Source
- Barnes-Holmes, Barnes-Holmes, McHugh and Hayes (2004), LIT-541, read in full in NOTE-430. The three selves are on pp. 367–369, and the deictic reading of defusion on pp. 370–371.
- Hayes, Luoma, Bond, Masuda and Lillis (2006), LIT-555, read in full in NOTE-431 from an early manuscript. Self as context is on p. 8.
The RFT book (LIT-549), to which every concept in both papers is cited, is filed unread.
What the sources say
Three selves. RFT defines “the conceptualized self (i.e., the self as the content of verbal relations); the knowing self (i.e., the self as the ongoing process of verbal relations); and the transcendent self (i.e., the self as the context of verbal relations)” (LIT-541, p. 368). The worked example is a woman deciding whether to end her marriage. Deciding from self as content, she stays because she is “the type of woman who simply cannot live without a man”. Deciding from self as process, she leaves because she has accurately noticed that she has long been unhappy, without asking what kind of person she wants to be. From self as context she can contact both, “but neither obtain absolute control over her decision”, and only from there can she see that protecting the first is feeding the hostility the second reports (pp. 368–369).
The context is a learned perspective. “As a result of deictic frames such as I-You, Now-Then, and Here-There human language leads to a sense of self as a locus or perspective.” “‘I’ emerges over large sets of exemplars of deictic relations, but since this sense of self is a context for verbal knowing, not the content of that knowing, it’s limits cannot be consciously known” (LIT-555, p. 8). The manuscript cites “growing evidence of its importance to language functions such as empathy, theory of mind, sense of self”, but the citation is an unfilled placeholder (“see xxx”), and its citation of trained perspective-taking is “(xx)” (NOTE-431, Corrections).
Fusion is content standing in for the context. “When we respond to our thoughts as I HERE NOW, we are in a sense caught up in them – we are not perceiving those thoughts and feelings from a perspective, but rather they are the perspective.” Defusion moves them to I THERE THEN, so that they are “thoughts and feelings that we had, rather than providing our current and only perspective” (LIT-541, pp. 370–371).
What is evidence and what is not. The three selves are definitions within a theory, and the marriage case is constructed. The claim that the perspective is learned from deictic relations is RFT’s empirical claim, and in the record it rests on studies cited and not read. Part 3 of the 2004 paper, which reads ACT’s processes through RFT, is labelled by its authors “no more than a heuristic” (p. 369). So the conceptual account is stated clearly enough to compare, and its empirical half is unsupported here.
Where it meets the record’s views of the self
- Self as content is the narrative self. Gallagher (LIT-453, NOTE-377) separates the narrative self, extended in time through memory and story, from the minimal self. Dennett (LIT-442, NOTE-364) makes the self a “centre of narrative gravity”, an abstractum the story coheres around, and Hume (LIT-451, NOTE-386) makes its identity “fictitious”, supplied by the imagination. RFT’s conceptualized self is this self described from the outside: a network of verbal relations whose subject is oneself. Where the record’s views differ is on its value. ACT treats domination by it (“defending a conceptualized self”, LIT-555, p. 7) as part of inflexibility. That converges with Strawson (LIT-452, NOTE-383), for whom self-narration tends to falsify the self it describes (C7), and with Dennett’s claim that self-interpretation can change the self (C4). It sits awkwardly with Gallagher’s richer narrative self, which includes self-deception by design, and with the ethical narrativity views Strawson attacks, of which the record holds Schechtman (LIT-468) unread.
- Self as context is what cannot be found as content. Hume, looking inward, “never can catch myself at any time without a perception” (NOTE-386), and concludes there is only the bundle. RFT accepts the observation and draws a different conclusion: what cannot be found among the contents is the perspective from which they are found, whose limits “cannot be consciously known”. Metzinger (LIT-450, NOTE-372) gives the same structure in self-model terms. A first-person perspective is a transparent model of the intentionality relation, one the system cannot recognise as a model (C8). On both accounts the perspective is not an object of experience, and on neither is it a substantial self. RFT calls it a behavioural context; Metzinger calls it a model. Neither cites the other, and the mapping is the record’s.
- The deictic “here” follows the point of view. In Dennett’s “Where Am I?” (LIT-405, NOTE-338), where a person is follows the point of view, and the narrator cannot think “here” of the vat that holds his brain. On RFT’s account the same deictic terms are what the perspective is learned from. Both treat “here” as fixed by where one sees from. RFT does not discuss the body, and the comparison is the record’s.
- The open seam is the minimal self. Gallagher’s minimal self is
pre-reflective and embodied. He reports neonates “less than an hour old”
imitating faces as evidence of a primitive self before language
(NOTE-377, C5), and Metzinger roots the self-model in non-linguistic motor
control (NOTE-372, C4). If RFT’s self as context were that minimal self,
its claim that “‘I’ emerges over large sets of exemplars of deictic
relations” would contradict this evidence. The sources support a
narrower reading. What RFT says is learned is a sense of self as
perspective, a context “for verbal knowing”, and the evidence it points
to concerns empathy and theory of mind, which are reflective,
self–other achievements. On that reading self as context is a verbal
layer that presupposes the minimal self and does not replace it, and the
two are compatible. The “transcendent, spiritual” language of LIT-555
reaches further than that reading, and nothing in the sources decides
between them.
promote_whensays what would. - Metzinger’s précis puts the perspective inside the model. The mapping above reads Metzinger’s first-person perspective as RFT’s context. The précis (LIT-568, read in NOTE-460; THEORY-064) makes the contrast sharper than that mapping suggests. What is experienced as a self is content: “the content of the conscious self”, a phenomenal self-model that cannot be recognised as a model (§1.2). And the perspective is itself modelled: a phenomenal model of the intentionality relation, “a model of the system as acting and experiencing” (§4). RFT says the self as context is “a context for verbal knowing, not the content”. The two agree only if “cannot be found as content” means what the précis calls transparency, a model whose earlier processing stages are unavailable to attention. They disagree if RFT means the perspective is not represented at all. The RFT sources do not say which, and the précis does not discuss RFT. Graded transparency would also give defusion a reading the 2004 paper lacks: holding a thought “from I THERE THEN” is making that part of the self-model available to attention as a model. That reading is the record’s.
- IFS’s Self is the opposite kind of construct. Schwartz’s Self (LIT-579, read in NOTE-453; the book’s first chapter in NOTE-457) is innate, “always right there if our parts let it in”, and it has qualities, the eight Cs: “calmness, curiosity, clarity, compassion, confidence, creativity, courage, and connectedness”. It is “an active healing presence”, not “a passive witness state”, and it leads the parts. RFT’s self as context is learned from deictic relations and has no content. So the two differ on both of this account’s axes: origin (innate against learned) and content (qualities against none). They share a function, a place from which thoughts and feelings are held without being fused with, and a theory that this shared function is a common active ingredient would need process studies for both, which the record does not hold.
- No substantial self is required. Like the bundle, the narrative abstractum and the self-model, RFT’s three selves are functional: what verbal behaviour is about, its ongoing activity, and the perspective it is emitted from. Nothing in the account needs a self as an entity. Shanahan (LIT-466) reads the dissolution of every candidate site of an AI’s self as Buddhist emptiness. ACT, which calls the perspective “transcendent” and fosters it by mindfulness, is the clinical relative of that reading, though neither paper says so.
What this does not say
- It does not say self as context is therapeutic. That is a separate, empirical claim, THEORY-051, with its own evidence.
- It does not say RFT has shown the perspective is learned. The sources assert it and cite studies the record has not read.
- It does not settle personal identity. The three selves say nothing about persistence over time, fission or what matters in survival which the record’s personal-identity readings address. Strawson’s Episodics, who do not experience the self they are as having been there in the past, could have all three selves.
- It does not answer the regress about the agent’s standpoint. Deciding “from self as context”, with neither content nor process in “absolute control”, looks like a candidate for the standpoint THEORY-029 argues no higher-order attitude can supply. A perspective is not an attitude, so the regress may not apply to it. Whether it can then make a decision the agent’s own is a question neither source asks.
Connections
- THEORY-050: fusion with self-content is one form of the inflexibility that account describes, and its note on literality as transparency is the same mapping as here.
- Zahavi’s Subjectivity and Selfhood (LIT-463) and Metzinger’s Being No One (LIT-467) are filed unread. Either could decide the minimal-self question above. Metzinger’s own précis of the book is now read (LIT-568); it roots the self-model in non-linguistic bodily input and leaves out the book’s case studies, so it does not decide the question.