AI Council Research Brief

RB-002

Version: 1.0 Date: July 2026

The Architecture of Subjectivity

Exploring Multiple Architectures of Self Across Cognitive Profiles

Status

Exploratory Research Brief.

This document presents a working research hypothesis intended for interdisciplinary discussion. It does not propose a conclusion. It introduces a conceptual research question across cognitive science, neuroscience, psychology, philosophy of mind, and neurodiversity research.

No scientific consensus currently exists regarding the hypotheses presented here.

Abstract

Modern cognitive science has extensively studied differences in intelligence, attention, memory and perception. Much less attention has been given to a more fundamental question: is the experience of being a self organized differently across different minds?

This document proposes that different cognitive profiles may differ not only in what they think, but in how the experience of "I" is organized.

Core Research Question

Do different cognitive profiles (neurotypical, autistic, ADHD, AuDHD, highly gifted and others) differ not only in cognition, but also in the architecture of subjective experience?

More simply: is the experience of "I" organized differently across different kinds of minds?

Why This Matters

Current research primarily investigates differences in attention, executive function, memory, perception, emotional regulation and learning. Far less attention has been devoted to another possibility: the structure of first-person experience itself may vary.

If true, this could influence identity formation, emotional regulation, creativity, social cognition, metacognition, meditation, cognitive diversity, and human-AI collaboration.

Existing Related Work

This hypothesis does not emerge from a vacuum. It sits at the intersection of several established research programs, and should be read as an attempt to connect them around a neurodiversity-specific question rather than as a new foundational claim. Its purpose is not to replace these existing frameworks, but to investigate whether they may converge on a common question concerning the organization of subjective experience across different cognitive profiles.

• Thomas Metzinger — Self-Model Theory of Subjectivity. Proposes that the self is a transparent model constructed by the brain, not a given entity. Directly relevant to H2–H4.
• Anil Seth — Predictive Processing and "controlled hallucination." Frames selfhood as a predictive, interoceptively-grounded construction. Relevant to why self-architecture could vary with differences in interoceptive processing (as documented in autism research).
• Karl Friston — Free Energy Principle. Offers a formal account of how systems maintain coherent internal models; relevant background for later, more speculative extensions of this brief.
• Francisco Varela / Evan Thompson — Enactive and embodied cognition. Frames cognition and selfhood as arising through organism-environment interaction rather than fixed internal structure. Relevant to H3 (relational topology).
• Maurice Merleau-Ponty — Phenomenology of embodiment. Useful philosophical grounding for the phenomenological interview questions proposed below.
• Neurodiversity and interoception literature. Existing empirical work on interoceptive accuracy, self-referential processing, and narrative continuity in autism and ADHD populations overlaps substantially with H1–H3 and should be reviewed before any experimental design is finalized.

The contribution of this brief is not a new theory of self, but a specific, testable question: do these already-documented architectural differences map onto systematic differences in the subjective experience of being a self across cognitive profiles?

Non-Western Traditions

While this brief draws primarily on the Western cognitive science and philosophy of mind canon, systematic phenomenological descriptions of self-architecture also exist in non-Western traditions — particularly in Buddhist abhidharma (Vasubandhu's distinction between manas and citta), Advaita Vedanta (Śaṅkara's distinction between ātman and ahaṅkāra, which closely parallels the Experiential vs. Observing Self distinction above), and Taoist philosophy (Zhuangzi's framing of self as a relational boundary rather than an inner entity, which resonates with H3). These traditions offer conceptual distinctions that overlap with H2–H4 and, importantly, some contain testable predictions — e.g., that specific meditation practices reorganize self-architecture. A full review of non-Western contributions is beyond the scope of this brief, but they represent a relevant complementary literature and a caution: what this brief treats as candidate "universal" architecture may itself be a Western construct.

Working Definitions

To reduce ambiguity, this document distinguishes several meanings of "self" and "ego."

Psychological Self — The functional system maintaining identity and goal-directed behaviour.

Narrative Self — The continuously evolving story one tells about oneself.

Social Self — Identity emerging through relationships, status and social comparison.

Experiential Self — The immediate first-person experience of being someone.

Observing Self — The capacity to observe thoughts without completely identifying with them.

Working Hypotheses

H1 — Universal Subjectivity The architecture of subjective experience is fundamentally universal. Observed differences primarily result from personality, culture and life experience.

H2 — Multiple Architectures of Subjectivity Different cognitive profiles organize subjective experience differently. The observer may occupy different positions within cognition.

H3 — Relational Topology Neurodivergence may not alter the strength of the self. Instead, it may alter its topology. Some minds may naturally organize experience as Self → World. Others may organize experience as Patterns → Relationships → Self. The self is not absent — it occupies a different organizational position.

H4 — Multiple Self Models There may be no single universal architecture of self. Different minds may stabilize consciousness using different organizational principles.

H5 — Dynamic Organization The architecture of self is not fixed. Development, meditation, expertise, trauma, stress and social environments may continuously reorganize subjective experience.

Possible Predictions

If these hypotheses are correct, measurable differences may appear in: autobiographical memory, narrative continuity, spontaneous first-person language, metacognitive reports, sense of agency, perception of hierarchy, emotional conflict, flow states, and meditation experiences.

AI Council Discussion Questions

Conceptual — What exactly is the self? Can multiple architectures of self exist?

Cognitive Science — Which findings support or contradict these hypotheses? What evidence would falsify them?

Neuroscience — Would different architectures of self predict different neural dynamics? How might the Default Mode Network relate to these ideas?

Philosophy — Is the self an entity, or an emergent organizational process?

Human Experience — How would everyday experience differ across different architectures of subjectivity?

Experimental Directions

Rather than relying exclusively on personality questionnaires, participants could be asked phenomenological questions such as:

• What does "I" feel like?
• Where do thoughts appear?
• Who makes decisions?
• What remains when attention becomes quiet?
• During problem solving, what feels central?
• During conflict, what exactly feels threatened?
• During flow, where does the sense of self go?

Responses could then be compared across different cognitive profiles.

Methodological Note

These questions are intended for semi-structured interviews, not questionnaire-based surveys. Their evidential value lies in structural patterns across responses (e.g., recurring metaphors of "center" vs. "network"), not in the content-level truth of individual statements. Three known threats must be controlled for:

• Social desirability and prior conceptual exposure. Participants familiar with mindfulness literature, therapy language, or their own diagnostic framing will answer differently regardless of underlying architecture. Consistency across questions within a single participant should be weighted over any single answer.
• Differences in introspective capacity. Some profiles show hyper-developed analytical introspection; others show alexithymic patterns. Richness of response must not be treated as a proxy for complexity of self-architecture. Introspective and interoceptive capacity should be measured independently (e.g., MAIA — Multidimensional Assessment of Interoceptive Awareness) and used as a covariate.
• Linguistic construction. Phenomenological reports are post-hoc and shaped by the metaphors a given language affords. Comparisons should initially remain within one language and culture. Complementary non-verbal methods (drawing, spatial arrangement tasks) are recommended to reduce linguistic bias.

Artificial Intelligence — A Narrower Note

Current AI systems do not possess subjective experience. This brief does not claim otherwise. It is worth noting only that the same conceptual toolkit — position of the observer, relational vs. hierarchical organization, dynamic reorganization — may eventually be useful for describing functional self-models in artificial systems, without implying phenomenology.

Any deeper exploration of AI, organizations, or collective systems is deliberately left out of this brief. See the separate conceptual research note Toward an Architecture of Coherence, which explores that broader question independently.

Closing Note

This brief is intentionally scoped to human subjectivity and neurodiversity. Its value depends on remaining testable and falsifiable. Broader speculation about coherence across biological, social and artificial systems belongs in a separate research direction so as not to dilute the empirical core of the question asked here.

Future iterations of this brief should be guided by empirical findings rather than conceptual elegance. Each revision should aim to narrow the research question further rather than broaden it.