NOTE-1498: Metzinger — Minimal phenomenal experience

Contribution

The paper names a target, minimal phenomenal experience (MPE), and a strategy, explanation by minimal models. The target is chosen because the contemplative reports describing it claim “minimal dimensionality” together with “maximal prototypicality” (pp. 2–3, n. 1). The paper gives a first six-dimensional concept of MPE and a hypothesis about what MPE represents: the organism’s own tonic alertness. Its evidence is quoted report and exegesis. It promises psychometric work, which became Metzinger 2024 (LIT-1943).

Key insight

What an “empty” consciousness represents is the capacity to know. Tonic alertness, “a graded functional property determining the capacity for sustained attention”, can exist without being represented, as in unresponsive wakefulness. When it is represented, the result is wakefulness, and “Consciousness is knowing that one is alert” (p. 33). The paper makes this model the minimal case of experience. It is a model of the system, but not of a self as subject: “MPE is non-egoic self-modelling” (p. 36).

Assumptions

• Naturalism, assumed “without having given an independent argument” (p. 7, n. 2).
• Representationalism. The reported contentlessness may itself be an appearance, and if there is a content “that can account for the reported experience of contentlessness, then Representationalism remains tenable. This is the line I will be pursuing” (p. 9).
• Probabilistic concepts. “Consciousness” has “not definitional but probabilistic structure”. MPE is “a cluster concept” referring to a “phenomenological prototype”, and membership is graded (p. 3, n. 1).
• Report as data, provisionally. Kriegel is quoted: “It is nonetheless open to the theoretician to challenge the practitioner’s own description of the state she is in” (p. 9). The three methodological obstacles are named in footnote 27 (p. 38): embodied theory contamination, the lack of a positive characterisation, and performative self-contradiction. They are discussed elsewhere (Metzinger 2019, not held), not here.

Key results

• From MPS to MPE (p. 2). Blanke and Metzinger’s minimal phenomenal selfhood (LIT-1274) has three features: identification with the body as a whole, spatiotemporal self-location and a weak first-person perspective. MPE is hypothesised to lack MPS and all other egoic self-consciousness, time representation and a spatial frame.
• Six constraints (pp. 7–23). PC1, Wakefulness, is “the phenomenal character of tonic alertness” (p. 9). PC2, Low Complexity, is “often described as the complete absence of intentional content” (p. 9). PC3 is Self-Luminosity, culturally variable and perhaps a metaphor for PC1 (pp. 17–20). PC4, Introspective Availability, PC5, Epistemicity, “the phenomenal experience of knowing … the subjective quality of confidence”, and PC6, Transparency/Opacity, are drawn from Moore’s “Refutation of Idealism” (pp. 20–23).
• The negative profile (pp. 10–11). Non-sensory, non-motor, atemporal, non-cognitive, non-egoic (“no localized unit of identification”), unbounded, and “aperspectival: no epistemic agent model”.
• No phenomenal oneness (p. 13). The space “has neither centre nor periphery and … cannot be introspectively individuated to create indivisible units of experience – not even a single one”. It “lacks the phenomenal quality of singularity or numerical identity”, so “non-oneness” is added to “non-duality”.
• No phenomenal contrast (p. 13). “The method of phenomenal contrast … cannot be applied to the phenomenology of tonic alertness”: without wakefulness “there simply is no conscious experience altogether”.
• Not remembered (p. 15). “The content of a full-absorption episode cannot be reported, because the self-referential mechanisms of forming an autobiographical memory are suspended”. Only “the process of entering into it or of emerging out of it” can be represented in the autobiographical self-model.
• The epistemic-space thesis (pp. 26–27). MPE “is neither an internal model of a self-as-subject, nor a model of an object”. The epistemic subject “simply is the organism or conscious system as a whole”. MPE is “an internal model of an epistemic space currently opened by it”.
• Transparency for a contentless state (pp. 30–31). If MPE had no intentional content and were transparent, “there would be nothing introspective attention could ever discover”. If it were opaque, PC1 and PC3 would report “properties of the physical carrier itself”, and some reports “may actually be reports about an entirely non-intentional process”. The theory sketched: “‘Appearance’ results from a transparent representation of meta-confidence” (p. 31).
• Three levels (pp. 32–33). Arousal is a physical property, tonic alertness a functional one and wakefulness a phenomenal one. Tonic alertness “if it is not represented or internally modelled, can exist without the conscious experience of wakefulness (as in some cases of unresponsive wakefulness syndrome)” (p. 33). Tonic alertness “could be realized by very different physical properties in machines that create an integrated internal model of their own epistemic space as such” (p. 32).
• Conclusion (pp. 35–38). MPE is “an inner representation of tonic alertness”. It is “non-egoic self-modelling: as such it is atemporal, selfless, and not tied to an individual first-person perspective” (p. 36). It “is aperspectival and it does not instantiate MPS, but it definitely has representational content” (p. 38). “Phenomenality is a transparent, temporally thin, and global representation of subjective confidence, relative to the abstract possibility of knowledge” (p. 38). Footnote 24 (p. 36) states that MPE is a counterexample to the “projective geometry” model of consciousness, since it is aperspectival.

Claims

id	claim	strength	support
C1	Experiences reported as “pure consciousness” exist	moderate as a claim about reports; weak as a claim about experience	sixteen quoted reports and texts (p. 38)
C2	MPE lacks egoic self-consciousness, time, a spatial frame and a first-person perspective	weak: phenomenological inference from retrospective reports, whose methodological problems are deferred (n. 27)	§2.1, pp. 9–15
C3	MPE has representational content: it is a model of the organism’s tonic alertness or epistemic space	weak: a hypothesis, argued by fit with the constraints and from naturalist and representationalist premises	§§2.2.2, 3, 4
C4	Tonic alertness can exist without wakefulness	moderate: unresponsive wakefulness is cited; whether those patients lack experience is itself inferred from behaviour	p. 33; Laureys et al. 2010
C5	MPE lacks phenomenal oneness	weak: conceptual analysis of “unbounded” in reports	p. 13
C6	Full-absorption episodes cannot themselves be remembered	weak: argued from the absence of a self-model in the episode, not shown	p. 15

Concepts

• Minimal phenomenal experience (MPE): the simplest conscious experience. The term is Windt’s (2015), and the paper uses it more widely (p. 2).
• Minimal model: a model including “only the core causal factors giving rise to the target phenomenon” (p. 3).
• Full-absorption episode: a state at maximal Wakefulness and minimal Low Complexity, after which only awareness can be reported (p. 14).
• Epistemic agent model (EAM): “a conscious internal model of the self as actively selecting targets of knowledge … the origin of our consciously experienced first-person perspective” (p. 11, n. 7).
• Unit of identification (UI): “whatever form of experiential content leads to phenomenological reports of the type: ‘I am this!’”. “If a conscious system has no UI … it has no sense of self” (p. 10, n. 6).
• Epistemic space model (ESM): MPE as an integrated model of the organism’s set of epistemic capacities (p. 27).

Connections

It builds on Blanke and Metzinger 2009 (LIT-1274) and on Windt’s coinage of MPE. It draws on Moore 1903 for transparency, and on Forman, Stace, Shear and Tibetan sources for the reports. LIT-1354 (Chandia-Jorquera et al.) adopts its definition of pure awareness. Woods, Windt and Carter (LIT-1351) test its premise that such goal states are contentless and find them not to be. Its active-inference reading is shared with Laukkonen et al. (THEORY-882). Metzinger 2024 (LIT-1943) is its sequel, with the survey.

Bearing on the record’s own argument

• THEORY-063 (unity without a self-model). This paper replaces the précis’s “Selfless Snapshot” with a minimal case that is selfless and not experienced as one. It has no “phenomenal quality of singularity” (p. 13). It is still a model, and a self-directed one: “non-egoic self-modelling” (p. 36). For THEORY-063 the bearing is twofold. (a) If MPE is real, then on Metzinger’s own later account there is consciousness that is not the integration of “one global world-model”, since MPE models the system’s capacity to know and not a world. So the first clause would be at best an account of ordinary waking experience, not of consciousness as such. (b) The second clause would hold only in the narrower form that phenomenal experience needs no egoic self-model. The inference is the record’s. The paper does not mention the globality constraint.
• QUESTION-048 (what each kind of evidence can establish). The paper argues that MPE’s own content is unreportable from memory (p. 15) and closed to the method of phenomenal contrast (p. 13). If both are right, its prototype is closed to the two standard routes of report-based evidence. It defers performative self-contradiction to other papers (n. 27). It also gives one dissociation the record can use. Tonic alertness without wakefulness, in unresponsive wakefulness, is a case in which a function is present and, on the paper’s reading, experience is absent (p. 33). The reading of experience as absent rests on behaviour, which is QUESTION-048’s problem again.
• QUESTION-254 and CLAIM-412 (what makes an organization experiential). The paper’s answer is a specific organization: a predictive model of the system’s own epistemic capacities, temporally thin and integrated. Without it, alertness is not experienced (p. 33). That is a candidate for CLAIM-412’s missing qualifying condition, in a form close to TERM-073’s interpretation (distinction, context, consequentiality). It is aperspectival, which CLAIM-412’s part 3 does not allow as worded. It is not stated independently of report, so it does not meet CLAIM-411’s defeater either.
• QUESTION-255 (non-gluing and unity). If MPE has no centre, no boundary and no oneness (p. 13), it is a case where the question of unity, glued or not, may not arise in phenomenal terms. One cannot ask whether local interpretations glue when there are no local interpretations. That is the record’s inference. It marks a limit of the sheaf framing, not a case of kind (2).
• TERM-082 (senses of consciousness). Its arousal, tonic alertness and wakefulness triad (pp. 32–33) is a three-way split of TERM-082’s sense 4, which runs the physical, the functional and the phenomenal together.
• TERM-045 (self, and which sense). “Non-egoic self-modelling” is a case for the “modeled” sense that TERM-045’s note of 2026-10-18 flags. It also shows that the modeled sense needs a further division, between a model of a self as subject and a model of the system’s capacities.
• THEORY filed. THEORY-1299, with Metzinger 2024.
• No instruction for machine-learning practice.

Limitations

• No new data. The sixteen case studies are quotations from published reports and from traditional and teaching texts. The author grants that “there are actually fewer paradigmatic reports of such experiences in the literature than previously assumed” (p. 38).
• The methodological problems are deferred (n. 27). The paper does not answer the performative worry, which is THEORY-063’s.
• A tension between PC2 and the conclusion. Low Complexity is first read as “the complete absence of intentional content” (p. 9), and the conclusion says MPE “definitely has representational content” (p. 38). The paper resolves this by treating the content as abstract (epistemic capacity) and the reported contentlessness as an appearance (p. 9). The resolution is a choice of theory, not a finding.
• Machine realisation is asserted as conceivable (p. 32), with no argument for what would make a machine’s model of its own epistemic space conscious rather than merely present.

Open questions

• Is there a report-independent marker of a represented tonic alertness, as opposed to tonic alertness itself? The paper’s account needs one to be testable on anyone not reporting.
• If full-absorption episodes cannot be remembered (p. 15), what are the reports of them reports of? The answer the paper implies, entry and exit, would make all the evidence evidence about transitions.