Akrasia¶
An agent intentionally acts against an available course they still judge better, all things considered.
Core Idea¶
Akrasia is the structure of an agent intentionally doing one thing while still judging that an available alternative would be better, all things considered. The question is not simply whether the action was unwise by an observer's standard. The agent's own practical judgment and action come apart at the relevant time. This is why akrasia presents a problem for accounts that treat considered judgment as sufficient to settle action. Aristotle's account of incontinence connects the puzzle with knowledge, perception and appetite, while later philosophy disputes exactly how the divergence should be explained.[1][2]
The distinction is diagnostic, not a ready-made causal theory. Desire, fear, aversion, habit or salience may contribute in different cases, but no one motive is part of the definition. Nor does a changed plan alone establish akrasia: an agent may rationally revise an intention after new information. Richard Holton argues that over-ready revision of intention—what he calls weakness of will—can be distinct from acting against a retained better judgment.[2]
Structural Signature¶
Sig role-phrases:
- Single agent: the same person evaluates and acts.
- Practical alternatives: at least two relevant courses of action are available under the case description.
- Better judgment: the agent judges one alternative better overall, not merely attractive in one respect.
- Persistence: the better judgment remains in force as the contrary action occurs; a sincere change of view is a different configuration.
- Intentional contrary action: the agent knowingly performs the lower-ranked action, rather than slipping, being forced, or failing from incapacity.
- Explanatory gap: the action reveals that the agent's evaluative ranking and effective motivation did not align.[1][2]
Condensed: retained better judgment + available better alternative + intentional contrary act → akratic divergence.
What It Is Not¶
- Not merely a bad choice. The better course must be better by the agent's own considered judgment.
- Not ignorance. If the agent did not know the relevant alternatives or consequences, the puzzle changes.
- Not coercion or inability. The agent must be able to pursue the judged-better option in the relevant sense.
- Not an accidental lapse. A slip of the hand or a forgotten appointment is not intentional contrary action.
- Not every broken resolution. A resolution can be abandoned after a defensible revision of judgment; Holton's analysis separates some intention failures from akrasia.[2]
- Not a diagnosis or fixed personality trait. A case concerns a relation among judgment, opportunity and action; the case alone does not identify a clinical condition or stable character.
Scope of Application¶
Classical discussions use cases of appetite or pleasure, but the structure is not confined to bodily indulgence. Aristotle's sweet-food illustration supplies a concrete clash between a rule against tasting and appetite directed at this sweet object. Holton's meat-eating friend supplies a different, contemporary ethical-practice case: he accepts the arguments against eating meat while continuing to eat it and acknowledging his inconsistency. Neither is evidence of how commonly people act this way. A case should describe what the agent judged, when they judged it, whether the alternative was available, and whether the contrary act was intentional.[1][2]
The philosophical boundary is disputed. Some accounts analyze an all-things-considered judgment as conditional on the considerations weighed, others examine the agent's attention at the moment of choice, and still others distinguish judgment from intention or desire. This entry preserves the common puzzle without making any one explanation part of the identity.[2]
Clarity¶
Ask whose judgment is at issue. An observer's claim that a student should work is not enough; the student must still judge work better overall. Ask when the judgment was held. A judgment on Monday and a sincerely revised judgment on Tuesday may describe changing preferences, not akrasia on Tuesday. Ask what action was intentional. Losing a battle with a technical barrier or missing a deadline because of an emergency differs from intentionally browsing while judging that work should begin. Also separate a judgment change from an intention change: justified reconsideration is not automatically weakness of will, while an unexamined inclination need not justify abandoning a resolution. These tests prevent the word from becoming a decorative synonym for procrastination or moral failure.[2]
Manages Complexity¶
The abstraction separates three questions often collapsed into one: what the agent regards as best, what the agent can do, and what the agent intentionally does. Once separated, a disagreement about akrasia can be localized. One theorist may deny that the better judgment remained intact; another may accept it but offer an account of contrary motivation; a third may question whether the better course was actually available. The pattern does not decide these disputes, but it makes their points of contact explicit.[2]
Abstract Reasoning¶
To test a purported case, reconstruct the agent's all-things-considered assessment at the time of action. Confirm that the alternative was feasible and that the action was chosen rather than accidentally produced. Then ask what explains the divergence without building that explanation into the definition. If the agent simply learned new facts and judged the once-inferior option best, the configuration has changed. If the agent merely failed to carry out an earlier resolution, test whether the old judgment persisted; failure to preserve an intention and action against better judgment are not interchangeable.[2]
Knowledge Transfer¶
The judgment–action distinction transfers among cases of consumption, work, interpersonal conduct and other choices. What transfers is the relational test, not a uniform psychological mechanism. A cue-driven snack and an anxious avoidance of a conversation might both meet the test yet involve different motives and remedies. Likewise, a domain may use “weakness of will” for failures to stick with resolutions. Holton's argument shows why those cases should be compared rather than automatically merged with akrasia.[2]
Examples¶
Aristotle's sweet-food case¶
In Book VII, chapter 3, Aristotle puts a prohibition on tasting beside the particular judgment that this object is sweet and an appetite for it. The appetite can move the person to taste despite the prohibitory rule. This is a source-attested philosophical illustration, not an observed participant. To apply the present signature, the person must still own the prohibition as the better practical conclusion at tasting, have abstention available, and taste intentionally. Aristotle's own explanation complicates that reconstruction: the relevant particular knowledge may be inactive or possessed only in a qualified sense under passion. The example therefore exposes, rather than settles, the disputed persistence premise. A later observer's claim that sweets are unhealthy would not establish it.[1]
Mapped back: Aristotle's eater → prohibitory rule, if still practically held → available abstention → appetite-led tasting. The contested arrow is whether the rule remains fully operative at the instant of action.
Holton's meat-eating friend¶
Holton describes a friend who accepts that the arguments favor not eating meat but continues to do so and calls himself inconsistent. Unlike the schematic student, the case records the agent's own avowed judgment; it is not inferred from Holton's disapproval. Eating meat is a deliberate practice, and abstention is presented as an alternative, though the paper does not establish each occasion's immediate options or the friend's occurrent judgment. Holton calls the friend akratic but not weak-willed in the intention-revision sense. He contrasts this with a hypothetical firm resolution to give up meat followed by roasting a pig, which would add an intention failure. The source thus tests why the akrasia classification does not depend on a prior resolution, while leaving the finer time-of-action premise open.[2]
Mapped back: Holton's friend → avowed all-things-considered judgment against meat → available abstention in the described practice → continued chosen meat eating; occasion-specific persistence remains an evidential qualification.
Justified intention revision¶
An agent planned to travel, then learns that travel would put another person at risk and decides not to go. The abandoned intention is real, but the all-things-considered judgment has changed. Holton's distinction makes this a useful counterexample to equating all intention failure with akrasia.[2]
Mapped back: same agent → old judgment replaced by a revised judgment → travel remains possible but is no longer judged best → chosen nontravel. The persistence role is absent; new evidence is not itself a better judgment.
Structural Tensions¶
Judgment's authority versus explanatory reach. Treating all-things-considered judgment as decisive for action preserves a parsimonious account of rational agency, but then Aristotle's eater or Holton's avowedly inconsistent friend must be reclassified by finding a change, qualification or inactivity of judgment. Allowing judgment and effective motivation to diverge preserves their apparent self-descriptions, but requires an independent account of why the agent's conclusion did not govern action. The practical cost is different: the first approach can dismiss testimony of conflict; the second can label an action akratic before the persistence premise is adequately checked. Diagnostic: what evidence locates the agent's own ranking at the instant of choice, rather than before or after it?[1][2]
Structural–Framed Character¶
Akrasia is framed, though its comparison of stated ranking with effective choice has a structural aspect. Evaluative weight is not incidental: the agent must still judge an alternative better all things considered. Human practical reasoning and intentional action are constitutive rather than optional applications. No school, institution or therapeutic practice created the pattern, although philosophical traditions supply different accounts of it. The vocabulary travels literally among domains of human action, such as consumption and work, when the same judgment/action test is met. Importing “akrasia” to an automated system that merely violates a programmed priority would be metaphor unless that system had an appropriately evidenced analogue of practical judgment; recognizing a generic policy/action mismatch is not yet recognizing this identity. Its character: a human-agency philosophical pattern, structurally clear but dependent on a normative, intentional frame.
Structural Core vs. Domain Accent¶
The skeletal relation is considered ranking versus effective choice, a possible future-prime question about policy/action divergence. Akrasia's actual mechanism requires an agent's retained all-things-considered judgment, an available better alternative, and intentional contrary action. Those conditions distinguish it from accidental inconsistency, later regret and ordinary plan revision. The named entry does not clear the prime bar simply because health, study and work supply multiple examples: those are still human-practical settings using the same philosophical vocabulary, not unrelated domains in which its diagnostics travel intact. The unparented root has no checked broader parent; a future genus needs separate identity review.
Instantiates / Related Primes¶
Self-control and rationality are relevant comparison concepts, not automatic parents. Akrasia can be described as a failure to act in accord with practical judgment, but neither the capacity for self-control nor the broad norm of rationality is a checked strict genus for this identity. Knowledge–Action Gap adds accurate knowledge, sincere intention and systematic nonperformance that a single akratic act need not have. The entry is an unparented root with no DAG edge; this does not foreclose a later, independently admitted broader genus.
Neighborhood in Abstraction Space¶
Akrasia sits in a moderately populated region (51st percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.
Family — Self, Identity & Social Cognition Biases (25 abstractions)
Nearest neighbors
- Appeal to Emotion — 0.86
- Emotive Conjugation — 0.86
- Performative Contradiction — 0.86
- Egoism — 0.86
- Door-in-the-Face Technique — 0.86
Computed from structural-signature embeddings · 2026-10-08
Not to Be Confused With¶
Weakness of will may refer to an over-ready reconsideration of intention even without action against better judgment. Temptation is a motivational circumstance, not proof of contrary action. Compulsion calls into question the available-alternative condition. Regret after an act is not evidence that the agent judged otherwise at the time. Procrastination can be akratic, but it need not be.[2]
References¶
[1] Aristotle, Nicomachean Ethics, Book VII, especially chapter 3 (sweet-object example and qualified knowledge) and chapter 9 (choice and judgment), primary historical discussion. registry ↩a ↩b ↩c ↩d ↩e
[2] Richard Holton, Intention and Weakness of Will, Journal of Philosophy 96 (1999), PDF pp. 9–11, especially the meat-eating friend and the distinction between akrasia and intention revision. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n