Contribution
A careful restatement of the evolutionary argument in Unto Others’ second half, with diagrams of the motivational structures at issue, and a rebuttal. It shows that the argument turns on an unstated assumption about cognitive architecture: that the beliefs a hedonist would need are ordinary, revisable beliefs.
Key insight
Whether a desire for someone else’s good is ultimate or instrumental is a question about where it sits in the agent’s practical reasoning, not about its object. A parent who wants to help her child “as an end in itself” is an altruist. A parent who wants it because she believes she will feel bad otherwise is a hedonist. The two act identically, and the debate is over which architecture selection built.
Assumptions
- The definitions (Preliminaries).
- A behaviour is evolutionarily altruistic if it lowers the actor’s inclusive fitness and raises another’s.
- It is psychologically altruistic if it is motivated by an ultimate desire for another’s well-being.
- An ultimate desire is one not produced or sustained by practical reasoning.
- Egoism says all ultimate desires are self-interested. Hedonism, its strongest form, which Sober and Wilson choose to meet, says they are only for one’s own pleasure and the absence of one’s pain.
- Motivational pluralism (Sober and Wilson’s altruism) allows both kinds.
- Practical reasoning as a causal process in which a desire and a belief produce a further desire (after Goldman 1970).
- Parental care as the test case, since it is not evolutionarily altruistic and so avoids the group-selection dispute.
Key results
- Evolutionary and psychological altruism are logically independent (Preliminaries). Kin-directed help is usually not evolutionarily altruistic on the inclusive-fitness definition. A mindless organism can be an evolutionary altruist, and an ultimate desire for one’s own offspring’s welfare is psychologically altruistic though fitness-enhancing.
- The three solutions Sober and Wilson consider (305): altruism, hedonism (pleasure and pain tied to beliefs about the child), and pluralism (both). Selection would weigh availability, reliability and energetic efficiency. They argue only reliability separates them.
- Stich’s models.
- Future Pain Hedonism: the parent believes she will feel bad later if she does not help.
- Current Pain Hedonism: the child’s need makes her feel bad now, and she believes helping will stop it.
- Arguments 1 and 2 (315–316): the belief–affect link is imperfect and must be strong. Stich replies that fear, anger and disgust are built the same way, so selection evidently tolerates such links. Neither argument touches Future Pain Hedonism.
- Argument 3 (320): two pathways (PLUR) are more reliable than one. Stich replies that redundancy has costs and the authors give no reason it pays here.
- Argument 4 (314–315): the hedonist’s linking belief is empirical and could be lost, and the parental care would then collapse. The instrumental desire stays only if the organism “is trapped by an unalterable illusion”. Stich calls this “a powerful argument … and an original one”.
- The reply (the closing section). If the linking representations are sub-doxastic, like Chomskyan grammar or Carey and Spelke’s core knowledge, they cannot be removed by evidence. The desire they sustain is still instrumental and as reliable as an ultimate one. Selection often keeps useful representations out of reach of belief revision.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Evolutionary and psychological altruism are logically independent | strong: by definition, with cases | Preliminaries |
| C2 | Sober and Wilson’s arguments from correlation, affect strength and redundancy do not favour altruism over hedonism | moderate: counterexamples from other emotion systems; no model of costs | the first three replies |
| C3 | The argument from the fragility of hedonism’s beliefs fails if the beliefs are sub-doxastic | moderate: depends on whether desires sustained by sub-doxastic states count as instrumental, which Stich takes as natural | the closing section |
| C4 | Sober and Wilson’s analysis does not provide evidence for psychological altruism | moderate: follows from C2 and C3; it does not show egoism true | Conclusion |
Concepts
- ultimate / instrumental desire: desired for its own sake, or produced by practical reasoning from another desire and a belief.
- psychological egoism / hedonism / altruism (motivational pluralism): as defined above.
- sub-doxastic state: a belief-like state that is not inferentially integrated, is hard to revise by evidence and is usually not open to introspection (Stich 1978).
Connections
- Unto Others (LIT-1683), unread. Stich quotes it at pp. 296–320, the psychological half; the evolutionary half on group selection is outside his scope.
- Batson’s objection (CLAIM-494). Stich does not discuss social psychology. Sober and Wilson, he reports, cast the contest as altruism against hedonism as “the version of egoism that is most difficult to refute” (297).
- Hamilton’s rule (THEORY-655). C1 bears on how the record uses it.
Bearing on the record
- The owner’s thesis (CLAIM-491). Stich’s terms show where the
thesis stands. These are inferences.
- The thesis is a form of egoism: every ultimate desire is self-directed. But it widens the self, not the motive. A parent’s ultimate desire “I do what will be most helpful for my child” is altruism to Sober and Wilson, and self-interest on a larger self to the thesis. The two may describe one architecture in two vocabularies. The debate Stich reports is about whether the desire is instrumental to the agent’s own feelings. The thesis gives no answer to that.
- So the thesis needs to say which it claims: that the desire for the child’s good is instrumental to some good of the agent (classical egoism, open to Sober and Wilson’s argument and to Batson), or that it is ultimate but its object counts as the agent’s self (a redescription that needs CLAIM-492’s independent criterion).
- The evolutionary support is of another kind. The claim cites Hamilton’s rule “as if the self’s genetic interest extended to kin”. By C1, kin selection says nothing about psychological motive. A desire for a kinsman’s welfare can be ultimate, and so psychologically altruistic, while kin-helping is not evolutionarily altruistic.
- A dated note is added to the claim.
- No THEORY filed.
- No ML instruction.
Limitations
- A critique, so the book’s argument is seen through it, though with long quotations. The book’s treatment of social psychology (Batson) and of group selection is not covered.
- Stich does not argue that egoism is true, only that this argument leaves it open (the last paragraph).
- Whether sub-doxastically sustained desires are instrumental is asserted as “the natural answer” and could be disputed.
Open questions
- Are the links between helping and the helper’s own affect revisable by evidence? If they are, Sober and Wilson’s fourth argument stands.