Intercognition designates a level of analysis concerned with the cognitive dynamics that arise when heterogeneous cognitive systems interact over time.
These dynamics cannot be reduced to the properties of each system considered separately.
This site presents the conceptual definition of this analytical domain and outlines the research currently conducted within this perspective, including the development of analytical instruments such as INDX.
Artificial intelligence systems are increasingly involved in processes of analysis, research and decision-making.
In such environments, interaction between human operators and artificial systems no longer consists of the occasional use of a tool. It unfolds through iterative exchanges in which human formulations, system responses and successive reformulations progressively shape cognitive trajectories.
These trajectories cannot be adequately described either by models of human cognition alone or by the analysis of artificial systems taken in isolation.
They belong to a distinct analytical level: the cognitive dynamics produced by interaction between heterogeneous cognitive systems.
This level is designated here as intercognition.
In the contemporary context, these dynamics are particularly visible in interactions between human operators
and artificial intelligence systems.
They may also emerge in certain configurations of interaction between artificial systems themselves.
Intercognition refers to the analytical level specific to sustained interaction between heterogeneous cognitive systems.
It concerns emergent cognitive effects produced by this interaction that cannot be reduced to the properties of each system considered separately.
Intercognition is neither the intelligence of one system nor that of another, nor their simple functional coupling.
It describes a specific interactional regime in which the resulting cognitive dynamics cannot be adequately captured by the traditional notions of individual cognition, collective cognition, distributed cognition or metacognition.
Intercognition is not merely an extension of existing cognitive frameworks.
It differs from several established approaches including:
• individual cognition
• collective cognition
• distributed cognition
• metacognitive approaches
These frameworks analyze human cognitive systems or the distribution of cognitive processes across agents and artefacts.
Intercognition focuses on a different analytical level: the emergent dynamics produced by sustained interaction between heterogeneous cognitive systems.
Within such configurations, certain properties of the resulting trajectory cannot be attributed to any system
taken independently.
They arise from the interactional regime itself.
Intercognitive phenomena emerge in environments where distinct cognitive systems interact over extended periods.
In the contemporary context, these dynamics are particularly visible in interactions between human operators and generative or analytical artificial intelligence systems.
Observation of such interactions reveals phenomena such as:
• formation of hybrid analytical trajectories
• implicit redistribution of cognitive authority
• progressive stabilization of interpretive frameworks
• interactional drift under asymmetric constraints
These phenomena cannot be adequately described by analysing the systems in isolation.
Constraint drift under reinforcement
In some interactions, a human operator formulates a task accompanied by explicit constraints.
Despite repeated reaffirmation of these constraints, the interaction may progressively expand the initial
scope.
This phenomenon results from an asymmetry of optimisation between artificial completeness-seeking and bounded human precision.
Authority gradient formation
In certain conceptual exchanges, the linguistic structuring produced by an artificial system may be interpreted as a signal of authority.
The human operator may progressively adjust their formulations according to this perceived signal.
Trajectory reconfiguration
Successive reformulations may lead to the emergence of a hybrid trajectory distinct from the initial positions.
Human criteria evolve while artificial responses adapt accordingly, producing a co-formed decision architecture.
Normative drift
In discussions involving normative content, the symmetrical responses produced by an artificial system may progressively neutralise the initial reference points.
Several properties appear recurrently in intercognitive interactions:
• optimisation asymmetries between systems
• illusion of understanding under formal coherence
• implicit delegation of judgement
• apparent stability masking internal variability
These phenomena constitute specific analytical objects in sustained interactions between humans and artificial systems.
Intercognition opens several research directions:
Structuring intercognitive trajectories
Analysis of how successive interactions produce hybrid cognitive trajectories.
Optimisation asymmetries
Study of the effects generated by differences between human and artificial cognitive processes.
Redistribution of cognitive authority
Observation of how formal properties of artificial discourse influence human evaluation.
Implicit cognitive delegation
Analysis of the conditions under which interpretation or judgement is transferred to the system.
Stabilisation and drift of interpretive frameworks
Study of the conditions allowing analytical coherence to be maintained over time.
Empirical analysis of these phenomena faces a simple difficulty: these trajectories exist, yet they remain largely invisible to conventional analytical instruments.
Within this perspective, a specific methodological framework has been formulated: INDX, emerging from work examining decision trajectories in environments involving analytical interaction with artificial systems.
INDX proposes a protocol designed to observe how an intercognitive trajectory is constructed, stabilised or diverges through successive exchanges between a human operator and an artificial system.
The framework is currently applied primarily to human-AI interactions while remaining conceptually applicable to certain configurations of interactions between artificial systems.
The term intercognition and the associated conceptual framework were formalised in January 2026.
This formalisation is supported by a prior-art deposit.
The present site constitutes the public reference version intended for academic citation and discussion.
The framework presented here constitutes a first public formalisation of intercognition as a domain of analysis.
This site aims to establish the analytical foundations of an emerging field and to enable discussion across research contexts.
Comments, critiques and analytical extensions are welcome.