Payton v. Abbott Labs

437 N.E.2d 171, 386 Mass. 540

• Court: Massachusetts Supreme Judicial Court
• Decided: June 22, 1982
• Status: Published

Opinions

Lynch, J.

This case comes before the court on certification from the United States District Court for the District of Massachusetts of four principal and several subsidiary questions involving Massachusetts tort law. See S.J.C. Rule 1:03, § 1, as amended, 382 Mass. 700 (1981).

The plaintiffs in the civil action in which these questions are certified seek redress for injuries allegedly caused by the prescription drug diethylstilbestrol (DES). They brought suit in the Federal District Court in April, 1976. In July, 1979, a judge of the Federal District Court conditionally certified the plaintiff class under Fed. R. Civ. P. 23 (c)(4)(A) to permit resolution of thirteen specific class-wide issues. Payton v. Abbott Labs, 83 F.R.D. 382, 386 (D. Mass. 1979). Several of these issues are factual; the four questions here certified involve issues of Massachusetts law which the judge believed “may be determinative of various aspects of this case and as to which it appears . . . that there is no controlling precedent” in the decisions of this court.

The questions are presented by the judge in the context of a motion to dismiss, based upon the plaintiffs’ allegations. The judge summarized those allegations as follows:

“The plaintiffs are all females whose mothers ingested a drug called diethylstilbestrol . . . while pregnant with the plaintiffs. DES was marketed by the defendants as a preventative for miscarriages, and was widely prescribed by physicians. DES is transmitted to the fetus, and has been identified as one cause of a relatively rare but extremely malignant cancer called clear-cell adenocarcinoma which attacks the reproductive organs of the female children of mothers who have ingested DES. DES has also been identified as one of the causes of more common benign changes in the female children’s reproductive organs, one of which is known as adenosis. The only corrective for clear-cell ade-nocarcinoma is timely radical surgery or radiation; if treatment is not successful the disease is likely to be fatal.

“Many of the plaintiffs have no symptoms of any of these conditions. As a result of their mothers’ ingestion of DES, these plaintiffs are statistically more likely to suffer one of several abnormalities of the reproductive organ than is the general population and are to a lesser degree more likely to contract clear-cell adenocarcinoma.

[*] They are anxious and emotionally upset by these possibilities. Some of the plaintiffs, on the advice of their physicians, are submitting to periodic medical examinations of the cervix and vagina so that symptoms of adenosis, other abnormalities, or clear-cell adenocarcinoma will be detected as early as possible, thus increasing the chance of successful treatment. These examinations may be expensive and traumatic.

“Among the various theories of liability asserted by the plaintiffs is that the defendants were negligent in marketing DES as a miscarriage preventative without adequate testing and without appropriate warnings.

“DES was manufactured according to substantially the same formula by all of the defendants and was marketed by some defendants under its generic description. The products of the various defendants were interchangeable and were sold interchangeably by pharmacists, a fact which the defendants knew or should have known. Most of the plaintiffs are unable to identify the specific manufacturer of the DES ingested by their mothers. Pharmaceutical companies other than the named defendants also manufactured and marketed DES under its generic description, and may have supplied some of the DES ingested by the plaintiffs’ mothers. In most cases, neither the plaintiffs nor the defendants will be able to identify whose product was ingested by which mothers. Furthermore, different defendants marketed DES at different times and under different circumstances which may bear on their negligence.

“Plaintiffs have alleged two bases for recovery against all the defendants by all the plaintiffs: (1) conspiracy or joint enterprise, and (2) ‘alternative liability.’ ‘Alternative liability’ is a term used to describe a theory of recovery adopted in some jurisdictions where there is not a joint tort, that is, where the tort was committed by only one of several possible tortfeasors, but there is no way to determine which one.[**] Different jurisdictions have developed different versions of this theory of liability with respect to the allocation of damages, the necessity of joining all possible tort-feasors as defendants, shifting the burden of proof to a defendant to establish the impossibility of that defendant’s responsibility, and right of contribution among defendants and against possible tort-feasors who are not parties. As far as the [judge] can determine, the [Supreme Judicial] Court has not addressed a claim of this kind in any form. While such a theory has never been recognized, it has never been rejected by a Massachusetts court.

“While the plaintiffs claim that DES was not efficacious in preventing miscarriages, the defendants claim that it was. If the evidence supports the defendants’ claim the trier of fact might be warranted in concluding that a particular plaintiff probably would not have been born had it not been for her mother’s ingestion of DES.

“The parties and the [judge] agree that in this diversity action the law of Massachusetts is the controlling law.”

If the plaintiffs prevail, and the defendants’ liability is established, there must be individual trials for members of the plaintiff class on the issue of damages and perhaps other issues as well.

Question One

“Does Massachusetts recognize a right of action for emotional distress and anxiety caused by the negligence of a defendant, in the absence of any evidence of physical harm, where such emotional stress and anxiety are the result of an increased statistical likelihood [that] the plaintiff will suffer serious disease in the future?” We answer, No.

Discussion

We note initially that neither the issue of negligence nor that of causation is before us. The certified question assumes both that the defendants were negligent, and that their negligence caused the plaintiffs’ emotional distress.

No Massachusetts case has yet concluded that a plaintiff who alleges that she was a direct victim of a defendant’s negligent conduct, but who does not allege that she has suffered resulting physical harm, can recover for emotional distress. In the absence of a specific factual context, the court has declined to decide this issue.³ This court has held that a parent who was not within the zone of physical danger created by the defendant’s negligence may recover for substantial physical injuries sustained as a result of emotional distress, resulting from injuries negligently inflicted on her child. Dziokonski v. Babineau, 375 Mass. 555, 568 (1978). Here, however, the plaintiffs are assumed to be direct victims without physical injuries, since the certified question assumes that the plaintiffs were threatened with physical harm by the defendants’ negligent conduct. We must decide whether this distinction should affect the result in this case.

Physical Harm Requirement

Both common law and policy considerations lead us to answer certified question one in the negative. There is ample support in the common law of this country for a negative answer. Jurisdictions allowing recovery for emotional distress without proof of physical harm⁴ in negligence cases are clearly in the minority.⁵ The most common justification for denying recovery for emotional distress in negligence cases absent physical harm is that that rule is necessary to prevent fraud and vexatious lawsuits. Those seeking to apply a more liberal rule argue that a jury are capable of distinguishing real from feigned injuries and that, therefore, the matter should be left to the jury. This response has appeal because in general it is for the jury, and not for an appellate court, to determine which injuries are real and which are contrived.

The task of determining whether a plaintiff has suffered purely emotional distress, however, does not fall conveniently into the traditional categories separating the responsibilities of the judge from those of the jury. A plaintiff may be genuinely, though wrongly, convinced that a defendant’s negligence has caused her to suffer emotional distress. If such a plaintiff’s testimony is believed, and there is no requirement of objective corroboration of the emotional distress alleged, a defendant would be held liable unjustifiably. It is in recognition of the tricks that the human mind can play upon itself, as much as of the deception that people are capable of perpetrating upon one another, that we continue to rely upon traditional indicia of harm to provide objective evidence that a plaintiff actually has suffered emotional distress.

Although this court has allowed recovery for emotional distress absent physical harm, it has done so only where the defendant’s conduct was extreme and outrageous, and was either intentional or reckless. See Simon v. Solomon, 385 Mass. 91, 95 (1982); Agis v. Howard Johnson Co., 371 Mass. 140 (1976); George v. Jordan Marsh Co., 359 Mass. 244 (1971). The outrageous and reckless or intentional nature of a defendant’s conduct permits a jury to infer that the plaintiff suffered genuine emotional distress. In the absence of these indicia of genuineness we require that plaintiffs alleging negligent infliction of emotional distress substantiate their claims by demonstrating, in order to recover, that they suffered physical harm.

It should be noted, in addition, that the retributive function of imposing tort liability is served by allowing recovery for emotional distress, without proof of physical harm, where a defendant’s conduct was either intentional or reckless. Where a defendant was only negligent, his fault is not so great as to require him to compensate the plaintiff for a purely mental disturbance.

1. Historical development. Since this area of tort law is in a state of change, a brief review of the history of recovery for infliction of emotional distress is appropriate.

a. Parasitic recovery for emotional distress. Where plaintiffs have suffered directly inflicted personal injuries as a result of a defendant’s negligence, courts have not been reluctant to allow recovery for emotional distress, occurring contemporaneously with those personal injuries, as an additional element of damages. Barney v. Magenis, 241 Mass. 268, 273 (1922). Driscoll v. Gaffey, 207 Mass. 102, 107 (1910). Homans v. Boston Elevated Ry., 180 Mass. 456, 458 (1902). In these cases, recovery for emotional distress was allowed as a claim “parasitic” to the “host” claim of damages for negligently inflicted physical injuries. See Dziokonski v. Babineau, supra at 559.

b. The impact rule. Massachusetts and other industrial States for many years limited recovery for negligent infliction of emotional distress to cases where plaintiffs could prove that a defendant’s negligence had caused a physical impact of some kind to the plaintiff’s person. The leading Massachusetts case, frequently cited by courts of other jurisdictions, was Spade v. Lynn & Boston R.R., 168 Mass. 285 (1897). The Spade court conceded that “ [a] physical injury may be directly traceable to fright, and so may be caused by it. We cannot say, therefore, that such consequences may not flow proximately from unintentional negligence, and, if compensation in damages may be recovered for a physical injury so caused, it is hard on principle to say why there should not also be a recovery for mere mental suffering when not accompanied by any perceptible [ directly-caused] physical effects.” Id. at 288. The court, while admitting that the rule rested mainly upon considerations of ease of administration, held that there could be no recovery for emotional distress “where there is no injury to the person from without.” Id. at 290.

The arbitrariness of the impact rule led courts to stretch. the boundaries of the term “impact” in order to allow recovery. See, e.g., Conley v. United Drug Co., 218 Mass. 238 (1914) (plaintiff bruised after falling to floor in faint following explosion); Steverman v. Boston Elevated Ry., 205 Mass. 508 (1910) (fire set plaintiff’s clothing to smoldering). The rule, assailed by commentators since its inception, has been abandoned by all but a small minority of jurisdictions.⁶ See W. Prosser, Torts 332 (4th ed. 1971).

c. Intentional infliction of emotional distress. English courts began to allow recovery for intentional infliction of emotional distress, without resulting physical harm, about the turn of the century. See, e.g., Wilkinson v. Downtown, [1897] 2 Q.B. 57. The commentators almost universally favored recognition of such claims. Magruder, Mental and Emotional Disturbance in the Law of Torts, 49 Harv. L. Rev. 1033, 1058 (1936). Prosser, Intentional Infliction of Mental Suffering: A New Tort, 37 Mich. L. Rev. 874 (1939). In 1952, an American court first allowed recovery for emotional distress absent physical harm. State Rubbish Collectors Ass’n v. Siliznoff, 38 Cal. 2d 330 (1952). In 1965, the Restatement (Second) of Torts acknowledged the existence of the tort of “outrageous conduct causing severe emotional distress,” Restatement (Second) of Torts § 46 (1965) (emphasis supplied), adding that “[t]he law intervenes only where the distress is so severe that no reasonable man could be expected to endure it.” Id. at Comment j.

An extensive intrusion was made upon the impact rule in Massachusetts when this court adopted (in essence) § 46 of the Restatement (Second) of Torts in 1971. George v. Jordan Marsh Co., 359 Mass. 244, 255 (1971). The George decision, however, allowed recovery only for intentionally inflicted severe emotional distress “with bodily harm resulting from such distress.” Id. The court expressly reserved the question whether recovery would be allowed on “allegations of distress without resulting bodily injury.”⁷ Id. Five years later, the court answered the question reserved in George. In Agis v. Howard Johnson Co., 371 Mass. 140 (1976), the court overruled the allowance of a motion to dismiss a complaint in which the plaintiff alleged that she had suffered severe emotional distress, intentionally inflicted, resulting in no more than weeping, “mental anguish, and loss of earnings and wages.” Id. at 141. The plaintiff’s employer dismissed her after announcing his intention to fire employees in alphabetical order until theft losses ceased.

d. Negligent infliction of emotional distress: the “zone of danger” rule. By 1965, when the American Law Institute published the Restatement (Second) of Torts § 436, the “strong majority” of American courts had ameliorated the harshness of the impact rule and permitted plaintiffs to recover for emotional distress, negligently inflicted, if those plaintiffs were within the “zone of danger” created by the defendant’s negligent conduct. Restatement (Second) of Torts §§ 313, 436, 436A (1965), and Reporter’s Notes. Courts adhering to this rule do not require plaintiffs to allege and prove that defendants have inflicted traumatic physical injuries directly upon them. See Fournell v. Usher Pest Control Co., 208 Neb. 684, 686-687 (1981).

e. Negligent infliction of emotional distress: bystander recovery. Perhaps the most persuasive cases for abolition of the limitation on recovery for emotional distress imposed by the “zone of danger” rule involved parents who suffered severe emotional distress, usually manifested by serious physical symptoms, as a result of witnessing their children being struck by defendants’ vehicles. The first Restatement of Torts contained a caveat on the issue whether such plaintiffs should be permitted to recover. Restatement of Torts § 313, at 851 (1934). The Restatement (Second), however, allowed a plaintiff within the zone of physical danger created by the defendant’s negligent act to recover for “shock or fright at harm or peril to a member of [the plaintiff’s] immediate family occurring in his presence.” Restatement (Second) of Torts § 436 (3) (1965). See id. at § 313 (2) (denying recovery to such plaintiffs who were not within the zone of danger) . The drafters of the Restatement (Second) expressed no opinion whether recovery should be allowed where the harmed or imperiled person was not an immediate family member, or was not harmed or imperiled while in the presence of the imperiled person. Restatement (Second) of Torts, caveat to § 436 (1965).

The California Supreme Court soon went beyond the boundaries set forth in the Restatement and allowed a mother to recover who, while not within the zone of danger, witnessed her minor daughter’s death in a motor vehicle accident caused by the defendant’s negligent conduct. The plaintiff suffered a shock to her nervous system resulting in mental pain and suffering. Dillon v. Legg, 68 Cal. 2d 728 (1968). Other courts similarly have allowed bystander plaintiffs to recover who suffered substantial physical injuries but were not within the zone of danger. See, for example, Barnhill v. Davis, 300 N.W.2d 104 (Iowa 1981); Corso v. Merrill, 119 N.H. 647 (1979); Tortee v. Jaffee, 84 N.J. 88 (1980); Sinn v. Burd, 486 Pa. 146 (1979); D’Ambra v. United States, 114 R.I. 643 (1975). The Court of Appeals of New York, on the other hand, determined that for public policy reasons no recovery should be allowed to bystanders for harm suffered as a result of witnessing harm or peril to another. Tobin v. Grossman, 24 N.Y.2d 609, 619 (1969) (“The problem for the law is to limit the legal consequences of wrongs to a controllable degree. ... It is enough that the law establishes liability in favor of those directly or intentionally harmed”).

This court considered the issue in 1978 and held that, under Massachusetts law, recovery by a bystander plaintiff would be allowed, regardless whether the plaintiff was within the “zone of danger,” on proof of the usual elements of a negligence claim (with emphasis on foreseeability) and proof that the plaintiff suffered a “substantial physical injury.” Dziokonski v. Babineau, supra at 568. Accord, Fer-riter v. Daniel O’Connell’s Sons, 381 Mass. 507, 517-518 (1980). One of the plaintiffs in Dziokonski died, allegedly as a result of the shock she suffered upon learning of the accident which injured her child: there could be no question that the plaintiff’s injury was “substantial.”

f. Negligent infliction of emotional distress: recovery without proof of physical harm. The Restatement (Second) of Torts § 436A (1965) sets forth what is still the rule adhered to by the majority of American courts: “If the actor’s conduct is negligent as creating an unreasonable risk of causing either bodily harm or emotional disturbance to another, and it results in such emotional disturbance alone, without bodily harm or other compensable damage, the actor is not liable for such emotional disturbance.” The cases generally hold that physical harm is required but (in accordance with the Restatement [Second] of Torts § 436[2] [1965]), the harm need not be caused by impact or trauma; physical harm resulting from emotional stress is sufficient.

Although there is still general agreement that recovery for emotional distress, negligently caused, will not be allowed absent proof of physical harm (see W. Prosser, Torts, supra at § 54), a minority of courts, in recent decisions, have concluded otherwise. See Taylor v. Baptist Medical Center, Inc., 400 So.2d 369 (Ala. 1981); Molien v. Kaiser Foundation Hosps., 27 Cal. 3d 916 (1980); Rodrigues v. State, 52 Haw. 156 (1970); Sinn v. Burd, 486 Pa. 146 (1979) (bystander case). See also Montinieri v. Southern New England Tel. Co., 175 Conn. 337, 345 (1978) (defendant liable if it “should have realized that its conduct involved an unreasonable risk of causing emotional distress and that that distress, if it were caused, might result in illness or bodily harm” regardless whether illness or bodily harm did result).

2. Analysis. The reasons for the majority rule are, at least, three. One is that emotional disturbance which is not so severe or serious as to have physical consequences is likely to be “so temporary, so evanescent and so relatively harmless” that the task of compensating for it would unduly burden defendants and the courts. The second is that, in the absence of the guarantee of genuineness provided by resulting bodily harm, such emotional disturbance can be too easily feigned or imagined. The third is that where the defendant’s conduct has been merely negligent, without any element of intent to harm, his fault is not so great that he should be required to make good a purely mental disturbance. Restatement (Second) of Torts, supra at § 436A, Comment b. We find merit in this reasoning.

Courts have been troubled for many years by the problem of how to deal with claims for damages for emotional distress. They have recognized that emotional distress can be both real and serious in some situations, while trivial, evanescent, feigned, or imagined in others. The various methods of treating such claims (summarized in the preceding section) have been attempts by the judicial system to formulate a means of separating the former from the latter. The underlying policy in most jurisdictions seems to be that of compensating plaintiffs with clearly recognizable serious injuries, while not burdening either the judicial system or individual defendants with the latter type of claim.

Instead of formulating a rule by means of which claims for emotional distress damages could be differentiated, courts could have rejected all such claims, or left to the fact finder the task of eliminating insubstantial claims. The first resolution would be unfair to plaintiffs who have suffered serious injuries, while the latter would burden defendants and the courts with the task of dealing with a myriad of claims that ought never to have been brought. Courts have sought to find a middle ground between these options, and we believe that still remains the advisable course.

The impact rule was the first attempt at making the desired distinction. As we have pointed out, this rule was crit-icised and modified almost from its inception. Where the impact rule has been abandoned, leading writers have indicated that in lieu of impact there must be, of necessity, some requirement of satisfactory proof and, further, that there should be no recovery where a normal individual would not have been affected under the circumstances. W. Prosser, Torts, supra at 332.

The parasitic recovery cases represent an attempt to define the required element of satisfactory proof as a contemporaneously inflicted physical injury. The logic of this approach is difficult to defend where the personal injury and the emotional distress alleged are not causally related. More recently, courts have found that plaintiffs met the requirement of satisfactory proof when, although they did not suffer direct physical injuries, they were within the “zone of danger” created by the defendant’s negligent conduct. See, e.g., Colla v. Mandella, 1 Wis. 2d 594 (1957). When a plaintiff has been subjected to the risk of serious bodily harm from an automobile or other object directed toward his person by the negligent conduct of a defendant, emotional damage may be expected to result, and the requirement of some additional element of satisfactory proof has been met.

The bystander cases are an extension of the zone of danger principle: recovery has been allowed, regardless whether the plaintiff was in the zone of danger, when emotional injuries resulted from the shock suffered on learning of serious physical injuries to a close member of the plaintiff’s family. Dillon v. Legg, 68 Cal. 2d 728 (1968). This court has permitted recovery in bystander situations, however, only when the bystander has suffered substantial physical injuries as well as emotional distress. Dziokonski v. Babineau, supra at 568. Again, the requirement of an additional element of satisfactory proof is satisfied by proof of serious physical injury to a close family member, and resulting serious physical injury to the plaintiff. See W. Prosser, Torts, supra at 330.

That additional element also can be found when the emotional distress has been intentionally or recklessly inflicted. Agisv. Howard Johnson Co., 371 Mass. 140, 143-145 (1976). In Agis, the court recognized that administrative difficulties do not justify denial of relief when serious invasions of mental and emotional tranquility have occurred, and concluded that greater proof of the plaintiff’s mental suffering is found in the defendant’s conduct designed to bring it about than in physical injury which may or may not have resulted from that conduct. The court was unwilling to hold that no cause of action for emotional distress without physical injuries exists merely because of difficulties of proof, but it opened the door to recovery cautiously by requiring the plaintiff to prove not only that (1) the defendant intended to inflict emotional distress, or knew or should have known that emotional distress was the likely result of his conduct, but also (2) the defendant’s conduct was extreme and outrageous, beyond all possible bounds of decency, and utterly intolerable in a civilized community, (3) the actions of the defendant were the cause of the plaintiff’s distress, and (4) the emotional distress suffered by the plaintiff was severe and of such a nature that no reasonable person could be expected to endure it. Id. at 145.

We conclude that when recovery is sought for negligent, rather than intentional or reckless, infliction of emotional distress, evidence must be introduced that the plaintiff has suffered physical harm. This requirement, like those set forth in Agis, will serve to limit frivolous suits and those in which only bad manners or mere hurt feelings are involved, and will provide a reasonable safeguard against false claims. We see no reason for abandoning such limitations.

Although cases may arise in which the emotional distress absent physical harm may not be temporary or slight, nothing before us indicates that most such claims are not of that character. We are unwilling, therefore, to impose upon the judicial system and potential defendants the burden of dealing with claims of damages for emotional distress that are trivial, evanescent, temporary, feigned, or imagined, in order to ensure that occasional claims of a more serious nature receive judicial resolution.

A careful analysis of the cases dealing with emotional distress discloses an additional reason for denying recovery when there has been only negligent conduct and no physical injury to the plaintiff. Professor Prosser has stated that, where the impact rule has been abandoned, there should be no recovery unless a normal person would have suffered severe emotional distress under the circumstances. W. Pros-ser, Torts, supra at 332. This concept is closely related to the traditional rule of negligence law that, to be compensa-ble, injuries to a plaintiff resulting from the conduct of a defendant must have been reasonably foreseeable. See Hill v. Winsor, 118 Mass. 251, 259 (1875). Where recovery has been allowed for emotional distress absent physical harm, the element of reasonable foreseeability usually exists. Certainly that is true when a defendant intentionally or recklessly inflicts emotional distress (Agis v. Howard Johnson Co., supra), as well as when a plaintiff has been subjected to the threat of serious bodily harm (Colla v. Mandella, supra), or has witnessed the infliction of serious bodily harm on a close member of the plaintiff’s family (Dziokonski v. Babineau, supra).

We believe, further, that emotional distress is reasonably foreseeable when there is a causal relationship between the physical injuries suffered and the emotional distress alleged. Many of the impact cases can be explained on the basis of the reasonable foreseeability of the plaintiff’s emotional distress when that distress occurred after an impact. See, for example, Petition of the U. S. as Owner of the U. S. Coast Guard Vessel CG 95321, 418 F.2d 264 (1st Cir. 1969). This analysis breaks down only with respect to those cases where the impact suffered was of such a nature that no emotional injuries could be expected to result, and those cases where the physical injuries suffered were unrelated to the emotional distress for which the plaintiff was allowed to recover. That these classes of cases exist is not a sufficient basis for allowing recovery, absent some additional element of satisfactory proof, for emotional distress which is not a reasonably foreseeable result of a defendant’s merely negligent conduct.

We therefore conclude, on the basis of the preceding analysis, that in order for any of these plaintiffs to recover for negligently inflicted emotional distress, she must allege and prove she suffered physical harm as a result of the conduct which caused the emotional distress. We answer, further, that a plaintiff’s physical harm must either cause or be caused by the emotional distress alleged, and that the physical harm must be manifested by objective symptomatology and substantiated by expert medical testimony. Finally, the emotional distress for which compensation is sought must be reasonably foreseeable: unless a plaintiff proves that the defendant knew or should have known of special factors affecting that plaintiff’s response to the circumstances of the case, the plaintiff can recover only for that degree of emotional distress which a reasonable person, normally constituted, would have experienced under those circumstances. Whether the emotional distress which a plaintiff is alleged to have experienced is reasonable, is to be determined by the finder of fact.

In summary, we hold that a plaintiff in order to recover for negligently inflicted emotional distress must prove the following: (1) negligence; (2) emotional distress; (3) causation; (4) physical harm manifested by objective symptoma-tology; and (5) that a reasonable person would have suffered emotional distress under the circumstances of the case.⁸

Due to our negative response to question one, we need not address questions one-A and one-B.

Question Two

“If the trier of fact concludes that a plaintiff would probably not have been born except for the mother’s ingestion of DES, is that plaintiff barred from recovery because of physical or emotional damage suffered as a result of the mother’s ingestion of DES?” We answer, Yes.⁹

Discussion

The defendants assert that the instant case most closely resembles the “wrongful life” cases: those in which severely defective infants sue physicians, alleging that “the physician negligently failed to inform the child’s parents of the possibility of their bearing a severely defective child, thereby preventing a parental choice to avoid the child’s birth” (citation omitted). Phillips v. United States, 508 F. Supp. 537, 538 n.1 (D.S.C. 1980). The gravamen of the infant plaintiff’s complaint in such cases is that, because of his defect, “he would be better off not to have been born.” Gleitman v. Cosgrove, 49 N.J. 22, 63 (1967) (Weintraub, C.J., dissenting in part), overruled in part, Berman v. Allan, 80 N.J. 421 (1979). Those courts which have entertained claims of this nature almost invariably have denied recovery, reasoning that “life is more precious than non-life,” id. at 429, and that the judicial system is incapable of assessing damages in a negligence case when that assessment must be based upon a comparison of the relative monetary values of existence and nonexistence. See id. at 427; Phillips v. United States, supra at 543; Becker v. Schwartz, 46 N.Y.2d 401, 411-412 (1978). But see Turpin v. Sortini, 31 Cal.3d 220 (1982).

The plaintiffs, on the other hand, argue that this case is most analogous to the “Good Samaritan” cases in which would-be rescuers have been held subject to a duty to exercise reasonable care in their efforts to save or protect the lives of others. See Black v. New York, N.H. & H.R.R., 193 Mass. 448 (1907); Indian Towing Co. v. United States, 350 U.S. 61 (1955); Restatement (Second) of Torts §§ 323, 324 (1965). When the rescuer’s act was not gratuitous, the plaintiffs insist that Massachusetts law imposes an even stricter duty of care, see Motta v. Mello, 338 Mass. 170, 172-173 (1958), and that the defendants have failed to meet that duty.

Neither of these analogies is compelling. The case before us differs from the “wrongful life” cases in at least one important particular: the plaintiffs here do not assert that they would be better off not to have been born, nor does the certified question make that assumption. The “Good Samaritan” analogy also breaks down when examined closely. The rationale for imposing a duty of reasonable care on one undertaking a rescue is that, if the rescue attempt is not effected in a reasonable manner, the person receiving assistance may be placed in a worse position than if the rescuer had not intervened. See Restatement (Second) of Torts § 323, Comment c (1965). In this case, if the only alternative to the defendants’ “rescue” is that the plaintiffs would not have been born, we cannot perceive how the plaintiffs would have been better off if the defendants had not acted.

We find the reasoning of the “wrongful life” cases more persuasive, however. Accordingly, we hold that if the trier of fact finds that a preponderance of the credible evidence supports the conclusion that a particular plaintiff would not have been born except for her mother’s ingestion of DES, the plaintiff is barred from recovery.¹⁰ In that case, assessment of damages for harm suffered as a result of the defendants’ negligence would require a comparison of the relative monetary values of existence and nonexistence. We agree that an attempt to make such a comparison would be beyond the competence of the judicial system.

The Chief Justice takes the view that, where there is only a probability that a plaintiff would not have been born but for her mother’s ingestion of DES, the manufacturer of the drug may be liable. The circumstances assumed in the second question do not involve situations in which it was probable that a perfect result could have been achieved without the defendants’ negligent involvement. In this case, a defendant’s involvement, even though negligent, is assumed to be the likely cause of the plaintiffs’ very existence. Thus, we find distinguishable those examples on which the Chief Justice relies. No doubt a dentist could be found who would not negligently drop a tooth down a patient’s throat, or a volunteer rescuer who would use reasonable care not to further imperil the subject of the rescue. An oxygen tent that contained no defect could have been provided. Here, however, question two assumes the unavailability of any other means by which a plaintiff probably would have been born: it is more probable than not that a plaintiff would not have been born but for the involvement of the defendant or defendants. The provider of the probable means of a plaintiff’s very existence should not be liable for unavoidable, collateral consequences of the use of that means.

The result of this ruling, however, is that the defense available to the defendants is narrow indeed. It requires that the defendants establish by a preponderance of the evidence that (1) the mothers of the plaintiffs would have suffered, in each instance, a miscarriage absent medical intervention; (2) no other means, method, or substance was available to prevent a miscarriage; and (3) DES did, in fact, prevent the miscarriage and cause the plaintiffs to be born.

Question Three

“Does Massachusetts recognize a right of action for injury to a plaintiff in útero resulting from ingestion of a drug by her mother?” We answer, Yes.

This certified question involves an area of the tort law which has undergone considerable change in recent years, both in Massachusetts and in other jurisdictions. We summarize the relevant Massachusetts cases before dealing with the issue raised by the question.

1. Background. In 1884, this court construed the Massachusetts wrongful death statute then in force as precluding recovery by the administrator of the estate of an infant whose death almost immediately after birth allegedly resulted from injuries suffered when the infant’s mother, because of the defendant’s negligence, slipped and fell on a defective roadway. Dietrich v. Northampton, 138 Mass. 14 (1884). Justice Holmes, speaking for the court, denied the relief requested on the grounds of a lack of precedent and that, “as the unborn child was a part of the mother at the time of the injury, any damage to it which was not too remote to be recovered for at all was recoverable by her.” Id. at 17. Dietrich was cited by the courts of other jurisdictions as “dispositive and controlling” with respect to recovery for prenatal injuries for almost a century. See Bonbrest v. Kotz, 65 F. Supp. 138, 139 & n.3 (D.D.C. 1946).

This court reaffirmed the Dietrich rule in 1950, and again in 1952. Bliss v. Passanesi, 326 Mass. 461 (1950). Cava-naugh v. First Nat’l Stores Inc., 329 Mass. 179 (1952). Both decisions noted the existence of a growing body of law from other jurisdictions reaching results contrary to that reached in Dietrich. Bliss v. Passanesi, supra at 462. Cavanaugh v. First Nat’l Stores Inc., supra at 180-181. The court remained convinced, however, that the problems created by allowing recovery, including the “practical difficulty of reliable proof’ of causation, Bliss v. Passanesi, supra at 463, tipped the balance in favor of retaining the rule against recovery for prenatal injuries.

In 1960, however, in light of the “growing body of precedent in favor of [allowing recovery] and the progress made in medical science,” the administratrix of the estate of an infant who died from injuries suffered in utero was permitted to seek recovery in a wrongful death action. Keyes v. Construction Serv., Inc., 340 Mass. 633, 636 (1960). The Dietrich decision was distinguished on the ground that there the plaintiff’s intestate was injured before becoming viable — able to live outside the mother’s body — while the plaintiff’s intestate in Keyes could have died of injuries sustained after becoming viable. Id. at 637.

Seven years later, in Torigian v. Watertown News Co., 352 Mass. 446 (1967), the administrator of the estate of an infant who was born alive, but died several hours later of injuries sustained while not yet viable, was permitted to reach the jury with a wrongful death claim. The defendant’s argument that no cause of action should be allowed because the harm alleged was too uncertain and speculative was dismissed: “The grounds which have been most frequently urged against allowing recovery are . . . the avoidance of speculation or conjecture, and the encouragement of fictitious claims. . . . The advancement of medical science should take care of most of these arguments. The element of speculation is not present to any greater extent than in the usual tort claim, where medical evidence is offered and the issue of causation must be weighed with great care. . . . The opportunity for fraudulent claims can be faced by the courts as in other types of cases.” (Citation omitted.) Id. at 448-449.

In 1972, wrongful death recovery was denied, to the administrator of the estate of an infant who was stillborn, in reliance upon a statement in Keyes v. Construction Serv., Inc., supra at 637, to the effect that “[i]f the child was stillborn the plaintiff would have no right of action.” Leccese v. McDonough, 361 Mass. 64 (1972). That decision was overruled three years later. Mone v. Greyhound Lines, Inc., 368 Mass. 354 (1975). In Mone, this court reemphasized that speculative damage awards were no more likely in prenatal injury cases than in other tort cases, and stated that it was “not persuaded that stillbirth renders the measure of damages recoverable any less capable of precise calculation than in cases of live birth.” Id. at 360.

2. Recovery for injuries sustained by a plaintiff in útero resulting from ingestion of a drug by the plaintiff’s mother. In answering certified question three, we make the following assumptions, which we find implicit in the questions certified by the judge of the Federal District Court: (a) the plaintiffs have suffered injuries which are legally compensa-ble if wrongfully inflicted; (b) the defendants’ conduct was negligent; and (c) the defendants’ conduct was the cause of the plaintiffs’ injuries. Given those assumptions, we conclude that recovery for personal injuries, negligently inflicted, should be allowed under the circumstances of this case.

Under Massachusetts law, ingestion of a drug which causes injury can constitute the basis for a tort claim. Diaz v. Eli Lilly & Co., 364 Mass. 153 (1973) (exposure to fungicide) . We see no reason to distinguish the case before us, in which the plaintiffs’ mothers ingested the drug, from cases in which plaintiffs themselves ingest a drug directly and suffer injury as a result.

The defendants argue that recovery should not be allowed because of the practical difficulty of proving causation and the consequent danger of fictitious claims. In answering certified question one, we have relied in part on the difficulty of proving causation and the existence and extent of purely emotional distress as a reason for denying recovery of damages for emotional distress in the absence of some objective evidence of physical harm. As we demonstrated earlier, the risk of speculation and fictitious claims is exacerbated, in the area of purely emotional distress, by the absence of any objective medical test or of other evidence that verifies the existence or extent of the plaintiff’s injury. Where the special circumstances of purely emotional injuries do not exist, however, this court has rejected a similar argument as a reason for denying recovery for injuries to a fetus (Mone v. Greyhound Lines, Inc., supra at 359-360). We remain convinced that the possibilities of fictitious claims and recovery of damages based on speculation should not bar the plaintiffs’ action for injuries that can be demonstrated to exist by medical evidence, and for all harm to the plaintiffs that is naturally and reasonably related to those injuries. The certified question assumes that the plaintiffs can prove negligence and causation. “If the tortious conduct and the legal causation of the harm can be satisfactorily established, there may be recovery for any injury occurring at any time after conception.” Restatement (Second) of Torts § 869, Comment d (1979). We believe that “[t]he element of speculation is not present to any greater extent [here] than in the usual tort claim.” Torigian v. Watertown News Co., supra at 448-449. Accordingly, we see no reason why the particular means by which the injuries allegedly were inflicted in this case should require the plaintiffs “to go through life carrying the seal of another’s fault. . . without any compensation therefor.” Montreal Tramways v. Leveille, [1933] 4 D.L.R. 337, 345.

The defendants argue, further, that any change in this area of the tort law should be made by the Legislature. This court dealt with that issue in Diaz v. Eli Lilly & Co., supra at 166-167, and in Mone v. Greyhound Lines, Inc., supra at 358-359, concluding in both cases that, since the proposed changes were not “drastic or radical incursion[s] . . . and would not seriously impair an existing interest, disappoint an expectation, or defeat a reliance,” judicial action was appropriate. Id. at 359. Diaz v. Eli Lilly & Co., supra at 167. We do not view the change wrought in Massachusetts tort law, if any, by our answer to certified question three to be a drastic or radical one and, as discussed below, any reliance or expectation based upon prior law would not be reasonable. Judicial action to clarify this aspect of Massachusetts tort law is therefore appropriate in the instant case.

In summary, we hold that a plaintiff who alleges that she suffered injury in útero as a result of her mother’s ingestion of a drug and that, but for the defendants’ negligence, her mother would not have ingested that drug, has stated a claim upon which relief can be granted under Massachusetts law.

Question Three-A

“If the answer to question 3 is affirmative, is such a right of action available to a plaintiff whose mother ingested the drug prior to your Honorable Court’s decision in Torigian v. Watertown News Co., Inc., 352 Mass. 446 (1967), assuming that it is not established that the fetus was probably viable at the time of the injury?” We answer, Yes.

Question Three-B

“If the answer to question 3 is affirmative, is such right of action available under any circumstances to a plaintiff whose mother ingested the drug prior to your Honorable Court’s decision in Keyes v. Construction Services, Inc., 340 Mass. 633 (1960)?” We answer, Yes.

Certified questions three-A and three-B inquire whether the rule we have stated above applies retroactively and whether, therefore, the plaintiffs in the instant case can take advantage of that rule. The issues raised by these two questions are virtually identical, and we consider them together.

Since the general rule is in favor of retroactive application of a change in decisional law, Tucker v. Badoian, 376 Mass. 907, 918-919 (1978) (Kaplan, J., concurring), Kuhn v. Fair-mont Coal Co., 215 U.S. 349, 372 (1910) (Holmes, J., dissenting) (“Judicial decisions have had retrospective operation for near a thousand years”), we would have to be satisfied that special circumstances exist to limit our answer here to only prospective application. No such special circumstances have been demonstrated to exist in this case.

Primarily because of concern for litigants and others who have relied on existing precedents, judicial changes in Massachusetts contract and property law have been given only prospective effect.¹¹ Johnson Controls, Inc. v. Bowes, 381 Mass. 278 (1980) (insurance contracts). Rosenberg v. Lipnick, 377 Mass. 666 (1979) (antenuptial contracts). Whitinsville Plaza, Inc. v. Kotseas, 378 Mass. 85 (1979) (covenants not to compete in deeds and leases). Cf. Tucker v. Badoian, supra at 918-919 (Kaplan, J., concurring) (“common enemy” rule of property law).¹²

In the area of tort law, however, reliance plays a much smaller part.¹³ See Dalton v. St. Luke’s Catholic Church, 27 N.J. 22, 26 (1958). Tort law, especially that dealing with negligence, does not concern itself with business decisions in the same way as does property and contract law: it would be unreasonable to assert that potential tortfeasors often reflect upon possible tort liability before embarking on a negligent course of conduct, or that they frequently are deterred from negligent conduct by the rules of tort law. See Fitzgerald v. Meissner & Hicks, Inc., 38 Wis. 2d 571, 578 (1968); Bielski v. Schulze, 16 Wis. 2d 1, 18 (1962); Note, Prospective Overruling and Retroactive Application in the Federal Courts, 71 Yale L.J. 907, 945-946 (1962).

The reliance interest that features most prominently in personal injury cases involves insurance against tort liability. R. Keeton, Venturing to Do Justice 42 (1969). See Molitor v. Kaneland Community Unit Dist. No. 302, 18 Ill. 2d 11, 29 (1959), cert. denied, 362 U.S. 968 (1960); Parker v. Port Huron Hosp., 361 Mich. 1, 28 (1960); Balts v. Balts, 273 Minn. 419, 431 (1966); Fitzgerald v. Meissner & Hicks, Inc., supra; Traynor, Quo Vadis, Prospective Overruling: A Question of Judicial Responsibility, 28 Hastings L.J. 533, 545-546 (1977) (hereinafter Traynor) (“Neither the tortfeasor nor the victim normally takes account of expanding or contracting rules of tort liability except tangentially in the course of routinely insuring against such liability”). The defendants in this case base their arguments against retroactive application of the rule we enunciate above largely on the ground that they relied on Massachusetts law, as it existed at the time the acts complained of by the plaintiffs occurred, in determining the appropriate amount of tort liability insurance to obtain. They assert that, regardless whether their conduct was negligent, it would be inequitable to subject them to possible tort liability in excess of the amount of liability insurance they choose to obtain.

The defendants point out, correctly, that the court has taken into account problems involving insurance coverage in determining whether decisions effecting changes in tort law should be applied only prospectively. Our review of these decisions, however, reveals that only where the discarded rule afforded complete immunity from suit to a large class of defendants has the court indicated that prospective application of a new rule would be warranted. See Whitney v. Worcester, 373 Mass. 208 (1977) (municipal immunity); Ricker v. Northeastern Univ., 361 Mass. 169 (1972) (charitable immunity).¹⁴ In these cases, the court assumed that the institutions involved reasonably might have relied on their immunity in making a choice not to obtain liability insurance. Other courts have made the same assumption. See cases collected in Note, The Retroactivity of Minnesota Supreme Court Personal Injury Decisions, 6 Wm. Mitchell L. Rev. 179, 186-187 n.34 (1980); Traynor, supra at 546; R. Keeton, Venturing to Do Justice, supra.

On the other hand, in tort cases not concerned with the issues of governmental or charitable immunity, this court has decided against only prospective application of the new tort rules on the ground that “the rule to be discarded may not reasonably be supposed to have determined the conduct of the litigants” with respect to, among other things, obtaining insurance against the liability imposed. Diaz v. Eli Lilly & Co., 364 Mass. 153, 167 n.46, quoting B. Cardozo, The Nature of the Judicial Process 151 (1921) (holding that wife can maintain action for loss of consortium against party who negligently inflicted injury on husband). See Pevoski v. Pevoski, 371 Mass. 358, 361 (1976) (abolishing inter-spousal immunity in motor vehicle accident cases); Bouchard v. DeGagne, 368 Mass. 45, 49 (1975) (retroactively applying rule that landlord owes duty of reasonable care to all lawful visitors).¹⁵

The defendants here are not absolutely immune from liability for negligent conduct causing injury to a fetus in útero: since 1884, it has been clear that a mother could recover for injuries to her child in útero as an element of damages in her own suit against a negligent tortfeasor. Dietrich v. Northampton, 138 Mass. 14, 17 (1884). Accordingly, the defendants do not claim that they decided not to obtain liability insurance in reliance upon their complete immunity from suit. Their claim, rather, is that it would be inequitable to subject them to the risk of a recovery by the plaintiffs in excess of the liability insurance coverage they have chosen to obtain. In addition, the defendants assert that allowing recovery here would be unfair to their insurers, who have set their rates in reliance on existing precedents.

We find neither of these reliance arguments convincing. The defendants have cited no case in which a new rule of tort law was given only prospective effect because parties had relied upon the old rule in determining the appropriate amount of liability insurance coverage to obtain, and our research has revealed none. The defendants admit that, in the late 1940’s and early 1950’s, when DES was most commonly used, suits were filed far less frequently and multimillion dollar tort recoveries were unknown. We infer that these facts, and not then-existing Massachusetts tort law, reasonably may be supposed to have determined the amount of liability insurance coverage obtained by the defendants. In any event, the defendants have not advanced reasons of public policy sufficient to prompt us to limit recovery in this case to the amount of their liability insurance coverage. Cf. Sorensen v. Sorensen, 369 Mass. 350 (1975).

With regard to possible reliance by the defendants’ insurers on Massachusetts tort law as it existed at the time of the defendants’ allegedly negligent conduct, we do not believe that such reliance reasonably may be supposed to have occurred. Certainly we have no basis for assuming, in this case, that such reliance occurred. Further, we agree with Judge Keeton that “general acceptance [of the argument] would in effect disable courts from creative decisions in accident law.” R. Keeton, Venturing to Do Justice, supra. See Lewis v. Lewis, 370 Mass. 619, 630 n.4 (1976); Pevoski v. Pevoski, supra at 363 (Quirico, J., concurring).

Although we conclude that any reliance on former Massachusetts law that may have existed on the part of the defendants and their insurers does not require that the rule allowing recovery for prenatal injuries be given only prospective effect, we do not believe that the inquiry should end there. We proceed to examine whether the purposes of the rule will be served by retroactive application, both in general and in the. case before us. See McIntyre v. Associates Fin. Servs. Co. of Mass., Inc., 367 Mass. 708, 712 (1975); Note, Prospective Overruling and Retroactive Application in the Federal Courts, 71 Yale L.J. 939, 942; Traynor, supra at 561. Since the purpose of the rule is to protect a child’s right to commence life unhampered and unimpaired by damage negligently caused to the body or mind by another, Sylvia v. Gobeille, 101 R.I. 76, 78 (1966), retroactive application should follow unless strong equity or policy reasons dictate otherwise.

The defendants’ strongest argument against retroactivity, as we see it, is that to allow retroactive application may discourage an activity which society views as desirable: the development of new and more efficacious drugs. The defendants, referring us to Ducharme v. Merrill-Nat’l Laboratories, 574 F.2d 1307,1310-1311 (5th Cir.), cert. denied, 439 U.S. 1002 (1978), where the court noted that the “collapse of the commercial liability insurance market” for insurance against tort liabilities arising out of the development of swine flu vaccine kept drug companies from entering the field until the Federal government intervened to limit their liability, assert that similar results will follow a retroactive decision in this case.

We are not prepared, on the basis of the record before us, to debate the defendants’ claims in this regard. On that record, we can find no rational justification for distinguishing between tort victims injured before and those injured after our decision in this case. If, as the defendants contend, sound public policy dictates that limitations be placed upon recovery by plaintiffs injured in útero as a result of their negligence, we believe that those limitations should be imposed by the Legislature. Accordingly, we answer, Yes, to certified questions three-A and three-B.

Question Four

“Assuming that the evidence does not warrant a conclusion that the defendants conspired together, or engaged in concerted action, or established safety standards through a trade association, may the defendant manufacturers, who probably supplied some of the DES ingested by the mothers of the plaintiff class, be held liable to members of the plaintiff class when neither the plaintiffs nor defendants can identify which manufacturer’s DES was ingested by which mothers?”

We cannot answer the question in the form stated. One of the difficulties presented by the form of the question is that, unlike question one, this question does not explicitly assume that the plaintiffs will be able to establish the negligence of these particular defendants. We suggest that whether a plaintiff can “identify” a defendant as the party who caused the harm is a question distinct from the question whether that party was negligent. In an effort to be of assistance, however, we set forth our general views. In so doing, we assume that the defendants will be shown to have been (a) negligent, and (b) actively in the DES market during all or a substantial part of the relevant period of time in which the mothers of the plaintiffs ingested DES.

Discussion

Identification of the party responsible for causing injury to another is a longstanding prerequisite to a successful negligence action. See Smith v. Ariens Co., 375 Mass. 620 (1978) (plaintiff alleging injury caused by negligent design of snowmobile must prove defendant manufactured same). This requirement serves two purposes: it separates wrongdoers from innocent actors, and also ensures that wrongdoers are held liable only for the harm that they have caused.

The plaintiffs seek to avoid the identification requirement, positing in its stead a “market share” theory of recovery. We believe that the plaintiffs’ market share theory fails adequately to protect either of the interests served by the identification requirement. It would be inappropriate, therefore, to allow recovery under this theory.

The theory advanced by the plaintiffs would require them to prove only: that (1) each named defendant acted negligently in marketing DES; (2) each plaintiff was injured, and the extent of that injury; and (3) DES caused the plaintiffs’ injuries. The plaintiffs assert that, if they are successful in making their proof, they should be allowed to recover 100% of their losses from the six named defendants: the plaintiffs would have us assume that these six defendants were responsible for all of the DES distributed in the “relevant” DES market, defined as the total sales in Massachusetts of DES for use in pregnancy. Liability for any damage award under this theory would be apportioned among named defendants in proportion to each named defendant’s share of the DES distributed by all named defendants. The plaintiffs also would have us prohibit exculpatory proof, that is, proof by particular defendants that the DES they marketed could not have been ingested, for whatever reason, by a particular plaintiff’s mother. To the defendants’ argument that other manufacturers than those named are responsible for marketing the DES ingested by at least some plaintiffs’ mothers, the plaintiffs reply that this problem can be solved by allowing the named defendants to seek contribution from unnamed marketers of DES. This means, of course, that the defendants would bear the risk that a particular marketer of DES to a plaintiff’s mother would turn out to be insolvent or otherwise immune from contribution.

The plaintiffs’ theory is based in large part upon the recent case of Sindell v. Abbott Laboratories, 26 Cal. 3d 588, cert. denied, 449 U.S. 912 (1980), where, by a four-to-three decision, the California Supreme Court adopted a form of market share liability.¹⁶ Both the plaintiffs’ theory and that of the California court can be traced to the principles developed in Summers v. Tice, 33 Cal. 2d 80 (1948). In Summers, the plaintiff was injured when two hunters each negligently shot in his direction. Although the plaintiff was unable to prove which of the two hunters had fired the shot that struck him, the hunters were held jointly and severally liable for the entire loss.

The plaintiffs’ analogy to Summers is not wholly compelling. In Summers, all of the possible tortfeasors were joined, and all had acted negligently. Here, in contrast, only six of what may be a very large number of potential tortfeasors have been joined. The plaintiffs make a quantum leap from the principles of Summers to their market share theory. To recover, the plaintiff in Summers was required to prove the negligence of all of the actors who might have caused his injury. Under their market share theory, the DES plaintiffs would need to prove only the negligence of those six actors they have chosen to sue. Acceptance of this theory of proof would create the risk of holding the named defendants liable in negligence for more harm than they caused. The plaintiffs’ market share theory does not protect tortfeasors from liability exceeding their responsibility, and therefore disserves one of the purposes of the identification requirement.

The plaintiffs’ theory also fails adequately to take account of the second purpose of the identification requirement: to separate tortfeasors from innocent actors. By prohibiting exculpatory proof — proof by a defendant that it could not have been responsible for injury to a particular plaintiff — the plaintiffs would practically ensure that defendants innocent of wrongdoing to a particular plaintiff would be held liable to her.

We cannot accept the plaintiff’s theories. Public policy favors the development and marketing of new and more efficacious drugs. The Restatement (Second) of Torts recognizes this policy by rejecting strict liability in favor of negligence for drug related injuries. Restatement (Second) of Torts § 402A, Comment k (1965). Under the plaintiffs’ market share theory each potential drug-marketing defendant would risk being held liable, not only for injuries that resulted from its own negligence, but also for injuries resulting from the negligence of other marketers and even for injuries caused by drugs marketed nonnegligently by others. Imposition of such broad liability could have a deleterious effect on the development and marketing of new drugs,¹⁷ especially those marketed genetically.¹⁸

In both their reply brief and in oral argument, the plaintiffs suggest that if particular aspects of their market share theory create difficulties, we should excise, reformulate, and rewrite to create a theory under which they could recover without meeting the identification requirement. The posture of the case and consequent state of the record, the magnitude of the ramifications of our decision with respect to this certified question, and our view of the judicial process combine to convince us that such a course of action is imprudent at this time.

That is not to say that on an adequate record this court would not recognize some relaxation of the traditional identification requirement in appropriate circumstances so as to allow recovery against a negligent defendant of that portion of a plaintiff’s damages which is represented by that defendant’s contribution of DES to the market in the relevant period of time.

Summary

In summary, therefore, we answer that (1) absent physical harm, Massachusetts does not recognize a right of action for emotional distress alleged to have resulted from an increased statistical likelihood that the plaintiffs will suffer a serious disease in the future, (2) if the trier of fact concludes that a plaintiff would probably not have been born except for the mother’s ingestion of DES, the plaintiff is barred from recovery for damages suffered as a result of her mother’s ingestion of the drug, (3) Massachusetts does recognize a right of action for injury to a plaintiff, in útero, resulting from ingestion of a drug by a mother and that right of action is not limited to prospective application, and (4) we are unable to give a definitive answer on this record whether the manufacturers of DES named as defendants (who probably supplied some of the DES ingested by the mothers of the plaintiffs) can or cannot be held liable to the members of the plaintiffs’ class when neither the plaintiffs nor the defendants can identify which manufacturers’ DES was ingested by which mothers and where the named defendants are only some of the manufacturers of DES ingested by the mothers of the plaintiffs. We have indicated, however, the view that we might permit recovery from those defendants shown to be negligent to the extent of their participation in the DES market, even though the plaintiffs cannot identify the particular source of DES which their mothers ingested.

The Reporter of Decisions and the clerk of this court are to follow the procedures set out in Hein-Werner Corp. v. Jackson Indus., 364 Mass. 523, 530-531 (1974), for furnishing copies of this opinion and transmitting them.

So ordered.

“Said to be of the maximum level of 1.4 women per thousand exposed” (footnote in original).

“A variant of this theory is sometimes referred to as ‘enterprise liability.’ In the leading case employing this term the defendants had delegated the formulation of safety standards to a trade association” (footnote in original).

3 See Dziokonski v. Babineau, 375 Mass. 555, 561 n.7 (1978); McDonough v. Whalen, 365 Mass. 506, 517 (1974); George v. Jordan Marsh Co., 359 Mass. 244, 255 (1971).

4 The Restatement (Second) of Torts § 7 (3) and Comment e (1965) distinguishes between “physical harm” and “bodily harm”: the former denotes “physical impairment of the human body, ... or of land or chattels,” while the latter is restricted to impairment of the body. The certified question uses “physical harm” in the sense of “bodily harm” as that term is used in the Restatement, and in contrast with purely mental harm. We use the two terms interchangeably in this opinion to denote harm to the bodies of the plaintiffs.

5 Many jurisdictions specifically require a showing of physical harm as a precondition to recovery for emotional distress. See M.B.M. Co. v. Counce, 268 Ark. 269 (1980) (dicta); Keck v. Jackson, 122 Ariz. 114 (1979) (en banc); Robb v. Pennsylvania R.R., 58 Del. 454 (1965). Accord, McClain v. Faraone, 369 A.2d 1090, 1095 (Del. Super. 1977); Gilper v. Kiamesha Concord, Inc., 302 A.2d 740 (D.C. Ct. App. 1973); Hatfield v. Max Rouse & Sons N.W., 100 Idaho 840 (1980); Charlie Stuart Oldsmobile, Inc. v. Smith, 171 Ind. App. 315 (1976), vacated in part, 175 Ind. App. 1 (1977); Clemm v. Atchison, T & S.F. Ry., 126 Kan. 181 (1928); Daley v. LaCroix, 384 Mich. 4,12 (1970) (“definite and objective physical injury”). Accord, Toms v. McConnell, 45 Mich. App. 647, 657 (1973) (bystander case) (“severe traumatic depressive reaction” constitutes “definite and objective physical injury”); Okrina v. Midwestern Corp., 282 Minn. 400 (1969); Sears, Roebuck b Co. v. Young, 384 So. 2d 69 (Miss. 1980); Fournell v. Usher Pest Control Co., 208 Neb. 684 (1981); Falzone v. Busch, 45 N.J. 559 (1965) (substantial bodily injury or sickness). Accord, Berman v. Allan, 80 N.J. 421, 433 (1979); Umbaugh Pole Bldg. Co. v. Scott, 58 Ohio St. 2d 282 (1979); fines v. Norman, 351 P.2d 1048 (Okla. 1960); Melton v. Allen, 282 Or. 731 (1978); D’Ambra v. United States, 114 R.I. 643 (1975) (bystander case); Padgett v. Colonial Wholesale Distrib. Co., 232 S.C. 593 (1958); Chisum v. Behrens, S.D. (1979) (283 N.W. 2d 235 [S.D. 1979]); Trent v. Barrows, 55 Tenn. App. 182 (1965); Farmers & Merchants State Bank v. Ferguson, 617 S.W.2d 918 (Tex. 1981) (dicta); Hughes v. Moore, 214 Va. 27 (1973); Vaillancourt v. Medical Center Hosp. of Vt., Inc., 139 Vt. 138 (1980); Hunsley v. Giard, 87 Wash. 2d 424 (1976) (en banc).

The courts of other jurisdictions require emotional distress to be physically manifested in order to be compensable. See Towns v. Anderson, 195 Colo. 517 (1978) (en banc) (physical manifestations or mental illness); Barnhill v. Davis, 300 N.W.2d 104 (Iowa 1981) (bystander case); Vicnire v. Ford Motor Credit Co., 401 A.2d 148, 155 (Me. 1979) (illness or bodily harm “substantial and manifested by objective symptomatology”); Vance v. Vance, 286 Md. 490, 500 (1979) (“capable of objective determination”); Corso v. Merrill, 119 N.H. 647, 653 (1979) (bystander case) (physical symptoms . . . susceptible to some form of objective medical determination and proved through qualified medical witnesses”). See also McDowell v. Davis, 33 N.C. App. 529 (1977) (physical impact or physical injury).

6 Five jurisdictions apparently still adhere to the impact rule. See Gilliam v. Stewart, 291 So. 2d 593 (Fla. 1974); Carlinville Nat'l Bank v. Rhoads, 63 Ill. App. 3d 502, 503 (1978) (bystander case, dicta); Indiana Motorcycle Ass’n v. Hudson, Ind. App. (1980) (399 N.E.2d 775 [Ind. App. 1980]); Deutsch v. Schein, 597 S.W.2d 141 (Ky. 1980); Williams v. School Dist. of Springfield, 447 S.W.2d 256, 266 (Mo. 1969).

7 The plaintiff in George suffered a heart attack, allegedly as a result of the defendant’s debt-collection practices.

8 We note that, by seeking recovery in this suit for emotional harm only, the plaintiffs have created a problem involving compensation for later-occurring physical harm which may not be dealt with adequately in any award of damages they receive. See Boyd v. Jamaica Plain Co-op. Bank, 7 Mass. App. Ct. 153, 166-167 (1979).

9 Due to our affirmative response to question two, we need not address part 2A of the question.

10 We assume that, by using the word “probably” in the certified question, the judge intended to indicate that the degree of proof required for that proposition to be established is a preponderance of the evidence, the usual standard in a civil case.

11 We are not here concerned with the retroactive application of overruling decisions in the field of criminal law.

12 See McIntyre v. Associates Fin. Servs. Co. of Mass., Inc., 367 Mass. 708, 713 (1975), where the court, after considering the effect of retroactive application of the due process decision of Fuentes v. Shevin, 407 U.S. 67 (1972), on property rights acquired by parties who had relied on prior Massachusetts prejudgment attachment procedures, decided that the Fuentes rule should be applied only prospectively.

13 Indeed, some commentators argue that considerations of reliance should play no part, and that all judicial decisions affecting tort law should be given retroactive effect. W. Seavey, Cogitations on Torts 69 (1954). Traynor, Quo Vadis, Prospective Overruling: A Question of Judicial Responsibility, 28 Hastings L.J. 533, 545-546 (1977). Former Chief Justice Traynor of the California Supreme Court, believes that “most judges would agree with [Justice Benjamin] Cardozo: ‘My impression is that the instances of honest reliance and genuine disappointment are rarer than they are commonly supposed to be by those who exalt the virtues of stability and certainty.’ Address by Justice Cardozo, New York State Bar Association, Jan. 22, 1932, in 55 N.Y. St. B. Ass’n Rep. 295 (1932).” Id. at 542 n.17.

14 In both cases, the Legislature acted to substitute, for these common law immunities, statutes allowing some recovery in tort against the institutions formerly protected by the immunities. St. 1978, c. 512, § 15 (municipal immunity). St. 1971, c. 785 (charitable immunity). The court was not required, in either instance, actually to apply a judicial change in tort law prospectively.

15 See also Sorensen v. Sorensen, 369 Mass. 350 (1975), in which this court, in abolishing parent-child tort immunity, limited recovery to the extent of the defendant’s insurance coverage. But see New Hampshire Ins. Co. v. Fahey, 385 Mass. 137, 138 (1982), which suggests that the availability of insurance may be irrelevant. In any case, the defendants here have argued for no such limitation on the plaintiffs’ recovery.

16 The New York Court of Appeals recently affirmed a decision allowing a plaintiff to recover, without identifying the defendant as the source of the DES ingested by the plaintiff’s mother, on a “concerted action [among manufacturers]” theory. Bichler v. Eli Lilly & Co., 55 N.Y.2d 571 (1982). That case provides little guidance with respect to the question asked of us, however. Question four assumes that the plaintiffs will not be able to prove that the defendants here engaged in concerted action. Further, the plaintiff in Bichler was permitted to recover primarily because the defendant failed to object to the trial judge’s instructions. The Court of Appeals held that the instructions became the law of the case and that, in light of those instructions, the jury reached a proper verdict. The Court of Appeals thus did not discuss the merits of the substantive arguments advanced, but affirmed the trial court’s decision on grounds essentially procedural in nature.

17 Cf. Polio Immunization Program, 1976: Hearing Before the Subcommittee on Health of the Senate Comm, on Labor and Public Welfare, 94th Cong., 2d Sess. 119 (Sept. 23,1976) (statement of Assistant Surgeon General David Sencer): “Largely because of court opinions dramatizing the inherent risks of vaccines and vaccine-associated disability, patterns of immunization programming by State and local health agencies and vaccination practices of private medical professionals are being modified. Called-for warnings of inherent risks in vaccines are likely to alarm potential vaccine recipients and result in diminished immunization program effectiveness. Manufacturer liability for vaccine-associated disability, regularly assigned by courts, threatens a predictable vaccine supply — especially of oral polio vaccine — and diminishes the chances of significant independent manufacturer-sponsored research and development of new biologies” (emphasis supplied).

18 Indeed, if a cure for clear-cell adenocarcinoma lies in the development and manufacture of some new drug, imposing market share liability might prevent the marketing of a cure for the very cancer threatening the plaintiffs.