Punishment via Kinsella’s dialogical estoppel — the puncher cannot coherently object to being punched — with proportionality as “an eye for an eye, not an eye for a tooth”; two objections (particularisation, changing one’s mind) answered; then the mens rea verdict — restitution on actus reus, retribution requires mens rea; then defense as exclusion by the means required and no more, the ball on the lawn, and Borer’s human-body sword and shield. Findings. (1) The estoppel argument is the weak retortion form the vault already flagged — doing X presupposes “I should do X” — and his smoker and umbrella passages are the attempt to save it, which fails on the praxeological/normative equivocation it needs. (2) On mens rea he and Chris agree on classification (actus reus in every trespass, knowing or not) and differ only on where mens rea lives — LZ codifies it for retribution, Chris puts it in the victim’s discretion; that is precisely the mens rea page’s resolution line. (3) Defense is the vault’s gentleness principle and causationism, nearly verbatim. (4) The course never defines “initiation” for threats — no brandishing, no imminence, nothing on firing near someone; Kinsella’s Limits of Armchair Theorizing: The Case of Threats is cited once (ch. 2) and the whole question is deferred to judges. Hub question 13 has no answer in his corpus.
Source: liquidzulu-wiki → src/content/libertarian-ethics/defensive-force-and-proportionality.md (n: 6, 2023-04-25, ~5.0k words) · live at liquidzulu.github.io/defensive-force-and-proportionality. CC0; quoted, not copied. Primary sources: Kinsella, Punishment and Proportionality: The Estoppel Approach (1996); Borer, The Human Body Sword (2010).
Links: LiquidZulu hub, Ch. 5 — Children, Ch. 2 — The NAP — where causation/foreseeability entered, Defensive Force and Collateral Damage — the vault twin: causationism, defense/retribution, gentleness, the imminence line, Mens Rea — Resolution: Where Flexibility Lives, Is-Ought — the repaired retortion form — the test estoppel fails, Restitution for Death — where “same form” breaks, The Objectivism Boundary — “the oughts are hypothetical” vs his “ought is ought”
LZ (via Kinsella): an unchallenged punishment is just per se (no conflict where wills agree); a challenged one must survive the nature of argumentation. A man is estopped — “his own act stoppeth his mouth” (Coke) — from claims inconsistent with his actions, not by a judge but “by the very nature of dialectic.” A punches B; B (or a hired champion C) punches back; A objects that punching is impermissible — but A “has pre-supposed that punching people sans-consent is just,” so his objection is incoherent and the punishment reverts to unchallenged. Proportionality falls out: A presupposed only that punching is proper, not murder or rose-tearing, so “the retribution must come in the same form as the aggression” — an eye for an eye, not an eye for a tooth.
(vault) — this is the weak retortion form, named on the is-ought page. The vault’s repaired form says the self-defeat lives in the authority of an assertion, not in the fact that an act occurred: “doing X does not entail you should do X — I breathe while asserting ‘one should not breathe’ and nothing follows.” Estoppel needs the act of punching to be the assertion “punching is just.” But an act reveals a preference (praxeological: A chose to punch rather than not), not a norm (A holds that punching is permissible for anyone). A thief need not believe theft is just; most do not. Kinsella’s estoppel works cleanly in one case — when the criminal argues “punching is fine” and then objects to being punched; there he is estopped by his assertion, and the vault agrees. LZ’s version extends it to the act itself, and that extension is the whole load. “Same form” proportionality inherits the problem and adds one: what is the same form of fraud, of rape, of murder? The restitution-for-death page calls that the framework’s hardest case; “an eye for an eye” does not reach it. The vault’s own rule is restitution automatic; retribution set by contract, common law, or the victim’s discretion — which is where “same form” belongs, as a default, not as a derivation.
LZ: Particularisation (“punching is fine for me, not for others”) fails because argued norms must be universalisable (Hoppe; the Golden Rule / categorical imperative), and denying that is polylogism, refuted in ch. 1; and if particularisation were allowed, B could particularise his punishment too. Changing one’s mind fails three ways: might-plus-regret would excuse everyone; denouncing your aggression concedes it is punishable; and the presupposition “punching is legitimate” was fixed by the act and cannot be dropped, so only the objection can be — “he is estopped by definition.” The smoker who says he should quit: acting is believing you should act, and holding the contrary belief too is just cognitive dissonance (“anyone who has encountered stupid people”). The umbrella: “it is raining, so you ought bring an umbrella” is either a non-sequitur (logical ought) or an is-statement in disguise (prudential advice) — “there is only ought; ought is ought. There doesn’t exist different flavours of ought in logic.”
(vault) — the second reply is good; the third is the weak form again, and the umbrella passage is the seam with Trunk 1. Reply 2 (to denounce your aggression is to concede it is punishable) is sound and sufficient on its own. Reply 3 restates that the act is an unretractable assertion, which is the equivocation of §1 made explicit: the smoker shows that choosing X is compatible with holding ¬X-ought, and LZ concedes this, then says the chosen-X “belief” still stands — but a chosen-X is not a belief about permissibility for others; it is a preference. Universalisability (a legitimate demand on argued norms) is being applied to an act. And “ought is ought” is the direct denial of the vault’s Trunk-1 position that all oughts are hypothetical, conditional on the structure of agency (Objectivism Boundary: “the oughts are hypothetical, and there are no unchosen duties”). His umbrella analysis is right — prudential ought reduces to a means-end is — and the vault says the “logical” ought has the same shape, only with the agent’s constitutive ends in the antecedent. He promises the categorical ought for ch. 7. That is the chapter to watch for it.
LZ: only one who chose the aggression is estopped, so retribution requires mens rea; but actus reus is present in every trespasser, knowing or not, so restitution is justified on actus reus alone (5 ounces back; a further 5 as retribution if the theft was knowing). Mens rea does not require knowing the act was wrong — the Nazi guard who shoots an escaping prisoner “knowingly invaded that Jew’s body,” and believing German law justified it does not absolve him.
(vault) — closer to Chris than either the video or ch. 2 suggested, and the residual difference is the mens rea page’s own resolution line. Compare the Mens Rea page: the law stays clean — aggression is binary, restitution + retribution are owed, no intent analysis; the people provide flexibility — victim’s discretion, community judgement, chosen arbitrators; mercy is voluntary, not codified. On classification LZ now agrees with Chris outright: the unknowing hiker is a trespasser; actus reus is intent-blind. On restitution, agreement. The only split is retribution: LZ writes mens rea into the law as its precondition; Chris keeps the law intent-blind and lets the victim’s discretion do what mens rea does. Same outputs in the ordinary case (the accidental trespasser is not punished either way); different institution. That is exactly the line the debate ended on — “the market would reinvent mens rea; yes, and that’s fine, codifying it is not.” Note the tension with ch. 2, where foreseeability governed whether an outcome is an action at all (the tornado-knife “is not a stabbing”); ch. 6’s “actus reus in every trespass, knowing or not” is the intent-blind rule ch. 2 did not use. The two can be reconciled (ch. 2 is about causation, ch. 6 about knowledge), but the reconciliation puts foreseeability — an intent-adjacent standard — back at classification by the causation door. The Nazi-guard clarification is good and matches Chris’s binary: the invasion is objective; belief in its lawfulness is irrelevant.
(Chris) The mens rea point lands, finally — and “retribution requires mens rea” is exactly where we disagree. There also has to be room for the punitive independent of the mind: take someone’s car and put it back while they are away, and the property is fully restored, yet restoration is plainly not enough. Whatever is owed beyond restoration is owed because of what was done to the owner (the taking, the risk, the time), not because of what the taker knew. (vault) On LZ’s scheme the joyrider has mens rea, so retribution applies — but the point generalises to the unknowing invader: restitution returns the object and nothing compensates the invasion itself. That is the gap his actus-reus/mens-rea split leaves, and the vault’s rule (retribution by contract, common law, or the victim’s discretion) fills it without asking what anyone knew.
(Chris) — and the appeal to a judge. I thought we could figure out law as a fact of the matter — where does judgement play a part? (vault) This is the ch. 1 seam surfacing where it costs him: the armchair-complete code needs a judge for every classification the axiom cannot make (threats, ch. 6 §6; whether a given act was aggression, ch. 2 §5; a child’s capacity, ch. 5). Each time, “the judge is necessary” is discovery-as-generation admitted and never named.
LZ: a property right in X is the right to exclude others from X, so defense = “whatever means are required to go about excluding people from X” — and no more: you may “possess him to the extent that is required to remove him from your lawn.” Informing suffices for the man who did not know; the stubborn one may be walked off by the arm; the needle-stabbing serves no exclusion and is aggression; the rocket only if required. The ball: if you do not try to exclude it but refuse to return it, you are forestalling; if you had fenced and the ball came over anyway, “antecedent rights must prevail” — you may exclude your neighbour from stepping on, and you gain a limited ownership of the invading ball (“not full-monte”; it reverts when removed) — this resolves Dominiak’s donut contradiction; not if you kicked it there.
(vault) — this is the gentleness principle, and one more patch. Required to exclude, and no more = the vault’s gentleness principle clause (b): the force used is the minimum necessary, assessed against actual alternatives. Agreement, and stated well. Two notes. (a) “Antecedent rights must prevail” is the concept he declined in ch. 3 §6 (“workable … a little bit tedious”) — now used by name to settle the ball. Add it to the tally of patches the theory forbids itself. (b) “Limited ownership of the ball, for the purpose of exclusion, reverting on removal” is the bundle view again — an incident of ownership allocated to a purpose — which his “ownership = just direction” cannot say without this ad hoc rider.
(Chris) — the ball, pressed. If your property ends up on another’s by accident, what happens? Just restoration — the ball goes back and that is it? No respect for the time and effort of the person invaded? And do we punish the trespass of someone recovering their own property? (vault) On the chapter’s rules: the ball is an invasion (actus reus, no mens rea) → restitution only, i.e. the ball’s removal; the lawn-owner’s time is uncompensated because nothing beyond the object was “taken”; and the neighbour who steps over to retrieve it commits a second invasion (“antecedent rights must prevail”), for which — if he knew the fence was there — mens rea attaches and he is punishable while the original ball-invasion was not. So the accidental invader’s recovery is worse in law than his invasion, and the invaded party is owed nothing for either. That inversion is what Chris’s punitive point fixes: retribution scaled to the invasion, not the mind.
LZ (via Borer): damage the stolen sword if required to stop the stab; damage the neighbour if the criminal wields him as the sword; shoot through the stolen quilt, and through the neighbour used as a shield — “your rights are antecedent to those of your neighbour. The aggressor brought his property into conflict”; the neighbour may not attack you to save his quilt or himself, because you are doing no wrong; the device-wired neighbour likewise. But bystanders are off limits — shooting one during the gunfight, or killing one under the criminal’s ultimatum, is fresh aggression; exclusion is only against the aggressor, and third-party damage is justified only when required to exclude him.
(vault) — the vault’s causationism, near-verbatim. The defensive-force page: the aggressor is the proximate causal agent of the entire causal structure they create; in the baby-armor case he bears responsibility for harm to the shield from legitimate defensive force. Same conclusion, same reason, same bystander limit, same defense/retribution split (his estoppel-proportionality confines retribution to the aggressor in the same form; the vault’s table says retribution “is not a license to harm bystanders”). Agreement across the board — and this is the section where his causation theory (persons as means, ch. 2) does its legitimate work, because here it assigns responsibility to the aggressor, not liability to a second free agent.
(vault) The chapter’s title is Defensive Force, and it contains no account of when defense may begin. Every case is an invasion in progress or completed: a trespasser on the lawn, a ball over the fence, a stab, a shot. There is nothing on brandishing, threats, imminence, or firing near someone — the vault’s imminence line (open question 1 there) and Chris’s hub question 13 (is shooting near someone aggression? if brandishing is initiation, “conflict” is wider than mutually exclusive use of a scarce means, or there is a second root). The course’s only touches on the subject are the stalemate rule in ch. 3 (“forcefully excludes” undefined) and a single ch. 2 footnote to Kinsella’s The Limits of Armchair Theorizing: The Case of Threats, with the comment that judges are necessary because the armchair cannot classify acts. So the corpus’s answer to “what is initiation?” is: the latecomer’s taking; the stalemate breaker’s force; and for everything else, ask a judge. That is discovery-as-generation conceded at the exact point where the axiom was supposed to do the work.
What holds. Reply 2 to the mind-changer (denouncing your act concedes it is punishable); mens rea not required for classification or restitution — Chris’s rule; the Nazi-guard point; defense as required to exclude and no more — the gentleness principle; the human-body sword and shield — causationism; bystanders excluded.
What does not. Estoppel-from-the-act is the weak retortion form, and the smoker passage concedes the distinction it needs to deny (§1–2); “same form” proportionality does not reach the cases that matter (§1); “ought is ought” is the seam with Trunk 1, promised for ch. 7 (§2); mens rea is codified for retribution where Chris leaves it to discretion — the debate’s own resolution line (§3); “antecedent rights” returns as a patch after being declined (§4); and there is no theory of threats (§6).
One-line read for the hub: on defense he and the vault are twins; on punishment his estoppel is the weak retortion; on mens rea the only difference is where it lives; and the course never says when defense may begin.
defensive-force-and-collateral-damage.md.