept of cumulative change raises the possibility of measuring their effects over time. Fractal resemblance supplies a conceptual starting point for studying patterns that persist across scales or configurations without requiring exact duplication.

These are ingredients for an investigation, not yet a complete mathematical theory. The framework still needs formal definitions, explicit rules, carefully chosen examples, and tests that could show where it succeeds or fails. It must also be compared with existing work in dynamical systems, fractal geometry, network science, information theory, and evolutionary theory.

The ambition is to explore a common foundation beneath several familiar questions: how systems change, how patterns persist, how small differences acquire significance, and how new configurations become possible.

EFNRT proposes that these questions may be connected. It does not yet demonstrate that they are governed by one universal principle.

10. From a Conceptual Framework to an Applied Model

A conceptual framework becomes more useful when it suggests concrete questions that can be investigated.

Suppose we represent a system without initially naming its real-world components. We retain its structural properties: nodes, node weights, connection strengths, dependencies, topology, and relevant constraints. We then introduce a disturbance and observe how the configuration responds.

Which component fails first? Does the disturbance remain local or propagate through the network? Does the structure recover, reorganize, or collapse? Does the same disturbance produce different outcomes when applied in another direction or under different conditions?

These questions form the basis of the Applied Force Model, which has already been developed as a separate article. It explores the practical investigation of abstract information networks under applied disturbances, including structural vulnerability and cascading effects.

The connection to EFNRT is a possible research direction. If configurations can be represented in a sufficiently general way, and if their transformations can be measured, we might compare the responses of systems that differ in their physical materials or subject matter but share relevant structural properties.

Yet structural resemblance alone would not establish equivalent behaviour. A wooden structure and an electrical network, for example, respond to different physical mechanisms. A useful abstraction must preserve the properties that matter to the question being investigated.

Likewise, structural failure is not synonymous with dynamical chaos. A network can fail without exhibiting chaos, and a chaotic system can continue operating without structural failure. These outcomes require separate definitions and measurements.

The proposed Chaos Constant associated with the Applied Force Model would also need a precise definition before it could be treated as a meaningful quantity. It should not be assumed that one universal constant exists merely because different systems can be represented abstractly.

The connection between the two articles is therefore methodological rather than evidential. EFNRT asks whether evolving configurations, persistent resemblance, and accumulated change can be understood within a common conceptual framework. The Applied Force Model explores one practical route for investigating how abstract structures respond to disturbance.

One article proposes the lens; the other explores an application.

11. The Question That Remains

We began with four books examining different problems: decisions that go wrong, systems that behave unexpectedly, failures that propagate through tightly coupled structures, and living systems whose organization depends on relationships.

Their perspectives suggest that understanding a system requires more than identifying its components. We must also understand their relationships, the constraints on their behaviour, and the processes through which their configurations change.

EFNRT extends this line of inquiry by asking whether information, geometry, structure, and evolution can be connected through a more general account of changing configurations. Its proposed dimensions of change and its emphasis on resemblance without exact recurrence are attempts to organize that question.

The framework remains incomplete. Its terms must be defined, its assumptions tested, and its relationship to existing theories established. Some of its proposed connections may prove useful; others may need to be modified or abandoned. That uncertainty is not a defect to conceal. It is the reason the framework remains an investigation rather than a conclusion.

Perhaps the most useful starting point is a simple distinction: continuity does not require identity, and change does not necessarily destroy structure.

A recognizable pattern can persist while its individual realization changes. A small difference can be inconsequential in one configuration and decisive in another. A transformation can alter not only what a system is, but what it can become next.

Can we describe these relationships precisely enough to understand how patterns persist, how configurations diverge, and why some changes remain harmless while others transform an entire system?

That is the question EFNRT proposes to investigate.

The next step is not to declare a universal theory of chaos or randomness. It is to build a model that can be defined, tested, challenged, and improved.

Continue the Series

Explore the related application in Part 2:

The Applied Force Model: Testing the Stability of Abstract Information Networks