Argued January 6, 2021; decided February 18, 2021

Greene v Esplanade Venture Partnership, 172 AD3d 1013, reversed.

This case begins with the heartbreaking death of a child. Our responsibility is to determine whether plaintiff grandparent Susan Frierson, who was in close proximity to the decedent grandchild at the time of the death-producing accident, may pursue a claim for bystander recovery under a "zone of danger" theory.

We have applied the settled "zone of danger" rule to "allow[ ] one who is . . . threatened with bodily harm in consequence of the defendant's negligence to recover for emotional distress" flowing only from the "viewing [of] the death or serious physical injury of a member of [that person's] immediate family" (Bovsun v Sanperi, 61 NY2d 219, 228 [1984] [emphasis added]). Unsettled at this juncture, however, are "the outer limits" of the phrase "immediate family" (id. at 233 n 13). Once again, we are not asked to fix permanent boundaries of the "immediate family." Instead, our task simply is to determine whether a grandchild may come within the limits of her grandparent's "immediate family," as that phrase is used in zone of danger jurisprudence.

We conclude that the grandchild comes within those limits. Consistent with our historically circumspect approach expanding liability for emotional damages within our zone of danger jurisprudence, our increasing legal recognition of the special status of grandparents, shifting societal norms, and common sense, we conclude that plaintiff's grandchild is "immediate family" for the purpose of applying the zone of danger rule.

On May 17, 2015, plaintiff Susan Frierson and her two-year-old granddaughter, decedent Greta Devere Greene, were in{**36 NY3d at 517} front of a building when they were suddenly struck by debris that fell from the facade of that edifice. Emergency measures taken to save Greta's life failed, and she died the next day.

Susan and Greta's mother, plaintiff Stacy Greene, subsequently commenced this action seeking damages for injuries sustained in that accident. The complaint was quickly superseded by an amended pleading in which plaintiffs alleged, among other things, that defendant Esplanade Venture Partnership owned the building, and that the remaining defendants were negligent with respect to the inspection of the facade of that structure. The amended complaint also alleged that the facade was in a dangerous condition, and that as a result, a piece of the facade broke, fell, struck Greta, and caused her to die.

Based on those allegations, plaintiffs asserted two causes of action; the first sounding in negligence, and the second in wrongful death. Nowhere in that amended pleading, however, did plaintiffs assert a cause of action for negligent infliction of emotional distress on behalf of Susan under the "zone of danger" doctrine.

Plaintiffs sought to cure that deficiency through a motion for leave to amend the amended complaint, and that application lies at the core of this appeal. In that motion, plaintiffs sought permission to "assert an additional cause of action on behalf of plaintiff Susan under the 'zone of danger' doctrine." That cause of action, plaintiffs contended, was appropriate in view of the "unique and special" nature of "the relationship between a grandparent and a grandchild."

To the extent the grandparent-grandchild relationship between Susan and Greta is not alone enough to bring Greta into Susan's "immediate family," plaintiffs maintained that the nature of the relationship warrants that classification. Susan, plaintiffs alleged, participated in Greta's birthing process, helped to care for Greta during the first few weeks of Greta's life, and subsequently developed a "powerful" "emotional bond" with Greta. By the time Greta was one year old, plaintiffs further alleged, Greta began to have overnight visits with Susan. It was during one of those visits that Susan was struck by debris that fell from the subject building, and Greta was struck and killed.{**36 NY3d at 518}

The motion to amend the amended complaint was granted (2016 NY Slip Op 33106[U] [2016]). Relying on the combination of our reasoning in Bovsun (61 NY2d at 232), this State's "specific recognition of the custody rights of grandparents with respect to their grandchildren," and the progression of zone of danger jurisprudence in other jurisdictions, Supreme Court subsequently denied defendants' motion to reargue, concluding that Susan "should be considered an 'immediate family member' and afforded a right to recover for her emotional injuries caused by this tragic accident" (2017 NY Slip Op 32335[U], *4-5 [2017]).

A divided Appellate Division reversed the 2016 order insofar as appealed from and denied the "branch of the plaintiffs' motion which was for leave to amend the amended complaint to add a cause of action sounding in negligent infliction of emotional distress" (172 AD3d 1013, 1014 [2d Dept 2019]). The majority ruled that leave to amend should have been denied (see id. at 1015) based on its interpretation of Bovsun (61 NY2d 219) and Trombetta v Conkling (82 NY2d 549 [1993]). Bovsun saw us hold "that a plaintiff may recover damages for emotional distress 'occasioned by [the plaintiff's] witnessing injury or death caused by the defendant's conduct to a member of the plaintiff's immediate family' (emphasis added)" (172 AD3d at 1015, quoting Bovsun, 61 NY2d at 224). That case, the Appellate Division believed, thus "stands for the proposition that spouses and their children are immediate family members" (172 AD3d at 1015, citing Bovsun, 61 NY2d at 233-234).

Bovsun was not an exercise in line-drawing. Although it identified certain relationships that come within the class of "immediate family members," Bovsun did not establish exhaustive boundaries with respect to the universe of "immediate family members." For that reason, the Appellate Division analogized this case—involving a grandmother and a granddaughter—to Trombetta (82 NY2d 549). There, we concluded that the plaintiff niece of a woman who was killed in the plaintiff's presence and with whom the plaintiff had a significant emotional bond was not entitled to "bring suit as a bystander for the negligent infliction of emotional injuries under the . . . 'zone of danger' rule" (id. at 550) because the decedent aunt was not part of the plaintiff niece's immediate family (see id. at 553; see also 172 AD3d at 1015-1016, citing{**36 NY3d at 519} Jun Chi Guan v Tuscan Dairy Farms, 24 AD3d 725, 725 [2d Dept 2005], appeal dismissed 7 NY3d 784 [2006]).

The dissenters at the Appellate Division would have affirmed Supreme Court's order. Though mindful of "the importance of precedent in our common-law system," the dissenters noted that the " 'living' " nature of the common law sometimes requires a " 'respon[se] to the surging reality of changed conditions' " (172 AD3d at 1016, quoting Gallagher v St. Raymond's R. C. Church, 21 NY2d 554, 558 [1968]) and maintained that the law should recognize that Greta was part of Susan's "immediate family" for the purpose of permitting a zone of danger claim (see 172 AD3d at 1028). The dissenters also rejected what they characterized as the majority's "use of consanguinity as a crude proxy for emotional harm" given the likelihood that "arbitrary and unjust results . . . will inevitably follow when, for instance, a child is denied recovery because [the child] does not live within a traditional family structure" (id. at 1030). "To be sure," the dissenters continued, "line drawing is often an inevitable element of the common-law process, but [it] . . . does not justify . . . clinging to a [boundary] that," as sketched by the Appellate Division, excludes Greta from the class of persons constituting Susan's immediate family (id. at 1031).

The Appellate Division subsequently granted leave to appeal to this Court and certified for our review the question whether its order was properly made. We now reverse and answer that question in the negative.

The past is always prologue. Our review of the merits begins with the underpinnings of our modern bystander zone of danger law.

In the nineteenth century, Mitchell v Rochester Ry. Co. (151 NY 107 [1896]) saw us conclude that "no recovery can be had for injuries sustained by fright occasioned by the negligence of another, where there is no immediate personal injury" (id. at 110). That now-outdated rule fittingly arose from antiquated circumstances; there the plaintiff was rendered unconscious and suffered a miscarriage during a "near miss" with the defendant's horse car (see id. at 108-109). Even assuming that the defendant was negligent in the management of its horses, however, we concluded that the plaintiff could not "recover for {**36 NY3d at 520}injuries occasioned by fright, as there was no immediate personal injury" (id. at 109). Central to that determination was the concern that recovery for emotional distress "would be contrary to public policy because that type of injury could be feigned without detection and it would result in a flood of litigation where damages must rest on speculation" (Battalla v State of New York, 10 NY2d 237, 240 [1961] [discussing Mitchell]).

By the 1930s, Mitchell had "been much discussed and frequently criticised by legal scholars" (Comstock v Wilson, 257 NY 231, 234 [1931]). In a point that carries weight today, we acknowledged that Mitchell had been shaped by views of policy rooted in the outlook of the times in which that case had been decided. We also noted that there are times in which "[p]ractical considerations must . . . determine the bounds of correlative rights and duties" (Comstock, 257 NY at 235). At that juncture, we believed that those practical concerns continued to require that a plaintiff be physically impacted in order to recover. Still, in Comstock we recognized that damages for "a mental disturbance" might be recoverable where the "fright" or "nervous shock" is produced by a concurrent physical impact (see id. at 237-239).

As time marched on, the law continued to drift away from Mitchell. By the early 1960s, we acknowledged that the core of the Mitchell conclusion—namely, that recovery for injury resulting from "mere fright" is not permitted—had been "demolished many times" (Battalla, 10 NY2d at 240). We also reconsidered the public policy underlying Mitchell (see Battalla, 10 NY2d at 240). The review was circumspect, to be sure; we balanced the possibility that "fraud, extra litigation and a measure of speculation are, of course, possibilities" in the event of a claim for fright against our awareness of the folly of denying "a logical legal right and remedy in all cases because in some a fictitious injury may be urged as a real one" (id. at 240-241). Against the backdrop of those competing considerations, and while mindful of factors including "justice," "experience[,] and logic" (id. at 239), we concluded that a plaintiff "subjected to the fear of physical injury as a direct result of the tortious conduct" (Howard v Lecher, 42 NY2d 109, 111 [1977]) may state a claim that he or she was "negligently caused to suffer 'severe emotional and neurological disturbances with residual physical {**36 NY3d at 521}manifestations' " when the defendant owed the plaintiff a direct duty, such as the one owed to the plaintiff by the ski-lift operator at issue in that case (Battalla, 10 NY2d at 238-239). By the late 1960s, we crystallized that point. There was "no longer any question" that, in some circumstances, "one may have a cause of action for . . . negligently induced mental trauma without physical impact" (Tobin v Grossman, 24 NY2d 609, 613 [1969]).

By the time Tobin was decided, we came to consider the issue whether a cause of action for negligent infliction for emotional distress might lie in favor of a bystander who did not suffer physical injury and who was owed no "direct duty" (see id.). Although we acknowledged developments in the law—including landmark developments in other areas in the tort field—we declined to recognize a cause of action for harm sustained by a "third person" parent as a result of injuries negligently inflicted directly upon her child when the parent was not in the zone of danger (see id. at 615-619; see also id. at 614, citing MacPherson v Buick Motor Co., 217 NY 382 [1916]). Stating the practical difficulties and the goal for the law to "limit the legal consequences of wrongs to a controllable degree," we observed that "[t]he risks of indirect harm from the loss or injury of loved ones is pervasive and inevitably realized at one time or another" (Tobin, 24 NY2d at 619).

In words that have failed the test of time, however, we also said that "[o]nly a very small part of that risk is brought about by the culpable acts of others" and that "[t]his is the risk of living and bearing children" (id.). That logic, and a concern with "[t]he problem of unlimited liability" (id. at 616), led us to conclude that it was "enough that the law establishes liability in favor of those directly or intentionally harmed" (see id. at 619).

Fifteen years later, we revisited the question and our policy consideration in Tobin to recognize a cause of action for bystander liability in part. Bovsun (61 NY2d 219) recognized that a plaintiff negligently

"expose[d] . . . to an unreasonable risk of bodily injury or death . . . may recover . . . damages for injuries suffered in consequence of the observation of the serious injury or death of a member of his or{**36 NY3d at 522} her immediate family—assuming, of course, that it is established that the defendant's conduct was a substantial factor [in] bringing about such injury or death" (id. at 230-231).

"Traditionally," we observed, "courts ha[d] been reluctant to recognize any [such] liability for . . . mental distress which may result from the observation of a third person's peril or harm" (id. at 227). However, by then, the "zone-of-danger rule, which" renders compensable emotional harm caused by the negligent infliction of injuries upon another person in certain cases, had "become the majority rule in this country" (id. at 228-229).^[FN1]

[*5]Consequently, we incorporated the zone of danger rule into our jurisprudence (see id. at 230-231; see also id. at 224). That rule "allows one who is . . . threatened with bodily harm in consequence of the defendant's negligence to recover for emotional distress" flowing only from the "viewing [of] the death or serious physical injury of a member of his or her immediate family" (id. at 228).

At the time Bovsun (61 NY2d 219) was decided, other courts had taken a more expansive view of liability for emotional distress resulting from viewing the death or serious physical injury of a relative. In Dillon v Legg (68 Cal 2d 728, 441 P2d 912 [1968]), for example, the Supreme Court of California held "that damages may be recovered for the emotional trauma caused when a plaintiff witnesses the injury or death of a close relative even though the plaintiff is not . . . within the zone of danger of physical injury, provided that the emotional injury is reasonably foreseeable" (Bovsun, 61 NY2d at 227 [discussing Dillon]). Bovsun's approach, of course, was more circumscribed. As in Tobin (24 NY2d at 618), the Bovsun Court expressly rejected adoption of the rule in Dillon, explaining that for those in the zone of danger,

"[r]ecognition of this right to recover for emotional distress attributable to observation of injuries suffered by a member of the immediate family involves{**36 NY3d at 523} a broadening of the duty concept but—unlike the Dillon approach—not the creation of a duty to a plaintiff to whom the defendant is not already recognized as owing a duty to avoid bodily harm" (61 NY2d at 229).

We emphasized that "the zone-of-danger rule provides a circumscribed alternative to the apparently sweeping liability recognized in Dillon . . . and does so within the framework of traditional and accepted negligence principles by using an objective test of whether the plaintiff was unreasonably threatened with bodily harm by the conduct of the defendant" (id. at 230).

Bovsun is also remarkable for what it did not say. While Bovsun recognized a zone of danger rule with an objectively defined class of bystanders as "immediate family," the Court did not list or enumerate "immediate family members." In fact, consistent with the caution with which we have historically approached this issue, we expressly declined to define the "outer limits" with respect to "the immediate family" element of the zone of danger rule (see id. at 233 n 13). "Inasmuch as all [of the] plaintiffs [at issue in Bovsun] were married or related in the first degree of consanguinity to the injured or deceased person," we left for another case the decision as to "where lie the outer limits of 'the immediate family' " (id.).

Our evolving zone of danger field jurisprudence is not the only development in the law relevant to our analysis.

In the 1990s we concluded that an unmarried, same-sex partner could adopt the partner's biological child (see Matter of Jacob, 86 NY2d 651, 655-656 [1995], citing Domestic Relations Law § 110, and 18 NYCRR 421.16 [former (h) (2)]). Same sex marriage was codified in 2011 (see Domestic Relations Law § 10-a [1] [added by L 2011, ch 95, § 3]) and affirmed in 2012 (see New Yorkers for Constitutional Freedoms v New York State Senate, 98 AD3d 285, 287 [4th Dept 2012], lv denied 19 NY3d 814 [2012]).

More recently we acknowledged that the definition of parent—which previously excluded a partner without a biological or adoptive relation to the subject child—had "become unworkable when applied to increasingly varied familial relationships" (Matter of Brooke S.B. v Elizabeth A.C.C., 28 NY3d 1, 14 [2016]). This analysis emphasized the point that roles and{**36 NY3d at 524} perspectives change, and that what once was accepted as a basic social premise has to be carefully examined in a way that reflects the realities of our lives.

Even more important in the context of this case is the legislative recognition of the changing nature of society's understanding of family and the special relationship between grandparents and grandchildren. In the 1960s the legislature established a vehicle for grandparents to obtain visitation rights with minor grandchildren (see Domestic Relations Law § 72; L 1966, ch 631, § 1). Although section 72 originally addressed only grandparent visitation, the statute was amended in 2003 "to provide guidance regarding the ability of grandparents to obtain [*6]standing in custody proceedings involving their grandchildren" (L 2003, ch 657, § 1). Following the amendment, section 72 now provides "that grandparents may demonstrate standing to seek custody based on extraordinary circumstances where the child has lived with the grandparents for a prolonged period of time, even if the child had contact with, and spent time with, a parent while the child lived with the grandparents" (Matter of Suarez v Williams, 26 NY3d 440, 444 [2015]).

The 2003 amendment was enacted based on an express legislative finding illustrative of the policy choice here. Namely, the legislature acknowledged that

"grandparents play a special role in the lives of their grandchildren and are increasingly functioning as care givers in their grandchildren['s] lives. In recognition of this critical role that many grandparents play in the lives of their grandchildren, the legislature finds it necessary to provide guidance regarding the ability of grandparents to obtain standing in custody proceedings involving their grandchildren" (L 2003, ch 657, § 1).

Thus, this Court recognized the "special status" of grandparents and that the unique path embodied in section 72 to establish extraordinary circumstances was enacted "in recognition of the important role of grandparents and the increasing number of grandparents raising their grandchildren" (Suarez, 26 NY3d at 446, 448). As we have recognized in other contexts (see e.g. Brooke S.B., 28 NY3d at 14), concepts of the creation and composition of family units have evolved beyond traditional legal notions of blood relation or consanguinity. What once was accepted as a basic social premise must be carefully examined{**36 NY3d at 525} in a way that reflects the realities of both our changing legal landscape and our lives.

It is here that the evolution of New York law with respect to bystander claims and the shifting understanding of varied familial relationships intersect.^[FN2] As noted, this case presents the question whether Susan, as the grandmother of Greta, may assert a viable cause of action for negligent infliction of emotional distress under the "zone of danger" theory based upon the emotional harm stemming from witnessing at close proximity the incident in which Greta was killed. There is no dispute that Susan was within the zone of danger at the time of that incident, meaning that the question whether she may assert that cause of action for negligent infliction of emotional distress turns on whether Greta was part of Susan's "immediate family" (see Bovsun, 61 NY2d at 228).

We have not established an outer boundary for "the immediate family" element of the zone of danger rule (see id. at 233 n 13). Here, we simply conclude that a grandchild is within our understanding of what is meant by "immediate family." That is, given the recognition by this Court and the legislature that the relationship of grandparent and grandchild enjoys a "special status" among familial relationships (Suarez, 26 NY3d at 448; see L 2003, ch 657, § 1), inclusion of grandparents in the common-law term "immediate family" under these circumstances is more than warranted.

This circumspect approach is consistent with our post-Bovsun decision in Trombetta v Conkling (82 NY2d 549 [1993]). There, we concluded that the decedent aunt was not "immediate family" of the plaintiff niece for the purpose of the zone of danger rule (see id. at 550). That determination was borne not{**36 NY3d at 526} of an investigation into the nature of the bond between the decedent and the plaintiff, but of our historical precision and prudence in this area (see id. at 553). We adhered to our articulated policy of "limiting the availability of recovery for the negligent infliction of emotional distress to a strictly and objectively defined class of bystanders" (id.). Critically, we also recognized that Bovsun did not "define the boundaries of 'the immediate family' " (id. at 552) and we did not do so in Trombetta either. Likewise, it is unnecessary to do so here, and that issue remains open.

This case calls upon us to blend the prudence we have shown in the course of many decades, including in Bovsun (61 NY2d 219) and that yielded Trombetta (82 NY2d 549) with recognition of reshaped societal norms and [*7]everyday common sense. We simply clarify that a discrete, limited class of persons that enjoys a special status under modern New York family law comes within the "narrow avenue to bystander recovery" (Trombetta, 82 NY2d at 552), and conclude that a grandchild is the "immediate family" of a grandparent for the purpose of applying the zone of danger rule.

The motion for leave to serve and file a second amended complaint should have been granted. "Permission to amend pleadings should be 'freely given' (CPLR 3025, subd [b])" (Edenwald Contr. Co. v City of New York, 60 NY2d 957, 959 [1983]), and the decision whether "to allow . . . the amendment is committed to the court's discretion" (id.). Here, inasmuch as the proposed amendment is not "patently devoid of merit" (172 AD3d at 1015), there was no reason to deny that application.^[FN3]

Accordingly, the order of the Appellate Division should be reversed, with costs, plaintiffs' motion to serve and file a second amended complaint granted in its entirety, and the certified question answered in the negative.

Rivera, J. (concurring).The Court has missed the moment. In Bovsun v Sanperi (61 NY2d 219 [1984]), the Court adopted the zone-of-danger rule that limits recovery for infliction of emotional distress to certain family members put within harm's{**36 NY3d at 527} way by defendants' conduct. On this appeal, we could have discarded that rule, which the Court knew then to be arbitrary and which served merely as a legally sanctioned excuse for "holding strict rein on liability" (id. at 230, quoting Tobin v Grossman, 24 NY2d 609, 618 [1969]). We could have explained that basing a right of recovery on whether a plaintiff was physically near to the injured or killed person leads to absurd results.

"[W]hile legislative bodies have the power to change old rules of law, nevertheless, when they fail to act, it is the duty of the court to bring the law into accordance with present day standards of wisdom and justice rather than 'with some outworn and antiquated rule of the past.' No reason appears why there should not be the same approach when traditional common-law rules of negligence result in injustice" (Woods v Lancet, 303 NY 349, 355 [1951], quoting Funk v United States, 290 US 371, 382 [1933]).

[*8]On this appeal, we could have discarded the Court's additional limitation that a person within the zone of danger must be the third party's "immediate family" member, defined strictly by marriage and degrees of consanguinity (see Bovsun, 61 NY2d at 233 n 13; Trombetta v Conkling, 82 NY2d 549, 553 [1993]). We would be justified in rejecting this outdated, patriarchal, and parochial definition, because families are formed not solely by matrimony and blood but also with bonds of friendship and love.

Lastly, we could have acknowledged that our rule was formed around the Court's assumption that a less stringent rule would lead to "unduly burdensome liability" (Tobin, 24 NY2d at 615), and then recognized that—as the experiences of other jurisdictions have proved—this concern is overstated. Regardless, such liability is merely "a kind of dollars-and-cents argument" (id. at 617) that neglects the dual purposes of tort law, which are to make wrongfully injured parties whole and provide a sufficient economic disincentive for injurious negligent conduct. We should have taken this opportunity to "reject[ ] as a ground for denying a cause of action that there will be a proliferation of claims" (id. at 615). "It suffices that if a cognizable wrong has been committed that there must be a remedy, whatever the burden of the courts" (id.). Having freed ourselves from the constraints of a retrograde and deeply flawed jurisprudence,{**36 NY3d at 528} the Court could have adopted a rule premised on the fundamentals of tort law—foreseeability, causation, and discernable harm—rather than a rule overwhelmingly concerned with assumed "unduly burdensome liability" (id.). Yes, the majority holding is a long overdue, though modest, expansion of the class of plaintiffs that are eligible to recover for emotional injuries sustained from watching loved ones suffer injury or death. And so I concur in the result. But the majority adheres to a legal framework that is arbitrary to the point of being contrary to public policy and blatantly unjust. The majority's small step is inadequate and solidifies an indefensible jurisprudence for the unforeseen future.

I write separately to explain why our rule should be that a person may recover for the emotional distress caused by perceiving the serious injury or death of any person with whom they shared a strong personal and loving bond. Alternatively, a person may recover if they contemporaneously observed the serious injury or death of another, regardless of their relationship, if they were at risk of immediate and serious physical harm from the defendant's conduct. This rule would eliminate the requirement that the plaintiff be related to the victim by marriage or consanguinity. The rule would also allow "strangers" to recover if the defendant violated a duty owed to them by putting them at risk of immediate and serious physical harm.

The majority has described with a broad brush the development of our common law that led to this Court's adoption of a standard that permits recovery for infliction of emotional distress caused by observing a defendant's harm to an immediate family member (majority op at 519-523). There are some additional observations and conclusions within these seminal cases worth noting, which illustrate why we should adopt a new rule as I have articulated.

In Battalla v State of New York (10 NY2d 237, 242 [1961]), the Court first recognized a claim for damages based on mental distress without physical injury. Battalla involved an infant that was placed in a chairlift by an employee of a ski resort and not properly secured (id. at 239). The infant "became frightened and hysterical" while riding on the chairlift and suffered severe emotional disturbance with residual physical manifestations as a result (id. 238-239). In finding that the{**36 NY3d at 529} plaintiff had stated a legally cognizable claim, the Court noted the well-understood principle, "fundamental to our common-law system[,] that one may seek redress for every substantial wrong" (id. at 240).

Battalla overruled Mitchell v Rochester Ry. Co. (151 NY 107, 110 [1896]), which prohibited recovery for emotional disturbance without physical injury. The Court explained that "a rigorous application of" the rule articulated in Mitchell "would be unjust, as well as opposed to experience and logic" (Battalla, 10 NY2d at 239).

Unlike the majority here, the Court in Battalla extolled our role in developing the common law. That role is not contingent upon legislative enactments, for as the Court explained, " '[w]e act in the finest common-law tradition when we adopt and alter decisional law to produce common-sense justice. . . . Legislative action there could, of course, be, but we abdicate our own function, in a field peculiarly nonstatutory, when we refuse to reconsider an old and unsatisfactory court-made rule' " (id., quoting Woods v Lancet, 303 NY 349, 355 [1951]).

Fidelity to the development of common law as the duty of an independent judiciary was well established long before its reaffirmation in Battalla and Woods. Aligning the common law with an evolving sense of justice is essential to the judicial exercise of this authority. Almost a century ago the United States Supreme Court explained that

" 'as it was the characteristic principle of the common law to draw its inspiration from every fountain of justice, we are not to assume that the sources of its supply have been exhausted. On the contrary, we should expect that the new and various experiences of our own situation and system will mould and shape it into new and not less useful forms.' . . .

"To concede this capacity for growth and change in the common law by drawing 'its inspiration from every fountain of justice,' and at the same time to say that the courts of this country are forever bound to perpetuate such of its rules as, by every reasonable test, are found to be neither wise nor just, because we have once adopted them as suited to our situation and institutions at a particular time, is to deny to the common law in the place of{**36 NY3d at 530} its adoption a 'flexibility and capacity for growth and adaptation' which was 'the peculiar boast and excellence' of the system in the place of its origin. . . .

"It has been said so often as to have become axiomatic that the common law is not immutable but flexible, and by its own principles adapts itself to varying conditions" (Funk, 290 US at 383).

In the same vein, the Battalla Court rejected the traditional arguments against expanding liability to include emotional injuries (10 NY2d at 240). Specifically, it questioned the Court's preoccupation in Mitchell with the possibility that emotional injury "could be feigned without detection" and the risk that expanding liability would lead to "a flood of litigation where damages must rest on speculation" (id.). The Court reasoned that, "[a]lthough fraud, extra litigation and a measure of speculation are, of course, possibilities, it is no reason for a court to eschew a measure of its jurisdiction" (id. at 240-241). In addition, the Court noted that "fraudulent accidents and injuries are just as easily feigned in the slight-impact cases and other exceptions wherein New York permits a recovery, as in no-impact cases which it has heretofore shunned" (id. at 241). Thus, the Court rejected the view that it was too difficult to discern the genuine from the fraudulent claim for nonphysical harm. That was the right choice because, as the Court observed, an "argument from mere expediency cannot commend itself to a Court of justice, resulting in the denial of a logical legal right and remedy in all cases because in some a fictitious injury may be urged as a real one" (id. [internal quotation marks omitted], quoting Green v T.A. Shoemaker & Co., 111 Md 69, 81, 73 A 688, 692 [1909]).

A few years later, California adopted a "bystander" recovery rule in Dillon v Legg (68 Cal 2d 728, 747, 441 P2d 912, 924-925 [1968]). In that case, California permitted a mother to recover for her emotional harm caused by observing the defendant's car hit her son. The court applied a foreseeability test, concluding that the emotional harm to the plaintiff was foreseeable even if the mother was not at risk of physical harm because the mother and victim shared a strong emotional bond and she was in close proximity to the events (68 Cal 2d at 741, 441 P2d at 920-921). The court rejected arguments that emotional harm was easily feigned, that its holding would encourage false claims, and that the defendant owed a duty of care solely to{**36 NY3d at 531} those in danger of physical harm (68 Cal 2d at 746-747, 441 P2d at 924-925).^[FN1]

The court noted legal commentators' observations that emotional injury should be treated as any other negligently inflicted injury (68 Cal 2d at 746, 441 P2d at 924, citing 2 Fowler V. Harper & Fleming James, Jr., The Law of Torts 1039 [1956], and Archibald H. Throckmorton, Damages for Fright, 34 Harv L Rev 260, 277 [1921]). For example, one commentator argued that

"[n]ervous shock resulting from fright is just one species of physical injury, and the rules of law, governing the right of recovery therefor, and the measure of recovery, are, or should be, the same as in all other cases of physical injury. The refusal to apply these general rules to actions for this particular kind of physical injury is nothing short of a denial of justice" (Throckmorton, 34 Harv L Rev at 277).

[*10]William L. Prosser described Dillon as a movement away from the zone-of-danger limitation, which he regarded as "no great triumph of logic," given the inconsistencies that it created within tort law (Prosser & Keeton, Torts § 54 at 366 [5th ed 1984]). Indeed, "[i]t is the business of the courts to make precedent where a wrong calls for redress, even if lawsuits must be multiplied" (id. at 360). And, "where the concern is to avoid imposing excessive punishment upon a negligent defendant, it must be asked whether fairness will permit leaving the burden of loss instead upon the innocent victim" (id. at 361; see also Leong v Takasaki, 55 Haw 398, 404, 520 P2d 758, 763 [1974] [describing limitations like the zone-of-danger requirement as "artificial devices" that "may actually foreclose relief to a genuine claim" that were unnecessary under the rule developed in Dillon]; Dunphy v Gregor, 136 NJ 99, 112, 642 A2d 372, 378 [1994] [noting that the need to put a limit on liability "does not outweigh the need to recognize claims that are legitimate and just"]).

{**36 NY3d at 532}New York did not follow Dillon's lead, although many other jurisdictions did.^[FN2] Less than a year after California's pathbreaking holding, our Court decided Tobin, which refused to allow recovery for emotional and mental damages incurred as a result of a defendant's injury to another (24 NY2d at 619). Like Dillon, Tobin involved a mother who sought recovery for her emotional and mental injuries caused by perceiving defendant's negligent harm to her child. Unlike in Dillon, the mother was not an eyewitness to the events; she heard the screech of automobile brakes and then discovered moments later that her infant child had been struck by a car and seriously injured (id. at 611).^[FN3] In rejecting the mother's claim for recovery, the Court acknowledged that "the common law is not circumscribed by syllogisms, however constructed out of precedents" (id. at 614).

However, the Court in Tobin erred in framing the issue as whether to create a new cause of action, which according to the Court required a "radical change in policy" that it was unwilling to make (id. at 615). Yet, all that was necessary was for the Court to employ the same logic of Battalla to extend the cause of action recognized therein to an individual who asserts emotional and mental injuries, leaving it to traditional principles of tort law to address that person's entitlement to recovery. As Judge Keating explained in dissent, the holding was unsupported by reason given that the majority "ha[d] shown that every element necessary to build a case for tortious liability in negligence is here present" (id. at 619 [Keating, J., dissenting]). Specifically, "[t]here is an important interest worthy of protection, there is proximate cause, there is injury, and there is foreseeability" (id.).

As both the Tobin majority and dissent recognized, withholding recovery from the mother was patently arbitrary (id. at 618, 620). In rejecting recovery nonetheless, the Tobin majority reviewed "several [policy] factors most often considered" in the development of the law governing recovery for emotional injury. These factors "overlap in various degrees" (id. at 615).

{**36 NY3d at 533}The Court had previously rejected the threat of a proliferation of claims and an increase in fraudulent litigation as outcome-determinative factors, and so handily rejected these as bases for deciding whether plaintiff should recover in Tobin (id. at 615-616). The Court concluded, however, that the remaining "factors"—foreseeability, the "inconsistency of the zone of danger rule, unlimited liability, unduly burdensome liability, and the difficulty of circumscribing the area of liability"—all weighed against plaintiff (id. at 615).

The Court noted that section 313 of the Restatement (Second) of Torts and several states had adopted the zone-of-danger rule, although it acknowledged that Dillon viewed that rule as a "rather arbitrary limiting rule which has the unpalatable consequence that a mother who also fears for herself may recover while, if she does not or has no such similar opportunity, she may not recover" (id. at 616, citing Dillon, 68 Cal 2d at 747, 441 P2d at 924-925). While [*11]unlimited and unduly burdensome liability were constants under any theory of recovery, the "most difficult factor" to resolve, according to the Court, was the "reasonable circumscription, within tolerable limits required by public policy, of a rule creating liability" (id. at 617).

As the Court recognized, a parent need not witness the events that caused harm to their child for the parent "to sustain grievous psychological trauma" (id.). "The sight of gore and exposed bones is not necessary to provide special impact on a parent" (id.). And apart from the irrationality of conditioning recovery on whether the parent was an eyewitness to the harm, there is "the logical difficulty of excluding the grandparent, the relatives, or others in loco parentis, and even the conscientious and sensitive caretaker, from a right to recover, if in fact the accident had the grave consequences claimed, rais[ing] subtle and elusive hazards in devising a sound rule in this field" (id.).

The Tobin dissent argued that arbitrariness alone warranted discarding the limitation on recovery. The majority justified its holding as necessary to limit defendants' liability (id. at 618). But the Court rejected the limitations adopted in Dillon because, according to the Court, "none of these standards are of much help if they are to serve the purpose of holding strict rein on liability and if the test is to be a reasonably objective one" (id.).

{**36 NY3d at 534}It was not until Bovsun that the Court finally held that third parties could recover "damages for injuries suffered in consequence of the observation of the serious injury or death" of another in certain circumstances (61 NY2d at 231). The Bovsun Court correctly viewed the issue as an expansion of the plaintiff's right to recover and not the creation of a new duty of the defendant (id. at 229). Nor did the Court create a new cause of action:

"There may be an enlargement of the scope of recoverable damages; there is no recognition of a new cause of action or of a cause of action in favor of a party not previously recognized as entitled thereto. In conformity with traditional tort principles, the touchstone of liability in these cases is the breach by the defendant of a duty of due care owed the plaintiff" (id. at 233).

Nevertheless, in its quest to limit liability, the Court adopted two requirements intended to circumscribe the class of plaintiffs who could assert this right of recovery (id. at 230-231). Under the first, a plaintiff may recover if they are within the zone of danger created by the defendant's conduct (id. at 223-224). This requirement does not require direct infliction of physical harm to the plaintiff, as the Court long ago rejected what had come to be known as the physical impact test (see Tobin, 24 NY2d at 613 [citation omitted]). As the analysis goes, a defendant who has placed a plaintiff in the zone of danger created by the defendant's conduct has violated a duty of reasonable care owed to the plaintiff, who may recover for emotional harm, even if the plaintiff is not physically impacted or injured (Bovsun, 61 NY2d at 229). Notwithstanding the rule's detractors, the Court viewed it "as comporting with the requirements set out in Tobin of a 'reasonably objective' standard which will 'serve the purpose of holding strict rein on liability' " (id. at 230, quoting Tobin, 24 NY2d at 618).

The second requirement limits recovery only to persons who are members of the third-party victim's "immediate family" (id. at 230-231). Because the plaintiffs in Bovsun and its companion case "were married or related in the first degree of consanguinity to the injured or deceased person," the Court had no occasion to opine at the time on "the outer limits of 'the immediate family' " (id. at 233 n 13).

It would not be long before the Court defined the boundary of "immediate family" in Trombetta. In that case, the Court {**36 NY3d at 535}denied a niece recovery as a bystander to her aunt's death, though the niece "shared a long and strong emotional bond" with her aunt who had cared for her and acted as her mother since she was a child (82 NY2d at 550). The Court held that the definition of "immediate family" was "confine[d] . . . to only the immediate family as surveyed in Bovsun" and "defined and limited by Bovsun" (id. at 553, 554). "Immediate family," limited to spouses, parents, and children, was "a strictly and objectively defined class" that was "discrete [and] readily determinable" (id. at 553-554).

The Court noted that, regardless of the quality of the plaintiff's bond with her aunt, and "[a]lthough plaintiff suffered a personal tragic loss," those facts could not "justify the significant extension of defendants' obligation to be answerable in damages for [plaintiff's] emotional trauma" (id. at 553). "On firm public policy grounds," bystander recovery [*12]was foreclosed to individuals "who may be able to demonstrate a blood relationship coupled with significant emotional attachment or the equivalent of an intimate, immediate familial bond" but were not "immediate family" as defined by Bovsun (id.).

The Court's limiting of the meaning of "immediate family" to the relationships "surveyed" in and "defined and limited" by Bovsun, and its further admonition that the definition did not include others "who may be able to demonstrate" the "equivalent of an intimate, immediate familial bond," was intended to close the door on further expansion of the plaintiff class. Thus, and contrary to the majority view, I agree with Judge Garcia that Trombetta is properly understood as holding that only spouses, parents, and children can recover for their emotional injuries caused by observing defendant's third-party harm (compare Garcia, J., concurring op at 548, with majority op at 525-526). However, the Bovsun/Trombetta approach, and the majority's inclusion of grandparents and grandchildren within the definition of "immediate family," removes any rational ground for excluding other close bonds that are functional equivalents.^[FN4]

As this discussion illuminates, the zone of danger and "immediate family" limitations suffer from analytic and practical {**36 NY3d at 536}deficiencies. I now turn to approaches broader than New York's to assist in designing a truly "objectively reasonable test."

Many jurisdictions have acknowledged the lack of sound policies justifying harsh limitations on a plaintiff's recovery for emotional distress and have adopted more flexible or permissive rules. Oregon has adopted bystander recovery exactly as it is articulated in the Third Restatement of Torts, which allows recovery by close family members who perceive accidents even if they are not within the zone of danger (Philibert v Kluser, 360 Or 698, 711, 385 P3d 1038, 1046 [2016]). And several other states have adopted tests with only slight differences from the Third Restatement rule (see e.g. La Chusa, 48 Cal 3d at 647, 771 P2d at 815 [allowing bystanders "closely related" to direct victims to recover for "emotional distress beyond that which would be anticipated in a disinterested witness"]; Jones v City of Houston, 294 SW3d 917, 920 [Tex App 2009] [allowing bystanders who have "suffered shock as a result of a direct emotional impact" and are "closely related" to the direct victim to recover]; Cameron v Pepin, 610 A2d 279, 284-285 [Me 1992] [requiring that the plaintiff be "closely related to the victim"]; Satchfield v R.R. Morrison & Son, 872 So 2d 661, 664 [Miss 2004] [same]; Portee, 84 NJ at 101, 417 A2d at 528 [requiring "a marital or intimate, familial relationship between plaintiff and the injured person"]; Clifton v McCammack, 43 NE3d 213, 218 [Ind 2015] [requiring the plaintiff to have witnessed the death or injury of "a loved one with a relationship to the plaintiff analogous to a spouse, parent, child, grandparent, grandchild, or sibling caused by the defendant's negligent or otherwise (tortious) conduct"]; Graves v Easterbrook, 149 NH 202, 209, 818 A2d 1255, 1261-1262 [2003] [requiring that the plaintiff and physically injured victim share a "relationship that is stable, enduring, substantial, and mutually supportive" and "cemented by strong emotional bonds and provid(ing) a deep and pervasive emotional security"]).

Some states that require strict familial relationships between bystander plaintiffs and injured third parties are still less restrictive than New York (see e.g. Barnhill v Davis, 300 NW2d 104, 108 [Iowa 1981] [requiring that the "bystander and the victim were husband and wife or related within the second degree of consanguinity or affinity"]; La Civ Code Ann art 2315.6 [A] [allowing spouses, children, siblings, grandparents,{**36 NY3d at 537} and grandchildren to recover as bystanders]; Grotts v Zahner, 115 Nev 339, 989 P2d 415 [1999] [employing a tripartite scheme in which nuclear family members have a sufficient relationship to assert a bystander claim, nonfamily members may not, and relationships between non-nuclear family members are submitted to the jury to decide if the relationship is sufficiently close to permit recovery]).

Other states use Dillon's foreseeability test, which is more permissive of recovery than the Third Restatement's test. "The touchstone" of that test "is not a rigid requirement of sensory and contemporaneous observance of the accident" or strict relational criteria, "but rather" whether there is a "reasonable foreseeability that the plaintiff-witness would suffer emotional harm" (Tommy's Elbow Room, Inc. v Kavorkian, 727 P2d 1038, 1043[*13][Alaska 1986]; accord Leong, 55 Haw at 408, 520 P2d at 764-765; Sinn v Burd, 486 Pa 146, 172-173, 404 A2d 672, 686 [1979]; Wages v First Natl. Ins. Co. of Am., 318 Mont 232, 239, 79 P3d 1095, 1100, 2003 MT 309, ¶ 25 [2003]; Paugh v Hanks, 6 Ohio St 3d 72, 78, 451 NE2d 759, 765 [1983]; James v Lieb, 221 Neb 47, 55, 375 NW2d 109, 115 [1985] [adopting the foreseeability rule but adding that plaintiffs must show a "marital or intimate familial relationship"]; Heldreth v Marrs, 188 W Va 481, 493, 425 SE2d 157, 169 [1992] [adopting a foreseeability test but limiting recovery to "close relatives"]; Hegel v McMahon, 136 Wash 2d 122, 136, 960 P2d 424, 431 [1998] [adopting a foreseeability test but requiring "objective symptoms of the distress that are susceptible to medical diagnosis and proved through qualified evidence"]).

While there are other holdouts throughout the country that have adopted rules that are as similarly harsh as New York's, or deny bystander recovery altogether,

"as is plain from approaches such as those employed by the New York courts, especially in Trombetta v. Conkling, the concentration on difficulties in 'reasonably circumscribing the area of liability' represents a policy perception betokening greater concern with the burdens of imposing liability than with compensating loss, even where clearly attributable to the fault of another. Separating out the underlying philosophical orientations, that concern seems more apparent than real, especially when, to date, there is no sense of explosive expansion of tort liability either in those jurisdictions that have{**36 NY3d at 538} recognized the bystander's claim or in those that have established a generally available tort for the negligent infliction of emotional distress" (Howard H. Kestin, The Bystander's Cause of Action for Emotional Injury: Reflections on the Relational Eligibility Standard, 26 Seton Hall L Rev 512, 542 [1996]).

New York's failure to reconsider and reject these unsupported policy justifications has rendered us an outlier. The Court has used those policies to shield its strict limitations and kept us from joining jurisdictions like California, Texas, and New Jersey in devising a rule better suited to address negligently inflicted harm. As jurisdictions adopt a more permissive rule without experiencing a flood of litigation, the more the policy foundation upon which New York's rule is built erodes.^[FN5]

The Second Restatement provided that if an actor's conduct is negligent and creates "an unreasonable risk of causing bodily harm to another," and where that harm "results from . . . shock or fright at harm or peril to a member of" the plaintiff's "immediate family occurring in" the plaintiff's "presence," the plaintiff could recover (Restatement [Second] of Torts § 436 [2], [3] [1965]).

The comments to this section noted that the rule applied "even though the plaintiff's shock or fright is not due to any fear for" the plaintiff's safety, "but to fear for the safety of" the plaintiff's "wife or child" (id., Comment f). In a caveat to the rule, the Restatement's commentators noted that, due to an absence of sufficient decisions, they "expresse[d] no opinion as to whether the rule" would allow for recovery for "shock or fright at harm or peril to a third person who is not a member" of a plaintiff's "immediate family, or where the harm or peril does not occur in" the plaintiff's "presence" (id., Caveat; Comment on Caveat h).

Similarly, section 46 of the Second Restatement provided that, where a person injures or kills a third person, the actor is{**36 NY3d at 539} subject to liability if they "intentionally or recklessly cause[ ] severe emotional distress (a) to a member of such person's immediate family who is present at the time, whether or not such distress results in bodily harm, or (b) to any other person who is present at the time, if such distress results in bodily harm" (Restatement [Second] of Torts § 46 [2] [1965]). Thus, under section 46, the Second Restatement recognizes two classes of "bystander" plaintiffs: the immediate family member and any person whose distress results in bodily harm.

[*14]As with the commentators' caveat to section 436, Comment l to section 46 of Restatement (Second) of Torts explains that the language "leave[s] open the possibility of situations in which presence at the time may not be required." Further, although the cases decided up to the point of the Second Restatement's publishing involved a plaintiff who was at least a close associate with the person harmed, the analysis could extend to strangers: "there appears to be no essential reason why a stranger who is asked for a match on the street should not recover when the [person] who asks for it is shot down before [their] eyes, at least where [the stranger's] emotional distress results in bodily harm" (Restatement [Second] of Torts § 46, Comment l [1965]).

The Third Restatement, Liability for Physical and Emotional Harm, first published in 2012 and updated in 2020, reflects the further evolution of the rules governing "bystander recovery" in the ways left open by the Second Restatement. Under the Third Restatement, "[a]n actor who negligently causes sudden serious bodily injury to a third person is subject to liability for serious emotional harm" caused to "a person who: (a) perceives the event contemporaneously, and (b) is a close family member of the person suffering the bodily injury" (Restatement [Third] of Torts: Liability for Physical and Emotional Harm § 48 [Oct. 2020 update]).

The bystander recovery rule articulated in the Third Restatement differs from the rule developed in Bovsun and Trombetta. It mandates that the plaintiff perceived the harm to the third party as it occurred. The comments to the Third Restatement explain that the contemporaneous-perception requirement "is not limited to sight" and give the example of a blind person hearing a close family member suffer harm as an instance where recovery would be permitted (id., Comment e).

In addition, the Third Restatement requires only that the plaintiff and injured or killed third party are "close family {**36 NY3d at 540}member[s]" as opposed to "immediate family." The Third Restatement commentators urged courts to adopt flexible definitions for "close family." As the Third Restatement commentators explained, "[s]ometimes people live functionally in a nuclear family without formal legal family ties" (id., Comment f). Thus, "[w]hen defining what constitutes a close family relationship, courts should take into account changing practices and social norms and employ a functional approach to determine what constitutes a family" (id.).

That view is supported by studies of human relations and modern family structures. For example, the percentage of children living with unmarried, cohabitating parents more than doubled between 1997, the first year that census data on cohabitation was available, and 2017; and in 2018 approximately one third of children were living with an unmarried parent (Gretchen Livingston, About One-Third of U.S. Children are Living with an Unmarried Parent, Pew Research Center, Apr. 27, 2018, https://www.pewresearch.org/fact-tank/2018/04/27/about-one-third-of-u-s-children-are-living-with-an-unmarried-parent/ [last accessed Feb. 5, 2021]; see also Rebecca L. Melton, Note, Legal Rights of Unmarried Heterosexual and Homosexual Couples and Evolving Definitions of "Family," 29 J Fam L 497, 499 [1991] [surveying statistical data indicating that American families take many varied forms, including group living and unmarried cohabitation]; Mary Patricia Treuthart, Adopting A More Realistic Definition of "Family," 26 Gonz L Rev 91, 91-92 [1991] [noting that most families are not organized around a married man and woman occupying traditional gender roles]; Michael J Rosenfled, Young Adulthood as a Factor in Social Change in the United States, 32 Population & Dev R 27, 40-47 [2006] [describing a trend away from traditional family structures dominant before and during the industrial revolution towards alternative family structures, including unmarried cohabitation, and attributing the shift to a less oppressive culture]; Benjamin Gurrentz, For Young Adults, Cohabitation is Up, Marriage is Down: Living with an Unmarried Partner Now Common for Young Adults, United States Census Bureau, Nov. 15, 2018, https://www.census.gov/library/stories/2018/11/cohabitaiton-is-up-marriage-is-down-for-young-adults.html [last accessed Feb. 5, 2021] [In 1968, only "0.2 percent of 25- to 34-year-olds lived with an unmarried partner"; in 2018, 14.8% did]).

Moreover, since this Court's decision in Tobin, multigenerational households have almost doubled in number; one in five{**36 NY3d at 541} Americans—20% of the United States population—live in these arrangements (D'Vera Cohn & Jeffrey S. Passel, A Record 64 Million Americans Live in Multigenerational Households, Pew Research Center, Apr. 5, 2018, https://www.pewresearch.org/fact-tank/2018/04/05/a-record-64-million-americans-live-in-multigenerational-households/ [last accessed Feb. 5, 2021]; see also Frank F. Furstenberg, Fifty Years of Family Change: From Consensus to [*15]Complexity, 654 Annals Am Acad Pol & Soc Sci 12, 13-15 [2014] [describing the increasing complexity in types of familial structures throughout the world]).

Unsurprisingly, the revolution in awareness of the diversity of family arrangements has worked its way into many areas of our jurisprudence. As a constitutional matter, for zoning purposes, unrelated individuals living together may be the "functional and factual equivalent" of a family (see e.g. Baer v Town of Brookhaven, 73 NY2d 942, 943 [1989] [finding that five unrelated elderly women who lived together in a home were a family]; McMinn v Town of Oyster Bay, 66 NY2d 544, 551 [1985] [holding unconstitutional a town zoning ordinance that defined "family" to exclude households consisting of more than two unrelated or unmarried persons younger than 62]; Group House of Port Washington v Board of Zoning & Appeals of Town of N. Hempstead, 45 NY2d 266, 272 [1978] [holding that a group home in which two surrogate parents cared for seven foster children was a family]; City of White Plains v Ferraioli, 34 NY2d 300, 303 [1974] [finding that, for purposes of a city's zoning regulations, a group home consisting of a married couple, their two biological children, and 10 foster children constituted a single family]).

Other cases similarly demonstrate that when tasked with recognizing the bonds that comprise family, the Court has adopted an approach that does not depend solely upon the rigid familial markers of birth or marriage. For example, in Braschi v Stahl Assoc. Co. (74 NY2d 201 [1989]), Judge Vito J. Titone, writing for the Court, explained the proper test for interpreting the word "family" in New York's rent-control code. The Court rejected the Appellate Division's holding that the rent-control code limited tenant succession rights to "family members within traditional, legally recognized familial relationships" (id. at 207). Instead, in order to serve the broad remedial aims of the rent-control law,

"the term family, as used in 9 NYCRR 2204.6 (d),{**36 NY3d at 542} should not be rigidly restricted to those people who have formalized their relationship by obtaining, for instance, a marriage certificate or an adoption order. The intended protection against sudden eviction should not rest on fictitious legal distinctions or genetic history, but instead should find its foundation in the reality of family life" (id. at 211).

To aid in recognizing the "reality of family life," the Court described a non-exhaustive list of factors for courts and juries to consider, including the "longevity of the relationship, the level of emotional and financial commitment, the manner in which the parties have conducted their everyday lives and held themselves out to society, and the reliance placed upon one another for daily family services" (id. at 212-213). Braschi acknowledged that the lower courts had already ably applied those factors in a variety of cases to determine whether "the totality of the relationship as evidenced by the dedication, caring and self-sacrifice of the parties" demonstrate that two people were family (id. at 213).

Scores of cases decided before and after Braschi prove that New York courts are well-equipped to apply a functional mode of analysis in order to identify strong and caring bonds, when the important remedial purposes of New York law so require (e.g. 2-4 Realty Assoc. v Pittman, 137 Misc 2d 898, 899-900 [Civ Ct, NY County 1987] [two men developed a father-son relationship after supporting each other through 25 "long years of life, poor health, and eventual death"], affd 144 Misc 2d 311 [App Term, 1st Dept 1989]; Zimmerman v Burton, 107 Misc 2d 401, 403 [Civ Ct, NY County 1980] [recognizing the tenant succession rights of unmarried partners and noting that "(t)he law must keep abreast of changing moral standards"]; Fleishman Realty Corp. v Garrison, 27 Misc 3d 1202[A], 2010 NY Slip Op 50527[U], *9 [Civ Ct, Bronx County 2010] [finding that over a decade a tenant formed "loving, committed, long-term relationships" with his boyfriend and boyfriend's mother]).

Thus, we come to the foundational error of Tobin that continues to taint this Court's analysis. The specter of unlimited liability is a weak justification for our tort rule, especially when it is unproven and when its application leads to irrational results. As the Tobin dissent observed,

"Not one piece of evidence is offered to prove that{**36 NY3d at 543} the 'dollar-and-cents' problem [of unlimited liability] will have the dire effects claimed. More important, the manner in which the argument about infinite liability is explicitly rejected one day only to be revived the next is indicative of what may be [*16]described as a rather erratic method of decision. One can easily recognize the ad hoc nature of the argument by comparing passages in our opinions where we have chosen to do what modern concepts of justice demand[ ] with those cases where we have hesitated to follow through on the basic currents in the law of torts for the past generation. The statements in these opinions cannot be reconciled. . . . [T]here has been an expanding recognition that the argument concerning unlimited liability is of no merit, yet the aberrations persist. One would imagine that we were here involved with a catastrophic loss. There have already been decisions imposing liability of far greater dimension than can ever arise if we should embark upon a search for 'essential justice' in the bystander class of cases" (24 NY2d at 620 [citations omitted]).

The dissent is not alone.

"Too many respectable commentators have dismissed the 'flood of litigation' argument for it to detain us. The problem of imposing 'unlimited and unduly burdensome liability' is no greater as the jurisprudence relating to this cause of action develops than it has classically been in the past as causes of action have been articulated and have matured. As we gain familiarity with the various elements of any cause of action, our concerns about the burdens of coping with them are usually allayed. Typically, it is not a new development expanding a cause of action that itself creates problems of understanding and administration but, rather, the sense of unfamiliarity with its requirements. By now, with respect to typical factual issues concerning the existence and extent of emotional injury, this is no longer a real problem" (Howard H. Kestin, The Bystander's Cause of Action for Emotional Injury: Reflections on the Relational Eligibility Standard, 26 Seton Hall L Rev 512, 542 [1996]).{**36 NY3d at 544}

Moreover, as others have observed, "rigid doctrinal limitations on liability to bystanders produce arbitrary, incongruous and indefensible results. Plaintiffs in substantially the same position have been treated differently" (Bowen v Lumbermens Mut. Cas. Co., 183 Wis 2d 627, 651-652, 517 NW2d 432, 442 [1994]). Consider the following examples.

A parent stands on a street corner with their child waiting for a car service, when the defendant negligently loses control of their car and fatally strikes the child. The parent suffers emotional trauma from losing their child and has nightmares of the accident in which the parent relives the moment when they saw the car hit the child.

As a person within the zone of danger created by the defendant's conduct and a person who was the "immediate family" member of the decedent, as defined in Trombetta, the parent may recover for the physical and emotional injuries suffered as a direct cause of the defendant striking them as well as the emotional trauma caused by observing the defendant strike and kill their child.

Incongruously, if the parent is a block away from the child waiting on the corner, and the child is at all times within the parent's line of sight, the parent cannot recover in New York because they were not within the zone of danger, although they are an "immediate family" member. Where is the fairness in this outcome? Why is the defendant in this example more worthy of protection from liability than the defendant in the first example? In both scenarios the defendant killed a child on a public street in plain view of the child's parent. The plaintiff's injury and the defendant's culpability are the same in both examples. The measure of deterrence required is also identical. There is no defensible justification for denying recovery to a plaintiff simply because of where they stood relative to the accident that they apprehended. Only blind adherence to a doctrine grounded in an unproven concern for limiting liability explains the difference in outcome.^[FN6]

Take the further example of a parent estranged from their minor child, who has not seen the child since the child's birth. The child lives with the other parent who has sole legal and{**36 NY3d at 545} physical custody pursuant to a court order, and who has raised and cared for the child since birth. The estranged parent seeks to reconnect and finds the child on the street as the child waits for a school bus, while the custodial parent watches from the window of their second-story apartment, directly above the bus stop. At that moment, the defendant's car fatally strikes the child.

Under New York's rule, the custodial parent may not recover, despite losing the child they nurtured since birth. This result, on its own, is indefensible. As with the prior example, this parent experiences injurious emotional distress over the loss of their child that is in no way lessened merely because the custodial parent was outside the zone of danger. But it further contravenes public policy by allowing the noncustodial parent to recover, in disregard of both society's and the law's recognition of the long-standing bond between the custodial parent and their child.

Or assume, as in Trombetta, a woman within the zone of danger created by the defendant's negligence watches a car strike and kill her aunt, who has been the woman's caretaker and a parental figure since childhood. The niece could not recover under Trombetta. And the majority's limited expansion of bystander recovery to grandparents and grandchildren would not expressly apply. Yet one cannot seriously argue that grandparents are intimate family members (based on presumptions of their role within a family), but the parental figure in this hypothetical should not be considered an intimate family member.^[FN7]

Marriage and blood relationships have served for too long as the primary measure for a strong human bond based on affection and connectedness. By its own terms, this measure is overinclusive, as it assumes a reaction that may not be accurate given the actual nature of the relationship. It is a strange principle that seeks to circumscribe liability but allows for recovery on such an assumption. But that is not the real problem because our system already disallows recovery without actual injury and the plaintiff must still provide proof of real emotional trauma.{**36 NY3d at 546}

The real problem is that this limitation is underinclusive because it assumes that only spouses and certain relatives have the type of emotional attachment to the third-party victim that justifies recovery. That assumption cannot stand the cold light of reality. Human beings have strong, loving connections to non-relatives and non-spouses that are at least as meaningful as those between spouses and to grandparents (Restatement [Third] of Torts: Liability for Physical and Emotional Harm § 48, Comment f [Oct. 2020 update] [urging a functional as opposed to formal definition of family]; Rebecca L. Melton, Note, Legal Rights of Unmarried Heterosexual and Homosexual Couples and Evolving Definitions of "Family," 29 J Fam L 497, 499 [1991]; Mary Patricia Treuthart, Adopting A More Realistic Definition of "Family," 26 Gonz L Rev 91, 91-92 [1991]). In the past we have not hesitated to reject outdated holdings and adopt standards and rules that better reflect life and family structures (Matter of Brooke S.B. v Elizabeth A.C.C., 28 NY3d 1, 26 [2016] [overruling a strict definition of "parent" that excluded non-biological and non-adoptive caregivers]).

We are also now well past the point of denying rights to persons who could not marry or who decide that it best serves them and their families not to marry (see e.g. Public Health Law § 2805-q [prohibiting hospitals from barring domestic partners from visiting one another]; 22 NYCRR 24.4 [a] [2] [defining "close family member" for purposes of paid sick leave for non-Family Medical Leave Act approved absences as an "employee's spouse; domestic partner; natural, foster or step child; natural, foster or step parent; or any relative residing with the employee or an individual for whom the employee is the primary caregiver"]; 12 NYCRR 380-4.4 [providing that proof of domestic partnership is sufficient in part to establish a non-birthing parent's right to take leave to bond with their child]). These legislative changes are no less powerful than the ones upon which the majority relies to expand tort liability to include grandparents, yet the majority offers no principle to distinguish legislative recognition of the familial relationship of grandparents from domestic partners.

The plaintiff in Trombetta is a prime example of the way many families are structured. But even without a blood relation, people form bonds, for example, based on a shared culture and a mutual experience born of shared struggle. We cannot ignore that unrelated children who grow up together consider themselves siblings in every sense, or that individuals who{**36 NY3d at 547} have grown close through shared experiences may be family in every sense of the word save [*17]for biology. Our rule should no longer preclude recovery by persons whose lives are intertwined but choose not to marry because their families would ostracize them for marrying someone of the same gender, or because they are of a different race or ethnicity, or from outside of their religious community, or because they simply choose not to marry. Our rule denying recovery in these situations cannot withstand the logic by which the Court extends bystander recovery today.

Further, the immediate family requirement should be eliminated for persons within the zone of danger whose emotional distress caused by observing the serious injury or death of another manifests in physical harm. In that case, the defendant owes a duty directly to the plaintiff and traditional rules of foreseeability and proof of actual injury should apply to all aspects of the claim and demands for recovery. Under those rules, it is foreseeable that a person within the zone of danger who fears for their life would suffer a particular serious emotional trauma if they survive but observe another's severe or fatal injuries. The third-party victim's harm is a constant reminder of the plaintiff's own close call. In contrast, a bystander across the street may be horrified by what they see, but they will not associate that memory with their own near-death encounter or harm.

Allowing recovery is not arbitrary in this circumstance because someone in physical proximity to the victim has a meaningfully different experience from that of someone who perceives no risk to themselves and suffers no injury (Restatement [Second] of Torts § 46, Comment l [1965]). Thus, there is a logical basis for allowing a stranger who is within the zone of danger to recover while denying liability to any other bystander who happens to observe the injury or death of another.

As a final point, a rule that excuses a defendant—like the ones sued by plaintiff here—of the consequences of their negligence is bereft of justice. The defendants failed to remedy the building façade's structural weaknesses and thus risked harm to any pedestrians that walked beneath the building. While the majority holding allows this plaintiff to recover, if she had stood in a slightly different relationship or was a stranger to the deceased child, the defendants would not have to compensate her for a part of the harm caused by their inexcusable negligence. Causing defendants like these to{**36 NY3d at 548} internalize the full cost of the harm that they cause, and making those harmed by them whole, promotes the important societal goal of public safety. Second, it encourages potential defendants to acquire appropriate insurance coverage; this, too, promotes public safety, as these individuals and entities will likely undertake risk-reduction measures to avoid hefty insurance premiums. For those concerned that society (renters in the building, for instance) will pay the price for increased insurance, the costs are unlikely to be greater than for any other type of tort recovery that we already permit, and defendant can contain increased insurance costs by taking the legally prescribed steps to reduce the likelihood that their buildings crumble and kill passersby.

The majority's expansion of the definition of "immediate family," while significant for this plaintiff and others like her, repeats the errors of the past. It has taken over a quarter century since Trombetta for the Court to get this far. "I can only hope that it will not take as long [for] New York" to finally "correct[ ] this error" (Tobin, 24 NY2d at 621 [Keating, J., dissenting]).

Garcia, J. (concurring). What's past is precedent. Here, that means this Court's decision in Trombetta v Conkling (82 NY2d 549, 554 [1993]) and by reversing the Appellate Division, we overrule, at least in part, that decision. While the majority is correct that this Court declined to set the "outer limits of the immediate family" in Bovsun v Sanperi (61 NY2d 219, 233 n 13 [1984] ["Inasmuch as all plaintiffs in these cases were married or related in the first degree of consanguinity to the injured or deceased person, we need not now decide where lie the outer limits of 'the immediate family' "]), we did so in Trombetta. In that case we held that "[s]ound policy and strong precedents justify our confinement and circumscription of the zone of danger rule to only the immediate family as surveyed in Bovsun" (82 NY2d at 553). That "survey"—covering only those married or related in the first degree of consanguinity—did not include the relationship at issue here.

Six of my colleagues agree that we should overrule our precedent and expand the pool of potential plaintiffs. Although I question the wisdom of that conclusion (see Bovsun, 61 NY2d at 234 [Kaye, J., dissenting] [noting that "sound policy {**36 NY3d at 549}considerations supported this court's decisions consistently denying" recovery for "emotional distress from observing physical injury to another" and that the "artificial and arbitrary" limitation imposed would eventually "give way to far-reaching liability affecting the public generally"]), I see no value in dissenting. Accordingly, acknowledging the proper context of what we do today—overturning in part our decision in Trombetta—I concur in the narrow holding that "a grandchild is the 'immediate family' of a grandparent for the purpose of applying the zone of danger rule" (majority op at 526).

Chief Judge DiFiore and Judges Stein and Feinman concur; Judge Rivera concurs in result in an opinion in which Judge Wilson concurs; Judge Garcia concurs in result in a separate concurring opinion.

Order reversed, with costs, plaintiffs' motion to serve and file a second amended complaint granted in its entirety and certified question answered in the negative.