Memorandum.

In Doerr v Goldsmith, the order of the Appellate Division should be reversed, with costs, defendant Smith's motion for summary judgment dismissing the complaint granted and the certified question answered in the negative. In Dobinski v Lockhart, the order of the [*2]Appellate Division should be affirmed, with costs.

 Under the circumstances of these cases and in light of the arguments advanced by the parties, Bard v Jahnke (6 NY3d 592 [2006]) constrains us to reject plaintiffs' negligence causes of action against defendants arising from injuries caused by defendants' dogs (see Bard, 6 NY3d at 596-599; see also Bloomer v Shauger, 21 NY3d 917, 918 [2013]; Smith v Reilly, 17 NY3d 895, 896 [2011], revg 83 AD3d 1492 [4th Dept 2011]; Petrone v Fernandez, 12 NY3d 546, 547-551 [2009]; Collier v Zambito, 1 NY3d 444, 446 [2004]). We decline to overrule our recently reaffirmed precedent (see Bloomer, 21 NY3d at 918; Petrone, 12 NY3d at 547-551). Furthermore, our holding in Hastings v Sauve (21 NY3d 122 [2013]) does not allow plaintiffs to recover based on defendants' purported negligence in the handling of their dogs, which were not domestic farm animals subject to an owner's duty to prevent such animals from wandering unsupervised off the farm (see Hastings, 21 NY3d at 124-126).

 Finally, in Dobinski, the Appellate Division properly granted summary judgment to defendants with respect to plaintiff's strict liability cause of action. Defendants carried their initial burden on summary judgment of establishing that they did not know of any vicious propensities on the part of their dogs. In response, plaintiff failed to demonstrate the existence of a triable issue of fact as to whether defendants had notice of the animals' harmful proclivities, and consequently, defendants were entitled to summary judgment on plaintiff's strict liability claim (see Petrone, 12 NY3d at 550; see generally Jacobsen v New York City Health & Hosps. Corp., 22 NY3d 824, 833 [2014]).

Abdus-Salaam, J. (concurring). In these two cases, we consider whether an individual injured by a domestic animal other than a farm animal may institute a negligence cause of action against the owner of the animal based on the owner's alleged misfeasance in supervising or directing the animal. Like the majority (see majority mem at 1116), I conclude that a negligence cause of action does not lie under our long-standing{**25 NY3d at 1117} precedent, and I reject the specific grounds advanced by plaintiffs here for declining to apply the controlling principles of law set forth in Bard v Jahnke (6 NY3d 592 [2006]). Therefore, I join the majority's memorandum opinion in full. I write separately to provide additional background on these cases, suggest further guidance for future cases, and respond to the particular contentions of the plaintiffs and my dissenting colleagues.

At about 7:00 a.m. on May 31, 2009, defendant Julie Smith and her boyfriend, defendant Daniel Goldsmith, were accompanied by Smith's dog in Manhattan's Central Park. [*3]The couple did not keep the dog on a leash, acting consistently with local regulations that permit a dog to remain off-leash in certain designated areas of the park from 9:00 p.m. to 9:00 a.m. (see Rules of City of NY Dept of Parks and Recreation [56 RCNY] § 1-04 [i] [2]). Smith, Goldsmith and the dog were near a section of the bicycle "loop" road that runs throughout Central Park. Specifically, Smith and Goldsmith were outside the roadway on opposite sides of the road, and Goldsmith was kneeling down and holding the dog in his arms, as if hugging it.

Meanwhile, plaintiff Wolfgang Doerr was riding his bicycle on the loop. As Doerr approached Smith and Goldsmith's location, Smith bent down and clapped her hands on her knees, and she allegedly called the dog over to her. Doerr called out for Smith and Goldsmith to control the dog, but it was too late; as the dog crossed the street, Doerr hit the dog and was thrown from his bike, resulting in significant injuries.

Doerr commenced this personal injury action in Supreme Court by filing a complaint asserting a negligence cause of action against Smith and Goldsmith based on their having negligently "controlled and directed their dog into the path of the plaintiff." Doerr did not set forth any strict liability cause of action or allege that the dog had a vicious propensity. As pertinent to this appeal, Smith answered and demanded discovery, including depositions. After discovery, Smith moved for summary judgment dismissing the complaint, arguing that Doerr could not bring a negligence cause of action based on injuries caused by a domestic pet because such claims are barred by the rule of Bard and{**25 NY3d at 1118} Petrone v Fernandez (12 NY3d 546 [2009]). Doerr opposed the motion on the ground that the Bard rule does not apply where, as here, liability is premised on the conduct of the owner in turning the animal into an instrumentality of harm. Supreme Court denied Smith's motion for summary judgment dismissing the complaint, largely adopting Doerr's position.

A divided panel of the Appellate Division reversed Supreme Court's order, granted Smith's motion for summary judgment dismissing the complaint and directed the clerk to enter judgment accordingly (see Doerr v Goldsmith, 105 AD3d 534, 534-535 [1st Dept 2013]). The majority determined that, because "New York does not recognize a common-law negligence cause of action to recover damages for injuries caused by a domestic animal," Doerr could not sue Smith for her allegedly negligent handling of her dog (id. at 534 [internal quotation marks and citation omitted]). One Justice dissented on the theory that Smith could be held liable for her negligent instruction to the dog, as distinct from the dog's own willful actions (see id. at 535-537 [Mazzarelli, J.P., dissenting]).

After the Appellate Division's decision in Doerr, we handed down our decision in Hastings v Sauve (21 NY3d 122 [2013]), wherein we held "that the rule of Bard v Jahnke does not bar a suit for negligence when a farm animal has been allowed to stray from the property [*4]where it is kept" (Hastings, 21 NY3d at 124 [citation omitted]). Doerr moved for reargument of his appeal in the Appellate Division or, in the alternative, for leave to appeal to this Court.

The Appellate Division issued an order granting the reargument motion, vacated and recalled its prior decision, substituted a new decision, and denied plaintiff's motion for leave to appeal (2013 NY Slip Op 87042[U] [1st Dept 2013]). On the same day it granted the reargument motion, the Appellate Division issued its new decision and order, which, by a split vote, affirmed Supreme Court's order and denied Smith's motion for summary judgment dismissing the complaint (see Doerr v Goldsmith, 110 AD3d 453, 453-455 [1st Dept 2013]). The majority recounted the Bard rule, and it cited favorably the Bard dissent's position that the rule is archaic and leads to unfair results (see id. at 453-454). The majority decided that this Court had recently ameliorated the harshness of the Bard rule in Hastings, which "recognized that an accident caused by an animal's 'aggressive or threatening behavior' is 'fundamentally distinct' from one caused by an animal owner's negligence{**25 NY3d at 1119} in permitting the animal from [sic] wandering off the property where it was kept" (id. at 454, quoting Hastings, 21 NY3d at 125). The majority continued:

"We recognize that the Hastings Court did not decide whether to apply the holding to dogs at that time. However, that should not be an impediment to denying summary judgment in this case. That is because this case is of an entirely different ilk than Hastings, Bard and Petrone. It is not about the particular actions of an animal that led to a person's injury. Rather, it is about the actions of a person that turned an animal into an instrumentality of harm. Here, the dog was in the control of defendants at all times in the split second before the accident occurred. Had Smith not called the dog, and Goldsmith not let it go, plaintiff would have ridden past them without incident.

"[Smith's and Goldsmith's] actions can be likened to those of two people who decide to toss a ball back and forth over a trafficked road without regard to a bicyclist who is about to ride into the ball's path. If the cyclist collided with the ball and was injured, certainly the people tossing the ball would be liable in negligence" (id. at 455).

Two Justices dissented, voting to reverse Supreme Court's order and grant Smith's summary judgment motion (see id. at 455-456 [Andrias, J., dissenting]). The dissent maintained that, in its previous decision in this case, the Court had not overlooked any material fact or principle of law that would warrant vacatur of the prior decision and order (see id. at 455). The dissent distinguished Hastings on the ground this Court had expressly limited its holding in that case to farm animals and left open the question of its application to domestic pets (see id. at 455-456). The dissent maintained that it would be inappropriate to extend Hastings's holding to this case absent further guidance from this Court (see id. at 456).

Smith moved the Appellate Division for leave to appeal to this Court from that Court's order denying her summary judgment motion, and the Appellate Division granted the motion and certified to us the question of whether its order was properly made (2014 NY Slip Op 63107[U] [1st Dept 2014]).{**25 NY3d at 1120}

On the morning of May 20, 2012, plaintiff Cheryl Dobinski and her husband were riding their bicycles on the shoulder of Route 98 in Franklinville. The Dobinskis reached an area of the road that was about 60 feet from the farm of defendants George and Milagros Lockhart. At approximately the same time, Milagros Lockhart released her husband's two German shepherds from her house and onto the outdoor portion of the property. The dogs barked at the Dobinskis and ran into the road. About 10 seconds after the dogs were released, Cheryl Dobinski struck one of them, which caused her to flip over the front of her bicycle and suffer severe injuries.

Cheryl Dobinski initiated this personal injury suit in Supreme Court by filing a complaint which, as amended, asserted causes of action for negligence and for strict liability against the Lockharts. Dobinski claimed that, because the Lockharts had allowed their dogs to leave the farm and collide with her on the road, the Lockharts had been negligent in their failure to adequately supervise and restrain the dogs. Dobinski also argued that the Lockharts were strictly liable for the injuries caused by the dogs, as they had actual or constructive knowledge of the dogs' harmful propensities, including the animals' purported tendency to run onto the road. Dobinski further asserted that, by failing to restrain the dogs on the property, the Lockharts had negligently created and maintained a dangerous condition on their premises and, eventually, on the roadway. The Lockharts joined issue, and discovery commenced.

During discovery, Dobinski and her husband acknowledged that they did not have any prior familiarity with the dogs or their propensities. For their part, the Lockharts maintained that they had never received any complaints about these specific dogs or observed any vicious behavior from them. In fact, a neighbor averred she had never heard of any complaints against those two dogs or seen them interfere with traffic. The Lockharts acknowledged that, on two occasions—one before and one after Dobinski's accident—two of their other dogs had run into the road and been struck by a car. George Lockhart further stated that he sometimes exercised the dogs involved in the accident by having them chase his four-wheeler at a relatively modest pace on his property, but never off the property. Additionally, while it was uncontested that German shepherds are a breed often used as guard dogs, there was no evidence that the Lockharts had used their dogs as guard dogs.{**25 NY3d at 1121}

Following discovery, the Lockharts moved for summary judgment dismissing the amended complaint. They argued that Petrone precluded a negligence action in this case and that [*6]the evidence adduced in discovery yielded no triable issue of fact as to strict liability because they had no actual or constructive notice of any dangerous propensity of the dogs. Dobinski opposed the motion on the ground that triable issues of fact allegedly existed with respect to the Lockharts' negligence and strict liability. In Dobinski's estimation, George Lockhart's use of a four-wheeler to exercise the dogs was tantamount to training the dogs to chase vehicles at high speed. Dobinski viewed the incident in which other dogs owned by the Lockharts had run into the road as evidence of the harmful propensities of the dogs at issue here.

Supreme Court denied the Lockharts' motion for summary judgment dismissing the amended complaint, finding that Dobinski had a viable negligence cause of action under Hastings. The court did not directly address Dobinski's strict liability claim. The Lockharts appealed.

The Appellate Division unanimously reversed Supreme Court's order, granted the Lockharts' motion for summary judgment dismissing the amended complaint and dismissed the amended complaint (see Dobinski v Lockhart, 121 AD3d 1601, 1601-1602 [4th Dept 2014]). Recognizing strict liability as the only viable theory of recovery, the Appellate Division determined that the Lockharts had "met their initial burden [on summary judgment] by establishing that they [had] lacked actual or constructive knowledge that the dog had a propensity to interfere with traffic" (id. at 1602). And, the Court concluded that, "[i]n opposition to the motion, [Dobinski] failed to raise a triable issue of fact in that respect" (id.). We granted Dobinski leave to appeal (24 NY3d 910 [2014]).

"For at least 188 years, the law of this state has been that the owner of a domestic animal who either knows or should have known of that animal's vicious propensities will be held liable for the harm the animal causes as a result of those propensities" (Collier v Zambito, 1 NY3d 444, 446 [2004] [citation omitted]). Under this rule, a "vicious propensity" is the " 'propensity to do any act that might endanger the safety of{**25 NY3d at 1122} the persons and property of others in a given situation,' " including behavior that is dangerous but not necessarily aggressive (id., quoting Dickson v McCoy, 39 NY 400, 403 [1868, op of Grover, J.]).

In Bard v Jahnke (6 NY3d 592 [2006]), we expounded upon the relationship between ordinary negligence principles and the strict liability cause of action permitted by Collier and other prior decisions, concluding that strict liability's traditional status as the predominant theory of recovery for injuries caused by a domestic animal logically forecloses recovery under a theory of negligence. In Bard, a farm worker was left alone in the farm owner's barn to watch the cows, but unbeknownst to the worker, a breeding steer was among the cows; the steer attacked [*7]and injured the worker (see Bard, 6 NY3d at 594-595). The worker sued in strict liability and negligence, and on appeal, we rejected both claims based on the worker's failure to demonstrate that the bull had a known vicious propensity (see id. at 596-599). Specifically, we held that "when harm is caused by a domestic animal, its owner's liability is determined solely by application of the rule articulated in Collier," for were we to allow a plaintiff to recover upon the typically less demanding showing of a prima facie case of negligence, we would "dilute our traditional rule under the guise of a companion common-law cause of action for negligence" (id. at 599).

Three Judges dissented, saying:

"Under the Restatement (Second) of Torts, the owner of a domestic animal who does not know or have reason to know that the animal is more dangerous than others of its class may still be liable for negligently failing to prevent the animal from inflicting an injury. This Court today becomes the first state court of last resort to reject the Restatement rule. I think that is a mistake. It leaves New York with an archaic, rigid rule, contrary to fairness and common sense, that will probably be eroded by ad hoc exceptions" (id. [R.S. Smith, J., dissenting]).

The dissent surveyed the nineteenth century cases creating the strict liability rule and maintained that none of them foreclosed a negligence action, and the dissent opined that Hyland v Cobb (252 NY 325 [1929]) and Dickson left open the possibility of a negligence action by declaring that " 'negligence by an owner, {**25 NY3d at 1123}even without knowledge concerning a domestic animal's evil propensity, may create liability' " (Bard, 6 NY3d at 602, quoting Hyland, 252 NY at 326-327). The dissent conceded that the majority's rule found support in our previous affirmances of Appellate Division decisions establishing strict liability as the sole viable theory of liability arising from animal-related injuries (see id., citing Kennet v Sossnitz, 260 App Div 759 [1st Dept 1940], affd 286 NY 623 [1941], Brown v Willard, 278 App Div 728 [3d Dept 1951], affd 303 NY 727 [1951]). The dissent nonetheless found that those cases should not control the matter in light of the Restatement's sounder rule (see id. at 602). The dissent predicted that we would eventually have to abandon the rule or recognize numerous unworkable exceptions to it in order to avoid harsh results, especially in cases involving exceptionally dangerous animal behavior directed at vulnerable or particularly sympathetic plaintiffs (see id. at 602-603).

We have consistently adhered to Bard in several subsequent cases, concluding that Bard's prohibition against negligence actions arising from harms caused by domestic animals served to foreclose the plaintiffs' negligence claims (see Bloomer v Shauger, 21 NY3d 917, 918 [2013]; Smith v Reilly, 17 NY3d 895, 895-896 [2011]; Petrone, 12 NY3d at 547-550; Bernstein v Penny Whistle Toys, Inc., 10 NY3d 787, 788 [2008]). Among those cases, Petrone and Smith [*8]have particular relevance to plaintiffs' claims here regarding a dog owner's negligence in failing to restrain his or her animal.

In Petrone, the plaintiff mail carrier saw that there was no fence around the yard in front of the defendant's house, and the defendant's Rottweiler was lying unrestrained in the yard (see id. at 547-548). Fearful of the dog, the plaintiff started walking away to her car, and when she turned around, she saw the dog running after her (see id.). The plaintiff ran away and jumped through the window of her car, injuring herself (see id.). The dog left the defendant's property and reached the plaintiff, but it did not harm her (see id.). The plaintiff sued the defendant for negligence and strict liability, asserting, among other things, that the defendant's failure to keep the dog contained on his property, as required by the local leash law, was proof of actionable negligence on his part (see id. at 549).

We rejected the plaintiff's claims as she failed to prove the defendant's prior knowledge of the dog's alleged vicious propensity, saying, "defendant's violation of the local leash law {**25 NY3d at 1124}is irrelevant because such a violation is only some evidence of negligence, and negligence is no longer a basis for imposing liability after Collier and Bard" (id. at 550 [internal quotation marks and citation omitted]). Two Judges joined in a concurring opinion, in which they stated that, while Bard was wrongly decided, the precedential force of that decision bound them to follow it (see id. at 551-552 [Pigott, J., concurring] ["(A)lthough I would not have joined the majority's opinion in Bard, I must, on constraint of that decision, concur in the majority's opinion in the present case"]).

In Smith, we similarly addressed an allegation of negligent supervision of a domestic animal other than a farm animal. There, the principal plaintiff was riding his bicycle when the defendant's dog ran into the road and collided with the plaintiff, propelling him over the handlebars of the bike and injuring him (see Smith v Reilly, 83 AD3d 1492, 1493 [4th Dept 2011]). The plaintiff commenced a personal injury suit against the defendant, pursuing strict liability and negligence theories of recovery (see id.). A divided panel of the Appellate Division denied the defendant's motion for summary judgment dismissing the complaint (see id. at 1493-1494). The majority found that the defendant's deposition testimony about her dog's tendency to "bolt" out the door of her home had created a triable issue of fact as to whether her dog had known vicious propensities (id.). The Appellate Division dissenters maintained that the plaintiff had not raised a triable issue of fact as to vicious propensity and that no other cause of action is available in this context under Bard (see id. at 1494-1495 [Scudder, P.J., and Smith, J., dissenting]). On further appeal, we reversed the Appellate Division's order, insisting that the plaintiff had to demonstrate that the defendant knew of the dog's vicious propensities before the plaintiff could recover for the injuries caused by the defendant's failure to restrain the dog (see [*9]Smith, 17 NY3d at 895-896). Thus, as in Petrone, we concluded that only a strict liability claim, and not a negligence cause of action, will lie based on a pet owner's failure to confine the animal to the owner's property or to restrain the animal from running into another person.

Subsequently, in Hastings, we considered an unusual variant of negligent animal control which implicated an individual's overlapping responsibilities as an owner of farmland and an owner of farm animals. There, the plaintiff was driving her van and hit the defendants' cow, which had wandered off their{**25 NY3d at 1125} farm and onto the public roadway without any apparent obstruction from the dilapidated fence around their property (see Hastings, 21 NY3d at 124). The plaintiff and her husband filed a complaint asserting negligence and strict liability causes of action against the defendants, and the defendants moved for summary judgment on the grounds that the cow's vicious propensity had not been shown and that negligence was not a viable theory of liability under Bard (see id. at 125).

We unanimously decided that the defendants were not entitled to summary judgment and "h[e]ld that the rule of Bard v Jahnke does not bar a suit for negligence when a farm animal has been allowed to stray from the property where it is kept" (id. at 124 [citation omitted]). We took care not to disturb the Bard rule, instead explaining that the Bard line of cases addressed claims of "aggressive or threatening behavior by an[ ] animal" (id. at 125)—in other words, claims that the owners of domestic animals engaged in negligent conduct by failing to adequately suppress the animals' harmful behavior. By contrast, we noted, the case before us involved a claim that was "fundamentally distinct from the claim made in Bard and similar cases: It is that a farm animal was permitted to wander off the property where it was kept through the negligence of the owner of the property and the owner of the animal" (id.). In our view, the Bard rule had no bearing on the issue of a farm owner's negligence in failing to keep his or her farm animals on the property, as applying the Bard rule in such circumstances would improperly "immunize defendants who take little or no care to keep their livestock out of the roadway or off of other people's property" (id.). We

"therefore h[e]ld that a landowner or the owner of an animal may be liable under ordinary tort-law principles when a farm animal—i.e., a domestic animal as that term is defined in Agriculture and Markets Law § 108 (7)—is negligently allowed to stray from the property on which the animal is kept," and hence the plaintiff had a viable negligence claim against the defendants (id. at 125-126).

However, we cautioned that we were "not consider[ing] whether the same rule applies to dogs, cats or other household pets" because "that question [had to] await a different case" (id. at 126).

On the same day we handed down our decision in Hastings, we faithfully applied the Bard rule in another case, rebuffing{**25 NY3d at 1126} the plaintiff's request to overrule Bard. In Bloomer v Shauger, the plaintiff was helping the defendant to control the defendant's distraught [*10]horse by holding the horse's halter (see Bloomer, 21 NY3d at 918). When the defendant was in the process of placing a lead line on the skittish animal, the plaintiff tried to assist him, and the horse jerked its neck back in response to the defendant's attempt to restrain it, causing the halter ring to sever the plaintiff's finger (see id. at 918; see also Bloomer v Shauger, 94 AD3d 1273, 1273-1274 [3d Dept 2012]; brief for plaintiff-appellant in Bloomer, 21 NY3d 917 [2013], available at 2012 WL 8681056, *10-12). The plaintiff sued the defendant on a theory of strict liability and for negligence (see Bloomer, 94 AD3d at 1273-1274). The Appellate Division dismissed the plaintiff's negligence claim on constraint of Bard, and it dismissed the strict liability cause of action based on the plaintiff's failure to show that the horse had a known vicious propensity (see id. at 1274-1277). One Justice dissented, urging that because the Bard rule was "extremely restrictive" and harsh, it should be applied narrowly in a manner that would allow the plaintiff to recover (id. at 1277 [Garry, J., dissenting]).

On appeal to us, the plaintiff contended that the Appellate Division had erred in finding his showing of vicious propensity inadequate (see brief for plaintiff-appellant in Bloomer, 21 NY3d 917 [2013], available at 2012 WL 8681056, *2-3, *13-17). In the alternative, the plaintiff argued that, while the Bard rule would ordinarily limit his ability to recover for the defendant's alleged negligence, that rule should not bar his negligence claim under the facts of his case. He said, "This case is significantly different from the four cases which have come before this Court since Collier" because "in none of the cases that have been before this Court has the owner caused the animal to engage in the injury[-]causing conduct" (id. at *18). Thus, the plaintiff asserted that the defendant was liable insofar as "the [defendant]'s conduct" in placing the lead line on the horse "caused [the horse] to injure [the plaintiff]" (id. at *18-19). To the extent Bard was to the contrary, the plaintiff "respectfully request[ed] that this Court reverse Collier and Bard and adopt the Restatement rule to allow appellant and those injured by domestic animals to recover under theories of common law negligence," asserting that those cases were wrongly decided for the reasons stated by the Bard dissent (id. at *21). The plaintiff reiterated this point in his reply brief (see{**25 NY3d at 1127} reply brief for plaintiff-appellant in Bloomer, 21 NY3d 917 [2013], available at 2013 WL 3818045, *1-3).

We were unmoved by the plaintiff's requests that we overturn Bard or permit his negligence action to proceed on the premise that the defendant's conduct, rather than the horse's instinctive behavior, caused the horse to injure the plaintiff. In a memorandum decision issued on the same day as Hastings, we unanimously affirmed the Appellate Division's order (see Bloomer, 21 NY3d at 918). Specifically, we declared:

"Under the rule of Bard v Jahnke (6 NY3d 592 [2006]), plaintiff cannot recover in the absence of a showing that defendant had knowledge of the animal's 'vicious propensity' or 'propensity to do any act that might endanger the safety of the persons and property of others' (6 NY3d at 596-597, quoting Collier [1 NY3d at 446]). No such showing was made [*11]here. A tendency to shy away when a person reaches for a horse's throat or face is, as the record shows, a trait typical of horses. The Appellate Division correctly held that a vicious propensity cannot consist of 'behavior that is normal or typical for the particular type of animal in question' (Bloomer v Shauger, 94 AD3d 1273, 1275 [2012])" (id. at 918).

Significantly, despite that a farm animal was involved in Bloomer, our decision there was fully consistent with both Bard and Hastings, as Bloomer did not involve Hastings's unique intersection of the defendant's obligations as a landowner and his obligations as the owner of a domestic animal.

Acknowledging that Bard and its progeny would otherwise preclude them from bringing negligence causes of action against defendants in these cases based on the injuries caused by defendants' animals, plaintiffs in the cases at bar propose two potential bases on which they might avoid Bard's prohibition against negligence liability. Plaintiffs' theories are: (1) our holding in Hastings logically extends to cover domestic pets, such that the failure to properly restrain a dog or confine it on the owner's property may create negligence liability; and (2) defendants' affirmative negligent acts in using their dogs as instrumentalities of harm created negligence liability because, unlike in Bard and similar cases involving animals instinctively{**25 NY3d at 1128} harming others without the direct control of their owners, such affirmative acts constitute the negligence of the owners rather than of their pets. In their briefs, plaintiffs do not ask us to overrule Bard, though they complain about some of its purported shortcomings. For the reasons that follow, I reject plaintiffs' arguments with respect to these two proposed exceptions and conclude that the Bard rule controls.

Contrary to plaintiffs' suggestion, our decision to impose liability on the owner of a farm animal in Hastings, based on the owner's failure to keep the animal from wandering off his or her property, does not logically create a negligence cause of action for the failure to restrain a dog on one's premises or in other locations. As we noted in Hastings itself, there is a "fundamental[ ] distinct[ion]" (Hastings, 21 NY3d at 125) between a domestic animal owner's failure to prevent his or her animal from behaving dangerously, which is generally not an actionable form of negligence, and a farm owner's decision to allow his or her farm animals to wander freely onto a public road or someone else's property, which as a commonsense matter violates societal expectations in a manner that gives rise to negligence liability. And, Hastings's treatment of the failure to keep farm animals on the owner's property as distinct from the negligent supervision of other types of domestic animals makes sense. In that regard, domestic [*12]farm animals—horses, cows, goats, sheep, swine, ducks and the like (see Agriculture and Markets Law § 108 [7])—either are difficult to train to remain on one's premises of their own accord (e.g., cows, goats, sheep, fowl), have a proclivity to bolt without confinement (e.g., horses) or have a particularly hazardous combination of a large size and a habit of extremely aggressive behavior related to establishing dominance with respect to humans and other animals (e.g., bulls, pigs). In recognition of the unique peril that arises from allowing farm animals to wander off a farm unsupervised and unconfined, the legislature requires such animals, by statutory definition, to be "raised in confinement" and under state license (id.).

In other words, in New York, society has long recognized that the owners of farm animals have a duty to exercise reasonable care in keeping such exceptionally dangerous or errant creatures on the owners' premises, and in Hastings, we recognized that duty in holding the defendant farm owner liable for his negligent failure to confine his cow on his property. Indeed, it is a matter of common expectation among people in{**25 NY3d at 1129} general that a 1,500-pound cow, a 400-pound pig or an unruly goat will not be permitted to wander freely into traffic or onto a neighbor's yard, mangling people and property alike.

By contrast, the ownership of most domesticated non-farm animals does not naturally necessitate such a duty of care and restraint on the part of their owners because non-farm animals pose different risks than farm animals. Domestic pets are more apt to stay on their owners' premises of their own volition as a result of the domestication and training that have been reinforced over the centuries of humanity's near-familial relationship with its pets. And, few lawfully kept domestic pets reach the levels of size, aggressiveness and desire to wander that many farm animals do.

Furthermore, at least since our decision in Bard, New York society has had no reasonable expectation that all domestic pets will be perpetually confined in their homes or physically restrained at all times. Pet owners often take their pets to public places in close proximity to other people, both on leashes and off of them depending on the location, and the people in those areas have come to expect this phenomenon and to approach such situations with caution. The average New Yorker knows or ought to know that he or she will encounter insufficiently restrained pets, which are not confined to the owner's premises and may harm others depending on the disposition of the pet and the degree of training it has received. In public parks, one regularly encounters dog owners with their unrestrained canine companions, whereas one almost never sees, and certainly never expects to see, someone taking his or her cow for a walk in the neighborhood. As Bard effectively holds, New York society has learned to adequately mitigate the perils of unrestrained pets by a combination of strict liability, statutory regulation and the prudence of those who approach these animals, without adopting a negligence [*13]cause of action based on the failure to supervise or restrain a domestic pet. Given the clear differences in the risks posed by farm animals and pets, there is no reason to extend the rule of Hastings to authorize negligence liability resulting from faulty supervision of a pet.

More fundamentally, plaintiffs' proposed expansion of liability under Hastings would run directly counter to our precedents, including Petrone and Smith. After all, in Petrone we rejected a claim that negligence should lie where a dog owner allowed his dog to stray from his property and chase a{**25 NY3d at 1130} mail carrier (see Petrone, 12 NY3d at 547-550). And in Smith, we concluded that a dog owner's failure to restrain her dog from running into the street and hitting a bicyclist did not give rise to a negligence cause of action (see Smith, 17 NY3d at 895-896). An extension of Hastings would clearly contradict those decisions because, under Hastings, the owner of a farm animal may be held liable for his or her negligence in permitting an animal to wander unrestrained and injure people, whereas in Petrone and Smith, we concluded that the defendants were relieved of negligence liability for engaging in essentially the same behavior with respect to their dogs, which were permitted to roam about and harm others. Rather than extending Hastings to create an inherent contradiction in our case law, I choose to harmonize Petrone, Smith and Hastings by finding that the Hastings rule is limited to cases involving farm animals. Thus, as the majority observes (see majority mem at 1116), the Hastings rule does not render defendants liable in the instant cases, both of which involve claims of negligence in the handling of domestic pets rather than the failure to keep farm animals from wandering off the farm.

Following the reasoning of the Appellate Division majority in Doerr, plaintiffs in the instant cases now proclaim that Bard and its progeny do not preclude their negligence claims because defendants engaged in affirmative acts of control over their pets, whereas in the Bard line of cases, the animals injured people of their own volition after their owners failed to restrain them. But this argument does not withstand close scrutiny. To see the flaws in this direct-control-versus-failure-to-restrain theory, one need only look at the Appellate Division majority's attempt to apply that theory in Doerr.

In Doerr, the Appellate Division majority tried to distinguish Bard and Petrone on the ground that defendant Smith, unlike the defendants in those cases, affirmatively used the dog as an instrumentality of harm by calling it to cross over the road to her; such a case "is not about the particular actions of an animal that led to a person's injury," the Appellate Division majority said, but rather "is about the actions of a person that turned an animal into an instrumentality of harm" (Doerr, 110 AD3d at 455). But, contrary to the Appellate Division's supposition, there is no material difference between the act of negligently failing to restrain an instrument of danger and the act of negligently ordering such an instrument to possibly cause harm because both forms of conduct have the same tendency to{**25 NY3d at 1131} injure others. Regardless of whether a person releases a [*14]dog to instinctively roam near a high-traffic bicycle path or urges the dog to cross the same path, there is some likelihood that the dog will run into a bicyclist and injure him or her. In both scenarios, then, the dog owner is equally capable of preventing harm to the cyclist and equally responsible for that harm, and consequently, the dog owner should face equal liability, or lack thereof, for his or her negligence.

Stated differently, neither the calling of a dog across the road nor the release of a dog near the road can cause injury without the volitional conduct of the dog, thereby making the two actions comparable for purposes of negligence law. Where the owner calls the dog across the road, the dog still will not cross the road and into the path of a bicyclist without instinctively and willfully choosing to do so. If the dog ignores the owner's call, as dogs are sometimes wont to do, no injury occurs. Likewise, if a dog is released from restraint at the side of a road, the dog may instinctively choose to wander into the way of a cyclist, to do nothing, or to wander away from the road, but only its choice, rather than the owner's act of releasing it from its bonds, will result in any injury. In both scenarios, then, the dog's volitional behavior ultimately creates the harm, and the owner's act or omission does not cause the injury.

This is why, in Doerr, the Appellate Division majority's assertion that the actions of Smith and her boyfriend in releasing and calling the dog across the biker's path were akin to those of a couple tossing a ball across a roadway must fall flat (see Doerr, 110 AD3d at 455). When a person tosses a ball, the object has no will of its own, and hence the object's nonexistent volitional behavior cannot cause any injury. The ball can do nothing other than obey the laws of physics, and the ball must move if the human actors carry out their wish for it to do so. While the individuals hurling the ball may not know the exact path it will follow, they are liable for negligence so long as it is reasonably foreseeable that the ball's travel through the air might place it on a collision course with another person. By contrast, dogs may deem their masters' commands considerably less compelling than the forces of acceleration and gravity. A dog owner's call may prompt the dog not to move at all, much less collide with someone, and therefore negligence cannot lie based on the owner's order{**25 NY3d at 1132}. This inability to predict or control how a dog will interpret, react, and respond to its surroundings is why negligence cannot lie based on the owner's order alone. With respect to the ball-tossing scenario, the more apt analogy would be between a person throwing a ball in someone's way and a person tossing a small dog in someone's way, causing the dog to hit the person regardless of its desire to avoid such a fate. Plainly, such a case is not presented in either of these appeals.

Not only does plaintiffs' proposed distinction between an omission of restraint and an act of control defy the practical realities of pet behavior, but it is also inconsistent with basic negligence principles. In that regard, the existence of liability for negligence generally does not [*15]depend on whether the negligent conduct of the tortfeasor is deemed an act or an omission, but rather on whether the individual violates a common-law duty to exercise reasonable care to prevent certain harms (see 79 NY Jur 2d, Negligence § 10; see generally Palsgraf v Long Is. R.R. Co., 248 NY 339, 342, 344 [1928]). Because Bard does not impose a duty on a pet owner to exercise reasonable care in the control of a pet that has no known vicious propensity, the owner's failure to exercise such care, whether by act or omission, does not furnish a basis for liability. It is presumably for this reason that we have never drawn the illusory act/omission distinction now urged by plaintiffs. In fact, as discussed above, in Bloomer, we declined the plaintiff's invitation to draw such a distinction.

One of my colleagues, who dissents in Doerr only, deems the Appellate Division's rationale in that case "not entirely satisfactory" (dissenting op in Doerr at 1142), yet offers a rephrased version of the same unsatisfactory rationale, declaring that "where a plaintiff sustains injury as the direct result of actions that a domestic animal took under the owner's direction and control, a cause of action in negligence should lie" (id.). However, this proposed rule proceeds from the same premise, and hence suffers from the same logical flaws, as the Appellate Division's decision. Again, where an owner calls to his or her animal, such that the animal is under the purported "direction and control" of the owner, the owner's call still merely sets off a "chain of events" (dissenting op in Doerr at 1142) rather than a certain and direct accident.^[FN1] Even in [*16]the face of a direct command, the animal may still decide to ignore the instruction,{**25 NY3d at 1133} follow the instruction perfectly, take the requested action for reasons unrelated to the instruction or carry out the requested action in an unanticipated way. Since there is no way to know why the animal did what it did, there is no principled basis on which to impose negligence liability.

To be sure, in Doerr, Smith seemingly attempted to control her dog by calling to it, and the dog did decide to cross the bike path after Smith made that call. But it does not follow that Smith had unquestioned command of the dog, or that the dog chose to cross the bike path at the moment of Doerr's approach as a direct result of Smith's command. For instance, Smith's dog might have crossed the bike path because it was attracted to the appearance of the grass on the other side. The dog could have completely failed to hear Smith's call over the din of a busy park, or misunderstood the order, and yet it still ran across the path because it was naturally inclined to run toward the area that it saw immediately in front of it. Perhaps the dog initially responded to Smith's call prior to Doerr's arrival on the scene, but then decided to move at a leisurely pace, lingering too long on the path to finish crossing before Doerr arrived.

In any of these scenarios, Smith might have hoped that the dog would choose to follow her command, but she did not know that it would, nor did she know that the dog would fail to dodge the oncoming bike or to otherwise proceed cautiously in response to her command. It is true that these various possibilities are speculative on this record, but so is the dissent's supposition that Doerr's injury was the "direct result" of Smith's "direction and control" of the dog (dissenting op in Doerr at 1142). The most that can be said is that Goldsmith's release of the dog enabled it to proceed unrestrained across the bike path, and that the dog crossed the path after Smith called to it. Ultimately, though, even if the dog decided to cross the path at its typical pace in order to please Smith, that was the dog's choice. As we have repeatedly made clear in the Bard line of{**25 NY3d at 1134} cases, the dog's choice does not result in negligence liability for the owner.

Were this analysis of Doerr not already enough to show that plaintiffs are taking a flawed approach by advancing this act/omission distinction, Dobinski's belated and perplexing attempt to apply this theory to her case confirms that plaintiffs' theory is unworkable. At oral argument before this Court, Dobinski's counsel contended for the first time that Milagros Lockhart's conduct in releasing the dogs onto the exterior portions of her property 10 seconds before the accident constituted an affirmative act on her part, which could support negligence liability. At first glance, this would be a plausible argument if one were to adopt the affirmative act theory advanced by the Appellate Division in Doerr. When a pet owner lets a dog outside to run near the highway, he or she plainly sets an instrumentality of harm in motion just as much as if he or she had called the dog to the highway. As in the case of an owner ordering a dog to come, it is ultimately the dog's decision to enter the roadway that causes harm to passersby, [*17]notwithstanding that the owner has enabled or encouraged the dog's movement. That being so, Dobinski's case stands on the same footing as Doerr's under the act/omission distinction, and if plaintiffs were correct, negligence would lie in both cases. Yet the possibility that plaintiffs' theory might create negligence liability in Dobinski shows that their proposal to distinguish the Bard line of cases in this way would soon lead to the abrogation of our precedent, for in Petrone and Smith, we rejected negligence liability in situations virtually identical to the one in Dobinski. Thus, the adoption of plaintiffs' proposed affirmative act theory threatens to swallow Petrone and Smith whole.

Accordingly, plaintiffs' efforts to extend the Hastings principle to this case and to distinguish the Bard line of cases are unavailing, and they do not assert any other grounds for declining to apply the Bard rule to their cases. Therefore, in my view, the rule of Bard controls these cases and bars plaintiffs' negligence claims arising out of the injuries caused by defendants' domestic pets.

Having disposed of plaintiffs' contentions, there is no need to go further to resolve their negligence claims. Nonetheless, I note that, while refraining from any direct attack on the precedential force of Bard, plaintiffs make some passing criticisms of our decision in Bard at various points in their briefs. My colleagues, who dissent in both of the instant appeals rather than{**25 NY3d at 1135} limiting their disagreement to Doerr, urge that we cast aside Bard (see dissenting op in both cases at 1142), notwithstanding that even plaintiffs do not ask us to go so far.^[FN2] A response is in order.

Plaintiffs and the dissent seem to lament various aspects of Bard. As did the Bard dissenters, the plaintiffs and the dissent imply that Bard is inconsistent with our older case law, such as Dickson, which recognized a cause of action for the negligent failure to restrain horses and cows (see dissenting op in both cases at 1142-1147). Citing the dissent in Bard, [*18]plaintiffs note the alleged harshness of the Bard rule in denying recovery to plaintiffs who suffer animal-related injuries, and like the Bard dissent, plaintiffs and the dissent point out that New York is an outlier in rejecting the sort of negligence action endorsed by the Restatement (Second) of Torts in this context (see dissenting op in both cases at 1147-1149).

However, even if the issue of Bard's continued status as precedent were properly before us—and it is not—I would not be convinced to overturn Bard based on the very arguments that we considered and rejected in that case. In general, we do not cast aside precedent unless it has become unworkable, increasingly irrational and/or increasingly unjust over time (see People v Peque, 22 NY3d 168, 194 [2013]; Policano v Herbert, 7 NY3d 588, 604 [2006]). Here, none of those things has occurred. No party to this litigation has suggested that the courts are incapable of consistently applying the Bard rule. To the contrary, even plaintiffs acknowledge that Bard provides an easy-to-apply bright-line rule that consistently proves fatal to negligence claims arising from injuries caused by certain animals. And, while plaintiffs claim the rule is harsh, they point to no evidence or scholarly study indicating that the supposed harm occasioned by Bard has increased since the issuance of Petrone, Smith and our other decisions upholding Bard. Indeed, the only thing that has changed since we last turned away a challenge to the Bard rule just two years ago (see{**25 NY3d at 1136} Bloomer, 21 NY3d at 918) is the composition of this Court, which is plainly not an appropriate basis on which to set aside precedent (see Peque, 22 NY3d at 194).

To be sure, plaintiffs' and the dissent's complaints about Bard are not baseless, but the same could be said of those criticisms when they were raised by the Bard dissent almost 10 years ago. Then, as now, we had issued decisions potentially permitting negligence actions against the owners of horses loosed from their confines—cases which can be squared with Bard and Hastings to the extent they involved farm animals (see Dickson, 39 NY at 401 [op of Dwight, J.]; cf. id. at 402-403 [op of Grover, J.]; see also Benoit v Troy & Lansingburgh R.R. Co., 154 NY 223, 225-227 [1897]) and which, as the Bard dissent conceded (see Bard, 6 NY3d at 601-602 [R.S. Smith, J., dissenting]), could be read as either implicitly supporting (see Hyland, 252 NY at 326-327) or rejecting (see Vrooman v Lawyer, 13 Johns 339, 339 [1816]; Kennet, 286 NY at 624; Brown, 303 NY at 728) the Restatement's rule. Thus, when Bard was decided, the Court was aware that there was some precedent to support a conclusion different from that which it reached.

Significantly, too, and contrary to the dissent's assertions (see dissenting op in both cases at 1147-1149), the Bard Court's evaluation of policy and precedent remains convincing. As noted, we decided Bard consistently with older decisions such as Dickson, which simply stand for the proposition, later reaffirmed in Hastings, that the owner of a farm animal—a horse in [*19]Dickson—is liable for negligently allowing such an animal to stray onto a public roadway (see Dickson, 39 NY at 401-403). The Bard Court remained faithful to the Dickson line of cases insofar as those cases do not dictate that owners of domestic animals may be sued for negligence based on other forms of misfeasance in the handling of their animals, and pre-Bard precedent suggests that no viable cause of action exists for general negligence in the control or supervision of a domestic animal (see e.g. Vrooman, 13 Johns at 339). Our post-Bard decisions applying the Bard rule, such as Petrone and Smith, flow logically from firmly rooted decisional law holding that "domestic favorites such as the family dog or cat, as emblematic of a suburban community as a cow or horse is to the standard farm, are, as a norm, frequently allowed to romp unguarded or unattended" because "[a]s a general proposition '[a] dog, unless vicious, has a right in the highway, and presumably, absent evidence of negligence, the dog's owner cannot be charged with li{**25 NY3d at 1137}ability for injury caused [merely] by its presence therein' " (Young v Wyman, 159 AD2d 792, 793-794 [3d Dept 1990], quoting 3 NY Jur 2d, Animals § 48 at 625-626, affd 76 NY2d 1009 [1990]; Kennet, 286 NY at 624 [affirming decision below that a dog owner could not be liable for failing to prevent his dog from running off his premises and knocking down the plaintiff on the sidewalk]). Even the Restatement rule, which we rejected in Bard, does not treat a domestic pet's untrammeled wanderings as actionable negligence (see Restatement [Second] of Torts § 518, Comment j). Consequently, the Bard dissent's suggestion that the Court was ignoring its past decisions rings as hollow now as it did then.

My colleague dissenting only in Doerr points out that, in some unusual cases, the facts of the animal-related injury-causing occurrence will not fit comfortably within the confines of the vicious propensity doctrine announced in the aforementioned precedent (see dissenting op in Doerr at 1140-1142).^[FN3] As we have recognized, however, "delineation of limits of [*20]liability in tort actions is usually determined on the basis of considerations of public policy" (Bovsun v Sanperi, 61 NY2d 219, 228 [1984]). When Bard was decided, we concluded that the benefits of a bright-line rule limiting recovery to a theory of strict liability outweighed the concerns of those few cases for which application of the rule may seem unsatisfactory. That determination remains reasonable today because there are legitimate policy reasons to retain the Bard rule, including its ability to keep liability within manageable limits, its bright-line guidance and its consistency with societal expectations. At its heart, the Bard principle is a commonsense rule of notice. Absent awareness of a domestic animal's previously demonstrated tendency to harm others, the owner should not bear the costs of the animal's instinctive decisions, notwithstanding that animals by their nature tend to pose some risk of harm, especially when poorly directed or left unrestrained. Notice rules such as this{**25 NY3d at 1138} are hardly oppressive or alien to our law, and they permit our citizens to manage their affairs based on the known risks of daily life.

Abandoning or eroding Bard's bright-line rule would harm pet owners and alter societal expectations. Pet owners and their insurers are currently entitled to rely on the Bard rule to plan their future conduct and their insurance needs, and changing the rule now would risk unfairly disrupting their expectations. Additionally, if Bard were overruled and negligence suits were permitted to proceed against pet owners, a violation of a local leash law may be proof of a pet owner's negligent failure to control his or her pet, and thus negligence suits might create a de facto private cause of action under local leash laws, effectively ignoring the will of any local legislature that has decided not to provide for such actions. While I acknowledge that out-of-state courts that have addressed the issue have nearly uniformly recognized a negligence cause of action arising from the handling of all domestic animals, the same was true at the time Bard was decided (see Bard, 6 NY3d at 600 [R.S. Smith, J., dissenting]), and that is not a sufficient reason to overrule our long-standing and recently reaffirmed precedent.

At bottom, regardless of our individual positions on the Bard rule, we should not revisit our adoption and consistent retention of the rule today, as neither plaintiffs' jabs at Bard nor the dissent's concerns are sufficiently weighty to overcome critical considerations of stare decisis. A state's highest court is, first and foremost, charged with creating a coherent body of settled law by which members of society may order their affairs. This mission inevitably reflects the policy choices of predecessor judges that decide an issue and thereby create a precedent, and is inevitably undermined if successor judges succumb to the very human impulse to cast aside or chip away at those rulings with which they simply disagree. Indeed, plaintiffs' veiled criticisms of Bard call to mind a variation of an observation made by the author of the Bard dissent in a [*21]different context: "Essentially this argument [for rejecting the current rule] has been ably made by three dissenting Judges [and two concurring Judges] in two of our prior cases" in Bard and Petrone, "[b]ut I respectfully suggest that, at this late date, the question should be considered settled" (People v Giles, 24 NY3d 1066, 1073 [2014, Smith, J., concurring]). Under our precedent, plaintiffs' negligence claims must fail because the particular{**25 NY3d at 1139} exceptions to the Bard rule proposed by plaintiffs are incompatible with Bard and its progeny. Having been presented with no alternative theory of recovery, I neither reject nor endorse any other potential legal theory or exception to the Bard rule not advanced by the parties in these cases.^[FN4]

In Dobinski, plaintiff Dobinski pleaded a strict liability cause of action in addition to her negligence claim. However, as the majority rightly concludes (see majority mem at 1116), Dobinski cannot proceed on her strict liability claim on this record because she did not sufficiently demonstrate that the Lockharts were aware that their dogs had a propensity to run into the road and pursue bicyclists. In their motion papers and depositions, the Lockharts stated, without contradiction, that they had never known their dogs to run into the road or approach bicyclists, and their neighbor confirmed that the dogs had no known tendency to interfere with traffic. Thus, the Lockharts carried their initial burden on summary judgment of showing that they did not know of any vicious propensities of their dogs.

In response, Dobinski failed to create a triable issue of fact by showing that the Lockharts had notice of the dogs' proclivity to harm others. Although Dobinski argues that the Lockharts trained the dogs to run after vehicles at high speed, the record shows that George Lockhart merely prompted the dogs to follow his four-wheeler from time to time at a slow pace for exercise purposes, and there is no evidence that the Lockharts trained their dogs to generally chase vehicles, including bicycles, outside their farmland or to chase vehicles at a high rate of speed. Nor does it matter that, in separate incidents before and after Dobinski's accident, the Lockharts' other dogs had run into the road, as those dogs' propensities cannot demonstrate that the entirely different dogs at issue here had a tendency to harm others. Likewise, even if one were to assume that the Lockharts violated the local leash law, such a violation would only be [*22]some proof of negligence and could not establish strict liability (see Petrone, 12 NY3d at 550). In the face of the Lockharts' deposition testimony and allegations, Dobinski raised no triable factual dispute about the Lockharts' knowl{**25 NY3d at 1140}edge of their dogs' propensities, and the Appellate Division properly granted the Lockharts summary judgment on that issue (see Smith, 17 NY3d at 896; see generally Jacobsen v New York City Health & Hosps. Corp., 22 NY3d 824, 833 [2014]). I further reject Dobinski's claim that the Lockharts were liable for their alleged negligence in creating a dangerous condition on their premises and on the roadway by mishandling their dogs.

Based on the foregoing observations, in Doerr, I vote to reverse the order of the Appellate Division, grant defendant Smith's motion for summary judgment dismissing the complaint and answer the certified question in the negative. In Dobinski, I vote to affirm the order of the Appellate Division.

Chief Judge Lippman (dissenting in Doerr v Goldsmith and concurring in Dobinski v Lockhart). In Hastings v Sauve (21 NY3d 122 [2013]), we held that the "vicious propensities" rule the Court had announced in Bard v Jahnke (6 NY3d 592 [2006]) was inapplicable to the situation created when a large farm animal was permitted to wander into a public roadway. I would conclude that the vicious propensities rule is similarly inapplicable to this case and that a cause of action can be maintained against defendant dog owner for her negligence.

As we observed in Hastings, Bard had adhered to the traditional vicious propensities rule, which requires a plaintiff to demonstrate that an owner had knowledge that the animal's behavior "reflect[ed] a proclivity to act in a way that puts others at risk of harm" (21 NY3d at 125 [internal quotation marks and citations omitted]). We further noted that our "vicious propensities" line of cases following Bard involved situations where the plaintiff had been bitten or somehow menaced by a dog (see Petrone v Fernandez, 12 NY3d 546 [2009][*23][plaintiff chased by dog]; Bernstein v Penny Whistle Toys, Inc., 10 NY3d 787 [2008] [plaintiff bitten by dog]; see also Smith v Reilly, 17 NY3d 895 [2011] [dog ran into the road and collided with plaintiff's bicycle]).

To the contrary, the fact pattern in Hastings did "not involve aggressive or threatening behavior by any animal" and we viewed its claim—that both the owner of the cow and the owner of the property were negligent in allowing it to wander away—as "fundamentally distinct" from the claims made in the Bard line of cases (see Hastings, 21 NY3d at 125). We held that{**25 NY3d at 1141} to apply the Bard vicious propensities rule "in a case like this would be to immunize defendants who take little or no care to keep their livestock out of the roadway or off of other people's property" (Hastings, 21 NY3d at 125).

Although the situation presented in Doerr is distinguishable from Hastings, in fundamental respects they are the same. The rule of vicious propensities should have no application here, where we are not faced with any aggressive or menacing animal behavior—quite the opposite, since defendant's dog allegedly did exactly as she was told (see e.g. Bloomer v Shauger, 21 NY3d 917, 918 [2013] ["a vicious propensity cannot consist of 'behavior that is normal or typical for the particular type of animal in question' "]). Moreover, application of the rule in this instance would serve only to immunize defendant from the consequences of her own negligent actions, for no reason other than that a dog happened to be involved in the accident.

I would continue to adhere to the vicious propensities rule where it appropriately applies, which would appear to be the great majority of cases involving injuries caused by domestic animals. Indeed, as noted above, we have reaffirmed that holding several times in the recent past. The rule reflects a policy decision that a pet owner is not required to anticipate and take steps to prevent aberrational, dangerous behavior from an apparently benign animal. For example, an owner will not be liable the first time a rambunctious dog welcoming a guest knocks him down the steps. Thus, in Dobinski, the absence of evidence that the defendants were aware their dogs had a penchant for running into the road dooms the plaintiff's case.

The situation presented by Doerr, however, is distinct. Defendant did in fact control her dog, but allegedly was negligent in directing it into the path of the oncoming bicycle. The concurring opinion emphasizes a dog's "volitional behavior" (concurring op at 1131) and I think we can all agree that a dog is not analogous to an inanimate object, as it has agency of its own. But people expend significant amounts of time and effort, and sometimes go to great expense, in an effort to train their dogs to be obedient. When those efforts are successful and the dog acts according to the owner's command, that is not a vicious propensity, but should not necessarily result in the owner's immunity from liability. As we recognized in Hastings, the vicious propensity rule does not cover every situation. By contrast to the above example of [*24]nonactionable canine greeting, were the owner to throw a ball towards the steps as the guest ascends{**25 NY3d at 1142} and the dog's inevitable chase propels the guest downward, it is clear to me that the owner's far greater culpability in the latter disaster should compel the availability of a recovery in negligence. I would place Doerr squarely in that category as well.

The rationale proffered by the Appellate Division below—that defendant should be held liable because her actions turned the dog into an instrumentality of harm—is not entirely satisfactory, as it could potentially encompass a much broader range of owner conduct. For instance, the owner in Dobinski could be seen as having launched an instrumentality of harm by letting her dogs run outside unleashed, thereby creating the opportunity for them to enter the road and injure the plaintiff. But there is a difference between setting in motion a chain of events, in the course of which an animal acts in a way that eventually injures someone, and directing the animal to engage in conduct that causes direct and immediate harm. In the first situation, the vicious propensities rule applies to the animal's dangerous behavior. In the second, the fault should be attributed to the owner.

I would hold that, where a plaintiff sustains injury as the direct result of actions that a domestic animal took under the owner's direction and control, a cause of action in negligence should lie. An exception to the vicious propensities rule is warranted under these circumstances. Indeed, a contrary rule automatically immunizing animal owners from the consequences of their own directions is too broad brush, as this situation is plainly distinguishable from our existing vicious propensities jurisprudence and contradicts any sensible logic.