The Initial Aggressor

The Initial Aggressor

The qualifying-threat screen asks what the defender faced, and the aggressor limit asks how the defender came to face it. A defense built on necessity assumes that the danger arrived from outside the defender's own conduct, and that assumption weakens when the defender helped bring the danger about. Model Penal Code § 3.04(2)(b) therefore withholds deadly force from an actor who provoked the encounter in the way the Code describes, even when the threat that finally arrives is one of the listed harms.¹ This part fixes who counts as an aggressor, using the Code's wording as the anchor and the common-law definition as contrast, and then shows the two ways an aggressor can earn the defense back.

Provoking the Encounter

The Code's limit reads this way: deadly force is not justifiable if the actor, "with the purpose of causing death or serious bodily harm, provoked the use of force against himself in the same encounter."² The clause has three working parts, and the justification for deadly force is lost only when all three are present. The first is provocation, meaning conduct by the actor that brings on the other person's use of force. The second is purpose: when the actor provoked, the aim was to cause death or serious bodily harm. The third is unity of time and event, because the provocation and the force that followed must belong to the same encounter. The clause also reaches only deadly force, so under this provision a person who provoked a quarrel may still answer with force below the deadly tier, within the ceiling of what is reasonably necessary.

Each part has its own reason, and the purpose requirement carries the incentive logic. Consider a person who wants to kill someone and also wants to avoid punishment. Without this limit, that person could engineer the very necessity the defense requires: goad the target into drawing a weapon or throwing a punch, kill in response, and plead self-defense. A defense meant to protect people from danger would become a method of arranging a killing, and the planner would be the one it rewarded. The limit closes that route by denying the justification to anyone who created the necessity in order to exploit it. The purpose element also keeps the limit narrow. A person who starts a quarrel out of anger or foolishness is at fault, but the fault is smaller than a plan to kill, and stripping such a person of all protection when the other side escalates to deadly force would make a bar-room argument cost the arguer any defense against being killed. The same-encounter element confines the forfeiture to the episode the provocation began, so a scheme that failed last month does not follow the actor into an unrelated confrontation today.

Common-law courts reach a similar place through the label aggressor, and they define it narrowly. The aggressor is generally the person whose own unlawful act, such as a physical attack or a threat of one, started the fight. Insults and other words alone ordinarily do not make a person an aggressor, and neither does conduct that was lawful. The two formulations look at different things. The Code looks to the actor's purpose and so reaches a schemer whose provocation was only a dare or a taunt, while the basic common-law definition looks to the kind of act and treats the person who threw the first punch as an aggressor even if no one planned a killing. Many courts also deny the defense to a person who deliberately provokes a fight in order to kill under its cover, which brings the common law close to the Code on the clearest case. At common law, a person who qualifies as an aggressor generally cannot use deadly force in self-defense until the defense is regained, as the next subpart explains.

Only Dale's purpose differs between the two scenes, and that fact decides the aggressor question. In the second scene the clause does not defeat him, although his claim still has to clear the other limits on deadly force, including whether he believed deadly force necessary against a listed harm and whether a safe retreat was available. Timing works the same way. Had Dale tried to provoke Rosa weeks earlier and failed, and had she attacked him today in an unrelated dispute, the earlier provocation would not belong to the same encounter, and the clause would not apply. Students over-apply the limit when they label as a forfeiting aggressor anyone who threw the first punch, because under the Code's wording starting a fight matters only when it was done with the purpose of killing or seriously injuring.

Regaining the Defense

Aggressor status describes how the encounter began and can change before it ends. The doctrine gives an aggressor two routes back to self-defense, and both rest on one idea: the necessity that justifies defensive force must arise from the other person's choices rather than from the aggressor's own conduct. The Code's text already builds part of this in, because the clause bars deadly force only for an actor who provoked with the purpose of causing death or serious bodily harm, so a person who started a fight with lesser aims is not barred by that clause when the other side makes a deadly threat.⁴ The two routes matter most under the common-law definition, which treats the person who struck first as an aggressor even without a deadly plan.

The first route is escalation. A nondeadly aggressor, meaning a person who began the confrontation with force or threats short of deadly force, who meets sudden and unexpected deadly force in return may defend with deadly force. The reason is proportion, the same principle that separates the two tiers of force. The aggressor's wrong was a scuffle, and a scuffle gives the other person no license to kill. Once the other side answers with deadly force, that person has gone past what the situation justified, and the aggressor now faces exactly the kind of threat that the qualifying-threat list describes. The word unexpected carries weight, because an aggressor who starts a fight knowing the other person is armed and ready to use the weapon has less claim to call the response an escalation and more reason to see it as the consequence the aggressor invited.

The second route is withdrawal. An aggressor who breaks off the encounter in good faith and makes that fact known to the other person regains the right to defend if the other person then continues the attack, because the continued attack is now a fresh danger of the other person's making. Each half of the requirement has its own reason. Good faith matters because a tactical pause, taken to catch breath or to wait for a better position, ends nothing and leaves the aggressor's original necessity in place. Communication matters because the other person cannot read the aggressor's mind and must act on how the situation appears. Someone who still faces an armed opponent and has heard no word of surrender is reasonably entitled to treat the fight as continuing. The communication requirement also removes an incentive: if silent withdrawal sufficed, any aggressor who was losing could claim afterward to have quit in his own head before the fatal blow. Withdrawal is the only route open to an aggressor whose first force was itself deadly, since that person cannot point to an escalation.

The two scenes turn on whether the other person had any way to know the fight was over. The usual error is to treat an aggressor's change of heart as enough, when the withdrawal must be made known to the other person before the law will say the necessity has changed hands. The checklist below runs the whole aggressor inquiry in order.

Clearing the aggressor limit is necessary and still leaves the defense unfinished, because the same subsection also withholds deadly force from an actor who knows that retreat in complete safety is available. The next part turns to that duty to retreat.

Provoking the Encounter

The Code denies deadly force to an actor who, "with the purpose of causing death or serious bodily harm, provoked the use of force against himself in the same encounter."⁷ Three conditions must be present together before the limit applies. The actor must have provoked, meaning the actor's own conduct brought on the other person's use of force. The provocation must have had a particular purpose, the aim of causing death or serious bodily harm. And the provocation and the force it drew must belong to the same encounter. The clause sits in the subsection governing deadly force, so it screens only force of that tier, which the Code defines as force the actor uses with the purpose of causing, or knows to create a substantial risk of causing, death or serious bodily harm.⁸ Each condition has its own job, and they are easiest to understand by asking what the limit is built to stop.

The purpose requirement carries the incentive logic. Self-defense rests on necessity: the defender faces a danger that someone else created and has no lawful way out except force. A person who wants to kill and also wants to escape punishment has a reason to counterfeit that necessity by goading the target into drawing a weapon or throwing a punch, killing in response, and then claiming self-defense. If the defense were open to that person, it would reward the very planning it ought to defeat. Denying the justification to anyone who provoked force for the purpose of causing death or serious injury removes the temptation at the source, because the necessity was manufactured by the actor. The same element also keeps the limit narrow. Losing the justification for deadly force is a severe consequence, since it leaves the actor undefended even against a genuinely lethal threat, and the Code reserves it for provocation aimed at producing a deadly outcome. A person who starts a shoving match out of temper has done wrong, but the other side's escalation to deadly force is a danger that person did not plan, and the Code leaves that person protected against it. The same-encounter requirement confines the loss in time, so an abandoned scheme from an earlier day or an earlier dispute does not follow the actor into a fresh confrontation.

Common-law courts reach a similar place through the label aggressor. The first aggressor, meaning the person who provoked the confrontation, cannot claim self-defense, and the court in Peterson, a District of Columbia decision, added that this holds for as long as the person keeps that role in the affray.⁹ The common law defines the role narrowly: the aggressor is one who intentionally provokes a violent encounter, ordinarily by attacking or threatening violence, and a person who merely trespasses, steals without violence, or insults another does not become an aggressor when the other party answers with force.¹⁰ The two formulations look at different things. The Code asks about the actor's purpose, and because its wording does not confine provocation to violent acts, it can reach a taunt designed to draw out a killing.¹¹ The common-law definition asks whether the actor deliberately started a violent encounter, so it reaches the person who threw the first punch even without any plan to kill, but it does not reach words alone. How a nondeadly aggressor gets back into the defense is the subject of the next subsection.

The narrow definition is what keeps ordinary wrongdoing from costing a person the right to defend life. Trespass, non-violent theft, and insults are wrongs, and the wrongdoer answers for them in the proper way, but none of them is a deliberate step into violence. Words and property offenses give the other party no license to use force against the person, so a knife attack in reply is itself unlawful force, and the person facing it stands where any other defender stands. The necessity in that scene was created by the attacker's choice to escalate, and the law does not strip the wrongdoer of the defense for a necessity the wrongdoer did not create.

The two scenes differ in whether the first conduct was a purposeful step toward violence. Garrett's provocation was aimed at producing a killing, and Imani's wrongs were not aimed at violence at all, so that single variable decides the aggressor question. The exam move is therefore to ask what the provocation was for. Students lose this issue when they treat any earlier wrongdoing as aggression, because trespass, theft, and insults standing alone do not forfeit the defense. Losing the perfect defense under this limit also does not always mean liability for murder, since some jurisdictions treat a killing by an aggressor who never intended a deadly encounter more leniently, a middle ground this chapter reaches under imperfect self-defense.

The third step does the real work, because provocation and timing are usually visible on the facts while purpose must be inferred from what the actor said, knew, and arranged beforehand. In a common-law jurisdiction the same sequence runs with a different second and third step: ask whether the actor intentionally provoked a violent encounter, and treat insults, trespass, and non-violent theft as insufficient.

Regaining the Defense

Aggressor status disables self-defense only while the necessity to use force remains the aggressor's own creation. The common-law formulation therefore carries two exceptions: a nondeadly aggressor who is met with unexpected deadly force may defend against it, and an aggressor who withdraws may defend against an attack the other person then continues.¹⁴ Each exception marks a moment when the danger stops being the aggressor's doing. In the first, the other person has chosen a level of force that the aggressor's conduct did not call for. In the second, the aggressor has ended the aggression, so whatever follows is a fresh choice by the other person. The two routes differ in who may use them and in what they demand, and the paragraphs below take them in turn.

The escalation route requires that the aggressor began with force short of deadly force, such as a punch or a shove, and that the other person answered with a weapon the aggressor had no reason to expect.¹⁵ Once the answer jumps a tier, the aggressor faces the kind of threat that would justify deadly force for anyone, and the person who escalated has become the source of the danger. Only a nondeadly aggressor can use this route. An aggressor who opened with deadly force cannot call a deadly reply an escalation, because that reply is the answer the first act invited.

Jurisdictions do not all give the escalation route the same effect. The exception as stated above restores the defense completely.¹⁶ Some courts instead hold that a person whose battery precipitated the fight lacks a perfect self-defense even when forced to kill, and they reduce the killing from murder to voluntary manslaughter under the doctrine of imperfect self-defense, which this chapter reaches in its last part.¹⁷ The Model Penal Code is more generous to the ordinary brawler. Its aggressor clause bars deadly force only for a person who provoked with the purpose of causing death or serious bodily harm, so a person who threw a punch without that purpose can claim perfect self-defense against a deadly reply while still answering for the original battery.¹⁸ The generosity stops at felonious aggression: a person who begins with a felonious knife attack, is turned on by the intended victim, and then shoots that person is guilty of murder rather than manslaughter.¹⁹

The second route turns on withdrawal, and the law asks for more than a change of mind. The aggressor must withdraw totally, not merely retreat as a matter of strategy, and the withdrawal must be communicated in a way the other party can perceive.²⁰ If the other party continues to attack after that, the continued attack is unlawful force against someone who has stopped fighting, and the former aggressor may answer it with force, deadly force included where necessary.²¹ Totality matters because a pause to catch breath, reach a weapon, or find better ground ends nothing, and the necessity the aggressor created is still running. An attacker's departure can be tactical in the same way. In one illustration, a person attacked by someone who then turns and leaves may no longer use force once the danger is over, but if that person reasonably believes the attacker is leaving only to bring back friends for a group assault, the departure is not a true withdrawal and the threat continues.²²

A standard illustration shows both requirements met. A patient in a therapist's waiting room draws a knife on a nurse and screams threats, and the nurse picks up a heavy replica sword and holds it in front. The patient hands the knife to the nurse, kneels, and begs forgiveness. The nurse, who should realize the danger has passed but does not, swings the sword and narrowly misses, and the patient grabs the knife back and stabs. Because the patient's aggression had ended before the swing, the nurse's response was unreasonable, and the patient may plead self-defense as though never an aggressor, provided the stab was proportionate to the threat the patient reasonably perceived.²³ These facts come from an illustration rather than a decided case, but they show each element at work: surrendering the weapon supplies the totality, and kneeling and speaking are what let the nurse perceive that the fight was over.

The two scenes differ in whether the retreat ended the fight in a way the other person could see. Students lose this issue when they treat any step backward as a withdrawal, since only a total and communicated withdrawal restores the defense. The communication element also guards against a predictable abuse. When the defendant has killed, the only other person who could say whether a surrender was communicated is dead, and the defendant has every reason to claim that the fight ended in the defendant's own mind before the fatal blow. That room for manipulation is why juries often reject withdrawal claims, and why the communication element carries so much weight.²⁴

Regaining the defense restores the ordinary right to defend and nothing more. Whichever route the actor used, the force is still tested by the limits that govern every defender: it may not exceed what is reasonably necessary, deadly force requires a deadly threat, and in some states a safe retreat must be taken first.²⁵ Initial-aggressor status also bears on that last limit, because the Model Penal Code's exception to retreat for the actor's own dwelling or workplace is withheld from the initial aggressor, a point developed in the discussion of the home exception below.²⁶