CLAIM-491: No action is selfless in the sense of aiming at a good that is no self's: what is called selflessness, collectivist or altruistic, is concern for the good of a self extended to include others or a group, so collectivism is self-interest on a larger self-identity

The claim

The owner’s position: there is no selflessness. A person who acts for a group or for others acts for a self that includes them. The difference between the “selfish” and the “collectivist” person is the size of the self each identifies with, not whether each pursues their self’s good.

The record holds four kinds of support:

• Self-categorization theory (THEORY-1114). Social identity is a self-categorization at a more inclusive level, as much a self as personal identity (Turner et al. 1994). Group-serving action is then action on that self’s behalf.
• Oneness and fusion. Cialdini et al. (LIT-1728) found helping tracked perceived self–other oneness, not empathy. Identity fusion (Swann et al., LIT-1735) predicts self-sacrifice for the group from felt oneness of personal and group self.
• Inclusive selves. Self-expansion (Aron et al., LIT-1706) and interdependent self-construal (Markus and Kitayama, LIT-1703) give selves that include others.
• Evolution. Hamilton’s rule (THEORY-655) values others’ fitness by relatedness, as if the self’s genetic interest extended to kin.

What it does not say

• Not that people are self-centred in the ordinary sense. The thesis locates collectivism within self-interest, not against it.
• Not that the extended self is a group subject. It is the individual’s self-identification that extends.
• Not settled by self-categorization theory. That theory gives levels of the self, but its depersonalization says the personal self recedes when the social one is salient. That is a change of self, not plainly a larger one. Identity fusion and León’s personalized “we” (THEORY-1115) fit “larger” better.
• Its sharpest objections are two. The empirical one: Batson et al. find empathic helping without self–other merging (CLAIM-494). The conceptual one: the thesis may be true by definition (CLAIM-492).
• Two neighbouring views draw the line elsewhere. Team reasoning (Sugden, LIT-1782) treats “what is good for us” as a third mode, neither self-interest nor altruism. Akerlof and Kranton (LIT-1752) put identity in the utility function, which makes the thesis true of an economist’s agent by construction.
• Where the owner’s question sits. It is in both territories. Self-categorization theory supplies the mechanism of a larger self; the phenomenology of the “we” (Zahavi, LIT-1758; León) asks whether a “we” is a self at all, and León argues that intimate “we”s are built by individualization. Johnston (LIT-1759) takes the same premise to the opposite moral: the good person’s self is as large as humanity.

Note of 2026-10-29: team reasoners and team benefactors

Salice and Miyazono (NOTE-1266) distinguish two people who both act for a group.

• The team benefactor reasons in the first person singular about the group’s good: “I intend that we J”.
• The team reasoner has undergone self-transformation and intends as part of the group: “We intend to J”.

The thesis fits the team reasoner best. Their good is the we’s, and if the we is their self, it is self-interest on a larger self. It fits the team benefactor least: they care about a group without counting themselves in it. Salice and Miyazono also offer what CLAIM-492 asks for, a mark of self-extension apart from caring: the activated social self, triggered by group cues and not under direct rational control. A sharpened thesis would say that collectivist action by team reasoners is self-interest on an extended self, and leave the team benefactor’s concern as the open case.

Note of 2026-10-29: identity fusion, read

Swann et al. (NOTE-1256) has been read. It supports the thesis in two ways and qualifies it in one.

• A larger self, not a changed one. Fused people keep a salient personal self and add the group to it. They say they would do more for the group than other members would, which is not depersonalization.
• The prediction holds. If the group is part of the self, defending the self should go through the group. Challenging fused participants’ personal self-views made them readier to fight and die for Spain. It had no effect on the nonfused.
• The self being served is a self-concept. The authors call dying for the group “not in the individual’s self-interest” and explain the effect by self-verification, the motive to keep one’s self-views true. If the thesis recruits fusion, “self-interest” has to cover keeping a self-concept true, not only furthering one’s welfare.

For CLAIM-492, fusion is measured by perceived overlap, is specific to one group and is separable from identification. But the paper never measures caring separately, and its gloss on the fused is that they “care as much about the outcomes of the group as their own”. So it supplies a candidate criterion without showing that the criterion comes apart from caring. The outcomes are self-reported willingness, not acts.

Note of 2026-10-29: Sugden, Stich and Johnston, read in place of unavailable works

Three works read in place of Sugden 2003, Unto Others and Surviving Death each press the thesis to say more exactly what it claims.

• A third shape of cooperation (Sugden 2015, NOTE-1288). Bacharach’s team reasoner pursues the team’s collective good, which can outweigh a member’s. That is close to self-interest on a larger self. Sugden’s cooperative team reasoner intends each member’s benefit, her own included, with no larger self in play. It is neither the thesis’s larger self nor sacrifice. Sugden also rejects building pro-social motives into an individual’s payoffs, which is what enlarging the individual does.
• Two readings of the thesis (Stich 2006, NOTE-1282). In the egoism debate’s terms the thesis must choose. Either the desire for a child’s or a group’s good is instrumental to some good of the agent, which is classical egoism and meets Sober and Wilson and Batson. Or it is an ultimate desire whose object counts as part of the self. That is a redescription of what Sober and Wilson call altruism, and has content only with CLAIM-492’s independent criterion. Stich also shows that kin selection is silent about motive, so Hamilton’s rule supports the thesis only by analogy.
• Johnston has changed his mind (Johnston 2023, NOTE-1257). The premise cited above, that the self’s extent is fixed by one’s deepest attitudes, is one he now rejects. His reason is that such selves multiply bearers of moral status. Whether an enlarged self inherits that problem is open.

Note of 2026-10-29: the psychology, read in place of unavailable works

Works read in place of the unavailable psychology papers bear on the thesis at five points.

• The thesis in a precise form (Miyazono and Inarimori, NOTE-1277). They read Cialdini’s “merging” as group identification: empathizing with someone makes her welfare a constituent of “our” welfare. Helping is then altruistic at the individual level and egoistic at the group level. That is this claim, stated exactly, and it leaves Batson’s individual-level altruism standing. On their reading empathy always brings identification with it, so the claim gains content only if identification can be found absent where empathy is present.
• Two routes into a group self (Otten, NOTE-1301). Projecting the personal self onto a new group (self-anchoring) is a larger self. Taking on the group’s traits (self-stereotyping) is a changed one. Which operates depends on how much is known about the group. When participants could also pay themselves, they took the most for themselves and favoured the in-group with the rest: the personal self comes first.
• Larger or switched (Spears and Postmes, NOTE-1280; Spears, NOTE-1268). The crowd member is not selfless: anonymity brings “the group in the self”, and behaviour follows group norms more closely. But anonymity amplifies whichever identity is salient, which is a switch of self, not an enlargement. Identity fusion (NOTE-1256) stays the record’s case for “larger”.
• Which larger self (Brewer, NOTE-1290). Close relationships and groups both extend the self, in different ways: the relational self takes in particular others, the collective self a group. Brewer separates relational collectivism, obligation to specific persons, from group-based collectivism. The thesis should say which it means, or that it means both.
• Helping follows the helper’s category (Dovidio et al., NOTE-1286). Recategorizing two groups as one removed in-group bias in helping: at a stadium, White fans complied with a Black interviewer’s request 59% of the time when he wore their university’s hat, against 36% when he did not. Shared identity raised empathic concern, which carried part of the effect. But White students helped a Black student who called himself a Black Colgate student no more than one who called himself Black. If helping follows an extended self, it is the helper’s version of the shared category.

The same extension in time, legacy, is QUESTION-301.

Note of 2026-10-29: the thesis in time

The readings on legacy and on obligations to the future (QUESTION-301, QUESTION-299) bear on the thesis in four ways.

• Its shape holds for prudence (Hershfield, NOTE-1259). Felt overlap with one’s future self predicts patience, so prudence is regard scaled by overlap. The thesis and its mirror, “no prudence is purely self-interested”, are claims about one scale.
• Scheffler is not the counterexample he was filed as (NOTE-1298). He reads our concern for humanity’s survival as dependence, not altruism. The self served would be one’s standing as a valuer, a third sense of “self-interest” the thesis would have to adopt explicitly.
• Where the far future is concerned, the evidence points elsewhere (Schubert et al., NOTE-1270; Caviola et al., NOTE-1267). Lay concern reaches the far future through a picture of it as good, an impersonal valuation that no measured self carries. That is the hardest case the record now holds, and it is thin.
• Bostrom grants the thesis its prediction (NOTE-1321). Concern for existential safety is undersupplied because people capture only their own share of it: concern reaches as far as the self’s stake.

Note of 2026-10-29: the thesis backward in time

The intergenerational-trauma readings (QUESTION-297) test the thesis’s reading that a descendant’s grief over an ancestral injury is self-regarding. They support its shape and move its boundary.

• The self an ancestral injury extends to is a category, not a line of descent (Eyerman, NOTE-1246). Slavery became the trauma of black Americans “whether or not they directly experienced slavery or even had ancestors who did” (p. 14). “Ancestral” grief, if self-regarding, is regard for a category’s past.
• What joins the self is an image, felt as grievance (Volkan, NOTE-1250). A chosen trauma is a shared image of an ancient injury, “real, fantasized or even mythologized”, called up when the group’s present identity is threatened. That fits a self-regarding reading well. Grief for particular, recent dead is the case Volkan keeps apart, and the thesis’s open one.
• Family transmission is not where the effect is (Fridman, NOTE-1291). Survivors’ daughters did not differ from daughters of families who shared the same national memory. If an ancestral injury is carried as part of a self, it is carried by the group, not handed down in the family.

Note of 2026-10-29: the remaining sources on the self, read

Four works the thesis was filed against have now been read. They agree on one point: what group membership changes is the self’s content or self-image, and none of them shows a self enlarged to take in the group’s good.

• Markus and Kitayama ground the premise, not the conclusion (NOTE-1275). For interdependent selves, “others are included within the boundaries of the self” (p. 245), and the authors name the thesis’s middle position, “an alternative to selfishness … besides selflessness” (p. 247). But this “does not result in a merging of self and other” (p. 228). The interdependent self is made of relations and roles, and its in-group may be narrower, not larger. The authors read the resulting action as possibly “genuine, other-oriented, altruistic” (p. 248). They also split collectivist motive in two (p. 229):
  • others’ goals taken as one’s own, which is the thesis;
  • others’ goals met as the price of watched reciprocity, which needs no extended self.
• Akerlof and Kranton make it true only in the trivial sense (NOTE-1311). Identity-driven sacrifice maximizes a utility that includes identity. The identity is a self-image, kept by following one’s categories’ prescriptions. No group’s welfare enters anyone’s utility, and a fellow member’s deviance is a loss one may relieve by harming them. Their criterion of the self’s extent, category assignment partly fixed by others, could come apart from caring (CLAIM-492). But its weights are free parameters read off the behaviour they explain.
• Zahavi: a “we” is “not an enlarged I” (NOTE-1289, p. 10). It keeps the plurality it unites. Identification changes how one apprehends oneself, as one of us seen through another’s eyes: a changed self more than a larger one. His account covers only the face-to-face “we”.
• The self-esteem hypothesis: benefit without motive (Rubin and Hewstone, NOTE-1278). This is social identity theory’s own version of the thesis. Favouring the group tends to raise self-esteem (18 experiments for, 13 against), but a need for self-esteem rarely drives it (7 for, 28 against). A benefit that follows group-serving action without motivating it does not show the action aimed at the self’s good. The group-level motive was nearly untested; the two studies that measured it both support it.

The thesis survives these readings only as the claim that the self served is a self-concept. That is the third sense of “self-interest” the note on Scheffler named. On that reading, CLAIM-492’s demand for a criterion apart from caring is still unmet.

Note of 2026-10-30: filed with its concessions

CLAIM-492 has been written into this claim’s defeated_if since filing: the thesis is answered only if it is tied to a criterion of the self’s extent, such as perceived overlap, fusion or salient self-categorization, that can come apart from caring. That conditional is conceded. The claim has not yet named such a criterion and shown it coming apart from caring, and until it does it says nothing that caring alone does not.

Note of 2026-10-30: Cialdini et al., read

Cialdini et al. (LIT-1728) are now read in full (NOTE-1444). The body’s account of them holds in direction and needs three corrections of scope.

• Willingness, not helping. Students chose how much costly help they would give an imagined near stranger, acquaintance, friend or family member. With oneness (the IOS circles and a “we” item) entered, empathic concern stopped predicting that choice in every test, near strangers included (fn. 5). No help was given, and empathic concern was rated after the choice (fn. 6).
• Others, not groups. The paper reserves “oneness” for a dyad with a specified other and keeps it apart from “we-ness” as group membership or social identity (fn. 1). It bears on the title’s “others”. On its “group”, and on collectivism, it offers only a speculative parallel (p. 492). Placing oneness beside identity fusion under one heading is the record’s grouping, not the paper’s.
• Not self-interest either. The authors call oneness-based helping “nonaltruistic” and decline to call it egoistic: the dichotomy “loses its meaning” (pp. 490–491). That supports “no action is selfless”, not “self-interest on a larger self-identity”.

On CLAIM-492, the paper separates oneness from empathic concern: severity of need raised empathic concern and not oneness (p. 491). It does not separate oneness from closeness or from caring in general, which is what CLAIM-492 asks.

Note of 2026-10-30: what holds it back

CLAIM-494 is unanswered. Cialdini et al. (LIT-1728), now read, found the opposite in their own paradigm, but the objection rests on Batson et al.’s reply (LIT-1700), which criticises that paradigm and is unread. CLAIM-492’s conceded condition is unmet: no reading shows a measure of the self’s extent coming apart from caring, and Cialdini et al. separate oneness only from empathic concern. And the collectivist half of the title has no direct evidence. Cialdini et al.’s oneness is dyadic, and the group readings (NOTE-1275, NOTE-1289, NOTE-1278) show a changed self-concept, not a self enlarged to the group’s good.