Concept
Provocation and Aggression
Self-defense and defense of others
Provocation and Aggression
Self-defense protects a person who answers an attack, not a person who arranged the occasion for force. Restatement (Second) of Torts § 63(1) privileges reasonable force against an unprivileged harmful or offensive contact or other bodily harm that the actor reasonably believes another is about to inflict intentionally.1 The earlier sections measured the threat, the actor’s belief, the purpose of the force, and its amount at the moment of confrontation. This section looks at what came before that moment. A hostile remark or a first blow can change what the actor may do when the other party responds.
Two ideas explain why. The first is necessity: a defensive killing is unnecessary if the occasion for it could have been averted,2 and a person who helped create the emergency owns part of its necessity. The second is pretext. The actor is not privileged when he uses the other’s attempt to inflict a contact as a mere excuse for applying force, although telling genuine defense from excuse may present a difficult question of fact.3
Earlier conduct does not run in only one direction, and a first wrong does not forfeit every protection. A person who commits the first invasion remains liable for it, and he keeps the normal privilege to defend against any excessive force or confinement the defender then uses.4 The history also bears on the defender’s side. That the other has already committed or attempted an aggression is a circumstance to weigh in deciding whether the other’s later conduct gives the actor reason to believe the aggression will be renewed.5 Depending on what the conduct was and who engaged in it, earlier conduct can support the privilege, defeat it, or leave it untouched.
The first subsection asks whether insults, abusive words, or threats provoke a physical response or justify one. The second takes up the person who started the physical confrontation and asks when that person regains the privilege.
Words as Provocation
A hostile remark is not a contact, and an insult by itself gives no sign that a contact is coming. Section 63(1) reaches only a contact or bodily harm the actor reasonably expects, so words alone provoke without privileging.6 Provocation, as used here, means conduct or language that angers another person and invites a response without itself threatening any touching. Mere provocation does not trigger the privilege: verbal abuse, insults, and threats of future harm ordinarily do not justify a physical attack in retaliation.7 The same limit covers force used to punish a past aggression or to warn against its repetition.8
Cases such as Crotteau v. Karlgaard and Prell Hotel Corp. v. Antonacci reflect that rule, which does not depend on how vile the words were. A test that did would send courts to rank insults by their sting and would leave every quarrel one remark away from a justified assault.
The reason lies in what the privilege is for. Self-help is tolerated when an imminent attack leaves no time for ordinary remedies, and an insult leaves that time. Nothing physical is pending, so the target has peaceful legal remedies and must use them instead of “taking the law into [his] own hands” by immediate physical force.9 Words injure feelings and reputation, and the law answers those injuries in other ways. A rule that measured the permissible response by the listener’s temper would let every remark justify a blow and every blow a counterblow, until each side could call its force defensive.
Threats of future harm fail on imminence. The Restatement allows force in self-defense only as a last resort, and not while there is a reasonable likelihood that the actor will have a later opportunity to protect himself.10 A person told of an attack at some later time can resort to peaceful legal remedies before then, and must.11
Words stop being mere provocation when the circumstances give the target reasonable grounds to expect an immediate contact. The privilege rests on that reasonable belief and not on a blow actually landing, and a reasonable mistake protects the defender even though force was not in fact necessary. In Keep v. Quailman, the plaintiff, who had a reputation for shooting people, suddenly put his hand in his pocket, and the defendant, believing he was about to draw a revolver, struck him on the head with a cane. The court held that the jury should have been told the defendant was liable only if he lacked reasonable grounds to fear an immediate attack.12
The Restatement’s illustrations show how earlier threats feed that judgment. A, a notorious desperado, has threatened to shoot B on sight, and B sees A approaching with a hand in a hip pocket. A is only reaching for a handkerchief, but B mistakenly and reasonably believes A is about to shoot, so B’s blow is privileged. An abnormally timid B who unreasonably believes A is his mortal enemy has no privilege.13 The desperado’s earlier words justified nothing by themselves; they gave meaning to what B saw. In the same way, statements can give another’s conduct a threatening appearance that the conduct would lack standing alone.14 See also Martin v. Estrella. Once the privilege arises, the force must still be proportionate to the danger threatened.15
Provocation that fails to justify a battery may still matter to damages. Some courts treat the plaintiff’s provoking words as no defense to liability but as relevant to the remedy, as in Manning v. Michael. The two kinds of damages answer different questions. Compensatory damages repay the injury, and a broken jaw is the same injury whatever the defendant was called beforehand. Punitive damages respond to blameworthiness, and a defendant who lashed out after being goaded is less culpable than one who struck without cause, though still a tortfeasor.
Aggressors and Withdrawal
An initial aggressor is the person whose own conduct started the confrontation, and the question here is when that person may defend against it. In tort, an original wrongdoer who meets excessive force keeps the normal privilege against the excess and remains liable for the original invasion.17 That rule does not settle the harder case, in which the victim used no excess and the aggressor simply wants out of the fight he began. Tort cases on the point are infrequent, and courts carry the criminal rules over with little variation, so the criminal formulation is the place to begin.18
In United States v. Peterson, a homicide case, the court stated the rule this way: one who is the aggressor in a conflict culminating in death cannot invoke the necessities of self-preservation, and only by communicating to his adversary his intent to withdraw and in good faith attempting to do so is he restored to the right of self-defense.19 The trial judge’s instruction in that case shows both halves. Mere words do not constitute provocation or aggression. If Peterson precipitated the altercation but then withdrew in good faith and told Keitt so by words or acts, he was justified in using deadly force against imminent danger of death or grave bodily harm. And even if Keitt was the aggressor, Peterson could use no more force than he reasonably and actually believed necessary.20
The rule has two parts, each with its own reason. The exclusion rests on necessity: the aggressor could have averted the occasion for force by not creating it. The route back rests on what the person attacked can know. A step backward looks the same whether it ends the fight or only pauses it, and an earlier aggression is a circumstance bearing on whether later conduct shows an intent to renew it.21 That is why withdrawal must be communicated, and why a feigned retreat earns nothing. The person who began a murderous assault has forfeited the privilege only for the moment, and cannot reacquire it by “retreat to the wall.” That person must bring the attack to an end, and if unable to get entirely away must convey in some manner that the fight is over. If circumstances do not permit this, the misfortune is the aggressor’s own, for beginning such a predicament.22
Allen v. State shows the rule defeating a tactical withdrawal. Leathers drove off after striking Allen in the face with a garden rake when Allen pursued her to her car, asking her to stay. Allen followed in her own car, parked, and walked toward Leathers’s vehicle, hoping, she claimed, to persuade Leathers to reconsider. When she saw Leathers coming toward her holding the rake, she went back to her car, took a gun from the glove compartment, turned, and fired, killing Leathers with a single shot.23 The court drew the line directly: there must be a distinction between retreating to avoid a confrontation and withdrawing a short distance to obtain a tactical advantage, here the acquisition of a deadly weapon. Allen reached her car, and instead of escaping she grabbed a weapon, turned, and confronted her attacker.24 The court also treated her pursuit of Leathers as potentially provocative, even assuming she did not intend to provoke an argument.25
Edgar v. Emily marks the other side of the line. Even a person who was initially an aggressor has a right of self-defense against the person he first threatened once he has retreated.26 The difference between Allen and a true retreat lies in where the step leads and what it is for. Allen’s step led to a weapon and back to the confrontation. A retreat that ends the encounter leaves the aggressor free to resist new force.
The formulations diverge when the other party escalates. The common-law account associated with Sir Michael Foster sorted these cases into three: the person who may stand and defend with deadly force, the person who must retreat first if a reasonably safe retreat exists (unless in his own “castle”), and the person who must withdraw before resorting to deadly force. As to the last, little disagreement exists.27 A nonlethal aggressor who meets a deadly response stands differently. In a few states that person can regain the privilege after exhausting every reasonable means of escape short of deadly force, a proviso that echoes § 65, which requires a person attacked with deadly force to retreat if she may do so safely.28 Most states deny even that. They hold that self-defense belongs only to a person free from fault, so the initial aggressor’s choices are to run if possible, to forgo self-defense and be killed, or to fight back unlawfully and face homicide charges if forced to kill.29 The jurisdiction therefore decides the result, and an answer should say which formulation it applies.
Because being an aggressor costs so much, the label turns on conduct, and words alone do not earn it. The charge in Peterson said as much, and the conduct need not aim at killing: a person who by provocative behavior initiates a confrontation loses the right of self-defense even without any intention of killing the other person.30
Step three is where most wrong answers begin, because any movement away from the other person tends to be described as retreat. A step toward the exit and a step toward a weapon are different acts, and the cases turn on that difference. Step four then asks whether the other person could tell the difference, which is why communication is required.