Research
The research program begins from one general idea developed in the Foundation:
System-level properties depend not only on what components exist, but on how they are causally organised, how that organisation operates through time, and how past consequences alter future organisation.
Once that claim is taken seriously, several more specific research questions follow.
This page is a map of those questions. The papers, essays, and concept pages below provide the detailed arguments.
One program, several levels of question
A compact way to see the current program is:
components and interactions
→ stable relations and organization
→ active organized dynamics
→ system-level capability or property
→ interaction with an environment
→ consequences and retained history
→ possible change in future organization
Different parts of the research program focus on different links in this chain.
1. Connection Theory — how does organisation generate a system?
Connection Theory (CT) is the parent framework.
It asks how interactions become feedback-bearing histories, how some influences stabilise into connections, how connections are coordinated into organisation, and how active organisation generates system-level dynamics and properties through time.
The current framework also separates active manifestation, retained capability, organisational continuity, and current applicability rather than treating persistence as a single question.
Current open problems include:
- how selective coordination should be characterised beyond a connection inventory;
- how organisational evidence is conditioned by the organisation that generates it;
- how update order and non-commutativity affect later state;
- where capability-relevant persistence state is localised or distributed;
- how lineage continuity should be tested across interruption;
- and how the framework can be formalised into stronger dynamical and empirical models.
Read the Connection Theory overview
Read the formal paper
2. Intelligence — what are we explaining when we call a system intelligent?
Once system-level capability is treated as an outcome of organised dynamics, intelligence becomes a more precise scientific question.
The current working view does not use intelligence as a primitive causal substance. It asks what lower-level organised dynamics generate learning, planning, flexible response, environmental transfer, and other behaviours that observers collectively classify as intelligent. AI remains the conventional technological label; intelligence is the theoretical attribution that requires separate justification. Read the terminology bridge.
This immediately creates several distinctions:
- stored capability is not the same as active adaptation;
- regulation is not the same as consequence-sensitive change;
- variable behaviour is not necessarily intelligent behaviour;
- performance in one environment does not justify unlimited generalisation;
- and intelligence attributed to a higher-level system requires evidence that the relevant adaptive behaviour is generated by that system's organisation rather than merely imported from one component or an external operator.
The public essay develops this argument at the research level without specifying protected engineering implementations. A separate bridge essay clarifies a point that is easy to miss: adaptation does not require foresight or permanent correctness; it requires informative consequences to be able to change later response while assessments remain open to revision.
Read: What Is Intelligence?
Read: Adaptation Is Not Foresight
3. Life, continuity and death — what persists when living dynamics change or stop?
Living systems make the distinction between active dynamics and retained organisation unavoidable.
The life and continuity work asks:
- what it means for life to be actively realised as an organised process;
- how a living boundary is generated and maintained rather than merely drawn geometrically;
- what makes a later organism the same organism after repair, turnover, dormancy, or bounded interruption;
- how recoverability differs from current activity;
- and how terminal loss should be defined relative to a specified lineage and intervention set.
Biology therefore functions here not merely as an analogy, but as a demanding reality test for a general theory of persistent organisation.
Read the Life & Continuity series
4. Flow Shaping Principle — how should consequences be connected to later change?
The Flow Shaping Principle (FSP) narrows the problem from general organisation to active updating.
If a system changes after a consequence, the key issue is not simply whether the later outcome improved. It is whether the observed consequence is causally informative about the configuration or pathway that should be changed.
FSP therefore asks whether consequence-sensitive updating should preserve correspondence between the evidence produced by an active configuration and the later pathway-specific change attributed to that evidence.
The proposal is explicitly falsifiable and remains under empirical development.
5. Evolutionary Relativity — how does history constrain the present?
If past causal history becomes retained in current organisation, then present state alone may not fully determine how a system should be interpreted.
Evolutionary Relativity (ER) develops this historical dimension. It asks how organisational path, observer scale, and reconstruction frame affect claims about history-dependent adaptive systems.
Two systems can look similar now while differing in the causal histories that generated their present organisation. Conversely, the same historical process can appear different when observed at different resolutions or timescales.
Read the Evolutionary Relativity paper
6. Effective system capability and collective function — what level of claim is actually supported?
A recurring danger in complex systems is to observe rich microscopic mechanisms and then infer a stronger macroscopic capability than the evidence warrants.
The canonical general term used on this site is Effective System Capability: what an organised system can reliably cause or accomplish under specified conditions. The methodological work From Microscopic Capability to Collective Function uses collective function more narrowly for a candidate system-level function whose dependence on interacting components is being tested.
Its evidential ladder separates mechanism availability, collective dynamical difference, response-class expansion, task-specific value after cost accounting, and stronger irreducibility claims. A claimed organisation-level capability should survive simpler competing explanations before it is attributed to organization.
This also supports a stronger regime hypothesis. Organization should add net capability when task demands for differentiated capability, information, verification, persistence, or coordinated action exceed what one node can efficiently supply by more than the coordination cost. When one node already covers the task, or when the organization is redundant or badly matched, organization can add little or make the result worse. The scientific problem is to characterize those boundaries before observing the outcome.
Read the collective-function attribution paper
Research Notes — empirical and methodological bridges
Research Notes are short scholarly notes that connect questions from the Connection Theory research program to specific empirical literatures, recent findings, methodological problems, and active research debates. They are narrower and more literature-focused than the Research Essays, while remaining distinct from formal Papers.
Research Note · September 2026
From Coordination Topology to Organization
A Cross-Domain View of Collective Problem Solving in Humans and AI
A cross-domain comparison of seventy-five years of human collective problem-solving experiments and recent multi-agent AI results, leading from node capability and topology to operating organization.
How the pieces fit together
The current program can be read as a sequence:
Connection Theory asks how organisation forms, persists, and produces system-level dynamics.
What Is Intelligence? asks when such dynamics justify a higher-level attribution of intelligence.
Life & Continuity tests the same framework against living organisation, persistence, recoverability, and identity.
FSP asks how consequences should be related to later adaptive change.
Evolutionary Relativity asks how retained history and observation frame affect the system we reconstruct.
Effective System Capability / Collective Function asks how strong a macro-level attribution the available evidence actually supports.
These are not presented as a completed unified theory. They are connected research problems generated by the same causal-dynamical starting point.
Current status
The program is at different stages across its components:
- several conceptual distinctions and public definitions are now stable enough to be used consistently across the site;
- the core papers provide formal or semi-formal frameworks and explicit failure boundaries;
- some empirical predictions and measurement strategies remain pre-validation;
- mathematical formalisation is incomplete;
- and canonical cross-domain experiments remain an open priority.
The intended standard is not internal coherence alone. The framework should ultimately survive discriminating comparison with competing explanations and produce useful predictions or designs that would not follow from a looser claim that “organisation matters.”
Continue through the research site
- Research Essays — accessible treatments of specific problems.
- Research Notes — short literature, empirical, and methodological bridges.
- Research Series — ordered reading paths through organisation, persistence, adaptation, life, and applied organisational questions.
- Papers — the current formal and supporting papers.
- Foundation — return to the guided derivation of the framework.