Michael E. Bratman (2018), “A Planning Agent’s Self-Governance over Time”, essay 11 in Planning, Time, and Self-Governance: Essays in Practical Rationality, New York: Oxford University Press — DOI-10.1093/oso/9780190867850.003.0011. Reprinted in Time in Action (Routledge, 2022), pp. 75–107. Read in the author’s draft of 10 July 2017.
Key takeaways
- What diachronic self-governance is. It needs synchronic self-governance along the way, which is guidance by a practical standpoint “sufficiently coherent to constitute a clear place where the agent stands”. It is “not merely a concatenation” of that, and needs “appropriate inter-connections across time”. Ulysses at the mast manages himself over time but does not govern himself.
- The glue is shared agency within one life. In planned activity, an intention at t1 to A at t2 interlocks with the intention at t2. The two intend mutual responsiveness and meshing sub-plans, depend on each other, expect success, and are out in the open. These are intrapersonal analogues of Bratman’s conditions (A)–(E) for shared intention. So in diachronic self-governance the agent at each time is acting “together” with herself at other times. The basic unity is this, not “cross-temporal narrative unity”.
- Willpower without a bargain. Willpower in temptation, and resistance to “shuffling” between options one judges non-comparable (Sartre’s young man), are self-governed when the agent’s standpoint includes a “conditional, reflexive end” of her own diachronic self-governance. That end is not a taste for “mere cross-temporal psychic tidiness”. The plan’s time-frame runs to “plan’s end” and includes one’s later reflection. That is why sticking to a resolution against temptation is supported and drinking Kavka’s toxin is not: one anticipates no regret in the first case and regret in the second.
Standing in the record
Filed on 2026-10-03 at the owner’s request, as the read statement of Bratman’s planning theory of self-governance over time. It stands in for Planning, Time, and Self-Governance (2018), which is closed and is not seeded separately: by its note 1 this essay develops the account of self-governance over time sketched in the book’s essay 10, “Rational Planning Agency”. It also stands in for the planning theory of Intention, Plans, and Practical Reason (1987, LIT-593, Deferred), which it summarises in its first paragraphs: plans that “settle practical issues, pose problems of means and the like, filter solutions to those problems, and guide action”.
It extends three earlier works in the LIT scheme’s sense, each of which it needs:
- LIT-593, for the planning theory itself: the settling, filtering and stability of intention, the snowball effect, and the rational non-reconsideration it cites.
- LIT-596 (“Three Theories of Self-Governance”, another reader’s filing in this batch), for its account of synchronic self-governance. The plan-infused standpoints and “self-governing policies” it relies on are cited to that essay.
- LIT-488 (Shared Agency, 2014), whose conditions for shared intention are the model for the intrapersonal analogues. Without them the central argument has no content.
It is a rival of Ainslie’s picoeconomics (LIT-586). Both explain willpower against temptation, and the accounts exclude each other. For Ainslie, will is a recursive self-prediction in an intertemporal bargaining game among successive interests. Bratman rejects the bargaining model. He cites Ainslie’s Picoeconomics (1992) together with McClennen’s bargain between “time-defined selves”, and says a bargain “seems to involve a kind of reciprocal inter-relation that is not possible here”, because the earlier self “will not be around” to respond. His alternative is an end of diachronic self-governance in the agent’s standpoint. Neither tests the other. The account built on each is filed: THEORY-076 on this essay, THEORY-067 on Ainslie’s précis.
Tags. self-governance and agency follow the owner’s distinction.
personhood is for cross-temporal agency: plans induce “psychological
connectedness and continuity of intention and plan that are in the spirit of
broadly Lockean models of personal identity over time”, and the chapter
borrows Shoemaker’s account of Lockean continuity for its first argument.
behavioral-integration is for temptation, willpower and intertemporal
conflict. social-ontology is because the chapter restates the theory of
shared intention it uses.
Where it bears on the record:
- THEORY-029 (no higher-order attitude makes a motive one’s own by its order alone). Bratman locates agential authority in a coherent plan-infused standpoint, not in the order of an attitude. This chapter adds a reflexive end in the standpoint (“a standpoint that includes this very end”). Whether that end is specifiable without presupposing the agent’s participation is exactly what THEORY-029’s promote_when asks Bratman’s self-governing policies to show. This chapter does not argue it: it takes the end as given and asks what it explains.
- Korsgaard (LIT-587). She reads Bratman as a naturalist and asks whether a post-hypnotically installed intention would be the agent’s own. She reports that he answered that it would. Her alternative is that the norms intentions commit us to are what constitute a self. The two accounts of the unity of agency over time meet there.
- Narrative identity (LIT-595). Bratman grants that narrative unity “may apply to many cases”, but holds that the “more basic” unity is acting together with oneself.
No anthology topic holds it, and it carries no instruction for machine-learning practice.