The majority refuses to expose the police to tort liability for what it describes as “inadequate or unreasonable assistance to suicidal individuals” because it believes such liability “could inhibit them from providing intervention at all.” (Maj. opn., ante, at p. 273.) This conclusion rests on distortions of both the facts and the law.
The conduct of the police in this case was not merely “inadequate” and “unreasonable,” but mindlessly reckless. The decedent, shown to be a caring person who never hurt others, suffered periodic bouts of depression and had a drinking problem. Though he possessed a weapon, he had never in the past or at the time of his death used it to threaten others. The decedent had been in his backyard for over an hour before the police found him. A few minutes *290after they found him he was killed in a hail of bullets. The officers at the scene did not, as my colleagues claim, merely fail to prevent this death; as the evidence abundantly shows, and the jury found, the death was aggressively provoked. Subjecting the gross misconduct in this case to tort liability will not unduly inhibit law enforcement intervention or burden local government. As the Supreme Court found in an analogous situation, imposing liability will simply “promote careful work.” (Johnson v. State of California (1968) 69 Cal.2d 782, 793 [73 Cal.Rptr. 240, 447 P.2d 352].)
The majority purports to assess only the question of appellants’ legal duty. Finding none, it declines to discuss the issue of immunity, which it deems moot. What the majority has really done, however, is to eliminate a duty clearly established in our jurisprudence by creating what amounts to a new form of governmental immunity. This is accomplished not just by overlooking the voluntary assumption of duty in this case but also the many ways in which the conduct of the police created a “special relationship” resulting in a duty to use due care. As a result, the majority has virtually wiped out the special relationship doctrine as it applies to law enforcement officials in a broad class of cases, repudiating the views expressed by the Supreme Court in Williams v. State of California (1983) 34 Cal.3d 18 [192 Cal.Rptr. 233, 664 P.2d 137]. The majority would subject police officers to liability only when their conduct “constitutes an intentional tort or a violation of an individual’s constitutional or other federally protected rights.” (Maj. opn., ante, at p. 274.) That is not the law. Moreover, this differential treatment of the police cannot be reconciled with the mandate of the Legislature that “a public employee is liable for injury caused by his .act or omission to the same extent as a private person,” unless the Legislature has “otherwise provided by statute.” (Gov. Code, § 820, subd. (a).)
For the foregoing reasons, I respectfully dissent.
I.
A.
Preliminarily, the majority fails to make it clear that the “duty” at issue here relates not to the reasonableness of appellants’ conduct, but whether, as a threshold matter, they had an affirmative duty to prevent respondents’ injuries. (On the distinction between these duty analyses, see Marois v. Royal Investigation & Patrol, Inc. (1984) 162 Cal.App.3d 193, 198-199 [208 Cal.Rptr. 384].) Like appellants, the majority does not dispute that, as the jury found, appellants failed to exercise due care and their negligence was the cause of respondents’ injury. The majority reasons that the fact that *291respondents’ injuries were caused by appellants’ negligence is beside the point, because the police had no duty to prevent the injuries that occurred.
If by this argument my colleagues mean that the police have no enforceable legal duty to assist persons in danger, I agree. In California, as in virtually all other common law jurisdictions, there is no duty to rescue. As set forth in the Restatement Second of Torts, “The fact that the actor realizes or should realize that action on his part is necessary for another’s aid or protection does not of itself impose upon him a duty to take such action.” (Rest.2d Torts, § 314.) “As a general rule one has no duty to control the conduct of another, and no duty to warn those who may be endangered by such conduct.” (Peterson v. San Francisco Community College Dist. (1984) 36 Cal.3d 799, 806 [205 Cal.Rptr. 842, 685 P.2d 1193], citing Rest.2d Torts, § 315; Davidson v. City of Westminster (1982) 32 Cal.3d 197, 203 [185 Cal.Rptr. 252, 649 P.2d 894]; Thompson v. County of Alameda (1980) 27 Cal.3d 741, 751 [167 Cal.Rptr. 70, 614 P.2d 728, 12 A.L.R.4th 701]; Tarasoff v. Regents of University of California (1976) 17 Cal.3d 425, 435 [131 Cal.Rptr. 14, 551 P.2d 334, 83 A.L.R.3d 1166]; see also 6 Witkin, Summary of California Law (9th ed. 1988) Torts, § 858, p. 220 et seq. and cases there cited; Weinrib, The Case for a Duty to Rescue (1980) 90 Yale L.J. 247; Landes & Posner, Salvors, Finders, Good Samaritans, and Other Rescuers: An Economic Study of Law and Altruism (1978) 7 J. Legal Stud. 83; and Note (1972) The Duty to Rescue, 47 Ind. LJ. 321.)
There are, however, some well-established exceptions to this general rule of no duty, and the genuine question is whether any apply in this case. It is on this issue that my colleagues and I part company, as I believe there are two applicable exceptions.
“First, even when one is not under a duty to act to protect or aid another, if one voluntarily undertakes to do so, he or she will generally be under a duty to exercise reasonable care. Second, a person may in some instances be obligated to take certain affirmative steps to protect or aid another if that person stands in some ‘special relationship’ to either the person endangered or the person whose conduct may injure the person endangered.” (1 Levy et al., California Torts (1998) § 1.10[2], p. 1-42.10, fns. omitted.) As later discussed, these exceptions are much more likely to apply where, as here, the defendant’s misfeasance, as opposed to nonfeasance, is the basis of the claim of negligence (Marois v. Royal Investigation & Patrol, Inc., supra, 162 Cal.App.3d 193, 198); although “negligence may also constitute an omission or failure to act.” (Williams v. State of California, supra, 34 Cal.3d at p. 24, citing Morgan v. County of Yuba (1964) 230 Cal.App.2d 938 [41 Cal.Rptr. 508].)
*292California courts have repeatedly held that the absence of a duty to take affirmative action to assist or protect another, no matter how great the danger in which the other is placed, has no application where . . there is some relationship between them which gives rise to the duty to act.” (6 Witkin, Summary of Cal. Law, supra, Torts, § 858, p. 220, italics in original; see also Nally v. Grace Community Church (1988) 47 Cal.3d 278, 293 [253 Cal.Rptr. 97, 763 P.2d 948] [“. . . we have imposed a duty to prevent a foreseeable suicide only when a special relationship existed between the suicidal individual and the defendant or its agents.”].) The imposition of tort liability on the basis of such a “special relationship,” or because the duty was voluntarily assumed, has nothing to do with Rowland v. Christian (1968) 69 Cal.2d 108 [70 Cal.Rptr. 97, 443 P.2d 561, 32 A.L.R.3d 496], as the majority claims,1 because that case does not concern exceptions to a general rule of no duty. (See discussion, post, at p. 308 et seq.)
As should be apparent, the conduct of the police in this case created a situation of dependency resulting in a “special relationship” between the respondents who sought and obtained their assistance and the decedent on the one hand and appellants on the other. The imposition of liability is, however, independently justified by the voluntariness of the police assumption of duty and the manner in which the police on the scene exacerbated the peril that previously existed.
Section 323 of the Restatement Second of Torts provides as follows: “One who undertakes, gratuitously or for consideration, to render services to another which he should recognize as necessary for the protection of the other’s person or things, is subject to liability to the other for physical harm resulting from his failure to exercise reasonable care to perform his undertaking, if ft]] (a) his failure to exercise such care increases the risk of such harm, or HQ (b) the harm is suffered because of the other’s reliance upon the undertaking.” (Rest.2d Torts, § 323, italics added.) The fundamental idea is that “. . . the undertaking to rescue, although not required, gives rise to the duty to exercise care not to leave the object of the rescue in worse condition than if the rescue had not been attempted.” (3 Harper et al., The Law of Torts (2d ed. 1986) § 18.6, p. 722.)
*293Application to the police of the legal principle embodied in section 323 of the Restatement Second of Torts, and the connection between that principle and the “special relationship” doctrine is best illustrated by Williams v. State of California, supra, 34 Cal.3d 18, even though the court found in that case that the plaintiff had failed to satisfactorily state a cause of action. The plaintiff in Williams was injured when a piece of a heated brake drum from a passing truck was propelled through the windshield of her automobile. She alleged that the police officers who arrived at the scene and investigated the accident negligently failed to test the brake drum part to determine whether it was still hot, failed to secure the identity of witnesses, and failed to attempt pursuit of the owner of the truck, virtually destroying the plaintiff’s ability to obtain compensation for her injuries and damages. The trial court granted the state’s motion for judgment on the pleadings. The Supreme Court reversed.
The court first observed that the state highway patrol has the right but not the duty to investigate accidents, or to come to the aid of stranded motorists. (34 Cal.3d at p. 24.) The chief issue in Williams was whether a legal duty could nonetheless be imposed because, by intervening in the situation, the police assumed the responsibility to act reasonably in the circumstances and thereby created a “special relationship.” The court stated that, “although ‘no special relationship may exist between members of the California Highway Patrol and the motoring public generally, or between the Patrol and stranded motorists generally’ [citation], when the state, through its agents, voluntarily assumes a protective duty toward a certain member of the public and undertakes action on behalf of that member, thereby inducing reliance, it is held to the same standard of care as a private person or organization. [Citations.] ftQ The breach of duty may be an affirmative act which places the person in peril or increases the risk of harm as in McCorkle v. Los Angeles (1969) 70 Cal.2d 252 . . . , where an officer investigating an accident directed the plaintiff to follow him into the middle of the intersection where the plaintiff was hit by another car. The negligence may also constitute an omission or failure to act, as in Morgan v. County of Yuba (1964) 230 Cal.App.2d 938 . . . , where a deputy sheriff promised to warn a decedent if a prisoner, who had made threats on her life, was released. The county was held liable when the sheriff failed to warn.” (Ibid., italics added.)
The Williams court cited Mann v. State of California (1977) 70 Cal.App.3d 773 [139 Cal.Rptr. 82], as an example of a case in which a special relationship was based on such dependency. In that case, “[h]ighway patrolmen, coming to the aid of a stranded motorist, placed their car with flashing lights behind two cars stalled on the freeway. After calling the tow truck, the officers withdrew without warning; they did not wait for the tow truck to *294line up behind the stalled car or provide the alternative protection of flares. Minutes later the stalled car was sideswiped by a passing car and the persons nearby were injured.” (Williams v. State of California, supra, 34 Cal.3d at p. 25.) The crucial factors in Mann, according to the Williams court, were that “. . . the officers’ conduct contributed to, increased, and changed the risk which would have otherwise existed. They stopped to investigate and they took affirmative steps to provide assistance, lulling the injured parties into a false sense of security and perhaps preventing other assistance from being sought.” (Ibid.)
Applying well-established principles, the Supreme Court concluded in Williams that the plaintiff there failed to establish a duty of care owed by the police officers who arrived at the scene, because the officers “took no affirmative action which contributed to, increased, or changed the risk which would have otherwise existed; there is no indication that they voluntarily assumed any responsibility to protect plaintiff’s prospects for recovery by civil litigation; and there are no allegations of the requisite factors to a finding of special relationship, namely detrimental reliance by the plaintiff on the officers’ conduct, statements made by them which induced a false sense of security and thereby worsened her position.” (34 Cal.3d at pp. 27-28, fn. omitted.) (However, the Supreme Court directed the trial court to permit the plaintiff to file an amended complaint in light of the lower court’s misperception of the legal question presented, and since it could not be said the plaintiff could never state a cause of action. (Id., at p. 28.))
The situation in the present case is, of course, completely different from that in Williams. The officers here—who, unlike the police in Williams, were witnessing the commission of felonies dangerous to human life (Pen. Code, §§ 417, subd. (c), 417.8)—immediately asserted almost complete control over the situation, thereby limiting respondents’ ability to deal with it themselves or take other measures.
Although five other officers were already there, Sergeant Osawa’s first act was to request additional units. Without requesting permission, Osawa and three other officers, all of whom had their weapons drawn, cocked and ready to be fired, searched the house, refusing to permit respondents to enter the premises. Finding nothing in the house, the officers turned their attention to the backyard. Osawa loudly identified himself and his colleagues as Fremont police, twice called out Patrick’s name, and ordered him to come out with his hands in the air. When Patrick was found sitting under a bush with a gun pointed at himself, Sergeant Osawa never considered even temporarily withdrawing. Nor, despite the fact he had no training in negotiations with disturbed persons, did Sergeant Osawa then consult Officer Tajima-Shadle, *295who had such training and was on the scene. Osawa also declined to invite respondents or others friendly to Patrick to participate in their efforts to induce him to put down his weapon, which respondents’ experts testified was the appropriate course. Instead, the officers flipped a picnic table on its side, knelt behind it for protection, and again directed Patrick to come out with his hands up. Sergeant Osawa then threatened to send in a police dog if Patrick did not immediately comply with his order. When Patrick remained silent, officers released the dog, commanded it to search, and followed the dog toward Patrick with weapons drawn. When they saw Patrick sitting with a gun cradled in his arms and pointed at his chest, the officers directed him to “Freeze,” and “Drop the gun.” After the dog became excited and continued barking loudly, Patrick uttered his first words: “Get the fucking dog out of here” and “What are you going to do, fucking shoot me?” The officers then went back behind the picnic table. One of the officers then went back to Patrick, pointed a shotgun at him, and started talking. Other officers in the backyard and at the windows of the residence had guns drawn and were pointing searchlights at Patrick, so he would be unable to see them. Several officers testified they had decided to shoot Patrick if he made any moves they considered threatening. During this time Patrick kept saying, in effect, “Leave me alone. Get out of my yard. I don’t want to talk to you.”
Sergeant Osawa then directed Officer Lopes to leave his station in the adjacent backyard, because he might get caught in a crossfire. About a minute later, at a time when eight officers had weapons pointed at Patrick, thirty-four shots were fired over a period of from five to ten seconds. Patrick’s body was pierced by 27 bullets, one of which came from his own gun. At all material times, respondents were restrained by the police from entering the backyard to intervene, as respondent Adams attempted to do.
By this conduct the police controlled the environment of the threatened suicide as completely as was possible. Respondents, who had been excluded from the premises and compelled to rely upon Sergeant Osawa and the numerous officers he was commanding, were not in a position to tell the police to leave and try to deal with Patrick themselves, solicit the intervention of friends, or simply do nothing and hope Patrick would recover his senses, as he had in the past. The peremptory assertion of such total control over the situation by the police, and the exclusion of respondents from any meaningful role in the attempt to dissuade Patrick from harming himself, clearly created the “situation of dependency” described by the Supreme *296Court in Williams v. California, supra, 34 Cal.3d at page 25, which “results in detrimental reliance [on the police] for protection.” (Ibid.)2
Not only did the police create a situation of dependency, which would be enough, as in Mann they also took affirmative action materially increasing the risk that previously existed. In addition, they changed the nature of the risk that previously existed by creating the possibility Patrick might be injured or killed by the police, or provoked into killing himself.
The evidence which most devastatingly establishes that the police significantly increased the risk of harm in this case was the testimony of respondents’ experts. Peter Reedy, a retired police officer trained by the FBI, who taught crisis management and been involved in “sixty to eighty” negotiations in hostage and suicide incidents involving persons under the influence of alcohol or drugs, testified at considerable length. In his view, Sergeant Osawa violated virtually every relevant law enforcement protocol, including those of the Fremont Police Department. By threatening the use of deadly force much too precipitiously and aggressively, the police dangerously increased the level of anxiety and tension, which is the opposite of what proper police practice calls for. This testimony was buttressed by that of Dr. Robert E. Litman, who specializes in the study of suicide prevention and *297lectures to law enforcement agencies. Dr. Litman explained at length why, in his opinion, “the police were a major cause, a substantial cause [of Patrick’s suicide.]”
There can be no doubt that the jury accepted the testimony of respondents’ experts and rejected the opposing views of Joseph Callahan, a “consultant” with associate of arts degrees in “mortuary science” and “police science” who “lectures to police and military groups on issues of tactics,” and Dr. Donald Lunde, a psychiatrist, who testified in behalf of appellants. The jury specified 13 ways in which Sergeant Osawa and his “SWAT” team unnecessarily inflamed the situation, increasing the danger Patrick might shoot himself and creating the new and different danger that he might unnecessarily be shot by the police: “[1] Lacked control of the officers. [2] Insufficient communications. [3] Lack of information. [4] Did not respond to suicide call as such. It was an assault response rather than assist. [5] Did not follow Fremont Police Dept, procedures for dealing with a critical incident. [6] Delayed calling in medical help. [7] Decision to use dog prior to using a negotiator. [8] Allowed untrained officers to attempt negotiation. [9] Did not evacuate all the neighbors. [10] Did not maintain the psychological sanctity of the family members at the scene. [11] The use of 7 armed officers left no option but force. [12] Once location of Pat was known, did not back down to allow calming of situation. [13] Yelling and shouting at Pat did not allow for calm.”
The majority unjustifiably attempts to brush this evidence aside. Claiming that the question of duty presents a pure “question of law to be determined by the court alone” (maj. opn., ante, at p. 265), the majority initially takes the position that the most important factual findings are irrelevant. According to the majority, we would “abdicate our distinct role” if we considered either “the inapposite findings of the jury” or the “testimony of respondents’ expert witnesses, who testified that the police caused Patrick’s suicide and violated the applicable standard of care by increasing the anxiety level at the scene or rushing the situation.” (Ibid.) By thus conveniently eliminating consideration of the findings and evidence which most powerfully shows that the conduct of the police created a “special relationship,” the majority concludes there was no such relationship and therefore no duty. But the facts cannot so easily be dismissed.
Legal rules are no more than conditional statements referring to supposed facts. The Restatement Second of Torts declares, for example, that the word “duty” is used “to denote the fact that the actor is required to conduct himself in a particular manner at the risk that if he does not do so he becomes subject to liability to another to whom the duty is owed for any *298injury sustained by such other, of which the actor’s conduct is a legal cause.” (Rest.2d Torts, § 4, p. 7.) The purely legal rule, which defines the “particular manner” in which an actor must ordinarily conduct himself, does not, however, always fully determine the existence of a duty. Whether the duty exists depends in part upon whether the actor conducted himself in the appropriate manner, which is, of course, a factual question. Thus, as has been stated, “[t]he duty issue frequently poses questions of the kind usually given to the jury. Under the prevailing rule duty to use due care is bounded by the foreseeable range of danger. Reasonable foreseeability of harm is the very prototype of the question a jury must pass on in particularizing the standard of conduct in the case before it.” (3 Harper et al., The Law of Torts, supra, § 18.8, p. 744.)3 Moreover, “. . . the question of foreseeability always involves more than the determination of simple facts—i.e., what the parties did or did not do, and what the surrounding circumstances were. It also involves a determination of what the parties should have perceived under those circumstances, i.e., whether the reasonably prudent person in the shoes of [the] party would have recognized unreasonable danger to the plaintiff from the source of harm or hazard that befell him. Within broad limits . . . this question is generally also one for the jury. [^Q The Restatement [Second of Torts] clearly recognizes that the jury may be called [upon] to make evaluations as well as to find simple facts—to decide what the parties should have done as well as what they did do.” (Id., at p. 747, italics in original, citing Rest.2d Torts, § 323C.)
The question of duty cannot be resolved in this case without resort to both the facts of the situation in which the parties found themselves and an evaluation of what the police on the scene should have perceived and should have done in the context of that situation.
As we have seen, the “special relationship” which gives rise to a duty may be created when police conduct “contributed to, increased, or changed the risk which would have otherwise existed.” (Williams v. State of California, supra, 34 Cal.3d at p. 27.) The evidence which most clearly shows what the police did or did not do in this case, the dangers they should have perceived, and the action, if any, they should have taken—all of which relate to whether *299the police unreasonably contributed to, increased or changed the preexisting risk, and therefore bear upon the question of duty—consists primarily of the testimony of the experts. The trial court acknowledged that the question of duty could not be resolved without a jury determination of these factual questions. So too did appellants, who never made any objection to the presentation of these issues to the jury. Nor did appellants ever object to the receipt in evidence of the testimony of the experts. The refusal of the majority to consider this testimony in connection with the question of duty, and to defer to the factual determinations made by the jury, which are supported by substantial evidence, is altogether unjustifiable.4
Equally unjustifiable is the majority’s assumption that a special relationship cannot be created without the collective presence of all of the factors which under Williams can create a “special relationship.” For example, allowing that “the officers’ conduct arguably increased the preexisting risk that Patrick would commit suicide, it did not change the preexisting risk that Patrick would do so.” (Maj. opn., ante, at p. 284.) In other words, according to the majority, the volunteered assistance of the police in this case must not only increase the preexisting risk but also change the nature of the preexisting risk and there must be detrimental reliance on the police conduct by the plaintiff. As indicated, I believe all these factors are present in this case, though the presence of only one would be sufficient to create a “special relationship.”
Citing section 323 of the Restatement Second of Torts, the Supreme Court explained in Williams that one who voluntarily comes to the aid of another “is under a duty to exercise due care in performance and is liable if (a) his failure to exercise such care increases the risk of such harm, or (b) the harm is suffered because of the other’s reliance upon the undertaking.” (Williams v. State of California, supra, 34 Cal.3d at p. 23, italics added.) In other words, it is sufficient if the actor either increases the risk (as by exacerbating a danger that already existed or creating a new danger) or the harm results from the plaintiff’s detrimental reliance on the assistance (as by foreclosing other forms of assistance). The fact that the Williams court found that none *300of the factors that can create a “special relationship” were present in that case does not suggest, as my colleagues believe, that all are necessary. As one authority has pointed out, the courts in Williams and Mann justify the imposition of a duty under the “special relationship” doctrine where “an individual officer had commenced a protective undertaking, and by his or her conduct either increased the risk to which the citizen was exposed during that episode, or induced the citizen to forego taking protective measures during the episode because the officer was apparently providing such limited protection.”5 (5 Levy et al., California Torts, supra, § 60.41 at p. 60-34.4, italics added.)
The majority is wrong in suggesting that the Supreme Court altered its attitude about the special relationship doctrine when in Williams it expressly disapproved Clemente v. State of California (1980) 101 Cal.App.3d 374 [161 Cal.Rptr. 799], (Maj. opn., ante, at p. 287.) Reaffirming the special relationship doctrine, the Williams court disapproved Clemente simply because unlike Mann (and the present case), where the police had actually “undertaken to protect the [injured party] from future physical harm,” the police in Clemente simply failed to investigate the cause and source of harm that had already occurred. (Williams v. State of California, supra, 34 Cal.3d at p. 26.)
In sum, even if (contrary to Allen v. Toten, supra, 172 Cal.App.3d 1079, 1090), appellants were Good Samaritans without any responsibility to enmesh themselves in the situation in the first place (as Williams compels me to conclude),6 their conduct in this case exposed them to liability because they voluntarily assumed responsibility to assist respondents and the decedent and their conduct substantially increased the preexisting risk. The fact that appellants’ conduct also changed the nature of the risk that already existed and compelled respondents to rely on their expertise is simply additional reason to find that a “special relationship” had been created.
*301The majority also says this case is different from Mann and other cases finding that the conduct of the police created a special relationship imposing a duty of care (i.e., McCorkle v. City of Los Angeles (1969) 70 Cal.2d 252 [74 Cal.Rptr. 389, 449 P.2d 453]; Wallace v. City of Los Angeles (1993) 12 Cal.App.4th 1385 [16 Cal.Rptr.2d 113]; Johnson v. County of Los Angeles (1983) 143 Cal.App.3d 298 [191 Cal.Rptr. 704]) because, unlike the situations in those cases, “. . . the responding officers made no express or implied promises that they would prevent Patrick’s suicide or that they would approach Patrick in a nonconfrontational manner.” (Maj. opn., ante, at p. 281.) The short answer to this objection is that, as emphasized by the Supreme Court in Williams, a “special relationship” can be created by conduct even without a promise and reliance thereon. “Such a relationship has also been found,” the court stated, “when the conduct of a police officer, in a situation of dependency, results in detrimental reliance on him for protection.” (Williams v. State of California, supra, 34 Cal.3d at p. 25, italics in original.)
The “situation of dependency” and resultant reliance on the police officers at the scene, by respondents as well as by the decedent, is much more clearly established by the evidence in this case than in the cases the majority relies upon. The control asserted by the police here also distinguishes this case from Natty v. Grace Community Church, supra, 47 Cal.3d 278, which addressed the duty of nontherapist counselors and a religious organization, not police officers. The Natty court observed that “. . . we have imposed a duty to prevent a foreseeable suicide only when a special relationship existed between the suicidal individual and the defendant or its agents.” (Id., at p. 293.) The court noted, as examples, cases in which such a duty was imposed on physicians or hospitals “after plaintiffs proved that the deceased committed suicide in a hospital or other in-patient facility that had accepted the responsibility to care for and attend to the needs of the suicidal patient.” (Ibid., citing Meier v. Ross General Hospital (1968) 69 Cal.2d 420 [71 Cal.Rptr. 903, 445 P.2d 519] and Vistica v. Presbyterian Hospital (1967) 67 Cal.2d 465 [62 Cal.Rptr. 577, 432 P.2d 193].) The Natty court refused to extend that duty of care “to personal or religious counseling relationships in which one person provided nonprofessional guidance to another seeking advice and the counselor had no control over the environment of the individual being counseled.” (Natty, supra, at p. 294, italics added.) The Natty court distinguished Meier and Vistica because unlike those cases, the plaintiff in Natty was not sufficiently under the control of the defendant, and the defendant did not fully accept responsibility: “Nally was not involved in a supervised medical relationship with defendants, and he committed suicide well over two weeks after he was released from the hospital against the advice of his attending psychiatrist and physician.” (Natty, supra, at p. 294.) *302The total control over Patrick the police exercised at all material times, which exceeded even the level of control found sufficient by the Supreme Court to justify the imposition of duty in Meier and Vistica, clearly distinguishes this case from Natty, even apart from the fact that we are dealing in this case with the duty of police officers, not personal or religious counselors.
This case is different from Natty (and the other cases the majority relies upon) in yet another important way. Respondents do not predicate liability upon appellants’ failure to intervene to save Patrick, but rather upon their affirmative acts that increased the preexisting risk, as respondents’ experts testified.
The majority correctly points out that “[a] long line of cases has held that a special relationship with a person in peril is not established simply because police officers responded to a call for assistance and took some action at the scene.” (Maj. opn., ante, at p. 279, citing Williams v. State of California, supra, 34 Cal.3d 18; M.B. v. City of San Diego (1991) 233 Cal.App.3d 699 [284 Cal.Rptr. 555]; Lopez v. City of San Diego (1987) 190 Cal.App.3d 678 [235 Cal.Rptr. 583]; Von Batsch v. American Dist. Telegraph Co. (1985) 175 Cal.App.3d 1111 [222 Cal.Rptr. 239]; Shelton v. City of Westminster, supra, 138 Cal.App.3d 610.) The police in the present case did not, however, merely respond to a call for assistance and take some inconsequential action which could not reasonably have induced reliance or otherwise created a special relationship. As described, the police not only preemptively asserted complete control, but initiated extreme measures involving the use of automatic weapons, guard dogs and searchlights, all of which were employed in a particularly aggressive manner and in violation of protocols of the Fremont Police Department. This conduct significantly increased the risk of harm, not just to Patrick, but to others, including the police themselves.
Throughout its opinion the majority describes the increase in the risk created by the police as merely “incremental.” Thus, for example, it says that “no authority exists imposing a duty [under the special relationship doctrine] where police conduct only incrementally increased the risk to which the injured person was already exposed.” (Maj. opn., ante, at p. 284.) The word “incremental” is notably inexact. If my colleagues used that term to refer to increases in risk that are insignificant I would agree. But that is clearly not their intent, as the increase in the risk of harm created by the conduct of the police in this case, to which they refer, is anything but insignificant. To be sure, the case law does not with precision mark the degree to which a person who volunteers assistance must increase preexisting risk in order to be liable under the special relationship doctrine. But if the degree to which appellants’ conduct increased the risk in this case is insufficient it is hard to imagine any *303increase in the risk of harm caused by noncriminal conduct that would suffice. As I have said, the majority’s assertion that there is “no evidence” that appellants acted with “reckless indifference to the consequences of their actions” (maj. opn., ante, at p. 271)—which is the foundation of the majority opinion—unjustifiably rejects findings of the trier of fact amply supported by the evidence. Indeed, the majority not only ignores certain critical findings of the jury but contradicts them. Thus, for example, ignoring the specific finding of the jury that the conduct of the police constituted “ ‘an assault response rather than [an] assist’” (maj. opn., ante, at p. 260), the majority contends that the “assaultive” party was not the police but Patrick (maj. opn., ante, at p. 270), so that he should bear 100 percent of the fault, not just the 25 percent determined by the jury.
My colleagues are simply unwilling to accept the determination of the jury that the danger to the police was more the result of their own conduct than that of Patrick. For example, the majority’s assertion that “. . . the majority of the disputed conduct in this case was the product of Sergeant Osawa’s deliberate tactical decisions designed to maximize the safety of the responding officers” (maj. opn., ante, at p. 276) conflicts with the extensive expert testimony—never objected to by appellants and accepted by the trier of fact—that Osawa’s conduct did not protect the safety of anyone and unnecessarily created the very danger (to the police themselves as well as to Patrick and others) the majority unfairly uses to exonerate Osawa. As earlier noted, Patrick had been sitting in the backyard for over an hour prior to the arrival of the police, during which time the effects of the alcohol he had earlier consumed was diminishing. As respondents’ experts persuasively explained, the danger to Patrick and others during that time was far less than that created by the arrival and provocations of Sergeant Osawa’s SWAT team.
While I certainly agree with my colleagues that there is a “social value of protecting the lives of police officers involved in a standoff with an armed individual”, (maj. opn., ante, at p. 276), I do not understand how the social value of protecting the police is advanced by refusing to impose liability on law enforcement officers whose unreasonable conduct unnecessarily endangers themselves (as well as those they purport to assist). Furthermore, as requested by appellants’ counsel, the trial court specifically instructed the jury that in situations involving threatened suicides the highest interest of the police is the protection of “[t]he physical safety of the community, including themselves, other citizens and family members.”7 (Italics added.) The jury must therefore be deemed to have taken this factor into consideration when it nonetheless rendered its verdict against the police.
*304It is also important to remember that the jury did not find the police wholly responsible for Patrick’s death, as it assigned 25 percent of the responsibility to Patrick’s own conduct, holding appellants responsible only for the remaining 75 percent.
In short, it is simply untrue that, as the majority claims, “imposing liability for the negligent handling of a threatened suicide improperly elevates the interests in preserving the life of the person threatening suicide over the interests of public safety and the physical safety of police officers.” (Maj. opn., ante, at p. 272.) The finder of fact in this case was asked to and presumably did take into account the factors my colleagues think can be considered only by constricting the legal duty of the police.
B.
Although the distinction between misfeasance and nonfeasance is sometimes tenuous, that is not true in this case. The majority ignores the significance our law attaches to this distinction.
As Professor Francis S. Bohlen pointed out in his classic 1908 essay on the duty to aid others, misfeasance differs from nonfeasance not only with respect to the character of the conduct complained of but as well “in the nature of the detriment suffered in consequence thereof.” (Bohlen, The Moral Duty to Aid Others as a Basis of Tort Liability (1908) 56 U. Pa. L.Rev. 217, 220.) Bohlen explained that th¿ difference between the results of nonfeasance and misfeasance, while “fundamental,” is not obvious. “In the case of active misfeasance the victim is positively worse off as a result the wrongful act. In cases of passive inaction plaintiff is in reality no worse off at all. His situation is unchanged; he is merely deprived of a protection which, had it been afforded him, would have benefited him. In the one case the defendant, by interfering with plaintiff or his affairs, has brought a new harm upon him, and created a minus quantity, a positive loss. In the other, by failing to interfere in the plaintiff’s affairs, the defendant has left him just as he was before; no better off, it is true, but still in no worse position; he has failed to benefit him, but he has not caused him any new injury nor created any new injurious situation. There is here a loss only in the sense of an absence of a plus quantity. It is this latter difference which in fact lies at the root of the marked difference in liability at common law for the consequences of misfeasance and non-feasance.” (Id., at pp. 220-221, italics added.)
The majority’s indifference to the affirmative nature of appellants’ unreasonable conduct is evident in its attempted analogy to certain other cases *305involving suicide. Citing Natty and two out-of-state cases (Lee v. Corregedore (1996) 83 Hawaii 154 [925 P.2d 324] and Donaldson v. YMCA (Minn. 1995) 539 N.W.2d 789, 792), the majority says that “[i]n cases involving suicide, courts have been extremely reluctant to impose liability based on the special relationship exception.” (Maj. opn., ante, at p. 277.) The cases the majority relies upon for this statement all involved acts of omission or nonfeasance, however, not unreasonably aggressive and unduly provocative acts of the sort that occurred in this case.
Cognizant that the distinction our Supreme Court has drawn between misfeasance and nonfeasance conflicts with its analysis, the majority ends up dismissing the distinction as merely “semantic,” because the same challenged conduct can almost always be characterized as either nonfeasance or misfeasance. (Maj. opn., ante, at p. 288.) The majority says, for example, that the conduct of the police in this case could be characterized as nonfeasance rather than misfeasance by describing it as the mere “fail[ure] to employ a sensitive approach.” {Ibid.) This is, of course, sophistry.
Though the significance of the misfe'asance/nonfeasance distinction has been repeatedly acknowledged by our Supreme Court (see, e.g., Williams v. State of California, supra, 34 Cal.3d 18, 23; Tarasoff v. Regents of University of California, supra, 17 Cal.3d 425, 435, fn. 5), my colleagues are correct that the distinction has been subjected to criticism. (See, e.g., 3 Harper et al., The Law of Torts, supra, § 18.6, p. 712 et seq.) But the burden of the criticisms is not that the distinction unjustifiably supports the imposition of liability for affirmative acts—which is the reason the majority dismisses it—but that the distinction may sometimes unjustifiably insulate a defendant from liability for the failure to act.8 Moreover, none of the commentators critical of the distinction urge that it be replaced by an analysis which would assist appellants. Consider, for example, the law review article upon which the majority relies. (Rowe & Silver, The Jurisprudence of Action and Inaction in the Law of Tort: Solving the Puzzle of Nonfeasance and Misfeasance From the Fifteenth Through the Twentieth Centuries (1995) 33 Duq. L.Rev. 807.)
*306The authors of that article complain that the misfeasance/nonfeasance distinction has been used “in a mindless, mechanical manner to countenance the statement that nonfeasance, which it equates with inaction, raises no liability.” (The Jurisprudence of Action and Inaction in the Law of Tort: Solving the Puzzle of Nonfeasance and Misfeasance From the Fifteenth Through the Twentieth Centuries, supra, 33 Duq. L.Rev. at p. 808, fn. omitted.) Analyzing two famous Cardozo opinions involving notions of nonfeasance and misfeasance,9 they state that Cardozo would distinguish an actionable negligent omission from inactionable nonfeasance “by reference to this question: Did the defendant’s action go forward to such a stage that inaction would produce an affirmative injury as opposed to the denial of a benefit? If the answer is ‘no,’ the defendant is an innocent nonfeasor. If it is ‘yes,’ then he might be a negligent misfeasor, depending, of course, on the prudence or imprudence with which the defendant acted.” (Id., at p. 839.)
The article goes on to make clear its repudiation of the view adopted by my colleagues in this case. While the article maintains that the misfeasance/ nonfeasance distinction is overly simplistic and has created confusion, it also contends that the distinction reflects a legitimate concern that could be better expressed. According to the authors of the article, the many courts that have employed the misfeasance/nonfeasance distinction “seem always to have been reaching for this proposition: One is duty bound to behave prudently only with respect to such risks as are attributable to him. They have been asking, sub silento, this simple question: Absent the defendant’s existence as a person (or entity), would the plaintiff have nonetheless suffered the damage of which he complains? If the answer is yes, (although this question can only be asked unconsciously) then the risk through which the plaintiff was damaged cannot be attributable to the defendant and the defendant is a nonfeasor only. In such cases, the damage may well be caused by the defendant’s behavior—his failure to act—which proposition is easily established by reference to a second question: Absent the defendant’s failure to act, would the plaintiff have nonetheless suffered the damage of which he complains? The answer to that question might easily be ‘no’—which means that the defendant’s failure to act has caused the damage at issue—even as the answer to the question previously asked is yes.” (The Jurisprudence of Action and Inaction in the Law of Tort: Solving the Puzzle of Nonfeasance and Misfeasance From the Fifteenth Through the Twentieth Centuries, supra, 33 Duq. L.Rev. at p. 851.) In other words, “[ajction and omission may both be negligent, but one has a duty to refrain from negligence only as to those risks created by one’s existence on earth.” (Id., at p. 854.)
A duty arises in this case under the foregoing tests even if (as is not the case) the conduct in this case could be considered a “not doing” rather than *307a “misdoing.” Appellants affirmatively intervened in the situation to such an extent that their negligent “omissions” produced affirmative injury, not merely the denial of a benefit, and their conduct was manifestly imprudent. Moreover, respondents’ experts provided evidence that this conduct was “a substantial cause” of Patrick’s death. Stated differently, absent appellants’ existence, the decedent’s body would not have been riddled with 27 bullets, and his survival far more likely, if not certain.
The basic idea was explained less elaborately in Williams v. California, supra, 34 Cal.3d 18, where the Supreme Court stated that “[ajbsence of duty [rather than statutory immunity] is a particularly useful and conceptually more satisfying rationale where, absent any ‘special relationship’ between the officers and the plaintiff, the alleged tort consists merely in police nonfeasance. [Citations.]” (Id., at p. 23, italics added.) By the same token, absence of duty is commensurately inappropriate in cases such as this, where the special relationship results, inter alia, from police misfeasance.
C.
The majority takes me to task for urging an expansion of the special relationship doctrine. Claiming I rely on “dated commentary,” they say I am “predicting a legal trend that never actually materialized.” (Maj. opn., ante, at p. 287.) This is demonstrably untrue. The expansion the majority abhors occurred in this and most other American jurisdictions long ago. As noted by our Supreme Court more than 20 years ago, this expansion was the salutary judicial response to the moral problem created by the absence in the law of a duty to rescue. In Tarasoff v. Regents of University of California, supra, 17 Cal.3d 425, the high court observed that the general common law rule that a person owes no duty to control the conduct of another “derives from the common law’s distinction between misfeasance and nonfeasance, and its reluctance to impose liability for the latter. [Citation.] Morally questionable, the rule [of no liability for nonfeasance] owes its survival to ‘the difficulties of setting any standards of unselfish service to fellow men, and of making any workable rule to cover possible situations where fifty people might fail to rescue . . . .’ [Citation.] Because of these practical difficulties, the courts have increased the number of instances in which affirmative duties are imposed not by direct rejection of the common law rule, but by expanding the list of special relationships which will justify departure from that rule. [Citation.]” (Id., at p. 435, fn. 5, italics added; accord, 3 Harper et al., The Law of Torts, supra, § 18.6, pp. 712-732, and authorities there cited and discussed.) As pointed out in Mann (which, as earlier noted, was cited with approval by the Supreme Court in Williams), “The California Supreme Court, Prosser and the Restatement Second of Torts all recognize that ‘special relationship’ is *308an expanding concept in tort law. (See Tarasoff v. Regents of University of California, supra, 17 Cal.3d at p. 435, fn. 5; Prosser, Law of Torts (4th ed. 1971) § 56, pp. 339-340; Rest.2d Torts (1965) § 314A, coins, a, b.) As the Restatement suggests, the law appears to be heading toward a recognition of the duty to aid or protect in any relation of dependence or of mutual dependence. (Id., com. b; see also Fleming, Law of Torts (4th ed. 1971) p. 143.)” (Mann v. State of California, supra, 70 Cal.App.3d 773, 779-780; see also, 3 Harper et al., The Law of Torts, supra, § 18.6, p. 712 et seq., and numerous authorities there cited; Weinrib, The Case for a Duty to Rescue, supra, 90 Yale LJ. 247; Shapo, The Duty to Act: Tort Law, Power and Public Policy (1977); Minor, The Moral Obligation as a Basis of Liability, 9 Va. L.Rev. 420 (1923); Ames, Law and Morals (1908) 22 Harv. L.Rev. 97; Bohlen, The Moral Duty to Aid Others as a Basis of Tort Liability, supra, 56 U. Pa. L.Rev. 217.)
There are situations, I readily concede, in which application of the special relationship doctrine might inappropriately punish a Good Samaritan and have adverse social consequences; but this is not such a case. The evidence relied upon by the trier of fact shows that the police intervention here significantly increased the risk of harm, not just to respondents and the decedent, but as well as to the police themselves and any others who may have been on or near the scene. The imposition of liability in these circumstances is not at all inappropriate; on the contrary, it is entirely consistent with the many cases, some of which have been decided by the Supreme Court, exposing the police to tort liability for negligent and intentional acts committed in the course of law enforcement activities. (See, e.g., Munoz v. Olin (1979) 24 Cal.3d 629 [156 Cal.Rptr. 727, 596 P.2d 1143]; Grudt v. City of Los Angeles (1970) 2 Cal.3d 575 [86 Cal.Rptr. 465, 468 P.2d 825].)
D.
Relying on a 1981 opinion from the District of Columbia (Warren v. District of Columbia (D.C.App. 1981) 444 A.2d 1), the majority in part justifies its refusal to impose a duty in this case on the theory that the duty police officers undertake by virtue of their employment does not relate to any particular individuals but' to the public at large. (Maj. opn., ante, at pp. 274-275.) In the view of the majority, “[permitting potential suicide victims and their families' to hold police officers personally liable for the negligent handling of a suicide crisis conflicts with the public nature of protection services police officers provide to the community at large.” (Maj. opn., ante, at p. 275.) This justification for the refusal to find a duty does not apply where the police have entered into a “special relationship,” and this is true even in the District of Columbia. (See, e.g., Rieser v. District of Columbia *309(D.C. Cir. 1977) 563 F.2d 462, 477-479 [183 App.D.C. 375].) Furthermore, the supposed conflict between the broad interest in public safety and the interest in preserving the life of a particular person does not exist, because the two goals are entirely compatible, and were in fact reconciled in this case by the trier of fact. This is one of the reasons the use of the public nature of law enforcement responsibilities to bar the imposition of liability has been widely criticized. As stated in a leading treatise: “It is frequently said that liability turns on a distinction between the police officer’s (or agency’s) ‘general’ or ‘public’ duties to prevent crime, for the breach of which there is no liability, and the officer’s ‘special’ duty owed to an individual, or a ‘special relationship’ with the crime victim. The distinction is quite unsatisfactory in terms of normal negligence theory. Some police decisions may deserve immunity as being nontortious because they require choices none of which is objectively unreasonable in the circumstances. Some may be treated as unsuitable for judicial review because adjudication would involve a court in unseemly interference with executive or legislative decisionmaking, e.g., the allocation of a municipality’s budget. There is also an understandable reluctance to subject municipalities to wholesale liability in negligence to all crime victims on generalized charges of insufficiency of care in law enforcement. But run-of-the-mill negligence in the conduct of routine activities should never be insulated from liability by the doubt that an actor owes a ‘duty’ of care to identifiable persons who will foreseeably be subjected, by such negligence, to unreasonable risk of bodily injury.” (5 Harper et al., The Law of Torts, supra, § 29.6A, pp. 640-641, fhs. omitted, italics added.)
The imposition of liability in this case would create none of the problems just described. As the jury explicitly found, the police made a choice that was not only objectively unreasonable in the circumstances but in violation of their own rules.
E.
The majority also endeavors to undermine the special relationship doctrine by creating a false conflict between that doctrine and Rowland v. Christian, supra, 69 Cal.2d 108, and using this as the excuse to virtually do away with the special relationship doctrine as it applies to law enforcement officers. According to the majority, “if a duty of care were imposed in each case where there was some progressive, increased chance of injury stemming from a preexisting harm, the special relationship doctrine would be in irremediable conflict with the traditional duty analysis derived from Rowland . . . .” (Maj. opn., ante, at p. 285.) This is so, the majority reasons, because “[i]n volatile situations, one can always argue that the arrival of police officers caused an incremental increase in tension at the scene, and thus increased the risk of injury occurring.” (Maj. opn., ante, at p. 284.)
*310In order to solve this putative problem, the majority creates a new rule: “Where police conduct results in some increase in a preexisting risk of harm, but an analysis of the traditional Rowland factors weighs against the imposition of a duty, we conclude that no special relationship duty may be imposed.” (Maj. opn., ante, at p. 285.)
Summoning higher authority for this emasculation of the special relationship doctrine, the majority claims that “[t]o expansively construe the special relationship doctrine to encompass such incremental increases in a preexisting risk would eviscerate our Supreme Court’s adoption in Williams, supra, 34 Cal.3d [18, 23,] of the public duty rule, that protects police officers from the burden of assuming greater obligations to others by virtue of their employment.” (Maj. opn., ante, at p. 285.) The majority mischaracterizes Williams, which, as earlier explained, articulates a much more balanced view than the majority implies. The court made it clear in that case that “when the state, through its agents, voluntarily assumes a protective duty toward a certain member of the public and undertakes action on behalf of that member, thereby inducing reliance, it is held to the same standard of care as a private person or organization.” (Williams v. State of California, supra, 34 Cal.3d at p. 24, italics added, citing Hartzler v. City of San Jose (1975) 46 Cal.App.3d 6, 10 [120 Cal.Rptr. 5]; and Mann v. State of California, supra, 70 Cal.App.3d at p. 780.) Elsewhere in Williams the court reiterated that while a law enforcement officer does not assume any greater obligation to others individually, “ ‘[a] person does not, by becoming a police officer, insulate himself from any of the basic duties which everyone owes to other people ....’” (34 Cal.3d at p. 24, fn. 3, italics added, quoting Warren v. District of Columbia, supra, 444 A.2d 1, 4-9.)
What the majority is saying is that the mere appearance of the police on the scene necessarily involves some (“incremental”) increase in the risk of harm (“or increased chance of injury”), and that it would be irrational to hold the police liable merely for appearing on the scene. The problem the majority sets out to solve does not exist. And the proof it does not exist is provided by the very cases the majority relies upon. (Williams v. State of California, supra, 34 Cal.3d 18; M.B. v. City of San Diego, supra, 233 Cal.App.3d 699; Lopez v. City of San Diego, supra, 190 Cal.App.3d 678; Von Batsch v. American Dist. Telegraph Co., supra, 175 Cal.App.3d 1111; Rose v. County of Plumas (1984) 152 Cal.App.3d 999 [199 Cal.Rptr. 842]; Shelton v. City of Westminster, supra, 138 Cal.App.3d 610; Clemente v. State of California, supra, 101 Cal.App.3d 374.) In all of those cases, as in many others, the police were relieved of any legal duty precisely because their involvement—which almost always consisted of more than mere appearance on the scene—nevertheless did not rise to the level of “affirmative action *311which contributed to, increased, or changed the risk which have otherwise existed.” (Williams, supra, at p. 27; Rose, supra, at p. 1005; accord, Von Batsch, supra, at p. 1124; see also Lopez, supra, 190 Cal.App.3d at pp. 682-683 [finding “inaction” is not misfeasance] and Shelton v. City of Westminster, supra, at p. 622 [no duty arises from police undertaking to investigate and take appropriate action to find missing person].)
As I have sought to emphasize, unlike Williams, Shelton, M.B., Lopez, Von Batsch, Rose, Clemente and virtually all the other cases the majority relies upon, the conduct complained of in the present case constitutes affirmative action which substantially (not “incrementally”) increased the danger that already existed and also created a new danger, which in fact materialized.
Rowland v. Christian appeals to my colleagues because, by inviting consideration of the “consequences to the community of imposing a duty to exercise care with resulting liability for breach” (69 Cal.2d at p. 113), it permits them to decide this case on the basis of value judgments—specifically, their view that, because the police did not “plan[] to precipitate Patrick’s suicide” nor act “with bad faith or a reckless indifference to the consequences of their actions,” there was “no moral blame” attendant to their conduct (maj. opn., ante, at pp. 270-271)—and to effectuate their belief that the imposition of liability would have a chilling effect on police activities generally. The chief reason I believe Rowland is irrelevant to the question of duty in this case, as I have said, is that the police, like everyone else, have no duty to rescue. The issue in this case is whether there is any applicable exception to a general no-duty rule, which is unrelated to the question presented in cases to which Rowland properly applies.10 But there is another important reason why Rowland does not and should not apply to determine whether a general duty applies in cases such as this.
*312What my colleagues dislike about the special relationship doctrine is that, by looking at conduct, it applies to a police officer the same as it applies to everyone else. Thus the doctrine conforms to the decree of our Legislature that “a public employee is liable for injury caused by his act or omission to the same extent as a private person,” unless the Legislature has “otherwise provided by statute.” (Gov. Code, § 820, subd. (a).) The policy factors that may be considered under Rowland, which are judicially developed, not statutory, cannot be applied so as to relieve only public employees (or a certain type of public employees, such as law enforcement officers) of liability to which similarly situated private persons would still be exposed. As will be discussed later, the Legislature has by statute limited the liability for injury of public entities generally (Gov. Code, §§ 820-822.2), and those engaged in police and correctional activities in particular (Gov. Code, §§ 844-846), through the grant of immunities. Those immunity statutes represent legislative resolution of the often competing policy considerations relating to whether all or certain public entities ought to be held responsible under our tort law to the same extent as others. Instead of deferring to the will of the Legislature with respect to this question my colleagues improperly employ Rowland to embark upon an independent policy review as if the Legislature had never spoken, thereby usurping its prerogatives.
If, as my colleagues say, the police must be relieved of a duty to use due care because such a legal responsibility would discourage them from intervening in life-threatening situations of the sort presented in this case, the tort liability of the police would be fundamentally different from that of private persons without regard to whether any governmental immunity applies. This *313is a revolutionary proposition. As earlier noted, our Supreme Court has agreed that a person does not, by becoming a police officer, assume any greater obligation than others, but neither, it has declared, does he “ ‘insulate himself from any of the basic duties which everyone owes to other people.’ ” (Williams v. State of California, supra, 34 Cal.3d at p. 24, fn. 3, italics added, quoting Warren v. District of Columbia, supra, 444 A.2d 1, 8.) The general rule in California, at least until now, is that “... a law enforcement officer is hable to the same extent as a private individual for lack of due care or [for] an intentional act which results in injury, unless the officer is protected by a statutory immunity.” (5 Levy et al., California Torts, supra, § 61.10, p. 61-62.6, italics added, citing, inter alia, Gov. Code, § 820 [“Except as otherwise provided by statute ... a public employee is liable for injury caused by his act or omission to the same extent as a private person.”].) The majority has so constricted the legal concept of duty applicable to law enforcement officers as to render statutory immunities almost irrelevant.
The trouble with the majority’s new rule is not just that it impermissibly invades the domain of the Legislature but that the underlying policy justification has been explicitly rejected by the California Supreme Court. Considering the question in the context of immunity, the issue to which I now turn, our high court has determined that subjecting law enforcement officers to liability for injury caused by their negligent acts will not, as my colleagues claim, inhibit them from providing assistance or unduly burden local government.
n.
Because it shares my colleagues’ concern about the adverse consequences of subjecting law enforcement officers to unlimited tort liability, the Legislature immunized certain specific police and correctional activities from liability. Thus, for example, neither a public entity nor a public employee is liable “for failure to provide sufficient police protection service” (Gov. Code, § 845) or “for injury caused by the failure to make an arrest or by the failure to retain an arrested person in custody.” (Gov. Code, § 846.) The police are additionally protected by statutory immunities generally applicable to public entities and their employees, including immunity for discretionary acts (Gov. Code, §§ 820.2, 815.2, subd. (b)) and for failure to enforce the law (Gov. Code, § 818.2, 821), as well immunities applicable to particular functions, such as confining or transporting certain persons. (Gov. Code, §§ 850.8, 856.)
Appellants claim the acts described by the jury in the special interrogatory as “negligent” were immunized from liability under Government Code *314section 820.2, which states as follows: “Except as otherwise provided by statute, a public employee is not liable for an injury resulting from his act or omission where the act or omission was the result of the exercise of the discretion vested in him, whether or not such discretion be abused.” The most pertinent and authoritative definition of the discretionary acts immunized under this statute was provided in the landmark opinion in Johnson v. State of California, supra, 69 Cal.2d 782.
In Johnson, the Supreme Court reversed a summary judgment for the state in an action for personal injuries sustained by a foster mother who had been attacked by a youth placed in her home for foster care by the California Youth Authority. She claimed Youth Authority employees knew of the youth’s homicidal tendencies but failed to provide her warning. Holding that the decision not to give warning was not an immune discretionary act, the court rejected a purely “mechanical” or “literal” approach to defining “discretionary actions,” and relied primarily “on policy considerations relevant to the governmental entity’s claim of immunity.” (69 Cal.2d at p. 789.) The chief policy consideration addressed in Johnson is precisely the one elevated by appellants in this case and by the majority; namely, that subjecting law enforcement officials to tort liability for negligent acts will inhibit them from providing any assistance at all. The Supreme Court resoundingly rejected this argument, stating that “[t]he danger that public employees will be insufficiently zealous in their official duties does not serve as a basis for immunity in California.” (Id., at p. 790, italics in original.)
The Supreme Court commenced its analysis by adverting to Judge Learned Hand’s classic articulation of the justification for attaching immunity to “discretionary” actions of public officials in order to protect them from the spectre of extensive tort liability: “ ‘It does indeed go without saying that an official, who is in fact guilty of using his powers to vent his spleen upon others, or for any other personal motive not connected with the public good, should not escape liability for the injuries he may so cause; and, if it were possible in practice to confine such complaints to the guilty, it would be monstrous to deny recovery. The justification for doing so is that it is impossible to know whether the claim is well founded until the case has been tried, and that to submit all officials, the innocent as well as the guilty, to the burden of a trial and to the inevitable danger of its outcome, would dampen the ardor of all but the most resolute, or the most irresponsible, in the unflinching discharge of their duties.’ ” (Id., at p. 790, quoting Gregoire v. Biddle (2d Cir. 1949) 177 F.2d 579, 581.) Our Supreme Court was unpersuaded by this concern for several reasons, explaining at considerable length why “California’s statutory provisions for indemnification of public officials largely remove the dangers that troubled Judge Hand and that any *315concern for the preservation of ardor in the performance of public duties need not constitute a substantial consideration in our definition of ‘discretionary’ action.” (69 Cal.2d at pp. 790-791.)
Nor did our Supreme Court deem an employee’s concern over the potential liability of his or her employer a sufficient reason to expansively define “discretionary,” and hence immune, acts. The court felt it “unlikely that the possibility of government liability will be a serious deterrent to the fearless exercise of judgment by the employee.” (69 Cal.2d at p. 792, citing Note, The Discretionary Function Exception of the Federal Tort Claims Act (1953) 66 Harv. L.Rev. 488, 495-496.) This statement of our Supreme Court cannot be reconciled with the majority’s belief “that the risk of liability will affect police conduct regardless of whether an adverse judgment is covered by insurance” (maj. opn., ante, at p. 274) and that “[ojnly the most irresponsible police officers would shrug off the possibility of a judgment holding them personally liable for another’s suicide solely because monetary damages would not be coming out of their own pocket.” {Ibid.)
Nor can the majority opinion be squared with the statement in Johnson that, “to the extent that [public employees are deterred by the imposition of liability], it may be wholesome. An employee in a private enterprise naturally gives some consideration to the potential liability of his employer, and this attention unquestionably promotes careful work; the potential liability of a government entity, to the extent that it affects primary conduct at all, will similarly influence public employees.”11 (69 Cal.2d at pp. 792-793, fn. omitted, citing James, Tort Liability of Governmental Units and Their Officers (1955) 22 U. Chi. L.Rev. 610, 652; Peck, The Federal Tort Claims Act: A Proposed Construction of the Discretionary Function Exception (1956) 31 Wash. L.Rev. 207, 224.)
The foregoing considerations were not the only reasons the Supreme Court imposed liability in Johnson. As the court explained, Government Code section 820.2 was designed to assure judicial abstention only “in areas in which the responsibility for basic policy decisions has been committed to coordinate branches of government. Any wider judicial review, we believe, would place the court in the unseemly position of determining the propriety of decisions expressly entrusted to a coordinate branch of government. Moreover, the potentiality of such review might even in the first instance *316affect the coordinate body’s decision-making process. [Citations.]” (69 Cal.2d at p. 793, italics in original; accord, Tarasoff v. Regents of University of California, supra, 17 Cal.3d at p. 445 [“section 820.2 affords immunity only for 'basic policy decisions.’ (Italics added.) [Citations.].”) With this guidepost in mind, the court found “no plausible reason for governmental immunity” in the situation presented in Johnson. (69 Cal.2d at p. 795.) The court found that the decision of Youth Authority employees to parole a youth to a given foster family was within the “discretionary function” language of section 820.2. “The decision to parole thus comprises the resolution of policy considerations, entrusted by statute to a coordinate branch of government, that compels immunity from judicial reconsideration.” (69 Cal.2d at p. 795, fn. omitted.) “Once an official reaches the decision to parole to a given family, however, the determination as to whether to warn the foster parents of latent dangers facing them presents no such reasons for immunity; to the extent that a parole officer consciously considers pros and cons in deciding what information, if any, should be given, he makes such a determination at the lowest, ministerial rung of official action. Judicial abstinence from ruling upon whether negligence contributed to this decision would therefore be unjustified; coupled with the administrative laxness that caused the loss in the first instance, it would only result in the failure of governmental institutions to serve the injured individual.” (69 Cal.2d at pp. 795-796.)
A police decision to assist in avoiding a potential suicide, like the Youth Authority decision to parole, “comprises the resolution of policy considerations, entrusted by statute to a coordinate branch of government, that compels immunity from judicial reexamination.” (69 Cal.2d at p. 795, fn. omitted.) But the decision to provide such public assistance is not at issue in this case. The decisions of the police at issue here, like the Youth Authority decision at issue in Johnson, were made after the police decided to assist and relate instead to the nature of the assistance provided. We are therefore not here concerned with the basic policy decision to assist in life-threatening situations involving a potential suicide, but the implementation of that policy. Moreover, the evidence shows that the decisions made by the police on the scene after they intervened violated the declared policy of the Fremont Police Department regarding the treatment of mentally distressed persons who threaten their own lives or those of others. The conduct which violated that policy is therefore not within the immunity for discretionary acts granted under Government Code section 820.2. As the Supreme Court reiterated in Johnson, " '\0\nce the determination has been made that a service will be furnished and the service is undertaken, then public policy demands (except when the Legislature specifically decrees otherwise) that government be held to the same standard of care the law requires of its private *317citizens in the performance of duties imposed by law or assumed.’ ” (69 Cal.2d at p. 796, italics added, quoting Sava v. Fuller (1967) 249 Cal.App.2d 281, 290 [57 Cal.Rptr. 312]; accord, McCorkle v. City of Los Angeles, supra, 70 Cal.2d 252 at p. 261.)
The inapplicability to this case of any statutory immunity is underscored by the existence of a special relationship between respondents and the police who responded to their call for assistance. As a leading treatise states, “when police officers are negligent in the performance of a duty which they have undertaken, and when there exists a special relationship between the public entity and plaintiff in which the public entity has voluntarily assumed a duty of police protection toward plaintiff, liability may be imposed irrespective of the immunity granted by Government Code section 845 [providing immunity for failure to provide sufficient police protective services]. Generally, immunity for failure to provide police protection does not apply to situations in which liability of the public entity is based not on its failure to provide police services, but rather on its breach of an affirmative duty to protect plaintiff.” (5 Levy et al., California Torts, supra, § 61.13 at pp. 61-70 to 61-71, fns. omitted, italics added, citing, inter alia, Wallace v. City of Los Angeles, supra, 12 Cal.App.4th 1385, 1402-1403; Carpenter v. City of Los Angeles (1991) 230 Cal.App.3d 923, 934-935 [281 Cal.Rptr. 500]; Hartzler v. City of San Jose, supra, 46 Cal.App.3d 6, 9-10; Hernandez v. Southern California Rapid Transit Dist. (1983) 142 Cal.App.3d 1063, 1067 [191 Cal.Rptr. 436]; see Lopez v. Southern Cal. Rapid Transit Dist. (1985) 40 Cal.3d 780, 792-793 [221 Cal.Rptr. 840, 710 P.2d 907].)
For the foregoing reasons, the police conduct challenged in this case is not within the immunity afforded discretionary acts under Government Code section 820.2.
III.
Appellants finally maintain that, as a matter of law, the evidence does not establish the requirements for recovery of damages based on negligent infliction of emotional distress because respondents did not directly observe the shooting of Patrick. The several theories they advance all rest on Thing v. La Chusa (1989) 48 Cal.3d 644 [257 Cal.Rptr. 865, 771 P.2d 814], in which the Supreme Court revisited its landmark decision in Dillon v. Legg (1968) 68 Cal.2d 728 [69 Cal.Rptr. 72, 441 P.2d 912, 29 A.L.R.3d 1316], concerning bystander recovery for damages for emotional distress. Appellants emphasize the language in Thing limiting recovery to situations in which, among other things, the plaintiff “is present at the scene of the injury-producing event at the time it occurs and is then aware that it is causing *318injury to the victim.” (Thing v. La Chusa, supra, 48 Cal.3d at p. 668.) Recovery for emotional distress is barred in this case, appellants maintain, because, as in Thing, respondents did not “contemporaneously witness” the conduct found to be negligent. According to appellants, “mere auditory perception” is insufficient. This is not an impressive argument.
First, no case called to our attention by appellants declares that the contemporaneous awareness requirement of Thing can only be satisfied when a plaintiff has visually witnessed the infliction of injury. The plaintiff’s problem in Thing was not that she did not “observe” the event, but that, in addition, she “was not aware that her son was being injured.” (48 Cal.3d at p. 669.)
“Awareness” can occur in a variety of sensory ways, not just visually. The case law provides many illustrations. Wilks v. Hom (1992) 2 Cal.App.4th 1264 [3 Cal.Rptr.2d 803], for example, was an action by multiple plaintiffs, including the mother of an injured minor, for wrongful death and personal injuries against landlords of a residence where an explosion occurred. The jury awarded damages to the mother for the emotional distress occasioned by the negligently caused injuries to her daughter. At the time of the explosion and injury, the mother and daughter were in different parts of the residence. The Court of Appeal determined that such damages were appropriate “because the mother was contemporaneously aware that the explosion was causing the injuries although she did not actually see or hear her daughter being injured.” (Id., at p. 1267, italics added.) The Wilks court justified this conclusion in part because the court in Thing pointed with approval to its earlier opinion in Krouse v. Graham (1977) 19 Cal.3d 59 [137 Cal.Rptr. 863, 562 P.2d 1022], holding that “the plaintiff need not visually perceive the third party injury in order to satisfy the Dillon guideline, suggesting only that he must suffer shock from ‘ “ ‘the sensory and contemporaneous observance of the accident . . . .”” ” (2 Cal.App.4th at p. 1269, quoting Thing v. La Chusa, supra, 48 Cal.3d at p. 656, quoting Krouse v. Graham, supra, 19 Cal.3d at p. 76.)
“In Krouse, the plaintiff sat in the driver’s seat of his car and knew that his wife was at the curb closing the door to the backseat when a car negligently driven by the defendant approached the rear of the plaintiff’s car, straddled the curb and hit and killed the plaintiff’s wife. The Krouse court ruled it was sufficient that the plaintiff knew his wife’s position an instant before she was struck, saw the defendant’s car coming toward her at high speed, and knew it must have hit his wife.” (2 Cal.App.4th at pp. 1269-1270.) The Wilks court concluded, and I agree, that the discussion of Krouse in Thing affirms “that bystander damages may be recovered only by a plaintiff who is present at the *319injury-producing event at the time it occurs and is then aware that it is causing injury to the victim. The court’s analysis did not indicate disapproval, however, of the holding in Krouse that the plaintiff need not visually perceive the injury while it is being inflicted.” {Id., at p. 1271, italics added.) All that Thing sought to make clear was that a plaintiff who arrived at the scene after the accident and neither saw nor heard the event that produced the injury, and was therefore not contemporaneously aware of it, cannot recover damages for emotional distress. That is not what happened in this case.12
Respondents were on the scene from the beginning and aware of all material events as they unfolded. They witnessed numerous officers search for Patrick in the house and enter the backyard accompanied by a trained dog and with shotguns and automatic weapons drawn. They were advised that the police had located Patrick under the bush in the rear of the house, and heard repeatedly shouted orders directing him to come out. Respondents were acutely aware of the risks created by the aggressive acts of the police. Respondent Gina Gohlston was so frightened by developments that, shortly before the police discharged their weapons, she sat down on a curb and vomited. Respondent Adams, Patrick’s wife, then ran toward the house but was stopped by officers. Moments later respondents heard a final shout from the rear of the house and then a barrage of 34 gunshots. A few moments later respondents saw Patrick carried to an ambulance on a stretcher. Clearly, respondents were “present at the scene” of the injury-producing event, within the meaning of Thing v. La Chusa, supra, 48 Cal.3d 644, and its progeny.
Appellants also maintain there can be no recovery for emotional distress because the special interrogatory did not specifically identify the discharge of weapons as negligent; therefore, appellants argue, any distress respondents may have suffered from hearing the fusillade was not negligently inflicted and cannot support the award of damages. This contention ignores many references in the special interrogatory to police conduct which inferentially included the use and discharge of weapons, such as the lack of *320control, violation of police procedures, the use of armed officers, which the jury felt left no option but force, and the “assault” mode of the police response to the call for assistance. Furthermore, appellants overlook the considerable expert testimony that the police were negligent in using and discharging weapons, which this court cannot ignore.
IV.
Relying on Elden v. Sheldon (1988) 46 Cal.3d 267 [250 Cal.Rptr. 254, 758 P.2d 582], which holds that an action for negligent infliction of emotional distress cannot be maintained by an unmarried cohabitant of the injured party, appellants argue that respondent Gohlston cannot recover damages for emotional distress because, as Patrick’s stepdaughter, she was too distantly related. Appellants initially raised this issue in a pretrial motion to dismiss respondent Gohlston’s action. In response, this respondent filed a lengthy declaration, which has never been disputed, describing the extremely close and loving relationship she had with her stepfather. The motion was denied after the trial court specifically found that respondent Gohlston was not too distantly related to Patrick to be able to maintain an action for negligent infliction of emotional distress.13 There is no basis upon which this court could set aside that factual finding, as the evidence of a close relationship is undisputed and no court has ever declared as' a matter of law that for purposes of determining whether a plaintiff may maintain an action for negligent infliction of emotional distress a stepchild must be treated differently than any other child.
V.
Appellants’ final contention regarding the claim for negligent infliction of emotional distress is that it should not be allowed at all because respondents failed to satisfy the claim-filing requirement of Government Code section 945.4. The argument fails. First, as respondents point out, their claims specifically sought damages for emotional distress as well as punitive damages, which cannot be recovered in an action that is merely for wrongful death. (Krouse v. Graham, supra, 19 Cal.3d 59, 72 [emotional distress]; Parker v. Superior Court (1985) 175 Cal.App.3d 1082, 1087 [223 Cal.Rptr. 292] [punitive damages].) Appellants were therefore well aware respondents *321intended to sue for more than just wrongful death, or should have been. Respondents’ claims provided sufficient information enabling appellants to adequately investigate the claim and settle the matter, if possible, without the expense of litigation, which is the purpose of the claims-filing requirement. (Phillips v. Desert Hospital Dist. (1989) 49 Cal.3d 699, 705 [263 Cal.Rptr. 119, 780 P.2d 349].) ¿deed, the trial court made such a finding. Furthermore, appellants conceded below that the reason they did not settle this case was not because they were inadequately advised of the nature of respondents’ claims, but because they did not believe they breached any legal duty. Finally, if appellants believed respondents’ claims were unclear in any particular—and it is difficult to believe there ever was any such uncertainty—they were statutorily obliged to file a notice of insufficiency, “stating with particularity the defects or omissions” of the claim presented. (Gov. Code, § 910,8.) By failing to provide such notice, appellants waived the defense that the claim was defective. (Gov. Code, § 911; Phillips v. Desert Hospital Dist., supra, 49 Cal.3d at p. 711.)
For the foregoing reasons, I would affirm the judgment.
A petition for a rehearing was denied January 4, 1999, and the opinion was modified to read as printed above. Kline, P. J., was of the opinion that the petition should be granted. The petition of plaintiffs and respondents for review by the Supreme Court was denied March 24, 1999. Mosk, J., and Kennard, J., were of the opinion that the petition should be granted.
I acknowledge that a few courts have applied the Rowland v. Christian analysis to determine whether the police have a responsibility to provide assistance. (See, e.g., Dutton v. City of Pacifica (1995) 35 Cal.App.4th 1171 [41 Cal.Rptr.2d 816]; Allen v. Toten (1985) 172 Cal.App.3d 1079 [218 Cal.Rptr. 725] and Shelton v. City of Westminster (1982) 138 Cal.App.3d 610 [188 Cal.Rptr. 205].) Those cases were, in my view, erroneous in that respect, although, as later explained, appellants would not be relieved of liability even if Rowland did apply. (See discussion, post, at fn. 10.) The discussion of Rowland in Natty v. Grace Community Church, supra, 47 Cal.3d at pages 296-299, was dicta, as the court previously found no “special relationship” in that case, which it indicated was the dispositive consideration. (Id., at p. 293.)
In a footnote, the majority takes issue with respondents’ statement at oral argument (in response to a question from the court) that the detrimental reliance requirement can be satisfied by decedent’s wife and stepdaughter’s reliance on the responding officers for assistance, and the officers’ refusal to permit familial participation in the situation. The majority claims this “contention” cannot now be made because it “was not properly raised in respondents’ appellate brief.” (Maj. opn., ante, at p. 281, fn. 29.) This statement is wholly unjustified. The rules of pleading certainly do not require such specificity; moreover, appellants never challenged the pleadings on this or any other ground, and do not now raise the matter on appeal, which is, of course, why respondents did not bother to address the issue in their reply brief.
Also unjustified is the majority’s assertion that there is “no evidence that [Patrick’s wife and stepdaughter] detrimentally relied on the conduct of the police officers by foregoing other means of assisting Patrick.” (Maj. opn., ante, at p. 281, fn. 29.) The evidence showed that Sergeant Osawa failed to inquire of Patrick’s wife and stepdaughter about the cause of his conduct, whether he had been using drugs or drinking and, if so, how he responded to such substances, whether he had a criminal history, or a history of past violent acts, or an aggressive or passive personality, or disliked the police, or had an aversion to dogs, or anything else that might shed light on his behavior and attitudes. Respondents’ experts testified that the failure to obtain this information—which respondents sought to provide— significantly diminished the ability of the police to respond intelligently to the situation, and thereby contributed to Patrick’s death. Furthermore, the wife and stepdaughter testified that the police placed them under strict constraints, establishing their inability to personally intercede, seek assistance from decedent’s friend Alan Kirshner (as one expert thought appropriate) or others, or see to it that Patrick remained undisturbed until he sobered up. The special interrogatories received from the jury indicate that the verdict was based in part on this evidence; we are therefore not free to ignore it, or to speculate as to how jurors could have interpreted the pertinent evidence.
The author of this treatise acknowledges the argument “that it is impossible in the nature of things for the duty problem to be decided by the jury, for if the court sends the issue to the jury this ‘necessarily operates as a ruling that there is a duty or else he would never have submitted the case to the jury at all.’ ” The author rejects this argument because, “[a]s in the case of any other issue, the judge will leave the question to the jury if it is a debatable one, but the jury may decide that (for example) plaintiff was beyond the apparent scope of danger from defendant’s conduct, and so beyond the scope of the duty to perform it carefully, even where they are quite ready to find defendant’s conduct clearly below the standard of reasonable care.” (3 Harper et al., The Law of Torts, supra, § 18.8, pp. 744-745.)
The majority alternatively justifies its indifference to the evidence that bears most directly on the question of duty by claiming that the problem is in the pleadings. For example, conceding that a “special relationship” may be established without an express or implied promise, the majority considers it fatal that respondents failed to “plead and prove that police conduct in a situation of dependency lulled [them] into a false sense of security, thereby inducing [their] detrimental reliance on the police for protection.” (Maj. opn., ante, at p. 282.) What the majority overlooks, however, is that appellants never asserted a defect in the pleadings prior to trial, when that issue should have been raised and any defect could have been cured by amendment (see Williams v. State of California, supra, 34 Cal.3d at p. 28), nor did they make such a claim at any other time either in the court below or now on this appeal. Any technical defect in the pleadings was waived.
It is on the basis of its erroneous belief that not just one but all of the foregoing factors must be present in order to create a “special relationship” that the majority attempts to distinguish this case from Mann v. State of California, supra, 70 Cal.App.3d 773. According to the majority, the “cornerstone” of Mann “was not simply police conduct that increased a preexisting risk of harm. Rather, the Supreme Court’s approval of the result in Mann rested on a variety of factors including: (1) police conduct that not only contributed to and increased the preexisting risk, but also changed the risk that would otherwise have existed; (2) the motorists’ situation of dependency; (3) the motorists’ detrimental reliance on the officers’ conduct that prevented them from seeking other assistance; and (4) the fact that the officers’ conduct lulled the motorists into a false sense of security.” (Maj. opn., ante, at p. 284, italics in original.) The majority’s belief that these factors must all be present, and that a “special relationship” cannot be created by any one of them is, as we have seen, contradicted by the opinion in Williams and by virtually all the other authorities.
But see Justice Mosk’s dissent in Williams v. State of California, supra, 34 Cal.3d at pages 28-30, concluding that the highway patrol officer in that case had a duty to assist the plaintiff and could not be considered a Good Samaritan.
The instruction was taken from language in the opinion in Allen v. Toten, supra, 172 Cal.App.3d 1079, 1089.
This is not a problem in California, however, because our Supreme Court has declared that “an omission or failure to act” may constitute a breach of duty. (Williams v. State of California, supra, 34 Cal.3d at p. 24.) For example, in Soldano v. O’Daniels (1983) 141 Cal.App.3d 443 [190 Cal.Rptr. 310, 37 A.L.R.4th 1183], the court recognized the possibility of liability for the wrongful death of a murder victim where the defendant bartender refused to permit the use of a public telephone to summon police help, when the victim was seen to be violently threatened. The distinction between misfeasance and nonfeasance remains important in this jurisdiction only because the imposition of liability is more likely in cases in which the defendant’s conduct consists of an affirmative act rather than a failure to act. (See Marois v. Royal Investigation & Patrol, Inc., supra, 162 Cal.App.3d 193, 198.)
MacPherson v. Buick Motor Co. (1916) 217 N.Y. 382 [111 N.E. 1050] and H. R. Moch Co. v. Rensselaer Water Co. (1928) 247 N.Y. 160 [159 N.E. 896, 62 A.L.R. 1199].
My belief that Rowland v. Christian does not apply to this case should not suggest I think it would justify a different result if it did apply. The easiest way to illustrate the point is to compare this case to Allen v. Toten, supra, 172 Cal.App.3d 1079, which was (erroneously, in my view) decided under Rowland, and is heavily relied upon by the majority.
Unlike the present case, the jury in Allen found that the police did not use excessive force and were not negligent in the manner in which they detained and arrested the husband, and that the police were not the cause of the husband’s injuries. (172 Cal.App.3d at p. 1084.) While the jury ruled against the husband and other relatives on their causes of action, it found in favor of the wife solely on her cause of action for emotional distress and awarded her $50,000 in damages. The only issue on appeal in Allen was whether the wife sufficiently stated a cause of action against peace officers for their alleged negligent infliction of emotional distress in bringing her to the scene of her husband’s threatened suicide. (Ibid.) Assuming there was no other basis upon which liability could be predicated, the court simply weighed the factors identified in Rowland v. Christian in order “to determine whether, as a matter of public policy, liability should be imposed upon peace officers and public entities for bringing a family member to the scene of a police standoff to aid in the surrender of an armed and suicidal relative.” (172 Cal.App.3d at p. 1087.) The Allen court determined that the interest in saving lives, which might be advanced by bringing a relative to the scene, was *312more important than “the interest of protecting some family members from the emotional trauma of viewing a suicide or wounding.” (Id., at p. 1089.) The court concluded that the highest priority was “the physical safety of the community, including [the police] themselves, other citizens, and family members,” and that “[d]issuading police, by imposing tort liability if things go awry, from exercising their best judgment in calling a family member to assist in disarming a suicidal person increases the burden on them by eliminating one means for peaceful resolution of a crisis.” (Id., at pp. 1089-1090.) The Allen court determined that the danger of bringing a family member to the scene was justified, because in any weighing of the competing considerations “preserving physical safety and life must be paramount.” (Id., at p. 1090.)
Applied to the very different facts of the present case, the policies considered transcendant in Allen justify the opposite result. Not only did the police in this case fail to take the action exonerated in Allen (which was among the reasons respondents’ experts believed they were negligent), but the action they took endangered not only respondents and the decedent, but the police themselves and anyone else who may have been on or near the scene. Neither the record before us nor the arguments of counsel provide a single policy justification for the gratuitously provocative acts of the police found negligent by the jury, certainly not the preservation of life. The only “policy” advanced by relieving the police of liability in this case is to free them from the ordinary consequences of even gross negligence, which does not protect but endangers life.
One of the strangest observations in the majority opinion is that the “imposition of a tort duty on public safety officers engaged in disarming suicidal persons is certainly likely to result in a more tentative police response to such crises.” (Maj. opn., ante, at p. 272.) I wholly agree with this statement and am at a complete loss to understand why the majority (which grudgingly concedes (maj. opn., ante, at p. 270) that appellants could have responded to the situation “in a less confrontational manner”) believes that result would not be salutary.
The cases relied upon by appellants are factually distinguishable because all involved situations in which, at the time of the injury-producing event, the plaintiffs were either not physically present or were then completely unaware of any danger to a family member. (See Justus v. Atchison (1977) 19 Cal.3d 564 [139 Cal.Rptr. 97, 565 P.2d 122]; Johnson v. County of Ventura (1994) 29 Cal.App.4th 1400 [35 Cal.Rptr.2d 150]; Evan F. v. Hughson United Methodist Church (1992) 8 Cal.App.4th 828 [10 Cal.Rptr.2d 748]; Fife v. Astenius (1991) 232 Cal.App.3d 1090 [284 Cal.Rptr. 16]; Golstein v. Superior Court (1990) 223 Cal.App.3d 1415 [273 Cal.Rptr. 270]; Wright v. City of Los Angeles (1990) 219 Cal.App.3d 318 [268 Cal.Rptr. 309]; Ebarb v. Woodbridge Park Assn. (1985) 164 Cal.App.3d 781 [210 Cal.Rptr. 751]; Jansen v. Children’s Hospital Medical Center (1973) 31 Cal.App.3d 22 [106 Cal.Rptr. 883].)
The Supreme Court noted in Thing v. La Chusa, that “[i]n most cases no justification exists for permitting recovery for NIED [(negligent infliction of emotional distress)] by persons who are only distantly related to the injured victim. Absent exceptional circumstances, recovery should be limited to relatives residing in the household, or parents, siblings, children, and grandparents of the victim.” (48 Cal.3d at p. 668, fn. 10.) In the present case, the trial court apparently determined that the stepdaughter, Gina Gohlston, was not too “distantly related” to Patrick and that the necessary “exceptional circumstances” were present.