Source
Metzinger (2020), LIT-1946, §§2–4, pp. 9–38, read in full in NOTE-1498. Metzinger (2024), LIT-1943, chapters 4, 25 and 34, pp. 36–44, 300–306 and 455–478, read in NOTE-1499.
What was actually shown
The three levels. Arousal is “a graded physical property of the human brain”. Tonic alertness is “a graded functional property determining the capacity for sustained attention”. Wakefulness is “a graded phenomenal property”, “a representation of tonic alertness” (2020, pp. 32–33; 2024, pp. 36–37). 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)”. “Consciousness is knowing that one is alert” (2020, p. 33; 2024, p. 37).
The minimal case. MPE lacks minimal phenomenal selfhood, time and a spatial frame (2020, p. 2). It is “aperspectival: no epistemic agent model” (2020, p. 11). It has no centre or periphery, and it “lacks the phenomenal quality of singularity or numerical identity” (2020, p. 13). Yet it “definitely has representational content” (2020, p. 38). It is “non-egoic self-modelling” (2020, p. 36), “an internal model of an epistemic space” whose epistemic subject “simply is the organism or conscious system as a whole” (2020, p. 27).
The 2024 model. In a nested active-inference sketch, the fourth level models “the system’s global ‘epistemic openness’”. “It is level 4 that really endows the content generated on all lower-order levels with phenomenality”. It is “a single, integrated model of its own epistemic openness in the form of a highest-level but nonegoic self-model”. A full-absorption episode is “inference at the highest level, without any perceptual inference at the levels below it” (2024, pp. 476–477). The epistemic agent model, the layer of the self-model that “creates subjectivity in the sense of a strong first-person perspective”, is “what is completely missing in all full-absorption episodes” (2024, pp. 302, 306).
The evidence. Quoted reports in 2020, and in 2024 a survey of meditators with free-text reports (THEORY-1295). Both texts call the model a hypothesis: “a hypothetical starting point” (2024, p. 477). The one dissociation offered independently of contemplative report is unresponsive wakefulness. There the absence of experience is itself inferred from behaviour.
What could have come out otherwise. Reports of pure awareness could have lacked wakefulness, which Metzinger finds “by far the most frequently mentioned kind of phenomenal character” (2020, p. 11). In the survey, the wakefulness items could have failed to cluster (2024, pp. 465–469). Neither happened. But both checks run on report, and the reports were collected under a definition that already named awareness of awareness.
What this does not say
- That consciousness needs no self-model. It needs no egoic self-model and no epistemic agent model. The minimal case is itself described as a model of the system’s own capacities, “non-egoic self-modelling”. Whether that counts as a self-model in Being No One’s sense is not settled in the texts. In that sense a self-model is one the system identifies with, and MPE is said to be able to become a “nonegoic unit of identification” (2024, pp. 470–471).
- That the minimal case is a unified world. It is neither a world nor experienced as one (2020, p. 13; 2024, p. 477). THEORY-063’s “Selfless Snapshot”, the presence of “one unified world”, is a different minimal case.
- That the model has been found. No neural or computational measure of a represented tonic alertness is reported. The Triple Triangulation Project (2024, pp. 220–221) is a plan.
- That a machine with such a model would be conscious. The 2020 paper says tonic alertness “could be realized by very different physical properties in machines” (p. 32). That is a statement of possibility, and no criterion follows from it.
Connections
- THEORY-063. That reading’s second clause, that unity needs no self-model, is narrowed by this account to “no egoic self-model”. Its minimal case, a selfless unified world, is not this account’s minimal case, which has no world and no oneness. The two can both be true only if THEORY-063’s first clause is about the integration of world-content in ordinary waking experience, not about phenomenality as such. The inference is the record’s.
- THEORY-064. The same transparency apparatus is applied to the model of epistemic openness: transparent, it yields a real world; opaque, an “explicit experience of emptiness and bare wakefulness” (2024, p. 369).
- THEORY-882 (Laukkonen, Friston and Chandaria). The same family: a hyper-model over the hierarchy that sets precision. They also read MPE as the model with no content. Neither gives data.
- CLAIM-412 and QUESTION-254. A candidate for the qualifying organization CLAIM-412 lacks: a temporally thin model of the system’s own epistemic capacities, stated in organizational terms. It is aperspectival, which CLAIM-412’s part 3 does not allow as worded. It is not stated independently of report, which CLAIM-411’s defeater requires.
- QUESTION-048. The account implies that its own prototype cannot be remembered (2020, p. 15) and has no phenomenal contrast class (2020, p. 13). If so, report evidence can reach only the entry to and exit from full absorption, and the account’s test must be architectural.
- TERM-082. The triad splits TERM-082’s sense 4 into a physical, a functional and a phenomenal property.