Which word this is about
This account is about the self: being a subject, the sense of being someone. It is not about personhood. Metzinger does say “you become a person by having the right kind of subpersonal self-model” (LIT-568, p. 25), but that is a further step this account does not take. It is not about phenomenal unity, which in the précis comes first and needs no self-model (THEORY-063). It is not about agency, although the précis says a self-model is where “the foundations for agency and autonomy are laid”.
Source
- Metzinger (2005), LIT-568, read in full in NOTE-460. Page references are to the PSYCHE PDF.
- Metzinger, “Self models” (LIT-450), read in NOTE-372: the same theory, with the biconditional “a phenomenal self exists iff a system operates under a transparent PSM” (C2) and the body-ownership evidence (C6).
What was actually shown
The account. A phenomenal self-model (PSM) is “the content of the conscious self”, “not a thing, but an integrated process” (p. 3). “We are systems that constantly confuse themselves with the content of their PSM”, and organisms “neither have nor are a self” (p. 3). “If an organism operates under a transparent self-model, then it possesses a phenomenal self.” Transparency is that “the representational character of the contents of self-consciousness is not accessible to subjective experience”, so selfhood “is a function realized by a lack of information” (pp. 3–4). On the representational level, “no such things as selves exist in the world” (p. 24).
Transparency is graded. “The degree of phenomenal transparency is inversely proportional to the introspective degree of attentional availability of earlier processing stages” (p. 11). The bodily self-model is fully transparent, reasoning is opaque, and emotions such as trust move between the two (NOTE-460, C3).
The perspective is modelled. A first-person perspective is a phenomenal model of the intentionality relation (PMIR), a model “of the system as acting and experiencing”, co-representing the relation itself (§4). So the point of view is not outside the model. It is part of what is modelled.
Why transparent. Self-modelling is logically an infinite process, and a transparent model is “an untranscendable object” that breaks the loop. This is offered as “my hypothesis”, the “principle of necessary self-reification” (pp. 24–25).
The prediction. No wholly opaque self-representation is known. “Were the PSM to become opaque in its entirety, the phenomenal property of selfhood would disappear” (p. 22).
The précis states the theory and its prediction. Its evidence is cited, and the case studies that would test it (agnosia, neglect, depersonalization) are left to the book (NOTE-460, Limitations).
What this does not say
- It does not say there is nothing there. “We are not things, but processes” (p. 26). The denial is of a substantial, indivisible self. A process view of the self is not refuted by it, and on that view the account is a theory of what the self is, not a denial that there is one.
- It does not say self-models are rare. “Any machine can do self-directed modeling” (p. 22). What makes a phenomenal self is a conscious, transparent one.
- It does not say the prediction has been tested. Reports of depersonalization meet the précis’s own worry about performative self-contradiction (NOTE-460, Open questions).
Connections
- The self as context in RFT and ACT (THEORY-044, THEORY-051). That
account holds that the self as context is a perspective that “cannot
itself be found as content”. This one holds that what is experienced as a
self is content, a transparent model, and that the perspective is itself
modelled. The two may not conflict: what cannot be found as content may
be what is transparent, the part of the model that cannot be attended to
as a model. They conflict if RFT means the perspective is not represented
at all. RFT’s sources do not say, so no
rivalsis declared. For the clinical claim (THEORY-051) the précis supplies a mechanism: making thoughts about oneself attentionally available as thoughts is making that part of the self-model less transparent, which is what graded transparency allows. - The Self in IFS (LIT-579, NOTE-453). IFS’s Self is innate, has qualities (the eight Cs) and leads. On this account, what clients report when they say “that’s my Self” is content of a self-model, and a state with qualities is a state of the model. That is the opposite of ACT’s contentless perspective and nearer to this account’s picture than to RFT’s. The record holds no study of either.
- Gallagher, Dennett and Hume (LIT-453, LIT-442, LIT-451). The précis treats Gallagher’s sense of ownership, “mineness”, as the property that binds contents into the self-model. Like Dennett’s centre of narrative gravity and Hume’s bundle, it denies a substantial self. Unlike them, it attaches a phenomenological prediction to the denial.
- Narrative identity (THEORY-060). That self is reflective and extended in time. This one is pre-reflective and present, and it attributes the experience of an enduring self to “the invariance of bodily self-awareness and autobiographical memory” (p. 26). They are accounts of different things and do not compete.
- Felt affect (THEORY-045). The précis anchors the self-model in continuous internal input (vestibular, visceral, brain stem) and likens the transparent PSM to Damasio’s core self. Consistent with that account; no evidence for it.
- The regress (THEORY-029). The précis’s reason for transparency is that it stops a regress of self-modelling with an object that cannot be transcended. THEORY-029’s regress is of authority, not of representation: what makes a motive the agent’s own. The shapes are alike and the questions differ; the précis does not touch ownership of motives.
Note of 2026-10-30: what holds it back
No reading in the record measures the attentional availability of self-model content together with reported selfhood, or reports a wholly opaque self-model either way. Metzinger 2013 (LIT-1871, NOTE-1434), read in full, narrows the test to the minimal case, “transparent spatiotemporal self-location is necessary and sufficient” for minimal selfhood. It names phenomenally opaque self-location in fully lucid bodiless dreams as the decisive case, and reports none. It also reads some meditative reports as identification with “phenomenality per se”, not loss of self, so THEORY-597’s “loss of basic self” cannot count as selfhood falling with opacity without a measure of opacity. Being No One (LIT-467) is Unavailable. CLAIM-156 cites this reading as corroboration, and CLAIM-355 still grounds on it.
Note of 2026-11-01: transparency carried to epistemic openness
The Elephant and the Blind (2024, LIT-1943, pp. 362–369) carries this reading’s transparency apparatus to the model of epistemic openness itself. It gives no test of the prediction that a fully opaque self-model would end selfhood. Meditation as “opacity control” names a mechanism but no measure. No status, title or warrant changes.