Edward L. Deci and Richard M. Ryan (2000), Psychological Inquiry 11(4):227–268 (target article) — DOI-10.1207/S15327965PLI1104_01

Key takeaways

  • What a need is. A need is not whatever moves people. It is a condition “for psychological health or well-being”, so that “psychological health requires satisfaction of all three needs; one or two are not enough” (pp. 228–229). A need is identified the way one shows that plants need water: by flourishing when it is met and breakdown when it is not.
  • Why and what. Autonomously regulated goal pursuit goes with better performance and well-being than controlled pursuit. Separately, placing relative importance on extrinsic aspirations (wealth, fame, image) goes with lower well-being even after the reasons for pursuing them are controlled for (Carver & Baird, pp. 244–245).
  • Autonomy is unique among the three. Competence and relatedness can be satisfied by controlled behaviour; autonomy “is essential for the goal-directed behavior to be self-determined” (p. 241).
  • Integration requires autonomy support. Relatedness and competence supports “can be sufficient to produce introjected values or compartmentalized (poorly integrated) identifications”. Integration also needs “an opportunity for the individual to freely process and endorse transmitted values” (p. 237).
  • Some values cannot be integrated. “Some goals are not integrateable because they are inherently inconsistent with human nature”, and can at best be introjected (pp. 246–247).

Standing in the record

Filed on 2026-10-03 at the owner’s request, beside Ryan and Deci’s American Psychologist paper (LIT-559), as the long statement of the needs. It is the paper the 2020 review (LIT-561) names as having “formally identified” the three needs, and it carries the material the shorter paper lacks: the concept of need against Hull and Murray, SDT’s account of the self, need substitutes, the evolutionary argument, and SDT’s comparison with social-learning, terror-management, control, achievement-goal, flow and attachment theories.

Its account of the self (pp. 247–248) is what the record’s agency holdings would have to test: whether “integration with the self” can be specified without presupposing the agent (THEORY-029). NOTE-439 draws that out.

No instruction for machine-learning practice; nothing here belongs in the anthology.